Ash Sap 2016

CHAPTER
1
Molecular basis of hematology
Laura G. Schuettpelz and Omar Abdel-Wahab
Basic concepts, 1 Clinical applications of Glossary, 19
Analytic techniques, 6 DNA technology in hematology, 14 Bibliography, 21
Basic concepts nucleotide and the 3′ position of the sugar moiety of the
adjacent nucleotide. The two strands are connected through
Advances in recombinant DNA technology over the past hydrogen bonds between strict pairs of purines and pyrimi-
several decades have substantially altered our view of dines; that is, adenine must be paired with thymine (A-T)
biologic processes and have immediate relevance to our and guanine must be paired with cytosine (G-C). This is
understanding of both normal hematopoietic cell function known as Watson–Crick base pairing. Consequently, the two
and hematologic pathology. A complete review of molecular strands of DNA are said to be complementary, in that the
genetics is beyond the scope of this chapter, but the following sequence of one strand determines the sequence of the other
is intended as a review of the concepts of the molecular through the demands of strict base pairing. The two strands
biology of the gene, an introduction to epigenetics and are joined in an antiparallel manner so that the 5′ end of one
genomics, an outline of noncoding RNAs, and an explana- strand is joined with the 3′ end of the complementary strand.
tion of the terminology necessary for understanding the role The strand containing the codons for amino acid sequences
of molecular biology in breakthrough discoveries. Emerging is designated as the sense strand, whereas the opposite strand
diagnostic and therapeutic approaches in hematology will be that is transcribed into messenger RNA (mRNA) is referred
reviewed. The concepts outlined in the following sections to as the antisense strand.
also are illustrated in Figure 1-1; in addition, boldface terms
in the text are summarized in the glossary at the end of
Structure of the gene
this chapter. Several examples of how these concepts and
techniques are applied in clinical practice are included. DNA dictates the biologic functions of the organism by the
flow of genetic information from DNA to RNA to protein.
The functional genetic unit responsible for the production of
Anatomy of the gene a given protein, including the elements that control the tim-
Structure of DNA ing and the level of its expression, is termed a gene. The gene
contains several critical components that determine both the
DNA is a complex, double-stranded molecule composed of amino acid structure of the protein it encodes and the mech-
nucleotides. Each nucleotide consists of a purine (adenine anisms by which the production of that protein may be
or guanine) or pyrimidine (thymine or cytosine) base controlled. The coding sequence, which dictates protein
attached to a deoxyribose sugar residue. Each strand of DNA sequence, is contained within exons; these stretches of DNA
is a succession of nucleotides linked through phosphodiester may be interrupted by intervening noncoding sequences, or
bonds between the 5′ position of the deoxyribose of one introns. In addition, there are flanking sequences in the
5′ and 3′ ends of the coding sequences that often contain
important regulatory elements that control the expression of
Conflict-of-interest disclosure: Dr. Schuettpelz declares no the gene.
competing financial interest. Dr. Abdel-Wahab declares no competing
Genes are arrayed in a linear fashion along chromo-
financial interest.
Off-label drug use: Dr. Schuettpelz: not applicable. Dr. Abdel-Wahab: somes, which are long DNA structures complexed with pro-
not applicable. tein. Within chromosomes, DNA is bound in chromatin,
| 1
BK-ASH-SAP_6E-150511-Chp01.indd 1 4/27/2016 6:35:25 PM

2 | Molecular basis of hematology
Nucleus
Coding Noncoding
sequence (intervening 3’ coding
5’ (exon) sequence, intron) strand
DNA
3’ 5’ noncoding
Transcription
strand
mRNA 5’ Exon Intron
3’
precursor
5’ CAP 3’ Poly (A),
modification and
Processing shortening of
transcript
Processed 5’ CAP Poly (A)-3’

mRNA
transcript
Nuclear “pore” 5’ CAP Poly (A)-3’ Cytoplasm
mRNA
Transport to
cytoplasm Figure 1-1 Flow of genetic information from
Initiation factors,
DNA to RNA to protein. DNA is shown as a
Translation tRNA, ribosomes double-stranded array of alternating exons (red)
and introns (pink). Transcription,
5’ CAP Poly (A)-3’ posttranscriptional processing by splicing,
polyadenylation, and capping are described in the
Completed apoprotein
text. The mature transcript passes from the nucleus
Cofactors and other subunits
to the cytoplasm, where it is translated and further
Microsomes, Golgi, etc.
modified to form a mature protein. Reproduced
with permission from Hoffman R et al. eds.
Hematology: Basic Principles and Practice. 6th
Completed functioning protein
Edition. Philadelphia, PA: Saunders Elsevier,
Inc; 2013:5.
a complex of DNA with histone and non-histone proteins protein from a gene is the synthesis of a pre-messenger RNA
that “shield” the DNA from the proteins that activate gene (pre-mRNA). The transcription of pre-mRNA is directed by
expression. The ends of chromosomes are capped by com- RNA polymerase II, which in conjunction with other pro-
plexes of repetitive DNA sequences and associated proteins teins generates an RNA copy of the DNA sense strand. The
known as telomeres. The DNA replication machinery can- introns are then removed by a complex process called mRNA
not effectively replicate the very ends of chromosomes, and splicing. This process involves the recognition of specific
thus shortening occurs with each cell division, ultimately sequences on either side of the intron that allow its excision
leading to chromosomal instability and cellular senscence. in a precise manner that maintains the exon sequence.
Telomeres are therefore critical for protecting chromosome The mRNA may then undergo modifications at the 5′ and 3′
ends from degradation and fusion, and mutations in genes ends (capping and polyadenylation, respectively). Although
encoding components of the telomere complex are asso RNA splicing is largely restricted to the nucleus, it also can
ciated with the bone marrow failure syndrome dyskerato- occur in the cytoplasm of platelets and neutrophils activated
sis congenita. Furthermore, inappropriate activation of by external stimuli.
the enzymes that maintain telomeres (“telomerase”) is Splicing of mRNA is a critical step in gene expression with
associated with numerous cancers, including hematologic important implications for understanding hematologic
malignancies. disease. Splicing is controlled by the spliceosome, a large
complex of proteins (100-300) and five small nuclear ribo-
nuclear proteins (snRNPs). mRNA splicing is an important
mechanism for generating diversity of the proteins produced
Flow of genetic information
by a single gene. Some genes exhibit alternative splicing, a
Transcription
process by which certain exons are included in or excluded
RNAs are mostly single-stranded molecules that differ from from the mature mRNA, depending on which splice
DNA in two ways: by a sugar backbone composed of ribose sequences are used in the excision process. For example, this
rather than deoxyribose, and by containing the pyrimidine is the means by which some erythroid-specific proteins of
uracil rather than thymine. The first step in the expression of heme synthesis (aminolevulinic acid [ALA] synthase) and

Basic concepts | 3
energy metabolism (pyruvate kinase) are generated, con- or they may change the amino acid sequence in a manner
trasting with the alternatively processed genes in the liver that has no effect on the overall function of the protein; these
and other tissues. This permits functional diversity of the are predicted to be phenotypically silent mutations. In other
products of the same gene and is one of several determinants cases, mutations may lead to a loss or gain of protein func-
of tissue specificity of cellular proteins. Mutations in the tion, or may result in the acquisition of a new function.
sequences of either introns or exons can derange the splicing Sickle cell disease is an example of a single base-pair change
process by either creating or destroying a splice site so that (point mutation) resulting in an amino acid alteration that
the intron sequence is not removed or the exon sequence critically changes the chemical characteristics of the globin
eliminated. If abnormal splicing results in a premature stop molecule. Other mutations may change a codon to a termi-
codon (nonsense mutation), then a surveillance pathway nation codon, resulting in premature termination of the
known as nonsense-mediated decay may result in degrada- protein (nonsense mutation). Finally, single or multiple
tion of the abnormal mRNA. This mechanism generally base-pair insertions or deletions can disrupt the reading
applies to stop codon mutations in the first one-third to one- frame of genes. These frameshift mutations render the gene
half of the mRNA and works to prevent synthesis of mutant incapable of encoding normal protein. These latter two
peptides. When mutations occur in the last one-third of the abnormalities account for some β-thalassemias and for poly-
mRNA molecule, abnormal peptides may be produced. cythemia due to a gain of function in the erythropoietin
Mutations affecting the spliceosome machinery are found receptor. Clinically important mutations also may occur in
in a variety of hematopoietic malignancies. For example, the noncoding region of genes, such as in the regulatory ele-
splicing factor mutations (SF3B1, U2AF1 and SRSF2) are ments upstream of the initiation codon or within intronic
found in approximately 50% of patients with myelodysplas- splicing sites.
tic syndrome (MDS) and about 14% of patients with acute
myeloid leukemia (AML). While the functional conse-
Control of gene expression
quences of these mutations are not fully understood, recent
reports suggest that they may lead to aberrant splicing of With the exception of lymphocytes (which undergo unique
select target genes associated with hematopoietic disease changes in the DNA encoding immunoglobulin or the T-cell
pathogenesis. receptor) and germ cells (which contain only half of the DNA
of somatic cells), each nucleated cell in an individual has the
same diploid DNA content. Consequently, biologic processes
Translation
are critically dependent on gene regulation, the control of
The mature mRNA is transported from the nucleus to the gene expression such that proteins are produced only at the
cytoplasm, where it undergoes translation into protein. The appropriate time within the appropriate cells. Gene regula-
mRNA is “read” in a linear fashion by ribosomes, which are tion is the result of a complex interplay of specific sequences
structures composed of ribonucleoprotein that move along within a gene locus, chromatin, and regulatory proteins
the mRNA and insert the appropriate amino acids, carried (transcription factors) that interact with those sequences to
by transfer RNAs (tRNAs), into the nascent protein. The increase or decrease the transcription from that gene.
amino acids are encoded by three base triplets called codons, DNA sequences that lie in proximity to and regulate the
the genetic code. The four bases can encode 64 possible expression of genes, which encode protein, are termed
codons; because there are only 20 amino acids used in pro- cis-acting regulatory elements. Nearly all genes have a site
tein sequences, more than one codon may encode the same for binding RNA polymerase II that is within the first
amino acid. For this reason, the genetic code has been termed 50 bases 5′ to the structural gene and is called the promoter
degenerate. An amino acid may be encoded by more than region. Other sequences that regulate the level of transcrip-
one codon; however, any single codon encodes only one tion of the gene are located at less predictable distances from
amino acid. The beginning of the coding sequence in mRNA the structural gene. Such sequences may increase (enhancers)
is encoded by AUG codon that has variable translation initia- or decrease (silencers) expression. A special type of enhancer
tion activity determined by the neighboring nucleotide is locus control region (LCR), which was first and best
sequences (Kozak sequence). In addition, there are three defined in the b-globin cluster of genes on chromosome 11.
termination codons (UAA, UAG, and UGA) that signal It is located approximately 50 kilobases (kb) upstream
the end of the protein sequence. from the β-globin gene, controls all genes in the b-globin
Single-base-pair alterations in the coding sequence of locus, and also has a strong tissue-specific activity (erythroid-
genes may have a range of effects on the resultant pro specific).
tein. Because the genetic code is degenerate, some single- Control of gene expression is exerted through the interac-
base-pair changes may not alter the amino acid sequence, tion of the cis-acting elements described previously with

proteins that bind to those sequences. These nuclear DNA processes such as X chromosome inactivation, imprinting,
binding proteins are termed trans-acting factors or tran- aging, and carcinogenesis.
scription factors. Most of these proteins have a DNA bind- Mendelian genetics is based on the principle that the phe-
ing domain that can bind directly to regulatory sequences notype is the same whether an allele is inherited from the
within the gene locus; many of them contain common mother or the father, but this does not always hold true.
motifs, such as zinc-fingers or leucine zippers, which are Some human genes are transcriptionally active on only
shared by many transcription factors. In addition, they fre- one copy of a chromosome (such as the copy inherited from
quently have unique domains that allow them to interact the father), whereas the other copy on the chromosome
with other transcription factors. Thus, a complex pattern inherited from the mother is transcriptionally inactive. This
exists whereby the expression of different transcription fac- mechanism of gene silencing is known as imprinting,
tors, which may interact both with one another and with and these transcriptionally silenced genes are said to be
specific regions of DNA to increase or decrease transcrip- “imprinted.” When genes are imprinted, they are usually
tion, determines the unique tissue and stage-specific expres- heavily methylated in contrast to the nonimprinted copy
sion of the genes within a given cell. of the allele, which typically is not methylated. A classic
example of imprinting is the inheritance of Prader-Willi
and Angelman syndromes, which are associated with a
Epigenetics
4-megabase (Mb) deletion of chromosome 15. This region
For a gene to be expressed, chromatin must be unwound contains the gene associated with Angelman syndrome,
and the DNA made more accessible to regulatory proteins. UBE3A, encoding a ligase essential for ubiquitin-mediated
This is controlled by epigenetic processes or modifications protein degradation during brain development. This gene
to the genome that regulate gene expression without alter- is imprinted on the paternal allele. In addition, the region
ing the underlying nucleotide sequence. These changes may contains multiple genes associated with Prader-Willi syn-
be modulated by environmental factors and may be herita- drome, which are imprinted on the maternal allele. Thus,
ble. Epigenetic modulation of gene expression was first rec- maternal inheritance of a mutation or deletion in UBE3A
ognized in studies of glucose-6-phosphate dehydrogenase removes the single active copy of the gene and results in
(G6PD), a protein encoded by an X-linked gene. Ernest Angelman syndrome, and paternal inheritance of deletions
Beutler deduced the principle of random embryonic in this region remove the only active copies of the Prader-
X chromosome inactivation from studies of G6PD defi- Willi–associated genes and result in Prader-Willi syndrome.
ciency. His observations and the studies of Mary Lyon and As DNA methylation modulates gene activity, acquired
Susumu Ohno on the mechanism of dosage compensation changes in methylation patterns are thought to contribute to
in mammals led to an understanding of X chromosome functional alterations in hematopoietic stem cells (HSCs)
inactivation in females. This was the first example of sto- during aging. HSCs display an age-related decline in func-
chastic epigenetic silencing in humans, demonstrating that tion and a relative loss of lymphoid differentiation potential,
human females are mosaics of the activity of X chromo- and these changes are associated with site-specific changes in
some–encoded genes. Using this principle in tumor tissue DNA methylation. Furthermore, aberrant methylation may
derived from females led to early demonstrations that neo- contribute to cancer. For example, mutations in enzymes
plastic diseases are, for the most part, clonal. Two common affecting DNA methylation, most notably the DNA methyl-
forms of epigenetic changes are DNA methylation and his- transferase DNMT3A, are common in acute myeloid
tone modifications. leukemia (AML). Mutations in DNMT3A are found in
approximately 20% of AML patients and are associated with
poor outcomes. These mutations appear to be acquired early
DNA methylation
in the evolution of the disease, thus likely serving as initiators
In addition to being complexed with protein, the DNA of of the disease. Other enzymes affecting DNA methylation,
inactive genes is modified by the addition of methyl groups including TET2 and IDH1/2, are also recurrently mutated in
to cytosine residues (resulting in 5-methylcytosine or 5mC). AML, further illustrating the role of aberrant methylation in
Methylation normally occurs throughout the genome. It is the pathogenesis of this disease. Of note, oxidation of meth-
generally a marker of an inactive gene, and changes in gene ylated cytosines (5mC) by members of the ten-eleven trans-
expression often can be correlated with characteristic location (TET) protein family results in the formation of
changes in the degree of methylation of the 5′ regulatory 5-hydroxymethylcytosine (5hmC), 5-formylcytosine (5fC),
sequences of the gene. This type of epigenetic modification and 5-carboxylcytosine (5caC). While the role of these epi-
is performed by enzymes called DNA methyltransferases genetic marks is not clearly understood, their levels are often
and is associated with alterations in gene expression and altered in cancers, including AML, suggesting that they

Basic concepts | 5
may contribute to the pathogenesis of disease. Given the deacetylases (HDACs), the enzymes that remove acetyl
prominent role of abnormal methylation in hematologic groups from histone tails, are common in hematopoietic
malignancies, small molecule inhibitors of DNA methyltrans- malignancies, and HDAC inhibitors are being tested in a vari-
ferases (eg, 5-azacitidine, decitabine) are used in the treatment ety of these diseases. The first of these inhibitors, Vorinostat,
of disorders that are characterized by aberrant DNA methyla- is used in the treatment of cutaneous T-cell lymphoma.
tion (eg, myelodysplastic syndrome [MDS], AML).
Noncoding RNAs
Histone modification
It has been estimated that only approximately 1%-2% of the
Histones are DNA packaging proteins that organize DNA genome encodes protein, but a much larger fraction is tran-
into structural units called nucleosomes. Octamers of the scribed. This transcribed RNA that does not encode protein
core histones—H2A, H2B, H3, and H4—make up the nucleo- is referred to as noncoding RNA (ncRNA) and is grouped
some around which 147 bp of DNA is wrapped, and histone into an increasingly large number of different classes,
H1 binds the “linker” DNA between nucleosomes. Histones including microRNAs, small nucleolar RNAs (snoRNAs),
are subject to multiple modifications, including methylation, small interfering RNAs (siRNAs), Piwi-interacting RNAs
acetylation, ubiquitination, phosphorylation, and others. (piRNAs), long non-coding RNAs (lncRNAs), and many
The particular combination of histone modifications at any others. Although each of these classes of ncRNAs differs
given locus is thought to confer a “histone code,” regulating in size, biogenesis pathway, and specific function, they
processes such as gene expression, chromosome condensa- share a common ability to recognize target nucleotide
tion, and DNA repair. Like methylation, histone modifica- sequences through complementarity and regulate gene
tions regulate gene activity, and therefore disruptions of the expression. The most well-described class of ncRNAs
normal pattern of these modifications can contribute to can- are microRNAs (miRs), whose biogenesis pathway is illus-
cer and other diseases. For example, hypoacetylation of his- trated in Figure 1-2. Following transcription, a portion of
tones H3 and H4 is associated with silencing of the cell cycle this RNA (the pri-microRNA) forms hairpin loops that are
regulator p21WAF1, a gene whose expression is reduced in cleaved by the enzymes Drosha and Dicer into short 21- to
multiple tumor types. Aberrant expression levels of histone 23-bp double-stranded RNAs. These short double-stranded
Nucleus Cleavage
“Cropping”
AAAAA
Cytosol
Cleavage
“Dicing”
Argonaute and
Figure 1-2 MicroRNA production. Production of microRNA other proteins RISC
3’ 5’
begins with transcription of the microRNA gene to produce a
stem-loop structure called a pri-microRNA. This molecule is Extensive match Less extensive match
processed by Drosha (“cropping”) to produce the shorter
AAAAA AAAAA
pre-microRNA. The pre-microRNA is exported from the mRNA mRNA
nucleus; the cytoplasmic Dicer enzyme cleaves the pre- “Splicing”
microRNA (“dicing”) to produce a double-stranded mature
microRNA. The mature microRNA is transferred to RISC AAAAA
(RNA-induced silencing complex), where it is unwound by a ATP

helicase. Complementary base pairing between the microRNA ADP RISC released
and its target mRNA directs RISC to destroy the mRNA (if
AAAAA Translation reduced
completely complementary) or halt translation (if a mismatch
Transfer of mRNA into
exists at the scissile site). Reproduced with permission from P-bodies and eventual
Alberts B, et al. Molecular Biology of the Cell. 5th Edition. Rapid mRNA degradation degradation
New York, NY: Garland Science; 2007.

RNAs contain both sense strands and antisense strands that This fusion results from a translocation between the BCR
correspond to coding sequences in mRNAs. These mature gene on chromosome 9 and the ABL1 proto-oncogene on
miRs then are incorporated into a larger complex known as chromosome 22 and confers constitutive activation of ABL1
a RISC (RNA-induced silencing complex). The miR is then and enhanced cell proliferation. Pharmacologic targeting of
unwound in a strand-specific manner, and the single- the activity of oncogenes, such as the use of the tyrosine
stranded RNA locates mRNA targets by Watson–Crick base kinase inhibitor Imatinib to treat CML, can be an effective
pairing. Gene silencing results from cleavage of the target therapeutic approach.
mRNA (if there is complementarity at the scissile site) or
translational inhibition (if there is a mismatch at the scissile
Tumor suppressors
site). This gene-silencing pathway is known as RNA interfer-
ence. As mediators of gene expression, miRs and other In contrast to oncogenes, tumor suppressors are genes that
ncRNAs are expressed in a tissue-specific manner and play encode for proteins or ncRNAs whose normal function is to
important regulatory roles in development and differentia- inhibit tumor development through the promotion of such
tion. Accordingly, dysregulation or mutations in ncRNAs are processes as apoptosis, DNA repair, cell cycle inhibition, cell
associated with various diseases, including cancer. In hemato- adhesion, and others. Loss of the expression or function of
poiesis, numerous miRs have been identified that influence these genes is associated with cancer, and generally both cop-
cell fate decisions. miR-223, for example, plays a central ies of the tumor suppressor gene must be altered to promote
role in myeloid differentiation, and reduced expression of neoplasia. Thus, most tumor suppressors follow the “two-hit
miR-223 is common in AML. Conversely, the expression of hypothesis” proposed by Alfred Knudson in his study of
miR-125b, which is expressed in normal hematopoietic stem the retinoblastoma-associated tumor suppressor gene RB1.
cells and whose targets include genes involved in regulating This gene encodes a protein that functions to regulate cell
cell proliferation, differentiation, and survival, is increased in cycling and survival. Because both copies of the gene must be
a variety of myeloid and lymphoid malignancies. mutated for retinoblastoma to manifest, individuals that
inherit a mutant allele (requiring just one more “hit” in the
remaining normal allele for loss of gene function) generally
Molecular basis of neoplasia
develop disease earlier than those that must acquire “hits” in
Normal cellular growth and differentiation depends on the both alleles. Familial cancer syndromes often result from the
precise control of gene expression, and alterations in the quan- inheritance of heterozygous mutations in tumor suppressor
tity or timing of gene expression can affect the survival and genes. For example, Li-Fraumeni syndrome results from
function of a cell. When such alterations occur in certain types inherited mutations in the cell cycle regulator TP53 and is
of genes known as oncogenes or tumor suppressor genes, the associated with the early onset of multiple tumor types,
cell may gain abnormal growth or survival properties, and including osteosarcoma, breast cancer, leukemia, and others.
accumulations of such mutations may lead to cancer. When mutations occur in the remaining normal allele,
termed “loss of heterozygosity,” tumor growth is initiated.
In some cases, loss of just one copy of a gene (“haploin
Oncogenes
sufficiency”) has been shown to contribute to cancer devel-
Oncogenes are genes that have the potential to cause cancer, opment. For example, loss of one copy of the ribosomal
and they arise from mutations in their normal counterparts gene RPS14 in patients with 5q- syndrome leads to aberrant
termed proto-oncogenes. Proto-oncogenes generally code ribosomal protein function and a block in erythroid
for proteins or ncRNAs that regulate such processes as differentiation.
proliferation and differentiation, and activating mutations
or epigenetic modifications that increase the expression or
enhance the function of these genes confer a growth or
Analytic techniques
survival advantage to a cell. The first described oncogene,
termed SRC, was discovered in the 1970s and is a member of
Digestion, amplification, and
a family of tyrosine kinases that regulate cell proliferation,
separation of nucleic acids
motility, adhesion, survival, and differentiation. Activating
mutations in the SRC family kinases are associated with the DNA may be cut, or digested, into predictable, small frag-
pathogenesis of multiple types of neoplasias, including can- ments using restriction endonucleases. Each of these bacte-
cers of the colon, breast, blood, head and neck, and others. rially derived enzymes recognizes a specific sequence of
Another classic example of an oncogene is the BCR-ABL1 4-8 bp in double-stranded DNA. These recognition sequences
fusion gene found in chronic myelogenous leukemia (CML). are usually palindromic (ie, they read the same sequence

Analytic techniques | 7
A Restriction enzyme digestion
Enzyme Recognition sequence Digestion products Overhang
5’ GAA T T C 3’ 5’ G 3’ 5’ A A T T C 3’
EcoRI 5’
3’ C T T AAG 5’ 3’ C T T AA 5’ 3’ G 5’
5’ GAG C T C 3’ 5’ G A G C T 3’ 5’ C 3’
SacI 3’
3’ C T C GAG 5’ 3’ C 5’ 3’ T C GAG 5’
5’ C AG C T G 3’ 5’ C A G 3’ 5’ C T G 3’
PvuII Blunt
3’ G T C GA C 5’ 3’ G T C 5’ 3’ GA C 5’
B Ligation of “sticky ends”

Figure 1-3 Restriction endonuclease GAA T T C GAA T T C
digestion. (A) Diagram of typical restriction
C T T AAG C T T AAG
enzyme recognition sequences and the
pattern of cleavage seen upon digestion with Cut with EcoRI
that enzyme. (B) Means by which restriction
enzyme can be exploited to form G AA T T C G AA T TC
recombinant proteins. Digestion of the two C T T AA G C T T AA G
fragments with the enzyme EcoRI results in
four fragments. Ligation with DNA ligase DNA ligase
can regenerate the original fragments, but

GAA T T C GAA T T C
can also result in recombinant fragments in
which the 5′ end of one fragment is ligated C T T AAG C T T AAG
to the 3′ end of the second fragment.
GAA T T C GAA T T C
This recombinant DNA can then be used
as a template for generation of recombinant C T T AAG C T T AAG
protein in expression vectors.
5′ to 3′ on opposite strands). The DNA is cleaved by the sequence. Two oligonucleotide primers are required: one is
enzyme on both strands at the site of the recognition complementary to a sequence on the 5′ strand of the DNA to
sequence. After restriction endonuclease digestion, DNA be amplified and the other is complementary to the 3′ strand.
fragments may be separated by size using agarose gel electro- The DNA template is denatured at high temperature; the
phoresis, with the smallest fragments running faster (closer temperature then is lowered for the primers to be annealed
to the bottom of the gel) and the largest fragments moving to the DNA. The DNA then is extended with a temperature-
more slowly (closer to where the samples were loaded). DNA stable DNA polymerase (such as Taq polymerase), resulting
can be visualized in the gel by staining with ethidium bro- in two identical copies of the original DNA from each piece
mide, a chemical that inserts itself between the DNA strands of template DNA. The products are denatured, and the pro-
and fluoresces upon exposure to ultraviolet light. A desired cess is repeated. The primary product of this reaction is the
fragment of DNA may be isolated and then purified from the fragment of DNA bound by the two primers. Thus, small
gel. Some restriction enzymes generate overhanging single- quantities of input DNA may be used to synthesize large
stranded tails, known as “sticky ends.” Complementary over- quantities of a specific DNA sequence. This technique has
hanging segments may be used to join, or ligate, pieces of superseded many blotting techniques for prenatal diagnosis
DNA to one another (Figure 1-3). These methods form the and cancer diagnostics. Using multiple primer pairs in the
foundation of recombinant DNA technology. same reaction, multiplex PCR can efficiently amplify several
fragments simultaneously.
Reverse transcriptase PCR (RT-PCR) is a modification of
Polymerase chain reaction
the PCR technique that allows the detection and amplifica-
The polymerase chain reaction (PCR) is a powerful tech- tion of expressed RNA transcripts. Complementary DNA
nique for amplifying small quantities of DNA of known (cDNA) is generated from RNA using reverse transcriptase,

an enzyme that mediates the conversion of RNA to DNA. present in the test sample. This technique is used widely to
The resultant cDNA is then subjected to routine PCR ampli- measure minimal residual disease or to monitor clearance of
fication. Because cDNA is generated from processed mRNA BCR/ABL transcripts in patients treated with tyrosine kinase
transcripts, no intronic sequences are obtained. RNA is inhibitors.
much less stable than DNA; thus, amplification of mRNA The power of PCR lies in its great sensitivity, but this
from tissue or blood requires careful preservation of source is also a potential weakness because small amounts of
tissue or blood samples. contaminating DNA or RNA from other sources can cause
Quantitative PCR is another modification of the PCR false-positive results. Clinical laboratories that use PCR for
technique. The most commonly used method is real-time critical diagnostic tests require elaborate quality assurance
PCR, in which a fluorogenic tag is incorporated into an oli- protocols to prevent inappropriate diagnosis. Equally trou-
gonucleotide that will anneal to the internal sequence of the blesome can be false-negative results that result from inap-
Taq DNA polymerase-generated PCR product. This tag con- propriate primer design, degraded RNA, or inappropriate
sists of a fluorescent “reporter” and a “silencing” quencher temperature parameters for the annealing of primers.
dye at opposite ends of the oligonucleotide. When annealed The amplified sequence of interest then can be rapidly
to the internal sequence of the PCR product, fluorescence evaluated for presence of mutation(s) by direct sequencing,
from the reporter is quenched because the silencer is in restriction enzyme digestion (if a suitable enzyme that
proximity. After completion of each cycle of PCR amplifica- discriminates between mutant and wild-type alleles is avail-
tion, the reporter is not incorporated in the product but is able), allele-specific PCR (discussed later in this chapter), or
cleaved by Taq DNA polymerase (because this enzyme also other techniques.
has exonuclease activity). This fluorogen tag is released,
generating a fluorescent signal (Figure 1-4). Real-time PCR
Hybridization techniques
detects the number of cycles when amplification of product
is exponential and expresses this as a ratio to standard house- DNA is chemically stable in the double-stranded form.
keeping RNA, such as ribosomal RNA or glyceraldehyde-3- This tendency of nucleic acids to assume a double-stranded
phosphate dehydrogenase (GAPDH) mRNA. This number structure is the basis for the technique of nucleic acid
can be converted to the number of molecules of mRNA hybridization. If DNA is heated or chemically denatured,
A R Q
Primer Probe
Denature
Figure 1-4 Real-time PCR. (A) Sample DNA (or cDNA) is
3’ 5’
denatured, and target-specific primers are annealed to begin
R the PCR amplification (shown for one strand). An
Q
Primer Probe oligonucleotide probe complementary to a sequence within
Anneal the PCR product is included in the reaction. The probe
3’ 5’
contains a fluorophore (R) covalently attached to the 5′ end
and a quencher (Q) at the 3′ end. As the Taq polymerase
R extends the nascent strand, its 5′-to-3′ exonuclease activity
Q degrades the probe, releasing the fluorophore from the
Primer
Taq quencher and allowing the fluorophore to fluoresce.
Extend An example of this fluorescent readout is shown in panel B,
3’ 5’
which depicts the relative fluorescence intensity from
amplification of the BCR-ABL fusion transcript (yellow line)
to an endogenous control transcript (green line) in a patient
B Diagnosis 1 month 3 months with chronic myelogenous leukemia before and after
Fluorescence
treatment with a tyrosine kinase inhibitor. The cycle number

at which fluorescence crosses a threshold (horizontal dotted
line) is inversely proportional to the amount of template DNA
or cDNA. Although the control template is consistently
detected throughout therapy, the BCR-ABL transcript
Cycles abundance is lower at 1 month (higher cycle threshold) and
undetectable at 3 months into therapy.

the hydrogen bonds are disrupted, and the two strands hybridized to total cellular DNA digested with that enzyme,
separate. If the denatured DNA is then placed at a lower tem- then the detected DNA fragment will be a different size.
perature in the absence of denaturing chemicals, the single- The β-globin point mutation resulting in hemoglobin S may
stranded species will reanneal in such a way that the be detected in this way. More commonly, genetic diseases
complementary sequences are again matched and the hydro- are not the result of single base-pair mutations that conve-
gen bonds reform. If the denatured DNA is incubated with niently abolish or create restriction enzyme sites. A similar
radioisotope- or fluorogen-labeled, single-stranded comple- technique may be used, however, to detect the presence of an
mentary DNA or RNA, the radiolabeled species will anneal RFLP that is linked to a disease locus within a family or
to the denatured, unlabeled strands. This hybridization pro- group, but that does not directly detect the molecular abnor-
cess can be used to determine the presence and abundance of mality responsible for the disease. This is because there are
an identical DNA species. The technique of molecular normal variations in the DNA sequence among individuals
hybridization is the basis for Southern blotting and many that are inherited but silent in that they do not cause disease.
other molecular techniques. These polymorphisms may be located in intronic sequences
Southern blotting (Figure 1-5) is used to detect specific or near the gene of interest. They are surrogates that can be
DNA sequences. In this procedure, electrophoretically- used to identify the region of DNA containing the genetic
separated DNA fragments are separated to a filter membrane variant in question. Because RFLPs are transmitted from
and subsequent fragment detection by probe hybridization. parent to offspring, they are extremely useful in the diagnosis
Restriction fragment-length polymorphism (RFLP) analy- of many genetic diseases.
sis is a Southern blot-based technique with many useful Hybridization techniques also can be applied to RNA.
applications in hematology. Using this technique, inherited Although RNA is generally an unstable single-stranded species,
disease-associated alleles may be identified and traced by the it is stabilized when converted to the double-stranded form.
presence of inherited mutations or variations in a DNA Therefore, if placed under hybridization conditions, RNA will
sequence that create or abolish restriction sites. Rarely, a complex with complementary, single-stranded nucleic acid
single-base, disease-causing DNA mutation will coinciden- species in the same fashion as DNA. Northern blotting is anal-
tally fall within a recognition sequence for a restriction ogous to Southern blotting, but it involves electrophoresis of
endonuclease. If a probe for the mutated fragment of DNA is RNA with subsequent transfer and hybridization to a probe.
A
Cleave with Transfer to
restriction enzyme nitrocellulose Hybridize to probe
x
Probe
DNA B A
Electrophorese x
Autoradiograph
on agarose gel
B Detection of protein
Separate proteins using immunostain
by size using Transfer proteins with antibody specific
electrophoresis to membrane to protein of interest
Protein
extracts
Autoradiograph
Figure 1-5 Common hybridization techniques in molecular biology. (A) Southern blot analysis of DNA. DNA is cleaved with a restriction
endonuclease, electrophoresed through an agarose gel, and transferred to nitrocellulose. The probe, as illustrated at the bottom of the figure, lies
on a piece of DNA of length x when DNA is digested with an enzyme that cleaves at sites ‘A’ and ‘B.’ Hybridization of the probe to the blot, with
appropriate washes and exposure to radiograph, show a single band of length x on the autoradiogram. (B) Western blot detection of protein.
Proteins are extracted from cells and then separated by gel electrophoresis to separate proteins based on length of denature polypeptide
(or occasionally based on 3D structure of native proteins). The separated proteins are then transferred to a membrane where they are stained
with antibodies to detect a protein of interest. Detection antibodies for Western blot are commonly conjugated to an enzyme which allows for
detection of the protein of interest in the membrane using a variety of methodologies such as colorimetric or chemiluminescent detection.

Whereas Southern blotting detects the presence of a gene or the two genes are juxtaposed, the two probes are in proxim-
its integrity, Northern blot analysis detects the level of expres- ity, and the fluorescent signals merge to generate a yellow
sion of a gene within a specific cell type. signal. The specificity of FISH is highly dependent on the
Protein can be detected by the blotting technique referred probes that are used. Numeric abnormalities, such as mono-
to as Western blotting (Figure 1-5). Proteins are detected by somy and trisomy, may be identified using centromere-
specific antibodies directed against the protein of interest. A specific probes.
labeled anti-immunoglobulin antibody raised in another The major advantage of FISH is that it can analyze known
species then can be used to detect the specific antibody cytogenetic abnormalities in nondividing cells (interphase
bound to the blot. nuclei); thus, peripheral blood slides can be directly pro-
cessed. FISH studies are most useful when assessing for the
presence of specific molecular abnormalities associated with
Cytogenetic techniques
a particular clinical syndrome or tumor type and are approx-
Uniform, nonrandom chromosomal abnormalities, termed imately 1 order of magnitude more sensitive than morphol-
clonal abnormalities, can be detected in malignant cellular ogy and conventional cytogenetic studies in detecting
populations by metaphase cytogenetics, or chromosomal residual disease. FISH panels are now available to detect
analysis. Conventional cytogenetic techniques can detect recurrent genetic changes in leukemias, lymphomas, and
numeric chromosomal abnormalities (too many or too few multiple myeloma. These panels are particularly useful in
chromosomes) as well as deletion or translocation of rela- predicting prognosis when conventional cytogenetic studies
tively large chromosomal fragments among chromosomes. are noninformative.
Certain chromosomal translocations are considered pathog- Since their introduction nearly 30 years ago, FISH tech-
nomonic of specific diseases, such as the t(15;17) in acute niques have evolved rapidly for use in hematologic disorders.
promyelocytic leukemia (APL). Normally, chromosomes For example, double-fusion FISH (D-FISH) uses differen-
cannot be seen with a light microscope, but during cell divi- tially labeled large probes that each span one of the two
sion, they become condensed and can be analyzed. To collect translocation breakpoints. This allows simultaneous visual-
cells with their chromosomes in this condensed state, bone ization of both fusion products and reduces false-negative
marrow or tumor tissue may be briefly maintained in culture results. Another technique known as break-apart FISH uses
and then exposed to a mitotic inhibitor, which blocks forma- differentially labeled probes targeting the regions flanking
tion of the spindle and arrests cell division at the metaphase the breakpoint. Thus, in normal cells, the signals appear
stage. Thus, cytogenetic studies require dividing cells. fused, but they split upon translocation. This technique has
Conventional cytogenetic studies have several limitations. been used to detect MYC translocations in Burkitt lym-
First, these studies require active cell division, which may not phoma and CCND1 translocations in mantle cell lymphoma.
be feasible for some clinical samples. Second, the technique is Labeling probes with unique combinations of fluorophores
insensitive to submicroscopic abnormalities. Finally, because in multiplex FISH (M-FISH) has not only permitted simul-
only a very small number of cells are analyzed, the technique taneous detection of every chromosome but also now has
is relatively insensitive for measurement of minimal residual been used to analyze specific chromosomal regions and can
disease burden. detect subtle rearrangements.
Fluorescence in situ hybridization (FISH) studies com-
plement conventional cytogenetic analysis by adding conve-
Array-based techniques
nience, specificity, and sensitivity. This technique applies the
principles of complementary DNA hybridization. A specific DNA microarrays are composed of oligonucleotide probes
single-stranded DNA probe corresponding to a gene or spanning sites of known single nucleotide polymorphisms
chromosomal region of interest is labeled for fluorescent (SNPs). Fluorescently labeled single-stranded DNA from a
detection. One or more probes are then incubated with the test sample is hybridized on the array to determine, for a spe-
fixed cellular sample and examined by fluorescence micros- cific region in the genome, which DNA sequence undergoes
copy. FISH probes have been developed that can identify complementary base pairing with the sample. The pattern of
specific disease-defining translocations, such as the t(15;17) hybridization signals is analyzed using computer software,
that characterizes APL. A probe corresponding to the PML providing a detailed profile of genetic variation specific to
gene on chromosome 15 is labeled with a fluorescent marker, an individual’s DNA. With current technology, a single
such as rhodamine, which is red. Another fluorescent microarray has sufficient density to analyze variation at
marker, such as fluorescein, which is green, is linked to a >1 million polymorphic sites. These data can be analyzed
probe corresponding to the RARa gene on chromosome 17. in several ways. First, the genotypes at each site can be used
When the t(15;17) chromosomal translocation is present, in genome-wide association studies (GWAS) in which the

allele frequencies at each SNP are compared in disease cases of class labels. Hierarchical clustering and self-organizing
and unaffected controls. Second, the intensity of fluorescent maps are two commonly used algorithms of unsupervised
signals from multiple adjacent sites can be used to infer learning. One potential application of unsupervised learning
changes in the abundance of DNA across the genome. is for discovery of previously unrecognized disease subtypes.
Changes in DNA content may include inherited copy num- The strength of this method is that it provides an unbiased
ber variants or somatically acquired deletions and amplifi- approach to identifying classes within a data set. A weakness
cations present in tumor samples. Finally, long stretches of is that these data sets are complex and the structure uncov-
homozygosity that reflect acquired partial uniparental ered by clustering may not reflect the underlying biology of
disomy (pUPD), a recurrent abnormality present in a variety interest. The second computational approach, supervised
of myeloid malignancies, can be identified. In a variation of learning, uses known class labels to create a model for class
this technique, the relative abundance of methylated versus prediction. For example, a training data set is used to create
unmethylated DNA can be detected in samples by pretreat- an expression profile for tumor samples from patients with
ing DNA with chemicals (eg, bisulfite) that convert methyl- “cured” versus “relapsed” disease. These profiles then are
ated cytosine bases before hybridization on an array. applied to an independent data set to validate the ability to
In comparative genomic hybridization (CGH), DNA make the prognostic distinction. In either method, it is
extracted from a test sample (eg, tumor) and a matched nor- important to demonstrate statistical significance and ensure
mal control (eg, buccal wash) is differentially labeled and that the tested samples are compared with the appropriate
hybridized to a microarray composed of oligonucleotide controls.
probes. The ratio of test to control fluorescence is quantified Once differentially expressed genes are defined, then it is
using digital image analysis. Similar to SNP arrays, amplifi- also often helpful to next determine whether the differen-
cations in the test DNA are identified as regions of increased tially expressed genes identified map to specific pathways of
fluorescence ratio, and losses are identified as areas of known biological significance. This is commonly done
decreased ratio. In array CGH, resolution of the analysis is through the use of the GO (Gene Ontology) or KEGG (Kyoto
restricted by probe size and the density of probes on the Encyclopedia of Genes and Genomes) Pathway analysis
array. These and other techniques permit high-resolution, database. Also, a statistical methodology known as GSEA
genomewide detection of genomic copy number changes. (Gene Set Enrichment Analysis) is also commonly used to
Careful analysis of AML genomes using these approaches determine if the differentially expressed genes are statistically
has revealed few somatic copy number changes that are not enriched in previously published and defined gene sets pub-
detectable by routine cytogenetics. In contrast, acute lym- lically deposited in prior microarray or RNA-seq datasets.
phoblastic leukemia (ALL) genomes are characterized by The main limitation of microarray expression technology
recurring copy number alterations, frequently involving the is that it analyzes only mRNA abundance. It does not reveal
loss of the genes required for normal lymphoid development important translational and posttranslational modifications
(eg, PAX5, IKZF1). and protein–protein interactions. Purity of the cell popula-
RNA expression arrays allow for comprehensive char- tion is also essential for these analyses, and one must ensure
acterization of the gene expression patterns within the cells that the control and analyzed cells are homogeneous, of the
of interest, referred to as a gene expression profile. This same cell type, and at comparable stages of differentiation.
technique has been used to classify disease, predict response It is advisable that any significant difference in mRNA
to therapy, and dissect pathways of disease pathogenesis. expression detected using microarray technology be con-
To perform these assays, mRNA is extracted from samples, firmed using an orthogonal approach (eg, real-time PCR).
and double-stranded cDNA is synthesized from the RNA
template. Then, biotinylated complementary RNA (cRNA)
Sequence-based studies
is generated from the cDNA template by in vitro transcrip-
tion using biotin-labeled nucleotides. The biotinylated Analysis of DNA sequence variation by conventional tech-
cRNA is fragmented and incubated with probes in a solu- niques (eg, Sanger sequencing) is being replaced rapidly by a
tion or hybridized to a microarray. Hybridization is then variety of novel high-throughput technologies (collectively
detected using a streptavidin-phycoerythrin stain, and the termed next-generation [next-gen] sequencing). These deve
fluorescence intensity of each feature of the array is lopments have greatly accelerated the pace and lowered the
quantified. cost of large-scale sequence production. At the core of each of
Two main computational approaches have been used to these technologies is the preparation of DNA fragment librar-
analyze microarray data: unsupervised and supervised learn- ies, which are then clonally amplified and sequenced by synthe-
ing. Unsupervised learning methods cluster samples based sis in multiple parallel reactions (Figure 1-6). Sequencing both
on gene expression similarities without a priori knowledge ends of the DNA templates (“paired-end reads”) improves the

A
Exome capture probes Hybridization of probes
and library fragments
Random
next-generation
sequencing library
Biotinylated probes Genomic DNA
Bind hybrids
to streptavidin
magnetic beads
Off-target capture
Apply magnetic field
Wash and elute
Amplify recovered library fragments and sequence
Normal
Tumor
Figure 1-6 Next-gen sequencing. (A) A library of genomic DNA fragments is shown at the upper right. This library can be used directly for
sequencing (whole genome), or can be hybridized with probes (shown in upper left) to enrich for targets in candidate genes or all coding genes
(whole exome). Reproduced with permission from Mardis E. Nat Rev Gastroenterol Hepatol. 2012;9:477-486. (B) Sequence reads from a tumor
sample (bottom panel) and matched normal tissue (upper panel) visualized in the Integrated Genome Viewer (Robinson JT, et al. Nat Biotechnol.
2011;29:24-26.). Each grey bar represents a single 75- to 100-bp sequence read. A nucleotide substitution shown in red at the position indicated by
the arrow is present in ~50% of tumor reads, suggesting that a heterozygous somatic mutation is present in the tumor sample.
efficiency of data production and facilitates the identification that encode for proteins (the exome) can be enriched by
of insertions, deletions, and translocations. With these hybridizing DNA to oligonucleotide probes before library
approaches, the search for inherited and somatic mutations construction. Exome sequencing is preferred for many
associated with hematologic malignancies and congenital studies (compared with whole genome sequencing) because
blood disorders has evolved from a candidate gene approach to the cost of sequence production is lower and the interpreta-
unbiased surveys of all coding and noncoding regions of the tion of sequence variants in protein-coding genes is more
genome. tractable. A limitation of exome sequencing is that it cannot
The DNA libraries used for these sequencing reactions detect structural variants (such as deletions, amplifications,
can be prepared from whole genomes or from selected and rearrangements). In addition, the use of DNA from tis-
regions of interest. For example, the regions of the genome sue uninvolved by the disease process is very important for

whole exome or genome sequencing to identify potential desorption/ionization time of flight (MALDI-TOF) mass
somatic alterations specific for the disease tissue given the spectrometry, the time of flight of the ions is detected and
large number of sequence variants that are often generated used to calculate a mass-to-charge ratio. The spectrum of
by sequencing the entire exome or genome. Also, some pro- mass-to-charge ratios present within a sample reflects the
tein-coding genes are not yet annotated and therefore will protein constituents within the sample. Supervised or unsu-
be excluded from commercially available reagents used to pervised learning approaches, as described previously, are
capture the target DNA. These assays can be restricted fur- then used to identify patterns within the data. More recently,
ther to panels of genes by first amplifying the genes of inter- protein microarrays have been developed. Analytical pro-
est (by PCR) or by hybridizing the DNA to oligonucleotide tein microarrays are composed of a high density of affinity
probes covering the region of interest, followed by next-gen reagents (eg, antigens, antibodies) that can be used to
sequencing. This is an efficient and cost-effective approach detect the presence of specific proteins in a mixture. Func-
to interrogate a number of targets in parallel from a single tional protein microarrays contain a large number of
sample (eg, genes that are recurrently mutated in hemato- immobilized proteins; these arrays can be used to examine
logic malignancies). Further modifications of the workflow protein–protein, protein–lipid, protein–nucleic acid, and
allow for detection of chromatin marks across the genome enzyme–substrate interactions. Although all of these tech-
(eg, transcription factor binding sites, histone modifica- nologies hold enormous potential, clinical applications have
tions) by using antibodies to immunoprecipitate the region yet to be realized.
of interest, followed by next-gen sequencing (ChIP-Seq).
Next-generation sequencing-based assays can also be used
Animal models
to identify various epigenetic marks. For example, bisulfite
sequencing is used to quantify the extent of cytosine meth- Analysis of both inherited and acquired diseases by reverse
ylation across the genome. In this method, sodium bisulfite genetics has resulted in the identification of many disease-
exposure converts unmethylated cytosines to uracil, leaving related genes for which the function is unknown. Once a
methylated cytosines unaffected. Finally, RNA templates disease-related gene has been identified, either by linkage
can be used to generate DNA libraries for sequencing mapping (eg, the gene for cystic fibrosis) or by identifying
(RNA-seq), allowing for the quantification of RNA rearranged genes (eg, the BCR gene at the breakpoint of the
abundance and for the detection of chimeric RNAs or alter- Philadelphia chromosome), the challenge lies in identifying
natively spliced products. the function of the protein encoded by that gene and char-
Although the cost of sequence production has fallen acterizing how changes in the gene can contribute to the
dramatically in recent years, the storage, analysis, and disease phenotype. Understanding the role of these genes
interpretation of these large data sets still pose significant and their encoded proteins has been aided greatly by the
challenges. development of techniques to alter or introduce these genes
in mice using recombinant DNA technology.
Mice can be produced that express an exogenous gene
Proteomics
and thereby provide an in vivo model of the gene’s function.
Proteins are the effectors of most cellular functions. Genetic Linearized DNA is injected into a fertilized mouse oocyte
defects perturb normal cellular functions because they result pronucleus, and the oocyte is then reimplanted into a pseu-
in changes in the level or the function of the proteins they dopregnant mouse. The resultant transgenic mice then can
encode. Many proteins undergo extensive posttranslational be analyzed for phenotypes induced by the exogenous gene.
modifications that influence their activity and function, Placing the gene under the control of a strong constitutive
including cleavage, chemical modification such as phos- promoter, which is active in all tissues, allows for the assess-
phorylation and glycosylation, and interaction with other ment of the effect of widespread overexpression of the gene.
proteins. These posttranslational events are not encoded by Alternatively, placing the gene under a tissue-specific pro-
the genome and are not revealed by genomic analysis or gene moter will elucidate the function of that gene in an isolated
expression profiling. Proteomics is the systematic study tissue. A third approach is to use the control elements of the
of the entire complement of proteins derived from a cell gene to drive the expression of a gene that can be detected by
population. chemical, immunologic, or functional means. For example,
Proteomic analysis relies on complex bioinformatic tools the promoter region of a gene can be joined to the green
applied to mass spectroscopy data. In general, these tech- fluorescent protein (GFP) cDNA, and expression of this
niques require some sort of separation of peptides, usually reporter can be assessed in various tissues in the resultant
by liquid chromatography, followed by ionization of the transgenic mouse. Use of such a reporter gene will show the
sample and mass spectrometry. In matrix-associated laser normal distribution and timing of the expression of the gene

from which the promoter elements are derived. These trans- somatic cells reprogrammed to become totipotent cells may
genic mice contain multiple copies of exogenous genes that overcome some of these obstacles. These cells, called induced
have inserted randomly into the genome of the recipient and pluripotent stem (iPS) cells, are produced by reprogram-
thus may not mimic physiologic levels or spatiotemporal ming adult somatic cells to become embryonic-like cells,
expression of the gene. In contrast, the endogenous genetic which, in turn, can be further differentiated along specific
locus of a gene can be manipulated in totipotent embryonic lineages. The concrete demonstration that iPS cells may be
stem (ES) cells by targeted recombination between the locus used to treat disease was replacement of the sickle globin
and a plasmid carrying an altered version of that gene that gene with a normal β-globin gene in mice. Corrected iPS
changes or disrupts its function. If a plasmid contains that cells from sickle mice were differentiated into hematopoietic
altered gene with enough flanking DNA identical to that of progenitors in vitro, and these cells were transplanted into
the normal gene locus, homologous recombination will irradiated sickle mice recipients. Erythroid cells derived from
occur at a low rate; however, cells undergoing the desired these progenitors synthesized high levels of human hemo-
recombination can be enriched by including a selection globin A and corrected the sickle cell disease phenotype.
marker in the plasmid, such as the neomycin resistance gene. Human iPS cells have been produced and hold great pro
The correctly targeted ES cell is then introduced into the mise as research tools and possibly as a source of tissue
blastocyst of a developing embryo. The resultant animals replacement.
will be chimeric, in that only some of the cells in the animal Given the time required for conventional gene targeting
will contain the targeted gene. If the new gene becomes part using homologous recombination, there has been great
of the germline, offspring can be bred to yield mice carrying interest in the recent development of genome editing using
the mutation in all cells. Knockout mice (homozygous for a zinc-finger nucleases, transcription activator-like effector
null allele) can illuminate the function of the targeted gene nucleases (TALENs), and CRISPR/Cas (clustered regulatory
by analyzing the phenotype of mice that lack the gene pro interspaced short palindromic repeat/Cas-based RNA-
duct. Similar approaches can be used to replace a normal guided DNA endonucleases). Each of these techniques takes
mouse gene in ES cells with a version containing a point use of a nuclease that induces DNA breaks and then stimu-
mutation, deletion, or other genetic variant to model abnor- lates DNA repair in a way that allows for creation of specific
malities detected in patients with hematologic disorders. mutations and/or inclusions of novel sequences of DNA. The
Many genes of interest participate in pathways that are nucleases are linked to sequence-specific DNA binding mod-
vital for viability or fertility; thus, constitutive knockout ules that allow for creation of mutations in specific locations
mice cannot be generated. Conditional gene modification of the genome. These techniques have allowed for rapid gen-
using Cre-loxP technology allows the gene of interest to be eration of knockout and knockin mice in embryonic or
altered in specific tissues or at specific times during develop- somatic stem cells to rapidly create genetically engineered
ment or postnatal life. This is accomplished by inserting the animal models.
altered gene with flanking DNA containing loxP sites. If mice
with paired loxP sites integrated into their genome are bred
with a second strain of mice that expresses an enzyme called Clinical applications of DNA
Cre recombinase, recombination will take place between the technology in hematology
loxP sites, removing or rearranging the desired portion of the
gene. Furthermore, expression of the Cre recombinase can Molecular biology has revolutionized the understanding of
be regulated in a tissue-specific manner by using an appro- molecular pathogenesis of disease in ways that have pro-
priate promoter or in a temporally restricted manner by foundly affected the diagnostic armamentarium of the
using a promoter that is induced by treatment of the mice hematologist. Several examples of how molecular studies are
with a drug (such as tetracycline). The use of transgenic, used for diagnosis and clinical decision-making in hematol-
knockout, and conditional knockout mice has been invalu- ogy are described in this section.
able in elucidating the function of large numbers of genes
implicated in the pathogenesis of both inherited and acqu
Applications to germline
ired diseases.
(inherited) mutations
Transgenic technology, however, is laborious, time con-
Hemoglobinopathies and thalassemias
suming, and expensive. Some of these disadvantages can be
circumvented by using rapidly reproducing and inexpensive One of the best examples of the use of molecular techniques
organisms, such as zebrafish or yeast. Like transgenic mice, in benign hematology is in the diagnosis of hemoglobinopa-
however, these models may not recapitulate human-specific thies and thalassemia. Although the most common hemo-
pathophysiology. Newer technology using dedifferentiated globin variants (ie, Hb S, Hb C, Hb D) typically are diagnosed

Clinical applications of DNA technology in hematology | 15
using nonmolecular methods, such as high-performance liq- resulting in an extensive repertoire of composite genes that
uid chromatography (HPLC) or protein electrophoresis, creates immense immunoglobulin and T-cell diversity. These
molecular testing can be useful in several settings, including somatic rearrangements result in deletion of intervening
the characterization of uncommon variants, family screen- DNA sequences between gene segments in the immunoglob-
ing studies, and prenatal diagnosis. Hemoglobin variants ulin and T-cell receptor loci. The details of this process in
may be detected by a variety of techniques, including PCR lymphocyte ontogeny are further outlined in Chapter 21.
using allele-specific primers designed to detect specific Rearrangements in immunoglobulin and T-cell receptor
mutations or sequencing studies of the HBA1/A2 and HBB genes can be detected by either Southern blotting or PCR;
loci. Molecular techniques are particularly valuable in the however, PCR-based approaches using standardized, com-
diagnosis of α-thalassemia, which usually is caused by one of prehensive primer sets such as those developed by the Euro-
several variably sized deletions that result in the loss of one Clonality consortium (so called BIOMED-2 primers) are
or both HBA genes in the α-globin locus. In the neonatal now preferentially used in the clinical setting due to the fact
period, α-thalassemia may be recognized by the presence of that they are more rapid, require less DNA, and can be per-
Hb Barts (4 tetramers) on electrophoresis or HPLC, but lab- formed on archived formalin-fixed, paraffin embedded
oratory diagnosis after the neonatal period requires molecu- (FFPE) tissue. PCR-based techniques targeting IGH and IGK
lar techniques. Deletions of the α-globin locus can be loci for B-cell rearrangements and TCRG and TCRB for
detected by gap-PCR, which uses PCR primers that bind to T-cell rearrangements are used to confirm the presence of
either side of a deletion breakpoint. In the absence of the clonal lymphocyte populations in the peripheral blood, such
corresponding deletion, the primers are too far apart to yield as in T-cell large granular lymphocyte disorders, and also are
an amplifiable product. When a deletion is present, however, powerful ancillary techniques for hematopathologists in the
an abnormal amplicon is detected. diagnosis of lymphoproliferative disorders from FFPE tissue
(Figure 1-7). Despite their power, molecular clonality studies
should be carefully interpreted in the context of the clinical,
Pharmacogenomics
morphologic, and immunophenotypic diagnosis. Clonal pro-
Pharmacogenomics is the study of how inherited genetic liferations may occur in some reactive conditions as well as in
variation affects the body’s response to drugs. The term malignant neoplasms. For example, clonal T-cell popula-
comes from the words pharmacology and genomics and is tions may be detected in the setting of viral infections, such
thus the intersection of both disciplines. For instance, germ- as with Epstein-Barr virus or cytomegalovirus, and clonal
line polymorphisms in the thiopurine methyltransferase B-cell populations may be detected in some benign lym-
(TPMT) gene result in loss of functional protein and predis- phoid proliferations, such as marked follicular hyperplasia.
pose ALL patients to severe hematologic toxicity unless the Furthermore, false-positive PCR results may occur in several
dose of mercaptopurine is reduced by 90%-95% of normal. circumstances; for example, when very small tissue samples
PCR-based studies may be performed to identify the presence are used, a few reactive T cells in the sample might result in
of alleles associated with decreased TPMT function. the appearance of oligoclonal bands. False-negative results
may occur as a result of tissue sampling, poor PCR amplifi-
cation, or lack of detection of specific rearrangements using
Applications to somatic (acquired)
standardized primer sets.
molecular abnormalities
The power of molecular biology to provide important

Identification of cryptic translocation in pediatric acute
insights into the basic biology of disease is perhaps most
lymphoblastic leukemia: prognostic significance
dramatically shown by the evolving concepts of malignancy.
Several examples of how molecular techniques have enhanced As many as 20% of children with pre–B-cell ALL have a
our understanding of the pathogenesis of hematologic malig- translocation that fuses the RUNX1 gene on chromosome
nancies, as well as their diagnosis and treatment, are pro- 21 with the ETV6 gene on chromosome 12 in the t(12;21)
vided in the following sections. translocation. Both of these genes encode transcription fac-
tors that regulate hematopoietic differentiation. Although
this translocation is common, it is rarely seen by standard
Gene rearrangement studies in lymphoproliferative
cytogenetic techniques. It can, however, be routinely identi-
disease: T-cell and B-cell rearrangements
fied with FISH or RT-PCR using specific ETV6-RUNX1
During the development of a mature lymphoid cell from probes. The identification of this translocation is important
an undifferentiated stem cell, somatic rearrangements of because it has been associated with a favorable outcome in
the immunoglobulin and T-cell receptor loci take place, pediatric ALL.

A VH DH JH
6 VH-FR1 primers JH primer
CTGTGCAAGAGCGGGCTATGGTTCAGGGAGTTATGGCTACTACGGTATGGACGTCTGG
CTGTGCAAGAGGACGAAACAGTAACTGCCTACTACTACTACGGTATGGACGTCTGG
CTGTGCAAGAGAGATAGTATAGCAGCTCGTACAACTGGTTCGACTCCTGG
B Heteroduplex analysis CTGTGCAAGAAGATCCGGGCAGCTCGTTTTGCTTTTGATATCTGG
Monoclonal
Monoclonal
CTGTGCAAGAGCCTCTCTCCACTGGGATGGGGGGCTACTGG
Polyclonal
Monoclonal cells CTGTGCAAGAGCAGCAGCTCGGCCCCCTTTGACATACTGG
Monoclonal in polyclonal Polyclonal CTGTGCAAGAGGACTTTGGATGCTTTGATATCTGG
cells background cells CTGTGCAAGAGGGTGGGAGCTACTAGACTACTGG
MW
H2O
CTGTGCAAGGGTAGCTAAACCTTTGACTACTGG
CTGTGCAATATCTACTTTGACTACTGG
Heteroduplexes C GeneScanning
Polyclonal
Denaturation (94°C)/renaturation (4°C)
Homoduplexes Monoclonal
Figure 1-7 Schematic diagram of heteroduplex analysis and GeneScanning of PCR products, obtained from rearranged Ig and TCR genes.
(A) Rearranged Ig and TCR genes (IGH in the example) show heterogeneous junctional regions with respect to size and nucleotide composition.
Germline nucleotides of V, D, and J gene segments are given in large capitals and randomly inserted nucleotides in small capitals. The junctional
region heterogeneity is employed in heteroduplex analysis (size and composition) and GeneScanning (size only) to discriminate between
products derived from monoclonal and polyclonal lymphoid cell populations. (B) In heteroduplex analysis, PCR products are heat denatured
(5 min, 94°C) and subsequently rapidly cooled (1 h, 4°C) to induce duplex (homo- or heteroduplex) formation. In cell samples consisting of
clonal lymphoid cells, the PCR products of rearranged IGH genes give rise to homoduplexes after denaturation and renaturation, whereas in
samples that contain polyclonal lymphoid cell populations the single-strand PCR fragments will mainly form heteroduplexes, which result in a
background smear of slowly migrating fragments upon electrophoresis. (C) In GeneScanning fluorochrome-labeled PCR products of rearranged
IGH genes are denatured prior to high-resolution fragment analysis of the resulting single-stranded fragments. Monoclonal cell samples will give
rise to PCR products of identical size (single peak), whereas in polyclonal samples many different IGH PCR products will be formed, which show
a characteristic Gaussian size distribution. Reprinted by permission from Macmillan Publishers Ltd: van Dongen J, et al. Leukemia.
2003;17:2257-2317.
Prognostically significant mutations in normal karyotype have suggested that AML with mutated NPM1 or mutated
acute myeloid leukemia CEBPA each represent distinct clinicopathologic entities.
Mutations in DNMT3A have emerged as powerful predictors
Up to 40% of AML cases have no chromosomal abnormalities
of poor prognosis in AML using multivariate analysis.
visible by conventional karyotyping. The prognosis in these
Currently, a limited number of genes are routinely tested in
cases can be further refined by molecular testing for mutations
AML patients by conventional (Sanger) sequencing or PCR
in various recurrently-mutated genes including NPM1, FLT3,
assays. The use of next-gen sequencing panels may allow for
CEBPA, DNMT3A, and others. NPM1 mutations, usually a
improved prognostic assessment and treatment selection
4-bp insertion in exon 12, are found in approximately 35% of
based on testing a larger number of recurring mutations
cases of AML. FLT3 mutations include variably sized duplica-
in myeloid neoplasms, including AML, myelodysplastic
tions (internal tandem duplications [ITD]) or point muta-
syndromes, and myeloproliferative neoplasms.
tions in the kinase domain and are found in approximately
one-third of AML cases. Mutations in CEBPA are diverse and
Minimal residual disease (MRD) monitoring
can be found in approximately 10% of AML cases. Cases of
AML with a normal karyotype and a mutant NPM1/wild-type The development of PCR has markedly increased the sensi-
FLT3 genotype or harboring biallelic CEBPA mutations are tivity of tests available for the monitoring of minimal resid-
associated with a favorable prognosis. Furthermore, studies ual disease in myeloid and lymphoid neoplasms. With the

Clinical applications of DNA technology in hematology | 17
availability of real-time PCR, the relative abundance of spe- Expression profiling: applications to
cific transcripts can now be monitored to assess trends of diagnosis and treatment
increase or decrease over time. For example, real-time quan-
Gene expression microarray studies have yielded important
titative RT-PCR is used routinely in CML to risk stratify
new insights into leukemia pathogenesis and may facilitate
patients based on transcript quantity rather than simply the
the development of therapeutic interventions. Early studies
presence or absence of a transcript (as discussed previously
demonstrated that the distinction between ALL and AML
in this chapter). The accuracy and reliability of real-time
could be made with 100% accuracy on the basis of gene
quantitative PCR as a measure of BCR-ABL1 transcript level
expression profiles alone. Later studies demonstrated that
depends on the quality control procedures carried out by the
distinct molecular subtypes of ALL and AML could be dis-
laboratory. Normalization of the results to an appropriate
criminated based on expression profiling. For example, the
control gene is required to compensate for variations in RNA
application of expression microarray technology to the eval-
quality and the efficiency of the reverse transcriptase reac-
uation of infant ALL with rearrangement of the MLL gene
tion. BCR and ABL1 have been used as control genes, and
located at chromosome 11q23 demonstrated a molecular
both seem to be suitable because they are expressed at low
signature with features of both myeloid and lymphoid lin-
level and have similar stability to BCR-ABL1. The introduc-
eages. In addition, this work identified the FLT-3 gene as one
tion of internationally recognized reference standards now
of the genes distinguishing infants with MLL-associated
has allowed for reporting of results on the International
ALL from those with non-MLL-rearranged, pre-B-cell ALL.
Scale, which allows for direct comparisons of results among
Microarray technology also may accurately delineate
laboratories, even those using different control genes.
subtypes of a heterogeneous disorder, such as pediatric ALL.
A major molecular response to Imatinib has been defined as
For example, hierarchical clustering has been used to distin-
a 3-log reduction in BCR-ABL1 transcripts (BCR-ABL1/
guish pediatric ALL subtypes: E2A-PBX, MLL, T-ALL, hyper-
reference gene) compared with a standardized baseline
diploid, BCR-ABL1, and ETV6-RUNX1.
obtained from patients with untreated newly diagnosed
Microarray studies are facilitating the classification of
CML, corresponding to 0.1% on the International Scale.
lymphomas and outcome prediction for specific patient
In similar fashion, PCR analysis of immunoglobulin or
populations with this disease. For example, expression pro-
T-cell receptor gene rearrangements allow the detection of
filing was used to create a prediction model that identified
residual disease in the blood or bone marrow of patients
two categories of patients with diffuse large B-cell lymphoma
who have undergone treatment of a lymphoid malignancy.
(DLBCL): germinal center and activated B-cell types, which
Because each gene rearrangement is unique, however, the
carry favorable and unfavorable prognosis, respectively.
PCR detection of gene rearrangements at this level of sensi-
In addition, expression profiling studies have identified
tivity is labor intensive. PCR of tumor tissue is performed
genes overexpressed in patients with poor prognosis, some
using primers based on consensus sequences shared by the
of which may represent potential therapeutic targets.
variable and joining regions of the appropriate locus
(immunoglobulin or T-cell receptor genes). The specific
rearrangement must then be sequenced so that an oligo-
Applications to stem cell transplantation
nucleotide specific to the unique rearrangement in that
Human leukocyte antigen typing
patient’s tumor can be synthesized. PCR can then be per-
for stem cell transplantation
formed using this allele-specific oligonucleotide, with
adequate sensitivity to detect 1 in 105 cells. As these assays Molecular techniques have been important to the further
become increasingly available, they will play an important understanding of the diversity of human leukocyte antigen
role in estimating prognosis and determining eligibility (HLA) genotypes. Serologic testing for HLA antigens often
for autologous transplantation and other therapeutic identifies broad groups of cross-reactive antigens. Because
modalities. there is an increased incidence of severe graft-versus-host
In addition to MRD monitoring by molecular monitor- disease (GVHD) in patients who receive transplantations
ing of DNA/cDNA retrieved from hematopoietic cells, flow from serologically compatible but genotypically incompat-
cytometric analysis to detect residual leukemic cells in ible unrelated donors, it is important to identify the
peripheral blood or bone marrow has also been heavily individual antigens within these cross-reactive groups.
used for MRD monitoring in acute lymphoblastic leukemia Genotypic HLA typing can be achieved by PCR amplifica-
(amongst other hematological malignancies). Further tion of the HLA locus followed by hybridization to specific
efforts to understand the sensitivity of flow cytometric oligonucleotides corresponding to the different alleles
MRD detection relative to molecular techniques are within a given cross-reactive group. Such genotyping is
underway. much more predictive of successful transplantation and the

risk of GVHD than serologic study or the mixed lympho- inhibition of ETV6-PDGFRB expression and inhibited
cyte assay, and it has supplanted these assays for the identi- proliferation of ETV6-PDGFRB–transformed cells. When
fication of optimal donors, especially unrelated donors. applied to mice, this strategy slowed tumor development and
Comprehensive genotyping using SNP arrays may improve death in mice injected with these cells compared with cells
HLA matching. This is discussed in detail in Chapter 12. not containing the siRNA. Stable siRNA expression sensi-
tized transformed cells to the PDGFRB inhibitor Imatinib,
suggesting that stable expression of siRNAs, which target
Analysis of bone marrow engraftment
oncogenic fusion genes, may potentiate the effects of con-
When donor and recipient are of opposite sex, the assess- ventional therapy for hematologic malignancies.
ment of donor engraftment is based on conventional cytoge-
netics and is relatively straightforward. When donor and
Gene therapy
recipient are of the same sex, RFLP analysis of donor and
recipient bone marrow allows the detection of polymorphic The application of gene therapy to genetic hematologic dis-
markers to distinguish DNA from the donor and recipient. orders has long been an attractive concept. In most cases,
After transplantation, RFLP analysis of recipient peripheral this involves insertion of normal genes into autologous
blood cells then can be used to document engraftment, chi- hematopoietic stem cells with subsequent transplantation
merism, graft failure, and disease relapse. In most centers, back into the patient. Candidate hematologic diseases for
PCR amplification and genotyping of short tandem repeat such therapy include hemophilia, sickle cell disease, thalas-
or variable number tandem repeat sequences that are poly- semia, and severe combined immune deficiency syndrome.
morphic between donor and recipient pairs are now used to Rapid advances in technology for the separation of hemato-
assess chimerism. poietic stem cells and techniques of gene transfer into those
cells have advanced efforts toward this goal, and many clini-
cal trials have been completed. Although significant meth-
Applications to novel therapies
odologic hurdles remain, research in this field continues to
Antisense and RNA interference therapy
move forward. It should be recognized, however, that correc-
The recognition that abnormal expression of oncogene tion of such diseases as hemophilia, sickle cell disease, and
products plays a role in malignancy has led to the proposal thalassemia requires efficient gene transfer to a large number
that suppression of that expression might reverse the neo- of hematopoietic stem cells with high levels of expression of
plastic phenotype. One way of blocking mRNA expression is the β-globin gene in erythroid precursors. Long-term repop-
through the use of antisense oligonucleotides. These are ulating stem cells have been relatively resistant to genetic
short pieces of single-stranded DNA or RNA, 17-20 bases modification; thus, many investigators have focused on gene
long, which are synthesized with a sequence complementary therapy applications in which low levels of expression could
to the transcription or the translation initiation site in the restore patients to health. A major impediment to successful
mRNA. These short single-stranded species enter the cell gene therapy has been the lack of gene delivery systems that
freely, where they complex to the mRNA through the com- provide safe, efficient, and durable gene insertion and that
plementary sequence. Investigation of the mechanism of can specifically target the cells of interest. An important
action of antisense oligonucleotides led to the discovery that safety concern with viral vectors that integrate into the host
naturally occurring double-stranded RNA molecules sup- genome is the potential to activate oncogenes or inactivate
press gene expression better than antisense sequences and tumor suppressor genes by insertional mutagenesis. Cur-
helped to unravel the mechanism of RNA interference. RNA rently used approaches include retroviral vectors, adenoviral
interference has significant advantages over antisense ther- vectors, other viral vectors, and nonviral vectors.
apy in that much lower concentrations are required. Numer- Recently, use of viral vectors to introduce cDNA encoding
ous studies are under way in hematologic diseases; however, artificial T cell receptors into T cells to stimulate T-cell medi-
methods for delivery of siRNAs are still far from perfect. In ated recognition of tumor-associated antigens has been suc-
one study, adult stem cells from sickle cell patients were cessful in the treatment of a variety of B cell malignancies.
infected with a viral vector carrying a therapeutic γ-globin These chimeric antigen receptor T cells (CAR T cells) are
gene harboring an embedded siRNA precursor specific introduced into autologous T cells from patients ex vivo and
for sickle β-globin. The newly formed red blood cells then re-administered to patients. CAR T cells targeting CD19
made normal hemoglobin and suppressed production of have been very promising, and clinical trials to define their
sickle β-globin. In another study, a retroviral system for long-term efficacy and safety in B-cell acute lymphoblastic
stable expression of siRNA directed to the unique fusion leukemia, chronic lymphocytic leukemia, and a variety of
junction sequence of ETV6-PDGFRB resulted in profound CD19-expressing lymphomas are underway.

Glossary | 19
Glossary complementary Sequence of the second strand of DNA

that is determined by strict purine–pyrimidine base
alleles Alternative forms of a particular gene. pairing (A-T; G-C).
allele-specific oligonucleotide Oligonucleotide whose complementary DNA (cDNA) Double-stranded DNA product
sequence matches that of a specific polymorphic allele. from an RNA species. The first strand is synthesized by
For example, oligonucleotides matching the sequence of reverse transcriptase to make a DNA strand complementary
unique immunoglobulin or T-cell receptor gene to the mRNA. The second strand is synthesized by DNA
rearrangements that are used for polymerase chain polymerase to complement the first strand.
reaction (PCR) detection of minimal residual disease. constitutive promoter A promoter that drives high-level
alternative splicing Selective inclusion or exclusion of expression in all tissues.
certain exons in mature RNA by utilization of a varied copy number variant A segment of DNA at least 1 kb in
combination of splicing signals. length that varies in copy number between individuals.
antisense oligonucleotides Oligonucleotides with a base CRISPR/Cas (clustered regulatory interspaced short
sequence complementary to a stretch of DNA- or palindromic repeat/Cas-based RNA-guided DNA
RNA-coding sequence. endonucleases) A technology which combines the Cas
capping Addition of the nucleotide 7-methylguanosine DNA nuclease with the sequence-specific DNA
to the 5′ end of mRNA. This is a structure that appears to recognition module of CRISPR to create targeted genetic
stabilize the mRNA. alterations in DNA.
chimera An organism containing two or more different cytogenetics Study of the chromosomal makeup of a cell.
populations of genetically distinct cells. As in chimeric degenerate Characteristic of the genetic code whereby
mice generated by microinjection of embryonic stem more than one codon can encode the same amino acid.
cells into a developing blastocyst or chimerism of Dicer A component of the processing mechanism that
donor and recipient cells after allogeneic stem cell generates microRNAs and siRNAs.
transplantation. Also used to describe transcripts that Drosha A component of the processing mechanism for
fuse coding sequences from different genes as a result of formation of microRNAs.
chromosomal rearrangements. enhancer A cis-acting regulatory sequence within a gene
chimeric antigen receptor T cells (CAR T cells) Genetically locus that interacts with nuclear protein in such a way as
modified T cells engineered to express an artificial T cell to increase the expression of the gene.
receptor that recognizes a specific tumor-associated epigenetics Changes in gene expression caused by
antigen. mechanisms other than alteration of the underlying
ChIP-Seq A combination of chromatin DNA sequence. Includes DNA methylation and histone
immunoprecipitation followed by next-gen sequencing modification. The changes are heritable in daughter
used to identify protein–DNA interactions. cells but can be modified pharmacologically
chromatin Complex of genomic DNA with histone and (eg, methyltransferase inhibitors, histone deacetylase
nonhistone proteins. inhibitors) or by normal enzymatic processes.
chromosome Large linear DNA structures tightly exome All protein coding portions of genes (exons) in the
complexed to nuclear proteins. genome.
cis-acting regulatory elements Sequences within a gene exon Portion of the structural gene that encodes for protein.
locus, but not within coding sequences, that are involved flanking sequences DNA lying 5′ and 3′ of a structural gene
in regulating the expression of the gene by interaction that frequently contain important regulatory elements.
with nuclear proteins. fluorescence in situ hybridization (FISH) High-resolution
clonal Arising from the expansion of a single cell. mapping of genes by hybridization of chromosome
coding sequence Portion of the gene contained within spreads to biotin-labeled DNA probes and detection by
exons that encodes for the amino acid sequence of the fluorescent-tagged avidin.
protein product. frameshift mutation Mutation within the coding sequence
codon The 3-nucleotide code that denotes a specific of a gene that results from deletion or insertion of a
amino acid. nucleotide that disrupts the 3-base codon structure of the
comparative genomic hybridization (CGH) A technique gene, thereby altering the predicted amino acid sequence
allowing for the detection of subtle chromosomal of the protein encoded by that gene.
changes (deletions, amplifications, or inversions that are gene Functional genetic unit responsible for the
too small to be detected by conventional cytogenetics production of a given protein, including the elements
techniques). that control the timing and the level of its expression.

gene expression profile Analysis of the global expression labeled cDNAs from a tissue of interest allows the
of a collection of cells using hybridization of mRNA to generation of a gene expression profile.
microarrays. microRNA Small RNA molecules encoded in the genomes
gene regulation Process controlling the timing and level of of plants and animals. These highly conserved,
expression of a gene. approximately 21-mer RNAs regulate the expression of
genetic code System by which DNA encodes for specific genes both by changing stability of mRNAs as well as by
protein through 3-nucleotide codons, each encoding a translational interference.
specific amino acid. next-gen sequencing Massively parallel sequence
genomics The study of the entire DNA sequence production from single-molecule DNA templates.
of organisms and interactions among various noncoding sequences DNA sequences that do not directly
genetic loci. encode for protein.
homologous recombination Alteration of genetic material nonsense-mediated decay Nonsense mutation (premature
by alignment of closely related sequences. In targeting stop codon) of one allele of an mRNA may result in
genes by homologous recombination, plasmids that degradation of the abnormal mRNA.
contain altered genes flanked by long stretches of DNA Northern blotting Analysis of RNA expression by gel
that match the endogenous gene are introduced into electrophoresis, transfer to nitrocellulose or nylon filter,
embryonic stem cells. A rare recombination event will and hybridization to a single-stranded probe.
cause the endogenous gene to be replaced by the mutated nucleic acid hybridization Technique of nucleic acid
gene in the targeting plasmid. This is the means by which analysis via association of complementary single-
knockout mice are obtained. stranded species.
imprinting Genetic process in which certain genes are nucleotide Basic building block of nucleic acids, composed
expressed in a parent-of-origin–specific manner. of a sugar moiety linked to a phosphate group and a
intron Intervening sequence of noncoding DNA that purine or pyrimidine base.
interrupts coding sequence contained in exons. oligonucleotide Short single-stranded DNA species,
induced pluripotent stem (iPS) cells A type of pluripotent usually composed of 15-20 nucleotides.
stem cell derived from a somatic cell that is generated by oncogene Cellular gene involved with normal cellular
exposing the somatic cell to factors that reprogram it to a growth and development, the altered expression of which
pluripotent state. has been implicated in the pathogenesis of the malignant
knockout mouse Mouse in which both of the copies of a phenotype.
gene have been disrupted by a targeted mutation. Such polyadenylation Alteration of the 3′ end of mRNA by
mutations are achieved by homologous recombination the addition of a string of adenosine nucleotides
using plasmids containing the mutated gene flanked by (“poly-A tail”) that appear to protect the mRNA
long stretches of the normal endogenous gene sequence. from premature degradation.
Mice that are heterozygous in the germline for the polymorphism Phenotypically silent mutation in DNA
targeted allele can be bred to generate mice that lack both that is transmitted from parent to offspring.
copies of the normal (wild-type) gene. pre-messenger RNA Unprocessed primary RNA transcript
leucine zipper Leucine-rich side chains shared by a group from DNA, including all introns.
of transcription factors that allow protein–protein and promoter Region in the 5′ flanking region of a gene that is
protein–DNA interactions. necessary for its expression; includes the binding site for
linkage mapping Analysis of a gene locus by study of RNA polymerase II.
inheritance pattern of markers of nearby (linked) loci. purine Two of the bases found in DNA and RNA: adenine
methylation DNA modification by addition of methyl and guanine.
groups to cytosine residues within genomic DNA. pyrimidine Bases found in DNA and RNA: cytosine and
Hypermethylation of clustered CpG groups in promoter thymine in DNA; cytosine and uracil in RNA.
regions (CpG islands) is a characteristic of quantitative PCR PCR in which the product is quantitated
transcriptionally inactive DNA; reduction in methylation in comparison to the PCR product resulting from a
is generally associated with increased transcriptional known quantity of template. This allows quantitation of
activity. the template in the reaction; it can, for example, allow an
microarray Glass slide or silicon chip on which cDNAs or estimate of the degree of contamination with tumor cells
oligonucleotides have been spotted to allow for the in a cell population.
simultaneous analysis of expression of hundreds to real-time PCR Automated technique for performing
thousands of individual mRNAs. Hybridization of quantitative PCR using a fluorogenic reporter to detect

Bibliography | 21
levels of target sequences during early cycles of the termination codon One of three codons that signal the
PCR reaction. termination of translation.
restriction endonucleases Enzymes produced by bacteria trans-acting factor Protein that interacts with cis-acting
that cleave double-stranded DNA at specific recognition regulatory region within a gene locus to regulate
sequences. transcription of that gene. Also called transcription factor.
restriction fragment-length polymorphism transcription Process by which pre-mRNA is formed from
(RFLP) Polymorphism in which a silent mutation the DNA template.
occurs within the recognition sequence for a restriction transcription factor Protein that interacts with cis-acting
endonuclease. This results in an alteration in the size of regulatory region within a gene locus to regulate
the DNA fragment resulting from digestion of DNA from transcription of that gene. Also called trans-acting
that DNA locus. factor.
reverse transcriptase Enzyme encoded by retroviruses transfer RNA (tRNA) Small RNA molecules that bind
that mediates conversion of RNA to complementary to the ribosome and covalently bind specific amino
DNA. acids, allowing translation of the genetic code into
reverse transcriptase polymerase chain reaction protein.
(RT-PCR) Amplification of RNA sequences by transgenic mouse A mouse that expresses an exogenous
conversion to cDNA by reverse transcriptase, followed by gene (transgene) introduced randomly into its genome.
the polymerase chain reaction. Linearized DNA is injected into the pronucleus of a
ribosome A ribonuclear protein complex that binds to fertilized oocyte, and the zygote is reimplanted. Resultant
mRNA and mediates its translation into protein by mice will carry the transgene in all cells.
reading the genetic code. translation Process by which protein is synthesized from
RISC RNA-induced silencing complex. A multiprotein an mRNA template.
complex that combines with microRNAs to target translocation breakpoint Site of junction of two
complementary mRNA for degradation or translation aberrantly juxtaposed (translocated) chromosomal
inhibition. fragments.
RNA expression array An array-based technique used to tumor suppressor Genes that promote tumor development
determine the abundance of each of the known mRNAs when deleted or inactivated.
(the gene expression profile) in a group of cells. uniparental disomy Occurs when two copies of a
RNA polymerase II Enzyme-mediating transcription of chromosome, or part of a chromosome, are derived from
most structural genes. one parent and no copies derive from the other parent.
RNA-seq Next-gen sequencing using RNA templates. In a somatic cell, this can result in progeny with
silencer A cis-acting regulatory sequence within a gene two copies of the wild-type allele or two copies of
locus that interacts with nuclear protein in such a way as the mutant allele.
to decrease the expression of the gene. Western blotting Detection of specific proteins via binding
single nucleotide polymorphism (SNP) Naturally of specific antibody to protein on a nitrocellulose or
occurring inherited genetic variation between individuals nylon membrane.
at the level of single nucleotides. zinc-finger Structural feature shared by a group of
siRNA Small interfering RNAs that act in concert with transcription factors. Zinc-fingers are composed of a zinc
large multiprotein RISC complexes to cause cleavage of atom associated with cysteine and histidine residues; the
complementary mRNA or prevent its translation. fingers appear to interact directly with DNA to affect
Southern blotting Analysis of DNA by gel electrophoresis, transcription.
transfer to nitrocellulose or nylon filter, and hybridization zinc-finger nucleases Artificial restriction enzymes
to single-stranded probe. generated by fusing a zinc-finger DNA binding domain
splicing Process by which intron sequences are removed to a DNA-cleavage domain for use in creating specific
from pre-mRNAs. genomic alterations in DNA.
transcription activator-like effector nucleases
(TALENs) Artificial restriction enzymes with sequence-
specific DNA binding activity which can be utilized to Bibliography
create specific genetic alterations in DNA.
telomeres Nucleoprotein structures at the ends of Alizadeh AA, Eisen MB, Davis RE, et al. Distinct types of diffuse
chromosomes that protect chromosome ends from large B-cell lymphoma identified by gene expression profiling.
degradation and fusion. Nature. 2000;403:503-511. One of the first papers to demonstrate

diversity in gene expression among diffuse large B-cell lymphomas showed that mice can be rescued after transplantation with
and the fact that gene expression reflects tumor proliferation rate, corrected hematopoietic progenitors obtained in vitro from
host response, and differentiation state of the tumor. autologous iPS cells.
Cazzola M, Della Porta MG, Malcovati L. The genetic basis of Hughes T, Branford S. Molecular monitoring of BCR-ABL as a
myelodysplasia and its clinical relevance. Blood. 2013;122:4021- guide to clinical management in chronic myeloid leukaemia.
4034. Blood Rev. 2006;20:29-41. A description of the use of real-time PCR
Hammond SM. Dicing and slicing: the core machinery or the RNA to monitor BCR-ABL transcripts in chronic myeloid leukaemia.
interference pathway. FEBS Lett. 2005;579:5822-5829. A review Plongthongkum N, Diep DH, Zhang K. Advances in the profiling
describing the discovery of the RNA interference pathway and of DNA modifications: cytosine methylation and beyond. Nat
discussion of future lines of work. Rev Genet. 2014;15:647-661.
Hanna J, Wernig M, Markoulaki S, et al. Treatment of sickle Volpi EV, Bridger JM. FISH glossary: an overview of the
cell anemia mouse model with iPS cells generated from fluorescent in situ hybridization technique. Biotechniques.
autologous skin. Science. 2007;318:1920-1923. Using a 2008;45:385-409. A succinct review of the many FISH strategies
humanized sickle cell anemia mouse model, these investigators available.

CHAPTER
2
Consultative hematology I:
hospital-based and selected
outpatient topics
Cindy Neunert and Anita Rajasekhar
The role of the hematology consultant, 23 Consultation for hematologic complica Hematology consultations in pediatric
Consultation for surgery and invasive tions of solid organ transplantation, 36 patients, 43
procedures, 24 Common outpatient hematology Bibliography, 52
Common inpatient consultations, 30 consultations, 38
The role of the hematology consultant can be inpatient or outpatient and the timing emergent,
urgent, subacute, or more planned. Imperative to an effec-
A hematology consultant provides expert advice about the tive and efficient consultation, both the referring clinician
diagnosis and management of benign or malignant hemato- and the hematology consultant must have a clear under-
logic disorders to requesting physicians and other health standing of the extent of the clinical questions being asked,
care providers. A consultation request might involve an which in turn will guide the aim and comprehensiveness of
adult general medical patient, a child or adolescent, a preg- the consult. In the era of rising health care costs, expert
nant woman, a perioperative patient, or an individual who is hematology consultation must seek to be cost-effective by
critically ill. Other consultative responsibilities of the hema- curtailing unwarranted diagnostic and therapeutic mea-
tologist may include serving on committees that maintain a sures. With that in mind, the American Society of Hematol-
formulary, developing clinical practice guidelines, establish- ogy (ASH) Choosing Wisely campaign seeks to identify and
ing policies and procedures for transfusion services, or mon- educate clinicians on commonly performed tests or proce-
itoring quality and efficiency. The setting of a consultation dures within the realm of hematology that are unnecessary,
not supported by evidence, duplicative, and potentially
Conflict-of-Interest disclosure: Dr. Neunert: Safety Advisory harmful (http://www.hematology.org/Clinicians/Guidelines-
Board: Genzyme. Dr. Rajasekhar: Consultancy: Alexion, Octapharma, Quality/502.aspx). A clinical hematologist must understand
Bayer, Baxter; Funding: American Society of Hematology. the principles of effective consultation and the extreme
Off-label drug use: Dr. Neunert: corticosteroids and rituximab for
use in TTP; ATIII and activated protein C for DIC; Cyclophospha importance of interphysician communication (Table 2-1).
mide, corticosteroids, intravenous immunoglobulin, and rituximab Consultants need to communicate effectively not only with
for CAPS; corticosteroids, intravenous immunoglobulin, rituximab, other staff physicians and consultants, but also with ancillary
and thrombopoietin receptor agonists for pediatric ITP; rituximab
for adult ITP; granulocyte-stimulating factor outside of severe con
members of the health care team, house staff, fellows, stu-
genital neutropenia. Dr. Rajasekhar: Desmopressin for use in platelet dents, and the patient and family. A commitment to effective
function disorders; desmopressin, recombinant FVIIa, prothrombin communication ensures maximal compliance with recom-
complex concentrate, activated prothrombin complex concentrate,
mendations and the highest quality of multidisciplinary
fibrin glue, and, ε-aminocaproic acid, and tranexamic acid for use in
surgical bleeding; conjugated estrogens for use in uremic bleeding; patient care.
Intravenous immunoglobulin and erythropoietin for use in parvovi This chapter discusses some of the most common hemato-
rus associated pure red cell aplasia; angiotensin converting enzyme logical consultations, including preoperative management of
inhibitors and angiotensin receptor blockers for use in post–renal
transplant erythrocytosis; rituximab for use in CD20+ posttransplant hematological disorders, inpatient and outpatient consulta-
lymphoproliferative disorder. tions, and specific issues pertaining to pediatric hematology.
| 23

24 | Consultative hematology I: hospital-based and selected outpatient topics
Table 2-1 Principles of effective consultation and interphysician communication
Principle Comment
Determine the question that is being asked The consultant must clearly understand the reason for the consultation
Establish the urgency of the consultation Urgent consultations must be seen as soon as possible (communicate any expected delays
and respond in a timely manner promptly); elective consultations should be seen within 24 hours
Gather primary data Personally confirm the database; do not rely on second-hand information
Communicate as briefly as appropriate Compliance is optimized when the consultant addresses specific questions with five succinct
and relevant recommendations
Make specific recommendations Identify major issues; limit the diagnostic recommendations to those most crucial; and
provide specific drug doses, schedules, and treatment guidelines
Provide contingency plans Briefly address alternative diagnoses; anticipate complications and questions
Understand the consultant’s role The attending physician has primary or ultimate responsibility; the consultant should not
assume primary care or write orders without permission from the attending
Offer educational information Provide relevant evidence-based literature or guidelines
Communicate recommendations directly to Direct verbal contact (in person or by phone) optimizes compliance and minimizes
the requesting physician confusion or error
Provide appropriate follow-up Continue involvement and progress notes as indicated; officially sign off the case or provide
outpatient follow-up
Adapted from Goldman L, Lee T, Rudd P. Arch Intern Med. 1983;143:1753-1755; Sears CL, Charlson ME. Am J Med. 1983;74:870-876;
and Kitchens CS, Kessler CM, Konkle BA. Consultative Hemostasis and Thrombosis. 3rd ed. Philadelphia, PA: Elsevier Saunders;2013:3-15.
Consultation for surgery and invasive bridging anticoagulation should be considered. In general,
procedures patients may be classified as having a high, moderate, or low
risk of perioperative thromboembolism. These categories cor-
respond to an estimated annual thrombotic risk of >10%,
Clinical case 5%-10%, and <5%, respectively. Individuals with mechanical
mitral valves, atrial fibrillation and CHADS2 scores of 5 or 6,
A 34-year-old female with systemic lupus erythematous (SLE) has
been referred to your hematology clinic for perioperative man
recent (within 3 months) stroke or venous thromboembolism
agement of her anticoagulation. She was recently diagnosed with (VTE), or severe thrombophilia (eg, deficiency of protein C,
antiphospholipid syndrome (APLS) during the workup for a large protein S, or antithrombin; antiphospholipid syndrome;
right middle cerebral artery (MCA) infarct 3 months ago. She is combination thrombophilias) are considered high risk. Those
on warfarin with an INR goal of 2.0-3.0 with approximately 80% with atrial fibrillation and CHADS2 scores of 0-2 or a remote
time in therapeutic range. She now requires a tooth extraction for history of VTE more than 12 months before surgery and no
an abscessed tooth that has not responded to medical therapy. other thrombotic risk factors typically are classified as low
In light of the patient’s APLS and history of cerebrovascular acci risk. Individual patient factors not captured in this classifica-
dent (CVA), you judge her thrombotic risk to be high if warfarin tion scheme as well as type of surgery should be considered in
is interrupted. Because of the risk of infection progressing, the
estimating an individual patient’s perioperative thrombotic
surgery cannot be delayed. The oral surgeon is concerned about
risk and whether bridging anticoagulation is necessary. This
the patient’s bleeding risk. You advise the patient to continue
thrombotic risk must be weighed against the risk of surgical
warfarin at the current dose and prescribe an adjuvant mouth
wash containing ε-aminocaproic acid to control local bleeding.
hemorrhage. For example, in patients with high risk of peri-
operative thrombosis, continuation of warfarin rather than
bridging with heparin in those requiring pacemaker or
Perioperative management of
implantable cardioverter-defibrillator surgery reduces clini-
antithrombotic therapy
cally-significant device-pocket hematomas without any differ-
Hematologists are often consulted to provide recommenda- ence in thromboembolic events. An assessment of hemorrhagic
tions on temporary interruption of antithrombotics for a sur- risk should take into account the propensity for bleeding asso-
gery or procedure. The perioperative management of patients ciated with both the procedure and antithrombotic agent in
taking antiplatelet or anticoagulant drugs is based on (i) an question. In addition, the patient’s prior history of bleeding,
assessment of the patient’s risk for thromboembolism and (ii) comorbidities that may affect bleeding (eg, renal function) as
an assessment of risk for perioperative bleeding. These consid- well as concomitant use of antiplatelet and nonsteroidal anti-
erations are used to determine whether antithrombotic ther- inflammatory medications is important in determining
apy should be interrupted prior to surgery and, if so, whether overall bleeding risk. Generally, procedures or surgeries

Consultation for surgery and invasive procedures | 25
Stop warfarin 5 days before surgery
Assess perioperative
thromboembolic risk
Low risk Moderate risk High risk
No bridging Consider bridging Bridging anticoagulation

• Resume warfarin • Based on assessment of • Use therapeutic dose
12-24 hours after individual patient- and SC LMWH or IV UFH
surgery and when surgery-related factors • Administer last dose of
hemostasis has LMWH 24 hours before
been achieved surgery or stop UFH 4-6
hours before surgery
• Resume therapeutic dose
UFH or LMWH 48-72 hours
after surgery and when
hemostasis has been achieved
Figure 2-1 Approach to perioperative management of patients on warfarin undergoing major surgery. Management should be informed by an
individualized assessment of host- and surgery-related risk factors for perioperative thromboembolism and hemorrhage as well as patient values
and preferences. LMWH = low–molecular weight heparin; UFH = unfractionated heparin. Based on Douketis JD et al. Chest. 141:e326S.
associated with the potential for intracranial, intraocular, may be performed without interruption of warfarin. Periop-
spinal, retroperitoneal, intrathoracic, or pericardial bleeding erative anticoagulation should be used with caution after cer-
are considered high-risk for bleeding. Procedures with a low tain procedures like (i) prostate or kidney biopsy, where
bleeding risk include non-major procedures (lasting <45 postoperative bleeding may be stimulated by the highly vas-
minutes) such as general surgical procedures (hernia repair, cular tissue and endogenous urokinase; (ii) large colonic pol-
cholecystectomy), dental, or cutaneous procedures. The ypectomies which can be associated with bleeding at the stalk;
American College of Chest Physicians (ACCP) updated guide- (iii) and cardiac pacemaker or defibrillator implantation
lines on the perioperative management of antithrombotic where a pocket hematoma may form. For intracranial or spi-
medications in 2012 (http://chestjournal.chestpubs.org). nal surgery, bridging therapy should be used with caution.
An evidence-based approach to the perioperative manage- Importantly, an INR <1.5 should be achieved the day prior to
ment of patients on warfarin undergoing major surgery is surgery. If LMWH bridging is deemed necessary, then the last
shown in Figure 2-1. Temporary discontinuation of warfarin, dose should be half the normal daily dose and administered
approximately 5 days until normalization of the INR, is rec- 24 hours before the procedure to avoid residual anticoagulant
ommended in all patients. Bridging anticoagulation with effect. Depending on the patient’s underlying thrombotic
therapeutic-dose LMWH or unfractionated heparin (UFH) risk, postoperative options include waiting 48-72 hours after
may be considered depending on the patient’s risk of throm- surgery before resuming full-dose LMWH bridging therapy,
boembolism. A recent study (BRIDGE trial) randomized using an intermediate or prophylactic LMWH, or utilizing
patients to bridging anticoagulation with LMWH versus no only mechanical prophylaxis if the bleeding risk is extremely
bridging anticoagulationin patients with atrial fibrillation high. For neuraxial anesthesia, the dosing and timing of peri-
that required warfarin interruption for a procedure or sur- operative LMWH follow the practice guidelines laid out from
gery. The authors found that forgoing bridging anticoagula- the American Society of Regional Anesthesia. Like warfarin,
tion was noninferior to perioperative bridging with LMWH the new oral anticoagulants dabigatran, rivaroxaban, apixa-
for the prevention of arterial thromboembolism with the ban, and edoxaban must be discontinued before major sur-
benefit of decreased major bleeding. In patients requiring gery. However, unlike warfarin the predictable short half-life
minor dental procedures, warfarin may be continued with of these new oral anticoagulants allows for relatively short-
co-administration of an oral antifibrinolytic agent, if needed, term cessation preoperatively without the routine need for
or warfarin may be stopped 2-3 days before the procedure. bridging anticoagulation. Most patients can safely undergo
Warfarin also may be continued in patients undergoing procedures within 24 to 48 hours of their last dose of these
minor dermatologic procedures with the use of adjunctive new oral anticoagulants, depending on surgical risk of bleed-
local hemostatic measures as necessary. Cataract surgery also ing. However, with renal impairment, hepatic impairment,

Table 2-2 Perioperative cessation and resumption of new oral anticoagulants
Type of NOAC* Dabigatran Rivaroxaban/apixaban
Bleeding risk of intervention Low High Low High Edoxaban

CrCL > 80 mL/min > 24h > 48h > 24h > 48h No published data
CrCL 50-80 mL/min > 36h > 72h > 24h > 48h
CrCL 30-50 mL/min > 48h > 96h > 24h > 48h
CrCL 15-30 mL/min Not indicated > 36h > 48h
CrCL <15 mL/min Not indicated Not indicated
Resumption after procedure† < 24h 24-48h < 24h 48-72h No published data
Adapted from Heidbuchel et al. Europace. 2013;15:625–651, with permission of Oxford University Press (UK); © European Society of Cardiology.
CrCL = creatinine clearance.
* For patients on dabigatran 150mg BID, apixaban 5mg BID, or rivaroxaban 20mg QD.
† Depending on if hemostasis achieved. If significant risk for perioperative thrombosis exists, prophylactic or intermediate
doses UFH or LMWH should be considered until full therapeutic anticoagulation is resumed.
older age, and concurrent antiplatelet medications, longer to cardiopulmonary bypass). Host factors include the pres-
cessation intervals may be necessary preoperatively (Table ence of an underlying congenital or acquired hemostatic
2-2). Given the short onset of action of these drugs, hemosta- defect and use of drugs that affect hemostasis.
sis must be achieved postoperatively before restarting these A focused medical history should include a detailed per-
new oral anticoagulants. Perioperative bridging protocols sonal history of abnormal bleeding; response to prior hemo-
with these agents have been proposed based on pharmacoki- static challenges, such as surgeries, trauma, and childbirth;
netic data but have not been investigated systematically. and comorbidities or use of medications that could affect
Perioperative management of antiplatelet therapy, like hemostasis. Patients should be queried specifically about
with oral anticoagulants, relies on an assessment of the indi- common procedures such as tooth extraction and tonsillec-
vidual patient’s thrombotic risk as well as the nature of the tomy, which they may not think to mention unless prompted.
planned procedure. In general, patients may remain on aspi- A careful family history of bleeding is crucial, particularly in
rin for minor dental or dermatologic procedures and cata- patients who may not have undergone extensive prior hemo-
ract surgery. For major noncardiac surgery, those at low risk static challenges themselves. A targeted physical examina-
of cardiovascular events should discontinue aspirin 7-10 tion for stigmata of bleeding and evidence of comorbid
days before surgery. Aspirin should be continued in patients conditions that may affect hemostasis, such as liver disease
judged to be at moderate or high risk. Patients who require or a connective tissue or vascular disorder, should be per-
coronary artery bypass grafting (CABG) should remain on formed as a complement to the history.
aspirin in the perioperative setting. If such patients are on Preoperative hemostatic laboratory testing (ie, PTT/PT) is
dual antiplatelet therapy, clopidogrel or prasugrel should be neither cost effective nor informative in patients without a
held beginning 5 days before surgery. personal or family history suggestive of a bleeding disorder.
In patients with a coronary stent who are receiving dual However, if the history or physical examination is suggestive
antiplatelet therapy and require surgery, surgery should be of a bleeding diathesis, preoperative testing should include a
deferred, if possible, during the period of highest risk for in- platelet count, prothrombin time (PT), and activated partial
stent thrombosis (6 weeks after placement of bare metal stents, thromboplastin time (aPTT). Normal initial testing does not
6 months after placement of drug-eluting stents). After this exclude a clinically important bleeding diathesis such as a
period has passed, clopidogrel or prasugrel may be suspended platelet function defect, von Willebrand disease, mild factor
temporarily for surgery. If surgery cannot be delayed, dual anti- deficiency, or a fibrinolytic disorder, and further testing
platelet therapy should be continued during and after surgery. should be guided by the clinical history and the results of the
initial laboratory evaluation. Once a diagnosis has been
established, a plan for perioperative hemostatic manage-
Preoperative assessment of bleeding risk
ment should be developed based on the nature and severity
Bleeding risk is related to both surgical and host factors. Sur- of the defect and the bleeding risk of the anticipated proce-
gical factors include the location and extent of the interven- dure. Although high-level evidence is lacking, a fibrinogen of
tion, the vascularity and fibrinolytic activity of the surgical at least 100 mg/dL and a platelet count of at least 50 × 109/L
bed, the compressibility of the site and the ability to achieve is desired for moderate- to high-risk procedures. For neuro-
surgical hemostasis, and the possibility that the procedure surgery and ophthalmologic procedures, it often is prudent
may induce a hemostatic defect (eg, platelet dysfunction due to target a platelet count of at least 100 × 109/L.

Management of perioperative hemorrhage surgical reexploration should be considered. Some systemic

bleeding diatheses (such as mild deficiency of factors VIII, IX,
Perioperative hemorrhage may be due to inadequate local
or XI; von Willebrand disease; qualitative platelet defects; or a
hemostasis or a systemic hemostatic defect. Potential hemo-
disorder of fibrinolysis) may not be identified by basic labora-
static defects include an unrecognized preexisting bleeding
tory testing. Patients with mild factor XI deficiency, for exam-
diathesis, drugs, uremia, dilutional coagulopathy, or dissemi-
ple, may have a normal or near-normal aPTT. Clinicians
nated intravascular coagulation (DIC). Not to be overlooked
should maintain a high index of suspicion for these disorders
is the increased risk of bleeding induced by acid-base distur-
in a patient with persistent unexplained surgical bleeding and
bances and hypothermia. Close attention should be paid to
test for specific coagulation factor levels as indicated.
the pattern of bleeding, specifically the timing in relation to Adjunctive agents may be used alone for minor bleeding
surgery, the location, and the tempo of the bleed. A struc- or as a complement to product replacement for major bleed-
tural defect is more likely with a single site (versus multiple ing in selected patients and clinical circumstances. DDAVP
sites) of bleeding, with sudden onset of bleeding (versus (desmopressin acetate) may be used for mild bleeding in
delayed bleeding following initial hemostasis), and/or with patients with mild hemophilia A, mild von Willebrand dis-
brisk bleeding (versus slow persistent oozing). ease, or a qualitative platelet defect. Ideally, response to this
Certain surgeries are associated with specific hemostatic agent should be documented before its use in the acute set-
defects. Excessive blood loss in patients undergoing cardio- ting. Mucocutaneous bleeding may respond to antifibrinol-
pulmonary bypass surgery may be due to the effects of the ytic therapy with tranexamic acid or e-aminocaproic acid.
bypass circuit on platelet function and fibrinolysis or the use Oral or intravenous conjugated estrogens, given for 5-7 days
of antiplatelet agents, heparin, or other anticoagulants. Liver preoperatively, may decrease platelet-related bleeding in
transplantation carries unique risks due to the temporary patients with chronic kidney disease. Topical fibrin sealants
loss of coagulation factor synthesis and enhanced fibrinoly- may be used to reinforce local hemostasis in patients with
sis. During reperfusion of the transplanted liver, tissue-type underlying bleeding disorders.
plasminogen activator is released into the circulation and Hemostatic agents have been used to prevent or treat surgi-
proteolysis of von Willebrand factor (vWF) occurs. cal bleeding in patients without known hemostatic disorders.
All patients with surgical bleeding should undergo an Tranexamic acid and e-aminocaproic acid have been shown to
immediate basic hemostatic laboratory evaluation including reduce blood loss and blood transfusion after cardiac surgery,
a platelet count, PT, aPTT, and fibrinogen. Blood must be liver transplantation, and prostatectomy. An observational
drawn from a fresh peripheral venipuncture site due to the study of 4,374 patients undergoing coronary revascularization
common contamination of blood samples with heparin surgery on cardiopulmonary bypass showed that use of these
flushes, saline, erythrocytes, or plasma. Significant abnor- agents was associated with a 30%-40% reduction in surgical
malities of any of these initial parameters suggest a systemic blood loss without an increased risk of thromboembolism.
hemostatic defect, which may require specific hemostatic Recombinant factor VIIa (rFVIIa) is approved in the
therapy. Clinically significant thrombocytopenia or fibrino- United States for the treatment of patients with congenital
gen deficiency in a bleeding surgical patient mandates appro- hemophilia A or B and inhibitors, patients with acquired
priate therapy and further testing to identify the cause of the hemophilia, congenital factor VII deficiency, and Glanz
deficiency. In general, cryoprecipitate and platelets should be mann thrombasthenia with platelet refractoriness. Despite
transfused to maintain a fibrinogen concentration of at least these limited indications, the majority of rFVIIa usage is off-
100 mg/dL and a platelet count of at least 50 × 109/L (100 × label, especially for the management of perioperative bleed-
109/L for organ- or life-threatening bleeding), respectively. ing. Controlled trials have shown rVIIa to be of no benefit in
Hypothermia and acid-base disturbances should be cor- reducing transfusion in cirrhotic patients undergoing partial
rected. Although the thromboelastograph (TEG) has tradi- hepatectomy or orthotopic liver transplantation. Recent
tionally been utilized more by the anesthesiologist than the studies have highlighted the potential thrombotic risk with
hematologist, this test can provide an accurate and rapid off-label use of rVIIa. In a meta-analysis of 35 randomized
method of diagnosing hyperfibrinolysis, as seen in cardio- controlled trials of rVIIa for unapproved indications, the
pulmonary bypass and orthotopic liver transplant. There is overall rate of thromboembolism in rVIIa-treated subjects
growing interest in the utilization of TEG to guide transfu- was 9.0%. The rate of arterial, but not venous, events was
sion replacement therapy in trauma-induced coagulopathy higher in subjects receiving rVIIa, particularly among those
and in surgical patients to predict thromboembolic events. 65 years and older. The indiscriminant use of rVIIa for the
If basic hemostatic laboratory parameters are normal or management of perioperative hemorrhage should be dis-
bleeding persists after correction of these parameters, inade- couraged; however, it may be useful for selected patients with
quate local hemostasis due to vessel injury is suggested and life-threatening bleeding despite conventional measures and

appropriate transfusion therapy. Advanced age and preexist- (eg, solid organ transplant rejection, myasthenia gravis, or
ing cardiovascular risk factors may increase the risk of arte- focal segmental glomerulosclerosis). Discussion of various
rial thromboembolic complications with rVIIa. indications for apheresis is beyond the scope of this chapter.
Prothrombin complex concentrates (PCCs) are plasma- The reader is referred to the 2013 American Society of
derived concentrates of the vitamin K–dependent clotting Apheresis “Guidelines on therapeutic apheresis” for practi-
factors. PCCs are classified as either 3-factor or 4-factor cal evidence-based recommendations on apheresis for spe-
depending on the amount of FVII included. These products cific diseases.
are approved for the treatment of hemophilia B. In 2013, the
4-factor PCC Kcentra was approved for the urgent reversal
Prevention and treatment of postoperative
of acquired coagulation factor deficiency induced by vitamin
venous thromboembolism
K antagonist therapy in adult patients with acute major
bleeding. Activated PCCs (APCCs) contain variable amounts VTE is a common and potentially lethal complication of sur-
of activated vitamin K–dependent clotting factors and are gery. Pulmonary embolism remains the leading cause of pre-
indicated for the treatment of patients with hemophilia ventable death in hospitalized patients. Despite contemporary
and inhibitors. The use of PCCs and APCCs for the mana thromboprophylaxis, postoperative VTE rates remain unac-
gement of perioperative hemorrhage has been reported but ceptably high, causing the Agency for Health Care Research
not prospectively investigated. Further studies are needed and Quality to cite prevention of VTE as the number one
before use of these agents to control surgical bleeding can be priority for improving patient safety in hospitals.
recommended. Risk factors for VTE in surgical patients include type
Fibrin sealant, also known as fibrin glue, consists of two and extent of surgery or trauma, general anesthesia for
main components: human fibrinogen and human thrombin. greater than 30 minutes, longer duration of hospitaliza-
When delivered together, the thrombin cleaves fibrinogen to tion, advanced age, cancer, personal or family history of
form a stable fibrin clot on the tissue surface. Although ran- VTE, obesity, immobility, infection, presence of a central
domized clinical trial data and evidence-based guidelines are venous catheter, pregnancy or the postpartum state, and
lacking, fibrin sealant is used for hemostasis in cardiac and thrombophilia. Several prediction models have been devel-
thoracic surgery, trauma, liver and spleen lacerations, and oped to estimate VTE risk in surgical patients, but all have
dental procedures. It also has been used on the liver surface important limitations. A general risk stratification schema
following orthotopic liver transplantation to augment local recommended by the ACCP for patients undergoing
hemostasis. Fibrin glue is also utilized as an adhesive to seal nonorthopedic surgery is shown in Table 2-3. The ACCP
dural leaks, repair otic ossicles and bony defects, and provide guidelines utilize two validated risk stratification models
adhesion for skin grafts. Fibrin glue products are safe and based on risk factor point systems (Rogers score and Cap-
effective, but rare side effects include hypotension, anaphy- rini score). These scoring systems estimate an individual’s
laxis, infection transmission, and air embolism. Since the use perioperative VTE risk as low, moderate, or high by assign-
of human thrombin in fibrin glue formulations, the previ- ing a point value for various patient- and procedure-related
ously-reported bleeding diatheses associated with anti- risk factors (eg, age, obesity, degree of immobility, specific
bovine factor V or factor II antibodies (bovine factor V is a comorbidities, type of surgery planned, known thrombo-
contaminant of bovine thrombin preparations) that cross- philia). In general, very low-risk (<0.5%) and low-risk
reacts with endogenous factor V/II no longer occur. (~1.5%) patients tend to be younger than 40 years old,
have no adverse patient- or surgery-related risk factors,
and require general anesthesia for less than 30 minutes.
Apheresis
Patients in the moderate risk (~3.0%) category include
Apheresis is an extracorporeal procedure that removes a those with risk factors who are undergoing minor surgery
selective component (eg, white blood cells, red blood cells, and those age 40-60 years who have no additional surgery-
or plasma) of a patient’s whole blood. When the removed or patient-related risk factors, but who will require general
volume is replaced with a replacement fluid (eg, red blood anesthesia for >30 minutes. High-risk patients generally
cells, plasma, saline), the procedure is termed an exchange. include individuals >60 years of age undergoing major sur-
Various hematologic indications for apheresis exist; for gery as well as those age 40-60 years with additional risk
example, red blood cell exchange in sickle cell disease, plasma factors who will be having major surgery.
exchange for idiopathic thrombotic thrombocytopenic pur- A strategy for thromboprophylaxis should be based on the
pura, and leukocytapheresis for leukostasis. However, often estimated risk of VTE and bleeding and the type of surgery.
the hematologist is consulted for this procedure as part of Prophylactic measures include ambulation; lower extre
the therapeutic management of non-hematologic conditions mity intermittent pneumatic compression (IPC); graduated

Table 2-3 General VTE risk stratification for patients undergoing nonorthopedic surgery
Type of surgery
Risk of VTE Major general, Gastrointestinal,
Risk (without thoracic or urological, vascular, Plastic and
category prophylaxis) vascular breast, or thyroid reconstructive Other surgical populations
Very low <0.5% Rogers score <7 Caprini score 0 Caprini score 0-2 Most outpatient or same-day surgery
Low ~1.5% Rogers score 7-10 Caprini score 1-2 Caprini score 3-4 Spinal surgery for nonmalignant disease
Moderate ~3.0% Rogers score >10 Caprini score 3-4 Caprini score 5-6 Gynecologic noncancer surgery
Cardiac surgery
Most thoracic surgery
Spinal surgery for malignant disease
High ~6.0% NA Caprini score ≥5 Caprini score 7-8 Bariatric surgery
Gynecologic cancer surgery
Pneumonectomy
Craniotomy
Traumatic brain injury
Spinal cord injury
Other major trauma
Adapted from ACCP guidelines.

VTE = venous thromboembolism; NA = not applicable.
compression stockings (GCS); and pharmacologic prophy- The timing of initiation of prophylaxis varies based on the
laxis with low dose UFH, LMWH, fondaparinux, or oral procedure and regional practice patterns. In Europe, LMWH
anticoagulation. In general, early ambulation alone is is usually started at half doses 12 hours before surgery,
recommended for very low-risk nonorthopedic surgery whereas in the United States, it is common to start full doses
patients; mechanical prophylaxis, preferably with IPC, for 12-24 hours after surgery. Bleeding rates are low with both
low-risk patients; pharmacologic or mechanical prophylaxis strategies and are greater when LMWH is started within
for moderate-risk patients; and a combination of pharmaco- four hours before or after surgery. Prophylactic warfarin
logic and mechanical prophylaxis for high-risk patients begun just before or immediately after surgery is less com-
(http://chestjournal.chestpubs.org). An important exception monly associated with hemorrhage, but it is also less effective
is patients judged to be at high risk for bleeding, in whom in preventing DVT. LMWH, fondaparinux, dabigatran, api
mechanical prophylaxis is favored over pharmacologic strat- xaban, rivaroxaban, and edoxaban should be avoided in
egies unless and until bleeding risk diminishes. Surveillance patients with renal failure.
compression ultrasonography to screen for DVT and inferior When VTE occurs postoperatively, the consultant may be
vena cava filter insertion for primary prevention of DVT are asked for treatment recommendations. For most low-risk
generally not recommended in surgical patients. procedures, full anticoagulation can be initiated safely within
In the absence of a heightened bleeding risk, most patients 12-24 hours after surgery. The agent of choice in the imme-
undergoing major orthopedic surgery should receive phar- diate postoperative period is continuous-infusion UFH
macologic thromboprophylaxis. LMWH, fondaparinux, because of its short half-life and rapid reversibility with prot-
low-dose UFH, warfarin, and aspirin (all GRADE 1B) are all amine if bleeding develops. Contraindications to immediate
reasonable options for patients undergoing hip fracture sur- postoperative anticoagulation include active bleeding and
gery. Any of these agents, as well as the new oral anticoagu- certain neurosurgical or ophthalmologic procedures in
lants, dabigatran, apixaban, and rivaroxaban (all GRADE 1B), which bleeding would risk permanent injury. In patients
may be used following total hip or total knee arthroplasty. with postoperative VTE and a contraindication to anticoag-
IPC is also reasonable in combination with pharmacologic ulation, insertion of a retrievable inferior vena cava filter
prophylaxis during the hospital stay or in lieu of pharmaco- may be required. Once it is deemed to be safe, anticoagula-
logic prophylaxis, particularly in patients at increased risk tion should be initiated and a plan should be made for
for bleeding. Pharmacologic prophylaxis should be contin- retrieval of the filter. The duration of anticoagulation after a
ued for a minimum of 10-14 days after major orthopedic first, uncomplicated postoperative VTE is generally 3
surgery. Extended prophylaxis for 4-5 weeks should be con- months. Longer treatment may be indicated for recurrent
sidered after major orthopedic surgery and after major VTE and in the setting of certain hypercoagulable condi-
abdominal or pelvic surgery for cancer. tions, such as active cancer or antiphospholipid syndrome.

thrombocytopenia [ITP] may be caused by both platelet

Key points destruction and platelet underproduction), and some critically
• Surgical bleeding risk is associated with both patient- and ill patients may also have more than one source. We propose
surgery-related factors. Patient factors include the presence of an the following practical approach to the diagnosis of thrombo-
underlying congenital or acquired hemostatic defect and use of cytopenia in the hospitalized patient tailored to specific ele-
drugs that affect hemostasis. Surgical factors include the nature ments of the history, physical examination, and laboratory
and extent of the intervention, the vascularity and fibrinolytic investigations (Figure 2-2): (1) exclude thrombocytopenic
activity of the surgical bed, the compressibility of the site, and the emergencies, (2) examine the blood film, (3) consider the clin-
ability to achieve surgical hemostasis.
ical context, (4) assess the degree of thrombocytopenia, (5)
• A focused medical history is the most important tool for
establish the timing of thrombocytopenia, and (6) assess the
assessing the risk of surgical bleeding.
patient for signs of bleeding and/or thrombosis.
• Perioperative management of patients receiving antiplatelet or
anticoagulant drugs depends on the patient’s risk of thromboem
bolism and the risk of surgical bleeding. 1. Exclude thrombocytopenic emergencies
• Since apheresis for both hematologic and non-hematologic
conditions is a common consult, hematologists should be aware of
the evidence-based indications for this procedure. Clinical case
• The type of postoperative thromboprophylaxis required
depends on the patient’s risk of VTE, the type of surgery, and the A 62-year-old man is in the ICU following complications from
patient’s risk of bleeding. cardiac surgery. You are consulted on postoperative day 6
• Management of acute VTE in a postoperative patient is similar because his platelet count is 30 × 109/L. His left leg is swollen,
to the approach in a nonsurgical patient; however, the risk of and one patch of skin around his left ankle is gangrenous. His
postoperative bleeding with anticoagulation and thrombolysis PT, PTT, fibrinogen, and D-dimers are all normal. Based on the
must be carefully considered. 4Ts, you decide he has a high probability of HIT and recom
mend changing all anticoagulation to non-heparin products and
sending specific HIT testing.
Common inpatient consultations

Any thrombocytopenic condition could become an emer-
This section focuses on two common hematological consulta- gency if it is severe and serious bleeding occurs (eg, intracra-
tions in hospitalized patients: thrombocytopenia and anemia. nial hemorrhage). But some thrombocytopenic disorders
are emergencies in themselves regardless of the degree of
thrombocytopenia because of their associated risk of signifi-
Thrombocytopenia
cant morbidity and mortality if not promptly recognized
Thrombocytopenia, defined as a platelet count less than and managed. These include: drug-induced immune throm-
150 × 109/L, is one of the most common reasons for hematol- bocytopenia (DITP), heparin-induced thrombocytopenia
ogy consultation in the hospitalized patient. In a registry of (HIT), thrombotic thrombocytopenic purpura (TTP), sepsis
>64,000 patients admitted to the hospital with acute coronary and DIC, catastrophic antiphospholipid antibody syndrome
syndromes, 6.8% had thrombocytopenia at baseline and 13% (CAPS), and posttransfusion purpura (PTP). These diagno-
developed it during their hospital stay. In a study of 2,420 hos- ses should be considered initially for any patient with throm-
pitalized medical patients receiving heparin for at least 4 days, bocytopenia. Consideration of the peripheral blood film and
36% developed thrombocytopenia. A systematic review of other laboratory values, clinical and medication history, and
6,894 critically ill patients reported that thrombocytopenia timing of thrombocytopenia can help with early recognition
occurred in 8%-68% of patients on admission to the intensive and treatment.
care unit (ICU) and developed in 13%-44% during their ICU
stay. The main challenges in the management of hospitalized
Drug-induced immune thrombocytopenia (DITP) and
patients with thrombocytopenia are to identify the underlying
heparin-induced thrombocytopenia (HIT)
cause and to recognize when urgent intervention is required.
A traditional approach to thrombocytopenia is to classify DITP is characterized by severe thrombocytopenia and may
etiologies into conditions of decreased platelet production, be associated with serious bleeding complications. It is usu-
increased platelet destruction, or sequestration. Although this ally an idiosyncratic reaction caused by drug-dependent
framework is comprehensive, it does not consider features platelet-reactive antibodies that cause rapid platelet clearance
related to the individual patient. Furthermore, many disor (see Chapter 10). An expanded list of drugs and the level of
ders have >1 mechanism of thrombocytopenia (eg, immune evidence for their association with thrombocytopenia has

Common inpatient consultations | 31
• Drug-induced immune thrombocytopenia (DITP)

• Heparin-induced thrombocytopenia (HIT)
• Thrombotic thrombocytopenic purpura (TTP)
Exclude thrombocytopenic
1 • Disseminated intravascular coagulation (DIC)
emergencies
• Catastrophic antiphospholipid antibody syndrome (CAPS)
• Primary immune thrombocytopenia (ITP) with bleeding
• Posttransfusion purpura (PTP)
• Platelet clumping: pseudothrombocytopenia

• Schistocytes: TTP, DIC
• Large platelets: inherited thrombocytopenia, ITP
2 Examine the blood smear • Small platelets: Wiskott-Aldrich syndrome, X-linked thrombocytopenia
• Poikilocytes, nucleated RBCs: myelophthisis
• Leukocyte abnormalities: hematologic malignancy or myelodysplasia
• Granulocyte Döhle bodies: MYH9-RD
• Postoperative patient: dilutional, HIT

• Medical history: HIV, hepatitis C
• Isolated thrombocytopenia: ITP
• ICU: sepsis, drugs, DIC, etc.
3 Consider the clinical context • Neonates: consider neonatal alloimmune thrombocytopenia,
HIT is uncommon
• Cancer patients: DIC, TTP
• Pregnancy: gestational thrombocytopenia, ITP, preeclampsia,
HELLP syndrome, TTP/HUS, acute fatty liver of pregnancy
• <20 × 109/L: Primary ITP (including DITP), TTP/HUS/DIC

Assess the degree of • 20–100 × 109/L: HIT (typical nadir is 60 × 109/L), gestational
4 thrombocytopenia thrombocytopenia (70 × 109/L)
• 100 × 109/L: splenomegaly/hypersplenism
Determine the timing of • Within 5–10 days: HIT or DITP; PTP

5 thrombocytopenia exposures • Within hours: tirofiban, eptifibatide, abciximab, HIT (rapid onset)
• Bleeding signs present: ITP, DITP, PTP

Assess for signs of bleeding
6 • Bleeding signs absent: HIT, TTP, APS
or thrombosis
• Thrombosis present: HIT
Figure 2-2 Practical approach to the patient with thrombocytopenia. Adapted with permission from Arnold DM, Lim W. Semin Hematol.
2011;48:251-258.
been reported online (http://www.ouhsc.edu/platelets). Clas- or LMWH compared with subcutaneous prophylactic doses.
sic DITP reactions, such as quinine-induced DITP, result in The 4Ts probability scale can be used to assess a patient’s
thrombocytopenia that occurs 5-10 days after first exposure likelihood of having HIT. With this scale, the degree and
to the drug, whereas the glycoprotein IIb/IIIa inhibitors timing of thrombocytopenia, presence of thrombosis, and
abciximab and eptifibatide can cause thrombocytopenia possible alternative causes of thrombocytopenia are each
within hours of the first drug exposure. independently considered on a scale of 0-2 and then summed
Heparin-induced thrombocytopenia (HIT) is a distinct together. A low score of 0-3 indicates a <1% probability of
clinical syndrome associated with thrombosis rather than HIT, an intermediate score of 4-5 indicates an approximate
bleeding. The prevalence is estimated to be between 0.1% 10% probability of HIT, whereas a high score of 6-8 is associ-
and 5.0%. HIT should be considered in patients with use of ated with an approximate 50% probability of HIT. The diag-
heparin who develop new onset of thrombocytopenia or nosis of HIT can be confirmed with either antigen or
thrombosis. Classically, patients present within 5-10 days of functional assays; however, the ASH Choosing Wisely cam-
heparin exposure; however, HIT can occur more rapidly paign recommends against testing or treating for suspected
(≤1 day) in patients who have had heparin exposure within HIT in patients with a low pre-test probability score. Anti-
30-100 days. The risk of HIT is highest with unfractionated gen assays for anti-PF4-heparin antibodies lacks specificity
heparin and less with low molecular weight heparin. Fur- and may lead to false-positive results in critically ill patients
ther, the risk of HIT is higher with therapeutic doses of UFH and functional platelet-activation tests, such as the serotonin

release assay, should be used to confirm the diagnosis. Treat- (diarrhea-negative HUS) and is associated with a higher
ment of HIT requires anticoagulation with a nonheparin incidence of end-stage kidney disease and mortality. This
alternative. Without proper treatment, up to 55% of patients form occurs more commonly in adults and often is caused
develop thrombosis, and approximately 5%-10% of patients by a dysregulation of the complement system. Mutations in
will die as a result of thrombotic complications. genes encoding complement proteins, including factor H,
membrane cofactor protein (CD46), factor I, and factor B
and C3, have been described. Like TTP, management of
Thrombotic thrombocytopenic purpura (TTP)
atypical HUS requires prompt recognition and empiric ini-
and hemolytic uremic syndrome (HUS)
tiation of plasma exchange. Prospective data suggest that
TTP and HUS are thrombotic microangiopathies character- complement inhibition with the monoclonal antibody eculi-
ized by microangiopathic hemolytic anemia and thrombocy- zumab that targets C5 may improve renal function and
topenia. These disorders should be considered in any patient hematologic parameters while allowing for discontinuation
with anemia, thrombocytopenia, and schistocytes on periph- of plasma exchange in patients with atypical HUS.
eral blood film in the absence of another identifiable cause
such as DIC. The clinical manifestations of these disorders
Disseminated intravascular coagulation and sepsis (DIC)
overlap, however; patients with TTP often have neurological
complications, whereas renal impairment predominates in DIC occurs in critically ill patients in the setting of a serious
HUS (see Chapter 10). TTP results from either a congenital underlying disease, such as sepsis, classical meningococcemia,
deficiency of ADAMTS13 (A Disintegrin And Metalloprotein- trauma, malignancy, and pregnancy catastrophes, including
ase with a ThromboSpondin type 1 motif), a vWF-cleaving placental abruption and amniotic fluid embolism. DIC also
protease, or an acquired antibody which can be either neutral- may complicate poisoning, major hemolytic transfusion reac-
izing or non-neutralizing against ADAMTS13 activity. While tions, and severe HIT. DIC is caused by enhanced thrombin
testing of both ADAMTS13 levels and antibodies is available, generation as a result of an imbalance in the normal procoag-
treatment should not be withheld while waiting on results ulant and anticoagulant pathways and results in a microangi-
if clinical index if suspicion is high. With proper treatment, opathic hemolytic anemia. As a result, many patients develop
survival of TTP patients is 85%; however, without it, survival significant thrombotic complications, including peripheral
drops to 10%. Management requires early institution of daily ischemia and skin gangrene. The clinical features are variable,
plasma exchange with 1.0-1.5 plasma volumes. If plasma and numerous tests of hemostasis become abnormal demon-
exchange cannot be initiated, then an infusion of FFP may be strating schistocytes, thrombocytopenia, increased fibrin deg-
given until such time that plasma exchange can occur. In radation products such as D-dimers, prolongation of the PT
patients with congenital ADAMTS13 deficiency or in those and aPTT, decreased fibrinogen concentration, and decreased
with acquired ADAMTS13 inhibition in whom plasma protein C concentration. A significant reduction in the level of
exchange cannot be initiated promptly, infusion of FFP to fibrinogen may indicate early or subclinical DIC even if it does
replace ADAMTS13 should be initiated. not result in hypofibrinogenemia. DIC is a dynamic process
Remission in TTP can be defined as either clinical or labo- requiring repeated measurements of hemostasis and careful
ratory. Some patients may experience a clinical remission but clinical monitoring.
continue to have reduced levels of ADAMTS13 and/or detect- Guidelines and consensus statements for the manage-
able antibodies. Approximately 35% of patients experience a ment of DIC highlight the importance of treating the under-
relapse following plasma exchange. Relapse rates are highest lying condition even though this can be challenging. While
among patients with ADAMTS13 levels of <10%, males, and evidence is lacking to clearly guide the use of prophylactic
during the first year of remission. In acquired TTP patients platelet transfusions, they should generally be reserved for
who are refractory or relapse despite plasma exchange, con- patients with a platelet count below 50 × 109/L, who are at
sideration can be given to the addition of immunosuppres- high risk of bleeding, or who have worsening thrombocyto-
sive agents such as corticosteroids and/or rituximab. penia. Similarly, plasma transfusions are primarily reserved
Unlike TTP, patients with HUS often do not require or for patients with an increased PT and bleeding, and cryopre-
respond to plasma exchange. The most common form of cipitate or fibrinogen concentrates are indicated for patients
HUS is associated with bloody diarrhea and is caused by with severe hypofibrinogenemia (fibrinogen <100 mg/dL).
enteric infection with strains of Escherichia coli that produce Prophylactic doses of UFH or LMWH are recommended
Shiga-like toxins (typical HUS or diarrhea-positive HUS). for prevention of venous thromboembolism, and therap
This variant accounts for up to 95% of all HUS in children, eutic doses should be considered for patients with throm
often occurs in epidemics, and generally is self-limited. The botic complications such as venous or arterial thrombosis,
atypical form of HUS occurs without a diarrheal prodrome severe purpura fulminans, or vascular skin infarctions.

Several coagulation factor concentrates have been investi- platelet-containing products and demonstrating circulating
gated for the treatment of severe sepsis and DIC. In clinical alloantibody to HPA-1a antigen in a patient whose own plate-
trials, antithrombin III (ATIII) showed some clinical benefit lets lack this antigen. Patients with PTP who require transfu-
in reducing the symptoms of DIC. Despite this, there was no sions for bleeding or severe thrombocytopenia should receive
survival advantage because of increased bleeding risk when HPA-1a-negative blood products if available.
used in combination with heparin; therefore, it is not con-
sidered standard of care. Further studies are needed, how-
2. Examine the blood film
ever, to investigate the use of ATIII alone.
Examination of the blood film is necessary for all patients
Catastrophic antiphospholipid antibody syndrome (CAPS) with thrombocytopenia. Platelet clumps are suggestive of
pseudothrombocytopenia, a laboratory artifact caused by
CAPS occurs in <1% of patients with the antiphospholipid
naturally occurring antibodies directed against the antico-
antibody syndrome. It is a life-threatening condition that
agulant ethylenediaminetetraacetic acid (EDTA). A repeat
requires prompt recognition and management. Diagnostic
sample collected in citrate or heparin tube usually resolves
criteria for CAPS are: (i) involvement of three or more
the platelet clumping. The size and morphology of the pla
organs, systems, or tissues; (ii) development of symptoms
telets can be assessed, and findings such as large platelets
simultaneously or in <1 week; (iii) confirmation by histopa-
can indicate a state of high platelet turnover, such as ITP.
thology of small vessel occlusion in at least one organ or tis-
The blood film also allows for morphological assessment of
sue; and (iv) laboratory confirmation of the presence of
erythrocytes and leukocytes, which may provide important
antiphospholipid antibodies (lupus anticoagulant; or anti-
clues to the underlying diagnosis: the presence of schisto-
cardiolipin or anti-b-2-glycoprotein I antibodies). A registry
cytes raise the possibility of a microangiopathic process such
of patients with CAPS has provided important information
as TTP or DIC; poikilocytes or nucleated RBCs may reflect a
on diagnosis and management (http://www.med.ub.es/
myelophthisic process; abnormal leukocytes may indicate a
MIMMUN/FORUM/CAPS.HTM). Infection is the most
hematologic malignancy or myelodysplasia; toxic granula-
commonly identified precipitant, but other triggers such as
tion of neutrophils are seen in sepsis; and neutrophilic inclu-
trauma, withdrawal of anticoagulation, and neoplasia have
sions known as Döhle bodies are associated with hereditary
also been described. Approximately 40% of patients with
forms of thrombocytopenia, such as the MYH9-related dis-
CAPS have no obvious underlying cause. Mortality can be
orders (MYH-RD).
up to 50%, and, therefore, most would recommend initia-
tion of aggressive multimodality therapy in addition to
plasma exchange to achieve the highest response rates. This 3. Consider the clinical context
approach is supported by data from the “CAPS Registry”
The clinical context in which the thrombocytopenia devel-
with use of multiple agents being reported in the manage-
oped is an important clue to the underlying diagnosis. Medi-
ment; anticoagulation (87%), corticosteroids (86%), cyclo-
cal history may reveal a source for thrombocytopenia such as
phosphamide (36%), immunoglobulin (IVIg) (22%),
medications, liver disease, or cause of secondary ITP such as
anti-platelet agents (10%). Rituximab has also been used
HIV or hepatitis C. Thrombocytopenia is a common occur-
with some success in more refractory cases.
rence among critically ill patients, particularly those with
underlying malignancies prone to DIC or thrombotic micro-
Posttransfusion purpura (PTP)
angiopathies. Age also helps narrow the differential diagno-
PTP is a syndrome characterized by severe thrombocytopenia sis; for example, neonatal alloimmune thrombocytopenia
and bleeding that develops 7-10 days after the transfusion of (NAIT) should be suspected in any newborn with severe
any platelet containing blood product (such as platelet or RBC unexpected thrombocytopenia, and HIT is distinctly rare in
concentrates). It typically affects women who have had a pre- children. Thrombocytopenia during pregnancy should lead
vious pregnancy or blood transfusion and most commonly is to consideration of gestational thrombocytopenia, ITP, or
due to antibodies against human platelet antigen (HPA)-1a. more severe conditions such as preeclampsia, HELLP syn-
The incidence of PTP is estimated at 1-2 per 100,000 transfu- drome, TTP/HUS, or acute fatty liver of pregnancy.
sions, and it appears to be less common with leukodepleted
blood products. The pathophysiology remains uncertain, but
Thrombocytopenia in patients admitted to the ICU
may involve the formation of immune complexes, adsorption
of soluble platelet antigens onto autologous platelets, or the Approximately 40% of critically ill patients have thrombocy-
induction of platelet autoantibodies. Diagnosis involves rec- topenia; however, the frequency varies based on case mix. In a
ognizing thrombocytopenia that occurs after transfusion of systematic review of medical, surgical, and mixed ICU studies,

prevalent thrombocytopenia (on ICU admission) occurred in of 60 × 109/L; whereas mild thrombocytopenia can be the
8.8%-67.6% of patients, and incidental thrombocytopenia result of splenomegaly, primary bone marrow failure, and
(occurring during the course of the ICU stay) occurred in cogenital thrombocytopenias. In patients with sepsis, plate-
13.1%-44.1% of patients. Thrombocytopenia was an inde- let counts are variable but thrombocytopenia tends to be
pendent risk factor for mortality. The association between mild or moderate. Gestational thrombocytopenia typically
thrombocytopenia and bleeding remains uncertain in this presents with platelet counts of greater than 70 × 109/L
population and is likely based on additional patient factors. which often helps distinguish it from ITP in pregnancy.
HIT in the ICU

5. Establish the timing of onset of thrombocytopenia
The frequency of HIT in ICU patients is 0.3%-0.5%, which
represents roughly 1 in 100 patients with thrombocytopenia The documentation of a normal platelet count before the
in this setting; thus, HIT is uncommon in this population. acute illness is helpful in narrowing the cause of thrombocy-
The diagnosis and management of HIT in critically ill patients topenia. A search for exposures to drugs or blood transfu-
can be challenging. After major surgery, a rapid decline in sion is important. Immune-mediated platelet disorders,
platelet count occurring by days 1-3 is expected due to con- including classic HIT, DITP, and PTP, typically occur after
sumption and dilution; however, delayed platelet count 5-10 days after exposure; however, certain drugs such as
recovery beyond day 7, worsening thrombocytopenia between tirofiban, eptifibatide, or abciximab may cause thrombocy-
days 5 and 14, or the development of new thrombosis in an topenia within hours of the first exposure. Rapid-onset HIT
already thrombocytopenic patient may indicate HIT. An can occur after reexposure to heparin when platelet-reactive
expanded discussion on HIT can be found in the section antibodies are already present, and delayed-onset HIT is
above and in Chapter 10. characterized by thrombocytopenia and thrombosis occur-
ring several weeks after heparin exposure.
Immune thrombocytopenia
Severe isolated thrombocytopenia in an otherwise well 6. Assess for signs of bleeding and/or thrombosis
individual may represent ITP. Many patients with ITP will Typical platelet-type bleeding presents as petechiae or bruis-
have minimal bleeding despite significant thrombocytope- ing; oral petechiae or purpura; and gastrointestinal, genito-
nia. Additional risk factors such as patient age, comorbidi- urinary, or intracerebral hemorrhage. Bleeding is common
ties, and medications may increase an individual patient’s in patients with DITP, severe primary ITP, and in newborns
risk of bleeding. First-line therapy for adults with ITP is a with NAIT. Despite the presence of thrombocytopenia,
course of corticosteroids. For those failing first-line ther- however, bleeding is rare in HIT and TTP, because these are
apy, additional treatment options include splenectomy, predominantly prothrombotic disorders and therefore the
rituximab, and thrombopoietin receptor agonists. Overall findings of thrombosis maybe more diagnostic.
treatment should be aimed at reducing bleeding symptoms
and improving health-related quality of life. The ASH
Choosing Wisely campaign recommends no treatment for Anemia
adults with ITP in the absence of bleeding or a very low plate- Perioperative transfusion and ICU setting
let count, usually defined as <30 × 109/L, in order to avoid the
unnecessary cost and side effects associated with treatment. Anemia is a common problem among hospitalized patients,
The decision to treat should be individualized for each patient especially in the ICU and the perioperative setting. Approxi-
and account for the patient’s symptoms, additional risk fac- mately 25%-30% of patients in the ICU will have a hemoglo-
tors for bleeding, social factors such as distance from the hos- bin (Hb) level <9 g/dL and approximately one-third of
pital, side effects of possible therapy, any upcoming critically ill patients will receive a RBC transfusion at some
procedures, and patient preferences. point during their ICU stay. Over the past decade, consider-
able debate has centered on the role of RBC transfusion in
critically ill and perioperative patients, largely based on the
4. Consider the severity of thrombocytopenia
realization that transfusion may be associated with an
The severity of thrombocytopenia is an important clue to the increase in infectious risk, postoperative complications, and
diagnosis. Significant thrombocytopenia, defined as platelet overall mortality. Therefore, the threshold for RBC transfu-
counts <20 × 109/L is typical of primary or secondary ITP, sion in ICU and surgical patients has changed over time.
DITP, and microangiopathic processes such as TTP/HUS/ A landmark trial investigating the benefit of a liberal or
and DIC. HIT generally causes a median platelet count nadir a restrictive transfusion strategy in the ICU was the

Transfusion Requirements in Critical Care (TRICC) trial. In overall transfusion requirements; and clerical error, resulting
this trial, 838 critically ill patients with hemoglobin values in inadvertent administration of an allogeneic product, it is
<9 g/dL were randomized to a transfusion strategy that main- recommended that PAD be restricted to healthy individuals
tained hemoglobin concentrations between 10 and 12 g/dL or requiring blood-intensive surgeries in which the likelihood
a transfusion strategy that maintained hemoglobin concen- of blood loss in excess of 500-1,000 mL is at least 5%-10%.
tration between 7 and 10 g/dL. Overall, there was no signifi-
cant difference in 30-day mortality; however, in-hospital Refusal of blood
mortality was significantly lower in the restrictive strategy
Not uncommonly, hematologists are asked to provide consul-
group. This study suggests that RBC transfusion is generally
tation for patients who refuse blood transfusions (eg, Jeho-
not required for hemoglobin concentrations >7 g/dL in the
vah’s Witnesses), including autologous blood transfusions.
ICU. The findings of a 2012 Cochrane meta-analysis show
Most Jehovah’s Witnesses do not accept any of the four major
that, in general, a liberal RBC transfusion strategy (transfu-
components of whole blood (ie, red blood cells, platelets,
sion for <10 g/dL) compared with a restrictive strategy (trans-
plasma, and white blood cells). Whether or not one would
fusion for Hb 7-8 g/dL) does not improve clinical outcomes,
accept blood subfractions, such as immunoglobulins, albu-
and a restrictive transfusion strategy is as safe, if not safer.
min, and coagulation factor concentrates, varies between
While different guidelines have recommended that transfu-
individuals. For this reason, it is vital that physicians engage
sion is not indicated for Hb >10 g/dL the lower threshold var-
Jehovah’s Witnesses in shared decision making and for
ies from 6 g/dL to 8 g/dL. The American Academy of Blood
patients to make clear what they will or will not accept, even if
Banks guidelines recommend that in hemodynamically stable
death is imminent. Treatment of anemia in this patient popu-
patients without active bleeding, a transfusion threshold of
lation is a challenge in the medical and surgical setting. Blood
7 g/dL to 8 g/dL should be adopted. In certain high-risk pop-
conservation and the use of adjunctive therapies remain the
ulations (eg, preexisting cardiovascular disease) or those with
mainstay of treatment in Jehovah’s Witnesses with anemia or
symptoms of chest pain, orthostatic hypotension, tachycardia
preoperatively in anticipation of a fall in hemoglobin. Blood
unresponsive to fluid resuscitation, or congestive heart fail-
conservation includes minimizing daily phlebotomy for rou-
ure, transfusion should be considered at hemoglobin concen-
tine labs, utilizing small volume sampling, and careful atten-
trations of <8 g/dL. The Transfusion Requirements in Septic
tion to minimizing blood loss intraoperatively. Supportive
Shock (TRISS) trial showed that a threshold of 7 g/dL com-
measures to prevent or treat anemia include swiftly stopping
pared to 9 g/dL was as safe in patients with septic shock. In
blood loss, stimulating erythropoiesis (eg, recombinant
postoperative surgical patients, including those with stable
human erythropoietin, intravenous iron, folic acid and vita-
cardiovascular disease, transfusion should be considered at a
min B12 supplementation), and maintaining blood volume.
hemoglobin concentration of <7 g/dL to 8 g/dL. In hospital-
ized stable patients with acute coronary syndrome, evidence
Blood storage
is lacking on the optimal transfusion strategy, although
experts suggest transfusion for Hb <8 g/dL and consideration One recent area of controversy is the effect of blood storage
of transfusion between 8 g/dL and 10 g/dL. time on clinical outcomes of transfusion recipients. A recent
Several subsequent studies have supported these observa- large randomized trial examined the effect of red blood stor-
tions, and recent surveys suggest that transfusion practices age time on change in multiple organ dysfunction (MOD)
have changed toward a more restrictive approach. The deci- scores in patients undergoing cardiac surgery. Patients were
sion to transfuse should be based on an individualized randomly assigned to received red cells for less than 10 days
assessment of the patient’s clinical status, oxygen delivery or greater than 21 days. The authors found no difference in
needs, and the pace of fall in hemoglobin rather than on a change in MOD scores, mortality, or major adverse events
predetermined hemoglobin trigger. Accordingly, the ASH between the two groups. This is in contrast to results from a
Choosing Wisely campaign recommends transfusion of the meta-analysis of 409,966 patients that included 21 studies,
smallest effective dose to relieve symptoms of anemia or to which found that older blood was associated with a signifi-
restore the patient to a safe hemoglobin range. cantly increased risk of death (odd ratio, 1.16; 95% confi-
Most RBC transfusions administered in the perioperative dence interval, 1.07-1.24). This represents data mostly
setting are allogeneic. Autologous RBCs, collected through derived from observational studies or small randomized tri-
preoperative autologous donation (PAD) or intraoperative als from select patient populations. Two additional large ran-
blood salvage, remain an option for some patients. However, domized control trials (RCTs), Age of Blood Evaluation trial
due to increased costs; risk of bacterial growth during liquid and Age of Blood in Children in Pediatric Intensive Care
storage; volume overload; hemolysis from improper han- Units trial, are currently are under way to address the impact
dling of stored units; lack of benefit with regards to decreased of age of red blood cells in different populations.

Drug-related complications
Key points
Immunosuppressant and antimicrobial drugs are prevalent
• Life-threatening causes of thrombocytopenia should be
causes of cytopenia after solid organ transplantation. Aza-
considered first in any patient presenting with thrombocytopenia:
thioprine is particularly problematic, causing cytopenias in
DITP, HIT, TTP, sepsis and DIC, CAPS, and PTP.
• Prompt treatment should be provided to any patient with a sug
approximately 10% of patients. Because azathioprine and its
gested life-threatening cause of thrombocytopenia while awaiting principal metabolite are cleared predominantly by the kid-
any definitive laboratory diagnosis. ney, azathioprine-induced marrow toxicity is common fol-
• The diagnosis of TTP should be considered in any patient with lowing rejection of a renal allograft. Azathioprine toxicity is
thrombocytopenia and microangiopathic hemolytic anemia. exacerbated by allopurinol, angiotensin-converting enzyme
• DIC is caused by enhanced thrombin generation. Management inhibitors, and trimethoprim/sulfamethoxazole, which fre-
is aimed primarily at treating the underlying cause. quently are prescribed in the posttransplant setting.
• Examination of the peripheral blood film should be part of the Thrombotic microangiopathy occasionally occurs within
investigations for any patient presenting with thrombocytopenia. the first few weeks after solid organ transplantation in
• HIT is an uncommon cause of thrombocytopenia in patients
patients treated with calcineurin inhibitors, such as cyclo-
admitted to the ICU.
sporine or tacrolimus. In renal transplant patients, this entity
• For most patients, RBC transfusions are not required for
may be difficult to distinguish from hyperacute humoral
non-bleeding critically ill patients with a hemoglobin concentra
tion >7 g/L or in surgical patients with a hemoglobin concentra
rejection of the allograft without a renal biopsy. Pathologic
tion of >8 g/dL. The decision to transfuse should be based on an evidence of thrombotic microangiopathy usually is restricted
individualized assessment of the patient’s clinical status, oxygen to the kidneys and often responds to switching, reducing, or
delivery needs, and the rate of decline in hemoglobin rather than withdrawing the offending drug. Solid organ transplanta-
on a predetermined hemoglobin trigger. tion-related TMA, in contrast to idiopathic TTP, is generally
not associated with a severe ADAMTS13 deficiency. Although
plasma exchange may be attempted in refractory cases, there
is little evidence to support its use in this setting.
Consultation for hematologic
Clinicians should also be aware that drugs not specific to
complications of solid organ
solid organ transplantation but used in supportive care can
transplantation
cause cytopenias; for example, folate deficiency from the
administration trimethoprim/sulfamethoxazole, hemolysis
This section offers an approach to the patient with hemato-
from dapsone or trimethoprim/sulfamethoxazole prophylaxis
logic complications after solid organ transplantation. One of
in patients with unrecognized glucose-6-phosphate dehydro-
the most common reasons for hematologic consultation in
genase deficiency, or drug-induced hemolytic anemia from
this setting, as illustrated by the clinical case, is single lineage
beta-lactam antibiotics or trimethoprim/sulfamethoxazole.
or multilineage cytopenia.
Infectious complications
Clinical case
While immunocompromised solid organ transplantation
You are consulted on a 33-year-old woman with thrombo recipients are at risk for various opportunistic infections,
cytopenia. She underwent renal transplantation 3 weeks ago those associated with posttransplant cytopenias include par-
for end-stage diabetic nephropathy. Over the past week, she
vovirus B19, cytomegalovirus (CMV), human herpesvirus 6
has developed abdominal pain, fever, and increased bruising.
(HHV6), and Epstein-Barr virus (EBV). Strategies for CMV
Her laboratory studies demonstrate a white blood cell count
surveillance posttransplant are well established. CMV vire-
of 5,000/mL, a hemoglobin of 7.5 g/dL, a platelet count of
52,000/mL, and a serum creatinine of 2.6 mg/dL. There is
mia is associated with leukopenia and thrombocytopenia.
evidence of nonimmune hemolysis with an elevated lactate Reactivation of latent HHV6 is common after transplant
dehydrogenase, reticulocytosis, reduced hemoglobin, and nega but is rarely associated with clinically significant disease.
tive direct Coombs test. She is taking prednisone and tacrolimus. Clinically significant HHV6 reactivation manifests as leuko-
Blood cultures and viral DNA testing are negative. A peripheral penia, although other cell lines can be affected. Peripheral
blood smear reveals 5-7 schistocytes per high power field. An blood quantitative polymerase chain reaction (PCR) is the
ADAMTS13 activity returns as 20%. A renal biopsy revealed preferable method of viral detection in immunocompro-
thrombotic microangiopathy without evidence of graft rejection. mised patients. The first-line treatment for both CMV and
You recommend discontinuation of tacrolimus and other nones HHV6 viremia includes ganciclovir or foscarnet. Parvovirus
sential medications.
B19 infection in immunocompromised patients specifically

Consultation for hematologic complications of solid organ transplantation | 37
targets erythroid-lineage cells causing a pure red cell aplasia been reported secondary to incompatibilities with the c, e,
associated with anemia, marked reticulocyopenia, and ery- JK(a), K, and Fy(a) antigens. Hemolysis is abrupt and occurs
throid maturation arrest at the pronormoblast stage. Diag- several days after transplantation. In addition to classic lab-
nosis can be confirmed by enzyme-linked immunosorbent oratory markers of hemolysis, the direct Coombs test is
assay for anti-B19 specific antibodies, quantitative PCR positive and serum antibodies against a target recipient red
for parvovirus B19 DNA, or by classic findings of giant cell antigen are detectable. Most cases can be treated with
pronormoblasts stage arrest in the bone marrow. Treat- red cell transfusions of organ donor ABO group compati
ment includes reduction of posttransplant immunosup- bility. If hemolysis persists, other treatments include escala-
pressive drugs, intravenous immunoglobulin (IVIg), and tion of immunosuppression, intravenous immunoglobulin,
erythropoietin. rituximab, red cell exchange to remove incompatible host-
origin red blood cells, or plasma exchange to remove donor
lymphocyte mediated antibodies. Passenger lymphocyte
Alloimmune complications syndrome is typically self-limited due to the short survival of
Graft-versus-host disease (GVHD) is a rare and often fatal donor lymphocytes in the circulation.
complication of solid organ transplantation. It is caused by
alloreactive passenger T lymphocytes in the transplanted Posttransplantation lymphoproliferative
organ. The risk of GVHD is related, in part, to the dose of disorders (PTLDs)
transplanted lymphocytes. Of all solid organ transplanta-
tion, patients receiving small bowel or liver transplantation PTLDs make up a group of predominantly B-cell neoplasms
receive the largest dose of passenger lymphocytes. As such, that occur in immunosuppressed individuals following solid
organ transplantation. In most cases, B-cell proliferation is
donors are typically treated with antilymphocyte antibodies
induced by Epstein-Barr virus (EBV) infection. PTLD affects
or corticosteroids before organ harvesting to minimize the
~1% of solid organ recipients and typically occurs within the
transplantation of donor T lymphocytes. GVHD in solid
first year. It is due to impairment of EBV-specific, cytotoxic
organ tranplanation patients presents similarly to acute
T-cell function by immunosuppression that allows for
GVHD after hematopoietic stem cell transplantation. Fever,
expansion of the latent EBV-infected B cells. Principal risk
rash, and diarrhea 2-6 weeks after transplantation are com-
factors for the development of this complication include
mon initial complaints. Cytopenias, due to GVHD directed
greater intensity of immunosuppression and receipt of a
against host hematopoietic cells, also may occur and must be
solid organ from an EBV-seropositive donor by an EBV-
distinguished from more common causes of cytopenias in
seronegative recipient.
the posttransplant setting, such as drugs and infection. The
Three types of EBV-related PTLD are recognized: benign
diagnosis may be confirmed by biopsy of the skin or other
polyclonal lymphoproliferation, which presents 2-8 weeks
affected organs and by peripheral blood chimerism studies, after initiation of immunosuppression and resembles infec-
which quantify the proportion of circulating lymphocytes tious mononucleosis in presenting symptoms; polyclonal
that are of donor and recipient origin. There is no standard lymphoproliferation with early evidence of malignant trans-
therapy for this rare disease. Management generally includes formation; and monoclonal B-cell proliferation with evidence
supportive care and immunosuppressive agents. Prognosis is of malignancy by cytogenetics and immunoglobulin gene
poor and death is typically due to infection from severe mar- rearrangements. Patients may present with constitutional
row aplasia and multiorgan system failure. symptoms, cytopenias, or lymphadenopathy. Extranodal dis-
Another alloimmune complication of solid organ trans- ease is common. Involved organs include the gastrointestinal
plantation is alloimmune hemolysis of host erythrocytes by tract, lungs, skin, liver, central nervous system, and the
antibodies produced by donor lymphocytes also known as allograft itself. The different types of PTLD are diagnosed by
passenger lymphocyte syndrome. Like GVHD, passenger a combination of histologic features (eg, underlying architec-
lymphocyte syndrome is more common in transplantations ture), clonality (polyclonal versus monoclonal), immuno-
containing greater numbers of lymphocytes. The syndrome globulin gene rearrangements, and EBV positivity within the
is most common after small bowel transplantation, followed context of the clinical scenario.
by heart-lung, liver, and kidney transplantation. Passenger Treatment depends on the type of PTLD. Benign poly-
lymphocyte syndrome occurs when donor memory B lym- clonal lymphoproliferation and polyclonal lymphoprolifera-
phocytes are stimulated after transplant by exposure to tion with early evidence of malignancy typically are managed
recipient or transfused red-cell antigens leading to antibod- with a reduction of immunosuppression and antiviral agents.
ies directed against these antigens. Most cases are due to Immunosuppression must be reduced cautiously to reduce
ABO or Rh(D) incompatibility, but the syndrome also has the risk of allograft rejection. Patients with monoclonal

PTLD rarely respond to reduction of immunosuppression Because of its rarity, irradiation of blood products is not
alone. If the PTLD expresses CD20, rituximab may be used routinely recommended for recipients of solid organ trans-
alone or in combination with chemotherapy. Single-arm plantations in contrast to recipients of hematopoietic stem
studies suggest response rates of 40%-70% with rituximab, cell transplants.
although RCTs have not been reported. Radiation therapy
may be used for treatment of local disease.
Posttransplantation erythrocytosis (PTE)
PTE is defined as an elevated hematocrit exceeding 51% that

Transfusion support
occurs following renal transplantation and persists for more
A hematology consultant may be asked to assist with trans- than 6 months in the absence of leukocytosis, thrombocyto-
fusion management in a patient undergoing solid organ sis, or another potential cause of primary or secondary eryth-
transplantation. Of all solid organ transplantations, RBC, rocytosis. PTE affects 8%-15% of renal transplant recipients;
plasma, and platelet transfusion is most commonly required however, the incidence appears to be decreasing. The patho-
for liver transplantation due to the underlying coagulopathy physiology of PTE is poorly understood, but it likely involves
of liver failure. Heart and heart-lung transplantations fre- dysregulation of the renin-angiotensin system.
quently require transfusion support, whereas kidney and PTE classically presents 8-24 months after transplanta-
kidney-pancreas transplantation generally do not require tion. Clinical manifestations include malaise, plethora,
blood product replacement. headache, and a propensity for both venous and arterial
Transfusion therapy for solid organ transplantation car- thromboembolism similar to patients with polycythemia
ries the potential risks of infection, human leukocyte antigen vera. First-line therapy in patients with a hemoglobin con-
(HLA) alloimmunization, and, rarely, transfusion-associ- centration between 17 and 18.5 g/dL is with an angiotensin-
ated GVHD. The most frequent transfusion-associated converting enzyme inhibitor or an angiotensin receptor
infection complicating solid organ transplantation is CMV. blocker. In patients who do not respond to medical therapy
Although CMV viremia usually is due to reactivation in a and in those with a hemoglobin concentration >18.5 g/dL,
seropositive immunocompromised recipient, seronegative therapeutic phlebotomy should be added.
recipients can acquire CMV through transfusion. To prevent
this complication, seronegative recipients should receive
transfusions that are CMV-negative or leukocyte reduced. Key points
In the past, transfusions were administered before trans- • Cytopenias occurring after transplantation of a solid organ may
plantation as a form of immunomodulation to reduce the be due to infection, drugs (most commonly azathioprine), GVHD,
risk of solid organ rejection. More recent randomized stud- or PTLD (if marrow involvement is present).
ies, however, have shown that modern immunosuppressive • Major risk factors for PTLD include greater intensity of immuno
agents are more effective at preventing graft rejection than suppression and receipt of a solid organ from an EBV-seropositive
pretransplantation transfusion. Moreover, exposure to allo- donor by an EBV-seronegative recipient.
geneic lymphocytes may induce anti-HLA antibodies, which • Hemolytic anemia in a solid organ transplantation patient may
increase the risk of acute and chronic rejection. To minimize be due to calcineurin inhibitor-associated thrombotic microangi
opathy or to the passenger lymphocyte syndrome.
this risk, patients expected to undergo kidney, heart, or lung
• Considerations for transfusion support of transplant patients
transplantation should receive blood that is leukocyte
include the risks of HLA alloimmunization; transmission of CMV;
reduced. Because of conflicting data, leukocyte reduction is
and, in rare cases, transfusion-associated GVHD.
considered optional for patients undergoing liver transplan- • PTE occurs 8-24 months after renal transplantation and
tation. Plasma exchange, IVIg, and rituximab have been used responds to medical therapy with an angiotensin-converting
in patients with a positive panel of reactive antibodies (PRAs) enzyme inhibitor or an angiotensin receptor blocker.
or major ABO incompatibility to minimize the risk of hyper-
acute rejection.
Transfusion-associated GVHD is rare among solid organ
transplantation patients, although it is associated with a
Common outpatient hematology
mortality of 90% or higher as a result of severe pancytope-
consultations
nia. The pathophysiology involves engraftment of donor-
This section focuses on some of the most common reasons
derived passenger leukocytes in an immunocompromised
for outpatient hematology consultations. Thrombocytope-
host unable to eliminate these passenger leukocytes. Presen-
nia, leukocytosis, and leukopenia are examined in detail.
tation is similar to transplant-associated GVHD and includes
Anemia is covered in other sections.
skin rash, diarrhea, and liver function abnormalities.

Common outpatient hematology consultations | 39
Mild thrombocytopenia Over-the-counter medications, in particular herbal supple-

ments, should be considered.
As with any hematologic disorder, examination of the
Clinical case peripheral blood film is an essential part of the evaluation.
A 12-year-old boy with a seizure disorder is referred to you be Clumped platelets, as seen with pseudothrombocytopenia;
cause of thrombocytopenia. His platelet count has gradually de large platelets, as seen with certain inherited macrothrombo-
creased from 180 × 109/L to 38 × 109/L over the past 4 months. cytopenic disorders; and small platelets, as seen with Wis-
His only medication is valproic acid, which started 12 months kott-Aldrich syndrome, may reveal important clues.
ago and which has led to good seizure control. He is otherwise Furthermore, abnormal leukocyte or red cell morphology
in good health. He reports no episodes of bleeding, and there can indicate an underlying disease. Hypersegmented neutro-
are no obvious bruises or petechiae on physical examination.
phils and macrocytosis may suggest vitamin B12 deficiency,
In conjunction with his neurologist, you recommend that he
lymphocytosis may suggest underlying chronic lymphocytic
reduce the dose of valproic acid.
leukemia, and circulating blasts are consistent with acute
leukemia. Dysmorphic red blood cells, hypogranulated neu-
Patients with platelet counts in the range of 80 × 109- trophils, or Pelger-Huët cells may suggest underlying myelo-
150 × 109/L are often referred for outpatient hematology con- dysplastic syndrome, which may present with isolated
sultation. Determining the onset of the thrombocytopenia is thrombocytopenia in up to 10% of patients.
important, which inevitably involves tracing back prior blood There are no guidelines as to when or whether the bone
counts. New-onset thrombocytopenia may represent a new marrow should be examined in patients with mild thrombo-
disease process (primary or secondary ITP, bone marrow cytopenia. Although the incidence of a primary bone mar-
infiltration, or myelodysplasia) or a complication of medica- row disorder such as myelodysplastic syndromes increases
tions or infections. Chronic thrombocytopenia or family his- with age, recent epidemiologic studies demonstrate that ITP
tory of thrombocytopenia may suggest the possibility of an is also common in elderly patients. For patients with typical
inherited process, such as a MYH9-related macrothrombocy- ITP (ie, isolated thrombocytopenia without other abnormal-
topenic disorder, which may be first discovered during preg- ities on the peripheral blood film or physical examination
nancy when women often have their blood tested for the first findings), bone marrow examination generally is not required
time. Other causes of thrombocytopenia include systemic (American Society of Hematology [ASH] guidelines). A bone
lupus erythematosus, chronic liver disease typically related to marrow examination should be performed if unexplained
underlying hepatitis C or alcohol with or without hypersplen- symptoms arise or other hematologic abnormalities appear
ism, or deficiency of nutrients required for hematopoiesis to rule out bone marrow pathology. In any case of thrombo-
(vitamin B12, folate, copper). Splenomegaly should be assessed cytopenia, close follow-up of repeat CBCs is warranted to
with a physical examination and ultrasound if appropriate. establish the trend and pace of the thrombocytopenia.
Mild thrombocytopenia itself is not dangerous, but it may
occur as a less severe presentation of a number of disorders
Leukocytosis
that can cause more pronounced thrombocytopenia. Thus,
the patient should be questioned carefully for signs or symp- Patients with unexplained leukocytosis frequently are
toms of infection, autoimmune disease, or malignancy, and referred to a hematologist because of the concern about an
the physical examination should focus on the assessment of underlying hematologic malignancy; however, most patients
lymphadenopathy, hepatosplenomegaly, skin rashes, stig- with unexplained leukocytosis do not have a hematologic
mata of bleeding, and musculoskeletal abnormalities. An malignancy. A common cause of unexplained leukocytosis is
underlying etiology often is not found. Most patients with benign neutrophilia in cigarette smokers.
mild thrombocytopenia (platelet count 100 × 109-150 × In addition to examining the peripheral blood film, a
109/L) that is thought to be due to an immune process can be careful history and physical examination are important.
reassured because after 10 years, the risk of developing more Unexplained fever or chills with a new heart murmur may
severe ITP or another autoimmune disease is low (approxi- suggest infection, such as bacterial endocarditis. A history of
mately 7% and 12%, respectively). DITP was discussed in diarrhea may suggest occult infection with Clostridium diffi-
reference to acutely ill patients but also should be considered cile. Lithium or corticosteroid use may indicate a drug-
in patients with mild thrombocytopenia. Although drugs induced leukocytosis. Examination of the skin, lymph nodes,
such as sulfa-containing antibiotics often cause severe liver, and spleen size is also important. Patients with exuda-
thrombocytopenia, others, including the anticonvulsant tive pharyngitis, splenomegaly, and lymphocytosis may have
drug valproic acid, may cause mild thrombocytopenia (and infectious mononucleosis. The cell type that is elevated lead-
other blood abnormalities) that may be dose dependent. ing to an increase in total leukocyte count also can provide a

Table 2-4 Hematology consultation for leukocytosis: etiologic considerations according to leukocyte subtype affected
Neutrophilia Monocytosis Eosinophilia Lymphocytosis

Eclampsia Pregnancy Allergic rhinitis Mononucleosis syndrome
Thyrotoxicosis Tuberculosis Asthma Epstein-Barr virus
Hypercortisolism Syphilis Tissue-invasive parasite Cytomegalovirus
Crohn disease Endocarditis Bronchopulmonary aspergillosis Primary HIV
Ulcerative colitis Sarcoidosis Coccidioidal infection Viral illness
Inflammatory/ rheumatologic Systemic lupus erythematosus HIV Pertussis
disease
Sweet’s syndrome Asplenia Immunodeficiency Bartonella henselae (cat scratch
disease)
Infection Corticosteroids Vasculitides
Bronchiectasis Juvenile myelomonocytic Drug reaction Toxoplasmosis
leukemia
Occult malignancy Adrenal insufficiency Babesiosis
Trauma/burn Occult malignancy Drug reaction
Severe stress (emotional Pulmonary syndromes Reactive large granular
or physical) lymphocytosis
Panic Gastrointestinal syndromes Chronic lymphocytic leukemia
Asplenia Hypereosinophilic syndrome Monoclonal B cell lymphocytosis
Cigarette smoking Postsplenectomy lymphocytosis
Tuberculosis
Chronic hepatitis
Hereditary neutrophilia
Medications
Corticosteroids
β-agonists
Lithium
G-CSF or GM-CSF
Myeloproliferative neoplasm
(CML, PV, ET)
HIV = human immunodeficiency virus; CML = chronic myelogenous leukemia; PV = polycythemia vera; ET = essential thrombocythemia.
clue to the underlying diagnosis. A concomitant increase in at increased risk of infection. Leukopenia can be further dif-
hemoglobin or platelet count may reflect a myeloprolifera- ferentiated by the specific cell type that is affected. Leukope-
tive neoplasm. Chronic persistent lymphocytosis with an nia results from either decreased marrow production of
absolute lymphocyte count of >5000/microL may be the first leukocytes or from decreased circulation of leukocytes due to
indication of an underlying chronic lymphocytic leukemia. destruction, margination, or sequestration. Neutropenia can
Table 2-4 lists specific causes of leukocytosis according to the be classified as either congenital or acquired. Congenital
predominant cell type that is elevated. Additional laboratory forms typically present in childhood with recurrent infec-
tests such as a bone marrow examination, flow cytometry, tions. For example, patients with cyclic neutropenia, due to
and cytogenetics may be required to detect an abnormal disorders with neutrophil elastase, typically have a 21-day
malignant clone if malignancy is suspected. periodicity associated with their neutropenia. A list of causes
of acquired leukopenias that affect neutrophils, lymphocytes,
or both is included in Table 2-5. Congenital neutropenias are
Leukopenia
reviewed below within the pediatric section.
Leukopenia is defined as a total leukocyte count that is A careful medication history is important because many
2 standard deviations below the mean. In evaluating a patient drugs, including antibiotics, anti-inflammatory drugs, and
with leukopenia, it is important to check previous complete anticonvulsants can cause leukopenia. Drug-induced leu-
blood counts (CBCs) to establish rate of changes. Some racial kopenia can be dose related, as is the case with phenothi-
groups such as Africans, African Americans, and Yemenite azines, or can be immune-mediated. Certain medications
Jews may have leukocyte counts that normally fall below the are classically associated with agranulocytosis from bone
reference range of many laboratories. Notably, these patients marrow suppression including clozapine, methimazole,
have adequate bone marrow neutrophil reserve and are not trimethoprim-sulfamethoxazole, among others. A wide

Common outpatient hematology consultations | 41
Table 2-5 Causes of acquired leukopenia with flow cytometry may be helpful to identify a malignant
Infection associated clone. A rheumatologic evaluation, including antinuclear
Postinfectious antibody (ANA) and rheumatoid factor may indicate a pre-
Active infection viously undetected collagen vascular disorder or systemic
Sepsis lupus erythematosus. Splenomegaly in this setting may sug-
Viral (HIV, CMV, EBV, hepatitis A, B, C, influenza, parvovirus) gest Felty’s syndrome, characterized by the triad of seroposi-
Bacterial (tuberculosis, tularemia, Brucella, typhoid) tive rheumatoid arthritis, neutropenia, and splenomegaly.
Fungal (histoplasmosis) Treatment of leukopenia depends on the specific etiology.
Rickettsial (Rocky Mountain spotted fever, ehrlichiosis)
Treatment with colony-stimulating factors should not be
Parasitic (malaria, leishmaniasis)
used without a definitive diagnosis requiring such an inter-
Drug-induced
Agranulocytosis
vention or if severe infection occurs in the setting of neutro-
Mild neutropenia penia. Basing initiation of colony-stimulating factors on
Autoimmune absolute neutrophil count (ANC) should be used with con-
Primary autoimmune sideration of the clinical context with treatment generally
Secondary autoimmune (systemic lupus erythematosus, given for patients with ANC <500, especially in the setting of
rheumatoid arthritis) an active infection or fever.
Felty syndrome
Malignancy
Acute leukemia Lymphadenopathy
Myelodysplasia
The peak mass of lymphoid tissue occurs in adolescence. In
Lymphoproliferative disorder
Large granular lymphocyte leukemia
adults, lymph nodes normally are not palpable except for the
Plasma cell dyscrasia inguinal region where small nodes up to 1.5 cm may be felt.
Myelophthisic process Although superficial enlarged nodes can be palpated, deeper
Nutritional nodes require imaging with computed tomography (CT),
Vitamin B12 or folate deficiency positron emission tomography, or magnetic resonance
Copper deficiency imaging (MRI) for detection. Lymph node enlargement can
Alcohol occur in a variety of disorders, including infections, malig-
Acute respiratory distress syndrome nancy, and collagen vascular disorders (Table 2-6).
Increased neutrophil margination (hemodialysis)
In the primary care setting, more than 98% of enlarged
Hypersplenism
Thymoma
lymph nodes are nonmalignant, whereas 50% of patients
Immunodeficiency referred to a specialist for lymphadenopathy are found to
Iatrogenic have malignant disease. A thorough exposure and travel his-
tory can reveal the source of underlying infections (eg, cat
CMV = cytomegalovirus; EBV = Epstein-Barr virus; HIV = human
immunodeficiency virus.
scratch and Bartonella henselae, undercooked meat and
toxoplasmosis, tick bite and Lyme disease, high risk behavior
and HIV). Constitutional symptoms such as fevers, night
variety of infectious disorders can cause leukopenia, including sweats, and weight loss may suggest infection or malignancy.
hepatitis, mononucleosis, HIV, typhoid, and malaria. Cocaine Localizing signs and symptoms of an infection should be
or heroin (contaminated with levamisole) is an increasingly elicited. Review of the medication list may reveal a drug (eg,
recognized cause of acquired leukopenia in young otherwise phenytoin) that is associated with lymphadenopathy. On
healthy individuals. As with cytopenias of red cell or platelet physical examination, large size, hard texture, fixed mobility,
lineage, autoimmune disorders, nutritional deficiencies, and asymmetry, and the lack of pain are features suggestive of
hypersplenism can lead to leukopenia. malignancy. The patient should be evaluated for splenomeg-
Patients with leukopenia may be asymptomatic and may aly as well. Additional laboratory investigations for patients
not require treatment. Patients who are profoundly leu with lymphadenopathy might include ANA, RPR, PPD,
kopenic may complain of fever, mouth sores, or myalgias. monospot, HIV, CBC, and review of the peripheral blood
Evaluation of patients with leukopenia includes a careful smear. Patients with localized lymphadenopathy can be
physical examination, including examination of the mucous observed for a few weeks provided no other concerning fea-
membranes and skin. The peripheral blood film should be tures on history or physical exam exist. Tissue biopsy is
evaluated for the presence of blasts, which would indicate required to determine the precise e tiology of lymphadenop-
acute leukemia, or Pelger-Huët cells, which are seen in athy. If a hematologic malignancy is suspected, an excisional
myelodysplastic syndrome. Evaluation of the bone marrow lymph node biopsy should be performed to preserve the

Table 2-6 Causes of persistent unexplained lymphadenopathy chest), or as multicentric, with generalized lymphadenopathy.
Localized Generalized Unicentric Castleman disease can be classified pathologically
into hyaline vascular variant, plasmacytoid variant, and
Bacterial infection Mononucleosis syndrome
Fungal infection Epstein-Barr virus human herpesvirus 8 (HHV8)-positive Castleman disease.
Tuberculosis Cytomegalovirus HHV8 encodes a viral IL-6 protein and has been implicated
Other mycobacterial infections Primary HIV especially in patients with HIV. Unicentric disease of the hya-
Bartonella henselae (cat scratch Chronic HIV line vascular variant is typically treated with radiation therapy
disease) or local resection. Mixed histology, plasmacytoid variants,
Sarcoidosis Other viral infections and multicentric disease can present with B-symptoms,
Langerhans cell histiocytosis Leptospirosis
organomegaly, and cytopenias. These aggressive subtypes may
Inflammatory pseudotumor Tularemia
Progressive transformation of Miliary tuberculosis
progress to lymphoma and require lymphoma-type treat-
germinal cetners ment. Antiviral agents, such as ganciclovir, have been investi-
Malignancy (eg, NHL, HD, Brucellosis gated in HIV-positive patients with HHV8-positive disease.
CLL, metastatic carcinoma) Lyme disease
Secondary syphilis
Toxoplasmosis Splenomegaly
Histoplasmosis, coccidiomycosis,
The normal adult spleen measures up to 13 cm in largest
crypotcoccus
diameter, weighs approximately 150 g, and is not palpable.
Systemic lupus erythematosus
Rheumatoid arthritis
Splenic enlargement frequently is not appreciated on physi-
Still’s disease cal examination unless the spleen size is increased by 40%.
Rosai-Dorfman disease Spleen size typically is quantified by measuring splenic
Sarcoidosis extension below the costal margin in centimeters. Splenic
Langerhans cell histiocytosis enlargement is best appreciated on physical examination
Phenytoin when there is percussive dullness in Traube’s semilunar tri-
Drug-induced serum sickness
angle bordered by the left sternal border, the costal margin,
Castleman disease
and lower border of the ninth rib. Ultrasonography can
Kikuchi disease
accurately determine the size of the spleen, and CT or MRI
Kawasaki disease
Angioimmunoblastic can be useful in assessing architectural changes due to infarc-
lymphadenopathy tion, infection, infiltration or tumor. Doppler should be
Atypical lymphoproliferative process obtained along with ultrasound to detect any changes in
(eg, Castleman disease) splenic and portal blood flow to account for splenomegaly.
Autoimmune lymphoproliferative Splenomegaly occurs in patients with cirrhosis, heart fail-
syndrome (ALPS) ure, or splenic vein thrombosis when increased portal pres-
Hemophagocytic
sure causes venous engorgement and disruption of the
lymphohistiocytosis
normal splenic architecture. Other causes are splenic infarc-
Malignancy (eg, indolent NHL, HD,
CLL, metastatic carcinoma)
tion, hematologic malignancy such as lymphoma, primary
myelofibrosis, infection, and infiltrative disorders such as
CLL = chronic lymphocytic leukemia; HD = Hodgkin disease; Gaucher disease. Splenomegaly can be seen in conditions of
HIV = human immunodeficiency virus; NHL = non-Hodgkin
ongoing hemolysis such as hereditary spherocytosis or when
lymphoma.
there is extramedullary hematopoiesis as is seen in severe
thalassemia. Solid tumor malignancies rarely metastasize
tissue architecture. Fine-needle aspirations often provide a to the spleen. Normally, about one-third of circulating plate-
sample of tissue that is inadequate for making the diagnosis lets are sequestered in the spleen, where they are in equilib-
of lymphoma. Lymph node biopsy specimens should be sent rium with circulating platelets; thus, splenomegaly can cause
for flow cytometry, cytogenetics, appropriate molecular cytopenias (termed hypersplenism) because of increased
genetic testing, and immunohistochemistry. splenic sequestration. In these instances, the apparent throm-
Castleman disease (angiofollicular lymph node hyperpla- bocytopenia rarely is associated with clinical bleeding or
sia) is a lymphoproliferative disorder characterized by poly- requirements for platelet transfusion since the total platelet
clonal expansion of plasma cells and B and T lymphocytes and mass and overall platelet survival remains relatively normal.
increased interleukin 6 (IL-6) levels leading to localized or However, in chronic liver disease where patients have multi-
systemic lymphadenopathy. The disease is categorized as uni- ple hemostatic issues (eg, true thrombocytopenia from
centric, involving one lymph node region (typically in the decreased thrombopoietin production, decreased production

Hematology consultations in pediatric patients | 43
coagulation factors and fibrinogen, hyperfibrinolysis, bone

marrow suppression from hepatitis C virus, and anatomical Key points (continued)
variceal bleeding) or require medical/surgical procedures, • An excisional lymph node biopsy is better than a fine needle
thrombocytopenia due to hypersplenism may contribute to aspiration for making a tissue diagnosis of lymphoma.
the risk of bleeding. The initial evaluation of a patient with • Patients requiring splenectomy should be vaccinated against
splenomegaly includes a detailed history to determine any of encapsulated bacteria to reduce the risk of overwhelming
the above underlying causes and a thorough physical exam postsplenectomy infection.
focusing on lymph nodes, spleen, and liver. Laboratory tests
may include a CBC, peripheral smear, liver function tests,
Hematology consultations
urinalysis, HIV test, and chest x-ray. Imaging to evaluate
in pediatric patients
malignancy or liver disease should be considered if the above
initial testing is unrevealing. Biopsy of affected tissues (eg,
Pediatric consultation requires evaluation based on knowl-
lymph nodes, liver, or bone marrow) may be pursued if the
edge of developmental hematology and distinct etiologies
cause of splenomegaly is not obvious, as splenic biopsy/aspi-
that are not present in other patient populations. These key
ration is typically not performed.
issues are discussed in this section.
Diagnostic and therapeutic splenectomy may be indicated
for patients with massive splenomegaly causing pain from
infarction or recalcitrant cytopenias. Splenectomy may be Anemia
indicated for patients with hereditary spherocytosis, ITP, or Following is an on overview of anemia in the pediatric popu-
warm antibody–mediated hemolytic anemia. Because of the lation. For additional information on individual conditions,
risk of rapidly progressive septicemia from encapsulated please refer to Chapters 5-7.
organisms, in patients with surgical, functional, or congenital
asplenia, these patients should be vaccinated for Streptococcus
Newborn
pneumoniae, Haemophilus influenzae, and Neisseria menin-
gitidis. Prophylactic antibiotics are recommended for asplenic Figure 2-3 illustrates the diagnostic approach to anemia in
children under the age of five, should be considered for 1 to 2 the newborn. At birth, infants are relatively polycythemic
years postsplenectomy in older children and adults, and and macrocytic, reflecting fetal RBC production in the
should be continued as lifelong prophylaxis for any asplenic hypoxic intrauterine environment. Mean hemoglobin and
individual who has a history of postsplenectomy sepsis. It is hematocrit on day 1 of life for a term newborn are elevated
critically important that despite these prophylactic measures, (Table 2-7) and, therefore, a hematocrit that would be con-
asplenic individuals who develop a fever should be treated sidered normal during childhood represents anemia in the
promptly with therapeutic antibiotics. Splenectomy may be newborn. Shortly after birth, erythropoietin declines, and by
associated with a long-term increased risk of vascular com- day 7, the reticulocyte count is 0.5% resulting in a physio-
plications and pulmonary hypertension, particularly when logic nadir hemoglobin concentration (10.7 ± 0.9 g/dL) at
performed for diseases with increased RBC turnover. In these approximately 7-9 weeks of age. This nadir can occur earlier
cases, aggressive VTE prophylaxis should be administered to and be more pronounced in premature infants.
prevent thromboembolic complications postoperatively. On Careful assessment of the obstetrical and birth history
the peripheral blood film, H owell-Jolly bodies (nuclear rem- along with review of the family history for jaundice, anemia,
nants within RBCs) most often indicate the absence of the splenectomy, or cholecystectomy can assist in identifying the
spleen from splenectomy or splenic hypofunction, as in sickle cause of anemia in the newborn. Physical examination
cell disease. After splenectomy, patients are often noted to should focus on findings such as jaundice, vital signs, and
have chronic leukocytosis and thrombocytosis. possible sources of internal blood loss. A review of the CBC,
red cell indices, reticulocyte count, and peripheral blood film
can narrow the broad differential. Additional laboratory test-
Key points
ing should be guided by the presence or absence of findings.
• Most patients with stable, mild thrombocytopenia (platelets Neonatal anemia can be classified as caused by blood loss,
100 × 109/L-150 × 109/L) do not develop worsening thrombocy increased RBC destruction, or decreased RBC production.
topenia or other autoimmune diseases. Blood loss can result from placenta previa or rupture of an
• Thrombocytopenia caused by medications may be immune abnormal umbilical cord. Acute or chronic fetal–maternal
mediated or dose dependent. hemorrhage and internal hemorrhage in the infant must also
• Hard, fixed, nontender, and enlarged lymph nodes may be
be excluded. Depending on the extent of blood loss, the infant
features suggestive of malignancy.
may have signs and symptoms of circulatory shock. In the

Anemia
(<13.5 g Hb/dL)
Decreased reticulocytes Increased

Congenital hypoplastic anemia reticulocytes
Drug-induced RBC suppression
Bone marrow replacement
DAT positive DAT

RH or ABO blood
group incompatibility DAT
Other blood group negative
incompatibility
RBC and
Low MCV/MCH reticulocyte MCV/MCH
Chronic fetal anemia

Normal or
Fetomaternal hemorrhage
high MCV/MCH
Twin-to-twin transfusion
Thalassemia
RBC and
High MCHC or HDW reticulocyte MCHC
Hereditary spherocytosis Normal or
Hereditary xerocytosis low MCHC
Abnormal Peripheral smear

Hereditary elliptocytosis
Hereditary stomatocytosis
Normal
Pyknocytosis
DIC, microangiopathy
Other hemoglobinopathies Blood loss
Congenital RBC Hemolysis

enzyme defects
Figure 2-3 Diagnostic approach to anemia in the newborn. From
PK Brugnara C, Platt OS. The neonatal erythrocyte and its disorders.
Hexokinase
G6PD In: Nathan DG, Orkin SH, Ginsburg D, Look AT, eds. Nathan and
Other Oski’s Hematology of Infancy and Childhood. 6th ed. Philadelphia,
PA: WB Saunders; 2003:19-55. CMV = cytomegalovirus;
Infection DAT = direct antiglobulin test; DIC = disseminated intravascular
Bacterial coagulation; G6PD = glucose-6-phosphate dehydrogenase;
Viral (CMV, rubella, HSV, Other HDW = hemoglobin distribution width; HSV = herpes simplex
coxsackievirus, adenovirus)
Toxoplasmosis Galactosemia
virus; MCHC = mean corpuscular hemoglobin concentration;
Fungal Hypothyroidism MCV/MCH = mean corpuscular volume/mean corpuscular
hemoglobin; PK = pyruvate kinase.
setting of chronic blood loss, the infant may be compensated defects, and hemoglobinopathies. The infant usually will
but exhibit pallor and in severe cases congestive heart failure. demonstrate a normocytic anemia with an increase in the
Fetal–maternal hemorrhage can be confirmed, and the quan- reticulocyte count. Immune causes of anemia are becoming
tity of blood loss estimated, by the Kleihauer-Betke test on increasingly rare in developed countries given the wides
maternal blood. An uncommon source of blood loss is the pread use of prenatal screening and Rh-immune globulin
twin–twin transfusion syndrome, defined as a 5 g/dL or more administration to Rh-negative women. Immune hemolysis
difference in hemoglobin concentration between twins. Hem- due to ABO incompatibility, currently the most common
orrhage can be acute or chronic, with variable presentations cause of hemolytic disease of the newborn in counties with a
and the potential for polycythemia in the reciprocal twin. high human development index, is most likely in the set-
Hemolytic anemia in the newborn may be classified as ting of an A infant and O mother, given that maternal iso-
either intrinsic or extrinsic. Extrinsic causes include alloim- hemagglutinin titers are usually higher for A than for B
mune-mediated destruction, infection, DIC, and severe aci- and that expression of the A antigen on neonatal RBCs is
dosis. Intrinsic causes include enzyme deficiencies, membrane usually higher than expression of the B antigen. If suspected

Table 2-7 Normal hematologic values for newborns
Red blood cell parameter Term newborn day 1 ± SD*

Hb (g/dL) 19.0 ± 2.2
Hct (%) 61 ± 7.4
MCV (fL) 119 ± 9.4
Reticulocytes (%) 3.2 ± 1.4
Coagulation/inhibitor parameter Healthy term newborn cord blood† Healthy preterm (30–38 weeks) cord blood†
PT (seconds) 16.7 (12-23.5) 22.6 (16-30)
INR 1.7 (0.9-2.7) 3.0 (1.5-5.0)
aPTT (seconds) 44.3 (35-52) 104.8 (76-128)
Fibrinogen (von Clauss; g/L) 1.68 (0.95-2.45) 1.35 (1.25-1.65)
Factor II activity (%) 43.5 (27-64) 27.9 (15-50)
Factor V activity (%) 89.9 (50-140) 48.9 (23-70)
Factor VII activity (%) 52.5 (28-78) 45.9 (31-62)
Factor VIII activity (%) 94.3 (38-150) 50 (27-78)
Factor IX activity (%) 31.8 (15-50) 12.3 (5-24)
Factor X activity (%) 39.6 (21-65) 28 (16-36)
Factor XI activity (%) 37.2 (13-62) 14.8 (6-26)
Factor XII activity (%) 69.8 (25-105) 25.8 (11-50)
Antithrombin III activity (%) 59.4 (42-80) 37.1 (24-55)
Protein C activity (%) 28.2 (14-42) 14.1 (8-18)
Protein C antigen (%) 32.5 (21-47) 15.9 (8-30)
Total protein S (%) 38.5 (22-55) 21.0 (15-30)
Free protein S (%) 49.3 (33-67) 27.1 (18-40)
From Reverdiau-Moalic P, Delahousse B, Body G, et al. Evaluation of blood coagulation activators and inhibitors in the healthy human fetus.
Blood. 1996:88;900-906.
aPTT = activated partial thromboplastin time; Hb =hemoglobin; Hct =hematocrit; INR = international normalized ratio; MCV = mean
corpuscular volume; PT = prothrombin time.
* Adapted from Matoth Y, Zaizov R, Varsano I. Acta Paediatr Scand. 1971;60:317-323.
† Values are means, followed by lower and upper boundaries, including 95% of population.
laboratory testing includes maternal and infant red cell and packed RBCs. Slow transfusions or exchange transfusion
Rh typing along with a direct antiglobulin test (DAT) in the should be considered in infants with severe anemia and car-
infant. The peripheral blood film shows a variable amount of diovascular compromise. Specific thresholds for transfu-
spherocytes depending on the degree of hemolysis. A nega- sions vary among centers and have been studied mostly in
tive DAT does not exclude the diagnosis of incompatibility premature and low-birth-weight infants. Studies comparing
because A antigen density may be too low to cause cross- restrictive (low) versus liberal (high) hemoglobin thresholds
linking in the test. Other than Rh and ABO, anti-Kell anti- showed only minimal differences in frequency of transfu-
bodies may produce severe disease in up to 40%-50% of sions and hemoglobin levels and did not have any impact on
affected fetuses. Common intrinsic red cell etiologies include combined outcomes of mortality or major morbidity. In a
hereditary spherocytosis (HS) and glucose-6-phosphate Cochrane review of transfusion thresholds in children with-
dehydrogenase (G6PD), which will be discussed in the fol- out respiratory support, hematocrits of 30%, 25%, and 23%
lowing paragraphs. were suggested as thresholds at 1, 2, and ≥3 weeks respec-
Impaired RCB production is less common, but it should tively. In cases of significant anemia from blood loss, supple-
be considered in any infant with isolated anemia and inap- mental oral iron should be provided for the first several
propriately low reticulocyte count. Causes include congeni- months of life. Additionally, premature infants will have
tal infections, particularly toxoplasmosis, rubella, CMV, and lower total-body iron stores than normal and should be sup-
herpes simplex (TORCH infections); drug-induced suppres- plemented with oral iron.
sion; and, rarely, Diamond-Blackfan anemia (DBA).
Management requires evaluation of the possible cause,
Children
severity, and hemodynamic status of the infant. Stable infants
with mild anemia may be followed with close observation. Asymptomatic anemia often is discovered incidentally at
Infants with more severe anemia can be managed with approximately 12 to 15 months of age when children

undergo a screening hemoglobin. This isolated value, how- 1- or 2-gene α-thalassemia will not be evident on hemoglo-
ever, does not identify the cause of anemia, and follow-up bin electrophoresis. b-Thalassemia trait or intermedia may
studies, including a CBC and reticulocyte count, are recom- not be detected on NBS; however, b-thalassemia major will
mended. This section provides an overview, and details of have a hemoglobin F only pattern. Later hemoglobin elec-
specific diagnoses are discussed in the specific chapters on trophoresis will reveal increased hemoglobin A2. It is impor-
anemia. Classification of anemia based on red blood cell size tant to make the correct diagnosis so that children with
(mean corpuscular volume [MCV]) and reticulocyte count thalassemia are not inappropriately treated with iron and
provides a practical approach to the child with anemia. genetic counseling can be provided. Another common and
Microcytic anemia most often is due to iron deficiency important and often unrecognized cause of microcytic or
anemia (IDA) or thalassemia. IDA is commonly diagnosed normocytic anemia is anemia of chronic inflammation (dis-
around 1-2 years of age. Maternal iron stores become cussed in Chapter 6).
exhausted after 6 months, and thereafter, the child must take Normocytic anemia poses a greater diagnostic dilemma
in enough dietary iron to maintain hematopoiesis. Although for the consulting physician. Common causes include:
the iron from breast milk is more bioavailable than from (i) early or rapid blood loss, (ii) hemolytic anemia, (iii) ane-
cow’s milk, it is generally inadequate as a sole source of mia of inflammation, and (iv) transient erythroblastopenia
iron beyond 4-6 months of life. In addition, at 1 year of life, of childhood (TEC). Information obtained from the history
children typically switch to iron-poor cow’s milk, have and physical may assist in the diagnosis, including onset of
inadequate intake of iron-containing foods, and develop symptoms, personal or neonatal history of jaundice or blood
gastrointestinal irritation with poor absorption and occult loss, or family history suggestive of hemolytic anemia (jaun-
blood loss secondary to cow milk proteins. A careful diet his- dice, splenectomy, transfusions, and cholecystectomy).
tory usually provides evidence that the child has IDA even Physical examination may reveal splenomegaly and jaundice
without laboratory studies. Older children or children within the setting of hemolytic anemia. Vital signs can provide a
out an obvious dietary explanation should be evaluated for clue to the duration of anemia based on hemodynamic com-
blood loss. Common sites include gastrointestinal, such as pensation. Finally, inclusion of the reticulocyte count will
inflammatory bowel disease or celiac disease, or menstrual help differentiate children with hemolytic anemia and a
loss in girls. Less common are anatomic abnormalities such review of the peripheral blood film often provides the
as a Meckel diverticulum or double uterus, pulmonary hemo diagnosis.
siderosis, or Wegener granulomatosis. Direct and repetitive Extrinsic causes of hemolytic anemia include immune-
questioning and specific testing may be required to elicit mediated destruction, microangiopathic destruction (DIC,
the cause. TTP, and HUS covered in the previous section), and medi-
A full discussion of the laboratory evaluation for IDA can cations. Primary autoimmune hemolytic anemia (AIHA)
be found in Chapter 6; in children, however, additional stud- can be caused by either IgG (warm-reactive) or IgM (cold-
ies are often not necessary if history, CBC, and indices are reactive) antibodies and presents with the acute onset of
highly suggestive. The best confirmatory test for IDA is uncompensated anemia. While management is similar to
response to a therapeutic trial of iron. Within 2 weeks of adults (see Chapter 7), unlike adults, children with AIHA
appropriate iron replacement (4-6 mg/kg/d of elemental have a good prognosis, with approximately 77% having an
iron), reticulocytosis and improvement of hemoglobin acute self-limited condition. Intrinsic causes can be further
should be observed. The most common reasons children fail classified by cause, including: (i) enzyme deficiencies
iron therapy include nonadherence, improper dosing, and a (G6PD), (ii) membrane defects (such as HS), or (iii) hemo-
diagnosis other than IDA. If there is no response to an ade- globinopathies (sickle cell disease). Each of these is reviewed
quate trial of iron and parents report adherence, this treat- in detail in Chapter 7. In all cases, the child usually will dem-
ment should be stopped and alternative causes, including onstrate a normocytic anemia with an increase in the reticu-
blood loss and malabsorption, should be sought. Recent locyte count; however, macrocytosis can occur in the setting
advances in the safety of IV iron makes this an option for of a robust reticulocyte count.
children who require ongoing iron replacement, have poor Special mention should be given to TEC, a normocytic
iron absorption, or do not tolerate oral iron. anemia with reticulocytopenia resulting from brief disrup-
The most common alternative diagnosis is thalassemia, par- tion of normal erythropoiesis in children. Spontaneous
ticularly in children of African American, Mediterranean, or recovery occurs with subsequent brisk recticulocyte response
Asian backgrounds. The gene deletions and corresponding that often mimics acute hemolytic anemia. TEC should be
nomenclature for thalassemia are discussed in Chapter 7. suspected in an otherwise-healthy child with acute onset of
Review of the newborn screening (NBS) is helpful in determin- anemia and no abnormalities on physical examination or
ing alpha-thalassemia; however, after hemoglobin switching, peripheral blood film.

Macrocytosis in childhood should always cause concern, acquired or inherited. Acquired causes include infection,
and a bone marrow evaluation should be undertaken to look drug-induced neutropenia, and autoimmune or chronic
for causes of a marrow failure. In early childhood, the diag- benign neutropenia. Autoimmune neutropenia (AIN) and
nosis of DBA, a congenital pure red cell aplasia, should be chronic benign neutropenia of childhood likely represent a
considered. A quarter of patients with DBA have macrocytic spectrum of disorders caused by immune destruction of
anemia at birth, and 25% of children will have at least one neutrophils. The condition usually presents in children less
congenital anomaly, including head or face, palate, limb, or than 3 years of age and for the most part is not associated
kidney abnormalities. Patients have elevated red cell adeno with serious infections. In the majority of children, antineu-
sine deaminase activity and fetal hemoglobin levels. Bone trophil antibodies can be detected; however, due to the poor
marrow evaluation shows a normocellular bone marrow sensitivity of antibody testing, a negative result does not
with striking paucity of erythroid precursors. Approximately exclude the diagnosis. Yield may be increased by repeating
25% of DBA patients have heterozygous mutations in the the test if clinical suspicion is high. Management is directed
ribosomal protein S19 (RPS19) gene, and mutations in at at treating infections with antibiotics, and G-CSF should be
least five other ribosomal protein genes now have been iden- reserved for severe or recurrent infections associated with a
tified. Treatment modalities include corticosteroids, chronic low absolute neutrophil count. Prognosis is excellent with
transfusions, and bone marrow transplant. spontaneous recovery occurring in almost all patients within
2 years of diagnosis.
A common cause of neutropenia is differences in ethnic
Neutropenia neutrophil norms. Certain ethnic populations, particularly
Newborn African Americans, may have lower normal limits. Usually
these children have mild neutropenia (absolute neutrophils
Neutropenia is relatively common, secondary to the limited
between 1.0 and 1.5 × 109/L), no history of infection or other
neonatal marrow capacity. Therefore, consumption in
concerning features on physical examination, and the value
response to sepsis, respiratory distress, or other acute pro-
will be relatively stable over time. Reassurance is all that is
cesses may exceed production. Neutropenia also may be seen
necessary in this setting. Inherited causes of neutropenia
in association with in utero stress due to pregnancy-induced
represent a rare group of disorders, including severe con-
hypertension (PIH). In both cases, the neutropenia is tran-
genital neutropenia (SCN), Shwachman-Diamond syn-
sient and resolves with resolution of the underlying illness
drome (SDS), and cyclic neutropenia. SCN, an autosomal
or, in the case of PIH, within 3-5 days of delivery.
recessive premalignant condition caused by mutations in the
Neonatal alloimmune neutropenia (NAIN) results from
ELA2 gene, is often diagnosed on the first day of life, and
the transplacental passage of maternal antibodies that react
patients have persistent neutropenia associated with fre-
with paternal antigens on the infant’s neutrophils. The diag-
quent episodes of infections. Bone marrow evaluation shows
nosis of NAIN generally can be made by confirming anti-
myeloid maturation arrest at the myelocyte stage. SDS
genic differences between maternal and paternal neutrophils,
includes neutropenia, pancreatic exocrine insufficiency,
most commonly the NA1 and NA2 alleles, and by demon-
metaphyseal chondrodysplasia, and short stature. Lastly,
strating maternal antibodies that bind to paternal neutro-
cyclic neutropenia is an autosomal dominant condition in
phils. Neutropenia can be profound, with the potential for
which patients experience severe neutropenia and associated
sepsis, oomphalitis, cellulitis, and other serious infections.
infections approximately every 21 days. Bone marrow will
Granulocyte colony-stimulating factor (G-CSF; 5 mg/kg/
look similar to SCN during the nadir, and it may be difficult
dose) is indicated in severe cases. The condition typically
to distinguish from other causes of neutropenia at first.
resolves in weeks to months once maternal antibodies are no
Careful monitoring with frequent blood counts 1-2 times a
longer present.
week for 6-8 weeks can help confirm the diagnosis. In all
NAIN must be differentiated from relatively rare inherited
cases, treatment with G-CSF is standard of care. Less clear is
causes of neutropenia that will be discussed in the following
the role of bone marrow transplant for those conditions that
section. In these conditions, the neutrophil count remains
are considered premalignant.
severely depressed and children are at risk for ongoing
infections.
Thrombocytopenia
Children Newborn
Neutropenia in children is defined as an ANC < 1.5 × 109/L. Thrombocytopenia in a neonate is defined as a platelet count
Outside of the neonatal period, it can be classified as either < 100 × 109/L with severe thrombocytopenia generally being

reserved for infants with a platelet count <50 × 109/L. As with count increments have been documented with this approach.
neutropenia, limited capacity of the neonatal marrow to IVIg (1.0 g/kg/d for 1-3 days depending on response) and
increase platelet production in the face of rapid consumption methylprednisolone also may decrease the rate of platelet
can result in thrombocytopenia in the sick newborn with destruction and can be used as adjunctive therapy. Regard-
estimates of almost 25% of neonates in the neonatal intensive less of treatment, NAIT usually resolves within 2-4 weeks.
care unit experiencing thrombocytopenia. Thrombocytope- Specific testing for NAIT, including platelet antigen typing
nia in the neonate can be classified as early or late. Within the and antibody identification can confirm the diagnosis; how-
first 72 hours, thrombocytopenia is usually the result in ante- ever, treatment should be instituted even if results of testing
natal or perinatal events such as perinatal asphyxia, intrauter- are unavailable. NAIT testing is important because of the
ine growth restriction, maternal hypertension, intrauterine implications for subsequent pregnancies where the risk of
infection, and intrauterine viral infections. It may also result severe thrombocytopenia is higher and can occur as early as
from immune destruction. After 72 hours, thrombocytope- the second trimester. Prenatal management, risk stratifica-
nia is more likely due to postnatal events including necrotiz- tion, and counseling of female family members is recom-
ing enterocolitis and late onset sepsis. mended and should be undertaken in conjunction with a
In an otherwise-well infant, autoimmune thrombocyto- high-risk obstetrician.
penia should be investigated. Knowledge of maternal medi- Outside of NAIT, which carries a high risk for bleeding,
cal history and platelet count is critical as management varies the role of prophylactic platelet transfusions and desired
depending on suspicion for alloimmune versus autoimmune thresholds for transfusion to prevent bleeding remain
thrombocytopenia. Autoimmune thrombocytopenia, either unclear. The majority of studies that have been conducted in
primary or secondary, presents early in infancy due to trans- this area have assessed platelet count as the primary outcome
placental passage of maternal platelet-reactive IgG, which and not bleeding events; therefore making conclusions about
binds to common antigens on the infant’s platelets. The true clinical utility difficult to draw. In one randomized trial,
mother may or may not have thrombocytopenia, as even a there was no reported increased risk for intraventricular or
remote history of resolved ITP in the mother can cause periventricular hemorrhage in neonates with moderate
transfer of antibodies to the infant. The risk of bleeding is thrombocytopenia defined as a platelet count of 50-150 ×
low, and infants often can be managed with observation 109/L. Further studies are needed in this area to determine
alone without need for treatment. If the infant does require best practice.
treatment, then IVIG can be given. Primary, ITP in a child
generally does not occur earlier than 6 months of age.
Children
NAIT should be suspected in an infant born with severe
thrombocytopenia, especially if maternal history is negative Causes of childhood thrombocytopenia generally can be
and platelet count is normal. NAIT results from the transpla- classified as either due to platelet destruction or impaired
cental passage of maternal antibody, which is reactive against platelet production. The most common cause of isolated
paternal-derived antigens expressed on the infant’s platelets. thrombocytopenia is ITP. ITP in children can be either pri-
This condition is analogous to Rh disease, in that the mother mary or secondary, and specific features of ITP in children
lacks the antigen and the infant inherits the antigen from the are outlined here.
father. Unlike Rh disease, however, first pregnancies may be ITP is a diagnosis of exclusion based on findings of iso-
affected by NAIT. The majority of NAIT cases (80%) arise as lated thrombocytopenia in an otherwise healthy child with-
a result of a maternal antibody against HPA-1a. Other anti- out abnormalities on physical examination or laboratory
gens, including HPA-5b and HPA-3b are less common. studies, including detailed evaluation of the peripheral blood
Thrombocytopenia caused by NAIT is associated with a high film. A bone marrow examination is not considered neces-
risk of intracranial hemorrhage (ICH; 10%-20%); therefore, sary for the diagnosis of ITP.
NAIT should be suspected in any healthy infant with severe Treatment of the child with ITP remains controversial.
thrombocytopenia and prompt management should be ini- Recently published guidelines by ASH recommend that chil-
tiated. All infants with NAIT should be investigated for ICH dren with no or mild bleeding do not require treatment
with either ultrasound or CT scan. Treatment is recom- regardless of the platelet count. This was based on evidence
mended for a platelet count <30 × 109/L or <100 × 109/L in that the majority of children will have spontaneous recovery
infants with severe hemorrhage. Optimal treatment includes of their platelet count, treatment is unlikely to alter the
transfusion of HPA-compatible platelets, which can be col- course of the disease, and severe hemorrhage is a rare event
lected and washed from the mother or from an antigen- even in children with severe thrombocytopenia. In addition
negative donor. Random donor platelets should be given if to bleeding symptoms, physicians need to consider quality of
antigen-negative platelets are unavailable since platelet life, access to care, and child behavior when determining

therapy. When drug therapy is indicated, prospective ran- coordination with an immunologist. Several conditions are
domized studies have demonstrated that IVIg and anti-D (in characterized by macrothrombocytopenia: MYH9-related
Rh-positive patients) lead to the most rapid increase in disease (autosomal dominant), Bernard-Soulier syndrome
platelet count. Although anti-D is easier to administer, it has (autosomal recessive), GATA1 mutations (X-linked reces-
been associated with fatal intravascular hemolysis and DIC, sive), and gray platelet syndrome (variable inheritance).
which led to a black box warning by the US Food and Drug Normocytic thrombocytopenia is seen in congenital amega-
Administration. Short courses of corticosteroids are effective karyocytic thrombocytopenia (autosomal recessive), throm-
and much less costly, but they take longer to increase the bocytopenia with absent radii (variable inheritance), and
platelet count. Long courses of corticosteroids are not rec- thrombocytopenia with radioulnar synostosis (autosomal
ommended in children. dominant). Unlike other inherited thrombocytopenias,
In contrast to adult ITP, the majority of children will have infants with thrombocytopenia-absent radius syndrome can
an acute course with 75% of patients achieving a complete demonstrate spontaneous resolution of thrombocytopenia
remission by 6 months from presentation. For patients with during childhood. Although supportive care with platelet
persistent or chronic disease treatment, options included transfusions commonly is used as initial management of
intermittent use of medications, splenectomy, or newer management for patients with congenital thrombocytope-
modalities such as rituximab, high-dose dexamethasone, nia, accurate diagnosis is important, as some conditions are
and thrombopoietin (TPO) receptor agonists. The benefit of associated with transformation to leukemia and may benefit
splenectomy is a high rate of durable remission, which from bone marrow transplant.
occurs in approximately 75% of patients; however, this must
be weighed against the risks associated with surgery, a life-
long risk of sepsis, and newly appreciated possible risk of
Coagulopathy
thrombosis. Rituximab and high-dose dexamethasone have
Newborn
been used in children with chronic ITP to avoid or delay
splenectomy with complete remission rates of approximately Accurate assessment of hemostasis in the newborn requires
20%-30%; however, remission duration is generally shorter knowledge of the normal range for coagulation parameters
than with splenectomy. The new TPO receptor agonists are (Table 2-7). The vitamin K–dependent factors II, VII, IX,
approved for the treatment of ITP in adults, but their role in and X and contact factors are physiologically low in neo-
pediatrics is not yet established. nates, despite the routine administration of vitamin K. Nota-
Additional causes of thrombocytopenia in children due to bly, the normal newborn range for factor IX activity,
destruction include microangiopathic conditions and HIT 15%-50%, occasionally has led to the misdiagnosis of mild
(rare in children), both discussed in the adult section. Auto- hemophilia B. By contrast, several factors are at adult levels
immune lymphoproliferative syndrome (ALPS) results from at birth, including factors VIII, V, and XIII; fibrinogen; and
impaired fas ligand–mediated apoptosis. Patients experience vWF. Because of these physiologic differences, both the
recurrent lymphadenopathy, organomegaly, and immune median and upper limit of the PT (median, 16.7 seconds;
cytopenias. Kasabach-Merritt syndrome is characterized by upper limit, 23.5 seconds) and aPTT (median, 44.3 seconds;
thrombocytopenia and giant hemangiomas during infancy. upper limit, 52 seconds) are higher than ranges established
Patients can develop a severe life-threatening consumptive for adult patients. Coagulation factor production gradually
coagulopathy, and many treatment modalities have been increases over the first few months of life, reaching adult lev-
described, including corticosteroids and vincristine. els by approximately 6 months of age. Therefore, compari-
Causes of decreased platelet production include aplastic son of obtained values to age-appropriate normal values is a
anemia, myelodysplastic syndrome, bone marrow infiltra- critical first step in evaluation of neonate with suspected
tion, and congenital thrombocytopenias. The congenital coagulopathy.
thrombocytopenias represent a diverse group of disorders In sick neonates, coagulation abnormalities can result
(see Chapter 10). In all cases, a detailed review of the family from sepsis, asphyxia, or other triggers of DIC. Unexpected
history, physical examination looking for additional anoma- bleeding in an otherwise-well newborn, such as hemorrhage
lies, and evaluation of platelet and white cell morphology on at circumcision, prolonged oozing from heelstick blood
the peripheral blood film provide important diagnostic draws, or more bleeding or bruising than expected from a
clues. Microthrombocytopenia in males should raise the difficult delivery, should raise the possibility of an inherited
possibility of Wiskott-Aldrich syndrome (WAS) or X-linked bleeding disorder. Screening can be undertaken with a PT
thrombocytopenia (XLT), caused by a mutation in the WAS and aPTT, with specific factor levels based on results.
gene. WAS, unlike XLT, is associated with immune deficiency, The most common inherited causes of an isolated aPTT in
and patients require early identification and management in an otherwise-healthy infant are factor VIII and factor IX

deficiency, with factor XI deficiency being significantly less specific information on individual disorders of coagulation
common. Family history may be suggestive of a bleeding dis- in Chapter 9.
order with X-linked inheritance; however, a negative family If an abnormality is identified, laboratory error or heparin
history does not exclude the diagnoses, as approximately contamination should be considered and eliminated as a
one-third of infants represent spontaneous mutations. possible cause. Otherwise, a common cause of isolated aPTT
Although also associated with an elevated aPTT, vWD rarely elevation in a healthy child with no bleeding history or
results in bleeding in the newborn unless it is severe (type 3). symptoms is a lupus anticoagulant. In this setting, transient
If there is an immediate need for treatment and the specific antibodies are interfering with the phospholipids that are
factor deficiency is unknown, fresh frozen plasma (FFP) will required to perform the aPTT. The condition usually is self-
provide adequate hemostatic coverage; however, it is impor- limited and does not represent a bleeding disorder in the
tant to draw a sample for specific factor testing prior to child. Confirmatory testing for a lupus anticoagulant can be
administration of FFP. undertaken if it is felt necessary or if there is a question of the
Elevation of both the PT and aPTT should prompt inves- diagnosis. Both the International Society of Haemostasis and
tigation global defects in hemostasis such as vitamin K defi- Thrombosis and the British Committee for Standards in
ciency. Although all infants born in the hospital receive Haematology have published guidelines to help guide addi-
supplemental vitamin K, home deliveries and parental desire tional testing methods. Patients with a concerning history
to avoid medical interventions has increased the incidence in should be evaluated for a factor deficiency. The child may
breastfed infants. Vitamin K deficiency may be classified as have a remote history of bleeding, such as hemorrhage with
early (within the first 24 hours of life), classic (between days circumcision, hematomas with immunizations, swelling to
of life 2 through 7), or late (beyond day 8 of life and as late as extremities with mild trauma, or previous bleeding with
6 months). Late deficiency is associated with a higher rate of even minor procedures. Family history may provide infor-
intracranial hemorrhage. Infants often present with diffuse mation to guide testing, with factor VIII and IX deficiency
severe hemorrhage that can be intracranial, gastrointestinal, having an X-linked inheritance. Testing for factor VIII and
umbilical, head or neck, at injection sites, or from circumci- factor IX deficiency as well as vWD should be considered in
sion. Treatment for infants with mild bleeding is administra- children with a prolonged aPTT. Very rarely, factor XI defi-
tion of 1-2 mg of vitamin K given either subcutaneously or as ciency can result in a prolonged aPTT and should be tested if
slow intravenous infusion. Rapid reversal of the coagulopa- no other abnormalities are identified.
thy begins within an hour of administration, but FFP should An isolated prolonged PT represents a deficiency of factor
be given to infants with severe bleeding. Additional defects VII. Inherited factor VII deficiency is a rare autosomal bleed-
that affect global hemostasis include DIC and liver disease. ing disorder with variable presentation and little correlation
Alternative causes of prolongation of both the PT and aPTT between bleeding rates and factor level. Beyond congenital
include rare deficiency in factors of the common pathway factor VII deficiency, consideration should be given to
such as afibrinogenia or dysfibrinogemia, prothrombin defi- acquired causes such as liver disease and vitamin K defi-
ciency, and factor V and factor X deficiency. ciency from malabsorption, cystic fibrosis, or medication
If there is a high suspicion for a coagulopathy, and both use. Given the extremely short half-life of factor VII, the PT
the PT and aPTT are normal, factor XIII deficiency should may prolong before the aPTT.
be considered. This condition is an autosomal recessive dis- Prolongation of both the PT and aPTT is seen in either
order caused by an inability to cross-link fibrin and com- common pathway factor deficiencies or in the setting of
monly presents with umbilical cord bleeding. A clot solubility multiple factor abnormalities. Common pathway factor
test can be performed to screen infants with concerning his- deficiencies are rare and include fibrinogen, prothrombin,
tory and a factor XIII activity used to confirm the diagnosis. factor V, and factor X. More commonly, this scenario is seen
with multiple factor deficiencies in the setting of liver dis-
ease, vitamin K deficiency, and DIC. Testing of factors VIII,
Children
V, and II often can provide information to distinguish these
The diagnostic work-up for a child with a suspected coagu- etiologies if it is not clinically apparent. In DIC, all three will
lopathy begins with a thorough history and screening for PT be decreased; in liver disease, factor VIII will remain normal;
and aPTT. Specific considerations for additional testing and in vitamin K deficiency, only factor II will be decreased.
depend on concerns identified on history and screening lab- In all cases, treatment should be aimed at reducing hem-
oratory examination. Samples should be drawn from a orrhage and at correcting the coagulopathy with manage-
peripheral venipuncture in order to avoid contamination ment of the underlying disease and replacement of deficient
from heparin. Here we will provide an overview to guide the factors. If the precise deficiency is identified, specific factor
initial evaluation based on laboratory findings, with more replacement should be provided; however, if specific factor

is not available, the deficiency is not known, or multiple fac- difficult to achieve a therapeutic aPTT, ATIII levels can be
tors are involved, then FFP can be given. checked and a supplement can be given if levels are low.
Warfarin dosing in infants can be complicated by several fac-
tors, including changing levels of coagulation proteins in the
Thrombosis
first months of life, disparate levels of vitamin K in breast
Newborn
milk and fortified formulas, and lack of a liquid warfarin
Similar to pregnancy, the balance between hemostasis and preparation. For these reasons, LMWH increasingly is pre-
fibrinolysis is shifted toward thrombosis in the newborn, ferred. Newborns have rapid metabolism of LMWH and
with ATIII levels being mildly lower in neonates and the thus higher starting doses are recommended in this popula-
vitamin K–dependent anticoagulants, proteins C and S, are tion, and dose adjustments should be made as needed to
strikingly lower (Table 2-7). Although evidence suggests that maintain anti-Xa activity levels of 0.5-1 U/mL 4 hours after
the fibrinolytic system is activated at birth, plasminogen lev- administration.
els are relatively low, so plasmin generation is somewhat
decreased in response to thrombolytic agents. When added
Children
to the physiologic stresses of labor and delivery, the newborn
period thus represents the greatest risk of thrombosis, espe- Recent evidence suggests that thrombosis in children is
cially in the sick neonate. Neonatal thrombotic complica- becoming a more common event, perhaps because of the
tions include those associated with umbilical venous or increased use of central venous catheters, greater recogni-
arterial catheters, renal vein thrombosis, arterial and venous tion, or improved imaging techniques. For the most part,
stroke, and cerebral venous sinus thrombosis. Clinically, it children with thrombosis have an identifiable secondary
may be difficult to determine whether the thrombotic event cause such as infection or central venous catheter, and spon-
occurred pre- or postnatally. taneous thrombosis are less common. Testing for thrombo-
Screening for inherited thrombophilia in a neonate with a philia in children with thrombosis or family history remains
first thrombotic event is controversial; although some rec- controversial; however, testing is generally recommended
ommend screening all such infants, others conclude that for children with spontaneous thrombosis. There are insuf-
unless it will alter acute management, screening is not cost ficient data to guide recommendations for routine testing in
effective. In addition, in neonates, age-related variation in children with an acquired risk factor such as a central cath-
normal factor levels may complicate interpretation of results. eter. If desired comprehensive testing includes protein C,
Lastly, in some cases, the mother may be screened for protein S, and ATIII levels along with factor V Leiden and
antiphospholipid antibodies, which can cross the placenta. prothrombin G20210A gene mutations. Additionally, one
Special mention should be made of the rare but potentially should consider lupus anticoagulant and antiphospholipid
devastating homozygous deficiencies of protein C and protein antibody testing in a child without other causes for sponta-
S. Infants classically present with purpura fulminans lesions at neous thrombosis. MTHFR and homocysteine testing has
birth without an obvious other cause for DIC. The level of been largely abandoned due to unclear significance. Ratio-
protein C or S in such patients is often undetectable. Genetic nale for testing is based on the notion that identification of
testing can be performed to confirm a congenital cause but thrombophilia may alter duration of anticoagulation ther-
should not delay immediate treatment with FFP along with apy and predict risk for recurrence. Treatment for children
anticoagulation with LMWH or UFH. Anticoagulation can be with thrombosis is similar to adults, and duration is based
transitioned to warfarin once therapeutic levels of LMWH or on the site and cause of thrombosis (see Chapter 8).
UFH are achieved. Protein C concentrates are approved for Given that spontaneous thrombosis is rare in children,
use in patients who have confirmed severe protein C defi- when it does occur, specific consideration should be given to
ciency. Generally, protein C or S replacement should be anatomical causes. May-Thurner syndrome caused by pres-
administered for 6-8 weeks, until all lesions have healed and a sure on the left common iliac vein by an overlying right
therapeutic international normalized ratio has been achieved. common iliac artery should be suspected in cases of left iliac
Beyond protein C and S deficiency, anticoagulation ther- vein thrombosis and should be evaluated with an MRI once
apy in infants with acute thrombosis can include thrombo- acute obstruction has resolved. Paget-Schroetter syndrome
lytic therapy, UFH, warfarin, and LMWH. Thrombolytic results from upper venous obstruction seen with thoracic
therapy can be considered in the newborn when thrombosis outlet syndrome. Patients may report activity that requires
poses risk to life, limb, or organ. Dosing of tPA may be frequent movement that raises the arm above the head lead-
somewhat higher in newborns compared with older patients ing to repeat compression. If present, proper management of
due to lower levels of plasminogen. UFH use may be compli- both conditions involves consultation with a vascular sur-
cated by low levels of ATIII in infants. Therefore, if it is geon or interventional radiology.

Consultation for hematologic complications of

Key points solid organ transplantation
• Attention must be given to age-appropriate normal values when
Smith EP. Hematologic disorders after solid organ transplantation.
performing a pediatric consult.
Hematology Am Soc Hematol Educ Program. 2010;2010:281-286.
• The sick newborn is particularly at risk to develop cytopenias
secondary to poor bone marrow reserve in the setting of stress.
• Ideal prophylactic transfusion thresholds for red cell and platelet Common inpatient consultations
transfusions in neonates and children remain unknown.
• During the newborn period, antigenic differences between the Arnold DM, Lim W. A rational approach to the diagnosis and
mother and the infant can result in alloimmune cytopenias. management of thrombocytopenia in the hospitalized patient.
• The majority of hematologic conditions during childhood Semin Hematol. 2011;48:251-258.
represent benign self-limited conditions and inherited causes are rare. Carson JL, Grossman BJ, Kleinman S, et al., Clinical Transfusion
• ITP in children, unlike in adults, usually is acute, and manage Medicine Committee of the AABB. Red blood cell transfusion:
ment with observation alone is appropriate only for children with a clinical practice guideline from the AABB. Ann Intern Med. July
ITP and cutaneous manifestations. 2012;157:49-58.
Hebert PC, Wells G, Blajchman MA. A multicenter, randomized,
controlled clinical trial of transfusion requirements in critical
care. N Engl J Med 1999;340:409-17.
Bibliography Levi M, Toh CH, Thachil J, Watson HG. Guidelines for the
diagnosis and management of disseminated intravascular
Consultation for surgery and invasive coagulation. British Committee for Standards in Haematology.
procedures Br J Haematol. 2009;145:24-33.
Linkins L. Heparin induced thrombocytopenia. Br J Haematol.
Bahl V, Hu H, Henke PK, et al. A validation study of a
2015; doi:10.1136/bmj.g7566.
retrospective venous thromboembolism risk scoring method.
Ann Surg. 2010;251:344-350.
Baron TH, Kamath PS, McBane RD. Management of
Common outpatient consultations
antithrombotic therapy in patients undergoing invasive
procedures. N Engl J Med 2013;368:2113-2124. Gibson C, Berliner N. How we evaluate and treat neutropenia in
Caprini JA, Arcelus JI, Hasty JH, Tamhane AC, Fabrega F. Clinical adults. Blood 2014;124:1251-1258.
assessment of venous thromboembolic risk in surgical patients. Neunert C, Lim W, Crowther M, et al. The American Society
Semin Thromb Hemost. 1991;17(suppl 3):304-312. of Hematology 2011 evidence-based guideline for immune
Douketis JD, Spyropoulos AC, Kaatz S, et al. Perioperative bridging thrombocytopenia. Blood 2011;17:4190-4207.
anticoagulation in patients with atrial fibrillation. N Engl J Med Rubin LG, Schaffner W. Care of the asplenic patient. N Engl J Med
2015;373:823-833. 2014;371:349-356.
Douketis JD, Spyropoulos AC, Spencer FA, et al., American
College of Chest Physicians. Perioperative management of
antithrombotic therapy: antithrombotic therapy and prevention Consultation in pediatric patients
of thrombosis. 9th ed. American College of Chest Physicians
Bizzarro MJ, Colson E, Ehrenkranz RA. Differential diagnosis and
evidence-based clinical practice guidelines. Chest. 2012;141(2
management of anemia in the newborn. Pediatr Clin North Am.
suppl):e326S-e350S.
2004;4:1087-1107.
Kitchens CS, Kessler CM, Konkle BA. Consultative Hemostasis
Bussel JB. Diagnosis and management of the fetus and neonate
and Thrombosis. 3rd ed. Philadelphia, PA: Elsevier Saunders;
with alloimmune thrombocytopenia. J Thromb Haemost.
2013:3-15.
2009;S1:253-257.
Rogers SO Jr, Kilaru RK, Hosokawa P, Henderson WG, Zinner
Journeycake J, Buchanan G. Coagulation disorders. Pediatr Rev.
MJ, Khuri SF. Multivariable predictors of postoperative venous
2003;24:83-91.
thromboembolic events after general and vascular surgery:
Whyte R, Jefferies A. Red blood cell transfusion in newborn
results from the patient safety in surgery study. J Am Coll Surg.
infants. Paediatr Child Health. 2014;19:213-217.
2007;204:1211-1221.
Schwartz J, Winters JL, Padmanabhan A, et al. Guidelines on the
use of therapeutic apheresis in clinical practice—evidence-
Additional references
based approach from the Writing Committee of the American
Society for Apheresis: The Sixth Special Issue. J Clin Apher. Hicks LK, Bering H, Carson KR, et al. The ASH Choosing Wisely®
2013;28:145-284. campaign: five hematologic tests and treatments to question.
Spyropoulos AC and Douketis JD. How I Treat anticoagulated Blood. 2013;122:3879-3883.
patients undergoing an elective procedure or surgery. Blood Hicks LK, Bering H, Carson KR, et al. Five hematologic tests and
2012;120:2954-2962. treatments to question. Blood. 2014;124:3524-3528.

CHAPTER
3
Consultative hematology II:
women’s health issues
Margaret V. Ragni, Peter A. Kouides, and Sarah H. O’Brien
The role of a multidisciplinary team Thromboembolism and thrombophilia Conclusion, 77
in managing women with blood in pregnancy, 64 Bibliography, 77
disorders, 53 Hematologic health issues in the
Hematologic health issues in premenopausal woman, 71
pregnancy, 54
The role of a multidisciplinary hospital, the role of a hematologist specifically trained in

team in managing women with women’s health issues is essential to ensure optimal out-
blood disorders comes. Furthermore, the plan of care should be formulated
with the multidisciplinary team, when available, utilizing
The diagnosis and management of women’s health issues in evidence-based guidelines from expert panels of the National
hematology require a multidisciplinary approach involving Heart, Lung, Blood Institute (NHLBI); National Hemophilia
some combination of hematologists, internists, family prac- Foundation (NHF); American College of Obstetrics and
tice physicians, obstetrician gynecologists, pediatricians, sur- Gynecology (ACOG); American College of Chest Physicians
geons, anesthesiologists, and other health care providers. (ACCP); and World Health Organization (WHO) and should
Because women and girls with blood disorders may be at be communicated in a timely manner with all consulting care
greater risk for bleeding, thrombosis, and reproductive preg- providers as well as the patient. This chapter summarizes the
nancy complications, their care requires a team of experts most recent evidence and guidelines available to minimize
with the availability of specialized laboratory, pharmacy, and risk in the woman with blood disorders, in particular in
blood bank support (Table 3-1). Whether the patient is an the pregnant woman and in the premenopausal female.
adolescent, pregnant, or a perioperative or critically ill female The hematologist may play a critical role, directly or indi-
patient, or whether the setting is inpatient, outpatient, or rectly, in the care of such patients, in a number of scenarios,
phone consultation with a nearby emergency room or whether serving on hospital committees, working groups,
and formulary committees, or developing clinical practice
guidelines, establishing policies and procedures for transfu-
Conflict-of-interest disclosure: Dr. Ragni: research support from sion services, monitoring quality of care and service effi-
Alnylam, ATHN, Baxalta, Biogen, CSL Behring, Dimension, Genen- ciency, developing practice guidelines, or consulting for the
tech Roche, Novo Nordisk, Novartis, Opko, Pfizer, Shire, SPARK, and federal government or pharmaceutical industry. Although
Vascular Medicine Institute; consultant for Baxter, Bio-Marin, Biogen,
Medscape, Opko, Tacere Therapeutics; member of the National He- these latter roles are not addressed specifically in this chapter,
mophilia Foundation Medical & Scientific Advisory Committee and the data management, organizational, and communication
Foundation for Women and Girls with Blood Disorders Advisory skills required for providing patient care or providing patient
Board. Dr. Kouides: consultant for Baxter, CSL Behring, Grifols; mem-
ber of the National Blood Clot Alliance Medical & Scientific Advisory
consultation are just as critical as those required when work-
Board. Dr. O’Brien: research support from Bayer and the Hemophilia ing on advisory groups. The clinical hematologist also serves
Thrombosis and Hemostasis Research Society of North America patients well when adhering to the principles of effective
(HTRS); serves on an advisory board for GlaxoSmithKline.
communication in work with other physicians and consul-
Off-label drug use: Dr. Ragni: not applicable. Dr. Kouides: low-
molecular-weight heparin, fondaparinux, aspirin, and warfarin in tants, house staff, fellows, students, and the patient and
pregnancy. Dr. O’Brien: not applicable. family. A commitment to effective multidisciplinary team
| 53

54 | Consultative hematology II: women’s health issues
Table 3-1 Principles of management of women’s hematologic health issues
Principle Comment
Women with blood disorders are at increased Establishing diagnosis of a blood disorder as early as possible, preferably preconception,
risk for complications of pregnancy and the is essential in optimal management of women with blood disorders. A prediagnosis
menopausal period, including bleeding and should be based on clinical and laboratory assessment. A personal and family
clotting in the antepartum, peripartum, and history of bleeding or clotting, drug history (including hormonal agents), previous
postpartum periods. pregnancy loss, or pregnancy complications should be determined.
Perform laboratory testing for presence of a For a woman with suspected blood disorder, a bleeding or thrombotic complication of
blood disorder (eg, a hemostatic disorder), as pregnancy, or disorder predating pregnancy, a thorough history, clinical assessment
indicated by personal and family history. should be obtained, and laboratory testing performed as indicated.
Involve a multidisciplinary team in the General management requires a multidisciplinary team, including hematologist,
management of women with blood disorders, internist, obstetrician-gynecologist, family practitioner, pediatrician, anesthesiologist,
and develop and discuss the plan with the and surgeon, and the availability of specialized laboratory, pharmacy, and blood
patient. bank support.
Develop a management plan with the Patient management should be coordinated with the multidisciplinary team. Specific
multidisciplinary team with patient-specific recommendations should be communicated regarding laboratory monitoring,
recommendations. treatment guidelines, including factor, blood products, or antithrombotic agents.
Communicate the management plan for delivery Ensure all management plans are communicated in a timely fashion with all members of
with the multidisciplinary team and discuss the the multidisciplinary team, and also with the patient. Resolve questions and readjust
plan with the patient. to optimize compliance.
Anticipate postoperative, postprocedure therapy, Initiate postprocedure, postdelivery, postoperative planning before the event, and
including duration and where and by whom it include plans for the dose, duration, and location of postprocedure care, involving
will be given. the patient in the decision making.
Offer educational information and guidance. Provide relevant evidence-based literature and guidelines to colleagues and to patients,
as requested.
Provide information to the multidisciplinary team Continue involvement and progress notes as indicated, particularly in vulnerable periods
and patient. such as postpartum. Provide for outpatient monitoring, treatment, and follow-up,
with ongoing communication with the multidisciplinary team and patient.
Adapted from Nichols WL et al. Haemophilia. 2008;14:171-232; MASAC Recommendations regarding girls and women with inherited bleeding
disorders. Medical and Scientific Advisory Committee, National Hemophilia Foundation. 2010;197; and Ragni M. Treatment Strategies-
Hematol. 2012;2:88-92.
collaboration and communication will ensure the highest intrauterine growth retardation and adverse effects on fetal
quality of patient care and optimal patient outcomes. growth when the hemoglobin falls below 8 g/dL.
Iron deficiency anemia

Hematologic health issues
in pregnancy Iron deficiency accounts for 75% of cases of nonphysiologic
anemia in pregnancy, and the incidence of iron deficiency
Anemia in pregnancy
anemia in the United States during the third trimester
During normal pregnancy, the plasma volume expands may exceed 50%. Clinical manifestations of iron deficiency
by 40%-60%, whereas the red blood cell mass expands by include fatigue, tachycardia, dyspepsia, poor exercise toler-
20%-50%. Thus, a physiologic anemia develops, leading to a ance, and suboptimal work performance. In addition, iron
normal hematocrit value of 30%-32%. Hemoglobin levels deficiency is associated with postpartum depression, poor
<10 g/dL suggest the possibility of a pathologic process, such maternal infant behavioral interaction, impaired lactation,
as nutritional deficiency. The prevalence of anemia in preg- low birth weight, premature delivery, intrauterine growth
nancy increases from 8% in the first trimester to 12% in the retardation, and increased fetal and neonatal mortality. The
second trimester and to 34% in the third trimester. The lat- total iron requirement during pregnancy is 1,190 mg, and,
ter is a major indicator of reproductive health. A goal of with a net iron balance during pregnancy of 580 mg, this
Healthy People 2010 was to reduce third trimester anemia equates to a requirement of 2 mg daily. Even with a normal
to 20% or less. At present, there is no definitive evidence diet, this is hard to maintain. Besides poor nutrition, other
whether the hemoglobin threshold for transfusion should factors impairing iron absorption include antacids and
be 7 or 8 g/dL, although some studies indicate greater risk of micronutrient deficiencies, including vitamin A, vitamin C,

Hematologic health issues in pregnancy | 55
zinc, and copper deficiency. In the absence of iron supple- 400 μg daily of folate to prevent neural tube defects. Folate
mentation, hemoglobin falls to 10.5 g/dL at 27-30 weeks of deficiency is most precisely diagnosed by measuring plasma
gestation; with iron supplementation, the nadir is less severe, levels of homocysteine and methylmalonic acid.
11.5 g/dL. By the third trimester, serum ferritin declines, Management: For pregnant women, daily folic acid 400 μg
erythropoietin levels surge, and maternal hepcidin levels are is recommended.
reduced to facilitate iron transfer and use at delivery.
Current recommendations suggest that pregnant patients
Aplastic anemia
receive 15-30 mg daily of supplemental elemental iron,
although studies examining the efficacy of iron supplemen- Aplastic anemia is rare in pregnancy. It may be either asso
tation during pregnancy have not shown a clear benefit to ciated with or precipitated by pregnancy. Some cases may
pregnancy outcomes. Ferrous gluconate is better tolerated either mimic or occur with idiopathic thrombocytopenia
due to fewer gastrointestinal effects than ferrous sulfate. For (ITP). The mechanism of the bone marrow aplasia that occurs
patients who do not tolerate oral iron, parenteral iron may in pregnancy is believed to be through the erythropoietic
be used. Iron sucrose is categorized as pregnancy class B suppressor effects of hormones during pregnancy. Alterna-
(presumed safe based on animal models) and is preferred tively, preexisting aplasia may be uncovered during preg-
over iron dextran or iron (fumoxytol), which are considered nancy. Aplastic anemia may lead to maternal death in up to
pregnancy class C (safety uncertain). Recent studies have 50% of cases, usually caused by hemorrhage or infection, and
demonstrated that for patients who do not respond well to in utero fetal complications may occur in one-third. Unfortu-
parenteral iron, the addition of recombinant erythropoietin nately, stem cell transplantation, which is the major therapy
may add benefit. In a study of 40 patients with gestational for nonpregnant aplastic anemia, is contraindicated in preg-
iron deficiency anemia who had unsatisfactory responses to nancy. Women with preexisting aplastic anemia have a bet-
oral iron, 20 patients were randomized to receive recombi- ter prognosis than those with pregnancy-induced aplastic
nant erythropoietin (rEPO) and parenteral iron sucrose anemia, although the treatment is similar, including tran
(group 1) and 20 were randomized to receive iron sucrose sfusion to maintain a platelet count >20,000/µL, growth
alone (group 2). Patients in group 1 displayed higher reticu- factors (eg, granulocyte colony-stimulating factor [G-CSF]),
locyte counts on day 4, greater increases in hemoglobin from and, in select cases, cyclosporine. Among women who survive
day 11, and a shorter duration of therapy to reach the target pregnancy-associated aplastic anemia, half may experience
hemoglobin of 11 g/dL. No abnormalities in fetal hemoglo- spontaneous remission, and the remainder are managed with
bin levels were observed, consistent with the belief that rEPO antithymocyte globulin, immunosuppression, or stem cell
does not cross the placenta. Thus, although rEPO may func- transplantation.
tion as an adjuvant to iron replacement therapy in pregnant Management: For pregnant women with aplastic anemia,
patients with iron deficiency anemia, it should be reserved transfusions to maintain a hemoglobin 7-8 g/dL, a platelet
for exceptional cases, particularly given the increased risk of count of >10,000/µL, and growth factors (eg, G-CSF), as
thrombosis during pregnancy and the fact that improved needed, are recommended. In pregnancy-induced aplastic
fetal outcomes have not been demonstrated. Alternative anemia, the role of termination or early delivery should be
causes of anemia should be sought in patients refractory to considered in management: case reports indicate improve-
standard iron therapy. Finally, although iron supplementa- ment of aplastic anemia following pregnancy.
tion improves hematologic parameters, it may not improve
neonatal outcomes.
Microangiopathic hemolytic anemias
Management: For pregnant women, daily 15-30 mg ele-
mental iron is recommended. For those not able to tolerate Microangiopathic hemolytic anemias are disorders charac-
oral iron, parenteral iron sucrose is preferred. terized by hemolytic anemia in association with thrombocy-
topenia and multiorgan failure. Hemolysis is caused by
microthrombi in small capillaries and is characterized by
Megaloblastic anemia
schistocytes, elevated lactate dehydrogenase (LDH) and
The majority of macrocytic anemias during pregnancy are indirect bilirubin, and reduced haptoglobin. Although they
due to folate deficiency, whereas vitamin B12 deficiency is rare. represent an uncommon cause of anemia in pregnancy (esti-
Multivitamin and folic acid supplementation reduce placental mates are >0.6%-1% of pregnancies are complicated by
abruption and recurrent pregnancy loss. Folate requirements microangiopathies), they may have devastating consequences
increase from 50 μg daily in the nonpregnant female to at least for both mother and child. These disorders, which include
150 μg daily during pregnancy, and the Centers for Disease thrombotic thrombocytopenic purpura (TTP); hemolytic
Control and Prevention recommend supplementation with uremic syndrome (HUS); preeclampsia; and hemolysis,

elevated liver function tests, low platelets (HELLP), are chal- control group, while another database of 14,000 deliveries
lenging to diagnose, given the wide overlap in clinical pre- found an increased risk of VTE comparable to that of preg-
sentation, and are difficult to treat, given disparate treatments. nant lupus patients.
These are discussed in the section “Thrombocytopenia in Management. The goal during pregnancy is to maintain
pregnancy.” Recommendations are provided for each pre-pregnancy hemoglobin and provide more frequent or
disorder. regular transfusions for increasing pain crises or other
complications, eg, chest syndrome. Alloantibody screen-
ing should be performed in early pregnancy, and in heav-
Sickle cell anemia
ily immunized pregnant women, phenotypically-matched
It is well established that pregnancy in women with sickle products should be given if possible. Pregnant SCD
cell anemia is high-risk related to underlying hemolytic patients with a prior history of VTE warrant antepartum
anemia and multiorgan dysfunction. As oxygen demand and postpartum thrombo-prophylaxis (see section
increases to meet the requirements of the growing fetus and “Thromboembolism and thrombophilia in pregnancy” in
placenta, along with the expanding blood volume, red cell this chapter). For SCD patients without a past history of
requirements increase. Further, pain crises and other com- VTE, while there are no clinical trials, antepartum and
plications may worsen if red cell production cannot keep postpartum LMWH prophylaxis should be considered if
up with oxygen demand. If possible, precipitating factors there are prothrombotic risk factors such as immobiliza-
should be avoided, such dehydration, stress, excessive exertion or obesity.
tion, and a cold environment. The problems of oxygenation
and pathophysiology of sickling may result in both mater-
Thrombocytopenia in pregnancy
nal and fetal morbidity. In addition, preeclampsia, eclamp-
sia, placental abruption, and antepartum bleeding may Thrombocytopenia affects 10% of pregnant women and
complicate pregnancy; and pre-term labor, intrauterine results from several disorders that may or may not be specific
growth restriction, and intrauterine fetal death may com- to pregnancy. Pregnant patients may present with isolated
plicate gestation. Nearly half of the women with sickle cell thrombocytopenia or may develop thrombocytopenia as a
disease (SCD) require an acute transfusion due to severe component of a systemic disorder that may be unique to
anemia or obstetric emergency. While there is no specific pregnancy. A summary of causes of thrombocytopenia in
transfusion trigger in pregnancy, the goal is to maintain pregnancy is presented in Table 3-2.
pre-pregnancy hemoglobin. If pain crises escalate, more
frequent or even regular (eg, monthly) transfusions may be
Gestational thrombocytopenia
required. Optimal management of other complications, eg,
acute chest syndrome, may also require more frequent Isolated thrombocytopenia most commonly results from
transfusion. Maternal mortality risk is up to 10% in women “gestational” or “incidental” thrombocytopenia of preg-
with SCD pulmonary hypertension. The 2014 NIH Guide- nancy. Gestational thrombocytopenia occurs in 5% of all
lines recommend discontinuing hydroxyurea in pregnancy pregnancies, usually during the second or third trimester,
and during breastfeeding, but few human data exist on and, rarely, in the first trimester, in otherwise-healthy preg-
potential harmful reproductive effects of hydroxyurea in nant women. Thrombocytopenia is usually mild and self-
males and females. It is suggested that iron chelation ther- limited, and by definition, the platelet count does not
apy be discontinued pre-conception. Early in pregnancy decrease below 70,000/µL. There is no diagnostic test for
supplemental folic acid at 4 g daily, a higher dose than stan- gestational thrombocytopenia, so it is a diagnosis of exclu-
dardly given in pregnancy should be initiated. Alloantibody sion. This disorder may represent an extreme example of
screening should also be performed early, and, if positive, the typical 10% decrease in platelet count that occurs dur-
phenotypic matching should be considered to avoid delayed ing normal pregnancy. Gestational thrombocytopenia does
hemolytic transfusion reactions or hemolytic disease of the not affect pregnancy outcome and does not result in
newborn. In addition, 10% or more of patients with SCD thrombocytopenia in the offspring of affected women;
will develop a venous thromboembolism (VTE) by adult- thus, no specific treatment is required. It is usually self-
hood, and increased risk is also typical in pregnancy. Con- limited and resolves postpartum, but may recur to the same
tributing risk factors for VTE in SCD may include degree in subsequent pregnancies. As gestational thrombo-
immobilization during hospitalization, vaso-oclusive crisis, cytopenia may not be distinguishable from ITP or more
intravenous access devices, and chronic hemolysis. In one serious disorders in late pregnancy, however, women with
database of 18,000 deliveries, SCD was associated with gestational thrombocytopenia should be monitored
higher rates of cerebral vein thrombosis and DVT than the throughout pregnancy.

Table 3-2 Differential diagnosis of thrombocytopenia in pregnancy

Severity of MAHA Coagulation Liver Renal CNS
Diagnosis thrombocytopenia defect defect Hypertension disease disease disease Time of onset
ITP Mild-severe – – – – – – Common in first
trimester
Gestational Mild – – – – – – Second and
third
trimester
Preeclampsia Mild-moderate Mild Absent-mild Moderate- – Proteinuria Seizures in Late second
severe eclampsia to third
trimester
HELLP Moderate-severe Moderate- May be Absent-severe Moderate- Absent- Absent- Late second
severe present severe moderate moderate to third
(mild) trimester
HUS Moderate-severe Moderate- Absent Absent-mild Absent Moderate- Absent-mild Postpartum
severe severe
TTP Severe Moderate- Absent Absent-severe Absent Absent- Absent-severe Second to third
severe moderate trimester
AFLP Mild-moderate Mild Severe Absent-severe Severe Absent-mild Absent-severe Third trimester
DIC Moderate-severe – May be Absent Absent Absent Absent Third trimester
severe
PNH Moderate-severe – Absent Absent Absent Absent Absent All trimesters
Proteinuria measured as total protein excreted in 24 hours is rated as follows: normal is <150 mg/24 h; mild is 150-300 mg/24 h; moderate is
300 mg to 1 g/24 h; severe is ≥1 g/24 h.
Thrombocytopenia is mild when platelets are >50,000/µL, moderate when >20,000/µL, and severe when <10,000/µL.
AFLP = acute fatty liver of pregnancy; CNS = central nervous system; HELLP = hemolysis, elevated liver function tests, low platelets; HUS =
hemolytic uremic syndrome; ITP = idiopathic thrombocytopenic purpura; MAHA = microangiopathic hemolytic anemia; TTP = thrombotic
thrombocytopenic purpura; DIC = disseminated intravascular coagulation.
Management: No treatment is recommended, as the disor- intravenous immunoglobulin (IVIg) are the first-line treat-
der generally resolves postpartum. ments for maternal ITP. Prednisone is given initially at
low dose, 10-20 mg/day, with adjustment to the minimum
dose providing a hemostatically effective platelet count.
Immune thrombocytopenic purpura
Although short-term, low-dose prednisone is considered
ITP affects approximately 1 in 10,000 pregnancies. In con- effective and safe in the mother, it may exacerbate hyperten-
trast to gestational thrombocytopenia, ITP is usually detected sion, hyperglycemia, osteoporosis, weight gain, and psycho-
in the first trimester. The diagnosis is a clinical one, as anti- sis; and, in the fetus, may increase cleft palate. Intravenous
body testing lacks specificity. A prior history of thrombocy- anti-D has been used successfully to treat ITP in Rh(D)-pos-
topenia or autoimmune disease preceding pregnancy or itive women, although only a few patients have been reported,
during previous pregnancies is useful in making a diagnosis and thus, the safety of this agent cannot be considered estab-
of ITP. Patients with ITP generally present with more severe lished. Similarly, there is little experience with the use of
thrombocytopenia than those with gestational thrombocyto- rituximab in pregnant individuals, although B-cell lympho-
penia, but the two disorders may be indistinguishable when cytopenia has been reported in the offspring of individuals
ITP is mild. Indications for treatment of ITP in pregnancy in treated with this agent, which is considered pregnancy class
the first two trimesters include: (i) when the patient is symp- C. The thrombopoietic agents romiplostim and eltrombopag
tomatic, (ii) when platelets fall <20,000-30,000/µL, or (iii) also are considered pregnancy class C, and a registry has been
to increase platelet count to a level considered safe for proce- developed for patients taking these agents who become preg-
dures. Although the lowest platelet count safe for spinal or nant. The use of cytotoxic therapy is associated with terato-
epidural anesthesia is controversial, most obstetric anesthe- genicity in many cases, although azathioprine commonly has
tists recommend a platelet count of 75,000/µL, and most been used in pregnancy with apparent safety.
hematologists recommend for at least 50,000/µL for cesar- Management: For pregnant women, IVIg is recommended
ean delivery. Therapy of ITP in pregnancy is similar to that for severe ITP. In those with severe ITP refractory to steroids
in patients who are not pregnant. Corticosteroids and and IVIg, splenectomy should be considered, optimally in

the second trimester, when the risk of inducing premature dysfunction occur in preeclampsia has led to studies of anti-
labor is minimized and the gravid uterus does not yet obscure platelet agents, primarily low-dose aspirin, in women with
the surgical field. preeclampsia. In a recent Cochrane review of 43 randomized
Up to 10% of the offspring of patients with ITP also will trials including over 32,000 patients, antiplatelet agents sig-
be thrombocytopenic, and 5% will have platelet counts nificantly reduced preeclampsia in both women at low and
<20,000/µL. There are no maternal laboratory studies that high risk for preeclampsia, if started before 20 weeks gesta-
reliably predict whether an infant of a mother with ITP will tion. Although the use of antithrombotic therapy, primarily
be born thrombocytopenic; perhaps the best indicator is a low–molecular weight heparin (LMWH), has been suggested
prior history of thrombocytopenia at delivery in a sibling. by some for management of high-risk preeclampsia (those
Moreover, no maternal interventions have been shown con- with past preeclampsia, a body mass index [BMI] >35 kg/m2,
vincingly to increase the fetal platelet count. The delivery of preexisting diabetes, twin pregnancy, family history of pre-
the offspring of mothers with ITP by cesarean delivery has eclampsia, chronic hypertension, renal disease, autoimmune
not been shown to reduce the risk of fetal intracranial hem- disease, or an underlying angiotensin-converting enzyme
orrhage, a rare complication affecting <1% of these infants at insertion or deletion polymorphism), its use remains equiv-
delivery; however, some continue to advocate this approach, ocal. In an observational study of women at high risk for pre-
particularly when a sibling previously has been found to be eclampsia, the addition of LMWH to aspirin resulted in no
severely thrombocytopenic at delivery. These considerations greater risk reduction than with aspirin alone, whereas a
and the appreciation that the risk of bleeding with fetal plate- more recent study of high-risk women with thrombophilia,
let count determination by percutaneous umbilical cord found dalteparin resulted in lower risk for severe preeclamp-
blood sampling (PUBS) is greater than that of fetal intracra- sia, low birth rate, or placental abruption than no dalteparin.
nial hemorrhage during vaginal delivery explains the aban- Caution should be taken in interpreting these results, as it is
donment of PUBS in recent years. not known whether findings in women with angiotensin-
Management: All offspring of patients with ITP should be converting enzyme (ACE) mutations apply to all women
monitored closely for the development of ITP within the with preeclampsia. Finally, disseminated intravascular coag-
first 4-7 days after delivery, and all thrombocytopenic neo- ulation (DIC) also may accompany severe preeclampsia and
nates should undergo cranial ultrasound. For severely may be initiated by such processes as retained fetal products,
affected offspring, IVIg is recommended. placental abruption, or amniotic fluid embolism. In these
settings, DIC can be severe, abrupt, and fatal if not managed
appropriately. Recent studies have indicated that LMWH
Preeclampsia and eclampsia
may reduce the rate of recurrent hypertensive disorders of
Thrombocytopenia also may occur in patients with pre- pregnancy before 34 weeks gestation.
eclampsia. Preeclampsia affects 6% of all first pregnancies Management: For pregnant women considered to be at
and usually develops in the third trimester; its diagnostic fea- risk for preeclampsia, and for those with a previous history
tures include hypertension and proteinuria (>300 mg/24 h). of preeclampsia, low-dose aspirin (but not thromboprophy-
Preeclampsia occurs most commonly in primagravidas <20 laxis) starting in the second trimester.
and >30 years of age. Eclampsia, defined by the presence of
grand mal seizures accompanying preeclampsia, complicates
Hemolysis, elevated liver function, low platelets
<1% of preeclamptic pregnancies. Some studies suggest that
preeclampsia may be more common in patients with throm- The HELLP (hemolysis, elevated liver function tests, low
bophilia, particularly those with antiphospholipid antibody platelets) syndrome affects 0.10%-0.89% of all live births
syndrome (APLAs). Up to 50% of patients with preeclamp- and is associated with a maternal mortality of 0%-4%.
sia develop thrombocytopenia, the severity of which gener- HELLP and preeclampsia share many clinical features,
ally is related to that of the underlying disease. The although HELLP occurs in a slightly older population (mean
pathogenesis of thrombocytopenia in preeclampsia is not age 25 years). It occurs predominantly in the third trimester,
well understood, but it has been hypothesized that a hypoxic between 28 and 36 weeks of gestation, and, in some cases,
placenta releases antiangiogenic factors, including soluble may occur postpartum, with up to 30% presenting within 48
Flt-1 and soluble endoglin, which impair capillary angiogen- hours of delivery, and even as late as 1 week postpartum.
esis, leading to endothelial dysfunction; and the clinical fea- Generalized edema precedes the syndrome in more than
tures of preeclampsia may evolve in response to endothelial 50% of cases. Approximately 70%-80% of patients with
dysfunction. The levels of sFlt1 and soluble endoglin in preg- HELLP coexist with preeclampsia, which by definition has
nant women are predictive of the severity of preeclampsia. hypertension and proteinuria. The major diagnostic criteria
The observation that endothelial dysfunction and platelet for HELLP include microangiopathic hemolytic anemia,

levels of serum aspartate aminotransferase exceeding 70 U/L, may develop in either the second or third trimesters. TTP
and thrombocytopenia, with a platelet count <100,000/µL. presenting during weeks 20-29 gestation is associated
Microangiopathic hemolytic anemia is accompanied by with severe fetal IUGR, and better outcomes have been
schistocytes on the peripheral blood film and an elevated observed when TTP presents <20 weeks or >30 weeks.
LDH; some experts suggest that a minimal LDH of 600 U/dL TTP is caused by severe deficiency of ADAMTS13, the von
is required for diagnosis. In some cases, HELLP may be dif- Willebrand factor (vWF)–cleaving protease, which may be
ficult to distinguish from TTP-HUS. Because many patients congenital (Upshaw-Schulman syndrome) or acquired
with HELLP may present with isolated right upper-quadrant (autoimmune). The hallmark of TTP is microthrombi in
and epigastric pain in the absence of hypertension and pro- small vasculature, which arise as a direct result of accu-
teinuria, patients may be misdiagnosed as having primary mulation of large superadhesive vWF multimers. Micro-
gastrointestinal disease and referred for surgical consider- thrombi lead to thrombocytopenia, microangiopathic
ation. HELLP is associated with significant maternal hemolytic anemia, and neurologic, renal, and central ner-
and fetal morbidity and mortality; therefore, prompt diag- vous system (hypothalamic) end-organ damage. Although
nosis and treatment are essential. In general, if there is congenital TTP accounts for only 5% of adult TTP, it
maternal hemodynamic instability or coagulation profile accounts for 24% of obstetric TTP. The manifestations of
abnormalities or the fetus is at least 32-34 weeks of gestation TTP in pregnant and non-pregnant women are similar,
at the time of presentation, prompt delivery is undertaken and pregnant patients respond equally well to plasma
(Table 3-3). If cesarean delivery is required, red cells, plate- exchange (Table 3-3), as do infants of affected mothers. A
lets, fresh frozen plasma (FFP), or cryoprecipitate (for hypo- major dilemma in the management of TTP is the difficulty
fibrinogenemia) may be necessary during and after delivery. in diagnosis, as overlap with other pregnancy-specific
Although coagulation and platelet abnormalities resolve disorders, such as HELLP, may delay diagnosis and lead to
usually within 48 hours after delivery, thrombocytopenia increased morbidity and mortality. Plasmapheresis is
may continue or become progressive, and thus careful post- the preferred therapy with a response rate reaching 75%.
partum monitoring is essential. If persistent, severe postpar- Corticosteroids have been used, but their effectiveness is
tum, HELLP may require steroids and plasmapheresis. The not established. Antiplatelet agents do not appear to be
offspring of patients with both preeclampsia and HELLP also helpful. TTP relapse and fetal loss in subsequent pregnan-
may become thrombocytopenic, although the thrombocyto- cies is common.
penia is usually mild. Therapy for HELLP and preeclampsia Management: For obstetric TTP, delivery of the fetus and
is directed toward stabilization of the mother, followed by supportive management of the mother is recommended,
expeditious delivery, after which these disorders usually and may include plasma exchange. The effectiveness of cor-
remit within 3-4 days in the majority of patients. HELLP, in ticosteroids is not established. ADAMTS-13 should be mon-
particular, occasionally may worsen or even develop post- itored during pregnancies in patients with non-pregnancy
partum. Prenatal or postnatal corticosteroids have been TTP, and when activity falls to <10%, plasma exchange
suggested in several small, randomized studies to hasten should be initiated.
resolution of the biochemical abnormalities and thrombocy-
topenia associated with HELLP, although these studies have
Hemolytic uremic syndrome
not been powered sufficiently to demonstrate an effect on
maternal or fetal mortality. One should consider the use of The incidence of HUS also is increased in association with
such adjunctive therapies if thrombocytopenia continues to pregnancy. Although some cases of HUS develop near term,
worsen or there is continuing clinical deterioration 5-7 days the majority of cases develop 3-4 weeks postpartum, and
after delivery. their clinical features most closely resemble atypical HUS,
Management: For HELLP, expeditious delivery of the fetus with renal failure as the predominant manifestation. The
and supportive care of the mother is recommended, includ- prognosis of postpartum HUS is poor, with persistent renal
ing transfusion support through delivery, and corticoste- failure in >25% of affected individuals. Although responses
roids and plasma exchange if platelet or coagulation to plasma exchange have been reported, the overall response
abnormalities persist postpartum. rate to this intervention is low; nevertheless, a trial of plasma
exchange is indicated, particularly given the difficulty in dis-
tinguishing TTP and HUS and the potential role of deficien-
Thrombotic thrombocytopenic purpura
cies of complement regulatory proteins in this syndrome
The incidence of TTP is increased during pregnancy. It is (Table 3-3). Treatment is similar to that for obstetric TTP.
estimated that 10%-30% of all adult TTP is obstetric, and There is no consensus on the risk of developing recurrent
7% of all adult TTP has its onset during pregnancy. TTP TTP or HUS in subsequent pregnancies: observational

Table 3-3 Guidelines for management of microangiopathic hemolytic anemias in pregnancy

Scenario Comments
Preeclampsia, eclampsia
For pregnant women considered at risk for preeclampsia, and those For women with a history of pregnancy complications, screening for
with a previous history of preeclampsia, low-dose aspirin is inherited thrombophilia is not recommended.
recommended throughout pregnancy, starting with the second
trimester.
Hemolysis, elevated liver function, low platelets (HELLP)
For women with HELLP, delivery of the fetus and supportive care of
the mother, which may include plasma exchange, is recommended.
Thrombotic thrombocytopenic purpura (TTP)
For women with TTP, delivery of the fetus and supportive care of the The effectiveness of steroids is not established.
mother, which may include plasma exchange, is recommended.
Hemolytic uremic syndrome (HUS)
For women with HUS, delivery of the fetus and supportive care of the Observational studies suggest eculizumab may reduce thrombosis and
mother, which may include plasma exchange, is recommended. fetal loss in women with HUS, but it is listed as category C and not
recommended in pregnancy but could be used postpartum. The
effectiveness of steroids is not established.
Acute fatty liver of pregnancy (AFLP)
For women with AFLP, delivery of the fetus and supportive Coagulation support is recommended for liver dysfunction and
management of the mother is recommended. disseminated intravascular coagulation if present.
Disseminated intravascular coagulation (DIC)
For women with DIC, delivery of the fetus and supportive Coagulation support especially if there is bleeding is recommended,
management of the mother is recommended including platelets, FFP, and cryoprecipitate.
Paroxysmal nocturnal hemoglobinuria (PNH)
For women with PNH and past thrombosis or high risk of Observational studies suggest eculizumab may reduce thrombosis and
thrombosis, antepartum and postpartum anticoagulation is fetal loss in women with PNH, but it is listed as category C and is
recommended (see guidelines, next section). not recommended in pregnancy but could be used postpartum.
Adapted from Bates SM et al. Chest. 2012;141(2 suppl):e691S–e736S; Woudstra DM, et al. J Thromb Haemost. 2012;10:64-72. Corticosteroids
for HELLP (hemolysis, elevated liver enzymes, low platelets) syndrome in pregnancy. Cochrane Database Syst Rev. 2010;9:CD008148; George
JN. Blood. 2010;116:4090-4099; Sanchez-Corral P, Melgosa M. Br J Hematol. 2010;150:529-542; Fesenmeir MF et al. Am J Obstet Gynecol.
2005;192:1416-1419. Vekemans MC et al. Blood Coagul Fibrinolysis. 2015;26:464-466.
studies suggest the risk may be 10-20%, but a recent registry Acquired antithrombin deficiency may also occur and, in
report did not confirm this. rare cases, could lead to thrombosis. Some cases of AFLP and
Management: For obstetric HUS, delivery of the fetus and possibly HELLP may result from fetal mitochondrial fatty
supportive management of the mother is recommended, acid oxidation disorders, most commonly a d eficiency of
which may include plasma exchange. The effectiveness of long-chain 3-hydroxyacyl-coenzyme A dehydrogenase.
steroids is not established. Eculizumab may also be consid- Management: For AFLP, delivery of the fetus and supportive
ered in the postpartum period for atypical HUS. management of the mother, and coagulation support for liver
dysfunction or DIC, if present, is recommended. If deficiency
of anti-thrombin III occurs, ATIII concentrate may be given.
Acute fatty liver of pregnancy
Acute fatty liver of pregnancy (AFLP) usually occurs in the

Disseminated intravascular coagulation
third trimester and affects primarily primiparas, and
although twins are a risk factor, only 1% of cases occur in Disseminated intravascular coagulation (DIC) may occur in
twins. Symptoms include nausea, vomiting, right upper- the third trimester secondary to amniotic fluid embolism,
quadrant pain, anorexia, jaundice, and cholestatic liver dys- retained dead fetus, or abruptio placenta. It may also compli-
function. Hypoglycemia is present in >50% of cases. cate severe preeclampsia/HELLP. DIC is a consumptive coag-
Thrombocytopenia is usually mild, but maternal bleeding is ulopathy characterized by activation of coagulation caused
common due to the accompanying coagulopathy resulting by entrance of thromboplastic or procoagulant substances
from diminished hepatic synthesis of coagulation proteins. into the circulatory system, eg, amniotic fluid. Typically,

there is consumption of coagulation factors in both intrinsic improved the treatment of PNH. Despite case series suggest-
and extrinsic pathways, prolonging both the PT and APTT, ing eculizumab is safe in pregnancy and prevents PNH-
and consumption of platelets resulting in thrombocytope- induced pregnancy complications, the package insert lists
nia, and presence of breakdown products of fibrin, including eculizumab as category C, with use in pregnancy not estab-
fibrin split products (FSP) and D-dimer. In general, manage- lished, eculizumab could be considered for use postpartum.
ment of DIC is supportive, with platelets, FFP, and cryopre- Management: For pregnant women with PNH, postpar-
cipitate should severe bleeding occur. tum prophylactic- or intermediate-dose LMWH is recom-
Management: For DIC, delivery of the fetus and support- mended, and, in those at high risk for thrombosis (eg, PNH
ive management of the mother is recommended, with plate- clone >50%, prior VTE, or recurrent fetal loss), antepartum
lets, FFP, and cryoprecipitate may be given to replace platelets prophylactic- or intermediate-dose LMWH is recom-
and factors consumed. Frequent monitoring of fibrinogen mended. Use of eculizumab during pregnancy is not estab-
and early replacement of fibrinogen with cryoprecipitate or lished and is classified category C, but eculizumab could be
fibrinogen concentrates are recommended. considered in the postpartum period (Table 3-6).
Paroxysmal nocturnal hemoglobinuria Bleeding disorders in pregnancy

Paroxysmal nocturnal hemoglobinuria (PNH) is a stem cell Postpartum hemorrhage (PPH) is a major cause of morbid-
disorder usually diagnosed in the early 30s. Thus, although ity and mortality in childbirth. Women with an underlying
rare, PNH affects females in their childbearing years. PNH is bleeding disorder are at greater risk for PPH: while several
characterized by hypoplastic anemia, bone marrow failure, single-center studies have reported PPH in up to a third
and hemolysis due to increased susceptibility of red cells to (most with von Willebrand disease), population-based stud-
complement-mediated lysis. The defect is a mutated phospha- ies indicate lower rates of PPH, about 1.5-fold greater than
tidyl inisitol gene (PIG-A), the anchor of the complement women without a bleeding disorder. Bleeding disorders and
regulatory CD55 and CD59 proteins, to the red cell mem- their management in pregnancy, including preferred agents,
brane. This defect results in loss of regulation of the terminal target levels, and dosing, is found in Table 3-4. PPH typically
complex C5b-9, leading to red cell lysis. In the case of the red occurs due to a failure of the uterus to contract after delivery.
cell, the absence of two GPI-linked complement regulatory Primary PPH is defined as an estimated blood loss of >500 mL
genes, CD55 and CD59, increases the sensitivity of red cells to at the time of vaginal delivery, or >1,000 mL at the time of a
activated complement and complement- mediated lysis. In cesarean delivery, and affects 4%-6% of all pregnancies. Sec-
addition to hemolysis, PNH is characterized by arterial and ondary PPH is excessive vaginal bleeding occurring between
venous thrombosis, which may occur due to depletion of 24 hours and 6 weeks after childbirth. The most common
nitric oxide which binds circulating free hemoglobin and may causes of PPH in the general obstetric population are uterine
occur at visceral sites, including the inferior vena cava (Budd atony, retained placenta, and genital tract trauma. Women
Chiari), splenic, hepatic, and cerebral veins. Thrombotic risk with inherited bleeding disorders have these same risk fac-
correlates with expression of GPI-linked proteins on the sur- tors, as well as the additional risk factor of their coagulation
face of granulocytes, with the greatest risk associated with a defect. In the general population, most PPHs are primary,
PNH clone >50%. When PNH occurs in pregnancy, up to 40% with <1% occurring more than 24 hours after delivery. In
end prematurely and only 30% deliver vaginally. Hemolysis women with bleeding disorders, delayed (or secondary) PPH
leads to smooth muscle dystonia, vasculopathy, and endothe- is much more common and has been reported in 20%-25%
lial dysfunction, increasing the risk for premature labor and of women with von Willebrand disease (vWD), 2%-11% of
fetal loss. There is an 8% maternal mortality rate in women hemophilia carriers, and 24% of women with factor XI defi-
with PNH, primarily related to postpartum thrombosis, and a ciency. Risk factors for uterine atony include prolonged
4% fetal mortality rate. It may be difficult to distinguish induced or augmented labor and expectant rather than
thrombotic complications of PNH from thrombotic compli- active management of the third stage of labor. Therefore, in
cations of pregnancy. Because of the high risk of VTE in preg- women with inherited bleeding disorders, these factors
nant women with PNH, at an incidence of 10%, should be minimized to reduce the risk of PPH. Active man-
antithrombotic therapy is recommended postpartum for all agement of the third stage of labor may include the adminis-
pregnant patients, and antepartum prophylaxis is indicated tration of prophylactic uterotonics to increase muscle
for patients with thrombosis-large PNH clones (>50%), prior contractility and controlled traction of the umbilical cord
history of VTE, or recurrent prior late fetal loss. In the during the delivery of the placenta. Hemostatic management
past decade, eculizumab, a monoclonal antibody that targets also may reduce the risk of PPH. Factor levels should be
the terminal component of complement, C5, has greatly assessed in the third trimester of pregnancy, and

Table 3-4 Management of bleeding disorders in pregnancy
Bleeding disorder Factor at delivery Dosing at delivery Target at delivery

von Willebrand disease
Type 1 vWD vWF concentrate 60-80 IU/kg, then 40-60 IU/kg q 8-12 h, then daily 3-5 days vWF:RCo >50 IU/dL
Type 2 vW D (2A, 2B, 2M, 2N) vWF concentrate 60-80 IU/kg, then 40-60 IU/kg q 8-12 h, then daily 3-5 days vWF:RCo >50 IU/dL
Type 3 vW D vWF concentrate 60-80 IU/kg, then 40-60 IU/kg q 8-12 h, then daily 3-5 days vWF:RCo >50 IU/dL
Hemophilia carrier
Hemophilia A carrier FVIII concentrate 25-50 IU/kg IV push, for up to 3-4 days FVIII:C >50 IU/dL
Hemophilia B carrier FIX concentrate 75-100 IU/kg IV, for up to 3-4 days FIX:C >50 IU/dL
Rare bleeding disorder
FI (fibrinogen) deficiency Cryo, FI concentrate 50-100 mg/kg twice weekly, then 3 day postpartum FI >1.0 g/L
FII deficiency FFP, FEIBA FFP 15-20 mL/kg or FEIBA 20-30 IU/kg FII >25 IU/dL
FV deficiency FEIBA FFP 20 mL/kg; FEIBA 75-100 IU/kg FV >15 IU/dL
FVII deficiency FVIIa concentrate FVIIa15-20 µg/kg every 3 hours, then taper FVII >40-60 IU/dL
FX deficiency FFP, FEIBA FFP 20 mL/kg; FEIBA 20-30 IU/kg every 48 h FX >10-20 IU/dL
FXI deficiency FFP FFP 15-20 mL/kg; and/or antifibrinolytics FXI >10-20 IU/dL
FXII deficiency NA (no bleeding) NA NA
FXIII deficiency Cryo, FXIII concentrate 10-40 IU/kg, then every 2-3 weeks FXIII >20 IU/dL
Adapted from Nichols WL et al. Haemophilia. 2008;14:171-232; and Kadir R, Chi C, Bolton-Maggs P. Haemophilia. 2009;15:990-1005. It should
be recognized that these represent expert opinion recommendations.
vWD = von Willebrand disease; vWF = von Willebrand factor; vWF:RCo = ristocetin cofactor activity; FVIII:C = factor VIII activity; FIX:C =
factor IX activity; cryo = cryoprecipitate.
prophylactic factor replacement given at delivery to those affected, only 0.1% are symptomatic, but many are unaware
with subtherapeutic levels (Table 3-3). Finally, care must be of their diagnosis. Clinically, the disease is characterized by
taken to minimize genital and perineal trauma to reduce the mucosal bleeding, including menorrhagia, bruising, epi-
risks of both PPH and perineal hematomas. Perineal (or vul- staxis, and postoperative bleeding. There are several variants.
var) hematomas, a rare complication of vaginal birth, occur Type 1 vWD is a partial, quantitative deficiency of vWF, and
with some frequency in women with bleeding disorders and accounts for 70-75% of all vWD cases. Type 2 vWD, account-
contribute to the increased incidence of PPH. In one patient ing for 35% of the disease, consists of four subtypes: type 2A
series, the prevalence of perineal hematoma was much higher is caused by a qualitative defect in vWF in which high–
in women with inherited bleeding disorders (1%-6%) as molecular weight vWF (HMW multimers) are reduced,
compared with a reported 0.2% in the general population. resulting in a more severe phenotype; type 2B is character-
Even after discharge from the hospital, women with inher- ized by a gain-of-function mutation resulting in increased
ited bleeding disorders require close follow-up during the affinity and binding of vWF to platelet GP1b, resulting in
postpartum period. In the general obstetric population, the thrombocytopenia and spontaneous platelet aggregation;
median duration of bleeding after delivery is 21-27 days. type 2M is characterized by decreased affinity of vWF for its
A recent case-control study revealed that women with inher- platelet receptor Glycoprotein 1b (GPIb); and type 2N is
ited bleeding disorders have significantly longer postpartum characterized by a loss-of-function mutation in which the
bleeding when compared with controls, even when they vWF binding site for factor VIII (FVIII) is mutated, resulting
receive appropriate hemostatic treatment. In vWD, the in greatly reduced factor VIII, which may be confused for
pregnancy-induced increase in coagulation factor levels
mild hemophilia A. Type 3 vWD accounts for <5% of the
starts to decline 3-7days after delivery and return to pre- disease and is characterized by a severe deficiency in vWF,
pregnancy levels within 7-21 days of delivery. Therefore, resulting in a corresponding deficiency of FVIII. Under the
close postpartum monitoring of women with bleeding disor- regulation of estrogen that occurs in pregnancy, the levels of
ders is recommended for, at minimum, 3 weeks. vWF, factor FVIII, and most other clotting factors increase,
generally beginning in the early second trimester and peak-
ing between 29 and 35 weeks. For this reason, a diagnosis of
vWD may be masked if VWF levels are performed when a
vWD is the most common inherited bleeding disorder. patient is pregnant or within 6-8 weeks of delivery, lactating,
Although approximately 1% of the general population is or is using a hormonal contraceptive. Thus, whenever

possible, the preconception vWF level and bleeding history time of caesarean delivery, and desmopressin may result in
should be established. During pregnancy, most patients fluid retention, life-threatening hyponatremia, and/or sei-
with type 1 vWD normalize their levels of vWF and FVIII, zures. Recent studies suggest postpartum replacement
although those with more severe disease may not. Given the should aim for VWF levels >50 IU/dL, with some experts
somewhat unpredictable nature of these responses, measure- suggesting a range of 100-200 IU/dL which are observed in
ment of vWF levels should be performed around 34-36 normal pregnant women, as well as postpartum adjuvant
weeks; levels generally remain fairly stable through the tranexamic acid, to reduce PPH. Monitoring for bleeding is
remainder of pregnancy, and thus, levels obtained at this recommended for at least 3 weeks, and, preferably, up to 6
time allow for a delivery plan to be developed. Although lev- weeks postpartum.
els of vWF may increase in patients with type 2 vWD, func-
tional levels may not be significantly enhanced due to the
Hemophilia carriers
production of a functionally defective protein. Levels of vWF
generally do not increase during pregnancy in patients with Postpartum bleeding may occur in 10% hemophilia carriers
type 3 vWD. and may lead to significant blood loss and anemia, in some
In most cases, the physiologic increase in vWF during cases requiring transfusion. The factor level does not predict
pregnancy exceeds the minimum 50 IU/dL vWF level recom- bleeding risk: up to 30% of hemophilia carriers even with
mended for epidural anesthesia and delivery; however, normal factor VIII and IX levels may have high bleeding
in women with a severe bleeding history or levels below scores, and up to 15% may be considered to be mild hemo-
50 IU/dL at the time of testing, vWF levels should be main- philia. In a small case series of 16 deliveries in five hemo-
tained at delivery at or above 50 IU/dL. This recommenda- philia B carriers, postpartum bleeding was reduced
tion is based on case series and expert opinion, as there are no significantly in those women receiving at least 4 days of factor
randomized trials regarding safe vWF levels for regional IX replacement, as compared with fewer than 4 days of
anesthesia. If epidural anesthesia is used, it is generally judi- replacement. In the hemophilia carrier expecting an affected
cious to remove the catheter as soon as possible after delivery infant, the risk of intracranial hemorrhage is 3-5%. Thus,
is completed. Postpartum, the decline in vWF levels generally although not proven, caesarian delivery is recommended over
occurs over 2-3 weeks, and it may be unpredictable and occa- vaginal delivery to reduce that risk. In high-risk infants, the
sionally precipitous; thus, the period of 3-6 weeks postpar- critical issue is the availability of a multidisciplinary team in
tum is considered a particularly vulnerable time for an obstetric unit with facilities for high-risk deliveries. Of
postpartum bleeding and close follow-up is recommended. course, the problem is that many carriers are not diagnosed
Not only is postpartum bleeding more common in pregnant until after delivery, and even in those who are known carriers,
women with vWD, so too is transfusion requirement, longer an affected infant may not be anticipated if they are not prop-
hospital length of stay, and mortality, which may be up to erly counseled. It should be recognized that pre-conception
1.2%. Several therapeutic options are available. Desmopressin counseling with genotyping is currently available as well as
(1-desamino-8D-arginine vasopressin [DDAVP]) is contra- pre- and postimplantation options, including pre-implantation
indicated with neuraxial anesthesia, as the latter requires fluid genetic diagnosis and postimplantation fetal DNA sex deter-
bolus replacement which is contraindicated with desmopres- mination, chorionic villus sampling, and amniocentesis.
sin. Thus, for patients with type 1 vWD, including those aller- Management: For hemophilia A (or B) carriers with FVIII
gic or unresponsive to desmopressin, and those with type 2 (or FIX) levels <50 IU/dL or severe past bleeding history,
and type 3 vWD, vWF concentrate is recommended and con- recombinant factor VIII (or IX) concentrate is recom-
tinued for up to 3-5 days postpartum, as required by disease mended at the time of neuraxial anesthesia and continued
severity and mode of delivery (Table 3-4). for up to 3-4 days postpartum. In women with hemophilia in
Management: Based on case reports and expert opinion, it whom an affected infant is anticipated, because of the poten-
is recommended that pregnant women with type 1 vWD and tial risk of central nervous system bleeding, caesarean deliv-
vWF <50 IU/dL in the eighth month of pregnancy with a ery should be offered. Vacuum and forceps should be
past severe bleeding history, or unresponsive or allergic to avoided because of the risk of cephalohematoma and intra-
desmopressin, and those with type 2 or 3 vWD receive vWF cranial hemorrhage. A team approach (Table 3-1), including
concentrate for neuraxial anesthesia and for up to 3-5 days the obstetrician, anesthesiologist, and hematologist, and
postpartum. Monitoring for postpartum bleeding is recom- communication regarding mode of delivery and factor
mended for at least 3 weeks and preferably 6 weeks postpar- replacement is critical in managing carriers. Factor VIII con-
tum. vWF concentrate is preferred over desmopressin at centrate is preferred over the use of desmopressin for deliv-
delivery. This is because women commonly receive 1-2 liters ery. Although mild hemophilia A carriers may prefer
of fluid at the time of vaginal delivery and 2-3 liters at the desmopressin for treatment of minor procedures, its use

is discouraged at delivery. Women commonly receive 1-2 those with afibrinogenemia, but less frequent in those with
liters of fluid at the time of vaginal delivery and 2-3 liters at hypofibrinogenemia or dysfibrinogenemia. Fibrinogen plays
the time of caesarean delivery, and desmopressin at delivery an important role in placental implantation and mainte-
may result in fluid retention and life-threatening hyponatre- nance of placental competency during pregnancy. When
mia and/or seizures. defects in fibrinogen conversion to fibrin occur during preg-
nancy, whether from deficient or defective fibrinogen, pla-
cental separation, miscarriage, spontaneous abortion, and
Rare bleeding disorders
hemorrhage may occur. The high rate of pregnancy compli-
The rare bleeding disorders include inherited deficiencies of cations may be reduced by fibrinogen replacement (Table
coagulation factors I, II, V, VII, X, XI, and XIII, which repre- 3-4). Several experts suggest that fibrinogen replacement
sent 5% of all inherited bleeding disorders. There is a wide should be initiated as early as possible in pregnancy.
spectrum of bleeding severity, from no bleeding to severe Management: For pregnant women with fibrinogen
bleeding, and it i.s difficult to predict bleeding risk. Thus, a <0.5 g/L and previous adverse pregnancy outcomes, prophy-
diagnosis of a rare bleeding disorder may not come to clini- laxis throughout pregnancy with fibrinogen concentrate ini-
cal attention until a woman, even with prior bleeding history, tially 50-100 mg/kg twice weekly adjusted to maintain
experiences postpartum bleeding. In general, risk is related fibrinogen activity >1 g/L to achieve a level of 1.5 g/L during
to factor levels, but not entirely. The key to optimal delivery labor and for 3 days postpartum. For pregnant women with
management is awareness of the diagnosis, testing the appro- thrombotic dysfibrinogenemia or with afibrinogenemia or
priate factor level at the eighth month of pregnancy, and hypofibrinogenemia and other risk factors for VTE, throm-
replacement therapy at delivery for factor deficiency. Because boprophylaxis should be considered.
coagulation factors generally increase during pregnancy, a
diagnosis may be masked during pregnancy and testing
Factor XIII deficiency
should precede pregnancy whenever possible.
In particular, factors I, VII, VIII, vWF, X, XII, and XIII Factor XIII deficiency is a rare disorder, occurring in 1 in 2
increase during pregnancy, whereas factors II, V, IX, and XI million, that is associated with pregnancy loss in over 90%.
show minimal or no increase. In general, fibrinogen levels Long-term prophylaxis is advised in all factor XIII deficient
>1.0 g/L, FII >25 IU/dL, FV >15 IU/dL, FVII >40-60 IU/dL, patients with a personal or family history of bleeding and
FX >10-20 IU/dL, FXI >10-20 IU/dL, and FXIII >30 IU/dL those with FXIII activity <0.1 IU/mL, starting at 10-40 IU/kg
are recommended, respectively, for each deficiency state, at every 2-3 weeks adjusted to maintain a trough XIII activity
the time of delivery. FVII and FXI may require no replace- above 20 IU/dL.
ment therapy (Table 3-4). When possible, preconception Management: For pregnant women with factor XIII defi-
counseling should be provided and genetics and reproduc- ciency, it is recommended the frequency of prophylaxis be
tive choices should be discussed. Although prenatal diagno- increased to 10-40 IU/kg every 2-3 weeks, aiming for a FXIII
sis with chorionic villus sampling and amniocentesis is activity above 20 IU/sL. At the time of delivery, an additional
possible, few obtain it, given the associated 1%-2% fetal loss. dose of 10-40 IU/kg is advised.
As noted, a team approach with a coordinated management
plan optimizes outcomes for affected women.
Management: Based on expert opinion, for rare bleeding Thromboembolism and thrombophilia
factor deficiency states, FFP or factor concentrate to bring fac- in pregnancy
tors to hemostatic levels (Table 3-4) is recommended at the
time of neuraxial anesthesia and for 3-4 days postpartum for Pregnant women are at increased risk for the development
vaginal delivery and up to 5-7 days for caesarean delivery. of thromboembolism. Venous thromboembolism (VTE)
affects 0.66 to 2.22 of 1,000 pregnancies, with a five- to ten-
fold increased relative risk relative to non-pregnant women.
Hypofibrinogenemia
Furthermore, VTE is one of the leading causes of maternal
Fibrinogen abnormalities, including afibrinogenemia, hypo- death in developed countries, accounting for 1.2 to 4.7
fibrinogenemia, and dysfibrinogenemia may be associated deaths per 100,000 pregnancies. VTE is symptomatic in 5-12
with hemorrhagic and thrombotic pregnancy complications, per 10,000 pregnancies and in 3-7 per 10,000 deliveries, for a
including postpartum hemorrhage, spontaneous abortion, seven- to tenfold increased risk for VTE in the antepartum
and placental abruption. Up to 30% of patients with con- period and a 15- to 35-fold increased risk in the postpartum
genital fibrinogen deficiency have thrombotic complications, period. The risk of VTE may be higher in the third than the
most commonly first-trimester abortion: this is common in first and second trimesters, but the increased VTE risk is

Thromboembolism and thrombophilia in pregnancy | 65
clearly present from the first trimester even before many of protein C resistance in the absence of factor V Leiden (FVL),
the anatomic changes of pregnancy occur. Approximately unexplained by the decrease in free proteins S, also is observed
80% of VTE events are deep venous thrombi (DVT), whereas in some pregnant patients, particularly in the third trimester.
20% are pulmonary embolism: one-third of pregnancy- Furthermore, there is also a decrease in tissue plasminogen
associated DVT and one-half of pulmonary emboli occur activator activity.
after delivery. The relative risk of pulmonary embolism is Management: The anticoagulant of choice in pregnancy is
four- to fivefold greater after caesarean than vaginal delivery. low molecular weight heparin, LMWH (Table 3-5).
The risk of arterial thromboembolism is also increased LMWH is preferred over VKA as well as over subcutaneous
approximately three- to fourfold in pregnant women, reflect- unfractionated heparin (SUH), as (compared with SUH)
ing the hypercoagulable state associated with pregnancy. it is associated with less bleeding risk, more predictable ther-
apeutic response, lower risk of HIT, longer half-life, less
bone density loss, and less local skin hemorrhage. LMWH
Pregnancy-associated venous
(as well as SUH) does not cross the placenta, and numerous
thromboembolism
studies have confirmed its use is safe for the fetus.
Pregnancy-associated DVT is more often proximal and mas- At the time of delivery, full-dose LMWH should be dis-
sive than in the nonpregnant setting and usually occurs in the continued at least 24 hours before the induction of labor or
left lower extremity. In contrast, distal DVT occurs with simi- caesarean delivery (or expected time of neuraxial anesthesia)
lar frequency in the left and right lower extremities. The left- and prophylactic dose LMWH should be discontinued at
sided predominance of VTE may reflect compression of the least 12 hours before. Alternatively, the patient can be
left iliac vein between the right iliac artery and lumbar verte- switched to SUH 2-4 weeks before the delivery date to allow
brae. A diagnosis of VTE is difficult in pregnancy because of a for option of an epidural, particularly in case of premature
lack of validation of standard diagnostic studies in this popu- delivery, although a prolonged anticoagulant effect has been
lation. Although an abnormal compression ultrasound (CUS) observed for up to 28 hours with every 12 hour dosing.
is considered sufficient for diagnosis of DVT during preg- LMWH or SUH can be resumed within 4-6 hours after a
nancy, a normal CUS does not reliably exclude DVT because vaginal delivery and 6-12 hours after a caesarean delivery, or
of the low sensitivity for isolated iliac DVT. MRI is the “gold longer if peripartum bleeding occurs: in the event that SUH
standard” test to diagnose iliac DVT, although an alternative or LMWH are held, sequential pneumatic compression
strategy is negative serial CUS with imaging of the iliac veins. devices should be used. If an epidural is in place, prophylac-
Chest radiography (CXR) is recommended by the American tic LMWH should be resumed no sooner than four hours
Thoracic Society (ATS), the Society of Thoracic Radiology after epidural removal, and full dose LMWH no sooner than
(STR), the ACCP, and ACOG as the first-line radiation-associ- 12 hours after epidural removal.
ated procedure for diagnosis of pulmonary embolism (PE) For pregnant women with acute VTE, adjusted full-dose
unless DVT signs/symptoms are present when CUS should be subcutaneous LMWH is recommended during pregnancy
performed. Lung scintigraphy is preferred if the CXR is nor- and continued or transitioned to a vitamin K antagonist tar-
mal. Computed tomographic pulmonary angiography (CTA) geted to an INR 2.0-3.0 for at least 6 weeks postpartum, for a
is preferred if the screening CXR is abnormal. As D-dimer is minimum total duration of 3 months. VKA is permissible
elevated in pregnancy, it should not be used to exclude PE in during breastfeeding and should be started the evening after
pregnant women. Enhanced hypercoagulability is the major delivery. In patients markedly prothrombotic, with multiple
cause of thrombosis in pregnant women. During pregnancy, risk factors for VTE in the prior 6 months, anticoagulation
hormonal changes, specifically increases in estrogen and pro- may be considered for up to 12 weeks postpartum: the odds
gesterone, lead to a transient hypercoagulable state. This tran- ratio for VTE is 10.8 for up to 6 weeks postpartum, as com-
sient procoagulant state was teleologically important, most pared with the same period one year later, while the odds
likely to protect against miscarriage or fatal hemorrhage at ratio is 2.2 for VTE from 7-12 weeks postpartum, compared
birth. Factor levels that increase during pregnancy include fac- with the same period one year later. While studies do not
tors I (fibrinogen), VII, VIII, X, vWF, and plasminogen activa- conclusively support routine Xa monitoring in pregnant
tor inhibitor (PAI-I and PAI-2), all of which return to normal women requiring full-dose anticoagulation, periodic Xa
beginning 2-3 weeks postpartum. In parallel, a substantial monitoring is reasonable in the obese patient since Xa levels
decrease in free protein S levels occur because of increased lev- fall by 25% in the third trimester as the blood volume rises.
els of C4b binding protein. It is important to recognize that Pharmacokinetics are optimized if dosing for an acute VTE is
protein S levels are as low as 30% in the second trimester and at 1 mg/kg twice daily, with an anti-Xa target of 0.6-1.0 IU/mL,
26% in the third trimester of normal pregnancies, to avoid or at 1.5 mg/kg/d once daily dosing, with an anti-Xa target
misdiagnosis of protein S deficiency. An increase in activated 0.8-1.6 IU/mL, after hospital discharge. For pregnant women

Table 3-5 American College of Chest Physicians (2012) Guidelines for Antithrombotic Therapy in Pregnancy*
Scenarios
Acute VTE
For pregnant women with acute VTE, adjusted-dose subcutaneous LMWH For pregnant women, LMWH is recommended for
is recommended during pregnancy and continued for at least 6 weeks antithrombotic therapy over adjusted-dose heparin and over
postpartum, for a minimum duration of 3 months. vitamin K antagonists.
For pregnant women receiving adjusted-dose LMWH, LMWH
should be stopped at least 24 hours before induction of labor,
caesarean delivery, or neuraxial anesthesia.
Prior episode of VTE
For pregnant women with prior VTE, postpartum prophylaxis for 6 For women on long-term vitamin K antagonists pre-pregnancy,
weeks with prophylactic- or intermediate-dose LMWH or vitamin adjusted-dose LMWH or 75% of a therapeutic dose of LMWH
K antagonists targeted at INR 2.0-3.0. is recommended during pregnancy and long-term postpartum.
For pregnant women with low risk of recurrent VTE (single-episode
VTE with transient risk factor unrelated to pregnancy or use of
estrogen), antepartum clinical vigilance and postpartum prophylaxis
with prophylactic or intermediate LMWH or Vitamin K antagonists
targeted to INR 2.0 to 3.0 is recommended.
For pregnant women with moderate to high risk of recurrent VTE
(unprovoked pregnancy- or estrogen-related VTE or multiple
unprovoked VTE on no long-term anticoagulation), antepartum
prophylactic- or intermediate-dose LMWH is recommended.
No prior VTE
High-risk thrombophilia: For pregnant women with no prior VTE For pregnant women with all other thrombophilias (“low-risk”)
who have homozygous factor V Leiden or the prothrombin 20210A and no prior VTE and no family history of VTE, antepartum
mutation and a family history of VTE [ACOG also Includes combined and postpartum clinical vigilance is recommended. [ACOG
FVL and PTGM heterozygotes or ATIII deficiency and does not require advises postpartum prophylaxis if there is a first-degree relative
a family history of VTE], antepartum prophylaxis with prophylactic- with a history of a thrombotic episode before age 50 years, or
or intermediate-dose LMWH and postpartum prophylaxis for 6 other major thrombotic risk factors, eg, prolonged immobility].
weeks with prophylactic- or intermediate-dose LMWH or vitamin K
antagonists targeted at INR 2.0 to 3.0 is recommended.
Low-risk thrombophilia: For pregnant women with all other For pregnant women undergoing caesarean delivery without
thrombophilias and no prior VTE who have a family history of VTE additional thrombosis risk factors, early mobilization but not
[ACOG defines family history as a first degree relative], antepartum thrombosis prophylaxis is recommended.
clinical vigilance and postpartum prophylaxis with prophylactic-
or intermediate-dose LMWH, or in women who are not protein
C or S deficient, vitamin K antagonists targeted to INR 2.0-3.0 is
recommended.
For pregnant women with no prior VTE who are known to be homozygous For women at increased risk of VTE after caesarean delivery
for factor V Leiden or the prothombin 20210A mutation, but no family because of the presence of one major and at least two
history of VTE [ACOG advises antepartum prophylaxis even if no family minor risk factors, pharmacologic thromboprophylaxis
history and includes FVL and PTGM heterozygotes or ATIII deficiency], (prophylactic LMWH) is recommended or mechanical
antepartum clinical vigilance and postpartum prophylaxis for 6 prophylaxis (elastic stockings or intermittent pneumatic
weeks with prophylactic- or intermediate-dose LMWH or vitamin K compression) for those with contraindications to
antagonists targeted to INR 2.0 to 3.0 is recommended. anticoagulants while hospitalized.
Adapted from Bates SM et al. Chest. 2012;141(2 suppl):e691S–e736S.
INR = international normalized ratio; LMWH = low-molecular-weight heparin; VTE = venous thromboembolism; PTGM = prothrombin gene
mutation (PT20210A).
*See American Congress of Obstetrics and Gynecology (2013) Practice Bulletin (#138) Guidelines.
with hemodynamically unstable PE, tissue plasminogen acti- For pregnant women with prior VTE, postpartum pro-
vator (TPA) has been given if the benefit outweighs the risk, phylaxis for 6 weeks is recommended, with prophylactic or
a reported 6-15% fetal death rate and 8-30% major maternal intermediate-dose LMWH or VKAs targeted to INR 2.0-3.0.
bleeding rate. For pregnant women with low risk of recurrent VTE

(eg, associated with transient risk factor unrelated to preg- intermediate-dose LMWH is recommended. Antepartum
nancy or use of estrogen), antepartum clinical vigilance and prophylaxis can be considered for pregnant women with
postpartum prophylaxis with low dose or intermediate dose no prior VTE and known ATIII deficiency. Women with ATIII
LMWH or VKAs targeted to INR 2.0-3.0, is recommended. deficiency may also be candidates for anti-thrombin concen-
For pregnant women at moderate to high risk of recurrent trates peripartum. For pregnant women with all other throm-
VTE (eg, single unprovoked VTE, pregnancy- or estrogen- bophilias and no prior VTE with a family history of VTE,
related VTE, or multiple prior unprovoked VTE not receiv- antepartum vigilance is r ecommended, and postpartum pro-
ing long-term anticoagulation), antepartum prophylaxis phylaxis with prophylactic- or intermediate-dose LMWH or
with prophylactic or intermediate-dose LMWH and post- VKAs targeted to INR 2.0-3.0. For pregnant women with prior
partum prophylaxis for 6 weeks, with low dose or interme VTE, postpartum prophylaxis for 6 weeks with prophylactic-
diate dose LMWH or VKAs targeted to INR 2.0-3.0, is or intermediate-dose LMWH or VKAs targeted to INR 2.0-3.0
recommended. A recent retrospective case series suggests low is recommended.
dose LMWH postpartum prophylaxis may be inadequate
compared to intermediate dose LMWH, and a randomized
Thrombophilia and pregnancy complications
trial is planned to address this issue.
A number of pregnancy complications have been linked to
thrombophilic states. Adverse pregnancy outcomes, how-
Pregnancy and thrombophilia, with and
ever, are not uncommon in the general population, with a
without thrombosis
15% rate of miscarriage and a 5% rate of two or more preg-
Between 20% and 50% of all thromboembolic events that nancy losses. The association between thrombophilia and
occur in pregnant women are associated with a thrombo- pregnancy loss has been confirmed in a number of case-
philic disorder. Published studies indicate the relative risk of control studies for women with thrombophilia, but it has
VTE associated with a thrombophilic disorder in pregnant not been confirmed in the methodologically stronger cohort
women without a family history of VTE is increased about studies. Although a single late fetal loss and severe pre-
34-fold in women who are homozygous for FVL mutation, eclampsia are associated with inherited thrombophilia, fetal
eightfold in women heterozygous for FVL mutation, 26-fold growth restriction and placental abruption are not. In a
for women homozygous for prothrombin G20210A gene meta-analysis of 25 studies, mostly case-control studies,
mutation, and nearly sevenfold in women heterozygous for other than homozygous FVL and homozygous prothrombin
prothrombin G20210A mutation. The positive predictive 20210G, the pooled risk for pregnancy loss was equivocal.
value for pregnancy-associated thrombosis is 1:400-1:500 for More rigorous studies that eliminate patients with previous
the FVL mutation, 1:200 for prothrombin G20210A, and VTE or VTE in pregnancy from the analysis do not support
4.6:100 for double heterozygotes of these mutations, assum- significant risk of recurrent pregnancy loss (Table 3-6).
ing an overall thrombosis rate of 0.66/1,000 pregnancies. Neither has there been any demonstrated association

The positive predictive value was 1:113 for protein C defi- between thrombophilia and preeclampsia, placental abrup-
ciency and 1:2.8 for type I antithrombin deficiency. Impor- tion, or fetal growth restriction.
tantly, regardless of presence of thrombophilia, a positive Management: For women with a history of pregnancy
family history increases the risk for VTE two- to fourfold. complications, screening for inherited thrombophilia is not
Cohort studies in women with protein C deficiency (with no recommended, and, for this group, neither is antithrombotic
past VTE) and a positive family history have a 1.7% risk of prophylaxis recommended. For women with inherited
developing a first antepartum or postpartum VTE, whereas thrombophilia and a history of pregnancy complications,
women who are homozygous carriers for FVL and a positive antithrombotic prophylaxis is not recommended, particu-
family history have a 14% risk. Given the differing absolute larly with the recent results of a randomized trial, the
risks of thrombosis, management should be individualized “TIPPS” trial, showing lack of benefit with antepartum
based on the type of thrombophilia, presence of homozy- dalteparin in pregnant women with thrombophilia and pre-
gous or heterozygous mutations, history of past VTE or vious placenta-mediated pregnancy complications.
pregnancy complications, and presence or absence of a fam-
ily history of VTE, as well as the presence of additional pro-
Antiphospholipid antibody syndrome
thrombotic conditions (Table 3-5).
Management: For pregnant women with no prior VTE and The strongest evidence of an association between thrombo-
known to be homozygous for FVL or prothrombin 20210GA philia and fetal loss comes from studies in patients with
mutation or a combined heterozygote, and a family history APLAs. A diagnosis of APLA requires both laboratory
of VTE, antepartum prophylaxis with prophylactic- or and clinical criteria. The laboratory criteria require the

Table 3-6 American College of Chest Physicians guidelines for thrombophilia-related fetal loss
Scenario Comments
Thrombophilia and pregnancy complications

For women with inherited thrombophilia and a history of pregnancy For women with a history of pregnancy complications, screening for
complications, antithrombotic therapy is not recommended. inherited thrombophilia is not recommended.
Antiphospholipid antibody syndrome (APLAs)
For women who fulfill the laboratory criteria for APLA syndrome For women with recurrent early pregnancy loss, ≥3 pregnancy
and meet the clinical APLA criteria based on ≥3 pregnancy losses before 10 weeks of gestation, screening for APLAs is
losses in the first trimester, or one beyond 9 weeks, or a recommended.
morphologically normal neonate before the 34th week due to
eclampsia or severe preeclampsia, antepartum prophylactic- or
intermediate-dose UFH or prophylactic LMWH combined with
low-dose aspirin, 75-100 mg daily, is recommended.
Artificial reproductive technology (ART)
For women developing ovarian hyperstimulation syndrome (OHSS),
there is a 3-fold risk of thrombosis, and LMWH is recommended
for three months postembryo transfer. Prophylaxis is also
appropriate in non-severe OHSS if there high-risk thrombophilia
present (homozygous FVL or PTGM) or low-risk thrombophilia
with a family history of thrombosis.
Paroxysmal nocturnal hemoglobinuria (PNH)
For women with PNH, antepartum and postpartum prophylactic- Eculizumab is listed as category C in pregnancy, and its use is not
or intermediate-dose LMWH is recommended, particularly in established in pregnancy.
those with PNH clones >50%, prior history of VTE, or recurrent
previous late fetal losses.
Heparin-induced thrombocytopenia (HIT)
In pregnant women with HIT, or in those with allergic reaction As safety has not been established in pregnancy, the use of
to heparin or LMWH, antithrombotic therapy with dabigatran (a new oral direct thrombin inhibitor) and
fondaparinux or lepirudin in place of heparin or LMWH is rivaroxaban or apixaban (new oral anti-Xa inhibitors) should be
recommended. avoided.
Adapted from Bates SM et al. Chest. 2012;141(2 suppl):e691S–e736S; Bornikova L et al. J Thromb Haemost. 2011;9:1687-1704; Brodsky RA.
Blood. 2009;113:6522-6527.
LMWH = low-molecular-weight heparin; UFH = unfractionated heparin.
presence of lupus anticoagulant or moderate to high titer proportion of APLA-positive patients experience late fetal
antibodies to immunoglobulin G (IgG) or immunoglobulin loss (after the 10th week of gestation). Inherited thrombo-
M (IgM) anticardiolipin (>99th percentile) or IgG or IgM philia is less strongly associated with pregnancy loss than
β-2-glycoprotein I (>99th percentile) on two occasions at APLA. Several randomized studies, none of which were pla-
least 12 weeks apart; and the clinical criteria require >3 first cebo controlled, have examined the effect of treatment of
trimester pregnancy losses, or >1 pregnancy loss at or beyond women with APLAs with aspirin, heparin, or both. These
10 weeks gestation, or >1 premature birth before 34 weeks of studies, which have been small and with heterogeneous crite-
gestation. The importance of β-2-glycoprotein I antibodies ria, generally have demonstrated an advantage of aspirin and
is less well established. Correlation of APLAs with fetal heparin over either aspirin or heparin alone, although a
growth restriction or placental abruption remains controver- recent, randomized trial was stopped early when it became
sial. Among women with recurrent fetal loss (>3 miscar- evident that LMWH and aspirin offered no advantage over
riages), 15% have APLA. In some studies, such women aspirin alone, with almost 80% of women in both arms hav-
experience fetal loss at a rate approaching 90% in subsequent ing successful pregnancies.
pregnancies, whereas in other older studies, normal out- Management: For women who fulfill laboratory and cli
comes may occur, even in the absence of treatment. Although nical criteria for APLAs, antepartum prophylactic- or
the majority of fetal losses in normal individuals and patients intermediate-dose unfractionated heparin (UFH) or pro-
with APLA occur early in the first trimester, an increased phylactic LMWH combined with low-dose aspirin, 75-100 mg

daily is recommended. For women with two or more miscar- disease (see Table 3-7), VKAs ideally should be avoided at
riages but no APLA or thrombophilia, no antithrombotic least during weeks 6-12 of pregnancy. Insufficient data exist
prophylaxis is recommended (Table 3-6). APLA screening is regarding safety or potential teratogenic effects of the new
recommended in women with recurrent pregnancy loss oral anti-Xa and new oral thrombin inhibitors to recom-
(>3 miscarriages before 10 weeks or one beyond 9 weeks), or mend their use in pregnancy.
with thrombophilia and recurrent pregnancy loss. Management: In pregnant women, LMWH is the pre-
ferred antithrombotic agent, as discussed above.
Assisted reproductive technology
Oral vitamin K antagonists
Approximately one in six couples experience infertility. In
achieving a subsequent successful pregnancy, assisted repro- Several toxicities of anticoagulant therapy unique to preg-
ductive technology (ART) is employed, primarily, in vitro nancy must be considered when developing anticoagulation
fertilization. ART is coupled with ovarian follicle stimulation treatment approaches. First, the oral VKA warfarin is terato-
by gonadotropins and gonadotropin-stimulating hormones. genic, causing an embryopathy consisting of nasal hypopla-
Despite this, successful pregnancy results in only of 35-40% sia or stippled epiphyses and limb hypoplasia. The frequency
of attempts (cycles). Thrombophilia has been proposed as a of these abnormalities is estimated to be between 5-12%.
potential mechanism in cases of failed reproduction. Case- The teratogenic effects occur primarily following exposure
control studies have indicated a threefold risk of failure in to warfarin during weeks 6-12 (primarily at 6-9 weeks) of
the presence of FVL or APLA, but cohort studies have not pregnancy, whereas warfarin is probably safe pre-conception
substantiated either finding. Furthermore, antithrombotic and during the first 6 weeks of gestation. VKAs used at any
therapy with aspirin or LMWH does not appear to apprecia- time during pregnancy have been associated with rare cen-
bly increase the success rate. On the other hand, ART is asso- tral nervous system (CNS) developmental abnormalities,
ciated with an increased relative risk of VTE in the threefold such as dorsal midline dysplasia and ventral midline dyspla-
range but with a low absolute risk. Arterial events have also sia, leading to optic atrophy. Finally, an increased risk of
been reported. The risk of thrombosis is increased if concur- minor neurodevelopmental abnormalities may occur in the
rent ovarian hyper-stimulation (OHSS) syndrome develops. offspring of women exposed to warfarin during the second
OHSS occurs in a third of cycles and is characterized by and third trimesters, although the significance of these prob-
abdominal pain, bloating, and fluid retention. OHSS is pres- lems is uncertain. Warfarin may cause a dose-dependent
ent in 90% of the arterial events, at a median of 11 days post- anticoagulant effect in the fetus, which may lead to bleeding
embryo transfer, and in ~80% of the venous event, at a at delivery. In at least one series, warfarin was found to
median of 42 days postembryo transfer. Most venous events increase the rate of miscarriage, leading to the recommenda-
occur in the neck or arm veins. tion that heart valve patients receiving VKA during the sec-
Management: Several guidelines advise prophylactic ond trimester of pregnancy should be switched to heparin
LMWH in severe OHSS, withholding LMWH for 12-24 beginning at 36 weeks of pregnancy.
hours before oocyte retrieval then resuming 6-12 hours after Management: In pregnant women, LMWH is the preferred
retrieval and continuing for three months. Prophylaxis is antithrombotic agent.
also appropriate in non-severe OHSS if there is high-risk
thrombophilia (homozygous FVL or PTGM) or low-risk
Heparin-induced thrombocytopenia
thrombophilia with a family history of thrombosis.
Heparin-induced thrombocytopenia (HIT) is an uncom-
mon problem in pregnancy, but with only case series and
Anticoagulation during pregnancy
anecdotal reports, the exact frequency is not known but esti-
A number of scenarios exist concerning the prevention and mated to be between 0.1 and 1% at first UFH (SUH) expo-
treatment of primary and recurrent thrombosis in pregnant sure. Whether platelet counts should be monitored in the
individuals. In general, when anticoagulation is indicated, pregnant patient on LMWH is not established. The mecha-
the agent of choice in pregnancy is LMWH or heparin, based nism by which HIT occurs is binding of antibodies directed
on evidence-based guidelines recently updated by the ACCP against the heparin-platelet-factor 4 complex, so-called hep-
(see the section “Bibliography” in this chapter). Among arin-platelet factor 4 (HPF4) antibodies, which result in
women with protein C or S deficiency, postpartum VKAs thrombocytopenia, with typically a greater than 50% drop
should be used very cautiously with adequate bridging with below baseline platelet count. The risk of thrombosis is as
LMWH, given the recognized low levels of protein C and S high as 50% in patients with HIT, and, thus, anticoagulation
due to pregnancy alone. In women with valvular heart is critical to prevent thrombosis or pregnancy-related

Table 3-7 Guidelines for pregnant patients with artificial heart valves
2012 American College of Chest Physician 2014 American Heart Association/American College of
guidelines for pregnant patients with artificial Cardiology recommendations for pregnant women with artificial
heart valves heart valves
Pregnant women with prosthetic heart valves

For women with mechanical heart valves one For women with mechanical heart valves, the following recommendations
of the following anticoagulant regimens is are made:
recommended: Class I (benefit >>> risk) ie, treatment should be given:
1. Adjusted-dose bid LMWH throughout 1. Therapeutic anticoagulation with frequent monitoring is recommended for
pregnancy, adjusted to achieve the all pregnant patients with a mechanical prosthesis.
manufacturer’s peak anti-Xa LMWH 4 hours 2. Warfarin is recommended in pregnant patients with a mechanical prosthesis
after injection; or to achieve an INR in the second and third trimesters.
2. Adjusted-dose UFH throughout pregnancy 3. Discontinuation of warfarin with initiation of intravenous UFH (with
administered SC every 12 hours in doses aPTT >2 times control) is recommended before planned vaginal delivery in
adjusted to keep the mid-interval aPTT at pregnant patients with a mechanical prosthesis.
least twice control or to attain an anti-Xa
Class IIa (benefit >> risk) ie, treatment is reasonable:
heparin level of 0.35–0.70 U/mL; or
4. Continuation of warfarin during the first trimester is reasonable for pregnant
3. UFH or LMWH (as above) until the 13th
patients with a mechanical prosthesis if the dose of warfarin to achieve a
week, with substitution by vitamin K
therapeutic INR is 5 mg/day or less after full discussion with the patient about
antagonists until close to delivery when UFH
risks and benefits.
or LMWH is resumed.
5. Dose-adjusted LMWH at least two times per day (with a target anti-Xa level
Pregnant women with prosthetic valves at high risk of 0.8–1.2 U/mL, 4–6 h post dose) during the first trimester is reasonable
of thromboembolism for pregnant patients with a mechanical prosthesis if the dose of warfarin is
1. For pregnant women with prosthetic heart >5 mg/day to achieve a therapeutic INR.
valves at very high risk of thromboembolism 6. Dose-adjusted continuous intravenous UFH (with an aPTT at least two
in whom concerns exist about the efficacy times control) during the first trimester is reasonable for pregnant patients
and safety of UFH or LMWH (eg, older with a mechanical prosthesis if the dose of warfarin is >5 mg/day to achieve a
generation prosthesis in the mitral position therapeutic INR.
or history of thromboembolism), vitamin
Class IIb (benefit > risk) ie, additional studies are needed:
K antagonists are suggested throughout
7. Dose-adjusted LMWH at least two times per day (with a target
pregnancy with replacement by UFH or
anti-Xa level of 0.8–1.2 U/mL, 4–6 h post dose) during the first trimester
LMWH (as above) close to delivery.
may be reasonable for pregnant patients with a mechanical prosthesis
2. For pregnant women with prosthetic valves
if the dose of warfarin is 5 mg/day or less to achieve
at high risk of VTE, additional low-dose
a therapeutic INR.
aspirin, 75-100 mg/d to one of the regimens
8. Dose-adjusted continuous infusion of UFH (with aPTT at least two times
is suggested.
control) during the first trimester may be reasonable for pregnant patients
with a mechanical prosthesis if the dose of warfarin is 5 mg/day or less to
achieve a therapeutic INR.
Class III (treatment is not effective, may be harmful):
9. LMWH should not be administered to pregnant patients with
mechanical prostheses unless anti-Xa levels are monitored 4-6 h after
administration.
Adapted from Bates SM, Greer IA, Middeldorp S, Veenstra DL, Prabulos AM, Vandvik PO. Chest. 2012;141(2 suppl):e691S-e736S; and
Nishimura RA, Otto CM, Bonow RO, et al. J Am Coll Cardiol. 2014;63:2438–2488.
aPTT = activated partial thromboplastin time; LMWH = low-molecular-weight heparin; SC = subcutaneous; UFH = unfractionated heparin;
VTE = venous thromboembolism.
thrombotic complications. Although up to 50% of cardiac considered to be the first-line therapy (Table 3-6).
surgical patients develop HIT, prospective case series evalu- Fondaparinux does cross the placenta, but the umbilical
ating HPF4 antibody in pregnancy have reported low rates anti-Xa levels are subtherapeutic in the fetus at one-tenth
of HPF4 seroconversion and low rates of thrombocytopenia the maternal level; but it must be stopped to allow safe anes-
(HIT). In pregnant patients with HIT, fondaparinux is thesia and delivery. However, the clinician should be

Hematologic health issues in the premenopausal woman | 71
reminded that safety data from first trimester exposure are was 9.2% in women receiving VKAs during pregnancy, with
lacking. Finally, data are insufficient regarding the safety of heparin substituted for VKAs from weeks 6-12, as compared
the new oral anti-Xa inhibitors (eg, rivaroxaban) and oral with a 33.3% risk of thromboembolic complications with
thrombin inhibitors (eg, dabigatran) to recommend their UFH alone throughout pregnancy.
use in pregnancy (Table 3-6). These findings are consistent with other retrospective
Management: In pregnant women with HIT, antithrom- studies demonstrating that VKAs are associated with a lower
botic therapy with fondaparinux in place of heparin or risk of valve thrombosis, maternal thromboembolism, and
LMWH is recommended. systemic emboli. LMWH in lieu of UFH has been used in
pregnant patients with mechanical heart valves, although
valvular thrombosis and maternal thromboembolism occur
Heparin-associated osteoporosis
in up to 22.4% of pregnancies, with the lowest risk, 10-11%,
Prophylactic UFH is associated with a substantial risk of associated with dose-adjusted LMWH, as compared with
osteoporosis and a 2%-3% incidence of vertebral fractures fixed-dose LMWH.
when administered throughout pregnancy. Several reports Management: For pregnant women with prosthetic heart
suggest that substantially fewer instances of osteoporosis valves, guidelines diverge between the 2012 ACCP and the
occur in patients who receive LMWH; however, at least one 2014 AHA/ACC (Table 3-7). AACP advises consideration of
randomized trial found no difference in bone density three approaches: (i) adjusted-dose LMWH throughout
between low-dose UFH and LMWH. pregnancy; (ii) adjusted-dose UFH throughout pregnancy;
Management: In pregnant women, LMWH is recom- or (iii) LMWH or UFH until the 13th week, with substitu-
mended as the preferred antithrombotic agent, as noted tion of VKAs until close to delivery, at which time LMWH or
above. UFH is resumed. For women judged to be at high risk for
thromboembolism, such as older generation prosthesis in
Heparin-associated skin reactions mitral position or previous VTE, VKAs are recommended
throughout pregnancy with replacement by LMWH or UFH
Skin reactions varying from type 1 urticarial eruptions to close to delivery. In women with prosthetic valves at high
type IV delayed hypersensitivity reactions have been reported risk of thrombosis, aspirin, 75-100 mg daily, is also recom-
in 0.3-0.6% of patients receiving heparin. In at least one pro- mended. The 2014 AHA/ACC guidelines advise as a Class 1
spective study of 66 pregnant women, 29% reported pruri- recommendation (ie, benefit >>>risk) the use of VKA in the
tus, local erythema, and less commonly subcutaneous second and third trimesters. For the first trimester, the
infiltrates and pain at the injection site. guidelines stress the complications of VKA for both mother
Management: In pregnant women without signs of a type and fetus are dose-dependent with fewer adverse events
1 reaction, switching to another LMWH preparation is rec- when doses of less than or equal to 5 mg warfarin are used.
ommended. In the approximately one-third in whom skin Consequently, as a Class IIa recommendation (ie, benefit >>
reactions recur after switching from one LMWH prepara- risk), continuation of warfarin is reasonable in the first tri-
tion to another, fondaparinux is recommended. mester when doses less than or equal to 5 mg warfarin are
used; otherwise LMWH should be used in the first trimester.
Artificial heart valves When LMWH is used, adjusted-dose LMWH, based on anti-
Xa levels to a target of 0.8-1.2 IU/mL 4-6 hours after a dose
Without anticoagulant therapy, patients with mechanical
is recommended with monitoring at least every 2 weeks.
heart valves have a high risk of arterial thromboembolism.
Weight-based dosing is not recommended. Although no
Warfarin appears to be more effective than heparin in pre-
guidelines exist for dose-adjustment based on trough anti-
venting valvular thrombosis in these patients. Debate con-
Xa levels, some studies suggest a target anti-Xa level of
tinues, however, as to whether the benefit in prevention of
0.6-0.7 IU/mL.
valvular thrombosis in the mother offsets the risk of
warfarin-induced embryopathy and neurodevelopmental

abnormalities in the fetus. A systematic review of observa-
Hematologic health issues in
tional studies from 1996 to 1997 that assessed outcomes with
the premenopausal woman
various anticoagulants in women with prosthetic heart
valves found that VKAs were associated with the lowest risk
Bleeding in the premenopausal woman
of valve thrombosis and systemic embolism, with no differ-
ence in fetal wastage and bleeding between groups. The inci- Bleeding disorders in women are an under-recognized and
dence of valve thrombosis was 3.9% and systemic embolism undertreated condition. Hemophilia, the most widely known

and studied bleeding disorder, is a disease of males. Women, consensus statement published by experts in the field of
however, are as likely as men to have bleeding disorders obstetrics and gynecology and hematology in 2009 states
other than hemophilia and are in fact disproportionately that an inherited bleeding disorder should be considered if
affected by these diseases due to the bleeding challenges of any of the following indicators are present: (i) HMB since
menstruation and childbirth. Because bleeding disorders in menarche, (ii) family history of a bleeding disorder, or (iii)
women tend to be less severe and specific than hemophilia, it a personal history of one or more additional bleeding
is more difficult for physicians and patients to recognize symptoms.
symptoms and diagnose these conditions. In one national As HMB is such a frequent problem, it would be cost-
survey of 75 women with vWD, the average time from onset prohibitive to screen all women with HMB for underlying
of bleeding symptoms to diagnosis was 16 years. This section bleeding disorders. Identifying patients with either “severe or
will review the most common gynecologic manifestations of significant” HMB is challenging given that actual measurement
bleeding as well as recommendations for the laboratory eval- of menstrual blood loss is not feasible in clinical practice.
uation and management of women presenting with exces- Therefore, an active area of research has been the develop-
sive bleeding. ment and validation of bleeding assessment tools in this
field. One of the first tools developed was the Pictorial Blood
Assessment Chart (PBAC), first published in 1990. To com-
Menorrhagia
plete the chart, women compare both the number and degree
The ACOG defines heavy menstrual bleeding (HMB), or of saturation of pads and tampons with those depicted on a
menorrhagia, as menstrual bleeding lasting for >7 days or chart. Lightly stained products obtain a score of 1, moder-
resulting in the loss of >80 cc of blood per menstrual cycle. ately stained a score of 5, and soaked a score of 20. A total
HMB is a common health problem, occurring in as many as score of >100 per menstrual cycle is associated with men-
10%-15% of women during their lifetime, and this excessive strual blood loss of >80 mL (definition of menorrhagia). A
bleeding can lead to iron deficiency and chronic anemia. major limitation of this tool is that it must be completed
Women with HMB have significantly lower perceived gen- prospectively, so results are not available at the time of initial
eral health and have a poorer quality of life in terms of their evaluation. Moreover, completion of the score after the eval-
ability to fully participate in school, work, athletics, and uation may be limited by subjective bias as well as poor com-
social activities. It is well established that HMB is the most pliance. In a study of 226 women who consented to formal
common bleeding symptom among women with bleeding measurement of menstrual blood loss, variables that pre-
disorders, occurring in up to 80%-90% of patients and that dicted blood loss >80 mL were changing a pad or tampon
bleeding disorders are common among women presenting more than hourly, passing clots >1 inch in diameter, and low
with HMB. Therefore, it is imperative that physicians screen ferritin. More recently, a screening tool developed by Philipp
for underlying bleeding disorders when evaluating an ado- et al. may identify women with HMB who are more likely to
lescent or woman with HMB. Up to 11%-16% of women have an underlying bleeding disorder. The tool contains
with heavy menstrual bleeding and a normal gynecologic eight questions in four categories: (i) severity of HMB,
exam will have von Willebrand factor deficiency. A recent (ii) family history of bleeding disorder, (iii) personal history
opinion issued by the Adolescent Health Committee sup- of excessive bleeding, and (iv) history of treatment for ane-
ports screening for vWD in adolescents presenting with mia (Table 3-8). The screen is considered positive if an affir-
severe menorrhagia. It is important to consider that von mative response is obtained in any one of the four categories.
Willebrand factor levels may be affected by stress, inflam- The sensitivity of this tool for underlying hemostatic defects
mation, anemia, pregnancy, oral contraceptives, phase of in adult women is 89%, which increases to 93%-95% with a
the menstrual cycle, and sample processing. Although some serum ferritin level of ≤20 ng/mL and a PBAC score of >185,
studies suggest that von Willebrand factor levels in healthy respectively. A variety of more general bleeding assessment
individuals are not significantly affected by oral contracep- tools, most modified based on the original Vicenza Bleeding
tive use, further research is needed in this area. Even in the Questionnaire, and including a consensus bleeding assess-
presence of gynecologic disease, such as anovulatory bleed- ment tool set forth by the International Society of Thrombo-
ing in adolescence or fibroids in perimenopause, an under- sis and Haemostasis (ISTH), have been developed. However,
lying bleeding disorder may be an additional contributing the sensitivity and specificity of these instruments in identi-
factor to HMB and should be considered in the evaluation. fying underlying bleeding disorders in women with HMB
In a study of 115 women 13-53 years of age presenting for have not been formally studied. It is important to recognize
evaluation of HMB, adolescents and perimenopausal that because of increased proliferation of the endometrium,
women were just as likely to have an underlying hemostatic menstrual bleeding may be even heavier during anovulatory
defect as women 20-44 years of age. An international cycles. For this reason, HMB in women with inherited

Table 3-8 Screening tool for inherited bleeding disorders in women presenting with heavy menstrual bleeding
Screening questions Score
Q1. How many days did your period usually last, from the time bleeding began until it completely 1 = ≥ 7 days
stopped? 0 = <7 days
Q2. How often did you experience a sensation of “flooding” or “gushing” during your period? 1 = Every or most periods
0 = Never, rarely, or some periods
Q3. During your period did you ever have bleeding where you would bleed through a tampon or napkin 1 = Every or most periods
in ≤2 hours? 0 = Never, rarely, or some periods
Q4. Have you ever been treated for anemia? 1 = Yes
0 = No
Q5. Has anyone in your family ever been diagnosed with a bleeding disorder? 1 = Yes
0 = No
Q6. Have you ever had a tooth extracted or had dental surgery? 1 = Yes, if had and bled
0 = No
Q7. Have you ever had surgery other than dental surgery? See 7a. below
Q7a. Did you have bleeding problem after surgery? 1 = Yes
0 = No
Q8. Have you ever been pregnant? See 8a. below
Q8a. Have you ever had a bleeding problem after delivery or after a miscarriage? 1 = Yes
0 = No
The screen is considered positive if an affirmative response is obtained in any one of the four categories covered by the eight questions,
including (i) bleeding severity, (ii) family history of bleeding disorder, (iii) personal history of excessive bleeding, and (iv) history of treatment
for anemia. Women with a positive screen should undergo comprehensive hemostatic testing to determine whether they have a bleeding
disorder.
Scores are adapted from Philipp CS et al. Am J Obstet Gynecol. 2011;204:209.e1-209.e7; and Philipp CS et al. Am J Obstet Gynecol. 2008;198:163.
e1-163.e38.
bleeding disorders often presents at menarche and may be unwilling to use the hormonal preparations recommended
particularly troublesome during the premenopausal years. as first-line therapy.
As a result, more aggressive or combination therapy may be Combined contraceptive agents containing both estro-
required during these time periods. gen and progestin are available in oral, transdermal, and
There are few published trials regarding the management vaginal ring formulations. These agents reduce menstrual
of HMB. In 2007, the NHLBI published guidelines on the loss by inducing changes that thin the endometrium. Sev-
diagnosis, evaluation, and management of vWD. The expert eral studies have demonstrated that combined OCs
panel recommended that the first choice of therapy for HMB increase fibrinogen, prothrombin, and factor VII, and con-
should be combined OCs, with alternative choices including sequently, promote hemostasis. It is unknown whether the
the levonorgestrel intrauterine device, antifibrinolytics, or increase in coagulation factors contributes to the clinical
intranasal desmopressin (DDAVP; Stimate). For women response, but these agents do reduce menstrual blood loss
who desire pregnancy, alternative options for controlling and increase hemoglobin in women with anemia. Addi-
HMB include desmopressin, antifibrinolytics, and vWF con- tional benefits of hormonal management of HMB include
centrate. In women with type 3 vWD or other severe factor cycle regulation, decreased dysmenorrhea, and improve-
deficiencies and severe menorrhagia unresponsive to OC, ment in acne. In a trial of combined contraceptive hor-
levonorgestrel-releasing intrauterine system (LNG-IUS, mones in 14 adolescents with vWD, menstrual blood loss
Mirena), antifibrinolytics, or desmopressin, VWF concen- measured by PBAC decreased in 12 of 14 patients. Although
trates or other concentrates may be considered. In women these agents are generally well tolerated in adolescents,
without bleeding disorders, tranexamic acid, an antifibrinol- there may be hesitancy, particularly in the families of
ytic agent given the first five days of menses, reduces men- young, sexually abstinent adolescents, to use them, and
strual bleeding by 30%-55%. One study in women with time should be allotted for thorough discussion and edu-
bleeding disorders demonstrated that menstrual bleeding cation. Extended cycling or continuous regimens of OCs
was reduced with the use of either intranasal desmopressin can be particularly helpful in reducing menstrual blood
or tranexamic acid. The management of HMB in adolescents loss, especially in the setting of anemia. Breakthrough
presents some additional challenges, as the etiology is often bleeding is a possible adverse effect of these regimens,
multifactorial, and patients or parents may be reluctant or especially in adolescents.

The LNG-IUS is a progestin-impregnated intrauterine residual follicle after an ovum is extruded. In the acute set-
device that reduces menstrual blood loss by opposing the ting, surgery, tranexamic acid, and clotting factor replace-
estrogen-induced growth of the endometrium. The short- ment have been used to manage hemorrhagic ovarian cysts.
term and long-term efficacy of this device has been demon- OCs, which reduce the likelihood of ovulation and increase
strated in a small cohort of adult women with inherited clotting factors, are used to prevent recurrences (Table 3-9).
bleeding disorders in the United Kingdom. These same Even among women with bleeding disorders but without
investigators have reported on the use of LNG-IUS in five documented hemorrhagic ovarian cysts, there is a high prev-
adolescents with bleeding disorders (16-19 years of age). alence of midcycle pain or “mittelschmerz,” a phenomenon
A recent small retrospective study, however, showed that is thought to be associated with bleeding at the time of
IUDs were removed in a significant proportion of women ovulation.
with bleeding disorders because of patient dissatisfaction, Women with bleeding disorders also are diagnosed more
malposition, or expulsion. Larger prospective studies of frequently with endometriosis. In one case-control study,
complication rates and effectiveness of pre-medications endometriosis was reported in 30% of women with vWD as
(eg, nonsteroidal anti-inflammatory agents) or prolonged compared with 13% in the control group. The etiology of
use of anti-fibrinolytics in preventing these complications this phenomenon is unclear, but one hypothesis is that
are needed. Physician-patient discussions regarding this women with HMB are at higher risk of retrograde menstrual
device also require substantial time for education, as patients bleeding (reflux of menstrual blood out of the uterine cav-
often have misperceptions that these devices cannot be ity), which then stimulates the development of endometrial
removed easily once placed, can be placed only in women tissue implants in the fallopian tubes or peritoneal cavity.
who have had children, or perhaps are even limited to those Alternatively, women with bleeding disorders may not be
who have completed their planned childbearing. Patients more likely to develop endometriosis but simply are more
should also be informed of a risk of prolonged spotting likely to present with symptomatic bleeding, or they are
and increased cyst formation. Other progestin-only contra- more likely to experience hemorrhagic cysts that are mis
ceptives, such as depot medroxyprogesterone acetate diagnosed as endometriosis. Similarly, the development of
(Depo-Provera), progestin-only pills, and the Implanon fibroids, polyps, and endometrial hyperplasia may unmask a
subcutaneous implant, also reduce endometrial prolifera- previously subclinical bleeding tendency and cause prob-
tion and therefore menstrual blood loss. Insertion of the lematic bleeding, so that the diagnosis of these conditions
Implanon implant might cause bleeding in a woman with a (but likely not the true frequency) becomes more common
bleeding disorder, and the use of a pre-procedure hemostatic in women with bleeding disorders. As a result of all of these
agent should be considered. manifestations, women with bleeding disorders are more
Management: Females with HMB since menarche or a fam- likely to undergo hysterectomy than their peers and more
ily or personal history of bleeding should be screened for a likely to undergo the procedure at an earlier age.
bleeding disorder. For women with menorrhagia, combined Management: For females who develop hemorrhagic ovar-
OCs containing both estrogen and progestin, are recom- ian cysts, clotting factor replacement alone, or together with
mended first-line therapy (Table 3-9). Alternative approaches tranexamic acid, is recommended for acute management,
include the levonorgestrel-releasing intrauterine system and OCs are recommended to prevent recurrence.
or other progestin-only contraceptives, including Depo-
Provera, progestin-only pills, or subcutaneous implants.
Thrombosis and oral contraceptives in the
Non-hormonal therapies include intranasal desmopressin
premenopausal woman
and antifibrinolytics. For patients with type 3 vWD or other
severe factor deficiencies, vWF or other clotting factor con- Hormonal agents are commonly used by >100 million pre-
centrates during menses may be considered. Based on case menopausal women in the United States in the form of con-
reports and expert opinion, vWF activity (ristocetin) and traceptive agents, which are available in oral, transdermal,
FVIII level >50 IU/dL is adequate for neuraxial anesthesia. and vaginal ring formulations. The most common formula-
tion is the oral combination of estrogen and progestin,
“combined” OC. Progestin-only agents are as effective as
Hemorrhagic ovarian cysts and endometriosis
estrogen-progestin combination agents and are available in
The second most common reproductive tract bleeding man- oral, intramuscular, intrauterine, and subdermal forms.
ifestation is hemorrhagic ovarian cysts, which occur more Over 40 case-control studies, prospective cohort studies, and
commonly in women with vWD, platelet function defects, randomized trials of women using OC provide estimates of
and rare bleeding disorders than in women without bleeding the risk of VTE to be two- to three-fold greater than in non-
disorders. Ovarian cysts develop when bleeding occurs in the users; although the absolute risk is low, 2-4 per 10,000

Table 3-9 Management of menorrhagia, ovarian cysts, and pregnancy in women with von Willebrand disease
Menorrhagia and ovarian cysts in women with vWD Pregnancy in women with vWD
Women with menorrhagia or abnormal vaginal bleeding should have a Women planning pregnancy should have a preconception
full gynecologic evaluation before therapy. evaluation with a hematologist and high-risk obstetrician skilled
in management of vWD.
In an adolescent or an adult woman who does not desire pregnancy Women with vWF levels <50 IU/dL or a history of severe bleeding:
but may desire future childbearing, the treatment of choice for 1. Should be referred to a center with high-risk obstetric
menorrhagia is combined oral contraceptives. capabilities and expertise in hemostasis
Factor replacement may be required for severe acute bleeding or blood 2. Should receive prophylaxis with desmopressin or vWD
loss until the above treatment is effective. concentrate before invasive procedures
3. Should achieve vWF:RCo and FVIII levels of at least
50 IU/dL before delivery and up to 3-5 days afterward
If a woman would otherwise be a suitable candidate for an intrauterine If vWF:RCo and FVIII levels can be monitored and maintained
device, the second choice of therapy for menorrhagia is the >50 IU/dL during labor and delivery, and no other coagulation
levonorgestrel intrauterine system. defects are present, then neuraxial anesthesia may be
considered (recommendation based on case series and expert
opinion)
For a woman who desires pregnancy, treatment with factor Because coagulation factors return to pre-pregnancy levels within
replacement, desmopression, or antifibrinolytics may be tried to 14-21 days after delivery, health care providers should be in
control menorrhagia. Dilation and curettage is not usually effective close contact with a woman during the postpartum period.
in managing excessive uterine bleeding in vWD.
If all efforts to reduce menorrhagia by factor replacement or hormonal
therapies fail, then surgical intervention may be required. For
women not desiring future pregnancy, uterine ablation, or
hysterectomy. For surgical procedures, factor to maintain vWF or
VIII >50 IU/dL is recommended.
In an adolescent or adult woman with acute hemorrhagic ovarian
cysts, treatment of choice is replacement therapy.
In an adolescent or adult woman who does not desire pregnancy
but may desire future childbearing, the first choice for long-term
prevention of hemorrhagic ovarian cysts should be combined oral
contraceptives.
Adapted from James AH et al. Am J Obstet Gynecol. 2009;201:12e1-8; James AH. Thromb Res. 2009;(123)(suppl 2):S124-128; and National
Heart, Lung, and Blood Institute. The Diagnosis, Evaluation, and Management of von Willebrand Disease. Bethesda MD: National Institutes of
Health. NIH Pub. No. 08-5832;2007.
vWF:RCo = ristocetin cofactor activity.
erson-years of OC use. This 2- to 3-fold increased risk is

p ral anticoagulants, such as protein S. Evidence has now
much lower than the 5- to 10-fold increased risk seen during accumulated that the negative influence of OCs on the anti-
pregnancy and the 15-35 fold increased risk during the post- coagulant protein C pathway leads to acquired protein C
partum period. The risk of VTE is highest in the first year of resistance, and this is thought to be a primary mechanism
use, especially in the first three months. Several studies, how- underlying the prothrombotic effect of these agents. Third-
ever, indicate that the VTE risk in heterozygous carriers of generation combined OCs (desogestrel) are associated with
the FVL or prothrombin G20210A is greater than would be higher VTE risk, as are drospirenone-containing prepara-
expected if the risks were additive, that is, 28-50 per 10,000 tions (Table 3-10). Progestin-only contraceptives appear to
woman-years of OC use. In women with protein C or anti- confer lower VTE risk, but there have been no clinical trials
thrombin III deficiency, the absolute VTE risk with OC use is to confirm this. The vaginal ring and the transvaginal patch
reported to be even higher, 400 per 10,000 patient-years of both contain estrogen and are associated with an increased
OC use. risk of thrombosis relative to nonusers.
The increased risk of thrombosis has been attributed to In addition to the small absolute risk of thrombosis with
the estrogen component of contraception preparations. OCs, women may have underlying thromboembolic risk fac-
Estrogen increases procoagulants, such as factor VIII, vWF, tors that augment the contraception-related thrombosis risk.
and fibrinogen, and decreases fibrinolytic activity and natu- For example, women who have hypertension, smoke, or are

Table 3-10 Risk of VTE associated with hormonal contraceptives
Contraceptive Odds ratio 95% CI
COC 30 μg, desogestrel 7.3 3.30-10.00

COC 30 μg, levonorgestrel 3.6 1.75-4.60
Depo-Provera 3.6 0.70-1.50
Transdermal patch* 2.2 0.70-3.80
Vaginal ring* 1.6 1.02-2.37
Progestin-only pills 0.6 0.33-3.41
Levonorgestrel IUD 0.3 0.10-1.26
Adapted from Manzoli L et al. Drug Safety. 2012;35:191-205; Ueng J, Douketis JD. Hematol Oncol Clin North Am. 2010;24:683-694; van
Hylckama Vlieg A et al. BMJ. 2009;339:b2921; Lidegaard O et al. BMJ. 2012;344:e2990; van Hylckama Vlieg A, Helmerhorst FM, Rosendaal
FR. Arterioscler Thromb Vasc Biol. 2010;30:2297-2300; van Vlijmen EF et al. Blood. 2011;118:2055-2061; WHO. Medical eligibility criteria for
contraceptive use. 2008 Update. http://whqlibdoc.who.int/publications/2010/9789241563888_eng.pdf; Mantha S et al. BMJ. 2012;345:34944.
CI = confidence interval; COC = combined oral contraceptive; IUD = intrauterine device.
>35 years of age have higher risks of myocardial infarction international study reported a possible increase in stroke risk
(MI) and stroke. Diabetes and hypercholesterolemia also in women with hypertension using injectable POCs. Most
increase the risk of MI, while migraines with aura raise the recently, a study from the Netherlands reported a 3.6-fold
risk of stroke. A history of prior VTE or complicated valvular increased risk of VTE in women using Depo-Provera, as
heart disease may increase the risk of contraception-related compared with nonusers, but no increased risk in women
thrombosis. Recently, an increased risk of VTE has been using LNG-IUS. A recent meta-analysis concluded that the
associated with polycystic ovarian syndrome. use of POCs was not associated with an increased risk of VTE
Among users of hormonal contraception, obese women, compared to nonusers. Thus, with a paucity of clinical trials,
smokers, and those with inherited thrombophilia are the many controversies remain regarding optimal contraception
patients at highest risk of VTE. WHO has categorized a in women at risk for thrombosis.
large number of medical conditions according to the level Given the increase in women using artificial reproduction
of risk associated with a variety of contraceptive agents. The technology, clinicians should be aware that in addition to
four categories established by WHO range from no restric- hormonal contraceptive use and pregnancy, another poten-
tions (category 1) to unacceptable health risks (category 4). tial time of increased risk of arterial and venous thrombosis
In 2003, the WHO added a new medical condition to the list for premenopausal women is during ovarian hyper-
of risk states for which OCs are non-preferred contracep- stimulation syndrome, the primary complication of ovarian
tive choices—any known inherited thrombophilia. Women stimulation during artificial reproduction. Thrombosis
who are considered to have an unacceptable level of throm- commonly occurs during the first cycle of treatment and
boembolic risk with OCs may still be candidates for proges- typically in unusual sites such as the jugular and subclavian
tin-only contraceptives. With a few exceptions, the WHO veins. The highest risk is seen in pregnant women with ovar-
classifies all of the risk states described above as category 3 ian hyper-stimulation syndrome requiring hospitalization
or 4 with regard to OCs, but as category 1 or 2 with regard (see the section “Anticoagulation during pregnancy” in this
to progestin-only products (POC). This remains uncon- chapter for additional information).
firmed with the lack of any trial assessing safety or efficacy Management: For premenopausal females with throm-
of POCs. bophilia seeking contraception, the potential VTE risks
Studies of coagulation factors during POC use have not associated with combined OCs should be weighed against
identified clinically meaningful changes. In case-control the risk of unplanned pregnancy. In the absence of defini-
studies investigating the association between oral POC and tive trials, it is recommended that premenopausal females
VTE, the risk of VTE was not significantly greater in POC with thrombophilia avoid combined OCs. If combined
users as compared with nonusers (Table 3-10), and in a OCs are used, co-administered antithrombotic therapy (eg,
recent meta-analysis there appeared to be no increased risk warfarin or LMWH) may reduce VTE risk. These agents,
of MI with POC use. The latter study included small num- however, are not without risks and their use in this setting
bers of non-healthy POC users. Concern regarding a pos- remains unproven. Alternatively, progestin-only contra-
sible risk of VTE with POC may stem from reports that ceptives or the levonorgestrel intrauterine system may pro-
higher-dose progestins, used for other than contraception vide lower VTE risk than with combined OCs, but this
purposes, have been associated with VTE. Furthermore, an awaits clinical trials. For the present, rather than strictly

Bibliography | 77
Table 3-11 Safety of medications during pregnancy and breastfeeding
Medication Pregnancy category Breastfeeding
Oral iron: gluconate (ferlecit) B Caution in nursing

Parenteral iron: succrose (venofer) B Caution in nursing
Parenteral iron: dextran (fumoxytol, dexferrum, INFeD) C Caution in nursing
Recombinant erythropoietin (epogen) C Caution in nursing
Folic acid (dietary) A No/minimal risk
Granulocyte colony-stimulating factor (G-CSF) C Caution in nursing
Cyclosporine C Safety unknown, caution advised
Hydroxyurea D Unsafe
Low molecular weight heparin (enoxaparin) B Caution in nursing
Low molecular weight heparin (dalteparin) B Caution in nursing
Low molecular weight heparin (fondaparinux) B Caution in nursing
Corticosteroids (prednisone) D Probably safe
Intravenous immunoglobulin (IVIg) C No/minimal risk
Rho (D) immune globulin (RhoGAM) C Probably safe, caution advised
Romiplostim (Nplate) C Safety unknown, caution advised
Eltrombopag (Promacta) C Safety unknown, caution advised
Azathioprine D Possibly unsafe
Antiplatelet agent (clopidogrel [Plavix]) B Safety unknown, caution advised
Antiplatelet agents (aspirin) D Possibly unsafe
Eculizumab (Soliris) C Caution in nursing
Antithrombin III concentrate (Thrombate) B Safety unknown, caution advised
Factor (VII, VIII, IX, VWF) C Caution in nursing
Recombinant factor (VII, VIII, IX) C Safety unknown, caution advised
Vitamin K antagonists (warfarin [Coumadin]) D (mech valves); X (all other) Caution in nursing
Vitamin K (phytadione) C Safe
For each drug, see package insert for drug-specific recommendations.
Pregnancy category A = Safe for use in pregnancy. Pregnancy category B = Animal studies show no risk or adverse fetal effects but controlled
human 1st trimester studies are not available or do not confirm; no evidence of 2nd, 3rd trimester risk; fetal harm unlikely. Pregnancy
category C = Animal studies show adverse fetal effect(s) but there are no controlled human studies OR no animal or human studies
exist; weigh possible fetal risk vs. maternal benefit. Pregnancy category D = Positive evidence of human fetal risk; maternal benefit may
outweigh fetal risk in serious or life-threatening situations. Pregnancy category X = Contraindicated; positive evidence of serious fetal
abnormalities in animals, humans, or both; fetal risks outweigh maternal benefit.
contraindicating OCs in asymptomatic women with findings will evolve to improve care and offer new and better
thrombophilia, particularly those with lower-risk condi- approaches. A commitment by the hematologist to continu-
tions such as factor V Leiden, individual counseling is pre- ally update and stay abreast of new evidence will be critical to
ferred to help women make an informed decision regarding ensure optimal work and interaction by the multidisciplinary
contraception. The absolute risk of VTE is low, and all team, which will translate into the highest quality of care and
combined hormonal contraceptive agents remain extremely best outcomes for women with blood disorders.
safe, with VTE risks much lower than those seen in For reference, Table 3-11 provides some guidance regard-
pregnancy. ing safety of medications during pregnancy and breastfeed-
ing, in the order the medications appear in the text.
Conclusion
Bibliography
This chapter has summarized the most recent evidence-
based guidelines available through the NHLBI, NHF, ACOG, Abdul-Kadir R, McLintock C, Ducloy AS, El-Refaey H, England
ACCP, and WHO to provide optimal care for women with A, Federici AB, Grotegut CA, Halimeh S, Herman JH, Hofer
blood disorders in pregnancy and premenopause. Although S, James AH, Kouides PA, Paidas MJ, Peyvandi F, Winikoff
this compilation is up to date, new therapies and clinical trial R. Evaluation and management of postpartum hemorrhage:

consensus from an international expert panel. Transfusion. thromboembolism and management of anticoagulation in
2014;54:1756-1768. This is a state-of-the-art document from a women with mechanical prosthetic heart valves. Best Pract
multidisciplinary group on the evaluation and management of Res Clin Obstet Gynaecol. 2014;28:519-536. This provides a
postpartum hemorrhage based on evidence. comprehensive review of thromboprophylaxis of pregnant women,
Bates SM, Greer IA, Middeldorp S, Veenstra DL, Prabulos AM, including those with artificial valves.
Vandvik PO. VTE, thrombophilia, antithrombotic therapy, Middeldorp S. How I treat pregnancy-related venous
and pregnancy: antithrombotic therapy and prevention of thromboembolism. Blood. 2011;118:5394-5400. This study
thrombosis (9th ed.): American College of Chest Physicians reviews drawbacks with D-dimer in diagnosing VTE in pregnancy
evidence-based clinical practice guidelines. Chest. 2012;141 and radiation risks with helical computed tomography scans and
(2 suppl):e691S-e736S. This provides updated guidelines pulmonary angiography, and provides practical therapy for VTE.
for the management of pregnancy-associated thrombosis, Nichols WL, Hultin MB, James AH, Manco-Johnson MJ,
thromboembolism, and pregnancy loss. Montgomery RR, Ortel TL, Rick ME, Sadler JE, Weinstein
Bates SM. Anticoagulation and in vitro fertilization and ovarian M, Yawn BP. Von Willebrand disease (VWD): evidence-
stimulation. Hematology. 2014;379-386. This review provides the based diagnosis and management guidelines. The National
latest evidence for managing women undergoing in vitro fertilization. Heart Lung and Blood Institute (NHLBI) Expert Panel
Committee Opinion No. 580. von Willebrand disease in women. Report (USA). Haemophilia. 2008;14:171-232. This is the first
Committee on Adolescent Health Care; Committee on NHLBI evidence-based clinical guideline for the diagnosis and
Gynecologic Practice. Obstet Gynecol. 2013;122:1368-1373. management of vWD, including recommendations regarding the
Gernsheimer T, James AH, Stasi R. How I treat thrombocytopenia laboratory monitoring and management during pregnancy and the
in pregnancy. Blood. 2013;121:38-47. This review provides a postpartum period.
practical approach to thrombocytopenia in pregnancy. Rodger M. Pregnancy and venous thromboembolism: ‘TIPPS’
Goonewardene M, Shehata M, Hamad A. Anemia in pregnancy. for risk stratification. Hematology. 2014;387-392. This
Best Pract Res Clin Obstet Gynaecol. 2012;26:3-24. This review review summarizes findings of the TIPPS study, regarding the
provides an excellent review of all anemias associated with advisability of anticoagulation for thrombotic complications of
pregnancy and suggestions regarding management. pregnancy.
Healthy People 2020; US Government 2000 website. http://2020. Scully M, Thomas M, Underwood M, Watson H, Langley K,
healthypeople.gov/2020/topic/nutrition-and-weight-status/ Camilleri RS, Clark A, Creagh D, Rayment R, McDonald V,
objectives=4956. Among low-income women the risk of anemia Roy A, Evans G, McGuckin S, Ainle FN, Maclean R, Lester W,
is higher, and thus a goal in this group is reduction of maternal Nash M, Scott R, OBrien P. Thrombotic thrombocytopenic
anemia to 20% or less in the third trimester. purpura and pregnancy: presentation, management, and
James AH, Jamison MG. Bleeding events and other complications subsequent pregnancy outcomes. Blood. 2014;124:211-219. This
during pregnancy and childbirth in women with von is a prospective study of TTP cases from the UK TTP Registry,
Willebrand disease. J Thromb Haemost. 2007;5:1165-1169. This the largest pregnancy-associated TTP cohort, with management
provides data from a national inpatient database of women with suggestions.
postpartum bleeding, including transfusion requirement, length of Venous thromboembolism in women: a specific reproductive health
stay, and mortality in pregnant women with vWD. risk. ESHRE Capri Workshop Group. Hum Reprod Update.
James AH, Kouides PA, Abdul-Kadir R, Edlund M, Federici AB, 2013;19:471-482. This report of the ESHRE Capri Workshop Group
Halimeh S, Kamphuisen PW, Konkle BA, Martinez-Pèrez, summarizes the VTE risks associated with infertility, combined
McLintock C, Peyvandi F, Winikoff R. Von Willebrand disease hormonal contraception, and hormone replacement therapy, based
and other bleeding disorders in women: consensus on diagnosis on Medline and other database reviews.
and management from an international expert panel. Am J Yawn BP, Buchanan GR, Afenyi-Annan AN, Ballas SK, Hassell
Obstet Gynecol. 2009;201:12 e1-12 e8. This international panel KL, James AH, Jordan LK, Lanzkron SM, Lottenberg R,
recommended that an inherited bleeding disorder should be Savage WJ, Tanabe PJ, Ware RE, Murad MH, Goldsmith JC,
considered in any female with heavy menstrual bleeding since Ortiz E, Fulwood R, Horton A, John-Sowah J. Management
menarche or a personal or family history of bleeding. of Sickle Cell Disease: Summary of the 2014 evidence-based
James AH, Konkle BA, Kouides P, Ragni MV, Thames B, Gupta S, report by expert panel members. JAMA. 2014;312:1033-1048.
Sood S, Fletcher SK, Philipp CS. Postpartum von Willebrand This special communication provides the 2014 expert panel
factor levels in women with and without von Willebrand disease evidence-based on review of MEDLINE, EMBASE, Cochran,
and implications for prophylaxis. Haemophilia. 2015;21:81-87. CINAHL, TOXLINE, and Scopus databases to address key
This study shows the rapid decline in coagulation levels 3-7 days questions regarding management of sickle cell disease, including
postpartum, and the potential delay up to 3 weeks until pre- stroke prevention, vaso-occlusive crisis, proliferative retinopathy,
pregnancy levels are attained. pulmonary hypertension, use of hydroxyurea, iron chelation, and
McLintock C. Thromboembolism in pregnancy: challenges and hemoglobin targets for preoperative transfusion as well as for long-
controversies in the prevention of pregnancy-associated venous term transfusion therapy.

CHAPTER
4
Hematopoietic growth factors
Aaron T. Gerds and Alan E. Lichtin
Introduction, 79 Erythroid growth factors, 85 Bibliography, 95
Myeloid growth factors, 79 Platelet growth factors, 92
Introduction The biological effects of G-CSF are mediated through the

G-CSF receptor expressed on both mature neutrophils and
The hematopoietic growth factors (HGFs) and their recep- neutrophil progenitors. G-CSF knockouts in mice with a tar-
tors play essential roles in regulating hematopoiesis. Spe- geted disruption of the G-CSF receptor develop severe neu-
cific factors for each hematopoietic lineage are critical for tropenia, whereas hematocrit levels and platelet counts are
producing and maintaining normal circulating levels of normal. Children with severe congenital neutropenia pro-
the cells. Granulocyte colony-stimulating factor (G-CSF) gressing to myelodysplasia or acute myeloid leukemia (AML)
regulates neutrophil production; granulocyte-macrophage often have acquired mutations in the G-CSF receptor, most
colony-stimulating factor (GM-CSF) enhances production of which consist of truncation of the cytoplasmic tail of the
of neutrophils, monocytes, and eosinophils; erythropoietin receptor (see Chapter 15, “Myeloid disorders and congenital
(EPO) regulates red blood cell production; and thrombopoi- marrow failure syndromes”).
etin (TPO) controls platelet production. This chapter focuses Available recombinant forms of G-CSF include filgrastim
on the results of clinical trials and approved uses for these produced in Escherichia coli by the introduction of the human
HGFs and provides a glimpse of other factors involved in the G-CSF gene. This form is identical to native human G-CSF
early stages in development. except for the addition of an amino-terminal methionine.
Filgrastim is licensed for use in the United States and in many
other countries (Table 4-1). An alternative non-glycosylated
Myeloid growth factors recombinant methionyl form of G-CSF, tbo-filgrastim, rec
ently garnered US Food and Drug Administration (FDA)
Granulocyte colony-stimulating factor approval (Table 4-2). Lenograstim is a glycosylated form of
(filgrastim, tbo-filgrastim, and lenograstim) G-CSF produced in a mammalian cell line and is not
G-CSF is a myeloid growth factor produced by monocytes, approved for clinical use in the United States.
macrophages, fibroblasts, endothelial cells, and a number of
other types of cells. G-CSF plays the central role of regulating Pegylated methionyl G-CSF (pegfilgrastim)
neutrophil formation and deployment. In healthy individu-
als, circulating levels of G-CSF are low or undetectable. A Pegfilgrastim is methionyl G-CSF with polyethylene glycol
dramatic increase in the circulating levels of G-CSF occur in covalently bound to the amino terminal methionine resi-
the setting of infection and inflammation and with the due. Importantly, pegylation reduces the renal clearance of
administration of endotoxin or mediators of inflammation, G-CSF through stearic hindrance and prolongs its circula-
such as interleukin-1 (IL-1) and tumor necrosis factor tion and the duration of its effects. Clinical trials compar-
(TNF). ing pegylated G-CSF and G-CSF demonstrated similar
biological activities and clinical benefits, including the
duration of chemotherapy-induced severe neutropenia and
Conflict-of-interest disclosure: Dr. Gerds declares no competing
financial interest. Dr. Lichtin declares no competing financial interest.
occurrence of febrile neutropenia. The pharmacokinetics
Off-label drug use: Dr. Gerds: Filgrastim in HIV. Dr. Lichtin: Epoetin of pegfilgrastim should not be affected by hepatic insuffi-
alfa and darbepoetin alfa in myelodysplastic syndromes. ciency, but it has not been evaluated adequately in this
| 79

80 | Hematopoietic growth factors
Table 4-1 FDA-approved indications for filgrastim Table 4-3 FDA-approved indication for pegfilgrastim
Accelerate neutrophil recovery in patients receiving myelosuppressive Decrease the incidence of infection as manifested by febrile
chemotherapy neutropenia in patients with nonmyeloid malignancies receiving
Accelerate neutrophil recovery after acute myeloid leukemia induction myelosuppressive anticancer drugs associated with a clinically
or consolidation chemotherapy significant incidence of febrile neutropenia
Accelerate neutrophil recovery in patients following a bone marrow
transplant
Mobilize peripheral blood stem cells substitution at position 23. Only sargramostim is approved
Severe chronic neutropenia (idiopathic, cyclic, congenital)
for clinical use by the US Food and Drug Administration
(FDA; see Table 4-4).
setting. Although less studied in children, the efficacy and Clinical use of G-CSF and GM-CSF
safety of pegfilgrastim appears similar to that in adults. The Prevention of chemotherapy-induced
FDA-approved indications for pegfilgrastim are shown in febrile neutropenia
Table 4-3.
The main clinical use of G-CSF and GM-CSF is for the pre-
vention of febrile neutropenia (temperature >38.3°C with
Granulocyte-macrophage colony-stimulating neutrophils less than 0.5 × 109/L) in patients receiving cancer
factor (sargramostim, molgramostim) chemotherapy. Febrile neutropenia represents the major
GM-CSF is a glycoprotein constitutively produced by mon dose-limiting toxicity of cancer chemotherapy and is associ-
ocytes, macrophages, endothelial cells, and fibroblasts. ated with considerable morbidity, mortality, and costs. The
GM-CSF production is enhanced by inflammatory cyto clinical use of G-CSF is based on results of numerous ran-
kines such as IL-1 or TNF. GM-CSF promotes the growth of domized controlled trials and meta-analyses of such trials
myeloid colony-forming cells (CFU-GM), increases the and supported by clinical practice guidelines. FDA approval
number of circulating neutrophils and monocytes, and of G-CSF for prevention of febrile neutropenia was based on
enhances the phagocytic function and microbicidal capacity two pivotal randomized controlled trials in patients with
of mature myeloid cells. GM-CSF also stimulates dendritic small-cell lung cancer receiving intensive combination che-
cell maturation, proliferation, and function, and it increases motherapy associated with prolonged severe neutropenia
antigen presentation by macrophages and dendritic cells. with a high risk of febrile neutropenia. Primary prophylaxis
That GM-CSF is not essential for hematopoiesis is confirmed with G-CSF initiated within the first 3 days after chemother-
by the demonstration of normal complete blood counts and apy and continued for up to 10 days reduced the duration of
normal number of marrow progenitor cells in GM-CSF severe neutropenia to about 3 days and reduced the occur-
knockout mice. Evidence exists, however, that GM-CSF rence of febrile neutropenia and documented infection by
plays a key role in the function of pulmonary macrophages. 50%. A pivotal randomized trial in patients with breast can-
Mice that lack GM-CSF have lung pathology consistent with cer found tbo-filgrastim to be superior to placebo, and
pulmonary alveolar proteinosis. Similarly, some cases of equivalent to filgrastim, in duration of severe neutropenia
human pulmonary alveolar proteinosis are related to a defect after chemotherapy.
in the common b-chain of the receptor for GM-CSF, inter- The results of these trials have been confirmed in multiple
leukin 3 (IL-3), and interleukin 5 (IL-5). Infants that are so other randomized controlled trials across a spectrum of
affected have decreased alveolar macrophage function and malignancies and chemotherapy regimens, consistently
accumulate surfactant in the alveoli.
Recombinant forms of GM-CSF available for clinical use Table 4-4 FDA-approved indications for GM-CSF sargramostim
include sargramostim derived from yeast and molgramostim
Reduce the risk of death due to infection in patients ≥55 years old
expressed by E. coli. The sequence of sargramostim differs
undergoing induction chemotherapy for AML
from that of native GM-CSF by a single amino acid Mobilize autologous peripheral blood stem cells and enhance
neutrophil recovery after transplantation
Promote neutrophil recovery after autologous or allogeneic bone
Table 4-2 FDA-approved indication for tbo-filgrastim
marrow transplantation
Reduction in the duration of severe neutropenia in patients with Improve neutrophil production in patients with delayed engraftment
non-myeloid malignancies receiving myelosuppressive anti-cancer or graft failure after autologous or allogeneic bone marrow
drugs associated with a clinically significant incidence of febrile transplantation
neutropenia
AML = acute myeloid leukemia.

Myeloid growth factors | 81
demonstrating a reduction in the risk of febrile neutropenia Clinical guidelines for the use of the myeloid
in the initial cycle as well as across repeated cycles of treat- growth factors
ment. At the same time, little or no benefit from G-CSF
The American Society of Clinical Oncology (ASCO), the
administration has been observed when treatment is delayed
National Comprehensive Cancer Network (NCCN), and
until neutropenia is already present. Although individual
other organizations have developed guidelines for the use of
studies were not sufficiently powered to assess any impact on
myeloid growth factors to prevent FN. In brief, current
infection-related or all-cause mortality, meta-analyses of
ASCO guidelines (Table 4-5) include the following:
these trials have demonstrated a significant reduction in
these complications with primary G-CSF prophylaxis in 1. Primary prophylaxis is recommended for patients at high
patients receiving conventional chemotherapy. These analy- risk (>20%) of FN due to age, medical history, disease
ses also have demonstrated that G-CSF prophylaxis enables a characteristics, or the myelotoxicity of the chemotherapy
greater percentage of patients to receive full-dose chemo- regimen.
therapy on schedule through the avoidance of neutropenic 2. Primary prophylaxis should be given with “dose-dense”
complications that lead to preemptive dose reductions or chemotherapy regimens.
treatment delays. Meta-analyses of randomized controlled 3. Secondary prophylaxis after a neutropenia-related event
trials also suggest that G-CSF support of patients receiving has occurred generally is recommended if reduced dosing
cancer chemotherapy may improve long-term outcomes, or dose intensity will compromise disease-free or overall
including survival, presumably most notably in patients survival or expected treatment outcome.
treated with curative intent.
Specific factors predisposing to FN and serving as current
Pegfilgrastim for prevention of febrile neutropenia indications to consider the use of a myeloid growth factors
are listed in Table 4-6.
A randomized phase 3 double-blind, placebo-controlled
clinical trial of primary prophylaxis with pegfilgrastim was
conducted in patients with breast cancer receiving doce Febrile neutropenia
taxel 100 mg/m2 every 3 weeks to determine the efficacy of All patients with FN should be treated empirically with anti-
pegfilgrastim when given with less myelosuppressive regi-
biotics after a thorough physical examination directed at
mens. Patients were randomly assigned to peg filgrastim
identifying a site of infection and after appropriate cultures
6 mg or placebo on the day following chemotherapy. Patients
are obtained. A number of studies have addressed whether
in the pegfilgrastim arm experienced significantly lower
patients with FN benefit from initiation of a myeloid growth
incidence of febrile neutropenia (FN) (1% vs. 17%), hospi-
factor in addition to broad-spectrum antibiotics. A meta-
talizations (1% vs. 14%) and anti-infective use (2% vs. 10%)
analysis of 13 randomized clinical trials compared the use of
(all P <0.001). Pegfilgrastim is FDA approved to reduce the
G-CSF or GM-CSF plus antibiotics with the use of antibiotics
risk of FN in patients undergoing chemotherapy with a
alone in patients with chemotherapy-induced FN. The meta-
17% or greater risk of FN without growth factor support
analysis showed that the use of a myeloid growth factor accel-
(Table 4-3).
erated the time to neutrophil recovery and shortened hospital
On the basis of the prolonged half-life of pegfilgrastim, it
stay but did not affect overall survival. ASCO guidelines rec-
has been recommended that chemotherapy not be given
ommend that the myeloid growth factors should not be used
sooner than 14 days after a dose of pegfilgrastim. Consider-
routinely as adjuncts to antibiotics for patients with FN.
able experience with pegfilgrastim in support of every
These guidelines recommend that the myeloid growth factors
2-week chemotherapy schedules, however, has demon-
should be considered for patients expected to have prolonged
strated acceptable efficacy and safety. Otherwise, the safety
(>10 days) and profound neutropenia (<0.1 × 109/L); use
profile of pegfilgrastim is similar to that of other forms of also should be considered for those >65 years old with pneu-
G-CSF. monia, hypotension, invasive fungal infections, or sepsis.
GM-CSF for prevention of FN

Acute myeloid leukemia
GM-CSF is approved to reduce the risk of death from infec-
tions in patients ≥55 years old undergoing induction therapy Neutropenia, anemia, and thrombocytopenia are common
for AML (Table 4-4). There is limited evidence from ran- presenting features of AML and also are important compli-
domized trials for the use of GM-CSF in non-myeloid malig- cations in its treatment. There are many studies of the
nancies, and it is not FDA approved for the prevention of FN myeloid growth factors to sensitize leukemic cells to
in this population. increase the effectiveness of chemotherapy and to prevent

Table 4-5 American Society of Clinical Oncology guidelines for use of myeloid growth factors to prevent FN
Setting/indication Recommended Not recommended

General circumstances FN risk in the range of 20% or higher
Special circumstances Clinical factors dictate use
Secondary prophylaxis Based on chemotherapy reaction among other factors
Therapy of afebrile neutropenia Routine use
Therapy of febrile neutropenia If high-risk for complications or poor clinical outcomes Routine use
AML Following induction therapy, patients >55 years old Priming prior to cytotoxic chemotherapy outside
most likely to benefit a clinical trial
After the completion of consolidation chemotherapy
MDS Routine use in neutropenic patients
Acute lymphocytic leukemia After the completion of initial chemotherapy or first
post remission course
Radiotherapy Consider if receiving radiation therapy alone and Patients receiving concurrent chemotherapy and
prolonged delays are expected radiation
Older patients If ≥65 years old with diffuse aggressive NHL and treated
with curative chemotherapy
Pediatric population Primary prophylaxis of pediatric patients with a G-CSF use in children with ALL
likelihood of FN and the secondary prophylaxis or
therapy for high-risk patients.
Source: Smith TJ, Khatcheressian J, Lyman GH, et al. J Clin Oncol. 2006;24:3187–3205.
ALL = acute lymphocytic leukemia; AML = acute myeloid leukemia; FN = febrile neutropenia; G-CSF = granulocyte colony-stimulating factor;
MDS = myelodysplastic syndrome; NHL = non-Hodgkin lymphoma.
infectious complications. Although G-CSF and GM-CSF neutropenia with an associated decrease in infections requir-
may shorten the duration of neutropenia during the inducing antibiotics with G-CSF therapy. No consistent favorable
tion phase of chemotherapy, neither consistently reduce the or detrimental impact of G-CSF or GM-CSF on treatment
occurrence of FN, infections, or the duration of hospitaliza- response and survival has been observed.
tion. Results for sensitization of the leukemic cells to chemo-
therapy also are inconsistent, and use of the myeloid growth
Acute lymphoblastic leukemia
factors in this way is not recommended except for research
studies. Neutropenia is a common consequence of treatment in
During the consolidation phase of treatment, the marrow patients with acute lymphoblastic leukemia (ALL). Eight
is more responsive, and two large randomized trials have randomized controlled trials, including more than 700
demonstrated significant decreases in the duration of severe adults and children, demonstrated that neutrophil recovery
is accelerated with myeloid growth factor therapy, mostly
utilizing G-CSF. No consistent therapeutic benefits in reduc-
Table 4-6 Risk factors for chemotherapy-associated neutropenia and ing infections, shortening hospitalizations, or improving the
its complications overall treatment outcomes were observed.
Age greater than 65 years
Previous chemotherapy or radiation therapy Mobilization of autologous peripheral blood stem cells
Bone marrow involvement of tumor and enhancement of neutrophil recovery after
Preexisting neutropenia, infections, open wounds, or recent surgery
autologous transplantation
Poor performance status
Decreased renal function Autologous peripheral blood stem cells are collected rou-
Decreased liver function, particularly increased bilirubin level tinely from cancer patients by leukapheresis after cytoreduc-
Adapted from Crawford J, Armitage J, Balducci L, et al. J Natl Comp tive chemotherapy or after cytoreductive chemotherapy
Cancer Netwk. 2013;11:1266–1290. followed by G-CSF or GM-CSF. Mobilization with G-CSF

Myeloid growth factors | 83
has been demonstrated to involve several steps. First, G-CSF accelerates neutrophil recovery by approximately 4-5 days. A
markedly enhances neutrophil production. G-CSF adminis- meta-analysis of 18 clinical trials totaling 1,198 patients
tration also releases neutrophil elastase and cathepsin G showed no change in the risk of acute or chronic GVHD af
from the granules of the developing marrow neutrophils. ter allogeneic stem cell transplantation with either GM-CSF
When released, these proteases cleave adhesion molecules or G-CSF when compared with patients who did not receive
expressed on the surfaces of the marrow stromal cells. Cleav- a myeloid growth factor.
age of the bond of chemokine receptor-4 (CXCR 4), Umbilical cord blood transplants have been able to extend
expressed on hematopoietic progenitor cells, and its ligand the benefits of allogeneic transplant to those without a matched
chemokine ligand 12 (CXCL12, also known as stromal donor. As a result of the size and composition of the graft,
derived factor 1 or SDF 1), expressed on marrow stromal hematopoietic recovery is prolonged, and recipients are at a
cells, is thought to be the principal mechanism for progeni- higher risk for infectious complications. In retrospective stud-
tor cell release into the circulation. ies, the use of G-CSF reduced the time to neutrophil recovery
As discussed in Chapters 13 and 14, transplantation of by approximately 10 days. Although prospective data is lack-
autologous peripheral blood stem cells results in the resto- ing, G-CSF is routinely used after cord blood transplant.
ration of hematopoiesis after high-dose (myeloablative)
chemotherapy. Clinical trials of autologous peripheral Improvement of neutrophil production in patients with
blood stem cell transplantation have shown that the use of delayed engraftment or graft failure after bone marrow
a myeloid cytokine after stem cell infusion accelerates neu- transplantation
trophil recovery by 2-4 days. However, neutrophil recovery
to >0.5 × 109/L is so rapid (median 11-14 days) without a Patients who do not achieve a neutrophil count of 0.1 × 109/L
myeloid growth factor that it has been difficult to demon- by day 21 after transplantation or whose neutrophil count
strate a meaningful clinical benefit of G-CSF or GM-CSF, drops below 0.5 × 109/L following engraftment in the absence
including reduced risk of sepsis or death due to infection in of relapse often respond to a myeloid growth factor with
patients receiving a peripheral blood stem cell product. improvement in neutrophil production.
Therefore, consensus on their use in this setting is lacking.
A few randomized studies have found no difference in Severe chronic neutropenia (idiopathic,
safety of pegfilgrastim as compared to filgrastim in this set- cyclic, congenital)
ting. Plerixafor, a CXCR4 antagonist, acts synergistically
with G-CSF to yield greater numbers of CD34+ stem cells Severe chronic neutropenia is a heterogeneous group of
and is FDA approved as an adjunct to G-CSF for stem cell inherited and acquired disorders characterized by a persis-
mobilization in certain conditions, particularly for patients tent neutrophil count of <0.5 × 109/L and recurrent bacterial
who are expected to mobilize poorly with G-CSF alone. infections including Kostmann syndrome, sporadic and
autosomal-dominant severe congenital neutropenia, and
cyclic neutropenia (see Chapter 15).
Mobilization of peripheral blood stem cells from Most patients with congenital and cyclic neutropenia
normal donors for allogeneic transplantation respond well to treatment with G-CSF. Treatment signifi-
cantly improves neutrophil counts, dramatically decreases
G-CSF treatment of normal donors effectively mobilizes the incidence and severity of bacterial infections, and appears
stem cells for use in subsequent allogeneic transplantation to improve survival. Responses can be maintained over
and has an excellent safety profile. many years with daily or alternate day G-CSF. Patients with
cyclic neutropenia maintained on G-CSF continue to have
regular fluctuations in the neutrophil count, but the depth of
Acceleration of neutrophil recovery after bone marrow
the nadir is reduced and lasts for fewer days. Patients with
and umbilical cord blood transplantation
severe congenital neutropenia attributable to mutations in
Peripheral blood stem cells are preferred over bone marrow ELANE, HAX1, or WAS or as yet unknown mutations are at
in some instances because of the ease of collection of periph- risk of developing AML. The lifetime risk is estimated to be
eral blood stem cells, lower risk of primary graft failure, and as high as 30%. In contrast, there is no apparent risk of AML
a more rapid neutrophil and platelet recovery. Nonetheless, in patients with cyclic neutropenia.
bone marrow is preferred for many recipients as the risk of The Severe Chronic Neutropenia International Registry
graft-versus-host disease (GVHD) is lower. When bone is a useful source for additional information about the diag-
marrow transplantation is performed, a myeloid growth nosis and treatment of severe chronic neutropenia (http://
factor after bone marrow stem cell infusion significantly depts.washington.edu/registry/).

Myelodysplasia Diabetic foot infections
The myelodysplastic syndromes (MDS), also discussed in The role of myeloid growth factors for the treatment of dia-
Chapter 17, are a group of acquired neoplastic hematopoi- betic foot infections is unclear. A meta-analysis summarized
etic stem cell disorders with the hallmark of ineffective the potential benefits of G-CSF as an adjunctive therapy. On
hematopoiesis. Both quantitative and qualitative defects in the basis of an analysis of five trials with a total of 167
neutrophils impair the ability to ward off bacterial infection patients, this review showed that G-CSF did not significantly
in these patients. A handful of clinical trials have investigated affect the likelihood of resolution of the infection or wound
treatment of myelodysplasia with the HGFs. Treatment with healing, but its use was associated with significantly reduced
G-CSF or GM-CSF can normalize the neutrophil count in likelihood of lower extremity surgical interventions, includ-
patients with myelodysplasia, but whether this translates ing amputation. G-CSF treatment appears to reduce the
into reduced mortality from bacterial or fungal infection duration of hospital stay but not the duration of systemic
is less clear. A randomized, phase 3 trial of 102 patients antibiotic treatment. The evidence suggests benefit, but it is
with high-risk MDS did not demonstrate a reduction in unclear exactly which patients will be helped by adjunctive
infectious complication but suggested an increase in non- G-CSF.
leukemic disease-related deaths associated with the routine
use of G-CSF to increase neutrophil counts. However, in Pneumonia
low-risk MDS, G-CSF or GM-CSF may enhance the effects
of erythropoietin in the treatment of MDS-related anemia. A number of clinical trials have explored the use of G-CSF in
There is no convincing evidence at present that growth fac- non-neutropenic adults with community-acquired pneu-
tor therapy accelerates progression from low-risk myelodys- monia or hospital-acquired pneumonia. In an evidence-
plasia to AML. based review, six studies with a total of 1,984 people were
identified. G-CSF use appeared to be safe, with no increase in
the incidence of serious adverse events. The use of G-CSF,
Other potential clinical uses of G-CSF
however, was not associated with improvement in mortality
HIV
at 28 days.
Neutropenia is common in HIV infection and is found in
5%-10% and 50%-70% of patients with early and advanced Myocardial infarction
disease, respectively. Furthermore, medications used in the
management of HIV, associated opportunistic infections, Studies have suggested that stem cells mobilized from the
and malignancies can lead to neutropenia. Treatment with marrow by myeloid growth factor may improve cardiac
G-CSF promptly increases the neutrophil count to the nor- function following myocardial infarction, presumably by
mal range in most patients. A large multicenter trial that ran- stimulating angiogenesis. However, a recent meta-analysis of
domized 258 HIV-positive patients with a low CD4 count seven trials involving 354 patients who received myeloid
(0.2 × 109/L) and absolute neutrophil count (ANC) 0.75-1.0 × growth factor or placebo for 4-6 days after acute myocardial
109/L showed that G-CSF-treated patients (dose adjusted to infarction found no difference in mortality and no improve-
increase the ANC to 2.0-10.0 × 109/L) had fewer bacterial ment in parameters of left ventricular function.
infections, less antibiotic use, and fewer hospital days, but no In one small prospective clinical study, G-CSF therapy
change in viral load, in comparison with the control group. with intracoronary infusion of peripheral blood stem cells
showed improved cardiac function and promoted angiogen-
esis in patients with myocardial infarction. Aggravation of
Leukapheresis
in-stent restenosis led to early termination of the study.
Large numbers of neutrophils can be collected by leukapher- Although studies such as these are intriguing for the utiliza-
esis from normal donors pretreated with G-CSF plus dexa- tion of G-CSF-mobilized stem cells for a variety of new
methasone, and these neutrophils exhibit normal function applications, no conclusive evidence exists at present sup-
in vitro. Transfusion of G-CSF-stimulated neutrophil leuka- porting these applications.
pheresis products into severely neutropenic leukemia
patients or stem cell transplant recipients can transiently
Side effects of G-CSF
raise the peripheral neutrophil count to the normal range
(<2.0 × 109/L). Whether neutrophil transfusions will increase The major side effect of G-CSF is bone pain in the hips,
survival in patients with profound sustained neutropenia which usually coincides with marrow recovery and may be
who have an active bacterial or fungal infection is under due to the expansion of hematopoiesis within the marrow
investigation. cavity. Medullary bone pain occurs in approximately 30% of

Erythroid growth factors | 85
patients treated with G-CSF, and osteoporosis has been the parent molecules. Several new products with a prolonged
observed in some patients who were administered G-CSF. duration of their stimulatory effects, similar to pegylated
Other side effects of G-CSF include headache and fatigue. G-CSF, are in development. A key issue is whether or not the
G-CSF should not be used in patients with sickle cell disease; new molecules are immunogenic. The development of anti-
case reports document the precipitation of sickling and bodies to a growth factor can be hazardous, as they can block
severe pain crisis in these individuals. Other rare side effects the activity of the administered drug and also can neutralize
include splenic rupture and adult respiratory distress the effects of the naturally produced, endogenous growth
syndrome. factors, thus worsening neutropenia.
The number of laboratories and biopharmaceutical com-
panies producing myeloid growth factors is also rapidly
Side effects of GM-CSF
increasing. Their products are molecularly similar to the
The major side effect of GM-CSF is a flu-like illness charac- approved products and are called “biosimilars.” Testing and
terized by fever (22% of patients) and myalgias and arthral- introduction of biosimilars is proceeding rapidly with the
gias (15%). A fraction of patients treated with GM-CSF first application for a biosimilar filgrastim being accepted by
experience fluid retention (8%) or dyspnea (13%). GM-CSF the FDA in mid-2014. As of note, the recently approved tbo-
should not be used concurrently with chemo-radiotherapy. filgrastim is not technically considered a filgrastim biosimi-
A case report detailed the abrupt onset of sickle cell pain cri- lar because it was filed as a Biologics License Application
sis in a patient who received GM-CSF injections around a since a biosimilars approval pathway had not been estab-
chronic leg ulcer. lished at the time of FDA submission.
Risk of leukemia with G-CSF and GM-CSF

Erythroid growth factors
Concerns have been expressed that G-CSF and GM-CSF
Erythropoietin
might cause leukemia as they are known to stimulate prolif-
eration of leukemic blasts. At present there is no convincing EPO is the principal HGF that regulates red blood cell pro-
evidence that treatment outcomes for AML are worsened by duction. The liver is the primary site of EPO production
myeloid growth factor treatments used in conjunction with during fetal development. In adults, EPO is produced pre-
appropriate chemotherapy. In patients receiving myelotoxic dominantly in the kidney, with a small amount produced in
chemotherapy agents for other types of cancer, there is a sig- the liver. Renal EPO production is under the control of an
nificant risk of secondary leukemias. This risk probably is oxygen-sensing mechanism involving transcriptional regu-
related directly to specific leukemogenic chemotherapy lation by hypoxia-inducible factor (HIF). HIF signaling and
agents and regimens. Recent analysis of data from random- local EPO production in osteoblasts in the hematopoietic
ized trials suggests that the risk of AML may be increased in stem cell niche have been reported. Plasma EPO levels are
those receiving chemotherapy supported by the myeloid measurable by a clinically available enzyme-linked immuno-
growth factors, but the interpretation of the results is made sorbent assay. In some patients with nonrenal anemia, the
difficult by the observation that myeloid growth factor– degree of plasma EPO elevation may assist in predicting
treated patients usually receive larger doses and longer response likelihood to recombinant human EPO (rhEPO)
courses of chemotherapy. The long-term risk of leukemia is therapy.
also of importance to normal stem cell donors, but little data EPO exerts its erythropoietic action by binding to its spe-
exists regarding donors mobilized with myeloid growth fac- cific high-affinity cell surface receptor (EPOR) expressed on
tors. It is estimated that it will require the observation of a erythroid progenitor and precursor cells in the bone mar-
little over 2,000 donors for a minimum of 10 years to detect row. EPOR does not possess intrinsic tyrosine kinase enzy-
a 10-fold increase in the incidence of leukemia. However, it matic activity. Its intracellular domain associates with a
is important to note that patients with idiopathic or cyclic cytoplasmic tyrosine kinase Janus kinase 2 (JAK2) to activate
neutropenia have received G-CSF for many years without downstream signaling that promotes the proliferation, sur-
progression to leukemia. vival, and differentiation of erythroid cells. Low levels of
EPOR expression have been found in neural tissues, endo-
thelial cells, and other nonhematopoietic cell types. Targeted
New formulations of G-CSF and GM-CSF
disruption of the genes encoding either EPO or EPOR in
Because of the potency and effectiveness of G-CSF and GM- mice leads to severe in utero anemia and embryonic death.
CSF, there have been many efforts to identify additional Cardiovascular and neural abnormalities also have been
myeloid growth factors and to make new derivatives from reported. These mice exhibit normal formation of early and

late erythroid progenitors, burst-forming unit-erythroid Table 4-7 FDA-approved indications for epoetin alfa
(BFU-e) and colony-forming unit-erythroid (CFU-e), indi- Anemia due to:
cating that commitment to erythroid lineage does not Chronic kidney disease in patients on dialysis and not on dialysis
require EPO but rather that terminal differentiation of The effects of concomitant myelosuppressive chemotherapy, and
CFU-e into mature red blood cells depends on intact EPOR upon initiation, there is a minimum of an additional 2 months of
signaling. planned chemotherapy
Naturally occurring, dominant gain-of-function EPOR Zidovudine in HIV-infected patients, and for reduction of allogeneic
gene mutations that disrupt down-regulation of JAK2 acti- red cell transfusions in patients undergoing elective, noncardiac,
nonvascular surgery
vation are associated with primary familial and congenital
polycythemia. An acquired, somatic, activating JAK2 V617F
mutation is encountered in 95% of polycythemia vera cases prolonged up to once every 4 weeks to maintain hemoglobin
and in about 50% of patients with other BCR-ABL1-negative levels in CKD patients on dialysis. CERA is approved for use
myeloproliferative neoplasms. Mutations in the genes in some European countries as well as by the FDA, but it is
encoding HIF, von Hippel-Lindau (VHL) proteins, and pro- not available for distribution in the United States because of
lyl hydroxylase domain (PHD) enzymes that regulate renal patent-related legal issues.
oxygen sensing and EPO production are found in some
patients with secondary familial and congenital polycythe-
FDA-approved clinical uses of rhEPO
mia due to inappropriate elevation of plasma EPO levels.
Chronic kidney disease
Normocytic, normochromic anemia associated with EPO

Recombinant human erythropoietins
deficiency occurs in the majority of patients with CKD dur-
Three main recombinant human erythropoietin (rhEPO) ing progression to end-stage renal disease. Anemia contrib-
preparations—epoetin alfa, epoetin beta, and darbepoetin utes to CKD-related symptoms and has been associated with
alfa—are available for clinical use in the United States and the presence of cardiac comorbidities, reduced quality of life,
Europe. The biologic activity and adverse effect profile of and increased risk of mortality. In patients with anemia due
these agents are comparable. The difference in the amount of to CKD, rhEPO therapy improves hemoglobin levels and
posttranslational glycosylation of each product modulates eliminates transfusion requirements; however, studies have
the pharmacokinetic properties. These agents are produced shown that targeting and maintaining near-normal or nor-
by recombinant DNA technology, by a mammalian cell line mal hemoglobin levels is associated with increased morbid-
into which the EPO gene has been introduced. Biosimilar ity and mortality risk.
products (“follow-on biologics”) for epoetins have been Following a safety review in 2011, the FDA mandated
available in some countries as the patent and exclusivity changes to the drug labels for epoetin alfa and darbepoetin
rights have expired. alfa warning that in controlled trials, patients experienced
Epoetin alfa was the first recombinant product approved greater risks for death, serious adverse cardiovascular reac-
by the FDA in 1989 for its indication in chronic kidney dis- tions, and stroke when they were administered rhEPO to tar-
ease (CKD) patients, followed by its approval in 1993 in the get a hemoglobin level >11 g/dL. It was noted that no trial
oncology supportive care setting for chemotherapy-induced has identified a hemoglobin target level, rhEPO dose, or dos-
anemia (Table 4-6). ing strategy that does not increase these risks.
Epoetin beta is available for clinical use in Europe. Darbe- Effective June 24, 2011, the FDA safety announcement
poetin alfa is a hyperglycosylated form and binds to the same indicated the following:
cellular receptor. The modification of two additional
N-linked oligosaccharide chains compared with EPO leads • Consider starting rhEPO treatment when hemoglobin
to a higher molecular weight than EPO and a threefold lon- level is <10 g/dL, without specifying how far below 10 g/dL
ger half-life in vivo. The advantage is that it can be adminis- is appropriate for an individual to initiate therapy. It is rec-
tered less frequently than epoetin alfa or epoetin beta to ommended to individualize dosing and use the lowest dose
achieve a comparable increment in hemoglobin. Darbepoe- sufficient to reduce the need for red blood cell transfu-
tin alfa was approved by the FDA for clinical use in 2001 sions. A target hemoglobin level is not specified.
(Table 4-7). • For patients with CKD not on dialysis, consider initiating
Continuous erythropoietin receptor activator (CERA) is a rhEPO treatment only when hemoglobin level is <10 g/dL
structurally distinct pegylated epoetin beta product contain- and if the rate of hemoglobin decline indicates the likeli-
ing a methoxy-polyethylene glycol polymer. This modifica- hood of requiring a red blood cell transfusion and reduc-
tion extends its half-life, allowing the dosing intervals to be ing allo-immunization or other transfusion-related risks is

a goal. If the hemoglobin level exceeds 10 g/dL, reduce or during the course of their illness. In this clinical setting,
interrupt rhEPO dose and use the lowest dose sufficient to epoetin alfa and darbepoetin alfa exhibit efficacy to increase
reduce the need for transfusions. hemoglobin and to reduce the requirement for red blood
• For patients with CKD on dialysis, initiate rhEPO treat- cell transfusions during chemotherapy. In a series of nine
ment when hemoglobin is <10 g/dL. If the hemoglobin meta-analyses, the relative risk for transfusion ranged from
level approaches or exceeds 11 g/dL, reduce or interrupt 0.58 to 0.67 in rhEPO-treated patients. Although the risks
the dose of rhEPO. associated with allogeneic transfusions are avoided in some
• For patients who do not respond adequately over a patients treated with rhEPO, the requirement for transfu-
12-week escalation period, increasing the rhEPO dose fur- sions is not completely eliminated.
ther is unlikely to improve response and may increase Several clinical trials and meta-analyses have reported that
risks. rhEPO therapy for chemotherapy-induced anemia may
improve quality of life as measured by Functional Assess-
The initial dose of epoetin alfa in predialysis CKD patients is ment of Cancer Therapy instruments. More recently, the
typically 50-100 U/kg administered subcutaneously once a presence, magnitude, and clinical significance of any poten-
week. Most patients respond to a regimen of 10,000 U/week. tial beneficial effect of rhEPO on quality of life has been con-
Darbepoetin alfa 60 mg every 2 weeks subcutaneously is an troversial, especially in the context of the accumulating
alternative regimen in predialysis patients. evidence of risks of rhEPO therapy in this patient popula-
For hemodialysis patients, the recommended initial dose tion, leading to use restrictions to minimize the potential for
of epoetin alfa is 50-100 U/kg three times per week. The harm.
weekly epoetin dose requirement was shown to be about In 2008, the FDA mandated changes to the labels of epo-
30% less with subcutaneous administration as compared etin alfa and darbepoetin alfa based on risks of shortened
with intravenous route in a randomized trial involving survival or increased risk of tumor progression in cancer
patients on hemodialysis. Most hemodialysis patients, how- patients, as well as the risks of cardiovascular complica-
ever, receive epoetin alfa intravenously because of discom- tions reported in other studies. Since 2010, prescribers and
fort with subcutaneous injections and the convenience of an hospitals must enroll in and comply with a risk manage-
intravenous route during dialysis. Darbepoetin alfa typically ment program REMS (risk evaluation and mitigation strat-
is started at 0.45 mg/kg administered intravenously once a egy), termed the ESA APPRISE Oncology Program
week. (Assisting Providers and Cancer Patients with Risk Infor-
Before and during rhEPO therapy, iron stores are assessed mation for the Safe Use of Erythropoiesis-Stimulating
and monitored to avoid development of iron deficiency and Agents) to prescribe or dispense rhEPO products to
to achieve maximum benefit from rhEPO. Ferritin levels patients with cancer. Over the past five years, the use of
typically are maintained ≥100 ng/dL and the transferrin sat- erythroid stimulating agents, especially in the setting of
uration at 20%. Many hemodialysis patients require intrave- chemotherapy-induced anemia, has dropped substantially.
nous iron infusions to ensure the adequacy of iron stores The FDA-approved label for rhEPOs currently recom-
during rhEPO therapy. mends the following:
Cancer patients receiving myelosuppressive • Use the lowest dose needed to avoid red blood cell
chemotherapy transfusions.
Patients with nonmyeloid malignancies receiving myelo- • Use rhEPO only for anemia from myelosuppressive che-
suppressive chemotherapy frequently develop mild to mod- motherapy.
erate anemia. To ameliorate cancer or chemotherapy-induced • rhEPO is not indicated for patients receiving myelosup-
anemia and its associated symptoms such as fatigue, about pressive chemotherapy when the anticipated outcome
50% of patients will require red blood cell transfusions is cure. The specific types of malignancies were not
indicated.
• Initiate rhEPO only if hemoglobin is <10 g/dL, and if there
Table 4-8 FDA-approved indications for darbepoetin alfa is a minimum of an additional 2 months of planned che-
Anemia due to: motherapy.
Chronic kidney disease in patients on dialysis and patients not on • Reduce dose by 25% if hemoglobin increases >1 g/dL in
dialysis any 2-week period or if hemoglobin reaches a level at
The effects of concomitant myelosuppressive chemotherapy, and which transfusion is not required.
upon initiation, there is a minimum of an additional 2 months of • Withhold dose if hemoglobin exceeds a level needed to
planned chemotherapy
avoid red cell transfusion.

• Discontinue use if there is no hemoglobin response or if • Starting doses and dose modifications should follow FDA-
transfusions are still required after 8 weeks of therapy. approved labeling.
• Discontinue following the completion of a chemotherapy • rhEPO should be discontinued after 6-8 weeks in nonre-
course. sponders.
• rhEPO should be avoided in cancer patients not receiving
The typical starting dose of epoetin alfa is 150 U/kg subcu- concurrent chemotherapy, except for those with lower risk
taneously three times per week or 40,000 U subcutaneously myelodysplastic syndromes (MDS).
weekly until completion of a chemotherapy course. The • Caution is recommended when using rhEPO with chemo-
starting dose for darbepoetin alfa is 2.25 mg/kg/week or 500 mg therapeutic agents in diseases associated with increased
every 3 weeks subcutaneously until completion of a chemo- risk of thromboembolic complications.
therapy course. An alternative darbepoetin regimen is
200 mg every 2 weeks with comparable efficacy to epoetin
Anemia associated with HIV infection
alfa 40,000 U weekly. Hemoglobin level is monitored weekly
until stable. Previous studies have not identified a specific The prevalence and severity of anemia in patients with HIV
plasma endogenous EPO level above which patients would disease have decreased in the era of highly active antiretrovi-
be less likely to respond to rhEPO therapy, though the higher ral therapy (HAART). In a cohort of 9,690 patients, anemia
the baseline EPO level, the less likely there will be a response (hemoglobin <14 g/dL in men; <12 g/dL in women) was
to exogenous EPO. observed in 36%. More severe anemia (hemoglobin <11 g/dL
Iron stores should be assessed before initiation of therapy in men; <10 g/dL in women) was infrequent, observed in 5%
and monitored periodically during therapy. Oral or paren- of patients.
teral iron supplementation may be required in some The pathogenesis of HIV-related anemia is often complex
patients to maximize response to rhEPO. In patients who and multifactorial, including myelosuppressive effects of
fail to respond to rhEPO, considerations include conco various drugs (notably zidovudine, co-trimoxazole, and
mitant iron deficiency, blood loss, vitamin deficiencies (B12 ganciclovir); coinfections; inflammation causing iron utili-
and folate), hemolysis, anemia associated with the malig- zation defect; HIV infection of marrow stromal cells, which
nancy (“anemia of cancer”), or an underlying hematologic limits their ability to support erythropoiesis; and mild rela-
disorder. tive EPO deficiency in some patients. Bleeding, autoimmune
or drug-induced hemolysis, iron or folate deficiency also
may contribute. Risk factors for anemia development include
American Society of Hematology/American Society of
zidovudine use, CD4 lymphocyte count <0.2 × 109/L, high
Clinical Oncology clinical practice guidelines
HIV viral load, African American ethnicity, and female sex.
The American Society of Hematology (ASH)/American Anemia in HIV infection is independently associated with
Society of Clinical Oncology (ASCO) Update Committee decreased survival, and retrospective analyses suggest that
reviewed data published between January 2007 and January recovery from anemia is associated with decreased risk of
2010 and presented the following recommendations for cli- death. Although rhEPO therapy has been reported to
nicians treating patients undergoing myelosuppressive che- increase hemoglobin level and reduce transfusions in some
motherapy who have a hemoglobin level <10 g/dL: patients, there is no evidence that survival is improved as a
result of rhEPO therapy.
• Identify alternative causes of anemia aside from chemo- In early studies, epoetin alfa (100-200 U/kg three times
therapy or underlying hematologic malignancy. per week) was reported to significantly improve hemoglobin
• Clinicians are advised to discuss potential harms (eg, throm- levels and reduce transfusion requirements in patients with
boembolism, shorter survival) and benefits (eg, decreased AIDS who were receiving zidovudine, with endogenous
transfusions) of rhEPO therapy compared with potential plasma EPO level <500 U/L. Epoetin alfa given once per
harms (eg, serious infections and immune-mediated week (40,000-60,000 U) for patients with hemoglobin <12 g/dL
adverse reactions) and benefits (eg, rapid hemoglobin was reported to be effective in raising hemoglobin level and
improvement) of red blood cell transfusions. improving quality of life. Previous studies have not addressed
• If used, rhEPO should increase hemoglobin to the lowest the issue of optimal target hemoglobin in this clinical setting.
concentration possible to avoid transfusions and it should Caution is advisable given the reported adverse effect profile
be administered at the lowest dose possible. in CKD and cancer patients associated with targeting normal
• Available evidence does not identify hemoglobin levels hemoglobin levels. In the HIV disease setting, the current
≥10 g/dL either as a threshold for initiating treatment or as FDA-approved label indicates to dose epoetin alfa to
targets for rhEPO therapy. achieve a hemoglobin level needed to avoid red blood cell

transfusions, to withhold therapy if hemoglobin exceeds their religious beliefs may benefit from preoperative epoetin
12 g/dL, and to discontinue therapy if no increase in hemo- therapy. One study randomized patients with mild anemia
globin is observed at 8 weeks at a dose level of 300 U/kg per (hematocrit ≤39%) to treatment with three different dosing
week. regimens of epoetin alfa or placebo beginning 25-35 days
before surgery. Iron supplementation was given intrave-
nously. A dose-dependent increase in the number of autolo-
Allogeneic blood transfusions in patients
gous units donated was observed.
undergoing surgery
Perioperative epoetin alfa administration reduces the risk of

Other clinical uses of rhEPO
allogeneic blood transfusions in patients undergoing major
Anemia in patients declining transfusion
elective, nonvascular, noncardiac surgery, primarily studied
in the orthopedic surgery setting. The FDA-approved regi- The published literature is dotted with small series and case
mens for this indication are 300 U/kg daily subcutaneously repost discussing using erythropoiesis stimulating agents in
for 14 days total, administered daily for 10 days before sur- patients who decline allogeneic or autologous blood transfu-
gery, on the day of surgery, and for 4 days after surgery. In sion. One such report reviewed the outcomes of 500 Jehovah’s
patients undergoing major orthopedic surgery with pretreat- Witness patients undergoing cardiac surgery at a single cen-
ment hemoglobin of 10-13 g/dL, significantly fewer epoetin- ter. This study compared an evolving bloodless surgical strat-
treated patients (23%) required transfusions compared with egy in two successive eras. In addition to blood-conserving
a placebo group (45%). In the cohort with baseline hemo- operative techniques, the backbone of this regimen was the
globin of 13-15 g/dL, there was no significant difference in administration of epoetin alpha 300 U/kg intravenously, plus
the number of patients transfused (9% for epoetin alfa and 500 U/kg subcutaneously, on admission. After surgery,
13% for placebo). An alternative approved epoetin alfa regi- 500 U/kg was given subcutaneously every second day, along
men is 600 U/kg/week subcutaneously administered 21, 14, with oral iron supplementation. Aminocaproic acid was also
and 7 days before surgery and on the day of surgery. Consid- given from the time of anesthesia induction to skin closure.
eration of antithrombotic prophylaxis is recommended dur- For the patients managed with this strategy, the 30-day mor-
ing perioperative epoetin alfa therapy. tality from the time of surgery ranged form 1-3%. Data on
Two modified epoetin alfa regimens were investigated in a thrombotic events was not reported. In light of the risk of
randomized, double-blind, placebo-controlled trial involv- venous thromboembolism associated with use of erythropoi-
ing 201 patients undergoing primary hip arthroplasty and esis-stimulating agents in patients with a hemoglobin level
hemoglobin level 9.8-13.7 g/dL. Four weekly doses (20,000 over 10 g/dL, as per the FDA’s black box warning, is it difficult
or 40,000 U) starting 4 weeks before surgery were adminis- to reconcile the potential risks and benefits of this approach.
tered along with oral iron supplementation. Both epoetin The ongoing Transfusion Indication Threshold Reduction 2
alfa regimens significantly reduced the requirement for allo- (TITRe2, ISRCTN70923932) randomized trial will provide
geneic blood transfusions (22.8% for the low-dose and insight to what is an acceptable transfusion threshold in
11.4% for the high-dose group) compared with the placebo patients undergoing cardiac surgery, the results of which will
group (44.9%). The incidence of thromboembolic events be directly applicable to the care of Jehovah’s Witness and
was not different between groups. other patients who decline transfusions.
In a trial of 680 patients undergoing spinal surgery who
did not receive thrombo-prophylaxis, patients were random-
Anemia in preterm infants
ized to preoperative epoetin alfa 600 U/kg for four doses (21,
14, and 7 days prior to surgery and on the day of surgery) or Anemia of prematurity in very low-birth-weight (<1,500 g)
standard care. There was an increased incidence of deep vein infants born before the third trimester of pregnancy is asso-
thrombosis (4.7%) in the epoetin alfa–treated cohort com- ciated with multiple factors, including rapid infant growth
pared with the standard care patient group (2.1%). and expansion of blood volume, shortened life span of neo-
Preoperative epoetin alfa treatment has been used to facil- natal red blood cells, and inadequate EPO production in
itate autologous blood donation, although routine applica- response to anemia. Iatrogenic factors, such as phlebotomies
tion for this indication is not justified in clinical practice for for laboratory tests during critical illness, exacerbate the
reasons of cost and safety; notably, an increased risk of post- problem. Many infants require red cell transfusions for
operative venous thromboembolism if hemoglobin levels are symptomatic anemia.
elevated at the time of surgery. Selected anemic patients who The physiologic decrease in circulating red cells that
are willing to donate autologous blood or those who decline occurs during the first weeks of life in all neonates is more
allogeneic or autologous red blood cell transfusions based on pronounced and rapid in low-birth-weight preterm infants.

The switch of the primary site of EPO production from the No randomized study to date has shown definitively that
liver to the kidney that normally occurs after birth has not rhEPO therapy affects the natural course of patients with
taken place in the preterm infant. EPO production in the MDS. A small, prospective randomized trial compared sup-
liver is less sensitive to anemia and hypoxia, leading to rela- portive care alone to epoetin alfa (with or without G-CSF) in
tively diminished EPO synthesis. anemic patients with lower-risk MDS. Epoetin alfa was
Although rhEPO therapy has been reported to reduce red administered at a daily dose of 150 U/kg. At 4 months, the
blood cell transfusions in very-low-birth-weight infants, erythroid response rate was 36% in the epoetin group com-
questions remain regarding the clinical significance of this pared with 9.6% for supportive care. The secondary objec-
beneficial effect in terms of the absolute reduction in trans- tives, including quality of life measures and overall survival,
fusion volume achieved and whether exposure to multiple were significantly better in epoetin responders compared to
blood donors and allo-immunization risk is prevented by nonresponders. AML transformation was not different
rhEPO therapy. Furthermore, the implementation of strin- between the groups.
gent transfusion criteria in clinical practice has reduced the Two retrospective studies have reported improved sur-
number and volume of transfusions independent of rhEPO. vival in responders to rhEPO therapy compared with non
For these reasons, rhEPO therapy in the setting of anemia of respon ders. The largest retrospective study involved 403
prematurity is not widely adapted into routine clinical patients with de novo MDS (303 patients IPSS low and int-1
practice. risk). The epoetin alfa or beta regimen was 60,000 U weekly,
Retrospective data from a few studies and a meta-analysis and darbepoetin alfa was 300 mg weekly for at least 12 weeks.
suggested a link between rhEPO therapy and exacerbation of Some patients (33%) also received G-CSF. The erythroid
retinopathy of prematurity, a disorder of vascular prolifera- response rate was 40% or 50% using different response ass
tion. At present, no conclusive data demonstrate a direct role essment criteria. Median duration of response was 20 weeks
for rhEPO in retinopathy of prematurity. The possibility of a from the onset of rhEPO therapy. Compared with a histori-
link, however, raises concerns in view of the reported asso- cal, untreated MDS cohort, rates of AML progression were
ciation between endogenous EPO and pathologic neovascu- similar. Overall survival was better in rhEPO responders
larization of proliferative diabetic retinopathy in adults. compared with nonresponders or compared with untreated,
matched, historical controls.
Myelodysplastic syndromes
Investigational uses of rhEPO
Anemia is the most common cytopenia encountered in
patients with MDS. rhEPO has been used as monotherapy or rhEPO was shown to exert neuroprotective and cardiopro-
in combination with G-CSF for treatment of anemia in tective effects in preclinical experimental models of tissue
MDS. Studies using darbepoetin alfa report erythroid injury and in clinical pilot studies. These findings consti-
response rates that are comparable to those with epoetin alfa tuted the rationale for randomized, placebo-controlled clini-
or beta. These drugs do not carry an FDA-approved indica- cal trials designed to investigate the safety and efficacy of
tion for anemia associated with MDS. rhEPO to improve outcomes in patients with acute stroke
The erythroid response rate, reported in single-arm studies, and coronary syndromes. In a clinical trial of patients with
varies widely between 20% and 50% depending on patient acute ischemic stroke, however, rhEPO treatment was not
selection and the response criteria used. Factors predicting associated with an improvement in clinical recovery. There
better response rate to therapy include a low transfusion was a higher death rate in rhEPO-treated patients as com-
requirement (<2 units/month), low endogenous pretreat- pared with patients receiving placebo, particularly in those
ment plasma EPO level (<500 U/L), <10% bone marrow who were treated with thrombolysis.
blasts, and low/intermediate-1 (int-1) risk International Prog- In a series of randomized, placebo-controlled clinical tri-
nostic Scoring System (IPSS). The addition of low-dose als involving patients with ST-segment elevation myocardial
G-CSF may augment the hemoglobin response to rhEPO infarction undergoing percutaneous coronary intervention,
therapy, although the role of G-CSF therapy on the biology rhEPO treatment did not reduce infarct size or improve left
and course of MDS has not been defined. Meta-analyses have ventricular ejection fraction. Higher rates of adverse cardio-
suggested that higher weekly epoetin or darbepoetin doses vascular events, particularly in older patients, were reported
may elicit better erythroid response rate; however, the optimal in some studies.
doses of these agents have not been studied in prospective, The safety and efficacy of rhEPO in reducing allogeneic
randomized studies. Therapy typically is maintained for 12 transfusions has been investigated in the intensive care set-
weeks to assess efficacy and then should be continued until the ting in patients with or without trauma. In randomized tri-
positive effect on anemia and transfusion requirements is lost. als, the effect of rhEPO on red blood cell transfusion

requirements was inconsistent. In a trial involving 1,460 myocardial infarcts or death associated with increased hema-
patients, epoetin alfa did not reduce the frequency of red tocrit leading to early termination of the study.
blood cell transfusions. There was a significant increase in In predialysis CKD patients, the CHOIR study involved
thrombotic events. There was a suggestion of reduced mor- 1,432 epoetin alfa–treated patients randomized to target a
tality in the subset of trauma patients; however, this outcome hemoglobin of 13.5 g/dL or 11.3 g/dL. This study was termi-
requires additional clinical investigation. nated early due to a significant (34%) increase in composite
The prevalence of anemia in patients with congestive cardiovascular outcome (death, myocardial infarction, hos-
heart failure ranges from 15% to 50%. The etiology is pitalization for congestive heart failure or stroke) in the nor-
thought to be multifactorial, including hemodilution, mal hemoglobin group. Post hoc analyses suggested that
inflammation, renal dysfunction, iron deficiency, and use of failure to achieve the target hemoglobin and a requirement
angiotensin converting enzyme inhibitors. Anemia in for higher doses of epoetin alfa were associated with increased
patients with heart failure is consistently associated with risk of adverse cardiovascular outcomes.
worse symptoms, functional impairment, and higher risk of The TREAT trial randomized 4,038 predialysis CKD
death compared with nonanemic patients. A series of small patients with diabetes and anemia to treatment with darbe-
clinical trials of rhEPO therapy reported increased hemoglo- poetin alfa, either to a hemoglobin target of 13 g/dL or to
bin levels associated with improved exercise capacity and left placebo with matching rescue darbepoetin when hemoglo-
ventricular ejection fraction. However, in 2013, the RED-HF bin was <9 g/dL. There was a doubling of the risk of stroke in
trial (2,278 subjects randomized to darbepoetin vs. placebo) patients assigned to darbepoetin compared with placebo. It
demonstrated that treatment with darbepoetin did not is noteworthy that in the subset of patients with a history
improve clinical outcomes in patients with systolic heart fail- of cancer at baseline, significantly more patients died of can-
ure and mild-to-moderate anemia. cer in the darbepoetin group compared with placebo. In a
follow-up analysis of the TREAT trial data, a poor initial
response to darbepoetin was associated with an increased
Adverse effects associated with rhEPO therapy
subsequent risk of death or cardiovascular events, as doses
The safety profile and adverse effects of epoetins and darbe- were escalated to meet target hemoglobin levels.
poetin alfa are considered to be comparable. Cardiovascular
adverse effects, venous thromboembolism, and increased
Venous thromboembolism
mortality or tumor progression in cancer patients constitute
the major concerns. Pure red cell aplasia due to the develop- In the supportive oncology setting, rhEPO therapy is associ-
ment of anti-EPO antibodies is rare and has been described ated with increased venous thromboembolism risk, observed
predominantly in patients with CKD. in both literature-based and individual patient data meta-
analyses as well as in randomized controlled trials. The over-
all rate of these events is relatively infrequent. For instance, a
Cardiovascular adverse effects
literature-based meta-analysis reported venous thromboem-
rhEPO use may be associated with exacerbation of hyperten- bolism in 7.5% of 4,610 patients treated with rhEPO com-
sion, particularly in patients with CKD, and therefore ther- pared with 4.9% of 3,562 control patients (relative risk, 1.57;
apy should not be initiated in individuals with uncontrolled 95% confidence interval [CI], 1.31-1.87). The mechanisms
hypertension. Blood pressure monitoring is essential and of venous thromboembolic events are not well defined and a
avoiding rapid rise of hemoglobin during therapy may ame- conclusive link between hemoglobin levels and increased
liorate the risk of hypertension. An increase of blood pres- thromboembolism risk has not been established. Increased
sure medication dose may be required during rhEPO risk of arteriovenous access thrombosis in hemodialysis
therapy. Hypertensive encephalopathy may be associated patients has been reported in association with higher hemo-
with a rapid rise in blood pressure. Seizures, usually related globin levels.
to uncontrolled hypertension, rarely may occur.
A series of randomized clinical trials raised concern for
Mortality or tumor progression in cancer patients
worse cardiovascular outcomes and mortality in CKD
patients treated with rhEPO to achieve and maintain normal A series of clinical trials since 2003 reported adverse effects,
or near-normal hemoglobin levels compared with lower including tumor progression or increased mortality in some
levels. The Normal Hematocrit Trial randomized 1,233 rhEPO-treated patients, across a diverse group of malignan-
hemodialysis patients with cardiac disease to epoetin alfa ther- cies, including head-neck, breast, non-small-cell lung, uter-
apy to achieve a hematocrit target of 30% or 42%. There ine cervix, lymphoproliferative malignancies, and mixed
was an insignificant trend toward an increase in nonfatal nonmyeloid cancers. The safety signals in these trials led to

implementation of rhEPO use restrictions and REMS to has been used successfully in the treatment of some patients.
minimize the potential for harm. Four of the eight trials However, this was removed from the US market in 2013 because
involved chemotherapy-treated patients, two trials included of increased deaths and cardiovascular events (see below).
patients treated with radiotherapy only, and two trials involved
patients with advanced cancer who did not receive antitumor
rhEPO biosimilars and other erythropoiesis-
therapy. In all eight trials, the target hemoglobin level during
stimulating agents
rhEPO treatment was >12 g/dL, higher than presently recom-
mended. In two trials, however, the achieved hemoglobin level The rationale for the development of epoetin biosimilars is
was <12 g/dL, therefore, raising concern that adverse rhEPO cost saving. These products are not fully identical to the orig-
effects may occur at lower hemoglobin levels as well. inal drugs, and documentation of their quality, safety, and
An individual patient data meta-analysis evaluating the efficacy is essential. Immunogenicity and the production of
effect of rhEPO therapy on mortality risk and survival autoantibodies induced by biosimilar epoetins have been
included 53 studies with 13,933 patients. There was a signifi- associated with pure red cell aplasia. Approved epoetin bio-
cantly increased mortality risk (hazard ratio: 1.17, 95% CI similars are available for clinical use in Europe.
1.06-1.30, P = 0.003) during the active study period associ- Peginesatide is a synthetic peptide-based erythropoiesis-
ated with rhEPO therapy. In the subgroup of patients receiv- stimulating agent (with no sequence similarity to EPO) that
ing chemotherapy, the observed increase in mortality risk stimulates the EPOR dimer and activates similar intracel
did not reach statistical significance (hazard ratio: 1.10, 95% lular pathways that are activated by rhEPO. The dimeric
CI 0.98-1.24, P = 0.12). In this meta-analysis, it was not pos- peptide is conjugated to a polyethylene glycol (PEG) moi-
sible to conclusively identify a subgroup of patients with ety, associated with a prolonged half-life of the PEGylated
either an increased or decreased mortality risk when receiv- product. Phase 3 clinical trials have been completed for the
ing rhEPO compared with other patients. rhEPO dosing fre- treatment of anemia in patients with CKD. The FDA ini-
quency three or more times a week compared with less tially approved it for use only in CKD patients on dialysis,
frequent schedules (once a week or once every 2 weeks) was with a warning and REMS implementation because of
associated with reduced mortality, although there were con- increased cardiovascular events compared with rhEPO
founding factors in this analysis and a dose-response associ- which were observed in two trials involving predialysis CKD
ation was not detected. patients. Subsequently, this product was withdrawn in the
US in 2013 due to studies showing greater rates of cardio-
vascular events and death with peginesatide compared with
Pure red cell aplasia
other forms of EPO.
Pure red cell aplasia is a rare complication that has been A novel class of erythropoiesis-stimulating agents in clini-
encountered primarily in CKD patients treated with subcu- cal development involves HIF stabilization by pharmaco-
taneous rhEPO and is mediated by neutralizing anti-EPO logic inhibition of the prolyl hydroxylation of HIF—the
antibodies that cross-react with endogenous EPO. The peak transcription factor that controls EPO gene expression—
incidence in 2001 was associated with a change in the formu- thereby preventing its degradation in the proteasome. An
lation of a specific epoetin alfa product (Eprex) containing a orally bioavailable PHD inhibitor FG-2216 was reported to
new stabilizing agent thought to induce increased immuno- increase the plasma EPO level in end-stage renal disease
genicity of the drug with subcutaneous administration. (ESRD) patients (even in anephric hemodialysis patients),
There have only been rare cases of red cell aplasia after the suggesting that abnormal oxygen sensing—not a loss of EPO
formulation problem was addressed and Eprex has been production capacity—plays a role in renal anemia.
administered by an intravenous route.
Loss of rhEPO response during therapy associated with
hemoglobin decline >0.5-1.0 g/dL/week and low reticulocyte Platelet growth factors
count <10 × 109/L leads to clinical suspicion of red cell apla-
Thrombopoietin
sia. Bone marrow examination reveals absent or severely
reduced erythroid precursor cells. Serum EPO antibody test- TPO is the major HGF that regulates megakaryopoiesis and
ing is required to confirm diagnosis. Discontinuation of platelet production. TPO is constitutively synthesized in
drug is indicated. Hematologic recovery occurs in the major- the liver and kidneys, released into the circulation, and
ity of patients treated with immunosuppressive therapy, binds to its receptor, MPL (myeloproliferative leukemia
such as corticosteroids, daily oral cyclophosphamide, calci- virus oncogene), expressed on platelets. Platelet-bound TPO
neurin inhibitors, or rituximab. A novel EPOR agonist is cleared from plasma, with the remaining TPO available to
peginesatide that does not cross-react with EPO antibodies bind MPL expressed on bone marrow precursors to activate

Platelet growth factors | 93
JAK2 tyrosine kinase and downstream intracellular signal- of a 14 amino acid peptide with no sequence homology to
ing. The disruption in mice of the gene encoding either TPO TPO, which binds to and stimulates MPL and downstream
or MPL leads to severe thrombocytopenia due to reduced signaling. In two parallel randomized placebo-controlled
number of megakaryocytes. trials involving splenectomized and nonsplenectomized
Naturally occurring mutations in the gene encoding TPO patients with ITP, a durable platelet response during the
that lead to increased plasma TPO levels have been found in 24-week study period was achieved in 38% of romiplostim-
families with hereditary thrombocytosis. Gain-of-function treated patients compared with 0% of placebo among sple-
mutations in the MPL gene also have been reported as the nectomized patients and 60% of romiplostim-treated
basis for congenital or inherited thrombocytosis. Acquired, patients compared with 4% placebo among nonsplenecto-
somatic mutations MPL W515L/K have been found in mized patients. In a subsequent randomized open-label trial
5%-10% of patients with essential thrombocytosis and pri- involving adults with ITP who had not undergone splenec-
mary myelofibrosis. Homozygous or compound heterozy- tomy, the rate of platelet response (>50 × 109/L) during the
gous inactivating mutations in MPL have been reported in 52-week study period was 2.3 times that in the standard-of-
association with decreased TPO response in congenital care group. Romiplostim-treated patients had a lower inci-
amegakaryocytic thrombocytopenia. dence of treatment failure and splenectomy, less bleeding,
fewer platelet transfusions, and a higher quality of life.
The recommended initial dose of romiplostim is 1 mg/kg
TPO receptor agonists
as a weekly subcutaneous injection with dose adjustments
The development of therapeutic agents to stimulate thrombo- weekly by increments of 1 mg/kg until the patient achieves a
poiesis has been of great interest to treat severe thrombocyto- stable platelet count ≥50 × 109/L. The maximum weekly
penia and bleeding associated with common hematologic dose is 10 mg/kg. Treatment goal is to achieve and maintain a
conditions, such as chemotherapy-induced thrombocytopenia, platelet count ≥50 × 109/L as necessary to reduce the risk for
MDS, and immune thrombocytopenia (ITP). First-generation bleeding by using the lowest dose of romiplostim. The devel-
recombinant TPOs were investigated in clinical trials involv- opment of romiplostim-binding antibodies is rare, and these
ing healthy individuals and patients with chemotherapy- antibodies are not cross-reactive with TPO.
induced thrombocytopenia. The emergence of antibodies that
cross-reacted with endogenous TPO prevented the further
Eltrombopag
development of these agents.
Second-generation agents termed TPO receptor agonists Eltrombopag is an orally bioavailable, nonpeptide, small
(or TPO mimetics) romiplostim and eltrombopag subse- molecule TPO receptor agonist that raises platelet counts in
quently were developed and studied in randomized clinical a dose-dependent manner. It activates MPL and downstream
trials in both splenectomized and nonsplenectomized adults signaling via JAK2 by association with specific amino acids
with ITP. The efficacy of these agents to increase platelet in the juxtamembrane and transmembrane regions of the
counts, durable responses as long as therapy is continued, receptor. In a randomized double-blind, placebo-controlled
and reduction in the need for other treatments led to FDA trial, once daily eltrombopag 50 mg was well tolerated and
approval of both agents in 2008. The approval indications in effective in improving thrombocytopenia. Platelet counts
Europe by the European Medicines Agency were more ≥50 × 109/L at 6 weeks was achieved in 59% of eltrombopag-
restrictive, indicated for splenectomized patients who are treated patients compared with 16% of placebo-treated
refractory to other treatments and considered as second-line patients. Eltrombopag-treated patients experienced signifi-
treatment for adult nonsplenectomized patients where sur- cantly less bleeding.
gery is contraindicated.
It currently is recommended that TPO receptor agonists be
considered only in patients with ITP whose degree of throm- Table 4-9 FDA-approved indications for romiplostim and
bocytopenia and clinical condition increases the risk for eltrombopag
bleeding. Long-term continuous therapy is required in the Thrombocytopenia due to:
great majority of patients to maintain the platelet response. Chronic ITP with an insufficient response to corticosteroids,
immunoglobulins, or splenectomy. (romiplostim and
eltrombopag)
Romiplostim Chronic hepatitis C to allow the initiation and maintenance of
Romiplostim is an injectable peptibody (antibody heavy interferon-based therapy (eltrombopag only)
chain linked to a therapeutic peptide) that consists of a Severe aplastic anemia with an insufficient response to
immunosuppressive therapy (eltrombopag only)
human immunoglobulin G1 Fc domain, linked to a dimer

The recommended initial dose for most adult patients is The true incidence of increased bone marrow reticulin
50 mg daily given orally on an empty stomach. Patients deposition and fibrosis is not known but appears to be infre-
with moderate to severe hepatic impairment (Child-Pugh quent and reversible following discontinuation of therapy.
score >7) and individuals of East Asian ethnicity (higher Retrospective analysis of a small number of bone marrow
plasma concentrations than white individuals) should be biopsies taken from romiplostim-treated patients in clinical
started on a lower dose of 25 mg daily. Response-guided dosing trials and a prospective trial involving pre- and on-therapy
involves biweekly dose adjustment to titrate the eltrombopag bone marrow biopsies showed reticulin increases in several
dose toward the target platelet level ≥50 × 109/L. The daily patients, without associated cytopenias. This usually occured
dose should not exceed 75 mg. Antacids, dairy products, and in patients receiving higher doses of romiplostim. In a report
mineral supplements (polyvalent cations such as calcium, of the extended eltrombopag study, 156 bone marrow biop-
iron, aluminum, magnesium) should not be taken within four sies were analyzed from 301 patients treated up to 4.5 years.
hours of drug ingestion because of reduced absorption. Four specimens (2.6%) exhibited increased reticulin grade
MF ≥2. No cytopenias were reported. While on long-term
therapy, periodic monitoring for the development of anemia
Monitoring and adverse effects in ITP patients
and leukoerythroblastic changes in peripheral blood is
Romiplostim and eltrombopag should not be used in an advisable.
attempt to normalize platelet counts. Platelet counts should
be measured weekly until stable at ≥50 × 109/L for at least
TPO receptor agonists in aplastic anemia
4 weeks without dose adjustment, and then monthly there
after. Dose reduction is recommended when platelets are In 2012, a study reported the results of a phase 2 study
>200 × 109/L. Rebound thrombocytopenia after drug dis- involving patients with aplastic anemia refractory to immu-
continuation, characterized by a transient worsening of nosuppression, treated with eltrombopag. Starting dose was
thrombocytopenia 10 × 109/L below the pretreatment base- 50 mg per day and could be titrated up to 150 mg/day for 12
line, may occur in 8%-10% of patients, and may be associ- weeks. Eleven of 25 (44%) had a hematologic response in at
ated with increased risk of bleeding. If treatment is held or least one lineage by 12 weeks. Six patients had improved
discontinued, it is advisable to monitor platelet counts twice hemoglobin levels, and three of them were previously red
a week for at least 2 weeks and reinitiate other treatments as cell transfusion dependent. Nine patients had improvement
indicated. Platelet counts usually recover to baseline after in neutrophil numbers. In an extension phase of this study,
several weeks. eight patients achieved a multi-lineage response. Serial bone
The potential adverse effects of these agents include head- marrow studies in eight of the total population in this study
ache, nausea, vomiting, diarrhea, fatigue, nasopharyngitis, developed new cytogenetic abnormalities, including five
and arthralgia. Eltrombopag may be associated with hepatic patients who had changes to chromosome 7. It is now rec-
injury and elevated alanine aminotransferase levels, observed ommended to discontinue eltrombopag for aplastic anemia
in 10% of patients compared with 7%-8% of placebo in clin- if this occurs. In 2014, the FDA approved this new indication
ical trials. Serum liver enzymes should be checked before ini- (refractory aplastic anemia) for eltrombopag.
tiation of eltrombopag therapy, every 2 weeks during the
dose titration period, and then monthly after establishment
Investigational uses of TPO receptor agonists
of stable dose.
Arterial or venous thromboembolic events were infre- Romiplostim and eltrombopag currently are not approved
quent in long-term studies of romiplostim and eltrombopag for the treatment of thrombocytopenia because of MDS or
in ITP, with an incidence ranging from 2% to 6% and no any cause of thrombocytopenia other than chronic ITP.
clear increase in placebo-controlled clinical trials. These Other potential indications are considered investigational at
events do not appear to correlate with platelet count and present. The published experience to date in chemotherapy-
tend to occur in patients with other risk factors for thrombo- induced thrombocytopenia is limited. In 12 patients with
sis. A recent study reported the absence of in vivo platelet MYH9 mutation-related inherited thrombocytopenia, plate-
activation associated with eltrombopag in ITP patients. let counts improved in 11 patients in response to eltrom-
Acute renal failure associated with eltrombopag ther- bopag treatment. The results of several larger clinical trials
apy was reported in two patients with ITP and antiphospho- involving patients with chronic liver disease and MDS have
lipid antibodies. Kidney biopsy showed acute thrombotic been reported. It appears that in patients with liver disease,
microangiopathy and tubular injury in one patient. Caution hypersplenism, and thrombocytopenia, treatment with
is required when considering TPO receptor agonist therapy eltrombopag to raise the platelet count can lead to more
in patients with ITP and antiphospholipid antibodies. rapid hepatic decompensation (see below).

Bibliography | 95
Thrombocytopenia in chronic liver disease placebo group patients. The AML rate through 58 weeks of
study was 6% for romiplostim compared with 2.4% for pla-
Eltrombopag was investigated in a randomized placebo-
cebo (hazard ratio: 2.51, 95% CI: 0.55-11.47). It is now
controlled trial for the treatment of thrombocytopenia asso-
advised not to use romiplostim in this setting because of this
ciated with hepatitis C–related cirrhosis to facilitate antiviral
concern over leukemia progression.
therapy by improving platelet counts. Eltrombopag therapy
increased platelet counts allowing for the initiation of antivi-
ral therapy and was well tolerated during the 20-week treat-
Acknowledgments
ment period.
A more recent placebo-controlled randomized trial involved
Much of the text in this chapter is similar to the previous
patients with thrombocytopenia resulting from chronic liver
edition’s description of hematopoietic growth factors, and
disease, treated for 14 days with eltrombopag before an inva-
we are indebted to those authors (Lyman and Arcasoy,
sive elective procedure. This trial was terminated because of
ASH-SAP 5th ed., 2013). We would also like to acknowledge
the occurrence of portal vein thrombosis in six patients in the
Molly W. Mandernach for her assistance in composing the
eltrombopag group compared with one patient in the placebo
questions for this chapter.
group. Five of the six patients treated with eltrombopag had
platelet counts >200 × 109/L. An association between an
increased risk of thrombotic events and platelet count ≥200 ×
Bibliography
109/L was identified in a post hoc analysis.
In an open-label study of eltrombopag involving 715 Imbach P, Crowther M. Thrombopoietin-receptor agonists
patients with thrombocytopenia complicating cirrhosis due for primary immune thrombocytopenia. N Engl J Med.
to hepatitis C virus infection, 97% of patients were reported 2011;365:734-741.
to respond with platelets ≥90 × 109/L. No thrombotic com- Kuderer NM, Dale DC, Crawford J, Lyman GH. Impact of
plications were reported to date. Studies investigating the primary prophylaxis with granulocyte colony-stimulating factor
efficacy and safety of eltrombopag for thrombocytopenia on febrile neutropenia and mortality in adult cancer patients
associated with chronic liver disease are ongoing. receiving chemotherapy: a systematic review. J Clin Oncol.
2007;25:3158-3167.
Olnes MJ, Scheinberg P, Calvo KR, et al. Eltrombopag and
Myelodysplastic syndromes improved hematopoiesis in refractory aplastic anemia. N Engl J
Med. 2012;367:11-19.
A phase 1/2 trial involved 44 patients with lower risk MDS Rizzo JD, Brouwers M, Hurley P, et al. American Society of
and platelets ≤50 × 109/L, treated with single agent weekly Hematology/American Society of Clinical Oncology clinical
romiplostim. A durable platelet response was achieved by practice guideline update on the use of epoetin and darbepoetin
46% of the patients. Increased bone marrow blasts were in adult patients with cancer. Blood. 2010;116:4045-4059.
observed in 9% and AML progression occurred in two Rodgers GM, Becker PS, Blinder M, et al. Cancer- and
patients. chemotherapy-induced anemia. J Natl Compr Canc Netw.
The initial results of a randomized, double-blind, placebo- 2012;10:628-653.
controlled clinical trial involving 250 patients with IPSS Swedberg K, Young JB et al. Treatment of anemia with
darbepoetin alfa in systolic heart failure. N Engl J Med.
low/int-1 risk MDS were reported. Patients were random-
2013;368:1210-1219.
ized 2:1 to romiplostim 750 mg/week or placebo for a median
Vogel C, Wojtukiewicz M, Carroll R, et al. First and subsequent
21 weeks. Bone marrow biopsies were analyzed after a cycle use of pegfilgrastim prevents febrile neutropenia in
4-week washout period. Romiplostim therapy was associ- patients with breast cancer: a multicenter, double-blind, placebo-
ated with increased platelet response, reduced bleeding controlled phase III study. J Clin Oncol. 2005;23:1178-1184.
events, and less platelet transfusions compared with placebo. Vaislic CD, Dalibon N, Ponzio O, et al. Outcomes in cardiac
Reversible increase in marrow blasts >10% was observed in surgery in 500 consecutive Jehovah’s Witness patients: 21 year
15% of romiplostim-treated patients compared with 3.6% of experience. J Cardiothorac Surg. 2012;7:95-102.

CHAPTER
5
Iron physiology, iron overload,
and the porphyrias
Heather A. Leitch and Jecko Thachil
Introduction, 97 The porphyrias, 109
Regulation of iron homeostasis, 97 Bibliography, 116
Hereditary hemochromatosis and other iron overload disorders, 102
The online version of this chapter contains an educational

regulatory hormone, hepcidin, leading to increased iron
multimedia component. absorption. This chapter focuses on iron physiology in the
normal host and in iron overload states, including hemo-
chromatosis and transfusional iron overload in acquired
Introduction anemias. Iron deficiency anemia is discussed with the under-
production anemias in Chapter 6. This chapter also discusses
Iron is vital for survival, but excess iron may be harmful, so the porphyrias as a model of disorders of heme synthesis.
iron balance must be tightly regulated. Essential functions of
iron include oxygen transport and exchange; cellular respi-
ration and electron transfer; metabolic reactions including Regulation of iron homeostasis
heme synthesis, production of oxygen radicals, and, con-
Body iron economy
versely, anti-oxidation; DNA synthesis and repair; ribosome
function and translation to polypeptides; cellular prolifera- Under normal conditions, dietary iron is usually 15-25 mg
tion; and inflammation. The ability of iron to accept and daily, of which 5%-10% (1-2 mg) is absorbed through the
donate electrons allows it to shuttle between the ferrous gastrointestinal (GI) tract and the same amount lost by des-
(Fe2+) and ferric (Fe3+) oxidation states, and is essential for quamation of GI epithelial cells (see Figure 5-1). The average
its participation in a number of enzymatic reactions. Iron, total body content of iron in men is 35-45 mg/kg; and lower
through Fenton chemistry, catalyzes the formation of free in menstruating women. Most iron (about 1800 mg) is pres-
radical ions, which may be harmful to cells, and therefore ent in hemoglobin. Men and women, respectively, have
under physiologic states, iron does not exist unbound to approximately 2 or 1.5 g of erythrocyte iron. Iron is stored in
proteins or heme. Causes of iron overload include repeated cells, predominantly macrophages of the spleen, bone mar-
blood transfusions, the ineffective erythropoiesis of certain row, and liver, but also in hepatocytes, as ferritin or hemo-
chronic anemias, and mutations in a number of genes that siderin (partially denatured ferritin). At steady state, the
lead to decreased production of or resistance to the iron serum ferritin level is a reasonably good reflection of total
body iron stores. Total storage iron is approximately 1 g in
men and 600 mg in women. Additional iron is found as
myoglobin in muscle and in cytochromes and other enzymes.
Conflict-of-interest disclosure: Dr. Leitch has received honoraria, The release of iron into the circulation is regulated by fer-
research funding, and/or served on advisory boards for Alexion, Apo- roportin, expressed on the basolateral GI epithelial cell surface
Pharma, Celgene, and Novartis. She is a member of the Exjade Speak- (and on cells of the reticuloendothelial system [RES] and hepa-
er’s Bureau. Dr. Thachil declares no competing financial interests.
Off-label drug use: Off-label use of iron chelation therapy is tocytes). Ferroportin is downregulated by hepcidin, and when
discussed. iron is low, hepcidin is low, allowing GI iron absorption to
| 97

98 | Iron physiology, iron overload, and the porphyrias
Bone marrow
~300 mg Liver
~1,000 mg
Other cells
and tissues
~400 mg
20–25 mg/d (Fe3+)2-Tf

1–2 mg/d
~3 mg
Red blood cells
~1,800 mg
Iron loss
Duodenum
~1–2 mg/d
Reticuloendothelial
macrophages
~600 mg
Figure 5-1 Body iron homeostasis. Adequate plasma iron levels must be maintained. Iron circulates in plasma bound to transferrin (Tf), which
maintains iron in a soluble form, serves as a major route of entry for iron into cells (via the transferrin receptor TfR1), and limits the generation
of toxic radicals. The homeostatic system responds to signals from pathways that consume iron (eg, erythropoiesis) and sends signals to the cells
that supply iron to the blood stream. Iron is released into the circulation from duodenal enterocytes, which absorb 1-2 mg of dietary iron per
day, and from macrophages, which internally recycle 20-25 mg of iron per day from senescent erythrocytes. While the body has processes by
which to absorb, transport, and release iron, there is no process for excreting excess iron. Redrawn from Hentze MW et al. Cell. 2004;117:285-
297, with permission.
increase and stores to be mobilized from the RES. When iron Intestinal iron absorption
is plentiful, hepcidin levels increase and result in decreased
Iron is found in food as inorganic iron and heme (iron com-
iron absorption and RES export. Hepcidin, in turn, is down-
plexed to protoporphyrin IX). The typical diet consists of
regulated by the recently described hormone erythroferrone
(ERFE), produced by erythroblasts during stress erythropoie- 90% inorganic and 10% heme iron, though diets in the
sis. Absorbed iron is transported by transferrin and taken up industrial world can contain up to 50% heme iron from
into cells via the transferrin receptor. Each molecule of trans- iron-rich meats. The bioavailability of inorganic but not
ferrin can bind two molecules of ferric (Fe3+) iron. Transfer- heme iron is influenced by multiple factors such as other
rin-bound iron turns over as iron is used, particularly by dietary constituents, for example, ascorbic acid (enhanced)
developing red blood cells in the bone marrow. The distribu- and phytates and polyphenols in cereals and plants (inhib-
tion of iron is influenced by multiple factors, and under nor- ited). Iron absorption is strongly inhibited by tea, and less so
mal conditions cells maintain a pool of labile iron by controlling by coffee. The rate of iron absorption is influenced by several
uptake via expression of transferrin receptors and storage via factors, including body iron stores, the degree of erythropoi-
ferritin. Most cells have no mechanism for iron efflux. etic activity, blood hemoglobin and oxygen content, and the
Iron balance is regulated such that the amount of iron presence of inflammation. Iron absorption increases when
absorbed equals the amount lost. There is, however, no phys- stores are low or when increased erythropoietic activity is
iologically regulated pathway for excretion of excess iron in required, such as during anemia or hypoxemia. Conversely,
iron overload. Over the past 15 years, considerable progress the physiologically appropriate response to iron overload is
has been made concerning the molecular mechanisms downregulation of intestinal iron absorption; this downreg-
underlying the absorption, transport, utilization, and storage ulation fails in patients with hereditary hemochromatosis or
of iron. The key proteins discussed are listed in Table 5-1. chronic iron-loading anemias.

Regulation of iron homeostasis | 99
Table 5-1 Major proteins involved in iron homeostasis
Protein Location Function Comments

Duodenal cytochrome b Duodenal lumen Reduces intestinal luminal ferric iron
(Dcytb) (Fe3+) to the ferrous form (Fe2+)
that can be transported by DMT1
across the luminal cell surface
Divalent metal transporter Duodenal luminal cell Absorption of nonheme iron Transports Fe2+ across the luminal cell
1 (DMT1) surface surface
Unknown transporter Absorption of heme iron
Hephaestin (HEPH) Duodenal basolateral Ferroxidase activity (catalyzes Fe2+ to Ceruloplasmin analog
membrane Fe3+); facilitates iron export via
ferroportin into the circulation
Ceruloplasmin (CP) Plasma and Ferroxidase activity (Fe2+ to Fe3+) Catalyzes the Cu-dependent oxidation of
macrophage enhances the export activity of iron; transferrin can transport only
macrophage ferroportin ferric (Fe3+) iron
Required to maintain ferroportin on
the cell surface
Ferroportin (FPN1) Duodenal basolateral Iron export out of enterocytes and Also referred to as solute carrier family 40
cell surface macrophages to Tf in the plasma member 1 (SLC40A1), iron-regulated
Hepatocyte cell surface transporter 1 (IREG1), and metal
Macrophage cell surface transport protein 1 (MTP1)
Transferrin (Tf) Plasma The major iron transport protein in Apotransferrin, no bound iron;
the circulation holotransferrin, 2Fe3+ bound*
Transferrin receptor Cell surface delivers iron from the plasma into
(TfR1) erythroid precursors and other cells
Ferritin Cell cytoplasm Major iron storage protein May sequester iron from infectious organisms
Circulating form Function unclear and malignant cells, limiting their growth
Iron regulatory proteins Cytoplasm Regulate synthesis of several proteins, Bind to iron responsive elements (IRE) on
(IRP)-1 and -2 including ferritin and TfR1 mRNA; binding to 3′ UTR of transferrin
mRNA stabilizes and increases mRNA
level; to 5′ UTR of ferritin decreases
translation
Hepcidin (HAMP) Produced by the liver Negatively regulates ferroportin; A 25-amino-acid peptide that regulates iron
causes internalization and absorption and release from macrophages
degradation of ferroportin
Erythroferrone (ERFE) Produced by many Negatively regulates hepcidin Suppresses hepcidin, allowing ferroportin
cells, particularly expression & absorption and mobilization
erythroid of iron
HFE Hepatocyte cell surface Participates in upregulation of A protein mutated in most cases of hereditary
hepcidin hemochromatosis
Hemojuvelin (HJV) Hepatocyte cell surface A BMP coreceptor
Transferrin receptor 2 Hepatocyte cell surface Upregulates hepcidin with HJV
(Tfr2) and HFE
Bone morphogenetic A group of soluble Upregulate hepcidin via BMP One or several BMPs are involved in the
proteins (BMPs) factors receptors hepatocyte response to serum iron
BMP receptors Hepatocyte cell surface Bind BMP’s with HJV, activate BMP receptors type I and type II form
SMAD pathway resulting in heterodimers
upregulation of hepcidin
ALK3 Component of BMP HFE interacts with this receptor C282Y inhibits expression of ALK3 on cell
receptors including resulting in upregulation of surface
BMP6 hepcidin H63D increases proteasomal degradation
Sons of mothers against A group of intracellular SMAD 1/5/8 complexing with
decapentaplegic signal transduction SMAD 4 promotes hepcidin gene
(SMAD) proteins and transcription transcription in response to iron
factors
(continued)

Table 5-1 Major proteins involved in iron homeostasis (continued)
Protein Location Function Comments

Transmembrane protease Hepatocyte cell Downregulates BMP signalling in Serine protease
serine S6 (TMPRSS6) membrane response to iron by cleaving HJV
IL-6, IL-6 receptor, JAK, Soluble factor, surface Upregulate hepcidin in response to Contributes to the anemia of inflammation
STAT receptor, signal inflammation
transduction
molecules
GDF15, TWSG1 Soluble factors Downregulate hepcidin, GDF15 via
HJV/BMP
Hypoxia-inducible factors Intracellular May downregulate hepcidin Hepcidin downregulation is likely indirect
(HIF) transcription factors through stimulation of erythropoiesis;
ERFE effect may prove to be relevant
* Once Tf is >85% saturated, non-transferrin bound iron (NTBI) is seen, a subset of NTBI is redox active and known as labile plasma/cellular
iron (LPI/LCI). LPI/LCI through Fenton chemistry result in the formation of reactive oxygen species (ROS), a measure of oxidative stress,
which has been demonstrated to damage cellular and subcellular components.
Iron is absorbed in the intestine via two pathways: one for response elements (IREs), conserved nucleo tide sequences
inorganic iron and the other for heme-bound iron. Little is with a stem-loop structure that binds iron regulatory proteins
known about heme iron absorption. Nonheme iron in the (IRPs)-1 and -2. The mRNAs for ferritin, DMT1, and FPN1
diet is largely in the form of ferric-oxyhydroxides (Fe3+, ie, have IREs in the 5′ untranslated region (UTR), and the mRNA
rust), but the intestinal epithelial cell apical iron importer, for the TfR has multiple IREs in the 3′ UTR. In low-iron states,
divalent metal transporter 1 (DMT1 or SLC11A2), trans- IRP-1 is in a conformation that allows it to bind to IREs; for
ports only ferrous iron (Fe2+). Iron must therefore be example, it binds the 3′ IRE of the TfR mRNA, stabilizing it
reduced to be absorbed, and this is facilitated by duodenal and allowing transcription of more TfR protein, and to the 5′
cytochrome B (Dcytb), a heme-dependent ferrireductase. UTR of ferritin mRNA, decreasing translation of ferritin for
Once transported across the apical border of the enterocyte, iron storage. Intracellular iron induces ubiquitination and
iron may be stored within the cell. For this purpose, iron is degradation of IRP-2. Iron deficiency by this mechanism
oxidized to Fe3+ by the H-subunit of ferritin and stored in upregulates IRP activity via increased IRE binding, resulting
this form. Eventually, the cell senesces and sloughs off into in increased cellular iron uptake and decreased iron storage.
the feces, and stored iron is lost to the system. Alternatively, Most iron in erythroid cells binds protoporphyrin to
iron may be transported across the basolateral membrane form heme, which complexes with globin proteins, form
into the portal circulation via ferroportin. Ferroportin 1 ing hemoglobin. Erythrocytes survive in the circulation for
(FPN1) is the only known iron exporter in mammals and, approximately 120 days, after which aging red blood cells are
like DMT1, transports only ferrous iron. Once reduced, fer- phagocytized by macrophages of the RES. Hemoglobin is
rous iron is transported across the basolateral membrane by catabolized and iron released to transferrin via ferroportin,
ferroportin, then oxidized to ferric iron by hephaestin. Intes- or stored within the RES as ferritin or hemosiderin.
tinal iron absorption is regulated by hepcidin, which binds The main form of cellular iron storage is ferritin, a com-
to ferroportin, inducing its internalization and degradation. plex of subunits that binds iron and renders it insoluble and
redox inactive. Circulating ferritin is present in a different
subunit form than cellular storage ferritin. The function of
Cellular iron uptake, storage, and recycling
circulating ferritin is incompletely understood.
Each molecule of transferrin binds two ferric (Fe3+) iron
atoms. Diferric transferrin (holotransferrin) binds to the
Regulation of systemic iron physiology
transferrin receptor (TfR1) on target cells and enters by recep-
tor-mediated endocytosis; it is then released from the TfR by Hepcidin is a 25-amino-acid peptide produced in the liver
acidification and transported into the cytoplasm by DMT1. and is the major regulator of iron absorption and storage.
The TfR is recycled to the cell surface. Regulation of the syn- Hepcidin regulates cellular iron egress by binding to ferro-
thesis of multiple proteins involved in iron physiology, includ- portin, leading to its internalization and degradation. In this
ing TfR1, DMT1, FPN1, and ferritin, is controlled at a way, elevated levels of hepcidin inhibit iron absorption from
posttranscriptional level by influencing mRNA stability and the GI tract and promote storage (inhibit release) of iron
translation. The mRNAs of these proteins contain iron within hepatocytes and macrophages (Figure 5-2). Hepcidin

Regulation of iron homeostasis | 101
Low hepcidin Increased circulating iron production is induced by interleukin (IL)-6 and IL-1 via the
JAK/STAT pathway, and can be increased more than 100-
Blood fold in inflammatory states. The dysregulation of iron bal-
Macrophage and Macrophages
ph
hag
age
age ance seen in the anemia of inflammation (anemia of chronic
hepatocyte iron
release Enterocytes disease) can be attributed to an inappropriate increase in
Tf
Fe2+
hepcidin levels, which leads to decreased circulating iron.
Fe2+
Cp
FPN
RBC FPN
Tf
Tf
CP Hepcidin levels are downregulated by anemia and iron defi-
HEPH
Tf ciency. Hepcidin agonists and antagonists are under clinical
FP
NTBI Intestinal iron

N
development for the treatment of disorders of inappropri-

CP
Tfr Tf absorption
ately low or high hepcidin levels, respectively.
Tf
Hepatocytes
Fe2+ Hepcidin Hepcidin, in turn, is negatively regulated by the recently
described hormone ERFE. ERFE is produced by erythro-
Erythroblasts blasts during stress erythropoiesis, feeding back on hepcidin
Erythroferrone and reducing its levels, thus allowing iron absorption and
export via ferroportin to proceed.
Figure 5-2 Regulators of iron balance. Dietary iron enters the enterocyte In hereditary hemochromatosis, defects in hepatocyte iron
after being reduced to the ferrous (Fe2+) state by duodenal cytochrome sensing lead to inappropriately low levels of hepcidin for the
B (Dcytb) and transported by divalent metal transporter 1 (DMT1). degree of iron present. Although details are still evolving, mul-
Hephaestin (HEPH) facilitates iron export by ferroportin (FPN). tiple hepatocyte cell surface proteins, including the hereditary
Hepatocytes take up either free (non-transferrin–bound iron/NTBI, hemochromatosis proteins HFE, hemojuvelin (HJV), and
which enters cells in an unregulated manner) or transferrin-bound iron transferrin receptor-2 (TfR2), are involved in communicating
(receptor mediated endocytosis) and release it back into the circulation
the transferrin saturation to a bone morphogenetic protein
via ferroportin. Iron is also released from macrophages via ferroportin.
(BMP)/sons of mothers against decapentaplegic (SMAD)-
Ferroportin-mediated release of iron is inhibited by hepcidin. Hepcidin
dependent signaling cascade that regulates hepcidin gene tran-
is inhibited by erythroferrone, produced by erythroblasts under
conditions of increased erythropoietic activity. CP, ceruloplasmin; TfR,
scription (Figure 5-3). Iron-dependent BMP signaling is
transferrin receptor. Adapted with permission from A. Piperno. downregulated by the hepatocyte cell surface serine proteinease
in
sferr
Tran 3+ BMP6
3+ Fe
Tra Fe
nsf
err
Fe 3+ in
Fe 3+
Transferrin Transferrin I II
Fe3+ Fe3+ Fe3+ Fe3+ ?
HJV
Cell
membrane
Cytoplasm TfR1 HFE TfR2 HFE TfR2
R-SMAD P P ?
AIK3
Hepatocyte
SMAD4
Smurf1
Ubiquitin
Nucleus
iSMADs
Hepcidin
Figure 5-3 Regulated protein-protein interactions among HFE, TfR2, HJV (proteins mutated in HH), BMP receptors, and BMP ligands play a
critical role in the “sensing” of transferrin-bound iron (Fe) to control hepcidin expression in hepatocytes. HFE binds to BMP receptor type I
(Alk3) to prevent its ubiquitination and proteasomal degradation. As a result, expression of ALK3 is increased on the cell surface, activating
BMP/SMAD signaling and hepcidin transcription. Redrawn from Muckenthaler MU. Blood. 2014;124:1212-1213, with permission.

TMPRSS6 (transmembrane protease serine S6), which is or labile cellular iron (LCI). This labile iron interacts with
thought to cleave HJV from the cell, rendering the receptor hydrogen peroxide and, through Fenton chemistry, leads to
inactive. Other molecules implicated in hepcidin regulation the production of oxygen free radicals or reactive oxygen
include GDF15 and TWSG1, as well as hypoxia inducible fac- species (ROS), which have been demonstrated in pre-clinical
tors (HIF), though details of these pathways are not yet clear. models to damage proteins, lipids, and nucleic acids, leading
It was recently demonstrated that HFE interacts with the to cellular consequences (Figure 5-4).
ALK3 component of the BMP receptors, including BMP6,
resulting in upregulation of hepcidin transcription in response
to increased iron. The C282Y mutation of hereditary HFE Hereditary hemochromatosis and
hemochromatosis inhibits trafficking of HFE to the cell surface, other iron overload disorders
affecting ALK3 cell surface stability, and the H63D mutation
increases its proteasomal degradation, interfering with hepci- Iron deposition in body tissues or organs is referred to as
din upregulation, thus allowing expression of ferroportin. iron overload (hemosiderosis). Hemochromatosis is the
The toxicity of excess iron may be mediated at least in part clinical expression of iron-induced injury in affected body
by its ability to transfer electrons. Once the transferrin satu- tissues. Hereditary hemochromatosis is a relatively common
ration is elevated (70% to 80%-85% depending on the congenital cause of iron overload from increased gastroin-
study), non-transferrin bound iron (NTBI) is seen, which testinal iron absorption. Other etiologies of hemochromato-
may enter cells and contributes to the labile iron pool (LIP) sis are discussed below (see Table 5-2).
A Cellular consequences of labile iron

HFE hemochromatosis
• Iron has an ability to transfer electrons
(Fenton reaction: Fe2+ + H2O2 Fe3+ + OH– + •OH) Epidemiology and genetics
• Production of free O2 radicals:
HFE hemochromatosis is the most common form of heredi-
ROS
tary hemochromatosis. It is prevalent in individuals of
Northern European descent because of the presence of the
autosomal-recessive founder allele, C282Y. It is distinctly
uncommon in other e thnicities. Significant variation exists
Mitochondrial Lipid DNA Protein Lysosomes
damage peroxidation damage damage between the genotypic and phenotypic expression of HFE
hemochromatosis because of the presence of genetic modi-
ROS may damage lipids, proteins, and nucleic acids
fiers or environmental factors.
A G-to-A mutation at nucleotide 845 of HFE leads to a
B Model: mitochondrial ROS signaling
dictates biological outcomes cysteine-to-tyrosine substitution at amino acid 282, the

Direct damage to C282Y mutation. Ten to 15 percent of white persons are het-
Signaling events DNA/protein/lipids
erozygous for this mutation (C282Y/WT). About 0.5% are
Proliferation/ Adaptive Senescence, homozygous (C282Y/C282Y), but homozygotes account for
differentiation genes death 60%-90% of clinical cases of hereditary hemochromatosis.
Although biochemical abnormalities such as an elevated
transferrin saturation or ferritin level rarely may be present
ROS
levels in heterozygotes, few will develop clinical features of iron
+ ++ +++ ++++ overload in the absence of other risk factors, such as alco-
“Stress” holic hepatitis (see Table 5-3).
A second mutation involves a G-to-C substitution at HFE
Figure 5-4 Cellular responses to oxidative stress. Once transferrin nucleotide 187, leading to a histidine-to-aspartic-acid substi-
saturation is elevated (70-85%), non-transferrin bound iron (NTBI)
tution at amino acid 63 (H63D). Up to 30% of Caucasians are
is seen, a subset of NTBI are redox active and known as labile plasma/
heterozygous for this allele. H63D is less penetrant than
cellular iron (LPI/LCI). LPI/LCI through Fenton chemistry result
in the formation of reactive oxygen species (ROS), a measure of
C282Y, and only a small minority of homozygotes (H63D/
oxidative stress, which has been demonstrated to damage cellular and H63D) will develop clinical features of iron overload. Hetero-
subcellular components (A). Cellular consequences may include cell zygotes for the H63D mutation (H63D/WT) rarely develop
death, or mutation and malignant progression (B). With permission biochemical or clinical evidence of iron overload. Compound
from I. Cabantchik (A) and redrawn from Hamanaka RB et al. Trends heterozygotes (C282Y/H63D) occasionally may develop mild
Biochem Sci. 2010;35:505-513, with permission (B). iron overload and should be evaluated for coexisting risk

Hereditary hemochromatosis and other iron overload disorders | 103
Table 5-2 Causes of iron overload
Condition Cause Mechanism Comments
1. Hereditary conditions Iron enters the enterohepatic circulation

and hepatocytes express transporters
for NTBI; therefore, hepatic iron
overload usually predominates
i) Hereditary hemochromatosis Affect the hepcidin/
ferroportin axis
HFE hemochromatosis Point mutations in the Inappropriately low hepcidin Amino acid substitutions
HFE gene expression; increased Found primarily in whites
intestinal iron absorption
TFR2 hemochromatosis Mutations in TFR2 Inappropriately low hepcidin Found in multiple ethnicities
expression
Hemojuvelin Mutations in HJV Inappropriately low hepcidin Juvenile hemochromatosis
hemochromatosis or compound expression
heterozygote with HFE
Hepcidin hemochromatosis Mutation in HAMP Loss of hepcidin function Juvenile hemochromatosis
Ferroportin disease
Classical Mutations in ferroportin Unable to export iron Loss of function, “macrophage type”
Nonclassical Unable to bind hepcidin Gain of function, “hepatic type”
ii) Other congenital iron
overload syndromes
African iron overload Possible polymorphism in Increases transferrin Hepatic and RES iron overload
ferroportin gene saturation and ferritin
Melanesian iron overload Unknown gene Autosomal dominant inheritance
Aceruloplasminemia Mutations in Affects ferroxidase activity Impairs ability to mobilize iron from
ceruloplasmin gene the RES
Neurological manifestations, DM
Atransferrinemia Mutations in Tf gene Unable to deliver iron to Increased GI iron absorption leads to
erythroid precursors increased cellular uptake of NTBI
iii) Congenital anemias Ineffective erythropoiesis Increased GI absorption +/- RES
+/- transfusions overload
β-thalassemia major & Discussed in Chapter 7
intermedia
Hemoglobin E/β-thalassemia Discussed in Chapter 7
Hemoglobin H disease Discussed in Chapter 7
Congenital dyserythropoietic Discussed in Chapter 15
anemias
Hereditary sideroblastic Discussed in Chapter 15
anemias
Diamond-Blackfan anemia Discussed in Chapter 15
2. Acquired clonal conditions Transfusions +/- ineffective RES overload +/- increased GI
erythropoiesis absorption
Aplastic anemia Discussed in Chapter 17
Myelodysplastic syndromes Discussed in Chapter 17
Myelofibrosis Discussed in Chapter 16
Other Any transfusion dependent anemia not
associated with blood loss
3. Iatrogenic Inappropriate iron Intravenous iron repletion for the
supplementation anemia of renal failure; oral iron
supplements for non-iron deficiency
causes of anemia
DM = diabetes mellitus; HFE = the gene affected in hereditary HFE hemochromatosis; GI = gastrointestinal; NTBI = non-transferrin-bound
iron; RES = reticuloendothelial system; Tf = transferrin.

Table 5-3 Prevalence of HFE genotypes among patients with identified during routine physicals, as part of screening when
hereditary hemochromatosis affected relatives are identified, or when iron panels are
Prevalence Gene drawn for other reasons. Early diagnosis is important to pre-
among patients frequency vent iron overload and avoid end-organ complications. The
with hereditary in the clinical presentation is varied and often nonspecific, such as
Genotype hemochromatosis population Penetrance fatigue, weakness, abdominal pain, arthralgias, and mild ele-
C282Y/C282Y 60%-90% 0.5%* 13.5%† vation of liver enzymes. Endocrine organs are commonly
C282Y/H63D 0%-10% Low affected, and diabetes, hypothyroidism, and gonadal failure
C282Y/WT Rare 10-15%* Low may occur. Both the mechanical and conduction systems of
H63D/H63D 0%-4% Lower the heart may be affected, resulting in heart failure or arrhyth-
H63D/WT Rare 20%‡ Lower still mias. However, iron-induced liver damage remains the most
WT/WT 15%-30% Unknown
recognized complication of untreated disease (Figure 5-5).
Private mutations Rare Unknown
The transferrin saturation in patients with hereditary
Adapted from Cogswell ME et al. Am J Prev Med. 1999;2:134-140. hemochromatosis is higher than in normal individuals but
WT = wild type. shows considerable variability. A transferrin saturation
C282Y refers to a cysteine to tyrosine substitution at amino
>50% in males or >45% in females should prompt a repeat
acid position 282. H63D refers to a histidine to aspartic acid
measurement, preferably fasting, and measurement of the
substitution at amino acid position 63.
*Caucasian population. serum ferritin level. Ferritin, though imperfect, is a reason-
†European; clinical iron overload in all but C282Y homozygotes able surrogate for total body iron stores. Ferritin can be ele-
should prompt a search for contributing factors to iron overload. vated in other conditions, including the metabolic syndrome,
‡Global population. inflammatory states, acute or chronic hepatitis, alcoholic
liver disease, and others. In a population-based screening
program performed through the US Centers for Disease
factors if hemochromatosis is clinically expressed. In the Control and Prevention, 11%-22% of individuals with an
United States, 15%-30% of patients with clinical hemochro- elevated serum transferrin saturation had a concurrent ele-
matosis have no identifiable HFE mutation. vation in serum ferritin level.
Although homozygosity for the C282Y allele accounts for Molecular genotyping of the HFE locus, now a readily
up to 90% of clinical hereditary hemochromatosis, the true available test, should be considered if the diagnosis remains
phenotypic penetrance of HFE mutations remains a matter in question after secondary causes of iron overload have
of debate. In a population screening study, 50% of C282Y been ruled out or if affected family members exist. Before
homozygotes developed disease expression, typically by age genotyping is performed, a detailed discussion with the
60. In a pedigree study of homozygous family members of patient concerning the possible clinical, emotional, and
known affected individuals, 85% of males and 65% of
females had biochemical evidence of iron overload. Despite
this, only 38% of males and 10% of females had disease- Clinical manifestations
related symptoms, and 15% had fibrosis or cirrhosis on liver
Fibrosis Cirrhosis
biopsy. Other studies suggested the clinical penetrance may
be lower; symptoms were no more prevalent in homozygotes Increased hepatic iron
than in an unaffected control population, and the penetrance
was estimated at less than 1%. The true clinical penetrance is Increased transferrin saturation
uncertain but probably between 1% and 25%. Much of the
Increased iron absorption
variability in estimates is a result of different populations
studied (blood donors vs. preventive care clinics vs. the gen-
eral population vs. family members of affected individuals) 0 10 20 30 40 50 60 70 80
Years of age
and how clinical penetrance was defined (iron studies vs.
liver function tests vs. clinical symptoms vs. liver biopsy).
Figure 5-5 The natural history of hemochromatosis in relation to the
liver. An increase in the percent saturation of transferrin can be detected
Clinical presentation and diagnosis in children homozygous for hemochromatosis. Increased liver iron stores
generally can be detected in homozygous men by the end of the second
The classic finding of a male with skin bronzing, hepatomeg- decade. The serum ferritin concentration increases as hepatic iron stores
aly, and diabetes is an advanced, and now rare, presentation. increase. Hepatic fibrosis can be detected early in the fourth decade.
Patients often present for evaluation of abnormal iron studies Clinical manifestations generally occur in the fifth decade or later.

financial implications of making a diagnosis of HFE hemo- enzymes. Arthralgias, diabetes, and hypogonadism may not
chromatosis should be undertaken. resolve, and cirrhosis or risk for hepatocellular carcinoma
Liver biopsy is the gold standard for diagnosis of hepatic may not be reversed.
iron overload. Biopsy provides information on iron content Phlebotomy usually is not indicated and only infrequently
and distribution and whether fibrosis or cirrhosis has devel- performed during adolescence. If an isolated increase in
oped. Liver biopsy has been recommended for C282Y homo- transferrin saturation is identified during screening, ferritin
zygotes with abnormal liver function tests or ferritin >1,000 values should be monitored at 3- to 6-month intervals and
ng/mL to evaluate for cirrhosis, and it can also be considered phlebotomy initiated when ferritin is >300 ng/mL in males or
if a strong suspicion of significant iron overload exists despite >200 ng/mL in nonpregnant females. Avoidance of alcohol
a negative evaluation for HFE mutations or other primary or and exogenous medicinal iron or iron-containing vitamins
secondary causes. If the serum ferritin is <1,000 ng/mL, cir- should be stressed. Dietary change aimed at avoiding iron-
rhosis is rare. Iron distribution is primarily within hepato- containing foods is often not necessary as long as patients are
cytes (parenchymal), sparing Kupffer cells. A Perls stain of compliant with phlebotomy. Patients should be warned
grade 3 or 4, a liver iron concentration (LIC) of 80 mmol/g about the risks of eating raw seafood because the incidence of
(4.5 mg/g) dry weight or greater, or a hepatic iron index score Vibrio vulnificus and Yersinia enterocolitica infections is
1.9 of greater (hepatic iron in mmol/g divided by patient age) increased in iron overload, as is the risk for mucormycosis,
all confirm the presence of increased body iron stores. especially if they begin chelation therapy. Iron chelation ther-
Another method of estimating storage iron is by phlebotomy. apy should be considered if phlebotomy is contraindicated.
If more than 4 g of iron (about 16 units of blood) can be Treatment of cardiac, hepatic, and other complications of
mobilized without the patient becoming iron deficient, iron iron overload is essential. With untreated iron deposition,
stores are at least four times normal. Liver biopsy is per- hepatic fibrosis and cirrhosis may develop. Once cirrhosis
formed less frequently now that confirmatory genotyping develops, there is a >200-fold increased risk of hepatocellular
has become readily available. Drawbacks of liver biopsy carcinoma compared with the general population. Serial
include its invasive nature and inhomogeneous distribution ultrasounds with or without measurement of α-fetoprotein
of storage iron, leading to inaccurate estimates of LIC. (AFP) may be employed to screen for hepatocellular carci-
Techniques including R2* or T2* magnetic resonance noma in at-risk individuals. Liver transplantation has been
imaging (MRI) or superconducting quantum interference performed for end-stage liver disease in these patients.
device (SQUID) susceptometry are noninvasive methods
increasingly used for documenting organ iron overload.
Screening
MRI is available in increasing numbers of centers and
assesses iron deposition in the liver and heart, and more Population screening for hereditary hemochromatosis is
recently the endocrine organs. SQUID is available in only a controversial and currently not recommended. However,
few centers worldwide. Newer techniques for measuring early screening of at-risk individuals or families by measure-
iron overload such as dual energy computed tomography ment of transferrin saturation, ferritin level, and HFE geno-
have also been described. typing should be discussed. The possibility of genetic
discrimination should be discussed before screening pro-
ceeds; for this reason, some authorities recommend against
Treatment
genetic screening before adulthood.
Prompt and aggressive treatment before end-organ complica-
tions occur is key to ensuring normal life expectancy.
Other autosomal-recessive forms
Phlebotomy of 1 unit of blood (200-250 mg iron) should be
of hereditary hemochromatosis
initiated at weekly intervals then tapered in frequency to main-
tain a ferritin level less than 50 ng/mL, provided the hemato- Patients with HFE hemochromatosis rarely present before
crit is maintained above 33%-35%. Normal adults become the fourth decade of life. Clinically significant iron overload
iron deficient after four to six phlebotomies because the typi- in the 20s and 30s is more likely a severe, early onset
cal 1 g of iron stores is depleted. Patients with 4 g of storage autosomal-recessive disorder, termed juvenile hemochroma-
iron do not become iron deficient until 16-20 phlebotomies tosis, which occurs due to recessive loss-of-function muta-
have been performed. In more advanced cases, total body iron tions in HJV or hepcidin. Juvenile hemochromatosis
burden can require weekly phlebotomy for over a year. characteristically presents with life-threatening heart failure
Phlebotomy is often effective at improving a patient’s and polyendocrinopathies (eg, hypogonadotropic hypogo-
overall sense of well-being, resolving fatigue and malaise, nadism and impaired glucose tolerance or diabetes mellitus)
normalizing skin pigmentation, and reducing elevated liver more frequently than liver dysfunction or other clinical

manifestations. Patients often require intensive manage- morbidity even in the absence of transfusion in patients with
ment of cardiac complications but may recover fully with an thalassemia intermedia and other anemias. Blood transfu-
aggressive iron depletion regimen. Recessive mutations in sions are the predominant cause of iron overload in patients
TfR2 are rare, and disease phenotype is indistinguishable with thalassemia major, aplastic anemia, pure red cell aplasia,
from HFE hemochromatosis other than a near-complete myelodysplastic syndromes (MDS), and sickle cell anemia.
penetrance and possible presentation at earlier age. Like HFE Less severe forms of iron overload have been described
hemochromatosis, a common feature of these disorders is a with alcoholic cirrhosis, hepatitis C virus infection, nonalco-
relative deficiency of hepcidin for the degree of iron over- holic steatohepatitis, and porphyria cutanea tarda. In some of
load; the severity of disease phenotype roughly correlates these disorders, the frequency of HFE mutations is higher
with the magnitude of hepcidin deficiency. than would be predicted by chance and likely contributes to
Neonatal hemochromatosis presents as perinatal liver fail- the risk of iron overload. Hereditary aceruloplasminemia
ure and widespread systemic parenchymal iron deposition, may mimic hemochromatosis, but is characterized by normal
but is likely not a primary disorder of iron balance, and transferrin saturation and the presence of neurologic deficits.
appears to be a consequence of alloimmune hepatitis from a As ceruloplasmin has ferroxidase activity that is important for
fetal–maternal antigen incompatibility. Treatment with the release of iron from macrophages, patients with a mutated
intravenous immunoglobulin (IVIg) beginning in midgesta- gene may accumulate excess iron. Finally, aggressive intrave-
tion mitigates the severity of iron overload in newborns born nous iron administration in conditions such as the anemia of
to mothers with a prior affected child. renal failure has been reported to result in iron overload.
Ferroportin disease Iron chelation therapy

Iron overload resulting from autosomal dominant muta- The management of secondary iron overload may be chal-
tions of FPN1 is known as ferroportin disease. Some patients lenging. Anemia often exists, requiring red blood cell trans-
with FPN1 mutations have clinical and histopathological fea- fusions and making phlebotomy impractical. In some cases,
tures similar to autosomal-recessive forms of hemochroma- erythropoiesis stimulating agents (ESA) such as erythropoi-
tosis. Characteristically, mutations affect the ability of etin can be used to increase the hematocrit to a range safe for
hepcidin to bind or induce the degradation of FPN1, leading phlebotomy. Splenectomy may decrease transfusion require-
to a hepcidin-resistant phenotype (gain of function). Other ments in some anemias. Treatment of the underlying condi-
patients have mutations that result in partial loss of the trans- tion, as in aplastic anemia, MDS, or myelofibrosis should be
port function of FPN1 (loss of function). Serum ferritin is undertaken if possible.
often increased in the presence of a low-normal transferrin In situations where offloading excess iron is desirable but
saturation or hemoglobin. These patients typically have sub- phlebotomy cannot be used, iron chelation therapy may be
stantial Kupffer cell iron storage early in their course. They considered. There is considerable experience with this treat-
often sustain an early decrease in serum iron and hemoglo- ment in the hemoglobinopathies, where offloading organ and
bin during phlebotomy, which may limit their tolerance of total body iron has been demonstrated to prevent and even
treatment. Patients with a gain of function mutation display reverse organ deposition and dysfunction. There is increasing
a spectrum of clinical phenotype, and though many require experience with iron chelation therapy in acquired anemias,
only careful monitoring, some may develop significant conditions in which at least some patients appear to benefit
hepatic iron overload or other complications such as from reduction of iron overload. There is a body of preclinical
arthropathy. It may be reasonable to assess tissue iron levels evidence suggesting that some benefit of iron chelation ther-
with imaging and institute treatment in affected individuals. apy in these conditions may be from suppression of labile iron
and its toxic effects, which takes minutes to hours with chela-
tion (Figure 5-4), rather than from, or in addition to, removal
Other causes of iron overload
of organ and total body iron, which takes months to years.
Many chronic anemias, particularly the thalassemias, are All chelators have potential side effects and require appro-
associated with clinically significant iron overload (Table 5-2). priate monitoring, as per the product monographs, and as
Iron overload in these patients can be due to transfusion, summarized in Tables 5-4 and 5-5. The first available iron
increased iron absorption, or both. Ineffective erythropoiesis, chelation agent was deferoxamine (Desferal), which has been
the intramedullary death of developing red blood cells, leads used extensively in hemoglobinopathy patients, and good
to inappropriately increased iron absorption through sup- compliance with chelation in patients with β-thalassemia
pression of hepcidin production, likely via erythroferrone. major improved their median survival from the teens to nor-
Ineffective erythropoiesis can lead to significant iron-related mal. Deferoxamine is administered by nightly continuous

Table 5-4 Iron chelation agents currently available for clinical use; properties and indications
Property Deferoxamine Deferiprone Deferasirox

Usual dose 20–60 mg/kg/day 75–100 mg/kg/day 20–30 mg/kg/day
Route sc, iv po po
>8–12 h, >5 d/wk 3 times daily Once daily
Half-life 20–30 min 3–4 h 8–16 h
Excretion Urinary, fecal Urinary Fecal
Side effects* Injection site rxn Agranulocytosis (rare) Renal insufficiency in up to 1/3‡
Potential ocular and/or otic toxicity† GI disturbance
Indications Acute iron intoxication IOL in BTM when DFO BTM ≥6 y with IOL from frequent RBC transfusion
Chronic IOL from TD-anemias contraindicated or IOL when DFO contraindicated or inadequate in:
inadequate Other anemias
Age 2-5 y
BTM with IOL from infrequent RBC transfusion
Reprinted from Leitch HA, Vickars LM, Hematology Am Soc Hematol Educ Program 2009;2009:664-672, with permission from the American
Society of Hematology.
BTM = b-thalassemia major; DFO = deferoxamine; IOL = iron overload; rxn = reaction; TD = transfusion dependent.
*Monitoring as per product monograph for all agents.
†Yearly monitoring recommended for all.
‡Usually reversible or non-progressive.
subcutaneous infusion (up to 40 mg/kg) over an 8- to 12-hour benefits. Pre-clinical studies aiming to increase the half-life of
period. Local injection site complications are frequent and deferoxamine (from 5-20 minutes) to 2-3 days by binding it
can be minimized by rotation of injection sites, addition of to a carrier molecule are in progress; this could potentially
hydrocortisone to the infusion, antihistamines, or local mea- make treatment with this agent more attractive to patients.
sures. The potential ocular and auditory complications of Deferasirox (Exjade) was the first oral iron chelator to
deferoxamine mandate annual audiologic and ophthalmo- receive approval from the US Food and Drug Administration.
logic evaluations. Chronic deferoxamine therapy may be In a prospective trial, 20-30 mg/kg of deferasirox daily reduced
arduous, and suboptimal compliance often limits potential LIC, serum ferritin levels, and transaminases that were
Table 5-5 Assessment of iron overload and common adverse events of chelators
Observation Frequency IOL assessment AE monitoring

Iron intake rate Each transfusion √
Chelation dose and frequency 3 monthly √ √
Renal function* As frequently as required √
Liver function 3 monthly √ √
Sequential serum ferritin, transferrin saturation† 3 monthly √
GTT, thyroid, calcium metabolism (BMD‡) Yearly in adults √
Liver iron (T2* MRI)§ At baseline where feasible and subsequently as √
clinically indicated
Cardiac function (echo, MRI, ecg) At baseline, then as clinically indicated √
Cardiac iron (T2* MRI) For patients receiving >50 U RBC prior to ICT, √
or with CHF or arrythmias
Slit lamp examination, audiometry Yearly √
Reprinted from Leitch H. Can Perspect Clin Hematol. 2015;1:4-10. With permission from Canadian Perspectives in Clinical Hematology.
Ideal assessments are listed, and mandatory assessments are shown in boldface type.
Abbreviations: AE = adverse event; BMD = bone mineral density; CHF = congestive heart failure; ecg = electrocardiogram;
echo = echocardiogram; GTT = glucose tolerance test; ICT = iron chelation therapy; IOL = iron overload; MRI = magnetic resonance imaging;
RBC = red blood cells; U = units.
*Creatinine should be measured at least every two weeks with each dose increase until stable.
†Transferrin saturation >80% may indicate the presence of oxidative stress (Sahlstedt L et al. Br J Haematol. 2001;113:836-838).
‡Based on early/suggestive data in transfusion dependent hemoglobinopathies [Ezzat H et al. Blood. 2012;120(21):abstract 3203].
§Up to 25% of hepatic IOL is underestimated by serum ferritin level (Gattermann N et al. EHA Annual Meeting 2013:poster 419).

elevated at baseline prechelation. Adverse events related to the independence is widely recognized to improve survival and
GI tract are frequent with deferasirox and may require dose quality of life in these patients. Multiple nonrandomized
reductions or other measures (published guidelines are avail- studies of iron chelation therapy in MDS, and fewer in the
able). Approximately one-third of patients experience an less common conditions aplastic anemia and myelofibrosis,
increase in serum creatinine. Rare decreases in neutrophil suggest superior survival in patients receiving chelation
count have been noted. Ocular and auditory disturbances have compared to patients not receiving chelation. While these
been reported, though with deferoxamine, these are more fre- data are controversial, what is clear is that in multiple stud-
quent at a ferritin level <1,000 ng/mL, and with deferasirox, ies of MDS, an erythroid response rate around 20% was
this does not seem to be the case. seen with chelation, including the achievement of transfu-
In the United States, the most recently approved oral iron sion independence. Similar responses have been reported
chelator is deferiprone (L1), which is dosed three times daily. in myelofibrosis, and may occur with more frequency in
Side effects include gastrointestinal upset, arthralgias, and aplastic anemia. Patient characteristics predicting for ery-
elevated hepatic enzymes. Drug-induced neutropenia or throid response are currently unclear. The results of a ran-
agranulocytosis requires weekly monitoring. Deferiprone domized controlled trial of chelation in MDS are awaited.
appears to be particularly effective in reducing cardiac iron Guidelines for chelation in MDS are extrapolated from the
overload, which may be a function of its ability to cross the experience with deferoxamine in hemoglobinopathies, and,
cell membrane. Experience with deferasirox, which also for example, suggest chelation once the ferritin level is >1,000
crosses cell membranes, for this indication, is accumulating. ng/mL or the tranfusion burden >20 units of packed red
In some circumstances, intensification of chelation may blood cells. In the future, it may be more appropriate to insti-
be desirable. For example, it has been shown in β-thalassemia tute (nondeferoxamine) chelation at lower doses to prevent
major that a ferritin level over 2,500 ng/mL portends inferior iron overload rather than trying to rescue damaged tissue and
cardiac disease-free survival. Continuous infusions of defer- being unable to dose increase appropriately to the degree of
oxamine or combination regimens should be considered in iron overload because of side effects. This approach, however,
this circumstance at least until cardiac iron status and left should be confirmed to be safe and effective in clinical trials
ventricular ejection fraction (LVEF) are documented as neg- before it can be considered for routine practice.
ative and normal, respectively, and preferably until the fer-
ritin level is consistently <2,500 ng/mL. For patients with
documented cardiac iron loading or decreased LVEF, inten- Key points
sive chelation may partially or fully reverse these complica-
• The absorption of iron is tightly regulated at the level of the
tions and combination therapy with deferoxamine and
enterocyte by hepcidin and ferroportin.
deferiprone, or 24-hour infusions of deferoxamine, should • Iron overload may be due to hereditary or acquired causes, or
be strongly considered. Combinations of deferoxamine and to repeated blood transfusions.
deferasirox are under study. Deferasirox as a single agent • The HFE C282Y/C282Y genotype is the most common and
does improve cardiac iron-related abnormalities; however, most penetrant mutation, leading to clinical iron overload in
intensification of the dose appropriate to the clinical situa- hereditary hemochromatosis.
tion may or may not be limited by side effects requiring dose • The clinical penetrance of the HFE C282Y/C282Y genotype is
interruptions and adjustments, and it is important to address probably <30%.
iron-related cardiac complications in a timely manner. Until • Some clinical manifestations of hemochromatosis are
the treating physician has accumulated experience and a reversible, but cirrhosis and the risk for hepatocellular carcinoma
are not.
comfort level with the use of these agents, the expected side
• Population screening is controversial, but high-risk individuals
effects, and the monitoring required, input from a hematol-
should be screened.
ogist with this expertise should be considered.
• Clinical manifestations of iron overload are similar regardless
Treatment of the acquired anemias is discussed in detail of etiology.
in Chapters 16 and 17, but a few words on chelation in • Phlebotomy to remove excess iron is the primary treatment
these disorders may be appropriate here. Transfusions and for conditions of iron overload not limited by anemia.
iron chelation therapy are generally considered to be sup- • Iron chelation therapy with deferoxamine or deferasirox is an
portive care for acquired anemias. For MDS, the goal of option when phlebotomy is not possible. Monitoring, including
active therapies such as ESA, immunomodulatory agents, regular creatinine levels and other chemistry, and annual
or immunosuppressive therapies for lower risk and hypo- audiologic and ophthalmologic examinations are required in
methylating agents for higher risk disease is hematologic individuals treated with these agents. Deferiprone is a more
improvement, including an erythroid response and trans- recently approved chelation agent for iron overload; monitoring
for agranulocytosis is mandatory.
fusion independence. The achievement of transfusion

The porphyrias | 109
The porphyrias create heme. Although these eight enzymatic steps are similar
in the erythroid cells and liver, control of heme production
Introduction differs between these two tissues, mainly due to the differ-
The porphyrias are a group of metabolic disorders that arise ences in rates of synthesis of ALA, coded by two different
from enzymatic defects in the heme biosynthetic pathway. genes, ALAS1 and ALAS2. In the liver, ALAS1 is the rate-
The name porphyria is derived from the Greek word porphu- limiting enzyme and is downregulated by high levels of heme
ros, or purple, which denotes the purple-red crystalline por- and upregulated by low heme levels. In erythroid cells, the
phyrins that exhibit characteristic red fluorescence on rate is limited by iron availability and not inhibited by heme.
exposure to ultraviolet light. Porphyrins complex with iron One practical implication of this difference is that heme can
to form heme, a molecule crucial for several biologic oxida- be used to treat an acute hepatic porphyria attack. This also
tion reactions. Although biosynthesis of heme occurs in all explains why drugs such as steroids, other chemicals, or
metabolically active cells containing mitochondria, the two stress can worsen hepatic porphyrias because they induce
predominant areas are the bone marrow (85%), where it is ALAS1 along with cytochrome P450 enzymes. On the con-
required for hemoglobin synthesis, and the liver (15%), trary, glucose can suppress gene expression of ALAS1 (the
where it helps in the formation of several enzymes, including glucose effect), accounting for the higher incidence of attacks
cytochromes and hemoproteins. The clinical presentation of while fasting and symptomatic response to glucose infusions.
porphyria is varied, gaining itself the term little imitator. In erythroid cells, regulation is more complex with the
ALAS2 gene expressed 30-fold more than the hepatic enzyme.
Biochemistry
Pathophysiology
Heme synthesis starts with the condensation of glycine and
succinyl-CoA into aminolevulinic acid (ALA), catalyzed by The different types of porphyria arise from a deficiency of one
the enzyme aminolevulinic acid synthase (ALAS), in the of the eight enzymes in the heme biosynthetic pathway (see
mitochondria (Figure 5-6). Another six enzymes are involved Table 5-6). This results in the accumulation of porphyrins and
in the reactions to convert ALA to protoporphyrin, with their precursors in a pattern specific to the enzyme involved,
some of the enzymatic reactions occurring in the cytoplasm which is reflected in the specific clinical manifestations. Dur-
and some others in the mitochondria. In the final step of this ing an acute attack, porphyrin precursors, ALA, and porpho-
enzymatic pathway, protoporphyrin is coupled with iron to bilinogen (PBG) are excreted in large quantities from the liver.
Mitochondria Cytosol
Glycine + succinyl-CoA
δ-Aminolevulinic acid synthase
5-Aminolevulinic acid Porphobilinogen

Aminolevulinic
acid dehydratase
Porphobilinogen deaminase
Hydroxymethylbilane
Heme Uroporphyrinogen III

synthase
Ferrochelatase
Uroporphyrinogen III
Protoporphyrin IX
Uroporphyrinogen decarboxylase
Protoporphyrinogen oxidase
Protoporphyrinogen IX Coproporphyrinogen III

Coproporphyrinogen
oxidase
Figure 5-6 The heme biosynthetic pathway.

Table 5-6 Classification of porphyrias
Inheritance Enzyme
Type of porphyria pattern affected Organs involved Symptoms Comments
Acute porphyrias
Acute intermittent AD Porphobilinogen Nervous system, Neurovisceral No cutaneous
porphyria deaminase/ liver symptoms
hydroxymethylbilane
synthase
Porphyria variegata AD Protoporphyrinogen Nervous system, Neurovisceral, Common in South
oxidase skin, liver cutaneous Africa
Hereditary AD Coproporphyrinogen Nervous system, Neurovisceral Skin lesions occur but
coproporphyria oxidase skin, liver Cutaneous not common
δ-ALA dehydratase AR ALA dehydratase Nervous system, Neurovisceral Very rare, chronic
porphyria liver neuropathy
Nonacute porphyrias
Porphyria cutanea tarda AD Uroporphyrinogen Skin, liver Cutaneous Sporadic and familial
decarboxylase forms exist
Erythropoietic AD Ferrochelatase Skin, red cells, liver Cutaneous Burning sensation in
protoporphyria photosensitive areas
Congenital erythropoietic AR Uroporphyrinogen Skin, red cells Cutaneous hemolytic Erythrodontia bone
porphyria III synthase anemia changes
Hepatoerythropoietic AR Uroporphyrinogen Skin, red cells, liver Cutaneous hemolytic Lab results similar to
porphyria decarboxylase anemia PCT
In all the conditions where liver is involved, chronic liver failure and hepatocellular carcinoma may develop.
AD = autosomal dominant; AR = autosomal recessive; PCT = porphyria cutanea tarda.
These compounds are extremely neurotoxic, especially for the phenotype. This is exemplified by the fact that despite the
autonomic and peripheral nervous systems, which lack blood– prevalence of 1 in 10,000 for the individuals who carry the
brain barrier protection. Although the blood–brain barrier gene for acute intermittent porphyria, only about 10% actu-
protects the brain from these agents, they may still cause vas- ally will present with symptoms.
cular injury and brain edema. The characteristic skin symp-
toms in porphyrias develop as a result of interaction of solar
Classification
radiation and high amounts of circulating porphyrins, which
accumulate in the skin. Once they absorb light, the porphyrin The porphyrias are classified as acute or nonacute (predomi-
molecules emit energy and cause cell damage by peroxidation nantly cutaneous) according to the presenting clinical features
of cell membranes and disruption of intracellular organelles. (Table 5-6). In the acute porphyrias, overproduction of all the
The principal site of light injury has been considered to be in porphyrins and porphyrin precursors proximal to the enzyme
the blood vessels of the papillary dermis. Skin biopsy in these defect occurs. The precursors are excreted in large amounts
cases is not informative and actually may cause harm. due to the decreased activity of PBG deaminase (PBGD),
either due to genetic mutation, as in acute intermittent por-
phyria, or by its feedback inhibition in variegate porphyria or
Inheritance
hereditary coproporphyria. In the nonacute porphyrias, there
Most forms of porphyria are autosomal dominant with is overproduction of all the porphyrins formed before the
incomplete penetrance, although some types are recessive. enzyme defect but not that of the porphyrin precursors, pos-
Rarely, X-linked or complex patterns of inheritance (com- sibly because of a compensatory increase in the activity of the
pound heterozygote) are observed. Concurrent inheritance enzyme PBGD. Additional classification as hepatic or erythro-
of more than one defect has been described in two situations: poietic porphyria is based on the organ in which accumula-
the Chester porphyria (acute intermittent porphyria and tion of porphyrins and their precursors primarily arises.
variegate porphyria) and dual porphyria (variegate por-
phyria and porphyria cutanea tarda). The penetrance of
Acute porphyrias
porphyrias varies with only about half of gene carriers dem-
onstrating clinical manifestations, suggesting additional fac- Four different porphyrias present with acute features,
tors including environmental influences in determining the including acute intermittent porphyria (the most common),

hereditary coproporphyria, variegate porphyria, and the been described in children with neurological symptoms and
extremely rare δ-ALA dehydratase porphyria. developmental delay as the predominant symptoms.
Acute intermittent porphyria Clinical features of acute porphyria

Acute intermittent porphyria (AIP; Swedish porphyria) The predominant symptoms of the acute porphyrias are
results from the deficient activity of the PBGD. It affects neurovisceral. Attacks can begin with minor behavioral
about 1 in 75,000 people in European countries, except in changes such as restlessness and insomnia and progress rap-
northern Sweden, where it is more frequent (1 in 1,000) idly. One typical presentation is with abdominal pain fol-
because of a founder effect. AIP is not associated with skin lowed by vomiting, constipation, and sometimes the
lesions. The mutations underlying AIP typically reduce the development of bladder paresis. Pain in the back or extremi-
activity of PBGD by 50%, which does not always result in ties is a common feature. The differentiating features from
symptoms unless there is an induction of the rate-limiting surgical causes of acute abdomen include poor localization,
hepatic enzyme ALAS1. The latter is induced by many non- absence of peritoneal signs or fever, and the absence of leu-
genetic factors, including certain drugs, endocrine factors kocytosis. The pathogenesis of pain is not yet well under-
(explaining pubertal onset and relation to menstrual peri- stood, although autonomic neuropathy, disturbances in
ods), and reduced calorie intake (linked with peroxisome smooth muscle function, intestinal angina, or lack of nitric
proliferator-activated receptor γ coactivator-1α). High levels oxide have all been suggested as possible reasons. Since the
of PBG contribute to the reddish or port-wine colored urine pain attacks characteristic of the acute porphyrias are also
in AIP. Erythrocyte PBGD activity is decreased in most seen in hereditary tyrosinemia and lead poisoning, all of
patients, although about 5% will have a genetic lesion only in which perturb heme synthesis, delta-aminolevulinic acid has
liver cells (de novo mutations), in which case detection of been suggested as the cause of pain, and the efficacy of hemin
the PBGD mutation is confirmatory. infusion may be due to its inhibition of the enzyme that
catalyzes delta-aminolevulinic acid formation. In support of
autonomic disturbances as the cause of pain attacks, the
Other acute porphyrias
most common clinical sign is tachycardia often accompa-
Variegate porphyria (VP; deficiency of protoporphyrinogen nied by hypertension, both signs of increased sympathetic
oxidase) and hereditary coproporphyria (HC; deficiency of activity. In progressive cases, the autonomic dysfunction can
coproporphyrinogen oxidase) differ from AIP in that they lead to arrhythmias and even occasionally cause cardiac
may present with cutaneous photosensitivity in addition to arrest.
the neurovisceral symptoms. The cutaneous symptoms are Peripheral neuropathy occurs in about 40% of cases of
related to the accumulation of photosensitizing porphyrins, acute porphyria attacks, developing usually after the onset of
which does not occur in AIP because the enzyme block in the abdominal symptoms. Motor neuropathy is more com-
AIP is upstream to the production of porphyrins. Skin mon, similar to cases of Guillain-Barré syndrome, which is
lesions develop in 60% of patients with VP and in about 5% an important differential diagnosis. Mental disturbances,
with HC. These usually develop many days after sun expo- recurrent episodes, and abdominal symptoms are unusual
sure and typically occur on the backs of the hands with easy with Guillain-Barré syndrome, however. Proximal muscles
fragility, blistering, and scarring occurring similar to patients are predominantly affected in the porphyria neuropathy
with porphyria cutanea tarda. VP has a predominance in with upper-limb involvement in 50% of cases. Sensory neu-
South Africa (characteristic mutation is Arg59Tryp), where ropathy, when it occurs, has a bathing-trunk distribution,
it is the most common porphyria seen because of a founder while cranial nerve involvement generally develops after the
effect from a mutation in a Dutch settler. δ-ALA dehydratase limbs and trunk are affected. Respiratory muscle weakness
porphyria (ALAD) (Doss porphyria or plumboporphyria) is may develop with progressive disease and respiratory failure.
the only acute porphyria inherited as an autosomal-recessive Central nervous system involvement, with encephalopa-
trait. Unlike the other acute porphyrias, the porphyrin pre- thy of varying severity, can develop with acute porphyria
cursor ALA and not PBG is increased substantially in the attacks. Cerebrospinal fluid examination is often normal in
urine in this condition. Marked deficiency of ALA dehydra- these cases. Seizures also may occur and often are associated
tase in the absence of lead poisoning suggests the diagnosis with severe hyponatremia. This metabolic disturbance is sec-
in mostly adolescent age-group patients. Chronic neuropa- ondary to a syndrome of inappropriate antidiuretic hor-
thy can develop in this condition. A late-onset type is seen in mone secretion, gastrointestinal loss, and dehydration. Early
association with myeloproliferative disorders. Recessive imaging in neurological porphyria has demonstrated
forms of AIP, VP, and HC (harderoporphyria) also have changes consistent with posterior reversible encephalopathy

s yndrome, linked to acute hypertensive episodes. The mech- potentially dangerous and the risk versus benefit for the use
anism of neural damage in acute porphyrias is not under- of these drugs should be considered in each situation.
stood, although vasospasm resulting from decreased nitrous
oxide production by nitrous oxide synthase (hemopro-
Diagnosis of acute porphyria
tein) or neurotoxicity from porphyrin precursors after its
uptake into cells have been suggested. ALA also can interact The starting point is to suspect acute porphyria. Although
with γ-aminobutyric acid receptors. Dramatic response of certain symptoms like abdominal pain in those with reddish
acute attacks to liver transplantation supports the hypothesis urine or muscle weakness typically occur after the start of a
that heme precursors from the liver cause the neurologic new drug or an oral contraceptive pill, approximately one-
manifestations. tenth of the patients may not have any abdominal symptoms
Psychiatric disturbances are features of acute attack of at all. Because delayed treatment can result in neurologic
porphyria. These include depression, hallucinations, and damage and sometime fatal consequences, a strong suspi-
delirium, while frank psychosis also may occur. Many of cion and early diagnosis is crucial.
these patients may be described as having a psychiatric disor- There are two stages to diagnosing an acute neuropathic
der, the best example being the British sovereign King porphyria. The first step is to obtain evidence that there is an
George III. Nonspecific symptoms like fatigue also are com- ongoing episode of one of the four acute neuropathic por-
mon in acute porphyria with reports of up to 50% of the phyrias (AIP, VP, HC, or ALAD). The fundamental test for
patients being affected by this symptom. this is a correctly collected 24-hour urine test. The second
Acute porphyrias commonly are associated with abnor- step is to determine the subtype of acute porphyria. This step
malities in liver function tests and have a significantly higher involves a review of 24-hour urine, stool, and selected blood
risk of advanced liver disease and hepatocellular carcinoma. tests (eg, red cell enzyme determinations for PBG deami-
Because serum α-fetoprotein may not always be raised in nase). The clinical status of the patient is important as well
these cases, regular screening using liver ultrasound or other because if a patient is hospitalized and critically ill, in addi-
methods is advisable in all adult patients. Chronic renal tion to definitive 24-hour quantitative urine tests, more
impairment may develop in acute porphyrias because of rapid qualitative screening tests should be ordered. These
hypertension, although repeated vasospasms during the qualitative tests may not be readily available, and empiric
recurrent attacks also have been implicated. therapy might have to be started for critically ill patients. Cli-
Acute attacks are much more common in women and nicians must understand the ordering system of the refer-
during the second to fourth decades, occurring rarely before ence laboratory with which their hospital or clinic works to
puberty and after menopause. Menstrual cycles are a com- ensure that appropriate tests are ordered and collected cor-
mon precipitant of the acute attack, with recurrent attacks rectly. For example, some reference laboratories include the
described typically in the late luteal phase (progesterone testing for the porphyrin precursor, PBG, in a 24-hour urine
implicated in increased heme catabolism). Although oral study, whereas others do not and the PBG must be ordered
contraceptive preparations aggravate the attacks, postmeno- separately. The same applies to ALA determinations. If one is
pausal hormone replacement therapy does not seem to have seeing a suspected acute porphyria patient for the first time,
a triggering effect. Other common aggravating factors are 24-hour determinations of both PBG and ALA should be
fasting, alcohol intake (induces or inhibits many enzymes in undertaken to exclude the rare ALAD presentation.
heme biosynthesis pathway), and infection, in addition to In this context, it is worth noting that one common clini-
several drugs. cal circumstance encountered by community physicians or
hematologists-oncologists is the observation of elevated cop-
roporphyrins in the urine of patients suspected of having a
Drugs and porphyrias
neuropathic porphyria. These patients usually have second-
Acute porphyrias have been termed pharmacogenetic diseases ary coproporphyrinuria with the critical diagnostic point
because many drugs can precipitate attacks. The different being that such patients will have normal PBG levels in the
mechanisms include induction of hepatic cytochrome P450 midst of their symptoms––a finding that excludes neuro-
and ALAS-1 and inhibition of other enzymes of heme syn- pathic porphyria, if the urine is collected correctly. These
thesis. Because heme biosynthesis is a complex biosynthetic patients often arrive at referral centers suspected of having
pathway, many drugs may be anticipated to trigger an acute HC because of elevated urinary coproporphyrins but with
attack. A list of safe and unsafe drugs has been compiled into no documentation of PBG levels in urine specimens. Another
databases such as those at http://www.porphyria-europe prevalent outpatient clinical circumstance is that some
.com and http://www.drugs-porphyria.org. Some drugs patients have secondary gain, and all of their prior 24-hour
definitely are contraindicated, but many others are only
urine tests were collected when they were not having

symptoms. In acute porphyria, PBG levels can normalize agent should be started early in the clinical course for better
between attacks. In these instances, the clinician should clinical outcome. The standard regimen is 1-4 mg/kg once
instruct the patient to collect a 24-hour urine sample at a daily of heme in the form of lyophilized hematin, reconsti-
time the patient determines he or she is experiencing clinical tuted with human albumin to avoid thrombophlebitis (Pan-
symptoms. True attacks of acute neuropathic porphyria are hematin; Recordati Rare Diseases Inc, Lebanon, NJ) and
diagnosed easily and have high abnormal levels of PBG or infused daily for 3-14 days, or heme arginate (Orphan
ALA. If a patient has been evaluated for long periods of time Europe), infused daily for 4 days. Hematin is safe in renal
but always with indeterminate results, acute neuropathic impairment. Adverse effects with this drug include fever,
porphyria is less likely. hemolysis, and, before the reconstitution with albumin,
Once the recognition of an acute porphyria episode has phlebitis. The response to therapy occurs often within 1-2
been made, confirmatory tests in the index case to determine days, especially if commenced early in an attack. Recom-
the subtype of acute porphyria are undertaken. Patients with mendations are to complete the full 4-day course of treat-
VP and HC have characteristic 24-hour stool findings even ment in the outpatient clinic.
between attacks. Biochemical confirmation of the type of Careful monitoring is advisable to detect complications
acute porphyria can be done by measuring erythrocyte PBG early (Table 5-7). At discharge, advice should be provided
deaminase levels (AIP) and urine, plasma, and fecal porphy- for various measures to prevent further attacks (Table 5-8).
rin levels by high-performance liquid chromatography or Because oral contraceptives are common precipitants,
fluorometric tests. DNA analysis or enzyme measurements gonadotropin-releasing hormone analogues can be used as
are useful for family members if a mutation is confirmed in alternatives given during the first few days of a cycle. Regular
the index case. Genetic councilors should be engaged for gynecologic assessment and bone density measurements are
familial studies. necessary in such cases. Although pregnancy increases levels
Differential diagnosis of acute porphyria includes lead of progesterone, women who have had acute porphyria
toxicity (where abdominal pain and neuropathy can coexist) should not be advised against pregnancy but rather should
and paroxysmal nocturnal hemoglobinuria (where abdomi- be managed in a specialist center that has experience in deal-
nal pain and discoloured urine occur in the absence of ing with porphyria. Heme arginate is safe in pregnancy and
peripheral neuropathy). The combination of peripheral neu- repeated use does not affect the outcome in pregnancy.
ropathy with central nervous system involvement is unusual About 10% of patients with acute porphyria have recur-
in other conditions and should alert the possibility of por- rent attacks. Once-weekly hematin infusions have been
phyria as the likely diagnosis. Hereditary tyrosinemia type 1,
which develops as a result of accumulation of succinyl ace- Table 5-7 Supportive measures and monitoring in acute porphyria
tone, an inhibitor of ALA dehydratase, can present in chil-
Supportive measures
dren with symptoms resembling acute porphyria.
• Nutritional support: oral, nasogastric, or intravenous
• Relief of pain: opiates
Treatment of acute porphyria • Dehydration: intravenous fluids
• Insomnia and restlessness: chloral hydrate or low doses of
Patients who present with acute porphyria attacks should short-acting benzodiazepines
be hospitalized. All contraindicated drugs should be stopped • Nausea and vomiting: chlorpromazine and prochlorperazine
or avoided. A multidisciplinary approach should be consid- • Tachycardia and hypertension: propranolol and β-blockers but
ered for all patients because the clinical manifestations care with hypovolumia
• Seizure prophylaxis (especially if hyponatremia coexists) and
encompass multiple organ systems. Mild attacks, without
control of seizures: gabapentin or vigabatrin; benzodiazepines
symptoms like severe abdominal pain, neuropathy, and
may be safe
hyponatremia, may be treated initially with high carbohy- • Anesthesia if required: nitrous oxide, ether, halothane, and
drate intake of 2,000 kcal/24 hours orally or through a naso- propofol
gastric tube. If this cannot be tolerated, intravenous 10% • Muscle relaxant: suxamethonium
dextrose should be given targeting at least 300 g/day glucose. • Bladder paresis: catheter
Precaution should be taken to avoid larger quantities of glu- Monitor
cose, which may lead to hyponatremia. Opioids and pheno- • Serum electrolytes, especially sodium and magnesium
thiazines can be given if necessary. Propranol can be used for • Kidney and liver tests
tachycardia and hypertension. • Vital capacity: consider intensive care management if
Severe attacks require treatment with intravenous hemin. deteriorating
Hemin binds to hemopexin and albumin in plasma and is • Neurological status
• Bladder distension
taken up by the liver where it suppresses the ALAS. This

Table 5-8 General and follow-up measures to be discussed with overload as a causative factor in this condition. In addition,
patients with acute porphyria chronic hepatitis C and HIV, and mutations in the hemo-
Counsel chromatosis HFE-gene (in 20% of cases), can contribute to
• Avoidance of alcohol pathophysiology of this disease by increasing liver iron. It is
• Stopping smoking of value to screen the patient and first-degree relatives of
• Information about drugs safe and not safe in porphyria patients with PCT for hereditary hemochromatosis.
• Avoidance of oral contraceptives
PCT usually presents in adults and is characterized by bul-
• Maintain adequate nutrition
• Arrange for medical bracelets
lous lesions, which start often as erythema and become con-
• Psychological input for depression and suicide risk fluent to form the blisters. They most often are observed on
• Genetic counseling for families the backs of the hands and other light-exposed areas. When
• Photoprotection* these blisters rupture, they can cause scarring. Small white
• Avoidance of sunlight exposure and easy skin trauma* papules termed milia are common in the same areas. Hyper-
Follow-up pigmentation and increased hair growth, particularly on the
• For liver problems, especially chronic liver failure and face, can cause disfigurement. Occasionally, the skin in sun-
hepatocellular carcinoma exposed areas becomes severely thickened, a condition
• Those with chronic hypertension needs close follow-up
termed pseudoscleroderma. Skin symptoms show seasonal
• Chronic pain management
• Chronic mental health issues management
variations, with more symptoms in the summer and autumn.
Rare ocular complications have been reported in PCT. Liver
*For porphyrias with cutaneous manifestations only.
dysfunction is common in PCT, especially with alcohol
excess and can vary from mild impairment to cirrhosis. The
suggested to be useful. Repeated infusions, however, may incidence of hepatocellular carcinoma is also higher in this
cause venous occlusion, necessitating central venous access cohort of patients as with acute porphyrias.
and iron overload because of their high iron content. Plasma porphyrin analysis is the best initial test for PCT,
Allogeneic liver transplantation has been performed in with very high levels of isocoproporphyrin noted in the feces.
AIP and VP patients with success. Posttransplantation, ele- In addition to avoiding precipitating factors like alcohol and
vated urinary ALA and PBG levels returned to normal in 24 iron supplements, phlebotomy to reduce hepatic iron is the
hours. Liver transplantation is dangerous, however, and cornerstone of treatment for this condition. Because iron
should be considered only in those who experience recur- overload is not marked, target ferritin can be achieved with
rent and severe attacks. Gene therapy (adeno-associated only few phlebotomies. The plasma porphyrin level can be
virus vector delivering the PBGD gene) and enzyme replace- followed as phlebotomies are undertaken, with expected
ment by recombinant human PBGD also have been control of skin lesions when elevations of the plasma por-
attempted. phyrins are no longer detected. Iron chelation therapy may
be considered if patient cannot tolerate phlebotomy. Low-
dose chloroquine (125 mg twice weekly) can mobilize liver
Cutaneous porphyrias
porphyrins to be excreted in the urine. This may be used in
These differ from acute porphryias mainly in the absence of conjunction with or as alternative to phlebotomy. Caution
neurological symptoms. should be exercised in those with severe liver impairment
because it may cause hepatitis. Any underlying disease like
hepatitis C should be treated while opaque sun-creams con-
Porphyria cutanea tarda
taining zinc oxide should be used as sun-blocking agents.
Porphyria cutanea tarda (PCT) is the most common cutane-
ous porphyria and can be either sporadic (type 1) or familial
Erythropoietic protoporphyria
(type 2). In the familial variety (20%), patients are heterozy-
gous for uroporphyrinogen decarboxylase (UROD) muta- Erythropoietic protoporphyria (EP), the most common por-
tions. The availability of half of the enzyme activity means phyria in children, results from mutations in the ferrochela-
many patients are asymptomatic unless other environmental tase gene and is inherited usually in autosomal-dominant
and precipitating factors intervene. In the sporadic form fashion. In EP, skin lesions begin in early childhood. A char-
(80%), in the absence of any mutations, clinical symptoms acteristic symptom is a burning sensation developing very
develop when the enzyme activity decreases to <20% of nor- quickly in areas exposed to the sun. These may turn erythem-
mal. Recently, uroporphomethene was identified to be an atous but rarely vesiculate. Chronic skin changes can develop,
inhibitor of UROD. Formation of this compound depends although severe scarring, hyperpigmentation, and hirsutism
on hepatic iron; emphasizing the importance of iron are rare. Some patients may have a microcytic, hypochromic

anemia. Late-onset EP has been described in association with Early diagnosis is necessary to avoid phototherapy pre-
myleodysplastic syndromes. Another unusual feature of this scription for neonatal jaundice. Red fluorescent urine in dia-
condition is the development of gallstones, in the absence of pers is suggestive. The management strategy of CEP is based
hemolysis, but this probably is due to excess amounts of pro- on suppressing excess erythropoiesis. In addition, sunlight
toporphyrin decreasing bile flow. Liver disease is common protection and avoidance of skin trauma are important.
but typically develops after age 30. The diagnosis of the con- Bone marrow transplantation has proved effective, whereas
dition is made by measuring total and fractionated porphy- splenectomy and hypertransfusion have shown no benefit.
rins and protoporphyrins. Close coordination with the
reference laboratory being used is important.
Hepatoerythropoietic porphyria
Management of EP mainly includes protection from sun-
light using special clothes, opaque topical sunscreens, or This rare condition is caused by homozygous or compound
ultraviolet-B phototherapy. Oral β-carotene (75-200 mg/day) heterozygous deficiency of uroporphyrinogen decarboxyl-
may help in solar sensitivity by working as an antioxidant, but ase. It tends to present in infancy or childhood and has clini-
it can cause a yellowish skin discoloration due to caroten- cal characteristics similar to CEP with red urine, skin lesions,
emia. Melanocyte-stimulating hormone analogue, which and scarring, whereas hemolytic anemia and splenomegaly
darkens the skin, also has been tried. Because biochemical also may develop. Laboratory tests are similar to PCT, how-
signs of iron deficiency and low vitamin D levels are frequent ever, while treatment is based on sunlight avoidance, with no
findings in this disease, replacement of the vitamin and close response to phlebotomy.
monitoring for anemia is necessary. Interruption of entero-
hepatic circulation of protoporphyrin with cholestyramine or
Pseudoporphyria
activated charcoal may prevent liver damage. Although liver
transplantation has been tried, its success is limited because of Pseudoporphyria is a bullous disorder with clinical and histo-
continued bone marrow production of protoporphyrin. logic features similar to those of PCT but without the charac-
Sequential liver and bone marrow transplantation also has teristic biochemical abnormalities. It was originally observed as
been described. During surgery, modification of surgical dermatological lesions in patients with renal failure on dialysis,
lighting is necessary to reduce injury to organs. earning it the name “bullous dermatosis of hemodialysis.” Sev-
An X-linked form of EP due to gain-of-function muta- eral drugs, however, have been noted to cause this condition,
tions in the ALAS2 gene recently has been described. In this including naproxen, the antibiotic nalidixic acid, dapsone,
condition, there is no ferrochelatase deficiency, but large amiodarone, and diuretics. It also has been noted in individuals
amounts of protoporphyrin accumulate in erythrocytes, who use tanning beds. The clinical features of pseudoporphyria
much of which is bound to zinc. Previously described muta- are identical to those of PCT except that the legs, upper chest,
tions in the ALAS2 gene are loss-of-function ones, which or face may be involved as well. In contrast to PCT, hypertri-
cause X-linked sideroblastic anemia. chosis and hyperpigmentation usually are not seen in pseudo-
porphyria. Treatment involves discontinuation of the suspected
drugs (or avoidance of salons) and sun protection. Alternative
Congenital erythropoietic porphyria
analgesics with less photosensitizing properties (eg, diclofenac)
Congenital erythropoietic porphyria (CEP; Günther disease), may be tried. Hemodialysis-associated pseudoporphyria has
the first-described porphryia, is unique among cutaneous por- responded to treatment with N-acetylcysteine in some reports.
phyrias in being an autosomal-recessive disorder due to defi-
cient activity of uroporphyrinogen III synthase. The severe
Porphyria in the 21st century
cutaneous photosensitivity in this case begins at birth or in
early infancy in most cases. In addition to blistering, the skin is A recent survey of 108 patients with porphyria from the
extremely friable and easily becomes infected. Repeated infec- American porphyria consortium provides some interesting
tions, hypertrichosis, and bone resorption in this condition findings. Most of the affected were females with less than half
lead to severe disfigurement. Because of the deposition of not reporting a parent known with porphyria. Approximately
excess porphyrins in the teeth, they become reddish brown 80% of patients with AIP and 77% and 100% of patients with
(erythrodontia) and fluorescent in ultraviolet light. Corneal HC and VP respectively reported their first symptoms in the
scarring and keratitis cause ocular problems. The excessive second to fourth decades of life. Symptoms of AIP were inter-
increase in red cell protoporphyrins (may be seen as needle- mittent but frequent in about half, present only during acute
like inclusions on blood smear examination) can cause non- attacks in about one-fourth, and constant in 18%.The trig-
immune hemolysis and subsequent splenomegaly. In certain gers for AIP attacks included medications (37%), weight-loss
cases, this may start in utero and manifest as hydrops fetalis. diets (22%), surgery (16%), and environmental toxins (7%).

Abdominal pain, nausea and vomiting, weakness, and consti- Bibliography

pation were the most frequent presenting symptoms of acute
attacks. Chronicity of AIP was manifested in this cohort of Anderson KE, Bloomer JR, Bonkovsky HL, et al.
patients with the development of peripheral neuropathy, sys- Recommendations for the diagnosis and treatment of the acute
temic hypertension, seizures, psychiatric conditions, chronic porphyrias. Ann Intern Med. 2005;142:439-450. A practical
kidney disease, and cirrhosis. guide for guiding physicians in dealing with patients with acute
In relation to the treatment of AIP, the majority of those porphyria written by experts in the field.
Andrews NC. Forging a field: the golden age of iron biology.
who received intravenous heme in the form of hematin
Blood. 2008;112:219-230. An easy-to-read narrative of the basic
found it to be very effective in improving their symptoms.
science of iron physiology.
However, among those who received opiates, less than half
Balwani M, Desnick RJ. The porphyrias: advances in diagnosis and
found them to be beneficial. Hematin was also successful in treatment. Blood. 2012;120:4496-4504. Most recent review on
the prevention of acute attacks. Oral contraceptives wors- porphyrias.
ened the symptoms of porphyria in just over one-third of Bonkovsky HL, Maddukuri VC, Yazici C, et al. Acute
women. Pregnancy was uneventful in 59/60 cases with the porphyrias in the USA: features of 108 subjects from
delivery of normal newborns. porphyrias consortium. Am J Med. 2014;127:1233-1241.
Genetic analysis can detect mutations in genes in AIP, HC, The largest and most recent survey of the clinical, laboratory
and VP, although there were no significant associations and genetic features of porphyria in the United States from the
between these mutations and clinical symptoms or labora- porphyria consortium
tory abnormalities. Fleming MD. The regulation of hepcidin and its effects on
systemic and cellular iron metabolism. Hematology Am Soc
Hematol Educ Program. 2008:151-158. A review discussing the
Key points mechanisms underlying the regulation of hepcidin levels, including
a discussion of HJV and the BMP/SMAD pathways.
• The most common porphyrias seen by clinicians are acute
Ganz T. Hepcidin and iron regulation, 10 years later. Blood.
intermittent porphyria, which is an acute porphyria without skin
2011;117:4425-4433. One of several recent reviews on hepcidin
findings, and porphyria cutanea tarda, which is primarily a
and its central role in regulating iron homeostasis.
cutaneous disease.
Kautz L, Jung G, Volore EV, et al. Identification of erythroferrone
• With acute attacks of porphyria, levels of the substrate PBG or
as an erythroid regulator of iron metabolism. Nat Genet.
rarely ALA are increased several log-fold. Mild elevations are not
2014;46:678-84. A landmark finding in the field of regulation
diagnostic of porphyria.
of iron homeostasis. Erythroferrone, produced by erythroblasts,
• It is important for clinicians to understand the reference
downregulates hepcidin.
laboratory instructions for the correct collection and handling of
Kwiatkowski JL. Real-world use of iron chelators. Hematology Am
specimens.
Soc Hematol Educ Program. 2011;2011:451-458. Practical guide
• Many more patients carry the diagnosis of porphyria than
to iron chelation in transfusional iron overload.
actually have the disease; thus, be critical of the diagnosis.
Wu XG, Wang Y, Wu Q, et al. HFE interacts with the BMP
Secondary coproporphyrinuria without PBG elevations is the
type I receptor ALK3 to regulate hepcidin expression. Blood.
most common clinical state confused with acute porphyria.
2014;124:1335-1343. Identification of the ALK3 component of the
• Many mutations are described in the PBG gene. Having a
BMP receptor as being affected by HFE mutations in hereditary
genetic defect alone does not equate with disease because of an
hemochromatosis, interfering with hepcidin upregulation in
extremely varied penetrance.
response to iron overload.

CHAPTER
6
Acquired underproduction anemias
Siobán B. Keel and Narla Mohandas
Microcytic anemias, 118 Macrocytic anemias, 126 Bibliography, 137
Normocytic anemias, 124 Other underproduction anemias, 133
Erythropoiesis is the process by which hematopoietic stem Heme is a complex of ferrous iron and protoporphyrin IX.
cells divide, differentiate, and mature into enucleated red There are eight enzymes in the mammalian heme synthetic
blood cells (RBCs). The earliest identifiable erythroid pro- pathway (Figure 6-1). The first step occurs within the mito-
genitor is the burst-forming unit-erythroid (BFU-E), which chondria where 5-aminolevulinate synthase (ALA-S2)
is defined functionally by its ability in vitro to form large catalyzes the condensation of glycine and succinyl coenzyme
“bursts” of erythroblast colonies after approximately 2 weeks A (CoA) to yield δ-aminolevulinic acid (ALA). This is the
in semisolid media. Each BFU-E generates >1,000 erythro- rate-limiting step in heme production, and it is regulated by
blasts. The next defined stage is the colony-forming unit- iron availability in erythroid cells. Of note, vitamin B6 is a
erythroid (CFU-E), which under low concentrations of cofactor for ALA-S2. ALA is then transported to the cytosol
erythropoietin give rise to 8-32 well-hemoglobinized eryth- where four additional enzymatic reactions occur, resulting in
roblasts after approximately 1 week in culture. The erythroid the production of coproporphyrinogen III, which is trans-
stages subsequent to CFU-E (proerythroblast to orthochro- ported back into the mitochondria for the remaining three
matic erythroblast) are defined by their light microscopic steps in the heme synthetic pathway, including the final step
appearance on marrow aspirate slides. Erythropoietin (Epo) catalyzed by the enzyme ferrochelatase, where iron is incor-
is the primary cytokine that controls erythropoiesis and acts porated into protoporphyrin IX. This is clinically relevant as
on erythroid progenitors in the stages of CFU-E to the earli- ferrochelatase inserts zinc into the porphyrin ring when iron
est basophilic erythroblasts. Clinically, this explains why the is unavailable, which can be measured in patients as erythro-
absolute reticulocyte count increases ~7 days after an Epo cyte zinc protoporphyin (ZnPP) and may indicate iron-
signal (eg, an acute bleed) because it takes ~7 days for a deficient red cell production.
CFU-E to differentiate into a reticulocyte. Epo is produced In adults, approximately 200 billion erythrocytes are pro-
primarily in the kidney, and its mRNA expression is increased duced each day to replace senescent red cells that are removed
by hypoxia via the transcription factor hypoxia-inducible from the circulation. This requires bone marrow stem cells,
factor (HIF; mutations in a protein required for the destruc- elemental iron, cytokines (including erythropoietin), vita-
tion of HIF under normoxia conditions that result in consti- mins, and a suitable marrow microenvironment. Deficiency
tutive Epo signaling, causes Chuvash polycythemia). Epo or unavailability of any of these key components can result
acts via dimerization of its receptor, resulting in activation of in decreased RBC production and anemia.
receptor-associated Janus kinase 2 (JAK2), which is mutated We define underproduction anemias clinically by the pres-
in the majority of patients with polycythemia vera (see ence of anemia and a corrected reticulocyte count [(reticulo-
Chapter 16) and this activation event initiates a sequence of cyte percent × patient’s hematocrit)/normal hematocrit] of
signaling reactions that prevents apoptosis and stimulates approximately <2%, indicating that the marrow is not
proliferation and maturation of erythroid cells. responding to the degree of anemia. The acquired and con-
genital (reviewed elsewhere) underproduction anemias can
Conflict-of-Interest disclosure: Dr. Keel declares no competing be further grouped by RBC size (ie, mean corpuscular vol-
financial interest. Dr. Mohandas declares no competing financial
ume [MCV]) into microcytic (e.g., iron deficiency anemia,
interest.
Off-label drug use: Dr. Keel: not applicable. Dr. Mohandas: not thalassemia), normocytic (eg, anemia of inflammation, ane-
applicable. mia associated with chronic kidney disease), and macrocytic
| 117

118 | Acquired underproduction anemias
Mitochondria Cytoplasm
Glycine Glycine
+
Succinyl-CoA
ALAS Nucleus
ALA ALA
4 steps
CoproIII
CoproIII
Heme Heme Hemoglobin Figure 6-1 Heme synthesis.

PPGIX FECH ALA = 5-aminolevulinic acid;
PPIX + Fe2+
CoproIII = coproporphyrinogen III;
PPIX PPGIX = protoporphyrinogen IX;
PPIX = protoporphyrin IX;
ALAS = 5-aminolevulinate synthase;
FECH = ferrochelatase.
anemias (eg, megaloblastic anemias, acquired pure red cell Laboratory evaluation reveals a microcytic anemia with a
aplasia, and myelodysplastic syndromes [MDS]). The nor- hemoglobin of 7.4 g/dL, mean corpuscular volume (MCV)
mal ranges of MCV vary by age and ancestry, and physicians of 74 fL, and reticulocyte count of 1%. White blood cell
should take into consideration that the same reference stan- count is normal, and the platelet count is slightly elevated at
dard for hemoglobin and MCV do not apply to all patients. 502,000/mL. Iron studies reveal a low serum iron, elevated
In African American subjects, some of this variability reflects total iron-binding capacity (TIBC), and a markedly reduced
the high prevalence of alpha thalassemia in this population.
A number of other acquired anemias with low corrected
reticulocyte counts are not routinely categorized by cell size Table 6-1 Select acquired underproduction anemias reviewed
(but are most often normocytic). These conditions can be in this chapter
complicated by multiple pathophysiologies, contributing to
Microcytic*
the suppressed RBC production, and are discussed in a sepa-
rate section in this chapter (ie, “Anemia of cancer,” “Myelo-
phthisic anemia,” “Anemia of malnutrition,” “Anemias Normocytic
Anemia of inflammation (~30% are microcytic)
associated with endocrine disorders and pregnancy,” “Ane-
Anemia associated with chronic kidney disease
mia of the elderly,” and “Anemia associated with HIV infec-
tion”). This chapter will focus only on the acquired (and not Macrocytic
Megaloblasticanemia (vitamin B12 and folate deficiencies)
congenital) underproduction anemias. A variety of primary
Acquired pure red cell aplasia
hematopoietic disorders can affect the bone marrow and
Anemia associated with liver disease
lead to acquired underproduction anemia as well other cyto-
Acquired sideroblastic anemias (often macrocytic)†
penias. Detailed discussion of these entities is included else-
Other
where (eg, aplastic anemia, acute leukemia, and MDS). An
Anemia of cancer
outline of the acquired underproduction anemias covered in
Myelophthisicanemias
this chapter is depicted in Table 6-1. Anemia from malnutrition/anorexia nervosa
Anemia associated with endocrine disorders
Anemia associated with pregnancy
Microcytic anemias Anemia of the elderly
*If we consider all low reticulocyte count microcytic anemias (not
Iron deficiency anemia case just those which are acquired), one can think of these broadly as
caused by heme (iron, many congenital sideroblastic anemias) or
A 72-year-old man presents to his primary care provider globin (thalassemia) deficiency.
complaining of increasing dyspnea on exertion and fatigue. † Many (but not all) congenital sideroblastic anemias are microcytic.

Microcytic anemias | 119
ferritin of 9 ug/L. A workup for gastrointestinal (GI) bleed- Table 6-2 Factors that affect dietary iron absorption
ing, including upper and lower endoscopy, reveals angiodys- Inhibit absorption Enhance absorption
plastic lesions of the large bowel. Intravenous iron gluconate
Calcium-rich foods Ascorbic acid
is administered with good clinical response.
Tannins in tea and coffee Heme iron
Iron deficiency is the most common cause of anemia
Phytates in cereals Germination and fermentation of
worldwide. In the United States, iron deficiency (defined by
legumes (removephytates)
the log ratio of transferrin receptor to ferritin) is common
Ferrous iron (Fe2+)
with prevalences in children ages 1-2 and 3-5 years old of
14% and 4%, respectively. The prevalence of iron deficiency
in women ages 12-19 and 20-49 years old is 9% for both age
groups (National Health and Nutrition Examination Study Iron deficiency anemia (IDA) occurs when iron supply is
[NHANES] data 2003-2006). insufficient to meet the iron requirement of developing
Consideration of total iron body content and trafficking RBCs. This occurs secondary to blood loss, increased iron
(see Chapter 5, Figure 5-1) is helpful when calculating the requirements, or inadequate iron supply (Table 6-4). A diag-
amount of iron needed to correct a patient’s iron deficit. The nosis of IDA or iron deficiency alone (without anemia)
vast majority of the body’s iron is contained in hemoglobin requires prompt investigation to determine the underlying
within erythroid cells. The iron in hemoglobin is entirely cause as it may represent the initial presentation of a number
recycled. Senescent RBCs are phagocytosed by reticuloendo- of serious diseases.
thelial macrophages, which degrade hemoglobin and export
the released iron into the plasma where it binds transferrin.
Transferrin-bound iron is then delivered to the bone marrow Table 6-3 Iron content of select foods
to support new RBC production or to the liver for storage as Milligrams/
ferritin (~1 g in men and ~300-600 mg in menstruating Food serving Percent DV*
women) or other sites. Every day ~25 mg of iron recycles Select breakfast cereals, fortified 18 100
through this loop. Furthermore, 1-2 mg of new iron enters with iron, 1 serving
the body each day from dietary intake and absorption, which White beans, canned, 1 cup 8 44
replaces the ~1-2 mg of iron lost daily through the sloughing Dark chocolate, 45-69% cacao 7 39
of skin and intestinal cells and menstrual blood loss. solids, 3 ounces
Intestinal iron absorption and the mobilization of iron Oysters, 6 medium 6 33
from stores in macrophages and hepatocytes are controlled
Beef liver, pan fried, 3 ounces 5 28
by both a stores-regulator and an erythroid-regulator. The
Blackstrap molasses, 1 tablespoon 3.5 19
stores-regulator maintains the body’s normal iron require-
Lentils, boiled and drained, ½ cup 3 17
ments and stores; the erythroid-regulator maintains iron
supply to the erythron regardless of the body’s iron balance. Spinach, boiled and drained, ½ cup 3 17
Hepcidin, a key regulator of iron metabolism, is likely the Firm tofu, ½ cup 3 17
final mediator of both the stores-regulator and the erythroid- Kidney beans, canned, ½ cup 2 11
regulator. Hepcidin acts by binding the iron export protein, Chickpeas, boiled and drained, ½ cup 2 11
ferroportin, leading to its degradation. This inhibits dietary Tomatoes, stewed and canned, ½ cup 2 11
iron absorption and macrophage iron recycling. Systemic and Beef, 3 ounces cooked 2 11
cellular iron homeostasis are described in detail in Chapter 5.
Baked potato, medium sized 2 11
Intestinal iron absorption depends on the amount of iron
Cashew nuts, oil roasted, 1 ounce 2 11
in the diet, its bioavailability, and physiological require-
Chicken, dark meat, 3 ounces cooked 1 6
ments. A typical Western diet contains ~10-20 mg of iron
(roughly 6 mg of iron per 1,000 calories), which exists mostly *DV = daily value recommended by the U.S. Food and Drug
Administration. The DV for iron is 18 mg for adults 19-50 years
as inorganic iron (cereals and legumes) and also as heme
old, 27 mg for pregnant women, and 8 mg for adults > _ 51 years old.
iron (red meats, fish, poultry). Inorganic iron is absorbed
Because iron from plants is less efficiently absorbed than that from
less readily than heme iron (in iron-replete patients, ~10%
animal sources, the recommended DV for iron in a completely
of inorganic iron vs. 30% of heme iron is absorbed). Review- vegetarian diet is ~1.8 times higher that for a nonvegetarian diet.
ing these and other factors that affect iron absorption Foods providing 20% or more of the DV are considered to be high
(Table 6-2) and the iron content of foods (Table 6-3) are sources of a nutrient. For more information on the iron content of
useful when counseling iron-deficient patients on their specific foods, search the USDA food composition database:
dietary habits. http://www.nal.usda.gov/fnic/foodcomp/search/.

Table 6-4 Causes of iron deficiency Accumulating evidence supports a significant role of Helico-
Blood loss bacter pylori infection in the pathogenesis of iron deficiency
Hookworm infections anemia. A number of mechanisms are proposed to account for
Gastrointestinal disorders (esophageal varices, hemorrhoids, peptic this association including occult gastrointestinal bleeding, com-
ulcer disease, malignancy) petition for dietary iron by the bacteria, and impaired absorp-
Menstruation tion due to the effect of H. pylori on the composition of gastric
Pulmonary (hemoptysis, pulmonary hemosiderosis), urologic, or juices. Eradication of H. pylori colonization, which can coexist
nasal disorders with autoimmune atrophic gastritis and may share a com-
Repeated blood donations or clinical blood draws or factitious
mon pathophysiologic mechanism, in infected individuals with
blood removal
refractory IDA has been shown to result in an appropriate
Dialysis
response to oral iron therapy and normalization of hemoglobin
Intravascular hemolysis with hemoglobinuria (paroxysmal
nocturnal hemoglobinuria, prosthetic heart valve)
levels.
Another acquired cause of iron deficiency worth highlight-
Increased iron requirements ing is iron deficiency in infants resulting from inadequate
Rapid growth during infancy and adolescence
dietary iron intake. This occurs in ~20%-40% of infants who
Erythropoietin therapy
drink cows’ milk or nonfortified formula as a sole source of
Pregnancy and lactation
nutrition. Several factors may act synergistically: (i) cows’
Inadequate iron supply milk has a low iron content, (ii) calcium and milk proteins
Poor dietary intake(generally not an independent cause in adults) can inhibit absorption of nonheme dietary iron, and (iii) con-
Malabsorption
sumption of cows’ milk is associated with intestinal blood
Duodenum and upper jejunum diseases (celiac disease, gastric
loss. Also of note, IDA is common during pregnancy (dis-
bypass surgery, inflammatory bowel disease)
cussed in the section “Anemia associated with pregnancy”).
Achlorhydria
Autoimmune atrophic gastritis/Helicobacter pylori colonization
Iron-refractory iron deficiency anemia (IRIDA) is a rare
Congenital disorders of iron transport (iron-refractory iron hereditary disease characterized by a congenital hypochro-
deficiency anemia, hereditary hypotransferrinemia, divalent mic, microcytic anemia, and low serum transferrin satura-
metal transporter 1 disease) tion, which is refractory to oral iron therapy and only
partially responsive to parenteral iron. IRIDA is caused by
mutations in TMPRSS6, a transmembrane serine protease.
In developing countries, hookworm infection leading to These mutations result in inappropriately high hepcidin
chronic intestinal blood loss is the most common cause of levels. Although this is congenital (rather than acquired) dis-
iron deficiency. In the industrial world, nonparasitic GI ease, we include mention of it here, as variants in mutations
blood loss is the most common cause of iron deficiency in in TMPRSS6 are associated with hemoglobin levels, mean
adult males and postmenopausal females. Among those with corpuscular volume measurements, and iron status and may
IDA, evaluation of the GI tract employing endoscopic and modify response to oral iron therapy in iron deficient patients.
radiographic methods identifies a causative lesion in ~60%
of cases. Menstrual blood loss is the most common cause of
Stages of iron deficiency and clinical manifestations
iron deficiency in premenopausal females.
Several additional common causes of iron deficiency are The manifestations of iron deficiency occur in several stages
worth noting. Approximately 5% of patients with IDA (Table 6-5), which are defined by the degree of iron deple-
referred for evaluation by a hematologist have subclinical tion. Initially, iron stores in the bone marrow, liver, and
celiac disease, and this number appears to be higher in those spleen are depleted, which is reflected in a decreased serum
patients with IDA who are unresponsive to oral iron therapy. ferritin. As iron stores become exhausted, the TIBC begins to
In IDA secondary to celiac disease, abnormal iron absorp- rise and serum transferrin saturation begins to fall. When
tion secondary to villous atrophy of the intestinal mucosa erythropoiesis becomes iron restricted, erythrocyte ZnPP
and a component of anemia of inflammation contribute to levels increase and the cells become microcytic. With further
the anemia. It is unclear whether intestinal blood loss also iron depletion, anemia develops.
contributes. Although folate and cobalamin deficiency are Iron-deficient individuals may have no symptoms, non-
known complications of celiac disease, IDA is the most com- specific symptoms of anemia, or other symptoms. Ice eating
mon nutritional deficiency associated with celiac disease. (pagophagia) and other forms of pica are seen in a minority
Autoimmune atrophic gastritis (defined as hypergastrinemia of cases. Findings on physical examination become more
and strongly positive antiparietal cell antibodies) is another pronounced as the iron deficiency worsens and include pal-
common cause of IDA seen by hematologists. lor, stomatitis, glossitis, koilonychia of the nails, and other

Table 6-5 Stages of iron deficiency

Normal Iron depletion Iron-restricted erythropoiesis Iron deficiency anemia
Marrow iron stores +++ None None None

Marrow sideroblasts Present Present Absent Absent
Ferritin (μg/L)* ~40-200 ~20 ~10 <10
MCV Normal Normal Slight microcytosis Microcytic
Anemia Absent Absent Absent Present
TIBC (μg/dL) Normal Normal Normal-mildly increased Increased
Fe (μg/dL), depends on diet ~60-150 ~<40 ~<20 ~<10
Transferrin saturation (%) 20-50 30 <15 <15
Erythrocyte zinc ~30-70 ~30-70 ~100 ~100-200
protoporphyrin (ng/mL)
Fe = iron; MCV = mean corpuscular volume; TIBC = total iron-binding capacity.

*These values represent pure iron deficiency uncomplicated by inflammatory diseases.
symptoms due to the effects of iron on rapidly dividing cells. reproductive age, a ferritin of <10 μg/L is diagnostic of iron
Plummer-Vinson syndrome describes the clinical triad of deficiency (defined as no stainable bone marrow iron stores)
dysphagia (due to esophageal webs), glossitis, and IDA. Sev- with a reported specificity of 98% and a sensitivity of ~75%.
eral studies have examined the relationship between iron A higher ferritin cutoff for suspecting iron deficiency may
deficiency and hair loss. Almost all have addressed women be appropriate in some populations; a serum ferritin of
exclusively and have focused on nonscarring hair loss. Some <30 μg/L in an anemic patient without inflammation was
suggest that iron deficiency may be related to various forms reported to be 92% sensitive and 98% specific in diagnosing
of hair loss, whereas others do not. IDA. Ferritin is an acute phase reactant, and its plasma level
is increased in liver disease, infection, inflammation, and
malignancy. IDA should be considered in these patient pop-
Diagnosis and treatment
ulations when anemia and a serum ferritin of ~60 μg/L are
In classical IDA, a patient presents with an obvious clinical present. Despite its limitations, serum ferritin is superior to
history of blood loss and a low reticulocyte count, microcytic serum transferrin, transferrin saturation, MCV, RDW, and
and hypochromic (pale) red cells, and an elevated platelet erythrocyte ZnPP in diagnosing iron deficiency. A serum
count. The mechanism of thrombocytosis in IDA is ferritin of <30 μg/L is useful in diagnosing iron deficiency in
unknown. Iron studies reveal a low serum ferritin, serum
iron, and transferrin saturation, and an elevated transferrin
(or TIBC). The peripheral blood smear confirms the micro-
cytosis and hypochromasia and may show increased aniso-
poikilocytosis (reflected in an increased red blood cell
distribution width [RDW]) and bizarrely shaped erythro-
cytes, including characteristic cigar-shaped or pencil-shaped
cells (Figure 6-2). Target cells may be seen; however, this
finding is difficult to interpret as it can reflect a drying arti-
fact in smear preparation. Table 6-6 compares the classical
laboratory results in iron-restricted erythropoiesis and ane-
mia of chronic inflammation.
Unfortunately, IDA rarely presents classically and routine
iron studies have limitations, which complicate the diagnos-
tic algorithm. Serum ferritin is a stable glycoprotein that
accurately reflects bone marrow iron stores in the absence of
inflammation. In healthy individuals, serum ferritin is Figure 6-2 Iron deficiency anemia. There are a variety of red
directly proportional to iron stores: 1 μg/L serum ferritin cell sizes and shapes. Included among these are hypochromic
corresponds to ~8-10 mg of tissue iron stores and is an excel- erythrocytes, microcytes, ovalocytes, and “pencil” cells.
lent outpatient screen for iron deficiency. In woman of Source: ASH Image Bank/Peter Maslak.

pregnant women (sensitivity of ~90% and specificity of simultaneously diagnostic and therapeutic and has a role in
~85%), who often have an elevated serum transferrin in the specific clinical scenarios.
absence of iron deficiency. Once iron deficiency or IDA is confirmed, a search for the
Serum iron and TIBC are unreliable indicators of iron underlying etiology (Table 6-4) should be initiated. The
availability to the tissues because of wide fluctuations in levels diagnostic work-up should focus on the likely pathologies
due to ingestion of iron, diurnal rhythm, and other factors. expected in a specific patient. Upper and lower GI endosco-
Transferrin is affected by nutritional status and transferrin pies should be pursued in all males and postmenopausal
saturation is a function of serum iron and transferrin. female patients with confirmed IDA. Capsule endoscopy to
A number of additional studies can support a diagnosis of evaluate the small bowel, repeat endoscopic exams, or other
IDA in situations in which serum ferritin is equivocal. diagnostic modalities at the discretion of a gastroenterologist
Increased RDW is sensitive for diagnosing IDA, but it lacks may be required to diagnose obscure GI bleeding (persistent
specificity. A trend of decreasing MCV and increasing RDW or recurrent bleeding from the GI tract after negative EGD
over time can be instructive. Erythrocyte ZnPP levels are and colonoscopy). Once it has been shown that there is not
increased in iron deficiency as zinc, rather than iron, is incor- an occult GI blood loss requiring diagnosis or intervention,
porated into the protoporphyrin ring when iron is unavail- close follow-up with replacement of iron may be a rational
able. ZnPP has a high sensitivity for detecting iron deficiency approach in some patients. If menstrual blood loss is signifi-
in adults but lacks specificity as it also is increased in lead cant and appears to be the primary source of IDA, a trial of
poisoning, anemia of chronic inflammation, and some iron therapy is reasonable before proceeding to GI evalua-
hemoglobinopathies. The percentage of circulating hypo- tion in carefully selected premenopausal female patients who
chromic erythrocytes and the reticulocyte hemoglobin con- are followed closely for response to therapy. In the clinic, it is
tent detect early iron-deficient erythropoiesis. They are difficult to discern between “normal” and “abnormal” men-
limited, however, by false normal values seen in patients with strual loss and knowledge of actual iron loss in women with
elevated MCVs or with thalassemia, and the requirement for “normal” and “abnormal” menstrual blood loss is scanty.
a specialized analyzer not available in most laboratories, For patients whose iron deficiency anemia remains unex-
respectively. Measurement of serum-soluble transferrin plained or refractory despite standard diagnostic work-up,
receptor (sTfR) also may be used to estimate body iron stores. some experts advocate serological or biochemical screening
The level of sTfR is directly proportional to the total amount for celiac disease with antiendomesial or antitransglutamin-
of cell surface-associated transferrin receptor 1 (TfR1) and ase IgA antibodies, for atrophic gastritis with gastrin and
thus reflects both the quantity of erythroid precursors and antiparietal cell antibody testing, and for H. pylori with IgG
the cell surface expression of TfR1 per cell. Levels are increased antibodies or fecal antigen followed by confirmatory testing
in iron deficiency and anemias with ineffective erythropoiesis with a urea breath test. The definition of refractory iron defi-
(eg, thalassemia, MDS, folate deficiency, and vitamin B12 ciency anemia is not standardized; one definition is a failure
deficiency) in which there is an imbalance between the to achieve a 1-g/dL increase in hemoglobin after 4 to 6 weeks
amount of iron that is endocytosed by marrow erythroblasts of at least 100 mg of elemental iron therapy per day. Cases of
and the amount of iron incorporated into circulating eryth- suspected celiac disease should be confirmed by duodenal
rocytes. The incorporation of sTfR into the sTfR-ferritin biopsy. In patients with iron-refractory or iron-dependent
index (sTfR/log10ferritin) has shown more promise in distin- anemia of unknown origin with confirmed H. pylori infec-
guishing IDA from anemia of chronic inflammation than tion, eradication of the infection with standard therapy is
sTfR1 alone. While commercially available, sTfR testing lacks reported to be curative and thus, should be considered.
standardization and is not widely available. An iron absorption test is useful in evaluating some
Serum hepcidin, the primary regulator of iron homeosta- patients with iron deficiency or IDA. This simple and mini-
sis in the body, is suppressed in iron deficiency. The utility of mally invasive test distinguishes an intestinal iron absorption
measuring serum hepcidin in the workup of iron deficiency defect from other causes of iron deficiency. Ideally, a patient
and other disorders of iron homeostasis, while potentially fasts for ~8 hours, and serum iron is measured at baseline
useful, remains undefined due to lack of a clinically validated and at 90 minutes after administration of ferrous sulfate
assay. However, further investigation of its usefulness in (5 mg/kg). In a patient with IDA with normal intestinal iron
widespread clinical practice is clearly warranted. absorption, the serum iron level is expected to increase by at
Evaluation of the bone marrow for stainable iron had been least 50 µg/dL 90 minutes after the oral iron challenge.
considered the gold standard for the diagnosis of iron defi- The test is less readily interpretable in a nonfasting patient.
ciency. High interobserver variability and the expense and The treatment of iron deficiency or IDA includes address-
invasiveness of the test limit its clinical utility. A trial of oral ing the underlying cause of iron deficiency and replacing the
iron therapy with close laboratory and clinical follow-up is iron deficit. Upfront, it is useful to calculate the patient’s

approximate iron deficit quantitatively. This includes the complicated by a low but significant risk of anaphylaxis
amount of iron required to normalize the hemoglobin plus (11.3/million) and thus no longer should be used. Low-
the amount of iron required to replete iron stores [the Gan- molecular-weight iron dextran is considerably safer than its
zoni equation: total iron deficit = weight {kg} × (target high-molecular-weight counterpart. The advantages of low-
Hb – actual Hb) {g/L} × 2.4 + iron stores {mg}]. This quan- molecular-weight iron dextran include its low cost and the
tity should be considered in the context of intestinal iron ability to give replacement doses of iron in a single infusion.
absorption when considering the likelihood of replacing the Iron sucrose and iron gluconate have very low incidences of
deficit by oral administration (constipation secondary to anaphylaxis (no fatal cases of anaphylaxis have been reported),
oral iron replacement can be dose limiting in some patients) and their administration does not require a test dose. Side
or to define the amount of parenteral iron to administer. effects of iron sucrose and iron gluconate include mild arthral-
Oral iron salts are a safe first-line treatment for iron defi- gias and myalgias. Disadvantages of these newer formulations
ciency. There is no evidence to suggest that a particular oral include greater cost and the inability to give a total replace-
iron preparation is more effective or better tolerated than ment dose in a single infusion because they cause GI and vaso-
another in equivalent doses. The typical replacement dose of active reactions at doses >200-400 mg. Newer iron preparations
elemental iron in adults is 100-200 mg/day and 3-6 mg/kg/day have been developed to enable more rapid high-dose bolus
in infants and children (doses split into divided doses). Ferrous injections. Ferumoxytol, which is only licensed for use in adult
sulfate is available in 325 mg tablets (each containing 65 mg of patients with chronic kidney disease and iron deficiency ane-
elemental iron) and ferrous gluconate is available in 320 mg mia, enables a bolus injection of 510 mg to be administered in
tablets (each containing 32 mg of elemental iron). Ferrous sul- 17 seconds. The US Food and Drug Administration has
fate elixir (a liquid formulation) is available for infants and recently issued a “black box warning” about the risk of severe
young children. Nausea, vomiting, epigastric discomfort, and and potentially fatal allergic reactions to ferumoxytol. Ferric
constipation are common dose-dependent side effects of iron carboxymaltose is now licensed in the US at a dose of 750 mg
salts; ~25% of patients cannot tolerate oral iron because of of elemental iron for patients weighing >50 kilograms and can
side effects. Patients should be alerted that iron might turn be given as an intravenous infusion over 15 minutes.
their stools a darker color. Oral iron salts are absorbed best on IDA patients receiving supplemental iron generally respond
an empty stomach but are better tolerated when taken with with a reticulocytosis within 7-10 days of initiating treatment.
foods. Ascorbic acid can facilitate iron absorption, but its addi- A hemoglobin response generally occurs within 2 weeks but
tion to the replacement regimen is not clearly cost effective and may take longer to fully correct, and serum ferritin should cor-
may increase the adverse effects of iron replacement therapy. rect once additional iron (beyond that to correct the hemoglo-
An alternative approach is to instruct patients to take oral iron bin) accumulates to replete body stores. The failure to respond
supplements with orange juice. Antacids, the tannins found in to oral iron should prompt consideration of ongoing bleeding,
tea, calcium supplementation, bran, and whole grains, if taken poor patient compliance, poor iron absorption, inadequate
concurrently with oral iron, can all decrease iron absorption. replacement dosing, or appropriateness of the diagnosis.
Treatment with oral iron should continue for ~3 months after
the hemoglobin normalizes to replete iron stores.
Oral iron supplementation is the preferred replacement Key points
route in most uncomplicated cases of iron deficiency. Paren-
• Iron deficiency is the most common cause of anemia
teral iron should be given intravenously (intramuscular is
worldwide, and its diagnosis requires a search for the underlying
painful) and is indicated when there is an absolute noncom-
etiology.
pliance with or intolerance to oral iron therapy, high iron
• Classical iron deficiency is characterized by a low corrected
requirements, or proven intestinal malabsorption. The long- reticulocyte count, hypochromic, microcytic anemia, and
term ramifications of IV iron therapy remain unstudied. elevated RDW.
Oral heme iron polypeptide (HIP), derived through the pro- • A ferritin of <10 μg/L in a premenopausal woman is diagnostic
teolytic digestion of porcine hemoglobin is another available of iron deficiency.
oral iron formulation. Heme-iron (derived from hemoglobin • Oral iron supplementation is the preferred replacement route
and myoglobin in animal food sources) is more efficiently in most uncomplicated cases of iron deficiency.
absorbed and via a different mechanism than elemental iron. • IDA is the most common nutritional deficiency associated with
Further study is warranted to determine whether HIP is more celiac disease.
efficacious and/or better tolerated than other oral iron for- • Failure to respond to oral iron should prompt consideration of
ongoing bleeding, inadequate replacement dosing, or poor
mulations; it is currently more expensive than oral iron salts.
absorption due to celiac disease, H. pylori infection, or atrophic
Multiple parenteral iron preparations are now available in
gastritis.
the United States. High-molecular-weight iron dextran is

Table 6-6 Iron studies in IDA versus anemia of chronic inflammation

IDA Anemia of chronic inflammation
Serum ferritin Decreased Normal or decreased

Serum iron Normal or decreased Normal or decreased
Total iron binding capacity or transferrin Increased Normal or decreased
% iron saturation Decreased (<10%-15%) Normal or decreased
MCV Decreased Normal or decreased
RDW Increased Normal
sTfR/log ferritin ratio >2 <1
Hepcidin (not currently clinically available) Suppressed Increased
IDA = iron deficiency anemia; MCV = mean corpuscular volume; RDW = red blood cell distribution width; sTfR = serum-soluble transferrin
receptor.
Normocytic anemias overlap and may not be discriminating, iron studies often are
used to differentiate AOCD from IDA. Typically, iron levels
Anemia of chronic inflammation are low to normal, iron-binding capacity is low to low nor-
(anemia of chronic disease) mal, and ferritin is normal or elevated in AOCD (Table 6-6).
In many, but not all, conditions, an elevated ESR or C-reac-
Anemia of chronic inflammation case
tive protein (CRP) may be a clue to the diagnosis of AOCD.
A 44-year-old woman is referred for evaluation of a hypop- Multiple processes are involved in the pathogenesis of AOCD.
roliferative normocytic anemia with a hemoglobin of 8 g/dL. Cytokines, such as tumor necrosis factor alpha (TNFα),
Her past medical history is significant for a mitral valve interleukin 1 (IL-1), interleukin 6 (IL-6), and interferons
replacement 1 year earlier. Recently, she has developed low- play a central role in the abnormalities observed in these
grade fevers, malaise, and generalized fatigue. Her examina- patients. These cytokines cause a reduction in the prolifera-
tion is remarkable for a temperature of 38.5 °C and a 2/6 tion of erythroid precursors in response to erythropoietin, a
systolic ejection murmur over the mitral valve. Laboratory decrease in the production of erythropoietin relative to the
evaluation reveals a decreased serum iron, low TIBC, and degree of anemia, and a moderate decrease in RBC survival.
low percent iron saturation. Serum ferritin is 55 ng/mL, and The hallmark of AOCD is an alteration in iron metabolism.
the erythrocyte sedimentation rate (ESR) is elevated at Inflammatory cytokines, especially IL-6, cause an increase in
92 mm/h. Blood cultures subsequently return positive for the hepatic synthesis of hepcidin, the key regulator of iron
α-hemolytic streptococci. Transesophageal echocardiogram transmembrane transport. Hepcidin acts by binding the iron
confirms subacute bacterial endocarditis of the prosthetic export protein, ferroportin, leading to its degradation; this
mitral valve. The patient is treated with penicillin and genta- inhibits dietary iron absorption and macrophage iron recy-
micin. Four weeks later, the hemoglobin increases to 11 g/dL. cling. This is reflected in low plasma iron levels and results in
Anemia is common in patients with chronic inflamma- iron-restricted erythropoiesis.
tory conditions such as malignancy, autoimmune diseases, In infants and children, inflammatory anemia does not
chronic infections, and chronic kidney disease. The resulting require the presence of an underlying chronic inflammatory
anemia is termed the anemia of chronic inflammation or the disorder. Even minor bacterial or viral infections, when recur-
anemia of chronic disease (AOCD). It is now recognized that rent, can cause a mild normocytic anemia with blunted retic-
patients with illnesses not traditionally thought to be inflam- ulocyte response within a few weeks. Their pathophysiologies
matory in origin, such as trauma, postsurgery, and prolonged likely mirror AOCD. This form of anemia of inflammation is
critical illness, also may develop AOCD. AOCD is generally a self-limited and resolves when the infant is free of infection.
sign of underlying disease activity and a thorough search for In most patients, AOCD is mild and improves with the
an underlying disorder should be performed when diagnos- treatment of the underlying disorder. Patients may have con-
ing AOCD as the cause of anemia. Patients with AOCD typi- comitant true iron deficiency or functional iron deficiency. If
cally have hemoglobin levels in the range of 7 to 11 g/dL. treatment becomes necessary, erythropoiesis stimulating
AOCD is characterized as a low corrected reticulocyte count, agents (ESAs) have been shown to be beneficial in some
normochromic, and normocytic anemia. Over time, how- patients +/− supplemental iron. Future treatment options for
ever, the anemia may become more severe with microcytic AOCD could involve decreasing hepcidin levels. The most
and hypochromic indices. Although laboratory values experience in using ESAs and supplemental iron to correct an

Normocytic anemias | 125
acquired underproduction anemia is in patients with chronic Multiple additional factors contribute to the anemia asso-
kidney disease. AOCD is one pathophysiology contributing ciated with CKD. Secondary hyperparathyroidism may
to their anemia (discussed further in the next section). suppress hematopoiesis. Uremia both impairs RBC deform-
ability, lending the cells more susceptible to mechanical
Key points destruction and clearance by macrophages and impairs RBC
enzyme activities (ie, enzymes involved in the hexose-
• AOCD is the most common cause of anemia in patients with monophosphate shunt and ATPase), resulting in increased
underlying inflammatory diseases.
susceptibility to oxidative stress and abnormal membrane
• AOCD is characterized by a low corrected reticulocyte count,
permeability. Hemodialysis may introduce RBC toxins (ie,
normocytic, or microcytic anemia with decreased serum iron, a
copper, formaldehyde, chlorine, nitrates, and chloramines),
decreased concentration of transferrin (or TIBC), and a de-
creased transferrin saturation with normal or increased serum
which can damage RBCs and decrease survival, and the
ferritin. dialysis procedure itself can cause RBC fragmentation.
• The pathophysiology of AOCD is multifactorial, but the Anemia associated with CKD is usually normochromic
sequestration of iron secondary to elevated hepcidin levels plays and normocytic unless complicated by iron deficiency or
a central role. vitamin deficiencies. The reticulocyte count is low. The
• Treatment should be directed at the underlying medical peripheral smear is often normal, but in rare patients with
condition. severe kidney failure, echinocytes can be seen (these cells are
characterized by irregular broad-based short blunt projec-
tions of the RBC membrane).
Anemia associated with chronic kidney disease The bone marrow usually is normocellular, but a hypo-
Anemia associated with chronic kidney disease case cellular marrow with relative erythroid hypoplasia and
marrow fibrosis (osteitis fibrosa) related to secondary
A 71-year-old woman presents to her primary care physician hyperparathyroidism has been described. Iron studies may
with increasing dyspnea on exertion. She is found to have a be normal but may show low serum iron levels accompa-
hypoproliferative, normocytic anemia (hemoglobin 9.5 g/dL), nied by low serum transferrin levels and high ferritin, as
and a creatinine of 2.2 mg/dL. Iron studies were normal. She seen in AOCD.
was started on an ESA along with an oral iron supplement. It is well established that recombinant human erythropoi-
Within 4 weeks, she had good clinical response; however, 2 etin and other ESAs improve the anemia associated with
months later she returns with recurrent exertional dyspnea. CKD. More recent studies have shown that the use of ESAs in
Laboratory values reveal a hemoglobin of 9.7 g/dL and an CKD patients to target normal or near-normal hemoglobin
MCV of 77 fL. Iron studies are consistent with IDA. Intrave- values is associated with increased risk of adverse cardiovas-
nous iron dextran is administered with good clinical response. cular events and mortality and modest or no improvement in
The anemia associated with chronic kidney disease (CKD) patient-reported fatigue compared to lower target hemoglo-
is primarily due to underproduction of erythropoietin sec- bin levels. Randomized control studies are needed to define
ondary to a decrease in the number of renal cortical cells the optimal hemoglobin level at which ESA therapy should
available to produce the hormone. The accumulation of ure- be initiated, the appropriate hemoglobin target, and the best
mic toxins also may cause reduced synthesis of erythro- use of IV iron to reduce ESA dose requirements. The National
poietin. Iron-restricted erythropoiesis also contributes Kidney Foundation Kidney Disease Outcomes Quality Ini-
significantly to the anemia in CKD and additionally accounts tiative (NKF KDOQI) and Kidney Disease: Improving Global
for the resistance to erythropoiesis stimulating agents (ESAs) Outcomes (KDIGO) provide nephrologists with guidelines
observed in these patients. The mechanism of iron-restricted on the management of anemia associated with CKD based
erythropoiesis in CKD is multifactorial, and contributing on available evidence and expert opinion.
factors include increased blood loss from phlebotomy or the
dialysis procedure, uremic platelet dysfunction and associ-
Key points
ated bleeding, and increased iron utilization from ESA ther-
apy. Hepcidin excess resulting in impaired dietary iron • Anemia associated with CKD is primarily due to an erythropoi-
absorption and impaired iron release from body stores is etin deficiency.
also present. Hepcidin excess in CKD patients may be • ESAs are effective treatment of anemia associated with chronic
due to reduced renal clearance of this small peptide hor- kidney disease.
mone and/or increased inflammatory-mediated hepcidin • ESAs and IV iron treatment of anemia associated with CKD
require careful consideration of the potential benefits and
expression. Hepcidin levels in CKD are further influenced
potential harms to the individual patient.
by iron and ESA administration.

Macrocytic anemias daily parenteral cobalamin replacement, with symptomatic

improvement and brisk reticulocytosis noted within 1
Megaloblastic anemias week.
Vitamin B12 (cobalamin) functions as an essential coen-
Megaloblastic anemias are characterized by a low corrected zyme for two enzymes in the human body: cytoplasmic
reticulocyte count, macrocytosis (often macro-ovalocytes) methionine synthase, which catalyzes methylation of homo-
and marked variation in the size and shape of red blood cysteine to methionine; and methylmalonyl-CoA mutase,
cells, hypersegmented neutrophils, megaloblastic changes in which catalyzes the conversion of methylmalonyl-CoA to
marrow morphology, and occasionally pancytopenia. These succinyl-CoA in the mitochondria. The succinyl-CoA
findings all reflect impaired DNA synthesis in hematopoi- formed in the latter reaction enters the Krebs cycle. The for-
etic cells. Megaloblastic morphologic changes in the mar- mer reaction is linked to folate metabolism, as the methyl
row result from the dyssynchrony between nuclear and group transferred to homocysteine is provided by the con-
cytoplasmic maturation and include a hypercellular mar- version of 5-methyl tetrahydrofolate to tetrahydrofolate.
row with an erythroid predominance (reversed Tetrahydrofolate is essential for purine and pyrimidine
myeloid:erythroid ratio) and the presence of giant pronor- synthesis.
moblasts and giant metamyelocytes. RBC development is Humans are completely dependent on a dietary source of
ineffective as defined by an imbalance between the amount cobalamin, namely animal food sources. In the stomach,
of iron that is endocytosed by marrow erythroblasts and the cobalamin is released from food components and is bound
amount of iron incorporated into circulating erythrocytes, to haptocorrin, a protein present in saliva and gastric fluids,
implying death of cells within the marrow. This is clinically which likely protects the vitamin from degradation in the
reflected by an elevated lactic dehydrogenase (LDH) and acidic stomach. In the duodenum, pancreatic enzymes
unconjugated bilirubin and low haptoglobin and occasion- degrade haptocorrin, and cobalamin subsequently binds to
ally the appearance of red cell fragments on peripheral intrinsic factor. Intrinsic factor is synthesized and secreted by
smear. There are case reports of severe megaloblastic ane- the parietal cells of the stomach. In the terminal ilium,
mia clinically mimicking thrombotic thrombocytopenic intrinsic factor-bound cobalamin is endocytosed by the
purpura (TTP); a higher platelet count and lower reticulo- receptor complex, cubam. Inside the ileal enterocyte, the
cyte count are two clinical clues that support a diagnosis of intrinsic factor is degraded and the released cobalamin exits
a megaloblastic anemia over TTP. Megaloblastic anemias the basolateral surface of the cell via a transporter. In the
and myelodysplastic syndromes also share a number of clin- plasma, cobalamin binds to transcobalamin II for delivery to
ical findings, and MDS should be included in the differen- tissues.
tial diagnosis during the work-up. Additionally, the Cobalamin deficiency results most commonly from
hypercellular bone marrow with giant pronormoblasts in abnormal intestinal absorption, or, in rare cases, from
megaloblastic anemias may be so alarming as to lead to an insufficient dietary intake or defects in bodily transport.
incorrect diagnosis of acute leukemia. Cobalamin and folate Select causes of cobalamin deficiency are listed in Table 6-7.
deficiency are the most common causes of megaloblastic Owing to efficient enterohepatic circulation, as well as
anemias. reuptake in the kidney, cobalamin is retained in the body
for long periods, and thus dietary deficiency requires years
Vitamin B12 deficiency to develop.
The most common cause of symptomatic cobalamin defi-
Cobalamin deficiency case
ciency is pernicious anemia (PA). PA is an autoimmune dis-
A 75-year-old man is referred by his urologist for investiga- order in which autoimmune gastritis leads to atrophy of the
tion of anemia. He has a diagnosis of transitional cell carci- mucosa of the body and of the fundus of the stomach that
noma of the bladder for which he has been treated by reduces the number of parietal cells that produce intrinsic
transurethral resection and an intravesical bacillus factor, necessary for the absorption of vitamin B12, which, in
Calmette-Guérin (BCG) vaccine. He has type 2 diabetes turn, is required for erythropoiesis and myelin synthesis. PA
treated with metformin. On examination, he is pale, has is frequently associated with other autoimmune disorders
mild peripheral edema and minimal loss of position and (eg, type 1 diabetes, autoimmune thyroiditis, primary hyper-
vibratory senses in the feet bilaterally. Laboratory evalua- parathyroidism, and vitiligo). The diagnosis relies on the
tion reveals a hemoglobin of 7.1 g/dL, MCV of 109 fL, neu- demonstration of a megaloblastic anemia, low serum vita-
trophil count of 960/mL, and platelet count of 35,000/mL. min B12 levels, and the presence of antibodies to gastric pari-
A serum cobalamin level is 72 pg/mL, serum folate is nor- etal cells or intrinsic factor. These antibodies cause the
mal, and RBC folate is mildly depressed. He is started on mucosal atrophy. Stomach biopsy or serum biomarkers

Macrocytic anemias | 127
Table 6-7 Select causes of vitamin B12 deficiency Diagnosis and treatment
Impaired absorption
Cobalamin deficiency can present insidiously with unex-
Deficiency of intrinsic factor or IF-bound vitamin B12 uptake
plained anemia, neuropsychiatric symptoms, or GI manifes-
Pernicious anemia
tations, including swollen or sore tongue (glossitis), anorexia,
Congenital intrinsic factor deficiency
Gastric bypass surgery
and diarrhea. Neurologic symptoms include paresthesias,
Decreased ileal absorption of vitamin B12 (Imerslund-Gräsbeck unsteady gait or clumsiness, motor weakness progressing to
syndrome) paralysis, and psychiatric symptoms such as mania, para-
Hypochlorhydia (impairs release of B12 from dietary proteins) noia, and irritability. In the nervous system, cobalamin defi-
Age ciency results in defective myelin synthesis, leading to central
Gastric atrophy (Helicobacter pylori or autoimmune gastritis) and peripheral nervous system dysfunction. The spinal cord
Medications (proton-pump inhibitors or H2 antagonists) manifestation, called subacute combined degeneration,
Inadequate pancreatic protease (vitamin B12 remains sequestered affects the posterior and lateral columns of the spinal cord
by haptocorrin) and results in loss of vibratory sense and proprioception and
Intestinal competition for host vitamin B12 (tapeworm
manifests on MRI imaging as a symmetrical abnormally
Diphyllobothrium latum)
increased T2 signal intensity in the posterior or posterior
Ileal resection or bypass
and lateral columns, commonly confined to the cervical and
Ileal dysfunction (Crohn disease, celiac disease, intestinal
lymphoma, bacterial overgrowth from blind loop syndrome)
thoracic spinal cord. Brain involvement in cobalamin defi-
Medications (metformin [mechanism unknown], nitrous oxide ciency on MRI has also been reported.
abuse) It is critical to recognize these signs and symptoms early
Insufficient dietary intake (strict vegans, some vegetarians, and
to avoid irreversible neurologic dysfunction. Folate replace-
breastfed infants of vitamin B12–deficient mothers) ment alone may improve the anemia in patients with
cobalamin deficiency, thereby masking that aspect of the
Defects in bodily transport
underlying cobalamin deficiency and allowing progres-
Congenital disorders of vitamin B12 transport (defects in cubam,
transcobalamin, others)
sion of the neurologic deficit. Therefore, cobalamin levels
always should be measured before initiation of folate in
patients at risk for concomitant cobalamin deficiency. Addi-
consistent with chronic atrophic gastritis are not required tionally, it should be noted that there is an inverse relation-
for the clinical diagnosis of PA. The gastric enzyme H+/K+- ship between neurological and hematological involvement in
ATPase is the target antigen recognized by parietal cell cobalamin deficiency. Why some patients develop one over
antibodies. the other remains unclear. Thus, macrocytic anemia cannot
Some data suggest an involvement of long-standing be used as the sole criterion for pursuing the diagnosis.
H. pylori infection in the pathogenesis of PA and atrophic No gold standard testing exists for diagnosing cobalamin
body gastritis. One hypothesis is that the active infectious deficiency, and each laboratory test has its disadvantages.
process is replaced gradually by an autoimmune process Adequate cobalamin treatment is the safest approach if the
mediated by autoreactive gastric T-cells that recognize H+/ clinical presentation and laboratory studies are confusing.
K+-ATPase and H. pylori antigens, which culminates in a Complete resolution of symptoms with therapy supports the
burned-out infection and irreversible damage to mucosa. It diagnosis. A serum cobalamin assay is the standard initial
remains subject to debate whether PA should be included routine diagnostic test. It quantifies all forms of cobalamin
among the long-term consequences of H. pylori gastritis, and in serum. It is a widely available, inexpensive, automated
thus the need to test for H. pylori in the work-up of PA method based on intrinsic factor binding of cobalamin and
remains controversial. immune-chemoluminescence. Results are highly variable
PA patients are at risk for the development of gastric ade- with significant intraindividual variation and large absolute
nocarcinoma and carcinoid tumors. Endoscopy at presenta- differences in results on repeat testing. Additionally, the assay
tion is commonly done; however, data are insufficient to lacks sensitivity and specificity. Spuriously high cobalamin
support routine subsequent endoscopic surveillance of levels have been reported in patients with PA, which has been
these patients. Follow-up should be individualized to the attributed to assay interference by high levels of antibodies
patient. against intrinsic factor. These authors use a cut-off level of
Nitrous oxide is associated with an acute megaloblastic <200 pg/mL or one derived from local reference range as evi-
anemia secondary to impaired cobalamin metabolism, and dence of cobalamin deficiency in the correct clinical context.
abuse of this compound has been associated with psychosis Methylmalonic acid (MMA) and total homocysteine (HCY)
and other neurologic defects. generally are considered to be more sensitive indicators of

early deficiency, and elevations in one or both have been increases to 60%-80% with increasing disease duration. As
shown to correlate with clinical response to cobalamin ther- cobalamin therapy can cause a false-positive result on intrin-
apy; serum levels of both MMA and HCY become elevated sic factor antibody testing, providers should wait at least a
before cobalamin levels fall below the lower limits of the nor- week after a cobalamin injection before sending this testing.
mal range. Measuring MMA (±HCY) levels may be useful in Parietal cell antibodies are present in 80%-90% of PA
patients with symptoms consistent with cobalamin defi- patients, especially in the early stages of the disease. In the
ciency and in whom serum cobalamin levels are equivocal, later stages of the disease, the incidence of parietal cell anti-
patients with atypical clinical findings in whom cobalamin bodies decreases due to the progression of the autoimmune
deficiency must be ruled out, and asymptomatic patients gastritis and a loss of gastric parietal cell mass. Parietal cell
accidentally found to have a low cobalamin level. HCY levels antibodies lack specificity and can also be found at low fre-
lack specificity and can be elevated in patients with folate quency in other autoimmune diseases (ie, Hashimoto’s thy-
deficiency, renal dysfunction, and other settings. roiditis or type 1 diabetes) or in elderly subjects. Historically,
There is debate on the clinical importance of identifying the Schilling test was used to measure cobalamin absorption,
patients without overt cobalamin deficiency symptoms (ie, but this test is no longer available in most centers.
neurologic or hematologic findings) who have laboratory Cobalamin deficiency is a rare and treatable cause of failure
tests suggesting cobalamin deficiency, so-called subclinical to thrive and delayed development in infants. Its long-term
cobalamin deficiency. The majority of patients with subclini- developmental consequences remain unknown. In developed
cal cobalamin deficiency do not progress to symptomatic countries, the deficiency can occur in infants exclusively
cobalamin deficiency. It remains uncertain whether these breastfed whose mothers are deficient (eg, unrecognized PA,
patients may have subtle and clinically unrecognized symp- vegetarians, or vegans), causing low cobalamin body stores in
toms of cobalamin deficiency and controversial whether they the infant at birth and inadequate amounts of cobalamin
should be treated or followed closely. This may reflect a lack in the breast milk. Signs and symptoms often present between
of uniform diagnostic criteria for subclinical cobalamin defi- the ages of 4 and 12 months and include failure to thrive, leth-
ciency and the limitations in laboratory testing for cobala- argy, hypotonia, and arrested or regressed developmental
min deficiency. Currently, routine screening of asymptomatic skills. It can rarely cause seizures or brain atrophy on imaging.
individuals for cobalamin deficiency is not recommended. Infants often demonstrate abnormal movements, including
Low cobalamin levels alone (without overt megaloblastic tremor, myoclonus, and choreoathetoid movements.
anemia or typical neurological symptoms) also can be seen Rare cases of cobalamin deficiency due to a congenital
in association with a variety of other clinical conditions, defect in intrinsic factor secretion from parietal cells (ie, con-
including pregnancy, because of changes in protein binding, genital PA) present around 18-36 months of age, after the
folate deficiency, and use of certain drugs (eg, oral contra- depletion of fetal liver stores. Typical acquired PA also may
ceptives and metformin). True cobalamin deficiency in these present in children. The Imerslund-Gräsbeck syndrome is a
situations can be confirmed by elevations in MMA and HCY rare congenital defect in cobalamin absorption resulting
levels. Other conditions can cause an elevated level of HCY from mutations in the cubam receptor complex. In some
alone (hypothyroidism, vitamin B6 deficiency), or MMA cases, this autosomal recessive disorder also causes protein-
alone (intestinal overgrowth), or of both (renal failure). uria, which is explained by the additional function of cubam
In patients in whom cobalamin deficiency is clinically sus- in the renal reabsorption of some filtered proteins. Transco-
pected, a serum cobalamin measurement is a reasonable ini- balamin II deficiency is inherited as an autosomal recessive
tial test and a level <200 pg/mL supports the diagnosis. It is trait that presents in early infancy with severe megaloblastic
important to note that given the significant diagnostic limi- anemia despite normal intrinsic factor secretion, cobalamin
tations of serum cobalamin measurements, values above this absorption, and cobalamin levels.
cut-off do not exclude the diagnosis and hematologists need Patients with cobalamin deficiency can be treated with
to carefully consider the clinical scenario of each case. If clin- parenteral or oral cobalamin replacement. Parenteral ther-
ical suspicion persists, metabolite testing with either MMA apy is recommended for patients with significant symptom-
or HCY may be reasonable. Once cobalamin deficiency is atology. Intramuscular cobalamin is given in doses of 1,000
diagnosed, patients should be evaluated for the underlying μg/day (up to 150 μg is retained from each injection by most
cause. Some authorities recommend testing for intrinsic fac- patients) for 1 week, and then 1,000 μg/week for 4 weeks, and
tor antibodies alone in patients with evidence of low cobala- then 1,000 μg/month or less frequently. Alternative dosing
min; others additionally send testing for nonspecific serum regiments can be used. Excess cobalamin is excreted in the
gastrin or pepsinogen levels in each cobalamin-deficient urine, so toxicity due to excessive vitamin replacement does
patient (predicted to be elevated and low in deficient patients, not occur. The observation of intrinsic factor-unrelated dif-
respectively). The incidence of intrinsic factor antibodies fusion of ~1.2% of any oral cobalamin dose provided the

rationale for studies that documented that oral cobalamin involved in one carbon metabolism and thus, is important in
may be a safe and effective treatment in some patients, even the synthesis of purines, pyrimidines, and the metabolism of
when intrinsic factor is present at low levels. The initial oral amino acids. Recently, a proton-coupled high-affinity folate
replacement dose begins at 1,000-2,000 μg/day. Patients transport protein expressed in the duodenum and jejunum
treated with oral cobalamin should be observed carefully to (PCFT/HCP1) was identified. A loss-of-function mutation
ensure that symptoms of anemia improve. After cobalamin in the gene encoding PCFT/HCP1 results in a syndrome of
replacement is commenced, some patients will become iron hereditary folate malabsorption.
deficient because of iron uptake by developing erythroid Folate deficiency may result from decreased dietary intake,
cells. This requires replacement. impaired absorption, or increased utilization (Table 6-8).
Following cobalamin replacement, the marrow shows res- The major cause of folate deficiency is decreased dietary
olution of megaloblastic changes within hours and reticulo- intake rather than impaired absorption, which is the major
cytes appear in the peripheral blood, usually peaking cause of cobalamin deficiency. Green leafy vegetables (eg,
approximately 1 week after starting replacement therapy. spinach and turnip greens), fruits (eg, citrus fruits and
Hypersegmentation of neutrophils may persist for up to 2 juices), and dried beans and peas are all natural sources of
weeks. Blood counts and MCV return to normal in 2-3 folate. The implementation of folic acid fortification of
months. Neurologic abnormalities usually resolve within 3 grains in many countries has drastically reduced the preva-
months; although in some patients, this may take as long as lence of folate deficiencies in those countries. The FDA-
6-12 months. In some individuals, the neurological deficits recommended daily dietary folate equivalent is 400 ug.
are irreversible. Folate deficiency due to inadequate dietary intake can
develop in months since body stores are not extensive. Folate
Key points supplementation should be part of routine prenatal care to
reduce the risks of neural tube defects. Folate supplementa-
• The most common cause of cobalamin deficiency is impaired tion should be considered in other patients with increased
absorption.
folate requirements (eg, chronic hemolytic anemia patients).
• PA is the most common cause of impaired cobalamin
absorption, resulting in symptomatic deficiency.
• Both cobalamin and folate deficiency cause a megaloblastic Diagnosis and treatment
anemia; however, neuropsychiatric symptoms are seen only in
cobalamin deficiency. The hematologic manifestations of folate deficiency are indis-
• Parenteral therapy is recommended for patients with any tinguishable from cobalamin deficiency. Folate deficiency
neuropsychiatric symptoms.
• Subclinical cobalamin deficiency (defined by elevated MMA
and HCY levels with no clinical signs or symptoms) is of Table 6-8 Select causes of folate deficiency
uncertain significance.
Impaired absorption
Intestinal dysfunction (Crohn disease, celiac disease)
Congenital abnormality in intestinal folate transporter (mutations
Folate deficiency in PCFT)
Folate deficiency case Insufficient dietary intake
Poor intake of fruits and vegetables or prolonged cooking of these
A 55-year-old man presents for routine physical examina- foods
tion. He complains of fatigue and shortness of breath. He Alcoholism (alcohol also increases renal folate excretion and
admits to daily excessive alcohol consumption since he lost impairs its intracellular metabolism)
his job 6 months ago. Physical examination reveals pallor, Increased requirements
glossitis, flow murmur, and a normal neurological examina- Increased cellular proliferation
tion. Laboratory evaluation reveals a hemoglobin of 7.1 g/dL, Pregnancy
MCV of 130 fL, neutrophil count of 1,000/mL, and platelet Lactation
count of 55,000/mL. A serum folate level is 1 ng/mL, cobala- Hemolytic anemia (sickle cell anemia, warm autoimmune
min level is 350 pg/mL. He is enrolled in an alcoholic treat- hemolytic anemia)
ment program and started on 2 mg of daily oral folic acid Malignancies (associated with a high proliferative rate)
replacement with symptomatic improvement and brisk Exfoliative dermatitis
reticulocytosis noted within 2 weeks. Hemodialysis
Medication that affect folate metabolism or possibly absorption
Folate exists in nature as a conjugate with glutamic acid
(methotrexate, phenytoin, carbamazepine)
residues. Folate, when reduced to tertrahydrofolate, is

does not cause subacute combined degeneration. Folate defi- Physical examination is significant for skin pallor and pale
ciency is strongly implicated in increasing the incidence of conjunctivae. Laboratory evaluation reveals hemoglobin of
neural tube defects in the fetus. Plasma (or serum) folate 6.4 g/dL, MCV of 99 fL, and corrected reticulocyte count of
undergoes diurnal changes related to recent food intake, 0.3%. White blood cell and platelet counts are normal. Bone
which limits the diagnostic usefulness of the assay. If the marrow examination reveals a maturation arrest at the pro-
serum folate is >4 ng/mL, folate deficiency is unlikely. If the erythroblast stage. Flow cytometry does not reveal a lympho-
serum folate concentration is <2 ng/mL, folate deficiency is proliferative disorder, and cytogenetic evaluation results are
likely. Alternatively, RBC folate levels remain constant from normal. Computed tomography (CT) scan of the chest fails
day to day and accurately reflect the average folate content of to identify a thymoma. Prednisone 1 mg/kg daily is pre-
the long-lived circulating RBC population. However, care scribed, and within 2 weeks, a partial response is seen. After
must be taken when interpreting this laboratory measure- 6 weeks, a complete response is seen, and a slow taper of
ment, as RBC folate levels may also be low in people with prednisone is begun. The patient relapses after prednisone
cobalamin deficiency, and additionally, the RBC folate assay withdrawal. She is begun on cyclosporine with a gradual, but
methodology is problematic. Folate deficiency also results in complete, response in her blood counts.
high levels of HCY (but not MMA). Folate-deficient people Pure red cell aplasia (PRCA) is characterized by severe
are treated with folic acid (1-5 mg/day) for 1-4 months, or normochromic, normocytic, or macrocytic anemia; reticu-
until complete hematologic recovery occurs. Folic acid can locytopenia; and the absence of hemoglobin-containing
partially reverse the hematologic abnormalities of cobalamin cells (some define as hemoglobinized cells comprise <3% of
deficiency while the neurologic symptoms resulting from the nucleated marrow cells) in an otherwise-normal mar-
cobalamin deficiency progress. Thus, cobalamin deficiency row aspirate or with maturation arrest at the proerythro-
should be excluded before initiating folate replacement ther- blast stage (Figure 6-3). If the PRCA is secondary to large
apy. Folate is inexpensive and effective even in persons with granular lymphocyte leukemia or another lymphoprolifer-
malabsorption. ative disorder, the marrow may show infiltration with
lymphocytes.
Key points PRCA occurs as either an acquired or congenital (Diamond-
Blackfan anemia; see Chapter 15) disorder. Acquired PRCA
• The most common cause of folate deficiency is decreased is further classified as primary or secondary, depending on
dietary intake.
the absence or presence of an associated disease, infection,
• Folate supplementation should be part of routine prenatal
or drug (Table 6-9). Alternatively, acquired PRCA can be
care.
classified by the pathophysiology of the anemia. There are
• Patients with chronic hemolytic anemia should receive daily
folate supplementation.
three distinct mechanisms by which erythropoiesis can fail.
• HCY is elevated and MMA is normal in folate deficiency. In most cases of PRCA, an aberrant immune response leads
• Cobalamin deficiency should be ruled out before treatment to suppression of RBC development: erythroid progenitor
with folate. cells are intrinsically normal, but their differentiation is
inhibited by a humoral or T-lymphocyte-mediated mecha-
nism. This can occur idiopathically or associated with an
Other causes of megaloblastic anemia underlying disorder. In about 10% of cases, PRCA results
from chronic parvovirus B19 infection, and in rare cases,
In addition to folate and cobalamin deficiency, there are
PRCA is the initial clinical manifestation of an MDS.
other rarer causes of megaloblastic anemias. Drugs that affect
DNA synthesis cause megaloblastic anemia (5-fluorouracil–
pyrimidine analog, azathioprine–purine analog, methotrexate–
(a) (b)
antifolate, and hydroxyurea, zidovudine, and some
anticonvulsants that also likely inhibit DNA synthesis).
Acquired pure red cell aplasia
Acquired pure red cell aplasia case
A 64-year-old female presents with fatigue and dyspnea on Figure 6-3 Pure red cell aplasia bone marrow aspirate with excess
exertion, which has been progressive over the last 2 months. proerythroblasts. Arrows indicate proerythroblasts. Wright-Giemsa–
She is on no medications and has no significant past medical stained marrow aspirates from a normal patient (a) and a pure red cell
history. Previous blood counts reportedly have been normal. aplasia patient (b) are shown.

Table 6-9 Classification of pure red cell aplasia When the anemia is severe or life threatening, treatment
Congenital PRCA (Diamond-Blackfan anemia) with plasma exchange using donor-type plasma and high
doses of recombinant human ESAs are effective in some
Acquired PRCA
patients.
Primary PRCA (likely immune-mediated mechanism)
Parvovirus B19 is a common infection in children and
Transient erythroblastopenia of childhood
Idiopathic
causes erythema infectiosum (ie, fifth disease). More than
Secondary PRCA (immune consequence of an underlying disorder) 75% of adults >50 years old have neutralizing antibodies
Thymoma: post-ABO and autoimmune against this virus. In all infected patients, the virus binds to
Hematologic malignancies (eg, chronic lymphocytic leukemia, large the blood group P antigen expressed on erythroid progeni-
granular lymphocyte leukemia, multiple myeloma) tors and is cytotoxic to the infected cells. In patients with
Solid tumors (eg, stomach, breast, lung, renal cell carcinomas) normal immunity, parvovirus persists at a high titer in the
Infections (eg, HIV, EBV, viral hepatitis) blood and marrow for 2-3 weeks and is then cleared. Because
Collagen vascular diseases the normal life span of the RBC is ~120 days, infection does
Drugs and chemicals (erythropoietin antibodies that developed in not result in a significant decrease in hemoglobin. Alterna-
some patients treated with ESAs) tively, clinically significant anemia develops in immunosup-
Post-ABO incompatible bone marrow transplantation pressed patients (eg, patients with HIV or organ transplant
Autoimmune chronic hepatitis or hypothyroidism recipients) whose immune systems are unable to clear the
Parvovirus B19 (virus is directly cytotoxic to red blood cell infection or in patients with shortened RBC survival (eg,
precursors) sickle cell anemia or hereditary spherocytosis) in whom the
Myelodysplastic syndrome (hematopoietic stem cell that is unable to anemic presentation is sometimes termed a “transient aplas-
differentiate along the erythroid lineage) tic crisis.”
EBV = Epstein-Barr virus; PRCA = pure red cell aplasia; Many different drugs have been reported to cause PRCA,
ESAs = erythropoietin-stimulating agents. and discontinuing the drug sometimes resolves the PRCA.
PRCA has been described as a result of the development of
antierythropoietin antibodies during treatment with recom-
Several causes of acquired PRCA are reviewed here. Tran- binant human ESAs. Although rare, this syndrome has
sient erythroblastopenia of childhood is an acquired PRCA occurred primarily after subcutaneous administration of
observed in infants and young children. Affected patients are Eprex (Janssen-Ortho, Toronto, Ontario, Canada), an
usually between 6 and 36 months of age and present only erythopoietin-α product marketed outside of the United
with the insidious onset of pallor. The degree of anemia is States. The number of ESA-associated PRCA cases peaked in
variable. The differential diagnosis includes Diamond- 2001 and has since declined because of changes in the manu-
Blackfan anemia and parvovirus B19 infection. Although the facturing, distribution, storage, and administration of Eprex.
pathophysiology is not well characterized, most cases appear Because LGL leukemia may be present even in the absence of
to be due to an antibody (IgG) directed against erythroblasts. significant lymphocytosis, it is recommended that patients
The condition resolves spontaneously within several months with idiopathic PRCA undergo lymphocyte immunopheno-
with no sequelae. typing to look for LGL leukemia.
Immunologic causes of acquired PRCA may be idiopathic
in origin or secondary to an underlying disease. PRCA devel-
Diagnosis and treatment
ops in ~5% of patients with thymoma, and, conversely, thy-
moma occurs in ~10% of patients presenting with PRCA. Acquired PRCA presents with symptoms related to the sever-
The response to thymectomy in these cases is variable; a ity of the anemia. Apart from pallor, physical examination in
minority of patients achieves a complete remission after thy- acquired primary PRCA often is normal. In acquired second-
moma resection. PRCA also may be found in patients with ary PRCA, findings related to the underlying disease such as
underlying lymphoproliferative disorders (eg, large granular hepatomegaly, splenomegaly, or lymphadenopathy may be
lymphocyte leukemia, chronic lymphocytic leukemia). present.
Large granular lymphocyte (LGL) leukemia may be the most Diagnosis of acquired PRCA is first suggested by finding a
common underlying cause of acquired secondary PRCA. normochromic, normocytic, or macrocytic anemia with
Approximately 20% of patients receiving ABO-mismatched reticulocytopenia. The white blood cell and platelet counts
bone marrow transplants develop a prolonged RBC aplasia are generally normal. Bone marrow biopsy and aspirate
due to recipient isohemagglutinins, especially anti-A, against establish the diagnosis. In parvovirus B19 infection, the mar-
the donor RBCs; generally, the condition improves over row aspirate may show giant pronormoblasts. Routine
time or with the development of graft-versus-host disease. karyotype and interphase fluorescence in situ hybridization

(IFISH) panel for MDS should be included as part of the ini- Anemia associated with liver disease
tial workup to evaluate for an underlying MDS. A careful
Anemia and other hematopoietic abnormalities frequently
history and physical exam should be used to guide further
are seen in patients with liver disease. The true incidence of
diagnostic testing. Additional studies to consider are a CT
anemia depends on the cause of liver disease, but it has been
scan of the chest to evaluate for thymoma, erythropoietin
reported in up to 75% of patients with chronic liver disease.
level, and parvovirus B19 DNA testing by polymerase chain
The etiology of anemia is multifactorial and likely reflects
reaction (serologies are not adequate).
underproduction, blood loss, and increased destruction of
PRCA caused by parvovirus B19 is treated with normal
RBCs. In alcoholic liver disease, concomitant folate defi-
pooled serum IgG, which provides specific antibodies to
ciency may be seen and should be evaluated. Alcohol-
clear the infection. PRCA associated with thymoma may
induced pancreatitis may lead to decreased vitamin B12
respond to thymectomy. There does not appear to be any
absorption. Ethanol and its metabolites have been shown to
benefit to the removal of a normal thymus in patients with
inhibit erythroid production directly and may lead to acute
PRCA who do not have a thymoma or thymic hyperplasia
or chronic anemia even in the absence of severe liver disease.
identified.
In addition, erythropoietin production and erythropoiesis
Presumed immunologically mediated PRCA is treated
are suppressed by alcohol. Viral hepatitis may be associated
with sequential trials of immunosuppressive therapies
with PRCA. Hematological complications of combination
(eg, prednisone, cyclosporine, oral cyclophosphamide, myco-
therapy for chronic viral hepatitis include clinically signifi-
phenolate mofetil, horse anti-thymocyte globulin (ATG),
cant anemia, secondary to treatment with ribavirin and/or
alemtuzumab, rituximab), which ultimately leads to remis-
interferon. Ribavirin-induced hemolysis can be reversed by
sion in ~60%-70% of patients. No prospective randomized
reducing the dose of the drug or discontinuing it altogether.
clinical trial data exist to support the use of one immunosup-
Interferons may contribute to anemia by inducing bone
pressive agent over another, and we generally select an agent
marrow suppression.
based on any identified underlying disorder, and in idiopathic
GI blood loss is common in liver disease, especially in
cases, use prednisone or cyclosporine as first-line agents. A
patients with esophageal varices. Shortened RBC survival
~3-month trial of each immunosuppressive agent is reason-
also is noted in chronic liver disease. The decreased life span
able to assess for response to therapy. Responding patients
of RBCs in liver disease is at least partially explained by con-
generally are treated for 3-6 months before immunosuppres-
gestive splenomegaly, abnormal erythrocyte metabolism,
sion is slowly tapered. Many patients will relapse after with-
and alterations in RBC membrane lipids. Changes in choles-
drawal of therapy and will require a long-term approach to
terol composition lead to alterations in RBC surface area
immunosuppression, particularly if an underlying disorder
characterized by the target cells typically seen on review of
(lymphoproliferative disorder or collagen vascular disease)
the peripheral blood smear. Marked hemolytic anemia may
persists. Causes of death in nonresponding patients include
develop in alcoholics with relatively mild liver disease. This
infection, iron overload, or cardiovascular events.
condition, when accompanied by jaundice and hyperlipid-
Patients with severe symptomatic anemia are treated with
emia, is termed Zieve syndrome. This syndrome may be self-
transfusion therapy and face the associated risks (ie, iron
limiting if the patient abstains from alcohol. Marked spur
overload and alloantibody formation). Supplemental ESAs
cells are noted on the blood smear in this condition, which
have been used in certain instances with variable success, such
has also been termed spur cell anemia (Figure 6-4). In the
as after ABO-incompatible bone marrow transplantation.
presence of underlying cirrhosis, spur cell anemia is likely
irreversible without liver transplantation.
Key points The typical anemia of liver disease is usually mild to mod-
• There are three pathophysiologic mechanisms of PRCA:
erate but may become more severe as cirrhosis, portal hyper-
immune m ediated, myelodysplasia, and parvovirus B19 infection tension, and splenomegaly develop. The anemia is often
in a susceptible host. macrocytic, but the MCV rarely exceeds 115 fL in the absence
• Parvovirus B19 infection causes PRCA in all patients infected of megaloblastic changes in the bone marrow. The reticulo-
with the virus, but anemia only manifests in immunosuppressed cyte count is often minimally to moderately increased in
patients or patients with shortened red cell survival. liver disease but may be suppressed by alcohol or concomi-
• PRCA secondary to parvovirus B19 infection is treated with tant iron deficiency. More marked reticulocytosis may be
intravenous immunoglobulin. seen with hemorrhage or in spur cell anemia. Bone marrow
• In the absence of clear myelodysplasia or parvovirus B19 cellularity often is increased and erythroid hyperplasia is
infection, PRCA is treated with courses of immunosuppressive
observed. The peripheral blood smear often will show acan-
agents.
thocytes and target cells as the disease severity increases.

Other underproduction anemias | 133
Other underproduction anemias

Spur Cell Anemia
The underproduction anemias discussed in this section gen-

erally are not categorized by MCV.
Copper deficiency anemia
Copper is an essential trace element that plays a role in

numerous physiologic processes, including proliferation and
differentiation. Copper deficiency is rare in humans and gen-
erally is thought to result from inadequate intake (eg, patients
on total parenteral nutrition without copper supplementa-
tion) or absorption (eg, postbariatric surgery, celiac disease,
excessive zinc intake, congenital defect in copper transport
called Menkes disease). Copper deficiency has been reported
Figure 6-4 Spur cell anemia. Note the acanthocytes (also known as
to result in anemia, neutropenia, and, less commonly, throm-
spur cells and indicated with arrows) and target cells.
bocytopenia. A review of 40 patients with copper deficiency
unrelated to Wilson’s disease documented that 10 patients
(25%) underwent bariatric surgery, 14 patients (35%) had
Megaloblastosis may be seen in up to 20% of subjects. The gastric resection, and no cause for copper deficiency could be
treatment of anemia in liver disease is primarily supportive. identified in 12 other cases. The anemia has been reported
If present, iron, vitamin B12, and folate deficiencies should variably as microcytic, normocytic, or macrocytic. Impor-
be corrected. If ongoing hemolysis is noted, folate should be tantly, copper deficiency can mimic an acquired MDS and
supplemented. Alcohol and other toxins should be avoided. can manifest with a macrocytic anemia and neutropenia and
variable marrow morphology including ringed sideroblasts,
Sideroblastic anemias dyserythropoiesis, dysmyelopoeisis, cytoplasmic vacuoliza-
tion of both erythroid and myeloid precursors, and hemo-
The sideroblastic anemias are a heterogeneous group of con- siderin-laden plasma cells. In addition to the hematologic
genital and acquired hematologic disorders characterized by manifestations, copper deficiency can cause neurologic
the presence of ringed sideroblasts. Ringed sideroblasts are symptoms resembling subacute combined degeneration due
erythroid precursors with excess mitochondrial iron that to vitamin B12 deficiency. Copper deficiency is an uncom-
surrounds or rings the nucleus. This iron is in the form of mon but very treatable cause of hematologic abnormalities.
mitochondrial ferritin. In both congenital and acquired sid- The mechanism by which copper deficiency results in
eroblastic anemia, the formations of ringed sideroblasts is anemia or other cytopenias is unknown. Copper is a cofactor
due to insufficient production of protoporphyrin to utilize for various redox enzymes, including hephaestin and cerulo-
the iron that is delivered to erythroblasts or due to faults in plasmin, which are required to convert ferrous iron to its
mitochondrial function that affect iron pathways and impair ferric form, a step necessary for the transport by transferrin
iron incorporation into protoporphyrin. Acquired sidero- in the intestine and liver, respectively. Cytochrome C oxi-
blastic anemias may be clonal (MDS; Chapter 17) or second- dase also requires copper as a cofactor and a decrease in this
ary to alcohol, drugs (eg, isoniazid, chloramphenicol, enzyme’s activity potentially contributes to the development
linezolid), or copper deficiency. The sideroblastic anemia of ringed sideroblasts (reflecting mitochondrial iron accu-
associated with alcohol is common and is often found in the mulation) in some cases of copper deficiency. To diagnose
most poorly nourished alcoholics and may be associated copper deficiency, serum copper level is measured. A cerulo-
with folate deficiency. The pathogenesis is multifactorial and plasmin level also can be checked but may lack specificity.
may in part be due to vitamin B6 deficiency and/or ethanol
induced abnormalities in vitamin B6 metabolism, and so a
Anemia of cancer
trial of vitamin B6 therapy is reasonable in this setting; vita-
min B6 is therapy is often effective treatment for X-linked The incidence of anemia in cancer patients is highly depen-
congenital sideroblastic anemia. As a general rule, congenital dent on many variables, including cancer type, stage, and
sideroblastic anemias are microcytic and acquired sidero- both present and past anticancer therapy. Its prevalence may
blastic anemias are macrocytic; however, this rule does not exceed 90% in patients with advanced disease receiving che-
always hold. motherapy. A large European survey of 15,000 patients

showed that 67% of patients with cancer are anemic at diag- MDS are an exception to this recommendation. Combining
nosis or become anemic (hemoglobin <12.0 g/dL) during an ESA with intravenous rather than oral iron increases the
the course of their treatment. The mechanisms underlying response rate with no increase in complications.
anemia of malignancy are complex, and numerous factors
contribute to its development. Cytokine-mediated changes Key points
cause a relative decrease in erythropoietin production and
a decrease in the response of erythroid precursors to erythro- • Anemia is frequent in cancer patients and leads to a decreased
poietin. As in AOCD, cytokines cause elevated hepcidin lev- quality of life.
• ESAs reduce transfusion requirements and improve quality of
els resulting in functional iron deficiency. Other causes of
life in cancer patients.
anemia in patients with cancer include effects of chemother-
• The use of ESAs in cancer patients requires careful patient
apy and radiotherapy, bone marrow infiltration, blood loss,
counseling on the potential benefits and risks, and their use
autoimmune and microangiopathic hemolysis, and nutri- should follow published guidelines.
tional deficiencies.
In another recent study involving 888 cancer patients
assessed at diagnosis prior to any cancer treatment, 63% were Myelophthisic anemia
anemic, with the lowest hemoglobin levels being found in
patients with most advanced cancer and, not surprisingly, with Myelophthisic anemia is a normochromic, normocytic ane-
significantly compromised performance status. Hemoglobin mia that occurs when normal marrow space is infiltrated and
levels were inversely correlated with inflammatory markers, replaced by nonhematopoietic or abnormal cells. The term
hepcidin, ferritin, erythropoietin, and reactive oxygen species. myelophthisic is not used commonly in practice and more
A major change in the supportive care of cancer patients often is referred to descriptively as a marrow infiltrative pro-
occurred with the availability of recombinant ESAs. Numer- cess. Causes include tumors, granulomatous disorders, bone
ous studies have demonstrated a decrease in transfusion marrow fibrosis (due to a primary hematologic or numerous
requirements among cancer patients receiving ESAs, with nonhematopoietic disorders), and lipid storage diseases. All
some studies also showing improvement in the quality of life of these causes may induce secondary marrow fibrosis. The
in treated patients. Data, however, suggest that ESAs may peripheral blood smear in myelophthisic anemia often shows
cause tumor growth and clinical studies have shown short- teardrop-shaped RBCs, immature leukocytes, nucleated
ened survival in patients with advanced breast, head and RBCs, and occasional myelobasts (ie, a leukoerythroblastic
neck, lymphoid, and non-small-cell lung cancer, especially process). Rarely, carcinocythemia, defined as cancer cells in
when a hemoglobin of 12 g/dL was targeted. The safety of the circulating blood, is seen. Bone marrow biopsy may show
ESAs for the treatment of the anemia of cancer was again frank tumor cells, Gaucher disease, or other infiltrating dis-
questioned in a recent large meta-analysis. This study ana- orders and marked marrow fibrosis. These conditions may
lyzed individual patient data from 13,933 patients with can- be accompanied by extramedullary hematopoiesis resulting
cer treated on 53 randomized controlled trials using ESAs in organomegaly due to marrow elements in the spleen, liver,
versus standard of care. The analysis demonstrated a 17% or other affected tissues. T1-weighted magnetic resonance
increase in mortality in ESA-treated patients during the imaging may demonstrate areas of abnormal signal, consis-
active study period. There was a 10% increase in mortality tent with marrow infiltration. Treatment is directed at the
when the analysis was restricted to the studies that included underlying disease.
patients treated with chemotherapy. A more recent analysis During infancy, anemia secondary to marrow fibrosis may
of the same data has shown that ESAs do not increase the risk be seen in the setting of osteopetrosis or marble bone disease.
of tumor progression if they are used according to guidelines, These conditions vary in their severity, but infants affected
but there is a small increased risk of venous thromboembolic with the autosomal recessive form present within the first few
disease. The American Society of Hematology/American months of life with pancytopenia, hepatosplenomegaly, and
Society of Clinical Oncology guidelines on ESA use in cancer cranial nerve palsies. Osteopetrosis is caused by failure of
recommend using the lowest possible dose of an ESA that osteoclast development or function, and mutations in at least
gradually will increase hemoglobin concentration to a level 10 genes have been identified, accounting for 70% of cases.
that avoids the need for transfusion and maintains hemoglo- Severe cases are treated by bone marrow transplantation.
bin levels <12 g/dL. ESAs are not recommended for patients
receiving chemotherapy with curative intent. ESAs should
Anemia from malnutrition or anorexia nervosa
not be given in the treatment of anemia associated with
malignancy in patients who are not receiving concurrent Prolonged starvation can lead to a normochromic, normo-
myelosuppressive chemotherapy; patients with low-risk cytic anemia. The bone marrows of such patients are often

Other underproduction anemias | 135
deficiency, folate deficiency, or hemolysis. Conversely,

Anorexia Nervosa - Marrow Morphology
microcytosis can occur in women with concomitant iron
deficiency from menorrhagia, which may be seen in myx-
edema. There is a well-recognized association between auto-
immune thyroid disease and PA, so patients with either
disorder should be screened for the other. The response to
thyroid replacement is typically sluggish, and it may take
months before the anemia resolves. Anemia in patients
with hyperthyroidism is also described, and it is often micro-
cytic; the anemia often corrects when patients are made
euthyroid.
Hypogonadism usually results in a decrease of 1-2 g/dL in
hemoglobin concentration. Androgens stimulate increased
erythropoietin production and can increase the effects of
erythropoietin on the developing erythron. Thus, men typi-
cally have higher hemoglobin concentrations than age-
Figure 6-5 A marrow biopsy is shown, illustrating almost complete matched women. The hemoglobin values of men treated
replacement of the marrow by hyaluronic acid extracellular matrix with antiandrogen therapy for prostate cancer typically falls
material. Hematopoietic elements and fat cells are markedly by 1-2 g/dL.
decreased. Hematoxylin and eosin stain; magnification ×4. Patients with Addison’s disease may have a normochromic
normocytic anemia that is responsive to ESAs or glucocorti-
coids. The mild decrease in RBC mass may be masked by a
hypocellular. Rarely, patients with severe starvation or concomitant decrease in plasma volume, leading to a normal
anorexia nervosa can have gelatinous transformation of the hemoglobin concentration. When replacement with gluco-
marrow with few marrow-derived cells seen histologically corticoid therapy initially is begun, the anemia may become
(Figure 6-5). apparent as the plasma volume is restored. Androgens
sometimes are used to correct anemias due to marrow
hypoproduction in such conditions as myelodysplasia and
Anemia associated with endocrine disorders
myelofibrosis.
In general, the anemia accompanying most endocrine disor- Anemia is more severe and occurs at an earlier stage in
ders is mild, often asymptomatic, and likely overshadowed patients with diabetic nephropathy compared with patients
by the clinical effects of the specific hormone deficiency. In with renal failure from other causes. The exact mechanism
fact, in some cases, the anemia actually may be considered for this finding remains uncertain.
physiologic due to the decreased oxygen requirements
accompanying some of these hormone deficiencies.
Anemia associated with pregnancy
Pituitary deficiency may be associated with a mild normo-
chromic normocytic anemia. The bone marrow of such Anemia is a common complication of pregnancy. Nor-
patients is usually hypoplastic and resembles that seen in mally, the RBC mass increases ~20%-30% during gesta-
other marrow failure states. Anemia most likely is due to section, while the plasma volume increases ~40%-50%,
ondary deficiency of hormones produced by endocrine resulting in a normochromic and normocytic anemia.
organs that normally are stimulated by the anterior lobe of This so-called physiologic anemia of pregnancy reaches a
the pituitary (thyroid hormone, androgens, or cortisol) and nadir at ~30 weeks. Because the RBC mass continues to
modulate erythropoietin production. The anemia responds increase after 30 weeks, when plasma volume expansion
to appropriate hormone replacement. has reached a plateau, the hematocrit may rise somewhat
Patients with primary hypothyroidism may be anemic. after 30 weeks.
The anemia is felt to be secondary to an absence of erythro- Definitions of pathologic anemia during pregnancy vary.
poietin-stimulated erythroid colony formation from lack of United Kingdom guidelines define anemia in pregnancy as a
triiodothyronine (T3), thyroxine (T4), and reverse triiodo- hemoglobin <11 g/dL in the first trimester, <10.5 g/dL in the
thyronine (rT3). The anemia is usually normochromic and second and third trimesters, and <10 g/dL in the postpartum
normocytic, and hemoglobin values usually are not <8 g/dL. period. The evaluation and workup of anemia in pregnant
Macrocytosis may occur in patients with autoimmune hypo- patients should mirror nonpregnant patients. Additionally,
thyroidism, particularly if there is coexistent vitamin B12 special consideration should be paid to any proposed

therapies since any treatment has effects on both the mother is an independent risk factor for cognitive decline and is
and developing fetus. associated with decreased bone density, decreased muscle
In a large study of anemia of pregnancy in Australia from strength, and decreased physical performance in elderly
1999-2005 involving over 124,000 individuals, 7.1% of patients. The presence of anemia, either with or without
women were noted to be anemic. Anemia of pregnancy was other chronic illnesses, is associated with increased hospital-
associated with a higher risk of fetal distress and perinatal ization, morbidity, and mortality.
complications but not with children’s development. Anemia Analysis of data from NHANES III showed that approxi-
of pregnancy can be exacerbated in individuals with sickle mately two-thirds of the cases of anemia were attributable to
cell disease and thalassemias and need to be carefully moni- iron deficiency, other nutritional deficiencies, chronic
tored and managed. inflammatory illnesses, and renal insufficiency. On the basis
Iron deficiency during pregnancy is common, especially in of information available in the database, however, 34% of
non-Western cultures. A full-term pregnancy requires a total cases were unexplained. Unexplained anemia of the elderly
of ~1 g of iron, which includes ~300 mg of iron for the fetus, appears to be a real entity characterized as a hypoprolifera-
~200 mg of maternal iron lost through various normal tive normocytic anemia. Several studies have attempted to
routes of excretion, and ~500 mg of iron for the expanding elucidate the mechanisms for this group of unexplained ane-
maternal RBC mass. Additionally, iron is secreted in breast mias. Investigators have suggested that a variety of age-
milk for use by the developing infant. Folate requirements related physiologic changes may contribute to this process,
also increase during pregnancy, and megaloblastic anemia including blunting of the hypoxia-driven ESA response,
has been reported, especially during the third trimester when decrease in sex steroids, and alterations in bone marrow stem
maternal folate stores become wasted. Prenatal vitamins cells and cellularity.
containing both iron and folate can help reduce, but not Attempts to identify suggested hemoglobin levels for
eliminate, these risks. blood transfusion therapy have been confounded for these
Vitamin B12 deficiency rarely occurs during pregnancy. Of elderly individuals due to their co-morbidities. Since no spe-
note, serum vitamin B12 levels may be less reliable during cific recommended hemoglobin level threshold for this age
pregnancy because of changes in protein binding, and MMA group has been established, some suggest maintaining a
levels should be checked to confirm true deficiency. There hemoglobin of 9 to 10 g/dL.
are reports of idiopathic acquired aplastic anemia patients Because it is likely that anemia is multifactorial in a given
developing worsening pancytopenia or relapse during individual, a thorough clinical and laboratory evaluation is
pregnancy. justified to identify those causes of anemia that are amenable
to therapy. A reasonable approach to evaluation is given in
Key points Table 6-10.
• Anemia in pregnancy is due in part to an imbalance between

the expansion of the plasma volume and the RBC mass. Key points
• Iron deficiency and folate deficiency are important causes of • Anemia is common in elderly patients and often multifactorial.
anemia in pregnancy. • In two-thirds of patients, the anemia is caused by hematinic
• The evaluation of anemia in pregnancy should mirror the deficiency or AOCD and is unexplained in about one-third of
evaluation of anemia in nonpregnant individuals. patients.
• Functional impairment and increased morbidity and mortality
have been demonstrated in elderly anemic patients.
Anemia of the elderly • Transfusion practice to maintain hemoglobin thresholds of 9
to 10 g/dL for the elderly population may be prudent.
Recently, findings from the National Health and Nutrition
Examination Study (NHANES III) indicated that 11.0% of
men and 10.2% of women over the age of 65 were anemic
(based on the definition of anemia as hemoglobin <13 g/dL
for men and <12 g/dL for women). Other studies have sup- Anemia is the most prevalent hematologic abnormality
ported this finding , as well as the higher prevalence rates of associated with HIV infection. Not surprisingly, anemia
anemia in elderly African Americans. In a cross-sectional prevalence increases with HIV disease progression. Several
retrospective study in a large European university hospital studies have shown that anemia is associated independently
cohort of 19,758 individuals >64 years, the overall incidence with decreased survival and decreased quality of life in HIV-
of anemia was 21.1% and it increased to 30.7% and 37% infected patients. Anemia in HIV-infected patients is usually
anemic at 80+ and 90+ years respectively. Anemia multifactorial, and the most likely etiologies depend on the

Bibliography | 137
Table 6-10 Evaluation of anemia in the elderly for the clinical of anemia in HIV patients should include correction of
hematologist: a practical approach. nutritional deficiencies and treatment of infections. ESAs
Always useful have been shown to reduce transfusion requirements in HIV
1. Anemia-oriented history and physical examination, with patients with baseline erythropoietin levels of <500 mU/mL,
particular emphasis on comorbid conditions drug history in whom nutritional deficiencies and other causes have been
associated with anemia corrected.
2. C BC/differential/platelet, absolute reticulocyte count, smear
review
Key points
3. Tests of iron stores (Fe, TIBC, ferritin)
4. Serum cobalamin and folate levels • HIV-related anemia is common and independently associated
5. Chemistry panel (including calculated creatinine clearance) with decreased survival.
6. TSH • Anemia in HIV is multifactorial and may reflect viral infection,
Sometimes useful malignancy, malnutrition, and medications.
• Macrocytosis is more common than anemia in patients treated
1. Serum testosterone
with zidovudine.
2. Serum erythropoietin
• HAART reduces the incidence and degree of anemia in
3. Tests of inflammation (ESR, C-reactive protein)
HIV-infected patients.
4. Methylmalonic acid, serum homocysteine
5. RBC folate level
6. Bone marrow aspiration and biopsy, cytogenetics
Bibliography
Modified from Guralnik JM, et al. Hematology Am Soc Hematol Educ
Program. 2005:531.
Andrews NC. Forging a field: the golden age of iron biology. Blood.
CBC = complete blood count; ESR = erythrocyte sedimentation rate;
2008;112:219-230. A concise review of iron metabolism and its
Fe = iron; RBC = red blood cell; TIBC = total iron-binding capacity;
clinical applications.
TSH = thyroid-stimulating hormone.
Bizzao B, Antico CA. Diagnosis and classification of pernicious
anemia. Autoimmun Rev. 2014;13:565-568. A concise review of
stage of the HIV infection and the medications the patient is pernicious anemia.
receiving. The pathophysiology of anemia in HIV is likely to Carmel R. How I treat cobalamin (vitamin B12) deficiency. Blood.
be a reflection of the underlying inflammatory pathways. 2008;112:2214-2221. A good review of cobalamin deficiency that
Zidovudine and trimethoprim-sulfamethoxazole are com- includes a discussion of quantitative cobalamin numbers that are
useful in understanding cobalamin physiology, depletion, and
monly associated with bone marrow suppression and mac-
therapy in adults.
rocytosis. Infections associated with anemia include
Charytan DM, Pai AM, Chan CT, et. al. Considerations and
Mycobacterium avium complex, tuberculosis, histoplasmo-
challenges in defining optimal iron utilization in hemodialysis.
sis, cytomegalovirus, Epstein-Barr virus, and human parvo- J Am Soc Nephrol. 2014;26:1-10. Thoughtful review and
virus (see the section “Acquired pure red cell aplasia”). commentary on published studies of intravenous iron therapy in
Malignant disorders, mainly non-Hodgkin lymphoma, can the treatment of anemia associated with chronic kidney disease.
infiltrate the bone marrow and cause anemia. Nutritional Cullis J. Diagnosis and management of anaemia of chronic disease:
deficiencies, including vitamin B12, folate, and iron, are current status. Br J Haematol. 2011;154:289-300. An excellent
common in HIV patients and are related to blood loss, review of the anemia of chronic disease.
malabsorption, and poor nutrition. Patients with HIV
Devalia V, Hamilton MS, Molloy AM, et al. Guidelines for
are at risk for hemolysis due to a variety of mechanisms, the diagnosis and treatment of cobalamin and folate disorders.
including microangiopathic hemolysis, antibody-mediated Br J Haematol. 2014;166:496-513. Review of cobalamin and
folate deficiency and guidelines for their diagnosis and treatment.
mechanisms, and drug-induced mechanisms, especially in
Goodnough LT, Schrier SL. Evaluation and management of
patients with glucose-6-phosphate dehydrogenase defi-
anemia in the elderly. Am J Hematol. 2014;89:88-96. An excellent
ciency. Hypogonadism is a frequent finding in patients with
review of anemia in the elderly.
advanced HIV and is associated with a mild anemia. The Hershko C, Camaschella C. How I treat unexplained refractory
HIV virus itself infects bone marrow cells and may interfere iron deficiency anemia. Blood. 2014;123:326-333. Recent review
with hematopoiesis. on the causes and management of unexplained refractory iron
The use of highly active antiretroviral therapy (HAART) deficiency anemia.
has been shown to reduce the prevalence of anemia in several Monzón H, Forné M, Esteve M, et al. Helicobacter pylori infection
population studies. Even the inclusion of zidovudine in the as a cause of iron deficiency anemia of unknown origin.
regimen does not appear to diminish the beneficial effect of World J Gastroenterol. 2013;19:4166-4171. Large study of adult
HAART on anemia. In addition to HAART, the management patients with iron-refractory anemia or iron-dependent anemia

of unknown causes that demonstrates the efficacy of H. pylori Rizzo JD, Brouwers M, Hurley P, et al. American Society
treatment in curing iron deficiency anemia. of Clinical Oncology/American Society of Hematology
Redig AJ, Berliner N. Pathogenesis and clinical implications of Clinical Practice Guideline Update on the use of epoetin
HIV-related anemia in 2013. Hematology Am Soc Hematol Educ and darbepoetin in adult patients with cancer. J Clin Oncol.
Program. 2013;2013:377-381. An excellent summary of HIV- 2010;28:4996-5010. Consensus guidelines on ESAs use in patients
related anemia. with cancer.

CHAPTER
7
Hemolytic anemias
Farzana A. Sayani and Sophie Lanzkron
Introduction, 139 Hemolysis due to extrinsic abnormalities
Hemolysis due to intrinsic abnormalities of the RBC, 167
of the RBC, 140 Bibliography, 182
Introduction (eg, sickle cell disease or hereditary spherocytosis) or

acquired (eg, autoimmune or microangiopathic). Alterna-
Hemolysis is the accelerated destruction of red blood cells tively, they can be characterized by the anatomic site of RBC
(RBCs), leading to decreased RBC survival. The bone mar- destruction: extravascular or intravascular. Extravascular
row’s response to hemolysis is increased erythropoiesis, hemolysis, in which erythrocyte destruction occurs by mac-
reflected by reticulocytosis. If the rate of hemolysis is modest rophages in the liver and spleen, is more common. Intra
and the bone marrow is able to completely compensate for vascular hemolysis refers to RBC destruction occurring
the decreased RBC life span, the hemoglobin concentration primarily within blood vessels. The distinction between
may be normal; this is called fully compensated hemolysis. If intravascular and extravascular hemolysis is not absolute
the bone marrow is unable to completely compensate for because both occur simultaneously, at least to some degree,
hemolysis, then anemia occurs. This is called incompletely in the same patient, and the manifestations of both can
compensated hemolysis. overlap. The site of RBC destruction in different conditions
Clinically, hemolytic anemia produces variable degrees of can be conceptualized to occur in a spectrum between pure
fatigue, pallor, and jaundice. Splenomegaly occurs in some intravascular and pure extravascular hemolysis. Some
conditions. The complete blood count shows anemia and hemolytic anemias are predominantly intravascular (eg,
reticulocytosis that depend on the acuity and severity of paroxysmal nocturnal hemoglobinuria), and some are pre-
hemolysis, and the degree of bone marrow compensation. dominantly extravascular (eg, hereditary spherocytosis).
Secondary chemical changes include indirect hyperbilirubi- Others have substantial components of both, such as sickle
nemia, increased urobilinogen excretion, and elevated lactate cell disease.
dehydrogenase (LDH). Decreased serum haptoglobin levels The hemolytic anemias can be classified according to
and increased plasma-free hemoglobin may also be detected. whether the cause of hemolysis is intrinsic or extrinsic to
Because free hemoglobin scavenges nitric oxide, e sophageal the RBC. Intrinsic causes of hemolysis include abnormali-
spasm or vascular sequelae such as nonhealing skin ulcers ties in hemoglobin structure or function, the RBC mem-
and pulmonary hypertension can occur in chronic hemo- brane, or RBC metabolism (cytosolic enzymes). Extrinsic
lytic anemia. Chronic intravascular hemolysis produces causes may be due to a RBC-directed antibody, a disor-
hemosiderinuria, and chronic extravascular hemolysis dered vasculature, or the presence of infecting organisms
increases the risk of pigmented (bilirubinate) gallstones. or toxins. In general, intrinsic causes of hemolysis are
The hemolytic anemias can be classified in different yet inherited and extrinsic causes are acquired, but there are
complementary ways (Table 7-1). They can be inherited notable exceptions. For example, paroxysmal nocturnal
hemoglobinuria (PNH) is an acquired intrinsic RBC dis-
Conflict-of-interest disclosure: Dr. Sayani declares no competing order, and congenital thrombotic thrombocytopenia pur-
financial interest. Dr. Lanzkron: research funding: Novartis, pura (TTP) is an inherited cause of extrinsic hemolysis. In
Glycomimetics, Emmaus. this chapter, hemolytic anemias will be divided into those
Off-label drug use: Rituximab, cyclophosphamide, azathioprine,
mycophenolate mofetil, cyclosporine and danazol in the treatment of
that are due to intrinsic or extrinsic abnormalities of the
autoimmune hemolytic anemia. RBC.
| 139

140 | Hemolytic anemias
Table 7-1 Methods of classification of hemolytic anemias sequence elements that bind erythroid-specific and nonspe-
Classification Example cific DNA binding proteins and serve as a “master switch,”
inducing expression within the downstream gene cluster. In
Inheritance
Inherited Sickle cell anemia
addition to binding specific transcriptional regulatory pro-
Acquired Autoimmune hemolytic anemia
teins, the LCRs also facilitate the binding and interaction of
Site of RBC destruction transcriptional regulatory proteins in proximity to the spe-
Intravascular Paroxysmal nocturnal hemoglobinuria cific genes within the downstream cluster. These regulatory
Extravascular Hereditary spherocytosis proteins influence the promoter function of the α-globin
Origin of RBC damage and β-globin genes to achieve a high level of erythroid- and
Intrinsic Pyruvate kinase deficiency development-specific gene expression.
Extrinsic Thrombotic thrombocytopenic purpura Figure 7-1 details the organization of the α- and β-clusters
with the associated upstream regulatory elements and the
normal hemoglobin species produced during the develop-
mental stages from intrauterine to adult life. Note that the
Hemolysis due to intrinsic genes are expressed developmentally in the same sequence in
abnormalities of the RBC which they are organized physically in these clusters (left to
right; 5′ to 3′). The process of developmental changes in the
Intrinsic causes of hemolysis include abnormalities of hemo- type and site of globin gene expression is known as hemoglo-
globin structure or function, the RBC membrane, or RBC bin switching. Switching within the cluster is influenced by
metabolism (cytosolic enzymes). Most intrinsic forms of differential enhancing and gene-silencing effects imparted
hemolysis are inherited conditions. by the combination of the LCR and local regulatory proteins,
Abnormalities of hemoglobin
Hemoglobin is the oxygen-carrying protein within RBCs. It Chromosome 11

is composed of four globular protein subunits, called glo- β LCR
bins, each with an oxygen-binding heme group. The two 5 4 321
ε Gγ Aγ ψβ δ β
main types of globins are the α-globins and the β-globins,
which are made in essentially equivalent amount in precur-
Chromosome 16
sors of RBCs. Normal adult hemoglobin (Hb A) has two
HS-40
α-globins and two β-globins (α2β2). Genes on chromo- ζ2 ψζ1 ψα2 ψα1 α2 α1 θ
somes 16 and 11 encode the α-globins and β-globins,

respectively. There are also distinct embryonic, fetal, and
Percent of total globin synthesis
Yolk Liver
minor adult analogs of the α-globins and β-globins encoded sac Spleen
Bone marrow
by separate genes.
50 α
γ β
Hemoglobin production
30
The β-globin gene cluster is on chromosome 11 and includes
an embryonic ε-globin gene, the two fetal γ-globin genes 10 ε
(Aγ and Gγ), and the two adult δ- and β-globin genes. The ζ δ
α-globin gene cluster is on chromosome 16 and includes the 6 16 30 6 18 30 42
embryonic ζ-globin gene and the duplicated α-globin genes Prenatal age
Birth
Postnatal age
(α1 and α2) which are expressed in both fetal and adult life. (weeks) (weeks)
Both clusters also contain nonfunctional genes (pseudo-
Figure 7-1 Hemoglobin gene clusters and developmental
genes) designated by the prefix ψ. The θ globin gene down-
hematopoiesis. The organization of the α- and β-globin gene clusters
stream of α1 has unknown functional significance.
are shown at the top of the figure. The bottom portion of the figure
The expression of each globin gene cluster is under the illustrates the developmental changes in Hb production, both by the
regulatory influence of a distant upstream locus control site of production of blood and changes in the proportions of the
region (LCR). The LCR for the β-cluster resides several kilo- different globins. Modified with permission from Stamatoyannopoulos
bases upstream. A similar regulatory region, called HS-40, G, Majerus PW, Perlmutter RM, et al, eds. The Molecular Basis of Blood
exists upstream of the α cluster. The LCRs contain DNA Diseases. 3rd ed. Philadelphia, PA: W. B. Saunders; 2001.

Hemolysis due to intrinsic abnormalities of the RBC | 141
but the entire process of regulatory determination remains oxygen delivery; and (ii) the higher pH level resulting from
incompletely defined. The ability to modulate the switch carbon dioxide elimination in the lungs increases oxygen
from the synthesis of γ- to β-globin chains has long been of affinity and oxygen loading of RBCs. An additional impor-
interest because “reversing the switch” to enhance expression tant influence on oxyhemoglobin dissociation is tempera-
of fetal hemoglobin (Hb F) could successfully treat sickle cell ture. Hyperthermia decreases affinity, providing the
disease. opportunity to deliver more oxygen at the tissue level. 2,3-
BPG, a metabolic intermediate of anaerobic glycolysis, is
another physiologically important modulator of oxygen
Hemoglobin structure
affinity. When 2,3-BPG binds in the pocket formed by the
Hemoglobin is a tetramer consisting of two pairs of globin amino termini of the β-chains, it stabilizes the deoxy confor-
chains. Heme, a complex of ferrous iron and protoporphy- mation of hemoglobin, thereby reducing its oxygen affinity.
rin, is linked covalently to each globin monomer and can The intraerythrocytic molar concentrations of 2,3-BPG and
reversibly bind one oxygen molecule. The molecular mass of hemoglobin are normally about equal (5 mM). When 2,3-
hemoglobin is approximately 64 kDa. The α-chains contain BPG levels increase during periods of hypoxia, anemia, or
141 amino acids, and the β-chains contain 146 amino acids, tissue hypoperfusion, oxygen unloading to tissues is
as do the β-like globins, δ and γ, which differ from β by enhanced.
10 and 39 amino acids, respectively. The compositions of Carbon dioxide reacts with certain amino acid residues in
normal Hb species throughout development are depicted in the β-chain of hemoglobin; however, this does not play a sig-
Figure 7-1. The postembryonic hemoglbins are Hb A (α2β2), nificant role in carbon dioxide transport. It recently has been
Hb A2 (α2δ2), and Hb F (α2γ2). reported that hemoglobin binds nitric oxide in a reversible
When hemoglobin is deoxygenated, there are substantial manner. The participation of hemoglobin in modifying
changes in the structures of the individual globins and the regional vascular resistance through this mechanism has
hemoglobin tetramer. The iron molecule rises from the been proposed.
plane of its heme ring, and there is a significant rotation of
each globin chain relative to the others. In the deoxy confor-
Disorders of hemoglobin
mation, the distance between the heme moieties of the
β-chains increases by 0.7 nm. This conformation is stabi- Disorders of hemoglobin can be classified as qualitative or
lized in a taut (T) conformation by salt bonds within and quantitative disorders. Qualitative abnormalities of hemo-
between globin chains and by the binding of allosteric modi- globin arise from mutations that change the amino acid
fiers such as 2,3-bisphosphoglycerate (2,3-BPG) and of pro- sequence of the globin, thereby producing structural and
tons. The binding of oxygen to hemoglobin leads to functional changes in hemoglobin. There are four ways in
disruption of the salt bonds and transition to a relaxed which hemoglobin can be qualitatively abnormal: (i) de
(R) conformation. creased solubility, (ii) instability, (iii) altered oxygen affinity,
and (iv) altered maintenance of the oxidation state of the
heme-coordinated iron. Hemolytic anemia results from
Hemoglobin function
decreased solubility and instability of hemoglobin. Qualita-
Hemoglobin enables RBCs to deliver oxygen to tissues by its tive hemoglobin disorders often are referred to as hemoglo-
reversible binding of oxygen. With the sequential binding of binopathies, even though the term technically can apply to
one oxygen molecule to each of the four heme groups, the both qualitative and quantitative disorders. Quantitative
salt bonds are progressively broken, which increases the oxyhemoglobin disorders result from the decreased and imbal-
gen affinity of the remaining heme moieties. Cooperation anced production of generally structurally normal globins.
between the heme rings results in the characteristic sigmoid- For example, if β-globin production is diminished by a
shaped oxygen affinity curve. This phenomenon accounts mutation, there will be a relative excess of α-globins. Such
for the release of relatively large amounts of oxygen with imbalanced production of α- and β-globins damages RBCs
small decreases in oxygen tension. and their precursors in the bone marrow. These quantitative
Deoxygenation of hemoglobin is modulated by certain hemoglobin disorders are called thalassemias. Both qualita-
biochemical influences. For example, deoxyhemoglobin tive and quantitative disorders of hemoglobin can be subdi-
binds protons with greater avidity than oxyhemoglobin, vided by the particular globin that is affected; for example,
which results in a direct dependence of oxygen affinity on there are α-thalassemias and β-hemoglobinopathies, among
pH over the physiologic pH range. This Bohr effect has two others. We will begin this chapter with a review of the thalas-
major physiologic benefits: (i) the lower pH of metabolically semias and end the section with a discussion of several of the
active tissue decreases oxygen affinity, which accommodates common qualitative hemoglobin disorders.

Thalassemia divides the thalassemias into the asymptomatic trait state

(thalassemia minor), severe transfusion-dependent anemia
Clinical case (thalassemia major), and everything in between (thalassemia
A healthy 48-year-old female of African descent is referred to
intermedia). The two systems can be used together, giving
you for evaluation of refractory microcytic anemia. She has been α-thalassemia minor or β-thalassemia intermedia, for
treated with oral iron formulations many times throughout her example.
life. Hemoglobin values have always ranged from 10-11 g/dL
with a mean corpuscular volume (MCV) ranging from 69-74
β-Thalassemias
fL. She has no other prior medical history. Her examination is
entirely unremarkable. Peripheral blood smear is significant for β-Thalassemia is prevalent in the populations of the Medi-
microcytosis, mild anisopoikilocytosis, and a small number of terranean region, the Middle East, India, Pakistan, and
target cells. The hemoglobin concentration is 10 g/dL with an Southeast Asia, and is somewhat less common in Africa. It is
MCV of 71 fL and mean corpuscular hemoglobin (MCH) of
rarely encountered in Northern European Caucasians.
23 pg. Additional laboratory studies include a transferrin saturation
of 32% and a normal ferritin of 285 ng/mL. Hemoglobin electro-
phoresis reveals hemoglobin A 98% and hemoglobin A2 1.8%. Molecular basis
β-Thalassemia results from >200 different mutations of the

Thalassemia occurs when there is the quantitatively decreased β-globin gene complex (Figure 7-2). Abnormalities have
synthesis of often structurally normal globin proteins. Muta- been identified in the promoter region, messenger RNA
tions that decrease the synthesis of α-globins cause α thalas- (mRNA) cap site, 5′ untranslated region, splice sites, exons,
semia; mutations that decrease the synthesis of β-globins introns, and polyadenylation signal region of the β-gene.
cause β-thalassemia. Gene deletions are infrequent except in δβ and egδβ thalas-
Heterozygous thalassemia (thalassemia trait) appears to semias. A variety of single–base pair mutations or insertions
confer protection against severe Plasmodium falciparum or deletions of nucleotides represent the majority of
malarial infection. As a result of this selective advantage, a described mutations. Thus, defects in transcription, RNA
wide variety of independent mutations leading to thalas- processing, and translation or stability of the β-globin gene
semia have arisen over time and have been selected for in product have been observed. Mutations within the coding
populations residing in areas where malaria is (or once was) region of the globin gene allele may result in nonsense or
endemic. In general, α-thalassemias are caused by deletions truncation mutations of the corresponding globin chain,
of DNA, whereas β-thalassemias are caused by point muta- leading to complete loss of globin synthesis from that allele
tions. If a mutation decreases the synthesis of one globin, (β0 thalassemia allele). Alternatively, abnormalities of tran-
α or β, it produces a relative excess of the other and an imbal- scriptional regulation or mutations that alter splicing may
ance between the two occurs. For example, if β-globin syn-
thesis is diminished by a mutation, there will be a relative
excess of α-globins. Such imbalanced production of α- and
Deletions
β-globins results in damage to precursors of RBCs in the
bone marrow. This damage occurs largely because the excess
unpaired globin is unstable, and it precipitates within early
RBC precursors in the bone marrow and oxidatively dam-
ages the cellular membrane. If the α- and β-globin imbal- 1 IVS 1 2 IVS 2 3
ance is severe, most of the RBC precursors in the bone
marrow are destroyed before they can be released into the P C I SP SP SP SP CL
FS FS FS FS
circulation. A severe microcytic anemia results. The body NS NS NS NS
attempts to compensate for the anemia by increasing eryth-
Point mutations 100 bp
ropoietic activity throughout the marrow and sometimes in
extramedullary spaces, although this effort is inadequate and
Figure 7-2 Common β-thalassemia mutations. The major classes and
compensation is incomplete. This pathophysiologic process
locations of mutations that cause β thalassemia are shown. Redrawn
is called ineffective erythropoiesis. from Stamatoyannopoulos G et al., eds. The Molecular Basis of Blood
The thalassemias can be described simply by two indepen- Diseases. 3rd ed. Philadelphia, PA: WB Saunders; 2001. C = CAP site;
dent nomenclatures: genetic and clinical. The genetic nomen- CL = RNA cleavage [poly(A)] site; FS = frameshift; I = initiation
clature denotes the type of causative mutation, such as codon; NS = nonsense; P = promoter boxes; SP = splice junction,
α-thalassemia or β-thalassemia. The clinical nomenclature consensus sequence, or cryptic splice site.

cause mild to markedly decreased, but not absent, globin

gene synthesis (β+ thalassemia allele). β-thalassemia major Reduced levels of Absence of
β-globin chains (β+) β-globin chains (β0)
(Cooley anemia) and β-thalassemia intermedia can be due
to various genotypes including homozygosity or compound
heterozygosity for two β0 alleles (β0/β0) or compound het- Excess free
α chains
erozygosity with a β0 and β+ allele (β0/β+). Patients with beta
thalassemia trait are generally heterozygous, carrying a single Precipitation in
RBC precursors within
β-thalassemia allele (β/β0, β/β+), but some patients who are Membrane damage to
bone marrow
peripheral RBC
homozygous or compound heterozygous for two very mild
β+ alleles may also have β-thalassemia minor phenotype. Premature death of
The clinical phenotype of patients with β-thalassemia is RBC precursors
determined primarily by the globin chain imbalance due to Hemolysis
the number and severity of the abnormal alleles inherited. Ineffective erythropoiesis
Additional factors that contribute to the phenotype include
Anemia
the number of alpha globin genes, genetic determinants
associated with increased gamma-chain production, and
secondary genetic modifiers such as uridine diphosphate-
glucuronosyltransferase (UDPG) gene polymorphisms. Erythropoietin Iron absorption Blood
increased increased transfusion
Pathophysiology
Bone marrow Iron
The defect in β-thalassemia is a reduced or absent produc- expansion loading
tion of β-globin chains with a relative excess of α-chains. The

decreased β-chain synthesis leads to impaired production of
Skeletal changes Liver cirrhosis
the a2β2 tetramer of Hb A, decreased hemoglobin produc- Hypermetabolic Endocrine
tion, and an imbalance in globin chain synthesis. The reduc- state dysfunction
Arrhythmias,
tion in Hb A in each of the circulating RBCs results in heart failure
hypochromic, microcytic RBCs with target cells, a character-
istic finding in all forms of β-thalassemia. Aggregates of
Figure 7-3 Pathophysiology of β-thalassemia. Effects of excess
excess α-chains precipitate and form inclusion bodies, lead- production of free α-globin chains in β-thalassemia. Adapted with
ing to premature destruction of erythroid precursors in the permission from Viprakasit V and Origa R. In: Guidelines for the
bone marrow (ineffective erythropoiesis) (Figure 7-3). In Management of Transfusion Dependent Thalassaemia (TDT). 3rd ed.
more severe forms, circulating RBCs also may contain inclu- Nicosia, Cyprus: Thalassaemia International Federation; 2014.
sions, leading to early clearance by the spleen. The precipi-
tated α-globin chains and products of degradation may also affecting α- or γ-globin synthesis or structural abnormalities
induce changes in RBC metabolism and membrane struc- of Hb (eg, Hb S).
ture, leading to shortened RBC survival. The response to β-Thalassemia minor (trait) is asymptomatic and is char-
anemia and ineffective erythropoiesis is increased produc- acterized by mild microcytic anemia. Most commonly, it
tion of erythropoietin leading to erythroid hyperplasia which arises from heterozygous β-thalassemia (β-thalassemia
can produce skeletal abnormalities, splenomegaly, extramed- trait). Neonates with β-thalassemia trait have no anemia or
ullary masses and osteoporosis. Ineffective erythropoiesis is microcytosis; these develop with increasing age as the transi-
associated with increased gastrointestinal iron absorption tion from Hb F to Hb A production progresses.
due to decreased hepcidin. RBC membrane damage with β-Thalassemia major (Cooley’s anemia) is characterized
increased surface expression of anionic phospholipids, plate- by absence of or severe deficiency in β-chain synthesis. Symp-
let activation, and changes in hemostatic regulatory proteins toms are usually evident within the first 6-12 months of life as
contribute to a hypercoagulable state in thalassemia. the Hb F level begins to decline. In the absence of adequate
RBC transfusions, the infant will experience failure to thrive
Clinical features
and a variety of clinical findings. Erythroid expansion leads to
widening of the bone marrow space, thinning of the cortex,
The clinical manifestations of β-thalassemia are quite het- and osteopenia, predisposing to fractures. Growth retarda-
erogeneous and depend on the extent of β-globin chain pro- tion, progressive hepatosplenomegaly, gallstone formation,
duction as well as the coinheritance of any other abnormalities and cardiac disease are common. Most homozygotes do not

survive without transfusions beyond the age of 5 years. RBC (δβ)0 thalassemia is due to the deletion of both δ and
transfusions ameliorate severe anemia and suppress ineffec- β genes, while (γδβ)0 is due to deletion of γ, δ and β genes.
tive erythropoiesis. A chronic transfusion program can allow Heterozygotes have a phenotype similar to β-thalassemia
for normal growth and development, prevent skeletal abnor- trait and may have splenomegaly. Homozygotes or com-
malities, and suppress extramedullary hematopoiesis. pound heterozygotes with a severe b+ or b0 mutation may
Iron overload is a major complication in β-thalassemia have thalassemia intermedia due to increased synthesis of
due to transfusions and increased gastrointestinal iron hemoglobin F.
absorption. Each milliliter of transfused blood contains 1 mg Heterozygotes for the rare egdβ deletion present with
of iron. Red cell transfusions are the major cause of iron moderately severe microcytic anemia in the neonatal period
loading in beta thalassemia major. Iron accumulates since with hemolysis, erythroblastosis, hepatosplenomegaly, and
the body does not have an active mechanism to excrete excess may need transfusions at birth or during the first 6 months
iron. Excess iron results in increased non-transferrin bound of life. In the adult, the hematologic findings are those of
iron, which generates harmful reactive oxygen species lead- β-thalassemia trait, except that the Hb electrophoresis shows
ing to lipid peroxidation, and organelle and DNA damage normal Hb F and A2 levels. The homozygous state is incom-
causing apoptosis, fibrosis and organ damage. Uncontrolled patible with fetal life.
transfusional iron loading leads to iron deposition in key
organs leading to an increased risk of liver cirrhosis, hepato- Laboratory findings
cellular carcinoma, heart failure, and endocrine complica-
tions including hypogonadotropic hypogonadism, diabetes, Patients with β-thalassemia trait may have a hemoglobin
hypothyroidism, osteoporosis, and hypoparathyroidism. ranging from 9 g/dL to a normal value. Peripheral smear
Over the last few years, patient survival has significantly shows microcytic, hypochromic RBCs and target cells. Baso-
improved due to improved iron chelation therapy, improved philic stippling is variable. The MCV is usually <70 fL, the
modalities to measure liver and cardiac iron load, and a com- MCH is reduced, and the reticulocyte count can be mildly
prehensive care approach. An increased frequency of Yersinia elevated. Hb A2 levels are elevated >3.5% (usually 4%-7%),
enterocolitica bacteremia is associated with iron overload and and Hb F levels may be mildly increased. A variable degree of
chelation therapy with deferoxamine. anemia with hypochromic, microcytic cells and target cells is
In the modern era, β-thalassemia intermedia is often observed in β-thalassemia intermedia. Laboratory abnormal-
grouped under the term non-transfusion dependent thalas- ities are similar to β-thalassemia trait, but more severe. A
semia (NTDT), which is used to describe patients with mod- child with β-thalassemia major who is not receiving transfu-
erate anemia who do not need life-long regular transfusions sions will have severe anemia. Peripheral blood smear findings
for survival, but need occasional or frequent transfusions in include anisopoikilocytosis, target cells, severe hypochromia,
certain clinical settings for short periods of time. NTDT nucleated red blood cells, and basophilic stippling. The reticu-
encompasses three clinically distinct forms of thalassemia locyte count is slightly increased, and nucleated RBCs are
including beta thalassemia intermedia, hemoglobin abundant. These findings are exaggerated after splenectomy.
E/b-thalassemia, and hemoglobin H disease. These patients Hemoglobin electrophoresis reveals persistent elevation of
exhibit a wide spectrum of clinical findings including hepa- Hb F (a2g2) and variable elevation of Hb A2 (a2d2). Hb A is
tosplenomegaly, extramedullary hematopoietic pseudotu- absent in homozygous b0 thalassemia. In (dβ)0 or (gdβ)0
mors, bone deformities, leg ulcers, thrombotic events, hemoglobin ranges between 8-13 g/dL, and the MCV may be
pulmonary hypertension, silent infarcts, gallstones, and iron normal or low-normal. Hemoglobin A2 is however normal,
overload. These complications, except for iron overload are and Hb F is increased between 5%-20%. Homozygotes for
generally limited in the well-transfused thalassemia patient (dβ)0 or (gdβ)0 thalassemia have 100% Hb F.
because transfusion interrupts the underlying pathophysiol-
ogy. An increased incidence of cerebral thrombosis, venous
Management of the β-thalassemias
thromboembolism, and pulmonary hypertension has been
reported in β-thalassemia major and β-thalassemia interme- Individuals with β-thalassemia trait do not require therapy
dia following splenectomy, and these risks should be consid- but should be identified to reduce the risk of inappropriate
ered before splenectomy. Iron overload in NTDT occurs iron supplementation. Individuals of child-bearing age
primarily due to increased gastrointestinal absorption in should be offered genetic counseling for informed reproduc-
the setting of ineffective erythropoiesis. Thus, even in the tive choices.
absence of transfusion, some patients may have iron over- Management of patients with β-thalassemia major and inter-
load. The iron overload significantly increases with increas- media involves a comprehensive multidisciplinary care approach.
ing number of transfusions. RBC transfusion has been the mainstay in the management

of β-thalassemia major and its complications are described studies exploring nonmyeloablative or unrelated donor
above. The goals of transfusions are to promote normal transplantation are encouraging, even in patients with prior
growth and development and to suppress ineffective eryth- iron loading (for whom chelation therapy before transplan-
ropoiesis. A lifelong chronic blood transfusion program to tation is advised) or concomitant hepatitis C virus (HCV)
maintain a pretransfusion Hb level between 9-10 g/dL infection. Many adults with thalassemia major have chronic
sufficiently suppresses bone marrow expansion while mini- HCV infection resulting from contaminated RBC products
mizing transfusional iron loading. that they received before the early 1990s. Treatment with
Monitoring iron load is key to establishing an individual- ribavirin-based regimens may be complicated by hemolysis
ized, effective iron chelation regimen for each patient. Iron resulting from ribavirin and have been limiting in thalas-
load is determined by serum ferritin, liver iron concentra- semia patients. New treatment regimens for HCV that do
tion, and cardiac iron load. Serum ferritin generally corre- not include ribavirin and are not associated with hemolysis
lates with body iron stores and is an easy, convenient, and are encouraging. Recent proof of principle gene therapy
inexpensive measure to trend. However, it has several limita- studies in β-thalassemia major have achieved transfusion
tions since it is an indirect measure of true body iron burden, independence and are promising for the future.
is an acute phase reactant, and it has a non-linear response to Individuals with NTDT may require intermittent transfu-
iron load at high levels. Liver iron concentration (LIC) can sions during acute episodes of worsening anemia, including
be determined by liver biopsy or by the new gold standard, infection or acute illness. Most indications for initiating a
liver MRI R2. Normal LIC is < 1.8 mg Fe/g dry weight. Car- chronic transfusion program in NTDT are similar to those
diac MRI T2* correlates with cardiac iron load and the risk in β-thalassemia major. However, these are generally initi-
of developing heart failure increases with T2* values <20 ms. ated later in childhood or in adulthood, depending on the
The risk for developing heart failure is highest when the car- severity of the phenotype. Iron overload in NTDT occurs
diac T2* is <8 ms. A complete iron load evaluation thus primarily due to increased gastrointestinal absorption but
includes at least serum ferritin every 3 months, yearly LIC by significantly increases with increasing transfusions. Iron-
MRI R2 starting at age 5, and yearly cardiac iron T2* starting associated complications are similar to those seen in beta
at 8-10 years of age. For young children, the risks of sedation thalassemia major, except cardiac siderosis is much less
should be weighed against the risks of severe liver iron over- common. Serum ferritin and LIC measurements show a gen-
load. The main goals of iron chelation therapy are to main- erally positive correlation and should be regularly evaluated
tain safe levels of body iron to prevent iron overload and its in all patients over 10 years of age. In NTDT, the total body
complications and to reduce accumulated iron. Iron chela- iron load may be higher than what the serum ferritin levels
tion therapy is tailored to each individual based on transfu- suggest. Thus a serum ferritin of >800 ng/mL warrants LIC
sion rates and iron burden. In β-thalassemia major, iron evaluation. Chelation therapy should be initiated if the LIC
chelation therapy with subcutaneous deferoxamine or oral is −
>5 mg Fe/g dry weight to reduce iron-associated morbid-
deferasirox is initiated when serum ferritin levels reach ity. Deferasirox has been well studied in NTDT with a good
approximately 1,000-1,500 ng/mL following approximately efficacy and safety profile.
12 months of scheduled transfusions or approximately 20
units of blood. Chelation is adjusted to maintain a ferritin
α-Thalassemias
<1,000-1,500 ng/mL, a LIC of 2-7 mg of iron/g dry weight,
and a cardiac T2* >20 ms. Monitoring for chelator-specific There is a high prevalence of α-thalassemia in Africa, the
complications should be performed. Mediterranean region, Southeast Asia, and, to a lesser extent,
In the modern era of improved transfusion support and the Middle East.
chelation, and in light of recognized post-splenectomy
complications of increased infection risk, venous thrombo-
Molecular basis
embolism, and pulmonary hypertension, splenectomy is
generally not recommended. Often, increasing the transfu- Duplicated copies of the α-genes are normally present on
sion targets is sufficient to reduce the degree of splenomeg- each chromosome 16, making the defects in α-thalassemia
aly. Allogeneic bone marrow transplantation from a more heterogeneous than in β-thalassemia. The α+ thalas-
histocompatibility (human leukocyte antigen [HLA]-com- semias result from deletion of one of the linked genes,
patible) sibling has been performed in >1,000 patients and –α/aα, or impairment due to a point mutation, designated
is now curative in most. The outcome has been influenced aTα/aα. The deletion type of α+ thalassemia is due
by the age of the patient, the presence of liver disease, and to unequal crossover of the linked genes, whereas the non-
the extent of iron overload. Graft-versus-host disease repre- deletion type includes mutations resulting in abnormal tran-
sents the most common long-term complication. Recent scription or translation or the production of unstable

α-globin. The –α/aα genotype (the “silent carrier” state) thalassemia (–α/aα), so-called silent carriers, are clinically
occurs in approximately one in three African Americans. normal. Thalassemia trait (two-gene deletion α-thalassemia)
Hemoglobin Constant Spring is a nondeletion α+ thalas- occurs in two forms: a0 thalassemia trait (– –/aα) or homo-
semia, common in Southeast Asia, resulting from a muta- zygosity for α+ thalassemia (–α/–α). Individuals with thalas-
tion that affects termination of translation and results in semia trait have a lifelong mild microcytic anemia. The
abnormally elongated α-chains. The – –/aα genotype of clinical manifestations are variable in Hb H disease (– –/–α),
α-thalassemia trait due to loss of linked α-genes on the same with severe forms demonstrating transfusion dependence
chromosome (cis configuration), is more common in indi- and other individuals having a milder course. As in
viduals of Asian descent, whereas the –α/–α genotype (dele- β-thalassemia, splenomegaly occurs commonly in the ane-
tions in the trans position) is more common in individuals mic patient. The homozygous state for Hb Constant Spring
of African or Mediterranean descent. results in moderate anemia with splenomegaly. Hb
H–Constant Spring (– – /αCSα) is typically more severe than
Pathophysiology classical Hb H disease (– –/–α). Homozygous a0 thalassemia
(– –/– –) results in the Hb Bart hydrops fetalis syndrome.
As in the β-thalassemias, the imbalance of globin chain syn-
thesis results in decreased hemoglobin synthesis and micro-
cytic anemia. Excess γ- and β-chains form tetramers termed
Hb Bart and Hb H, respectively. These tetramers are more
soluble than unpaired α-globins (as in β-thalassemia) and
α0 Thalassemia trait α0 Thalassemia trait
form RBC inclusions slowly. Consequently, although
α-thalassemia is associated with a hemolytic anemia, it does
not lead to significant ineffective erythropoiesis. The homo- X
zygous inheritance of a0 thalassemia (– –/– –) results in the
total absence of α-chains, death in utero, or hydrops fetalis.
Unpaired γ-chains form Hb Bart (γ4), and there may be per-
sistence of embryonic hemoglobins. Hb Bart is soluble and
does not precipitate; however, it has a very high oxygen affin-
ity and is unable to deliver oxygen to the tissues. This leads to Normal α0 Thalassemia α0 Thalassemia Hb Bart
severe fetal tissue hypoxia, resulting in edema, congestive trait trait hydrops
heart failure, and death. Hb H disease results from the coin-

heritance of a0 thalassemia and α+ thalassemia (– –/ –α) or α0 Thalassemia trait α+ Thalassemia trait
a0 thalassemia and a nondeletional form of α-thalassemia
(– –/aTα) such as Hb Constant Spring (– –/αCSα). The excess X
β-chains form Hb H (β4) that is unstable, causing precipita-
tion within circulating cells and hemolysis. Patients have
moderately severe hemolytic anemia.
Hb H also can be produced as an acquired phenomenon
in the setting of myelodysplastic syndromes and some
myeloid leukemias, in which somatic mutations of the ATRX
Normal α0 Thalassemia α+ Thalassemia Hb H
gene downregulate α-globin production and cause globin trait trait disease
chain imbalance. This condition is called the α-thalassemia–
myelodysplastic syndrome (ATMDS). The X-linked ATRX
gene encodes a chromatin-remodeling factor (X-linked heli- Figure 7-4 Genetics of α-thalassemia. The α-globin genes are
case 2) that regulates α-globin production. Constitutional represented as boxes. The red α-globin genes represent deletions or
deletions of this gene produce the α-thalassemia–mental otherwise inactivated α-genes. The open boxes represent normal
retardation syndrome. α-genes. The terms α0 and α+ thalassemia are defined in the text. The
potential offspring of two parents with α0 thalassemia trait is shown in
Clinical features the upper panel. The potential offspring of one parent with α0
thalassemia trait and the other with α+ thalassemia trait is shown in
In contrast to β-thalassemias, α-thalassemias can manifest the lower panel (note the lack of hemoglobin Bart hydrops fetalis in
in both fetal and postnatal life. The clinical manifestations of these offspring). Redrawn from Stamatoyannopoulos G et al., eds. The
α-thalassemia are related to the number of functional Molecular Basis of Blood Diseases. 3rd ed. Philadelphia, PA: W. B.
α-globin genes (Figure 7-4). Heterozygotes for α+ Saunders; 2001.

The lack of Hb F due to the absence of α chains produces program or splenectomy may prove useful. A fetus with
intrauterine hypoxia, resulting in marked expansion of bone homozygous α0 thalassemia can be rescued with intrauterine
marrow and hepatosplenomegaly in the fetus and enlarge- transfusion, followed by postnatal chronic transfusions or
ment of the placenta. In utero death usually occurs between stem cell transplantation. Because of the high prevalence of
30 and 40 weeks or soon after birth. Moreover, the mother the α0 genotype in Southeast Asian and certain Mediterra-
often experiences morbidity from polyhydramnios. nean populations, screening programs and genetic counsel-
ing can reduce the occurrence of births resulting in Hb Bart
Laboratory features hydrops fetalis and Hb H disease.
Minimal or no anemia or morphologic abnormalities of

RBCs are observed in silent carriers (heterozygous α+ thalas- Clinical case (continued)
semia [–α/aa]). The hemoglobin electrophoresis is normal. The patient presented in this case likely has two copies of
In newborns who are heterozygous for a0 thalassemia (– –/aα), alpha deletions in the trans position (–α/–α). Patients with this
the hemoglobin electrophoresis reveals 2%-5% Hb Bart and condition usually have mild microcytic, hypochromic anemia.
microcytosis (<95 fL). Children and adults heterozygous for Targeted RBC forms suggest the presence of thalassemia in an
a0 thalassemia (– –/aα) or homozygotes for α+ thalassemia otherwise healthy person. With single or double α-gene dele-
(–α/–α) have mild anemia with hypochromic, microcytic tions, the hemoglobin electrophoresis is typically normal, unlike
RBCs and target cells. The RBC indices are similar to those of in β-thalassemia. α-Thalassemia is often a diagnosis of exclusion,
β-thalassemia trait, but the hemoglobin electrophoresis is and identification of similar findings in family members supports
normal (or shows a reduction in Hb A2). Molecular testing is the diagnosis. Molecular testing for specific α gene deletions
confirms the diagnosis. Iron deficiency should be ruled out.
required to confirm the diagnosis in alpha thalassemia due to
Exogenous iron should not be prescribed because it is unneces-
one or two gene deletions. The high prevalence of the –α/–α
sary and potentially harmful. Patients are generally asympto-
genotype in African Americans is noteworthy. About 2%-3%
matic, require no treatment, and have a normal life expectancy.
of all African Americans in the United States have asymp-
tomatic microcytosis and borderline anemia (often mistaken
for iron deficiency) as a result of this condition. Hb H disease Key points
(– –/–α) is characterized by anisopoikilocytosis and hypo-
chromia with elevated reticulocyte counts. Hemoglobin • The thalassemias are characterized by a reduced rate of
electrophoresis reveals 5%-40% of the rapidly migrating synthesis of one of the globin subunits of the hemoglobin
molecule.
Hb H. Supravital staining with brilliant cresyl blue will reveal
• The intracellular precipitation of the excess, unpaired globin
inclusions representing in vitro precipitation of Hb H. The
chains in thalassemia damages red cell precursors and circulat-
blood smear in Hb Bart hydrops fetalis syndrome (– –/– –)
ing red cells, resulting in ineffective erythropoiesis and
reveals markedly abnormal RBC morphology with anisopoi- hemolysis.
kilocytosis, hypochromia, targets, basophilic stippling, and • The β-thalassemias are caused by >200 different mutations,
nucleated RBCs. The hemoglobin electrophoresis in a neo- usually point mutations, with a wide variety of genetic abnor-
nate reveals approximately 80% Hb Bart and the remainder malities documented.
Hb Portland (z2g2). • Patients with thalassemia major require transfusion support,
and develop iron overload associated complications. A spectrum
of clinical manifestations is observed in thalassemia intermedia,
Management of the α-thalassemias whereas the carrier state has no associated symptoms.
• The hemoglobin electrophoresis in β-thalassemia reveals
Individuals with 1 or 2 alpha gene deletions (–α/aα, – –/aα,
increased levels of hemoglobin A2 and variably increased
–α/–α) do not generally require any specific treatment.
hemoglobin F.
Patients with Hb H disease are categorized as NTDT as • The α-thalassemias are primarily due to DNA deletions. Four
described earlier and usually require no specific interven- α-genes are normally present, so multiple phenotypes are
tions. However, since the phenotype can be variable, close possible when gene deletions occur.
observation and follow up is important. Some patients, espe- • α-Thalassemia trait is characterized by mild anemia with
cially those with Hb H–Constant Spring, require intermit- microcytic indices and a normal hemoglobin electrophoresis.
tent or chronic RBC transfusions. Intermittent transfusions • The clinical manifestations in hemoglobin H disease are
may be required in acutely worsening anemia due to infec- variable, with some affected individuals requiring transfusions
tion or acute illness, or to allow for normal growth and and others less symptomatic.
development in childhood. For those patients with signifi- • Homozygous a0 thalassemia manifests in fetal life with the
formation of hemoglobin Bart (γ4) and hydrops fetalis.
cant anemia and splenomegaly, a chronic transfusion

Sickle cell disease
HBGR
HBG1
HBG2
Centromere
HBD
HBB
HBE
Clinical case
pter
3’ 5’
A 17-year-old African American male with homozygous sickle
HBBC
cell anemia (HbSS) is admitted to the hospital with a 4-day
history of a typical painful episode involving his arms and Chromosome 11
legs. There is no recent febrile illness. Past medical history is
16
14
11
13
22
23
p 1
2
remarkable for few hospital admissions for pain crises and red
q
cell transfusion once as a young child. He is in severe pain and
appears ill, and vital signs are remarkable for a pulse of 129 and
temperature of 38.5°C. Scleral icterus and moderate jaundice are
noted. Laboratory data include hemoglobin 7.2 g/dL (baseline
ε GγAγ ψβ δ β
9.1 g/dL), corrected reticulocyte count of 2%, and platelet count β-like genes 5’ 3’
72,000/µL. Liver function tests are elevated above baseline and
0 10 20 30 40 50 60 Kilobases
include a direct bilirubin of 4.8 mg/dL, aspartate aminotrans-
ferase (AST) of 1,200 U/L, and alanine aminotransferase (ALT)
1,550 U/L. His creatinine is elevated at 4.3 mg/dL. Abdominal βS6(Glu Val)
ultrasound is nondiagnostic. He is immediately started on intra- βC6 (Glu Lys)
venous fluids and opioid analgesics. Broad-spectrum antibiot- β E26(Glu Lys) βD
121(Glu Gln)
ics are empirically administered. Over the next 24 hours he
5’ Exon 1 Exon 2 Exon 3 3’
becomes tachypneic and slightly confused. Hypoxemia develops Intron 1 Intron 2
despite oxygen supplementation, and anuria ensues. Serum
creatinine has increased to 6.4 mg/dL, direct bilirubin to
7.8 mg/dL, AST to 2,725 U/L, and creatine phosphokinase (CPK) Figure 7-5 Common β-globin variants. The locations of the
to 2,200 IU/L and hemoglobin has decreased to 5.8 g/dL. The mutations within the chromosome (top), the β-globin cluster
patient undergoes simple transfusion and subsequently red cell (middle), and the β-globin gene itself (bottom) are shown for four
exchange. Acute dialysis is required. He slowly improves during common β-globin variants.
a prolonged 3-week hospitalization. No infectious etiology was
identified.
(including sickle cell trait). In contrast, the term sickle cell
disease includes only those genotypes associated with vary-
Sickle hemoglobin (HbS) was the first hemoglobin variant ing degrees of chronic hemolytic anemia and vaso-
discovered. It has been well characterized on a biochemical occlusive pain (not sickle trait): homozygous sickle cell
and molecular level. Heterozygosity for the sickle cell anemia (HbSS), sickle-Hb C disease (HbSC), sickle-b0
gene (bS), called sickle cell trait, occurs in >20% of indi- thalassemia (HbSb0), and sickle-β+ thalassemia (HbSβ+).
viduals in equatorial Africa; up to 20% of individuals in the Less common hemoglobin mutants, such as OArab, DPunjab,
eastern provinces of Saudi Arabia and central India; up to or E, may be inherited in compound heterozygosity with bS
6.3% in Hispanic populations; and approximately 5% of to result in sickle cell disease.
individuals in parts of the Mediterranean region, the Mid- Sickle cell trait (Hb AS) occurs in 8%-9% of the African
dle East, and North Africa. In HbS, a hydrophobic valine is American population. It is associated with the rare compli-
substituted for the normal, more hydrophilic glutamic acid cations of hematuria, renal papillary necrosis, pyelone-
at the sixth residue of the β-globin chain (Figure 7-5). This phritis during pregnancy, and risk of splenic infarction at
substitution is due to a single nucleotide mutation (GAG/ high altitude. Sickle trait also is associated with the
GTG) in the sixth codon of the β-globin gene. Heterozy- extremely rare medullary carcinoma of the kidney and an
gous inheritance of HbS offers a degree of protection from increased risk of sudden death during extreme conditions
severe malaria infection. This has been offered as an expla- of dehydration and hyperthermia. Recent publications
nation for the evolutionary selection of the HbS gene have shown that individuals with sickle trait are at higher
despite the devastating effects of the homozygous state. The risk of chronic kidney disease and venous thromboembo-
bS gene is inherited in an autosomal codominant fashion. lism. This simple heterozygous state generally has an
That is, heterozygous inheritance does not cause disease hemoglobin A:S ratio of approximately 60:40, because of
but is detectable (sickle cell trait); homozygous inheritance the greater electrostatic attraction of α-chains to βA rather
or compound heterozygous inheritance with another than bS chains. When the availability of α chains is limited
mutant β-globin gene results in disease. The sickle cell syn- by coinherited α-thalassemia, the A:S ratio is further
dromes include all conditions in which bS is inherited increased.

Pathophysiology themselves but rather is mediated through linked differences

in γ-chain (Hb F) production. The bS gene has been found to
The hallmark of sickle cell pathophysiology is the intraeryth-
be associated with five distinct haplotypes, referred to as the
rocytic polymerization of deoxyhemoglobin S. When deoxy-
Benin (Ben), Senegal (Sen), Central African Republic (CAR
genation of Hb S occurs, the normal conformational change
or Bantu), Cameroon (Cam), and Arab-Indian (Asian) hap-
of the tetramer exposes on its external surface a hydrophobic
lotypes. This is evidence that the bS gene arose by five sepa-
β6 valine (instead of the hydrophilic glutamate of Hb A),
rate mutational events. In general, the Asian and Sen
resulting in decreased solubility and a tendency of deoxyhe-
haplotypes are associated with a milder clinical course, and
moglobin S tetramers to aggregate or polymerize. The rate CAR is associated with a more severe course.
and degree of this polymerization determines the rheologic Although the deoxygenation-polymerization-sickling
impairment of sickle erythrocytes and the change in mor- axiom provides a basic understanding of sickle cell disease,
phology for which the condition was named. Polymerization there is an increasing appreciation that interactions of sickle
rate and extent are related to the intracellular concentration cells with other cells and proteins contribute to the hemo-
of Hb S, the type and fractional content of other hemoglo- lytic and vaso-occlusive processes. In vitro data show that
bins present (particularly Hb F), and percent oxygen satura- sickle erythrocytes exhibit abnormally increased adherence
tion. These variables correlate with the rate of hemolysis in to vascular endothelial cells as well as to subendothelial
sickle cell syndromes. extracellular matrix proteins. Apparent endothelial damage
Multiple factors determine the clinical manifestations of is demonstrated by the increased number of circulating
sickle cell disease. In addition to physiologic changes such as endothelial cells in sickle cell disease patients and by the
tissue oxygenation and pH, multiple genetic polymorphisms increase in such cells during vaso-occlusive crises. The dis-
and mutations may modify the presentation of the disease. ruption of normal endothelium results in the exposure of
This is best appreciated by examining the influence of the extracellular matrix components, including thrombospon-
coinheritance of other hemoglobin abnormalities on the din, laminin, and fibronectin. Endothelial cell receptors
effects of Hb S. For example, the coexistence of α-thalassemia include the vitronectin receptor aVb3 integrin and the
reduces the hemolytic severity as well as the risk of cerebro- cytokine-induced vascular cell adherence molecule-1
vascular accidents. High levels of fetal hemoglobin (Hb F) (VCAM-1). RBC receptors include CD36 (glycoprotein IV),
may substantially reduce symptoms as well as clinical conse- the αIVb1 integrin, the Lutheran blood group glycoproteins,
quences. Compound heterozygosity for a second abnormal and sulfatides. Vaso-occlusion thus may be initiated by
hemoglobin (eg, Hb C, D, or E) or β-thalassemia also modi- adherence of sickle erythrocytes to endothelial cells and
fies some of the manifestations of disease (discussed later in extracellular matrix molecules exposed during the process of
this section) (Table 7-2). endothelial damage and completed by trapping of sickled,
Several restriction fragment-length polymorphisms nondeformable cells behind this nidus of occlusion. Activa-
(RFLPs) may be identified in the vicinity of a known gene tion of blood coagulation resulting in enhanced thrombin
and define the genetic background upon which a disease- generation and evidence for platelet hyperreactivity have
causing mutation has arisen. For example, the coinheritance been demonstrated in patients with sickle cell disease during
of a defined set of RFLPs around the β-globin gene can steady-state and vaso-occlusive episodes. It has been sug-
define a disease-associated “haplotype” that marks the sickle gested that the exposure of RBC membrane phosphatidyl-
mutation within a specific population. These β-globin hap- serine and circulating activated endothelial cells in sickle cell
lotypes also have been associated with variations in disease disease patients contribute to the hypercoagulability by pro-
severity. This association is probably not related to the RFLPs viding procoagulant surfaces. The correlation of elevated
Table 7-2 Typical clinical and laboratory findings of the common forms of sickle cell disease after 5 years of age
Disease Clinical severity S (%) F (%) A2 (%)* A (%) Hemoglobin (g/dL) MCV (fL)
SS Usually marked >90 <10 <3.5 0 6-9 >80
Sβ0 Marked to moderate >80 <20 >3.5 0 6-9 <70
Sβ+ Mild to moderate >60 <20 >3.5 10-30 9-12 <75
SC Mild to moderate 50 <5 0† 0 10-15 75-85
S– HPFH Asymptomatic <70 >30 <2.5 0 12-14 <80
2 can be increased in the presence of Hb S, even in the absence of β-thalassemia. The classical findings are shown here.
*Hb A
†There will be 50% hemoglobin C that migrates near hemoglobin A on alkaline gel electrophoresis or isoelectric focusing.
2
HPFH = hereditary persistence of fetal hemoglobin.

white blood cell counts to increased mortality and adverse

outcomes identified by epidemiologic studies of sickle cell
disease patients suggest that neutrophils also participate in
vaso-occlusion. This concept has been further supported by
the precipitation of vaso-occlusive episodes with markedly
increased neutrophil counts associated with the administra-
tion of granulocyte colony-stimulating factor (G-CSF).
These findings together support the concept that the prod-
ucts of multiple genes as well as inflammatory cytokines
contribute to the pathology of sickle cell disease.
Laboratory features
The diagnosis of the sickle cell syndromes is made by the

identification of Hb S in erythrocyte hemolysates. Histori-
cally, cellulose acetate electrophoresis at alkaline pH was
used to separate Hb A, Hb A2, and Hb S; and citrate agar Figure 7-6 Irreversibly sickled cell. This peripheral blood film shows
electrophoresis at acidic pH was used to separate comigrat- an irreversibly sickled cell (ISC) that occurs in sickle cell anemia
(SS), Sβ0 thalassemia (double arrow). ISCs are rare in hemoglobin
ing Hb D and Hb C from Hb S and Hb A2, respectively. Cur-
SC and Sβ+ thalassemia. Also note the Howell-Jolly bodies in this
rently, high-performance liquid chromatography (HPLC)
view (single arrow). Source: ASH Image Bank/John Lazarchick
and isoelectric focusing are used in most diagnostic labora-
(image 00003961).
tories to separate Hbs. In both Hb SS and Sb0 thalassemia,
no Hb A is present. In Hb SS, however, the MCV is normal,
whereas in Hb Sb0 thalassemia, the MCV is reduced. Hb A2
approximately 10-25 days, resulting in moderate to severe
is elevated in Sb0 thalassemia, but it also can be nonspecifi-
hemolytic anemia, with a steady-state mean hemoglobin
cally elevated in the presence of Hb S, so an elevation of A2
concentration of 8 g/dL (ranging from 6 to 9 g/dL) in Hb SS
alone cannot reliably distinguish Hb SS from Sb0 thalas-
disease. The anemia is generally well tolerated because of
semia. In sickle cell trait and Sβ+ thalassemia, both Hb S and
compensatory cardiovascular changes and increased levels of
Hb A are identified. The A:S ratio is 60:40 in sickle trait (more
2,3-BPG. Several conditions are associated with acute or
A than S) and approximately 15:85 in Sβ+ thalassemia (more
chronic declines in the hemoglobin concentration, which
S than A). Microcytosis, target cells, anemia, and clinical
may lead to symptomatic anemia (Table 7-3). The transient
symptoms occur only in Sβ+ thalassemia and not in sickle
aplastic crisis resulting from erythroid aplasia is caused by
trait (Table 7-2). Review of the peripheral smear will reveal
human parvovirus infection, which may result in severe or
the presence of irreversibly sickled cells in Hb SS and Hb Sb0
life-threatening anemia. Lesser degrees of bone marrow
thalassemia (Figure 7-6), but only rarely in Sβ+ thalassemia
and Hb SC. Turbidity tests (for Hb S) are positive in all sickle
cell syndromes, including Hb AS (sickle trait). The classic
sickle cell slide test or “sickle cell prep” (using sodium Table 7-3 Causes of acute exacerbations of anemia in sickle cell disease
metabisulfite or dithionite) and the turbidity test detect only Cause Comment
the presence of Hb S, so they do not differentiate sickle cell Aplastic crisis Caused by human parvovirus; does not
disease from sickle cell trait. Therefore, they have limited recur
utility. Sickle cell disease can be diagnosed by DNA testing of Acute splenic Often recurrent in childhood before
the preimplanted zygote in the first trimester of pregnancy sequestration crisis splenic involution
using chorionic villus sampling, in the second trimester Acute chest syndrome Anemia may precede the onset of
using amniocentesis, or after birth using peripheral blood. respiratory signs and symptoms
Vaso-occlusive crisis Minimal decline only
Hypoplastic crisis Mild decline; accompanies many infections
Clinical manifestations
Accelerated hemolysis Infrequent; accompanies infection of
Two major physiologic processes, shortened RBC survival concomitant G6PD deficiency
(hemolysis) and vaso-occlusion, account for most of the Hepatic sequestration Rare
clinical manifestations of sickle cell disease. The erythrocyte Folate deficiency Rare, even in the absence of folate
(megaloblastic crisis) supplementation
life span is shortened from the normal 120 days to

“suppression” are associated with other infections. Sudden because of deficient production of natural antibodies. The
anemia may be caused by acute splenic sequestration in chil- risk for such infections persists into adulthood.
dren with Hb SS or Sb0 (and in adults with Hb SC or Sβ+ There are many important clinical differences among the
thalassemia) or, less frequently, hepatic sequestration, con- genotypes that cause sickle cell disease (Table 7-2). Hemo-
comitant glucose-6-phosphate dehydrogenase (G6PD) defi- globin SS is associated with the most severe anemia, most
ciency, or superimposed autoimmune hemolysis. Chronic frequent pain, and shortest life expectancy (median age,
exacerbations of anemia may be the result of folate or iron 42 years for men and 48 years for women in one large, but
deficiency or reduced erythropoietin levels due to chronic old, study), although there is tremendous heterogeneity in
renal insufficiency. Because of the chronic erythrocyte these variables even within this genotype. Hemoglobin Sb0
destruction, patients with sickle cell disease have a high thalassemia can closely mimic Hb SS, despite the smaller red
incidence of pigmented gallstones, which are often blood cells, lower MCH concentrations, and higher levels of
asymptomatic. Hb F and Hb A2 associated with this genotype. Patients with
The acute painful “vaso-occlusive crisis” is the stereotypi- Hb SC generally live longer lives (median age, 60 years for
cal and most common complication of sickle cell disease. men and 68 years for women) and have less severe anemia
These often unpredictable events are thought to be caused by (~20% are not anemic at all), higher MCH concentrations
obstruction of the microcirculation by erythrocytes and and less frequent pain, but they have more frequent ocular
other blood cells, leading to painful tissue hypoxia and and bone complications. Although Hb C does not enter into
infarction. They most commonly affect the long bones, back, the deoxyhemoglobin S polymer, patients with Hb SC have
chest, and abdomen. Acute pain events may be precipitated symptoms, whereas those with sickle cell trait (AS) do not.
by dehydration, cold temperatures, exercise (in particular This is thought to be caused by two important consequences
swimming), pregnancy, infection, or stress. Often no precipi- of the presence of Hb C: the Hb S content in Hb SC is 10%-
tating factor can be identified. Painful episodes may or may 15% higher than that seen in sickle trait (Hb S of approxi-
not be accompanied by low-grade fever. mately 50% vs. 40%), and the absolute intraerythrocytic
One of the first manifestations of sickle cell disease, acute concentration of total Hb is increased. The latter phenome-
dactylitis (hand-foot syndrome), results from bone marrow non results from persistent loss of cellular K+ and water from
necrosis of the hands and feet. The first attack generally these cells induced by the toxic effect of Hb C on cell mem-
occurs between 6 and 18 months of life, when the Hb F level branes. Another effect of this dramatic cellular dehydration
declines. Dactylitis is uncommon after age 3 years, as the site is the generation of target cells, which are far more prevalent
of hematopoiesis shifts from the peripheral to the axial skel- on the peripheral smear than sickled forms (Figure 7-7).
eton. Long-bone infarcts with pain and swelling may mimic Finally, in Hb SC disease, the increased hematocrit combined
osteomyelitis. Other skeletal complications of sickle cell dis- with the higher MCH concentration (MCHC) and cellular
ease include osteomyelitis, particularly due to Salmonella dehydration results in higher whole blood viscosity, which
and staphylococci, and avascular necrosis, especially of the
femoral and humeral heads.
Sickle cell disease is a multisystem disorder. Organ systems
subject to recurrent ischemia, infarction, and chronic dys-
function include the lungs (acute chest syndrome, pulmo-
nary fibrosis, pulmonary hypertension, hypoxemia), central
nervous system (overt and covert cerebral infarction, sub-
arachnoid and intracerebral hemorrhage, seizures, cognitive
impairment), cardiovascular system (cardiomegaly, conges-
tive heart failure), genitourinary system (hyposthenuria,
hematuria, proteinuria, papillary necrosis, glomerulone-
phritis, priapism), spleen (splenomegaly, splenic sequestra-
tion, splenic infarction and involution, hyposplenism),
eyes (retinal artery occlusion, proliferative sickle retinopa-
thy, vitreous hemorrhage, retinal detachment), and skin (leg
ulcerations). The risk of life-threatening septicemia and
Figure 7-7 Sickle-hemoglobin C disease. This peripheral blood film
meningitis because of encapsulated organisms, such as shows no irreversibly sickled cells, as expected for hemoglobin SC, but
Streptococcus pneumoniae, is increased markedly in children shows instead a large number of target cells and several dense,
with sickle cell disease. This susceptibility is related to func- contracted, and folded cells containing aggregated and polymerized
tional and anatomic asplenia and decreased opsonization hemoglobin.

Table 7-4 Important completed randomized clinical trials in sickle cell disease
Clinical trial Year Outcome

Penicillin Prophylaxis in Sickle Cell Disease (PROPS) 1986 Oral penicillin greatly reduces the incidence of invasive pneumococcal
infections in children.
Penicillin Prophylaxis in Sickle Cell Disease II 1995 Penicillin prophylaxis can be discontinued at 5 years of age.
(PROPS II)
Multicenter Study of Hydroxyurea in Patients With 1995 Hydroxyurea reduces the frequency of painful episodes and appears to reduce
Sickle Cell Anemia (MSH) the frequency of acute chest syndrome, transfusions, and hospitalizations.
National Preoperative Transfusion Study 1995 Simple transfusion to increase the Hb concentration to 10 g/dL is as effective
as exchange transfusion to reduce Hb S to <30%.
Stroke Prevention Trial in Sickle Cell Anemia (STOP) 1998 First overt stroke can be prevented with red blood cell transfusions in high-
risk children identified by transcranial Doppler (TCD) ultrasonography.
Optimizing Primary Stroke Prevention in Sickle Cell 2005 Discontinuation of prophylactic red blood cell transfusions after 30 months
Anemia (STOP 2) results in a high rate of reversion to abnormal TCD velocities and stroke.
Hydroxyurea to Prevent Organ Damage in Very Young 2011 Hydroxyurea starting at 9-18 months of age did not prevent splenic and renal
Children With Sickle Cell Anemia (BABY HUG) damage (the trial’s primary endpoints), but it did decrease the frequency
of dactylitis and painful episodes (secondary outcomes).
Stroke with Transfusions Changing to Hydroxyurea 2012 Terminated early due to futility for the primary composite endpoint of
(SWiTCH) recurrent stroke and resolution of iron overload. There was an excess
of recurrent strokes in the hydroxyurea arm (N = 7) compared with
continued transfusions (N = 0).
may increase the likelihood of vaso-occlusion. Patients with by some to prevent depletion of folate stores and megalo-
Hb Sβ+ thalassemia have less severe anemia and pain than blastic anemia related to chronic hemolysis, but this is prob-
patients with Hb Sb0 thalassemia. This is the result of smaller ably unnecessary in industrial countries where diets are
cells, lower MCHC, increased content of Hb F and Hb A2, better and flour is fortified with folate. Screening transcra-
and, most important, the presence of significant amounts nial Doppler (TCD) ultrasonography to determine risk of
(10%-30%) of Hb A that interferes with polymerization. overt stroke should be performed at least yearly for children
of age 2-16 years with Hb SS or Sb0 thalassemia (see further
discussion of TCD in the sections “Central nervous system
Treatment
disease” and “RBC transfusion” in this chapter). Ophthal-
Treatment of sickle cell disease includes general preventative mologic examinations should be performed periodically
and supportive measures, as well as treatment of specific beginning around age 10 years. Genetic counseling services
complications. The National Institutes of Health recently by trained individuals should be available for families with
published “Evidence-based management of sickle cell disease members having sickle cell syndromes.
expert panel report, 2014: guide to recommendations” which
is an excellent resource for addressing the spectrum of treat-
Painful episodes
ment issues. Table 7-4 summarizes the results of major clini-
cal trials influencing current clinical practice. Acute pain unresponsive to rest, hydration, and oral analge-
sics at home requires prompt attention and is the leading
cause for hospitalization. Painful episodes can be associated
Preventive interventions
with serious complications, and a high frequency of pain is a
Children should receive the 13-valent pneumococcal conju- poor prognostic factor for survival. It is essential to consider
gate vaccine (PCV-13), the 4-valent meningococcal conju- infectious and other etiologies of pain in the febrile patient.
gate vaccine (MCV-4), the 23-valent pneumococcal A complete blood count should be obtained. Because some
polysaccharide vaccine (PPV-23), and vaccines against Hae- degree of negative fluid balance often is present, oral or intra-
mophilus influenzae and hepatitis B virus, in addition to venous hydration is important. Caution must be used with
twice-daily penicillin prophylaxis at least until the age of intravenous hydration of adults, especially, who may have
5 years. Vaccinations against influenza on an annual basis decreased cardiac reserve. Prompt administration of analge-
and the pneumococcal vaccine at 5-year intervals (after the sics is a priority, and the selection of agents should be indi-
childhood PCV-13 and PPV-23 vaccinations) should be vidualized based on previous experience. Parenteral opioids,
administered to all patients. Folic acid supplements are used preferably morphine or hydromorphone, are often necessary

for both children and adults. The addition of nonsteroidal correction of any dehydration). Pain control to avoid exces-
anti-inflammatory drugs (NSAIDs), such as ibuprofen or sive chest splinting and the use of incentive spirometry are
ketorolac, may decrease the requirement for opioid analge- key adjunctive measures. Bronchodilator therapy is effective
sics but should be used with appropriate vigilance in sickle if there is associated reactive airway disease, which is particu-
cell disease because of potential nephrotoxicity. Maintenance larly common in children. Transfusion of RBCs should be
analgesia can be achieved with patient-controlled analgesia considered if there is hypoxemia or acute symptomatic, exac-
(PCA) pumps or with administration at fixed intervals. Con- erbation of anemia. Exchange transfusion should be per-
stant infusion of opioids requires close monitoring because formed for hypoxemia despite oxygen supplementation,
the hypoxia or acidosis resulting from respiratory suppres- widespread (bilateral, multilobar) infiltrates, and rapid clini-
sion is particularly dangerous. Meperidine is discouraged cal deterioration. Patients with acute chest syndrome are at
because of its short half-life and the accumulation of the risk for recurrences as well as subsequent chronic lung dis-
toxic metabolite normeperidine, which lowers the seizure ease. Preventive measures include hydroxyurea therapy and
threshold. Use of pain assessment instruments and attention chronic RBC transfusions.
to the level of sedation at regular intervals are necessary. Oxy-
gen supplementation is not required unless hypoxemia is
Central nervous system disease
present. The use of incentive spirometry has been shown to
reduce pulmonary complications in patients presenting with Without primary prevention, overt stroke may occur in 11%
chest or back pain. It has been demonstrated that the number of young sickle cell anemia patients (but is much less com-
of hospitalizations for painful events can be reduced by mon in SC disease and Sβ+ thalassemia), accounting for sig-
prompt intervention in an outpatient setting dedicated to nificant morbidity and mortality. The more frequent use of
sickle cell disease management. Nonpharmacologic manage- neuroimaging has identified a substantial incidence of sub-
ment techniques should be considered as well as evaluation clinical cerebrovascular disease, with 25%-40% of children
for depression for the patient with frequent episodes or having covert or silent strokes. The majority of overt strokes
chronic pain. Blood transfusion is not indicated in the treat- result from ischemic events involving large arteries with
ment of uncomplicated painful episodes. associated vascular endothelial damage, including intimal
and medial proliferation. Hemorrhagic events are more
common in adults and may result from rupture of collateral
Acute chest syndrome
vessels (moya moya) near the site of previous infarction. Sus-
The diagnosis of acute chest syndrome is based on a new picion of a neurologic event requires emergent imaging with
radiographic pulmonary infiltrate associated with symp- computed tomography (CT) to assess for hemorrhage fol-
toms such as fever, cough, and chest pain. As the nonspecific lowed by MRI. The acute management of overt stroke
term implies, various insults or triggers can lead to acute includes transfusion, usually by an exchange technique, to
chest syndrome. Young age, low Hb F, high steady-state reduce the Hb S percentage to <30%. Chronic transfusion
hemoglobin, and elevated white blood cell count in steady therapy to maintain the Hb S <30% decreases the chance of
state have been identified as risk factors. In a multicenter recurrent overt stroke but does not eliminate it. After
prospective study, bacterial (often atypical) or viral infec- 3-5 years of such transfusions and no recurrent neurologic
tions accounted for approximately 30% of episodes, whereas events, some physicians “liberalize” the transfusion regimen
fat emboli from the bone marrow were responsible for to maintain the Hb S <50%. The optimal duration of trans-
approximately 10% of events, with pulmonary infarction as fusions is not known, and they often are continued indefi-
another common suspected cause. In children, fever is a nitely. A randomized controlled trial (the SWiTCH study) of
common presenting symptom, whereas chest pain is more continued chronic transfusions versus hydroxyurea for long-
common in adults. Acute chest syndrome often develops in term secondary stroke prevention was stopped early due to
patients who initially present only with an acute painful futility, and there was an excess of recurrent strokes in the
event. Early recognition of the condition is of utmost impor- hydroxyurea arm (N = 7) compared with continued transfu-
tance because acute chest syndrome has become the leading sions (N = 0).
cause of death for both adults and children with sickle cell An abnormally increased TCD blood flow velocity can
disease. Management includes maintaining adequate oxy- identify children with Hb SS at high risk of primary overt
genation and administration of antibiotics to address the stroke. A randomized controlled trial of prophylactic trans-
major pulmonary pathogens and community-acquired fusions versus observation for children with abnormal TCD
atypical organisms. Fluid management needs particular velocities showed a reduced risk of the first stroke in patients
attention to prevent pulmonary edema by limiting oral and receiving transfusions (the STOP study). The use of hydroxy-
intravenous hydration to 1.0-1.5 times maintenance (after urea currently is being explored as means of primary stroke

prevention in a phase 3 multicenter randomized controlled prevented by chronic transfusion therapy. Transfusion has
trial for children with abnormal TCD velocities (the also been advocated for patients with severe pulmonary
TWiTCH study). hypertension and chronic nonhealing leg ulcers and to pre-
Silent cerebral infarcts are the most common neurologic vent recurrences of priapism, but clinical trial data are lack-
complications in children with sickle cell anemia. A random- ing. When chronic transfusion is indicated, RBCs may be
ized clinical trial assigned children ages 5-15 with sickle cell administered as a partial exchange transfusion, which may
anemia to receive regular blood transfusions or observations. offer a long-term advantage of delaying iron accumulation.
Regular blood-transfusion therapy significantly reduced the The goal of chronic transfusion is usually to achieve a nadir
incidence of the recurrence of cerebral infarct (6% in treat- total hemoglobin level of 9-10 g/dL with the Hb S under
ment arm versus 14% in observation arm). 30%-50%. It is important to avoid the hyperviscosity associ-
ated with hemoglobin levels >11-12 g/dL in the presence of
30% or more Hb S. Patients with Hb SC requiring transfu-
Pregnancy
sion pose special challenges, with the need to avoid hypervis-
Pregnancy poses some risk to the mother as well as to the cosity usually necessitating exchange transfusion (goal Hb A
fetus. Spontaneous abortions occur in approximately 5% of >70%) to ensure the hemoglobin concentration does not
pregnancies in sickle cell anemia, and preeclampsia occurs at exceed 11-12 g/dL.
an increased frequency in sickle cell disease. Preterm labor Preoperative transfusion in preparation for surgery under
and premature delivery are common. All patients should be general anesthesia may afford protection against periopera-
followed in a high-risk prenatal clinic, ideally at 2-week inter- tive complications and death but is probably not indicated in
vals with close consultation with a hematologist. Patients all cases, particularly minor procedures in children. In a mul-
should receive folic acid 1 mg/d, in addition to the usual pre- ticenter study, simple transfusion to a total hemoglobin level
natal vitamins, and should be counseled regarding the addi- of 10 g/dL afforded protection equal to partial exchange and
tional risks imposed by poor diet, smoking, alcohol, and was associated with less red cell alloimmunization. Another
substance abuse. Data do not support the routine use of pro- recently completed randomized trial (TAPS) included
phylactic transfusions. Simple or exchange transfusions, how- patients with sickle cell anemia undergoing low- or medium-
ever, should be instituted for the indications outlined risk surgery. Subjects were randomized to either preopera-
previously, as well as for pregnancy-related complications (eg, tive transfusion or no transfusion. Thirty-nine percent of
preeclampsia). Close follow-up is indicated postpartum when 33 patients in the no-preoperative-transfusion group had
the patient is still at high risk for complications. The option of clinically important complications, compared with 15% in
contraception with an intrauterine device, subcutaneous the preopeative-transfusion group (P = 0.023). Patients
implant, progesterone-only contraceptives, or condoms undergoing prolonged surgery or with regional compromise
should be discussed with all women of childbearing age. of blood supply (eg, during orthopedic surgery), hypother-
mia, or a history of pulmonary or cardiac disease may do
better with preoperative exchange transfusion. Transfusions
RBC transfusion
also may be useful for some patients preparing for intrave-
Patients with sickle cell disease often receive transfusions nous ionic contrast studies, dealing with chronic intractable
unnecessarily. RBC transfusions, however, may be effective pain, or facing complicated pregnancy. Transfusions are not
for certain complications of the disease. Transfusion is indi- indicated for the treatment of steady-state anemia, uncom-
cated as treatment of specific acute events, including moder- plicated pain events, uncomplicated pregnancy, most leg
ate to severe acute splenic sequestration, symptomatic ulcers, or minor surgery not requiring general anesthesia.
aplastic crisis, cerebrovascular accident (occlusive or hemor- Up to 30% of patients with sickle cell disease who repeat-
rhagic), acute ocular vaso-occlusive events, and acute chest edly undergo transfusion will become alloimmunized to
syndrome with hypoxemia. Although the first two events RBC antigens (especially E, C, and Kell), and this risk
only require correction of anemia and thus are treated with increases with increasing exposure. Alloimmunization pre-
simple transfusion, stroke, ocular events, and severe acute disposes patients to delayed transfusion reactions. Severe
chest syndrome are best treated with exchange transfusion painful crises with a decrease in the hemoglobin level within
aimed at decreasing the percentage of Hb S to <30% and days to weeks of a transfusion should alert the clinician to
increasing the Hb level to 9-10 g/dL. In addition, transfu- consider this diagnosis. Identification of a new alloantibody
sions are indicated for the prevention of recurrent strokes as may not be made acutely, and reticulocytopenia can be an
well as for the treatment of high-output cardiac failure. As associated finding. In this situation, additional transfusions
mentioned, an abnormal TCD velocity can identify children are hazardous and should be avoided if at all possible. Uni-
with Hb SS at high risk of primary overt stroke, which can be versal RBC phenotyping and matching for the antigens of

greatest concern (eg, C, D, E, and Kell) can minimize engraftment and minimal graft-versus-host disease in adults.
alloimmunization. These efforts are undergoing further development and, as
the procedure becomes more refined, stem cell transplanta-
tion for patients with sickle cell disease could be greatly
Modifying the disease course
expanded.
In addition to chronic transfusions, two other disease-
modifying treatments currently are available: (i) hydroxy-
urea, which is ameliorative; and (ii) hematopoietic stem cell Clinical case (continued)
transplantation, which is curative. On the basis of knowledge The case in this section describes a patient with sickle cell
that patients with high hemoglobin F levels have less severe anemia who has experienced pain episodes but no other major
disease, many investigators tested a variety of experimental complications related to his disease. He is admitted for a pain
strategies for pharmacologic induction of hemoglobin F crisis, and multiorgan failure ensues. Acute multiorgan failure is a
production and identified hydroxyurea as efficacious and well-described complication of sickle cell disease. High base-
practical. A multicenter, randomized, placebo-controlled line hemoglobin levels may represent a key risk factor. Acute
trial then found that daily oral administration of hydroxy- multiorgan failure often is precipitated by a severe acute pain
crisis and is thought to be secondary to widespread intravascu-
urea significantly reduced the frequency of pain episodes,
lar sickling, fat embolization, and subsequent ischemia within
acute chest syndrome, and transfusions in adult Hb SS
affected organs. Aggressive transfusion therapy can be lifesaving
patients (MSH study). No serious short-term adverse effects
and result in complete recovery.
were observed, although monitoring of blood counts was
required to avoid potentially significant cytopenias. Interest-
ingly, the therapeutic response to hydroxyurea sometimes
Key points
precedes or occurs in the absence of a change in Hb F levels,
suggesting that a reduction in white blood cell count and • The clinical manifestations of sickle cell disease are primarily
other mechanisms may be beneficial. Laboratory studies due to hemolysis and vaso-occlusion.
revealed that hydroxyurea reduced adherence of RBCs to • Multiple cellular and genetic factors contribute to the
phenotypic heterogeneity of sickle cell disease.
vascular endothelium, improved RBC hydration, and
• The hemoglobin F level is a major determinant of clinical
increased the time to polymerization. Follow-up at 9 years
manifestations and outcomes.
indicates that patients taking hydroxyurea seem to have
• Pneumococcal sepsis is now uncommon, but it remains a
reduced mortality without evidence for an increased inci- potential cause of death in infants and young children, so
dence of malignancy. Classical indications for hydroxyurea universal newborn screening, compliance with penicillin
include frequent painful episodes, recurrent acute chest syn- prophylaxis, and vaccination remain a priority.
drome, severe symptomatic anemia, and other severe vaso- • Human parvovirus infection is the cause of aplastic crisis.
occlusive events. Given the safety of hydroxyurea and that • Splenic sequestration is a consideration in the differential
Hb SS is a morbid condition, many clinicians use hydroxy- diagnosis of a sudden marked decrease in the hemoglobin
urea more liberally even when the classical indications for concentration.
hydroxyurea therapy are not present. There are now guide- • There are differences in frequency of clinical events and
line recommendations to strongly consider the use of survival among the various genotypes of sickle cell disease.
• Sickle cell disease is a leading cause of stroke in young
hydroxyurea in patients with sickle cell anemia who have
individuals, and a substantial incidence of covert or silent
daily chronic pain that interferes with quality of life. Clinical
infarctions recently has been appreciated.
trials of hydroxyurea in children also show a reduction in the
• A randomized clinical trial has demonstrated efficacy of red
frequency of painful episodes, but no convincing evidence cell transfusion in preventing first stroke in children with
yet indicates that early hydroxyurea therapy prevents or abnormal TCD velocity.
delays the onset of organ damage. Pregnancy should be • A randomized clinical trial demonstrated that preoperative
avoided while taking hydroxyurea. Hematopoietic stem cell simple transfusion was as effective as exchange transfusion. The
transplantation has been used primarily for children with preoperative management of the older patient, particularly with
stroke or other severe disease manifestations, with an event- cardiac or pulmonary dysfunction, has not been defined.
free survival rate of >80%. In most centers, few patients meet • A randomized, placebo-controlled clinical trial has established
the usual eligibility criteria, which includes an HLA-matched the efficacy of hydroxyurea in reducing the frequency of painful
sibling donor. Alternative donor sources such as umbilical episodes and acute chest syndrome. A follow-up study suggests
a reduction in mortality for patients taking hydroxyurea.
cord blood are now being used. Novel approaches such as
• The causes of acute chest syndrome include infection, fat
nonmyeloablative conditioning regimens and haplotype
embolism, and pulmonary infarction.
transplants have shown early success with up to 60%

Hemoglobin E Hemoglobin C
Hb E is a β chain variant with highest frequency in Southeast Hb C is the third most common mutant hemoglobin, after
Asia. The highest prevalence occurs in Myanmar and Thailand, Hb S and Hb E. The Hb C mutation arose in West Africa. The
where the gene frequency may approach 70% in certain prevalence in African Americans is 2%-3%. The hemoglobin
regions. The gene frequency is also high in Laos, Cambodia, mutant results from the substitution of lysine for glutamic
and Vietnam. It is also found in Sri Lanka, North East India, acid as the sixth amino acid of β-globin, the consequence of
Nepal, Bangladesh, Malaysia, Indonesia, and the Philippines. It a single nucleotide substitution (GAG/AAG) in the sixth
has become more common in the United States during the past codon (Figure 7-5). The resultant positive-to-negative charge
20-30 years as a result of immigration. The structural change is difference on the surface of the Hb C tetramer results in a
a substitution of glutamic acid by lysine at the 26th position of molecule with decreased solubility of both the oxy and deoxy
the β-globin chain (Figure 7-4). The mutation is also thalasse- forms that may undergo intraerythrocytic aggregation and
mic because the single-base GAG/AAG substitution creates a crystal formation. Hb C stimulates the K:Cl cotransport sys-
cryptic splicing site, which results in abnormal mRNA process- tem, promoting water loss and resulting in dehydration and
ing and reduction of mRNA that can be translated. Hb E is also poorly deformable RBCs that have a predilection for entrap-
slightly unstable in the face of oxidant stress. Hb E is some- ment within the spleen. Consequently, patients with Hb CC
times referred to as a “thalassemic hemoglobinopathy.” and patients with Hb C-β thalassemia have mild chronic
Individuals with hemoglobin E trait are asymptomatic hemolytic anemia and splenomegaly. Patients may develop
with or without mild anemia (hemoglobin >12 g/dL), and cholelithiasis, and the anemia may be more exaggerated in
mild microcytosis. Peripheral smear may be normal or may association with infections. Heterozygous individuals (Hb C
show hypochromia, microcytosis, target cells, irregularly trait) are clinically normal, however, identifying the diagno-
contracted cells, and basophilic stippling. HbE usually makes sis is important for genetic counseling. The coinheritance of
up 30% or less of total hemoglobin. The HbE concentration Hb S and Hb C results in a form of sickle cell disease, Hb SC
will be lower with the coinheritance of α-thalassemia. (see the section “Sickle cell disease” in this chapter).
Homozygotes (Hb E disease) are usually asymptomatic with Laboratory studies in Hb CC show a mild hemolytic ane-
no overt hemolysis or splenomegaly. Individuals may have mia, microcytosis, and slightly elevated reticulocyte counts.
mild anemia, microcytosis (MCV approximately 65-69 fL in The MCHC is elevated because of the effect of Hb C on cel-
adults and 55-65 fL in children), and reduced MCH. Periph- lular hydration. The peripheral blood smear shows promi-
eral smear will show hypochromia, microcytosis, and a vari- nent target cells, microcytosis, and irregularly contracted red
able number of target cells and irregularly contracted cells. cells. RBCs containing hemoglobin crystals also may be seen
HbE plus HbA2 will make up 85%-99% of the total hemo- on the blood smear, particularly in patients who have had
globin. The compound heterozygous state, HbE-β thalas- splenectomy. Individuals with Hb C trait have normal hemo-
semia, results in a very variable phenotype ranging from globin levels, and microcytosis is common. The peripheral
thalassemia trait, thalassemia intermedia to thalassemia smear may be normal or may show microcytosis and target
major depending on the β mutation. It is now one of the cells. Confirmation of the diagnosis requires identification of
more common forms of thalassemia in the United States. It Hb C; which comigrates with Hb A2, Hb E, and Hb OArab on
is characterized by microcytic anemia, with mildly increased cellulose acetate electrophoresis and isoelectric focusing.
reticulocytosis. The peripheral smear will include anisocyto- Thus, Hb C must be distinguished by citrate gel electropho-
sis, poikilocytosis, hypochromia, microcytosis, target cells, resis or HPLC.
nucleated red blood cells, and irregularly contracted cells. Specific treatment for patients with Hb CC is not gener-
HbE-b0 thalassemia is associated with a mostly HbE electro- ally necessary.
phoretic pattern, with increased amounts of HbF and HbA2.
The electrophoretic pattern in HbE-β+ thalassemia is similar
Hemoglobin D
except for the presence of approximately 15% HbA. HbE
comigrates with HbC and HbA2 on cellulose acetate electro- Hb D is usually diagnosed incidentally. Hb DPunjab (also
phoresis and isoelectric focusing. called HbDLos Angeles) results from the substitution of gluta-
Patients with HbE disease are usually asymptomatic and mine for glutamic acid at the 121st position of the β-chain
do not require specific therapy. However, patients who coin- (Figure 7-4). This mutant has a prevalence of approximately
herited HbE and β-thalassemia, especially those with 3% in the Northwest Punjab region of India but also is
HbE-b0, may have significant anemia. Some need intermit- encountered in other parts of the world. Patients who are
tent or chronic RBC transfusions, and some may benefit homozygous (Hb DD) may have a mild hemolytic anemia.
from splenectomy. Individuals who are heterozygous (Hb AD) are clinically

normal with normal blood counts and a peripheral smear due to denaturation and subsequent pitting), “bite cells,”
with the occasional target cell. The major clinical relevance and basophilic stippling may occur. The evaluation includes
of Hb D is with compound heterozygous inheritance with hemoglobin electrophoresis (which is often normal), crystal
HbS, resulting in a form of sickle cell disease, perhaps as a violet Heinz body staining, and the isopropanol stability
result of the low-affinity of Hb D promoting hemoglobin test. The isopropanol test may be falsely positive in the neo-
deoxygenation. The diagnosis of Hb AD (D trait) or Hb DD nate due to high fetal hemoglobin levels, so the heat-
is made by hemoglobin electrophoresis. Hb S and Hb D have stability test should be used during the first months of life.
similar electrophoretic mobility on alkaline cellulose acetate Management includes avoidance of oxidant agents, and
electrophoresis and isoelectric focusing. They can be differ- some recommend supplementation with folic acid. Splenec-
entiated by acid citrate agar electrophoresis, HPLC, or solu- tomy may be useful for patients with severe hemolysis and
bility studies. This distinction is important for genetic and splenomegaly. The risk of thrombosis is high after splenec-
prognostic counseling. tomy in individuals with a severely unstable hemoglobin,
and thus, patients should be educated and closely evaluated
in this regard.
Key points Hemoglobin M includes various α, β, and γ chain variants
that are associated with an increased tendency to oxidation
• Hemoglobins C, D, and E are common hemoglobin variants
to methemoglobin resulting in cyanosis and hemolysis.
that can have significant consequences when coinherited with
Some unstable hemoglobins also may have altered oxygen
hemoglobin S.
affinity, which could exacerbate (decreased oxygen affinity)
• Homozygosity for hemoglobin E (EE) is a mild condition, but
compound heterozygosity for HbE and β-thalassemia can be a
or ameliorate (increased oxygen affinity) the degree of
clinically significant thalassemia syndrome. anemia.
Abnormalities of the RBC membrane

Unstable hemoglobin
Clinical case
Unstable hemoglobin variants are inherited in an autosomal
A 36-year-old woman is referred for evaluation of moderate
dominant pattern, and affected individuals are usually het-
anemia. She has been told she was anemic as long as she can
erozygotes. Unstable hemoglobins constitute one of the larg- remember, and she has intermittently been prescribed iron. She
est groups of hemoglobin variants, although individually, occasionally has mild fatigue but is otherwise asymptomatic.
each is rare. In both Hb Köln (β98 Val/Met substitution) and Her past history is significant only for intermittent jaundice and
In Hb Zurich (β63 His/Arg), the amino acid substitution a cholecystectomy for gallstones at age 22 years. She takes no
destabilizes the heme pocket. Other mechanisms that desta- medications. A cousin and an aunt have also had anemia and
bilize hemoglobin include (i) alteration of the a1b1 interface jaundice. Her examination is significant for mild splenomegaly.
region (eg, Hb Tacoma, b30 Arg/Ser); (ii) distortion of the Prior laboratory data reveal hematocrit values between 29% and
helical configuration of structurally important regions 33%. Today’s hematocrit is 27%, MCV 98 fL, MCHC 38 g/dL.
(eg, Hb Bibba, a136 Leu/Pro); and (iii) introduction of the Corrected reticulocyte count is 7%. Review of the peripheral
blood smear reveals numerous spherocytes.
interior polar amino acid (eg, Hb Bristol, β67 Val/Asp).
Unstable γ-chain variants (eg, Hb Poole, g130 Trp/Gly) can
cause transient hemolytic anemia in the neonate that will Hereditary spherocytosis (HS), hereditary elliptocytosis
spontaneously resolve. (HE), and hereditary pyropoikilocytosis (HPP) are a hetero-
These abnormal hemoglobins precipitate spontaneously geneous group of disorders with a wide spectrum of clinical
or with oxidative stress. Precipitated hemoglobin inclusions manifestations. This group of disorders is characterized by
(Heinz bodies seen using a supravital stain) impair erythro- abnormal shape and flexibility of RBCs because of a defi-
cyte deformability, resulting in premature erythrocyte ciency or dysfunction of one or more of the membrane pro-
destruction by macrophages of the liver and spleen. The teins, which leads to shortened RBC survival (hemolysis).
severity of the hemolysis varies with the nature of the muta- Multiple genetic abnormalities, including deletions, point
tion but may be accelerated by fever or ingestion of oxidant mutations, and defective mRNA processing, have been iden-
drugs. tified as underlying causes. The HS syndromes generally are
An unstable hemoglobinopathy should be suspected in a due to private mutations unique to each kindred. In contrast,
patient with hereditary nonspherocytic hemolytic anemia. some HE syndromes are due to specific mutations in indi-
The hemoglobin level may be normal or decreased. Hypo- viduals from similar locales (eg, Melanesian elliptocytosis),
chromia of the RBCs (resulting from loss of hemoglobin suggesting a founder effect.

RBC membrane protein composition and assembly involve the assembly of α- and β-spectrin chains into het-
erodimers, which self-associate to form tetramers. Because
The RBC membrane consists of a phospholipid-cholesterol
the distal ends of spectrin bind to actin, with the aid of pro-
lipid bilayer intercalated by integral membrane proteins such
tein 4.1 and other minor proteins, a contractile function of
as band 3 (the anion transport channel) and the glycophorins
the cytoskeleton may be important for normal RBC survival.
(Figure 7-8). This relatively fluid layer is stabilized by attach-
Conceptually, HS is caused by defects in vertical protein–
ment to a membrane skeleton. Spectrin is the major protein
protein interactions in the RBC membrane, whereas HE is
of the skeleton, accounting for approximately 75% of its mass.
caused by defects in horizontal interactions (Figure 7-8).
The skeleton is organized into a hexagonal lattice. The hexa-
gon arms are formed by fiber-like spectrin tetramers, whereas
the hexagon corners are composed of small oligomers of Hereditary spherocytosis
actin that, with the aid of other proteins (4.1 and adducin), HS is common in individuals of Northern European descent
connect the spectrin tetramers into a two-dimensional lattice. with an occurrence of approximately 1 in 2,000. Penetrance
The membrane cytoskeleton and its fixation to the lipid-pro- is variable, and the prevalence of a clinically recognized dis-
tein bilayer are the major determinants of the shape, strength, order is much lower. In 75% of cases, the inheritance pattern
flexibility, and survival of RBCs. When any of these constitu- is autosomal dominant with sporadic cases representing the
ents are altered, RBC survival may be shortened. remaining 25%, half of which represent an autosomal reces-
A useful model to understand the basis for RBC membrane sive inheritance pattern and the other half de novo muta-
disorders divides membrane protein–protein and protein– tions. HS is characterized by spherocytic, osmotically fragile
lipid associations into two categories. Vertical interactions are RBCs and is both clinically and genetically heterogeneous.
perpendicular to the plane of the membrane and involve a
spectrin–ankyrin–band 3 association facilitated by protein 4.2
Pathophysiology
and attachment of spectrin–actin–protein 4.1 junctional
complexes to glycophorin C. Horizontal interactions, which The pathophysiology of HS generally involves aberrant
are parallel to and underlying the plane of the membrane, interactions between the skeleton and the overlying lipid
3 CD47 3
RhAG
4.2
Hereditary Spectrin
spherocytosis
Ank 4.2
4.1
β
α
4.1
β
α
Spectrin-ankyrin- Spectrin
band 3 self-association
Hereditary
elliptocytosis
Spectrin-actin-4.1
Figure 7-8 Membrane defects in hereditary spherocytosis and elliptocytosis. Redrawn with permission of Elsevier from Lux S. In: Orkin S et al.
eds. Nathan and Oski’s Hematology and Oncology of Infancy and Childhood, 8th ed., Philadelphia, PA: W.B. Saunders; 2015:515-579.

Table 7-5 Defects of red blood cell membrane proteins in hereditary spherocytosis, elliptocytosis, and pyropoikilocytosis
Class of defect Hereditary spherocytosis Hereditary elliptocytosis and pyropoikilocytosis
Protein deficiency Spectrin Spectrin†

Ankyrin* Protein 4.1
Band 3 Glycophorin C
Protein 4.2
Protein dysfunction β-spectrin abnormality affects Defective spectrin dimer self-association due to spectrin mutations
β-spectrin-protein 4.1 interaction* Protein 4.1 abnormality affects b-spectrin-protein 4.1 interaction
*Red cells of the patients are also partially deficient in spectrin.
†Seen in patients with hereditary pyropoikilocytosis in cases in which it coexists with a spectrin mutation that affects spectrin self-association.
bilayer (“vertical interactions”). A common epiphenomenon evaluation for unrelated conditions. Patients with moderately
in HS RBCs is a varying degree of spectrin loss, which is usu- severe disease may present with several additional complica-
ally due to a defect in one of the membrane proteins involved tions. Aplastic crisis, which may be the initial presentation
in the attachment of spectrin to the membrane rather than a for some patients, may require urgent attention. The cause
primary defect in the spectrin molecule itself. Spectrin as the of aplastic crisis is human parvovirus infection, which pro-
major protein of the skeleton forms a nearly monomolecular duces selective suppression of erythropoiesis, resulting in
submembrane layer that covers most of the inner-membrane reticulocytopenia and inability to compensate for ongoing
surface; therefore, the density of this skeletal layer in HS RBC destruction. In contrast, the “hyperhemolytic crisis” is
erythrocytes is reduced. Consequently, the lipid bilayer is characterized by accelerated hemolysis, leading to increased
destabilized, leading to loss of membrane lipid and, thus, jaundice and splenic enlargement, which is a common prob-
surface area through microvesiculation. The result of these lem in children. Other complications include the rare mega-
changes is a progressively spheroidal RBC. The inherent loblastic crisis secondary to acquired folic acid deficiency
reduced deformability of spherocytes makes it difficult for usually associated with high-demand situations, such as
them to traverse the unique constraining apertures that pregnancy. Leg ulcerations have been rarely reported.
characterize splenic vascular walls. The spleen “conditions” Patients with severe hemolysis and resulting expansion of the
RBCs, enhancing membrane loss. Retained and further dam- erythroid compartment in the bone marrow can develop
aged by the hypoxic and acidic environment in the spleen, maxillary hyperplasia interfering with dentition or extra-
they ultimately are destroyed prematurely. medullary hematopoietic masses that may mimic malig-
The molecular basis of HS is heterogeneous (Table 7-5). nancy. Patients may manifest a variety of issues attributable
A deficiency or defect of the ankyrin molecule represents the to splenomegaly, including early satiety, left upper-quadrant
most common cause of dominant HS. In 30%-45% of cases, fullness, and hypersplenism. HS may be diagnosed in the
the defect includes both ankyrin and spectrin deficiency, in neonatal period based on a positive family history or marked
30% spectrin only, and in 20% band 3 mutations. Various jaundice. The diagnosis also should be considered in patients
mutations of the ankyrin gene have been identified. Multiple of all ages with intermittent jaundice, mild “refractory” ane-
band 3 mutations have been described. Although less fre- mia, or splenomegaly. Rare associated syndromes suggest
quent, mutations of the β-spectrin gene have been found in that mutant RBC membrane proteins may reside in other tis-
autosomal dominant HS, whereas α spectrin gene abnor- sues. For example, distal renal tubular acidosis may occur in
malities have been identified only in recessively inherited HS. HS patients harboring mutant band 3 (the anion channel
Mutations in the protein 4.2 gene have been found primarily protein).
in Japanese patients with autosomal recessive HS.
Laboratory evaluation
In addition to the usual laboratory abnormalities indicating
The clinical expression ranges from an asymptomatic and hemolysis, the principal diagnostic feature is the identifica-
often undiagnosed condition with nearly normal hemoglo- tion of spherocytes on the peripheral blood smear
bin levels (compensated hemolysis) to severe hemolysis and (Figure 7-9a). The extent of spherocytosis is variable and, in
anemia. Patients with mild HS have a relatively uneventful mild cases, it may be missed even by the experienced clini-
course, although some may develop pigmented (bilirubi- cian. Additional morphologic abnormalities, including cells
nate) gallstones in childhood or adult life. Mildly anemic with membrane extrusions and elliptocytes, may be observed.
patients may be diagnosed later in life as adults during The RBC indices may provide a clue, with an increase in the

from HE, HPP, and stomatocytosis. A different test for HS

and related cytoskeleton-associated hemolytic anemias is the
eosin-5-maleimide (EMA) binding assay. EMA binds to
band 3 on RBCs, and a reduction in binding, measured by
fluorescence intensity, corresponds to a quantitative reduc-
tion in erythrocyte band 3, consistent with HS. The EMA-
binding test has a higher predictive value for HS than
standard osmotic fragility testing. Review of the complete
blood count, reticulocyte count, and peripheral smear from
family members may prove helpful. The differential diagno-
sis for spherocytes and increased osmotic fragility include
autoimmune hemolytic anemia, so a direct antiglobulin test
(DAT) should be performed as part of the evaluation when
the family history is negative. Likewise, HS should be consid-
ered in the differential diagnosis of DAT-negative hemolytic
anemia.
Treatment
As with other hemolytic anemias, folic acid supplementation

should be considered for patients with severe anemia, even
though overt folic acid deficiency rarely is encountered in the
industrial nations due to supplementation in grain products.
Patients need to be aware of the signs and symptoms of
aplastic and hyperhemolytic crises to seek prompt medical
attention. The definitive treatment of HS is splenectomy,
which ameliorates the hemolytic anemia in almost all
patients, although the underlying intrinsic defect of the cir-
culating RBCs is not altered. In rare patients with HS and
severe hemolysis, the procedure markedly diminishes the
hemolytic rate but may not fully correct the anemia. Con-
trolled clinical trial data are not available to provide guide-
lines in making the decision to recommend splenectomy.
Figure 7-9 Peripheral blood findings in inherited disorders of the red Thus, the indications for splenectomy are somewhat contro-
cell membrane: (a) numerous spherocytes (arrows), (b) numerous versial, but the prevailing view advocates surgery for patients
elliptocytes and a rod-shaped cell (arrow), and (c) marked with symptomatic hemolytic anemia or its complications,
poikilocytosis. particularly gallstones. Additional considerations for sple-
nectomy in the pediatric population include failure to thrive,
recurrent hyperhemolytic episodes, or complications of
MCHC (due to cellular dehydration) even in the context of chronic anemia, including a hypermetabolic state. The lapa-
minimal anemia. The osmotic fragility test using increas- roscopic technique often is preferred to open splenectomy.
ingly hypotonic saline solutions will support the diagnosis Accessory spleens are common, so a thorough search should
with the finding of increased RBC lysis compared with nor- be performed at the time of splenectomy. The patient should
mal RBCs. Sensitivity of the test is enhanced by 24-hour receive pneumococcal, H. influenzae type b, and meningo-
incubation at 37°C, but mild cases still can be missed by the coccal vaccines before the procedure, and pediatric patients
test. Osmotic fragility testing is a test for spherocytes of any usually receive prophylactic penicillin for at least several
cause, not a specific test for HS. Related to standard osmotic years thereafter to reduce the risk of bacterial sepsis. Throm-
fragility testing is the newer osmotic gradient ektacytometry, boembolic events may occur following splenectomy,
which measures deformability of whole RBCs measured as a although data are limited. Because of the increased fre-
function of osmolality using a laser-diffraction viscometer. quency of postsplenectomy infections in young children,
Ektacytometry appears to be more sensitive than standard splenectomy should not be performed before the age of
osmotic fragility testing, and it can help differentiate HS 5 years except in patients with particularly severe disease.

Partial splenectomy has been advocated to resolve the ane- severe forms, rod-shaped cells, poikilocytes, and fragments
mia of HS yet maintain some residual splenic phagocytic (Figure 7-9b); (ii) spherocytic HE, a phenotypic hybrid
function. Long-term results of partial splenectomy (4-6 between HE and HS; and (iii) Southeast Asian ovalocytosis
years) in small observational studies are promising, but the with unique spoon-shaped erythrocyte morphology. Clini-
spleen may increase in size and the hemoglobin concentra- cal manifestations range from asymptomatic carrier state to
tion may fall after splenectomy. Markers of splenic function severe transfusion-dependent hemolytic anemia with poi-
indicate variable degrees of residual activity, but postopera- kilocytosis and erythrocyte fragmentation. In most cases, the
tive penicillin is recommended. inheritance of HE is autosomal dominant. The exception is
HPP, a rare and severe variant of common HE that is reces-
Clinical case (continued) sively inherited (Figure 7-9c).
The patient presented in this section should be suspected of

having HS. It is not uncommon for the diagnosis to be made in Pathophysiology
adulthood, as patients with mild or moderate disease are often
well compensated. An elevated reticulocyte count, elevated
The underlying defects involve horizontal interactions
MCHC, intermittent jaundice, history of gallstones, and sphero- between proteins of the membrane skeleton, especially
cytes on peripheral smear all support the diagnosis. The diag- spectrin–spectrin and spectrin–protein 4.1 interactions.
nosis should be confirmed by demonstrating increased osmotic These defects weaken the skeleton. Under the influence of
fragility of RBCs, especially if the spherocytosis is not obvious on shear stress in the microcirculation, the cells progressively
the peripheral blood film, and by a negative DAT. Family mem- lose the ability to regain the normal disc shape and are stabi-
bers should be evaluated for anemia. lized in the elliptocytic or poikilocytic shape. In severely
affected patients, the weakening of the skeleton grossly dimin-
ishes membrane stability, leading to RBC fragmentation.
Different underlying molecular defects have been identi-
Key points
fied in common HE, consistent with the heterogeneous
• HS is the most common inherited hemolytic anemia of nature of the disorder (Table 7-5). In the majority of cases,
individuals from Northern Europe. patients have mutant α- or β-spectrin, resulting in defective
• Abnormalities in ankyrin, spectrin, band 3, and protein 4.2 self-association and an increased percentage of spectrin het-
(“vertical interactions”) that result in a reduction in the quantity erodimer in the membrane. A partial or complete absence or
of spectrin account for the red cell membrane loss characteristic
dysfunction of protein 4.1 occurs in some patients with mis-
of HS.
sense and deletion mutations. Patients with HPP appear to
• HS should be suspected in cases of direct antiglobulin
be compound heterozygotes. Coinheritance of a mutation
test-negative hemolytic anemia when spherocytes are identified
on the peripheral blood smear. A positive family history is
leading to spectrin deficiency and a mutation of spectrin
supportive of the diagnosis. resulting in a qualitatively defective molecule has been iden-
• Clinical manifestations of HS vary from a lack of symptoms to tified in some patients with the condition. Southeast Asian
severe hemolysis. ovalocytosis is prevalent among certain ethnic groups in
• The diagnosis of HS can be supported by the osmotic fragility Malaysia, the Philippines, Papua New Guinea, and probably
test, the sensitivity of which is increased with incubation at 37°C. other Pacific countries as well. It is an asymptomatic condi-
Osmotic gradient ektacytometry and the EMA-binding assay are tion characterized by rigid RBCs of a unique spoon-shaped
newer, often more sensitive, tests for membranopathies like HS morphology. Affected individuals are heterozygous for a
and HE. mutation of band 3.
• Splenectomy decreases hemolysis and reduces gallstone
formation, but it should be reserved for symptomatic or severe
patients. Clinical manifestations, laboratory evaluation,
and treatment
HE must be differentiated from a variety of other conditions

Hereditary elliptocytosis and hereditary
in which elliptocytes and poikilocytes commonly are found
pyropoikilocytosis
on the peripheral blood smear, including iron deficiency,
The clinical presentation, inheritance, and alteration in RBC thalassemia, megaloblastic anemia, myelofibrosis, and
shape and physical properties and the underlying molecular myelodysplasia. As opposed to HE, however, the percentage
defects are considerably more heterogeneous in HE than in of elliptocytes in these other conditions usually does not
HS. Three distinct subtypes are distinguished: (i) common exceed 60%. The presence of elliptocytes and evidence of
HE, characterized by biconcave elliptocytes and, in more dominant inheritance of elliptocytosis in other family

members differentiate HE from the previous conditions.

Whereas most patients with common HE are asymptomatic,
occasional patients who are homozygotes or compound het-
erozygotes for 1 or 2 molecular defects have more severe
hemolytic disease. African American neonates with common
HE may have severe hemolysis, with striking RBC abnormal-
ities similar to HPP, which abates during the initial months
of life. Approximately 10% of HE patients and all HPP
patients have mild-to-moderate anemia with clinical fea-
tures of pallor, jaundice, anemia, and gallstones. The most
severe form of elliptocytosis, HPP, typically is inherited
recessively and is characterized by a striking micropoikilo-
spherocytosis and fragmentation with some elliptocytes. A
markedly low MCV, typically in the range of 50 to 60 fL, may
be observed. In HPP, RBCs are thermally unstable and frag-
ment at temperatures of 46°C-48°C, reflecting the presence
Figure 7-10 Acanthocytes.
of mutant spectrin in the cells. Additional specialized labora-
tory investigation includes separation of solubilized mem-
brane proteins by polyacrylamide gel electrophoresis, which Of significant clinical importance is the recognition that
may reveal either an abnormally migrating spectrin or a defi- patients with hereditary stomatocytosis have an increased risk
ciency or abnormal migration of protein 4.1. An increased of developing thrombotic events after splenectomy. Acquired
fraction of unassembled dimeric spectrin in the extract can stomatocytosis can be seen in acute alcoholism and hepatobi-
be detected by electrophoresis of extracts under non- liary disease (although target cells are more common) and
denaturing conditions. occasionally in malignant neoplasms and cardiovascular dis-
Treatment is not necessary for most individuals with com- orders. Stomatocytes also may occur as an artifact.
mon HE. Splenectomy may be of benefit for patients with
symptomatic hemolytic anemia or its complications (see ear-
Acanthocytosis
lier discussion of splenectomy for hereditary spherocytosis).
Spur cells, or acanthocytes (from the Greek acantha, or
thorn; Figure 7-10), are erythrocytes with multiple irregular
Key points projections that vary in width, length, and surface distribu-
tion. Several conditions are associated with this morphology.
• HE is due to defects in the interactions of red cell cytoskeleton
In severe liver disease, acanthocyte formation is a two-step
proteins (“horizontal interactions”), with spectrin abnormalities
process involving the transfer of free nonesterified choles-
accounting for most of the cases.
• The majority of patients with HE are not symptomatic and
terol from abnormal plasma lipoproteins into the
require no therapy.
• HPP is a rare condition with apparent coinheritance of spectrin
defects leading to markedly abnormal red cells characterized by
increased thermal instability.
Other RBC membrane disorders
Stomatocytosis
Stomatocytes have a wide transverse slit or stoma toward the

center of the RBC (Figure 7-11). A few stomatocytes (between
3% and 5%) are found on blood smears of healthy individuals.
Several inherited and acquired disorders are associated with
stomatocytosis. The inherited forms are associated with
abnormalities in erythrocyte cation permeability and
volume, which is either increased (hence, the designation
hydrocytosis), decreased (xerocytosis), or near normal. Figure 7-11 Stomatocytes.

erythrocyte membrane and then the subsequent remodeling anemia. Three proteins (RhCE, RhD, and Rh50) comprise
of abnormally shaped erythrocytes by the spleen. Rapidly the Rh protein family. This disorder arises through autoso-
progressive hemolytic anemia is seen in association with mal recessive inheritance of either a suppressor gene unre-
advanced and often end-stage alcoholic cirrhosis, sometimes lated to the Rh locus or a silent allele at the locus itself. The
referred to as Zieve syndrome, or other conditions such as normal, complexed structure forms an integral membrane
metastatic liver disease, cardiac cirrhosis, Wilson disease, protein; its loss disrupts membrane architecture. Rhnull cells
and fulminant hepatitis. have increased rates of cation transport and sodium–
In abetalipoproteinemia, the primary molecular defect potassium membrane adenosine triphosphate (ATP)-ase
involves a congenital absence of apolipoprotein B in plasma. activity that results in dehydrated RBCs. This dehydration
Consequently, all plasma lipoproteins containing this apo- results in stomatocytes and occasional spherocytes on the
protein as well as plasma triglycerides are nearly absent. peripheral blood smear. Laboratory evaluation shows
Plasma cholesterol and phospholipid levels also are mark- increased RBC osmotic fragility, reflecting a marked reduc-
edly reduced. The role of these lipid abnormalities in pro- tion of the membrane surface area. The relationship between
ducing acanthocytes is not well understood. The most the absence of the Rh antigen proteins and RBC alterations
striking abnormality of the acanthocyte membrane in abet- leading to hemolysis presumably involves membrane microve-
alipoproteinemia is an increase in membrane sphingomy- siculation, leading to diminished erythrocyte flexibility. Sple-
elin. Abetalipoproteinemia is an autosomal recessive disorder nectomy results in improvement of the hemolytic anemia.
that manifests in the first month of life with steatorrhea.
Retinitis pigmentosa and progressive neurologic abnormali-
Abnormalities of RBC enzymes
ties, such as ataxia and intention tremors, develop between
5 and 10 years of age and progress to death by the second or Clinical case
third decade of life.
Acanthocytes have also been described in patients with the A 23-year-old African American male who recently underwent
McLeod phenotype, a condition in which the erythrocytes cadaveric renal transplant for end-stage renal disease is referred
have reduced surface Kell antigen. The affected red cells lack for evaluation of anemia. His past history is significant for an epi-
sode of hemolytic uremic syndrome (HUS) that led to renal fail-
the Kx protein which is a membrane precursor of the Kell
ure 2 years prior to referral. He had no further relapses of HUS or
antigen and needed for its expression. The Kx antigenic pro-
thrombotic thrombocytopenic purpura (TTP). His posttransplant
tein is encoded by the X chromosome, so males are affected
course has been unremarkable with good graft function and no
with mild compensated hemolysis with variable acanthocy- rejection. When he left the hospital, his hematocrit was 31%.
tosis (8%-85%). Due to lyonization, female carriers are His discharge medications included prednisone, cyclosporine,
asymptomatic with occasional acanthocytes and may be trimethoprim/sulfamethoxazole, and acyclovir. He complains of
identified by flow cytometric analysis of Kell blood group increasing fatigue and dyspnea over the 10 days since discharge.
antigen expression. In some ethnicities, the frequency of the Friends have noted yellowing of his eyes. He denies any fever or
Kx antigen is >99%, and thus, individuals with the McLeod infectious symptoms. On physical examination, he has a heart
phenotype can develop major problems with alloimmuniza- rate of 112, blood pressure (BP) of 89/45, and scleral icterus.
tion after immunizing events such as a transfusion. As such, Otherwise, the examination is unremarkable. Current hemato-
autologous donation should be considered where possible. crit is 20%, corrected reticulocyte count 10%, LDH 1,543 U/L.
Serum creatinine is 1.8 mg/dL and the platelet count is
The McLeod phenotype is a key feature of McLeod syn-
302,000 /mm3, similar to hospital discharge. On review of the
drome, a rare multisystem disease characterized by neuro-
peripheral blood smear, polychromatophilia is noted. A moder-
psychiatric, neuromuscular, cardiac, and hematological
ate number of bite and blister cells are identified.
abnormalities. The subtle hematological abnormalities may
precede the neurological complications for decades until
patients develop premature dementia, cognitive impairment, Normal metabolism of the mature RBC involves two princi-
social retraction, personality changes, a choreatic movement pal pathways of glucose catabolism: the glycolytic pathway
disorder, or dystonia. McLeod phenotype has also been asso- and the hexose-monophosphate shunt. The three major func-
ciated with X-linked granulomatous disease. tions of the products of glucose catabolism in the erythrocyte
are (i) maintenance of protein integrity, cellular deformability,
and RBC shape; (ii) preservation of hemoglobin iron in the
Rh deficiency (null) syndrome
ferrous form; and (iii) modulation of the oxygen affinity
This term is used to designate rare cases of either absent of hemoglobin. These functions are served by the regulation
(Rhnull) or markedly reduced (Rhmod) expression of the of appropriate production of five specific molecules:
Rh antigen in association with mild to moderate hemolytic ATP, reduced glutathione, reduced NADH, reduced NADPH,

and 2,3-BPG. Maintenance of the biochemical and structural

integrity of the RBC depends on the normal function of >20
enzymes involved in these pathways as well as the availability
of five essential RBC substrates: glucose, glutathione, NAD,
NAD phosphate (NADP), and adenosine diphosphate (ADP).
The primary function of the glycolytic pathway is the gen-
eration of ATP, which is necessary for the ATPase-linked
sodium–potassium and calcium membrane pumps essential
for cation homeostasis and the maintenance of erythrocyte
deformability. The production of 2,3-BPG is regulated by the
Rapoport–Luebering shunt, which is controlled by bisphos-
phoglyceromutase, the enzyme that converts 1,3-BPG to
2,3-BPG. Concentration of 2,3-BPG in the RBC in turn reg-
ulates hemoglobin oxygen affinity, thus facilitating the trans-
fer of oxygen from hemoglobin to tissue-binding sites. The Figure 7-12 Pyruvate kinase deficiency. The peripheral blood film
major function of the hexose-monophosphate shunt is pres- shows many small dense crenated cells (echinocytes).
ervation and regeneration of reduced glutathione, which
protects hemoglobin and other intracellular and membrane These disorders are rare; patients often present in childhood
proteins from oxidant injury. with mild to moderate anemia and reduced exercise toler-
ance. A form of acquired hexokinase deficiency occurs in
Wilson disease, in which elevated copper levels in the blood
Abnormalities of the glycolytic pathway
inhibit hexokinase in a fluctuating fashion that may lead to
Defects in the glycolytic pathway lead to a decrease in the intermittent brisk intravascular hemolysis. Phosphofructo-
production of ATP or a change in the concentration of kinase deficiency was first described as a muscle glycogen
2,3-BPG. Deficiencies of erythrocyte hexokinase, glucose storage disease; some patients with this deficiency have a
phosphate isomerase, phosphofructokinase, and pyruvate chronic hemolytic anemia. In phosphofructokinase defi-
kinase (PK) all lead to a decrease in ATP concentration. ciency, low levels of erythrocyte ATP lead to low-grade
Although genetic disorders involving nearly all of the hemolysis, but the limiting symptoms are usually weakness
enzymes of the glycolytic pathway have been described, PK and muscle pain on exertion. Children with phosphoglycer-
accounts for >80% of the clinically significant hemolytic ate kinase have associated neuromuscular manifestations,
anemias from defects in this pathway. With the exception of including seizures, spasticity, and mental retardation.
phosphoglycerate kinase deficiency, which is X-linked, all These enzymopathies are associated with anemia of vari-
other glycolytic enzyme defects are autosomal recessive. able severity. Peripheral blood smears from patients with PK
PK deficiency is the most common congenital nonsphero- deficiency show small dense crenated cells (echinocytes or
cytic hemolytic anemia caused by a defect in glycolytic RBC “prickle cells”) (Figure 7-12). In the most severe cases,
metabolism. The syndrome is both genetically and clinically marked reticulocytosis, nucleated RBCs, and substantial
heterogeneous. PK deficiency has a worldwide distribution anisopoikilocytosis can be seen. The MCV is usually normal
but is more common among those of northern and eastern or increased, reflecting the contribution of reticulocytes. A
European heritage. Severe cases can present either with neo- marked increase in the reticulocyte count (up to 70%) occurs
natal jaundice or in early childhood with jaundice, spleno- after splenectomy in PK deficiency.
megaly, and failure to thrive. Alternatively, a mild presentation Patients with severe hemolysis should receive folate supple-
with fully compensated hemolytic anemia has been described. mentation. Splenectomy generally is reserved for patients with
Osmotic fragility of the patient’s RBCs is typically normal poor quality of life, chronic transfusion requirements, need for
and may be helpful in differentiating this condition from HS. cholecystectomy, and persistent severe anemia. The response is
Several screening tests have been developed to diagnose PK variable, but most patients with PK deficiency benefit with an
deficiency, but often they lack sensitivity to diagnose specific increase in the hemoglobin level. Splenectomy may be compli-
PK variants. Reference laboratories can perform quantitative cated by postoperative thromboembolic phenomena.
measurement of the erythrocyte enzyme level necessary to
diagnose this condition accurately.
Abnormalities of the hexose-monophosphate shunt
Both glucose phosphate isomerase and hexokinase defi-
ciencies produce nonspherocytic hemolytic anemia associ- G6PD deficiency is the most frequently encountered abnormality
ated with decreased erythrocyte ATP and 2,3-BPG content. of RBC metabolism, affecting >200 million people worldwide.

A survival advantage has been noted in G6PD-deficient Table 7-6 Agents that cause clinically significant hemolysis in patients
patients infected with P. falciparum malaria, possibly with G6PD deficiency
accounting for its high gene frequency, especially in endemic Acetanilide Pentaquine
regions. The gene for G6PD is carried on the X chromosome Dapsone Phenylhydrazine
and exhibits extensive polymorphism. Enzyme deficiency is Dimercaptosuccinic acid Phenazopyridine
observed in males carrying a variant gene. Females with a Furazolidone Primaquine
variant gene have two clonal RBC populations, one normal Glibenclamide Sulfacetamide
and one deficient; the clinical presentation depends on the Isobutyl nitrite Sulfamethoxazole
extent of inactivation (“lyonization”) of the affected X chro- Methylene blue Sulfanilamide
mosome bearing the abnormal gene. Worldwide, >300 Nalidixic acid Sulfapyridine
genetic variants of G6PD have been described and are cate- Naphthalene Thiazolesulfone
gorized according to whether the defect leads to normal Niridazole Toluidine blue
activity, moderately deficient activity, or severely deficient Nitrofurantoin Trinitrotoluene (TNT)
activity, and whether it is associated with hemolytic anemia. Pamaquine Urate oxidase
G6PD enzyme variants are distinguished based on electro- Data from Beutler E. In: Starbury JB et al. eds. The Metabolic Basis
phoretic mobility. G6PD B, the wild-type enzyme, and of Inherited Disease. 5th ed. New York: McGraw-Hill, 1983. Updated
G6PD A+, a common variant in the African American popu- from Beutler E. In: Beutler E et al. eds. Williams Hematology. 6th ed.
lation, demonstrate normal enzyme activity and are not New York: McGraw-Hill, 2001.
associated with hemolysis. G6PD A– is present in approxi-
mately 10%-15% of African American males. This variant is
an unstable enzyme, which results in a decrease in enzyme manifest anemia until they are exposed to an oxidant drug or
activity in aged RBCs. In contrast, other G6PD variants have other oxidant challenge. Such an exposure may provoke an
reduced catalytic activity and marked instability or are pro- acute hemolytic episode with intravascular hemolysis. In the
duced at a decreased rate, rendering both reticulocytes and G6PD A– variant, an adequate reticulocyte response can
older cells susceptible to hemolysis. Enzymatic deficiency of result in restoration of the hemoglobin concentration even if
this type is seen in up to 5% of persons of Mediterranean the offending drug is continued because the newly formed
or Asian ancestry, as well as Ashkenazi Jews. The com- reticulocytes are relatively resistant to oxidant stress given
mon example of this deficiency is the G6PD B variant, their higher G6PD levels. Women heterozygous for G6PD A–
G6PD-Mediterranean. usually experience only mild anemia upon exposure to oxi-
Hemolysis in G6PD-deficient RBCs is due to a failure to dant stress because a population of G6PD sufficient (normal)
generate adequate NADPH, leading to insufficient levels of cells coexists. The G6PD-Mediterranean variant is more
reduced glutathione. This renders erythrocytes susceptible to severe than the African G6PD A– variant and is thus prone
oxidation of hemoglobin by oxidant radicals, such as hydro- to more severe hemolytic episodes. Men and heterozygous
gen peroxide. The resulting denatured hemoglobin aggre- women with the G6PD-Mediterranean variant can experi-
gates and forms intraerythrocytic Heinz bodies, which bind ence severe hemolysis in the face of oxidant stress, and the
to membrane cytoskeletal proteins. Membrane proteins are offending agent must be removed because the reticulocytes
also subject to oxidation, leading to decreased cellular have low enzyme levels and are prone to hemolysis.
deformability. Cells containing Heinz bodies are entrapped Certain G6PD variants may result in a congenital nons-
or partially destroyed in the spleen, resulting in loss of cell pherocytic hemolytic anemia with persistent splenomegaly.
membranes through pitting of Heinz bodies and leading to Affected individuals are extremely susceptible to the oxidant
hemolysis. stress associated with the drugs and disorders mentioned
The severity of hemolytic anemia in patients with G6PD previously and also may exhibit severe hemolysis (“favism”)
deficiency depends on the type of defect, the level of enzyme after ingestion of fava beans. Hemolytic anemia due to
activity in the erythrocytes, and the severity of the oxidant favism may be severe or even fatal, particularly in children.
challenge. Ingestion of an oxidant drug is sometimes the pre- G6PD deficiency predisposes to neonatal jaundice, and it
cipitating cause (Table 7-6). Hemolytic anemia in patients may be the result of impairment of hepatic function, hemo-
with G6PD deficiency may first be recognized during an lysis, or both.
acute clinical event that induces oxidant stress, such as infec- When hemolytic anemia occurs after the ingestion of an
tion, diabetic ketoacidosis, or severe liver injury. In children, oxidant drug or in association with the clinical states leading
infection is a common precipitating event. Hemolysis trig- to oxidant stress, G6PD deficiency should be considered.
gered by exposure to naphthalene (moth balls) is now much Significant anemia, hyperbilirubinemia, elevated plasma Hb,
less common in children. Individuals with G6PD A– do not and hemoglobinuria may be due to brisk intravascular

hemolysis. G6PD deficiency should be considered in an indi- the activity of a normal erythrocyte enzyme. The excessive
vidual with evidence of chronic DAT-negative hemolysis. deaminase activity prevents normal salvage of adenosine and
The peripheral blood smear reveals RBCs with the Hb con- causes subsequent depletion of ATP and hemolysis.
fined to one side of the cells, with the remainder appearing as
an Hb-free ghost (eccentrocytes) (Figure 7-13). The mor-
phology previously has been described as bite or blister cells, Clinical case (continued)
interpreted as the result of removal of denatured Hb by the
The patient presented in this section should be suspected of
spleen; however, it appears that the accumulated oxidized
having G6PD deficiency. Patients with the African American
Hb actually remains and is adherent to the RBC membrane.
variant (G6PD A–) are often asymptomatic until they ingest
Brilliant cresyl blue staining may reveal Heinz bodies. medication or experience an infection, which leads to oxidant
Screening or quantitative biochemical assays can be used to stress of the RBCs. Trimethoprim/sulfamethoxazole may be
make the diagnosis. In the G6PD A– variant, during an acute an offending agent. During the early phases of hemolysis,
hemolytic episode, an elevated reticulocyte count will raise eccentrocytes can be seen on review of the peripheral blood
the mean level of erythrocyte G6PD and render a false-nega- smear. A Heinz body preparation will show the typical inclu-
tive result. G6PD levels, therefore, should be checked several sions, which consist of denatured hemoglobin. G6PD levels may
months after the acute event when there will be RBCs of be misleading in the acute setting, as values may be normal due
varying ages. Although defects of other hexose-monophos- to reticulocytosis. Treatment is primarily supportive. Offending
phate shunt enzymes (eg, phosphogluconate dehydrogenase, drugs should be discontinued and alternative agents chosen.
If the prescribed agent is necessary and cannot be substituted,
glutathione reductase) are rare, they should be considered in
however, a trial of continuation is reasonable, as hemolysis often
cases of oxidant-induced hemolysis in which G6PD levels
is compensated in the G6PD A– variant even if drug administra-
are normal.
tion is continued.
Abnormalities of nucleotide metabolism
Pyrimidine-5′-nucleotidase deficiency is an enzymatic Key points

abnormality of pyrimidine metabolism associated with
• The glycolytic pathway provides the erythrocyte with the ATP
hemolytic anemia. The peripheral blood smear in patients
necessary for maintenance of membrane integrity, preservation
with this defect often shows RBCs containing coarse baso- of ferrous hemoglobin, and oxygen affinity. Additional products
philic stippling. Lead intoxication also inactivates the of the pathway are NADH for methemoglobin reduction and
enzyme, leading to an acquired variant of pyrimidine- 2,3-BPG for regulating the oxygen affinity of hemoglobin.
5′-nucleotidase deficiency. • Glucose metabolism through the hexose monophosphate
Adenosine deaminase (ADA) excess is an unusual abnor- shunt produces NADPH to maintain the antioxidative activity of
mality. It is caused by a genetically determined increase in the red cell.
• Enzymopathies represent a major consideration in the differen-
tial diagnosis of inherited DAT-negative nonspherocytic
hemolytic anemias.
• PK deficiency is the most common defect of the glycolytic
pathway. The echinocyte is the characteristic abnormality
observed on the peripheral blood smear.
• The most common enzyme deficiency is G6PD with >300
genetic variants. Oxidant stress in the presence of deficient
G6PD activity results in hemolysis with the generation of blister
cells and bite cells (eccentrocytes).
• In the G6PD A– deficiency, a quantitative measurement of the
enzyme levels should be delayed until after the acute hemolytic
episode. In the G6PD B variant (eg, G6PD-Mediterranean), levels
are low in red cells of all ages.
• Defects of purine and pyrimidine metabolism are infrequent.
The peripheral blood smear in pyrimidine-5’-nucleotidase
deficiency shows red cells with coarse basophilic stippling.
Figure 7-13 G6PD deficiency. The peripheral blood smear shows
• Iron overload can occur in non-transfused inherited chronic
several red cells with the hemoglobin confined to one side of the cells
hemolytic anemias due to increased gastrointestinal iron
with the remainder appearing as a hemoglobin-free ghost
absorption.
(eccentrocytes).

Hemolysis due to extrinsic abnormalities of the RBC | 167
Hemolysis due to extrinsic Hemolytic anemia due to immune injury

abnormalities of the RBC to RBCs
In autoimmune hemolytic anemia (AHA), shortened RBC

Clinical case survival is mediated by autoantibodies. AHA is classified by
the temperature at which autoantibodies bind optimally to
A 68-year-old male is admitted to the hospital with complaints
of weakness, shortness of breath, and chest pain. Over the prior
the patient RBCs. In adults, the majority of cases (80%-90%)
year, he has experienced weight loss and intermittent night are mediated by antibodies that bind to RBCs at 37°C (warm
sweats, and has generally felt poorly. His prior history is signifi- autoantibodies). In the cryopathic hemolytic anemias, the
cant for diet-controlled diabetes and elevated cholesterol. He is autoantibodies bind most avidly to RBCs at temperatures
taking no medications. On examination, he appears chronically <37°C (cold autoantibodies). Some patients exhibit both
ill and pale. Scleral icterus is noted. Axillary adenopathy and warm and cold reactive autoantibodies. These cases are clas-
splenomegaly are appreciated. His fingertips are mildly cyanotic sified as mixed AHA.
appearing. Laboratory data are significant for a spun hematocrit The warm- and cold-antibody classifications are further
of 24% and an MCV of 143 fL. LDH is elevated at 2,321 U/L, divided by the presence or absence of an underlying related
indirect bilirubin at 2.1 mg/dL, and reticulocyte count at 13%.
disease. When no underlying disease is recognized, the AHA
The peripheral blood smear shows agglutinated RBCs. The blood
is termed primary or idiopathic. Secondary cases are those in
bank reports a direct Coombs test positive for complement
which the AHA is a manifestation or complication of an
(3+) but negative for immunoglobulin G (IgG). Serum protein
electrophoresis reveals a monoclonal IgM. Abdominal CT scan
underlying disorder. In general, the secondary classification
reveals splenomegaly and diffuse adenopathy. should be used in preference to idiopathic only when the
AHA and the underlying disease occur together more often
Table 7-7 Classification of immune injury to red blood cells

I. Warm-autoantibody type: autoantibody maximally active at 37°C
A. Primary or idiopathic warm AHA
B. Secondary warm AHA
1. Associated with lymphoproliferative disorders (eg, Hodgkin disease, lymphoma)
2. Associated with the rheumatic disorders (eg, SLE)
3. Associated with certain nonlymphoid neoplasms (eg, ovarian tumors)
4. Associated with certain chronic inflammatory diseases (eg, ulcerative colitis)
5. Associated with ingestion of certain drugs (eg, a-methyldopa)
II. Cold-autoantibody type: autoantibody optimally active at temperatures <37°C
A. Mediated by cold agglutinins
1. Idiopathic (primary) chronic cold agglutinin disease (usually associated with clonal B-lymphocyte proliferation)
2. Secondary cold agglutinin hemolytic anemia
a. Postinfectious (eg, Mycoplasma pneumoniae or infectious mononucleosis)
b. Associated with malignant B-cell lymphoproliferative disorder
B. Mediated by cold hemolysins
1. Idiopathic (primary) paroxysmal cold hemoglobinuria
2. Secondary
a. Donath-Landsteiner hemolytic anemia, usually associated with an acute viral syndrome in children (relatively common)
b. Associated with congenital or tertiary syphilis in adults
III. Mixed cold and warm autoantibodies
A. Primary or idiopathic mixed AHA
B. Secondary mixed AHA
1. Associated with the rheumatic disorders, particularly SLE
IV. Drug-immune hemolytic anemia
A. Hapten or drug adsorption mechanism
B. Ternary (immune) complex mechanism
C. True autoantibody mechanism
D. Nonimmunologic protein adsorption (probably does not cause hemolysis)
Adapted from Packman CH. In: Lichtman, MA. Williams Hematology. 8th ed. New York, NY: McGraw-Hill; 2010.
AHA = autoimmune hemolytic anemia; SLE = systemic lupus erythematosus.

than random and when the AHA resolves with successful with Mycoplasma infection, as well as with Epstein-Barr
treatment of the underlying disease. The connection is virus-related disease. In addition, cold agglutinin disease is
strengthened when the underlying disease has a component typically seen in the elderly, almost always associated with
of immunologic aberration. Using these criteria, primary B-cell lymphoproliferative disorders; it is caused by a mono-
(idiopathic) AHA and secondary AHA occur with approxi- clonal IgM antibody that binds to carbohydrate I antigens or
mately equal frequency. i antigens at temperatures below body temperature. Cold-
Certain drugs also may cause immune destruction of reacting IgG (Donath-Landsteiner) autoantibodies, seen in
RBCs by three different mechanisms. Some drugs induce for- paroxysmal cold hemoglobinuria, may cause significant
mation of true autoantibodies directed against RBC antigens. intravascular lysis of RBCs as a result of their ability to fix
The hapten-drug adsorption mechanism is characterized by complement. Donath-Landsteiner hemolytic anemia fre-
the presence of antidrug antibodies in the blood. These anti- quently was associated with congenital syphilis when that
bodies bind only to RBC membranes that are coated with disease was common. Now, it is almost always idiopathic.
tightly bound drug. In a third type of drug-immune hemo- Donath-Landsteiner hemolytic anemia accounts for almost
lytic anemia, antibodies recognize a neoantigen formed by a one-third of AHA cases in children. The responsible autoan-
drug or its metabolite and an epitope of a specific membrane tibodies bind to antigens in the P blood group system.
antigen. This is termed ternary or immune complex mecha-
nism. In some, if not all, cases mediated by the ternary Mixed AHA
(immune) complex mechanism, antibodies may recognize
both a drug or its metabolite and an epitope of a specific RBC Some cases of AHA are associated with the presence of both
antigen. The classification of the immune hemolytic anemias IgM and IgG autoantibodies. Hemolysis is generally more
is shown in Table 7-7. severe in these cases. AHA due to IgA antibodies is rare. IgA
autoantibodies usually are accompanied by IgG autoanti-
Pathophysiology bodies. The mechanisms for RBC destruction appear to be
similar to those for IgG.
Warm AHA
The most common type of AHA is mediated by warm-

Drug-induced immune hemolytic anemia
reactive autoantibodies of the IgG isotype. Warm-reacting
IgG antibodies bind optimally to antigens on RBCs at 37°C The clinical and laboratory features of drug-induced and
and may or may not fix complement, but they typically do idiopathic hemolytic anemia are similar, so a careful his-
not cause direct agglutination of RBCs because of their small tory of drug exposure should be obtained in the initial
size. Enhanced destruction of antibody-coated RBCs is evaluation. The number of drugs that can cause immune
mediated by Fc receptor–expressing macrophages, primarily hemolytic anemia is large and encompasses a broad spec-
located in the spleen. Partial phagocytosis results in the for- trum of chemical classes (Table 7-8). Three basic mecha-
mation of spherocytes that may circulate for a time but even- nisms of drug-induced immune RBC injury are recognized.
tually become entrapped in the spleen, resulting in enhanced A fourth mechanism may lead to nonimmunologic deposi-
RBC destruction. tion of multiple serum proteins, including immunoglobu-
lins, albumin, fibrinogen, and others, on RBCs, but RBC
Cold AHA injury does not occur. The mechanisms of drug-induced
immune-hemolytic anemia and positive DATs are summa-
In contrast to warm-reactive autoantibodies, cold-reactive rized in Table 7-9. Second- and third-generation cephalo-
autoantibodies bind optimally to RBCs at temperatures sporins account for about 88% of drug-induced immune
<37°C. Cold autoantibodies are typically of the IgM isotype, hemolytic anemia.
and because of their large, pentameric conformation, they
are able to span the distance between several RBCs to cause
Hapten or drug adsorption mechanism
direct agglutination. Their ability to injure RBCs depends on
their ability to fix complement. The consequence of comple- Hapten or drug adsorption mechanism applies to drugs that
ment fixation is clearance of C3b-coated cells by attachment bind firmly to proteins on the RBC membrane. The classic
to complement receptors on macrophages, primarily in the setting is very high-dose penicillin therapy, but other drugs
spleen, and Kupffer cells in the liver. Direct lysis by comple- such as cephalosporins and semisynthetic penicillins also are
tion of the terminal complement sequence may also occur. implicated. The antibody responsible for hemolytic anemia
Cold autoantibodies are characteristic of AHA associated by this mechanism is of the IgG class and is directed against

Table 7-8 Drugs associated with immune injury to RBCs or a positive Hemolytic anemia typically occurs 7-10 days after the drug is
direct antiglobulin test started and ceases a few days to 2 weeks after the patient dis-
Hapten or drug adsorption mechanism continues taking the drug.
Carbromal Oxaliplatin
Cephalosporins Penicillins
Ternary or immune complex mechanism: drug
Cianidanol Tetracycline
antibody-target cell interaction
Hydrocortisone Tolbutamide
6-Mercaptopurine Drugs in this group exhibit only weak direct binding to
Ternary-immune complex mechanism blood cell membranes. A relatively small dose of drug is
Amphotericin B Nomifensine capable of triggering destruction of blood cells. Blood cell
Antazoline Oxaliplatin injury is mediated by a cooperative interaction among three
Cephalosporins Pemetrexed
reactants to generate a ternary complex consisting of
Chlorpropamide Probenecid
the drug or a drug metabolite, a drug-binding membrane
Diclofenac Quinine
site (an antigen) on the target cell, and a drug-dependent
Diethylstilbestrol Quinidine
antibody. The drug-dependent antibody is thought to bind,
Doxepin Rifampicin
through its Fab domain, to a compound neoantigen consist-
Etodolac Stibophen
Hydrocortisone Thiopental
ing of loosely bound drug and a blood group antigen i ntrinsic
Metformin Tolmetin to the RBC membrane. The pathogenic antibody recognizes
Autoantibody mechanism the drug only in combination with a particular membrane
Cephalosporins Lenalidomide structure of the RBC (eg, a known alloantigen). Binding of
Cianidanol Mefenamic acid the drug to the target cell membrane is weak until the attach-
Cladribine α-Methyldopa ment of the antibody to both drug and cell membrane is sta-
Diclofenac Nomifensine bilized. Yet the binding of the antibody is drug dependent.
l-DOPA (levodopa) Oxaliplatin RBC destruction occurs intravascularly after completion of
Efalizumab Pentostatin the whole complement sequence, often resulting in hemo-
Fludarabine Procainamide globinemia and hemoglobinuria. The DAT is positive usu-
Glafenine Teniposide ally only for complement.
Latamoxef Tolmetin
Nonimmunologic protein adsorption
Carboplatin Cisplatin
Autoantibody mechanism
Cephalosporins Oxaliplatin Several drugs, by unknown mechanisms, induce the formation
Uncertain mechanism of immune injury of autoantibodies reactive with RBCs in the absence of the
Acetaminophen Melphalan
instigating drug. The most studied drug in this category has
p-Aminosalicylic acid Mephenytoin
been α-methyldopa, but levodopa and other drugs also have
Carboplatin Nalidixic acid
been implicated. Patients with chronic lymphocytic leukemia
Chlorpromazine Omeprazole
treated with pentostatin, fludarabine, or cladribine may have
Efavirenz Phenacetin
severe and sometimes fatal autoimmune hemolysis, although
Erythromycin Streptomycin
Fluorouracil Sulindac
the mechanisms of autoantibody induction are likely different,
Ibuprofen Temafloxacin most likely involving dysregulation of T lymphocytes.
Insecticides Thiazides
Isoniazid Triamterene
Nonimmunologic protein adsorption
epitopes of the drug. Other manifestations of drug sensitiv- A small proportion (<5%) of patients receiving cephalospo-
ity, such as hives or anaphylaxis, usually are not present. The rin antibiotics, cisplatin and carboplatin, develop positive
antibody binds to drug molecules attached to the RBC mem- antiglobulin reactions because of nonspecific adsorption of
brane. Antibodies eluted from patients’ RBCs or present in plasma proteins to their RBC membranes. This process may
their sera react in the indirect antiglobulin test (IAT) only occur within 1-2 days after the drug is instituted. Multiple
against drug-coated RBCs, which distinguishes these drug- plasma proteins, including immunoglobulins, complement,
dependent antibodies from true autoantibodies. Destruction albumin, fibrinogen, and others, may be detected on RBC
of RBCs coated with drug and IgG antidrug antibody occurs membranes in such cases. Hemolytic anemia resulting from
mainly through sequestration by splenic macrophages. this mechanism does not occur. This phenomenon, however,

Table 7-9 Immune hemolytic anemia and positive direct antiglobulin reactions caused by drugs
Hapten-drug Ternary-immune Autoantibody Nonimmunologic

adsorption complex formation formation protein adsorption
Prototype drug Penicillin Third-generation a-Methyldopa Cephalothin
cephalosporins
Role of drug Binds to red cell Forms three-way complex Induces antibody to Possibly alters red
membrane with antibody and red cell native red cell antigen cell membrane
membrane component
Drug affinity to cell Strong Weak None demonstrated Strong
Antibody to drug Present Present Absent Absent
Antibody class predominating IgG IgM or IgG IgG None
Proteins detected by direct IgG, rarely Complement IgG, rarely complement Multiple plasma
antiglobulin test complement proteins
Dose of drug associated with High Low High High
positive antiglobulin test
Mechanism of red cell Splenic Direct lysis by complement Splenic sequestration None
destruction sequestration plus splenic sequestration
Modified from Packman CH. In: Lichtman, MA. Williams Hematology. 8th ed. New York, NY: McGraw-Hill; 2010.
may complicate cross-match procedures unless the drug his-

tory is considered.
Clinical manifestations and laboratory findings
Several clinical features of AHA are common to both warm-

and cold-antibody types. Patients may present with signs
and symptoms of anemia (eg, weakness, dizziness), jaundice,
abdominal pain, and fever. Mild splenomegaly is common.
Hepatomegaly and lymphadenopathy may be evident at pre-
sentation depending on the etiology. Anemia may vary from
mild to severe, usually with either normocytic or macrocytic
cells. Patients most frequently present with reticulocytosis.
Reticulocytopenia, however, initially may be present up to
one-third of the time as a result of intercurrent folate defi- Figure 7-14 Warm-antibody autoimmune hemolytic anemia. Note
ciency, infection, involvement of the marrow by a neoplastic the small round spherocytes and the large, gray polychromatophilic
erythrocytes.
process, or unidentifiable causes. Indirect bilirubin and LDH
are elevated to varying degrees, and the haptoglobin is
depressed. The blood smear often demonstrates spherocytes
(Figure 7-14). Nucleated RBCs also may be present. MCHC measurements and decrease in the RBC count are
The onset of warm-antibody AHA may be rapid or insidi- observed due to simultaneous passage of two or three agglu-
ous, but rarely is it so severe as to cause hemoglobinuria. Pre- tinated RBCs through the aperture of the automated cell
senting symptoms usually are related to anemia or jaundice. counter.
In secondary cases, the presenting complaint usually is Drug-immune hemolytic anemia due to the hapten or
related to the underlying disease. true autoantibody mechanism is usually mild. In contrast,
Patients with idiopathic or primary cold agglutinin dis- hemolysis due to the ternary or immune complex mecha-
ease usually have mild to moderate chronic hemolysis. Acute nism can be acute in onset, severe, and sometimes fatal.
exacerbations can be associated with cold exposure. Sponta- The DAT (Coombs test) is usually positive in AHA but
neous autoagglutination of RBCs at room temperature may may be negative in some patients. The threshold of detection
be seen as clumps of cells on the blood smear (Figure 7-15). of commercial antiglobulin reagents, which detect mainly
Occasionally spurious marked elevations in the MCV and IgG and fragments of C3, is approximately 200-500 antibody

the diagnosis sometimes can be confirmed by using more

sensitive assays for RBC-bound immunoglobulin, such as an
enzyme-linked immunoadsorbent assay (ELISA) or radiola-
beled anti-immunoglobulin. Specific assays for cell-bound
IgA also may be worthwhile. In cold agglutinin disease, the
DAT is positive with anti-C3 only.
Approximately 1 in 10,000 healthy blood donors have a
positive DAT. The positive DAT in these individuals usually is
due to warm-reacting IgG autoantibodies, indistinguishable
from those occurring in AHA. Many of these individuals
never develop AHA, but some do. It is not known how many
of these normal individuals with a positive DAT eventually
may develop AHA.
Figure 7-15 Cold agglutinin disease.

Treatment
Asymptomatic patients develop anemia over a period suffi-

molecules per cell. However, <100 molecules of IgG per cell cient to allow for cardiovascular compensation and do not
may significantly shorten RBC survival in vivo. IgM cold require RBC transfusions. For patients with symptomatic
agglutinins are usually removed from RBCs during washing coronary artery disease or patients who rapidly develop
and usually are not detected. Most commercial reagents do severe anemia with circulatory failure, as in paroxysmal cold
not detect IgA. When monospecific anti-IgG and anti-C3 hemoglobinuria or ternary (immune) complex drug-
reagents are used, 30%-40% of patients with AHA will have immune hemolysis, transfusions can be lifesaving.
only IgG on their RBCs; a slightly larger number will have Transfusion of RBCs in immune hemolytic anemia is
both IgG and C3; and only approximately 10% will have C3 often problematic. Finding serocompatible donor blood is
alone. The major reaction patterns of the DAT and their dif- rarely possible because, in most cases, the autoantibody is a
ferential diagnosis are summarized in Table 7-10. panagglutinin. It is most important to identify the patient’s
The strength of the direct Coombs test has poor clinical ABO type to find either ABO-identical or ABO-compatible
correlation with severity of hemolysis among patients, but in blood for transfusion to avoid a hemolytic transfusion reac-
a given patient over time, the degree of hemolysis correlates tion. The difficult technical issue relates to detection of RBC
fairly well with the current strength of the antiglobulin alloantibodies masked by the presence of the autoantibody.
reaction. In the rare case of direct Coombs test–negative Clinicians and blood bank physicians speak of identify-
hemolytic anemia suspected of having an immune etiology, ing “least incompatible” blood for transfusion, but this is a
misnomer because all units will be serologically incompat-
ible. Units incompatible because of autoantibody are less
Table 7-10 Differential diagnosis of reaction patterns of the direct dangerous to transfuse, however, than units incompatible
antiglobulin test because of alloantibody. Patients with a history of preg-
Reaction pattern Differential diagnosis nancy, abortion, or prior transfusion are at risk of harbor-
ing an alloantibody. Patients who have never been pregnant
IgG alone Warm antibody autoimmune hemolytic
anemia
or transfused with blood products are unlikely to harbor an
Drug-immune hemolytic anemia: hapten/drug alloantibody. Consultation between the clinician and the
adsorption type or autoantibody type blood bank physician should occur early to allow for
Complement Warm antibody autoimmune hemolytic anemia informed discussion and confident transfusion of mis-
alone with subthreshold IgG deposition matched blood if the situation demands. Clinicians must
Cold-agglutinin disease understand that the dropping hemoglobin often in the set-
Paroxysmal cold hemoglobinuria ting of reticulocytopenia is a life-threatening situation, and
Drug-immune hemolytic anemia: ternary- delay in transfusion over concerns about red cell incompat-
immune complex type ibility can lead to a patient’s demise.
IgG plus Warm antibody autoimmune hemolytic The selected RBCs should be transfused slowly while the
complement anemia patient is monitored carefully for signs of a hemolytic trans-
Drug-immune hemolytic anemia: autoantibody
fusion reaction. Even if transfused cells are rapidly destroyed,
type (rare)
the increased oxygen-carrying capacity provided by the

transfused cells may maintain the patient during the time cyclophosphamide have been beneficial in selected cases.
required for other modes of therapy to become effective. Chemotherapy is indicated if the disorder is associated with
In AHA, therapy is aimed at decreasing the production of a lymphoproliferative disorder. Splenectomy usually is not
autoantibody and at decreasing clearance of RBCs from the indicated because cells typically are cleared by intravascular
circulation. For warm-antibody IgG-mediated hemolysis, hemolysis or hepatic Kupffer cells. Intravenous immuno-
glucocorticoids such as prednisone usually are the first-line globulin does not have a role in management. Plasmaphere-
treatment in all but drug-induced syndromes (for which sis may be temporarily effective in acute situations by
removal of the offending agent is the principal treatment). removing IgM cold agglutinin from the circulation. Recently,
Glucocorticoids decrease the ability of macrophages to clear a combination of fludarabine and rituximab has been used
IgG- or complement-coated erythrocytes and reduce auto- with success, but toxicity is a concern.
antibody production. After remission is achieved with pred- AHA during childhood tends to occur suddenly, during,
nisone at approximately 60-100 mg/d (or 1 mg/kg/d), the or after an acute infection. As many as one-third of cases are
dose should be decreased by 10 mg/d each week until a dose associated with intravascular hemolysis because of a Donath-
of 30 mg/d is reached. Subsequent dose reduction should Landsteiner antibody directed against the erythrocyte P anti-
then proceed more slowly (at 5 mg/d per week), with the gen. Usually these patients exhibit only a single paroxysm of
goal of either maintaining remission with prednisone at hemolysis. In warm-antibody hemolytic anemia, acute man-
20-40 mg every other day or complete weaning of predni- agement is similar to that for adults. Approximately two-
sone if the DAT becomes negative; this goal is not always thirds of children recover completely within a matter of
achievable. Approximately two-thirds of adult patients weeks. Only a small percentage of children (but a larger pro-
respond to prednisone, with approximately 20% achieving portion of adolescents) exhibit more chronic refractory dis-
complete remission. Pulses of high-dose glucocorticoids ease that warrants consideration of other pharmacologic
(eg, 1 g methylprednisolone intravenously) are effective in agents or splenectomy.
some patients in whom standard therapy has failed.
Splenectomy is often considered if hemolysis remains
severe for 2-3 weeks at prednisone doses of 1 mg/kg, if remis- Clinical case (continued)
sion cannot be maintained on low doses of prednisone, or if The patient presented in this section has cold agglutinin disease,
the patient has intolerable adverse effects or contraindica- likely secondary to underlying lymphoma. Automated tech-
tions to glucocorticoids. Removing the spleen results in a niques reveal the red cell count is artifactually low, and the MCV
reduced rate of clearance of IgG-coated cells. Although not and MCHC are falsely elevated secondary to red cell agglutina-
usually recommended in children, splenectomy in patients tion. Warming of the blood tube with immediate measurement
past adolescence appears relatively safe. Patients should and slide preparation will minimize agglutination. The direct
receive pneumococcal, H. influenzae, and meningococcal antiglobulin test is positive only for complement. Lymphopro-
vaccines before splenectomy. Approximately two-thirds of liferative disorders are well-identified underlying etiologies. The
patient should be maintained in a warm environment. Ameliora-
patients will have complete or partial remission with sple-
tion of the anemia can be anticipated with cytotoxic therapy for
nectomy, but relapses are common.
the lymphoma.
Other therapies may be effective for patients with refrac-
tory hemolysis or for those who relapse after glucocorticoids
or splenectomy. Standard-dose (375 mg/m2) and low-dose
(100 mg/m2) monoclonal anti-CD20 (rituximab) has been Key points
useful in refractory cases. Adults and children respond equally
• Warm-antibody-induced immune hemolytic anemia is
well with response rates ranging from 40% to 100%. Immu-
typically IgG mediated and results in spherocytic red cells.
nosuppressive drugs, such as cyclophosphamide, azathio-
• Cold agglutinin disease is IgM mediated with associated
prine, mycophenolate mofetil, and cyclosporine, as well as the complement activation. The peripheral blood smear reveals red
nonvirilizing androgen, danazol, have been used with varying cell agglutination and spherocytes.
degrees of success. Intravenous immunoglobulin has been • A variety of drugs cause immune hemolytic anemia. Clinical
less successful in treatment of AHA than in immune TTP. laboratory support of the diagnosis may not be available.
For patients with idiopathic cold agglutinin disease, main- Discontinuation of the suspected offending drug is indicated.
taining a warm environment may be all that is needed to • Symptoms resulting from autoimmune hemolytic anemia are
avoid symptomatic anemia. Cold agglutinin disease usually typically indistinguishable from other causes of hemolysis.
does not respond to glucocorticoids. Recently, rituximab has • The direct antiglobulin test is the primary tool for diagnosing
demonstrated efficacy in treating cold agglutinin disease, autoimmune hemolytic anemia. It is rarely positive in healthy
individuals and may be negative in autoimmune hemolytic anemia.
with response rates approaching 50%. Chlorambucil and

lyse more readily in the presence of activated complement.

Key points (continued) Earlier tests to diagnose PNH (eg, Ham test or acid hemoly-
• Warm-antibody-mediated autoimmune hemolytic anemia is sis test; sucrose hemolysis test) were based on this property
treated with glucocorticoids, other immunosuppressive agents of PNH RBCs. It is now known that PNH granulocytes and
such as rituximab, and splenectomy. platelets are sensitive to complement as well.
• Avoidance of cold environments may be sufficient to avoid The biochemical basis of complement sensitivity was ini-
complications of cold agglutinin disease. Chemotherapy and tially elusive. Early on, PNH blood cells were found to be
rituximab have a role, and plasmapheresis occasionally can be deficient in leukocyte alkaline phosphatase and erythrocyte
helpful in the acute and temporary management of symptom-
acetylcholinesterase. Neither of these deficiencies, however,
atic cases by physically removing the antibody.
explained the complement sensitivity or the clinical manifes-
• Immune-mediated hemolytic anemia is uncommon in
tations in PNH. Subsequently, two complement regulatory
children. Most cases are acute and transient, following viral
infection.
proteins, CD55 (decay accelerating factor [DAF]) and CD59
• Transfusion therapy can be difficult in patients with autoim- (homologous restriction factor or membrane inhibitor of
mune hemolytic anemia. Consultation with the blood bank is reactive lysis [MIRL]), were found to be missing from PNH
important. A history of prior pregnancy, abortion, or transfusion blood cells, helping to explain the unusual sensitivity of
of blood products should be obtained, as these patients are at RBCs to the hemolytic action of complement. Of these,
risk to harbor dangerous alloantibodies. No patient with AHA CD59, whose action is to inhibit the terminal complement
should be allowed to die because serologically “compatible” sequence leading to hemolysis, seems to be the most
RBCs are not available. important.
The approximately 20 proteins missing from the hemato-
poietic cells in PNH are all attached to the membrane by a
Paroxysmal nocturnal hemoglobinuria glycosylphosphatidylinositol (GPI) anchor. Defective syn-
thesis of the GPI anchor is due to somatic mutations in the
Clinical case
pig-A gene in hematopoietic stem cells. Whereas a pig-A gene
A previously healthy 37-year-old female is admitted to the hospi- mutation appears to be necessary for the development of
tal for evaluation of severe abdominal pain. Workup reveals mes- PNH and its clinical manifestations, it is not sufficient
enteric vein thrombosis. The patient is treated with thrombolytic because pig-A mutations can be found in small numbers of
therapy and anticoagulated with heparin, leading to clinical hematopoietic stem cells in normal individuals. Patients
improvement. She has no prior or family history of thrombosis. with aplastic anemia exhibit a larger proportion of stem cells
She currently is taking an oral contraceptive. Her examination
with pig-A mutations.
is significant for mild scleral icterus and jaundice. There is no
A multistep process seems necessary for PNH to develop.
abdominal tenderness. Mild splenomegaly is noted. Laboratory
It is thought that in aplastic anemia and likely in PNH that
studies are significant for a hematocrit of 32% with a corrected
reticulocyte count of 8%. White count and platelet count are
immunologic processes suppress proliferation of normal
slightly depressed. Indirect bilirubin is elevated at hematopoietic precursors more efficiently than proliferation
4 mg/dL, but AST, ALT, and alkaline phosphatase are normal. of precursors lacking GPI-anchored proteins. Resistance to
LDH is also increased at 1,024 U/L. Blood bank evaluation apoptotic death may partly explain the survival advantage of
confirms a Coombs-negative hemolytic anemia. A bone marrow these GPI-negative cells. The abnormal clones thus are able
aspirate and biopsy are hypocellular and reveal findings concern- to expand until the numbers of abnormal progeny are suffi-
ing for early myelodysplasia. cient to cause the clinical manifestations of PNH.
Two missing GPI-linked proteins may contribute to the
increased incidence of thrombosis in PNH: (i) urokinase
Paroxysmal nocturnal hemoglobinuria (PNH) should be plasminogen activator receptor, the lack of which may
considered in the patient with unexplained hemolysis, decrease local fibrinolysis; and (ii) tissue factor pathway
pancytopenia, or unprovoked thrombosis. PNH is an inhibitor, the lack of which may increase the procoagulant
acquired clonal disorder of hematopoietic stem cells occur- activity of tissue factor. PNH platelets, which are sensitive to
ring in both children and adults with no apparent familial the lytic activity of complement, are hyperactive. RBC phos-
predisposition. pholipids released during intravascular hemolysis also may
initiate clotting.
Most of the clinical manifestations of the disease are
Pathophysiology
due to the lack of the complement-regulating protein
Hemolysis in PNH is due to the action of complement on CD59. The monoclonal antibody eculizumab, which binds
abnormal RBCs. Compared with normal RBCs, PNH RBCs the complement component C5, thereby inhibiting terminal

complement activation, decreases hemolysis of RBCs and of most cell lineages. FLAER is most useful to assay the GPI
the tendency to thrombosis as well. The drug does not anchor on granulocytes because aerolysin binds weakly to
seem to alter the defect in hematopoiesis. Thus, although glycophorin on RBCs.
deficient hematopoiesis is probably related to deficiency of
GPI-anchored proteins, it is not related to complement
sensitivity.
Although chronic hemolytic anemia is a common manifesta-
Laboratory findings tion, only a minority of patients report nocturnal hemoglo-
binuria. The degree of anemia seen in PNH varies in affected
There are no specific morphologic abnormalities of the individuals from minimal to quite severe. Symptoms related
RBCs in PNH. RBCs may be macrocytic, normocytic, or to episodes of hemolysis include back and abdominal pain,
microcytic, the latter occurring when iron deficiency devel- headache, and fever. Exacerbations of hemolysis can occur
ops because of chronic urinary iron loss from intravascular with infections, surgery, or transfusions. Several symptoms
hemolysis. With or without iron deficiency, the reticulocyte in PNH may be related to the ability of free plasma hemoglo-
count may not be as elevated as expected for the degree of bin to scavenge nitric oxide. These include esophageal spasm,
anemia. This is due to underlying bone marrow dysfunction male erectile dysfunction, renal insufficiency, thrombosis,
that often accompanies the PNH. Leukopenia and thrombo- and pulmonary hypertension.
cytopenia often are present. Serum LDH usually is elevated Aplastic anemia has been diagnosed both before and after
and may suggest the diagnosis in the patient with minimal the identification of PNH. PNH clones are present in
anemia. Iron loss may amount to 20 mg/d, and urine hemo- approximately 20% of patients with severe aplastic anemia.
siderin often is identified. Bone marrow examination reveals Approximately 20% of patients with myelodysplastic syn-
erythroid hyperplasia unless there are associated bone mar- dromes have PNH clones. Hemolysis in the setting of bone
row disorders. marrow hypoplasia or myelodysplastic or myeloproliferative
disorders should suggest the diagnosis of PNH. Infections
Laboratory diagnosis associated with leukopenia and bleeding due to thrombocy-
topenia contribute to increased mortality. An increased inci-
The laboratory diagnosis of PNH formerly relied on the dence of acute leukemia also has been reported.
demonstration of abnormally complement-sensitive eryth- Patients frequently have thrombotic complications that
rocyte populations. Ham first described the acidified serum can be life threatening and may represent the initial manifes-
lysis test in 1938. In that test, acidification of the serum acti- tation of PNH. In addition to venous thrombosis involving
vates the alternative pathway of the complement, and an extremity, there is a propensity for thrombosis of unusual
increased amounts of C3 are fixed to RBCs lacking comple- sites such as hepatic veins (Budd-Chiari syndrome), other
ment regulatory proteins. Complement sensitivity of PNH intra-abdominal veins, cerebral veins, and venous sinuses.
RBCs also can be demonstrated in high-concentration Thus, complaints of abdominal pain or severe headache
sucrose solutions, the basis for the “sugar water” or sucrose should alert the clinician to the consideration of thrombosis
hemolysis test. These tests are primarily of historical interest in the patient with PNH. The thrombotic tendency is par-
and are not used routinely in the clinical laboratory because ticularly enhanced during pregnancy.
flow cytometry techniques aimed specifically at demonstrat-
ing the deficiency in expression of GPI-anchored proteins in
Treatment
PNH are readily available. Using commercially available
monoclonal antibodies, blood cells can be analyzed for The clinical manifestations of PNH are highly variable
expression of the GPI-anchored proteins CD55 (DAF) and among patients. For patients with PNH clones numbering
CD59 (MIRL). These methods have the sensitivity to detect <10%, no clinical intervention is needed. Because expansion
small abnormal populations; because monocytes and granu- of the clone may occur, however, the size of the clone should
locytes have short half-lives and their numbers are not be monitored every 6-12 months. Anemia is often the domi-
affected by transfusion, analysis of GPI-anchored proteins nant issue in PNH. Glucocorticoids can reduce complement
on neutrophils or monocytes rather than RBCs is preferred. activation and decrease the hemolysis; however, high doses
A new assay is being used increasingly to detect GPI- are frequently necessary, and every-other-day administra-
deficient blood cells. The fluorescein-labeled aerolysin tion has been recommended to reduce the adverse effects.
(FLAER) assay exploits a property of aerolysin, the principle Iron may be required to replace the large urinary losses seen
virulence factor of the bacterium Aeromonas hydrophila, in PNH. Folate supplementation usually is recommended as
which binds selectively with high affinity to the GPI anchor well. Erythropoietin (10,000-20,000 U three times weekly)

may be helpful for patients with inadequate reticulocyte who are at particularly increased risk for thrombosis; low–
responses. Transfusion may be necessary when these mea- molecular weight heparin may be useful in these patients
sures fail to maintain adequate hemoglobin levels. during pregnancy and the puerperal period. Eculizumab is a
Eculizumab is a humanized monoclonal antibody that pregnancy category C pharmaceutical; however, there are
was engineered to reduce its immunogenicity; importantly, recent case reports of its apparent safe use in pregnancy.
it is unable to bind Fc receptors on cells or to activate com- Also, patients with PNH undergoing surgery should receive
plement. It binds to C5 and blocks the terminal complement prophylactic anticoagulation in the perioperative period.
sequence. The US Food and Drug Administration approved The recommended duration of either prophylactic or thera-
its use in PNH to treat hemolysis based on efficacy in two peutic anticoagulation has not been established.
phase 3 clinical trials. Eculizumab reduces intravascular but
not extravascular hemolysis, eliminates or reduces transfu-
Prognosis
sion requirement in almost all patients, improves quality of
life, improves pulmonary hypertension, and decreases the The median survival for PNH is 10-15 years. Thrombotic
risk of thrombosis. It does not treat the marrow failure. It events, progression to pancytopenia, and age >55 years at
must be used indefinitely because it does not treat the under- diagnosis are poor prognostic factors. The development of a
lying cause of PNH. myelodysplastic syndrome or acute leukemia markedly
Although eculizumab generally is well tolerated, its most shortens survival. Patients without leukopenia, thrombocy-
serious complication is sepsis due to Neisseria organisms. topenia, or other complications can anticipate long-term
Patients congenitally lacking one of the terminal comple- survival.
ment components, C5 to C9, are known to be at risk for
Neisseria infection. Patients receiving eculizumab are at risk
because of its inhibition of the terminal complement Clinical case (continued)
sequence. Vaccination against Neisseria meningitidis is rec-
The patient presented in this section likely has PNH. She has
ommended 2 weeks before starting therapy. Revaccination evidence of hemolysis and marrow failure. The diagnosis can
every 3-5 years may be important because eculizumab is be confirmed by flow analysis for CD55 and CD59 on granulo-
given for an indefinite period. Because vaccination does not cytes, revealing a population of cells with absence of GPI-linked
eliminate the risk completely, patients should be told to seek proteins. Treatment is aimed at the major clinical presentation.
medical attention for any symptoms consistent with Neisse- Eculizumab is effective in decreasing hemolysis and thrombosis,
ria infection. but not marrow failure. Thrombosis is treated with anticoagula-
Allogeneic hematopoietic stem cell transplantation is the tion; thrombolytic therapy may be employed if the thrombosis is
only known cure for PNH. Because of the high risk for seri- acute. There are no randomized studies to support anticoagula-
ous complications including death, however, such treatment tion for prophylaxis of thrombosis, but it is prudent to employ
prophylaxis in high-risk situations for thrombosis, such as preg-
should be reserved for patients with severe pancytopenia or
nancy or surgery. If pancytopenia is marked, immunosuppressive
the rare individuals whose hemolysis or thrombosis is not
therapies, such as antithymocyte globulin and cyclosporine,
controlled by eculizumab. For patients with PNH and mar-
have been used. Allogeneic marrow transplantation has been
row failure who lack an HLA-matched sibling donor, immu- performed in selected cases, primarily those with severe marrow
nosuppressive therapy may be attempted. failure and an HLA-matched sibling donor. Marrow transplanta-
Thrombosis is the leading cause of death in PNH patients. tion is the only potentially curative therapy of PNH.
Thrombosis should be treated promptly with anticoagula-
tion. Thrombolytic therapy may be considered as well,
depending on the extent and location of the clot. In contrast
to anticoagulation as treatment, prophylactic anticoagula- Key points
tion is controversial. In one large, nonrandomized trial, pri- • PNH is an acquired clonal hematopoietic stem cell disorder
mary prophylaxis with warfarin decreased the risk of caused by a somatic mutation of the pig-A gene that results in
thrombosis in patients with large PNH clones (>50% PNH hematopoietic cells lacking GPI-linked proteins.
granulocytes). Because eculizumab also decreases the risk of • Patients may experience chronic hemolytic anemia, cytope-
thrombosis, however, prophylactic anticoagulation is not nias, or a thrombotic tendency.
indicated in these patients based on the current state of • Flow cytometric techniques to identify cell populations lacking
knowledge. The bigger question concerns prophylaxis in GPI-linked proteins (CD55 and CD59) have replaced the sucrose
hemolysis and Ham tests.
patients who do not require eculizumab; in general, lacking
• PNH clones have been identified in individuals without
a randomized trial, it is probably not indicated until further
hematologic abnormalities.
studies are available. The exception may be pregnant women

Key points (continued)

• Bone marrow failure often precedes or follows clinical PNH.
• Steroid therapy along with supportive measures can amelio-
rate the hemolytic anemia.
• Eculizumab, a monoclonal antibody directed against C5, elimi-
nates or reduces hemolysis, improves quality of life, and
decreases the risk of thrombosis.
• Neisseria sepsis is a potentially fatal complication of eculizum-
ab therapy. Vaccination against Neisseria should be given
2 weeks before initiation of eculizumab.
• Prompt evaluation is indicated for symptoms of thrombosis,
particularly at unusual sites. Anticoagulation is indicated for
documented thrombosis and thrombolytic therapy may be
useful, depending on the location and size of the clot.
• Prophylactic warfarin seems to prevent thrombosis in patients
with large PNH clones, but its use for this purpose is controver-
sial, at least in patients who respond to eculizumab.
Figure 7-16 Schistocytes.
• Allogeneic hematopoietic cell transplantation has curative
potential. Because of the risk of serious or fatal complications, its
use should be reserved for those patients with severe cytopenias Pathophysiology
or patients with severe hemolysis or thrombosis refractory to
eculizumab. Among the several causes of fragmentation hemolysis, the
common thread is mechanical damage to RBCs, resulting in
the presence of fragmented RBCs or schistocytes on the
Fragmentation hemolysis blood smear. When microvascular or endothelial injury is
present, the process is termed microangiopathic hemolytic
Clinical case anemia. When thrombosis is part of the picture, the term
thrombotic microangiopathy is used. In disseminated intra-
A 63-year-old male is referred for evaluation of anemia.
vascular coagulation (DIC), the microangiopathic hemolytic
His past history is significant for oxygen-dependent chronic
obstructive pulmonary disease, coronary artery disease, a
anemia is accompanied by activation and consumption of
mechanical aortic valve placed in 1986, and mild heart failure. soluble clotting factors, resulting in prolongation of the pro-
On examination, he has distant breath sounds and a thrombin time and activated partial thromboplastin time,
grade III/VI systolic ejection murmur heard at the left upper- whereas TTP-HUS is associated with activation of platelets
sternal border. Mild scleral icterus is noted. Laboratory data but not soluble clotting factors.
are significant for a hematocrit of 21% (normal 2 years Injury to blood vessel endothelium, intravascular clotting,
prior). Corrected reticulocyte count is elevated at 3%, LDH and primary platelet activation all result in formation of
1,686 IU/dL, and indirect bilirubin 3.4 mg/dL. Examination of fibrin strands in the circulation. The shearing force gener-
the blood smear reveals schistocytes, hypochromic RBCs, and a ated as the RBCs pass through the fibrin strands causes the
few cigar-shaped RBCs.
RBCs to be cut into small irregular pieces. RBCs may be bro-
ken into pieces by direct mechanical trauma as may occur in
Fragmentation hemolysis takes place within the vasculature.

Laboratory features common to both intra- and extravascu- Table 7-11 Differential diagnosis of microangiopathic hemolytic
lar hemolysis include increased concentrations of plasma anemia
bilirubin and LDH and decreased concentration of plasma Thrombotic thrombocytopenia purpura
haptoglobin. Additional features characteristic of intravas- Atypical hemolytic uremic syndrome
cular as opposed to extravascular hemolysis include the pres- Disseminated intravascular coagulation
ence of free Hb in the plasma and urine, resulting in red HELLP syndrome
urine and pink plasma. If the hemolysis is chronic, urine Cardiac valve disease
hemosiderin may be present. In fragmentation hemolysis, Malignancy
schistocytes are a prominent feature of the blood smear Vasculitis
(Figure 7-16). The differential diagnosis of microangiopathic Malignant hypertension
Scleroderma renal crisis
hemolytic anemia (MAHA) is summarized in Table 7-11.

march hemoglobinuria or with a dysfunctional mechanical Thrombotic thrombocytopenic purpura

heart valve in which high-velocity jets of blood strike an
TTP is due to the deposition of platelet microthrombi along
unendothelialized surface. The resulting small RBC frag-
the endothelium of small vessels of multiple organs. The
ments are self-sealing and continue to circulate, albeit with
classic clinical presentation consists of microangiopathic
shortened survival. This is due in part to their decreased
hemolytic anemia and thrombocytopenia. In advanced
deformability, which results in accelerated removal by the
stages, fever, renal failure, and CNS involvement are seen.
spleen.
TTP may be confused with eclampsia, HUS, the HELLP
(hemolysis, elevated liver enzymes, low platelets) syndrome,
and acute fatty liver of pregnancy (AFLP), all of which can
Etiology
present with microangiopathic anemia. A critical distin-
Cardiac valve hemolysis guishing feature between TTP and DIC is the presence of
consumptive coagulopathy in the latter. Malignant hyperten-
Hemolysis may occur with calcific or stenotic native heart
sion and renal crisis in scleroderma may resemble TTP, pre-
valves, although it is usually very mild and well compensated
senting with microangiopathic hemolysis, thrombocytopenia,
in the absence of severe valvular disease. Mechanical heart
and renal insufficiency. Rapid control of the hypertension is
valves have a smaller diameter than the native heart valve.
important in these patients. (TTP and the HUS are covered
Normally, the hemodynamic consequences are minimal.
in detail in Chapter 10.)
Prosthetic valve dysfunction or perivalvular regurgitation
Certain drugs, especially antineoplastic agents, can cause
may result in intravascular hemolysis, however. An aged or
microangiopathic hemolysis that resembles TTP. Mitomycin,
damaged valve surface may become irregular, leading to
a chemotherapeutic agent used in the treatment of gastroin-
thrombus formation. In a high-flow state, such as exists
testinal malignancies, has been best described. Gemcitabine,
across the aortic valve or across a regurgitant mitral valve,
another chemotherapeutic agent, also has been implicated.
the formation of jets and turbulent flow results in high shear
The mechanism has been proposed to be direct endothelial
stress that may exceed the stress resistance of the normal
injury. Both tacrolimus and cyclosporine used to prevent
RBC. Hemolysis may be made worse with concomitant car-
and treat graft-versus-host disease can cause a similar syn-
diac failure or high-output states. Recently designed biopros-
drome. Both ticlopidine and clopidogrel, antiplatelet agents,
thetic heart valves have a significantly decreased risk of
have been associated paradoxically with TTP.
thrombus formation and a lower rate of traumatic hemoly-
sis. A recent prospective study reported a 25% rate of mild
Hemolytic uremic syndrome
subclinical hemolysis with a mechanical prosthesis and a 5%
rate with a bioprosthesis. HUS is characterized by red cell fragmentation, thrombocy-
Ruptured chordae tendinae, aortic aneurysm, and patch topenia, and renal failure. HUS has been divided into typical
repair of cardiac defects, as well as intraventricular assist and atypical forms. The typical form is usually caused by
devices and aortic balloon pumps used in the management infection with Escherichia coli. The atypical form of HUS
of severe heart failure, have been associated with traumatic (aHUS) is due to dysregulation of the complement alternative
hemolysis. Intravascular hemolysis has been described after pathway. aHUS is becoming increasingly recognized with the
cardiopulmonary bypass and is thought to be secondary to ability to distinguish this cause of MAHA and thrombocyto-
both physical damage and complement activation. penia from TTP with the use of ADAMTS13 testing. Indi-
Anemia is variable in patients with prosthetic valve hemo- viduals with aHUS do not always have thrombocytopenia as
lysis. The blood smear may include abnormal erythrocytes seen in a French registry of patients where 16% did not have
with schistocytes and cells with abnormal membrane a low platelet count at presentation. It is now thought that
projections. aHUS occurs in individuals with an underlying complement
With chronic hemolysis, hemoglobin is lost in the urine, risk factor who have a secondary trigger. Triggers can include
leading to iron deficiency. Iron-deficient RBCs are mechani- infection, pregnancy, drug exposure, and malignancy.
cally fragile, which can worsen hemolysis, exacerbate ane- Distinguishing TTP or other MAHA from aHUS is cru-
mia, and lead to further hemodynamic compromise that cially important, as therapy with eculizumab has been shown
may increase the rate of hemolysis. At times, this cycle may to be more effective than plasma exchange in the manage-
be abated by correction of iron deficiency or by RBC transfu- ment of aHUS. Although mostly based on case reports, ecu-
sion. The addition of erythropoietin to increase RBC pro- lizumab has been shown to result in full recovery of baseline
duction may compensate for ongoing hemolysis. If anemia is renal function in 80% of children and 31% of adults. With
severe or fails to respond to the conservative measures, valve the significant morbidity associated with aHUS, a French
replacement may become necessary. working group recommended the use of eculizumab as a

first-line therapy in children with aHUS. In a patient with dexamethasone in HELLP, previously supported by small stud-
MAHA and thrombocytopenia with normal ADAMTS13 ies, has not proven helpful in subsequent randomized trials.
levels, eculizumab therapy should be considered.
Kasabach-Merritt syndrome
Disseminated intravascular coagulation Kasabach-Merritt syndrome is characterized by consumptive

coagulopathy occurring in the capillaries of a large kaposi-
DIC is associated with many disorders, including sepsis, form hemangioendothelioma. Microangiopathic hemolytic
obstetrical catastrophes, and malignancy. The disorder is anemia accompanies evidence of DIC. A number of treat-
characterized by activation of coagulation and generation of ments, including glucocorticoids, chemotherapy, interferon-
excess thrombin leading to deposition of fibrin strands in alfa, embolization, and surgical removal have been tried with
arterioles, venules, and capillaries. MAHA may be present some success.
but often is not severe enough to cause morbidity. Dissemi-
nated malignancy presents with microangiopathic anemia Foot strike hemolysis
and DIC in approximately 5% of cases. Fibrin deposition
and vascular disruption by the malignancy itself have both Foot strike hemolysis, also known as march hemoglobinuria,
been noted. Mucin-producing adenocarcinomas are fre- has been described in soldiers subjected to long foot-stomping
quent offenders. Promyelocytic leukemia characteristically marches in rigid-soled boots, long-distance runners, conga
presents with DIC due, at least in part, to the release of tissue drummers, pneumatic hammer operators, and karate enthusi-
factor from promyelocytic granules. If treatment is effective asts. Hemoglobinuria occurs shortly after the episode of exer-
at reversing the underlying condition causing DIC, hemoly- cise. The hemolysis is caused by direct trauma to RBCs in the
sis and the coagulopathy often resolve. blood vessels of the extremities. This condition has become
much less common as shoe technology has improved. Cessa-
HELLP syndrome
tion of the activity always leads to resolution of the hemolysis.
The HELLP syndrome (microangiopathic hemolytic ane- Clinical case (continued)

mia, elevated liver enzymes, and a low platelet count) is a
serious complication of late pregnancy that is part of a spec- The patient presented in this section has evidence of a moder-
trum with preeclampsia. Thrombocytopenia and MAHA ate hemolytic anemia. The blood smear is consistent with both
traumatic hemolysis and iron deficiency, as schistocytes and
with or without renal failure may also occur in pregnancy
hypochromic and cigar-shaped cells were noted on review of the
due to TTP-HUS and AFLP. It is important to distinguish
peripheral blood smear. Valve structure and function should be
TTP, HUS, and AFLP from HELLP and preeclampsia for
investigated with an echocardiogram or other imaging stud-
therapeutic reasons. The clinical features are quite similar, ies. Other causes for hemolysis should be ruled out. The patient
however, and the correct diagnosis is often elusive. should be evaluated for iron deficiency. If further evaluation
Although not absolute, the timing of onset during the confirms iron deficiency, the patient should receive oral iron.
pregnancy may be helpful. In general, TTP, and aHUS occurs Erythropoietin administration also should be considered. He
earlier in gestation than AFLP, preeclampsia, or HELLP; appears to be a poor surgical candidate, but valve replacement
approximately two-thirds of TTP cases in pregnancy occur may become necessary if conservative treatment fails.
in the first or second trimester. Most cases of AFLP, pre-
eclampsia, and HELLP occur after 20 weeks of gestation, the
great majority in the third trimester. A history of proteinuria Hemolytic anemia due to chemical
and hypertension before the onset of hemolysis, liver abnor- or physical agents
malities, and thrombocytopenia favors the diagnosis of pre-
eclampsia or HELLP, whereas a high LDH level with only Clinical case
modest elevation of AST favors TTP. Severe liver dysfunction A 23-year-old female is referred for evaluation of mild anemia
or liver failure favor AFLP. noted during a workup of liver function test abnormalities. Her
The characteristics of the coagulopathy are different as well. recent history has been significant for bizarre schizophrenic-like
Whereas both TTP and HELLP are characterized by thrombo- behavior and arthritis. She has not had a menstrual period in sev-
cytopenia in HELLP and more so in AFLP, DIC may also be eral months. Recent slit-lamp examination by an ophthalmolo-
present with evidence of consumptive coagulopathy. In TTP, gist revealed golden brown pigmentation of the cornea. Physical
only thrombocytopenia is seen without evidence of consump- examination is otherwise unremarkable. Laboratory data suggest
a Coombs-negative hemolytic anemia. Liver enzymes are moder-
tion of soluble clotting factors. Treatment of HELLP and
ately elevated. A ceruloplasmin level returns low at 11 mg/dL.
AFLP consist of prompt delivery of the fetus. The use of

The use of primaquine and dapsone to prevent or treat RBC membrane through damage to protein band 3 and
Pneumocystis jirovecii in acquired immunodeficiency syn- other integral proteins. In the northwestern United States,
drome (AIDS) patients has become fairly common. Both hemolysis has been noted after hobo spider (Tegenaria
drugs may cause methemoglobinemia in high doses in nor- agrestis) bites. Microangiopathic hemolysis may occur after
mal patients and may precipitate hemolysis in patients with the bite of pit vipers (eg, rattlesnakes, cottonmouth mocca-
G6PD deficiency. Most AIDS clinics screen their patients for sins, copperheads) associated with DIC induced by the
G6PD deficiency before starting either of these drugs. Met- venom. Cobra venom contains phospholipases that may
hemoglobinemia and G6PD deficiency are covered in detail cause hemolysis. Massive bee and wasp stings rarely have
earlier in this chapter. been associated with intravascular hemolysis.
Ribavirin, used to treat HCV infection, is a frequent cause Fragmentation hemolysis has been described after injury
of hemolysis by an unknown mechanism. The hemolysis is from a variety of physical agents. Thermal injury can lead to
dose dependent and decreases or resolves with decreased severe intravascular hemolysis. This is best described in
ribavirin dose or discontinuation of the drug. The rate of patients suffering from extensive third-degree burns. At
sustained HCV response, however, also decreases with dose temperatures above 47°C, irreversible injury occurs to the
reduction. Erythropoietin has been used as an adjunct to RBC membrane. Shortened RBC survival has been noted
maintain ribavirin therapy at full dose. after ionizing radiation exposure.
Phenazopyridine is a bladder analgesic that is used to treat
the symptoms of cystitis. In high doses, it has been associated
Clinical case (continued)
with oxidative hemolysis. It is recommended that patients be
treated for no more than 2 days. Overdoses, prolonged The patient presented in this section displays the classic historical
administration, and renal insufficiency have led to methe- and physical findings of Wilson disease. The low ceruloplasmin
moglobinemia and severe hemolysis, occasionally severe level is diagnostic. Hemolytic anemia has been well described
in this disease. Once the severity of her liver disease is further
enough to induce acute renal failure.
evaluated, treatment with penicillamine should be considered.
Lead intoxication can lead to a modest shortening of
The hemolytic anemia is likely to resolve as excess copper is
RBC life span, although the anemia more often is due to
removed.
an abnormal heme synthesis and decreased production
of erythrocytes. On the blood smear, RBCs are normocytic,
hypochromic, with prominent basophilic stippling in young Hemolytic anemia due to infection
polychromatophilic cells.
Copper causes hemolysis through direct toxic effects on Clinical case
RBCs and has been observed in association with hemodialy-
A 21-year-old man went to the emergency department of his
sis. Copper accumulates in RBCs and disrupts normal meta-
local hospital complaining of fever and shaking chills. He had
bolic function through a variety of mechanisms, including
just returned from a 6-month deployment in eastern Afghani-
oxidation of intracellular reduced glutathione, hemoglobin,
stan with the US Army. He has been home for 2 weeks on leave
and NADPH and inhibition of multiple cytoplasmic before reporting for his next duty assignment in the United
enzymes. Wilson disease, due to a mutation of the ATP7B States. He states that he faithfully took his malaria prophylaxis
gene, leads to absence or dysfunction of a copper-transporting consisting of mefloquine 250 mg weekly while in Afghanistan;
ATPase encoded on chromosome 13. This subsequently he was instructed to continue the weekly mefloquine for four
results in lifelong copper accumulation. Hemolytic anemia more doses postdeployment, plus primaquine 15 mg daily
may be an early manifestation. The hemolytic process in for the first 2 weeks. On examination, he appeared acutely
Wilson disease varies in severity and duration. Kayser- ill. His vital signs were BP 126/66, pulse 110, respirations 20,
Fleischer rings due to the deposition of copper around the and temperature 39°C. The remainder of the examination was
periphery of the cornea are a key diagnostic finding. Diagno- unremarkable. There was no splenomegaly. A Wright-Giemsa
stained thick blood smear confirmed the diagnosis of Plasmo-
sis can be made by quantitative ceruloplasmin measurements
dium vivax malaria.
or by liver biopsy with assessment of the copper concentra-
tion. Treatment consists of penicillamine, which mobilizes
copper stores. Acute hemolysis in Wilson disease has been Infection may lead to hemolysis through a variety of mecha-
treated successfully with plasmapheresis. nisms. Parasites may directly invade RBCs, leading to pre-
Certain spider bites may be associated with traumatic mature removal by macrophages of the liver and spleen.
RBC fragmentation. In the southern United States, the Alternatively, hemolytic toxins may be produced by the
brown recluse spider (Loxosceles reclusa) is the most com- organism and lead to damage of the RBC membrane. Devel-
mon species causing hemolysis. The toxin proteolyzes the opment of antibodies to RBC surface antigens has been well

described with certain viral and bacterial illness, especially Immune hemolysis associated with infections
infectious mononucleosis and Mycoplasma pneumoniae
Pneumonia caused by M. pneumoniae can be associated with
infections. Hypersplenism may ensue, which can further
production of cold agglutinins, IgM antibodies directed
decrease RBC life expectancy. In addition, the antibiotic
against the RBC I antigen. Hemolytic anemia associated with
drugs used to treat a variety of these infections may lead to
M. pneumoniae may occur during the second or third week
further hemolysis in G6PD-deficient individuals. Anemia
of the illness. The onset of the hemolysis may be rapid, usu-
that occurs with concomitant acute or chronic infection is
ally occurring after recovery from respiratory symptoms.
likely to be multifactorial, with the anemia of chronic inflam-
The clinical presentation often includes dyspnea or fatigue
mation often coexisting and predominating.
and the presence of pallor and jaundice. The blood smear
shows RBC agglutination with or without spherocytosis and
RBC membrane injury caused by bacteria with polychromatophilia (Figure 7-14). When ethylenedi-
aminetetraacetic acid–anticoagulated blood is cooled in a
Clostridial sepsis test tube, RBC agglutination can be seen; disagglutination
Clostridial sepsis is seen in patients with anaerobic subcu- occurs when the blood is warmed. Cold agglutination titers
taneous infections, in body areas of impaired circulation, at the onset of hemolysis usually exceed 1:256 and may reach
after trauma, after septic abortion or postpartum sepsis, higher levels, although they are typically lower than in
and in patients with acute cholecystitis with gangrene monoclonal cold agglutinin disease. The DAT is positive for
of the gallbladder or bowel necrosis. Severe neutropenia of complement deposition on RBCs. The hemolytic anemia
any cause may be a significant risk factor. The α toxin associated with Mycoplasma pneumonia is self-limited, tran-
of Clostridium is a lecithinase (phospholipase C) that dis- sient, and usually mild, although severe cases requiring cor-
rupts the lipid bilayer structure of the RBC membrane, ticosteroid therapy or plasmapheresis have been reported.
leading to membrane loss and hemolysis. Brisk intravascu- Infectious mononucleosis caused by Epstein-Barr virus
lar hemolysis with spherocytosis seen on the peripheral infection may be associated with hemolytic anemia due to
blood smear is accompanied by hemoglobinemia, hemo- cold agglutination. The cold agglutinin in this case is an IgM
globinuria, and severe anemia. The plasma may be a bril- antibody directed against the i antigen. Severe hemolytic
liant red color, and there may be dissociation between the anemia associated with infectious mononucleosis is unusual,
hemoglobin and hematocrit values because of the plasma although anti-i antibodies frequently are present. When
hemoglobin levels reaching several grams per deciliter. hemolytic anemia occurs, the mechanism involves fixation
Acute renal failure may ensue secondary to excessive of complement on the RBC membrane by IgM antibodies.
hemoglobinuria, but the exact mechanism remains dis- Hemolysis proceeds either by completion of the comple-
puted. Renal failure and hepatic failure contribute to the ment cascade through C9 or by opsonization of RBCs with
high mortality in clostridial sepsis. fragments of C3 leading to phagocytosis of RBCs by macro-
phages in the liver or spleen.
Several other viral infections have been associated with
Hemolytic anemias with Gram-positive and
AHA. These include cytomegalovirus, herpes simplex, rube-
Gram-negative organisms
ola, varicella, influenza A, and HIV. Postviral acute hemolytic
Septicemia and endocarditis caused by Gram-positive bacte- anemia in children may be due to the formation of a cold-
ria, such as streptococci, staphylococci, S. pneumoniae, and reactive hemolytic IgG antibody of the Donath-Landsteiner
Enterococcus faecalis are often associated with hemolytic ane- type, which induces complement lysis of RBCs.
mia. The anemia in patients with infections due to Gram- Microangiopathic hemolytic anemias associated with
positive cocci appears to result from the direct toxic effect of infection include bacteremia with Gram-negative organisms,
a bacterial product on erythrocytes. Salmonella typhi infec- staphylococci, meningococci, and pneumococci, all of which
tion may be accompanied by severe hemolytic anemia with can lead to DIC with endothelial damage and fibrin thrombi
hemoglobinemia. In typhoid fever, the onset of hemolysis within the microcirculation. RBC injury results from mechan-
may occur during the first 3 weeks of illness, with anemia ical fragmentation by fibrin strands in the vasculature. Micro-
lasting from several days to 1 week. Salmonella and other vascular damage induced by meningococcal and rickettsial
microorganisms can cause direct agglutination of RBCs in infections (eg, Rocky Mountain spotted fever) may be associ-
vitro, but it is not known whether this phenomenon contrib- ated with DIC, thrombocytopenia, microvascular thrombi,
utes to in vivo hemolysis. In approximately one-third of and fragmentation hemolytic anemia. Patients with either
patients with typhoid fever, a positive DAT develops, but congenital or tertiary syphilis may develop paroxysmal cold
hemolytic anemia is not manifest in all cases. hemoglobinuria. Whereas paroxysmal cold hemoglobinuria

used to be fairly common in the late 19th and earlier 20th

centuries, it is rare in the 21st century due to the disappear-
ance of congenital and tertiary syphilis.
Hemolytic anemia associated with

parasitic infestation of RBCs
Malaria
Malaria is the most common cause of hemolytic anemia

worldwide. Transmitted by the bite of an infected female
Anopheles mosquito, sporozoites that are injected from the
mosquito make their way to liver cells. Merozoites enter into
the bloodstream 1-2 weeks later. Hemolysis in malaria results
directly from erythrocytic infestation by Plasmodium organ-
isms (Figures 7-17 and 7-18). Noninfected RBCs may be Figure 7-18 Intraerythrocyte parasite Plasmodium vivax: trophozoite
hemolyzed by poorly understood mechanisms. Infested (ring form) and female gametocyte.
erythrocytes are selectively removed from the circulation by
the spleen, with some RBCs reentering circulation after
With P. falciparum infection, intravascular hemolysis may
splenic pitting of parasites. Previously infested erythrocytes
be severe and associated with hemoglobinuria (blackwater
manifest membrane and metabolic abnormalities along with
fever). Another potentially lethal complication of P. falci-
decreased deformability. In addition, the Plasmodium species
parum infection, cerebral malaria, results from expression of
digests the host RBC hemoglobin for its own use as a
P. falciparum erythrocyte protein 1 on the membranes of
nutrient.
infected RBCs. These RBCs adhere to receptors on vascular
The severity of the hemolytic process is often out of pro-
endothelium in various organs, including the central ner-
portion to the degree of parasitemia. P. vivax and Plasmo-
vous system, resulting in vaso-occlusion and neurologic
dium ovale invade only reticulocytes, whereas Plasmodium
manifestations.
malariae invades only mature erythrocytes. P. falciparum
Diagnosis of malaria is based on identification of parasite-
invades erythrocytes of all ages and is associated with more
infected RBCs on a thick Wright-stained blood smear. The
severe hemolysis. In areas where malaria has been a frequent
distinction of P. falciparum infection from the other species
cause of death for many centuries, a number of genetic poly-
is important because its treatment may constitute a medical
morphisms are prevalent, including G6PD deficiencies, thal-
emergency. The findings of two or more parasites per RBC
assemias, and hemoglobinopathies. These polymorphisms
and infestation of >5% of RBCs are characteristic of P. falci-
have in common the ability to interfere with the ability of the
parum infection.
malaria parasites to invade RBCs.
Chemoprophylaxis should be offered to all people plan-
ning travel to known endemic areas. The hemolytic anemia
of malaria resolves after successful therapy with quinine,
chloroquine, artemisinin, and other drugs, depending on the
species of malaria. Many of these agents may be associated
with drug-induced hemolysis in patients with G6PD
deficiency.
Babesiosis
Babesiosis is a protozoan infection caused by Babesia

microti. Once thought to be rare, outbreaks have been
described with increasing frequency on Nantucket Island, in
Cape Cod, in northern California, and in several other
North American locations. The organism is transmitted by
the bite of the Ixodes tick, which infects many species of wild
birds and domestic animals and occasionally humans. Babe-
Figure 7-17 Intraerythrocyte parasite Plasmodium falciparum. siosis rarely may be transmitted by transfusion with fresh or

Oroya fever. On Giemsa-stained blood films, red-violet rods

of varying lengths can be identified on RBCs and represent
the bacteria. Effective treatment exists and consists of peni-
cillin, streptomycin, chloramphenicol, or tetracycline.

The patient was admitted for treatment. The CDC Malaria
Hotline (1-770-488-7788) was called, and the regimen of chlo-
roquine and primaquine was recommended for vivax malaria
acquired in Afghanistan. He made a full recovery. He ultimately
admitted that he had forgotten to take his prophylactic medica-
tions after leaving Afghanistan. The most common cause of
failure of malaria prophylaxis in military or civilian populations
is noncompliance. Because of the importance of primaquine in
terminal prophylaxis and treatment of vivax malaria, it is cur-
rently the policy of the US military to screen all personnel for
G6PD deficiency.
Figure 7-19 Intraerythrocyte parasite Babesia microti.

Key points
• Red cell fragmentation syndromes are diverse in etiology.
• In all suspected cases of hemolytic anemia, the blood smear
frozen-thawed RBCs. Infection leads to a clinical syndrome should be examined carefully for schistocytes. Their presence
of fever, lethargy, malaise, and hemoglobinuria 1-4 weeks can direct differential diagnosis.
after the bite. Hemolytic anemia due to intravascular hemo- • Red cell destruction can be at the macrovascular or microvas-
lysis occurs, and renal and liver function tests are frequently cular (microangiopathic) level of the circulatory system. Classic
abnormal. The disease is often asymptomatic in people with examples include heart valve hemolysis, DIC, and TTP.
intact spleens; patients who have undergone splenectomy • Various chemical exposures or physical agents can cause
are at high risk for severe symptomatic infection. Babesia fragmentation hemolysis.
infection can be diagnosed by demonstrating typical • Infection can cause accelerated RBC destruction through a
variety of mechanisms, including direct invasion, toxin produc-
intraerythrocytic parasites on a thin blood smear (Figure
tion, and immune mechanisms.
7-19). Standard treatment has consisted of clindamycin and
• Malaria, the most common infectious disease worldwide,
quinine. Recent studies have suggested that atovaquone plus
causes hemolysis through both direct parasitic invasion of RBCs
azithromycin is an equally efficacious regimen, yet better and through alterations in noninfected cells. Malaria can be
tolerated. diagnosed by thorough review of a thick Wright-stained
peripheral blood smear.
Bartonellosis
Bartonellosis, caused by Bartonella bacilliformis, manifests in

two clinical stages: an acute hemolytic anemia and a chronic Bibliography
granulomatous phase. The microorganism enters the blood
following the bite of an infected sand fly. The infective Bar- Brodsky RA. How I treat paroxysmal nocturnal hemoglobinuria.
Blood. 2009;113:6522-6527. An excellent, well-referenced review
tonella agent adheres to the membrane of RBCs that are then
on the management of PNH.
removed by the spleen. The hemolytic anemia of bartonello-
Cappellini MD, Cohen A, Porter J, et al., eds. Guidelines for
sis develops rapidly and may be severe, with hemoglobinemia the Management of Transfusion Dependent Thalassaemia (TDT).
and hemoglobinuria. When untreated, this disorder is associ- 3rd ed. Nicosia, Cyprus: Thalassaemia International Federation;
ated with high mortality. Survivors manifest a second stage of 2014. http://www.ncbi.nlm.nih.gov/books/NBK269382/.
the disease with cutaneous granulomas. Bartonellosis is com- A well written TIF publication on guidelines for the diagnosis and
mon in South America and has been reported in the Peruvian management of patients with thalassemia including transfusions,
Andes and parts of Brazil, where it is also known as iron overload and associated complications.

Bibliography | 183
Gallagher PG. Abnormalities of the erythrocyte membrane. National Heart, Lung, and Blood Institute and Division of
Pediatr Clin North Am. 2013;60:1349-1362. An excellent, easy Blood Diseases and Resources. Evidence based management
to read, review of red cell membrane disorders. of sickle cell disease: Expert Panel Report. Bethesda, MD:
Garratty G, Petz LD. Approaches to selecting blood for transfusion National Institutes of Health. 2014. http://www.nhlbi.nih.gov/
to patients with autoimmune hemolytic anemia. Transfusion. health-pro/guidelines/sickle-cell-disease-guidelines. A useful
2002;42:1390-1392. Outlines the approach to transfusion in monograph addressing diagnosis, counseling, health maintenance,
patients with immune hemolytic anemia, with emphasis on and treatment of sickle cell disease.
selection of blood units and early consultation between the Taher A, Vichinsky E, Musallam K, Cappellini MD, Viprakasi
clinician and the blood bank physician. V; Weatherall D ed. Guidelines for the Management of Non
George, J, Nester C. Syndromes of thrombotic microangiopathy. Transfusion Dependent Thalassemia (NTDT). Nicosia,
N Engl J Med. 2014;371:654-666. A well written overview of Cyprus: Thalassemia International Federation; 2013. http://
approach to thrombotic microangiopathy. www.ncbi.nlm.nih.gov/books/NBK190453/ A well written TIF
Lux SE. Disorders of the red cell membrane. In: Orkin SH, Fisher publication on guidelines for the management of non-transfusion
DE, Ginsburg D, Look AT, Fisher DE, Lux SE, eds. Nathan depedent thalassemia patients. The E-book is available in its
and Oski's Hematology and Oncology of Infancy and Childhood. entirity. The above link will lead the reader to subsequent chapters.
Philadelphia, PA: Saunders; 2015:515-579. Van Wijk R, van Solinge WW. The energy-less red blood cell is
Miller LH, Baruch DI, Marsh K, Doumbo OK. The pathogenic lost: erythrocyte enzyme abnormalities of glycolysis. Blood.
basis of malaria. Nature. 2002;415:673-679. A good review of 2005;106:4034-4042. A comprehensive review of RBC glycolytic
malaria pathogenesis. enzyme abnormalities.

CHAPTER
8
Thrombosis and thrombophilia
David Garcia, Saskia Middeldorp, and Anjali Alatkar Sharathkumar
Introduction, 185 Pathophysiology of thrombosis, 195 Antithrombotic drugs, 210
Venous thromboembolism, 185 Thrombophilias, 196 Bibliography, 217
Introduction and children, with the highest rates found in neonates and
adolescents.
This chapter gives an overview of the pathophysiologic con-
tributors to thrombosis; describes the mechanisms, epidemi-
Deep vein thrombosis and
ology, testing issues, and clinical relevance of inherited and
pulmonary embolism
acquired thrombophilias; discusses the drugs used as anti-
thrombotics; and reviews various clinical, diagnostic, and Symptom
therapeutic aspects of thrombosis. The pediatric sections are
DVT of the pelvic and leg veins presents with varying degrees
authored by Dr. Sharathkumar.
of leg swelling, pain, warmth, and skin discoloration. Symp-
toms are typically diffuse. It is important to recognize that
Venous thromboembolism proximal DVT is defined as involving the popliteal or more
proximal veins, whereas distal DVT involves vessels distal to
Venous thromboembolism (VTE) is inconsistently defined. the trifurcation of the popliteal vein. Localized symptoms are
For some, it encompasses any thrombosis in veins (no mat- more suggestive of a superficial thrombophlebitis. A palpable
ter whether superficial or deep veins) plus pulmonary embo- subcutaneous cord-like firmness is indicative of a superficial
lism; for others, it only signifies deep vein thrombosis (DVT) thrombophlebitis. The onset of symptoms of DVT can be
and pulmonary embolism (PE). The term DVT refers to sudden or subacute over days to weeks. DVT is not infre-
thrombosis involving deep veins of either the leg (popliteal, quently missed or misdiagnosed, as the symptoms can be
femoral, iliac) or arm (brachial, axillary, subclavian, and bra- nonspecific. PE presents with varying degrees of severity of
chiocephalic). With an incidence of 2 to 3 per 1,000 per year, shortness of breath, chest pain that is classically respiratory
estimates suggest that at least 300,000, and as many as dependent, nonproductive cough, and hemoptysis. A massive
600,000, people in the United States develop DVT/PE each PE can lead to sudden death. Small PEs are often asymptom-
year, and that at least 60,000-100,000 deaths each year are atic. There is no uniform definition for the severity or degree
due to DVT/PE. Approximately half of DVT/PE episodes are of PE. The definition can be either anatomic or physiologic.
hospital associated. VTE in children is uncommon, but hos- The physiologic one is preferred for treatment decision mak-
pitalization increases the risk: DVT/PE occurs in 1 in 200- ing, as it is a better predictor of mortality. Any PE that causes
300 hospitalized children. There is a bimodal peak in infants hemodynamic instability (hypotension) is referred to as mas-
sive PE. Submassive PE is the term for PE associated with nor-
mal arterial blood pressure but right ventricular dysfunction.
Conflict-of-interest disclosure: Dr. Garcia: consultancy: Bristol-
Myers Squibb, Portola, Pfizer, Boehringer Ingelheim, and Daiichi
Sankyo. Dr. Middeldorp: consulting fees from Bayer, Boehringer Diagnosis
Ingelheim, Bristol-Myers Squibb, Pfizer, and Daiichi Sankyo and
research support from GlaxoSmithKline, Aspen, the alliance of Bristol- Since VTE is confirmed with objective testing in a rela-
Myers Squibb and Pfizer, and Sanquin Blood Supply. Dr. Sharathkumar; tively small percentage of patients presenting with possible
consulting fees from Baxter, CSL Behring, and Biogen Idec. DTV/PE, clinical scoring systems (eg, the Wells score) have
Off-label drug use: not applicable.
| 185

186 | Thrombosis and thrombophilia
been validated; by defining the pretest probability of disease, Acute therapy

these scores help determine which diagnostic tests will be
Outpatient management of patients with DVT and selected
most appropriate. Selected whole blood or plasma D-dimer
low-risk patients with PE has been shown to be safe, feasi-
tests are well evaluated and useful in the diagnostic workup
ble, cost effective, and, if possible, is the preferred treat-
for DVT and PE. In outpatients with a low pretest probabil-
ment of choice. Hospital admission is appropriate if either
ity for DVT or PE, a negative test with a sensitive D-dimer
the patient is too sick to be managed at home or if social
assay reliably rules out VTE, and no further imaging study is
and financial circumstances make this the safer and more
needed. Outpatients with a low pretest probability for DVT
feasible option.
or PE and a positive D-dimer test and any patient with mod-
Patients with acute VTE need to be anticoagulated to pre-
erate or high pretest probability for DVT or PE needs to
undergo imaging studies. The generalized application of vent the extension of thrombus and decrease mortality.
D-dimer testing, however, is limited by the large number of Apixaban (10 mg orally BID for 7 days; 5 mg orally BID
different assays available, some highly sensitive and others thereafter) and rivaroxaban (15 mg orally BID for 21 days
less sensitive, and a lack of standardization of assays. Because and 20 mg daily thereafter) can be used to treat acute DVT or
clinicians often are not aware of the type of D-dimer assay PE without prior or overlapping parenteral therapy. Intrave-
used by their laboratory or the predictive value of the par- nous unfractionated heparin (UFH), subcutaneous low-
ticular assay available to them, reliance on D-dimer results molecular-weight heparin (LMWH), and fondaparinux are
for clinical decision making for the exclusion of VTE can be all effective and acceptable treatment options and need to
unsafe unless the test has been validated locally. In children, be given for at least 5 days (overlapping with warfarin until
the D-dimer test as a diagnostic tool for VTE has not been the INR is ≥ 2.0, or prior to starting dabigatran or edoxaban,
well studied. The D-dimer test is of limited diagnostic utility if one of these agents is used). In high-risk patients who may
in a variety of conditions (e.g. pregnant patients, patients require thrombolysis, UFH is preferable to direct oral anti-
with cancer) where it is known to be elevated at baseline. coagulants (DOACs), LMWH, or fondaparinux because it
Venous compression ultrasound (CUS) is the most widely has a shorter half-life and easily can be dose-adjusted, dis-
used imaging study to look for DVT of the legs. Sensitivity continued, or reversed (with protamine). In selected patients
and specificity of the test are operator dependent, and an with extensive acute femoral or iliac DVT with symptom
experienced ultrasound technician or physician is key in duration of <14 days and with low bleeding risk, catheter-
obtaining reliable results. It can be challenging, even for an directed thrombolysis with or without mechanical thrombus
experienced operator, to distinguish between acute vs. chronic fragmentation and aspiration can be considered to reduce
thrombus solely based on the CUS. Magnetic resonance (MR) acute symptoms. Whether pharmacomechnical thromboly-
venography of leg or pelvic veins is a sensitive test to detect sis of femoral or iliac DVT confers a net clinical benefit is the
DVTs, but it is expensive and not widely available. Imaging subject of ongoing clinical trials (eg, the ATTRACT trial).
with MR or computer tomography (CT) venography may be May-Thurner syndrome is the term used for the chronic
necessary for upper-extremity DVT, particularly catheter- compression of the left common iliac vein between the
related events, because ultrasound may miss occlusion within overlying right common iliac artery and the fifth lumbar
the superior vena cava and brachiocephalic and subclavian vertebral body posteriorly. Varying degrees of vein narrow-
veins due to interference of the clavicles and ribs. ing with this anatomic variant are common in the general
To diagnose PE, several imaging modalities exist: population. If a May-Thurner syndrome is demonstrated on
ventilation/perfusion (VQ) scanning, PE-protocol chest CT venography or MR imaging in the patient with left-leg femo-
angiography (also known as spiral CT, helical CT or PE-pro- ral or iliac DVT who successfully has received thrombolytic
tocol CT), chest MR angiography, and conventional intrave- therapy, correction of the stenosis using balloon angioplasty
nous contrast pulmonary angiogram. The VQ scan is a and stenting can be considered, although there is no high-
well-validated imaging study. PE-protocol chest CTs have quality evidence that such interventions reduce the risk of
replaced VQ scans as the diagnostic method of choice, how- recurrent VTE or PTS.
ever, because they are easier and faster to perform and have Thrombolytic therapy in PE is indicated for massive
good performance characteristics. Conventional intravenous life-threatening PE (ie, PE with hypotension due to right
contrast pulmonary angiography, once considered the gold ventricular dysfunction). Whether high-risk PE patients
standard for the diagnosis of PE, now is rarely done because without hypotension benefit from thrombolytic therapy
the test is invasive, not widely available, and has a strong neg- remains controversial. Concerns about the risk of cata-
ative predictive value. There is ongoing debate about the clin- strophic bleeding associated with thrombolytic therapy has
ical significance of isolated tiny pulmonary artery filling limited its use in many centers. If thrombolytic therapy is
defects that can be seen on chest CT scans (subsegmental PE). given to a patient with PE, it is recommended that it be given

Venous thromboembolism | 187
systemically via a peripheral vein and with short infusion Table 8-1 Considerations when discussing time-limited versus
time, such as 2 hours. long-term anticoagulation therapy in patients with unprovoked VTE
Clinical factors that favor extended anticoagulant therapy
Duration of anticoagulant therapy History of recurrent VTE

Male sex
The risk of recurrent VTE in patients with a VTE secondary Patient had a PE, not just a DVT
to a major transient (reversible) risk factor is low. Therefore, D-dimer on anticoagulant therapy elevated at 3 or 6 months*
time-limited anticoagulation for 3 months with a DOAC or D-dimer elevated after having been off anticoagulants for 4 weeks*
vitamin K antagonist (VKA) is recommended. For patients Anticoagulant therapy well tolerated (with good INR control, if on
with unprovoked proximal leg DVT or unprovoked PE in warfarin)
Little or no impact of anticoagulant therapy on patient’s lifestyle
whom risk factors for bleeding are absent and for whom
Patient’s preference is to continue treatment
good anticoagulation control is achievable, long-term Patient has a known, strong thrombophilia (either congenital or
(extended) anticoagulation therapy (with a DOAC or VKA) acquired)
should be strongly considered. Unselected patients who stop
Factors favoring limited duration of anticoagulation
anticoagulants after some initial (eg, 3-12 month) period of Female sex
treatment for unprovoked VTE have a 3-year recurrence rate Distal DVT only
of 20%-30%. Several parameters can be used by the clinician D-dimer negative after having been off anticoagulation for
in effort to individualize the risk of recurrence (Table 8-1). 4 weeks (this is most relevant to females)
This, along with the patient’s bleeding risk factors and the Occurrence of bleeding complications or significant
patient’s personal preferences, should be used to help the risk for bleeding
patient make an informed decision about whether to Patient’s preference is to be off anticoagulants
continue therapy (Figure 8-1). When extended anticoagula- DVT = deep vein thrombosis; INR = international normalized ratio;
tion therapy is chosen, the risks, benefits, and burdens of PE = pulmonary embolism; VKA = vitamin K antagonist; VTE = venous
long-term anticoagulant therapy should be reevaluated peri- thromboembolism.
odically (eg, once a year). For patients with a first episode of *Cutoff is assay specific (not all D-dimer assays have been studied for
unprovoked distal (ie, below the trifurcation of the popliteal this purpose; clinicians should check with their local laboratory).
vein) leg DVT, 3 months of VKA therapy is recommended.
To aid decision making in which patients to discontinue unprovoked VTE who have access to one of the DOACs
and in which to continue anticoagulation, risk scores have discussed below may be especially good candidates for

been created using data from VTE trials in which the rate of extended anticoagulation since these medications not only
recurrent VTE was recorded and subgroup analyses were are more convenient than VKA but also provide excellent
performed. Male sex is associated with a higher risk of protection against recurrent thrombosis with a compara-
recurrence; in women, the absence of post-thrombotic syn- tively small increase in the risk of major bleeding.
drome, a low D-dimer (measured after stopping anticoagu- Whether thrombophilia testing should be performed to
lation for 4 weeks), and possibly a body mass index of determine the duration of anticoagulant treatment for a patient
<30 kg/m2 predict a lower risk of recurrence. Patients with with unprovoked VTE is discussed later in this chapter.
VTE due to transient risk factor

3 months
Woman, DVT or PE, hormone associated
Woman, unprovoked DVT

Figure 8-1 Management strategy
regarding length of anticoagulation Woman, unprovoked PE
therapy decisions in patients after a first
Man, unprovoked DVT
episode of provoked or unprovoked Long term
proximal VTE. DVT = deep venous Man, unprovoked PE
thrombosis; INR = international
normalized ratio; PE = pulmonary
Other considerations: bleeding, fluctuating INRs, lifestyle impact, patient preference
embolism.

If a patient presents with recurrent VTE despite thera- UFH, LMWH, or fondaparinux. Both the optimal dosing of
peutic anticoagulation, treatment options are either to these drugs (full dose, intermediate dose, or prophylactic
increase the target INR (for patients on VKA), to increase low dose) and the duration of therapy are not well defined;
LMWH dose by 25%, to add another anticoagulant, or to 45 days of prophylactic dose fondaparinux (2.5 mg daily), in
switch to another anticoagulant (particularly from VKA to comparison to placebo, has been shown to reduce the risk of
LMWH). In patients who fail anticoagulation with VKA, DVT extension, but the number needed to treat (to prevent
the clinician should remain vigilant for evidence of cancer, one clinically important event) is large, and the cost-
antiphospholipid syndrome (APS), or an anatomic cause of effectiveness of routinely anticoagulating patients with
thrombosis. superficial thrombophlebitis strategy is debated. Extension
Patients who choose to discontinue anticoagulation but of superficial thrombophlebitis into the deep venous system
are still at some risk for recurrent VTE should be informed occurs in about 1 in 6 patients with extensive superficial
that daily low-dose aspirin is relatively safe and appears to thrombophlebitis and often is present at time of diagnosis.
reduce the likelihood of DVT/PE by 30%-40%. Indeed, for To rule out concomitant DVT or extension, follow-up ultra-
some patients, low-dose aspirin may achieve the best balanc- sonography should be considered in all such patients for
ing of risk, benefit, and cost considerations. whom anticoagulation is not prescribed. The term Trousseau
syndrome often is used for migratory thrombophlebitis in
patients who subsequently are diagnosed with cancer, but
Cancer-associated VTE
the term is not well or uniformly defined.
For most patients treated with a VKA a target INR of 2.0-
3.0 virtually eliminates the risk of recurrent VTE. However, Pediatric considerations
in patients with cancer, the risk of recurrent VTE on VKA is
high, even if the INR is in the 2.0-3.0 range. LMWH has The current management of acute VTE in children is extrap-
been shown to be more effective than VKAs in preventing olated from adult patients. LMWH is the preferred antico-
recurrences in these patients and is the preferred treatment agulant in children for the treatment of VTE because of its
for the first six months after the acute VTE (thereafter, it is predictable pharmacokinetics, lack of interference with diet,
unknown as no studies have compared LMWH with VKA), and easy availability of anti-Xa assays for its monitoring. The
if feasible, from a financial and insurance reimbursement pediatric doses are calculated according to age and weight of
point of view. DOACs are highly effective and very safe for the patient, as both influence the volume of LMWH distri-
extended VTE treatment in patients without cancer, but bution. Young infants (age <2 months) require higher doses
their efficacy and safety have not been established in of LMWH and UFH. Oral VKA therapy is an acceptable
patients with cancer-associated thrombosis. option in children, but its management in young infants
could be challenging for several reasons: (i) Physiologic
reduction of VK dependent coagulant proteins; (ii) excessive
Superficial thrombophlebitis
sensitivity to VKAs in breast fed infants; (iii) resistance
Superficial thrombophlebitis refers to the cephalic, pero- to VKA therapy due to vitamin K intake in infant formula;
neal, posterior tibial, and saphenous veins. It may either be (iv) lack of availability of liquid formulation in many coun-
unprovoked (also called idiopathic) or occur in the setting tries; (v) vascular access issues for INR monitoring. The
of varicose veins, trauma, intravenous catheters or phlebot- experience of using fondaparinux is limited in children, but
omy, underlying hypercoagulable states, cancer, or as septic dosing regimens are available in children older than 1 year of
thrombophlebitis with infections. Superficial vein throm- age. The 2012 American College of Chest Physicians (ACCP)
bosis also occurs in association with inflammatory bowel guidelines provide details on dosing regimens and monitor-
disease, thrombangiitis obliterans (Buerger disease), and ing for specific anticoagulants. The majority of VTE events
Behçet disease. in children are caused by transient risk factors, such as
Superficial thrombophlebitis may or may not require central venous catheters, and these children are treated for
anticoagulant treatment. Thrombophlebitis that is not very 6 weeks to 3 months with anticoagulation. The 2012 ACCP
extensive (ie, <5 cm in length and not close to the deep guidelines suggest that children with unprovoked (idio-
venous system) requires only symptomatic therapy, consist- pathic) VTE receive anticoagulant therapy for 6-12 months.
ing of analgesics, anti-inflammatory medications, and warm Children with inherited higher risk thrombophilias are
or cold compresses for symptom relief. Patients with exten- treated like adult patients and receive long-term anticoagu-
sive or recalcitrant superficial thrombophlebitis may benefit lation depending on their risk of recurrence. DOACs for
from a short course of out-of-hospital anticoagulant ther- treatment of acute VTE are currently being evaluated in
apy, such as up to 6 weeks of subcutaneously administered pediatric populations.

Postthrombotic syndrome for chronic PE. Echocardiography can be used to estimate

pulmonary artery pressure. Right-heart catheterization with
Postthrombotic syndrome (PTS) may be caused by several
pulmonary artery pressure measurements then defines the
factors, including incompetent venous valves damaged by
degree of hypertension, and pulmonary arteriography helps
the thrombus, associated inflammatory mediators, and
define the etiology and allows assessment of whether poten-
impairment of venous return due to residual venous obstruc-
tially curative pulmonary endarterectomy is indicated.
tion from incompletely cleared thrombus. Fewer than 5% of
Long-term anticoagulant therapy is indicated. Pharmaco-
DVT patients will develop severe PTS (sometimes referred to
logic therapy specific for pulmonary hypertension, such as
as postphlebitic syndrome), in most cases within 1-2 years of
bosentan, an endothelin receptor antagonist, can be consid-
the acute DVT. Symptoms and signs include chronic extrem-
ered in the inoperable patient.
ity swelling, pain, heaviness, fatigue, paresthesias, skin indu-
ration, dryness, pruritus, erythema and chronic dark
pigmentation, and, in more severe cases, skin ulcers. The risk Primary prevention of VTE
for developing postthrombotic syndrome had previously
Prophylaxis against VTE should be considered in every
been thought to be decreased by wearing graduated com-
hospitalized patient based on an individual patient’s risk
pression stockings (40 mm at ankle, 30 mm at mid-calf) for
stratification. Detailed prophylaxis guidelines for all types of
2 years after an acute DVT, but a recent blinded, randomized
patients have been published in the medical literature, most
trial showed no impact of this intervention on the risk of
notably the 2012 ACCP guidelines. Formal DVT prophylaxis
PTS. Treatment options for patients with postthrombotic
guidelines should be in use in all hospitals. The most con-
syndrome are limited. Compression stockings should be
vincing evidence of net benefit from VTE prophylaxis comes
worn if they provide symptom relief. In patients with signifi-
from surgical populations.
cant leg swelling, imaging of leg veins with Doppler ultra-
Mechanical methods of prophylaxis with graduated com-
sound and of the pelvic veins with CT or MR venography
pression stockings or intermittent pneumatic compression
can be considered to evaluate for focal pelvic vein obstruc-
devices typically are recommended for patients who are at
tion or narrowing due to May-Thurner syndrome or post-
high risk for bleeding or as an adjunct to anticoagulant-based
thrombotic scarring. Although these sorts of anatomic
prophylaxis. They often are not suggested as a first choice for
abnormalities might be amenable to pelvic vein angioplasty
primary prevention because they have been studied less
and stenting, there is no high-quality evidence to support the
intensely than anticoagulant-based methods. Although there
efficacy or safety of such interventions in this challenging
is some evidence that aspirin and other antiplatelet agents
clinical situation. Finally, a home compression pump with
provide some protection against VTE in hospitalized patients
compression sleeve for the affected leg can be considered for
at risk, they are probably less effective than other pharmaco-
patients with significant symptoms. Similar to adults, almost
logic methods of VTE prophylaxis.
20%-30% of children are diagnosed with postthrombotic
If anticoagulation for VTE prophylaxis is appropriate,
syndrome. Management is extrapolated from adults.
several options are available: (i) UFH at every 8- or 12-hour
dosing intervals; (ii) LMWHs at once- or twice-daily inter-
vals; (iii) fondaparinux once daily; (iv) VKAs; or, in patients
Pulmonary hypertension
undergoing total knee or hip replacement, apixaban or riva-
Pulmonary hypertension due to VTE, termed chronic throm- roxaban can be used. FDA-approved indications vary
boembolic pulmonary hypertension (CTEPH), is defined as an between the different pharmacological options. Prophy-
elevated mean pulmonary artery pressure of >25 mm Hg laxis may be given only during the hospitalization or, if the
and occurs in 1% of patients with acute PE after 6 months. VTE risk persists after discharge home, for an extended
One study has reported CTEPH in 3.8% of PE patients after period of time. The net benefit and cost effectiveness of
2 years; the authors’ experience suggests that clinically appar- postdischarge prophylaxis (up to 5 weeks) are well estab-
ent CTEPH may be less common. CTEPH can be the result lished in patients after hip fracture, hip replacement, and
of a single episode of PE that did not resolve appropriately or major cancer surgery.
the result of recurrent episodes of PE. The patient who expe-
riences chronic shortness of breath or significant generalized
Pediatric consideration
malaise after PE should be evaluated for pulmonary hyper-
tension. A formal walk test with pre- and postexercise pulse There is growing concern about rising prevalence of VTE in
oximetry measurements is appropriate. It is important to hospitalized children, but the role of pharmacological throm-
realize that chest CT angiogram findings may be minimal boprophylaxis in preventing hospital acquired CVC-related
with chronic distal PE. A VQ scan is probably more sensitive thrombosis is controversial. Anticoagulation prophylaxis

with oral VKA therapy is currently recommended for chil- Unusual site venous thromboses
dren receiving long-term home total parental nutrition
Upper-extremity DVT (and catheter-related
(TPN); the strength of this recommendation is limited by the
thrombosis)
lack of high-quality evidence.
The superficial veins of the arm include the antecubital,
cephalic, and basilic veins. The deep venous system includes
Inferior vena cava filters
the brachial vein, which becomes the axillary vein, followed
A clear indication for an inferior vena cava (IVC) filter exists by the subclavian and brachiocephalic veins, and finally the
only when a patient has acute proximal leg vein thrombosis superior vena cava. Upper extremity DVTs make up 1%-4%
or PE and cannot be anticoagulated. It is not clear whether of all DVT in adults (but a higher proportion of pediatric
an IVC filter is beneficial in the patient with recurrent pelvic DVT). Roughly 80% are secondary to central venous cathe-
or proximal leg DVT despite therapeutic anticoagulation. ters and cancer, and 20% are primary events; however, these
The 2012 ACCP guidelines suggest that IVC filters not be data depend largely on what patient population is studied.
used for primary VTE prevention in patients with trauma or Doppler ultrasound (sensitivity 78%-100% and specificity
undergoing major abdominal or pelvic surgery. When an 82%-100%), contrast venography (gold standard), and MR
IVC filter must be placed, a retrievable filter should be con- venography are the diagnostic tools used to diagnose upper-
sidered. Retrievable filters can be left in place for weeks to extremity thrombosis. The risk of PE with upper extremity
months or can remain permanently, if necessary. For patients DVT is not well defined, but seems to be low (especially if the
with acute VTE who have an IVC filter inserted as an alter- clot is catheter-associated). Postthrombotic syndrome is
native to anticoagulation, anticoagulant therapy should be common; residual thrombosis and axillo-subclavian vein
initiated or resumed as soon as the patient’s risk of bleeding thrombosis appear to be associated with a higher risk of
permits. The presence of an IVC filter increases a patient’s upper extremity PTS, whereas catheter-associated DVT may
risk for recurrent DVT and confers a risk of vena caval be associated with a lower risk.
thrombosis. When making a decision on the length of anti- Management for DVT of the upper-extremity consists of
coagulant therapy in a patient with a permanent IVC filter, the following: (i) LMWH, UFH, or fondaparinux in the
the presence of the IVC filter should be viewed as a risk fac- acute setting; (ii) continued anticoagulation (vitamin K
tor for recurrent VTE. antagonist [VKA] or LMWH if a cancer patient) for at least
3 months for unprovoked DVT or catheter-associated DVT;
and (iii) no catheter removal in patients with DVT associ-
Venous stents
ated with a central venous catheter if the catheter is func-
Although there are no clinical trials to determine their effi- tional and still needed. A DOAC would probably be effective
cacy, venous stents are sometimes placed in various locations in this setting but, since these agents have not been studied in
of the venous system, most commonly into the left common catheter-associated thrombosis, it may be prudent to use
iliac vein due to May-Thurner syndrome, the right and left VKA or LMWH pending further evidence. Decisions about
pelvic veins due to postthrombotic vessel narrowing and scar- duration of therapy for upper-extremity DVT usually are
ring, and the superior vena cava and central arm veins in cen- based on information extrapolated from studies of patients
tral venous catheter-associated strictures. The best long-term with lower-extremity DVT or PE. For line-associated DVT,
management of patients who have venous stents is not known a brief (4-12 weeks) period of anticoagulation after line
due to a lack of quality prospective studies examining their removal is likely sufficient. There is little or no direct evi-
long-term patency with and without antiplatelet drugs or dence to support any particular duration of anticoagulant
anticoagulants. Because stents are foreign bodies in the venous therapy after a first unprovoked (or catheter-associated)
system and may lead to flow disturbances, it is possible that upper extremity DVT.
they have some prothrombotic risk. In addition, endothelial Upper-extremity DVTs may be due to thoracic outlet
cell proliferation within stents is known to occur, potentially syndrome, also referred to as effort thrombosis, thoracic out-
leading to stent stenosis and occlusion. In view of the limited let syndrome, or Paget-Schroetter syndrome. This is due to
data available, it may be best to view the presence of a venous compression of the axillary vein by pressure from the clavi-
stent as a potential risk factor for recurrent VTE. After venous cle, an extra rib, or enlarged or aberrantly inserted muscles,
stent placement, it may be reasonable to keep a patient on often provoked or potentiated by abduction of the arm and
anticoagulants for 3 months and then make an assessment on repetitive arm movements. There is no uniform approach to
the need for long-term anticoagulation versus no further anti- treatment of these patients. Management options include
coagulation based on a comprehensive assessment of the anticoagulation, thrombolytic therapy, angioplasty with or
patient’s risk factors for recurrent VTE and bleeding. without stent placement, thoracic outlet surgery with rib or

soft tissue resection, and surgical resection of the focally nar- It occurs in up to 26% of patients with cirrhosis of the liver.
rowed vein with vein reconstruction. Individual treatment As with other venous thromboembolic disorders, multiple
decisions need to be made, and a team approach that includes contributors often are identified. In a number of cases, no
vascular medicine, vascular surgery, and interventional predisposing factor is found. Diagnosis typically is made by
radiology may be appropriate. Doppler ultrasonography. CT or MR venography also can
provide evidence that portal vein thrombosis is present.
Cavernous transformation of the portal vein reflects old por-
Hepatic vein thrombosis
tal vein thrombosis, as do collaterals in the porta hepatis.
Hepatic vein thrombosis, also referred to as Budd-Chiari In the patient with acute portal vein thrombosis, extension
syndrome, has varied clinical presentations, ranging from of thrombus into the mesenteric veins may occur and lead to
asymptomatic to fulminant liver failure. A cause can be intestinal infarction and the need for surgical bowel resec-
identified in approximately 84% of patients. Similar to other tion. The patient with acute portal vein thrombosis typically
venous thromboembolic disorders, Budd-Chiari syndrome is anticoagulated for at least 3-6 months to prevent progres-
also often has a multifactorial etiology. Most patients (84%) sion of thrombosis. Regarding long-term anticoagulation
have at least one thrombotic risk factor, and many (46%) therapy in these patients, as well as in patients with inciden-
have more than one risk factor; the most common are tally discovered portal vein thrombosis, the risk of bleeding
myeloproliferative neoplasms (MPNs) (49% of patients). has to be balanced individually against the risk of re-throm-
Polycythemia vera accounts for 27% percent of cases; essen- bosis. The net benefit of anticoagulation for a patient with
tial thrombocytosis (ET) and primary myelofibrosis are less asymptomatic, cirrhosis-associated portal vein thrombosis is
prevalent causes. The JAK2 mutation is present frequently in uncertain. Patients with cirrhosis-associated portal vein
patients with the syndrome (29% of cases), even if no thrombosis appear to be at high risk of both bleeding and
hematologic abnormalities suggestive of an MPN are pres- thrombotic events. Thus, management has to be tailored to
ent. This is discussed in detail in the “Thrombophilias” sec- the individual patient. The factors to be considered before
tion in this chapter. Any of the inherited and acquired long-term anticoagulation is prescribed include identifica-
thrombophilias can contribute to the development of Budd- tion of the triggering factor for the thrombotic event, the
Chiari syndrome, as can estrogens and pregnancy. Paroxys- extent of thrombosis, the presence of persistent prothrom-
mal nocturnal hemoglobinuria (PNH), although an botic factors, the extent of esophageal and gastric varices, the
uncommon disorder, can be detected in almost one-fifth of presence and degree of thrombocytopenia due to hyper-
patients with Budd-Chiari syndrome. splenism, and of the risk of bleeding.
The diagnosis is made by Doppler ultrasonography,
contrast-enhanced CT scanning, or magnetic resonance
Mesenteric vein thrombosis
imaging (MRI). In the acute setting of fulminant thrombo-
sis, thrombolytic therapy can be considered. Angioplasty of Venous drainage of the intestine is via the superior mesen-
narrowed or occluded hepatic veins can be performed, shunt teric vein (SMV) and inferior mesenteric vein (IMV) into
procedures may be required, and liver transplantation may the portal vein. The SMV drains the small intestine and
be necessary. Anticoagulation is usually appropriate and ascending colon, whereas the IMV drains mostly the sigmoid
often is given long term, typically with VKAs. INR monitor- colon. The transverse and descending colon can drain
ing may be problematic, however, because liver synthetic through the middle and left colic veins either into the SMV
dysfunction may lead to a baseline elevation of INR even or IMV. SMV thrombosis, if diagnosed late, leads mostly to
before VKA therapy. Alternative monitoring tests for VKAs, small bowel ischemic changes. The very rare IMV thrombo-
such as factor II or X activity, may have to be used. Also, sis may lead to ischemia in the sigmoid colon. Mesenteric
treatment with LMWH or fondaparinux instead of VKAs vein thrombosis may be caused by trauma, surgery, intra-
can be considered. abdominal infections, inflammatory bowel disease, pancre-
atic disease, and progression of portal vein thrombosis, but
also may occur spontaneously, particularly in patients with
Portal vein thrombosis
inherited or acquired thrombophilias, MPNs, presence of the
Portal vein thrombosis often is silent and may be discovered JAK2 V617F mutation, and PNH. Symptoms are vague, often
only upon evaluation of a variceal gastrointestinal bleed. It is leading to a delay in diagnosis. Nonspecific abdominal pain
associated with the inherited and acquired thrombophilias, is common, and nausea may be present. Gastrointestinal
the MPNs, JAK2-positive status without overt MPN, PNH, bleeding and peritonitis are seen when transmural ischemia
intra-abdominal neoplasia or inflammation, infection, has occurred. Symptoms may be present for days to weeks
trauma, surgery, and neonatal umbilical vein catheterization. before a diagnosis is made, which often may occur only when

the patient presents as a surgical emergency with ischemic and perinatal complications are contributors. In adults, the
bowel. The principal cause of a high mortality rate in mesen- most frequent but least specific symptom is severe headache,
teric vein thrombosis is a delay in diagnosis. The surgical either of subacute or acute onset, present in 90% of patients.
findings are typically those of a dusky but not frankly gan- In children, seizures, focal neurological signs, and headache
grenous intestine, unless full bowel wall infarction already are the most common manifestations. Routine noncontrast
has occurred. Areas of viability of intestine are not as sharply and contrast head CT scans and brain MRI scan often are
demarcated as they are in arterial mesenteric ischemic dis- unrevealing, resulting in missed diagnoses, unless CT veno-
ease. A mesenteric artery pulse is typically felt. Preoperative gram or MR venogram are requested specifically.
diagnosis is made by CT angiography. Doppler ultrasound Approximately 40% of patients with cerebral and sinus
may be diagnostic, but it is operator dependent and may vein thrombosis have a hemorrhagic infarct, which is a con-
have limited sensitivity in the obese patient. Once diagnosed, sequence of the venous occlusion. Currently, no available
patients are managed with anticoagulation alone or in com- evidence from randomized controlled trials exists regarding
bination with surgical intervention. Most patients improve. the efficacy or safety of systemic or local thrombolytic ther-
Decisions on length of VKA therapy depend, as with most of apy in cerebral and sinus vein thrombosis. Heparin is often
the other VTE disorders, on the triggers for the thrombotic used in acute cerebral and sinus vein thrombosis, even if
episode and the presence of thrombophilias or other perma- some parenchymal hemorrhage is present. The 2012 ACCP
nent risk factors. Length of treatment is at least 3 months but guidelines do not make any reference to anticoagulant man-
may have to be long-term. The role of anticoagulation and/ agement of patients with cerebral and sinus vein thrombosis,
or antiplatelet therapy in the long-term secondary preven- but the 2010 European Federation of Neurological Societies
tion of mesenteric thrombosis in a patient with a confirmed recommends LMWH or UFH therapy in the acute treat-
MPN is not well established. ment. The optimal duration of anticoagulant therapy is
unknown. After a first episode of cerebral and sinus vein
thrombosis, expert guidelines recommend anticoagulation
Splenic vein thrombosis
for: (i) 3 months if the thrombosis was associated with a
Because of the intimate anatomic contact of the splenic vein transient risk factor, (ii) 6-12 months if the event was unex-
with the pancreas, the main causes of splenic vein thrombo- plained and no high-risk thrombophilia has been detected,
sis are pancreatitis and pancreatic malignancies. Similar to and (iii) long-term if a high-risk thrombophilia is detected
mesenteric vein thrombosis, intra-abdominal problems or the event is recurrent.
(infection, surgery, and trauma) and thrombophilias also In children, use of anticoagulant therapy with cerebral
play a role in the etiology. Symptoms often are subtle, and and sinus vein thrombosis has not been well studied in clini-
the diagnosis is not infrequently a coincidental discovery on cal trials, but it is not associated with serious hemorrhage in
abdominal imaging studies done for other reasons. Length of selected patients.
anticoagulant treatment depends on the triggering factors
and the persistent thrombophilic risk factors.
Renal vein thrombosis
In adults, the classical symptom triad of acute renal vein

Cerebral and sinus vein thrombosis
thrombosis, namely, acute flank pain, hematuria, and sudden
Blood from the brain drains via cerebral and cortical veins deterioration of renal function, is uncommon. More com-
into the dural sinuses that then drain into the internal jugu- mon is a chronic course with subtle worsening of renal func-
lar veins. Thrombosis of the cerebral, cortical, and sinus tion, progressive proteinuria, and edema, often without pain
veins often is referred to as cerebral and sinus vein thrombosis. or hematuria. As many as 30%-50% of patients with chronic
It is seen much more commonly in the neonatal period than nephrotic syndrome have evidence of renal vein thrombosis,
in any other age group. It occurs in 1-2 cases per 100,000 in and it is not uncommonly bilateral and often protrudes into
the general population, in up to 7 cases per 1 million in chil- the IVC. Nephrotic syndrome leads to hypercoagulability,
dren, and in approximately 120 cases in the peri- and post- which may be the result of urinary AT loss, free protein S
partum period per 1 million deliveries. The clinical outcome deficiency secondary to an increase in C4b-BP, or unknown
in adults is good in most cases. As with other venous throm- causes. Diagnosis is made by Doppler ultrasound or by
boembolic events, cerebral and sinus vein thrombosis often MR venography. Thrombolytic therapy should be consid-
is multifactorial, and the inherited and acquired thrombo- ered in case of acute thrombosis, particularly if there is bilat-
philias play a role in its etiology, as do estrogen therapy and eral disease or impending renal failure. Anticoagulation
pregnancy. Infections such as mastoiditis, otitis, sinusitis, therapy is indicated. The length of anticoagulant therapy
and meningitis are risk factors, and in neonates dehydration depends upon whether the thrombotic event was associated

with a transient prothrombotic risk factor or whether the thrombotic event occurs in, a number of risk factors and
patient has permanent risk factors or a higher risk thrombo- associated disorders should be investigated, to clarify the eti-
philia. Renal vein thrombosis in children typically presents ology of the event (Table 8-2). As for specific arterial territo-
within the first month of life. In neonates, it may be associ- ries, in the case of upper-extremity arterial thromboembolism,
ated with acute dehydration, perinatal asphyxia, and polycy- thoracic outlet syndrome should be considered; in lower-
themia. Infants of diabetic mothers are particularly at risk extremity claudication or arterial thromboembolism, popli-
for renal vein thrombosis. Outside of the neonatal period, teal artery entrapment syndrome, cystic adventitial disease of
nephrotic syndrome is the most common risk factor for the popliteal artery, fibromuscular dysplasia of the lower-
thrombosis in children as well as in adults. extremity arteries, and endofibrosis of the iliac artery should
be considered; and in the case of stroke, s pontaneous or trau-
matic cervical artery dissection should be considered.
Retinal vein thrombosis
Relatively little is known about thrombophilias predispos-
Thrombosis can occur as central retinal vein occlusion ing to arterial thrombosis. Arterial thrombosis is a classifying
(CRVO) or as branch retinal vein occlusion (BRVO). CRVO clinical criterion for antiphospholipid syndrome (APS).
has a prevalence of 1 in 250 to 1,000 in individuals ≥ 40 years Whether young patients with otherwise unexplained arterial
of age. The presence of cardiovascular risk factors, such as thromboembolic events and an inherited “strong” thrombo-
hypertension, hyperlipidemia, and especially diabetes, has philia (such as protein C, protein S, or AT deficiency), would
been associated with RVO. An association with inherited or benefit from taking an anticoagulant (in addition to or
acquired thrombophilia has not been convincingly demon- instead of antiplatelet therapy) is not known. This, along
strated. Unfortunately, there is very little high-quality evi- with the lack of high-quality evidence that these inherited
dence on which to judge the utility of antiplatelet or thrombophilias are linked to arterial thrombosis, leads many
anticoagulant therapy for CRVO. One small (67-patient) clinicians to avoid searching for inherited thrombophilia in
randomized trial indicates that 90 days of LMWH treatment patients with arterial events.
may be more effective than aspirin for the prevention of
visual loss in patients with RVO; however, the optimal dura- Pediatric considerations
tion of anticoagulant therapy is not known because long-
term comparisons of anticoagulant vs. antiplatelet vs. no Indwelling arterial catheters are widely used in neonatal and
therapy have not been performed. pediatric intensive care units for hemodynamic monitoring
and blood sampling purposes. Insertion through the umbili-
cal artery often is used for preterm and term neonates,
Arterial thromboembolism whereas in infants and older children, the preferred site is the
radial artery. Alternative insertion sites include the ulnar,
General comments
brachial, axillary, dorsalis pedis, and tibialis posterior arter-
The hematologist typically does not get called upon for the ies. The most frequent complication is development of arte-
management of patients with ischemic disease that is due to rial thrombosis. The overall incidence of indwelling arterial
arteriosclerosis. Therefore, this chapter does not discuss the catheter related thrombosis is approximately 3.2%; 90% of
pathophysiology of arteriosclerosis and its role in arterial these events occur in the iliofemoral artery, usually after car-
occlusive disease or the management approaches aimed at diac catheterization in a child with congenital heart disease.
modifying an individual’s arteriosclerosis risk factors, such The most serious complications of arterial thrombosis
as weight reduction, cessation of smoking, increased physical include loss of limb, organ, or life. Close clinical monitoring
activity, and treatment of diabetes mellitus, hypertension, for skin discoloration, loss of pulses distal to the catheter,
and hyperlipidemia. References to the major treatment and organ dysfunction is required. Depending on the sever-
guidelines are listed, for the interested reader, in the Bibliog- ity of the complications, management ranges from removal
raphy. Considerations unique to a person who is young, has of arterial catheters to the use of thrombolytic therapy fol-
no significant arteriosclerosis risk factors, or has a personal lowed by standard anticoagulation. Neonates with umbilical
or family history of thrombophilia will be discussed below. artery catheters should be monitored for extension of throm-
bosis to renal vessels.
Arterial thrombosis in the absence of arteriosclerosis

Atrial fibrillation and stroke prevention
Arterial thromboembolic events in the young person
(<50 years of age) are rare, unless significant arteriosclerosis The hematologist is occasionally asked about the risk-benefit
risk factors are present. No matter which territory the arterial tradeoffs associated with anticoagulation in a patient with

Table 8-2 “Unexplained” arterial thromboembolism: suggested approach to structured evaluation
A. Is arteriosclerosis the underlying problem?

• Arteriosclerotic changes demonstrated on imaging studies or pathology specimens?
•• Arteriosclerosis risk factors present?
 Cigarette smoking
 High blood pressure
 High low-density lipoprotein (LDL) cholesterol
 Low high-density lipoprotein (HDL) cholesterol
 High lipoprotein(a)
 Diabetes mellitus
 Obesity
 Family history of arterial problems in young relatives (< 50 years of age)
B. Has the heart been thoroughly evaluated as an embolic source?

•• Atrial fibrillation – EKG, Holter, or event monitor
•• Patent foramen ovale – obtain cardiac echo: transthoracic echo with bubble study and Valsalva maneuver; if negative, consider
transesophageal echo with bubble study
C. Other causes
•• Is the patient on estrogen therapy (contraceptive pill, ring, or patch; hormone replacement therapy)?
•• Does the patient use cocaine or anabolic steroids?
•• Is there evidence for Buerger’s disease (does patient smoke tobacco or use cannabis)?
•• Does patient have symptoms suggestive of a vasospastic disorder (Raynaud’s)?
•• Were anatomic abnormalities seen in artery leading to the ischemic area (web, fibromuscular dysplasia, dissection, vasculitis, external
compression)?
•• Does patient have evidence of a rheumatologic or autoimmune disease (arthritis, purpura, or vasculitis)? Consider laboratory workup
for vasculitis and immune disorder.
•• Is there a suggestion of an infectious arteritis?
•• Could the patient have hyperviscosity or cryoglobulins?
D. Thrombophilia workup for arterial events
•• Hemoglobin and platelet count
•• Antiphospholipid antibodies
 Anticardiolipin IgG and IgM antibodies
 Anti–b -glycoprotein-I IgG and IgM antibodies

2
 Lupus anticoagulant
•• Flow cytometry to exclude PNH (if any evidence of hemolysis or cytopenias are present)
•• Homocysteine
•• Do not test for MTHFR polymorphisms, PAI-1 or tPA levels or polymorphisms, fibrinogen or factor VIII activities.
•• A benefit of testing for FVL mutation, prothrombin gene mutation, protein C/S activity and AT activity is not established
EKG = electrocardiogram; PAI-1 = plasminogen activator inhibitor-1; tPA = tissue plasminogen activator.
* Uncertain clinical utility.
nonvalvular atrial fibrillation. Detailed information relevant is the most common clinical presentation. Many infants do
to this clinical decision can be found elsewhere (Table 8-3), not present for several months, however, until they are noted
but—for most patients with AF—the risk of anticoagulation to have hemiparesis or early hand preference. It is often dif-
with either a DOAC or a VKA will be outweighed by the ficult to determine whether the stroke occurred in utero, at
benefit. The rare exceptions will be patients who are either the time of delivery, or within the first week. Most neonatal
at very low risk of stroke or at exceptionally high risk of stroke occurs in the distribution of the left-middle cerebral
anticoagulation-related major bleeding. artery. MRI and angiography is the best test to determine
extent of disease. There is no standard approach for the eval-
uation and treatment of perinatal stroke. At the time of diag-
Neonatal stroke
nosis, though, it is important to determine whether the
Neonatal stroke, defined as a cerebrovascular event that thrombotic event was related to an underlying disorder, such
occurs between 28 weeks gestation and 7 days of age, occurs as congenital heart disease or so-called TORCH (toxoplas-
in 1 in 250 live births. There is a male predominance. Seizure mosis, syphilis, herpes, cytomegalovirus) infections, which

Pathophysiology of thrombosis | 195
Table 8-3 Key resources for use of antithrombotic drugs in arteriosclerotic occlusive arterial disease, atrial fibrillation and valvular heart disease
Antithrombotic therapy guidelines

Disease/condition ACCP (Chest February 2012) AHA/ACC
Peripheral arterial disease Alonso-Coello P et al. Smith SC et al., 2011
TIA and stroke Lansberg MG et al. Furie KL et al., 2011
Coronary artery disease Vandvik PO et al. Smith SC et al., 2011
Myocardial infarction Vandvik PO et al. Wright RS et al., 2011
Atrial fibrillation You J et al. Fuster V et al., 2011
Valvular and other heart disease Whitlock R et al. Bonow RO et al., 2008
ACCP = American College of Chest Physicians; AHA/ACC = American Heart Association/American College of Cardiology; TIA = transient
ischemic attack.
are passed in utero from the mother to the developing fetus, (eg, congenital homozygous protein C deficiency or
systemic bacterial infections, or metabolic diseases. Mater- antiphospholipid syndrome [APS]) appears to benefit
nal drugs and medical conditions, placental disorders, peri- from anticoagulation. Warfarin, UFH, or LMWH has been
natal asphyxia, and birth trauma all have been associated successful in treating and preventing recurrence of acute
with neonatal cerebrovascular events. Several studies have stroke in children with these underlying disorders. For
demonstrated an association between inherited prothrom- strokes of other etiologies, anticoagulation does not
botic conditions and neonatal stroke. The incidence of improve outcome better than treatment with antiplatelet
recurrent stroke is extremely low (<5%); therefore, antico- agents. Although the use of thrombolytic agents within 3
agulation is not indicated unless there is evidence of embolic hours of initiation of the signs of stroke can be successful
heart disease. Children with a cardioembolic cause of stroke in improving outcomes in adults, the safety and efficacy of
should be referred to a pediatric cardiologist for evaluation, this strategy in children has not been demonstrated.
management, and correction of the heart defect. Fifty per-
cent of infants with perinatal events will be neurologically
normal by 12-18 months of age. Long-term sequelae, such as Pathophysiology of thrombosis
mild hemiparesis, speech or learning problems, behavioral
problems, and seizures, are more likely to persist in patients Thrombosis, defined as excessive clotting, has three main
who present outside the newborn period. causes, referred to as Virchow’s triad: reduced blood flow
(stasis), blood hypercoagulability, and vascular wall abnor-
malities. Under normal circumstances, if blood vessel integ-
Childhood stroke
rity is interrupted, coagulation takes place and a blood clot
Stroke affects 1 in 7,000 to 35,000 children per year, with a forms to prevent excessive bleeding. On the other hand,
male predominance. As many as 65% of affected children blood in the intact vasculature is kept in a fluid state by mul-
will have lifelong disabilities, such as neurological defects tiple endogenous anticoagulant factors. These natural anti-
and seizures, and the risk of a second stroke is 20%. Despite coagulants, such as antithrombin (AT), protein C, and
therapy, mortality rates as high as 10% have been reported. protein S, prevent excess thrombin formation. Once a throm-
Certain comorbid conditions are highly correlated with bus has formed, its growth is limited by clot lysis, which
stroke, including congenital or acquired cardiac disease eventually leads to thrombus resolution.
that can cause embolic phenomena and sickle cell anemia. Venous thromboembolism (VTE) is believed to be
Numerous systemic disorders can contribute to stroke, and multicausal, with more than one factor (genetic or
inherited prothrombotic states are causative as well. A rare environmental) needed for thrombosis to occur. A patho-
cause of stroke in childhood is severe iron deficiency ane- physiologic model suggests that each individual has a base-
mia. Fifty percent of children with stroke will have no iden- line (or background) thrombosis risk that increases with age
tifiable disorder. There is limited published information (Figure 8-2A). Transient risk factors, such as major surgery
about the etiology and outcome of childhood stroke and or estrogen therapy, temporarily increase a person’s throm-
very little evidence in the literature about appropriate bosis risk, but the threshold of thrombosis formation often
management and prevention approaches. Embolic stroke is not reached (Figure 8-2B). Most people, therefore, never
resulting from cardiac disease or carotid dissection and develop symptomatic VTE. However, the individual with
stroke associated with severe prothrombotic conditions a higher baseline thrombosis risk, such as the individual

A
Arterial clots usually form in areas of atherosclerotic vas-
cular damage. The events leading to atherosclerosis, mainly
VTE lipid disturbances, oxidative stress, and inflammation,
Thrombosis potential
Persons with have been relatively well studied. The composition and vul-
congenital
Thrombosis threshold thrombophilia
nerability of plaque rather than the severity of stenosis are
the most important determinants for the development of
arterial ischemic syndromes. Disruption of the fibrous cap
General or endothelium overlying an atheromatous plaque exposes
population collagen and tissue factor to the circulating blood, leading
to platelet adhesion and aggregation and local thrombin
formation, with subsequent partial or complete vessel

80 years
Age
occlusion.
B
VTE
Persons with
congenital
Thrombophilias
Thrombosis potential
thrombophilia
The terms thrombophilia and hypercoagulable state refer to
hereditary or acquired predispositions to develop thrombo-
sis. Although the role of testing for these conditions has
General diminished somewhat in recent years, the hematologist
population
must be familiar with the nature, limitations, and interpre-
tation of such testing. It is important to note that first-
degree relatives of patients who have experienced VTE
Contraceptives, 80 years
major surgery, etc.
(provoked less so than unprovoked) are at an increased risk
Age
of venous thrombosis, irrespective of thrombophilia test
results. When assessing risk, selective testing in families
Figure 8-2 Threshold model of thrombosis risk. VTE = venous with a strong history of VTE and, consequently, cosegrega-
thromboembolism. Modified from Rosendaal FR. Lancet. 1999;353: tion of known and unknown genes in the early days of
1167-1173. thrombophilia research has resulted in a perceived stronger
risk increase than more contemporary studies have estab-
lished. This is particularly true for antithrombin, protein C,
with a known or unknown inherited or acquired intrinsic and protein S deficiency. Table 8-4 lists the prevalence and
predisposition to clotting (thrombophilia), may cross the association with various clinical manifestations. Table 8-5
thrombosis threshold while exposed to a transient risk factor lists the risk of a first VTE in asymptomatic first-degree
and, thus, present with symptomatic VTE (Figure 8-2B). relatives of patients with VTE.
In general, venous thrombosis is caused by disturbances
in the plasma coagulation system with platelet participa-
tion playing a minor role, whereas in arterial thrombosis Inherited thrombophilias
platelets play the predominant role, with some participa-
Family history of VTE
tion of the plasma coagulation system. This paradigm helps
explain why coagulation protein abnormalities, such as fac- Simply having a family history of VTE is a risk factor for
tor V Leiden (FVL), the prothrombin 20210 mutation, and first-time VTE, no matter whether or not a thrombophilia is
deficiencies of protein C, protein S, and AT are associated detectable in the family. This additional risk is due to
with an increased risk of VTE but have not been linked unknown or unmeasured risk factors. Having one first-
consistently to a higher likelihood of arterial events, such as degree relative with a history of VTE increases an individu-
myocardial infarction or stroke. Thrombus formation in al’s risk of VTE two- to fourfold. Young age of the affected
the cardiac ventricles and atria often is caused by stagnant relative, an unprovoked clot in the affected relative, and having
blood flow in dyskinetic, or aneurysmal parts of the heart >1 affected first-degree relative all increase the likelihood to
chambers or in fibrillating atria. These intracardiac thrombi develop a first VTE. Whether a strong family history
arise in a low-flow environment and are thus pathophysio- of VTE is a risk factor for recurrent VTE, and thus should
logically thought to be similar to the thrombi that lead to be used in decision making on length of anticoagulation
venous thrombosis. therapy after a first episode of VTE, is not known.

Thrombophilias | 197
Table 8-4 Prevalence of thrombophilia and relative risk estimates for various clinical manifestations
Factor V Leiden Prothrombin

Antithrombin Protein C Protein S mutation 20210A mutation
deficiency deficiency deficiency (heterozygote) (heterozygote)
Prevalence in the general population 0.02% 0.2% 0.03-0.13% 3-7% 0.7-4%
Relative risk for a first venous thrombosis 5-10 4-6.5 1-10 3-5 2-3
Relative risk for recurrent venous thrombosis 1.9-2.6 1.4-1.8 1.0-1.4 1.4 1.4
Relative risk for arterial thrombosis No association No consistent No consistent 1.3 0.9
association association
Relative risk for pregnancy complications 1.3-3.6 1.3-3.6 1.3-3.6 1.0-2.6 0.9-1.3
Factor V Leiden suggests that it has led to evolutionary advantages, perhaps

including protection against massive postpartum hemor-
General information
rhage, increased fecundity, and increased male sperm
Activated protein C (APC) is a potent inhibitor of the coag- count.
ulation system, cleaving the activated forms of factor V and
VIII (FVa and FVIIIa) (Figure 8-3A and B). The factor
Prevalence
V Leiden mutation (FVL), one of the most commonly
identified inherited thrombophilias, is a point mutation The prevalence of heterozygous FVL is 3%-8% in Caucasian
(G1691A) in the factor V gene, leading to a factor V mole- populations and 1.2% in African Americans. It rarely is
cule with an arginine-to-glutamine substitution at position found in native African and Asian populations. Homozygous
506 (Arg506Gln, R506Q). This abolishes a cleavage site for FVL occurs in 1 in 500 to 1,600 Caucasians.
APC and makes factor Va less susceptible to inactivation
(Figure 8-3C). Based on the initial observation that APC
Laboratory aspects
did not appropriately prolong the activated partial throm-
boplastin time (aPTT) in a dose-dependent fashion, this The diagnosis of FVL is made by genetic testing (ie, poly-
defect was termed activated protein C resistance (APC resis- merase chain reaction [PCR]); some laboratories screen for
tance). FVL accounts for >90% of APC resistance. Other FVL with an APC resistance assay). The currently used
causes of APC resistance include less common genetic second-generation APC resistance assays, which are aPTT-
mutations of factor V (factor V Cambridge, factor V Liver- based coagulation assays using factor V–deficient plasma, are
pool) and acquired causes of APC resistance, including very sensitive and relatively specific for the detection of the
antiphospholipid antibodies (APLA), pregnancy, and can- FVL mutation. An abnormal APC resistance test result, how-
cer. FVL is inherited in an autosomal-dominant fashion. ever, may be due to causes other than FVL and, therefore,
The high prevalence of FVL in the general population should be followed by the genetic FVL test.
Table 8-5 Estimated incidence of a first episode of VTE in carriers of various thrombophilias (data apply to individuals who have at least one
symptomatic, first-degree relative)
Antithrombin, Factor V Prothrombin Factor V
protein C, or protein S Leiden, 20210A Leiden,
deficiency heterozygous mutation homozygous
Overall (%/year, 95%CI) 1.5 (0.7-2.8) 0.5 (0.1-1.3) 0.4 (0.1-1.1) 1.8 (0.1-4.0)*
Surgery, trauma, or immobilization (%/episode, 95%CI)† 8.1 (4.5-13.2) 1.8 (0.7-4.0) 1.6 (0.5-3.8) –
Pregnancy (%/pregnancy, 95%CI) (includes postpartum) 4.1 (1.7-8.3) 2.1 (0.7-4.9) 2.3 (0.8-5.3) 16.3‡
During pregnancy, %, 95%CI 1.2 (0.3-4.2) 0.4 (0.1-2.4) 0.5 (0.1-2.6) 7.0‡
Postpartum period, %, 95%CI 3.0 (1.3-6.7) 1.7 (0.7-4.3) 1.9 (0.7-4.7) 9.3‡
Oral contraceptive use (%/year of use, 95%CI) 4.3 (1.4-9.7) 0.5 (0.1-1.4) 0.2 (0.0-0.9) –
* Based on pooled OR of 18 (8-40) and an incidence of 0.1% in noncarriers.
† These risk estimates mostly reflect the situation before thrombosis prophylaxis was routinely used.
‡ Data from family studies, risk estimates lower in other settings.

A B C
XII VII Normal Factor V
factor V Leiden
X Protein S
Va Va
Va Act. protein C
506 Act. protein C 506 506 Act. protein C 506
Thrombin Antithrombin
Fibrin clot
Figure 8-3 (A) Sites of action of the natural anticoagulants; (B) method of inactivation of factor V; (C) demonstration of the inability of
activated (Act.) protein C to inactivate factor Va when the factor V Leiden mutation is present.
females who may consider avoiding some oral contraceptives

Risk for thrombosis
in case they are heterozygous for the FVL mutation; further-
Heterozygosity for FVL is mildly thrombophilic, leading to a more, the risk for pregnancy-related VTE may justify the use
three to fivefold increased risk of first-time VTE. Homozy- of postpartum (and, in some cases, antepartum) prophylaxis
gosity confers an 18-fold increased risk compared with indi- with LMWH (discussed elsewhere in ASH-SAP). As with
viduals without the FVL mutation. Additional VTE risk other inherited thrombophilias, a patient with a strong fam-
factors, such as age, smoking, obesity, and particularly use of ily history who is contemplating discontinuation of antico-
estrogens and pregnancy, increase the risk further. Among agulant therapy may wish to undergo FVL testing, since
FVL carriers with a first degree relative with VTE, the inci- homozygosity for the mutation may significantly alter the
dence of first VTE increased from 0.25% (95% confidence estimated future risk of recurrence.
interval [95%CI], 0.12% to 0.49%) in the 15- to 30-year-old
age group to 1.1% (CI, 0.24% to 3.33%) in persons older Pediatric considerations
than 60 years of age. Half of the episodes of VTE occurred
spontaneously, 20% were related to surgery, and 30% were As in adults, heterozygosity for FVL is also associated with an
associated with pregnancy or use of oral contraceptives. increased risk of first VTE. FVL alone, however, does not
The risk for recurrent VTE in FVL heterozygous carriers is increase the risk of recurrent VTE. In the homozygous state
only modestly increased risk (odds ratio [OR] 1.56; 95% or if combined with other thrombophilias, the risk for
confidence interval 1.14-2.12) compared to individuals with recurrent events is increased. Although FVL is not a clinically
a history of VTE without FVL. The risk of recurrence in indi- relevant risk factor for arterial disease in adults, several pedi-
viduals with homozygous FVL compared with those without atric studies have demonstrated an association with stroke,
FVL was estimated to be 2.65-fold (95% confidence interval particularly perinatal stroke.
1.2-6.0) increased, although there are wide ranges in risk
estimates. For practical purposes, there is no clinically Prothrombin 20210 mutation
meaningful association between FVL and arterial thrombo-
General information
embolic events in adults: a meta-analysis demonstrated
the risk to be 1.21-fold increased (95% confidence interval A point mutation in the factor II gene in the noncoding
0.99-1.49) in FVL carriers compared with noncarriers. region in nucleotide position 20210 (G20210A) is the second
most common known inherited risk factor for venous throm-
bosis. Individuals who are heterozygous for this polymor-
Management
phism have slightly higher levels of circulating prothrombin.
Because heterozygosity for FVL confers only a mildly It is inherited in an autosomal dominant fashion.
increased risk of VTE recurrence compared with individuals
without FVL, its finding alone typically does not alter antico-
Prevalence
agulation treatment decisions. Furthermore, asymptomatic
family members of persons with FVL heterozygosity gener- The mutation is found most commonly in individuals of
ally need not be tested. The possible exception is young southern European ancestry, with a prevalence throughout

Europe of 0.7%-4%. In the United States, it occurs in 2% of homozygous prothrombin 20210 mutation do not exist.
the general population and in 0.5% of the African American Although some studies suggest a relationship between the
population. The prothrombin 20210 mutation is rare prothrombin 20210 mutation and stroke and myocardial
in non-Caucasian populations. Homozygosity for the pro- infarction risk in younger patients, meta-analysis has not dem-
thrombin 20210 mutations occurs, by calculation, in approx- onstrated any clinically meaningful association between the
imately 1 in 4,000 individuals of Caucasian heritage. prothrombin mutation and arterial thromboembolism.
Laboratory aspects Management
Testing is done using genetic testing (PCR). Although the Because heterozygosity for the prothrombin 20210 mutation
mutation leads to higher circulating factor II levels, it is not does not confer a statistically significant or clinically relevant
helpful in individual patients to use factor II activity or anti- increased risk of VTE recurrence, its finding does not alter
gen levels as screening tests, because there is a wide overlap length of anticoagulation treatment decisions. Furthermore,
of levels between people with and without the mutation. similar to the discussion about FVL, first-degree relatives of
people who are heterozygous need not be tested routinely
because they are known to be at increased risk if their relative
Risk for thrombosis has experienced symptomatic VTE. Again, the exception
Heterozygosity for the prothrombin 20210 mutation is may be young women in the reproductive phase of their lives
mildly thrombophilic, conferring a three-fold increased risk (Tables 8-5 and 8-6).
of first-time VTE compared with noncarrier status. The
effect of this mutation on the risk for first and recurrent Pediatric considerations
VTE is very similar to FVL.
A recent meta-analysis showed that presence of the
The risk for recurrent VTE in carriers of the prothrombin
prothrombin 20210 mutation leads to an increased risk
20210 mutation compared with the absence of the mutation is,
of recurrent VTE in children, with an odds ratio of 2.15
at most, modestly increased (OR 1.45; 95% confidence interval
(95% confidence interval 1.1-4.1). It is unclear at this point if
0.96-2.2). Thus, treatment decisions on length of anticoagu-
it should alter the duration of therapy.
lant therapy are not based on the presence or absence of the
heterozygous prothrombin 20210 mutation. Population-based
data regarding the risk of thrombosis for homozygotes for the Protein C deficiency
prothrombin gene mutation are not available. A summary of
General information
70 cases of homozygous individuals published in the medical
literature indicates a marked phenotypic heterogeneity. Data Protein C is a vitamin K–dependent protein, converted dur-
on the risk of recurrence of VTE in individuals with ing the coagulation process to APC. APC acts as a natural
Table 8-6 Estimated number of asymptomatic thrombophilic women or women with a positive family history for VTE who would have to
avoid using oral contraceptives to prevent one VTE, and estimated number needed to test
Risk on OC Risk difference N not taking OC N of female relatives
Thrombophilia per year, % per 100 women to prevent 1 VT to be tested
Antithrombin, protein C, or protein S deficiency
Deficient relatives 4.3* 3.6 28 56
Non-deficient relatives 0.7*
Factor V Leiden or prothrombin 20210A mutation
Relatives with the mutation 0.5* 0.3 333 666
Relatives without the mutation 0.2*
Family history of VTE
General population, no family history 0.04† 0.03 3333 none
General population, positive family history 0.08† 0.06 1667 none
Based on family studies as outlined in Table 8-5.
† Based on a population baseline risk of VTE in young women of 0.01% per year, a relative risk of VTE by use oral contraceptives of 4, and a
relative risk of 2 of VTE attributable to positive family history.

anticoagulant. In complex with the cofactor protein S, it level at a separate time point is also necessary. Confirmation
inactivates coagulation factors Va and VIIIa, making them of a hereditary defect by testing a parent or other relative is
unavailable as cofactors during the coagulation process recommended.
(Figure 8-3A). Inherited protein C deficiency as a cause of
thromboembolism was first described in 1981. Two types of
Risk for thrombosis
deficiency are known, but their distinction is not clinically
important with regard to the thrombotic risk they confer. Protein C deficiency is considered to be one of the higher-
Type I deficiency is defined as a quantitative deficiency with risk thrombophilias. It is a risk factor mainly for VTE
low functional protein C (activity) and immunologic (anti- (Table 8-4). Rates of thrombosis vary widely among indi-
gen) level; type II is defined as a qualitative deficiency with viduals and families with protein C deficiency. For asymp-
low activity but normal antigen level. Approximately 85% of tomatic relatives of probands with protein C deficiency and
the reported cases have type I deficiency, whereas 15% have a first VTE, most studies suggest the risk of first VTE is
type II deficiency. More than 160 mutations causing protein increased between four- and sevenfold (Table 8-5).
C deficiency have been described. It is inherited in an auto- The annual incidence of a first VTE is 1.52% in protein
somal dominant fashion. C-deficient individuals compared with approximately 0.1%
in the general population.
Protein C deficiency is only modestly associated with a
Prevalence
risk of recurrent VTE (Table 8-4). In adults, a link between
The prevalence of inherited protein C deficiency in the gen- protein C deficiency and risk of arterial thrombosis has not
eral population is approximately 1 in 500 to 600. By calcula- been firmly established.
tion, homozygous or double heterozygous protein C
deficiency occurs in approximately 1 in 1 million
Management
pregnancies.
Patients with protein C deficiency initiated on VKAs are at
Laboratory aspects risk for warfarin-induced skin necrosis. This transient hyper-
coagulable state is related to abrupt declines in protein C
When evaluating an individual for protein C deficiency, a activity (which was low to begin with) after the initiation of
protein C functional (activity) test should be performed, VKA. Any patient with acute VTE who is initiated on VKA
because obtaining only an antigen level will miss type II needs concurrent anticoagulation with a parenteral antico-
deficiencies. Outside of research studies, there is no need to agulant for at least 5 days and until the international normal-
obtain protein C antigen levels. Because laboratory reports ized ratio (INR) is >2.0, but this is particularly important in
may report results only as “protein C normal,” leaving it the person with known protein C or S deficiency. With
unclear whether an activity or antigen test was done, to regard to the need for testing, similar considerations apply as
avoid missing a type II deficiency, a physician may wish to for FVL. Given the somewhat higher risk increase associated
clarify which test was actually performed. Falsely low pro- with protein C deficiency, particularly in families with a
tein C activity values may be seen with high levels of factor strong tendency to develop VTE, the presence of protein C
VIII and with lupus anticoagulants. The most common deficiency may shift the decision towards extended duration
reason for low protein C levels is treatment with vitamin K of anticoagulation.
antagonists (VKAs) (Tables 8-1 and 8-2). Patients should
stop warfarin for 2-3 weeks (longer for other VKAs) before
Pediatric considerations
protein C activity testing is performed. It is not known how
many patients who carry a diagnosis of protein C defi- Levels of natural anticoagulant proteins change with devel-
ciency truly have a congenital deficiency and how many opment. Protein C activity is very low when compared with
have an erroneous diagnosis of protein C deficiency due to adults at the time of birth. This reduction of activity can per-
testing at an inappropriate time (eg, while on VKAs). Thus, sist until adolescence. When interpreting the results of
the hematologist should always question the diagnosis thrombophilia screening in children, it is imperative that
until review of records and laboratory results has clarified they are compared with pediatric normative ranges and not
that the timing of testing was correct and no confounding adult ranges. Homozygous or double heterozygous protein C
issues led to a transient decrease in protein C. A normal deficiency is associated with catastrophic thrombotic com-
PT at the time of protein C testing is important to exclude plications at birth, manifested by purpura fulminans (exten-
vitamin K deficiency as a cause of decreased protein C sive microvascular thrombosis of the skin) and, less
activity. Repeat confirmatory testing of a low protein C commonly, massive DVT. Intrauterine thrombotic events

also have been reported in these infants. For confirmation of Prevalence

homozygous protein C deficiency in a neonate with purpura
Reported prevalence in the general population varies
fulminans or massive venous thrombosis, the infant should
between 1 in 800 and 1 in 3,000, but due to difficulties in
have undetectable protein C activity (<5 IU/dL) and both
establishing the normal range of protein S concentrations in
parents should be heterozygous for protein C deficiency.
the general population and the difficulties in making an
These newborns need initial treatment with protein C con-
accurate diagnosis, the true prevalence of protein S defi-
centrates (plasma-derived protein C concentrates) along
ciency is not known (Table 8-4).
with anticoagulants to control and prevent the progression
of thrombosis. Neonates and children with severe inherited
protein C deficiency have an ongoing risk of purpura fulmi- Laboratory aspects
nans and, therefore, require long-term antithrombotic ther-
By measuring either free protein S antigen or protein S
apy. The risk of thrombosis is very low in children with
activity, one will detect most cases of protein S deficiency.
heterozygous protein C deficiency. Counseling these families
However, because these individual tests, if done in isola-
about avoiding exposure to transient risk factors of VTE and
tion, can occasionally yield falsely normal results, it is
its signs and symptoms is critical. Adolescent girls with
advisable to include both functional testing (protein S
known protein C deficiency should be discouraged from
activity) and immunologic testing (free protein S antigen)
using hormonal therapy for birth control and should be
if the clinical suspicion for protein S deficiency is high.
offered birth control options without thrombotic risk.
High factor VIII levels, the presence of the FVL mutation,
or the presence of a lupus anticoagulant may give falsely
Protein S deficiency low protein S activity values.
Protein S levels are low in the setting of estrogen therapy,
General information pregnancy, liver disease, nephrotic syndrome, disseminated
intravascular coagulation (DIC), and therapy with VKAs
Protein S is also a vitamin K–dependent protein. Forty per-
(Tables 8-7 and 8-8). Congenital protein S deficiency cannot
cent of protein S exists in a free form, and the remaining 60%
be diagnosed in these circumstances. A patient needs to have
in a complex with the transport protein called C4b-binding
been off warfarin for 3 weeks (longer for some other VKAs)
protein (C4b-BP). It is mostly free protein S that functions as
before protein S levels can be considered reliable. Thus, as
a natural anticoagulant, by being a cofactor for APC to inac-
with protein C deficiency, timing of the testing is essential to
tivate FVa and FVIIIa (Figure 8-3A).
make a correct diagnosis, and repeat confirmatory testing
Protein S deficiency was first described in 1984. More
(including both free antigen and activity) on a new plasma
than 131 different mutations have been identified leading to
sample is advisable. A normal PT at the time the sample is
protein S deficiency. Protein S deficiency is an autosomal
obtained will exclude vitamin K deficiency as a cause for
dominant disorder. Severe protein S deficiency due to
abnormal protein S activity or antigen levels. Critical ques-
homozygous or double heterozygous mutations can lead to
tioning as to whether a patient said to have protein S defi-
early onset of VTE or severe neonatal purpura fulminans
ciency truly has the disorder is appropriate.
and death.
Protein S deficiency is classified into type I, a quantitative
deficiency, in which both free and total protein S antigen lev-
Risk for thrombosis
els are decreased; type II, a qualitative defect due to a dys-
functional protein, in which protein S activity is low, but free Protein S deficiency is considered to be one of the higher risk
and total antigen levels are normal; and type III, a quantita- thrombophilias. Because of the genetic diversity of muta-
tive deficiency, in which free protein S antigen level is low tions associated with protein S deficiency, rates of thrombo-
and the total antigen level is normal. Type III deficiency is sis vary widely among individuals and families with known
either due to a high C4b-BP plasma concentration or to an defects. Although there is considerable variability among
abnormal binding of protein S to this carrier protein. The reports, most family cohort studies have found a relatively
basis for type III deficiencies is not known, but it appears to weak association between protein S deficiency and VTE risk
encompass genetic and environmental factors. The majority (Table 8-4). Interestingly, protein S deficiency seems to have
of the known mutations (approximately 93%) lead to quan- no association with increased VTE risk in some population-
titative deficiencies (ie, type I and III). Protein S deficiency is based case-control studies. The annual incidence of first VTE
inherited in an autosomal dominant fashion. Confirmation is 1.9% in protein S–deficient individuals from families with
of a hereditary defect by testing a parent or other relative is thrombosis, compared with approximately 0.1% in the gen-
recommended. eral population.

Protein S deficiency is only modestly associated with a risk of (ii) the heparin-binding region, and (iii) a variety of other
recurrent VTE (Table 8-4). A link between protein S defi- AT molecule regions. More than 130 different genetic muta-
ciency and increased risk for arterial thrombosis has not tions are known. AT deficiency is inherited in an autosomal
been well established. The heterogeneity in the clinical phe- dominant fashion.
notype of patients with protein S deficiency must be taken
into consideration when making decisions on anticoagulant
Prevalence
treatment and family counseling.
Inherited AT deficiency occurs in 1 in 500 to 5,000 people.
Deficiencies are typically heterozygous, as homozygous defi-
Management
ciencies are almost always incompatible with life. In the gen-
The implications of finding inherited protein S deficiency in eral population, type II deficiencies are the more prevalent
an individual are similar to those discussed for the person subtype, accounting for 88% of all AT deficiencies. A major-
found to have protein C deficiency. Diligent overlap of parity of these type II deficiencies are heparin-binding defects,
enteral anticoagulants upon initiation of VKAs for at least which are not very thrombogenic. Causes of acquired
5 days and until the INR is >2.0 is important to avoid AT deficiency can be found in Table 8-7.
warfarin-induced skin necrosis. Individuals with a first
unprovoked episode of VTE who have a strong family his-
Laboratory aspects
tory of VTE (and are contemplating discontinuation of anti-
coagulants) may wish to undergo protein S testing because a Testing for AT deficiency should be performed using a func-
decrease in protein S activity may shift the decision towards tional assay to detect both quantitative and qualitative
extended duration of anticoagulation. defects. Heparin therapy can decrease AT levels by 30%
(Table 8-8). Testing is best performed a few weeks after the
initial thrombotic event and may best be done when a patient
is not on heparin. No one should be diagnosed as having
Purpura fulminans can occur in the rare newborn with AT deficiency on the basis of one single abnormal test result,
severe protein S deficiency because of homozygous or dou- and a familial deficiency should be confirmed in a relative.
ble heterozygous mutations. Children are born with physio- An abnormal result should lead to repeat testing on a new
logically lower levels of total protein S than adults. Because
the amount of C4b BP also is reduced, however, the free pro-
Table 8-7 Conditions associated with acquired coagulation factor
tein S level is almost the same as found in adults. Any reduc- deficiencies
tion of protein S level or activity in healthy newborns should
Factor Conditions associated with decreased factor levels
normalize by early childhood (after 6 months of age).
Screening asymptomatic children for thrombophilia should Protein C •• Acute thrombosis
thus be delayed until after 6 months of age. Clinical presen- •• VKA therapy
tation of homozygous or double-heterozygous protein S Vitamin K deficiency
deficiency is similar to severe protein C deficiency. The man- •• Liver disease
agement principles of purpura fulminans are similar to •• Protein-losing enteropathy
Protein S •• Acute thrombosis
homozygous protein C deficiency, except that no protein S
•• VKA therapy
concentrate exists. Therefore, fresh frozen plasma (FFP) is
Vitamin K deficiency
the treatment of choice.
•• Liver disease
•• Inflammatory states
•• Estrogens (contraceptives, pregnancy,
Antithrombin deficiency postpartum state, hormone replacement therapy)
General information •• Protein-losing enteropathy
Antithrombin •• Acute thrombosis
AT is an enzyme that interrupts the coagulation process •• Heparin therapy
mostly by inhibiting thrombin (Figure 8-3A), activated fac- •• Liver disease
tor X (factor Xa), and activated factor IX (factor IXa). •• Nephrotic syndrome
It used to be referred to as antithrombin III (ATIII). AT defi- •• Protein-losing enteropathy
ciency was first described in 1965. Qualitative (type I) DIC
and quantitative (type II) defects exist. Type II deficiencies Sepsis
Asparaginase chemotherapy
consist of defects affecting: (i) the thrombin-binding region,

Table 8-8 Influence of acute thrombosis, heparin, vitamin K antagonists, and DOACs* on thrombophilia test results
Acute Unfractionated Low-molecular- Vitamin K Direct oral

Test thrombosis heparin weight heparin antagonists anticoagulant*
Factor V Leiden genetic test Reliable Reliable Reliable Reliable Reliable

APC† resistance assay Reliable‡ ?‡ ?§ Reliable‡ Likely not reliable§
Prothrombin 20210 genetic test Reliable Reliable Reliable Reliable Reliable
Protein C activity ?|| Reliable Reliable low ?§
Protein S activity May be low Reliable Reliable low ?§
Antithrombin activity May be low May be low May be low Occasionally elevated** Probably reliable if
chromogenic assay
is used
Lupus anticoagulant Reliable¶ ?# ?# May be false positive False positive likely
Anticardiolipin antibodies Reliable¶ Reliable Reliable Reliable Reliable
Anti–β2-glycoprotein I antibodies Reliable¶ Reliable Reliable Reliable Reliable
Homocysteine Reliable Reliable Reliable Reliable Reliable
* Dabigatran, rivaroxaban, apixaban, edoxaban.
† APC = activated protein C.
‡ Reliable
if the assay is performed with factor V depleted plasma; thus: clinician needs to inquire how the individual laboratory performs the
assay.
§ Depending
on the way the assay is performed results may be unreliable; health care provider needs to contact the laboratory and ask how the
specific test performs in the presence of the drug in question.
|| Probably reliable, but limited data in literature.
¶ Test often positive or elevated at time of acute thrombosis, but subsequently negative.
# While
many test kits used for lupus anticoagulant testing contain a heparin neutralizer making these tests reliable on UFH and possibly
LMWH, clinicians need to inquire with their laboratory how their individual test kit performs in samples with UFH and LMWH.
**A few case reports show that VKA can lead to an increase in AT levels in selected families.
blood sample. Because type II AT deficiency due to a showed no association between AT deficiency and arterial
heparin-binding defect appears to be much less thrombo- thromboembolism.
genic than type I and other type II subtypes, differentiation
of the AT deficiency subtype may be important for clinical
Management
purposes. Specialized AT assays (AT activity in the absence
of heparin) or gene sequencing need to be used for that Long-term anticoagulation is usually recommended for
purpose, but they are not widely available. patients with AT deficiency who have had a symptomatic
VTE, although this may be inappropriate in patients with a
provoked VTE and/or absence of a strong family history for
VTE. Asymptomatic individuals with AT deficiency typically
Risk for thrombosis
should receive VTE anticoagulant prophylaxis in high-risk
AT deficiency overall is considered to be one of the higher situations. AT concentrate is available, either derived from
risk thrombophilias. Type I and type II mutations affecting the plasma of human donors or transgenically produced in
the thrombin-binding domain can be associated with VTE in goat milk. Because of a lack of high-quality evidence, the role
nearly 50% of affected family members, although this may of AT concentrate in clinical practice is not established.
also be the result of selective testing of thrombophilic fami-
lies. The prevalence of VTE in individuals with a defect in the
heparin-binding site is much lower; only 6% of such indi-
viduals will develop a VTE. AT levels are reduced at birth and normalize by approxi-
Once anticoagulation is stopped, the risk of recurrent mately 6 months of age. Screening of asymptomatic children
VTE in individuals with AT deficiency is considered to should be delayed until this time. Although risk of thrombo-
be high, although later studies found a much weaker sis is very low in children with confirmed deficiency of AT, it
association with recurrent VTE. A large family study is critical that these families should be counseled about

avoiding exposure to transient risk factors of VTE and its mutation, is found in nearly 100% of patients with polycy-
signs and symptoms. themia vera and in 50% of those with ET. Some studies
show that the presence of the JAK2 V617F mutation is asso-
Acquired thrombophilias ciated with an increased risk of thrombosis, either arterial or
venous, in patients with ET. At present, however, there are
Cancer no data to suggest that therapeutic anticoagulation deci-
General information sions should be based on the presence or absence of the
mutation. These disorders are discussed in more detail in
Approximately 20% of all VTEs occur in patients with can- Chapter 16.
cer. About 6% of patients with unprovoked VTE have a pre-
viously undiagnosed cancer at the time of the VTE, and
Splanchnic vein thrombosis and JAK2 V617F mutation
another 10% of patients with unprovoked VTE will be diag-
nosed with a cancer in the year following the VTE diagnosis. The JAK2 V617F mutation commonly is found in patients
Evaluation for occult cancer should be considered in with splanchnic vein thrombosis (Budd-Chiari syndrome
selected patients, such as those with recent weight loss and and portal, mesenteric, and splenic vein thrombosis), occur-
other unexplained symptoms or abnormalities on routine ring in approximately a third of such patients. Only about
laboratory testing, such as anemia. Patients presenting with half of these JAK2 V617F mutation-positive patients have an
unprovoked VTE who are not up to date on age- and gen- overt myeloproliferative neoplasm (MPN) at the time of the
der-appropriate cancer screening (colorectal cancer screen- diagnosis of their thrombotic event. JAK2 V617F mutation–
ing, mammography, pap testing, etc.) should be encouraged positive patients with splanchnic vein thrombosis are more
to become so. It is uncertain whether extensive screening likely to develop an MPN during follow-up than patients
(eg, computed tomography of the chest/abdomen/pelvis) with splanchnic vein thrombosis without the mutation.
for cancer in all patients with unprovoked VTE would result Thus, patients with splanchnic vein thrombosis who are
in decreased cancer-associated morbidity or improved sur- found to have the JAK2 V617F mutation should be followed
vival. Similar to adults, children with cancer are at increased very closely to facilitate early detection of the development of
risk for the development of VTE, but the majority of these clinical signs of an MPN. One can similarly argue that the
VTEs are related to central venous catheters or cancer ther- JAK2 V617F mutation–negative patients should be followed
apy, such as asparaginase. just as closely, because up to 10% of these patients also will
develop an MPN.
Management
Other VTEs and JAK2 V617F mutation
Based on superior efficacy in more than one prospective,
randomized comparison with warfarin, LMWH is the stan- In patients with nonsplanchnic vein thrombosis, the preva-
dard of care for cancer-associated VTE. Guidelines from lence of the JAK2 V617F mutation was found to be around
both the American College of Chest Physicians and the 2% in a Dutch case-control study. The presence of the JAK2
National Cancer Center Network recommend the patients V617F mutation without symptoms of an MPN is not
with cancer-associated VTE receive LMWH monotherapy significantly associated with increased risk of first (OR 4.5;
for at least the first 6 months after diagnosis. The initial 95%CI 0.5–40.9). More importantly, none of the carriers
total daily dose of LMWH mimic that used for treatment of had progression to an MPN over a 6-year follow-up period.
non-cancer acute VTE; after one month, the daily dose can This argues against screening patients with nonsplanchnic
probably be reduced by 20-25%. The DOACs have not been vein thrombosis for the JAK2 V617F mutation.
compared head-to-head to LMWH in cancer-associated
VTE and should not be used routinely in this setting until Paroxysmal nocturnal hemoglobinuria
additional evidence becomes available.
General information
Myeloproliferative disorders PNH is a clonal bone marrow disorder resulting from an
acquired mutation of the phosphatidylinositolglycan class
General information
A gene in a hematopoietic stem cell, leading to absent
Essential thrombocythemia (ET) and polycythemia vera are or decreased cell surface expression of glycoprotein (GP)
associated with a substantial risk for thrombosis (arterial I–anchored proteins on the surface of blood cells. PNH is
more commonly than venous). A gain-of-function muta- associated with increased risk of venous and arterial throm-
tion of the Janus kinase-2 (JAK2) enzyme, the JAK2 V617F bosis, which most often occurs in intra-abdominal veins,

particularly the hepatic veins (Budd-Chiari syndrome). Hormonal therapy and pregnancy
Cerebral and peripheral vein thromboses also occur but
The increased VTE risk associated with hormonal contra-
less commonly. The pathophysiology of thrombosis is
ceptives and pregnancy is discussed in Chapter 3.
not well understood and no consistent abnormalities have
been found.
Antiphospholipid antibodies
Management General information
Screening for PNH by peripheral blood flow cytometry for Antiphospholipid antibodies (APLAs) are acquired
CD55 and CD59 is warranted in thrombophilia evaluations autoantibodies directed against phospholipids and
of patients with venous or arterial thrombosis plus unex- phospholipid-binding proteins, such as b2-glycoprotein-I
plained hemolysis or peripheral blood cytopenias. Anti- and prothrombin. They are associated with arterial throm-
thrombotic therapy (eg, VKA with or without aspirin) can boembolism, VTE, and pregnancy complications. A variety
be used for the treatment and secondary prevention of of different mechanisms leading to thrombosis have been
PNH-associated thrombosis, but “breakthrough” clotting proposed, but the precise pathophysiologic explanation for
events are well described. Long-term treatment with the the clinical phenomena is not known. Diagnosis of the
complement inhibitor eculizumab (Soliris) appears to antiphospholipid syndrome (APS) requires a history of
reduce the risk of thromboembolism (and improve life venous or arterial thrombosis, unexplained recurrent early
expectancy) in patients with PNH. pregnancy loss or one or more late pregnancy losses, or
pregnancy complications associated with placental insuffi-
ciency together with persistent laboratory evidence of
Abnormalities in fibrinolysis
APLA, tested at least 12 weeks apart. The syndrome can
A variety of parameters of fibrinolysis (Figure 8-4) have occur either as primary APS (not associated with any other
been investigated as potential causes of thrombophilia. diseases) or as secondary APS (associated with autoimmune
Investigation of these parameters has been challenging diseases, malignancy, or drugs).
because coagulation assays do not reliably reflect fibrinolysis
of formed thrombi. Studies often have yielded conflicting or
Prevalence
indecisive results regarding an association of antigen levels,
enzyme activity, or certain polymorphisms and the risk for The prevalence of APS is poorly defined, but APLAs are
arterial or venous thrombosis. Given the variability of data found in nearly 50% of patients with systemic lupus erythe-
associating impaired fibrinolysis to arterial and venous matosus (SLE) and up to 5% of the general population.
thrombosis, workup for abnormalities in the fibrinolytic Nearly 40% of patients with SLE will meet diagnostic criteria
pathway (ie, testing for plasminogen, tissue plasminogen for the APS.
activator [tPA], plasminogen activator inhibitor-1 [PAI-1],
and thrombin-activatable fibrinolysis inhibitor [TAFI]),
Testing
with the knowledge we have at present, is not meaningful.
Results do not explain the etiology of a thrombotic event in Laboratory evidence of an APLA is defined as: (i) moderately
an individual patient, and they do not influence decision or highly positive immunoglobulin G (IgG) and immuno-
making regarding length of anticoagulant therapy. globulin M (IgM) anti-b2-glycoprotein I antibodies; (ii) mod-
erately or highly positive IgG and IgM anticardiolipin
antibodies; and (iii) evidence of a lupus anticoagulant

(sometime called a lupus inhibitor) (Figure 8-5). Lupus
Plasminogen
anticoagulants are detected when phospholipid-dependent
Urokinase
tPA PAI-1
clotting times (eg, aPTT, Russell viper venom time) are pro-
Streptokinase
longed. False-positive lupus anticoagulant test results are not
TAFI uncommon, occurring frequently in patients who are on
oral anticoagulants (including the newer thrombin and Xa
Plasmin inhibitors). False-negative results may occur if the blood
Cross-linked Fibrin degradation
fibrin polymer products sample was suboptimally centrifuged and the prepared
plasma was not platelet poor. APLA titers at the time of an
Figure 8-4 Fibrinolysis. TAFI = thrombin-activatable fibrinolysis acute thrombotic event may be decreased temporarily,
inhibitor; tPA = tissue plasminogen activator. thought to be due to consumption, but also may be

transiently positive. Thus, the time of the acute thrombotic APLA tests (ie, lupus anticoagulant, anticardiolipin, and
event is a suboptimal time for testing, and testing may better anti-b2-glycoprotein I antibody tests) is associated with the
be delayed for a few weeks. Because APLA can be transient, highest risk for thrombosis (and pregnancy loss). There is a
guidelines suggest that repeatedly positive tests (at least 5%-15% failure rate of warfarin therapy in preventing recur-
12 weeks apart) be documented, along with corresponding rent thrombosis in patients with APS. Patients with APS are
clinical phenomena, to confirm a diagnosis of APS. thought to be at high risk of recurrent thrombosis, but the degree
A number of other APLA tests are not part of the revised to which the recurrence risk is increased (compared to a similar
Sapporo criteria, as their association with thrombosis patient who tests negative for APLA) is not well established.
or pregnancy loss has not been established, including
immunoglobulin A (IgA) anticardiolipin and IgA anti-
Management
b2-glycoprotein I antibodies, antiphosphatidylserine anti-
bodies, antiphosphatidylethanolamine antibodies, and Because of the previously mentioned challenges related to
antiphosphatidylinositol antibodies. There is presently no laboratory APLA testing and interpretation, as well as the
clear indication for testing for these additional APLA in transient nature of antibodies in many patients, it is advis-
routine clinical practice. The different anticardiolipin and able to always question a diagnosis of APS until the previous
anti-b2-glycoprotein I antibody test kits available for clini- laboratory test results have been reviewed and, if necessary,
cal practice are suboptimally standardized. Also, lupus repeat testing has been performed. Because of the high rate
anticoagulant reporting is not standardized, and labora- of recurrent VTE, patients with true APS with a history of
tory reports can be difficult to read and interpret. Thus, unprovoked VTE should be maintained on anticoagulation
familiarity with the methods of a particular laboratory is indefinitely. Randomized trials have shown that a target INR
especially desirable for APLA testing. range of 2.0 to 3.0 is equally effective in preventing recurrent
The INR determined from plasma occasionally is invalid thrombosis as a target range of 3.0 to 4.0. This probably
in APS patients on VKAs because of a lupus anticoagulant holds true as long as the INR is reliable and indicates a
effect on the prothrombin time. Furthermore, for patients patient’s true level of anticoagulation. If the aPTT is pro-
with APLAs, INR determinations by point-of-care INR longed at baseline due to a lupus anticoagulant, then anti-
monitors are often inaccurate and significantly overesti- factor Xa levels need to be used to monitor UFH therapy. If
mate a patient’s level of anticoagulation. Alternative tests, the PT is prolonged at baseline, then the validity of the
such as chromogenic factor X activity can be used to measure patient’s INR should be checked once the patient is on VKA
the VKA effect when laboratory-based or point-of-care INR by comparing the INR to a chromogenic factor X assay.
testing may be inaccurate. The target ranges for these tests It then can be determined whether the INR is a reliable mea-
depend on the reagents and instruments used for their deter- sure of that patient’s anticoagulation and can be used for
mination, but an INR range of 2.0-3.0 typically corresponds to VKA monitoring. Also, if whole blood point-of-care (POC)
a chromogenic factor X activity of approximately 40%-20%. INR testing is planned for a patient with APS, results of the
POC instrument should be correlated with phlebotomy
plasma-based INR results tested in the clinical laboratory. As
Risk for thrombosis
APLA titers can fluctuate over time, a recorrelation between
True APS is highly thrombophilic and is associated with the INR measured by POC and from a phlebotomy plasma
both arterial and venous thrombosis. Positivity for all three sample should be performed every so often, such as every 4-6
months. It is not known, however, what the optimal frequency
of such recorrelation is. Whether the DOACs will be more,
APLA less, or equally effective compared to VKAs in patients with
APS has not been studied.
LA ACA
It is not known whether patients with arterial thrombosis
and APS are more effectively treated with antiplatelet or VKA
anticoagulation therapy. Some evidence, from patients with
APLA and noncardioembolic stroke, suggests that ASA and
β2-GP I
VKA therapy may be equally effective. In the absence of pro-
spective randomized trial data, this clinical question remains
unanswered. Rituximab has been shown to decrease APLA
Figure 8-5 Antiphospholipid antibodies (APLAs) with their different titers in some patients, but whether lowering (or spontane-
subtypes. ACA = anticardiolipin antibody; b2-GPI = anti-b2-glycoprotein I ous disappearance) of APLA leads to a decreased thrombosis
antibodies; LA = lupus anticoagulant. risk is not known with certainty. The management of pregnant

women with APLA is discussed elsewhere in this self- a simple Mendelian inheritance pattern. Among first-degree
assessment program. relatives of patients with thrombosis and persistently
elevated levels of FVIII, 40% had elevated levels of FVIII.
Prevalence
APLA can be found in a high percentage of children without
any underlying disorder, with an estimated frequency that Elevated factor VIII levels have been defined operationally as
ranges from 3% to 28% for anticardiolipin antibodies and values found in the top decile of a given population. Factor
from 3% to 7% for anti-b2GPI antibodies. The reason of such VIII is an acute-phase reactant, and baseline levels vary con-
frequent occurrence in comparison with adults has been siderably. In the population-based Leiden Thrombophilia
related to the frequent exposure of children to infectious pro- study, 25% of patients with a first episode of VTE had eleva-
cesses. The majority of these antibodies are transient and dis- tions in factor VIII without elevations in C-reactive protein.
appear within a few weeks to few months (~3-6 months). Elevations in factor VIII are seen commonly in patients of
Studies of healthy children who present for surgery, especially African ancestry with VTE.
tonsillectomy, show a 2% prevalence of transient lupus anti-
coagulant with no apparent pathologic consequence due to
Laboratory aspects
the fact that these postinfectious APLAs more commonly
bind cardiolipin in a non-b2-glycoprotein-I–dependent man- Factor VIII clotting (functional) assays are available, but
ner. The prognostic significance of the transient lupus antico- have not been standardized to define the top decile of the
agulant in children who present with thrombosis in the setting local reference population.
of concurrent infection is probably similar to that of children
who have an asymptomatic lupus anticoagulant. It is difficult
to estimate the prevalence of APS in the pediatric population Risk for thrombosis
because there are no validated criteria, and the diagnosis rests Population-based, controlled studies have demonstrated ele-
on extension of adult guidelines and clinical judgment. Trans- vations in factor VIII >150% confer a 4.8 times greater risk for
placental transmission of maternal APLA has been reported first-episode VTE than if levels are <100%. In a large family
in the newborn period. Registry data suggest that these anti- study of first degree relatives of patients with VTE or prema-
bodies are not associated with thromboembolic events. ture arterial disease and elevated levels of FVIII, the absolute
annual incidence in the youngest age group with elevated lev-
els of FVIII:C was 0.16%(0.05–0.37) and gradually increased
Other thrombophilias
to 0.99% (0.40–2.04) in those older than 60 years of age,
Lipoprotein(a) although the odds ratios were not statistically significant.
Some studies have shown that elevated factor VIII levels
Lipoprotein(a) [Lp(a)], which is involved in cholesterol
are also a risk factor for recurrent VTE, but this has not been
metabolism, competes with plasminogen for binding to
found uniformly.
fibrin because of its structural similarity with plasminogen.
This impairs plasminogen activation, plasmin generation,
and fibrinolysis. Lp(a) also binds to macrophages and pro- Management
motes foam cell formation and the deposition of cholesterol
The role of elevated factor VIII levels in recurrent VTE is con-
in atherosclerotic plaques. Elevations in Lp(a) are associated
troversial. Since factor VIII is an acute phase reactant, its levels
with coronary heart disease and stroke in adults, as well as
can vary substantially over time; furthermore, there
ischemic stroke in children. Individual studies in adults have
is no consensus about the level of factor VIII activity at which a
not shown consistent association between elevated Lp(a)
meaningful increase in recurrence risk would be seen. For
and the risk of either first or recurrent VTE.
patients with unprovoked VTE who are contemplating discon-
tinuation of anticoagulant treatment, a FVIII activity determi-
Factor VIII elevation nation may be reasonable if a very high value (>200%-250%)
would lead them to remain on treatment that they otherwise
General information
would have discontinued. However, as discussed in the "Dura-
Elevated plasma levels of factor VIII are an independent and tion of Anticoagulation" section of this chapter, the most
dose-dependent risk factor for VTE. Elevations in factor VIII important clinical predictor of recurrence risk is the nature
have a familial-inherited component, but they do not follow (provoked vs. unprovoked) of the original thrombotic event.

Pediatric considerations Risk for thrombosis
There is suggestion that persistently elevated FVIII activity, Meta-analyses show that the MTHFR polymorphisms in
particularly when associated with persistently elevated North America, where food is supplemented with folic acid,
D-dimers, has prognostic significance in children. These are not risk factors for venous and arterial thromboembo-
children have higher rates of post-thrombotic syndrome and lism or for pregnancy complications.
recurrent thrombotic events. It may be that it is beneficial to
treat these patients with extended anticoagulation, but this
Management
has not yet been investigated.
Because the presence of MTHFR polymorphisms is not a
Homocysteine and MTHFR thrombophilic state, there is no indication to screen for
these mutations. Because lowering of homocysteine levels
General information has no demonstrated clinical benefit on thrombotic risk,
Homocystinuria is a rare autosomal recessive defect in the there is no indication for treatment of elevated homocyste-
homocysteine pathway, most commonly in the cysta ine levels with B vitamin or folic acid supplementation.
thionine-β-synthase enzyme and is associated with mark- Finally, because finding elevated homocysteine levels has no
edly elevated homocysteine levels (>100 µM/L). The clinical consequences, there is no rationale to routinely
worldwide prevalence of cystathionine-β-enzyme deficiency measure homocysteine levels in thrombophilia evaluations.
based on newborn screening is reported at 1 in 344,000 The exception may be in the younger individual (<30 years
live births. Affected individuals have a high rate of arterial of age) with arterial thromboembolism or VTE in whom
and venous thrombotic events before the age of 30 years. there is a suspicion for homocystinuria.
A number of associated symptoms and signs occur, most
commonly dislocation of the lens. Mild to moderately ele- Others
vated homocysteine levels, on the other hand, are common,
and are referred to as hyperhomocysteinemia. Elevated lev- Thrombosis may occur as a complication of systemic or
els may be due to deficiency of vitamin B6, vitamin B12, or local infection. Head and neck infections may trigger cere-
folate; renal impairment; polymorphisms in the genes bral and sinus vein thrombosis. Liver disease not only leads
involved in the synthesis of the enzymes of the homocyste- to a coagulopathy with bleeding diathesis due to decreased
ine metabolism; or unknown causes. synthesis of procoagulant factors but also can lead to an
Modestly elevated levels of plasma homocysteine have increased risk for thrombosis because of decreased synthesis
been shown to be associated with an increased risk of of anticoagulants (eg, AT, protein C, and protein S) and
venous and arterial thrombosis. However, a number of fibrinolytic factors. In children, complex congenital heart
prospective, controlled studies have demonstrated that disease is highly associated with both venous and arterial
lowering a patient’s homocysteine level does not decrease thrombotic events, either because of the disorders or because
the risk of either first or recurrent thromboembolism of the need for cardiac catheterizations, hospitalization, and
(venous or arterial). The methylenetetrahydrofolate reduc- major surgeries.
tase (MTHFR) enzyme is a regulator of homocysteine
metabolism. Polymorphisms in the MTHFR gene may lead
Thrombophilia: reasons to test or not test
to elevated plasma homocysteine levels, but do not neces-
sarily do so. Thrombophilia testing often is considered for patients who
(i) experience unprovoked thrombosis at a young age (<50
years), (ii) experience unprovoked thrombosis at an unusual
Prevalence
site, (iii) have a history of VTE in one or more first-degree
A common MTHFR mutation is the C677T or “thermolabile” relatives, and (iv) remain uncertain about whether to con-
mutation, for which approximately 34%-37% of US whites tinue anticoagulant therapy after estimating the risk of
are heterozygous and 12% are homozygous. The A1298C recurrence with other available information (sex, posttreat-
polymorphism occurs in most ethnic groups and is present in ment D-dimer concentration, family history).
the heterozygous state in 9%-20% of the population. Elevated A variety of reasons for and against thrombophilia test-
homocysteine levels may be seen in an individual with homo- ing exist (Table 8-9). Importantly, negative thrombophilia
zygous C677T mutation or double heterozygous C677T plus testing does not necessarily correlate with a low risk of
A1298C mutation but also may occur in the absence of these VTE recurrence. In asymptomatic relatives, the presence of
polymorphisms. inherited thrombophilia may alter decisions regarding

Table 8-9 Reasons for and against thrombophilia testing thrombophilias as a group were assessed to not be strong or
Reasons for testing consistent enough risk factors to meaningfully predict recur-
•• Patient with thrombosis
rence of VTE. The UK-based National Institute for Clinical
Influence on length of anticoagulation therapy Excellence guidelines recommend that hereditary thrombo-
Possible explanation (for patient and physician) philia testing be considered “in patients who have had
why thrombosis occurred unprovoked DVT or PE and who have a first-degree relative
Reasons against testing
who has had DVT or PE if it is planned to stop anticoagula-
•• Lack of therapeutic consequences even if test positive/abnormal tion treatment.” The same guideline suggests that antiphos-
•• Suboptimal performance of tests (false-positive and false- pholipid antibody (APLA) testing be done only “in patients
negative results) or misinterpretation of tests who have had unprovoked DVT or PE if it is planned to stop
•• Poor medical advice based on test results anticoagulation treatment.”
•• Anxiety, if test is positive
•• False sense of security that thrombosis risk is low, if test result
Thrombophilia testing: pediatric considerations
normal/negative
•• Cost of testing In asymptomatic healthy children or girls considering
•• Lack of impact for asymptomatic first-degree relatives (possible hormonal contraception, thrombophilia testing is not rec-
exception is females contemplating estrogen use) ommended. For children with documented thromboem-
bolic disease, thrombophilia testing is controversial and no
general consensus exists for when and what to test. Based on
contraceptive measures or postpartum prophylaxis in current understanding, thrombophilia testing is not sug-
young women. An important requisite is that a test result gested for children who develop VTE due to exposure to
indeed dichotomizes carriers and noncarriers in terms of transient VTE risk factors. Although a recent meta-analysis
their risk for a first episode of VTE. For women who wish in children with inherited thrombophilia demonstrated that
to use oral contraceptives and who have a positive first- inherited thrombophilia, with the exceptions of FVL and
degree relative with VTE and a known thrombophilic Lp(a) elevation, increases a child’s risk for recurrent VTE,
defect, one can estimate the effect of avoidance of oral con- there is no evidence that screening and identifying the
traceptives on the number of prevented episodes of VTE by inherited thrombophilia rather than practicing standard
means of thrombophilia testing, or alternatively, by using a secondary VTE prophylaxis in any person who has had one
positive family history without thrombophilia testing. The VTE event will reduce the rate of recurrence or postthrom-
results are listed in Table 8-6, in which the first column botic syndrome. Thrombophilia testing should be consid-
shows the observed incidence of VTE during 1 year of oral ered in children who present with unprovoked DVT;
contraceptive use in carriers and noncarriers from throm- thrombosis at unusual site; extension of thrombosis despite
bophilic families. From the risk difference between carriers adequate therapy; and strong suspicion of a higher risk
and noncarriers (second column) the number of women thrombophilia, specifically APS and homozygous protein C
that need to refrain from oral contraceptive use to prevent and S deficiency and AT deficiency. This information can
one episode of VTE can be calculated (third column). Table influence treatment decisions and VTE outcome: (i) a
8-6 clearly indicates that women with antithrombin, pro- patient with AT deficiency may need AT concentrates to
tein C, or protein S deficiency have a high absolute risk of achieve therapeutic anticoagulation with heparin deriva-
VTE provoked by use of oral contraceptives. However, in tives; and (ii) a patient with severe protein C deficiency will
these families, women without a deficiency also have a need replacement with protein C concentrates for the treat-
markedly increased risk of oral-contraceptive-related VTE ment of thrombosis. Although infants who present with
compared to pill users from the general population (0.7% stroke in newborn period have a higher incidence of throm-
versus 0.04% per year of use), reflecting a selection of fami- bophilic traits, the risk of thrombosis recurrence is negligi-
lies with a strong thrombotic tendency in which yet ble; thus, thrombophilia testing is not suggested for these
unknown thrombophilias have cosegregated. Thus, patients.
although selective avoidance of oral contraceptive use pre-
vents VTE episodes in deficient women, for women from
Interpreting test results and educating patients
these families a negative thrombophilia test may lead to
false reassurance. When interpreting thrombophilia laboratory test results, it is
In the 2012 ACCP guidelines, the absence or presence of important to be aware of the circumstances that lead to
thrombophilia did not influence recommendations on dura- abnormal test results without a true thrombophilia being
tion of anticoagulant therapy in patients with VTE because present. Several results are temporarily abnormal in the

patient with acute thrombosis and therapy with heparin and acceptable alternative, it is not known whether one method
VKAs (Tables 8-7 and 8-8). When a thrombophilia is identi- over the other leads to superior safety or efficacy of heparin
fied, educating the patient and the patient’s family members therapy. UFH is mostly cleared by the reticuloendothelial
is important. Online education and support resources on a system and to a smaller degree by the kidney. The half-life of
variety of thrombophilias and the genetic aspects of family heparin in plasma depends on the dose given. It is 60 min-
testing exist (eg, see http://www.clotconnect.org; http:// utes with a 100 U/kg bolus. A patient on continuous infusion
www.stoptheclot.org). intravenous UFH at therapeutic doses likely will have a
return to the baseline aPTT within 3-4 hours after discon-
tinuation of heparin.
Antithrombotic drugs Weight-based heparin dosing nomograms achieve thera-
peutic aPTTs faster than other approaches to selecting an
Anticoagulants
UFH dose. In many patients at average risk for bleeding, a
Heparins loading dose of 80 U/kg of intravenous heparin, followed by
a continuous infusion of 18 U/kg/h is appropriate for full
Mechanism of action
anticoagulation. This dosing, however, may have to be modi-
Heparins are extracted from porcine intestine or bovine fied in the patient at higher risk for bleeding. The aPTT or
lung and consist of glycosaminoglycans of different lengths. anti-Xa level should be determined 6 hours after initiation
UFHs have a mean length of 40 monosaccharide units. of heparin and each dose change, and once every 24 hours
LMWHs are made from UFH through chemical and physi- once the aPTT or anti-Xa level is in the therapeutic range. In
cal processes and have a mean of 15 monosaccharide units. the occasional patient in whom the aPTT is invalid, such as
A pentasaccharide structure within these polysaccharide a patient with a lupus anticoagulant, anti-Xa levels need to
molecules binds to and enhances the action of AT, which be used for heparin monitoring. Neonates may require
inactivates thrombin and factor Xa. Molecules of 18 mono- higher doses of heparin because the clearance is more rapid
saccharide units or more are required to bind thrombin and secondary to a large volume of distribution and they have
AT simultaneously (ie, to enhance heparin’s AT effect on lower AT levels. Long-term use of UFH leads to an increased
thrombin). The five sugars of the pentasaccharide structure, risk of osteoporosis.
however, are sufficient to lead to a conformational change of
AT that can then inactivate factor Xa. Therefore, LMWHs Low-molecular-weight heparin
inactivate mostly factor Xa, whereas UFH acts against
thrombin and factor Xa. Fondaparinux (Arixtra) is a syn- The various LMWH drugs differ in their composition and,
thetic pentasaccharide that binds to AT, leading to specific thus, in their degree of inhibition of thrombin and factor Xa.
inactivation of factor Xa. Therefore, dose recommendations for VTE prophylaxis and
for full-dose treatment vary for the various LMWHs. The
lack of significant binding of LMWHs to plasma proteins
Unfractionated heparin
gives them a more predictable anticoagulant effect than
UFH at therapeutic doses is typically monitored with the UFH, so that fixed or weight-adjusted dosing is possible
aPTT. The therapeutic aPTT range depends on the hepa- without the need for routine anticoagulant laboratory moni-
rin sensitivity of the aPTT reagent and the instrument toring. The peak plasma effect is reached 3-4 hours after
used by a laboratory. A therapeutic aPTT is considered injection. The half-lives of the various agents differ, ranging
that which corresponds to a plasma anti-Xa heparin level between 3 and 7 hours. Once- or twice-daily dosing regi-
of 0.3-0.7 U/mL. Optimally, a coagulation laboratory mens are available for the different drugs. Since the LMWHs
should provide clinicians with the therapeutic aPTT range are—to varying degrees—renally cleared, anti-Xa activity
for the reagent-instrument combination used in that labo- measurement at steady state is suggested in patients with
ratory. If a laboratory has not provided a therapeutic aPTT renal impairment. Because the pharmacokinetic effect of
range for aPTT determinations, then an aPTT ratio of impaired renal function differs among LMWHs, however,
1.5-2.5 times the mid-point of the normal range is often there is not a single creatinine clearance cutoff value below
considered to be therapeutic. With some aPTT reagents, which dose reduction or assessment of anticoagulant effect is
however, this range is subtherapeutic and underdosing of needed. Below a glomerular filtration rate of 30 mL/minute,
a patient may occur. caution with LMWH dosing is appropriate and consultation
UFH therapy also can be monitored with anti-Xa levels, of the package insert for the individual LMWH being
and a number of laboratories have switched to routinely used appears advisable to determine FDA recommenda-
using this method for UFH monitoring. Although this is an tions on dose. In severe renal impairment (creat clearance

Antithrombotic drugs | 211
<15 mL/minute) and dialysis dependence, UFH should be own. Therefore, the minimal amount of protamine to neu-
chosen over LMWH. It may be appropriate to increase the tralize heparin should be given. LMWH is only partially
LMWH dose for patients with morbid obesity (body mass reversed by protamine. In case of significant bleeding on
index of >35 kg/m2). For full-dose LMWH use, dosing LMWH, however, protamine should be considered. Fresh
should be based on actual body weight, and anti-Xa mea- frozen plasma (FFP) likely has little, if any, effect on bleeding
surement generally is not necessary for patients weighing up associated with heparin, LMWH, and fondaparinux and is
to 150 kg. Anti-Xa activity measurement and twice- (rather not indicated unless there is also evidence of a coagulopathy
than once-) daily dosing should be considered in patients resulting in factor depletion.
with morbid obesity.
An expected anti-Xa level (obtained 3-4 hours after sub- Heparin-induced thrombocytopenia
cutaneous injection) is in the order of 1.0-2.0 U/mL for
once-daily dosing; for twice-daily dosing, it is 0.6-1.2 U/mL. Heparin-induced thrombocytopenia is a rare but important
Anti-Xa levels might be useful if a patient on LMWH has a complication that can occur with both unfractionated as
recurrent thrombosis or a significant bleed to document well as LMWH administration. It is discussed in detail in
whether the patient had sub- or supratherapeutic anti-Xa Chapter 10.
levels, which could explain the clotting or bleeding event.
Anti-Xa activity might also be advisable in patients using Heparin resistance
LMWH in the setting of severe renal impairment. That being Heparin resistance is a term used when patients require
said, neither “high” nor “low” levels of anti-Xa activity have unusually high doses of UFH to prolong the aPTT into the
been well correlated with the risk of adverse clinical out- therapeutic range or to prolong the activated clotting time
comes. Many neonates (especially preterm) have minimal above the value (typically >400-450 second) at which extra-
subcutaneous tissue making injection impractical, so that corporeal circulation on heparin is thought to be safe from
intravenous anticoagulation with LMWH can be considered. an anticoagulant point of view. Causes include AT defi-
In children treated with therapeutic doses of LMWH, rou- ciency, increased heparin clearance, significantly low base-
tine monitoring with anti-factor Xa levels is advocated, par- line aPTT (eg, due to elevations of factor VIII and fibrinogen),
ticularly as LMWH therapy often is instituted in critically or or increased nonspecific heparin-binding proteins.
chronically ill children.
Thrombin inhibitors
Fondaparinux
This section discusses only parenteral thrombin inhibitors;
Fondaparinux (Arixtra) is a synthetic pentasaccharide, is dabigatran, an oral thrombin inhibitor, is discussed in the
AT dependent, and consists of the five key monosaccha- section “Direct oral anticoagulants” in this chapter.
rides of heparin that bind to AT and magnify AT-mediated
inhibition of factor Xa. It is specific against factor Xa and
Hirudins
does not inhibit thrombin. It is given subcutaneously,
reaches its peak plasma level in 2 hours, and because of a Natural hirudin is a 65-amino-acid direct thrombin inhibi-
half-life of approximately 17 hours, it is dosed once daily. tor derived from the saliva of the leech Hirudo medicinalis.
Because it does not bind significantly to plasma proteins, it It does not require the presence of AT to exert its anticoagu-
can be given without laboratory monitoring as a fixed lant effect. Several derivatives and recombinant products
dose for prophylaxis of VTE or in body weight–adjusted have been developed. Desirudin (Iprivask) is also a 65 amino
fashion for therapy of VTE. It is cleared by the kidney acid recombinant hirudin, administered subcutaneously.
and, thus, should not be used in patients with creati- Peak plasma levels are reached 1-3 hours after injection. It is
nine clearance <30 mL/minute. Fondaparinux does not metabolized primarily by the kidney, and dose reductions are
cause (and is sometimes used to treat) heparin-induced needed in patients with moderate and severe renal impair-
thrombocytopenia (HIT). ment. It is FDA-approved for postsurgical VTE prophylaxis.
Bivalirudin (Angiomax) is a synthetic, 20-amino-acid poly-
peptide that directly binds to and inhibits thrombin. It is
Management of bleeding
given intravenously and has a half-life of 25 minutes. Dose
If bleeding occurs in a patient on UFH, intravenous prot- adjustment for severe renal impairment is necessary. It is
amine can be given, which binds to and neutralizes heparin. FDA-approved for use during percutaneous transluminal
Protamine can impair platelet function and interact with coronary angioplasty (PTCA), including patients undergo-
coagulation factors, causing an anticoagulant effect of its ing PTCA who have HIT.

Argatroban (Novastan) anticoagulant should be given for at least 5 days; thereafter it

can be stopped when the INR is >2.0.
Argatroban (Novastan) is a small synthetic molecule that
binds to and inhibits thrombin at its catalytic site. It is given
intravenously. Since it is metabolized in the liver, dose reduc- Monitoring and dose requirement
tions in patients with impaired liver function are necessary. VKAs are monitored with the prothrombin time (PT),
Serum tests for liver function always should be obtained which is standardized between laboratories as an interna-
before its use. Its half-life is 40-50 minutes. The drug can be tional normalized ratio (INR). Coumarin VKAs are metabo-
started without the need for an initial bolus. The dosing is lized by the cytochrome P450 enzyme complex, mostly the
adjusted to an aPTT of 1.5-3 times the initial baseline value enzymes CYP2C9 and CYP1A2 (Figure 8-6). Because of a
(not to exceed 100 seconds). It is FDA-approved for the high degree of inter-individual variability in the activity of
treatment of HIT. these enzymes, there is a high degree of variability in the
daily drug dose that patients need to maintain their INR in
Vitamin K antagonists the (narrow) therapeutic range. Polymorphisms in the genes
transcribing enzymes involved in the metabolism of VKAs,
Mechanism of action such as CYP2C9 (cytochrome P2C9 enzyme) and VKORC1
All coagulation factors are synthesized in the liver, although (vitamin K epoxide reductase complex-1) contribute to the
von Willebrand factor and factor VIII also are produced in inter-individual variability in dose requirements. Finger
extrahepatic sites. Factors II, VII, IX, X, protein C, and pro- stick (point-of-care) whole blood INR monitors are avail-
tein S need to be carboxylated in a final synthetic reaction to able and, up to an INR of 4.0, yield results comparable
become biologically active. This step requires the presence to plasma-based measurements performed on a laboratory-
of vitamin K (Figure 8-6). The half-lives of the vitamin based instrument. INR home monitoring by appropriately
K-dependent coagulation factors are 4-6 hours for factor selected patients is safe and effective and a good treatment
VII, 24 hours for factor IX, 36 hours for factor X, 50 hours for option. In some patients with fluctuating INRs, daily supple-
factor II, 8 hours for protein C, and 30 hours for protein S. mentation with micro-dose oral vitamin K, such as 100-300 mg/d,
Because of the long half-lives of some of these factors, par- has been shown to decrease INR fluctuations.
ticularly factor II, the full antithrombotic effect of VKAs is
not reached until several days after having started these
Available VKAs
drugs. Because protein C has a relatively short half-life and
decreases early, its lowering renders the patient hypercoagu- Two classes of VKAs exist: coumarin derivates (warfarin,
lable during the first few treatment days, before factor II, phenprocoumon, acenocoumarol, and tioclomarol), which
with its longer half-life, decreases and protects the patient are the most widely used VKAs; and the indandione deriva-
from thrombosis. Thus, VKAs may create a paradoxical pro- tives (fluindione, anisindione, and phenindione), which are
thrombotic state in the first 5 days, putting the patient at risk used in some countries outside the United States. The only
for coumarin-induced skin necrosis and progression of FDA-approved VKAs are warfarin (approved in 1954) and
thrombosis, unless a parenteral anticoagulant is given over- anisindione (approved in 1957). Warfarin (Coumadin, Jan-
lapping with the VKA in these first few days. The parenteral toven) has a pharmacodynamic half-life of 1-2.5 days, with a
mean of approximately 40 hours.
The typical loading dose of warfarin in the hospitalized
Glutamate patient is 5 mg daily on day 1 and 2, with subsequent dosing
O2 CO2 based on the INR measurement after the first two doses.
Reduced Nonfunctional In children, this will equate to initial doses of 0.1-0.2 mg/kg. A
vitamin K factors frail or elderly patient, one who has been treated with
II, VII, IX, X
prolonged antibiotics, has liver disease, or has undergone
γ-Glutamyl carboxylase intestinal resection, will need a lower dose in the first few
VKOR
Vitamin K
days. Women generally need lower doses. Some clinicians
Factors
epoxide II, VII, IX, X prefer using higher loading doses of 7.5-10 mg, particularly in
(oxidized)
a young, nutritionally replete outpatient. For maintenance
Warfarin
γ-Carboxyglutamate dosing, the highest dose requirements for keeping a patient in
the therapeutic range are in men <50 years old (median dose,
Figure 8-6 Role of vitamin K, point of activation of warfarin, and 6.4 mg/d) and the lowest requirements are in women >70
enzymes involved in vitamin K and warfarin metabolism. years of age (median dose 3.1 mg/d). Occasionally, patients

Table 8-10 Thrombophilia tests to consider if decision to look for and may make re-anticoagulation more difficult. Fresh fro-
thrombophilia is made zen plasma (FFP) can lower the INR some but not com-
Venous thromboembolism pletely or markedly because the amount of any particular
CBC clotting factor in a unit of plasma is small. If complete or
Factor V Leiden mutation immediate INR reversal is needed, such as when treating a
Prothrombin 20210 mutation major bleeding episode, a prothrombin complex concen-
Protein C activity trate (PCC) is preferred over FFP, if available. PCCs are
Protein S activity, free protein S antigen plasma-derived products from human donors that contain
Antithrombin activity high concentrations of the vitamin K–dependent factors (ie,
Anticardiolipin IgG and IgM antibodies II, VII, IX, and X). They exist as so-called four-factor PCCs
Anti–b2-glycoprotein-I IgG and IgM antibodies
(eg, KCentra, Octaplex Cofact), containing all vitamin
Lupus anticoagulant
K-dependent coagulation factors, and as three-factor PCCs
JAK2 V617F and PNH (in splanchnic vein thrombosis)
(eg, Bebulin, Profilnine), which contain relatively low con-
Lipoprotein(a) (in pediatrics)
centrations of factor VII. The four-factor products are capa-
Arterial thromboembolism, unexplained
ble of restoring individual clotting factor activity to nearly
See Table 8-2
100% within minutes of administration of a low-volume
CBC = complete blood count. intravenous infusion. KCentra is the only four-factor PCC
available in the United States as of February 2015. Recombi-
nant factor VIIa is not recommended in the management of
need doses as high as 20-30 mg per day. Genetic testing for
VKA-associated hemorrhage.
polymorphisms of the CYP2C9 and VKORC1 enzyme genes is
available and helps predict, to some degree, warfarin doses
Periprocedural interruption of VKA therapy
needed to reach therapeutic INR ranges; but, despite extensive
clinical trial testing, pharmacogenetic testing has not been Whether there is a need to stop oral anticoagulant therapy
shown to reduce the risk of thrombosis or bleeding. before a surgical or radiological procedure depends on the
bleeding risk associated with the procedure. How far in
Management of elevated INRs and bleeding advance of the procedure to stop VKAs depends on the INR,
the age of the patient, and the half-life of the VKA. Bridging
Several options exist to manage elevated INRs and bleeding therapy with a subcutaneous or intravenous anticoagulant is
that occurs on VKAs, and they depend on the degree of INR typically unnecessary but may be beneficial in patients whose
elevation and the presence or absence of risk factors for thrombosis risk is very high (Table 8-12).
bleeding and of active bleeding itself. A general management
strategy is presented in Table 8-11 and encompasses holding
the next anticoagulant dose(s) and giving vitamin K. Giving
too high a dose of vitamin K should be avoided if there is no Warfarin is the most commonly used VKA in children. Age
major bleeding, because it will reverse the INR completely is an independent predictor of warfarin dose, more so than
Table 8-11 Management
INR Bleeding? Risk factors for Intervention
strategy for elevated INRs in
bleeding?
patients on VKAs.
Supratherapeutic, No No/yes Lower or omit next VKA dose(s); reduce
but <5.0 subsequent dose(s)
5.0-9.0 No No Omit next VKA doses; reduce subsequent
dose; low-dose oral vitamin K will
accelerate INR drop but is not likely to
improve clinical outcome
5.0-9.0 No Yes Vitamin K 1-2.5 mg p.o.
>9.0 No No/yes Vitamin K 2.5-5 mg p.o.
Serious bleed at Yes Vitamin K 10 mg i.v. + FFP or PCCs
any INR
FFP = fresh frozen plasma; INR = international normalized ratio; PCCs = prothrombin complex
concentrates; VKA = vitamin K antagonist.

Table 8-12 Recommendations when interrupting warfarin therapy for invasive procedures*
Risk of thrombosis Before surgery After surgery
Low d/c warfarin 5 d pre-op Restart warfarin 12-24 h after surgery

No LMWH or low dose LMWH No LMWH or low dose LMWH
Intermediate d/c warfarin 5 d pre-op Restart warfarin 12–24 h after surgery
No LMWH or low dose LMWH No LMWH or low dose LMWH
High d/c warfarin 5 d pre-op Restart warfarin 12–24 h after surgery
Full-dose LMWH or iv UFH Full-dose LMWH or iv UFH
d/c = discontinue; pre-op = preoperatively.

*These recommendations are “grade C” recommendations (ie, very weak recommendations based on little or no high-quality evidence).
Other alternatives may be equally reasonable.
the VKORC1 and CYP2C9 genotypes. Children ≤1 year of rapidly expanding as clinical trial data become available and
age require higher warfarin doses, longer overlap with hepa- are being reviewed by the FDA, the reader is encouraged to
rin, longer time to achieve target INR ranges, and more fre- obtain up-to-date approval status information when reading
quent INR testing and dose adjustments, and they have fewer this section of this chapter. The names, molecular targets, and
INR values in the target range. Other factors that affect VKA other pharmacologic properties of the four new oral antico-
therapy in children include diet, drugs (specifically antibiot- agulants furthest along in development are listed in Table
ics), infections, underlying disease state, and weight gain. 8-13 and include dabigatran (Pradaxa), rivaroxaban (Xarelto),
Because newborns have reduced levels of the vitamin apixaban (Eliquis), and edoxaban (Lixiana or Savaysa).
K-dependent proteins, VKA therapy is challenging in very
young children. Infant formulas containing vitamin K sup- Management issues
plements can cause resistance to VKAs, whereas breastfed
infants will be sensitive to VKAs, as there is negligible Several issues are important when managing patients who
amount of vitamin K in breast milk. Infants and even some are being treated with DOACs.
older children have inadequate venous access for frequent First, although routine monitoring of the anticoagulant
INR monitoring. The major complication of VKA use in effect of these drugs is not necessary, in selected clinical situ-
children is bleeding. Reported nonhemorrhagic complica- ations, measurement of their anticoagulant effect will be
tions in children treated with VKA therapy for >1 year necessary. For example, laboratory measurement of antico-
include hair loss, tracheal calcification, and loss of bone den- agulant effect may be helpful for: a bleeding patient, a patient
sity. Point-of-care whole blood monitors have made regulat- in whom treatment failure is suspected, or a patient for
ing VKA therapy more convenient for families, as they can whom the risks and benefits of urgent surgery are being
perform the test regularly at home. considered. Data on expected therapeutic plasma drug lev-
els determined by clinical bleeding and clotting events
and the performance of the various coagulation tests have
Direct oral anticoagulants
been published elsewhere. The ideal test for dabigatran is
Several DOACs have been approved for a variety of indica- the dilute thrombin time or an ecarin-based assay. For
tions in recent years. Most of them are small molecule inhibi- FXa inhibitors (apixaban, rivaroxaban, and edoxaban), an
tors of coagulation factor Xa (anti-Xa drugs) or thrombin anti-Xa activity—calibrated to the drug being measured—is
(anti-IIa drugs). They share several desirable attributes: preferred over other options. Many, but not all, aPTT assays
(i) rapid onset of action; (ii) lack of need for routine monitor- will be prolonged by clinically relevant concentrations of
ing of anticoagulant effect in most patients; (iii) relatively few dabigatran. The same is true of the PT for rivaroxaban. For
clinically important interactions with medications; both medication-test combinations, the clinician should be
(iv) no dietary restrictions; and (v) short half-lives that sim- aware of the sensitivity of his or her own laboratory’s assays
plify perioperative anticoagulation management. On the before interpreting the results.
other hand, the dependence of some of these drugs on renal Second, a growing body of evidence suggests that for
clearance limits their use is some patients. Some of these drugs patients whose thrombosis risk is low or moderate, DOACs
are FDA-approved and available for certain indications, some can be interrupted for brief (<5 days) periods with a very
are awaiting FDA approval, and some are still undergoing low risk of thrombotic complications. The duration of pre-
clinical testing. As the approved indications are relatively procedural interruption will typically be 24-72 hours but will

Table 8-13 New oral anticoagulants: selected pharmacologic properties and approval status
Generic name Apixaban Dabigatran Edoxaban Rivaroxaban
Brand name Eliquis Pradaxa Lixiana, Savaysa Xarelto
Target FXa FIIa FXa FXa
tmax (hours) 1-3 1.25-3 1-2 2-4
Half-life (hours) in 8-15 12-14 8-10 9-13
patients with normal
renal function
Effect of hepatic Mild-moderate hepatic Moderate hepatic Moderate hepatic Moderate hepatic
impairment insufficiency (Child insufficiency (Child insufficiency impairment (Child
Pugh A or B): Pugh B): No evidence of (Child Pugh B): Pugh B): increased
No evidence of a a consistent change in No evidence of a mean exposure by
consistent change in exposure consistent change in 2.3-fold
exposure exposure
Renal excretion (%) 25 80 35-40 66
Effect of renal CrCL 30–50: 1.29-fold CrCL 30–50: 2.7-fold not reported CrCL 30–49: 1.5-fold
impairment greater exposure greater exposure greater exposure
CrCL 15–29: 1.44-fold CrCL 10–30: 6-fold greater CrCL 15–29: 1.6-fold
greater exposure exposure (2-fold increase greater exposure
in the plasma half-life)
Dosing frequency Twice daily Twice daily Once daily Once daily†
Drug interactions P-gp, CYP3A4 P-gp P-gp P-gp, CYP3A4
Approval status as of stroke prevention in AF; stroke prevention in AF; stroke prevention in stroke prevention in
March 2015 (United acute VTE treatment* acute VTE treatment‡ AF; acute VTE AF; acute VTE
States) and secondary VTE and secondary VTE treatment‡ and treatment†* and
prevention; primary prevention secondary VTE secondary VTE
VTE prevention after prevention prevention; primary
total knee or hip VTE prevention after
replacement total knee or hip
replacement
AF = atrial fibrillation, VTE = venous thromboembolism, CrCL = creatinine clearance (mL/min).
*Apixaban is given 10 mg twice daily for the first 7 days in patients with acute VTE.
†Rivaroxaban is given 15 mg twice a day for the first 21 days in patients with acute VTE.
‡For dabigatran and edoxaban a 5-day “lead-in” with heparin or LMWH is required in the treatment of acute VTE.
depend on renal function and the risk of bleeding inherent benefit (if any) in patients with DOAC-associated major
to the planned surgery. bleeding is not established. Specific antidotes for the DOACs
Although both major intracranial bleeding and fatal are being developed, but as of February 2015 are not yet
bleeding occur less frequently with DOACs than with war- available. Despite the lack of antidotes for DOACs, there is
farin, major bleeding in patients taking these drugs can no evidence from either the pooled analyses of the random-
occur. Therapy with oral charcoal is appropriate in the ized phase 3 trials or post-approval registry studies that
patient who ingested drug within 2 hours of presentation major bleeding outcomes are worse in patients taking
with major bleeding. FFP would not be expected to have any DOACs than in patients taking warfarin.
efficacy. Dabigatran is partly dialyzable, unlike rivaroxaban, Very few patients with antiphospholipid syndrome, can-
apixaban, and edoxaban, which cannot be removed with cer, or warfarin failure were included in the VTE treatment
dialysis because they are highly bound to plasma proteins. trials of DOACs; thus, they should be avoided or used with
Preclinical data from animal models, healthy volunteers, caution in these subpopulations, pending further evidence.
and ex vivo coagulation experiments suggest that PCCs or Because dabigatran was inferior to warfarin in a study of
recombinant VIIa (NovoSeven) may be of some benefit, but patients with mechanical prosthetic heart valves, the DOACs
these interventions should be reserved for truly dire cir- should not be used to replace VKA treatment in a patient
cumstances since they can cause thrombosis, and their with a mechanical prosthetic heart valve.

Thrombolytic agents Phosphodiesterase inhibitors
A number of different thrombolytic (fibrinolytic) drugs are in Dipyridamole (Persantine)

clinical use, including streptokinase, urokinase, recombinant
Dipyridamole (Persantine) leads to an increase in intraplate-
tPA, and tPA variants. All of them activate plasminogen to
let cyclic adenosine monophosphate (cAMP) levels, which
plasmin, which can then exert its thrombolytic effect on fibrin
inhibits platelet aggregation to several agonists. By itself,
(Figure 8-4). In clinical practice, these drugs are used rela-
however, dipyridamole has little or no effect as an antithrom-
tively rarely for venous thromboembolism because the associ-
botic drug. Its platelet aggregation inhibitory effect is revers-
ated risk of major bleeding is often not justified by the
ible. The combination of aspirin 25 mg and dipyridamole
potential benefit. Streptokinase is derived from the culture of
200 mg in a sustained-release formulation is available as
beta-hemolytic streptococci and urokinase is derived from the
Aggrenox. Dipyridamole also has vasodilatory effects and,
tissue culture of human neonatal kidney cells. Alteplase
therefore, should be used with caution in patients with severe
(Cathflow, Activase) is a recombinant full-length wild-type
coronary artery disease in whom episodes of angina may
human tPA molecule of 527 amino acids. By deletion or sub-
increase due to the steal phenomenon. Aggrenox has its
stitution of functional domains or alteration of the molecules’
major indication in secondary stroke prevention.
carbohydrate composition, mutants of tPA have been pro-
duced. Reteplase (Retavase) is such a mutant tPA molecule,
modified to be only 355 amino acids long. This leads to a lon- Cilostazol (Pletal)
ger half-life and better penetrance into clots. Tenectaplase Cilostazol (Pletal) is a selective inhibitor of the phosphodies-
(TNKase) is a recombinant full-length tPA molecule with terase-3 isoenzyme and leads to inhibition of agonist-induced
three modifications, leading to increased binding of the mol- platelet aggregation, granule release, and thromboxane A2
ecule to thrombus-bound plasminogen compared with native production. It also has vasodilator effects and should not be
tPA, as well as greater resistance to inactivation by its endoge- used in patients with congestive heart failure. Cilostazol has its
nous inhibitor (PAI-1). In neonates and children <12 months major indication in disabling claudication, particularly when
of age, there is a need for plasminogen supplementation revascularization cannot be performed.
(using FFP) before administration of tPA, as plasma concen-
trations of plasminogen in the first year of life are 50% lower
than adults, making thrombolytic therapy less effective. Pentoxyphylline (Trental)
Pentoxyphylline (Trental) is a phosphodiesterase inhibitor

Antiplatelet agents that has been shown to have some beneficial effects in isch-
emic disease states. Its inhibitory action on phosphodiester-
Aspirin ase in erythrocytes leads to increased cAMP levels and
Aspirin (acetylsalicylic acid) inhibits the enzyme cyclooxy- improved erythrocyte flexibility, and reduction of blood vis-
genase-1 (COX-1), which is needed to form thromboxane A2 cosity may be the result of decreased plasma fibrinogen con-
in platelets. Thromboxane A2 normally is released from centrations and inhibition of red blood cell and platelet
platelet granules upon platelet adhesion and during platelet aggregation. The major indication for pentoxyphylline is
aggregation and serves as an agonist to activate and, thus, peripheral arterial disease with claudication.
recruit other platelets to the platelet plug. Because platelets
do not synthesize new cyclooxygenase and aspirin binds irre- Adenosine diphosphate receptor antagonists
versibly to COX-1, full recovery of thromboxane production
Clopidogrel (Plavix) and ticlopidine (Ticlid)
of the platelet pool after stopping aspirin takes approximately
10 days (ie, the platelets’ life span). Recovery of platelet aggre- Clopidogrel (Plavix) and ticlopidine (Ticlid) inhibit the
gation is quicker, however, occurring within 4 days of stop- adenosine diphosphate (ADP) receptor P2Y12 by irreversibly
ping aspirin, because thromboxane from newly synthesized altering its structure. Both drugs are closely related, but clopi-
platelets can activate aspirin-affected platelets. Complete dogrel has a more favorable side-effect profile with less fre-
inactivation of platelet COX-1 typically is achieved with a quent thrombocytopenia and leukopenia and, therefore, has
daily dose of aspirin of 160 mg. When used as an antithrom- replaced ticlopidine in clinical use. Because maximal inhibi-
botic drug, aspirin is maximally effective at doses between 50 tion of platelet aggregation is not seen until days 8-11 after
and 325 mg per day. In most clinical situations, higher doses starting therapy, loading doses of these drugs often are given
increase the likelihood of toxicity (gastric ulceration and to achieve a more rapid onset of action. Inhibition of platelet
bleeding) but have not been consistently shown to improve aggregation persists for the life span of the platelet. In all indi-
efficacy. cations, clopidogrel appears to be as effective as aspirin, except

Bibliography | 217
in peripheral arterial disease, where it has been shown to be occupying the IIb/IIIa receptor and preventing binding of
slightly more effective for the prevention of ischemic events. fibrinogen, and, thus, preventing platelet aggregation. It is
Clopidogrel is a prodrug, activated in the liver by cytochrome given as a bolus, followed by a continuous infusion for up to
p450 enzymes, including CYP2C19. Genetic polymorphisms 3 days. The platelet aggregation inhibitory effect lasts for
in CYP2C19 lead to decreased clopidogrel metabolism and, 6-12 hours after cessation of infusion.
thus, to a decreased antiplatelet effect. It is unclear, however,
whether switching patients who are poor clopidogrel metabo- Tirofiban (Aggrastat)
lizers to a different antiplatelet agent is clinically beneficial.
Tirofiban (Aggrastat) is a nonpeptide (peptidomimetic),
small-molecule inhibitor of the IIb/IIIa receptor, which also
Prasugrel (Effient), ticagrelor (Brilinta), and cangrelor
binds to the RGD receptor site, similar to eptifibatide.
(Kengreal)
Prasugrel (Effient), ticagrelor (Brilinta), and cangrelor (Keng-

real) also are inhibitors of the platelet P2Y12 receptor. In com-
parison with clopidogrel, they are more rapid in onset, lead to Aspirin dose in children is generally 1-5 mg/kg daily, but
less variable platelet response, and lead to more complete there is variability in the dose required to inhibit platelet
inhibition of platelet function. Prasugrel-mediated inhibition aggregation. The primary side effect of long-term aspirin
is irreversible while the P2Y12 inhibition induced by ticagrelor therapy is bleeding, but it is rarely seen except in neonates
is reversible. who have slower clearance, patients with concurrent bleed-
ing disorders, or children receiving anticoagulation therapy.
There is also a theoretical risk of developing Reye syndrome
Glycoprotein IIb/IIIa receptor antagonists in children with intercurrent influenza or varicella infection,
but this complication usually is not seen unless the dose of
The platelet glycoprotein IIb/IIIa (GPIIb/IIIa) receptors are
aspirin is >40 mg/kg, which is a high dose necessary for an
the sites where fibrinogen binds during platelet aggregation,
anti-inflammatory effect. Dipyridamole in doses of 2-5 mg/
resulting in cross-linking of platelets and platelet plug for-
kg is an alternative to aspirin therapy. Although there is
mation. Several inhibitors of this receptor have been devel-
growing use of antiplatelet agents (eg, clopidogrel) that
oped and are in clinical use.
selectively inhibit ADP-induced platelet aggregation, these
drugs have not been well studied in children.
Abciximab (ReoPro)
Abciximab (ReoPro) is the Fab fragment of a chimeric human-

murine monoclonal antibody against the IIb/IIIa receptor. Bibliography
The drug is given as a bolus, followed by a continuous infusion
for 12 hours or longer. The unbound drug is cleared from Pathophysiology of thrombosis
the circulation with a half-life of about 30 minutes. Drug
Furie B, Furie BC. Mechanisms of thrombus formation. N Engl
bound to the IIb/IIIa receptor inhibits platelet aggregation for
J Med. 2008;359:938-949.
18-24 hours, measured in vitro, but the bound drug is
Mackman N. Triggers, targets and treatments for thrombosis.
demonstrable in the circulation for up to 10 days. Ex vivo Nature. 2008;451:914-918.
platelet clumping in ethylenediaminetetraacetic acid (EDTA)–
containing blood tubes can be seen in patients treated with the
drug, leading to pseudothrombocytopenia when platelets are Thrombophilias and thrombophilia testing
counted by an automatic blood cell counter. This phenome- Howard LS, Hughes RJ. NICE guideline: management of venous
non is clinically irrelevant and does not require discontinua- thromboembolic diseases and role of thrombophilia testing.
tion of the drug. True thrombocytopenia also occurs and, if Thorax. 2013;68:391-393.
severe enough, can require drug discontinuation. Kearon, C. Influence of hereditary or acquired thrombophilias on
the treatment of venous thromboembolism. Curr Opin Hematol.
2012;19:363-370.
Eptifibatide (Integrilin)
Eptifibatide (Integrilin) is a synthetic peptide inhibitor of the

Antiphospholipid antibodies
arginine-glycine-aspartic acid (so-called RGD) binding site
of the IIb/IIIa receptor. It mimics the geometric and charge Keeling D, Mackie I, Moore GW, Greer IA, Greaves M; British
characteristics of the RGD sequence of fibrinogen, thus Committee for Standards in Haematology. Guidelines on the

investigation and management of antiphospholipid syndrome. clinical decision rule. J Thromb Haemost. 2012;10:1326-1334.
Br J Haematol. 2012;157:47-58. The first attempt of developing a comprehensive pediatric specific
VTE risk-prediction tool.
Tormene D, Simioni P, Prandoni P, et al. The incidence of venous
Pediatrics thromboembolism in thrombophilic children: a prospective
Albisetti M, Schmugge M, Haas R, Eckhardt BP, Bauersfeld U, cohort study. Blood. 2002;100:2403-2405. Report on the
Baenziger O, Hug MI. Arterial thromboembolic complication in prevalence of VTE in children with thrombophilia.
critically ill children. J Crit Care. 2005;20:296-300. Provides and Vidal E, Sharathkumar A, Glover J, Faustino EV. Central venous
overview of arterial thromboembolic events in children. catheter-related thrombosis and thromboprophylaxis in
Goldenberg NA, Donadini MP, Kahn SR, Crowther M, Kenet children: a systematic review and meta-analysis: reply. J Thromb
G, Nowak-Göttl U, Manco-Johnson MJ. Post-thrombotic Haemost. 2015;13:161-162. Metaanalyses of studies related to role
syndrome in children: a systematic review of frequency of of thromboprophylaxis for CVL related thrombosis. Literature does
occurance, validity of outcome measures, and prognostic not support pharmacological thromboprophylaxis prevents CVL
factors. Haematologica. 2010;95:1952-1959. Provides related VTE.
an overview of current understanding of post-thrombotic Young G, Albisetti M, Bonduel M, et al. Impact of inherited
syndrome in children. thrombophilia on venous thromboembolism in children: a
Holzhauer S, Goldenberg NA, Junker R, et al. Inherited systematic review and meta-analysis of observational studies.
thrombophilia in children with venous thromboembolism Circulation. 2008;118:1373-1382. Meta-analysis of studies on
and the familial risk of thromboembolism: an observational the risk of first and recurrent VTE with various thrombophilias in
study. Blood. 2012;120:1510-1515. Study arguing that screening children.
for protein S, protein C, and antithrombin deficiency in children
with VTE is warranted and to allow their relatives to be tested,
DOACs
as they have a high rate of VTE if they also have the identified
thrombophilia. Cuker A, Siegal DM, Crowther MA, Garcia DA. Laboratory
Monagle P, Chan AK, Goldenberg NA, et al. Antithrombotic measurement of the anticoagulant activity of the non-vitamin K
therapy in neonates and children: antithrombotic therapy oral anticoagulants. J Am Coll Cardiol. 2014;11:1128-1139.
and prevention of thrombosis, 9th ed.: American College of Siegal DM, Garcia DA, Crowther MA. How I treat target-specific
Chest Physicians Evidence-Based Clinical Practice Guidelines. oral anticoagulant-associated bleeding. Blood. 2014;123:1152-
Chest. 2012;141(2)(suppl):e737S-e801S. Evaluation of published 1158.
evidence regarding antithrombotic therapy and attempt to give
treatment recommendations.
Nowak-Göttl U, Strater R, Heinecke A, et al. Lipoprotein(a) and Venous thromboembolism
genetic polymorphisms of clotting factor V, prothrombin,
Kearon C, Akl EA, Comerota AJ et al. Antithrombotic therapy
and methylenetetrahydrofolate reductase are risk factors of
for VTE disease: antithrombotic therapy and prevention of
spontaneous ischemic stroke in childhood. Blood. 1999;94:3678-
thrombosis, 9th ed.: American College of Chest Physicians
3682. Investigation of association of thrombophilias with ischemic
evidence-based clinical practice guidelines. Chest. 2012;141(2)
stroke in children.
(suppl):e419S-e494S. Detailed evidence-based recommendations
Sharathkumar AA, Mahajerin A, Heidt L, et al. Risk-prediction
on VTE treatment of various patient populations.
tool for identifying hospitalized children with a predisposition
for development of venous thromboembolism: peds-clot

CHAPTER
9
Bleeding disorders
Christine L. Kempton and Jorge A. Di Paola
Overview of hemostasis, 219 Disorders of primary hemostasis, 223 Bibliography, 246
Approach to the patient with excessive Disorders of secondary hemostasis, 232
bleeding, 221 Disorders of fibrinolysis, 245
The online version of this chapter contains an educational integrin a2b1 and glycoprotein (GPVI), resulting in cal-
multimedia component. cium mobilization, granule release, activation of the fibrin-
ogen receptor, integrin αIIbb3, and subsequent platelet
aggregation (Figure 9-1). For a detailed discussion of plate-
Overview of hemostasis let function, please see Chapter 10.
The plasma phase of coagulation can be further subdi-
Hemostasis is the process through which bleeding is con- vided into initiation, priming, and propagation (Figure 9-2).
trolled at a site of damaged vascular endothelium and rep- Initiation begins when vascular injury also leads to exposure
resents a dynamic interplay between the subendothelium, of TF in the subendothelium and on damaged endothelial
endothelium, circulating cells, and plasma proteins. The cells. TF binds to the small amounts of circulating activated
hemostatic process often is divided into three phases: the factor VII (FVIIa), resulting in formation of the TF:FVIIa
vascular, platelet, and plasma phases. Although it is helpful complex, also known as the extrinsic tenase complex; this
to divide coagulation into these phases for didactic pur- complex binds to and activates factor X (FX) to activated FX
poses, in vivo, they are intimately linked and occur in a (FXa). The TF:FVIIa:FXa complex converts a small amount
continuum. The vascular phase is mediated by the release of of prothrombin to thrombin. This small amount of throm-
locally active vasoactive agents that result in vasoconstric- bin is able to initiate coagulation and generate an amplifica-
tion at the site of injury and reduced blood flow. Vascular tion loop by cleaving factor VIII (FVIII) from vWF, activating
injury exposes the underlying subendothelium and proco- clotting factors FVIII, XI (FXI), and platelets, which result in
agulant proteins, including von Willebrand factor (vWF), exposure of membrane phospholipids and release of par-
collagenn and tissue factor (TF) that then come into con- tially activated factor V (FV). At the end of the initiation and
tact with blood. During the platelet phase, platelets bind to priming phases, the platelet is primed with an exposed phos-
vWF incorporated into the subendothelial matrix through pholipid surface with bound activated cofactors (FVa and
their expression of glycoprotein Iba (GPIba). Platelets FVIIIa). During the propagation phase, FIXa, generated
bound to vWF form a layer across the exposed subendothe- either by the action of FXIa on the platelet surface or TF-VIIa
lium, a process termed platelet adhesion, and subsequently complex on the TF bearing cell, binds to its cofactor, FVIIIa,
are activated via receptors, such as the collagen receptors to form the potent intrinsic tenase complex. FX is then
bound and cleaved by the tenase complex (FIXa:FVIIIa)
leading to large amounts of FXa, which in association with
its cofactor, FVa, forms the prothrombinase complex on the
Conflict-of-interest disclosure: Dr. Di Paola: Consultancy: CSL
Behring; DSMB Pfizer. Dr. Kempton Consultancy: Baxter Healthcare,
activated platelet surface. The prothrombinase complex
Biogen Idec, Kedrion Biopharma; Research support: Novo Nordisk. (FXa:FVa) then binds and cleaves prothrombin leading to an
Off-label drug use: Dr. Di Paola and Dr. Kempton: rFVIIa for ultimate burst of thrombin sufficient to convert fibrinogen
management of bleeding in hemophilia at doses and regimens
to fibrin (Figure 9-3) and to result in subsequent clot forma-
that are not approved and for other off-label indications; and pro-
thrombin complex concentrates for treatment of factor II and X tion. The formed clot is stabilized by the thrombin-mediated
deficiency. activation of factor XIII (FXIII), which acts to cross-link
| 219

220 | Bleeding disorders
GPIb-V-IX
(VWF, collagen) PAR P2Y12 Aggregation X II
(thrombin) (ADP)
GPVI Va IIa
(collagen) VIIa Xa
αIIbβ3
Xa
TF
Dense tubular Mitochondria Outside-in
α2β1 structure signaling Tissue factor–bearing cell
(collagen)
Spreading TF
Calcium
Fibrinogen IX IXa
Factor V Initiation
VWF P-selectin
Procoagulant
TFPI
Granule release ADP, 5-HT activity VWF
Ca2+, Mg2+ IIa
VIIa Xa
XI VIII/VWF
Figure 9-1 Platelet activation. Platelets can undergo activation through TF
stimulation by soluble agonists, such as thrombin, or by contact V Platelet
Tissue factor–bearing cell VIIIa
(adherence) to the subendothelial matrix. This simplified cartoon shows XIa
Va
several platelet components, including receptors and granules as well as
the pathways of activation and the effect on platelet responses, such as VIIIa
aggregation, spreading, granule release, and procoagulant activity.
Va
XIa
Activated platelet
fibrin, and thrombin-activatable fibrinolysis inhibitor
Priming
(TAFI), which acts to remove lysine residues from the fibrin
clot, thereby limiting plasmin binding. Ultimately, the clot
undergoes fibrinolysis, resulting in the restoration of normal IXa
blood vessel architecture. The fibrinolytic process is initiated X II IIa
by the release of tissue plasminogen activator (tPA) near the IX
site of injury. tPA converts plasminogen to plasmin, which VIIIa Va
(via interactions with lysine and arginine residues on fibrin)
IXa
cleaves the fibrin into dissolvable fragments. XIa Xa
Both the hemostatic and fibrinolytic processes are regu- Activated platelet
lated by inhibitors that limit coagulation at the site of injury
and quench the reactions, thereby preventing systemic acti- Propagation
vation and pathologic propagation of coagulation. The Figure 9-2 Thrombin generation occurs on two distinct cellular surfaces.
hemostatic system has three main inhibitory pathways; The first is the TF bearing cell at the site of vascular injury. Initiation of
antithrombin (AT), the protein C:protein S complex, and coagulation occurs on the TF bearing cell through generation of a small
tissue factor pathway inhibitor (TFPI). AT released at the amount of thrombin that then goes on to prime the system by activating
margins of endothelial injury binds in a 1:1 complex with platelets, releasing FVIII from vWF and activating it, and activating factor
thrombin, inactivating thrombin not bound by the develop- XI. At the end of the priming step, the activated platelet with bound FXIa
ing clot. Antithrombin also rapidly inactivates FXa; thus, and cofactors FVa and FVIIIa are ready to form essential complexes, tenase
any excess FXa generated by the TF:VIIa complex during (FVIIIa:FIXa) and prothrombinase (FVa:FXa) and through an
initiation is inactivated and unable to migrate to the acti- amplification loop can propagate thrombin generation forming a burst of
thrombin capable to form a hemostatic fibrin clot.
vated platelet surface. Excess free thrombin at the clot mar-
gins binds to thrombomodulin, a receptor expressed on the
surface of intact endothelial cells that when complexed with
thrombin activates protein C; activated protein C complexes results in reduced subsequent thrombin generation and
with its cofactor protein S and inactivates FVa and FVIIIa. quenching of fibrin generation. The action of both AT and
TFPI is a protein produced by endothelial cells that inhibits TFPI inhibits FXa during the initiation phase leading to
the TF:FVIIa complex and FXa. Binding to FXa is required dependence of platelet-surface FXa generation during the
for the inhibitory effect on TF:FVIIa. This negative feedback propagation phase for adequate hemostasis. The fibrinolytic

Approach to the patient with excessive bleeding | 221
COOH Approach to the patient with

H2N NH2 excessive bleeding
D E D Fibrinogen
Excessive bleeding may occur in both male and female
patients of all ages and ethnicities. Symptoms can begin as
COOH early as the immediate newborn period (uncommonly even
Thrombin
in utero) or anytime thereafter. The bleeding symptoms
experienced are related in large part to the specific factor
D E D Fibrin monomer
and level of deficiency; bleeding can be spontaneous; that
Polymerization and is, without an identified trigger, or may occur after a hemo-
cross-linking
2-stranded
static challenge, such as delivery, injury, trauma, surgery, or
D E D D E D the onset of menstruation. Furthermore, bleeding symp-
protofibril
D D E D D toms may be confined to specific anatomic sites or may
occur in multiple sites. Finally, bleeding symptoms may be
Plasmin present in multiple family members or may occur in the
absence of a family history. All of this information is impor-
D D (DD)E complex
tant to arrive at a correct diagnosis rapidly and with mini-
E
mal yet correctly sequenced laboratory testing. Thus, a
Plasmin detailed patient and family history is a vital component of
the approach to each patient with a potential bleeding
disorder.
D-dimer D D E Fragment E
Importance of medical history
Figure 9-3 Fibrin formation and degradation. Fibrinogen has a
Obtaining a detailed patient and family history is crucial
trinocular structure with a central E and two D domains. Thrombin
regardless of prior laboratory testing. The history includes a
cleaves fibrinopeptides A and B (not depicted), located in the E
domain. The resultant fibrin monomers polymerize nonenzymatically
detailed discussion of specific bleeding and clinical symp-
forming fibrin polymers. Factor XIIIa cross-links the D domains of toms. Information regarding bleeding symptoms should
nearby fibrin monomers. Plasmin degrades cross-linked fibrin, thereby include location, frequency, and pattern as well as duration
generating (DD)E complexes composed of an E fragment both in terms of age of onset and time required for cessation.
noncovalently bound to D-dimer. With further plasmin attack, the The location may suggest the part of the hemostatic system
(DD) E complex is degraded into fragment E and D-dimer. affected; patients with disorders of primary hemostasis
(platelets and vWF) often experience mucocutaneous bleed-
system also includes two inhibitors, principally plasmino- ing, including easy bruising, epistaxis, heavy menstrual
gen activator inhibitor-1 (PAI-1), and a2-antiplasmin bleeding, and postpartum hemorrhage in women of child-
(a2AP), which inhibit tPA and plasmin, respectively. bearing age; whereas patients with disorders of secondary
This chapter is devoted to a discussion of the pathophysi- hemostasis (coagulation factor deficiencies) may experience
ology, clinical presentation, diagnosis, prognosis, and treat- deep-tissue bleeding, including the joints, muscles, and cen-
ment of hemostatic abnormalities, hereafter referred to as tral nervous system. The bleeding pattern and duration of
bleeding disorders. The first section will review the approach each episode, particularly for mucus membrane bleeding,
to a patient with excessive bleeding followed by a discussion assist in the determination of the likelihood of the presence
of the specific disorders. of an underlying bleeding disorder. The onset of symptoms
can suggest the presence of a congenital versus acquired dis-
order. Although congenital conditions can present at any
Key points age, it is more likely that patients with a long history of
• Hemostasis is a complex and highly regulated process symptoms or symptoms that begin in childhood have a con-
involving the subendothelium, endothelial cells, circulating cells, genital condition, whereas patients whose onset occurs at an
and plasma proteins that include both positive and negative older age are more likely to have an acquired condition.
feedback mechanisms. Congenital clotting factor deficiencies that do not present
• The generation of thrombin is dependent on specific protein until later in life do occur and include mild factor deficien-
complexes occurring on cellular surfaces: TF:VIIa complex at the cies and coagulation factor deficiencies associated with vari-
site of injury and FIXa:FVIIIa (tenase complex) and FXa:FVa
able bleeding patterns, most notably FXI deficiency.
(prothrombinase complex) on the activated platelet surface.
Additional important information to be collected includes

the current use of medications and herbal supplements, as factor or a global inhibitor as best exemplified by a lupus
these may affect the hemostatic system; the presence or anticoagulant.
absence of a family history of bleeding; a history of hemo- Screening tests also are utilized to identify individuals
static challenges, including surgery, dental procedures, and with a high likelihood of vWD or platelet disorders. The
trauma; and a menstrual history in females. The goal at the bleeding time, once widely used, has become obsolete
end of the history is to establish the likelihood of a bleeding because of the lack of sensitivity and specificity. The PFA-100
disorder, as this will guide the direction of the laboratory (platelet function analyzer) has been proposed to have a role
investigation. Quantification of clinical bleeding is a chal- in screening individuals with suspected platelet dysfunction
lenge, particularly in the outpatient setting. In recent years, or vWD. Initial studies demonstrated the efficacy of the PFA-
several bleeding assessment tools (BAT) have been devel- 100 in the evaluation of patients with known severe platelet
oped to more accurately differentiate bleeding phenotypes disorders or vWD. The PFA-100 induces high shear stress
in healthy individuals and in patients with bleeding disor- and simulates primary hemostasis by flowing whole blood
ders. These tools, which were originally designed for assess- through an aperture with a membrane coated with collagen
ing bleeding in von Willebrand disease (vWD) do not appear and either adenosine diphosphate (ADP) or epinephrine.
to be diagnostic and are in the process of being validated for Platelets adhere to the collagen-coated surface and aggregate
the ability to screen other bleeding disorders. However, it is forming a platelet plug that enlarges until it occludes the
increasingly clear that a normal bleeding score rules out the aperture, causing cessation of blood flow. The time to cessa-
presence of a bleeding disorder. Therefore, if the bleeding tion of flow is recorded as closure time (CT). The sensitivity
score is indicative of excessive bleeding, it should be followed and specificity of the CT of the PFA-100 were reported as
by an evaluation of a hematologist to evaluate the need for 90% for severe platelet dysfunction or vWD, with vWF
further laboratory tests. plasma levels below 25%. The utility of the PFA-100 as a
screening tool, however, has been challenged based on the
reported low sensitivity (24%-41%) of the device in indi-
Screening tests
viduals with mild platelet secretion defect, mild vWD or
The laboratory evaluation for bleeding includes performance storage pool disorders. Additionally, a significant limitation
of initial screening tests. The most common screening tests of the PFA-100 is the fact that the platelet count and hemo-
utilized include the platelet count, prothrombin time (PT), globin levels affect the CT. The CT will be abnormal if the
and activated partial thromboplastin time (aPTT). When the platelet count is less than 100,000/mL and the hemoglobin is
PT or aPTT is prolonged, mixing studies are required via a <10 g/dL.
one-to-one mix of patient plasma with known normal stan- It is likely that by the time patients are referred to a hema-
dard plasma. Test correction in the mixing study indicates a tologist that some, if not all, of the previously mentioned
deficiency state, whereas lack of correction indicates an tests may have been performed. Screening tests are sensitive
inhibitor, either one directed against a specific factor or a to specimen handling, may vary in reliability based on labo-
global inhibitor as best exemplified by a lupus anticoagulant. ratory, and may be influenced by medications. Repeating
Specific factor analyses are performed after mixing studies these laboratory tests often is required; if possible, it is best
reveal a correction of prolonged coagulation screening test(s) to discontinue medications known to affect their results.
indicative of a deficiency state or in the face of normal Therefore, although screening tests are used widely to iden-
screening tests with a positive history. Screening tests are not tify hemostatic abnormalities associated with bleeding, they
sensitive and do not evaluate all abnormalities associated are not perfect. Therefore, the clinical suspicion for a bleed-
with bleeding including vWF, FXIII, PAI-1, and a2AP defi- ing disorder is critical to determine extent of the laboratory
ciencies and may be insensitive to mild FVIII and FIX defi- investigation.
ciencies; therefore, a patient history strongly suggestive of a
bleeding disorder may warrant testing for such deficiencies,
Key points
including rare abnormalities regardless of screening test
results. The most common screening tests utilized include • Patients with bleeding disorders occasionally may present for
the platelet count, prothrombin time (PT), and activated evaluation before symptom onset, especially in the presence of a
partial thromboplastin time (aPTT). When the PT or aPTT known family history or abnormal screening laboratory tests.
• Patients with bleeding disorders can present at any age with
is prolonged >10 seconds, mixing studies are required via a
bleeding in a variety of sites. The more severe disorders tend to
one-to-one mix of patient plasma with known normal
present earlier in life and with bleeding symptoms that often are
standard plasma. Test correction in the mixing study
spontaneous or in such areas as the joints, muscles, or central
indicates a deficiency state, whereas lack of correction nervous system.
indicates an inhibitor, either one directed against a specific

Disorders of primary hemostasis | 223
become activated through binding of agonists such as ADP

Key points (continued) or thrombin, and secrete granular contents that enhance
• The approach to patients with a potential bleeding disorder vasoconstriction and further platelet aggregation. Finally, the
requires a detailed personal and family history and involves the use platelet membrane exposes negatively charged phospholip-
of screening laboratory tests, mixing studies when results are abnor- ids, the surface upon which the plasma clotting factors bind
mal, and subsequent further specific coagulation factor testing. and form the fibrin meshwork.
• Some patients with a history or physical examination
indicative of a bleeding disorder may have a normal laboratory
evaluation. A study by Quiroga et al. showed the diagnostic Etiology
efficacy of laboratory testing in patients with hereditary
Although this section briefly encompasses some of the most
mucocutaneous bleeding is approximately 40%.
well described defects, a full review of platelet function
defects is included in Chapter 10, and a number of excellent
review articles addressing this topic are available.
Disorders of primary hemostasis
Defects at any stage of the platelet activation process can
Platelet function disorders result in platelet dysfunction and subsequent bleeding. For
example, absence or functional defects in GPIba results in
Pathophysiology
Bernard-Soulier syndrome, whereas a gain of function muta-
Platelets play a key role in primary hemostasis both by con- tion in the same receptor is associated with excess binding of
stituting the cellular structure for the primary hemostatic vWF, resulting in platelet-type vWD, a rare bleeding disor-
plug and by providing a phospholipid surface upon which der. Defects in the production, storage, and secretion of
plasma coagulation proteins bind and form complexes. Low vasoactive and hemostatic molecules result in excessive
platelets or impaired platelet function may result in bleeding; bleeding. Such disorders are exemplified by the storage pool
thrombocytopenic and platelet function defects are reviewed defect, which is associated with reduced secretion of ADP,
in detail in Chapter 10. Abnormalities in platelet function and the gray platelet syndrome, a defect in α-granule forma-
can occur in any of the multitude of processes required for tion. A defect in or absence of αIIbb3 results in Glanzmann
normal platelet function, including defects in receptor num- thrombasthenia, the most severe platelet function defect.
ber or function, signaling, and granule content and secre- Most platelet function defects are diagnosed via platelet
tion. An overview of platelet pathophysiology is important aggregation. Identification of the causative defect or its
to the understanding of described platelet function defects. presence in multiple family members implies a genetic

A simplified cartoon with the platelet major receptors and abnormality.
activation responses is shown in Figure 9-1. Platelet activa- Acquired platelet defects are most commonly the result of
tion is the result of multiple signaling pathways that culmi- medications or herbal supplements, chronic medical condi-
nate into the activation of the fibrinogen receptor integrin tions such as uremia, or the result of medical interventions
αIIbb3, an integrin that normally exists in a resting (low- such as cardiopulmonary bypass. The list of medications
affinity) state but that transforms into an activated (high- associated with platelet dysfunction is large. The most com-
affinity) state when stimulated by the appropriate signal monly used medications that result in platelet dysfunction,
transduction cascade. Activated αIIbb3 then mediates platelet many of which are over the counter, include aspirin and
aggregation and promotes stable thrombus formation. This other nonsteroidal anti-inflammatory drugs, antihistamines,
activation occurs following vascular injury when subendo- guaifenesin, certain anticonvulsants (valproic acid in partic-
thelial collagen engages a2b1 and GPVI receptors, and tur- ular), antibiotics, and antidepressants, including, most com-
bulent shear stress promotes vWF binding to GPIba-IX-V. A monly, selective serotonin reuptake inhibitors. Commonly
process known as inside-out signaling follows this platelet used supplements, such as garlic, ginger, omega-3 fatty acids,
surface receptor stimulation, leading to activation of αIIbb3 vitamin E, and ginkgo biloba, also have been reported to
and resulting in affinity modulation during thrombus initia- affect platelet function. Thus, when obtaining a medical his-
tion. This conformational change allows engagement of tory, it is imperative to ask not only about prescribed medi-
fibrinogen by multiple αIIbb3 integrins, resulting in platelet cations but also over-the-counter and herbal supplements.
aggregation. Subsequently, outside-in signaling is initiated Most of these medications and supplements will not lead to
when ligand-occupied αIIbb3 integrins cluster during aggre- a clinically apparent bleeding disorder, but instead they often
gation by binding fibrinogen, fibrin, or vWF and trigger sig- exacerbate clinical bleeding associated with a mild disorder
nals that stabilize the aggregate leading to activation or confound results of platelet function tests. Therefore,
responses, including granule release, platelet spreading, and knowledge of all medications and supplements is critical to
clot retraction. During this multistep process, platelets also interpret laboratory tests. It is also important to mention

that severe anemia can exacerbate bleeding likely due the Platelet aggregometry
impaired rheological effect of red cells that otherwise push
The most specific assay of platelet function is platelet
platelets onto the damaged blood vessel wall.
aggregation by light transmission aggregometry. This
assay uses platelet-rich plasma (PRP) and evaluates plate-
Clinical presentation let aggregation via light transmission after the addition of
a variety of agonists, such as ADP, epinephrine, ristocetin,
Patients with platelet function disorders will present with
arachidonic acid, collagen, and thrombin-related activa-
similar symptoms regardless of the specific defect. The sever-
tion peptide. Recent recommendations by the Interna-
ity of symptoms is dictated by the specific condition and
tional Society of Thrombosis and Haemostasis (ISTH)
clinical situation. Patients with platelet function defects
Subcommittee in Platelet Physiology expanded the use of
exhibit mucocutaneous bleeding similar to patients with
agonists to the thromboxane A2 mimetic U46619. Patients
vWD. Severe hemorrhage can occur in patients with pro-
with a variety of both severe and mild platelet function
found thrombocytopenia or Glanzmann thrombasthenia.
disorders exhibit abnormal platelet aggregation profiles,
Patients may present to the hematologist as a result of abnor-
and, furthermore, the spectrum of abnormalities can be
mal bleeding, a known family history of bleeding either with
diagnostic of specific disorders. For example, if results
or without a personal bleeding history, or an abnormal
demonstrate absent aggregation to all agonists except ris-
screening test such as the PFA-100 obtained before a planned
tocetin, the pattern is diagnostic of Glanzmann throm-
procedure.
basthenia, whereas normal aggregation to all agonists and
absent response to ristocetin is consistent with Bernard-
Diagnosis Soulier syndrome. In addition, a pattern of aggregation
The diagnosis of platelet disorders is covered in Chapter 10. followed by disaggregation with ADP is consistent with
Briefly, the platelet count must be determined and the smear secretion defects. Luminometry, commonly used in com-
reviewed; platelet aggregation assays will be abnormal in the bination with platelet aggregation, provides a sensitive
setting of significant thrombocytopenia (ie, <100 × 109), and evaluation of ATP release from dense granules. ATP
the PFA-100 will be abnormal with significant thrombocyto- released by the platelets provides energy for the added
penia or anemia. Thus, a complete blood count (CBC) light-producing enzyme luciferase, and a light burst is
should be performed before obtaining platelet-specific stud- recorded. In patients with a dense granule deficiency or
ies. The two commonly available tests to screen for platelet platelet release defect, this burst is impaired. A more
function disorders both have limitations. The original detailed discussion of platelet aggregation can be found in
screening test was the bleeding time; as previously stated, the reviews of platelet function disorders.
bleeding time has fallen out of favor because of its limita- As with the PFA-100 CT, several preanalytical variables
tions, particularly its inability to predict clinical bleeding. may affect the results of the test. Many medications and
supplements have been reported to affect platelet aggrega-
tion studies; therefore, the assay should be performed
PFA-100
when patients are no longer receiving these medications
The PFA-100 is a widely available laboratory test that may or supplements for approximately 10 days. It is recom-
be abnormal in some congenital and acquired platelet func- mended that individuals to be tested refrain from con-
tion disorders. Patients with severe platelet function defects, suming alcohol, caffeine, tobacco, or high fat content
such as Bernard-Soulier syndrome and Glanzmann throm- meals several hours before the test. Ideally platelet aggre-
basthenia, also will have abnormal results. The CT is often gation should be performed in fasting state and abstaining
abnormal in patients on aspirin, NSAIDs, and clopidogrel. form flavonoids for several days. The assay can be per-
The effects of other medications known to affect platelet formed in anemic and even thrombocytopenic patients (if
function, such as valproic acid, are not clear. The utility of one suspects a platelet function defect in addition to
the CT is limited by insufficient sensitivity, such that it thrombocytopenia) as it is performed on PRP. For throm-
rarely obviates the need for further testing, and its inability bocytopenic patients, the amount of blood required may
to distinguish between the two most common bleeding dis- be prohibitive, and consultation with the coagulation lab-
orders (ie, platelet function defects and vWD). The CT may oratory is recommended before ordering the assay in this
be abnormal in mild disorders, such as common platelet circumstance. Although most laboratories in the United
secretion defects; however, its sensitivity for these disorders States use PRP for aggregometry studies, whole blood
is insufficient to rule out such defects in the face of a nor- aggregometry is also available in some centers with
mal result. reported reliable results.

Flow cytometry before and approximately 90 minutes after administration;

however, it is recognized that a poor correlation between the
Flow cytometry may be employed to quantify levels of plate-
results of platelet function tests and clinical outcomes exists,
let surface receptors and can confirm the diagnosis of Ber-
and thus, the value of this approach is uncertain. Desmo-
nard-Soulier syndrome and Glanzmann thrombasthenia. In
pressin is a relatively safe agent, although its use can lead to
some institutions, these assays are available and have become
vasomotor symptoms resulting in headache, tachycardia,
the method of choice for diagnosis but have not been stan-
and facial flushing, with rare reductions in blood pressure
dardized for widespread use.
sufficient to result in clinical symptoms. Moreover, as an
analog of an antidiuretic hormone, desmopressin can result
Electron microscopy in water retention and hyponatremia. Although this rarely
occurs in adults and older children, the risks are increased in
Some platelet function defects lead to easily identifiable young children and in those receiving intravenous fluids.
platelet ultrastructural changes visualized by electron Therefore, an experienced care provider should oversee its
microscopy. In particular, patients with a deficiency or use. Repeated use at short intervals should be limited because
absence of dense bodies (d-storage pool deficiency) or of the development of tachyphylaxis. Desmopressin should
α-granules (gray platelet syndrome) can be diagnosed by not be used in children under 2 years of age because of the
this method. high risk of hyponatremic seizures.
Finally, and because most of the genes responsible for Antifibrinolytic agents (aminocaproic acid [EACA] and
these disorders have been identified, genetic testing is avail- tranexamic acid [TXA]) are commonly used adjunctive
able for selected families and may guide future therapeutic hemostatic therapies. These agents, which are lysine ana-
strategies as well as provide information for genetic counsel- logues, inhibit plasmin-mediated thrombolysis and exert
ing. Additionally, it is important to remember that some their effect through clot stabilization and prevention of early
platelet disorders may have systemic manifestations that dissolution. Thus, these agents may be effective in preven-
should be explored such as the presence of oculocutaneous tion of rebleeding, a common problem in individuals with
albinism or pulmonary fibrosis in patients with Hermansky- bleeding disorders especially in areas with increased fibrino-
Pudlak syndrome. lysis, such as the gastrointestinal tract. These agents may be
administered intravenously, orally, or topically in amenable
circumstances. These agents are used either therapeutically
Treatment
for bleeding or prophylactically as part of perioperative
Congenital platelet function defects may benefit from medi- management. Treatment of mucosal bleeding commonly
cal modalities for hemostatic control, although ultimately, includes the use of antifibrinolytic agents in conjunction
platelet transfusions may be required if medications or local with desmopressin; this combination is also effective in
measures are ineffective. In acquired conditions, treatment bleeding from other sites, for example, in the management
or reversal of the underlying condition will resolve the plate- of heavy menstrual bleeding. Antifibrinolytic agents have
let dysfunction; however, this is not always possible. In such been used widely for many years, have a documented safety
situations, the approach to management of bleeding is simi- profile, and are well tolerated in most patients. Commonly
lar to that for congenital disorders. reported side effects include headache and abdominal dis-
Several medications may enhance hemostasis nonspecifi- comfort; however, these symptoms do not preclude its con-
cally and may be useful in the face of platelet dysfunction. tinued use if ameliorated with other agents, such as
These include desmopressin, antifibrinolytic agents, estro- acetaminophen. Antifibrinolytic agents should be used with
gen, and recombinant FVIIa (rFVIIa). Desmopressin may caution in patients with a history of thrombosis or athero-
improve platelet function in many congenital disorders, sclerosis and are contraindicated when hematuria is present
uremia, and during cardiopulmonary bypass; the specific as obstructive uropathy secondary to ureteral clots may
mechanism of action is not clear. Desmopressin may be develop.
administered intravenously, subcutaneously, and, for home Estrogens have documented effectiveness in the manage-
management, intranasally. Intranasal use requires the highly ment of excessive menstrual bleeding. The mechanism of
concentrated solution (Stimate; CSL Behring, King of Prus- action is not well elucidated, although their use is associated
sia, PA) as the intranasal formulation commonly used to with an increase in procoagulants, including vWF and FVIII,
manage diabetes insipidus or enuresis is ineffective as a and a decrease in naturally occurring coagulation inhibitors,
hemostatic agent. In some circumstances, it may be useful to particularly protein S. Conjugated estrogens also are used for
perform a desmopressin challenge test before its clinical use. the management of severe heavy menstrual bleeding with
The challenge test entails assessment of platelet function both the previously mentioned hemostatic effects and with

the additional local effect of reduced uterine blood flow. quality of life. In some patients, bleeding is so severe that
Estrogen in combination with progestins, as in oral contra- bone marrow transplantation has been undertaken to cor-
ceptive agents, is useful for home management of heavy rect the defect by replacing the population of megakaryo-
menstrual bleeding in patients with bleeding disorders, cytes. This extreme approach is reserved only for the most
including platelet function disorders and vWD. The positive severe patients in whom an unaffected human leukocyte
effects of these agents are likely similar to conjugated estro- antigen (HLA)–compatible sibling is available. Patients with
gens in conjunction with progestin-induced stabilization of platelet function disorders should receive thorough educa-
the endometrial lining. The Mirena intrauterine device tion regarding their condition and its management so that
(IUD) is also an emergent therapy for heavy menstrual bleeding episodes are recognized early, managed at home, or
bleeding. The risks associated with estrogens include throm- prevented through appropriate measures or interventions.
bosis; thus, these agents should be avoided in patients with a Important: patients should be advised to report their condi-
history of thrombosis or who are deemed at high risk for tion to physicians before undergoing invasive procedures so
thrombosis. that appropriate prophylactic measures are used to prevent
Although rFVIIa has been shown anecdotally to be effec- bleeding; in addition, all new medications should be checked
tive for the management of severe bleeding in patients with for their ability to interfere with platelet function.
platelet function defects, its value in this setting is not clearly
defined. rFVIIa is costly and may be associated with adverse
Gaps in knowledge
events, including thrombosis; therefore, its use should be
supported by evidence of its efficacy and judicious utiliza- The complexity of establishing a correct diagnosis cannot be
tion. rFVIIa is licensed in the European Union and more underestimated as the first and most important step in the
recently in the United States for the management of bleeding appropriate management of patients with platelet function
in patients with Glanzmann thrombasthenia refractory to disorders. Although current laboratory assays are helpful,
platelet transfusions. For severe bleeding, especially in patients may be left without a more specific diagnosis other
patients with Bernard-Soulier syndrome and Glanzmann than the broad category of a platelet function defect. The
thrombasthenia, platelet transfusion should be administered complexities of platelet structure and function make identi-
to provide normally functioning platelets. The general risks fication at a molecular or cellular level impractical or impos-
associated with platelet transfusion common to all patients sible in many patients outside of specialized research centers.
include the risk of transfusion reactions and potential trans- Therefore, an important area for future research is the devel-
mission of infectious agents (see Chapter 11 for details on opment of widely available laboratory assays with increased
risk of platelet transfusions). A more important specific risk sensitivity and specificity that are able to unravel platelet
associated with Bernard-Soulier syndrome and Glanzmann function defects into better defined categories. Some prom-
thrombasthenia is alloimmunization because of the forma- ising approaches, such as the use of platelet proteomics and
tion of antibodies against the absent receptor. Once antibod- platelet adhesion assays under flow conditions, are being
ies develop, future platelet transfusions are likely to be developed and improved. Although these assays presently
ineffective and may be associated with unusual reactions. are used only in a research setting, it is feasible that further
Thus, judicious use of platelet transfusions is imperative in work will allow development of clinically useful versions. In
these patients. addition, the ongoing development of global hemostatic
The benefits of local measures in the management of assays may allow for identification of a patient’s defect
bleeding episodes for which these approaches are applicable despite their previous evaluations being poorly defined or
should be emphasized. Application of direct pressure is an unrevealing. At present, a number of assays are under evalu-
effective measure for epistaxis, oral bleeding, and cutaneous ation; it is hoped that in the relatively near future, these may
bleeding. For accessible bleeding sites, including the nose, become a part of the armamentarium available in the coagu-
mouth, and skin, the use of topical adjunctive agents are also lation laboratory.
effective and safer than systemic therapy.
Key points
Prognosis and outcomes
• Platelet function disorders can be congenital or acquired and
The majority of commonly encountered platelet function typically present with mucocutaneous bleeding symptoms.
disorders are associated with mild intermittent bleeding epi- • Screening test for platelet disorders have limited value. The
sodes that do not significantly interfere with daily life. Disor- gold standard laboratory evaluation for platelet function disorder
ders like Glanzmann thrombasthenia, however, can be involves platelet aggregation studies.
associated with significant bleeding that profoundly affects

chaperone for FVIII in plasma, protecting it from proteolytic

Key points (continued) degradation. vWF undergoes dimerization in the endoplas-
• Glanzmann thrombasthenia is the most severe platelet mic reticulum (ER), glycosylation in the ER and Golgi com-
function defect and has the potential to result in significant plex, and multimerization in the Golgi complex and is
bleeding requiring blood transfusion. Platelet transfusions in this packed into storage granules after cleavage of the vWF pro-
disorder are reserved for life-threatening bleeding because of the peptide (vWFpp). The vWFpp is released in equimolar con-
risk of developing alloantibodies that render further transfusions centrations to the mature vWF molecule. The vWFpp
ineffective. therefore is useful to measure the rate of clearance of mature
• Secretion defects are among the most common platelet
vWF. When multimers are secreted into the blood, the larg-
function defects and typically cause mild to moderate mucocuta-
est (also called ultralarge) vWF multimers are cleaved by the
neous bleeding symptoms that are managed with desmopressin,
metalloprotease ADAMTS13 (adisintegrin and metallopro-
antifibrinolytic agents, and hormonal therapy for heavy
menstrual bleeding.
tease with thrombospondin). Recent data suggest that vWF
clearance is led in part by macrophages in the liver and
spleen.
Pathophysiology Classification of vWD

vWD is the most common bleeding disorder, with a reported vWD often is categorized into quantitative or qualitative
prevalence of symptomatic disease that ranges from 1/100 to vWF defects. Although vWD type 1 and type 3 represent
1/10,000. The transmission of vWD is autosomal dominant partial and absolute quantitative defects respectively, vWD
for most types but also can be rarely inherited in a recessive type 2 is characterized by a qualitative vWF defect. Follow-
manner. ing is a brief description of the different subtypes and the
vWD is caused by the quantitative or qualitative defi- molecular mechanisms that define them. Figures 9-4 and
ciency of vWF, which is a large, multimeric glycoprotein 9-5 illustrate these mechanisms and how they lead to the
produced in megakaryocytes and endothelial cells. There- current classification. Table 9-1 describes the subtypes in
fore, two pools of vWF are available for normal hemostasis. more detail.
Plasma vWF, mostly released from stored vWF in Weibel-
Palade bodies in endothelial cells, and platelet vWF that is
vWD type 1
stored in α-granules and released upon platelet activation.
The main roles of vWF in hemostasis are to promote platelet vWD type 1 is defined by partial quantitative deficiency of
adhesion to the exposed subendothelium and to serve as a vWF and bleeding symptoms. A family history of
Type 1 VWD Type 3 VWD Type 1C VWD

VWF:Ag VWF:Ag VWF:Ag
VWF:RCo VWF:RCo VWF:RCo
FVIII:C FVIII:C FVIII:C
α granules Multimers Normal Multimers Absent VWFpp/VWF:Ag ratio
Multimers Abnormal
Higher MWM
Decreased secretion Increased clearance
Resting
platelets VWF propeptide VWF
(VWFpp)
Endothelial cell Weibel-Palade body
Figure 9-4 Mechanisms of disease for vWD types 1 and 3. Note that in boxes are shown the most common laboratory findings for these types.
Redrawn from Branchford BR, Di Paola J, Hematology Am Soc Hematol Educ Program. 2012;2012:161-167.

Type 2A VWD Type 2B VWD Type 2M VWD

VWF:Ag VWF:Ag VWF:Ag
VWF:RCo VWF:RCo VWF:RCo
FVIII:C FVIII:C FVIII:C
Multimers Abnormal Multimers Abnormal Multimers Normal
LD RIPA Enhanced
Decreased Increased Decreased
secretion susceptibility Resting Increased binding binding to Type 2N VWD
to ADAMTS13 platelet to platelets platelets VWF:Ag
VWF:RCo
FVIII:C
Multimers Normal
α granules
Decreased binding
to FVIII
VWF
Activated
platelet
Endothelial cell ADAMTS13 Weibel-Palade body Factor VIII
Figure 9-5 Mechanisms of disease for vWD types 2. Note that in boxes are shown the most common laboratory findings for the different
subtypes. Redrawn from Branchford BR, Di Paola J, Hematology Am Soc Hematol Educ Program. 2012;2012:161-167.
the disease or others with clinical symptoms is usually Approximately 75% of cases of vWD type 1 result from
present, yet their absence does not preclude the diagnosis. mutations that exert a dominant negative effect by impair-
Those patients with vWF levels <30 IU/dL usually have ing the intracellular transport of vWF subunits and caus-
identifiable mutations in the vWF gene (vWF) and com- ing subsequent decrease in vWF secretion. A second
monly are associated with significant bleeding symptoms. recently identified mechanism is the rapid clearance of
Table 9-1 Classification and diagnosis of von Willebrand disease
Condition Description vWF:RCo (IU/dL) vWF:Ag (IU/dL) FVIII vWF:RCO/ vWF:Ag
Type 1 Partial quantitative vWF deficiency (75% of <30* <30* ↓ or normal >0.6
symptomatic vWD patients)
Type 2A Decreased vWF-dependent platelet <30* <30-200*† ↓ or normal <0.6
adhesion with selective deficiency of
high-molecular- weight multimers
Type 2B Increased affinity for platelet GPIb <30* <30-200*† ↓ or normal Usually <0.6
Type 2M vWF-dependent platelet adhesion without <30* <30-200*† ↓ or normal <0.6
selective deficiency of high-molecular-
weight ↓ multimers
Type 2N Markedly decreased binding affinity for 30-200 30-200‡ ↓↓ <0.6
FVIII
Type 3 Virtually complete deficiency of vWF <3 <3 ↓ ↓ ↓ (<5 IU/dL) Not applicable
(severe, rare)
Low vWF 30-50 30-50 Normal <0.6
Normal 50-200 50-200 Normal <0.6
↓ refers to a decrease in the test result compared to the laboratory reference range.
* <30 IU/dL is designated as the level for a definitive diagnosis of vWD; some patients with type 1 or type 2 vWD have levels of vWF:RCo or
vWF:Ag of 30-50 IU/dL.
†The vWF:Ag in the majority of individuals with type 2A, 2B, or 2M vWD is <50 IU/dL.
‡This does not preclude the diagnosis of vWD in patients with vWF:RCo of 30-50 IU/dL if there is supporting clinical or family evidence for
vWD, nor does this preclude the use of agents to increase vWF levels in those who have vWF:RCo of 30-50 IU/dL and who may be at risk for
bleeding.

vWF from the circulation because of specific mutations in causing reduced binding to FVIII allowing for increased
the vWF gene. Therefore, impaired secretion and increased clearance).
clearance are likely the two most common molecular
mechanisms that lead to vWD type 1. The variant of vWD vWD type 3
type 1 that is due to increased clearance is called type 1C.
Because vWF is synthesized on a 1:1 ratio with vWFpp, an vWD type 3 is inherited in an autosomal recessive mode and
alteration of the ratio in favor of the propeptide suggests is characterized by complete lack of vWF protein with unde-
increased vWF clearance. This, plus the presence of tectable vWF antigen assay (vWF:Ag) and ristocetin cofactor
unusually large multimers, is indicative of vWD type 1C. assay (vWF:RCo) levels, and resultant very low FVIII:C levels
Patients with type 1C vWD have a robust initial response (<5%), representing the steady state of factor VIII in the
to desmopressin, but they exhibit an abrupt vWF level absence of its vWF chaperone. Multimers are absent and the
decrease within 2-4 hours, placing them at high risk for bleeding pattern is usually severe.
delayed post-operative hemorrhage. The clinical presentation of vWD includes mucocutane-
A consistent diagnostic criterion is difficult to achieve, as ous bleeding—specifically easy and excessive bruising and
not all individuals that inherit a mutation in vWF show bleeding from mucosal surfaces, including the nose, mouth,
signs of clinical disease (phenomenon known as low pene- and gastrointestinal tract. The extent, location, and nature of
trance), and not all individuals that inherit the same muta- bruising are important clinical points. Multiple bruises of
tion show the same clinical signs (known as variable various ages in a variety of locations are suggestive of a dis-
expressivity). More than 50% of individuals with vWF lev- order of primary hemostasis. Epistaxis or oral-pharyngeal
els in the mildly decreased range (30-50 IU/dL) are asymp- bleeding sufficient to result in anemia suggests the presence
tomatic or have minimal bleeding symptoms. Therefore, of a hemostatic disorder. Heavy menstrual bleeding, particu-
the presence of plasma vWF levels between 30 and 50 IU/dL larly at onset of menarche, also is suggestive of a mucocuta-
does not automatically define vWD type 1. Individuals with neous bleeding disorder. Excessive bleeding following
blood group O have 25%-30% lower vWF levels as com- procedures involving the mucus membranes may unmask a
pared with those who have blood group A, although this previously unknown bleeding disorder. The most common
variability should not affect the way that the disease is diag- of these events include childbirth, oral surgery, including
nosed. However, due to the fact that mild bleeding com- dental work, tonsillectomy or adenoidectomy, and sinus sur-
monly is reported in the healthy population, it is possible gery. Some patients may present to the hematologist as a
that many individuals diagnosed with mild vWD type 1 result of a documented family history of bleeding without an
may not have genetically inherited vWD but rather an asso- individual specific bleeding event. Less commonly, patients
ciation of mild decreased vWF levels (within the estab- may present because of abnormal screening tests ordered
lished range for blood group O) and mild bleeding before a planned procedure. Clinical manifestations may
symptoms. Regardless of causality, most individuals with range from mild to severe. Type 3 vWD may be associated
vWF levels 30-50 IU/dL may benefit from similar therapeu- with similar bleeding events observed in severe hemophilia,
tic measures used for patients with vWD type 1 depending likely because of the extremely low FVIII levels. Severe heavy
on their personal and family history of bleeding and the menstrual bleeding resulting in early hysterectomy has been
severity of hemostatic challenge. observed in women with all subtypes.
vWD type 2 Diagnosis
vWD type 2 is characterized by qualitative defects in vWF As mentioned previously, screening laboratory tests have
due to mutations in vWF that affect the interactions of vWF limited value when a diagnosis of vWD is suspected. There-
with many of its ligands. vWD type 2 is subclassified into fore, in clinical practice, in the face of a significant history of
type 2A (loss of intermediate- and high-molecular-weight mucocutaneous bleeding, specific laboratory assays for vWD
multimers because of decreased secretion or increased sus- are required.
ceptibility to ADAMTS 13), type 2B (gain-of-function muta- Diagnostic assays for vWD include quantitative measure-
tion resulting in spontaneous vWF-platelet binding under ment of vWF (vWF:Ag), the platelet-binding function
physiologic shear conditions, resulting in clearance of the (vWF:RCo, in which the agglutination of fixed platelets in
highest-molecular-weight multimers and mild thrombocy- response to patient plasma is measured in the presence of
topenia), type 2M (loss of function mutations that decrease ristocetin), and the FVIII coagulation (FVIII:C). Also, the
the interaction of vWF with its platelet receptor and decreases vWF multimers distribution is used to differentiate subtypes.
ristocetin cofactor activity), and type 2N (mutations in vWF A potential role for assays that measure the binding of vWF

to collagen has been described and has recently gained atten- mature vWF molecule. Multimeric analysis is required to
tion by the fact that several families have been reported to differentiate between various subtypes of vWD type 2, and
have mutations in the collagen binding site of vWF that can- their absence easily identifies vWD type 3 (Figure 9-6). The
not be easily identified by traditional laboratory assays, yet FVIII activity level and FVIII-binding assay provide a more
are thought to have contributed to abnormal bleeding. Limi- accurate diagnosis of vWD type 2N. Finally, the collagen-
tations exist with several of these assays. Both vWF and FVIII binding assay measures binding of large vWF multimers to
are acute-phase reactants and may increase two to five times collagen and represents an additional method to assess vWF
above baseline because of a variety of conditions or circum- functional activity. The collagen-binding assay does not
stances, including (among others) infection, stress, and preg- require the use of ristocetin, but studies have reported that
nancy. These increased levels elevate low baseline levels to the type of collagen employed influences the results.
within the normal range, obscuring diagnosis. Therefore, Laboratory test results are compatible with vWD type 1 if
normal levels do not completely rule out vWD, especially in the levels of both vWF:RCo and vWF:Ag < 30 IU/dL and the
the face of a suspicious clinical history, and must be inter- plasma vWF multimer distribution is normal, though the
preted with caution. Performance of these assays requires an intensity may be reduced due to lower amount of protein.
experienced coagulation laboratory, ideally with on-site pro- Additionally, the vWF:RCo/vWF:Ag ratio approximates 1. In
cessing and analysis as opposed to an experienced coagula- patients with vWD type 1C, the vWF:Ag and vWF:RCo levels
tion laboratory analyzing a “send-out” sample often drawn are low, and in most cases the multimer assay is characterized
thousands of miles away. Because of the difficulty in ruling by the presence of abnormally large high-molecular-weight
out this disorder with one normal evaluation, it is not
uncommon for patients to undergo repeated testing. When
local laboratory results are inconsistent, a useful strategy is to
perform testing in a reference hemostasis laboratory. Finally,
many pre-analytic variables must be considered to accurately
interpret laboratory testing. For example, refrigeration of
whole blood samples before separation can result in reduced
plasma vWF levels; in addition, platelet contamination of the
separated plasma may result in protease-induced vWF alter-
ations, causing decreased activity. Given these variables, the
laboratory diagnosis should not be made on just one set of
subnormal levels but at least two unless there are clearly
identified acute phase reactant effects present “normalizing”
at least one set of vWF levels.
vWF:RCo is widely used and has been accepted for decades
as the gold standard for vWF activity. The assay can be diffi-
cult to interpret, as it exhibits a high coefficient of variability
when the vWF:RCo is lower than 15 IU/dL. A recent report
suggests that the vWF:RCo assay can be abnormal in a subset
of otherwise-healthy African Americans due to the presence
of the common single nucleotide polymorphism (SNP)
0.65% agarose
D1472H in exon 28 of vWF. This SNP affects ristocetin bind-
ing without conferring an evident hemorrhagic risk and Figure 9-6 Representation of a vWF multimer analysis. The third
potentially may lead to a false diagnosis of vWD type 2. The column from the left represents normal plasma as indicated by the NP
collagen binding assays should also be normal in these indi- at the top of the column. In type 2A vWD, there is a loss of high- and
viduals. Newly developed functional assays utilizing the intermediate-weight multimers as indicated by the loss of the bands in
the gel under the heading. In type 2B vWD, there is a loss of HMWM.
platelet ligand for vWF, GPIb, may allow for assessment of
In type 1, all the multimers are present but in reduced amounts as can
vWF activity without the need for ristocetin, thus improving
be seen by the presence of all the bands but with more faint staining
discrimination of type 2 variants without the noted pitfalls.
than seen in normal plasma. In type 3 disease, there is a complete
Low-dose ristocetin-induced platelet aggregation (LD-RIPA) absence of multimers, and no staining of bands is visible. The labeled
is used to identify abnormally increased binding of vWF to columns VSD and TTP stand for ventricular-septal defect, a condition
platelets, as occurs in type 2B and platelet-type vWD. vWF that results in AWS with the loss of multimers of all sizes, and
multimers usually are run on an agarose gel to evaluate the thrombotic thrombocytopenic purpura in which ultralarge multimers
full range of molecular weight multimers present within the can be observed.

forms. As this subtype is characterized by rapid vWF clear- of a homologous partial pseudogene in chromosome 22.
ance, a vWFpp level allows for discrimination of vWD type Therefore, gene sequencing for diagnosis currently is
1C through the vWFpp/vWF:Ag ratio. reserved for specific cases in which these test results will
vWD type 2 is a qualitative defect caused by mutations in likely contribute significantly to diagnosis and management,
vWF that result in abnormal interactions with several of particularly in cases in which treatment options vary based
its ligands. The diagnosis of type 2A is made in the presence on diagnosis. Genetic testing can be used to differentiate
of a low vWF:Ag and a disproportionately low vWF:RCo types 1 from type 2M, type 2B from pseudo-vWD, and mild
with pronounced loss of high-molecular-weight multimers hemophilia from vWD type 2N. Genetic testing may be also
(HMWM). The vWF:RCo/vWF:Ag ratio approximates 0.5. justified in vWD type 3, as large deletions may predispose to
Type 2M is caused by mutations in the platelet glycoprotein the development of inhibitory antibodies and anaphylactic
1bα (GPIb) binding site, with resultant decreased binding of reactions.
vWF to GPIb and subsequent impairment of platelet-depen-
dent function. The multimer structure and distribution in
Acquired von Willebrand syndrome
vWF is normal. Type 2B results from gain-of-function muta-
tions in the binding site for GPIb, leading to the formation Acquired von Willebrand syndrome (AvWS) is a rare disor-
of rapidly cleared platelet–vWF complexes. LD-RIPA is der in which, in most cases, vWF is synthesized normally but
employed to confirm this subtype. A level of ristocetin insuf- cleared from the circulation more rapidly. One exception is
ficient to promote platelet binding with normal vWF causes the association of hypothyroidism and AvWS caused by
enhanced platelet agglutination in these gain-of-function decreased secretion of vWF that normalizes when patient are
mutations. This phenomenon is also seen in patients with given thyroid hormone replacement. Three mechanisms are
platelet-type vWD (also known as pseudo-vWD), a rare disassociated with the observed increased clearance: (i) auto
order caused by mutations in platelet GPIb. It is important antibodies against vWF and immune complex formation
to differentiate these two entities, as treatment approaches (eg, systemic lupus erythematosus, Hashimoto thyroiditis),
are significantly different. vWD type 2B is treated with vWF (ii) vWF binding to cancer cells (eg, Wilms tumor, lympho-
concentrates, as the molecular defect is in vWF; whereas proliferative disorders), and (iii) increased proteolytic
pseudo-vWD is treated with platelet transfusions, as it is activity of HMWM under pathological high-shear-stress
caused by mutations in platelet GPIb. For the evaluation of conditions (eg, congenital heart disease, aortic stenosis,
pseudo-vWD, the patient’s platelets are tested with a normal extracorporeal devices). Also, ciprofloxacin has been associ-
exogenous vWF substrate in a ristocetin-induced platelet ated with AvWS, likely by induced proteolysis of high-
agglutination-based mixing study. Enhanced binding con- molecular-weight multimers. The laboratory diagnosis and
firms the diagnosis. Finally, type 2N is characterized by management of AvWS does not differ significantly from the
mutations in the FVIII-binding site of vWF, disturbing the congenital forms. Treatment of the underlying disorder
normal interaction of these two proteins. Patients with vWD leading to AvWS often resolves the defect.
type 2N may exhibit normal or decreased vWF:Ag and
vWF:RCo with disproportionately decreased FVIII:C.
Treatment
Patients with this diagnosis may be misclassified as mild
hemophilia. Specific FVIII-binding assays are used to con- The principle of management of vWD is to increase or
firm the diagnosis of type 2N. Symptomatic patients are replace vWF to achieve hemostasis. This is accomplished
either homozygous or compound heterozygous for muta- with either medications that cause the release of endogenous
tions in the vWF. Patients with a prior diagnosis of mild stores of vWF into the circulation (desmopressin) or the use
FVIII deficiency who do not respond well to FVIII infusions of vWF-containing concentrates derived from human
or belong to families for whom the inheritance appears to be plasma. Mild to moderate bleeding associated with type 1
autosomal dominant should be evaluated for vWD type 2N. vWD often is managed with desmopressin, most commonly
vWD type 3 is characterized by undetectable vWF:Ag and with the intranasal preparation, and antifibrinolytic agents
vWF:RCo levels, FVIII:C levels commonly <5%, and lack of as required for mucosal based surgery or bleeding. Desmo-
multimers. A description of the laboratory pattern for each pressin’s mechanism of action is based on the secretion of
subtype is shown in Table 9-1. stored vWF from Weibel-Palade bodies in endothelial cells
into the plasma. A desmopressin challenge test, as described
in the “Platelet Function Disorders” section in this chapter,
Genetic testing
should be performed to document a hemostatic response; in
Sequencing of the vWF gene is challenging due to its vWD, the vWF:Ag, vWF:RCo, and FVIII levels are performed
large-size, highly polymorphic structure, and presence before and 60-90 minutes after the dose, depending on the

route of administration. Repeat laboratory evaluation at 4 type 1 disease. This can be especially difficult because vWF
hours post dose may be appropriate when an altered half-life levels may appear to be normal because of the associated
of the native protein is suspected, as observed in type 1C. clinical circumstances, despite a clinical history suggestive of
Approximately 90% of patients with type 1 vWD respond this disorder. Recently published data used a Bayesian analy-
with hemostatic levels; however, the response varies and sis of laboratory data and personal and family history to pre-
should be measured to determine its adequacy for specific dict the probability of diagnosis of vWD. Future research
hemostatic challenges. Repeated administration of desmo- aimed at the development of laboratory assays with improved
pressin in proximity may lead to tachyphylaxis, with performance characteristics to decrease variability and diag-
decreased response levels with repeated use likely resulting nostic dilemmas is needed. A wide variation in bleeding
from depletion of the storage pool. Repeated doses also symptoms exists among patients within the same disease
increase the risk of hyponatremia. Thus, use of desmopressin subtype, likely because of genetic modifiers of the bleeding
no more than once daily and no more than on 2-3 consecu- phenotype. Overall, currently available therapies are effec-
tive days serves as an acceptable clinical guideline for home tive; however, it is not completely clear under what circum-
use. There are some reports of the benefits of desmopressin stances specific therapies are best applied to achieve an
in type 2 vWD; in general, it is less effective in these subtypes optimal outcome. There are few prospective comparative
and has been reported to precipitate thrombosis or result in therapy studies to guide physicians in determining the risks
significant thrombocytopenia as a result of in vivo platelet and benefits of available therapies. Published treatment
aggregation in type 2B or platelet-type vWD. However it can guidelines published by the National Heart, Lung, and Blood
be useful in some patients with types 2M and 2A and in some Institute as well as more recent ones published by the United
genotypes of Type 2B. Desmopressin is ineffective in type 3 Kingdom Haemophilia Centre Doctors Organization are
vWD, and treatment is dependent on the use of replacement based on the best available evidence and expert opinion.
therapy via concentrate.
Several products available in the United States contain Key points
intact vWF, including Humate-P (CSL Behring, King of
• vWD is one of the most common inherited bleeding disorder
Prussia, PA), Alphanate (Grifols Biologicals, Los Angeles,
in the general population.
CA), Koate DVI (Talecris, Research Triangle Park, NC), and
• vWD is divided into several subtypes. Type 1 is the most
Wilate (Octapharma, Lachen, Switzerland), with other simi- common, encompassing two-thirds of cases.
lar products available in other countries. These plasma- • Laboratory diagnosis of vWD may be difficult, especially in
derived concentrates contain vWF and FVIII in varying type 1.
ratios and with variable amounts of multimer size or distri- • vWD treatment is based on the subtype; the most common
bution. Humate-P, Alphanate, and Wilate are approved by agents used for treatment include desmopressin, antifibrinolyt-
the US Food and Drug Administration (FDA) for the treat- ics, hormonal therapy for heavy menstrual bleeding, and vWF
ment of vWD. Although these products are manufactured concentrates for severe bleeding or in types 2 and 3.
via processes that include viral attenuation and inactivation
steps, a theoretic risk of transmission of infectious agents
Disorders of secondary hemostasis
exists. A new recombinant vWF is under FDA licensure
review. Congenital hemophilia A and B (FVIII and FIX
Antifibrinolytic agents are useful adjunctive therapies and deficiency)
are used in a similar fashion as described for platelet defects.
Pathophysiology
Contraceptive agents including oral and levonorgestrel-IUD
can be effective therapies for the management of heavy men- The previous review of the physiology of hemostasis reveals
strual bleeding. Topical measures also are useful in some situ- the critical roles played by FVIII and FIX in thrombin gen-
ations. The benefits and risks of these agents are identical to eration and ultimately normal fibrin clot formation. Absence
those described in the “Treatment” section of the “Platelet or decreased amounts of either FVIII or FIX results in
function disorders” section in this chapter. Case reports exist reduced thrombin generation on the surface of activated
in the literature regarding the use of rFVIIa in vWD; these are platelets at injured sites. Inadequate thrombin generation
limited to patients with type 3 disease with inhibitors to vWF. leads to fibrin clots with poor structural integrity as visual-
ized by electron microscopy; specifically formation of large,
coarse fibrin strands as opposed to normal thinner strands
Gaps in knowledge
that form a tight network are observed. In addition, reduced
The most challenging aspect in the management of vWD is thrombin generation results in decreased activated FXIII,
the establishment of an accurate diagnosis, particularly in which is required for cross-linking of fibrin monomers and

Disorders of secondary hemostasis | 233
decreased TAFI generation, both of which result in a clot less Severe hemophilia, defined as a factor activity level <1%,
resistant to normal lysis. Therefore, deficiencies of FVIII or may present in the newborn period with intra- or extracra-
FIX result in poorly formed clots that are more susceptible to nial bleeding; prolonged bleeding from venipuncture or heel
fibrinolysis, clinically observed as the bleeding manifesta- stick or after circumcision; or with excessive bruising. Infants
tions in hemophilia. with severe hemophilia who do not develop symptoms in the
newborn period often present during the first year of life
with abnormal bruising, muscle hematoma (especially with
Etiology
immunization), or bleeding in the joint or muscle due to
Congenital deficiencies of FVIII and FIX occur as a result of activity or intercurrent injury. Although the precise preva-
genetic mutations in F8 and F9, respectively, both located on lence of intracranial hemorrhage is not known, it likely
the long arm of the X chromosome. Accordingly, these defi- approximates 1%-3%. Assisted delivery is associated with
ciencies are most commonly observed in males due to their bleeding in the neonatal period and maternal awareness
hemizygous state. In women and girls, a range of factor levels of carrier status is associated with lower use of assisted
can be observed, though rarely, levels in the severely or delivery devices (forceps/vacuum) and, in turn, lower rate
moderately deficient range can occur because of skewed of intracranial hemorrhage. Moderate hemophilia (factor
X-chromosome inactivation or the presence of other genetic activity levels between 1% and 5%) has a variable age of pre-
abnormalities such as Turner syndrome or X-autosomal sentation; diagnosis may be established due to a known
translocations. A wide range of mutations result in hemo- family history, in the newborn period due to bleeding, or
philia, and the mutation type (deletion, inversion, missense, later in life, even as an adult, with a bleeding event associated
or nonsense) and specific area of the protein affected deter- with intercurrent injury or invasive procedure. Bleeding
mines the severity of disease. In approximately 25% of cases, symptoms include deep tissue, muscle, or joint bleeding;
no family history is identified. In such cases, the affected mucocutaneous bleeding is a common presentation due to
individual’s mother is either not a carrier and the de novo fibrinolysis in the oropharynx and the inability to form a
mutation arose after conception of the affected male child or stable clot. Mild hemophilia (factor activity levels between
is a carrier as a result of a germline mutation at the time of 5% and 40%) may be diagnosed at ages similar to moderate
the mother’s conception. hemophilia. In the absence of a family history, patients with
Although over 2,100 unique mutations have been associ- mild hemophilia typically present later in childhood or dur-
ated with hemophilia A, the most common mutation in ing the teenage or adult years with bleeding associated with
patients with severe hemophilia A, occurring in up to 45% of injury or surgery rather than spontaneous.
patients, is the intron-22 inversion. The inversion is caused
by homologous recombination between the 9.5 kb sequence
Diagnosis
within intron 22 of the F8 gene and one of two extragenic
homologous regions. As a result, exons 1 to 22 are inverted In the absence of a known family history of hemophilia, in
and separated from exons 23 to 26. A wide variety of caus- which case direct assessment of FVIII or FIX in accordance
ative F8 gene mutations have been reported including small with the family history is most appropriate, the laboratory
and large deletions, and missense, nonsense, and splice site diagnosis of hemophilia begins with screening coagulation
mutations in nearly all of the coding areas of F8. Over 1,100 studies, including the PT and aPTT; the aPTT is almost
unique mutations have been associated with hemophilia B. In always abnormal. However, it is important to be cognizant
contrast to hemophilia A, there is not one predominant gene of circumstances in which the aPTT may be normal, espe-
defect in hemophilia B; missense mutations predominate. cially in mild deficiencies (Figure 9-7). After identification of
a prolonged aPTT, a mixing study with normal plasma is
performed. Correction of the prolongation into the normal
Clinical presentation
range points to a factor deficiency. Specific factor assays are
The clinical presentation of congenital hemophilia is highly used to identify the deficient factor. In the setting of an iso-
variable and is correlated with the level of deficiency. In lated prolonged aPTT, FVIII, FIX, and FXI should be assayed.
infants born to known carriers, the diagnosis most often can There are two methods by which the factor activity can be
be established at birth by assaying FVIII or FIX from umbili- measured; a one-stage assay and a two-stage chromogenic
cal cord blood. Prenatal testing is available if the genetic assay. The one-stage assay is based on the principles of the
defect has been identified within the family. The presenta- aPTT in which plasma is combined with phospholipid, cal-
tion of symptoms leading to diagnosis in patients either cium, and a contact activator (kaolin, silica, ellagic acid, etc.)
without a family history or not tested at birth is quite vari- and the time it takes to form a fibrin clot is measured. When
able and dependent on the severity of disease. measuring specific factor activity in a one-stage assay, the test

discussed, appropriate specimen procurement and handling

XII are critical to obtain accurate results. In newborns where
Prekallikrein cord blood is tested due to a known family history, levels
HMW-kininogen
Prothrombin may be altered based on sample procurement, level of defi-
time
XI ciency, and neonatal variations, as seen with decreases in
vitamin K-dependent clotting factors. Therefore, repeat test-
IX
VII VIII
ing may be required based on cord blood results and their
concordance with expected results and clinical symptoms. In
X
Partial
addition, assaying factor activity levels at the lowest range of
V
thromboplastin the curve is technically difficult, and sample analysis through
II time a reference laboratory may aid differentiation of the severe
from moderate forms. Finally, because FVIII is an acute-
Fibrinogen Fibrin
phase reactant, obtaining a true baseline level may be diffi-
Thrombin time
cult in patients with moderate and mild deficiencies when
tested in the setting of inflammation.
Clot XIII
stability Stable fibrin
test clot Treatment
The mainstay of hemophilia treatment is replacement of

Figure 9-7 Plasma coagulation reactions in in vitro laboratory assays.
Factor XII, prekallikrein, and high-molecular-weight kininogen are
the deficient coagulation factor. There are a number of
required for a normal activated partial thromboplastin time but not commercially available factor concentrates to treat both
for normal in vivo hemostasis. This diagram outlines the coagulation FVIII and FIX deficiency (Tables 9-2 and 9-3). The choice
factors required for each of four basic tests that characterize the of the specific product used includes consideration, among
coagulation cascade (PT, aPTT, thrombin time, and FXIII assay). other things, of availability, cost, method of manufacture,
delivery system, and half-life. Theoretical concerns include
plasma is serially diluted in plasma that is deficient in the infectious agent transmission and inhibitor development.
clotting factor of interest (so that all other clotting factors are Both recombinant and plasma-derived products are
in excess), and an aPTT is performed. The results should available, and decisions of which product to use should be
form a line that is parallel to a line made for the standard made in consultation with the patient and family. Recently,
reference sample from which the level of the factor of interest FVIII and FIX products have undergone modifications to
is determined. The two-stage method involves an incubation extend the half-life such as fusion to FC fragment, albumin,
step to generate FXa and a second step to measure the FXa, or polyethylene glycol (PEG). Fusion with an FC fragment
typically by its cleavage of a chromogenic substrate rather or albumin takes advantage of the FC receptor and leads to
than formation of a clot. The amount of FXa generated is recirculation of the fused FVIII or FIX. These alterations
proportional to the amount of FVIIIa/FIXa available. Dis- have led to an approximately 1.5-fold increase in FVIII
crepancies between these two measures have been identified half-life, though interindividual variation exists, and a
in up to a third of patients with mild and moderate hemo- four- to fivefold increase in the FIX half-life. Pegylation
philia A; both low one-stage assay compared with the two- increases the size of the molecule in circulation, which has
stage assay and vice versa. The reason for the discrepancy lies also been a more effective strategy for FIX than FVIII.
in the underlying F8 mutation. Mutations that affect FVIII Efforts to extend the half-life of clotting factors have been
cleavage by thrombin will result in higher FVIII levels in the less successful with FVIII than FIX because FVIII is already
two-stage assay because the incubation period allows for a large molecule that is predominantly intravascular and
more FVIII cleavage by thrombin to occur. Conversely, because FVIII clearance is dictated by vWF. This group of
mutations that reduce FVIII stability have lower two-stage products will be referred to as extended half-life (EHL)
FVIII levels compared with one-stage levels due to degrada- products.
tion of the FVIII during the incubation period. As a result, For all products 1 IU/kg of FVIII will typically increase the
some have urged that both assays are performed for the diag- FVIII level by 2%. Infusions of nonmodified FVIII products
nosis of hemophilia A. At a minimum, one needs to consider can be repeated as needed approximately every 8-12 hours.
performing the alternative FVIII assay if the clinical picture Dosing of EHL products during acute bleeding events will
is not consistent with the assay result. vary and should be performed according to prescribing
Ultimately, the type and level of severity of hemophilia are information. With FIX, dosing depends on the product used.
established based on the FVIII or FIX assay. As previously Plasma-derived FIX (pdFIX) and FIX-FC both increase the

Table 9-2 Factor VIII concentrates currently available in the United States
Brand name Generation Pd vs R Presence of human proteins Stability at RT
Monoclate NA Pd Albumin No
Hemophil M NA Pd Albumin No
Recombinate 1st R Albumin No
Kogenate FS/Helixate FS 2nd R Albumin in processing, not final product Yes: 3 months
Advate 3rd R None Yes: 6 months
Xyntha (B-domain deleted) 3rd R None Yes: 6 months
Eloctate (FVIII-FC) R None Yes: 6 months
Other FVIII concentrates approved for use in FVIII deficiency; these may also contain vWF and are not in general use.
NA = not applicable; RT = room temperature; PD = plasma derived; R = recombinant.
FIX level by 1% for each IU/kg infused, whereas with rFIX, disease seen on examination or on imaging studies. Studies
the level increases by 0.6%-0.8%, with children exhibiting a of tertiary prophylaxis have demonstrated improvements in
lower recovery compared with adults. FIX (rFIX and pdFIX) bleeding frequency, quality of life, and joint examination.
doses can be repeated every 12-24 hours as needed for acute Joints with repeated bleeding develop acute or chronic syno-
bleeds. vitis, followed by articular damage; the process of repeated
Treatment approaches are divided into two main cate- bleeding (three or more bleeds during a 6 month period) in
gories: prophylaxis and on-demand. Prophylaxis is the a joint is termed target joint. The bleeding pattern in target
regular infusion of factor replacement therapy to prevent joints has been documented to be amenable to prophylaxis.
bleeding events. Primary prophylaxis is factor replace- For those unable to complete every other day or thrice weekly
ment therapy started before or shortly after the first hem- prophylaxis, options include initiation of once weekly pro-
arthrosis and before the age of 3 years, has been proven to phylaxis with a plan to escalate in response to bleeding.
be the most effective approach to prevent the develop- Administration of factor concentrates to prevent bleeding
ment of joint disease. Therefore, primary prophylaxis only prior to circumstances that places patients at high risk
should be considered the optimal therapy for severe for bleeding, such as before sports, may be useful in those
hemophilia. Prophylaxis is time and resource intensive unwilling to do continuous prophylaxis or for those with
and requires adequate venous access often necessitating a non-severe disease who historically bleed when participating
central venous catheter. Once primary prophylaxis has in these types of activities. In other situations, limited pro-
been instituted, most will need to continue indefinitely. A phylactic therapy is reasonable and is reviewed in cited
subset of patients (~25%) may be able to discontinue pro- references.
phylaxis in adulthood while maintaining a low rate of Although primary prophylaxis is used most frequently in
bleeding. patients with severe disease, some individuals with moderate
Secondary prophylaxis is the regular infusion of factor hemophilia require this therapy because of their bleeding
replacement initiated after the second hemarthrosis but pattern. Issues related to prophylaxis include adherence,
before the presence of joint disease on physical examination cost, and the need for adequate venous access; prophylaxis
or imagining studies. The goal is to interrupt a bleeding pat- and the associated issues have been reviewed. Several pro-
tern or prevent joint damage through suppression of bleed- phylactic and general treatment-dosing approaches exist and
ing episodes. Tertiary prophylaxis is when regular infusions are detailed in Table 9-4. Although prophylaxis is effective in
of factor replacement are started after the onset of joint the prevention of the majority of spontaneous bleeding
events, patients who experience breakthrough-bleeding epi-
sodes require immediate treatment according to the recom-
Table 9-3 FIX concentrates currently available in the United States mendations in Table 9-4.
Episodic treatment for bleeding episodes is referred to as
Presence of
Brand Pd vs R human proteins Stability at RT
on-demand therapy (ie, the use of factor replacement ther-
apy after bleeding occurs). This treatment approach does not
Mononine Pd FIX and others No require regular infusions with their associated issues and is
Alphanine Pd FIX and others No less expensive in the short run, but is ineffective at prevent-
Benefix R None Yes (for 6 months)
ing joint disease. This mode of therapy now is used primarily
Alprolix R (FC-fused) None Yes (for 6 months)
for patients with moderate and mild hemophilia due to the
Pd = plasma derived; R = recombinant. infrequency of bleeding events and the associated low risk of

joint disease. Rarely, patients with severe hemophilia will render standard treatment with replacement therapy ineffec-
have infrequent bleeding events and can be managed with tive and result in hemorrhagic episodes that are prolonged
on-demand therapy with or without prophylactic infusions and more difficult to control, with associated increased risk
prior to specific activities. The typical initial dosing for of morbidity and mortality. The incidence of inhibitors is
bleeding episodes can be found in Table 9-4. Infusion ther- between 20% and 35% in severe, previously untreated,
apy for hemophilia, regardless of the regimen used, is best FVIII-deficient patients, 13% in non-severe FVIII-deficient
delivered in the home setting to allow for prompt therapy patients who receive FVIII replacement therapy, and <5% in
(within 2 hours of bleeding onset). Family members and severe FIX-deficient patients. The present inhibitor preva-
patients should be trained to administer the factor replace- lence is approximately 10% in FVIII deficiency and 3%-5%
ment therapy at home via peripheral venipuncture without in FIX deficiency. Risk factors for inhibitor development are
the need for a medical facility. both characteristics of the patient and how factor replace-
Adjunctive therapy for hemophilia is similar to that dis- ment therapy is delivered. Because of the greater prevalence
cussed for platelet defects and vWD. Patients with mild of hemophilia A and higher incidence of inhibitors amongst
hemophilia A may be treated with desmopressin after a chal- patients with hemophilia A, more is known about risk fac-
lenge dose demonstrates a hemostatic response; the response tors for inhibitor development in this population. Among
level dictates the type of bleeding events that may be treated the patient-specific risk factors, the most important is hemo-
with this agent. Antifibrinolytic agents are efficacious for philia severity, with patients with severe disease at highest
mucosal bleeding and commonly are used in conjunction risk. The specific genetic mutation, ethnicity, and family his-
with factor replacement or desmopressin. For women with tory of inhibitors also have been shown to affect the expres-
hemophilia who experience heavy menstrual bleeding, hor- sion of this complication. Mutations resulting in major
monal therapy including the levonorgestrol IUD can be used disruptions of the F8/F9 genes, such as large deletions, are
as well as antifibrinolytic therapy. associated with increased risk. Genetic polymorphisms of
immune response genes (IL10, TNF-alpha, and CTLA-4)
Complications of treatment: inhibitors
have also been associated with inhibitor risk. In addition,
A significant complication of hemophilia after exposure to patients of African or Hispanic ethnicity have a significantly
replacement therapy is the development of neutralizing higher rate of inhibitor development. Treatment-related risk
antibodies that bind FVIII/FIX termed inhibitors. Inhibitors factors have been purported to include the source of the
Table 9-4 Typical dosing for FVIII and FIX deficiency in different clinical circumstances*
Factor Joint/muscle Life or limb threatening Preoperative Prophylaxis
FVIII 25 IU/kg; repeat as needed 50 IU/kg; multiple doses required 50 IU/kg 20-40 IU/kg three times
weekly or every other
day
FVIII-FC 20-30 IU/kg; repeat every 40-50 IU/kg; repeat every 12-24 40-60 IU/kg followed by a repeat 50 IU/kg every 4 days
24-48 hours (12-24 for hours (8-24 hours for patients dose of 40-50 IU/kg after 8-24
patients <6 years of age) less than 6 years of age) until hours (6-24 hours for patients
as needed bleeding resolved less than 6 years of age)
and then every 24 hours to
maintain FVIII activity within
target range
pdFIX 50 IU/kg; repeat as needed 100 IU/kg; multiple doses 100 IU/kg 50 IU/kg twice weekly
required
rFIX 26-70 IU/kg; repeat as 130-140 IU/kg; multiple doses 120-140 IU/kg 60-70 IU/kg twice weekly
needed required
FIX-FC 30-60 IU/kg; repeat every 80-100 IU/kg; repeat dose after 60-100 IU/kg; repeat dose after 50 IU/kg once weekly, or
48 hours as needed 6-10 hours, then every 24 6-10 hours, then every 24 100 IU/kg once every
hours for 3 days, then every 48 hours for 3 days, then every 48 10 days
hours until healing achieved hours until healing achieved
*Represent general dose guidelines, practice varies. The volume of distribution of infused rFIX is ~1.2 in adults and ~1.4 in children.
Therapy duration for intracranial hemorrhage varies but is minimally ~2 weeks; in children intracranial hemorrhage should prompt strong
consideration for ongoing prophylactic therapy.

factor product used (plasma derived vs. recombinant), in nonactivated and activated forms. The mechanism of
though these data remain controversial. Other treatment- action of APCC largely is ascribed to the presence and action
related risk factors include receipt of intensive replacement of FXa and prothrombin, although FIXa and FVIIa also are
therapy (5 or more consecutive days) or surgery during early contained; small quantities of nonactivated FVIII may be
factor exposure. Inhibitor development in hemophilia B is present. rFVIIa contains FVIIa as its sole agent and is geneti-
far less common and has associated unusual complications. cally engineered. The main mechanism of action of rFVIIa in
Patients with FIX deficiency may develop anaphylactoid patients with hemophilia is through tissue-factor indepen-
reactions to infused FIX concentrate before or at the time of dent thrombin generation on the surface of activated plate-
inhibitor emergence. lets. Both APCC and rFVIIa have been demonstrated to be
Inhibitors are divided into two categories: low-titer (also safe and effective, with variable response rates ranging from
known as low-responding inhibitors) and high-titer (high- 70% to 90%. Two prospective studies compared these prod-
responding inhibitors). A low-responding inhibitor is char- ucts and revealed essentially similar response rates. Both
acterized as one with a titer, measured in the Bethesda assay products have considerable data supporting their safety (>30
of <5 Bethesda units (BU)/mL despite repeated exposure to years for APCC and >10 years for rFVIIa) with few reported
factor replacement, whereas high-responding inhibitors are thrombotic events in hemophilic inhibitor patients. In addi-
those that achieve a titer >5 BU/mL at any time regardless of tion, APCC as a plasma-derived product has an excellent
present titer. Patients with high-responding inhibitors may safety record without documented viral transmission.
exhibit a decrease in or an undetectable inhibitor titer with The most important consideration when choosing a prod-
complete withdrawal of the specific clotting factor. Despite uct in an inhibitor patient is its ability to achieve rapid bleed
this, with subsequent exposure to the deficient factor, these control and thereby limit morbidity and mortality. Thus,
patients mount a memory response and will demonstrate an product choice is individualized. Because APCC is an FIX-
increase in inhibitor titer in 7-10 days after exposure. Stimu- based product, its use in FIX inhibitor patients with infusion-
lation and increase of inhibitor titer is termed anamnesis. associated reactions is contraindicated. APCC may contain
Therefore, it is clear that high-responding inhibitor patients small quantities of FVIII and result in continued stimulation
who achieve an undetectable inhibitor titer have not had the of the inhibitor titer in FVIII-deficient patients. Accordingly,
inhibitor response ablated and should not be challenged rFVIIa which does not contain FVIII or FIX will not lead to
again unless experiencing life- or limb-threatening bleeding anamnesis and may be preferred if trying to allow the inhibi-
episodes or there is a plan for inhibitor eradication with tor to reach a low level before initiation of ITI. Management
immune tolerance induction (ITI). of acute bleeding is critical; therefore, inhibitor stimulation is
Patients with low-responding inhibitors commonly are not an absolute contraindication to APCC use during this
managed with higher doses of factor replacement therapy to time if any bleeding episode is unresponsive to rFVIIa. Dos-
overcome the inhibitor titer and achieve a hemostatic factor ing regimens for both products have been established (Table
level. Approximately 10% of low-titer inhibitors resolve 9-5). Occasionally, patients present with bleeding events
without intervention (often within a few weeks) and are refractory to both agents. In such cases, the use of combina-
termed transient inhibitors; therefore, ongoing measurement tion APCC and rFVIIa has been reported using an alternative
of titers is important to document persistence. sequential regimen. Alternatively, another approach is to
The three important strategies for the management of adjust APCC or rFVIIa dosing based on results of global
patients with high-responding inhibitors include: (i) man- hemostatic assays (thrombin generation assay and throm-
agement of bleeding episodes, (ii) prevention of bleeding, boelestography). Although these approaches have been dem-
and (iii) eradication of the inhibitor. The management of onstrated to be effective and safe in a small number of young
bleeding episodes in inhibitor patients is challenging, with children, the reports remain anecdotal.
the majority of hemophilia-related morbidity in the United Historically, the prevention of bleeding in inhibitor
States occurring in patients with high-responding inhibitors. patients was confined to prevention of bleeding during
Bypassing agents are used to treat bleeding episodes in invasive procedures. Because of obvious concerns for
patients with high-responding inhibitors, as they are not hemostatic control during surgery and postoperatively
able to achieve hemostatic clotting factor levels with factor with bypassing agents and concern for thrombotic events
concentrates. Two bypassing agents are available for the with repeated use in a high-risk setting, inhibitor patients
management of bleeding in inhibitor patients: activated pro- were not offered elective surgery until fairly recently. Few
thrombin complex concentrate (APCC) (FEIBA; Baxter, studies demonstrating successful hemostatic strategies for
Deerfield, IL) and rFVIIa (NovoSeven; Novo Nordisk, Bags- inhibitor patients in the surgical setting have been per-
vaerd, Denmark). APCC is a plasma-derived concentrate formed. Over the past decade, several prospective studies
consisting of the vitamin K-dependent clotting factors both have demonstrated the successful use of rFVIIa for both

Table 9-5 Typical dosing for currently available bypassing agents

Agent Joint/muscle Life or limb threatening Preoperative Prophylactic
APCC* 50-75 U/kg 75-100 U/kg 50-100 U/kg‡ 85 U/kg every other day
Repeat every 8-12 hours as needed Repeat every 12 hours
rFVIIa 90-120 μg/kg 90-120 μg/kg 90-120 μg/kg 90 μg/kg/day
Standard dose† Repeat every 2-3 hours as needed Repeat every 2-3 hours Repeat every 2 hours‡
rFVIIa 270 μg/kg 270 μg/kg No data 270 μg/kg/day
High dose Data not available on follow-up Data not available on follow-up
doses required doses required
*Doses >200 U/kg/day contraindicated per prescribing information.
†The licensed dose of rFVIIa in the United States is 90-120 μg/kg for treatment and prevention of bleeding during surgery; not approved for
prophylaxis.
‡Frequency and duration vary according to the type of surgery. Refer to prescribing information.
minor and major surgery (see Table 9-5 for dosing recom- 100 IU/kg given once daily. Clinical studies have identified
mendations). This has led to an increased availability of several factors associated with ITI success, including the
required surgical procedures in inhibitor patients, most historical peak inhibitor titer (<200 BU/mL), titer at start
notably orthopedic procedures for amelioration of hemo- of therapy (<10 BU/mL), peak titer after the start of ITI
philic arthropathy. APCCs have been used in the surgical (<100 BU/mL), age at initiation (<8 years), and time from
setting, but the body of reports supporting their use, dos- inhibitor development to ITI start (<2 years). It is best to
ing, and safety is smaller compared with rFVIIa. It is impor- initiate ITI when the titer is <10 BU/mL, although this must
tant to mention that the risk of thrombosis may increase be balanced against the risk of delaying tolerance and per-
with the sequential use of rFVII and APCCs. sistent risk of bleeding. For those who fail an initial course
Recently, prophylaxis with bypassing agents to prevent of ITI, the rate of success with a second ITI course is
bleeding episodes in inhibitor patients has become more unknown. Options to consider include using a vWF-
routine. Several studies have demonstrated its utility. APCC containing FVIII product or adding rituximab, though
(85 U/kg 3-3.5 times per week) has demonstrated a 62%- clear evidence to guide treatment decisions is lacking. The
72.5% reduction in the frequency of bleeding events. How- best approach to inhibitor eradication in patients with
ever, the response was variable amongst patients with up to nonsevere hemophilia is also unclear. In general, patients
38% having minimal to no change in bleeding frequency. with nonsevere hemophilia do not respond to ITI as well as
rFVIIa prophylaxis was studied at two doses, 90 μg/kg daily patients with severe disease. Rituximab without ITI has also
and 270 μg/kg daily, and led to a 47% and 68% reduction in been used and may lead to more rapid inhibitor eradication
bleeding frequency. Although prophylaxis with bypassing than observation alone.
agents has demonstrated benefit, its benefit is less than that Because of the associated risk of allergic reactions in
seen with tertiary prophylaxis using factor replacement ther- patients with hemophilia B, ITI may not be possible or, if
apy in non-inhibitor patients and is much more difficult to undertaken, requires desensitization to FIX. FIX-deficient
achieve given that APCC is typically a large infusion volume patients with inhibitors undergoing ITI are at risk for devel-
and rFVIIa is dosed frequently. For these reasons, inhibitor oping nephrotic syndrome. ITI-associated nephrosis is more
eradication is important and new therapies are needed for likely to occur in patients with a history of an anaphylactoid
those that fail ITI. reaction. The etiology of nephrosis in these patients is
Inhibitor eradication with ITI requires regular adminis- unclear, although it is thought to be related to immune com-
tration of the deficient factor to reset the immune system plex formation. The overall success rate of ITI in FIX defi-
by inducing peripheral tolerance. An international pro- ciency is 35%, far lower than the 75% achieved in FVIII
spective ITI study in good-risk patients recently was com- deficiency. Thus, although fewer FIX inhibitor patients exist,
pleted and published (Hay et al., 2012). This study they are a significant treatment challenge for practitioners.
compared daily high-dose FVIII (200 IU/kg) daily to low-
dose FVIII (50 IU/kg) three times weekly. The study was
Prognosis and outcomes
stopped before reaching the planned endpoint because of
an increased rate of bleeding observed in patients receiving Currently, patients with severe hemophilia without inhibi-
FVIII 50 IU/kg thrice weekly. Typical ITI regimens may tors, HIV, or HCV treated on a prophylactic regimen have an
include either of these infusion schedules or a regimen of excellent prognosis and lead near-normal lives possibly even

without the development of hemophilic arthropathy. The

Swedish cohort followed for nearly 40 years substantiates Key points
these outcomes. For patients with inhibitors, the outcome is • Hemophilia is an X-linked disorder resulting from deficiencies
more variable and the risk of morbidity is significant. When of FVIII or FIX and is categorized as mild, moderate, and severe
ITI is successful, the outcome can be converted to that of a depending on the factor level.
noninhibitor patient, yet the morbidity experienced depends • Patients with severe hemophilia are at risk for the develop-
on the amount of joint disease and other bleeding events that ment of joint disease termed hemophilic arthropathy that can be
occurred before ITI success. It is likely that many of these prevented by regular prophylactic factor infusions begun at an
early age.
patients will have experienced hemarthroses, muscle, or
• Factor replacement therapy is available to treat bleeding
even intracranial hemorrhage and that some of these bleed-
episodes and is highly effective.
ing events will be associated with permanent sequelae. For
• Patients with hemophilia, most notably those with severe
inhibitor patients in whom ITI was not successful or not per- disease, may develop neutralizing antibodies directed against
formed, significant musculoskeletal morbidity is common, the deficient factor-termed inhibitors; inhibitors are divided into
resulting in permanent disability and poor quality of life. high- and low-responding types, and the presence of an
With improved hemostatic coverage available for surgical inhibitor may render replacement therapy ineffective.
interventions, even hemophilic patients with inhibitors now • Inhibitors can be eradicated with ITI in approximately 70% of
may undergo procedures to reduce pain and increase func- patients with hemophilia A.
tionality. Combined with the increased use of prophylaxis, it • Patients with high-responding inhibitors are infused with
is possible now to develop treatment strategies to ameliorate bypassing agents to treat or prevent bleeding episodes; overall,
the consequences of recurrent bleeding and allow patients to bypassing products are not as effective as standard factor
replacement in noninhibitor patients, and as such, inhibitor
lead more productive lives.
patients have an increased risk of hemorrhage-associated
morbidity and mortality.
Gaps in knowledge
The greatest challenge with the potential for significant

reward lies with gene therapy, a potentially curative Acquired hemophilia
approach. Early clinical trials using adenoviral vectors have Pathophysiology and etiology
demonstrated the ability of FIX gene transfer to modestly
increase FIX levels and reduce bleeding and factor consump- Rarely, hemophilia can be acquired as a result of the develop-
tion. Development of improved therapeutic approaches for ment of autoantibodies most commonly directed against
inhibitor patients who still experience increased morbidity FVIII and is referred to as acquired hemophilia. It has been
and mortality compared with noninhibitor patients are associated with a variety of conditions, including pregnancy,
required. One approach deserving of future work is the pre- malignancies, and autoimmunity. In ~50% of cases, no
vention of inhibitor formation. An improved understanding known associated disorder can be identified. Overall, the
of the immunologic pathways involved in inhibitor forma- annual incidence is 1.4 per million, though the frequency
tion and development of tolerance would open avenues to increases with age with the median age of onset approxi-
prevent inhibitor development or increase the rate of toler- mately 77 years. The anti-FVIII autoantibodies inhibit the
ance achieved. It is conceivable that an approach could be functional activity of endogenous FVIII, resulting in a bleed-
developed to program the immune system to induce toler- ing diathesis. Although some bleeding symptoms are similar
ance before or in association with exposure to exogenous to congenital hemophilia, the incidence of hemarthroses in
normal factor concentrate. Future research efforts could lead acquired hemophilia is low, whereas soft tissue, abdominal,
to the development of replacement products that are less or and retroperitoneal hemorrhage are more frequent. Addi-
perhaps not at all immunogenic. In inhibitor patients, meth- tionally, bleeding in patients with acquired hemophilia may
ods to perform ITI in FIX deficiency lag behind those for be more severe than is seen in congenital hemophilia despite
FVIII deficiency. For patients with anaphylactoid reactions, similar FVIII levels.
options for desensitization and subsequent ITI are limited,
with an overall poor outcome, although rare success has
been reported. The FIX-deficient inhibitor population with
anaphylactoid reactions represents a small vulnerable popu- Acquired hemophilia may present with the dramatic onset of
lation with only one therapeutic agent presently available for either mucocutaneous or internal bleeding. Life-threatening
the management of bleeding episodes; new approaches and bleeding, such as gastrointestinal, intracranial, or large mus-
treatments clearly are required. cle hematoma with associated compartment syndrome is

observed. Hemarthroses are uncommon. In the era of bypass Rare factor deficiencies
therapy, fatal bleeding is reported in 3%-9% of patients.
Deficiencies of other coagulation factors that play a role in
thrombin generation, cross-linking, and stabilization of the
Diagnosis fibrin clot or down-regulation of fibrinolysis may lead to a
bleeding diathesis. Deficiencies of fibrinogen, factor II (FII),
Acquired hemophilia should be suspected in patients that
FV, FVII, FX, and FXIII result in bleeding disorders in cases
present with bleeding symptoms and a prolonged aPTT.
in which the severity of the bleeding is loosely related to the
Rarely, patients will present with an asymptomatically pro-
factor levels (Table 9-6). Although FVIII and FIX deficiency
longed aPTT. Thus, evaluation of a prolonged aPTT in an
are defined as rare disorders affecting approximately 20,000
adult should include a mixing study and FVIII level regard-
Americans, deficiencies of these other coagulation factors
less of the presence or absence of bleeding. Acquired FVIII
are far less common. Therefore, the clinical presentation
inhibitors are typically time- and temperature-dependent
related to any specific level and the range of symptoms expe-
which translates into a normal immediate mixing study that
rienced are less well described than in hemophilia A and B.
then fails to correct with mixing after incubation.
For detailed discussion of these disorders, see the special
issue of the British Journal of Haematology (volume 167,
Treatment issue 3, November 2014).
The management of bleeding episodes in acquired hemophilia is
similar in many respects to that of congenital hemophilia with Fibrinogen deficiency
inhibitors, and the principles outlined earlier largely apply. An
Pathophysiology
exception of note is that patients with acquired hemophilia often
are elderly and at increased risk for thrombosis; thus, bypassing As discussed in the overview of hemostasis, fibrinogen is
agents, although often required for control of bleeding, may have cleaved by thrombin to form fibrin, which then polymerizes
an associated higher rate of thrombotic complications. Recom- to form tight thin strands in a meshwork that form the clot
binant porcine FVIII (Obizur; Baxalta, Westlake Village, CA) (Figure 9-3).
was recently approved for treatment of bleeding in patients with
acquired FVIII inhibitors. It was shown to be effective at reduc-
Etiology
ing bleeding in 28 patients treated as part of a phase 2/3 study.
The starting dose is 200 IU/kg with subsequent dosing and fre- Congenital deficiencies of fibrinogen are due to defects in
quency titrated based on FVIII levels. FVIII levels are recom- the genes (FGA, FGB, FGG) that code for of one of three
mended to be performed 30 minutes and 3 hours after the initial fibrinogen protein chains (Aα, Bβ, and γ) and can be inher-
dose and 30 minutes after subsequent doses. Responses will vary ited both an autosomal dominant or recessive patterns.
according to the amount of anti-FVIII inhibitor that is present Deficiencies can be complete (afibrinogenemia) or partial
and the degree to which it is cross-reactive with porcine-FVIII. (hypofibrinogenemia) and associated with dysfunction
Strategies to promote inhibitor eradication in acquired hemo- (hypodysfibrinogenemia). Acquired causes of hypofibrino-
philia are different than in congenital hemophilia with inhibitor. genemia include liver disease, use of chemotherapeutic
Because acquired hemophilia is due to the development of auto- agents such as l-asparaginase, and the Kasabach-Merritt
antibodies that result from loss of self-tolerance, they tend to syndrome (hemangioma with consumptive coagulopathy).
respond to immunosuppressive medications effective in auto- Other consumptive processes such as disseminated intra-
immune disorders in general. Corticosteroids are considered vascular coagulation lead to multiple coagulation factor
first line therapy and should be used even in patients without deficiencies in addition to fibrinogen.
current bleeding symptoms. Patients with detectable FVIII levels
and inhibitor concentrations <20 BU/mL may respond to corti-
costeroids alone. Patients with inhibitor titers >20 BU/mL are
less likely to respond to corticosteroids alone and cyclophos Bleeding can be variable with potential sites being mucocutane-
phamide should be added up front. The role of rituxmab in up ous, soft tissue, intracranial, umbilical stump traumatic, post-
front therapy remains controversial. Since patients with acquired surgical bleeding, and recurrent miscarriages. Particularly in the
hemophilia continue to produce their own FVIII, exogenous setting of a dysfibrinogenemia and hypodysfibrinogenemia,
administration, as is done with ITI, is not required for inhibitor thrombosis may be coexistent with bleeding. Although patients
eradication in this setting. Relapses occur in approximately with severe symptoms are more likely to have lower levels, the
10%-20% of patients, and most often occur during the first year, levels of those that are symptomatic and asymptomatic
thus ongoing monitoring is essential. overlap.

Table 9-6 Bleeding sites and symptoms and factor replacement choices for rare factor deficiencies
Factor deficiency (level

associated with major
bleeding)* Bleeding sites Other symptoms Factor replacement Acquired deficiencies
Fibrinogen (<10 mg/dL) No typical sites Splenic rupture Fibrinogen concentrate: Liver disease
Miscarriage RiaStap Asparaginase therapy
Thrombosis Cryoprecipitate DIC
Factor II (<10%) No typical sites None PCC Vitamin K deficiency
Liver disease
Vitamin K antagonists
Antiphospholipid syndrome
Factor V (<1%) No typical sites None FFP Topical bovine thrombin exposure,
Platelet transfusion antibiotics
Factor VII (<10%) Intracranial Thrombosis rFVIIa Vitamin K deficiency
Liver disease
Factor X (<10%) Intracranial None PCC Vitamin K deficiency
Liver disease
Amyloidosis
Factor XI (no clear Surgery or injury None FFP Autoantibodies (rare)
association between related FXI concentrates available in
levels bleeding) some countries
Factor XIII (<5%) Intracranial Poor wound healing pdFXIII concentrate: Corifact Cardiopulmonary bypass
Umbilical stump Miscarriage rFXIII: Tretten Inflammatory bowel disease
RiaStap is licensed for congenital afibrinogenemia. Recombinant factor VIIa is licensed for the treatment of congenital FVII deficiency. Corifact
and Tretten are licensed for congenital FXIII deficiency. Prothrombin complex concentrates (PCC) not licensed for the treatment of rare factor
deficiencies and contain variable amounts of factors II, VII, and X with dosing based on FIX units.
DIC = disseminated intravascular coagulation; FFP = fresh frozen plasma; PCC = prothrombin complex concentrate.
* Official Communication of the Scientific Subcommittee on Rare Bleeding Disorders of the International Society of Thrombosis and
Haemostasis (ISTH).
Diagnosis every 2-4 days to maintain a fibrinogen level of >100-150

mg/dL as needed.
Diagnosis is suspected in the setting of both a prolonged PT
and aPTT. The thrombin time will also be prolonged and the
fibrinogen activity reduced. Once a low fibrinogen activity is Prothrombin deficiency
measured, a fibrinogen antigen should be measured. Afibrin-
Pathophysiology
ogenemia and hypofibrinogenemia have similar fibrinogen
activity and antigen levels, whereas patients with dysfibrino- As discussed in the overview of hemostasis, a small amount
genemia have higher antigen than activity levels. Molecular of thrombin is produced by the TF:FVIIa complex and is
testing can be performed to confirm that diagnosis, but it is needed for initiation of coagulation. A large burst of throm-
not routinely available. bin is produced by cleavage of prothrombin to thrombin by
the thrombinase complex on the activated platelet surface.
Treatment Thrombin has both pro- and anticoagulant functions.
Thrombin’s main procoagulant function is to cleave fibrino-
Fibrinogen concentrate is approved for treatment of fibrinogen to fibrin.
gen deficiency (afibrinogenemia and hypofibrinogenemia).
For treatment of bleeding or prior to invasive procedures,
Etiology
the dose of fibrinogen concentrate can be calculated as (tar-
get − baseline)/1.7 × weight in kg. The half-life of fibrinogen Both quantitative and qualitative defects of FII can be
is approximately 80 hours, thus repeat dosing can be given inherited as an autosomal disorder and its prevalence is

estimated to be 1 in 2,000,000. Acquired deficiencies of FII Clinical presentation

can occur in association with antiphospholipid syndrome
Bleeding symptoms are diverse with the most severe bleeding
as a result of anti-prothrombin antibodies that lead to
occurring in those with levels <1%, though there is overlap
increased prothrombin clearance, liver disease, or vitamin
of FV levels between those that are symptomatic and those
K deficiency.
that are asymptomatic.
Clinical presentation Diagnosis

There is a poor correlation between clinical symptoms and The laboratory investigation will demonstrate a prolonged
FII levels. Bleeding symptoms are reported more commonly PT and aPTT, and normal thrombin time. Other common
with FII levels <10% and may include mucocutaneous, soft pathway clotting factors (fibrinogen, FII, and FX) will be
tissue, joint, surgical, and menstrual bleeding. Heterozygous normal. FV activity using a one-stage PT-based assay will be
carriers typically are asymptomatic. low. FVIII activity should also be measured to differentiate
between combined FV and FVIII deficiency and isolated FV
Diagnosis deficiency.
The diagnosis should be considered in the setting of a pro- Treatment

longed PT and aPTT with a normal thrombin time (defi-
ciencies of FV and FX will have a similar pattern). FII activity There is currently no FV concentrate available, thus replace-
can be measured using a one-stage PT based assay. The pres- ment of FV at the time of bleeding or prior to invasive pro-
ence of antiphospholipid antibodies, liver failure, and vita- cedures requires infusion of fresh frozen plasma (FFP).
min K deficiency should also be evaluated when FII deficiency Additionally, platelet alpha-granules contain 20% of circu-
is suspected. lating FV; thus platelets can be a source of FV and used in
combination with FFP when FFP alone has been ineffective.
Treatment
In general, 15 mL/kg of FFP is estimated to raise the FV
activity 15%. The half-life of FV following FFP infusion is
Management of bleeding in patients with FII deficiency is 16-36 hours.
with prothrombin complex concentrates (PCC); 20-30 IU/
kg is expected to increase the plasma FII concentration by
Factor VII deficiency
40%-60%. Since FII’s half-life is 60 hours, doses of PCC can
be repeated every 2-3 days as needed Pathophysiology
Factor VII is a vitamin K–dependent protein, and 1% circu-

Factor V deficiency lates in the active form (FVIIa) available to bind exposed TF
Pathophysiology at sites of vascular injury.
Factor V acts as a cofactor for FX to potentiate FXa cleavage

Etiology
of prothrombin.
FVII deficiency is an autosomal recessive disorder with an
Etiology
estimated worldwide prevalence of 1 in 500,000.
Deficiency of FV is an autosomal disorder and is estimated

to have a prevalence of 1 in 1,000,000. FV deficiency can
also occur in combination with FVIII deficiency. Combined Patients with severe FVII deficiency, FVII <1%, are those
FV and FVIII deficiency is the result of mutations in two most likely to have significant bleeding, including intracra-
genes, LMAN1 and MCFD2. These genes encode for pro- nial hemorrhage. In an international registry, cases with
teins that participate in transporting FV and FVIII from the severe bleeding had levels 0%-21%, and those that were
ER to the Golgi necessary for normal secretion of FV and asymptomatic had levels 15%-35%.
FVIII into the circulation. Very rarely, acquired FV defi-
ciency can occur after exposure to bovine thrombin found
Diagnosis
in topical thrombin preparations and result in formation of
antibodies against bovine thrombin that cross-react with The laboratory picture of FVII deficiency is that of an iso-
human FV. lated prolonged PT with normal aPTT and thrombin time.

An isolated prolonged PT should prompt measurement of The half-life of FX is approximately 30 hours and thus repeat
FVII activity. Early vitamin K deficiency may predominately doses of PCC can be given every 1-2 days as needed to main-
affect the FVII activity. tain hemostasis. In the setting of AL amyloidosis, treatment
of the AL amyloid typically normalizes the FX level. If hemo-
Treatment static support is needed prior to treatment of the underlying
AL amyloidosis, as may occur with emergent surgery, PCC
Treatment of FVII deficiency is with rFVIIa 15-20 μg/kg with
20-30 IU/kg can also be used, though strong evidence to sup-
doses repeated every 6 hours as needed. Antifibrinolytics
port treatment decisions is lacking.
(aminocaproic acid or tranexamic acid) may also be useful
for minor bleeding, particularly in those with no or minimal
personal history of bleeding. FVII levels >30% are typically Factor XI deficiency
adequate for surgery. Pathophysiology
Factor XI is activated by thrombin after initiation of coagula-

Factor X deficiency tion. FXIa is then available to activate FIX on the activated
Pathophysiology platelet surface providing an amplification loop. FIX activa-
tion by FXIa is required for generation of a burst of throm-
Factor Xa, in conjunction with its cofactor, FVa, cleaves pro- bin that is adequate to activate TAFI. In the absence of
thrombin for thrombin. activated FXI to promote an adequate burst of thrombin,
clots are more susceptible to fibrinolysis.
Etiology
Congenital FX deficiency is an autosomal recessive disorder Etiology

with an estimated worldwide prevalence of 1 in 1,000,000. FXI deficiency can occur as both autosomal dominant and
Acquired FX deficiency can be seen in the setting of AL- recessive. Prevalence has been difficult to estimate due to
amyloidosis secondary to binding of FX to AL amyloid, variability of the clinical phenotype, but it is known to be
effectively removing it from circulation. higher in Jewish populations, where 1 in 11 are heterozygous
and 1 in 450 are homozygous or compound heterozygous.
Clinical presentation FXI is activated during the initiation phase by thrombin on
the platelet surface.
Abnormal bleeding in patients with FX deficiency can mani-
fest as mucocutaneous, soft tissue, or gastrointestinal bleed-
ing. Importantly, intracranial bleeding was reported in up to
21% of symptomatic cases. Severe bleeding symptoms are Bleeding after surgery or trauma is the most common mani-
more likely to occur in the setting of FX activity <10%. festation of FXI deficiency, as well as in sites where fibrinoly-
sis is active such as the gastrointestinal tract and urogenital
Diagnosis system. Spontaneous bleeding is uncommon. The clinical
phenotype is quite variable, and there is a weak correlation
Laboratory testing in patients with FX deficiency will show a between FXI level and bleeding.
prolonged PT and aPTT, but normal thrombin time similar
to FII and FV deficiency. FX activity is typically measured
Diagnosis
using a one-stage PT-based assay. The choice or thrombo-
plastin used in the assay may influence the FX activity result. Laboratory testing in the setting of FXI deficiency demon-
Congenital FX deficiency is distinguished from acquired FX strates a prolonged aPTT and normal PT and thrombin
deficiency secondary to AL-amyloidosis on clinical grounds. time. FXI activity is measured using a one-stage aPTT assay.
Mixing studies in the setting of AL amyloidosis will demon-
strate correction and appear consistent with a deficiency. Treatment
Since the FXI level is such a poor predictor of bleeding, the

Treatment
presence or absence of bleeding with prior traumatic events
Treatment of FX deficiency is with prothrombin complex or invasive procedures should be considered when determin-
concentrates (FX concentrates may be available in some ing bleeding risk and need for treatment. Antifibrinolytic
countries). A typical dose of PCC is 20-30 IU/kg and would therapy with tranexamic acid or aminocaproic acid should
be expected to increase the plasma FX activity 40%-60%. be a mainstay of treatment. For persons with FXI levels

<10% or with a personal history of bleeding, replacement of plasma-derived FXIII (pdFXIII) (Corifact; CSL Behring)
FXI using FFP 15-25 mL/kg can be considered for severe and a recombinant FXIII (rFXIII) (Tretten; Novo Nordisk).
bleeds or major surgery. FXI concentrate is available in some pdFXIII contains both subunits, whereas rFXIII contains
countries, but not in the United States. Alloantibodies against only subunit A. Accordingly, patients with subunit B defi-
FXI (FXI inhibitor) have been reported to occur following ciency should be treated with pdFXIII and not rFXIII.
replacement therapy.
Vitamin K-dependent coagulation factor deficiency
Factor XIII deficiency
Pathophysiology
Pathophysiology
Factors II, VII, IX, and X are vitamin K dependent clotting
FXIII circulates as a heterotetramer with two catalytic A sub- factors. During synthesis, they undergo γ-glutamyl carboxyl-
units and two carrier B subunits. FXIII is activated by throm- ation by γ-glutamyl carboxylase and the cofactor vitamin K
bin and once activated covalently crosslinks fibrin. hydroxyquinone (KH2). During γ-carboxylation, KH2 is
oxidized to vitamin K 2,3 epoxide that then undergoes de-
Etiology epoxidation by VKOR to restore KH2.
Congenital factor XIII is an autosomal recessive disorder
with a worldwide prevalence estimated to be 1 in 2,000,000. Etiology
Acquired FXIII deficiency can occur in the setting of cardiac
Vitamin K-dependent coagulation factor deficiency (VKD-
surgery, malignancy, infection, and inflammatory bowel dis-
CFD) is an autosomal recessive disorder that has been
ease. FXIII can be caused by mutations in the genes that code
reported to occur in fewer than 30 families worldwide. It is
for either the catalytic A subunit or the B carrier subunit,
caused by defect in the γ-glutamyl carboxylase protein or in
though mutations in subunit B are reported to account for
subunit 1 of VKOR protein and leads to deficiencies of vita-
<5% of cases of congenital factor XIII deficiency.
min K-dependent clotting factors; FII, FVII, FIX, and FX.
The clinical phenotype is similar regardless of the subunit
affected. The most common sites of bleeding are umbilical Clinically, VKDCFD presents at birth with intracranial or
stump, soft tissue, surgical, and intracranial. In addition to umbilical bleeding or early childhood with joint, mucocuta-
bleeding, poor wound healing is often present and preg- neous, or soft tissue bleeding.
nancy loss can occur. Heterozygous carriers may have FXIII
activity levels as low as 20% and may display mild bleeding Diagnosis
symptoms.
Factors II, VII, IX, and X will be reduced. Distinguishing
VKDCFD from acquired vitamin K deficiency requires dem-
Diagnosis
onstration of a normal fasting KH2 concentration.
Laboratory diagnosis requires measurement of FXIII activ-
ity, as typical screening tests such as the PT, aPTT, and Treatment
thrombin time will be normal. Qualitative assays for FXIII
activity (clot solubility) are only abnormal with levels <5%. Treatment with oral or parenteral vitamin K1 has been
Quantitative assays are also available and can detect abnor- shown to partially or completely restore coagulation factor
mal FXIII levels despite a normal clot solubility test. Genetic activities and is the mainstay of long-term therapy for pre-
analysis is the most effective means to determine if subunit A vention of bleeding.
or B is affected.
Gaps in knowledge
Treatment
Large, well-designed prospective studies of congenital rare
Given the high rate of intracranial hemorrhage, prophylaxis factor deficiencies are not possible due to the low disease
with FXIII concentrate is recommended in all patients with a prevalence. Much of current knowledge of these conditions
FXIII level <10%. Since FXIII has a half-life of 7 days, hemo- is derived from registry data and small interventional stud-
static FXIII levels can be maintained by administering FXIII ies. There is a need for both epidemiologic and therapeutic
concentrates every 28 days. Available concentrates include studies in these disorders. Development of international

Disorders of fibrinolysis | 245
databases is required to establish the natural history and described. Both conditions are inherited as autosomal reces-
treatment outcomes of these disorders from which mini- sive traits. Additionally, hyperfibrinolysis may occur due to a
mally active hemostatic levels can be established. variety of acquired conditions, including liver disease and
A major limitation in some of these conditions is the lack disseminated intravascular coagulation (DIC); after surgery,
of availability of a specific replacement concentrate for treat- particularly cardiac surgery; and some prostatic diseases and
ment. Presently in the United States, three licensed products cases of acute promyelocytic leukemia. Although these con-
for rare disorders are available, specifically for afibrinogen- ditions also contribute to bleeding for other reasons (factor
emia, FVII, and FXIII deficiency. A specific concentrate for deficiencies due to liver disease, consumption of clotting fac-
FXI deficiency is available in the European Union. In the tors in DIC, and platelet dysfunction in cardiac surgery), the
United States, off-label use of products continues, including possibility of a contributing hyperfibrinolytic state should be
use of prothrombin complex concentrates for deficiencies of considered, as specific therapies are available.
FX and FII. In FV and FXI deficiency, fresh frozen plasma
remains the mainstay of therapy; in addition, platelet trans-
fusions are sometimes used in FV deficiency as platelets also
contain FV. Even when a concentrate is available, its use in The clinical presentation of hyperfibrinolysis is highly vari-
these rare disorders often is guided by personal experience or able. Hyperfibrinolytic bleeding may occur in isolation as a
anecdotal reports. For example, determination of appropri- result of a congenital deficiency; but most commonly, it
ate patients for whom prophylaxis is indicated and the occurs as a part of a complex coagulopathy in an acquired
appropriate dosing regimen is largely poorly defined. Also, disorder. Congenital deficiencies of the fibrinolytic pathway
the peri- and postoperative care of patients with rare disor- may present with delayed bleeding after injury or interven-
ders is not founded on evidence-based data. There is a clear tion and may include mucus membrane, cutaneous, or deep
need for consistent data collection and studies on the clinical tissue bleeding; however, intracranial hemorrhage has been
management of rare factor deficiencies. reported in PAI-1 and a2AP deficiency. Acquired hyperfibri-
nolysis presents with bleeding at a variety of sites, and in
Key points patients with recent surgery, delayed postoperative hemor-
rhage often occurs at the surgical site.
• Rare factor deficiencies occur as a result of genetic mutations
and acquired disorders.
• Treatment of an associated underlying disorder may lead to Diagnosis
the resolution of the acquired deficiency.
• Rare factor deficiencies result in highly variable bleeding symp-
Laboratory investigation of the fibrinolytic system is diffi-
toms, ranging from injury or interventional bleeding (FXI) to cult. The euglobulin clot lysis time (ELT) currently is not
severe spontaneous intracranial bleeding (FX and FXIII). available in all laboratories, and interpretation of results is
• Few specific factor replacement concentrates are available for not always straightforward. The ELT assesses the capacity of
patients with rare factor deficiencies. plasma to lyse a clot formed in patient plasma. Under assay
conditions, a clot is expected to dissolve within a set period
of time, commonly approximately 2-6 hours, and a short-
Disorders of fibrinolysis ened ELT suggests hyperfibrinolysis. Several new global
hemostatic assays are under evaluation for their ability to
Pathophysiology
more accurately detect hyperfibrinolysis. A currently avail-
The fibrinolytic system provides orderly clot remodeling and able global assay is the thromboelastogram and most com-
dissolution. Imbalances in fibrinolysis may lead to excessive monly is used in surgical settings; thromboelastography is a
fibrinolytic activity through a variety of mechanisms, includ- method to assess global hemostasis and can detect hyperfi-
ing increased tPA activity or inadequate inhibition as the brinolysis in cases in which the use of antifibrinolytic agents
result of PAI-1 or a2AP deficiencies, and may result in exces- may be helpful to control excessive bleeding.
sive bleeding. It is possible to measure a few individual components of
the fibrinolytic system, including a2AP and plasminogen.
Although it is possible to measure antigenic levels of PAI-1,
Etiology
the activity assay is problematic as the normal range
Hyperfibrinolysis may result from congenital deficiencies of includes levels of zero, thereby making detection of a dys-
PAI-1 or a2AP. PAI-1 deficiency is extraordinarily rare, proteinemic deficiency state impossible. Elevated PAI-1
and in only a few cases has the genetic alteration causing levels have been associated with atherosclerosis and are not
the disorder been identified. Defects in a2AP also have been associated with bleeding. PAI-1 levels also exhibit diurnal

variation, and any one level may not represent either the difficult than diagnosis. The fibrinolytic pathway remains
highest or lowest physiologic level. A deficiency of a2AP is the most problematic both in terms of diagnosis of a defi-
measurable; however, the correlation of level of deficiency ciency state and clearly attributable clinical manifestations.
and risk for bleeding is poorly established. It also is possible Improved and specific laboratory methods are required. A
to measure the fibrinolytic proteins tPA and plasminogen, reliable, easily performed, reproducible screening assay
with a hyperfibrinolytic state expected to result in increased would represent an important first step in the diagnosis of
tPA and decreased plasminogen. Again, the correlation these disorders, followed by development of specific factor
between specific levels and the degree of hyperfibrinolysis assays for all components of the fibrinolytic system. Levels of
has not been established. deficiency correlated with clinical bleeding could then be
Therefore, laboratory diagnosis of the fibrinolytic system established. An improved understanding of the genetics of
presently is not optimal, requiring the clinician to rely on congenital fibrinolytic deficiencies and the associated spec-
clinical suspicion, including the presence of delayed bleed- trum of clinical manifestations would assist clinicians in the
ing, the clinical context, and, at times, response to therapeu- diagnosis of these rare disorders.
tic interventions.
Key points
Treatment • Fibrinolytic disorders are the least well-defined hemorrhagic
The treatment of hyperfibrinolytic bleeding is fairly straight- diatheses.
• Hyperfibrinolytic disorders should be suspected in the setting
forward except when it occurs as a complex coagulopathy
of delayed bleeding.
when treatment requires careful consideration of throm-
• Hyperfibrinolytic disorders are most often acquired, although
botic risk. The control of fibrinolytic bleeding is based on the
rare congenital defects have been documented.
use of antifibrinolytic agents; although several agents • Laboratory diagnosis of fibrinolytic disorders is difficult and
are available, two are most widely used: aminocaproic acid inconsistently precise.
and tranexamic acid. The mechanism of action of both • Treatment of hyperfibrinolytic bleeding is based on the use of
agents involves competition with negatively charged lysine- antifibrinolytic agents, including aminocaprioic acid and
rich residues in the kringle domain of plasminogen, which tranexamic acid.
render plasminogen resistant to activation by tPA. Thus,
these agents are effective in tissues rich in tPA. Both are avail-
able for intravenous and oral administration. Adverse effects
and precautions were described previously. When using anti- Bibliography
fibrinolytic therapy, it is important not to discontinue ther-
apy prematurely because of the risk of delayed bleeding. It is Overview of hemostasis
recommended to continue therapy up until the hyperfibri- Hoffman M, Monroe DM III. A cell-based model of hemostasis.
nolysis is felt to have resolved, or possibly on an ongoing Thromb Haemost. 2001;85:958-965.
basis if a congenital defect is confirmed and ongoing therapy
is warranted.
Hemophilia
Prognosis and outcomes Berntorp E, Shapiro A. Modern hemophilia care. Lancet.

2012;379:1447-1456.
Most commonly encountered causes of hyperfibrinolysis are Bowyer AE, Van Veen JJ, Goodeve AC, Kitchen S, Makris M.
acquired; with trigger resolution, the patient’s hemostatic Specific and global coagulation assays in the diagnosis of
system should return to normal, and, provided that cata- discrepant mild hemophilia A. Haematologica. 2013;98:
strophic bleeding has not occurred, patients should recover 1980-1987.
without sequelae. For the rare patient with a confirmed con- Collins PW. Therapeutic challenges in acquired factor VIII
deficiency. Hematology Am Soc Hematol Educ Program.
genital disorder, management with antifibrinolytic agents,
2012;2012:369-374.
even as prophylaxis, can minimize or reduce bleeding
Feldman BM, Pai M, Rivard GE, et al. Tailored prophylaxis in
symptoms.
severe hemophilia A: interim results from the first 5 years of the
Canadian Hemophilia Primary Prophylaxis Study. J Thromb
Gaps in knowledge Haemost. 2006;4:1228-1236.
Hay CR, DiMichele DM. The principal results of the international
The major gap in knowledge in these conditions is the ability immune tolerance study: a randomized dose comparison.
to establish an accurate diagnosis because treatment is less Blood. 2012;119:1335-1344.

Bibliography | 247
Kempton CL, Meeks SL. Toward optimal therapy for inhibitors in Laffan, MA, Lester W, O’Donnell JS, et al. The diagnosis and
hemophilia. Blood. 2014;124:3365-3372. management of von Willebrand disease: a United Kingdom
Manco-Johnson MJ, Abshire TC, et al. Prophylaxis versus Haemophilia Centre Doctors Organization guideline approved
episodic treatment to prevent joint disease in boys with severe by the British Committee for Standards in Haematology.
hemophilia. N Engl J Med. 2007;357:535-544. Br J Haematol. 2014;167:453-465.
Manco-Johnson MJ, Kempton CL, Reding MT, et al. Randomized, Neff AT, Sidonio RF Jr. Management of vWD. Hematology Am Soc
controlled, parallel-group trial of routine prophylaxis vs. on- Hematol Educ Program. 2014;2014:536-541.
demand treatment with sucrose-formulated recombinant factor Ng C, Motto DG, Di Paola J. Diagnostic approach to von
VIII in adults with severe hemophilia A (SPINART). J Thromb Willebrand disease. Blood. 2015;125:2029-2037.
Haemost. 2013;11:1119-1127. Nichols WL, Hultin MB, James AH, et al. von Willebrand disease
Nathwani AC, Tuddenham EG, Rangarajan S, et al. Adenovirus- (vWD): evidence-based diagnosis and management guidelines.
associated virus vector-mediated gene transfer in hemophilia. The National Heart, Lung, and Blood Institute (NHLBI) expert
N Engl J Med. 2011;365:2357-2365. panel report (USA). Haemophilia. 2008;14:171-232.
Oldenburg J, Albert T. Novel products for haemostasis—current Sadler, JE. Low von Willebrand factor: sometimes a risk factor
status. Haemophilia 2014;20(suppl 4):23-28. and sometimes a disease. Hematology Am Soc Hematol Educ
Peyvandi F, Kunicki T, Lillicrap D. Genetic sequence analysis of Program. 2009;2009:106-112.
inherited bleeding diseases. Blood. 2013;122:3423-3431.
Richards M, Lavigne-Lissalde G, Combescure C, et al; European
Haemophilia Treatment and Standardization Board. Neonatal Platelet disorders/bleeding
bleeding in haemophilia: a European cohort study. Br J Hayward CP. Diagnostic evaluation of platelet function disorders.
Haematol 2012;156:374-382. Blood Rev. 2011;25:169-173.
Srivastava A, Brewer AK, Mauser-Bunschoten EP, et al. Quiroga T, Mezzano D. Is my patient a bleeder? A diagnostic
Guidelines for the management of hemophilia. Haemophilia. framework for mild bleeding disorders. Hematology Am Soc
2013;19:e1-e47. Hematol Educ Program. 2012;2012:466-474.
Tiede A, Klamroth R, Scharf RE, et al. Prognostic factors for
remission of and survival in acquired hemophilia A (AHA):
results from the GTH-AH 01/2010 study. Blood. 2015;125: Rare factor deficiencies
1091-1097.
Mumford AD, Ackroyd S, Alikhan R, et al. Guideline for the
Trossaert M, Boisseau P, Quemener A, et al. Prevalence, biological
diagnosis and management of the rare coagulation disorders: a
phenotype and genotype in moderate/mild hemophilia A with
United Kingdom Haemophilia Centre Doctors’ Organization
discrepancy between one-stage and chromogenic factor VIII
guideline on behalf of the British Committee for Standards in
activity. J Thromb Haemost. 2011;9:524-530.
Haematology. Br J Haematol. 2014;167:304-326.
van Dijk K, Fischer K, van der Bom JG, Scheibel E, Ingerslev J,
Peyvandi F, Bolton-Maggs PHB, Batorova A, De Moerloose P. Rare
van den Berg HM. Can long-term prophylaxis for
bleeding disorders. Haemophilia. 2012;18(suppl 4):148-153.
severe haemophilia be stopped in adulthood? Results
Peyvandi F, Di Michele D, Bolton-Maggs PHB, Lee CA, Tripodi
from Denmark and the Netherlands. Br J Haematol.
A, Srivastava A; Project on Consensus Definitions In Rare
2005;130:107-112.
Bleeding Disorders of the Factor VIII/Factor IX Scientific and
Standardization Committee of the International Society of
Thrombosis and Haemostasis. Classification of rare bleeding
disorders (RBDs) based on the association between coagulant
Branchford BR, Di Paola J. Making a diagnosis of vWD. factor activity and clinical bleeding severity. J Thromb Haemost.
Hematology Am Soc Hematol Educ Program. 2012;2012:161-167. 2012;10:1938-1943.

CHAPTER
10
Disorders of platelet number
and function
Adam Cuker and Andreas Greinacher
Platelet biology: structure and Immune causes of thrombocytopenia, Disorders of platelet function, 268
function, 249 252 Bibliography, 275
Regulation of platelet number, 251 Other causes of thrombocytopenia, 262

proteins. The plasma membrane is a bilayer of phospholipids
multimedia component. in which cholesterol, glycolipids, and GPs are embedded. The
phospholipids are asymmetrically organized in the plasma
membrane; negatively charged phospholipids (such as phos-
Platelet biology: structure and phatidylserine [PS]) are present almost exclusively in the
function inner leaflet, whereas the others are more evenly distributed.
Platelets have an elaborate channel system, the open canalic-
Hemostasis encompasses a series of interrelated and simulta- ular system, which is composed of invaginations of the
neously occurring events involving the blood vessels, platelets, plasma membrane and extends throughout the platelet and
coagulation system, and the fibrinolytic pathway. Defects opens to the surface. The discoid shape of the resting platelet
affecting any of these major participants may lead to a is maintained by a well-defined cytoskeleton consisting of the
hemostatic defect and a bleeding disorder. This chapter spectrin membrane skeleton, the marginal microtubule coil,
focuses on disorders related to platelet number and function. and the actin cytoskeleton. The microtubule coil, present
below the platelet membrane, is made up of α-β-tubulin
dimers and plays a role in platelet formation from megakary-
Platelet structure
ocytes, in addition to maintaining the discoid platelet shape.
Blood platelets are anucleate fragments derived from bone In proximity to the open canalicular system is the dense
marrow megakaryocytes. The platelet diameter ranges from tubular system, a closed-channel network derived from the
1.5 to 3.0 mm, roughly one-third to one-fourth that of smooth endoplasmic reticulum; it is considered the major
erythrocytes. Mean platelet volume is approximately 7 fL. site of platelet prostaglandin and thromboxane synthesis.
Electron microscopy reveals a fuzzy coat (glycocalyx) on the Platelets contain a variety of organelles: mitochondria and
platelet surface composed of membrane glycoproteins (GPs), glycogen stores, lysosomes, peroxisomes, dense granules, and
glycolipids, mucopolysaccharides, and adsorbed plasma α-granules. The lysosomes contain acid hydrolases; the dense
granules contain calcium (which gives them high electron
density), adenosine triphosphate (ATP), adenosine diphos-
phate (ADP), magnesium, serotonin (5-hydroxytryptamine),
Conflict-of-interest disclosure: Dr. Cuker: Consultant: Amgen, and polyphosphates (which promote coagulation through
Bracco, Genzyme; research support: Diagnostica Stago, T2 Biosys-
tems. Dr. Greinacher: Consultant: Bayer Healthcare, Boehringer-
various means including activation of the intrinsic pathway).
Ingelheim, ASPEN, Instrumentation Laboratory; honoraria: Bristol Serotonin is taken up by platelets from plasma and incorpo-
Myers Squibb. rated into the dense granules. The α-granules contain a large
Off-label drug use: Desmopressin for inherited platelet function
number of proteins, including β-thromboglobulin (bTG)
defects and uremic platelets. Recombinant VIIa for inherited platelet
function defects. Rituximab for ITP and TTP. Fondaparinux and bi- and platelet factor 4 (PF4), which are considered platelet
valirudin for HIT. specific; several coagulation factors (eg, fibrinogen, factor
| 249

250 | Disorders of platelet number and function
V, factor XIII); von Willebrand factor (vWF); growth factors stimulus for platelet activation. Activated platelets release the
(eg, platelet-derived growth factor [PDGF], vascular endo- contents of their granules (secretion), including ADP and
thelial growth factor [VEGF]); vitronectin; fibronectin; serotonin from the dense granules, which causes the recruit-
thrombospondin; the factor V binding protein multimerin; ment of additional platelets. These additional platelets form
P-selectin; albumin; and immunoglobulin G (IgG). Some of clumps at the site of vessel injury, a process called aggrega-
these (eg, vWF, PF4, bTG) are synthesized by megakaryo- tion (cohesion). Aggregation involves binding of fibrinogen
cytes, whereas others (eg, albumin, IgG) are incorporated to specific platelet surface receptors, a complex composed of
into the alpha granules from plasma. GPIIb-IIIa (integrin αIIbb3). GPIIb-IIIa is platelet specific
and has the ability to bind vWF as well. Although resting
platelets do not bind fibrinogen, platelet activation induces a
Platelet function in hemostasis
conformational change in the GPIIb-IIIa complex that leads
Following injury to the blood vessel, platelets interact with to fibrinogen binding. Moreover, platelets play a major role
collagen fibrils in the exposed subendothelium by a process in coagulation mechanisms; several key enzymatic reactions
(adhesion) that involves, among other events, the interaction occur on the platelet membrane lipoprotein surface. During
of a plasma protein, vWF, and a specific glycoprotein (GP) platelet activation, the negatively charged phospholipids,
complex on the platelet surface, GP Ib-IX-V (GPIb-IX) especially PS, become exposed on the platelet surface, an
(Figure 10-1). This interaction is particularly important for essential step for accelerating specific coagulation reactions
platelet adhesion under conditions of high shear stress. After by promoting the binding of coagulation factors involved in
adherence to the vessel wall via vWF and the long GP Ib-IX- thrombin generation (platelet procoagulant activity).
V receptor, other platelet receptors interact with proteins of A number of physiologic agonists interact with specific
the subendothelial matrix. Hereby collagen provides not receptors on the platelet surface to induce responses, includ-
only a surface for adhesion but also serves as a strong ing a change in platelet shape from discoid to spherical
von Willebrand disease Bernard-Soulier GPIIb-IIIa

syndrome Aggregation
Fibrinogen
vWF Thrombasthenia
GPIb
Adhesion
cAMP P Afibrinogenemia
Gi AC Pleckstrin
ADP (P2Y12) ATP
Receptor PKC TxA2
defects Secretion
TS
ADP (P2Y1) PLC PIP2 DG PGG2/PGH2
Gq CO Disorders of
secretion/signal
Thrombin Arachidonic transduction
IP3
Thromboxane acid
PAF PLA2 IIa Platelet
TK coagulant
Phospholipids
activities
Ca2+
Collagen P Ca
Va II
MLC MLC Xa
MLCK
Ca
2+
Ca VIIa IXa
X
Figure 10-1 Schematic representation of selected platelet responses to activation and inherited disorders of platelet function. The Roman
numerals in the circles represent coagulation factors. Modified with permission from Rao AK. Am J Med Sci. 1998;316:69-76. AC = adenylyl
cyclase; ADP = adenosine diphosphate; BSS = Bernard-Soulier syndrome; CO = cyclooxygenase; DAG = diacylglycerol; G = guanosine
triphosphate–binding protein; IP3 = inositol trisphosphate; MLC = myosin light chain; MLCK = myosin light chain kinase; PAF = platelet
activating factor; PIP2 = phosphatidylinositol bisphosphate; PKC = protein kinase C; PLC = phospholipase C; PLA2 = phospholipase A2;
TK = tyrosine kinase; TS = thromboxane synthase; TxA2 = thromboxane A2; vWF = von Willebrand factor; vWD = von Willebrand disease.

Regulation of platelet number | 251
(shape change), aggregation, secretion, and thromboxane expressed on mature platelets, which bind and clear TPO
A2 (TxA2) production. Other agonists, such as prostacyclin, from the circulation. TPO is secreted constitutively from the
inhibit these responses. Binding of agonists to platelet liver; although its synthesis may increase slightly during
receptors initiates the production or release of several intra- thrombocytopenic states, its overall production is relatively
cellular messenger molecules, including products of constant. As a consequence, the level of free TPO is regulated
hydrolysis of phosphoinositide (PI) by phospholipase C
primarily by the number of circulating platelets, the platelet
(diacylglycerol and inositol 1,4,5-triphosphate [InsP3]), life span, and the megakaryocyte mass. Recent mouse studies
TxA2, and cyclic nucleotides (cyclic adenosine monophos- have challenged this paradigm. The Ashwell-Morell receptor
phate) (Figure 10-1). These induce or modulate the various on murine hepatocytes binds platelets that have lost sialic
platelet responses of Ca2+ mobilization, protein phosphory- acid residues on their surface. Binding activates a JAK-STAT
lation, aggregation, secretion, and thromboxane production. signaling pathway, resulting in increased hepatic TPO
The interaction between the platelet surface receptors and mRNA and TPO production. The relevance of this pathway
the key intracellular enzymes (eg, phospholipases A2 and C, to normal human thrombopoiesis is not yet known.
adenylyl cyclase) is mediated by a group of proteins that In conditions such as aplastic anemia, which is character-
bind and are modulated by guanosine triphosphate (G pro- ized by a low platelet count and decreased bone marrow
teins). As in most secretory cells, platelet activation results in megakaryocyte mass, free TPO levels are high. In immune
an increase in cytoplasmic ionized calcium concentration; thrombocytopenia, the megakaryocyte mass may be ex -
InsP3 functions as a messenger to mobilize Ca2+ from intra- panded and platelet clearance is accelerated. This results in
cellular stores. Diacylglycerol activates protein kinase C enhanced TPO clearance and plasma TPO levels that usually
(PKC), resulting in phosphorylation of several proteins. fall within the normal range despite thrombocytopenia. The
PKC-activation is considered to play a major role in platelet role of TPO as the principal physiologic regulator of platelet
secretion and in the activation of GPIIb-IIIa. Numerous production has been confirmed in studies of TPO and c-Mpl
other mechanisms, such as activation of tyrosine kinases and deficient mice, which have 5%-15% of normal levels of cir-
phosphatases, also are triggered by platelet activation. Either culating platelets, megakaryocytes, and megakaryocyte pro-
inherited or acquired defects in these platelet mechanisms genitor cells. TPO alone, however, does not fully support
may lead to impairment of platelet function in hemostasis. megakaryocyte polyploidization in vitro, suggesting that
additional factors, such as stem cell factor, interleukin
3 (IL-3), interleukin 6 (IL-6), and interleukin 11 (IL-11), are
Regulation of platelet number required for optimal megakaryocyte development.
Overview
Normal platelet production
The platelet count is regulated by the relative rates of platelet
production and clearance. Kinetic studies have demon- Megakaryocyte proliferation and differentiation involve
strated that the average platelet life span is 7-10 days. Plate- endomitosis and polyploidization, a process in which the
lets that are lost through senescence, activation, or other nucleus divides but the cell does not. In the process of matu-
processes are replaced by new platelets derived from bone ration, megakaryocytes form secretory granules and a demar-
marrow megakaryocytes. Platelet production from mega- cation membrane system that permeates the cytoplasmic
karyocytes, in turn, is driven by the hormone thrombopoi- space. This extensive membrane system eventually projects
etin (TPO) and its cellular receptor, c-Mpl. multiple filamentous pseudopodial structures called pro-
platelets. This process utilizes the entire repertoire of cyto-
plasmic granules, macromolecules, and membranes.
Thrombopoietin and the thrombopoietin
Ultimately, fragmentation of the pseudopodial projections
receptor c-Mpl
leads to the release of new platelets. The exact steps leading
A healthy adult produces 1 × 1011 to 3 × 1011 platelets per from megakaryocytes to mature platelets are still not fully
day, although production can increase tenfold during times resolved. Different mechanisms are proposed and have been
of high demand. The number of circulating platelets is regu- shown in mouse models: (i) at the sinusoids of the bone mar-
lated chiefly by TPO, which binds to megakaryocytes and row the megakaryocytes produce long proplatelet strings,
hematopoietic stem cells via c-Mpl. c-Mpl is a member of from which individual platelets rupture; (ii) larger fragments
the class I hematopoietic growth factor receptor superfamily of megakaryocytes consisting of proplatelets are released into
and activates several signaling pathways in megakaryocytes, the sinusoids, which then divide into individual platelets in
resulting in megakaryocyte proliferation and differentiation, the circulation; (iii) the entire megakaryocyte migrates into
ultimately resulting in platelet production. c-Mpl is also the sinusoids, is transported with the blood stream into the

lung, where the shear forces in the lung arterioles cause release Table 10-1 International Working Group proposed definitions of
of platelets. It is likely that all three models are at least par- disease
tially true and together contribute to platelet production from Primary ITP Isolated thrombocytopenia
megakaryocytes. Each megakaryocyte produces 1,000-3,000 Platelets <100,000/µL
platelets before the remaining nuclear material is phagocy- No other apparent causes of thrombocytopenia
tosed by resident macrophages. Released platelets circulate for No secondary cause of ITP present
7-10 days before undergoing senescence and clearance by Secondary ITP All other forms of immune-mediated
phagocytic cells in the reticuloendothelial system. thrombocytopenia except primary ITP
Designate with presumed cause, in parentheses,
following secondary ITP, eg, secondary ITP
(lupus-associated)
Key points
Phases of the Newly diagnosed: within 3 months of diagnosis
• The primary mediator of platelet production is TPO, produced disease Persistent: between 3 and 12 months of diagnosis
primarily by the liver. Chronic: lasting >12 months
• TPO production is largely constitutive; TPO levels are Severe: presence of bleeding at presentation
regulated by the platelet and megakaryocyte mass through sufficient to mandate treatment, or
binding of TPO to its receptor, c-Mpl. occurrence of new bleeding symptoms
• TPO levels are typically normal in immune thrombocytopenia requiring additional intervention
(ITP) but are elevated in bone marrow failure syndromes. Adapted from Rodegheiro F et al. Blood. 2009;113:2386-2393.
• The normal platelet life span is 7-10 days. ITP = immune thrombocytopenia.
Immune causes of thrombocytopenia thrombocytopenia in which a predisposing condition can be

identified. ITP is also classified according to disease duration.
Within 3 months of presentation, ITP is termed newly diag-
Clinical case nosed. ITP lasting 3-12 months and >12 months is denoted
A 68-year-old man is referred for evaluation of increased bruis- as persistent and chronic, respectively (Table 10-1). This ter-
ing, primarily on his forearms, for the last 3 months. He restores minology was adopted in the ITP guideline developed by the
old cars as a hobby and believes that trauma associated with this American Society of Hematology.
work may have caused his bruises, although he cannot recall ITP is a relatively common cause of thrombocytopenia in
specific instances during which he injured himself. He denies adults and children. Estimates of prevalence vary widely,
epistaxis, melena, or other evidence of systemic bleeding. He is ranging between 3 and 20 per 100,000 persons, with an esti-
in otherwise good health other than mild hypertension treated
mated incidence of 2-10 cases per 100,000 patient-years. In
with an angiotensin converting enzyme inhibitor; he does not
childhood, the highest incidence is in children <5 years old,
take other prescription medications but takes fish oil and vitamin C
with a gradual decrease toward adolescence. Most studies
supplements. On physical examination, he looks well but several
2.0-cm bruises are noted on the distal upper extremities and
find the incidence to be equal in girls and boys, although
backs of the hands. Complete blood count reveals a hemoglobin some reports suggest a higher incidence in boys <5 years old.
of 13.8 g/dL, white blood cell (WBC) count of 6.9 × 109/L, and In adults, the incidence and prevalence of ITP is greatest in
platelet count of 22,000/µL. the elderly, with a female preponderance in the middle-adult
years and a slight male preponderance in patients >70 years
of age. In children, ITP often occurs after an antecedent viral
infection and is self-limited in 80% of cases. In contrast, pri-
ITP is an autoimmune disorder characterized by thrombo- mary ITP assumes a chronic course in approximately 75% of
cytopenia and a variable risk of bleeding. An international adult patients. Although patients with more severe throm-
working group proposed standard terminology and defini- bocytopenia may present with mucocutaneous bleeding,
tions for ITP. The term immune is now used instead of idio- those diagnosed with thrombocytopenia on a routine blood
pathic and the term purpura has been abandoned because count are often asymptomatic. There is no gold-standard
bleeding symptoms, including purpura, are not present in all laboratory test for ITP. Although detection of glycoprotein-
cases. Thus, the working group recommended the term specific antiplatelet antibodies on the patients' autologous
immune thrombocytopenia, although the abbreviation ITP is platelets suggests the diagnosis, these antibodies are detect-
preserved. In this classification scheme, primary is used to able in only about 60% of patients with ITP. The diagnosis of
denote ITP with no apparent precipitating cause, while sec- ITP is primarily made by excluding nonimmune causes of
ondary ITP refers to other forms of immune-mediated thrombocytopenia and investigating potential secondary

Immune causes of thrombocytopenia | 253
Table 10-2 Differential diagnosis of immune thrombocytopenia
Previously diagnosed or high risk of conditions that may be associated with autoimmune thrombocytopenia (eg, HIV, hepatitis C virus, or
other infection; other autoimmune or immunodeficiency disorders; malignancy; recent vaccination)
Liver disease, including cirrhosis from any cause
Drugs (prescription or nonprescription), alcohol abuse, consumption of quinine (tonic water), environmental toxins
Bone marrow disorders, including myelodysplastic syndromes, leukemias, other malignancies, fibrosis, aplastic anemia, and megaloblastic
anemia
Recent transfusions (posttransfusion purpura) and recent immunization
Inherited thrombocytopenia: thrombocytopenia-absent radii syndrome, radioulnar synostosis, congenital amegakaryocytic thrombocytopenia,
Wiskott-Aldrich syndrome, MYH9-related disease, type IIb von Willebrand disease, Bernard-Soulier syndrome
Adapted from Rodegheiro F et al. Blood. 2009;113:2386-2393.
causes. The most compelling evidence supporting the pres- quality of life. Fatigue correlates with a platelet count
ence of ITP is a platelet response to ITP-specific therapy. <100,000/µL and treatment with steroids but not with dura-
Secondary ITP occurs in the setting of lymphoproliferative tion of ITP, age, or gender. It usually resolves with successful
disorders; systemic lupus erythematosus, antiphospholipid treatment of ITP. Mounting epidemiologic evidence suggests
syndrome, or other autoimmune disorders; infections such as that ITP is associated with an increased risk of venous
hepatitis C, HIV, and Helicobacter pylori; and immune defi- thromboembolism. The mechanism of thrombosis is not
ciency states such as common variable immune deficiency. well established but may relate to underlying disease patho-
Drug-induced immune thrombocytopenia is described in the physiology and/or effect of treatment.
section “Drug-induced immune thrombocytopenia” later in Physical examination should focus on typical bleeding
this chapter. Nonimmune causes of thrombocytopenia sites. Dependent areas and skin underneath tight clothing
including hypersplenism, hereditary thrombocytopenias, should be examined for petechiae and purpura, and oral
thrombotic thrombocytopenic purpura (TTP), and type 2B mucous membranes should be examined for hemorrhagic
von Willebrand disease (vWD) should be included in the dif- bullae, which may be associated with an increased risk of
ferential diagnosis of ITP (Table 10-2). Occasional patients severe bleeding at other sites. In a patient with primary ITP,
with myelodysplastic syndromes or aplastic anemia may the remainder of the general physical examination is normal.
present with isolated thrombocytopenia. The presence of lymphadenopathy or splenomegaly should
prompt investigation for other etiologies of thrombocytope-
nia. Skeletal, renal, or neurologic abnormalities suggest a
Clinical features of ITP
familial cause of thrombocytopenia.
Clinical features of primary and secondary ITP are generally
similar, although in secondary ITP clinical manifestations
Pathophysiology of ITP
related to the underlying disorder may be prominent. A
platelet count below 1100,000/µL is required for the diagno- Primary ITP is a syndrome that results from several different
sis of ITP because mild thrombocytopenia may occur in nor- pathophysiologic mechanisms. Classic experiments per-
mal individuals and rarely results in development of more formed in the 1950s and 1960s demonstrated a critical role
severe thrombocytopenia or other autoimmune disease. The for antiplatelet antibodies in mediating the enhanced clear-
most common symptom of ITP is mucocutaneous bleeding, ance of platelets in patients with ITP. These antibodies
which may manifest as petechiae, ecchymosis, epistaxis, recognize GPs on the platelet surface, most commonly
menorrhagia, oral mucosal, or gastrointestinal bleeding. In a GPIIb-IIIa and GPIb-IX. Antibody-coated platelets are
recent systematic review, intracranial hemorrhage was cleared from the circulation by phagocytes in the reticuloen-
reported in 1.4% of adults and 0.4% of children. Spontane- dothelial system, primarily the spleen. Antiplatelet antibod-
ous bleeding is uncommon at platelet counts >30,000/µL. ies may recognize the same targets on megakaryocytes,
There is significant variability in bleeding among leading to impairment of megakaryocyte proliferation and
patients with similar platelet counts, however, and some differentiation and proplatelet production. As noted above,
individuals with counts <10,000/µL bleed infrequently. The plasma levels of TPO generally are not elevated in patients
risk of fatal bleeding is greatest in elderly patients with per- with ITP due to an expanded megakaryocte mass and accel-
sistent and severe thrombocytopenia (platelets <20,000/µL). erated platelet clearance.
Non-hemorrhagic clinical manifestations common among Not all patients with ITP have detectable platelet antibod-
patients with ITP include fatigue and reduced health-related ies. Dysregulated T cells may have a direct cytotoxic effect on

platelets and impair platelet production by megakaryocytes. enlarged platelets (as a rule of thumb, five normal platelets fit
Recent interest has focused on decreased levels of regulatory into one red blood cell). If the blood smear shows more than
T cells (Treg cells) in patients with ITP; successful ITP treat- 60% large or even giant platelets (less than two platelets fit
ment has been associated with restoration of Treg levels. into one red blood cell), hereditary macrothrombocytopenia
The pathogenesis of secondary ITP may share similar (see section “Hereditary thrombocytopenia” in this chapter)
mechanisms with primary ITP. For example, the thrombo- is more likely the underlying cause.
cytopenia that occurs in patients with antiphospholipid Bone marrow examination is not required routinely and is
antibodies may reflect the concurrent presence of antibod- generally not useful for diagnosing ITP, but should be per-
ies against platelet GPs. Unique pathogenic mechanisms, formed to exclude other causes of thrombocytopenia when
however, have been identified in some types of secondary atypical features such as unexplained anemia, lymphade-
ITP. For example, antigen mimicry, in which antibodies nopathy, or splenomegaly are present. Because approxi-
directed to a foreign (viral) protein cross-react with specific mately 80% of patients with ITP respond to initial therapy
epitopes on platelet GPIIb-IIIa has been observed in hepati- with corticosteroids, intravenous immunoglobulin (IVIg),
tis C–associated ITP. A similar pathophysiology may under- or Rh-immune globulin (anti-D), failure to respond to these
lie the pathogenesis of ITP in patients with H. pylori agents should prompt consideration of bone marrow exami-
infection and HIV. nation and other causes of thrombocytopenia. Bone marrow
examination may also be warranted in elderly patients in
whom myelodysplasia is suspected. Megakaryocyte number
Diagnosis of ITP
is typically normal or increased in the marrow of patients
The diagnosis of ITP rests on a consistent clinical history, with ITP. In a blinded study, hematopathologists were not
physical examination, and exclusion of other causes of able to reliably distinguish ITP marrows from those of non-
thrombocytopenia. The leukocyte count is characteristically thrombocytopenic controls.
normal. The hemoglobin concentration is typically normal With appreciation that secondary causes of ITP may be
as well unless thrombocytopenic bleeding has resulted in more common than previously believed and may influence
anemia. Examination of the peripheral blood film should be management, additional laboratory studies such as screen-
performed to exclude pseudothrombocytopenia (ethylenedi- ing for hepatitis C and HIV should be considered. Table 10-3
aminetetraacetic acid-dependent platelet agglutinating anti- contains a list of suggested screening studies proposed by the
bodies), microangiopathic hemolytic anemia (fragmented ITP International Working Group. In our practice, we
red cells), or abnormalities suggestive of other disorders. perform a basic evaluation including a history, physical

Identification of unexpected abnormalities should prompt examination, complete blood cell and reticulocyte count,
an evaluation for other etiologies of thrombocytopenia. examination of the peripheral blood smear, ABO-Rh blood
The mean platelet volume (MPV) may be increased in type, and HIV and hepatitis C testing in all patients. Addi-
patients with ITP. However, ITP patients always show a het- tional tests are requested in selected patients based on the
erogeneous platelet population with not more than ~30% findings of the basic evaluation.
Table 10-3 International Working Group recommendations for the diagnosis of ITP in adults
Tests of uncertain
Basic evaluation Test of potential utility benefit
Patient and family history Glycoprotein-specific antibodies TPO levels

Physical examination Antiphospholipid antibodies Reticulated platelets
CBC and reticulocyte count Antithyroid antibodies and thyroid function Platelet-associated IgG
Peripheral blood film Pregnancy test in women of childbearing potential Platelet survival study
Bone marrow exam PCR for parvovirus and CMV Bleeding time
Blood group (Rh) Complement levels
Direct antiglobulin test
H. pylori, HIV, HCV (suggested by majority regardless of
geographic region)
Quantitative immunoglobulins (consider in children with ITP,
recommend in children with persistent or chronic ITP)
Adapted from Provan D et al. Blood. 2010;115:168-186.

CBC = complete blood count; CMV = cytomegalovirus; HCV = hepatitis C virus; IgG = immunoglobulin G; ITP = immune thrombocytopenia;
PCR = polymerase chain reaction; TPO = thrombopoietin.

Management of primary ITP in children by some experts as a cutoff for considering treatment of ITP.
However, there is significant variability in bleeding among
Because spontaneous recovery is expected in most children
patients and therapy should be individualized. Treatment
with primary ITP, families of children generally need coun-
decisions should not be dictated by the platelet count alone,
seling and supportive care rather than specific drug therapy.
but should take into account other factors including individ-
Severe hemorrhage occurs in ~1 in 200 children with newly
ual patient’s bleeding phenotype, the need for concomitant
diagnosed ITP, and intracerebral hemorrhage occurs in
antithrombotic therapy or other medications that affect
<1 in 500, usually in the first month after diagnosis. For
hemostasis, the need for an invasive procedure or surgery, and
those in whom treatment is considered necessary, a short
patient values and preferences including a desire to participate
course of corticosteroids, IVIg, or anti-D (in Rh-positive
in sports or other activities associated with bleeding risk. Even
individuals) generally results in rapid recovery of the platelet
in asymptomatic of minimally symptomatic patients, an ini-
count. Adverse effects of therapy in children include behav-
tial short-term treatment course is reasonable to support the
ioral changes from corticosteroids, headache from IVIg, and
diagnosis of ITP and to identify a treatment to which the
hemolysis from anti-D, which rarely may be severe. Patients
patient responds in case of worsening symptoms or need for
(adults and children) with a positive direct antiglobulin test
an invasive procedure.
should not receive anti-D because of an increased risk of
Although several first-line therapies are available, predni-
severe hemolysis.
sone (1 mg/kg daily) remains the initial treatment of choice
Recovery of the platelet count ultimately occurs in 80% of
because of its efficacy and low cost. Approximately 75% of
children even without therapy, usually within 3-6 months
patients initially respond to corticosteroids, although taper-
but occasionally over a year or more since presentation. The
ing usually precipitates relapse, and ultimately only 20%-
remaining 20% have chronic thrombocytopenia, yet even in
25% of patients are able to maintain a durable platelet
this group, major bleeding is uncommon. Splenectomy is
response after steroid discontinuation. High-dose dexa-
generally reserved for severe persistent or chronic thrombo-
methasone (40 mg daily for 4 days given over 1 to 6 cycles
cytopenia and bleeding and results in complete remission in
repeated every 2 to 4 weeks) provides an alternative for the
~75% of children. The risk for overwhelming sepsis after
initial treatment of patients with ITP. A single cycle of high-
splenectomy is greater in young children, and, therefore,
dose dexamethasone does not improve overall or long-term
splenectomy generally is deferred until at least 5 years of age.
response rates compared with standard-dose prednisone.
Vaccination against Streptococcus pneumoniae, Neisseria
Whether multiple cycles of high-dose dexamethasone are
meningitidis, and Haemophilus influenzae type b should be
superior to prednisone is the subject of an ongoing random-
given before splenectomy in children and adults, and peni-
ized controlled trial. Two controlled trials have compared
cillin prophylaxis is recommended until adulthood. Ritux-
high-dose dexamethasone alone or in combination with
imab is another effective therapy in children, with an initial
rituximab in newly diagnosed ITP. Combination therapy was
response rate of 40%-50% and a long-term response rate of
associated with superior response rates at 6 and 12 months,
~25%. Small clinical trials of thrombopoietin receptor ago-
but the difference waned with longer-term follow-up, and
nists (TRAs) suggest that they are effective and safe in refrac-
grade 3 and 4 adverse events were greater in the combination
tory childhood ITP. Larger trials are ongoing. These agents
therapy arms. Approximately 25% of patients with ITP may
are not FDA-approved for childhood ITP, and their use in
achieve a durable remission after treatment with corticoste-
this context should be considered investigational.
roids, usually within the first year after presentation. This
observation has led to a recommendation by the Interna-
tional Working Group that splenectomy be deferred until at
Management of primary ITP in adults
least 1 year after presentation, if possible.
In contrast to children, ITP in adults evolves into a chronic For patients who do not achieve a response with cortico-
disease in approximately 75% of patients. The goal of ITP steroids, therapy may be supplemented with intermittent
management in adults is to maintain a hemostatic platelet IVIg or anti-D. Both agents are associated with response
count while minimizing the toxicity of therapy. There are no rates similar to those of corticosteroids; however, the dura-
controlled studies demonstrating the superiority of any spe- tion of response is generally only 2-4 weeks and thus fre-
cific treatment algorithm, and significant variability exists quent, intermittent dosing is required if these agents are used
among treatment approaches advocated by different experts. for chronic therapy. One uncontrolled study of
Asymptomatic patients with mild or moderate thrombocyto- 28 Rh-positive, nonsplenectomized adults reported that
penia and no bleeding require no specific treatment. Platelet repeated dosing of anti-D for platelet counts <30,000/µL was
counts <30,000/µL may be associated with an increased an effective maintenance therapy and that 43% of patients
bleeding risk. This platelet count threshold has been suggested treated in this manner ultimately entered a durable remission.

Nevertheless, both IVIg and anti-D generally are considered 2 years. An appealing aspect of rituximab therapy is the poten-
to be bridging agents used to maintain platelet counts in a tial induction of long-term responses in a subset of patients,
hemostatic range until more definitive therapy can be though long-term remission rates have generally been disap-
initiated. pointing. In a long-term follow-up study, only 21% of adults
Second-line therapy is indicated for patients who do not treated with rituximab remained free of relapse at 5 years. It
respond to first-line therapy or relapse after it is tapered. is unknown whether rituximab was responsible for inducing
Options for second-line therapy include rituximab, TRAs, or long-term remission in these patients or whether their dis-
splenectomy. Splenectomy has been used to treat ITP for ease would have spontaneously remitted irrespective of
decades, although the availability of alternative treatments, therapy. Adverse effects of rituximab include infusion reac-
concerns about adverse events, and the realization that some tions (eg, hypotension, chills, and rash), serum sickness,
patients with newly diagnosed ITP ultimately may improve and cardiac arrhythmias. Reactivation of latent JC virus
over time has led to decreased utilization in contemporary causing progressive multifocal leukoencephalopathy has
cohorts compared with older series. Although both the ITP been reported, but it appears to be extremely uncommon.
International Working Group and the revised American Reactivation of hepatitis B after rituximab has been
Society of Hematology (ASH) guidelines consider splenec- described, and active hepatitis B infection is a contraindica-
tomy an acceptable second-line therapy for ITP, the former tion for treatment. Rituximab also interferes with the
group weighs splenectomy equally to other medical options, response to polysaccharide vaccines. This is of potential con-
whereas the ASH guidelines recommend splenectomy (grade cern in patients who may subsequently undergo splenectomy
1B evidence) for patients who fail corticosteroids while and supports the practice of administering immunizations
suggesting rituximab or TRAs (grade 2C evidence). Splenec- prior to rituximab.
tomy leads to a high rate of durable remission. In a system- The TPO receptor agonists romiplostim and eltrombopag
atic review, 1,731 (66%) of 2,623 adults with ITP achieved a are approved in many countries for patients with ITP who
complete response following splenectomy at a median fol- have had an insufficient response to corticosteroids, immu-
low-up of 28 months (range 1 to 153 months), and this noglobulins, or splenectomy. These agents bind and activate
response rate was maintained at ≥10 years after splenectomy. the TPO receptor, c-Mpl, leading to increased platelet pro-
Splenectomy does not jeopardize subsequent responses to duction; however, they have no structural similarity to
other ITP therapies (other than anti-D) and may reduce endogenous TPO and do not stimulate cross-reactive TPO
long-term costs of ITP management. Disadvantages of sple- antibodies. The response rates to these agents range from
nectomy include a lack of validated predictors of response; 59% to 88%, and loss of response while on continued ther-
surgical risk with a 30-day mortality, and complication rate apy is uncommon. These agents are effective before and after
of 0.2% and 9.6% for laparoscopic splenectomy and 1.0% splenectomy and usually allow decreases in dosage or dis-
and 12.9% for open splenectomy; an increased risk of post- continuation of concomitant ITP therapy. A disadvantage of
splenectomy infection; and an increased risk of vascular these agents is the need for indefinite therapy and the associ-
thrombosis. The incidence of infection may be reduced by ated costs, although anecdotal reports describe patients in
presplenectomy pneumococcal, meningococcal, and Hae- whom these drugs have been discontinued with maintenance
mophilus influenzae b vaccination; repeat pneumococcal of hemostatic platelet counts. Increased bone marrow retic-
vaccination 5 years after initial vaccination; and antibiotic ulin develops in approximately 5% of patients treated with
prophylaxis for fever. TPO receptor agonists, but there is no evidence for develop-
Rituximab, an anti-CD20 monoclonal antibody that rap- ment of progressive or irreversible bone marrow fibrosis.
idly depletes CD20+ B lymphocytes, may be used in lieu of Eltrombopag carries a boxed warning because of the poten-
splenectomy or in patients who have failed splenectomy. The tial for hepatotoxicity. Patients treated with either agent may
usual dose is 375 mg/m2 weekly for 4 weeks, although an develop severe thrombocytopenia following discontinuation
optimal dosing regimen has not been defined and lower of treatment. There may be an increased risk of thrombosis
doses have shown similar efficacy. In a systematic review of in patients with preexisting risk factors. Thrombotic risk
313 ITP patients, half of whom were not splenectomized, does not appear to be linked to the platelet count.
62.5% achieved a platelet count response (platelet increment For patients who do not respond to or are intolerant of
of 50,000/µL), with a median time to response of 5.5 weeks second-line therapy, various immunosuppressant medica-
(range, 2 to 18 weeks) and a median duration of response of tions are available including azathioprine, cyclosporine,
10.5 months (range, 3 to 20 months). In a single-arm study mycophenolate mofetil, cyclophosphamide, vinca alkaloids,
of 60 nonsplenectomized ITP patients, 40% achieved a plate- dapsone, and danazol. Evidence on use of these agents is lim-
let count –>50,000/µL with at least a doubling from baseline ited to uncontrolled case series. Thrombocytopenia in
at 1 year, and in 33.3%, this response was sustained for patients with secondary ITP may respond to treatment of the

underlying disease. For example, treatment of HIV with Drug-induced immune thrombocytopenia
highly active antiretroviral therapy (HAART) induces a
More than 300 drugs have been implicated in drug-induced
platelet response in most patients. Eradication of H. pylori
immune thrombocytopenia (DITP), including quinine and
has led to resolution of ITP in >50% of cases in certain coun-
quinidine (present in tonic water, bitter melon, and certain
tries, including Japan, although it has generally not been
medications), nonsteroidal anti-inflammatory agents, trim-
effective in North America. This may reflect differences in
ethoprim-sulfamethoxazole, vancomycin, rifampin, anti-
endemic H. pylori strains in different geographic regions.
convulsants, sedatives, and acetaminophen as well as the
Iron-deficiency anemia is common in ITP, particularly in
platelet GPIIb-IIIa inhibitors tirofiban, eptifibatide, and
menstruating women. In addition to platelet-raising ther-
abciximab. A systematic review of individual patient data
apy, an important element of management is correction of
found that the most commonly reported drugs with a def-
iron-deficiency and anemia because a normal red blood cell
inite or probable causal relation to thrombocytopenia
count improves hemostasis, probably through rheological
were quinidine, quinine, rifampin, and trimethoprim-
factors that bring platelets into closer proximity to the endo-
sulfamethoxazole. A database of implicated drugs is avail-
thelium in flowing blood.
able online and periodically updated (Platelets on the Web;
available at http://www.ouhsc.edu/platelets). Heparin-induced
Emergency treatment of ITP thrombocytopenia (HIT) is discussed separately because of its
unique clinical manifestations and pathophysiology.
Patients with new-onset, severe thrombocytopenia (20,000/
µL) and bleeding should be hospitalized. Examination of the
peripheral blood smear to exclude thrombotic microangi-
Mechanisms of DITP
opathy and a careful medication history to exclude drug-
induced thrombocytopenia should be undertaken. Once a DITP characteristically occurs approximately 1-2 weeks after
presumptive diagnosis of ITP has been reached, manage- initial drug exposure, a timeframe consistent with produc-
ment of bleeding may require platelet transfusions in tion of drug-dependent or drug metabolite-dependent IgG
combination with high doses of parenteral corticosteroids antibodies. An exception is thrombocytopenia induced by
(methylprednisolone 1 g intravenously daily for 2-3 days) the GPIIb-IIIa antagonists, eptifibatide, tirofiban, and abcix-
supplemented with IVIg (1 g/kg for 1-2 days). Increases in imab, which may present within hours of exposure due to
the platelet count may become apparent within 3-5 days, naturally occurring antibodies. Several mechanisms specific
although complete responses may require 1-2 weeks. A TRA, for individual drugs underlie the development of DITP.
vinca alkaloid, or emergency splenectomy may be required Quinine-induced thrombocytopenia was first described
for patients with refractory thrombocytopenia and persis- more than a century ago and serves as a prototype. In this
tent bleeding. In case of life threatening bleeding, massive disorder, the binding of antibodies to platelet GPs is greatly
platelet transfusion can control hemorrhage, but typically enhanced in the presence of sensitizing drug. This may result
multiple platelet units are required. from binding of the drug to specific GPs, such as GPIIb-IIIa
or GPIb-IX. Affinity maturation of B-cells producing low-
Key points affinity antibodies reacting with the neoepitope induced by
the complex of the drug and the platelet GP may result in the
• ITP may occur as a primary disorder or secondary to a generation of antibodies that can destroy platelets in the
predisposing illness. presence of the drug. Another potential mechanism is modi-
• The diagnosis of primary ITP is made by excluding other
fication of the hypervariable region of the antibody when a
causes of thrombocytopenia.
small molecule drug binds to the antigen recognition site,
• ITP in children is usually self-limited; conversely, ITP in adults
thereby modifying the specificity of the antibody.
develops into a chronic disease in ~75% of patients.
• The pathogenesis of ITP involves accelerated platelet destruc-
A much rarer mechanism of DITP involves the induction
tion and decreased platelet production. of autoantibodies by drugs such as gold and interferon-α or -β,
• Corticosteroids, supplemented as needed with IVIG or anti-D, leading to development of a syndrome that resembles ITP. An
are first-line therapy for ITP. often-overlooked cause of DITP is that which follows vacci-
• Second-line therapy (splenectomy, rituximab, TRA) is nations, including diphtheria-pertussis-tetanus (DPT) and
indicated in patients who do not respond to first-line therapy or measles-mumps-rubella (MMR), which reflects the develop-
relapse after it is tapered. ment of true autoantibodies similar to those described in ITP.
• Emergency treatment of severe ITP includes a combination of Tirofiban and eptifibatide (“fibans”) are small molecule
steroids and IVIG and, in case of life threatening bleeding, mimetics of the RGD region of fibrinogen that inhibit fibrinogen
massive platelet transfusion.
binding to activated GPIIb-IIIa and block platelet aggregation.

Thrombocytopenia may occur because of preexisting anti- Treatment for DITP involves discontinuation of the offend-
bodies that recognize conformation-dependent neoepitopes ing drug. A practical approach in hospitalized patients on
(mimetic induced binding sites [MIBS]) induced in GPIIb- multiple medications is to stop all drugs started within the
IIIa following drug binding (rapid onset) as well as by induc- last two weeks (excluding electrolytes and nutrients), when
tion of new antibodies towards the neoepitope induced by feasible, and to switch antibiotics. The platelet count starts to
fiban binding to the GPIIb-IIIa complex (delayed onset). recover after 4-5 half-lives of the culprit drug or drug metab-
Abciximab, a chimeric (mouse–human) Fab fragment to olite, which can last several days. Patients with severe throm-
GPIIb-IIIa, causes acute profound thrombocytopenia in bocytopenia and bleeding, as well as those judged to be at
0.5%-1.0% of patients on their first exposure because of particularly high risk of bleeding, may be treated with IVIg,
preexisting antibodies that recognize the murine portion of corticosteroids, or plasma exchange, though there is only
abciximab. About 50% of cases of “fiban-induced throm- limited evidence to support these interventions. Platelet
bocytopenia” are due to pseudothrombocytopenia. The com- transfusion is generally ineffective. Patients should be
bination of fiban binding and calcium depletion by the instructed to avoid the culprit drug in the future, and it
anticoagulant enhances binding of natural IgM antibodies in should be added to their allergy list.
the test tube, resulting in platelet clumping. Fiban-induced
pseudothrombocytopenia frequently also manifests in citrated
Key points
blood. As fibans are given in patients with high-risk coronary
interventions, recognition of pseudothrombocytopenia is of • Many drugs have been implicated as causes of DITP.
major importance. Inappropriate cessation of anti-platelet • Quinidine, quinine, and antibiotics such as trimethoprim-
therapy and potentially even platelet transfusion or other pro- sulfamethaxazole and vancomycin are common culprits.
hemostatic measures may subject the patient to an increased • Thrombocytopenia caused by tirofiban, eptifibatide, and
risk of thrombosis including in-stent thrombosis. abciximab may occur soon after exposure in patients not
previously exposed to these drugs.
• DITP can be confirmed in some cases by demonstration of
Diagnosis of DITP drug- (or drug metabolite) dependent platelet-reactive antibod-
ies in vitro.
Clinical criteria have been proposed that may be used to
judge the likelihood of a given drug causing DITP. These
include a temporal association between drug exposure and
Heparin-induced thrombocytopenia
thrombocytopenia, the exclusion of other causes of throm-
bocytopenia, and recurrence of thrombocytopenia upon HIT is an idiosyncratic drug reaction caused by antibodies
drug rechallenge. In practice, particularly in hospitalized against multimolecular complexes of PF4 and heparin. Bind-
patients, a multitude of potential culprit drugs and concur- ing of HIT antibodies to Fc receptors on monocytes and
rent illnesses, such as infections, may make the diagnosis of platelets causes cellular activation; HIT antibodies also acti-
DITP difficult. An important diagnostic clue is the timing of vate endothelial cells by binding endothelial cell-associated
initiation of the drug. Typically, the causal drug has been PF4. The net result is elevated levels of circulating micropar-
started 1-2 weeks before the onset of thrombocytopenia. In ticles and an intensely prothrombotic state. HIT occurs most
hospitalized patients, the most frequent cause of DITP are commonly in patients receiving unfractionated heparin
antibiotics. Specialized laboratory testing for antibodies that (UFH).The incidence of HIT varies widely between patient
bind to platelets in the presence of drug or drug metabolite groups, with reported incidences of 0.2%-5.0%. The risk of
have been developed. However, such assays are only avail- HIT associated with low-molecular-weight heparin (LMWH)
able for a limited number of drugs and drug metabolites, are is five- to tenfold lower, and use of LMWH instead of UFH is
not standardized, and are only performed at a small number the most efficient measure to prevent HIT in any patient
of reference laboratories around the world. They may pro- group (but LMWH must not be used when HIT has devel-
vide useful confirmation of DITP, but because there is a oped). Thrombosis develops in 40%-50% of patients with
several-day turnaround time for these “send-out” tests, cli- HIT despite the occurrence of thrombocytopenia; bleeding is
nicians are forced to make critical initial decisions about rare. Although the diagnosis of HIT in the acute setting is
whether to suspend suspicious medications without the ben- clinical, confirmation depends on correlative laboratory test-
efit of laboratory results. ing. Transient thrombocytopenia following the administra-
DITP is characteristically severe with median nadir plate- tion of heparin (previously called type I HIT, or nonimmune
let counts of 20,000/µL and a high risk of hemorrhage. A HIT) is an innocuous syndrome caused by binding of hepa-
review of 247 case reports of DITP found an incidence of rin to platelet GPIIb-IIIa, thereby inducing a signal, which
major and fatal bleeding of 9% and 0.8%, respectively. lowers the threshold for platelet activation by other agonists.

Clinical features heparin exposure (delayed-onset HIT). In these patients, the

antibodies have gained autoreactivity (i.e. they recognize
HIT is a clinicopathological syndrome, which requires both
PF4 bound to endogenous glycosaminoglycans on platelets
the presence of platelet activating antibodies, usually directed
and therefore activate platelets even in the absence of hepa-
towards PF4/heparin complexes, and clinical symptoms,
rin). Rarely, HIT can manifest as an autoimmune disorder
which include a decrease in the platelet count by >50%
without any exposure to heparin, so called spontaneous or
and/or new thromboembolic complications. As a general
autoimmune HIT. The autoantibodies in this disease are
rule, the clinical symptoms manifest between day 5 and 14
often triggered by a preceding infection or by orthopedic
after initiation of heparin. An exception is rapid-onset HIT,
surgery. PF4 also binds to polyanions other than heparin
in which patients with recent heparin exposure (usually
such as lipopolysaccharide on bacteria or RNA/DNA and
within the last 30 days) and preexisting HIT antibodies may
undergoes the same changes in its conformation as when
manifest clinical HIT within hours of heparin re-exposure. A
platelet count decrease or a new thrombosis without corre- binding to heparin. These endogenous PF4-polyanion com-
sponding antibodies is not HIT. Similarly, a positive assay plexes are likely the trigger for autoimmune HIT. The result-
for platelet activating antibodies or a positive PF4-heparin ing antibodies are typically of very high titer and, in contrast
enzyme-linked immunoadsorbent assay (ELISA) without to typical HIT antibodies, can persist for months.
corresponding clinical symptoms is not HIT. Several clinical scoring systems have been developed to
HIT is uncommon in children and is more common in assist with determining the pretest probability of HIT. The
females (odds ratio 2.37). The incidence of HIT is approxi- most commonly used is the 4T system (thrombocytopenia,
mately threefold greater in surgical than in medical patients. timing, thrombosis, and other; see Table 10-4). This system
While patients receiving thromboprophylaxis with UFH has been shown to have a high negative predictive value (ie, a
after major orthopedic surgery had the highest incidence of low score is useful in ruling out HIT), but its effectiveness is
HIT (5%) in the 1990s, today HIT is rare in this patient limited by modest interobserver agreement and a relatively
group. Whether this is related to the widespread use of low positive predictive value. Recent studies have demon-
LMWH or to other changes in surgical practice is unknown. strated that this system may be of less utility in intensive care
Today, patients with cardiac assist devices and those under- patients, a setting in which HIT is often suspected due to the
going cardiac surgery have the highest incidence of HIT high prevalence of thrombocytopenia but relatively uncom-
(1%-3%). Absolute thrombocytopenia (platelet count mon with an incidence of ~0.5%. Another system, the HIT
<150,000/µL) is not required for a diagnosis of HIT; rather, expert probability score, has also been developed, although
a substantial (>50%) decrease in the platelet count from the the clinical experience with this system is not as extensive. The
highest platelet count after initiation of heparin is required. impact of either scoring system on patient outcomes has not
This is particularly relevant to the postoperative setting, in been determined. While a low 4T score (<4 points) makes
which platelet count values typically rise to 20%-30% above HIT unlikely, HIT may nevertheless be the underlying cause
the preoperative baseline at day 8-10 after major surgery. in 2%-3% of patients. However, a combination of a low score
Uncommonly, HIT may develop 2-3 weeks after prior and a negative PF4-heparin ELISA essentially rules out HIT.
Table 10-4 4Ts scoring system for HIT
4Ts 2 points 1 point 0 point
Thrombocytopenia Platelet count decrease of >50% and Platelet count decrease of 30%-50% or Platelet count fall of <30% or
platelet nadir ≥20,000/µL platelet nadir of 10,000-19,000/µL platelet nadir <10,000/µL
Timing of platelet Clear onset of thrombocytopenia 5-10 days Consistent with day 5-10 decrease Platelet count decrease
count fall after heparin administration; or platelet but not clear (eg, missing platelet <4 days without recent
decrease within 1 day, with prior counts) or onset after day 10; or exposure
heparin exposure within 30 days decrease within 1 day, with prior
heparin exposure 30-100 days ago
Thrombosis or other New thrombosis (confirmed); skin necrosis Progressive or recurrent thrombosis; None
sequelae (lesions at heparin injection site); acute nonnecrotizing skin lesions;
systemic reaction after intravenous suspected thrombosis (not
unfractionated heparin bolus proven)
Other causes for None apparent Possible Definite
thrombocytopenia
Adapted from Lo G et al. J Thromb Haemost. 2006;4:759-765.

Thrombosis is present in ~50% of newly diagnosed cases detects IgG, IgA, and IgM antibodies, also increases specific-
of HIT, and it develops in ~40% of patients with asymptom- ity, as may the addition of a confirmatory step performed in
atic thrombocytopenia resulting from HIT within the first the presence of high heparin concentrations.
10 days following heparin discontinuation. Venous thrombosis Functional assays have improved specificity compared
occurs twice as frequently as arterial thrombosis, although with the ELISA. These assays are technically difficult, how-
limb artery thrombosis, myocardial infarction, and micro- ever, requiring washed donor platelets, and for the SRA,
vascular thrombosis have been described. HIT-associated radioisotope. Because of these considerations, the perfor-
thrombosis occurs with increased frequency at sites of vessel mance of functional assays is limited primarily to specialized
injury (eg, central venous catheter-associated deep vein reference labs, and their results generally are not available at
thrombosis in intensive care patients). For this reason, vas- the time the diagnosis of HIT must be considered. They are,
cular interventional procedures (other than arterial throm- however, important for confirming the diagnosis and for
bectomy) and placement of intravascular devices such as long- term management since patients without HIT may be
vena caval filters should generally be avoided. Adrenal harmed by being incorrectly labeled as having a history of
infarction secondary to adrenal vein thrombosis, skin necro- HIT with consequent avoidance of heparin and unnecessary
sis at heparin injection sites, and anaphylactoid reactions use of alternative anticoagulants.
after an intravenous heparin bolus also may occur as a result
of PF4/heparin antibodies. Thrombosis in unusual sites,
Treatment of HIT
such as cerebral sinuses, vascular grafts, and fistulas, and vis-
ceral vessels may also develop. Phlegmasia due to occlusion Although previously underdiagnosed, increased apprecia-
of the lower-extremity venous system resulting in arterial tion of HIT and the frequent use of highly sensitive tests has
insufficiency is a typical complication when vitamin K led to overdiagnosis in the current era, with the attendant
antagonists are started too early in HIT (ie, prior to platelet costs and increased bleeding risks associated with inappro-
count recovery). The resulting protein C deficiency triggers priate anticoagulation therapy. Current guidelines of the
microvascular thrombosis distal to large vessel thrombosis, American College of Chest Physicians suggest that routine
which may have occurred as an initial manifestation of HIT. monitoring of the platelet count in patients on heparin ther-
Very severe HIT can be associated with disseminated intra- apy should be performed every 2-3 days for patients with a
vascular coagulation (DIC). Patients with DIC often present risk of HIT of >1% and that routing monitoring is unneces-
with platelet counts below 20 × 109/L, whereas otherwise a sary for those in whom the risk of HIT is <1% (Table 10-5).
platelet count nadir of 40 × 109 to 80 × 109/L is the more typi- Because typical-onset HIT begins no earlier than day 5 of
cal range for HIT. heparin treatment and because reactive thrombocytosis after
surgery needs to be considered in order to detect a >50%
HIT testing
Table 10-5 Incidence of HIT according to patient population and
Two types of tests are available for detection of HIT antibod-
type of heparin exposure
ies: quantitative PF4/heparin immunoassays (PF4/heparin
ELISA) and functional assays demonstrating the ability of Patient population (minimum 4 days’
HIT antibodies to activate washed platelets, such as the sero- exposure) Incidence of HIT (%)
tonin release assay (SRA), generally considered the gold Postoperative patients
standard for diagnosis, or heparin-induced platelet activa- Heparin, prophylactic dose 1-5
tion (HIPA). Heparin, therapeutic dose 1-5
The sensitivity of the PF4/heparin ELISA approaches Heparin, flushes 0.1-1.0
LMWH, prophylactic or therapeutic dose 0.1-1.0
100%, and thus a negative test is useful in excluding HIT.
Cardiac surgery patients 1-3
Difficulties concerning its use include long turnaround time
Medical
in institutions in which it is not performed daily, and its low
Patients with cancer 1.0
specificity and positive predictive value, particularly in the Heparin, prophylactic or therapeutic dose 0.1-1.0
post-cardiac surgery setting. Specificity may be increased by LMWH, prophylactic or therapeutic dose 0.6
considering the level of positivity. High ELISA reactivity cor- Intensive care patients 0.4
relates closely with the presence of platelet-activating HIT IgG, Heparin, flushes <0.1
whereas positive platelet activation studies are uncommon Obstetric patients <0.1
in patients with weakly positive ELISA optical density val- Adapted from Linkins LA et al. Chest. 2012;141(2)(suppl):e495S-e530S.
ues (0.4-0.9). The use of an ELISA that detects only anti- HIT = heparin-induced thrombocytopenia; LMWH = low-molecular-
PF4/heparin IgG, as opposed to the polyspecific ELISA that weight heparin.

relative fall in the platelet count, monitoring the platelet argatroban to warfarin should be performed by following
count on days 5, 7 and 9 after surgery in high-risk patients is the guidelines suggested by the manufacturer. Bivalirudin is
generally sufficient. If the platelet count fall begins on or approved for percutaneous coronary interventions in
before day 9, it can be detected by these monitoring time patients with HIT or a history of HIT and has the advantage
points. When a new thrombosis occurs after day 9, compari- of a short half-life of only 25 minutes. A common issue of
son of the platelet count at the time of thrombosis with the both direct thrombin inhibitors is that they are subject to
one obtained on day 9 facilitates recognition of HIT, while aPTT confounding, a phenomenon in which patients with
comparison with the preheparin baseline platelet count clotting factor deficiency (due to liver impairment, warfarin
could underestimate the magnitude of the platelet count fall. treatment, or consumptive coagulopathy) have resultant
However, platelet count monitoring rarely helps to prevent prolongation of the APTT, leading to underdosing of antico-
initial thrombosis in HIT, as the time between the fall in agulation. Use of an ecarin-based assay or dilute thrombin
platelet count and onset of thrombosis can be very short, or time assay provides more reliable results, but these tests are
both may occur concomitantly. not available in many centers. Other anticoagulants such as
The cornerstone of HIT therapy is immediate discontinu- danaparoid and lepirudin are no longer available in the
ation of heparin when the disease is suspected, usually before United States. A number of reports have described the use of
laboratory diagnosis. Some experts recommend 4-limb the synthetic pentasaccharide fondaparinux in patients with
ultrasound in patients with HIT because silent DVT is com- HIT, although this agent has not been studied in a controlled
mon and may influence the duration of anticoagulation. manner. The direct oral FXa inhibitors (rivaroxaban, apixa-
Anticoagulation using a nonheparin anticoagulant in thera- ban, edoxaban) or the direct oral thrombin inhibitor dabiga-
peutic dose should be initiated even in patients with no tran may be an option in HIT, but apart from a few case
thrombosis because of the continued high risk of thrombosis reports, no systematic data are available. Because the plasma
after heparin discontinuation. Alternative anticoagulation in levels of these drugs change considerably between peak and
patients without thrombosis should be continued until the trough, there is a risk that the highly prothrombotic state of
platelet count has normalized (ie, it has reached a stable pla- acute HIT could lead to breakthrough thrombosis at drug
teau for two consecutive days). Some advocate a longer trough levels. A similar explanation may underlie the fail-
duration of anticoagulation (eg, 30 days), although no con- ure of dabigatran in patients with mechanical heart valves.
trolled data demonstrating the benefit of this approach are The oral direct thrombin and FXa inhibitors may be used
available. Patients with HIT and thrombosis should receive once the acute phase of HIT is resolved (as signified by plate-
at least 3 months of therapeutic dose anticoagulation. let count recovery) or in patients with a history of HIT.
LMWH must not be used because of cross-reactivity with HIT antibodies are transient and typically vanish within
most heparin-dependent antibodies. Warfarin must not be 3 months after discontinuation of heparin. Once antibodies
given in acute HIT. It may be started once the platelet count disappear (ie, HIT laboratory testing becomes negative), it is
has reached a stable plateau, indicating that the acute pro- safe to re-expose patients to heparin during a cardiovascular
thrombotic process is under control, but only with appro- procedure or surgery. Heparin must be limited to the intraopera-
priate overlap with an alternative parenteral anticoagulant. tive setting and scrupulously avoided before and after surgery.
Warfarin leads to hypercoagulability because of the inhibi-
tion of protein C γ-carboxylation, which increases the risk Key points
for microvascular thrombosis. Patients who develop HIT
while on warfarin or who have been started on warfarin • HIT occurs in 0.2%-5% of adults exposed to UFH, approxi-
mately 40%-50% of whom develop thrombosis.
alone should be treated with vitamin K in addition to a non-
• HIT antibodies are directed against large multimolecular
heparin anticoagulant.
complexes of PF4/heparin (or other polyanions).
Currently available nonheparin anticoagulants in the
• Systematic scoring systems facilitate estimation of the pretest
United States include argatroban and bivalirudin, both of probability of HIT. A low 4T score (<4 points) makes HIT very unlikely
which are direct thrombin inhibitors. Argatroban is hepati- and, together with a negative PF4-heparin ELISA, rules out HIT.
cally cleared and approved for treatment of HIT with or • Functional assays, including the 14C-serotonin release assay,
without thrombosis, as well as percutaneous coronary inter- and the HIPA test are required to confirm the diagnosis of HIT.
ventions in patients with HIT or at risk for HIT. The use of • When HIT is suspected, heparin must be discontinued and a
argatroban in HIT is associated with a hazard ratio of 0.3 for nonheparin anticoagulant initiated in therapeutic dose (unless
the development of new thrombosis. Argatroban is moni- there is a substantial risk for bleeding).
tored using the activated partial thromboplastin time • Warfarin must not be started in acute HIT but may be given
(aPTT); but in conjunction with warfarin, it may have sig- for long-term anticoagulation once the platelet count has
normalized at a stable plateau for two consecutive days.
nificant effects on the PT. Thus, transitioning patients from

Other causes of thrombocytopenia diarrhea, however, whereas up to 30% of aHUS patients

may provide such a history; thus, the presence or absence of
Thrombotic microangiopathies diarrhea does not always distinguish these disorders. Renal
insufficiency is usually the most prominent component of
Clinical case typical HUS.
A 17-year-old female is referred for evaluation of renal insuf- Distinguishing between different TMAs may be difficult
ficiency and anemia. She and her siblings were placed in foster because of extensive overlap in symptoms. Although neuro-
care while they were very young, and she has no information on logic manifestations may be more frequent in TTP and renal
the health of her parents or older relatives. Her renal function failure is more common in HUS and aHUS, these character-
was first noted to be abnormal 1 year ago and over the last istics alone do not allow discrimination. Recent scientific
2 months she has developed profound fatigue. Her 22-year-old advances have led to new information concerning the patho-
sister is married and in good health. Her 15-year-old brother genesis of these diseases and development of diagnostic
also has been noted to have mildly abnormal renal function as
tests. For example, TTP is associated with deficiency of
well as significant anemia. On examination, she appears fatigued
ADAMTS13, while mutations in complement regulatory
and pale. There is no organomegaly. The complete blood
proteins can be identified in 50%-70% of cases of aHUS.
count reveals hemoglobin of 8.5 gm/dL, WBC of 9.1 × 109/L,
and a platelet count of 77 × 109/L. The lactic dehydrogenase
These findings have facilitated the development of patho-
(LDH) is elevated at 632 IU/L. The peripheral blood film reveals genesis-based classification schemes for TMAs; an example
1-2 schistocytes per high-power field. Subsequent evaluation of one scheme developed by the British Committee for Stan-
including sequencing of complement regulatory genes reveals a dards in Haematology and the British Transplantation Soci-
mutation in factor H. ety is depicted in Table 10-6.
Clinical features Pathogenesis

The thrombotic microangiopathies discussed in this chapter TMAs cause microvascular thrombi in critical organs, lead-
include thrombotic thrombocytopenic purpura (TTP) and ing to ischemia and organ damage. These thrombi induce
the typical and atypical hemolytic uremic syndrome (HUS shearing of red blood cells, leading to the characteristic
and aHUS, respectively). Each of these disorders is charac- schistocytic anemia. Endothelial cell activation or damage
terized by microangiopathic hemolytic anemia (MAHA) also promotes TMA, leading to the elaboration of unusually
and thrombocytopenia, with a variable component of neu-
rologic or renal dysfunction and fever. This pentad of symp-
Table 10-6 Classification scheme for thrombotic microangiopathies
toms was once common at the time of presentation, but
increased awareness of these disorders has led to earlier diag- Disorders in which etiology is established
nosis. Currently, the presence of schistocytic anemia and ADAMTS13 abnormalities
thrombocytopenia is sufficient for the diagnosis of throm- ADAMTS13 deficiency secondary to mutations
Antibodies against ADAMTS13
botic microangiopathy (TMA).
Disorders of complement regulation
TTP occurs in both a rare inherited form called Upshaw- Genetic disorders of complement regulation
Schulman syndrome due to mutations in the vWF-cleaving Acquired disorders of complement regulation (eg, factor
protease, ADAMTS13 (a disintegrin and metalloprotease H antibody)
with thrombospondin-1-like repeats), as well as a more Infection induced
common acquired form in which ADAMTS13 deficiency is Shiga toxin- and verotoxin- (Shiga-like toxin) producing bacteria
caused by autoantibodies. Patients with TTP generally pres- Streptococcus pneumoniae
Defective cobalamin metabolism
ent acutely or subacutely with fatigue and malaise, with
Quinine induced
variable neurologic symptoms that may range from mild
Disorders in which etiology is not well understood
personality changes to obtundation. Renal insufficiency
HIV
may or may not be present. aHUS presents in a similar Malignancy
manner, but it may demonstrate a more chronic presenta- Drugs
tion with progressive renal insufficiency, low-grade MAHA, Pregnancy
and thrombocytopenia. Neurologic defects are less com- Systemic lupus erythematosus and antiphospholipid antibody
mon in aHUS than in TTP. Typical HUS follows infection syndrome
with enteropathogenic Escherichia coli, may occur in epi- Adapted from Taylor CM et al. Br J Haematol. 2009;148:37-47.
demics, and often is preceded by bloody diarrhea and ADAMTS13 = a disintegrin and metalloprotease with
abdominal pain. Not all patients with typical HUS have thrombospondin-1-like repeats.

Other causes of thrombocytopenia | 263
large vWF multimers that enhance platelet aggregation and

microvascular occlusion.
TTP results from an inherited or acquired deficiency of
ADAMTS13, leading to elevated levels of unusually large
vWF multimers that induce platelet aggregation in the ADAMTS13
microvasculature. ADAMTS13 regulates vWF activity by
cleaving high-molecular-weight multimers; failure to do so Blood
flow
may result in the microvascular thrombosis and ischemia
characteristic of TTP (Figure 10-2). The observation that Normal hemostasis
some patients with ADAMTS13 deficiency do not have clini-
cal manifestations of TTP suggests that factors other than
ADAMTS13 deficiency, such as endothelial damage or acti-
vation, are also needed to trigger TTP. Other TTP-like syn- Vessel wall No
dromes can be caused by drugs, including quinine, ADAMTS13
ticlopidine, clopidogrel, cyclosporine, tacrolimus, mitomy-
cin C, and gemcitabine, or may occur in the setting of bone
marrow transplantation, systemic lupus erythematosus, dis-
seminated malignancy, and HIV infection. The pathogenesis TTP
of these syndromes is diverse; whereas some are associated
with antibodies to ADAMTS13, others are not and may Figure 10-2 Pathogenesis of idiopathic TTP caused by ADAMTS13
result from direct endothelial cell toxicity. deficiency. Multimeric vWF adheres to endothelial cells or to
Typical HUS results from infection by enteropathogenic connective tissue exposed in the vessel wall. Platelets adhere to
E. coli, most commonly serotype O157:H7. The capacity of vWF through platelet membrane GPIb-IX. In flowing blood,
organisms to cause HUS reflects their production of two vWF in the platelet-rich thrombus is stretched and cleaved by
ADAMTS13, limiting thrombus growth. If ADAMTS13 is absent,
70-kDa bacterial exotoxins called verotoxins. Verotoxin-1 is
vWF-dependent platelet accumulation continues, eventually causing
homologous to a Shigella toxin and therefore generally is
microvascular thrombosis and TTP. Redrawn from Sadler JE. Blood.
referred to as Shiga-like toxin 1 (SLT-1 or Stx1). Most strains
2008;112:11-18. ADAMTS13 = a disintegrin and metalloprotease
of pathogenic E. coli produce a second toxin, Stx2, which is with thrombospondin; GP = glycoprotein; TTP = thrombotic
associated with a higher risk of developing HUS. The intact, thrombocytopenic purpura; vWF = von Willebrand factor.
70 kD Stx holotoxin consists of a 32-kDa A subunit and five
7.7-kDa B receptor-binding subunits that bind globosyltri-
aosylceramide (Gb3; CD77) receptors expressed on capillary proteins is required to prevent complement-mediated injury.
endothelium. Following binding to Gb3, the toxin is inter- AP activation leads to the generation of the C5b-C9 lytic
nalized. The A subunit is proteolyzed to a 27-kDa A1 subunit complex on cell surfaces, and in the case of aHUS, endothe-
that binds the 60s ribosomal subunit, inhibiting protein syn- lial cell damage is the primary consequence, resulting in
thesis and inducing endothelial cell apoptosis. Recent studies characteristic microvascular thrombotic lesions. Comple-
have demonstrated that signal transduction initiated through ment activation is regulated primarily by the plasma protein,
cross-linked Stx B subunit/Gb3 complexes induce the release factor H, and the membrane-associated membrane cofactor
of vWF from endothelial cells. Finally, Stx acts in concert protein (MCP; CD46); each of which binds membrane-
with lipopolysaccharide to trigger a procoagulant state that bound C3b and promotes its inactivation by factor I. Several
involves platelet activation, tissue factor induction, and the mutations in complement regulatory proteins underlie the
release of unusually large vWF multimers. development of aHUS. Most common are mutations in fac-
The pathogenesis of aHUS reflects increased activation of tor H, which impair the interactions of factor H with mem-
the alternative complement pathway (AP) because of muta- brane-bound C3b, and account for 30% of cases; an
tions or autoantibodies resulting in loss or functional additional 5%-10% of cases of aHUS result from acquired
impairment of complement regulatory proteins, or, less fre- antibodies to factor H. Mutations in CD46, usually impair-
quently, activating mutations in complement proteins them- ing membrane expression, are observed in 15% of patients
selves. Most hereditary forms of aHUS are transmitted in an with aHUS. Factor I mutations occur in 12% of aHUS
autosomal dominant manner, although penetrance is only patients. Activating mutations in factor B or C3 occur in
50%. Under normal conditions, the AP is constitutively acti- 5%-10% of patients with aHUS. Mutations in thrombomod-
vated because of ongoing C3 hydrolysis (Figure 10-3), and ulin, another complement regulatory protein, have been
thus tight regulation of the AP by complement inhibitory described.

A C
Protease Alternative pathway
C3 C3b + C3a C3 convertase
C3bBbP
C3
B C3
AP
AP
,
P
, B, D
P
B, D C3b
C3b
C3a and C5a
B, SP C3b + B C3bB AP
(Anaphylatoxins)
D, SP C3bB + D C3bBb C5b-C9
P, stabilizer C3bBb + P C3bBbP Endothelial cell membrane
Figure 10-3 The alternative pathway of complement activation. (A) The AP of the complement system originally consisted of a serine protease
that cleaved C3 to the opsonin C3b and the proinflammatory anaphylatoxin C3a. (B) An amplification loop was next evolved to more efficiently
deposit C3b on a target and liberate C3a into the surrounding milieu. B indicates factor B, D indicates factor D, a serine protease; P, properdin,
a stabilizer of the enzyme. (C) Development of a C5 convertase. The same enzyme that cleaves C3 (AP C3 convertase) can cleave C5 to C5a and
C5b with the addition of a second C3b to the enzyme complex (AP C5 convertase). Redrawn from Liszewski MK, Atkinson JP. Hematology Am
Soc Hematol Educ Program. 2011;2011:9-14. AP = alternative complement pathway; GP =glycoprotein.
Diagnosis therapy should not be withheld from such individuals. More-

over, recovery of ADAMTS13 levels during initial plasma
The diagnosis of TMA requires clinical awareness and prompt
exchange may lag behind clinical response and is not useful in
recognition of symptoms. TTP is more common in females,
determining the duration of plasma therapy.
with a peak incidence in the fourth decade; other risk factors
Patients with aHUS may present acutely, mimicking TTP,
include obesity and African ancestry. The diagnosis of TTP
or in some cases more insidiously with renal insufficiency as
should be suspected in patients with MAHA and thrombocy-
the primary symptom. Thrombocytopenia may be less severe
topenia without another apparent etiology, such as malignant in aHUS than TTP. A family history of similar disease may
hypertension, vasculitis, scleroderma renal crisis, tumor be apparent, although the low penetrance of complement
emboli, or disseminated intravascular coagulation. Fever and inhibitor mutations may make such a history difficult to dis-
neurologic symptoms may be present but are less common sect. Exacerbations of disease may follow infections and may
than they once were due to earlier diagnosis; evidence of renal be accompanied by fatigue and malaise. aHUS may present
involvement even in the absence of renal insufficiency some- in association with pregnancy, most commonly at 3-4 weeks
times can be obtained through examination of the urinary postpartum. Complement levels in patients with aHUS may
sediment. Schistocytes are invariably present and are accom- be decreased, but normal levels do not exclude aHUS.
panied by elevation of the LDH, which may be striking; levels Sequencing of complement inhibitor proteins are useful for
of unconjugated bilirubin also may be increased. Nucleated confirming a clinical impression of aHUS, but this sequenc-
red blood cells are frequently present. The PT, aPTT, and ing is performed only in specialized laboratories with a turn-
fibrinogen levels are typically normal, and the D-dimer is around time of several weeks.
normal or only mildly increased. The direct antiglobulin test Typical HUS is more common in the pediatric population
is negative. Consideration of secondary causes of TTP should than in adults, and it is the most common cause of acute
include a detailed drug history, HIV testing, and a focused renal failure in children. The disease begins with abdominal
search for autoimmune disease and malignancy. TTP may pain and watery diarrhea 2-12 days after toxin exposure.
present during pregnancy, particularly in the second and Bloody diarrhea generally ensues on the second day, though
third trimesters. ADAMTS13 assays may be useful in con- up to one-third of patients do not report blood in the stool.
firming the diagnosis of TTP when severe deficiency (<10%) Fever is typically absent or mild. The presentation may be
is present in the appropriate clinical setting. ADAMTS13 test- difficult to differentiate from inflammatory bowel disease,
ing may also provide prognostic information, with lower lev- appendicitis, ischemic colitis, or intussusception. Definitive
els of ADAMTS13 and higher levels of anti-ADAMTS13 diagnosis is made by culture of E. coli on sorbitol-MacCon-
antibodies associated with higher relapse rates. Many patients key agar. The presence of Shiga toxin or its structural genes
with TMA and detectable or even normal ADAMTS13 levels, may be detected by enzyme immunoassay or PCR of the
however, also respond to plasma exchange, and thus, this stool. Serologic studies demonstrating an increase in conva-

lescent antibody titer to Shiga toxin or E. coli lipopolysac- likelihood of dialysis-dependence. Therefore, eculizumab
charide may be useful in confirming the diagnosis. should be initiated promptly in patients with thrombotic
microangiopathy who do not have severe deficiency of
ADAMTS13, Shiga toxin-producing E. coli, or another
Management
apparent cause of TMA and who do not respond to plasma
Plasma exchange is the standard of care for treatment of exchange. Treatment should not be delayed until comple-
TMAs, particularly TTP. Untreated, TTP is associated with a ment mutation results are available because turnaround
mortality of approximately 85%, although 90% of patients time for this testing is several weeks and because patients
with TTP treated with plasma exchange survive. The superi- with clinical aHUS without identifiable mutations benefit
ority of plasma exchange over infusion was demonstrated in from eculizumab. The current standard of care is to con-
a randomized Canadian trial of 103 adults with TTP, tinue eculizumab indefinitely in patients with aHUS. Fur-
although patients randomized to the plasma exchange arm ther study is needed to determine whether it may be safe to
received more plasma. The exchange of 1 to 1.5 plasma vol- discontinue treatment in selected patients.
umes is standard initial treatment; however, larger volume Treatment of E. coli-associated typical HUS is generally
exchanges may have additional benefit in patients with an supportive; it was long assumed that the use of antibiotics may
inadequate response. Plasma exchange is continued daily lead to increased toxin release and worse outcome; however,
until the platelet count reaches normal levels (>150,000/µL), during an epidemic outbreak of typical HUS, antibiotic treat-
LDH normalizes, and symptoms have resolved. Neurologic ment was associated with reduced morbiditiy. Some patients
symptoms improve most rapidly. No evidence suggests a may require transfusion support and/or dialysis during the
benefit of either abrupt discontinuation or tapering of acute phase of their illness. A benefit for plasma exchange in
plasma exchange. Antiplatelet agents have not been shown to typical HUS has not been demonstrated. Immunoadsorption
be beneficial and may increase bleeding, although some using anti-IgG columns resulted in rapid reversal of severe
guidelines advocate their use in patients in whom the platelet neurological symptoms and normalization of renal function
count increases rapidly during plasma exchange. Corticoste- in patients with E. coli O104:H4-associated HUS.
roids are used initially in most patients with TTP because of
the presence of ADAMTS13 antibodies, although a signifi-
Key points
cant benefit has not been demonstrated consistently in ran-
domized studies. In recent years, the potential utility of • TTP, atypical HUS (aHUS), and typical (Shiga-like toxin; Stx)
rituximab in TTP has been revealed. In a single-arm study, HUS share many common features and may be difficult to
the use of rituximab in patients who did not respond rapidly distinguish from one another.
to plasma exchange (with plasma exchange continued), led • The pathogenesis of TTP involves deficiency of ADAMTS13,
usually because of acquired autoantibodies against ADAMTS13.
to more rapid resolution of TTP and a lower incidence of
This leads to accumulation of ultralarge vWF multimers that
relapse compared with historical controls. Other studies
induce platelet aggregation in the microcirculation.
have demonstrated the apparent efficacy of rituximab in
• aHUS involves excessive activation of the AP, leading to
relapsed TTP and the disappearance of ADAMTS13 anti- complement-mediated damage to vascular cells.
bodies following treatment. Other adjunctive therapies for • The pathogenesis of typical HUS reflects the toxic effects of
refractory TTP include immunosuppressive agents, such as Shiga toxin on vascular endothelium and other cell types.
cyclosporine and vincristine, as well as splenectomy, which • The treatment of choice for TTP is plasma exchange.
may decrease relapse rates. Platelet transfusion has been • Plasma exchange is effective in some cases of aHUS. Eculizumab
associated with a rapid decline in clinical status in occasional should be used in patients who do not respond to plasma
patients and is relatively contraindicated, although a retro- exchange.
spective analysis could not identify a clear association of • Plasma exchange is not effective in typical HUS, which is
platelet transfusion with poor outcomes. usually self-limited. Treatment is supportive.
Plasma exchange historically has been the treatment of

choice for aHUS as well as TTP, and it remains so in patients
Splenic sequestration
with TMA in whom a clear diagnosis of TTP or aHUS can-
not be established. Response rates to plasma exchange in Splenic enlargement, most commonly from cirrhosis and
patients with aHUS are not as robust as in TTP. Eculizumab, portal hypertension, results in sequestration of platelets in
an antibody against complement C5, has shown efficacy in the splenic vascular network, leading to mild to moderate
patients with aHUS, leading to its approval for aHUS treat- thrombocytopenia. Typical platelet counts in patients with
ment in 2011. Delays in initiation of eculizumab are associ- splenic sequestration are 60,000-100,000/µL. Other mecha-
ated with worse long-term renal function and a greater nisms associated with liver disease that may induce

thrombocytopenia include hepatitis C-induced secondary film demonstrates Döhle bodies in neutrophils (which are
ITP, suppression of platelet production by megakaryocytes best detected by immunofluorescence, Figure 10-4). Associ-
resulting from direct viral infection, and decreased produc- ated clinical features including hearing loss, cataracts, and
tion of TPO by the cirrhotic liver. Interferon-based therapy renal failure are present in some patients. Bleeding symp-
for viral hepatitis may also induce thrombocytopenia. toms are mild to moderate as platelet function is nearly
normal apart from a reduction in platelet cytoskeleton con-
traction with resulting reduced clot stability. About 30% of
Hereditary thrombocytopenia
patients have a de novo mutation and therefore a negative
Hereditary thrombocytopenic syndromes are uncommon but family history. Similar large platelets are found in a sub-
not as rare as often assumed. It is critical that treating physi- group of von Willebrand disease type IIB and in the autoso-
cians maintain a high index of suspicion for these disorders, mal recessive Bernard-Soulier syndrome (BSS), which is
as patients are often misdiagnosed as having ITP, resulting in characterized by the absence of the platelet GPIb-IX com-
unnecessary, ineffective, and potentially harmful treatments plex, lack of platelet aggregation by high-dose ristocetin, and
such as immunosuppression and splenectomy. In about 50% bleeding. Furthermore, mutations in the platelet cytoskele-
of affected families, at least one family member has been sple- ton proteins beta tubulin, filamin, and alpha actinin result in
nectomized to treat “ITP.” The diagnosis should be considered macrothrombocytopenia.
in any patient with a family history of thrombocytopenia, or Hereditary thrombocytopenias also occur in association
in patients with long-lasting “ITP” who do not respond to with mutations in specific transcription factors that regulate
standard therapy. Whenever possible, physicians should megakaryocyte and platelet production, including GATA1
attempt to document a historical normal platelet count in a (X-linked inheritance, dyserythropoiesis). Patients with the
patient with thrombocytopenia to exclude a hereditary Paris-Trousseau/Jacobsen syndrome, an autosomal domi-
thrombocytopenic disorder. The presence of anatomic nant macrothrombocytopenia, have psychomotor retarda-
defects, including absent radii (thrombocytopenia-absent tion and facial and cardiac abnormalities. This syndrome
radius [TAR] syndrome) or right-heart defects (DiGeorge arises because of deletion of a portion of chromosome 11,
syndrome), high-tone hearing loss, cataracts before age 50, or 11q23-24, that encompasses the gene encoding the tran-
interstitial nephritis support the diagnosis of hereditary scription factor friend leukemia integration 1 (FLI-1). Gray
thrombocytopenia. The blood smear also often helps to iden- platelets, often enlarged, result from mutations in the growth
tify patients with potential hereditary thrombocytopenia. For factor independent 1B (GFI1B-mutation, autosomal domi-
example, large platelets and neutrophil inclusions may indi- nant) and the NBEAL2 gene (recessive trait).
cate the presence of a MYH9-related disorder. Genetic testing Normal sized platelets are found in inherited thrombocy-
may be used to confirm the diagnosis. Although about 40% of topenias due to RUNX1 (autosomal dominant, increased
families with inherited thrombocytopenia do not have an risk for AML) and the ANKRD26 mutations.
identifiable gene defect, this is a rapidly evolving area. Small platelets are typical of Wiskott-Aldrich syndrome
Thrombocytopenia with large platelets: Many inherited (WAS), an X-linked disorder characterized by severe immu-
thrombocytopenias involve defects in platelet production, nodeficiency, small platelets, and eczema; and congenital
while megakaryocytopoiesis is largely normal. The platelet amegakaryocytic thrombocytopenia (CAMT), a recessive
mass is distributed to fewer platelets, which results in macro- disorder due to mutations in the c-Mpl receptor character-
thrombocytopenia. Automated particle counters often ized by severe thrombocytopenia, absence of megakaryo-
underestimate the platelet number by counting the large cytes in the bone marrow, and a risk of trilineage failure.
platelets as red cells or leukocytes. While platelet size may be Establishing the diagnosis of hereditary thrombocytope-
increased in ITP, the platelet population is typically heterog- nia may be difficult. Historically, demonstration of decreased
enous with large- and normal-sized platelets. As a rule of expression of platelet GPIb-IX using flow cytometry has
thumb, about five normal sized platelets fit into one red cell. been used to diagnose BSS. Clustering of myosin in granulo-
Any blood smear showing >60% large platelets is highly sus- cytes using an immunofluorescent antibody against non-
picious for a hereditary macrothrombocytopenia. muscle myosin heavy chain type IIa may aid in screening for
Autosomal dominant MYH9-related macrothrombocyto- MYH9-related disorders. Improvements in sequencing tech-
penic disorders are probably the most frequent of the nologies have allowed for the expansion of genetic analyses
inherited thrombocytopenias. They consist of the for- for BSS, MYH9-related thrombocytopenia, CAMT, GATA1-
merly distinguished May-Hegglin, Fechtner, Sebastian, and related thrombocytopenia, TAR syndrome, and WAS-
Epstein syndromes and are caused by mutations in the associated thrombocytopenia, and several laboratories in
MYH9 gene, which codes for nonmuscle myosin IIA. In the United States and Europe now provide these services
addition to macrothrombocytopenia, the peripheral blood (see http://www.genetests.org).

Y11K R702H N1424D E1841K
Figure 10-4 MYH9-associated macrothrombocytopenia. Heterozygous mutations in the MYH9 gene encoding nonmuscular myosin IIa are the
most frequent cause of hereditary macrothrombocytopenia. The mutated protein clusters in the cytoplasm of neutrophils. The inclusion bodies
differ in size and shape depending on the mutation. Large inclusion bodies result from mutations in the downstream part of the gene and are
visible by both light microscopy and immunofluorescence. More upstream mutations result in smaller inclusion bodies that are only visible with
immunofluorescence. (A) Peripheral blood film from a patient with a downstream mutation. A giant platelet is visible. Adjacent to the platelet
is a neutrophil containing a large blue Döhle body. Source: ASH Image Bank/Julie Braza. (B) Immunofluorescence stain from patients with
4 different mutations.
Infection-associated thrombocytopenia and

Key points hemophagocytic lymphohistiocytosis
• Splenic sequestration is a common cause of thrombocytopenia
Mild and transient thrombocytopenia occurs with many sys-
in patients with liver disease.
temic infections. Thrombocytopenia may be caused by a com-
• Failure to respond to standard ITP therapy (corticosteroids, IVIg)
should prompt consideration of a hereditary thrombocytopenia.
bination of mechanisms, including decreased platelet
• Genetic diagnosis of hereditary thrombocytopenia should be production, increased destruction, and increased splenic
obtained when possible. sequestration. In viral infections, infection of megakaryocytes
may lead to suppression of platelet production; in rickettsial

infections, platelets may be consumed in vasculitic lesions; in concomitant hemorrhagic or thrombotic risk factors the
bacteremia, platelet consumption may result from DIC or patient may harbor. High-quality evidence linking a platelet
enhanced clearance of immune complex-coated platelets. HIV, count threshold with bleeding risk in ICU patients is lacking.
hepatitis C virus, and H. pylori are causes of secondary ITP. A Prophylactic platelet transfusion is generally given when the
rare and unusual cause of infection-related thrombocytopenia platelet count decreases to 10,000-20,000/µL. Patients with
is the hemophagocytic syndrome, also known as hemophago- bleeding or a planned invasive procedure may require a
cytic lymphohistiocytosis (HLH). This disorder may be inher- higher platelet count.
ited, occur in conjunction with rheumatologic disease or
malignancy, or occur in response to infection, with Epstein-
Barr virus (EBV) being the most common associated patho- Key points
gen. HLH is more common in children and is characterized by • Infection is a common cause of thrombocytopenia, particularly
persistent activation of macrophages and cytotoxic T-cells, in ICU patients, and can be induced by a variety of organisms.
leading to damage of multiple organ systems. Thrombocyto- • Thrombocytopenia in the ICU may arise from a number of
penia occurs in most patients, but usually in the context of etiologies. Treatment should be individualized based on the
bicytopenia or pancytopenia. Characteristic clinical and labo- etiology of thrombocytopenia and individual patient factors.
ratory features in addition to cytopenias include fever (due to
hypercytokinemia), hepatosplenomegaly (accumulation of
macrophages), elevated triglyceride levels (inhibition of lipo- Disorders of platelet function
protein lipase by elevated tumor necrosis factor-alpha levels),
elevated ferritin levels, elevated levels of the circulating α-chain Disorders of platelet function are characterized by variable
of the interleukin-2 (IL-2) receptor (increased secretion from mucocutaneous bleeding manifestations and excessive
macrophages), and low fibrinogen levels (release of tissue plas- hemorrhage following surgical procedures or trauma. Spon-
minogen activator [tPA] from activated monocytes), and low taneous hemarthrosis and deep hematomas are unusual in
or absent natural killer cell activity. Treatment of HLH includes patients with platelet defects. Most patients have mild to
suppression of hyperinflammation and reduction of activated moderate bleeding manifestations. A majority of patients,
cells by immunosuppressive therapy. Demonstration of hemo- but not all, have a prolonged bleeding time. Platelet aggre-
phagocytosis on tissue or bone marrow biopsies is useful but gation and secretion studies provide evidence for the defect
not required. In cases associated with EBV, therapy is directed but generally are not predictive of the severity of clinical
toward eradication of EBV-infected cells. manifestations. Defects in platelet function may be inher-
ited or acquired, with the latter being far more commonly
encountered.
Thrombocytopenia in the critically ill
This topic is covered in greater detail in Chapter 2. Approxi-
Inherited disorders of platelet function
mately 40% of patients in medical or surgical intensive care
units (ICUs) develop a platelet count <150,000/µL; 20%-
25% develop a platelet count <100,000/µL; and 12%-15% Clinical case
develop severe thrombocytopenia, with a platelet count A 9-year-old girl is referred by her pediatrician for evaluation
<50,000/µL. The development of thrombocytopenia in of long-standing easy bruising and recurrent epistaxis. She
patients in the ICU is a strong independent predictor for has not had any surgery. The physical examination reveals
ICU mortality. The spectrum of disorders that cause throm- scattered bruises on the lower extremities. The platelet count
bocytopenia in this setting is extensive and includes DITP, is 190 × 109/L, and the hemoglobin is 11 g/dL. Plasma levels
infection, DIC, surgery, hemodilution, extracorporeal cir- of factor VIII, vWF antigen, and ristocetin cofactor are within
cuitry/intravascular devices (eg, cardiopulmonary bypass, normal range. The hematologist recommends platelet aggrega-
intra-aortic balloon pumps, extracorporeal membrane oxy- tion studies. These studies reveal abnormal platelet aggregation
responses upon activation—a primary wave but no secondary
genation), HIT, among others. Management is highly depen-
wave in response to ADP and epinephrine and decreased aggre-
dent on the etiology of thrombocytopenia. For example,
gation with collagen. The response to ristocetin is normal. The
whereas platelet transfusion and suspension of anticoagulant
patient is diagnosed with a platelet secretion defect.
prophylaxis may be indicated in a patient with DITP, HIT
requires prompt initiation of an alternative nonheparin anti-
coagulant and constitutes a relative contraindication to Table 10-7 provides a classification of inherited disorders
platelet transfusion. Treatment must therefore be individual- associated with impaired platelet function (Figure 10-1). Of
ized to the underlying cause of thrombocytopenia and any note, not all of these disorders are due to a defect in the

Disorders of platelet function | 269
Table 10-7 Inherited disorders of platelet function subendothelium is abnormal. There are three disorders in
1. Defects in platelet-vessel wall interaction (disorders of adhesion) this group: vWD, resulting from a deficiency or abnormality
a. von Willebrand disease (deficiency or defect in plasma vWF) in plasma vWF and BSS, in which platelets are deficient in
b. Bernard-Soulier syndrome (deficiency or defect in GPIb) GPIb (and GPV and GPIX) produce a bleeding phenotype
c. GPVI deficiency because the platelet–vWF interaction is compromised. Rare
2. Defects in platelet-platelet interaction (disorders of aggregation) patients with deficiency of GPVI, the collagen receptor on
a. Congenital afibrinogenemia (deficiency of plasma fibrinogen) platelets, have a mild bleeding disorder due to impaired
b. Glanzmann thrombasthenia (deficiency or defect in binding of platelets to the subendothelium. Binding of
GPIIb-IIIa)
fibrinogen to the GPIIb-IIIa complex is a prerequisite for
3. Disorders of platelet secretion and abnormalities of granules
platelet aggregation. Disorders characterized by abnormal
a. Storage pool deficiencies (δ, α, aδ)
b. Quebec platelet disorder
platelet–platelet interactions (aggregation disorders) arise
4. Disorders of platelet secretion and signal transduction because of a severe deficiency of plasma fibrinogen (congen-
a. Defects in platelet-agonist interaction (receptor defects) ital afibrinogenemia) or because of a quantitative or qualita-
(ADP, thromboxane A2, collagen, epinephrine) tive abnormality of the platelet membrane GPIIb-IIIa complex,
b. Defects in G-proteins (Gaq, Gas, Gai abnormalities) which binds fibrinogen (Glanzmann thrombasthenia). Patients
c. Defects in phosphatidylinositol metabolism and protein with defects in platelet secretion and signal transduction are
phosphorylation a heterogeneous group lumped together for convenience of
• Phospholipase C-b2 deficiency classification rather than an understanding of the specific
• PKC-θ deficiency underlying abnormality. The major common characteristics
d. Abnormalities in arachidonic acid pathways and thromboxane in these patients, as currently perceived, are abnormal aggre-
A2 synthesis
gation responses and an inability to release intracellular
• Phospholipase A2 deficiency
granule (dense) contents upon activation of platelet-rich
• Cyclooxygenase deficiency
• Thromboxane synthase deficiency
plasma with agonists such as ADP, epinephrine, and colla-
5. Disorders of platelet coagulant-protein interaction gen. In aggregation studies, the second wave of aggregation is
(Scott syndrome) blunted or absent.
6. Defects related to cytoskeletal/structural proteins The patient described in the clinical case at the beginning
a. Wiskott-Aldrich syndrome of this section falls in this heterogeneous large group of
b. b1-Tubulin deficiency “platelet secretion defects”; the platelet dysfunction may
c. Actinin deficiency arise from a variety of mechanisms. A small proportion of
d. Nonmuscular myosin IIa mutations these patients have a deficiency of dense granule stores (stor-
e. Filamin deficiency age pool deficiency). In other patients, the impaired secre-
7. Abnormalities of transcription factors leading to functional tion results from aberrations in the signal transduction
defects
events or in pathways leading to thromboxane synthesis that
a. RUNX1 (familial platelet dysfunction with predisposition to
govern end-responses, such as secretion and aggregation.
acute myelogenous leukemia)
b. GATA1
The findings on aggregation studies are nonspecific, and it is
c. FLI-1 (Dimorphic dysmorphic platelets with giant α-granules not possible to pinpoint a specific molecular abnormality
and thrombocytopenia; Paris-Trousseau/Jacobsen syndrome) from the tracings. Another group consists of patients who
d. Growth factor independent 1B (GFI1B) have an abnormality in interactions of platelets with proteins
of the coagulation system; the best described is the Scott syn-
Modified with permission from Rao AK. Am J Med Sci. 1998;316:69-77.
FLI-1 = friend leukemia integration 1; GATA1 = sex-linked drome, which is characterized by impaired transmembrane
inheritance; GPIb = glycoprotein Ib; PKC = protein kinase C; migration of procoagulant–PS during platelet activation.
RUNX1 = autosomal dominant; vWF = von Willebrand factor. Defects related to platelet cytoskeletal or structural proteins
also may be associated with platelet dysfunction. Recent
studies document impaired platelet function associated with
platelet per se. Some, such as vWD and afibrinogenemia, mutations in transcription factors (eg, RUNX1, GATA1, and
result from deficiencies of plasma proteins essential for FLI-1) that regulate the expression of important platelet pro-
platelet adhesion or aggregation. Some of these disorders are teins. In addition to these groups, some patients have abnor-
distinctly rare but shed light on platelet physiology. In many mal platelet function associated with systemic disorders,
patients with inherited abnormalities in platelet aggregation such as Down syndrome, in which the specific aberrant
responses, the underlying molecular mechanisms remain platelet mechanisms are unclear. The prevalence and relative
unknown. In patients with defects in platelet–vessel wall frequencies of the various platelet abnormalities remain
interactions (adhesion disorders), adhesion of platelets to unknown.

Disorders of platelet adhesion and fibrinogen. In Glanzmann thrombasthenia, platelet

aggregation is impaired. Clot retraction, a function of the
Bernard-Soulier syndrome
interaction of GPIIb-IIIa with the platelet cytoskeleton, is
BSS, a rare autosomal recessive platelet function disorder, also compromised.
results from an abnormality in the platelet GPIb-IX complex, Binding of fibrinogen to the GPIIb-IIIa complex on plate-
which mediates the binding of vWF to platelets and thus let activation is required for aggregation in response to all
plays a major role in platelet adhesion to the subendothelium, physiologic agonists. Thus, the diagnostic hallmark of
especially at high shear rates. GPIb exists in platelets as a thrombasthenia is the absence or marked decrease of platelet
complex consisting of GPIb, GPIX, and GPV. There are aggregation in response to all platelet agonists (except risto-
approximately 25,000 copies of GPIb-IX on the surface of an cetin), with absence of both the primary and secondary wave
individual platelet, and these are reduced or abnormal in BSS. of aggregation. The shape change response is preserved.
Although GPV also is decreased in BSS platelets, it is not Platelet-dense granule secretion may be decreased with weak
required for platelet surface GPIb-IX expression. The bleed- agonists (eg, ADP) but is normal on activation with throm-
ing time is markedly prolonged, the platelet count is moder- bin. Heterozygotes have approximately half the number of
ately decreased, and platelets are markedly increased in size platelet GPIIb-IIIa complexes, but platelet aggregation
on the peripheral smear. In platelet aggregation studies, responses are normal. Although congenital afibrinogenemia
responses to ADP, epinephrine, thrombin, and collagen are is also characterized by absence of platelet aggregation, it can
normal and response to ristocetin is decreased or absent, a be distinguished from thrombasthenia by a prolonged PT,
feature shared with severe vWD. This is because ristocetin- aPTT, and thrombin time and absent fibrinogen. The diag-
induced platelet agglutination is mediated by binding of vWF nosis of thrombasthenia can be confirmed by demonstrating
to the GPIb complex. Unlike in vWD, however, plasma vWF decreased platelet expression of the GPIIb-GPIIIa complex
and factor VIII are normal in BSS, and the addition of normal using flow cytometry. A subgroup of disorders of the GPIIb-
donor vWF does not restore ristocetin-induced agglutination IIIa complex are inherited dominantly. These patients have
of platelets because of GPIb deficiency on the patient’s plate- moderately decreased platelet numbers and large platelets.
lets. Dense granule secretion on activation with thrombin The underlying cause is a mutation in the transmembrane or
may be decreased in these patients. The diagnosis of BSS is intracellular part of the integrin, which results in permanent
established by demonstrating decreased platelet surface GPIb, activation of the GPIIb-IIIa complex.
which can be performed using flow cytometry.
Disorders of platelet secretion and signal
von Willebrand disease transduction
See the section titled “von Willebrand disease” in Chapter 9. As a unifying theme, patients lumped in this remarkably het-
erogeneous group generally are characterized by impaired
dense granule secretion and the absence of a second wave of
Disorders of platelet aggregation aggregation upon stimulation of platelet-rich plasma with
ADP or epinephrine; responses to collagen, thromboxane
Glanzmann thrombasthenia
analog (U46619), and arachidonic acid also may be impaired.
Glanzmann thrombasthenia is a rare autosomal recessive Conceptually, platelet function is abnormal in these patients
disorder characterized by markedly impaired platelet aggre- either when the granule contents are diminished (storage
gation, a prolonged bleeding time, and relatively severe pool deficiency [SPD]) or when the mechanisms mediating
mucocutaneous bleeding manifestations compared with or potentiating aggregation and secretion are impaired
most other platelet function disorders. It has been reported (Table 10-7). Identification of the underlying defect is often
in clusters in populations in which consanguinity is com- very difficult and in many patients it is caused by a combina-
mon. Normal resting platelets possess approximately 50,000- tion of several minor abnormalities.
80,000 GPIIb-IIIa complexes on their surface. The primary
abnormality in Glanzmann thrombasthenia is a quantitative
Deficiency of granule stores
or qualitative defect in the GPIIb-IIIa complex, a heterodi-
mer consisting of GPIIb and GPIIIa whose synthesis is gov- SPD refers to deficiency in platelet content of dense granules
erned by distinct genes located on chromosome 17. Thus, (δ-SPD), α granules (α-SPD), or both types of granules
thrombasthenia may arise due to a mutation in either gene, (αδ-SPD).
with decreased platelet expression of the complex. Platelet Patients with δ-SPD have a mild to moderate bleeding dia-
aggregation is mediated through interaction of GPIIb-IIIa thesis associated with a prolonged bleeding time. In platelet

studies, the second wave of aggregation in response to ADP patients, aggregation responses to thrombin, collagen, and
and epinephrine is absent or blunted, and the collagen ADP are impaired.
response is impaired markedly. Normal platelets possess 3-8 The Quebec platelet disorder is an autosomal-dominant
dense granules (each 200-300 nm in diameter). Under the disorder associated with delayed bleeding and abnormal pro-
electron microscope, dense granules are decreased in δ-SPD teolysis of α-granule proteins (including fibrinogen, factor V,
platelets. By direct biochemical measurements, the total vWF, thrombospondin, multimerin, and P-selectin) result-
platelet and granule ATP and ADP contents are decreased ing from increased amounts of platelet urokinase-type plas-
along with other dense granule constituents including cal- minogen activator. Patients have normal to reduced platelet
cium, pyrophosphate, and serotonin. counts, proteolytic degradation of α-granule proteins, and a
δ-SPD has been reported in association with other inher- defective aggregation response with epinephrine.
ited disorders, such as Hermansky-Pudlak syndrome (HPS)
(oculocutaneous albinism and increased reticuloendothelial
Defects in platelet signal transduction and platelet
ceroid), Chediak-Higashi syndrome, WAS, TAR syndrome,
activation
and Griscelli syndrome. The simultaneous occurrence of
δ-SPD and defects in skin pigment granules, as in HPS, point Signal transduction mechanisms encompass processes that
to commonalities in the pathways that govern synthesis and are initiated by the interaction of agonists with specific plate-
trafficking of platelet-dense granules and melanosomes. let receptors and include responses such as G-protein activa-
In northwest Puerto Rico, HPS affects 1 of every 1,800 tion and activation of phospholipase C and phospholipase A2
individuals. There are at least nine known HPS-causing (Figure 10-1).
genes, with most patients having HPS-1 and being from Patients with receptor defects have impaired responses
Puerto Rico. HPS is autosomal recessive and heterozygotes because the platelet surface receptors for a specific agonist
have no clinical findings. In addition to albinism, most are decreased. Such defects have been documented in the
patients have congenital nystagmus and decreased visual P2Y12 ADP receptor (the receptor targeted by thienopyri-
acuity. Two additional manifestations associated with cer- dines such as clopidogrel), TxA2 receptor, collagen receptors
tain HPS subtypes are granulomatous colitis and pulmonary (GPIa-IIa and GPVI), and epinephrine receptor. Because
fibrosis. ADP and TxA2 play a synergistic role in platelet responses to
Chediak-Higashi syndrome is a rare autosomal recessive several agonists, patients with defects in the receptors for
disorder characterized by δ-SPD, oculocutaneous albinism, these agonists manifest abnormal aggregation responses to
immune deficiency, cytotoxic T and natural killer (NK) cell multiple agonists.
dysfunction, neurologic symptoms, and the presence of G proteins serve as a link between surface receptors and
giant cytoplasmic inclusions in different cells. It arises from intracellular effector enzymes, and defects in G protein acti-
mutations in the lysosomal trafficking regulator (LYST) gene vation can impair platelet signal transduction (Gaq, Gai1,
on chromosome 1. and Gas). As G proteins are required in many cell types,
Patients with gray platelet syndrome have an isolated defi- affected patients typically have syndromic phenotypes, often
ciency of platelet α-granule contents. The name refers to the associated with neural deficiencies.
initial observation that the platelets have a gray appearance Downstream abnormalities in platelet function have also
with a paucity of granules on the peripheral blood smear. been identified such as defects in phospholipase C activation
These patients have a bleeding diathesis, mild thrombocyto- (phospholipase C-b2 and PKC-θ), calcium mobilization,
penia, and a prolonged bleeding time. The inheritance pat- and pleckstrin phosphorylation. A major platelet response to
tern is variable; autosomal recessive, autosomal dominant, activation is liberation of arachidonic acid from phospholip-
and sex-linked patterns have been described. Causative ids and its subsequent oxygenation to TxA2, which plays a
mutations have been identified in the growth factor indepen- synergistic role in the response to several agonists. Patients
dent 1B (dominant) and NBEAL2 genes (recessive). Under have been described with impaired thromboxane synthesis
the electron microscope, platelets and megakaryocytes reveal because of congenital deficiencies of phospholipase A2,
absent or markedly decreased α-granules. The platelets are cyclooxygenase, and thromboxane synthase.
severely and selectively deficient in α-granule proteins,
including PF4, bTG, vWF, thrombospondin, fibronectin, fac-
Disorders of platelet procoagulant activities
tor V, and PDGF. In some patients, plasma PF4 and bTG are
raised, suggesting that the defect is not in their synthesis by Platelets play a major role in blood coagulation by providing
megakaryocytes but rather in their packaging into granules. the surface on which several specific key enzymatic reactions
Platelet aggregation responses are variable. Responses to occur. In resting platelets, there is an asymmetry in the dis-
ADP and epinephrine are normal in most patients; in some tribution of some of the phospholipids such that PS and

phosphatidylethanolamine are located predominantly on platelet transfusions is administration of desmopressin

the inner leaflet, whereas phosphatidylcholine has the oppo- (DDAVP), which shortens the bleeding time in some patients
site distribution. Platelet activation results in a redistribution with platelet function defects, depending on the platelet
with expression of PS on the outer surface, mediated by abnormality. Most patients with thrombasthenia do not
phospholipid scramblase. The exposure of PS on the outer show a shortening of the bleeding time following DDAVP
surface is an important event in the expression of platelet infusion, whereas responses in patients with signaling or
procoagulant activities. A few patients have been described secretory defects have been variable, with a shortening of the
in whom this process is impaired, and this is referred to as bleeding time in some patients. Recombinant factor VIIa has
Scott syndrome. In these patients, who have a bleeding dis- been approved for the management of bleeding events in
order, the bleeding time and platelet aggregation responses patients with Glanzmann thrombasthenia and has been used
are normal. in some other inherited defects. Hormonal contraceptives
and/or antifibrinolytic therapy are often effective for man-
agement of menorrhagia. Other minor mucosal bleeding
Other abnormalities
may respond to intranasal DDAVP or antifibrinolytic agents.
Platelet function abnormalities have been described in asso-
ciation with other entities such as WAS, an X-linked Key points
inherited disorder affecting T-lymphocytes and platelets
• Patients with inherited platelet defects typically have mucocu-
characterized by thrombocytopenia, immunodeficiency,
taneous bleeding manifestations.
and eczema. The bleeding manifestations are variable. Sev-
• Patients with BSS have thrombocytopenia, large platelet size,
eral platelet abnormalities, including dense granule and a defect in the platelet GPIb-V-IX complex, leading to
deficiency and deficiencies of platelet GPIb, GPIIb-IIIa, and impaired binding of vWF and adhesion.
GPIa, have been reported in WAS. WAS arises from • Patients with Glanzmann thrombasthenia have absent or
mutations in the gene coding for the WAS protein, which decreased platelet GPIIb-IIIa, leading to impaired binding of
constitutes a link between the cytoskeleton and signaling fibrinogen and absent aggregation to all of the usual agonists
pathways. Platelet dysfunction also occurs with mutations in except ristocetin.
b1-tubulin, a cytoskeletal protein. More recently, abnormal • Patients with δ-storage pool deficiency have decreased dense
platelet function has been documented in patients with granule contents; some patients may have associated albinism,
mutations in transcription factors RUNX1 (also called core- nystagmus, and neurologic manifestations.
• Patients with the gray platelet syndrome have decreased
binding factor A2), GATA1, and FLI-1. Patients with RUNX1
α-granule contents.
mutations have hereditary thrombocytopenia, platelet dys-
• In a substantial number of patients with abnormal aggrega-
function, and predisposition to acute leukemia.
tion responses, the underlying mechanisms are unknown. Some
patients have defects in platelet activation and signaling
Treatment of inherited platelet mechanisms.
function defects
Because of the wide disparity in bleeding manifestations,

Acquired disorders of platelet function
management needs to be individualized. A basic principle
in all patients with platelet disorders is to prevent iron- Alterations in platelet function occur in many acquired dis-
deficiency anemia. Red blood cells are required for sufficient orders of diverse etiology (Table 10-8), of which drugs are
hemostasis. Iron-deficiency anemia is common in this popu- the most frequent. Besides the typical antiplatelet drugs,
lation, especially in women of childbearing age. Oral iron nonsteroidal anti-inflammatory drugs, serotonin reuptake
supplementation may be insufficient to normalize iron inhibitors (and other antidepressants), and anticonvulsive
stores, and intravenous iron therapy is often required. Plate- drugs are the most widely prescribed. In most of the non-
let transfusions are indicated in the management of signifi- drug-induced acquired platelet function disorders, the spe-
cant bleeding and in preparation for surgical procedures. cific biochemical and pathophysiologic aberrations leading
Platelet transfusions are effective in controlling bleeding to platelet dysfunction are poorly understood. In some, such
manifestations but come with potential risks associated with as the myeloproliferative neoplasms (MPN), there is pro-
blood products, including alloimmunization in patients duction of intrinsically abnormal platelets by the bone mar-
lacking platelet GPs. For example, patients with Glanzmann row. In others, the dysfunction results from an interaction of
thrombasthenia and BSS may develop antibodies against platelets with exogenous factors, such as artificial surfaces
GPIIb-IIIa and GPIb, respectively, which compromise the (cardiopulmonary bypass), compounds that accumulate in
efficacy of subsequent platelet transfusions. An alternative to plasma due to impaired renal function, and antibodies, while

Table 10-8 Disorders in which acquired defects in platelet function by irreversibly binding to the platelet P2Y12 receptor at the
are recognized ADP-binding site. In contrast, ticragrelor is a reversible
Uremia inhibitor of platelet function that binds to P2Y12 at a site dif-
Myeloproliferative disorders ferent from ADP and allosterically blocks access of ADP to
Acute leukemias and myelodysplastic syndromes the receptor. These drugs prolong the bleeding time and
Dysproteinemias inhibit platelet aggregation responses to several agonists,
Cardiopulmonary bypass including ADP, collagen, epinephrine, and thrombin, to var-
Acquired storage pool deficiency
ious extents depending on agonist concentrations. The active
Acquired von Willebrand disease
drug/metabolites of the irreversible antiplatelet drugs (aspi-
Antiplatelet antibodies
Drugs and other agents rin, clopidogrel, prasugrel) disappear from the circulation
within relative short time (aspirin 1 hour; clopidogrel
~5 hours; prasugrel ~8-10 hours). However, ticagrelor
in liver disease platelet function is often secondarily impaired reaches such high plasma levels that it is present in the circu-
due to preactivation of platelets by low levels of thrombin. lation for 72-96 hours despite its relatively short half life of
As with inherited disorders of platelet function, in acquired 7-8 hours. This has major implications for patient manage-
disorders, bleeding is usually mucocutaneous with a wide ment. While hemostasis can be normalized after cessation of
and unpredictable spectrum of severity. The usual labora- the irreversible platelet function inhibitors within a reason-
tory tests that suggest platelet dysfunction include a pro- able time by platelet transfusion, they are rather ineffective in
longed bleeding time and abnormal results in studies of case of ticagrelor and any invasive procedures with increased
platelet aggregation or the platelet function analyzer (PFA)-100. bleeding risk should be postponed by 96 hours, if possible.
The bleeding time and the PFA-100 are not reliable discrimi- GPIIb-IIIa receptor antagonists are compounds that
nators, because these tests may be variably abnormal or nor- inhibit platelet fibrinogen binding and platelet aggregation.
mal, even in individuals with impaired platelet aggregation These include the Fab fragment of a monoclonal antibody
responses. In patients with acquired platelet dysfunction, the against the GPIIb-IIIa receptor (abciximab), a synthetic pep-
correlation between abnormalities in platelet aggregation tide containing the RGD sequence (eptifibatide), and a pep-
studies and clinical bleeding remains weak. tidomimetic (tirofiban). They are potent inhibitors of
aggregation in response to all agonists (except ristocetin)
and prolong the bleeding time. DITP (secondary to drug-
Drugs that inhibit platelet function
dependent antibodies) occurs in 0.2%-1.0% of patients on
Many drugs affect platelet function. Moreover, the impact of first exposure to GPIIb-IIIa antagonists (see the section
concomitant administration of multiple drugs, each with a “Drug-induced immune thrombocytopenia” earlier in this
mild effect on platelet function, is unknown. Because of their chapter).
widespread use, aspirin and nonsteroidal anti-inflammatory A host of other medications and agents, including onco-
agents are an important cause of platelet inhibition in clini- logic drugs (eg, mithramycin) and food substances, inhibit
cal practice. Aspirin ingestion results in inhibition of platelet platelet responses, but the clinical significance for many is
aggregation and secretion upon stimulation with ADP, epi- unclear. β-Lactam antibiotics, including penicillins and
nephrine, and low concentrations of collagen. Aspirin irre- cephalosporins, inhibit platelet aggregation responses and
versibly acetylates and inactivates platelet cyclooxygenase may contribute to a bleeding diathesis at high doses. The
(COX-1), leading to the inhibition of synthesis of endoper- platelet inhibition appears to be dose dependent, taking
oxides (prostaglandin G2 and H2) and TxA2. Typically, it is approximately 2-3 days to manifest and 3-10 days to abate
recommended to wait approximately 1 week after cessation after drug discontinuation. The clinical significance of the
of aspirin ingestion to perform studies intended to assess effect of antibiotics on platelet function remains unclear.
platelet function and elective invasive procedures to ensure The general context in which bleeding events are encoun-
that the antiplatelet effect is gone. Nonsteroidal anti-inflam- tered in patients on antibiotics prevents identification of the
matory drugs also impair platelet function by inhibiting the precise role played by the antimicrobials because of the pres-
cyclooxygenase enzyme and may prolong the bleeding time. ence of concomitant factors (eg, thrombocytopenia, DIC,
Compared with aspirin, the inhibition of cyclooxygenase by infection, vitamin K deficiency). Avoidance or discontinua-
these agents generally is short-lived and reversible (1-2 days). tion of a specifically indicated antibiotic usually is not
Cyclooxygenase-2 inhibitors do not inhibit platelet aggrega- necessary.
tion responses. Evidence is growing that selective serotonin reuptake
Ticlopidine, clopidogrel, and prasugrel are orally adminis- inhibitors (SSRIs) inhibit platelet function in vivo. Serotonin
tered thienopyridine derivatives that inhibit platelet function in plasma is taken up by platelets, incorporated into dense

granules, and secreted upon platelet activation. SSRIs inhibit Acute leukemias and myelodysplastic syndromes
the uptake of serotonin and platelet aggregation and secre-
The major cause of bleeding in these conditions is thrombo-
tion responses to activation. In epidemiologic studies,
cytopenia. In patients with normal or elevated platelet
patients on SSRIs have had increased gastrointestinal bleed-
counts, however, bleeding complications may be associated
ing and increased bleeding with surgery. Last, given the
with platelet dysfunction and altered platelet and mega-
increasing use of herbal medicines and food supplements,
karyocyte morphology. Acquired platelet defects associated
their role and interaction with pharmaceutical drugs needs
with clinical bleeding are more common in acute myeloge-
to be considered in the evaluation of patients with unex-
nous leukemia but also have been reported in acute lympho-
plained bleeding.
blastic and myelomonoblastic leukemias, hairy cell leukemia,
and myelodysplastic syndromes.
Myeloproliferative neoplasms
Bleeding tendency, thromboembolic complications, and Dysproteinemias

qualitative platelet defects are all recognized in MPNs, which
include essential thrombocythemia, polycythemia vera, Excessive clinical bleeding may occur in patients with dys-
chronic idiopathic myelofibrosis, and chronic myelogenous proteinemias, and this appears to be related to multiple
leukemia. Platelet function abnormalities result principally mechanisms including platelet dysfunction, specific coagu-
from development from an abnormal stem cell clone, but lation factor abnormalities, hyperviscosity, and alterations in
some alterations may be secondary to enhanced platelet acti- blood vessels because of amyloid deposition. Qualitative
vation in vivo. The clinical impact of in vitro qualitative platelet defects occur in some patients and have been attrib-
platelet defects, which are observed even in asymptomatic uted to coating of platelets by the paraprotein. The bleeding
patients, is unclear. tendency may be aggravated by concomitant inhibition of
Numerous studies have examined platelet function and vWF by the paraprotein.
morphology in patients with MPNs. Under the electron
microscope, findings include reduction in dense and
Uremia
α-granules, alterations in the open canalicular and dense-
tubular systems, and reduction of mitochondria. The bleed- Patients with uremia are at an increased risk for bleeding
ing time is prolonged in a minority (17%) of MPN patients complications. The pathogenesis of the hemostatic defect in
and does not correlate with an increased risk of bleeding. uremia remains unclear, but major factors include platelet
Platelet aggregation responses are highly variable in MPN dysfunction and impaired platelet–vessel wall interaction,
patients and often vary in the same patient over time. comorbid conditions, anemia, and the concomitant use of
Decreased platelet responses are more common, although medications that affect hemostasis. The bleeding time may
some patients demonstrate enhanced responses to agonists. be prolonged.
In one analysis, responses to ADP, collagen, and epineph- Multiple platelet function abnormalities are recognized
rine were decreased in 39%, 37%, and 57% of patients, in uremia, including impaired adhesion, aggregation, and
respectively. The impairment in aggregation to epinephrine secretion. These hemostatic defects may be linked to the
is more commonly encountered than with other agonists; accumulation of dialyzable and nondialyzable molecules in
however, a diminished response to epinephrine is not the plasma of patients with renal failure. One such com-
pathognomonic of a MPN. Platelet abnormalities described pound, guanidinosuccinic acid, accumulates in plasma,
in MPN include decreased platelet a2-adrenergic receptors, inhibits platelets in vitro, and stimulates generation of nitric
TxA2 production, and dense granule secretion and abnor- oxide, which inhibits platelet responses by increasing levels
malities in platelet surface expression of GPIIb-IIIa com- of cellular cyclic guanosine monophosphate.
plexes, GPIb, and GPIa-IIa. Aggressive dialysis ameliorates the uremic bleeding dia-
Platelets from patients with polycythemia vera and myelo- thesis in most patients. Hemodialysis and peritoneal dialysis
fibrosis, but not essential thrombocythemia or chronic are equally effective. Platelet transfusion is indicated in the
myelogenous leukemia, have been shown to have reduced management of acute major bleeds. Other treatments includ-
expression of the TPO receptor (Mpl) and reduced TPO- ing DDAVP, cryoprecipitate, and conjugated estrogens have
induced tyrosine phosphorylation of proteins. MPN patients also been shown to be beneficial. Elevation of the hematocrit
have been reported to have defects in platelet-signaling with packed red blood cells or recombinant erythropoietin
mechanisms. An acquired decrease in plasma vWF has been may shorten bleeding times, improve platelet adhesion, and
documented in some MPN patients with elevated platelet correct mild bleeding in uremic patients. The beneficial effect
counts and may contribute to the hemostatic defect. of red blood cells has been attributed to rheologic factors

Bibliography | 275
whereby the red blood cells exert an outward radial pressure supranormal platelet function. However, some may have
promoting platelet–vessel interactions. Other factors pre impaired platelet function due to antiplatelet antibodies that
disposing to bleeding in patients with renal failure include interfere with platelet function. Other patients may have
concomitant administration of antiplatelet agents or antico- autoantibodies which impair platelet function but do not
agulant medications. induce accelerated platelet clearance or thrombocytopenia.
Acquired SPD Key points
Several patients have been reported in whom dense granule • Alterations in platelet function are described in many disorders
SPD appears to be acquired. This defect probably reflects the of diverse etiologies; the clinical significance in terms of
release of dense granule contents because of in vivo platelet relationship to bleeding manifestations remains unclear in many.
• A careful drug history should be taken in any patient suspect-
activation or production of abnormal platelets. Acquired SPD
ed to have platelet dysfunction.
has been observed in patients with antiplatelet antibodies, sys-
• Aspirin, nonsteroidal anti-inflammatory agents, and other
temic lupus erythematosus, chronic ITP, DIC, HUS, renal
medications are a major cause of acquired platelet dysfunction.
transplant rejection, multiple congenital cavernous hemangi- • Patients with MPN may have altered platelet function that
omas, MPN, acute and chronic leukemias, severe valvular dis- contributes to the bleeding manifestations.
ease, and in patients undergoing cardiopulmonary bypass. • High platelet counts, as observed in MPN patients, may be
associated with a loss of high-molecular-weight multimers of
vWF in plasma.
Acquired von Willebrand disease
• Patients with renal failure may have impaired platelet function
Acquired vWD (AvWD) is an uncommon bleeding disorder. related to accumulation of substances in plasma that inhibit
Most patients are older (median age 62 years) and do not platelet function. Vigorous dialysis is a major part of manage-
have previous manifestations or a family history of a bleed- ment of the platelet dysfunction in these patients.
ing diathesis. Predisposing conditions include lymphoprolif-

erative disorders, plasma cell dyscrasias, cardiac disease,
MPN, and autoimmune disorders. Patients with MPN and Bibliography
reactive thrombocytosis demonstrate an impressive correla-
tion between the plasma vWF abnormalities and elevated
platelet counts. AvWD has been documented in patients Neunert C, Lin W, Crowther M, Cohen A, Solberg L Jr, Crowther
with severe aortic stenosis and congenital valvular heart dis- MA. The American Society of Hematology 2011 evidence-
ease and in those with left- ventricular assist devices (LVAD), based practice guideline for immune thrombocytopenia. Blood.
due to shear-stress induced loss of the high-molecular- 2011;117:4190-4207.
Provan D, Stasi R, Newland AC, Blanchette VS, Bolton-Maggs, P,
weight multimers of vWF from plasma. Laboratory findings
Bussel JB, Chong BH, Cines DB, Gernsheimer TB, Godeau B,
in AvWD include a prolonged bleeding time and decreased
Grainger J, Greer I, Hunt BJ, Imbach PA, Lyons G, McMillan
plasma levels of vWF and factor VIII. Most patients exhibit a R, Rodegheiro F, Sanz MA, Tarantino M, Watson S, Young J,
type 2 vWD pattern with a selective reduction in large vWF Kuter DJ. International consensus report on the investigation
multimers and a decreased vWF activity to antigen ratio. The and management of primary immune thrombocytopenia.
goals of treatment in AvWD are to raise plasma vWF levels Blood. 2010;115:168-186.
(using DDAVP or vWF-containing factor VIII concen-
trates), to treat or prevent bleeding, and to address the
underlying condition. Drug-induced immune thrombocytopenia
Arnold DM, Kukaswadia S, Nazi I, et al. A systematic
Antiplatelet antibodies and platelet function evaluation of laboratory testing for drug-induced immune
thrombocytopenia. J Thromb Haemost. 2013;11:169-176.
Binding of antibodies to platelets may produce several effects George JN, et al. Platelets on the Web. http://www.ouhsc.edu/platelets.
including accelerated destruction, platelet activation, cell
lysis, aggregation, secretion of granule contents, and out-
ward exposure of phosphatidylserine. In ITP, antibodies are Heparin-induced thrombocytopenia
directed against specific platelet surface membrane GPs that Linkins LA, Dans A, Moores LK, Bona R, Davidson BL,
play a major role in normal platelet function including GPIb, Schulman S, Crowther M; American College of Chest
GPIIb-IIIa, GPIa-IIa, GPVI, and glycosphingolipids. Patients Physicians. Treatment and prevention of heparin-induced
with ITP are generally assumed to have normal or thrombocytopenia: Antithrombotic therapy and prevention

of thrombosis, 9th ed: American College of Chest Physicians Inherited disorders of platelet function
evidence-based clinical practice guidelines. Chest. 2012;141
(2 suppl):e495S-e530S. Diz-Küçükkaya R. Inherited platelet disorders including
Warkentin TE, Greinacher A. Heparin-Induced Thrombocytopenia. Glanzmann thrombasthenia and Bernard-Soulier syndrome.
5th ed. Boca Raton, FL: CRC Press; 2013. Hematology Am Soc Hematol Educ Program. 2013;2013:
268-275.
Gresele P; Subcommittee on Platelet Physiology. Diagnosis of
Thrombotic microangiopathy inherited platelet function disorders: guidance from the SSC of
the ISTH. J Thromb Haemost. 2015;13:314-322.
Legendre CM, Licht C, Muus P, et al. Terminal complement
inhibitor eculizumab in atypical hemolytic-uremic syndrome.
N Engl J Med. 2013;368:2169-2181. Acquired disorders of platelet function
Scully M, Hunt BJ, Benjamin S, Liesner R, Rose P, Peyvandi F,
Cheung B, Machin SJ; British Committee for Standards in Abrams CS, Shattil SJ, Bennett JS. Acquired qualitative platelet
Haematology. Guidelines on the diagnosis and management of disorders. In: Kaushansky K, Lichtman MA, Beutler E, Kipps TJ,
thrombotic thrombocytopenic purpura and other thrombotic Seligsohn U, Prchal JT, eds. Williams Hematology. 8th ed. New
microangiopathies. Br J Haematol. 2012;158:323-335. York, NY: McGraw-Hill; 2010:1971-1991.
Diz-Küçükkaya R, López JA. Acquired disorders of platelet
function. In: Hoffman R, Benz EJ Jr, Silberstein LE, Heslop H,
Inherited thrombocytopenias Weitz J, Anastasi J, eds. Hematology: Basic Principles and Practice.
6th ed. Philadelphia, PA: Elsevier; 2013:1867-1882.e7.
Noris P, Balduini C. Inherited thrombocytopenias in the era of
personalized medicine. Haematologica. 2015;100:145-148.

CHAPTER
11
Laboratory hematology
John L. Frater and Anne M. Winkler
General concepts, 277 Specific laboratory tests, 277 Bibliography, 301
Terminology, 277 Hemostasis and thrombosis testing, 288

risk of an FP result increases (increasing sensitivity comes at
multimedia component. the cost of decreasing specificity). Very sensitive tests are
helpful for screening, by ruling out a diagnosis or disease
when the test is negative (high negative predictive value).
General concepts Specificity [TN/(TN + FP) × 100] is the ability of a test to
detect a normal result if the abnormality is not present; as
Hematology laboratory tests are ordered and interpreted the specificity increases, the risk of an FN result increases.
within the context of a specific patient; for example, a rou- Specific tests are useful for confirmation, by ruling in a
tine screening or preoperative assessment, or in the setting of diagnosis or disease when the test is positive (high positive
an illness for diagnosis or follow-up. Clinical judgment is predictive value).
applied in both the selection of tests and in their interpreta- Precision is reproducibility of a value during repeated testing
tion. Some unexpected results may require confirmation, of a sample.
particularly if there is a question about the integrity of the Accuracy is the ability of a test to obtain the assigned value
specimen (eg, heparin contamination, wrong collection tube of an external standard (run as though it were a clinical
or volume of blood, delayed processing). Additional causes sample).
of inaccurate laboratory results include sample mislabeling, Predictive value is the likelihood that an abnormal test indi-
analytical mistakes, and reporting errors. cates a patient with the clinical abnormality (positive pre-
dictive value [TP/(TP + FP) × 100]) or the likelihood that
a normal test indicates a patient without the abnormality
Terminology (negative predictive value [TN/(TN + FN) × 100]). Positive
and negative predictive values depend on the frequency of
Sensitivity, specificity, and positive or negative predictive val- the abnormality being sought in the population as well as
ues are defined using the following clinical variables: true the sensitivity and specific of the laboratory methods.
positive (TP; assay correctly identifies a condition in those Reference ranges are derived from a sample of a well population
who have it), false positive (FP; assay incorrectly identifies dis- and typically reflect the results of 95% (mean ±2 standard
ease when none is present), true negative (TN; assay correctly deviations [SDs]) of disease-free individuals. The reference
excludes a disease in those without it), and false negative (FN; ranges of some assays are determined by the results of 98%–
assay incorrectly excludes disease when it is present). 99% of disease-free individuals (mean ±3–5 SD).
Sensitivity [TP/(TP + FN) × 100] is the ability of a test to detect

a true abnormality; as the sensitivity of a test increases, the Specific laboratory tests
Automated blood cell counting
Conflict-of-interest disclosure: Dr. Frater declares no competing In addition to complete blood counts (CBCs) and five-part
financial interest. Dr. Winkler declares no competing financial interest.
Off-label drug use: Dr. Frater: not applicable. Dr. Winkler: not leukocyte differential counts, hematology analyzers provide
applicable. quantitative and qualitative information about reticulocytes,
| 277

278 | Laboratory hematology
nucleated red blood cells (NRBCs), and platelets. Additional

information, such as platelet immaturity, extended leukocyte
counts, and reticulocyte-specific indices, is available only
from selected instrument manufacturers. Because of the
large number of cells counted from each blood sample and
analysis using multiple physical principles and sophisticated
software, hematology analyzers generally produce accurate
and precise CBCs and leukocyte differential counts, with the
exception of basophils, because of their low frequency. Many
laboratories no longer report band neutrophils, because
accurate and precise identification by automated and mor-
phologic techniques is poor and their clinical significance, if
any, appears minimal with the possible exception of neonatal
sepsis. Hematology analyzers provide excellent sensitivity to
distinguish between normal and abnormal samples via oper-
ator alerts (flags) prompting microscopic review of a stained
peripheral blood smear for selected samples. As a result, a
variable percentage of hospital patients’ samples require
review of a stained blood smear.
Automated blood cell counters use various technologies
to enumerate and classify blood cells (Figure 11-1). Most
platforms available for clinical use utilize at least two of the
following techniques.
Aperture impedance (Coulter principle)
Cells diluted in a conducting solution are counted, and their Figure 11-1 Data and histograms performed on a Beckman-Coulter
LH 750 automated hematology analyzer from a healthy adult. The
volume is determined by measuring change in electrical
white blood cell (WBC), red blood cell (RBC), and platelet (PLT)
resistance as they flow through a narrow aperture and inter-
histograms represent cell volumes determined by impedance. The
rupt a direct electrical current. Software analysis defines second histogram from the top displays WBC light scatter in a flow
RBCs, white blood cells (WBCs), and platelets based on vol- cell; the y-axis indicates forward scatter and volume, and the x-axis
ume limits and calculates RBC and platelet indices. indicates side scatter due to granularity and nuclear features.
Basophils (BA) are detected by an alternative physical property not
displayed. EO = eosinophil; HCT = hematocrit; Hb = hemoglobin;
Optical absorbance
LY = lymphocyte; MCH = mean corpuscular hemoglobin; MCHC =
This technique exploits the cytochemical reaction of an mean corpuscular hemoglobin concentration; MCV = mean
intracellular enzyme, such as myeloperoxidase, to absorb corpuscular volume; MO = monocyte; MPV = mean platelet volume;
white light from a tungsten light source after the addition of NE = neutrophil; RDW = red blood cell distribution width.
a substrate. Light absorbance is proportionate to the inten-
sity of the enzyme-catalyzed reaction. This technique may be such factors as cytoplasmic granules and nuclear shape.
used to detect and distinguish peroxidase-containing cell All of the major hematology analyzer manufacturers use
types (neutrophils, eosinophils, monocytes) from peroxi- light-scattering technology.
dase- negative lymphocytes and basophils.
Fluorescence
Optical light scatter
In addition to the physical properties of cells, fluorochrome-
This method monitors the light-scattering properties of labeled antibodies recognizing cell surface or intracellular
blood cells, using a technique similar to that employed by epitopes further refine the separation of individual cell types.
flow cytometers. Cells pass in single file across the path of a A variety of reagents can be used to distinguish platelets (thi-
unifocal laser. The amount of light scattered at a low angle azole orange, anti-CD41, anti-CD42b, anti-CD61), reticulo-
from the incident light path is proportional to cell volume. cytes (thiazole orange, anti-CD4K, RNA dyes), fetal RBCs
The amount of light scattered at a wide angle depends on (antihemoglobin F/D), nucleated RBCs (propidium iodide,

Specific laboratory tests | 279
Draq 5, other DNA-binding dyes), neutrophils, lympho- Reticulocyte counts

cytes, and blasts.
Automated hematology analyzers use dyes or fluorecence
techniques to detect residual mRNA in young erythrocytes,
Erythrocyte analysis and all provide accurate reticulocyte counts expressed as a
percentage of RBCs or as an absolute number. Some blood
Automated blood cell counters measure the number (RBC
cell counters provide reticulocyte indices that are analogous
count, reported in units of 106/mL) and volume (mean cor-
to the standard RBC indices, including reticulocyte hemo-
puscular volume [MCV], reported in units of 10-15 L) of
globin content (CHr) on Advia analyzers (Siemens, Tarry-
RBCs, and hemoglobin concentration (reported in units
town, NY) and reticulocyte MCV (MCVr) on several others.
of g/dL) after lysing red blood cells; all other parameters are
Reductions in CHr and MCVr reflect inadequate hemoglo-
calculated. Hemoglobin is converted by potassium ferricy-
bin synthesis in real time, providing immediate information
anide to cyanmethemoglobin, and absorbance is measured
about functional iron deficiency when other biochemical
by a spectrophotometer at 540 nm. Some analyzers use
markers of iron availability may be difficult to interpret due
a c yanide-free method. RBCs may be spuriously increased in
to inflammatory conditions. CHr is particularly useful for
patients with hyperleukocytosis and giant platelets, and
assessing response to erythropoiesis-stimulating agents and
decreased in the presence of RBC agglutinins, cryoglobulins,
iron therapy in renal dialysis patients.
and in vitro hemolysis. Hemoglobin measurement can be
elevated artifactually by increased sample turbidity because
of leukocytosis, paraproteinemia, carboxyhemoglobinemia, Red blood cell fragments
hyperbilirubinemia, or hyperlipidemia.
The reliable identification of red blood cell fragments (schis-
MCV is calculated from the distribution of individual
tocytes) is important in the diagnosis of microangiopathies
RBC volumes. This measurement can be elevated artifactu-
such as hemolytic uremic syndrome, thrombotic thrombo-
ally by agglutination of RBCs, resulting in measurement of
cytopenic purpura, transplant associated thrombotic micro-
more than one cell at a time; hyperglycemia, causing osmotic
angiopathy, and disseminated intravascular coagulation.
swelling of the RBC; and spherocytes, which have decreased
The Sysmex XE and Siemens Advia systems both take advan-
deformity.
tage of the small size of red blood cell fragments to distin-
Automated hematocrit (%) is calculated by multiplying
guish them from normal red blood cells. Although both
the MCV by the RBC number: hematocrit = MCV (10-15 L)
platforms are noted to overestimate the number of red blood
× red blood cells (1012/L) × 100. Some analyzers directly
cell fragments in a specimen, this parameter can be used by
measure hematocrit.
the laboratory for identification of specimens needing
The mean corpuscular hemoglobin (MCH) is expressed
microscopic examination.
in picograms (10-12 g). The MCH is calculated by dividing
hemoglobin (g/L) by red blood cell count (1012/L). An
elevated MCH can be an artifact of increased plasma Nucleated red blood cells
turbidity.
The MCH concentration (MCHC) is expressed in grams Circulating NRBCs occur in newborns; however, beyond this
of hemoglobin per deciliter of packed RBCs. The MCHC is period, the presence of NRBCs is abnormal and is associated
calculated by dividing the hemoglobin concentration (g/dL) with various hematopoietic stresses, including hemolytic
by the hematocrit (%) × 100. Any artifact affecting the hema- anemias, myeloproliferative disorders, metastatic cancer to
tocrit or hemoglobin determinations can alter the MCHC. bone marrow, and hypoxia. All major hematology analyzer
The red blood cell distribution width (RDW) is the coef- brands enumerate NRBCs and correct WBC and lymphocyte
ficient of variation of RBC size (anisocytosis): standard devi- counts for interference from NRBC analysis.
ation/MCV. The RDW is used in the evaluation of anemia.
The RDW is more frequently elevated with microcytic ane-
Leukocyte analysis
mias due to iron deficiency anemia than to thalassemia or
anemia of chronic disease, and also is elevated more fre- To differentiate lymphocytes, monocytes, neutrophils, eosin
quently with macrocytic anemias due to vitamin B12 or folate ophils, and basophils, most instruments use impedance and/
deficiency than to liver disease, hypothyroidism, or a reticu- or light scattering, plus additional physical properties. Coul-
locytosis. Myelodysplastic syndromes, such as refractory ter and Sysmex use radiofrequency conductivity, and Advia
anemia, or RBC transfusions to patients with severe micro- (Siemens) uses peroxidase staining. Differentials typically
cytic or macrocytic anemias can produce a dimorphic RBC are reported as percentages of WBC and as absolute counts.
pattern with a very wide RDW. Automated blood cell counters provide sensitive flags and

warnings for immature granulocytes and monocytes and ×100), the optimal area to examine RBC, platelet, and leuko-
abnormal lymphocytes. Instrument manufacturers continue cyte morphologies and to perform WBC differentials is the
to refine technologies to report extended differentials to transitional area between the thick part of the smear and the
quantify neutrophil precursors, including metamyelocytes, leading edge (Table 11-1), where there are only a few over-
myelocytes, promyelocytes, and blasts. Some Sysmex analyz- lapping RBCs and central pallor of normal red blood cells is
ers identify a subset of WBCs called hematopoietic progeni- evident. Hematologists should review a patient’s peripheral
tor cells, which correlate with CD34 counts. smear as part of any consultation potentially involving quali-
tative or quantitative blood cell abnormalities.
The accuracy of manual WBC differentials suffers from
Platelet analysis
small sample size (typically 100 cells), distributional bias of
Automated blood cell counters enumerate platelets, measure WBCs on the smear, and variable interobserver agreement
volume, and calculate mean platelet volume (MPV). Asso- regarding cell classification. Advances in digital microscopy
ciations between MPV and acquired mechanisms of throm- and image analysis can improve the accuracy of WBC clas-
bocytopenia suggest that MPV increases with peripheral sification while reducing technical time. For example, the
destruction of platelets because of increased megakaryocyte CellaVision DM96 (CellaVision, Lund, Sweden) scans a
ploidy and production of larger platelets, whereas MPV stained blood smear, makes digital images of WBCs, classi-
decreases when platelet production is suppressed. Platelets fies them, and presents the sorted WBC images to an opera-
undergo time-dependent shape changes when exposed to tor to confirm or reclassify. When compared with manual
ethylenediaminetetraacetic acid (EDTA), however, leading differentials, automated morphologic differentials demon-
to inaccurate MPV results and thus diminishing its clinical strate excellent routine differential accuracy and sensitivity
utility. Inaccurate automated platelet counts can result from to detect blasts. In addition, stored images can be reviewed at
fragmented RBCs, congenital (inherited macrothrombocy- remote locations, such as outpatient clinics.
topenia disorders such as May-Hegglin anomaly) or acquired Supravital stains are used to detect RBC inclusions. C
rystal
(myeloproliferative disorders or idiopathic thrombocytope- violet detects denatured hemoglobin inclusions (Heinz bod-
nic purpura) macrothrombocytes, and EDTA-mediated ies) because of enzymopathies such as glucose-6-phosphate
platelet clumping because of immunoglobulin M (IgM) dehydrogenase (G6PD) deficiency; brilliant cresyl blue is
autoantibodies. Hematology analyzers provide sensitive used to precipitate and detect unstable hemoglobins (hemo-
warnings for abnormal platelet populations requiring man- globin H cells in α-thalassemias).
ual smear review to confirm or revise platelet counts. Analo-
gous to reticulocytes, young platelets (also called reticulated
The bone marrow aspirate and biopsy
platelets) contain detectable mRNA. Currently, only certain
analyzers (eg, the Sysmex XE) provide an immature platelet Following are the most frequent indications for bone mar-
fraction (IPF), with a reference range of 1.1%-6.1%, based row biopsy: unexplained cytopenias; unexplained leukoc
on the analysis of cell volume and fluorescence intensity of ytosis, erythrocytosis, or thrombocytosis; staging of
mRNA binding dye. Potential applications include differen- lymphoma and some solid tumors (particularly in patients
tiating thrombocytopenia due to megakaryopoiesis failure with cytopenias or other findings suggestive of bone mar-
from peripheral destruction and determining earlier evi- row involvement); diagnosis of plasma cell neoplasms
dence of marrow regeneration following stem cell transplan- (myeloma and monoclonal gammopathy of uncertain sig-
tation or response to a thrombopoietin mimetic drug. nificance); evaluation of systemic iron levels; and unex-
plained splenomegaly. Bone marrow aspirate and biopsy
commonly are performed by collecting specimens from the
Examination of peripheral blood smears
posterior iliac crests. Bone marrow aspirates also can be
Blood smears are typically stained with either the Wright or obtained from the sternum. In newborns and young
the May-Grünwald-Giemsa stains and can be prepared by infants, marrow aspirates often are obtained from the ante-
automated slide maker or strainers, which can be interfaced rior tibia. Quality smears require adequate spicule harvest-
with hematology analyzers. Microscopic examination of ing because perispicular areas are the most representative
stained blood smears can identify morphologic abnormali- areas to examine.
ties that automated hematology analyzers nonspecifically The bone marrow aspirate and touch prep are usually
flag or, rarely, miss. The examination begins at low power stained with either Wright or May-Grünwald-Giemsa stain.
(×10), scanning for platelet clumps or abnormal, large, The aspirate is used for cytologic examination of the bone
nucleated cells that may be located along the lateral and lead- marrow cells and for performing the differential. Bone mar-
ing edges of the smear. At higher magnification (×50 and row core biopsies are most commonly fixed in formalin, and

Table 11-1 Red blood cell abnormalities*
Abnormality Description Cause Disease association

Acanthocytes Irregularly spiculated red cell Altered membrane lipids Liver disease, abetalipoproteinemia,
(spur cells) postsplenectomy
Basophilic stippling Punctate basophilic inclusions Precipitated ribosomes Lead toxicity, thalassemias
Bite cells (degmacyte) Smooth semicircle taken from Heinz body pitting by spleen G6PD deficiency, drug-induced
one edge oxidant hemolysis
Burr cells (echinocytes) Short, evenly spaced spicules May be related to abnormal Usually artifactual; also uremia,
membrane lipids bleeding ulcers, gastric carcinoma
Cabot ring Circular, blue, threadlike Nuclear remnant Postsplenectomy, hemolytic anemia,
inclusion with dots megaloblastic anemia
Howell-Jolly bodies Small, discrete basophilic dense Nuclear remnant Postsplenectomy, hemolytic anemia,
inclusions; usually single megaloblastic anemia
Pappenheimer bodies Small dense basophilic granules Iron-containing siderosomes Sideroblastic anemia,
or mitochondrial remnant postsplenectomy
Schistocytes Distorted, fragmented cell, with Mechanical distortion in Microangiopathic hemolytic anemia,
(helmet cells) 2-3 pointed ends, loss of central the microvasculature by prosthetic heart valves, severe burns
pallor fibrin strands; damage by
mechanical heart valves
Spherocytes Spherical cell with dense Decreased membrane Hereditary spherocytosis,
appearance and absent central redundancy immunohemolytic anemia
pallor; usually decreased
diameter
Stomatocytes Mouth- or cuplike deformity Membrane defect with Hereditary stomatocytosis,
abnormal cation immunohemolytic anemia
permeability
Target cell (codocyte) Targetlike appearance, often Increased redundancy of cell Liver disease, postsplenectomy,
hypochromic membrane thalassemia, HbC
Teardrop cell Distorted, drop-shaped cell Myelofibrosis, myelophthisic anemia
(dacrocyte)
Modified from Kjedsberg C, ed. Practical Diagnosis of Hematologic Disorders. 5th ed. Chicago, IL: ASCP Press; 2010.
G6PD = glucose-6-phosphate dehydrogenase; HbC = hemoglobin C.
*Blood smear abnormalities can be artifacts of poor slide preparation or viewing the wrong part of the smear.
the biopsy specimen is decalcified and embedded in paraffin; Immunohistochemical stains

3- to 4-mm sections are then cut and stained with hematox-
A large array of specific antibodies detected by enzymatic
ylin-and-eosin or Giemsa stain.
formation of a colored product linked to the antigen– anti-
When examining pediatric marrows, it is understood that
body complex are now available for use on blood smears,
erythroid hyperplasia is present at birth because of high lev-
marrow aspirates, and bone marrow biopsies or other tis-
els of erythropoietin. Lymphocytes may compose 40% of the
sues. Many cytochemical stains, such as tartrate-resistant
marrow cellularity in children <4 years of age, and eosino-
acid phosphatase (TRAP) and myeloperoxidase, have been
phils are present in higher numbers than in adults.
converted into immunohistochemical (IHC) reactions.
Perls or Prussian blue reactions are used to identify hemo-
Immunohistochemistry (IHC) is used on marrow aspi-
siderin in nucleated red blood cells (sideroblastic iron) and
rates and blood smears as an alternative or adjunct to flow
histiocytes (reticuloendothelial iron). See Table 11-2 for
cytometry. The advantage of IHC is the ability to correlate
other cytochemical stains.
morphology with phenotype. IHC can be used to phenotype
Ring sideroblasts are abnormal nucleated red cells with
undifferentiated tumors, lymphoproliferative disorders, and
blue-staining iron granules surrounding at least two-thirds
atypical lymphoid infiltrates. In patients whose marrow can-
of the nucleus. These iron granules are present in mitochon-
not be aspirated (dry tap), IHC can be performed on the
dria surrounding the nuclear membrane. Iron staining of the
biopsy section. IHC also can be used on sections of lymph
biopsy can underestimate the marrow iron stores because of
nodes or other tissues when there is concern about lym-
the loss of iron during decalcification.
phoma or some other neoplastic disease.

Table 11-2 Cytochemical stains
Cytochemical stain Substrate and staining cells

Myeloperoxidase Primary granules of neutrophils and secondary granules of eosinophils. Monocytic lysosomal
granules stain faintly.
Sudan black B Stains intracellular phospholipids and other lipids. Pattern of staining is similar to
myeloperoxidase.
Naphthol AS-D chloroacetate esterase Granulocytes stain; monocytes do not stain. Can be used in biopsies to stain granulocytes
(specific esterase) and mast cells.
α-Naphthyl butyrate (nonspecific esterase) Stains monocytes, macrophages, and histiocytes. Does not stain neutrophils.
α-Naphthyl acetate (nonspecific esterase) Megakaryocytes stain with α-naphthyl acetate but not α-naphthyl butyrate.
Terminal deoxynucleotidyl transferase (TDT) Intranuclear enzyme. Stains thymocytes and lymphoblasts. Some myeloblasts stain positively.
Tartrate-resistant acid phosphatase (TRAP) Stains an acid phosphatase isoenzyme. Positive staining in hairy cell leukemia, Gaucher cells,
activated T-lymphocytes.
Periodic acid-Schiff (PAS) Detects intracellular glycogen and neutral mucosubstances. Positive staining in acute
lymphoblastic leukemia, acute myeloid leukemia, erythroleukemia, and Gaucher cells.
Toluidine blue Detects acid mucopolysaccharides. Positive in mast cells and basophils.
Preparation of bone marrow samples for Table 11-3 Specimen allocation for ancillary studies
ancillary studies
Clinical problem Ancillary techniques
Bone marrow collected in ethylenediaminetetraacetic acid Pancytopenia Flow cytometry (LGL, hairy cell leukemia,
(EDTA) is adequate for both flow cytometry and molecular PNH clone, AML)
analysis. Bone marrow collected for cytogenetic studies Cytogenetics (AML, MDS)
should be collected in a sterile tube. Molecular genetics
Paraffin-embedded tissue can be used for polymerase
Myeloid leukemia Flow cytometry (phenotyping)
chain reaction (PCR) of genomic DNA sequences. Reverse
Cytogenetics and FISH
transcriptase PCR (RT-PCR) assays require that RNA prepa-
Molecular genetics (BCR-ABL,
rations be performed as early as possible to prevent digestion
PML/RARA, AML1/ETO)
by ubiquitous nucleases.
Lymphoproliferative Flow cytometry (phenotyping, prognostic
disorder markers such as ZAP-70)
Flow cytometry Cytogenetics: t(1;19) in pre–B-cell
ALL, t(14;18) in follicular
The most common applications of flow cytometry in hema- lymphomas, etc.
tology include the detection of cell surface or cytoplasmic FISH
proteins using fluorescent-labeled monoclonal antibodies Molecular genetics (clonality,
and the assessment of DNA content using DNA-binding dyes. specific markers such as BCL2,
Flow cytometry is used for phenotyping populations of BCL6, etc.)
cells, enumerating early progenitors for stem cell transplan- Immunohistochemistry (phenotyping,
tations, detecting minimal residual disease, detecting targets prognostic markers)
for immunotherapy, and assessing the presence of prognos- Myeloproliferative disorders Cytogenetics
tic markers. See Table 11-3 for a summary of clinical uses of FISH (BCR-ABL)
flow cytometry in ancillary studies. Molecular genetics (BCR-ABL, JAK2)
Gating is necessary to identify cells of interest in a mixed Plasmaproliferative Flow cytometry (phenotyping, labeling
population of cells. Three major leukocyte populations disorders index)
(lymphocytes, monocytes, and neutrophils) can be defined Cytogenetics
using light scatter. Forward-angle scatter (FS; low angle)
ALL = acute lymphoblastic leukemia; AML = acute myelogenous
measures cell size, and side-light scatter (SS) measures inter-
leukemia; CLL = chronic lymphocytic leukemia; FISH = fluorescence
nal cellular granularity. Lymphocytes have the lowest FS and
in situ hybridization; LGL = large granular lymphocyte leukemia;
SS signals, monocytes have intermediate FS and SS signals, MDS = myelodysplastic syndrome; PNH = paroxysmal nocturnal
and neutrophils have high SS and slightly lower FS signals. hemoglobinuria.
Blasts generally have low FS and SS.

The most common method for gating different cell popu- B- and T-cell precursors express terminal deoxynucleotidyl
lations is by plotting right-angle SS against CD45. Cells can transferase (TdT), human leukocyte antigen (HLA)-DR, and
be separated based on the intensity of staining they display CD34. B-cell differentiation is indicated by the expression of
with the conjugated antibody that is classified as either bright CD19 followed by CD10. As B-cells mature, they lose cell sur-
or dim. Lymphocytes are bright CD45 and have a low SS sig- face expression of CD34 and CD10 and express IgM on the
nal, neutrophils are dim to moderately bright CD45 and cell surface. Expression of surface IgM is associated with the
have a high SS signal, and monocytes are bright CD45 and expression of mature B-lymphocyte markers (CD20, CD21,
have an intermediate SS. Blasts have low SS and dim to nega- CD22, and CD79b). Mature B-cells express an immunoglob-
tive CD45 expression, the latter being more common in ulin heavy chain, such as IgM, and either the κ- or a λ-light
blasts of lymphoid lineage. chain. A predominant expression of one type of light chain on
Flow cytometry also can be used to detect populations of the cell surface of a population of B-cells is known as light-
natural killer (NK) cells. NK cells express CD2, CD7, CD16, chain restriction and is indicative of a monoclonal process.
and CD56 and show variable expression of CD57 and CD8. T-cell precursors express TDT, HLA-DR, and CD34. Dif-
NK cells do not express CD3, and the absence of CD3 expres- ferentiation is indicated by the expression of cytoplasmic
sion can be used to differentiate NK cells from T-cells. CD3 and CD7, followed by the expression of CD2 and CD5.
In addition to determining cell lineage, flow cytometry The common thymocyte also expresses CD1, CD4, and CD8.
can be used to detect prognostic markers. For example, flow The mature helper or inducer lymphocyte expresses CD2,
cytometric analysis of the tyrosine kinase ZAP-70 can be CD3, CD4, and CD5 and may express CD7. The mature sup-
used to subdivide chronic lymphocytic leukemia (CLL) into pressor or cytotoxic T-lymphocyte expresses CD2, CD3,
prognostic groups. Positivity for ZAP-70 is highly correlated CD4, CD5, and CD8 and may express CD7. T-cell neoplasms
with unmutated immunoglobulin heavy chain variable may be associated with abnormal expression patterns of
region (IgVh), a feature of CLL arising from pre–germinal T-cell antigens, and the abnormal pattern may be detected by
center cells, and patients with pre–germinal center CLL have flow cytometric analysis. See Tables 11-4 through 11-10 for
a decreased overall survival when compared with patients useful CD markers.
with CLL arising from post–germinal center cells. Positivity Flow cytometry can be used to diagnose paroxysmal noc-
for CD38 by flow cytometric analysis also is correlated with turnal hemoglobinuria (PNH). PNH is associated with the
unmutated IgVh, but the correlation is not as strong as it is absence of glycosylphosphatidylinositol (GPI)-anchored
with ZAP-70. In addition, expression of CD49d, an integrin membrane proteins, including two complement regulatory
alpha subunit, by CLL cells is associated with a more aggres- molecules: decay accelerating factor (DAF, CD55) and pro-
sive disease course. tectin (MIRL, CD59). The absence of these proteins from the
Uncommitted hematopoietic progenitors are CD34+ and cell surface of erythrocytes can be detected by flow cytometry
CD38−; expression of CD38 is evidence of lineage commit- using antibodies to CD55 and CD59, respectively. Alterna-
ment. Myeloid maturation is characterized by the following tively, PNH granulocytes are detected by the absence of GPI
maturational sequence: colony forming units–erythroid gran- anchor binding by FLAER, an Alexa 488 labeled variant of
ulocyte, macrophage, and megakaryocyte (CFU-GEMM, aerolysin. Flow cytometry technology can discriminate
CD34+, MHC class II+, CD33 −/+); and followed by colony between fetal and adult red cells or Rh+ and Rh− red cells
forming unit–granulocyte, macrophage (CFU-GM, CD34+, during pregnancy and postpartum and can identify red cell
major histocompatibility complex [MHC] class II+, CD33+, skeletal disorders, such as hereditary spherocytosis.
CD13 −/+, CD15 −/+). Neutrophil precursors then progres-
sively lose MHC class II and CD33 and gain CD11b, CD16,
Cytogenetics
and CD32. Monocytes retain expression of MHC class II and
CD33 and also gain expression of CD14 and CD64. Cytogenetic analysis can be performed from cultured (indi-
Appearance of CD71, loss of CD34 and CD33, and rect) and uncultured (direct) preparations. In the indirect
decreased expression of CD45 characterize erythroid matura- assay, cells are grown so that mitotic forms can be visualized
tion. With further differentiation, CD71 expression decreases, and distinct chromosomal banding patterns can be assessed
glycophorin expression increases, and CD45 disappears. (conventional cytogenetics). Growing or culturing the cells
Megakaryocytic differentiation is indicated by the increases the mitotic rate and improves chromosome mor-
expression of glycoprotein (GP) IIb (CD41). GPIIb-IIIa
phology. Mitogens may be useful in improving the yield of
(CD61) expression increases as CD34 expression decreases. karyotyping abnormal cells and are particularly useful when
GPIb (CD42b) is expressed at the promegakaryocyte stage. analyzing mature B- or T-cell processes.
GPV (CD42d) expression increases with megakaryocyte A cytogenetic clone is defined by a minimum of two
differentiation. mitotic cells with gain of the same chromosome or with the

Table 11-4 Clinically useful CD markers
Marker Lineage association

Progenitor cells
CD34 Progenitor cells, endothelium
CD38 Myeloid progenitors, T, B, NK cells, plasma cells, monocytes, CLL subset
B-cell markers
CD10 Pre-B-lymphocytes, germinal center cells, neutrophils
CD19 B cells (not plasma cells or follicular dendritic cells)
CD20 B cells (not plasma cells)
CD21 Mature B cells, follicular dendritic cells, subset of thymocytes
CD22 Mature B cells, mantle zone cells (not germinal center cells)
CD23 B cells, CLL
CD79b B cells (not typical CLL)
CD 103 Intraepithelial lymphocytes, hairy cell leukemia, T-cells in enteropathic T-cell lymphoma
FMC7 B cells (not typical CLL), hairy cell leukemia
T-cell markers
CD2 Pro- and pre-T-cells, T-cells, thymocytes, NK cells, some lymphocytes in CLL and B-ALL
CD3 Thymocytes, mature T-cells, cytoplasm of immature T-cells
CD5 Thymocytes, T-cells, B cells in CLL, B-cells in mantle cell lymphoma
CD4 Helper T-cells, monocytes, dendritic cells, activated eosinophils, thymocytes
CD7 Pro- and pre-T-cells, T-cells, thymocytes, NK cells, some myeloblasts
CD8 Suppressor T-cells, NK cells, thymocytes
CD25 Activated T- and B-cells, adult T-cell leukemia/lymphoma
NK/cytotoxic T-cell markers
CD16 NK cells, monocytes, macrophages, neutrophils
CD56 NK cells, myeloma cells
CD57 NK cells, T-cell subset
Myeloid and monocytic markers
CD13 Monocytes, neutrophils, eosinophils, and basophils
CD14 Monocytes, macrophages, subset of granulocytes
CD33 Myeloid lineage cells and monocytes
CD117 Immature myeloid cells, AML
Monocytes
CD11c Monocytes, macrophages, granulocytes, activated B- and T-cells, NK cells, hairy cell leukemia
CD15 Myeloid lineage cells and monocytes
CD64 Monocytes, immature myeloid cells, activated neutrophils
Megakaryocytic markers
CD41 Platelets and megakaryocytes (GPIIb)
CD42 Platelets and megakaryocytes (CD42a: GPI; CD42b: GPIb)
CD61 Platelets, megakaryocytes, endothelial cells (GPIIb-IIIa)
Erythroid markers
CD71 Transferrin receptor is upregulated upon cell activation
CD235a Glycophorin A
AML = acute myelogenous leukemia; B-ALL = B-lineage acute lymphoblastic leukemia; CLL = chronic lymphocytic leukemia; GP =
glycoprotein; NK = natural killer.
same structural abnormality or three mitotic cells with loss are determined on peripheral blood T-lymphocytes grown in
of the same chromosome. culture with phytohemagglutinin (PHA), a T-cell mitogen.
Constitutional chromosome abnormalities, associated Fluorescence in-situ hybridization (FISH) is a cytoge-
with either congenital genetic syndromes or normal variants, netic technique that uses specific fluorescently-labeled DNA

Table 11-5 Acute myeloid leukemia phenotyping
HLA-DR CD34 CD33 CD13 CD11c CD14 CD41 CD235a

M0 + + + +/− +/− − − −
M1 + + + + +/− +/− − −
M2 +/− +/− + + +/− +/− − −
M3 − − + + +/− − − −
M4 + +/− + + + + − −
M5 + − + + + + − −
M6 +/− − − − +/− − − +
M7 +/− +/− +/− − − − + +
Table 11-6 B-lineage acute lymphoblastic leukemia phenotyping
TDT CD19 CD10 CD20 Cyto-m Surface Ig

Pro-B + + − − − −
Pre-pre-B (common ALL) + + + − − −
Pre-B + + + 1/− + −
Early B (Burkitt) − + + + − +
Cyto-m = cytoplasmic m; Ig = immunoglobulin; TDT = terminal deoxynucleotidyl transferase.
Table 11-7 T-lineage acute lymphoblastic leukemia phenotyping
Surface TDT CD7 CD2 CD5 CD1 CD3 CyCD3 CD4/CD8
Prothymocyte + + + − − − + −/−
Immature thymocyte + + + + − − + −/−
Common thymocyte + + + + + + /− + +/ +
Mature thymocyte − + + + − + + CD4 or CD8+
Mature T-cell − + + + − + + CD4 or CD8+
Cy CD3 = cytoplasmic CD3; TDT = terminal deoxynucleotidyl transferase.
Table 11-8 Common B-cell neoplasms
CD20 CD5 CD10 CD23 CD43 CIg SIg Cyclin D1 Other
CLL/SLL + ++ − ++ ++ 5%+ + − FMC7−, CD79b+

LPL ++ − − − +/− + + −
PLL ++ +/− − ++ −
HCL ++ − − − − − + +/− CD11c+, CD25+, CD103+
MCL ++ ++ − − ++ − ++ ++ FMC7+
MZL ++ − − − +/− +/− ++ −
FCL ++ − 60% + − − − ++ − BCL2+
LCL ++ 10%+ 25%-50%+ + +/− +/− +/] − BCL2+ in 30%-40%
Burkitt ++ − + − − + + −
Myeloma − − Occ + − + ++ − 15%-20%+ CD56+, CD38+, CD138+
CIg = cytoplasmic immunoglobulin; CLL = chronic lymphocytic leukemia; FCL = follicular center cell lymphoma; HCL = hairy cell leukemia;
LCL = large-cell lymphoma; LPL = lymphoplasmacytic lymphoma; MCL = mantle cell lymphoma; MZL = marginal zone lymphoma; PLL =
B-cell prolymphocytic leukemia; SIg = surface immunoglobulin; SLL = small lymphocytic lymphoma.
probes to identify each chromosomal segment. FISH can be marrow or peripheral blood smears or fixed and sectioned
performed using either cultured or uncultured prepara- tissues.
tions. In the uncultured technique, the assay is performed Hybridization of centromere-specific probes is used to detect
using nuclear DNA from interphase cells that are affixed monosomy, trisomy, and other aneuploidies. Chromosome-
to a microscope slide. FISH can be performed using bone specific libraries, which paint the chromosomes, are useful in

Table 11-9 Common mature T-cell and NK-cell neoplasms
CD3S CD3C CD5 CD7 CD4 CD8 CD30 CD16 CD56 EBV
T-PLL + + − + 4.8 4.8 − − + −

T-LGL + + − + − + − + − −
NK-leukemia − − − +/− − +/− − − + +
EN-NK/T − + − +/− − − − + + +
HS-gd lym + + - + − − − + +/− −
Ent-T lym + + + + − +/− +/− − − −
SC pannic T lym + + + + − + +/− − − −
PTCL-NOS + − +/− +/− +/− +/− +/− − + +/−
AILT + + + + +/− − − − + +/−
ALCL + − +/− +/− +/− +/− + − − −
AILT = angioimmunoblastic T-cell lymphoma; ALCL = anaplastic large cell lymphoma; CD3C = cytoplasmic CD3; CD3S = surface CD3; EBV =
Epstein-Barr virus; Ent-T lym = enteropathy-associated T-cell lymphoma; EN-NK/T = extranodal natural killer/T-cell lymphoma; HS-gd
lym = hepatosplenic T-cell lymphoma; NK-leukemia = natural killer cell leukemia; PTCL-NOS = peripheral T-cell lymphoma, not otherwise
specified; SC pannic T lym = subcutaneous panniculitis-like T-cell lymphoma; T-LGL = T-cell large granular lymphocyte leukemia; T-PLL =
T-prolymphocytic leukemia.
Table 11-10 Immunohistochemical diagnosis of Hodgkin disease for the target sequence, are used. The primers are added to
CD45 CD30 CD15 CD20 CD3 Pax5 denatured single-stranded DNA. A heat-stable DNA poly-
merase and the four deoxynucleotide triphosphates are used
Hodgkin − + + LPHD(+) − weak +
to initiate the synthesis of new DNA strands. The newly syn-
(R-S cells)
thesized DNA strands are used as templates for further cycles
B-lymphoma + +/− − + − +
of amplification. The amplified DNA sequence can be
T-lymphoma + +/− +/− − + −
detected by electrophoresis on an agarose gel, and visualiza-
LPHD = nodular lymphocyte predominant Hodgkin lymphoma; tion can be accomplished by the use of a DNA dye; alterna-
R-S = Reed-Sternberg.
tively, the amplified DNA can be sequenced directly in an
automatic sequencer.
identifying marker chromosomes or structural rearrangements, Uses of PCR in clinical laboratories include detection of
such as translocations. Translocations and deletions also can be the break cluster region-Abelson tyrosine kinase (BCR-ABL)
identified in interphase or metaphase by using genomic probes translocation in chronic myeloid leukemia, detection of pro-
that are derived from the breakpoints of recurring transloca- myelocytic leukemia-retinoic acid receptor alpha (PML-
tions or within the deleted segment. Multiplex FISH (spectral RARA) in acute promyelocytic leukemia, and detection of
karyotyping) consists of simultaneously painting all chromo- the Janus kinase-2 (JAK2) mutation in polycythemia vera,
somes in a cell using different colors for each chromosome. essential thrombocythemia, and primary myelofibrosis. PCR
Cytogenetics is particularly useful in the subclassification is of increasing importance in the diagnosis of acute myeloid
of acute myeloid leukemias and in confirming the diagnosis leukemia, particularly for the detection of internal tandem
and prognosis of B-cell neoplasias. CLL, acute leukemias, duplications in the FMS-like tyrosine kinase 3 (FLT-3) locus,
B-cell lymphomas, and multiple myeloma all have cytoge- mutations in CCAAT/enhancer binding protein α CCAAT
netic abnormalities that can be detected using either conven- enhancer binding protein alpha (CEBPA), and point muta-
tional cytogenetics or FISH. tions in the nucleophosmin 1 (NPM1) and Wilms tumor 1
(WT1) genes.
DNA sequencing is important in the identification of
Molecular diagnostics point mutations. The earlier Sanger (chain termination)
technique rapidly is being eclipsed by the next-generation
Polymerase chain reaction
sequencing technology, which has a high throughput capac-
PCR is designed to permit selective amplification of a spe- ity and thus makes the parallel analysis of many genes pos-
cific target DNA sequence within total genomic DNA or a sible. The potential clinical uses, including diagnosis,
complex complementary DNA (cDNA) population. Partial predictors of response to therapy, and risk stratification are
DNA sequence information from the target sequences is being explored for a variety of malignancies, including
required. Two oligonucleotide primers, which are specific myeloma, leukemias, and lymphoma.

Miscellaneous laboratory hematology methods detergent to lyse red blood cells, and high-concentration salt
buffer. Deoxygenated Hb S forms tactoids that defract and
Erythrocyte sedimentation rate
reflect light; whereas nonsickling hemoglobins remain solu-
The erythrocyte sedimentation rate (ESR) measures a physi- ble, allowing the detection of lines or letters when viewed
cal phenomenon—the opposing forces of gravity and buoy- through the hemolysis solution. A positive solubility test
ancy on RBCs when blood is suspended in an upright cannot discriminate between Hb S trait, homozygous Hb S,
tube—and is expressed in millimeters per hour. Elevated Hb S/β-thalassemia, or other combinations that include
plasma proteins, primarily fibrinogen and immunoglobins, hemoglobin S. FP results can occur due to paraprotein or
neutralize red blood cell membrane negative charge, facili- cryoglobulin precipitation, and FN results can occur in ane-
tating rouleaux formation and more rapid sedimentation mic (hemoglobin <7.0 g/dL) sickle trait individuals or when
because of increased mass per surface area. The clinical util- Hb S concentration is <2.6 g/dL. Because the concentration
ity of ESR generally is poor except for selected rheumato- of Hb S in affected newborns is low, sickle solubility testing
logic disorders, and it is not an appropriate screening test in should not be performed on infants <6 months old because
asymptomatic patients. Conditions associated with elevated of the risk of FN results. If used as a screening test, a positive
ESR include malignancies; infections; and inflammatory solubility test requires evaluation by an alternative method
conditions, particularly polymyalgia rheumatic and tempo- to confirm and quantify Hb S and to identify coexisting
ral arteritis; as well as hematologic conditions such as nonsickling hemoglobinopathies or thalassemias. Other
cold agglutinin disease, cryoglobulinemia, and plasma cell rare hemoglobinopathies produce a positive solubility test,
dyscrasia–related M proteins. ESR reference ranges increase including hemoglobin C Harlem, and, if coinherited with
with age and are higher for women. Additional variables Hb S, they will produce a sickle cell disease phenotype.
affect ESR; anemia and macrocytosis can cause faster sedi-
mentation, whereas sickle cells, by impeding rouleaux for-
Hemoglobin electrophoresis
mation, and microcytosis cause slower sedimentation. The
modified Westergren method (EDTA blood diluted 4:1 in Methods to separate normal (hemoglobins A, A2, and F) and
sodium citrate and put in a 200 mm vertical glass tube) is the abnormal hemoglobins, primarily based on differences in
preferred manual method. Automated ESR devices monitor charge, include alkaline and acid gel electrophoresis, isoelec-
sedimentation for shorter periods, extrapolate to millimeters tric focusing, high-performance liquid chromatography
per hour, and correlate reasonably well with the Westergren (HPLC), and capillary electrophoresis (Figure 11-2). No
method. method can definitively identify and quantify all hemoglobin
variants, and any abnormal hemoglobin identified by the
method chosen for screening must be confirmed by an alter-
Solubility testing for hemoglobin S
native method (including solubility test for presumed Hb S).
Manual qualitative methods to detect hemoglobin S (Hb S) HPLC and capillary electrophoresis instruments are fully
rely on visual detection of turbidity when blood containing automated, provide precise measurements of normal and
Hb S is added to a solution containing a reducing agent, variant hemoglobins, and are well suited for laboratories
Alkaline gel Acid gel A SC disease B SS disease

C S FA C SA F
30% 30%
F
A
20% 20%
10% 10%
F
A2 A2
0 0
B
0 1 2 3 4 5 6 0 1 2 3 4 5 6
Figure 11-2 Examples of alkaline and acid gel electrophoresis and high-performance liquid chromatography patterns for a patient with
hemoglobin SC disease (a) and a patient with homozygous SS sickle cell disease (b).

performing many analyses to diagnosis hemoglobins S, C, members are unavailable, laboratory studies are indicated to
and E and other uncommon hemoglobinopathies and confirm loss of red cell membrane, anchoring proteins, and
β-thalassemia trait (elevated Hb A2, microcytic anemia). For spectrin. Although spherocytes are more susceptible to lysis
optimal genetic counseling, DNA analysis may be appropri- when suspended in hypotonic saline solutions because of a
ate to completely characterize α-thalassemias and some decreased surface area or volume, increased osmotic fragility
uncommon thalassemias and hemoglobinopathies. (OF) is an insensitive screening test for mild and compen-
sated HS, and OF can produce FP results. A more sensitive
and specific method is detection of decreased eosin-5-
G6PD testing
maleimide (EMA) binding by flow cytometry due to loss of
Evaluation for inherited RBC enzymopathies is appropriate red cell membrane proteins. Hereditary elliptocytosis causes
in patients with nonspherocytic, nonimmune-mediated minimal, if any, anemia and is a morphologic diagnosis
hemolytic anemia. X-linked inheritance of G6PD deficiency (normal OF and EMA binding). Pyropoikilocytosis caused
is the most common RBC enzyme defect and is associated by inheritance of both qualitative and quantitative red cell
with hemolysis during oxidative stresses because of acute ill- skeletal defects, which produce severe hemolytic anemia,
ness, medications, or, rarely, ingestion of fava beans. Decreased deranged red cell morphologies, and decreased EMA.
G6PD activity diminishes nicotinamide adenine dinucleotide
phosphate (NADPH) production and prevents reduction of
methemoglobin by reduced glutathione, leading to denatured Hemostasis and thrombosis testing
hemoglobin (Heinz bodies) and shortened RBC survival.
Sensitive qualitative screening tests for G6PD deficiency Hemostasis involves multiple molecular and cellular interac-
include dye decolorization and fluorescent spot tests, which tions to initiate and regulate platelet aggregation (primary
monitor NADPH-dependent chemical reactions. FN results hemostasis) and coagulation (secondary hemostasis) at the
may occur if testing is performed during or shortly after a site of vascular injury to produce a durable “patch” without
hemolytic event in individuals (typically African and African occluding blood flow. Laboratory evaluation of hemostasis is
American males) with the A-mutation, however, because performed in several clinical settings, including screening of
enzyme activity is near normal in reticulocytes. Pyruvate asymptomatic patients before selective invasive procedures
kinase deficiency is the second most common RBC enzyme and patients with underlying disorders associated with
defect, presenting with chronic hemolysis of variable severity bleeding complications, evaluation of patients with personal
and an autosomal recessive inheritance pattern. In patients or family histories of abnormal bleeding or bruising, assess-
with hemolysis, a suspicion for an RBC enzymopathy, and ment of inherited and acquired thrombosis risk factors, and
normal G6PD screening, blood should be sent to a reference antithrombotic drug monitoring.
laboratory that offers a panel of additional RBC enzyme tests. Hemostasis screening typically consists of a prothrombin
time (PT), activated partial thromboplastin time (aPTT),
and platelet count. Abnormal screening test results require
Hereditary red cell skeletal disorders
additional clinical and laboratory investigation to determine
The unique flexibility of a red cell depends on its lipid bilayer the etiologies. Mucosal bleeding, menorrhagia, petechiae,
attachment to an underlying scaffold of α- and β-spectrin and ecchymoses suggest primary hemostatic disorders such
dimers via transmembrane proteins and other linking mol- as von Willebrand disease (vWD) and qualitative platelet
ecules. Inherited quantitative and qualitative red cell cyto- disorders, whereas hematomas, hemarthroses, and delayed
skeleton defects are an infrequent cause of nonimmune bleeding suggest a coagulation factor defect.
chronic hemolysis, but these defects are relatively more com- Testing for thrombophilia is usually performed when a
mon in people of northern European ancestry. The most patient has a venous thromboembolic event (VTE) in the
common phenotype is hereditary spherocytosis (HS), with an absence of compelling acquired risk factors, such as major sur-
estimated incidence of 1 in 2,000 whites (see Chapter 7 for gery or trauma, cancer and its treatment, and immobility. The
more structural details). The intensity of hemolysis can vary decision to test for a predisposition to VTE also depends on
from severe anemia to a completely compensated state. In the patient’s age, history of thrombosis, family history of
about 75% of HS cases, there is an autosomal-dominant thrombosis, and whether the results would influence duration
inheritance pattern, and diagnosis can be made on the basis of of anticoagulant therapy. Laboratory testing for inherited defi-
family history, a negative direct antiglobulin test (DAT), ane- ciencies of coagulation regulatory proteins should be done
mia with reticuloctyosis, mild splenomegaly, and spherocytes after a patient has completed treatment for a VTE and is in
on blood smear (typically <10% of red cells). In suspected stable health. The levels of protein C (PC), protein S (PS), and
cases of HS that appear to be sporadic, or if data on family antithrombin (AT) can decrease during the acute phase of a

Hemostasis and thrombosis testing | 289
VTE, and also be reduced in association with anticoagulation; such as bleeding, breakthrough thrombosis, suspected non-
PC and PS levels are reduced by warfarin and AT levels are compliance, populations at risk for drug accumulation,
lower during unfractionated heparin therapy. In addition, the and prior to urgent surgery or administration of thrombo-
direct oral anticoagulants (DOACs) dabigatran, rivaroxaban, lytic therapy.
apixaban, and edoxaban, can also interfere with thrombo- It is important to be aware of the differences in hemostasis
philia testing. Lupus anticoagulant (LAC) testing should ide- factor reference ranges among infants, older children, and
ally be performed before anticoa gulation is initiated, in adults to avoid mislabeling newborns with inherited defects
conjunction with serologic assays (anti-cardiolipin and anti- (see Chapter 2 for details). The following sections provide
β-2-glycoprotein I IgM and IgG antibodies), and abnormal specific information regarding hemostasis laboratory meth-
results should be repeated at least 12 weeks later to determine ods as they apply to the aforementioned clinical situations.
whether they are persistent to fulfill the laboratory criteria for
antiphospholipid syndrome (APS). Genetic thrombophilia
Screening coagulation testing and associated
testing (factor V Leiden and prothrombin 20210 gene muta-
abnormalities
tions) can be ordered at any time and is not affected by clinical
status or medications. Heparin-induced thrombocytopenia Most in vivo coagulation reactions are believed to be initi-
(HIT) and thrombotic thrombocytopenic purpura (TTP) are ated by tissue factor activation of factor VII. Important inter-
unique acquired thrombocytopenia disorders with the poten- actions occur between the extrinsic and intrinsic pathways in
tial for thrombotic complications. Laboratory test results can physiologic in vivo hemostasis. Although the division into
provide subsequent support for these diagnoses, but immedi- two separate pathways, as shown in Figure 11-3, does not
ate therapeutic interventions largely should be based on clini- reflect complex interactions among coagulation factors, it
cal assessment. does provide a useful way to interpret screening coagulation
Two major forms of anticoagulation therapy, warfarin test results when evaluating for potential abnormalities of
antagonism of vitamin K–dependent γ-carboxylation of hemostasis or thrombosis.
coagulation factors II, VII, IX, and X, proteins C and S; and
unfractionated heparin, require therapeutic drug monitor-
Preanalytical variables
ing because of unpredictable anticoagulant activities. Efforts
to harmonize interlaboratory monitoring of warfarin with The aPTT and, to a lesser degree, the PT are sensitive to changes
the PT ratio and heparin with the aPTT have led to the in the ratio of sodium citrate in the collection tube. Filling a
international normalized ratio (INR) and heparin activity tube with less than the recommended volume to obtain a
(chromogenic anti-Xa) assays, respectively. The DOACs, sodium citrate-to-blood ratio of 1:9 or collecting blood in the
dabigatran, rivaroxaban, apixaban, and edoxaban, do not proper proportions from a polycythemic patient increases the
require therapeutic drug monitoring; however, assays based concentration of citrate in the plasma compartment, leading
upon anti-IIa or anti-Xa are available in some laboratories to to incomplete recalcification when a fixed v olume of CaCl2 is
measure drug concentrations in special clinical situations added, and results in an artifactual prolongation of the aPTT
Contact factors:
prekallikrein, HMW kininogen
XII Tissue factor
XI
IX VII
VIII
Intrinsic Ca2+, PL Extrinsic

Figure 11-3 Simplified coagulation cascade
X
indicating the intrinsic pathway measured by the aPTT PT
activated partial thromboplastic time (aPTT), the V, Ca2+, PL
extrinsic pathway measured by the prothrombin
Common
time (PT), the common pathway (factor X, factor V, Prothrombin Thrombin
prothrombin, and fibrinogen) measured by PT and
aPTT, and the conversion of fibrinogen to fibrin
TT Fibrinogen Fibrin clot
measured by the thrombin time (TT). HMW = high
molecular weight.

or PT. Heparin contamination due to blood collection from 37°C for 1-2 hours followed by repeating the aPTT. There is no
central lines can also cause a prolonged aPTT. A prolonged consensus approach for interpretation of mixing study results,
thrombin time (TT) that c orrects when repeated after treat- and inflexible requirements such as correction to within the
ment of plasma with a h eparin-neutralizing agent confirms laboratory’s PT or aPTT reference ranges to rule out inhibitor
heparin contamination. Alternatively, a prolonged TT and a activity can be misleading. One must consider the clinical con-
normal reptilase time, which uses a snake venom not neutral- text (bleeding or thrombotic event) and the initial extent of PT
ized by heparin-accelerated AT, will confirm the presence of and/or aPTT prolongation when assessing the 1:1 mix results.
heparin. Most PT reagents contain heparin-neutralizing agents Sometimes mixing studies will not be definitive, especially
such as polybrene, making this screening clotting test insensi- when an aPTT is mildly prolonged and corrects with mixing,
tive to heparin contamination. Many coagulation tests per- in which case performing both selected factor activity assays
formed on plasma from patients taking oral direct factor-IIa and LAC screening will be necessary.
(dabigatran) and factor Xa (rivaroxaban, apixaban, edoxaban)
anticoagulants are at risk for either positive or negative biases,
Coagulation factor activity assays
which can be clinically important (Table 11-11). The purpose
of a mixing study is to determine whether a prolonged aPTT Determination of a coagulation factor activity in a patient’s
or, occasionally, a prolonged PT is more likely due to a defi- plasma typically is performed on automated instruments
ciency of one or more coagulation factors or to an inhibitor. and requires two reagents: PNP and plasma completely defi-
The first step is to exclude contamination with heparin or cient in the factor of interest. Combining equal volumes of
direct thrombin inhibitors by performing a TT, heparin neu- plasma from a large number of healthy adults averages the
tralization, or anti-Xa assays. Next, the aPTT or PT is repeated interindividual variability for coagulation factors, which
on a 1:1 mixture of patient plasma and pooled normal plasma typically ranges from 50% to 150%, to produce PNP with
(PNP), which should provide at least 50% activity for all coag- 100% activity for all factors. Mixing PNP and factor-defi-
ulation factors and substantial correction if a factor deficiency cient plasma in different ratios produces calibrators of
is the cause. Because factor VIII inhibitors and some LACs known factor activities. PTs are performed on the calibration
manifest their effect in prolonging the aPTT in a time- and samples for factors VII, X, V, and II, and aPTTs are performed
temperature-dependent manner, 1:1 mixtures are incubated at for the intrinsic pathway factors. When the factor activities
Table 11-11 Coagulation tests interference caused by direct oral anticoagulants
Rivaroxaban, apixaban,
Dabigatran, oral factor edoxaban, oral factor
Test IIa inhibitor Xa inhibitors Comments
APCr ratio, aPTT based +bias +bias Risk of missing FVL

Antithrombin, anti-Xa method +bias Risk of missing AT deficiency
Antithrombin, anti-IIa method +bias Risk of missing AT deficiency
Factors X, VII, V, II (PT based) −bias −bias Possible inhibitor pattern
Factors PK, HMWK, XII, XI, IX, −bias −bias Possible inhibitor pattern
VIII (aPTT based)
dRVVT screen prolonged prolonged
dRVVT screen/confirm prolonged prolonged Risk of false-positive LAC
Protein C clotting assay +bias +bias Risk of missing PC deficiency
Protein S clotting assay +bias +bias Risk of missing PS deficiency
PT and aPTT prolonged prolonged
PT 1:1 mix prolonged prolonged Inhibitor pattern
aPTT 1:1 mix prolonged prolonged Inhibitor pattern
Thrombin time prolonged unaffected
Fibrinogen based on clot formation −bias with some methods unaffected
Chromogenic anti-Xa unaffected positive bias Not a quantitative test for
Monitoring of heparin/LMWH rivaroxaban, apixaban or edoxaban
unless calibrated with the drug
APCr = activated protein C resistance; aPTT = activated partial thromboplastin time; AT = antithrombin; HMWK = high-molecular-weight
kininogen; dRVVT = dilute Russell’s viper venom; LAC = lupus anticoagulant; LMWH = low-molecular-weight heparin; PC = protein C;
PK = prekallikrein; PS = protein S.

of the calibrators are plotted against the corresponding PT or people), whereas the estimated prevalence of homozygous
aPTT results on logarithmic axes, a line or curve is generated. factor VII deficiency is 1 in 300,000 people. Some factor VII
Then, a PT or aPTT is performed on patient plasma mixed mutations produce greater PT prolongations with rabbit or
with factor-deficient plasma, and the corresponding activity bovine tissue factor than with human tissue factor. Therefore,
is determined from the calibration curve. it is important to confirm a suspected congenital factor VII
Additionally, factor levels are determined at a minimum deficiency by measuring factor VII activity with recombinant
of three serial dilutions of a patient’s plasma, and the results, human thromboplastin. Dysfibrinogenemia occasionally
corrected for the dilution factor, are compared. If an inhibi- causes a prolongation of the PT without a prolongation of the
tor is present, the factor activity appears to increase with aPTT, and factor VII inhibitory autoantibodies are extremely
dilution and results are nonparallel to the calibration curve. rare (Figure 11-4).
To determine whether the inhibitor interference is specific Warfarin causes prolonged PTs and variably, prolonged
for one factor, such as factor VIII, or nonspecific like a LAC aPTTs depending on the degree of factor IX, X, and II defi-
may require performance of additional factor activities. ciencies. Therapeutic monitoring of warfarin depends on the
The end point for most coagulation tests is detection of a PT. Thromboplastins, however, have different sensitivities
fibrin clot. Factor VIII and IX chromogenic activity assays to the effects of warfarin. To account for this variability, and
exist but are not widely used. The end point of this assay is to obtain an international sensitivity index (ISI), reagent
cleavage of a small peptide by an activated coagulation factor manufacturers compare PTs obtained with commercial
that generates a change in color (optical density) propor- thromboplastin lots to a World Health Organization refer-
tional to the activity of the factor. Chromogenic assays are ence thromboplastin, with the behavior of human tissue fac-
more precise than clotting assays, but they may not detect tor, performed on plasma samples from healthy controls and
some defects in a factor that disrupt the binding of the factor stable, anticoagulated patients. A sensitive thromboplastin
to its natural (larger) substrate. However, chromogenic with an ISI of 1.0 is equivalent to human tissue, whereas a
assays can also be susceptible to interference from colored thromboplastin with an ISI of 2.0 is relatively insensitive to
substances such as hemoglobin or bilirubin that absorb light depletion of vitamin K-dependent clotting factors. The INR
at the assay’s defined wavelength of a photo-optical detec- is the ratio of the patient’s PT to the laboratory’s PT geo-
tion method or suspended particles such as lipoproteins that mean raised to the exponent of the thromboplastin ISI.
alter the pathway of the light. As a result, most chromogenic The INR is designed to accurately monitor patients who
assays have established tolerance limits for hemolysis, hyer- have been stabilized on warfarin. It is not intended for assess-
lipidemia, and hyperbilirubinemia. While it is recommended ing coagulopathies due to liver disease or DOACs because
to recollect samples to avoid these interfering substances, it the ISI is not validated for these conditions.
can be impractical or physiologically impossible, and the
coagulation laboratory may have to modify assay settings or
utilize an alternative method if available.
Isolated prolonged PT mixing study
prothrombin time
Prothrombin time
The PT measures the time to form a fibrin clot after adding

Not corrected Corrected
thromboplastin (source of tissue factor combined with
phospholipid) and CaCl2 to citrated plasma and assesses
three of the four vitamin K-dependent factors (factors II, Lupus anticoagulant Perform specific
testing assay for factor VII
VII, and X) plus factor V and fibrinogen. Commercial
thromboplastins contain either recombinant human tissue
factor combined with phospholipid or thromboplastins
derived from rabbit or bovine tissues. Almost all PT reagents
Positive Negative
contain a heparin-neutralizing additive to allow for moni-
toring of warfarin during concurrent heparin therapy.
Isolated prolongation of the PT most often reflects a defi-
Lupus Inhibitor of clotting factor:
ciency of vitamin K-dependent factors resulting from poor anticoagulant perform factor VII assay;
nutrition, inadequate absorption of vitamin K, antagonism of present if low, perform inhibitor assay
γ-carboxylation of the vitamin K-dependent factors by warfa-
rin, or decreased hepatic synthesis. Congenital deficiencies of Figure 11-4 Algorithm for evaluation of an isolated prolonged
factors X, V, and II and fibrinogen are rare (1 in 1-2 million prothrombin time.

Activated partial thromboplastin time deficiency while patients with type 3 vWD typically have a
severe deficiency of FVIII.
The aPTT is a two-step assay to measure the time to form a
Factor XI deficiency should be investigated when a pro-
fibrin clot after incubation of citrated plasma with phospho-
longed aPTT is encountered in a person of Ashkenazi Jewish
lipid and negatively charged particles followed by the addi-
ancestry. Bleeding risk is variable and does not correlate par-
tion of CaCl2. The negative surface and phospholipid activate
ticularly well with the severity of factor XI deficiency.
the contact factors (factor XII, prekallikrein [PK], and high-
LACs can cause a prolonged aPTT (see section on throm-
molecular-weight kininogen [HMWK]) and factor XI. The
bophilia testing). If a prolonged aPTT does not substantially
addition of CaCl2 permits activation of factor IX and the
shorten when repeated on a 1:1 mix with PNP, perform LAC
remaining reactions to proceed to form a fibrin clot.
testing or specific factor activities, depending on the clinical
Causes of an isolated prolonged aPTT include preanalyti-
context.
cal artifacts, congenital factor deficiencies, acquired inhibi-
Inhibitors to factor VIII are detected in 25%-30% of males
tors, and anticoagulation therapies (Figure 11-5).
with severe hemophilia A due to the development of alloan-
Deficiencies of factors VIII, IX, XI, XII, PK, and HMWK
tibodies to infusions of foreign factor VIII. Alloantibody for-
prolong the aPTT. Severe deficiencies of factor XII, PK, and
mation to factor IX in males with severe hemophilia B occurs
HMWK are rare, typically produce aPTTs >100 seconds and
less often. Acquired hemophilia caused by autoantibodies to
do not cause a bleeding disorder. Depending on the coagula-
factor VIII is the most common acquired specific factor
tion reagents and instrument in use, for an isolated intrinsic
inhibitor. A factor VIII antibody is suspected in patients
factor deficiency to prolong the aPTT, activity is usually
without a significant bleeding history who present with
30%-40%.
severe bleeding symptoms and coagulation testing shows a
Factors VIII and IX deficiencies, or hemophilia A and B,
prolonged aPTT that fails to fully correct immediately after
respectively, are X-linked inherited disorders that often are
1:1 mixing and subsequently prolongs after a 1- to 2-hour
diagnosed early in life due to spontaneous bleeding or a posi-
incubation of the 1:1 mixture at 37°C. A very low or unde-
tive maternal family history of hemophilia. Occasionally,
tectable factor VIII activity and mild inhibitor patterns for
diagnosis is delayed until adulthood if it is a mild deficiency
factors IX and XI due to partial inhibition of factor VIII in
(5%-40%).
these aPTT-based activity assays confirm the presence of a
Patients with type 1 vWD may have a slightly prolonged
specific factor VIII inhibitor. The Bethesda assay determines
aPTT if the factor VIII level is low, as von Willebrand factor
the potency of a factor VIII inhibitor by incubating dilutions
(vWF) serves to stabilize FVIII. Patients with the type 2 Nor-
of patient plasma combined 1:1 with PNP at 37°C for 2
mandy variant of vWD can have a moderate factor VIII
hours, followed by determination of residual factor VIII
activity. The antibody titer is expressed in Bethesda units
(BU) equal to the reciprocal of the patient plasma dilution
Isolated aPTT mixing study required to obtain recovery of 50% of the expected factor
prolonged aPTT
VIII activity in the incubated 1:1 mixture. A titer of 0.5-5.0
BU/mL is a low titer and may be overwhelmed with larger
infusions of factor VIII, whereas a titer >10 BU/mL will
Not corrected Corrected
require treatment of bleeding episodes with a factor VIII
bypassing agent, such as recombinant factor VIIa or acti-
Lupus anticoagulant • Do factor VIII, IX, XI vated prothrombin complex concentrate. The Nijmegen
testing activities. If all normal, modification of the 1:1 mix conditions improves accuracy
then
• Do factor XII, PK, and and precision of the Bethesda assay for low-titer inhibitors.
HMWK activities Most hospitals use aPTT-based nomograms to guide ther-
apeutic anticoagulation with unfractionated heparin. A ther-
Positive Negative apeutic aPTT range typically is determined by collecting
plasma samples from patients on heparin and comparing
aPTTs to heparin activity using the chromogenic anti-Xa
Lupus Inhibitor of clotting factor. Perform: assay. The aPTT therapeutic range in seconds will corre-
anticoagulant • specific factor assay for VIII, IX, XI
present • inhibitor assay for factor that is decreased spond to an anti-Xa range of 0.3-0.7 IU/mL. The aPTT also
is used to monitor the parenteral direct thrombin inhibitor
Figure 11-5 Algorithm for evaluation of an isolated prolonged argatroban, and the therapeutic target recommended by the
activated partial thromboplastin time (aPTT). HMWK = high- manufacturer is 1.5-3.0 times the baseline aPTT. Therapeutic
molecular weight kininogen; PK = prekallikrein. infusions of direct thrombin inhibitors also prolong the

PT/INR, and the intensity depends on the specific direct Combined abnormalities of PT and aPTT
thrombin inhibitor and the thromboplastin reagent. The
Deficiency or inhibition of a factor in the common pathway
DOACs can prolong the aPTT and/or PT (Table 11-11);
(factors X, V, II, and fibrinogen), hypofibrinogenemia, acq
however, these assays cannot be used to predict plasma con-
uired dysfibrinogenemia, disseminated intravascular coagu-
centrations. As a result, assays based upon anti-IIa or anti-Xa
lation (DIC), and a LAC can cause combined prolongation
are available in some laboratories to measure drug concen-
of the PT and aPTT. Advanced liver disease can cause
trations in special clinical situations such as bleeding, break-
decreased hepatic synthesis of all coagulation factors, except
through thrombosis, suspected noncompliance, populations
for factor VIII, and acquired dysfibrinogenemias are sug-
at risk for drug accumulation, and prior to urgent surgery or
gested by a low fibrinogen level in a functional assay com-
administration of thrombolytic therapy.
bined with a normal or high level of immunologic fibrinogen
The anti-Xa assay is a variation of a chromogenic AT
(see the section “Fibrinogen assays” in this chapter). See
assay (see section on thrombophilia testing) comparing an
Figure 11-6 for evaluation of a prolonged PT and aPTT.
unknown concentration of heparin in the patient plasma to a
Inhibitors to factor V can develop following exposure to
calibration curve prepared with an unfractionated heparin,
bovine thrombin, which also contains bovine factor V, when
low-molecular weight heparin, or hybrid standard curve.
combined with fibrinogen to produce “fibrin glue” during
Following addition of activated factor Xa to the test plasma,
surgical procedures to control bleeding. Bovine factor V
the rate of factor Xa neutralization by AT is positively corre-
antibodies may cross-react with human factor V to cause
lated with the heparin concentration, and the rate of chro-
bleeding in some patients. Low factor V activity and specific
mogenic substrate cleavage by factor Xa is correlated inversely
in vitro inhibition of factor V confirm the diagnosis. Fortu-
with the heparin concentration. Directly monitoring heparin
nately, fibrin glue therapeutics containing either plasma-
anticoagulation with the anti-Xa assay is the preferred
derived or recombinant human thrombin are now available.
approach in some hospitals and is an alternative to the aPTT
Acquired prothrombin deficiency rarely accompanies
when unusually high rates of heparin infusion are required
LACs, causes moderately prolonged PTs, and can cause
or when a patient’s baseline aPTT is prolonged because of an
abnormal bleeding. The autoantibodies do not produce an
LAC or deficiency of a contact activator (factor XII, PK, or
inhibitor pattern in mixing studies because they are not
HMWK). Low-molecular weight heparins (LMWHs) may
directed against the active site of the molecule. Rather, they
minimally prolong the aPTT at therapeutic concentrations.
form immune complexes, increasing the clearance rate and
LMWHs typically do not require monitoring. Under certain
lowering prothrombin activity.
situations, however, including patients of extremely low and
high weights, pediatric patients, pregnant patients, and
patients with impaired renal function, monitoring plasma
Thrombin time
LMWH activity approximately 4 hours after a subcutaneous
injection using a chromogenic anti-Xa assay calibrated The TT measures the time required to convert fibrinogen to
against an LMWH is recommended. a fibrin clot, bypassing the intrinsic, extrinsic, and common
Screening
thrombin time (TT)
Normal TT: Prolonged TT:

perform PT and aPTT anticoagulant
mixing studies interference?
Not corrected: Yes: stop and No: begin

Corrected
do LA testing obtain new fibrinogen
sample evaluation
Figure 11-6 Algorithm for evaluation

of a prolonged prothrombin time (PT) Positive: Negative:
LA present possible Perform factor X, Fibrinogen clot–based
and activated partial thromboplastin
specific factor V, II assays assay for FDP, D-dimer
time (aPTT). FDP = fibrin degradation inhibitor Fibrinogen antigen assay
product; LA = lupus anticoagulant.

pathways. TT requires sufficient amounts of normal fibrino- (Durham, NC) are currently available in the United States.
gen and an absence of thrombin inhibitors and substances The change in viscosity of blood as it clots in a cup is trans-
that impede fibrin polymerization. The only reagent is mitted through a pin immersed into the blood through a
bovine or human thrombin, and the test sample is undiluted mechanical-electrical transducer, producing a tracing of clot
citrated plasma. firmness over time. Certain patterns correlate with coagu-
lopathies, hypofibrinogenemia, thrombocytopenia, and
• Unfractionated heparin, LMWH, argatroban, bivalirudin, hyperfibrinolysis. Most experience with viscoelastic testing
and dabigatran inhibit thrombin and prolong the TT. has been in liver transplantation, trauma, and cardiopulmo-
• Dysfibrinogenemias usually prolong the TT and are sus- nary bypass surgery settings, where rapid point-of-care
pected if the functional test (fibrinogen activity) is dis- hemostasis information is used to select blood component
proportionately low compared with an immunologic replacement products.
measurement of fibrinogen (fibrinogen antigen).
• Hypofibrinogenemia usually prolongs the TT when levels of
von Willebrand factor assays
fibrinogen are below approximately 90 mg/dL. l-Asparagi
nase can cause hypofibrinogenemia by inhibiting synthesis. Endothelial cells and megakaryocytes synthesize vWF, which
• Fibrin degradation products in very high concentrations undergoes dimerization and subsequent linkage to form
and M proteins can inhibit fibrin polymerization and pro- vWF multimers before secretion into the blood. Once
long the TT. released, large multimers undergo remodeling to smaller
• Heparin-like anticoagulants (heparan sulfates) have molecules via cleavage by the protease adisintegrin and
occurred in patients with multiple myeloma and other metalloprotease with thrombospondin (ADAMTS13). vWF
tumors. They prolong the TT by interacting with AT in a has multiple domains with specific functions to support its
manner similar to heparin. The reptilase time will be nor- two activities: adhesion to connective tissue and platelets and
mal in these patients. binding factor VIII. Although most deficiencies of vWF are
congenital, vWD deficiency also can be acquired—a condi-
tion known as the acquired von Willebrand syndrome
Fibrinogen assays
(AvWS). AvWS is often associated with lymphoprolifera-
The Clauss method is a modified TT in which fibrinogen rather tive disorders, particularly monoclonal gammopathy of
than the thrombin is limiting. The time to clot formation is unknown significance (MGUS), autoimmune disorders,
inversely proportional to fibrinogen activity calibrated against hypothyroidism, and severe aortic stenosis, as well as with
a standard of known concentration and expressed as milli- left-ventricular assist devices (LVAD). Laboratory testing for
grams per deciliter. The thrombin concentration usually is suspected vWD is challenging because of the variability of
high enough to not be affected by therapeutic concentrations personal and family bleeding histories, multiple types of
of heparin but can be affected by direct thrombin inhibitors. vWF defects, physiologic variables affecting vWF levels, and
Fibrinogen also can be measured in immunologic tests (radial analytical imprecision of certain vWF test methods. Repeated
immunodiffusion) to evaluate for possible dysfibrinogenemia. testing frequently is indicated to confirm abnormal results
before diagnosing a patient with vWD. See Chapter 8 for
additional information regarding clinical presentation, clas-
Reptilase time
sification, and management of vWD.
Reptilase is snake venom that cleaves only fibrinopeptide A
from fibrinogen (in contrast to thrombin, which cleaves both
Initial testing for vWD
fibrinopeptide A and fibrinopeptide B) and results in fibrin
clot formation. This assay is prolonged by hypofibrinogen- Global tests of primary hemostasis, including bleeding time
emia and most dysfibrinogenemias but is not prolonged by and PFA-100 closure times, lack both sensitivity and specific-
heparin because the reptilase enzyme is not inactivated by ity for vWD, and aPTT is an indirect and potentially insensi-
AT or direct thrombin inhibitors. tive screening test for low factor VIII activity. vWF antigen
concentration (vWF:Ag), vWF-mediated agglutination of
platelets (vWF:RCo) or vWF binding to collagen
Global hemostasis tests
(vWF:collagen binding activity), and factor VIII activity
Thromboelastography/thromboelastometry involves moni- measurements are sufficient initial screening tests. Reference
toring the viscoelastic properties of whole blood during clot intervals for these analytes vary based on blood type, with
initiation, contraction, and lysis. Two commercial instru- type O individuals having mean values approximately 25%
ments: TEG (Haemonetics, Braintree, MA) and ROTEM lower than nontype O controls. Some laboratories provide

blood type-specific reference intervals, whereas other labo- Specialized testing to classify vWD
ratories provide a single reference range (with lower limits of
Dismissing a diagnosis of vWD or confirming a diagnosis of
approximately 50%) and note that asymptomatic type O
type 1 or type 3 vWD usually can be accomplished by review-
individuals may have vWF antigen, activity, and factor VIII
ing vWF:Ag, vWF:RCo, and factor VIII activity results.
levels as low as 35%-40%. It is reasonable to consider vWF
vWF:RCo or factor VIII activity much lower than vWF:Ag,
levels in the range of 30% to 50% as risk factors for mild
however, is an indication for more specific testing. vWF
bleeding tendency and not necessarily as an indication of
multimer analysis provides qualitative information by iden-
inheritable disease. Fluctuations of vWF in patients during
tifying structural abnormalities that correlate with qualita-
physiologic alterations associated with acute stresses, the
tive defects in vWF adhesion (Figure 11-7). Electrophoresis
menstrual cycle, or pregnancy make the interpretation of
of plasma through low-concentration agarose gel separates
these analytes problematic, and patients may require repeat
vWF multimer bands by size, which are detected with radio-
testing. Several equivalent and accurate methods can be used
labeled, enzyme-linked, or fluorescent vWF antibodies.
to quantify vWF:Ag. Measuring vWF functional activity is
Analysis of the band patterns can distinguish normal or
another matter. The most widely used method and the cur-
subtly abnormal patterns (consistent with type 1 and 2N or
rent gold standard is the ristocetin cofactor assay (vWF:RCo),
2M vWD, respectively) from major losses of high- and
which can be performed by automated immunoturbidity
intermediate-size bands (consistent with type 2A, type 2B,
assays using lyophilized platelets and ristocetin or by platelet
and platelet-type vWD).
aggregometry, and assesses vWF binding to the platelet
The ristocetin-induced platelet aggregation assay (RIPA)
GPIb/IX/V complex. Ristocetin, an antibiotic, binds to vWF
is a variation on the vWF:RCo activity to investigate platelet
causing a confirmational change that mimics the effect of
adhesion defects. Several ristocetin concentrations (ranging
high shear stress in vivo to expose the platelet-binding
from 0.6 to 1.5 mg/mL) are added to separate aliquots of a
domain. The vWF:RCo activity is sensitive to both quantita-
patient’s platelet-rich plasma, while change in light trans-
tive deficiencies of vWF (type 1 and type 3 deficiencies) and
mission is monitored as platelets bind to vWF and aggregate
to mutations causing reductions in high- and intermediate-
(Figure 11-8). Normal and mild type 1 vWD platelet-rich
weight vWF multimers or defects in platelet binding (types
plasma typically produces no or minimal aggregation at low
2A, 2B, and 2M vWD). A vWF:RCo/vWF:Ag ratio of <0.7
ristocetin concentrations and increasing aggregation at
supports a qualitative, or type 2 vWF defect and warrants
specialized confirmatory testing (Tables 11-11 and 11-12).
The vWF:RCo assay is labor intensive, poorly standardized
and imprecise, leading to the development of alternative
methods to assess adhesive activity, including ristocetin
induced binding to recombinant wild type GPIb fragments
(vWF:GPIbR), spontaneous binding of vWF to a gain of
function mutant GPIb fragment (vWF:GPIbM), or by bind-
ing of a monoclonal antibody to a vWF A1 domain epitope
(vWF:Ab).
Table 11-12 Assays for vWD classification
vWD vWF vWF

type activity antigen RIPA FVIII Multimers
Type 1 ↓ ↓ ↓ ↓ Normal pattern

Type 2A ↓↓ ↓ ↓↓ ↓ ↓ High and
intermediate
Type 2B ↓↓ ↓ ↑↑↑ ↓ ↓ High
Type 2M ↓↓ ↓ ↓↓ ↓ Normal
NP 1 2A 2B
Type 2N* NI NI NI ↓ Normal
Figure 11-7 von Willebrand multimer patterns. NP = normal plasma;
Type 3 ↓↓↓ ↓↓↓ ↓↓↓ ↓↓↓ Undetectable
1 = type 1 von Willebrand disease (vWD) with normal bands but
RIPA = ristocetin-induced platelet aggregation; decreased staining intensity; 2A = type 2A vWD with loss of high and
vWD = von Willebrand disease; 2N = Normandy variant of vWD. intermediate multimers; 2B = type 2B vWD with loss of high-
*FVIII low. molecular weight multimers.

A Type 2B B Normal control

–20 –20
0 0
10 10
20 20
30 30
Aggregation (%)
Aggregation (%)
40 40
50 50
60 60
70 70
80 80
90 90
100 100
0 1 2 3 4 5 6 0 1 2 3 4 5 6
Minutes Minutes
Figure 11-8 Examples of platelet-rich plasma aggregation responses to a range of ristocetin concentrations [1 = 1.5, 2 = 1.2, 3 = 0.9,
4 = 0.6 (mg/mL)]. (A) Type 2B vWD patient showing >50% aggregation with all ristocetin concentrations; (B) normal control
demonstrating concentration-dependent aggregation.
higher concentrations. Platelet-rich plasma from severe Bleeding disorders with normal screening
type 1 and types 2A and 2M vWD patients produces attenu- hemostasis tests
ated aggregation at high ristocetin concentrations, whereas
Abnormal, typically delayed bleeding due to severe factor XIII
platelet-rich plasma from type 2B or platelet-type vWD
deficiency and fibrinolytic pathway defects is rare, yet it should
patients shows an enhanced aggregation response to low
be considered when evaluations for coagulopathies and pri-
ristocetin concentrations. Estimates of the relative fre-
mary hemostasis defects are negative. Thrombin activates factor
quency of type 2B vWD to platelet-type vWD range from
XIII, and factor XIIIa cross-links fibrin monomers to produce a
8-10 to 1. Although the disorders have similar clinical pre-
durable clot. The urea clot lysis test is a qualitative screening test
sentations and inheritance is autosomal dominant, they
for severe factor XIII deficiency. Thrombin is added to plasma,
require different types of hemostasis replacement products
and the clotted fibrin is added to a high-molar solution of urea
(vWF concentrate versus platelet transfusion, respectively).
that will disrupt the clot if fibrin has not been cross-linked by
Mixing studies using normal washed platelets plus patient
factor XIIIa. Alternative quantitative assays are available to
plasma, or vice versa, can distinguish whether the patient’s
directly quantify factor XIII antigen and activity. Global screen-
vWF or platelet receptor is abnormal. In addition, some ref-
ing tests of the fibrinolytic system include the euglobulin clot
erence laboratories perform platelet-vWF binding assays
lysis time (ELT), which measures the time to lyse a fibrin clot in
using a vWF monoclonal antibody to assess the ability of a
the absence of plasmin inhibitors, and the whole blood clot lysis
patient’s vWF to bind formalin fixed platelets in the pres-
time (see section “Global Hemostasis Tests” in this chapter).
ence of low-dose ristocetin. Genotyping to detect known
Congenital hyperfibrinolysis is due to deficiencies of tissue plas-
mutations associated with each disorder is offered by a few
minogen activator (tPA) or plasmin natural inhibitors, and
reference laboratories. Rarely, men and women with mild
laboratory evaluation requires a panel of analytes, including
or moderate factor VIII deficiencies lacking X-linked inher-
plasminogen, plasminogen activator inhibitor 1 (PAI-1) activ-
itance pattern consistent with hemophilia A may be homo-
ity, and antigen, tPA antigen, and a2-antiplasmin activity typi-
zygous for type 2N vWD (decreased vWF binding affinity
cally performed in reference laboratories.
for factor VIII) or compound heterozygous (type 1/2N).
Causes of acquired hyperfibrinolysis resulting in circulat-
Decreased binding of recombinant factor VIII to the
ing plasmin overwhelming a2-antiplasmin inhibition include
patient’s immobilized vWF in an enzyme-linked immuno-
decreased hepatic clearance of tPA due to advanced cirrhosis
adsorbent assay (ELISA) and equivalent vWF:Ag and vWF
or during liver transplantation, increased release of tPA from
activity results are consistent with type 2N vWD. Genotyp-
endothelial cells during cardiopulmonary bypass, amyloido-
ing specific for type 2N mutations is offered by a few refer-
sis, envenomization from several species of snakes, and as a
ence laboratories.

component of DIC associated with acute promyelocytic leu- including uremia, liver failure, and myeloproliferative and
kemia and rarely with solid tumors, including bladder or myelodysplastic disorders, but formal aggregation studies are
prostate cancer. Laboratory evidence for primary fibrinolysis usually not informative in these cases. Platelet function testing
includes reduced fibrinogen levels due to cleavage by plasmin, is technically demanding, time consuming, and poorly stan-
elevated fibrin(ogen) degradation products, and no signifi- dardized, although efforts are under way to develop guidelines
cant elevation of D-dimer levels because lysis of cross-linked for performing and interpreting these studies. The hematolo-
fibrin clot is not the dominant process. DIC is the result of a gist should be aware that labs use two different platforms to
primary disease process that leads to the release of tissue fac- analyze platelet aggregation: instruments that are used to test
tor or other coagulation-activating factors into the blood (see platelet-rich plasma and instruments that use whole blood
Chapter 2 for more details). Because of variations in the (whole blood aggregometry [WBA]). Testing is performed on
amount and rate of procoagulant material released deter- aliquots of citrated whole blood or platelet-rich plasma with
mined by the underlying disease, there are no diagnostic pat- different concentrations of agonists, such as adenosine diphos-
terns of laboratory results. In acute, overwhelming DIC, initial phate (ADP), epinephrine, and collagen; arachidonic acid,
platelet counts and fibrinogen levels are low, or serial testing which platelets metabolize to the agonist thromboxane A2 via
shows a downward trend. PT, aPTT, and TT may be pro- the cyclooxygenase pathway; and ristocetin to screen for plate-
longed, depending on the severity of consumption, and let GPIb/IX/V deficiency. Formation of platelet aggregates
D-dimer levels are markedly elevated, indicating unregulated causes an increase in light transmission over time (light trans-
thrombin activity and secondary fibrinolysis. mission aggregometry [LTA]). Figure 11-9 shows a normal
Vessel wall defects, such as collagen diseases (eg, Ehlers- aggregation response of platelet-rich plasma to collagen and
Danlos and Marfan syndromes), also can cause abnormal ADP, and a clear first and second wave with epinephrine, indi-
bleeding. In addition to physical examination and imaging cating initial aggregation in response to exogenous epineph-
information, genetic testing is becoming more readily avail- rine followed by additional, irreversible aggregation because
able for some of these syndromes. of a release of ADP from platelet-dense granules. The platelet
release reaction can be assessed in a lumi-aggregometer, which
simultaneously monitors WBA through changes in electrical
Platelet function tests
impedance as platelets aggregate and platelet activation when
In vitro assessment of platelet activation and aggregation released adenosine triphosphate combines with luciferin-
in response to selected platelet agonists should be reserved luciferase enzyme-releasing light. Certain patterns of platelet
for patients with convincing bleeding histories in whom aggregation responses to a panel of agonists are sensitive to
evaluations for coagulopathies, vWD, and moderate to
specific inherited and rare qualitative platelet disorders,
severe thrombocytopenia are negative. In addition, prescribed including Glanzmann thrombasthenia, Bernard-Soulier syn-
and over-the-counter medications that can inhibit platelet drome, and collagen receptor defects. Platelet secretion defects
function must be discontinued before testing. Many disease resulting from abnormal signal transduction and qualitative
processes can produce acquired qualitative platelet defects, and quantitative granule disorders are more common,
–20
0
10
20
30
Aggregation (%)
40
50
60
70
80
90
100
Figure 11-9 Representative platelet aggregation curves
performed on normal platelet-rich plasma. 1 = collagen 0 1 2 3 4 5 6 7 8 9
5 mg/mL; 2 = ADP 5 mg/mL; 3 = epinephrine 5 µM. Minutes

produce variable aggregation patterns, and require additional of laboratory methods to detect platelet-specific antibodies.
diagnostic tests that are not readily available for clinical use. Assays detecting total or surface-bound platelet immuno-
These tests, including platelet electron microscopy, may be globulins are nonspecific and are not recommended.
accessible through research laboratories.
Assays for heparin-induced thrombocytopenia
Global primary hemostasis screening tests
Heparin-induced thrombocytopenia (HIT) is a clinical diag-
The template bleeding time is an invasive test, fraught with nosis supported by serologic and functional assays. In vitro
difficult-to-control technical and patient variables, that lacks functional assays monitor activation of control platelets by
specificity and sensitivity for detection of primary hemosta- patient serum in the presence of therapeutic concentrations
sis disorders. Prolonged bleeding times performed on of heparin and at high heparin concentrations. Activation
asymptomatic patients do not predict a risk of abnormal with a low heparin concentration and no activation at high
bleeding during surgery or other invasive procedures. The heparin concentration are considered to be both specific and
test is performed by making a standard incision in the fore- sensitive for detection of platelet factor-4 (PF4) heparin-
arm using a spring-loaded blade while maintaining a blood immune complexes, which are capable of causing in vivo
pressure cuff at 40 mm Hg. Blood oozing from the incision is platelet activation, thrombocytopenia, and thrombosis. In
wicked away with filter paper every 30 seconds until bleeding North America, selective laboratories perform the serotonin
stops. The typical reference range in adults is approximately release assay to monitor carbon-14-labeled serotonin secre-
5-10 minutes. tion from control platelets. In Europe, heparin-induced
Most laboratories have discontinued performing bleed- platelet aggregation performed in microtiter wells with
ing times and substituted automated in vitro screening meth- visual detection of platelet aggregation is the preferred
ods, which do not require an incision and provide precise method. Both assays are technically difficult, labor-intensive,
results from samples of blood collected in citrate, yet have and not readily available.
similar limitations. The PFA-100 instrument monitors vWF- Commercial ELISA assays detect antibodies recognizing
dependent platelet adhesion and aggregation under condi- immobilized PF4 bound to heparin or polyvinyl sulfonate
tions that mimic the shear forces in the arterial circulation. complex. Although sensitive, HIT ELISA results are nonspe-
Citrated blood is aspirated through a minute hole in a mem- cific, detecting antibodies incapable of activating platelets in
brane coated with collagen and ADP (COLL/ADP) or collagen vitro or causing thrombocytopenia and thrombosis in vivo.
and epinephrine (COLL/EPI). vWF multimers bind to colla- The positive predictive value of a positive PF4 ELISA result
gen and platelets adhere to vWF, are activated by COLL/ADP alone to confirm a diagnosis of HIT is low, and if used as the
or COLL/EPI, aggregate, and occlude the aperture, which is only criterion, a positive PF4 ELISA results in the overdiag-
recorded as closure time in seconds. Each laboratory must nosis of HIT. Growing evidence supports three approaches
determine reference intervals, although typical ranges are to improving the specificity of PF4 ELISA testing. First, clini-
55-137 seconds and 78-199 seconds for COLL/ADP cians can improve the pretest likelihood that thrombocyto-
and COLL/EPI cartridges, respectively. Prolonged PFA-100 penia is due to HIT by applying a validated clinical scoring
closure time is not sufficiently sensitive for all congenital system called the 4Ts (thrombocytopenia, timing, thrombo-
qualitative platelet disorders and types of vWD to be used as sis, and other more likely causes of thrombocytopenia).
a general screening test. In addition, as anemia and throm- Patients with low 4T scores are unlikely to have HIT, even
bocytopenia worsen, closure times increase, and these vari- with a positive PF4 ELISA, removing the need for testing.
ables should be considered when interpreting prolonged This is especially true for patients who have an increased
closure times in the setting of hematocrit <30% and platelet likelihood of having a false positive test, such as patients
count <100 × 106/µL. Prolonged COLL/EPI closure time is a who have had a cardiopulmonary bypass procedure.
sensitive test for aspirin inhibition of platelets, but the COLL/ Second, identifying only IgG instead of a combination of
ADP closure time is insensitive to blockade of the platelet IgG/IgM/IgA PF4 antibodies improves the specificity of a
P2Y12 ADP receptor by thienopyridines. positive PF4 ELISA with a slight impact on sensitivity. Finally,
ample evidence suggests that the higher a PF4 ELISA optical
density (OD) is, the more likely a functional HIT assay will
Specialized testing for acquired
be positive and the clinical presentation and course will be
thrombocytopenia
consistent with HIT. No cutoff point, however, completely
Assays for platelet antibodies
segregates all platelet-activating antibodies from nonactivat-
Immune-mediated thrombocytopenia remains a clinical ing antibodies. Conversion from viewing PF4 ELISA results
diagnosis of exclusion due to the general poor performance as simply positive or negative to considering OD as a

continuous variable, with increasing probability for HIT as surgery, trauma, immobilization, cancer, or other acquired
OD increases, is still evolving as clinical research continues. risk factors. The likelihood of identifying a deficiency is
increased if thrombosis is recurrent or in an unusual loca-
tion, the patient is young (<45 years old), or the patient has a
Assays for TTP and vWF-cleaving protease (ADAMTS13)
positive family history of thrombosis. To avoid misleading
In sporadic cases of TTP, ultralarge forms of vWF initiate the low results due to temporary conditions related to acute ill-
formation of platelet aggregates and lead to thrombi and ness, thrombosis, and anticoagulation therapy, testing for
thrombocytopenia. In these cases, the activity of the vWF- non-genetic based assays ideally should be delayed until sev-
cleaving protease, ADAMTS13, typically is low (ie, <5%- eral weeks after completion of treatment when a patient has
10%), and in many cases, in vitro evidence of an inhibitory returned to baseline. The biologic and analytical variability
autoantibody is present. In hereditary forms of TTP, there associated with phenotypic diagnoses of these deficiencies
are mutations in the gene encoding the enzyme, and the requires verification of an abnormal test result on a new
activity of ADAMTS13 is absent or markedly decreased; sample. Because of the large number of mutations associated
however, no inhibitor is present. with deficiencies of AT, PC, and PS, genotyping is not rou-
The main methods that are currently used employ a tinely performed.
recombinant 73-amino-acid peptide from the A2 domain of
vWF containing the Y1605-M1606 bond recognized by
Antithrombin deficiency
ADAMTS13 to detect substrate cleavage by either ELISA or
fluorescence resonance energy transfer (FRET) methods. The most sensitive screening tests for AT deficiency are
Two amino acids in the peptide substrate are modified in the chromogenic activity assays designed to quantify AT inhibi-
FRET assay; one fluoresces when excited, and the other tion of factor Xa or IIa in the presence of unfractionated
absorbs or quenches the released energy. When ADAMTS13 heparin. Abnormal low-AT activity results require measure-
cleaves the substrate and separates the modified amino acids, ment of AT antigen to classify the deficiency as type I (activ-
emitted energy is detected in a fluorescent plate reader. The ity = antigen) or type II (activity < antigen). Type I AT
method for ADAMTS13 neutralizing antibody detection is deficiency is more common than type II deficiency in symp-
similar to the Bethesda assay for factor VIII inhibitors; dilu- tomatic kindreds. Subclassification of type II deficiency
tions of patient serum and plasma are mixed with PNP fol- requires performance of the chromogenic activity assay
lowed by measurement of residual enzyme activity using the without heparin to differentiate type IIa resulting from reac-
synthetic substrate. Typical reference values are ADAMTS13 tive site defects and IIb resulting from AT heparin-binding
activity >67% and inhibitor titer <0.4. Measuring ADAMTS13 defects. Although type IIb is associated with a low risk of
antigen is not necessary when evaluating a patient for spo- thrombosis, progressive AT activity assays are not readily
radic or idiopathic TTP. available and typically are not performed.
The decision about whether to initiate plasma exchange is
made on the basis of clinical assessment and should not be
Protein C deficiency
delayed by ADAMTS13 testing. Importantly, samples for
assessment of ADAMTS13 activity and inhibitor should be The preferred screening tests for PC deficiency are chromo-
obtained prior to transfusion with fresh frozen plasma or genic assays. PC is activated with a snake venom and PC
plasma exchange. ADAMTS13 testing may also be obtained activity correlates with hydrolysis of a synthetic peptide and
following completion of a course of plasma exchange as per- change in OD. Clot-based PC activity assays are an alterna-
sistently low ADAMST13 activity and positive inhibitor tive, but potentially inaccurate results may occur due to vari-
titers are predictors of relapse during remission. ations in factor VIII and PS levels, factor V Leiden, inhibitory
antibodies, and presence of some anticoagulants. An abnor-
mal low PC activity result requires measurement of the PC
Assays for thrombophilia
antigen to classify the deficiency as type I (activity = antigen)
Inherited deficiency of one or more of the identified natural or type II (activity < antigen).
inhibitors of coagulation (AT, PC, and PS) is a risk factor for
venous thrombosis, and functional and immunologic assays
Protein S deficiency
are available to measure these inhibitors. The use of these
assays generally should be restricted to patients in whom the PS assays are challenging because of the unique biology of
result may affect prognosis and duration of anticoagulant PS. Total plasma PS is partitioned between free and bound
treatment. This generally includes patients who present with forms. The protein is nonfunctional when bound to comple-
spontaneous thrombosis not temporally related to recent ment 4b binding protein and functional when it is free. In its

unbound form, the protein can serve as a cofactor for acti- correction in a 1:1 mix of patient and normal pooled plasma,
vated PC (aPC). The typical PS bound-to-free ratio of 60:40 and (iii) confirming phospholipid dependence by shortening
varies under different physiologic and pathologic conditions. the clotting time with the addition of more phospholipid.
Clot-based PS activity assays are the most sensitive screening Although some LACs are discovered when a routine aPTT is
tests for PS deficiency but suffer from potential inaccuracy prolonged, a normal aPTT is generally not a sensitive LAC
because of the same variables that can affect PC activity test- screening test and should not prevent performance of more
ing. An alternative screening assay is free PS antigen concen- sensitive LAC testing based on the clinical circumstances.
tration to avoid confounding variables. Free PS testing, There is no gold-standard LAC method. Recent updated
however, is insensitive to type II PS deficiency (low activity consensus expert guidelines from the International Society
but normal free antigen level). Some laboratories screen with of Thrombosis and Hemostasis Scientific Subcommittee on
PS activity, some screen with free PS antigen, and other labo- Lupus Anticoagulant/Phospholipid Antibodies recommend
ratories use both assays. performing two sensitive LAC tests in parallel—one aPTT-
based test and one Russell’s viper venom (activation of factor
Factor V Leiden and prothrombin gene mutations Xa)–based test—and accepting a positive result from either
or both as evidence of an LAC. Preanalytical variables requir-
Two autosomal inherited coagulation factor variants increase ing attention include platelet contamination (>10,000/µL)
the risk for VTE; these are factor V G1691A (factor V Leiden) due to inadequate centrifugation, which can produce false
and prothrombin G20210A. Several sensitive commercial negative LAC results because of the neutralizing effect of
clot-based screening assays for factor V Leiden mutation platelet-derived phospholipid, and concurrent anticoagula-
demonstrate a resistance of factor Va cleavage by aPC in the tion therapy. The presence of a direct thrombin inhibitor or
presence of factor V Leiden mutation. Coagulation testing, factor Xa inhibitor in the test plasma nullifies the validity of
activated with aPTT, PT, or Russell’s viper venom reagents, is LAC testing. Heparin can be neutralized by additives in the
performed with or without added aPC, and the clotting LAC test reagents or in a separate step before testing, and the
times are expressed as a ratio. Abnormally low ratios repre- mixing step can compensate for mild to moderate coagu-
sent aPC resistance (aPCr). Specificity is improved by repeat lopathies due to liver disease or vitamin K antagonists like
testing of positive plasmas after dilution with factor warfarin. The preferred time, however, for LAC testing is
V-depleted plasma to minimize impact of inhibitors, antico- before or after anticoagulation treatment. Rarely, a specific
agulants, and high factor VIII levels. Genotyping should be factor inhibitor can cause a false positive LAC result, typi-
performed on all aPCr-positive patients determine whether cally with an aPTT-based LAC test due to a factor VIII inhib-
they are heterozygous or homozygous for factor V Leiden. itor. A more frequent occurrence, however, is the appearance
Although prothrombin G20210A mutation is associated of multiple coagulation factor deficiencies when the true
with elevated prothrombin levels, measuring factor II activ- coagulation factor levels are within normal limits; this mis-
ity is not a sensitive screening test, and genetic testing is the leading picture occurs because the same antibodies respon-
primary method. sible for the LAC effect also interfere with coagulation factor
assays. The hematologist should be aware that rare patients
concurrently may have both an LAC and a true factor VIII
Antiphospholipid syndrome
inhibitor. Abnormal bleeding likely would be present, and
APS is an important acquired thrombotic condition. specific factor assays would confirm an isolated factor defi-
Consensus-based criteria have been developed for the inves- ciency. LAC tests are either positive or negative, and evidence
tigational diagnosis of APS. These criteria require a combi- is insufficient to support reporting gradations of positive
nation of clinical conditions (unexplained venous or arterial results. Because of differences in test methods, reagents,
thromboembolic events, pregnancy complications) and per- instrumentation, preanalytical variables, and approaches to
sistent laboratory evidence of autoantibodies that recognize analyzing and reporting results, there is substantial interlab-
epitopes on selected proteins associated with phospholipids oratory variability of LAC results based on external profi-
and identified by coagulation-based (LACs) or serologic- ciency testing surveys.
based (anti-cardiolipin and anti–b2-glycoprotein IgM and LAC can cause reagent-dependent prolongations of PT
IgG antibodies) testing. LACs are heterogeneous antibodies results. Although this is usually mild, occasionally LAC-
that interfere with in vitro clotting assays. Indirect evidence positive patients will have elevated INRs before starting war-
for the presence of a LAC requires: (i) prolongation of a farin. Chromogenic factor X activity is an alternative to the
screening clotting assay designed to be sensitive to the INR for therapeutic anticoagulation monitoring (target
phospholipid-dependent behavior of LAC, (ii) ruling out 20%-40%); however, availability of the test is limited.
prolongation due to a coagulopathy by showing incomplete Another option is to measure PT-based factor II, VII, and X

Bibliography | 301
activities and observe whether the LAC produces an inhibi- Bibliography

tor pattern on the serial dilutions of plasma. If one or more
factor assays appear unaffected by the LAC, then suppression General hematology
of a specific clotting factor can serve as the therapeutic target Briggs C. Quality counts: new parameters in blood cell counting.
for warfarin anticoagulation. A markedly prolonged PT in Int J Lab Hematol. 2009;31:277-297.
the setting of LAC may be a result of acquired factor II defi- Buttarello M, Plebani M. Automated blood cell counts. Am J Clin
ciency due to a nonneutralizing prothrombin autoantibody Pathol. 2008;130:104-116.
that increases the clearance rate. These patients are at risk for Old JM. Screening and genetic diagnosis of haemoglobinopathies.
spontaneous bleeding. To recognize this rare condition, a Scand J Clin Lab Invest. 2007;67:71-86.
factor II activity level should be obtained in an LAC-positive
patient with a prolonged PT/INR. Hematopathology
Performance of ELISA testing for anti-cardiolipin (aCL)
and anti–b2-glycoprotein I (ab2GPI) IgM and IgG antibodies Betz BL, Hess JL. Acute myeloid leukemia diagnosis in the 21st
century. Arch Pathol Lab Med. 2010;134:1427-1433.
should accompany LAC testing to maximize sensitivity
Foucar K. Non-neoplastic disorders of lymphoid cells. In: Foucar
because persistently positive (arbitrarily defined as >12 weeks
K, Reichard K, Czuchlewski D, eds. Bone Marrrow Pathology.
apart) results from serologic tests or LAC, or both, fulfill the 3rd ed. Chicago, IL: American Society of Clinical Pathology.
laboratory criteria for APS. Commercial ELISA kits for aCL 2010:450-471.
and ab2GPI lack standardization, and interlaboratory agree-
ment is poor for weakly positive sera. To improve specificity,
some experts consider only medium and high titer-positive Hemostasis testing
IgG and IgM aCL and ab2GPI results to be clinically impor- Kamal AH, Tefferi A, Pruthi RK. How to interpret and pursue an
tant. In addition, significant immunosuppression (especially abnormal prothrombin time, activated partial thromboplastin
humoral) may lead to false negative results with ELISA based time, and bleeding time in adults. Mayo Clin Proc. 2007;82:
testing. While other antiphospholipid antibody specificities 864-873.
are currently not included in the classification criteria, anti- Ortel TL. Antiphospholipid syndrome: laboratory testing and
bodies to ab2GPI and anti-phosphatidylserine/prothrombin diagnosis strategies. Am J Hematol. 2012;87:S75-S81.
(aPS/PT) antibodies have been shown to be predictive of
thrombotic risk.

CHAPTER
12
Transfusion medicine
William Savage and Suzanne Bakdash
Introduction, 303 Transfusion of plasma Transfusion support in special clinical
Red blood cell transfusion, 303 products, 315 settings, 321
Platelet transfusion, 310 Pretransfusion testing, 316 Transfusion risks, 329
Granulocyte transfusion, 313 Apheresis, 318 Bibliography, 335
Introduction system as the ABH system. Blood group O is most common

in the white US blood donor population (45%), followed by
Transfusion medicine encompasses blood collection, pretrans- group A (40%), group B (11%), and group AB (4%). In Afri-
fusion compatibility testing, transfusion of blood components can Americans, the order of frequency is similar, but there
for the appropriate indications, and recognition and evalua- are fewer group A (27%) and more group B (20%).
tion of adverse reactions to transfusion. Specific hematology The ABO gene locus is on chromosome 9. The A and B
populations, such as sickle cell disease patients and hemato- genes encode transferase enzymes that covalently attach the
poietic stem cell transplant (HSCT) recipients, pose unique specific terminal saccharide moiety to the subterminal galac-
transfusion-related challenges that are of particular relevance tose. The large number of A and B transferase polymor-
to hematologists. Apheresis is another integral component of phisms has precluded the development of reliable genotyping
transfusion medicine that includes therapeutic apheresis, methods for the determination of ABO blood group. In most
which removes or modifies a constituent of whole blood con- cases, the gene underlying the group O phenotype is identi-
tributing to disease pathogenesis, and peripheral blood stem cal to the A transferase gene except for a single base-pair
cell (PBSC) harvesting for allogeneic or autologous HSCT. deletion that leads to premature termination of translation.
Interestingly, blood group O provides a selective advantage
against severe malaria. Individuals with blood group O also
Red blood cell transfusion have lower levels of von Willebrand factor (vWF), which
needs to be taken into consideration in the diagnosis of mild
ABO system type I von Willebrand disease. Conversely, nongroup O indi-
The ABO system is a group of carbohydrate antigens viduals have a greater risk of venous thromboembolism
defined by their terminal saccharide moiety. The subtermi- attributable to higher levels of vWF and FVIII.
nal galactose, in association with a constitutively expressed Healthy people past infancy nearly always produce
fucose moiety, defines the H antigen. The addition of immunoglobulin M (IgM) anti-A or anti-B antibodies, also
N-acetylgalactosamine or galactose to the subterminal galac- known as isohemagglutinins, directed against the respective
tose yields red blood cells (RBCs) of group A or group B, ABO antigens that are not present on their own cells. Thus,
respectively. Individuals who express both sugars are group group O individuals have so-called naturally occurring anti-
AB, whereas individuals who express neither of these sugars A and anti-B antibodies, group A individuals have anti-B
are group O. As the H antigen remains unmodified in these antibodies, group B individuals have anti-A antibodies, and
group O individuals, some authors refer to the ABO antigen group AB individuals have neither. ABO compatibility is
the most important factor in determining whether blood
from a specific donor can be transfused to a specific reci
Conflict-of-interest disclosure: Dr. Savage received research pient. Preformed recipient isohemagglutinins predictably
funding from Fresenius-Kabi. Dr. Bakdash declares no competing induce acute hemolysis if ABO-incompatible RBCs are
financial interest.
Off-label drug use: not applicable.
tra
nsfused. Because anti-A and anti-B isohemagglutinins
| 303

304 | Transfusion medicine
are predominantly of the IgM isotype, and thus efficient at in donor ABO antibody-mediated hemolysis of the recipi-
fixing complement, the ensuing hemolysis is intravascular ent’s RBCs. For example, transplantation of a liver from a
and can be severe leading to shock, renal failure, dissemi- group O donor would be acceptable for a group A recipient
nated intravascular coagulation, and death. In blood group because the recipient’s anti-B antibodies will not cause
O individuals, anti-A and anti-B antibodies may be of both humoral rejection of the transplanted organ. However, if the
the IgM and IgG isotypes. An additional antibody, anti-A,B, solid organ contains passenger lymphocytes from the group
which cross-reacts with both type A and type B RBCs, also is O donor which are producing anti-A and anti-B antibodies,
found in type O individuals and is predominantly of the IgG the anti-A antibody may cause hemolysis of the recipient’s
isotype. Because IgG antibodies may cross the placenta, circulating type A RBCs.
whereas IgM antibodies cannot, the presence of IgG isohem-
agglutinins in blood group O individuals explains the higher
Rh system
frequency of ABO hemolytic disease of the fetus and new-
born (HDFN) in blood group O mothers with non–blood Clinically, the Rh blood group system is second in impor-
group O fetuses and newborns. tance to the ABO system. Rh antigens are proteins and,
Variations in the strength of ABO antigen expression are unlike the ABO system, antibodies to Rh antigens are rarely
occasionally clinically significant. For example, group A sub- present unless a person has been immunized by pregnancy,
groups, such as A1 and A2, demonstrate differences in the transfusion, or stem cell transplantation. The Rh system is a
activity of their glycosyltransferase enzyme activity resulting collection of 52 different antigens encoded by two genes, des-
in quantitative and qualitative difference in A antigen expres- ignated RHD and RHCE. The RHD and RHCE genes are
sion. Type A2 individuals manifest substantially weaker A 97% identical, include 10 exons, and evolved from a gene-
antigen expression and occasionally can develop antibodies duplication event on chromosome 1. Individuals who are
directed against the more common type A RBCs (referred to referred to as Rh positive express the D antigen; approxi-
as the A1 subgroup, which accounts for approximately 80% mately 85% of whites and 92% of blacks are RhD positive.
of group A individuals) despite the presence of detectable A Individuals who are Rh negative do not express D, either
carbohydrate. As the differences in A antigen expression because they have a complete deletion of the RHD gene
among RBCs of differing A subgroups are qualitative as well, (European descent) or have nonfunctioning RHD resulting
the anti-A1 antibodies made by non–A1-expressing individ- from premature stop codons, gene insertions, or other causes
uals are not self-reactive. Anti-A1 antibodies typically bind (Asian and African descent). On a protein level, D and CE are
only to A1-positive RBCs at nonphysiologic temperatures 417-amino acid, nonglycosylated transmembrane proteins
often resulting in ABO typing discrepancies. When anti-A1 that are predicted to span the membrane bilayer 12 times.
antibodies are reactive at 37°C or the antihuman globulin Whereas one protein carries the D antigen, and identifies an
(AHG or Coombs phase), however, only A2 or O RBCs individual as Rh positive or Rh negative, the protein prod-
should be used for transfusion. Subgroups of B antigen exist ucts of the two RHCE alleles carry four CE antigens (C, c, E,
as well but are encountered much less frequently. e) that are inherited in various combinations (eg, ce, cE, Ce,
In addition to being expressed on RBCs, ABO antigens or CE). D differs from CE by 32-35 amino acids, depending
also are expressed on endothelial cells. ABO compatibility on which of the four possible forms of CE protein is present.
typically is required for solid organ transplantation to avoid This magnitude of difference between the two Rh proteins
the risk of ABO antibody-mediated acute humoral rejection. may explain the relatively high degree of immunogenicity of
For example, if a blood group O recipient is transplanted the D antigen to the Rh-negative individual when compared
with a solid organ from a blood group A donor, the recipi- with the immunogenicity of other blood group antigens in
ent’s anti-A antibody can mediate humoral rejection and which single amino acid changes distinguish their polymor-
destruction of the transplanted organ. There are few excep- phic alleles.
tions to the requirement for ABO compatibility in solid Inheritance of the Rhnull phenotype, in which none of the
organ transplantation, most of which involve donors of the previously listed Rh antigens are expressed on the cell sur-
A2 subgroup or infants who have not yet begun to produce face, is associated with stomatocytic erythrocytes and low-
isohemagglutinins. ABO compatibility is not required for grade hemolytic anemia. The Rh polypeptides D and CE
hematopoietic stem cell (HPSC) transplantation because facilitate the assembly of major transport proteins in the
ABH antigens are not expressed on HPSCs. Passenger lym- RBC membrane, such as band 3.
phocyte syndrome may be seen in both solid organ and Blood is routinely typed and matched for the presence of
minor ABO incompatible HPSC transplants. In these cases, the D antigen for two primary reasons. First, the D antigen is
passenger donor B-lymphocytes may continue their isohem- highly immunogenic and approximately 80% of D-negative
agglutinin production in the recipient, potentially resulting individuals exposed to D become alloimmunized to D, which

Red blood cell transfusion | 305
can result in hemolysis if RhD positive RBCs are transfused. significant in terms of their ability to cause hemolysis or
Second, anti-D antibodies can cause significant HDFN. HDFN. Therefore, such individuals should be transfused
Before the advent of prophylaxis with Rh(D)-immune glob- with Rh-negative RBCs and, if pregnant, should be given
ulin (eg, RhoGAM; Ortho-Clinical Diagnostics, Raritan, Rh(D)-immune globulin. In practice, these individuals are
NJ), anti-D commonly caused HDFN. Prevention of immu- often identified only after they have formed anti-D despite
nization to the D antigen in women of childbearing potential typing as Rh positive; they may also present with inconsis-
is extremely important, and failure of prophylaxis continues tent RhD typing results with different reagents that recog-
to account for a significant percentage of cases of HDFN. nize different epitopes of the D polypeptide. Molecular
Rh(D)-immune globulin is 99.9% effective in preventing genetic test methods may be utilized to distinguish weak and
maternal alloimmunization to D when administered to an partial D.
RhD negative woman at 28 weeks of pregnancy (typically as For blood donors, reagents and techniques to detect weak-
a single dose of 300 mcg), and again at delivery if the new- ened D expression are paramount so that units from donors
born is Rh positive (post-delivery dose is calculated based on with weak or partial D antigens are labeled Rh positive to
the estimated volume of fetal-maternal hemorrhage). The avoid sensitizing patients who truly are Rh negative. As it is
exact mechanism by which Rh(D)-immune globulin pre- usually not possible to distinguish between weak and partial
vents sensitization in the Rh-negative individual when D phenotypes by serologic testing alone, most transfusion
exposed to Rh-positive RBCs remains unknown. It has been service laboratories choose Rh-typing reagents and methods
proposed that Rh-positive fetal RBCs coated with Rh(D)- that will not detect weak D phenotypes when testing patients.
immune globulin in the maternal circulation serve to cross- This creates a situation in which an individual may be classi-
link surface immunoglobulin to inhibitory Fc receptors on fied as Rh negative as a patient but Rh positive as a blood
maternal naïve B-cells to render them anergic, although this donor. Enhanced testing for RhD is also performed on
mechanism has yet to be proven experimentally. The other newborns of Rh negative women to determine if maternal
major antigens of the Rh system—C, c, E, and e—are also post-delivery prophylaxis with Rh(D)-immune globulin is
relatively potent immunogens and can cause HDFN of vary- indicated.
ing severity, albeit at lower frequencies than D. No immune
globulin preparations are available for the prevention of
Other protein antigen blood group systems
alloimmunization to Rh antigens other than D.
Variations in the strength of expression of Rh antigens can Outside the ABO and Rh systems, most clinically significant
be clinically significant. In the case of D, clinical significance blood group alloantibodies are directed against protein-
depends on the etiology of its reduced expression. Two types based antigens, particularly antigens in the Kell, Kidd, Duffy,
of scenarios are worthy of discussion. The first is strictly and MNSs systems (Table 12-1). These systems are defined
quantitative in nature—that is, RBCs express a structurally by protein (as opposed to carbohydrate) antigenic determi-
normal D protein, but they do so in reduced amounts. Indi- nants and, in general, antibodies to these antigens are
viduals with this weak-D phenotype can be considered Rh acquired only after exposure by transfusion, pregnancy, or
positive; that is, they are immunologically tolerant to D and via HSCT. Some patients appear predisposed to develop
thus can receive wild-type Rh-positive RBCs. Likewise, if antibodies and may form several antibodies simultaneously,
pregnant with an Rh-positive fetus, they would not need to which can limit the availability of donor blood. In the acute
receive Rh(D)-immune globulin to prevent D sensitization. phase of alloimmunization to nonself protein antigens,
The second scenario is one in which an individual’s D anti- T-cell-independent IgM antibodies may appear first, which
gen is qualitatively different than wild-type D, which, in turn, subsequently isotype switch to IgG. As is the case with anti-
can lead to a quantitative reduction in the level of cell surface bodies directed against antigens of the Rh blood group sys-
expression. In many of these cases, the qualitative alteration tem, antibodies directed against other protein antigen
is the result of the replacement by homologous recombina- systems are typically of the IgG isotype when discovered
tion of one or more exons of the RHD gene by the corre- during pretransfusion testing.
sponding exon(s) of the nearby RHCE gene. As a consequence, Antibodies to certain blood group antigens are seen in
the RBCs of such individuals express D proteins that are chi- patients more commonly than others. This is due to differ-
meric in nature and are made up of pieces of CE, referred to ences in antigenicity and relative frequencies of antigens in
as partial D antigens. Such individuals may type as Rh posi- the population. For example, the K antigen of the Kell blood
tive (depending on the particular formulation of anti-D group system is expressed on the RBCs of approximately
typing reagent used); however, after being transfused with 10% of individuals. Therefore, there is a sizable proportion
true Rh-positive RBCs, may make anti-D alloantibodies to of individuals who are capable of mounting an immune
the D epitopes they lack. Such antibodies may be clinically response to K (90%) and a reasonable chance (10%) of one

Table 12-1 Commonly-occurring red blood cell (RBC) antigens of clinical significance
RBC Molecule Hemolytic Hemolytic disease of the

antigen expressing Antibody immune/ transfusion reaction fetus and newborn (HDFN)
system antigen Function of molecule naturally occurring from antibody from antibody
ABO Glycoprotein or Unknown Naturally occurring Yes, acute Yes, usually mild (IgG anti-
glycolipid A,B generally present in
blood of group O mothers)
Rh Protein Ammonium ion transport Immune Yes, delayed Yes, can be severe
Kell Glycoprotein Member of neprilysin Immune Yes, delayed Yes, often severe;
(M13) family of zinc hypoproliferative
metalloproteases component
Kidd Glycoprotein Urea transport Immune Yes, delayed Yes
Duffy Glycoprotein Chemokine receptor DARC Immune Yes, delayed Yes
(Duffy antigen receptor
for chemokines)
MNSs Glycoprotein Structural role in Naturally occurring Rare (anti-M/N); Rare (anti-M/N);
RBC membrane (anti-M/N); yes (anti-S/s) yes (anti-S/s)
(glycophorins A and B) immune (anti-S/s)
P Glycolipid Unknown Immune (anti-P); Yes (anti-P); rare Yes, mild
naturally occurring (anti-P1)
(anti-P1)
of these individuals receiving a unit of K-positive cells dur- Another important protein antigen system is the Kidd sys-
ing a transfusion. Consequently, anti-K antibodies are com- tem, which is located on the erythrocyte urea transporter.
monly identified antibodies. Antibodies directed against antigens in the Kidd system are
Anti-K antibodies can cause accelerated clearance of trans- notorious for their involvement in delayed hemolytic trans-
fused cells as well as significant HDFN. Unlike HDFN due to fusion reactions (DHTRs). The pathophysiology of DHTRs
anti-Rh blood group antibodies, the anemia associated with will be discussed later, but briefly note the following: (i) an
HDFN due to anti-K appears to have a hypoproliferative individual is sensitized via transfusion, pregnancy, or HSCT;
component as well, because of the expression of Kell antigens (ii) the antibody titer decreases over time, such that the anti-
on fetal hematopoietic progenitor cells. From a clinical man- body becomes undetectable by standard serologic techniques
agement standpoint, the effects of maternal anti-K antibod- at the time that the antibody screen is performed; (iii)
ies on the fetal RBC precursors that express K render the use because of the apparently negative antibody screen, the
of the maternal anti-K antibody titers less reliable as an indi- patient is transfused with an ABO and RhD-compatible unit;
cator of fetal risk than titers in Rh antibody–associated (iv) upon re-exposure to the Kidd antigen, the recipient
disease. The Kell blood group system is interesting for several develops a rapid anamnestic antibody response, which
other reasons as well. The protein bearing the Kell system results in clinically significant hemolysis several days after
antigens is related structurally to a number of metalloprote- the transfusion. Although antibodies to other blood group
ases, including CALLA (common acute lymphoblastic leuke- antigens can cause DHTRs, the severity of DHTRs resulting
mia antigen, or enkephalinase) and endothelin-converting from antibodies of the Kidd blood group system is com-
enzyme. In addition, weakened expression of inherited Kell pounded by the fact that Kidd antibodies, although IgG, are
antigens may be associated with a rare phenotype, called the excellent at fixing complement resulting in the more clini-
McLeod phenotype, because of a deficiency of a protein cally significant intravascular form of hemolysis.
called Kx. Unlike Kell, Kx is encoded by the XK gene on the Alloantibodies that develop to antigens in the Duffy blood
X chromosome, near the locus affected in the X-linked form group system may cause mild to severe acute or delayed
of chronic granulomatous disease (CGD); therefore, in some hemolytic transfusion reactions and HDFN. The Duffy glyco-
families, the McLeod phenotype is associated with CGD. protein itself is structurally related to chemokine receptors
McLeod phenotype cells are classified as spur cells (acantho- that bind interleukin (IL)-8, monocyte chemotactic protein-1
cytes) because they contain sharp, irregular cytoplasmic pro- (MCP-1), and other chemokines, although its function on
jections. Absence of the Kx protein is responsible for most RBCs is not clear. It may impart RBCs with the ability to scav-
instances of congenital acanthocytosis associated with neu- enge excess chemokines from the circulation. The Duffy gly-
rologic dysfunction. coprotein also serves as a receptor for the malarial parasite

Plasmodium vivax, which explains the decreased expression antigen. The I antigen is also the predominant specificity for
of Duffy antigens in blacks, compared to whites. There is RBC autoantibodies responsible for IgM-mediated autoim-
some evidence that the Duffy glycoprotein is expressed on mune hemolytic anemia (AIHA; cold agglutinin disease).
nonerythroid tissue and represents a minor histocompatibil- As is the case with antibodies directed against ABO anti-
ity antigen in kidney transplantation. gens, antibodies directed against other carbohydrate antigens
The MNSs blood group system is highly complex, includ- tend to be of the IgM isotype. One exception to this rule is
ing 46 antigens that reside on one or both of the major RBC found in paroxysmal cold hemoglobinuria (PCH), in which
membrane glycoproteins—glycophorin A (GPA) and gly- Donath-Landsteiner antibodies, which are cold-reacting IgG
cophorin B (GPB). The red cell antigens M and N reside on autoantibodies directed against the P antigen, can fix com-
GPA, which is critical to the invasion of RBCs by Plasmo- plement on circulating RBCs resulting in intravascular
dium falciparum. Alloantibodies to the M and N antigens are hemolysis. Of note, the DAT (direct antiglobulin or Coombs
generally IgM antibodies that are not reactive at 37°C and test) is usually paradoxically positive for complement and
rarely are clinically significant. In contrast, alloantibodies to negative for IgG in PCH because the Donath-Landsteiner
the S and s antigens, which reside on GPB, are clinically sig- IgG autoantibodies usually detach from circulating RBCs
nificant IgG antibodies that can cause hemolytic transfusion after fixing complement. PCH is sometimes associated with
reactions and HDFN. syphilis or nonspecific childhood viral infections.
Other carbohydrate antigen blood Blood group genotyping

group systems
Over the past two decades, the molecular bases of almost all
Carbohydrate antigen systems other than the ABO system the major blood group antigens have been elucidated. It is
are rarely significant in clinical transfusion practice, but now known that the majority of blood group polymor-
some issues of pathophysiologic importance are of interest. phisms are caused by simple single-nucleotide polymor-
These systems include Lewis, P, and Ii. phisms in the genes encoding the protein antigens or genes
The Lewis antigens (Lea and Leb) and the P1 antigen are encoding the glycosyltransferases for the carbohydrate anti-
common targets of cold-reacting IgM alloantibodies. Lewis gen systems. The initial application of blood group genotyp-
antigens are technically not blood group antigens because ing was in the prenatal management of iso-immunized
they are not intrinsic to RBCs, but rather they are acquired pregnancies. Fetal DNA extracted from amniocytes allowed
passively by absorption from the plasma. Lewis antigens for the determination of fetal RhD status in a mother known
expressed on gastric mucosa serve as receptors for Helico- to be sensitized to RhD while avoiding the much riskier pro-
bacter pylori. People who lack all P system antigens (pp phe- cedure of cordocentesis. The technology then evolved to per-
notype) may produce a clinically significant antibody mit the same analysis on the basis of free fetal DNA in
directed against the P antigen. These individuals also are maternal plasma, eliminating the risk of amniocentesis alto-
resistant to parvovirus B19 infection because the P antigen gether. In some European countries, pregnant women who
on RBCs acts as the receptor for this virus. The Pk antigen is are Rh negative (with a partner who is Rh positive) and are
a receptor for Shiga toxins and Pk expression may also mod- not known to be sensitized undergo such noninvasive
ulate host resistance to HIV infection. molecular testing to determine fetal RhD status, which then
The expression of the Ii antigen system is age dependent. dictates whether Rh immune globulin prophylaxis is given at
In newborns, the predominant allele is the i antigen, which 28 weeks gestation. In North America, such testing is less
includes linear repeats of N-acetylglucosamine and galactose widely available and universal prenatal prophylaxis with Rh
(N-acetylgalactosamine). In older individuals, the predomi- immune globulin in Rh-negative women remains the stan-
nant allele is the I antigen, which includes the same polysac- dard of care. As noted above, in white populations, the RhD-
charides but is arrayed in a branched configuration rather negative phenotype results from a deletion of the RHD gene
than a linear configuration. Activity of the “branching on chromosome 1. In black populations, a significant por-
enzyme” that forms the branched structure is absent in fetal tion of RhD-negative individuals do not have a complete
erythrocytes but appears at about 6 months of age. Fetal and gene deletion. This difference is important for genotype-
cord blood cells thus express strong i and weak I antigens, phenotype correlation of DNA-based RhD typing assays. In
whereas adult erythrocytes express i weakly and I strongly. addition, D-variant alleles are more common in people of
Individuals with infectious mononucleosis sometimes African descent. Molecular classification of D variants allows
develop cold agglutinins directed against the i antigen, for the recognition of partial D phenotypes—that is, indi-
whereas people with Mycoplasma pneumoniae infections viduals who type as RhD positive but are at risk of develop-
sometimes develop cold agglutinins directed against the I ing anti-D if transfused with wild-type Rh-positive blood or

pregnant with a wild-type Rh-positive fetus. Black sickle cell date, there is no difference in morbidity and mortality out-
patients have a higher prevalence of RHD and RHCE vari- comes when “fresh” versus standard-age blood was used.
ants, which can lead to anti-Rh antibodies in almost half of
chronically transfused patients, despite the provision of phe-
Clinical transfusion of RBCs
notypically matched blood.
In the past several years, array-based high-throughput
approaches to blood group genotyping have become com- Clinical case
mercially available. Blood group genotyping is invaluable in A 29-year-old man with chronic renal failure has a hemoglobin
multitransfused patients, such as patients with sickle cell dis- (Hb) of 6.7 g/dL and is seen in the emergency department
ease, whose erythrocytes represent a mixture of donor and (ED) for a shoulder injury. The patient has normal heart rate
recipient and as such cannot be accurately phenotyped using and blood pressure. He states that he usually has this degree
conventional serologic techniques. Patients with warm auto- of anemia, has recently begun therapy with darbepoietin and
antibodies whose erythrocytes are coated with IgG (leading iron through his nephrologist, and is able to conduct his daily
to a positive direct antiglobulin or Coombs test) are also dif- routines without difficulty. The attending ED physician orders a
ficult to antigen type with conventional antisera for many blood transfusion.
protein antigens and may benefit from blood group genotyp-

ing. Availability of blood group genotypes in these and other Clinically, the starting point for transfusion is deciding
patient populations allows for provision of phenotypically whether it is indicated (further discussion below). The next
or genotypically “antigen matched” blood units (beyond most important consideration for ensuring safe administra-
ABO/Rh) to patients to prevent additional alloimmunization tion of blood products is definitive identification of the patient.
or reduce the risk of hemolytic transfusion reactions when Specifically, it is imperative that the labeling of the type and
the reactivity on the antibody screen cannot be resolved. crossmatch sample, as well as the definitive identification of
the unit to be transfused, occur at the patient’s bedside.
After a request for RBCs and a properly labeled specimen
Collection and storage of RBCs
have been obtained with meticulous identification, the next
Most RBCs collected in the United States are obtained from step in the transfusion of RBCs is selection of the appropriate
healthy volunteer donors. Collection of autologous RBCs unit(s). The steps involved in the selection of an RBC unit
and RBCs from directed donors is possible, although they include A-, B-, and D-antigen typing of the patient’s RBCs;
are becoming less common. Most units of whole blood col- screening of the patient’s serum for antibodies to clinically
lected from volunteer donors are fractionated into one or significant RBC antigens (called the antibody screen); and
more transfusable components, including RBCs, platelets, performing a crossmatch, in which immunologic compati
plasma, and others. bility between the patient and the prospective RBC unit
RBCs can be stored in anticoagulated plasma with or is assessed (further details of this process are presented in
without additive solutions (AS). One commonly used AS for the section “Pretransfusion Testing”). Finding crossmatch-
RBC storage is AS-1 which contains glucose, adenine, and compatible blood for individuals who have been alloimmu-
mannitol. RBCs are stored routinely for up to 42 days at 4°C nized to RBC antigens after prior pregnancies or transfusions,
in currently available storage media. Techniques for freezing such as multitransfused sickle cell patients, may take hours to
RBCs impart a shelf life of 10 years or greater in special situ- days. Close communication with the blood bank about antic-
ations, such as rare blood types. ipated need for transfusion is critical. Accessing rare blood
Cold storage of RBCs has long been known to induce types through major regional or national blood centers that
biochemical changes in the RBCs, such as decreased 2,3- maintain collections of frozen erythrocytes may be required.
diphosphoglycerate (2,3-DPG) levels, which are mostly
Whole blood is not widely available in the United States;
reversible in vivo after transfusion. Retrospective studies in therefore RBCs units are far more frequently transfused.
surgical patients suggested that transfusion of RBCs stored Table 12-2 summarizes the most common available RBC
for >2 weeks was associated with increased postoperative products and their respective indications. For acute blood
complications and mortality. These and studies in other pop- loss, RBCs are used either alone or in combination with crys-
ulations sparked several randomized controlled trials (RCT) talloid and/or colloid solutions or plasma. Washed RBCs are
to prospectively investigate differences in outcomes after the indicated for patients who have had severe allergic or ana-
transfusion of fresher versus older stored RBCs in cardiac sur- phylactic reactions to blood transfusion. Although washed
gery (the RECESS trial), in intensive care unit (ICU) patients erythrocytes have been recommended to avoid excessive
with respiratory failure (the ABLE and TRANSFUSE trials), complement exposure in patients with paroxysmal noctur-
and preterm neonates (the ARIPI trial). In results reported to nal hemoglobinuria (PNH), it is not known to what degree

Table 12-2 Characteristics and indications for various red blood cell and platelet products
Product Characteristics Indication(s)
Whole blood 450 mL; coagulation factors adequate; platelets To provide increased oxygen-carrying capacity and blood volume
low in number; not widely available
Red blood cells (RBCs) 250-300 mL; can be stored up to 42 days To provide increased oxygen-carrying capacity
Leukocyte-reduced Contain < 5 × 106 leukocytes per unit To reduce the incidence of febrile nonhemolytic reactions, CMV
RBCs transmission, HLA alloimmunization, and platelet transfusion
refractoriness
Leukocyte-reduced, Leukoreduced and irradiated To reduce the risk of transfusion-associated graft-versus-host
irradiated RBCs disease, in addition to the benefits of leukoreduction listed above
Washed RBCs Saline-suspended red cells, 200-250 mL To provide RBC support to patients with severe or recurrent allergic
or anaphylactic reactions (including IgA deficiency with allergic
reactions); washed RBCs are also provided for intrauterine
transfusions
Deglycerolized frozen 200 mL; rare RBCs are frozen in glycerol (to To support patients with antibodies requiring transfusion of rare
RBCs prevent hemolysis) and need to be washed blood group phenotypes
and deglycerolized prior to transfusion
Pooled platelets* 300-325 mL, 4-6 whole blood donors Prophylaxis and treatment of bleeding in the setting of
thrombocytopenia or platelet dysfunction
Single-donor apheresis 150-350 mL, 1 apheresis donor Same as pooled platelets; limits donor exposure
platelets*
HLA-matched Apheresis platelet from a donor with known Immune-mediated platelet transfusion refractoriness with
platelets* HLA type, matched to patient documented anti-HLA antibodies
* Platelet products should be subjected to leukoreduction or irradiation for the same indications as discussed for red blood cells.
this is necessary. Washing rarely is indicated to reduce the The primary goal of erythrocyte transfusion is to improve
extracellular potassium load in adult patients, even those the oxygen-carrying and delivery capacity of blood in patients
with renal insufficiency, but it may be indicated in the pedi- with anemia. Erythrocyte transfusion can also aid in the overall
atric setting particularly if large volumes of older RBC units management of hypovolemia in patients with intravascular
need to be transfused to an infant or neonate. Cryopreserved volume depletion because of massive acute blood loss. It is
erythrocytes primarily are used for multiply alloimmunized important to note, however, that concomitant medical issues
patients who require units of rare blood type. may stymie increased oxygen delivery despite increased oxygen
Cellular blood products, including RBCs and platelets, carrying capacity; for example, arterial thrombosis or ongoing
are contaminated with small numbers of leukocytes sometimes massive hemorrhage. Numerous compensatory mechanisms
referred to as passenger leukocytes. Evidence suggests that pas- exist to maintain oxygen delivery in the face of anemia. These
senger leukocytes play an important role in alloimmunization include increased pulse rate and cardiac contractility, periph-
to human leukocyte antigens (HLAs), transmission of cytomeg- eral vasodilatation, increased oxygen delivery to tissues result-
alovirus (CMV) infection, cytokine-mediated febrile nonhe- ing from decreased oxygen affinity of hemoglobin because of
molytic transfusion reactions, transfusion-associated increased erythrocyte 2,3-DPG concentration and decreased
graft-versus-host disease (TA-GVHD), and other adverse plasma pH, and altered oxygen consumption and utilization
events. Reduction in the number of passenger leukocytes (leu- within the tissues. Studies in healthy people indicate that a
koreduction) results in clinically important reductions in the shift to anaerobic metabolism occurs at hemoglobin levels of
incidence of platelet transfusion refractoriness, alloimmuniza- approximately 7.5 g/dL or lower when the blood hemoglobin
tion to HLA antigens, and transfusion-transmitted CMV infec- concentration is reduced rapidly. Below this level, compensa-
tion. As a result, there has been a trend towards the universal use tory mechanisms to enhance oxygen transport are likely to be
of prestorage leukoreduction of both RBCs and platelets, par- inadequate in patients with relatively rapid-onset anemia.
ticularly in patients who are likely to require prolonged transfu- Studies of Jehovah’s Witness patients, who refuse allogeneic
sion support. Leukoreduction does not provide protection blood transfusion, show that mortality increases with hemo-
against TA-GVHD, so irradiation of all cellular blood products, globin levels under 6 g/dL. As a result of these factors, there is
in addition to leukoreduction, is necessary for patients at no fixed hemoglobin target for the transfusion of RBCs. An
increased risk of TA-GVHD (see discussion in the section otherwise-healthy individual may tolerate a blood hemoglobin
“Graft-versus-host disease [TA-GVHD]” later in the chapter). concentration of 6 g/dL or less, as long as the anemia was

gradual in onset. In contrast, someone with a severe cardiac or Platelet transfusion

pulmonary disorder or an individual with acute blood loss
may require a higher blood hemoglobin concentration to The ABO system
maintain clinical stability. RCT evidence from the TRICC trial Platelets express A, B, and H antigens to varying degrees.
indicates that hemodynamically stable patients in the ICU About 10% of group A and B people have high antigen
should not be transfused unless their hemoglobin is <7 g/dL. expression, which can impact platelet increments in major
A 2013 RCT showed that a restrictive transfusion strategy ABO incompatible transfusion.
improved outcomes in selected patients with acute upper-
gastrointestinal bleeding. The FOCUS trial, which compared
The HLA system
hemoglobin triggers of 8 g/dL vs. 10 g/dL in high-risk patients
after hip surgery indicated that the restrictive strategy did not Alloimmunization to HLA antigens is the major cause of
improve functional outcome (walking across the room), nor immune-mediated refractoriness to platelet transfusion in
did it result in an increase in cardiac events, even though the patients undergoing chronic platelet transfusion therapy. It
mean age was 82 years and a majority had a history of cardio- is important to note, however, that nonalloimmune causes of
vascular disease. The TRACS (Transfusion Requirements after platelet refractoriness are more common (eg, ITP, hyper-
Cardiac Surgery) trial established that a restrictive transfusion splenism, and consumptive coagulopathy).
strategy (8 g/dL transfusion trigger) is noninferior to a liberal Although only HLA class I antigens at the HLA-A and
(10 g/dL transfusion trigger) strategy. The Transfusion Indica- HLA-B loci have been shown to be important in causing
tion Threshold Reduction Trial (TITRe2) also established immune-mediated refractoriness to platelet transfusion,
equivalence of infectious and ischemic outcomes when com- given the high degree of polymorphism in the HLA system,
paring 7.5 to 9 g/dL transfusion triggers (35.1% vs. 33.0%, large numbers of HLA-typed donors need to be available to
respectively; P = 0.3). Overall mortality was higher in the blood centers to provide HLA-compatible platelets to indi-
restrictive group (4.2% vs. 2.6%), even though 30 day mortal- vidual patients. If HLA-matched platelet donors are not
ity was more similar (2.6% vs. 1.9% for restrictive and liberal available, identification of the specificity of the patient’s
arms, respectively). The evidence regarding a restrictive or lib- HLA antibodies may allow blood centers to provide antigen-
eral transfusion strategy in patients with acute coronary syn- negative platelets for transfusion (ie, platelets that do not
dromes is unclear. A randomized trial comparing red cell express HLA antigens against which the patient has known
transfusion thresholds in HSCT patients is under way in Can- antibodies). HLA antigens can be categorized into groups
ada, and bleeding is one prespecified outcome. with common epitopes that may cross-react with the same
In the clinical case previously described, the attending phy- HLA antibodies; these groups of HLA antigens are referred
sician’s initial decision to administer RBCs in response to a to as cross-reactive groups (CREGs). When an exact HLA-
hemoglobin value, without taking the patient’s overall presen- identical platelet donor is not available, blood centers also
tation into account, was incorrect because it failed to consider can use CREGs to locate platelet donors in whom the risk of
the fact that young, otherwise healthy individuals with cross-reactivity between the recipient’s antibodies and the
gradual-onset anemia often tolerate low hemoglobin levels donor’s antigens may be minimized.
without difficulty. Therefore, the case illustrates the impor-
tance of using bedside clinical judgment in making transfusion
Human platelet antigens
decisions rather than relying on arbitrary hemoglobin cutoffs.
In addition to anti-HLA antibodies, antibodies to platelet-
specific antigens may also cause platelet transfusion refracto-
Key points riness. The human platelet antigens (HPAs) arise as a result
• The ABO system is the most important determinant of
of polymorphisms involving various platelet membrane gly-
transfusion compatibility. coproteins. Differences in HPA allelic frequencies in differ-
• Rh compatibility is necessary because of the high immunoge- ent ethnic populations may partially account for differences
nicity of the Rh D antigen and role of anti-D antibodies in in the rates of alloimmunization to HPA antigens reported
hemolytic disease of the fetus and newborn, and delayed by different investigators.
hemolytic transfusion reactions. There are a number of well-characterized HPA antigen
• Other frequently relevant blood group systems include Kell, systems, but alloimmunization is most commonly due to
Kidd, Duffy, and MNSs. polymorphisms involving HPA-1a/1b system (previously
• There is no fixed threshold for transfusion of RBCs. RCT known as the PLA1/A2 system). The HPA-1a/1b system
evidence to date supports restrictive RBC transfusion strategies in
arises from a polymorphism on the b3 subunit of the platelet
many clinical settings.
fibrinogen receptor, GPIIb/IIIa, also known as integrin a2bb3

Platelet transfusion | 311
or CD41/CD61. In addition to ethnic differences in allelic platelets must be stored at room temperature, which increases
frequencies, alloimmunization to HPA-1a is strongly associ- the risk of bacterial growth and currently limits the storage of
ated with expression of HLA-DRB3*0101 and HLA- platelets to 5 days. Clinical studies indicate that there is rela-
DQB1*0201 in the recipient. tively little loss of platelet function and viability during this
Alloimmunization to HPAs can cause neonatal alloim- time. The storage lesion primarily involves platelet activation,
mune thrombocytopenia (NAIT) and posttransfusion pur- which is reflected in platelet shape change, adhesion, aggrega-
pura (PTP) and accounts for a small proportion of tion, secretion of platelet granular contents, and the expression
immune-mediated platelet transfusion refractoriness in of activation antigens.
multiply transfused platelet transfusion recipients. In addi-
tion, there are case reports of alloimmune thrombocytope-
Clinical transfusion of platelets
nia after HPA-mismatched allogeneic HSCT. PTP is a
syndrome in which transfused platelets are destroyed by
HPA alloantibodies through a process analogous to a delayed Clinical case
hemolytic transfusion reaction. However, following expo- A 56-year-old multiparous female develops acute myeloid leuke-
sure to the HPA antigen in question through RBC or platelet mia and receives induction therapy. Her platelet count decreases
transfusion, what then follows is the apparent immune to <10,000/µL. The patient initially responds well to prophylactic
destruction of the patient’s own antigen-negative platelets, transfusion with pooled platelet concentrates. Later in the hos-
in addition to any transfused antigen-positive platelets. The pitalization, however, her 1-hour post-transfusion platelet count
mechanism by which autologous platelets are destroyed in increments are persistently <5,000/µL. Having obtained HLA
PTP is unclear, although a process involving cross-reactivity typing on the patient before induction, the attending physician
of HPA alloantibodies to patient platelets is a favored expla- asks the blood bank for HLA-matched platelets.
nation. From a transfusion perspective, for the patient with a

history of PTP, RBC units should be washed to remove any
contaminating platelets that could incite an additional epi-
Prophylactic platelet transfusion
sode of PTP. For platelet transfusions, alloantigen-negative
platelets should be selected. Bleeding in thrombocytopenic patients occurs at all platelet
counts, but several studies indicate that the rate of spontane-
ous bleeding does not dramatically increase until the platelet
Collection and storage of platelets
count is 5,000/µL or less. For many years, a platelet count of
Two basic types of platelet products are available for clinical ≤20,000/µL was considered to be the appropriate threshold
use: pooled and single-donor apheresis platelet products. for prophylactic platelet transfusion. Several prospective ran-
Pooled platelets are obtained by pooling individual platelet domized transfusion trials, however, showed no differences
concentrates derived from whole blood units obtained from in hemorrhagic risks between prophylactic platelet transfu-
4-6 volunteer whole blood donors. The platelet content of sion triggers of ≤10,000 and ≤20,000/µL. A randomized, con-
pooled platelet products varies depending on the number of trolled noninferiority trial of no platelet prophylaxis versus
units in the pool and various technical factors. prophylaxis (TOPPS trial) in patients with hematologic
The second basic type of platelet products is single-donor malignancies demonstrated no prophylaxis was statistically
platelets (SDPs). Unlike pooled platelet concentrates, which inferior to prophylaxis, with a trigger of <10,000/µL, although
are derived from multiple volunteer donors’ whole blood, not by a large margin: WHO grade 2 or higher bleeding was
SDPs are collected from single donors using continuous cen- 50% with no prophylaxis versus 43% with prophylaxis.
trifugation plateletpheresis techniques in which RBCs and Common indications for raising the prophylactic platelet
plasma are returned to the donor. Plateletpheresis collection transfusion target include blast crisis or acute promyelocytic
techniques have been refined such that a minimum of leukemia during induction; recent or imminent invasive
3 × 1011 platelets—that is, approximately the same number procedures; qualitative platelet dysfunction due to uremia,
of platelets contained in a pool of 6 whole-blood derived drugs, or other causes; concurrent coagulopathy; fever;
platelets—can be collected from a single donor in a single hypertension; and acute pulmonary processes. In patients
session. with significant active bleeding, most clinicians target the
As with RBCs, leukoreduction of platelet products can result platelet count to 50,000 or up to 100,000/µL in patients with
in reductions in the incidence of platelet transfusion refra definite or suspected central nervous system bleeding. Realis-
ctoriness, alloimmunization to HLA antigens, transfusion- tic target counts should be set in patients who do not respond
transmitted CMV infection, and febrile nonhemolytic well to platelet transfusion, such as those with splenomegaly
fusion reactions. For optimal viability and function, or immune-mediated platelet transfusion refractoriness.
trans

The survival of endogenously produced platelets in healthy may be used to prevent alloimmunization to RhD, particu-
individuals is in the range of 7 to 9 days. The half-life of plate- larly in females of childbearing potential.
lets in individuals undergoing induction chemotherapy for
acute leukemia, HSCT, or other acute medical situations, how-
ever, is typically 1-3 days. Checking an immediate (<1 hour) Platelet transfusion dose
postinfusion platelet count is a necessary screen for platelet The dose of platelets administered to a thrombocytopenic
refractoriness. As detailed later in “Diagnosis and manage- patient depends on the therapeutic goal. If the primary goal
ment of platelet transfusion refractoriness,” poor immediate is to administer a sufficient number of platelets to prevent
postinfusion counts may reflect the presence of alloantibodies bleeding in an uncomplicated patient, the target typically
to HLA antigens, which is an indication for requesting HLA- would be to transfuse when the platelet count drops below
matched products. Table 12-2 summarizes the major platelet 10,000/µL. To select the appropriate platelet dose factors that
preparations and their respective indications. affect the response to individual transfusions should be
taken into consideration, including the size of the patient
Choice of platelet product and the presence of splenomegaly, active bleeding, dissemi-
nated intravascular coagulation, anti-HLA or other anti-
The current evidence indicates that apheresis platelets and platelet antibodies, and other factors. In addition, it appears
pooled platelets can generally be used interchangeably for that thrombocytopenia itself decreases platelet survival such
most platelet transfusions. An argument that often has been that more severely thrombocytopenic patients may require
proposed in favor of apheresis platelets over pooled platelets is higher doses or more frequent transfusions to maintain a
the theoretical reduction in the incidence of transfusion- particular peripheral blood platelet count.
transmitted infectious diseases. Because of the very low abso- FDA standards dictate that SDPs (single donor apheresis
lute magnitude of the infectious risk associated with platelets) must contain at least 3 × 1011 platelets and that
transfusion of blood products (discussed in the section “Infec- individual platelet concentrates prepared from single units
tious complications” later in this chapter) in comparison with of whole blood must contain at least 5.5 × 1010 platelets, that
the magnitude of the other treatment-related and disease- is, the equivalent of approximately 3 × 1011 platelets per 5- or
related risks to which most platelet transfusion recipients are 6-pool. In an average-size patient, in the absence of any of
subject, the cost effectiveness of requiring single-donor trans- the risk factors for poor platelet transfusion response listed
fusions for all platelet transfusion recipients is questionable. previously, approximately 3 × 1011 platelets is considered an
In contrast to the availability of universal RBC donors appropriate adult dose, and it is expected to increase the
(blood group O negative) and universal plasma donors platelet count by 20,000-50,000/µL. If a patient is being
(blood group AB), universal donors for platelets do not managed as an outpatient, larger doses of platelets may
exist because platelet products contain both platelets and a extend the interval between transfusions. A multicenter RCT
substantial quantity of plasma (typically ~300 mL). For (Platelet Dose Trial) compared low-, typical-, and high-
example, group O platelets contain anti-A and anti-B isohe platelet doses of prophylactic platelets for a platelet count of
magglutinins that would react against the RBCs of all but 10,000/µL in patients undergoing chemotherapy or HSCT.
type O recipients, whereas group AB platelets likely will yield WHO grade 2 or higher bleeding was the same in all groups,
a decreased platelet count increment when transfused into and the low dose group received significantly fewer platelets,
all but type AB recipients because the recipient’s anti-A and/ albeit over more transfusion episodes.
or anti-B antibodies may result in immune-mediated clear-
ance of platelets expressing the A and B antigens. Ideally,
patients should receive ABO-identical platelets; in reality, Diagnosis and management of platelet
platelets are in short supply, mainly because of their short transfusion refractoriness
shelf life (5 days). As there is no absolute requirement for
A commonly used bedside definition of platelet transfusion
ABO compatibility of this product, blood banks have vary-
refractoriness is two consecutive postinfusion platelet count
ing procedures and policies for managing out of group plate-
increments ≤10,000/µL. A more formal definition of refrac-
let product transfusions. Transfusion of a platelet product
toriness, which adjusts for both the size of the patient and
from an RhD-positive donor to an RhD-negative recipient
the number of platelets actually infused, uses the corrected
uncommonly (<1% incidence) may result in anti-D anti-
count increment (CCI), which is based on a platelet count
body formation because of exposure to the minimal volume
obtained within 1 hour of transfusion, calculated as follows:
of residual RhD positive RBCs in the platelet product. In
situations in which Rh negative platelets are unavailable and CCI = body surface area (BSA; m2) × platelet count
platelet transfusion is required, Rh immune globulin (RhIG) increment × 1011/number of platelets transfused

Granulocyte transfusion | 313
For example, if 3 × 1011 platelets (standard dose, as described cross-reactivity are selected for transfusion. It is not clear
above) are transfused to a patient with a BSA of 1.8 m2, and which method (HLA matching or platelet crossmatching) is
the posttransfusion increase in platelet count is 23,000/µL, superior, and some centers use a combination of both meth-
then the CCI =1.8 m2 × 23,000/µL × 1011/3 × 1011 = 13,800. ods. Even when a suitable HLA-matched donor is identified,
Platelet transfusion refractoriness often is defined as two or it can take several days to obtain a product for transfusion, as
more consecutive postinfusion CCIs of <5,000-7,500. the donor typically has to be called in to donate specifically
A trial of fresh, ABO-matched platelets may increase the for the patient in question, and the subsequent donation
post transfusion increment modestly. As noted previously, must pass all infectious disease testing before release.
alloimmunization to HLA antigens accounts for some cases A variety of approaches have been taken when no com
of platelet transfusion refractoriness. In the absence of obvi- patible platelets can be found for a patient who is alloim
ous nonimmune causes of platelet transfusion refractori- munized to HLA antigens. Platelet transfusion refractoriness
ness, such as splenomegaly and disseminated intravascular in HSCT recipients can be managed by obtaining platelets
coagulation, the clinician should request an anti-HLA anti- from the original stem cell donor. In other settings, the
body evaluation. Anti-HLA antibodies, and their specifici- rapeutic modalities used have included corticosteroids,
ties, typically are detected on high-throughput platforms plasmapheresis, intravenous γ-globulin, frequent platelet
such as Luminex microbeads coated with HLA class I and II transfusion, continuous-infusion platelet transfusion, and
antigens. In patients whose panel-reactive antibody (PRA) ε-aminocaproic acid. Clinical data do not clearly support
screen is positive, platelets should be selected based on HLA any one of these modalities over the others. Realistic targets
matching, avoiding the antibody specificities found in the should be set in alloimmunized patients who are not
responding well to platelet transfusion or those for whom
patient, or platelet crossmatching, although these methods
HLA-matched products are unavailable. Transfusion of
do not always guarantee improved platelet responses. There
multiple units of platelets from random donors, whether
is no evidence that the use of single-donor or HLA-matched
pooled or apheresis, with no realistic expectation of an
platelets enhances response to platelets in the absence of
increase in platelet count or cessation of bleeding, exposes
documented alloimmunization to HLA antigens. Alloim-
the patient to all the risks of transfusion with no benefit. In
munization sometimes appears to resolve spontaneously;
addition, it may result in reduced availability of platelet
thus, the requirement for HLA-matched products may not
products for other patients when supplies are limited.
persist indefinitely.
Platelets express all HLA class I antigens, but HLA-A and
HLA-B antigens are the clinically significant antigens in HLA Key points
immune platelet refractoriness. Therefore, most blood cen-
• Antibodies directed against HLA antigens can develop
ters optimize matching only at the HLA-A and HLA-B loci. following blood transfusion or pregnancy and are the most
Depending on the HLA type of an individual, one may have important cause of immune-mediated platelet transfusion
little or great difficulty in locating platelets that are HLA iden- refractoriness.
tical, or at least within the same CREG as the patient. A grad- • Human platelet antigens are polymorphisms on platelet
ing system has been developed to semiquantitatively define surface glycoproteins that may also mediate platelet transfusion
the degree to which the platelet donor and the platelet reci refractoriness, as well as NAIT, PTP, and alloimmune thrombocy-
pient are matched at these loci, although the predictive value topenia following HSCT.
of this system is modest. The relatively low-stringency, • Prophylactic platelet transfusion should be considered when
the peripheral blood platelet count decreases below 10,000/µL
serologic, four-loci, HLA-matching protocols that transfu-
in uncomplicated patients. The platelet count target should be
sion medicine specialists typically use to select platelet prod-
increased in the presence of additional risk factors for bleeding
ucts is quite different from the relatively high-stringency, or platelet consumption, including invasive procedures,
molecular-level, 10- to 12-loci, HLA-matching schemas that coagulopathy or anticoagulation, fever, hypertension, or acute
HSCT specialists typically use to select HSCT donors. Never- pulmonary processes.
theless, for some patients with unusual HLA types locating an
appropriate HLA-matched platelet donor may still be diffi-
cult, and relying solely on HLA matching has certain short- Granulocyte transfusion
comings. For these reasons, there has been interest in adopting
Granulocyte antigen systems
a platelet crossmatching approach similar to that used in RBC
compatibility testing. In the latter approach, a sample of the There are five human neutrophil antigen (HNA) systems
patient’s serum is incubated with aliquots of platelets from that represent polymorphisms on a variety of neutrophil cell
candidate donor units, and those units that manifest the least surface proteins, although expression of HNA-3 (CTL2),

HNA-4 (CD11b), and HNA-5 (CD11a) is not restricted to place by which the physician of the intended granulocyte
neutrophils. HNA-1 (FcgRIIIb, CD16b) appears to be the recipient can authorize transfusion of the product before
most commonly antigenic. FcgRIII is linked to the outer leaf- infectious disease testing has been completed. In these situa-
let of the cell membrane bilayer by a glycosylphosphati- tions, the physician is given the opportunity to weigh the
dylinositol (GPI) anchor. As a result, HNA-1 antigens are potential benefit of granulocyte transfusions with the puta-
poorly expressed on neutrophils in patients with PNH as tive risk of infectious disease transmission by the blood prod-
well as in a proportion of patients with a variety of other uct. Granulocyte donors are typically selected from regular
clonal myeloid disorders, including some patients with apheresis platelet donors who have had documented negative
myeloid leukemia, in which the expression of GPI-linked infectious disease testing within the prior month, or the PBSC
proteins has been reported to be absent or reduced. Com- donor (in matched-related transplants) if applicable. Some
mon properties of HNA systems include their absence on blood centers also may bring the donor in on the day before
early myeloid precursors and the acquisition of expression the granulocyte donation to collect samples for infectious
during neutrophil differentiation. Donor–recipient or fetal– disease testing to ensure that results are available before
maternal mismatches involving the HNA antigens appear to release of the granulocyte product.
be responsible for a significant percentage of reported cases
of neonatal alloimmune neutropenia (NAIN), granulocyte
Clinical transfusion of granulocytes
transfusion refractoriness, transfusion-related acute lung
injury (TRALI; see section “Transfusion-related acute lung Most cases of prolonged marrow aplasia can be treated ade-
injury” this chapter), and delayed neutrophil recovery or sec- quately without granulocyte transfusion. The initial treat-
ondary graft failure following HSCT. ment of patients with neutropenic fever consists of
HNA alloantibodies appear to play an important role in broad-spectrum antibiotics and recombinant growth fac-
some cases of febrile nonhemolytic transfusion reactions tors. Granulocyte transfusions should be considered only in
and TRALI. In one study, more than one-third of patients patients with a realistic expectation of marrow recovery who
undergoing HSCT acquired antibodies directed against neu- have ongoing neutropenia with persistence or progression of
trophils in the posttransplantation period; the presence of bacterial or fungal infection despite appropriate antibiotic
such antibodies was independently correlated with both and antifungal therapy. Although underpowered, the RING
delayed neutrophil engraftment and postengraftment neu- trial on the efficacy of high-dose granulocyte transfusion
tropenia, that is, secondary graft failure. The latter observa- therapy in neutropenic patients with infection did not show
tion is important because such patients often respond to a benefit of granulocytes in neutropenic patients, as com-
steroids or granulocyte colony-stimulating factor (G-CSF) pared to conventional therapy.
and thus may be able to avoid retransplantation. In some Once the decision to use granulocyte transfusions has
patients alloimmunized to neutrophil-specific antigens, been made, an adequate dose should be given. A minimum
transfused granulocytes do not migrate to sites of infection, dose of 2 × 1010-3 × 1010 neutrophils should be given to
which suggests that some neutrophil-specific antibodies can adults. Achieving this dose requires transfusing multiple
interfere with qualitative neutrophil function. units from unstimulated donors or using a collection method
that increases the granulocyte yield from a single donor, such
as pretreatment of the donor with corticosteroids or G-CSF.
Collection and storage of granulocytes
Because of the high volume of contaminating RBCs, ABO-
Approximately 1010 granulocytes can be harvested from a compatible donors need to be used unless effective RBC
healthy donor during a single leukapheresis session. Pretreat- sedimentation is performed. Granulocyte transfusions are
ment with corticosteroids induces neutrophilia in donors, continued, as they are available from donors, until the infec-
increasing the granulocyte yield. Pretreatment of granulocyte tion is controlled; until the patient’s neutrophil count has
donors with small doses of G-CSF significantly increases the increased to >500/µL; or until significant toxicity, particu-
granulocyte yield. Several studies suggest that administering larly pulmonary toxicity, occurs. Patients with alloantibod-
G-CSF to healthy donors does not lead to an increased inci- ies to granulocyte-specific antigens may not achieve a
dence of hematologic disorders. Because of the short half-life satisfactory therapeutic response to granulocyte transfusions
of granulocytes and 24-hour expiration time of the compo- and are at higher risk of pulmonary toxicity. Granulocyte
nent, granulocytes should be harvested, transported, and transfusions should be separated temporally from ampho-
infused into the intended recipient within a matter of hours. tericin administration because case series evidence suggests
This conflicts with the time required for infectious disease that pulmonary toxicity otherwise is increased. Serologic
screening of the donor, which can take 24-48 hours to com- testing for antineutrophil antibodies is not performed rou-
plete. Consequently, some institutions have procedures in tinely, but it is indicated if significant transfusion reactions

Transfusion of plasma products | 315
develop. If antibodies are found, leukocytes from compatible blood loss, or plasma exchange for such indications as
donors may be used. Leukocyte reduction filters obviously thrombotic thrombocytopenic purpura (TTP). These prod-
should not be used with granulocytes. If the potential for ucts (FFP, PF24, and thawed plasma) often are used inter-
CMV transmission is a concern, then granulocytes collected changeably; however, because of the decrease in levels of the
from CMV-seronegative donors should be used. Unlike stem labile factors V and VIII over time, thawed plasma should
cells and donor lymphocyte infusions (DLIs), however, gran- not be used as the sole source of factor replacement in
ulocytes should undergo irradiation. patients who are significantly deficient in either of these fac-
tors. ADAMTS13 (a disintegrin and metalloprotease with
thrombospondin) protease activity is stable up to 5 days of
Key points thawed storage and, as such, thawed plasma can be used for
• Antibodies directed against HNA system antigens can mediate
plasma exchange in TTP. Four-factor prothrombin complex
TRALI, refractoriness to granulocyte transfusions, NAIN, concentrates (PCCs) with adequate FVII content (in addi-
alloimmune neutropenia following HSCT, and qualitative tion to prothrombin, FIX, and FX) increasingly are being
neutrophil dysfunction. used for urgent warfarin reversal, particularly in the setting
• Transfusion of granulocytes can be considered in patients with of intracranial hemorrhage. For hemophilia, optimal man-
severe prolonged neutropenia and antibiotic-refractory agement is with recombinant factors.
infections as a bridge to endogenous granulocyte recovery. Prophylactic plasma transfusions to correct mild prolon-
gations of coagulation values before an invasive procedure
usually are not indicated. RCTs to determine the appropriate
Transfusion of plasma products indications and dosing of plasma therapy have not been
completed in part because of the low baseline bleeding risk
Plasma
associated with minor coagulopathies, making appropriately
Units of plasma most commonly are obtained from units of powered trials prohibitively large. However, when clinically
whole blood donated by volunteer blood donors. The tradi- indicated, plasma is typically dosed at 10-20 mL/kg.
tional nomenclature of fresh frozen plasma (FFP) applies to
plasma frozen within 8 hours of collection and used within
Cryoprecipitate
24 hours of thawing. Other types of plasma commonly used
include plasma frozen within 24 hours of collection and used Cryoprecipitate is prepared by thawing FFP at 4°C and then
within 24 hours of thawing (PF24), and thawed plasma, removing the supernatant from the cryoprecipitable proteins
which is made from FFP or PF24 and kept refrigerated for up following centrifugation at 1°C-6°C. Cryoprecipitate is a
to 5 days after thawing. A standing inventory of thawed concentrated preparation of procoagulant factors, including
plasma is typically available quickly in emergency bleeding fibrinogen, factor VIII, vWF, factor XIII, and fibronectin.
situations in large centers. New viral inactivation methods to Although cryoprecipitate contains a subset of procoagulants,
pathogen reduce plasma have recently been approved. The unlike plasma, it does not contain appreciable quantities of
most common of these techniques uses a solvent detergent physiologic anticoagulants, such as protein C or proteins S.
method that disrupts lipid-containing viruses. Methylene Cryoprecipitate alone is not indicated in patients with dis-
blue is another method of pathogen inactivation, commonly ease processes that deplete both procoagulants and antico-
used in Europe, in addition to the use of ultraviolet-activated agulants, such as disseminated intravascular coagulation or
psoralen derivatives. Psoralen or riboflavin-based UV treat- severe hepatic failure. In theory, cryoprecipitate could be
ment systems are new methods in the US. used to treat von Willebrand disease, hemophilia A, or con-
In theory, plasma could be used to treat acquired or congenital fibrinogen disorders, when recombinant factors or
genital deficiencies of virtually any circulating pro- or anti- virally inactivated factor concentrates are not available.
coagulant factor. It is standard practice, however, to use Cryoprecipitate has also been used to treat qualitative plate-
recombinant or purified pharmaceutical preparations of let dysfunction due to uremia, and life-threatening hemor-
coagulation-related proteins when available and replace- rhage secondary to thrombolytic therapy. The supernatant
ment of a single factor is indicated. Thus, the most common plasma (sometimes referred to as cryosupernatant or cryo-
indications for plasma transfusion therapy include situations poor plasma), which lacks the high-molecular-weight multi-
in which multiple factor deficiencies are present simultane- mers of vWF, can be used in the treatment of TTP but does
ously, such as patients with liver disease, disseminated intra- not appear to be superior to plasma for this indication.
vascular coagulation, vitamin K deficiency (nutritional or Cryoprecipitate is not pathogen inactivated, and a pool of
due to warfarin therapy requiring urgent therapy), dilutional 8–10 units of cryoprecipitate is needed to correct hypofi
coagulopathy of massive transfusion secondary to acute brinogenemia in an adult, resulting in multiple donor

exposures. Fibrinogen concentrate that has undergone viral pools. Overt, acute alloimmune hemolytic anemia can also
inactivation is available in the United States and is indicated develop, especially for blood group A and AB recipients.
when fibrinogen replacement alone is needed; however, Fever is a relatively common sequela of IVIg administration
approval is currently limited to congenital fibrinogen and does not necessarily preclude the administration of
deficiency. additional IVIg.
Immunoglobulin Key points
Commercially available intravenous immunoglobulin (IVIg) • The most common indications for plasma transfusion include
products are typically prepared by cold ethanol fractionation rapid reversal of warfarin effects; treatment of deficiencies of
coagulation factors for which specific coagulation replacement
of large pools of human plasma followed by viral inactiva-
products are not available; and plasma exchange in patients
tion procedures, such as solvent detergent treatment or heat
with TTP.
pasteurization. As is the case with virally inactivated plasma,
• The most common indication for transfusion of cryoprecipitate
the risk of transmission of hepatitis B virus (HBV), hepatitis is hypofibrinogenemia in the context of complex coagulopathy
C virus, or HIV appears to be negligible, although concerns (eg, DIC). Fibrinogen concentrates are available for selective
remain regarding the potential transmission of certain fibrinogen replacement.
difficult-to-inactivate pathogens, such as parvovirus B19 and
prions. There have been reports of acute renal failure occ
urring in association with the administration of IVIg, par- Pretransfusion testing
ticularly in patients with preexisting renal insufficiency,
hypovolemia, diabetes, or other risk factors. Most of the The term pretransfusion testing refers to the series of labora-
immunoglobulin in commercially available preparations of tory tests that blood banks and transfusion services perform
IVIg is IgG itself, and the IgG immunoglobulin subtype dis- to provide immunologically compatible blood products to
tribution (ie, IgG1 through IgG4) is similar to that found in patients. It is important for hematologists to have a general
normal human plasma. Relatively small amounts of IgA and working knowledge of what takes place behind the scenes in
IgM also are present. IVIg has been used to treat a variety of the blood bank between the time when blood is ordered and
hematologic disorders, including congenital immunodefi- when it is received.
ciency syndromes, ITP, autoimmune neutropenia, and
recurrent bacterial infections occurring in association with
ABO/Rh(D) typing
chronic lymphocytic leukemia, multiple myeloma, and other
immune dysregulation conditions. In autoimmune cytope- Determining a patient’s ABO blood group includes two
nias such as ITP, IVIg is considered first-line emergency independent sets of tests that are expected to yield comple-
intervention when a rapid response is required, although the mentary results. In the forward typing, patient RBCs are
effect is transient. mixed with IgM anti-A or anti-B reagent typing sera. Agglu-
The mechanism by which IVIg ameliorates autoantibody tination of cells with either reagent indicates the presence of
destruction of blood cells is unknown. Historically, it has the A or B antigen, respectively, on the patient’s RBCs.
been assumed that the infused IgG blocked Fc receptors on Because of the importance of determining a patient’s ABO
phagocytic cells of the reticuloendothelial system. Numerous blood type with absolute certainty, a second test known as
other theories have been proposed, including autoantibody reverse typing is performed for confirmation. This test takes
neutralization by anti-idiotypic antibodies, cytokine modu- advantage of naturally occurring isohemagglutinins to the A
lation, and complement neutralization. Recent studies have or B antigens that an individual’s RBCs lack. The patient’s
provided experimental evidence that IVIg may serve to cre- serum or plasma is mixed with reagent RBCs known to be
ate soluble immune complexes that interact with activating either blood group A or B and agglutination is assessed. Table
Fcγ-receptor on dendritic cells, which leads to the inhibition 12-3 illustrates the expected forward- and reverse-typing
of macrophage phagocytic activity. Recent studies in murine results for the four possible ABO blood types.
models of ITP have demonstrated that the immunomodula- Discrepancies between the forward- and reverse-typing
tory effects of IVIg can be reproduced by the adoptive trans- reactions occur and can sometimes be explained by evaluat-
fer of CD11c-positive dendritic cells that have been primed ing the patient’s recent transfusion history. For example, a
with IVIg. blood group B individual given type O RBCs in an emer-
A significant proportion of patients receiving IVIg develop gency situation could continue to demonstrate only the
a positive DAT because of the presence of anti-A or anti-B appropriate anti-A antibodies by reverse typing but show a
antibodies derived from type O individuals in the donor mixed field of RBCs, that is, both agglutinated (the patient’s

Pretransfusion testing | 317
Table 12-3 ABO blood group typing reaction results screening consists of performing agglutination screens with
Forward typing Reverse typing patient serum or plasma and 2-3 reagent RBCs whose
extended phenotype (ie, antigenic composition across many
Reaction of patient’s Reaction of patient’s
Patient’s blood groups) is known and includes all major common,
RBCs with: serum with:
ABO clinically significant RBC antigens. If the patient’s serum
type Anti-A Anti-B A1 RBC B RBC does not react with the screening cells, then the patient is
O 0 0 + + said to have a negative antibody screen and ABO-compatible
A + 0 0 + units can be selected for crossmatching.
B 0 + + 0 If the patient’s antibody screen is positive, further testing
AB + + 0 0 is required to determine the specificity of the antibody (or
antibodies) present. To accomplish this, the patient serum is
tested against a larger set of reagent red cells (typically 11-16,
blood group B cells) and unagglutinated (the transfused referred to as an RBC panel). By comparing the resulting
blood group O cells) RBCs upon forward typing with anti-B pattern of reactivity (ie, which cells agglutinate and which do
reagent typing sera. Forward- and reverse-typing discrepan- not) with the phenotype of each of the reacting and nonre-
cies are expected in newborns because isohemagglutinin acting cells, alloantibody specificities can be identified to an
production is delayed for several months while their immune acceptable degree of certainty. All reagent RBCs used as
systems mature. Thus, the ABO type for newborns would be screening cells and panel cells are selected purposely to be
interpreted as blood group AB if determined based solely on blood group O so that the presence of anti-A or anti-B iso-
their reverse-typing reactions. For this reason, only forward hemagglutinins does not affect the results.
typing is performed on newborns. Other settings for forward Based on results of the antibody identification panel,
and reverse ABO typing discrepancies include patients who ABO/Rh-compatible RBC units are selected from inventory;
have undergone ABO-mismatched HSCTs, particularly dur- RBCs from attached segments from each of the units are
ing their transition from one blood type to another. What- tested against reagent antisera to identify antigen-negative
ever the cause, it is important to determine the etiology for units.
ABO typing discrepancies in order to select the appropriate Although agglutination reactions are the end point of all
ABO type blood components for transfusion. currently availably methods for antibody screening, various
Typing for the presence or absence of the Rh(D) antigen on test methods are available, including tube testing, gel-based
RBCs is also an important part of determining a patient’s testing and solid phase testing. Differences in the sensitivity,
blood type. Typing for D does not involve a reverse typing specificity, and interfering substances in the detection of
similar to ABO typing because anti-D is not normally expected clinically insignificant antibodies exist among the available
to be present in the sera of Rh-negative individuals. Please methods. For example, gel and solid phase methods are for-
refer to the section on Rh antigens for discussion of weak D mulated specifically to identify IgG antibodies and not detect
versus partial D phenotypes and their clinical significance. IgM antibodies, which are typically insignificant for transfu-
sion purposes, but may be important to identify in patients
with suspected AIHA, for example. Therefore, if the purpose
Antibody screen and specificity identification
of testing is to evaluate for the presence of a cold agglutinin
In general, a patient who has never been pregnant or trans- or other IgM antibody, consultation with a blood bank phy-
fused is expected to have only the appropriate naturally sician may be required to ensure that the test method that is
occurring isohemagglutinins based on his or her ABO type. used does not ignore such antibodies. The antibody screen,
For such patients, transfusion with blood products selected indirect antiglobulin test (IAT) and indirect Coombs test are
solely on the basis of ABO and RhD compatibility should be all different names for the same test; in other words, if the
acceptable. However, as patient transfusion and pregnancy patient has an available antibody screen result, there is no
histories can be unreliable or not readily obtained, and pos- added utility to ordering a separate IAT or indirect Coombs
sible exposure and sensitization to non-ABO blood group test.
antigens cannot be ruled out with absolute certainty, it is antibody identification may take several hours to several
required to test all patient sera for the possible presence of days to complete, depending on the complexity of the reactiv-
RBC alloantibodies. If any clinically significant alloantibod- ity. Patients with warm autoantibodies or AIHA present a sig-
ies are detected, then ABO-compatible RBCs lacking the cor- nificant challenge to transfusion services because of the
responding antigen(s) must be selected for transfusion. presence of panreactive antibodies (ie, anti-RBC antibodies
Alloantibody exclusion and identification are the most that not only bind to the patient’s own RBCs but also to all
time-consuming of the pretransfusion tests. Antibody other RBCs, including reagent screening and panel cells).

As a result, the presence of additional alloantibodies to

specific blood group antigens may be masked by the autoan- Key points
tibodies. Time-consuming absorption techniques must be • For blood products to be issued to a patient, the patient’s
used in these cases and are discussed in the section “Autoim- ABO/Rh blood type must be determined and the patient’s
mune hemolytic anemia” in this chapter. plasma must be screened for the presence of red cell alloanti-
bodies that may have formed following a previous transfusion,
HSCT, or pregnancy.
Crossmatching • If a patient’s plasma lacks clinically significant red cell
alloantibodies, then an immediate spin or computer crossmatch
There are two basic types of crossmatch procedures, and the
is performed with prospective RBC units to ensure ABO blood
one used depends on the results of the patient’s antibody
group compatibility.
screen. If the antibody screen is negative and the blood bank
• If a patient’s plasma demonstrates the presence of clinically
has historical records indicating no alloantibodies in the significant red cell alloantibodies, then ABO/Rh-compatible RBCs
patient, then a crossmatch between the donor unit and the that lack the corresponding antigen(s) must be identified. These
patient’s blood type to confirm ABO compatibility is per- prospective units must then undergo a full or Coombs cross-
formed. Classically, this is performed as an immediate spin match with the patient’s plasma.
crossmatch, in which the patient’s plasma is mixed at room
temperature with an aliquot of RBCs from the prospective
ABO-compatible unit and the absence of agglutination due to Apheresis
IgM isohemagglutinins is verified. It is now acceptable prac-
Plasmapheresis
tice, however, for blood banks to perform an electronic or
computer crossmatch, in which the laboratory information Common indications for therapeutic apheresis are given in
system runs through an algorithm to ensure that both patient Table 12-4. For a more comprehensive list and discussion
and prospective RBC unit are compatible with regard to ABO using an evidence-based medicine approach, a recent special
and RhD. This type of crossmatch is thus a virtual crossmatch issue of the Journal of Clinical Apheresis is cited in this chap-
because no physical mixing of cells and sera takes place. ter’s bibliography. Plasma exchange typically involves cen-
The other type of crossmatch procedure is known as a full trifugation of whole blood removed from the patient, selective
or Coombs crossmatch. This type of crossmatch is required removal of plasma which is replaced with defined volumes of
for cases in which the patient has a historical or currently replacement fluid (5% albumin, plasma, saline, or various
positive antibody screen, with or without a specifically iden- combinations of these fluids), and return of cellular blood
tified RBC alloantibody. Availability of antigen-negative elements from the extracorporeal circuit to the patient. For a
units varies significantly depending on the specificity of the standard procedure, one plasma volume of patient plasma is
antibody (or antibodies) identified in the patient’s plasma. removed and replaced, typically with 5% albumin. Plasma is
After identifying prospective ABO/Rh-compatible units that used when humoral factors are needed, as in the case of TTP
are negative for the antigen(s) against which the patient has or perioperative plasma exchange. C entrifugal apheresis typi-
alloantibody(ies), a full crossmatch is performed. This con- cally is performed in a continuous-flow fashion so that the
sists of incubating patient plasma with RBCs from the patient remains euvolemic throughout the procedure.
selected RBC units and carrying the testing through from the Frequency and duration of therapy depend on the indica-
immediate spin to the Coombs (IgG or antihuman globulin) tion. Besides clinical trial data, the principles of apheresis
phase to ensure that the selected units are not only ABO that determine a treatment course include the theoretical
compatible but also are compatible at the IgG phase. When efficiency of immunoglobulin removal. Approximately half
the patient’s antibody is reactive in the current sample, the of total IgG is intravascular. Because subsequent procedures
Coombs crossmatch additionally ensures that the units lack are removing plasma that has already had immunoglobulin
the antigens for which the patient’s serum contains pre- removed, the efficiency of subsequent procedures is theo-
formed alloantibodies. retically less than the first one. Allowing time in between
Incompatible crossmatches with multiple or all selected procedures allows for redistribution of IgG back into circula-
RBC units may be seen in a number of situations, most com- tion which increases the available IgG for removal. IgM is
monly in the presence of warm autoantibodies or panagglu- approximately 80% intravascular and is more efficiently
tinins. Understanding the reason for the incompatible removed than IgG. The adverse effects of plasma exchange
crossmatch is critical to determining the risk versus the ben- are primarily driven by complications related to the central
efit of proceeding with transfusion of a crossmatch- venous catheter, if needed; the risk of reactions with plasma
incompatible RBC unit. Consultation with a blood bank replacement; vagal reactions; and reactions to the citrate or
physician is warranted in these situations. heparin used for anticoagulation.

Apheresis | 319
Table 12-4 Abbreviated list of therapeutic apheresis procedures grouped by ASFA indication category
Disease/disorder Procedure
Category 1. Accepted as first-line therapy, stand-alone or adjunctive

Chronic inflammatory demyelinating polyradiculoneuropathy (CIDP) Plasmapheresis
Cryoglobulinemia Plasmapheresis
Cutaneous T-cell lymphoma; mycosis fungoides (erythrodermic) Extracorporeal photopheresis
Familial hypercholesterolemia (homozygotes) Selective absorption
Goodpasture syndrome plasmapheresis
Guillain-Barré syndrome Plasmapheresis
Hyperleukocytosis/leukostasis Leukapheresis
Hyperviscosity in monoclonal gammopathies Plasmapheresis
Myasthenia gravis Plasmapheresis
Sickle cell disease (acute stroke) Red blood cell exchange
TTP Plasmapheresis
ANCA-associated rapidly progressive glomerulonephritis Plasmapheresis
Babesiosis, severe Red blood cell exchange
Antibody-mediated renal transplant rejection Plasmapheresis
Category 2. Accepted second-line therapy, stand-alone or adjunctive
ABO-incompatible hemopoietic progenitor cell transplantation Plasmapheresis
Cold agglutinin disease, life-threatening Plasmapheresis
Familial hypercholesterolemia Plasmapheresis
Familial hypercholesterolemia (heterozygotes) Selective absorption
Graft-versus-host disease (skin) Extracorporeal photopheresis
Catastrophic antiphospholipid syndrome Plasmapheresis
Malaria Red blood cell exchange
Sickle cell disease (stroke prophylaxis, acute chest syndrome) Red blood cell exchange
Category 3. Role of apheresis is not established. Decision making should be individualized
Aplastic anemia; pure red blood cell aplasia Plasmapheresis
Coagulation factor inhibitors Plasmapheresis
Graft-versus-host disease (nonskin) Extracorporeal photopheresis
Hyperleukocytosis/leukostasis (prophylaxis) Leukapheresis
Myeloma and acute renal failure Plasmapheresis
Posttransfusion purpura Plasmapheresis
Warm autoimmune hemolytic anemia Plasmapheresis
Category 4. Evidence indicates apheresis to be ineffective or harmful
Diarrhea-associated HUS Plasmapheresis
Dermatomyositis/polymyositis Plasmapheresis
ITP (refractory) Plasmapheresis
Rheumatoid arthritis Plasmapheresis
SLE nephritis Plasmapheresis
ANCA = antineutrophil cytoplasmic antibody; ASFA = American Society for Apheresis; HUS = hemolytic-uremic syndrome; ITP = immune
thrombocytopenic purpura; SLE = systemic lupus erythematosus; TTP = thrombotic thrombocytopenic purpura.
Extracorporeal photochemotherapy (ECP or photophere- 2 consecutive days every 4 weeks; the median time to res
sis) involves collecting peripheral blood mononuclear cells ponse is 4-6 months. Response correlates with the pres-
by apheresis (processing about one-third of the blood vol- ence of circulating clonal tumor cells and a CD8-mediated
ume), adding a photoactivating agent (8-methoxypsoralen) antitumor response. ECP also is used to treat acute and
into the mononuclear cell suspension, treating the mono- chronic GVHD after allogeneic stem cell transplantation.
nuclear cells with ultraviolet A light, and returning the The best response rates to ECP (~70%) are seen with
treated cells to the patient. The process takes about 2-4 hours. chronic cutaneous GVHD in steroid refractory cases.
ECP is an adjunctive therapy for erythrodermic cutane- LDL apheresis selectively removes LDL from plasma and is
ous T-cell lymphoma; patients typically are treated on used to treat homozygous familial hypercholesterolemia or

the heterozygous carrier refractory to maximal lipid- chemotherapy with growth factors. Although earlier studies
lowering drug therapy. The most common method used in suggested that leukapheresis should commence when the
the United States uses a dextran sulfate column to bind LDL white cell count reaches 1 × 109/L, more recent data suggest
and very low-density lipoprotein, while sparing immuno- that a more optimal time to start collecting is when the white
globulins and high-density lipoprotein (HDL). A typical cell count exceeds 10 × 109/L. Cyclophosphamide followed by
procedure reduces LDL by 50%-75% and is performed every G-CSF is a commonly used protocol. The white cell count
2 weeks. Some experts believe that homozygous children reaches 1 × 109-10 × 109/L around day 11-13 after chemo-
should start LDL apheresis around age 7 years to prevent therapy. Leukapheresis usually is scheduled for day 10-12
premature atherosclerosis. after chemotherapy. A mobilization regimen that has a pre-
dictable rebound phase allows for more efficient use of
apheresis and stem cell processing staff. The use of growth
RBC Exchange transfusion
factor alone for mobilization avoids the risk of neutropenia
RBC exchange transfusion therapy is performed most often with chemotherapy and is used in allogeneic donors. G-CSF
in patients with sickle cell disease to prevent and treat acute at 10 mg/kg/d has been the mobilization regimen of choice for
complications of the disease, such as stroke and acute chest allogeneic PBSC donors. With this regimen, leukapheresis
syndrome. Preoperative exchange transfusion is indicated begins on day 5, when the white cell count is 20-50 × 109/L.
prior to extensive surgery or when the preoperative hemo- The correlation is excellent between the number of CD34+
globin is too high to permit simple transfusion. Patients cells in the peripheral blood on the day of leukapheresis (or
should be transfused with blood that is known to be negative the preceding day) and the number of CD34+ cells that can
for hemoglobin S. During RBC exchange utilizing an aphae- be collected by apheresis. For instance, for a target collection
resis machine, the patient’s erythrocytes are removed and of 2 × 106/kg CD34+ cells, the preceding day CD34+ cell
replaced with donor erythrocytes while the patient’s own count in the peripheral blood should exceed 20 × 106/L. If
plasma is continually returned to minimize disturbance of collections are planned to take place over a number of con-
hemodynamic and coagulation parameters, although plate- secutive days, many centers begin collections when the
lets are removed during erythrocytapheresis. Another option peripheral CD34+ cell count is lower (eg, 10 × 106/L).
is manual RBC exchange, in which manual phlebotomy is Although the administration of mobilizing doses of
followed by infusion of donor RBCs. This is performed par- G-CSF can induce seemingly worrisome degrees of
ticularly in small children who cannot tolerate the volume leukocytosis—transient peripheral blood leukocyte counts
shifts associated with apheresis and/or those do not have the of 80,000/µL or higher are not uncommon—follow-up stud-
required vascular access to be exchanged chronically using ies reported to date suggest that administration of short
an apheresis machine. When manual exchange is done, care- courses of G-CSF to healthy donors is not associated with
ful attention must be paid to the potential for volume deple- any adverse long-term consequences. A rare complication of
tion. The goal of exchange transfusion in most situations, G-CSF mobilization is splenic rupture, which has been
such as acute chest syndrome, stroke, or preoperative reported in healthy adult PBSC donors, most commonly
exchange, whether performed manually or via automated after five daily doses of G-CSF.
RBC apheresis, is to achieve a hematocrit of 30% with a Large-volume leukapheresis (LVL) refers to the processing
hemoglobin S of ≤30%. of large volumes of blood (15-30 L over 5 hours); data sug-
gest that committed progenitor cells are recruited into the
circulation during LVL. Although the magnitude of recruit-
PBSC harvesting
ment from LVL is small relative to the effects of chemother-
Mobilization refers to the technique of increasing the number apy and growth factor mobilization, the two techniques can
of circulating progenitor cells in the peripheral blood. It was be combined for maximal benefit. LVL requires good venous
noted in the 1970s that progenitor cells in peripheral blood access that would permit sufficient flow rates. This may
increased up to 20-fold after chemotherapy for ovarian can- necessitate a central venous catheter. To minimize the risks
cer. The introduction of hematopoietic growth factors in the of citrate toxicity, heparin may be added to the citrate; cal-
late 1980s shortened the period of neutropenia after chemo- cium supplementation is an alternative to heparin use. Plate-
therapy and was noted to increase circulating hematopoietic let depletion is another predictable consequence of LVL,
progenitors up to 1,000-fold. G-CSF downregulates the although newer cell separators may be able to collect PBSCs
expression of adhesion molecules on the surface of HPSCs, with less platelet loss.
progenitor cells, precursor cells, and mature neutrophils and A relatively common problem with PBSC harvesting is
mobilizes clinically significant numbers of HPSC into the inadequate collection. Although multiple definitions of
peripheral blood. Many mobilization regimens combine inadequate collection have been used, it is clear that the

Transfusion support in special clinical settings | 321
incidence of inadequate collection is much higher in heavily Transfusion support in patients with aplastic anemia mer-
pretreated patients than in healthy donors. In healthy PBSC its special discussion because several studies have supported a
donors, increasing age, white ethnicity, and female sex were strong association between the number of blood product
associated with lower post–G-CSF peripheral blood CD34+ exposures before allogeneic HSCT and the risk of the often-
counts, which correlate with lower CD34+ yields from col- fatal complication of graft rejection in such patients. It is
lection. Risk factors for an inadequate autologous collection believed that recipient-derived T-cells directed against a vari-
include multiple prior chemotherapeutic regimens, exten- ety of hematopoietic targets contribute to the genesis of the
sive prior radiation therapy, or administration of certain aplastic anemia itself and also may be responsible for the par-
chemotherapeutic agents, such as fludarabine, lenalidomide, ticularly high incidence of graft rejection that has been
melphalan, chlorambucil, and nitrosoureas. observed among patients undergoing HSCT for aplastic ane-
Plerixafor is a small-molecule reversible inhibitor of the mia. Administration of irradiated, leukoreduced blood prod-
chemokine receptor CXCR4 on stem cells; this inhibition ucts may help reduce the risk of graft rejection; however,
facilitates HPSC egress from the bone marrow and is syner- serious attention must be paid to minimizing the number of
gistic with the mobilizing effects of G-CSF. One dose of blood product exposures in patients with aplastic anemia.
plerixafor given with G-CSF has been shown to successfully Menstrual suppression in young women can be helpful. RBCs
mobilize CD34+ cells in patients with multiple myeloma, should be transfused only for significantly symptomatic ane-
Hodgkin disease, and non-Hodgkin lymphoma who failed mia. Single-donor (apheresis-derived) platelet products are
previous mobilization attempts; plerixafor as a single agent preferred to pooled random-donor platelets when possible.
is being studied in healthy PBSC donors. The adverse effect antifibrinolytics are a useful adjunct in thrombocytopenia
profile of plerixafor (mostly gastrointestinal) does not patients for prophylaxis and treatment of hemorrhage.
appear to overlap with that of G-CSF.
Hematopoietic stem cell infusion
Key points
In the setting of allogeneic HSCT, PBSCs and bone marrow
• Apheresis selectively removes plasma, erythrocytes, or typically are infused “fresh,” without cryopreservation. RBC or
leukocytes for therapeutic benefit in a variety of hematologic plasma depletion of the PBSC component may be required if
diseases (eg, TTP, sickle cell disease, an HPSC collection). the donor is major or minor ABO incompatible, respectively.
• A variety of nonhematologic, antibody-mediated disorders can Autologous PBSCs or bone marrow are nearly always cryopre-
be successfully treated with apheresis, to remove the causative
served before use because most transplantation preparative
antibodies, including Goodpasture syndrome, Guillain-Barré
regimens require at least several days to administer before the
syndrome, and humoral rejection in organ transplantation.
stem cells can be infused. Optimal viability of the stem cells is
achieved by controlled-rate freezing, using dimethyl sulfoxide
(DMSO) as the cryopreservative. PBSCs typically are stored in
Transfusion support in special clinical the vapor phase of liquid nitrogen. Frozen aliquots of 50-75
settings mL are thawed sequentially during the infusion, at the bed-
side, or in the laboratory. This approach allows the mainte-
Patients who are candidates for HSCT
nance of a relatively slow infusion rate while simultaneously
As autologous and nonmyeloablative HSCTs are being offered maximizing PBSC viability by minimizing the interval
to a wider population of patients with hematologic malignan- between thawing and infusion of each aliquot. DMSO toxicity
cies, clinicians must take into account the possibility that commonly manifests as flushing, nausea, vomiting, and blood
many patients newly diagnosed with hematopoietic malig- pressure fluctuations; to minimize toxicity, the volume of
nancies are likely to become potential candidates for HSCT at DMSO infused should be limited to no more than 1 mL/kg at
some point in their clinical course. Therefore, it is important one sitting (which translates to 10 mL/kg of PBSCs for com-
to avoid administering transfusion products obtained from ponents that were cryopreserved with 10% DMSO).
family members because they may increase the risk of graft
rejection via alloimmunization to minor histocompatibility
Key points
antigens. For newly diagnosed patients with acute leukemia, it
is useful to perform HLA typing earlier in the course of induc- • Blood transfusion prior to HSCT can hinder engraftment and
tion therapy and to anticipate problems in platelet support in lead to graft failure via alloimmunization.
patients who are at risk of HLA alloimmunization, such as • For optimal cell viability, frozen aliquots of HPSCs must be
multiparous females. HLA typing results obtained upfront thawed rapidly and infused into the patient without delay,
sometimes leading to DMSO toxicity.
will also be useful for potential allogeneic HSCT.

RBCs, platelets, and granulocytes must be irradiated

Clinical case before transfusion in HSCT recipients to prevent TA-GVHD.
A 56-year-old woman is being evaluated for matched HSCT from Some institutions with high-volume oncology and HSCT
her brother, for high-risk acute myeloid leukemia in first remis- caseloads have elected to irradiate all platelets and RBCs to
sion. She is A positive, and he is O negative. She is enrolled in a avoid the disastrous consequence of omitting this step as
nonmyeloablative conditioning protocol. On day 0, the PBSC is irradiated components are safe for patients who are not at
plasma depleted and infused without incident. On day 8, she is risk of TA-GVHD. Irradiation shortens the shelf life of RBCs
noted to have a hemoglobin of 6 g/dL (down from 9 g/dL the (but not platelets), necessitating attention to inventory man-
day before). She is asymptomatic and without any evidence of
agement. Most centers recommend that HSCT survivors
bleeding.
receive irradiated blood components indefinitely, in the
absence of data that show the safety of non-irradiated com-
ponents in long-term HSCT survivors.
Transfusion support after HSCT
Prevention of CMV infection is an important part of
The intensity of transfusion support varies among condi- transfusion management in CMV-seronegative HSCT
tioning regimens. In general, the transfusion needs are less in recipients. Leukoreduction filters achieve a 3- to 4-log
autologous transplantation and nonmyeloablative allogeneic reduction of leukocytes in blood products. A landmark ran-
conditioning regimens compared with allogeneic myeloab- domized comparison of leukoreduced versus CMV-
lative regimens. In a hemodynamically stable patient with- seronegative blood components in CMV-seronegative
out underlying cardiovascular disease, it is common practice HSCT recipients (with seronegative donors) found no sig-
to transfuse RBCs for a hemoglobin of 7-8 g/dL, although nificant difference in the incidence of CMV infection and
there has been no randomized clinical trial conducted to CMV disease as a composite outcome. With the improve-
determine the optimal RBC transfusion threshold in this ment in CMV surveillance followed by preemptive antiviral
patient population. The landmark studies that support a therapy, which has further reduced the incidence of CMV
platelet transfusion threshold of 10 × 109/L were conducted disease after HSCT, this question will never be answered
in patients undergoing leukemia induction. Risk factors for definitively. Most transplantation centers, in practice, will
platelet refractoriness such as fever, infection, bleeding, use prestorage leukoreduced blood components for CMV
amphotericin, and vancomycin are common occurrences in prevention.
HSCT patients. Veno-occlusive disease increases platelet
consumption from endothelial damage and activation of
ABO-incompatible HSCTs
vWF; portal hypertension and hypersplenism further
increase platelet transfusion requirements. Allogeneic HSCTs do not require ABO matching because
Despite new antifungal agents, fungal infections in patients ABO antigens are not expressed on pluripotent stem cells.
who have prolonged neutropenia remain problematic. There Because the genes for HLA are encoded on chromosome 6
are case series of patients who received granulocyte transfu- and the genes for ABO are on chromosome 9, two siblings
sions as adjunctive therapy for refractory fungal (and bacte- can have an identical HLA type but different ABO types. A
rial) infections after HSCT and as secondary prophylaxis report compiled from multicenter data reported to the Inter-
during HSCT after a prior episode of fungal infection. The national Blood and Marrow Transplant Research group
use of granulocyte transfusions as primary prophylaxis after included 3,000 patients with early stage leukemia who
allogeneic HSCT produced a modest decrease in febrile days underwent transplantation between 1990 and 1998 with
and antibiotic usage but no difference in treatment-related bone marrow from an HLA-identical sibling donor. There
mortality in one study. Patients known to be HLA alloimmu- was no difference in overall survival, transplantation-related
nized are at risk of greater pulmonary toxicity from granulo- mortality, and grades 2-4 acute GVHD in the ABO-identical
cyte transfusions, although routine screening for HLA or ABO-mismatched groups. A single-institution study that
antibodies before granulocyte transfusions is not universal. focused exclusively on nonmyeloablative regimens, however,
As noted in the “Granulocyte transfusion” section earlier in found that ABO incompatibility was associated with
the chapter, the initial report from the most recent RCT of increased nonrelapse mortality within the first year after
granulocytes (RING study) does not demonstrate a benefit HSCT. Similarly, the Japanese Marrow Donor Program has
for granulocyte transfusion, although the trial was under- reported increased acute GVHD in ABO-mismatched unre-
powered. Based on available data to date, routine use of gran- lated donor transplantations and increased transplantation-
ulocytes is not warranted given the minimal advantages to the related mortality in the subset that received nonmyeloablative
recipients and potential risks of subjecting healthy donors to conditioning. In the unrelated donor setting, there may be
G-CSF and corticosteroids. multiple potential HLA matches for any given patient, and in

light of these findings, ABO compatibility is a secondary recipient lymphocytes may resume production of RBC allo-
consideration (in addition to other such factors as donor sex, antibodies with donor specificity. It is important to con-
age, CMV status) that ultimately dictates donor choice. sider alloimmune hemolysis in the differential diagnosis of
A major ABO mismatch occurs when the recipient’s hyperbilirubinemia in the posttransplantation patient.
plasma contains antibodies against the donor’s RBCs. An Finally, HSCT patients whose disease relapses may revert to
example would be a group O recipient–group A donor pair. the recipient ABO/Rh type. The transfusion service must be
In such cases, RBC depletion of the donor marrow is usu- alert to subtle changes in mixed-field agglutination in
ally performed; apheresis collections typically do not ABO blood grouping during these situations. Table 12-5
require RBC depletion. Major ABO-mismatched HSCT provides useful guidelines for the selection of the appropri-
recipients have a slightly slower neutrophil recovery; how- ate blood group type for RBCs, platelets, and plasma
ever, the main concern in these cases is red cell engraft- for various scenarios of donor–recipient ABO-HSCT
ment. Mature erythroid progenitors do express ABO incompatibility.
antigens, and immune-mediated delayed erythropoiesis
with reticulocytopenia can occur, leading to prolonged Key points
transfusion dependence up to 1 year after transplantation.
The incidence is approximately 10%, and there is an inverse • RBC, platelet, and granulocyte products administered to HSCT
correlation between ABO isohemagglutinin titers and retic- recipients must be irradiated to minimize the risk of potentially
fatal TA-GVHD and leukoreduced to minimize the risks of CMV
ulocyte counts.
transmission and alloimmunization to HLA antigens.
A minor ABO mismatch occurs when the donor’s plasma
• Donor–recipient mismatches involving the ABO system usually
contains antibodies against the patient’s RBCs (group A
are well tolerated but occasionally can cause delayed alloim-
recipient–group O donor, such as in the illustrative case). mune hemolytic anemia or pure red cell aplasia.
Plasma depletion of the HPSC product is performed by
some centers, but acute hemolysis from donor plasma is
uncommon. A more important concern is the phenomenon
Pediatric transfusion issues
of immune hemolysis from mature, competent passenger
Hemolytic disease of the fetus and newborn (HDFN)
lymphocytes transfused with the HPSC component, espe-
cially with T-cell-depleted marrows, PBSCs versus marrow, HDFN (or erythroblastosis fetalis) is most commonly due to
the use of cyclosporine alone (without methotrexate) for maternal–fetal mismatches involving Rh or ABO antigens,
GVHD prophylaxis, and reduced-intensity conditioning which can cause an antigen-negative mother to mount an
regimens. Hemolysis of the residual recipient RBCs in circu- antibody response against the antigen-positive fetal RBCs.
lation due to passenger lymphocyte syndrome can occur
abruptly, from days 7-14 after HSCT, and can be severe or
even fatal. Some centers perform periodic DAT screening in Table 12-5 Peri-transplant guidelines for blood component selection
minor ABO–mismatched HSCT recipients, although the in ABO-incompatible HSCT
effectiveness of this strategy is not clear. It may be prudent to Recipient Donor RBC Platelet/plasma
maintain a higher transfusion threshold in minor-mismatch blood type blood type transfusion transfusion
recipients during the at-risk period after transplantation.
O A O A or AB
Massive hemolysis may be treated by erythrocyte exchange
O B O B or AB
transfusion using RBCs compatible with both donor and
O AB O AB
recipient types.
A B O AB
HSCT recipients with non-ABO RBC antibodies such A AB A or O AB
as anti-D have undergone transplantation with antigen- A O O A or AB
positive grafts using the same principle of RBC depletion if B A O AB
marrow is the HPSC component. An Rh(D)-positive recip- B AB B or O AB
ient who undergoes HSCT from an Rh-negative donor may B O O B or AB
develop anti-D as the donor lymphocytes respond to the AB A A or O AB
residual Rh-positive RBCs. Patients with sickle cell disease AB B B or O AB
undergoing HSCT may present a challenge if they have AB O O AB
developed multiple RBC alloantibodies or antibody to a Rh neg Rh pos Rh neg Rh pos or Rh neg
high-incidence antigen. The optimal time to discontinue Rh pos Rh neg Rh neg Rh pos or Rh neg
antigen-negative blood is not known, but one strategy is to HSCT = hematopoietic stem cell transplantation; neg = negative;
wait until lymphocyte type is 100% donor because residual pos = positive; RBC = red blood cell.

Maternal IgG antibodies can then cross the placenta and Neonatal exchange transfusion
cause passively-acquired immune-mediated hemolytic ane-
Advances in phototherapy and antenatal care have made
mia in the fetus, potentially leading to profound anemia and
exchange transfusion for HDFN, or hyperbilirubinemia due to
hydrops fetalis with fetal demise in severe cases. Maternal
other causes, an uncommon occurrence. Appropriate unit
antibodies formed in response to sensitization by red cell
selection follows the same principles for IUT described previ-
transfusion may also behave in a similar fashion. The inci-
ously (ie, fresh O-negative unit, negative for any offending anti-
dence of severe HDFN has been reduced dramatically with
gen, crossmatched against maternal serum, leukoreduced,
the use of antenatal and peripartum administration of anti-
irradiated, and hemoglobin S negative). In addition, the RBCs
D to Rh(D)-negative mothers, which abrogates the maternal
are concentrated and reconstituted with group AB plasma, typi-
immune response to primary exposure to the D antigen. As
cally in a 1:1 ratio to produce a unit of reconstituted “whole
a result, most cases of HDFN now are attributed to other
blood” (hematocrit 50%) for the exchange. A double-volume
RBC antigens, including Rh antigens other than D, as well as
exchange removes approximately 85% of the neonate’s antigen-
K (Kell blood group system) and ABO. Although severe
positive RBCs but is less efficient in lowering plasma bilirubin.
examples may rarely be seen, ABO HDFN is usually charac-
Complications of exchange transfusion include hypocalcemia,
terized by hyperbilirubinemia with mild anemia (if any); the
dilutional thrombocytopenia, and catheter-related complica-
mother is typically group O with IgG anti-A,B alloantibodies
tions such as thrombosis, infection, or bleeding.
(an antibody with cross-reactivity to both A and B antigens),
and the infant is typically group A.
Alloimmune cytopenias in the fetus or newborn
Intrauterine transfusion
Analogous to HDFN, maternal–fetal mismatches involving
Due to wide-spread use of Rh immune globulin for prophy- platelet-specific or neutrophil-specific antigen systems may
laxis against anti-D antibody formation in association with result in neonatal alloimmune thrombocytopenia (NAIT) or
pregnancy, intrauterine transfusion (IUT) is much less com- neonatal alloimmune neutropenia (NAIN), respectively. The
monly needed and technical expertise is currently concen- target antigens are quite diverse but are often membrane gly-
trated in centers that specialize in high-risk obstetrics. In a coproteins. The most common antibody specificity in NAIT
sensitized pregnancy, Doppler ultrasound and amniotic in whites targets HPA-1a (PLA1), which resides on the plate-
fluid studies guide the need for fetal blood sampling, which let fibrinogen receptor GPIIb/IIIa, although numerous other
may be performed after 20 weeks of gestation. Blood is pre- platelet antibody specificities have been reported. It is worth
pared for IUT if the fetal hematocrit is <25%-30%. Group O, noting that while HDFN due to maternal sensitization
D-negative RBCs lacking the implicated RBC antigen are through pregnancy typically occurs with the second preg-
selected; some centers match the extended maternal RBC nancy, NAIT may occur during a first pregnancy. NAIN
phenotype beyond the implicated antigen. Maternal serum often is due to fetal-maternal mismatches involving the
or plasma is used for crossmatching. A fresh RBC unit neutrophil-specific NA-1/NA-2 system. No prophylactic

(≤5 days old) is typically used, either citrate-phosphate- therapies currently are available for NAIT or NAIN.
dextrose-adenine (CPD-A) unit (without additive solution) Management of these disorders often includes antenatal
or an additive solution unit with the supernatant removed. maternal IVIg to reduce antibody levels, decrease placental
The unit must be γ-irradiated to prevent TA-GVHD because transfer of antibodies, and reduce cellular destruction in the
the fetal immune system is immature; leukoreduction or a fetus. Transfusion support of NAIT is initiated with random
CMV-seronegative donor is used to provide a CMV-safe donor platelets, which produce an adequate platelet incre-
component. The unit should be negative for sickle hemoglo- ment in the majority of cases, however short lived, in the
bin. The selected RBC unit is usually washed and concen- setting of life-threatening bleeding such as intraventricular
trated to the volume and hematocrit specified by the hemorrhage, which can occur in utero. IVIg is a therapeutic
obstetrician performing the procedure. Once IUT is initi- option if the bleeding is mild to moderate. If subsequent
ated, it is repeated every 3-4 weeks until 35 weeks gestation platelet transfusion is needed, washed irradiated maternal
to maintain fetal hematocrit at approximately 25%. Neo- platelets can be obtained. Maternal platelets are negative for
nates who have undergone IUT will type as O negative; such the target antigen in question; their use abrogates the need to
neonates may have suppressed erythropoiesis and/or persis- wait for the time-consuming identification of platelet allo-
tent maternal antibody, which necessitate postnatal transfu- antibody specificity. Some blood centers have registries of
sion support for up to 3 months. Complications of IUT are specific platelet antigen–negative donors available for plate-
related primarily to the technical complexity of vascular let donation if the mother is unable to donate or obtaining
access. maternal platelets is a logistical challenge.

Maternal ITP or autoimmune neutropenia can cause pas- irradiation criteria on gestational age and birth weight. In
sively acquired immune thrombocytopenia or immune neu- addition, leukoreduced cellular components are used to
tropenia in the fetus, respectively. It is important to screen reduce the risk of CMV transmission. Some centers may
for these disorders in the mother before using the mother as use CMV-seronegative components for specific subgroups,
a source for platelets or neutrophils. The currently available such as neonates weighing <1,200 g. The quantity of
assays for antiplatelet antibodies and antineutrophil anti- additives in stored RBCs, such as citrate, adenine, and
bodies are not highly sensitive or specific and the diagnosis mannitol, is far less than levels believed to be toxic.
or exclusion of immune-mediated thrombocytopenia or Washing to reduce the potassium load is not indicated in
neutropenia should not be based solely on the results of anti- small-volume transfusions; however, use of fresh RBCs or
body assays. washed RBCs may be necessary for this purpose with large
volume transfusion. RBCs for neonates requiring large-
volume transfusion should be irradiated as close as
Red blood cell transfusion in preterm neonates
possible to the time that they will be transfused to avoid
The physiologic anemia of infancy occurs at 8-12 weeks, and significant increases in extra-cellular potassium levels.
the nadir hemoglobin is rarely <9 g/dL. Among preterm Although 2,3-DPG is depleted in stored RBCs, it is rapidly
infants, this decline occurs at an earlier age, and the nadir is regenerated after transfusion; infants given stored RBCs
7-8 g/dL; the physiologic nadir is compounded by iatrogenic have stable 2,3-DPG levels after small-volume RBC
phlebotomy. The blood loss through cumulative phlebot- transfusions. In the ARIPI double-blind RCT, use of fresh
omy in a preterm infant’s first weeks of life commonly RBCs compared with standard blood bank practice did
exceeds the entire blood volume. Delaying umbilical cord not improve outcomes in premature, very-low birth
clamping for 30-60 seconds for infants who do not require weight infants requiring transfusion.
immediate resuscitation has been advocated by some to be
the first step in counteracting the anemia of prematurity.
Other component therapy in neonates
Limiting phlebotomy is a crucial part of minimizing trans
fusion in a preterm infant. Erythropoietin has limited effi- Newborns may require plasma transfusion, most commonly
cacy at best and appears to increase the risk of retinopathy of for disseminated intravascular coagulation secondary to sep-
prematurity. There should be a program to limit donor sis; 10-15 mL/kg produces a 15%-20% increase in factor
exposure in neonatal transfusions. Typically, a fresh a dditive- level, assuming ideal recovery. If cryoprecipitate is required
solution O-negative RBC unit (≤7 days old) is dedicated to for persistent hypofibrinogenemia despite plasma transfu-
one or two preterm infants and is used exclusively for all sion, a dose of 1 unit should produce a 100 mg/dL increase in
transfusions for those one or two infants for up to 42 days of fibrinogen (in older infants, the cryoprecipitate dose is 1 unit
allowable storage. per 5-10 kg of body weight).
Two randomized clinical trials of restrictive versus lib- Neonatal thrombocytopenia is common in preterm neo-
eral transfusion criteria used transfusion thresholds that nates, occurring in 22% of infants in one series. It is frequently
varied with patients’ respiratory and other status and post- a sign of sepsis, severe inflammation, and perinatal asphyxia/
natal age. A stable older infant in the restrictive arm, for placental insufficiency. Prophylactic transfusions often are
instance, would be transfused at a hemoglobin level of recommended in neonates with platelet counts <20,000-
approximately 7.5 g/dL; a younger mechanically ventilated 30,000/µL if otherwise stable; in unstable neonates or those
preterm infant would be transfused at a hemoglobin level requiring invasive procedures, platelets are often transfused to
of approximately 11.5 g/dL. Significantly, in both trials, the maintain a count of ≥50,000/µL. The usual platelet dose in
number of donor exposures from RBC transfusions alone neonates is 10-15 mL/kg, which is one whole-blood derived
was not reduced by restrictive transfusion criteria, presum- platelet concentrate (allowing for tubing volume). Similar to
ably reflecting the efficacy of using dedicated donor units; RBCs, infants requiring significant platelet transfusions may
only one of the two trials demonstrated that a restrictive also receive aliquots from a dedicated apheresis platelet unit to
transfusion threshold increased the percentage of infants reduce donor exposures, although the shelf-life of such units is
who avoided transfusion altogether (from 5%-11%). An quite short. Platelets should be ABO identical to avoid the
ongoing RCT is designed to evaluate potential differences transfusion of minor incompatible plasma into the small
in neurologic outcomes between restrictive and liberal blood volume of a neonate. If ABO-identical (or group AB)
transfusion strategies. platelets are not available, platelets can be washed to remove
Most US centers routinely irradiate all cellular incompatible plasma. Routine washing or volume reduction
components for neonates for a variable period of time of platelets is not necessary or recommended because the pro-
after birth (typically 4-6 months). Other centers base cedure can jeopardize platelet quality.

Key points Clinical case (continued)

• The immune system in the fetus and in neonates up to the of breath, a red cell transfusion is ordered. Multiple red cell
age of 4 months is immature and typically not capable of crossmatches are incompatible. Two units of crossmatch-
generating antibody responses to transfusions. Thus, the most incompatible leukoreduced red cells are transfused. The periph-
crucial compatibility issues involve the passive transfer of eral blood hemoglobin concentration increases to 8 g/dL, and
antibodies from the mother to the fetus, as well as maintaining she experiences no untoward reactions.
ABO compatibility between the donor and the fetus or neonate.
• Current blood banking practice attempts to limit the number
Transfusion in patients with AIHA can be challenging.
of donor exposures to fetal and neonatal patients by using
Autoantibodies to RBCs can result in multiple incompatible
multiple transfusion aliquots from single blood products.
RBC crossmatches, which may lead blood banks to inform
clinicians that no compatible RBC units are available. FDA
Pediatric transfusion beyond the neonatal regulations require the patient’s physician to provide written
period consent to release incompatible units, which makes many cli-
nicians uncomfortable. If the patient has not been previously
The posttransfusion long-term survival rate in pediatric transfused or pregnant, alloantibodies to non-ABO antigens
transfusion recipients is much higher than in adults, so the are unlikely to be present, and patients can usually be trans-
principle of minimizing donor exposure, which carries risks fused safely with ABO-compatible blood. Even in patients
of transfusion-transmitted disease (involving known and who have been previously transfused or pregnant, withhold-
unknown infectious agents), continues beyond the neonatal ing transfusions because the crossmatches are incompatible
period. A multicenter trial of restrictive versus liberal trans- may preclude the administration of lifesaving transfusions.
fusion thresholds (7 g/dL vs. 9.5 g/dL) in pediatric ICUs Several studies have examined the incidence of clinically
found that a restrictive transfusion strategy was noninferior significant alloantibodies in multiply transfused patients
in the primary outcomes (28-day mortality and new or pro- with AIHA and found that up to 30%-35% of such patients
gressive multiorgan dysfunction) and successfully avoided develop alloantibodies. Thus, if a patient has received a
transfusion in 54% of patients (compared with 2% in the transfusion or been pregnant, the transfusion service must
liberal transfusion group). The older child or adolescent perform specific testing to determine whether alloantibodies
undergoing elective surgery may benefit from judicious use are present concurrently with the panagglutinating autoanti-
of autologous blood donation and intraoperative cell salvage bodies typically associated with AIHA. The term panaggluti-
in an integrated blood-conservation approach. Pediatric nating refers to the fact that most autoantibodies that cause
patients with sickle cell disease and thalassemia benefit from AIHA will agglutinate most RBCs, including reagent red cells
leukoreduced blood transfusion to reduce HLA alloimmuni- and RBC units for transfusion, because the antigenic target is
zation and febrile nonhemolytic transfusion reactions; some typically a “public” antigen; that is, an antigen present on the
evidence suggests that development of HLA antibodies is RBCs of a large proportion of the population. The public
associated with the development of RBC antibodies in antigen is often a common epitope on the Rh protein.
patients with sickle cell disease. Such patients also benefit Some transfusion services routinely perform extended
from the use of extended RBC phenotypic matching (for C, RBC phenotyping—that is, the patient’s RBCs are typed
c, E, e antigens within the Rh system and Kell) to reduce allo- with regard to antigen systems in addition to ABO and
immunization. Delayed hemolytic transfusion reactions Rh(D), at the time that a diagnosis of AIHA is first rendered,
from RBC alloantibodies can precipitate vaso-occlusive cri- to facilitate the identification of alloantibodies that may
ses, hyperhemolysis, or autoantibody development. appear subsequently. DNA-based methods are preferable,
when available, due to the potential interference of a positive
Autoimmune hemolytic anemia DAT with serologic typing, and the additional information
provided by these methods. In rare situations, when the
presence of underlying alloantibodies cannot be excluded,
Clinical case
transfusion of RBC units phenotypically similar to the
A 69 year-old woman presents with a Hb of 6 g/dL. The DAT is patient’s own RBCs may help reduce the risk of hemolysis
positive for IgG and negative for complement, indicating that due to alloantibodies in this setting.
circulating RBCs are coated with IgG. Her reticulocyte count The technique for detecting alloantibodies in the presence
is <1%. She has never been transfused and has never been of autoantibodies is called adsorption. With the autoadsorp-
pregnant. The patient is started on prednisone for treatment
tion technique, an aliquot of the patient’s plasma is adsorbed
of presumed warm (IgG-mediated) AIHA. Because of shortness
repeatedly with the patient’s own RBCs and then tested for

alloreactivity with a panel of donor RBCs in a standard anti- such tests can give the clinician a sense of the potential clini-
body screen. If the patient has undergone transfusion cal significance of the autoantibodies in vivo at body
recently, autoadsorption cannot be reliably interpreted temperature.
because the transfused RBCs present in the patient’s circula- In the clinical case described above, the patient’s reticulo-
tion could adsorb the very same alloantibodies that the labo- cyte count was low. A substantial minority of patients will
ratory is attempting to detect. In this situation, a method manifest at least transient reticulocytopenia early in the
called differential alloadsorption is used. Differential alload- course of AIHA, a phenomenon that may be due to the fact
sorption, sometimes called triple adsorption, involves that autoantibody titers may increase more quickly than the
adsorbing aliquots of patient serum against RBCs of defined bone marrow can generate a reticulocyte response or due to
phenotypes to produce several adsorbed sera that give dif- rapid destruction of reticulocytes by the autoantibody.
ferential reactivity in standard antibody screens. The differ- Reticulocytopenia with brisk AIHA is an emergent situation
ential reactivity results from the fact that alloantibodies are and transfusion should not be delayed.
left behind in the serum following the adsorption if the
adsorbing cells are negative for the antigen in question. Key points
Because most warm-reacting autoantibodies react with
RBC-surface determinants that do not vary among individu- • Red cell transfusions in patients with life-threatening AIHA
als (ie, public antigens), adsorption with RBCs of different should not be withheld simply because all available units are
crossmatch incompatible. Consulting with a transfusion
phenotypes removes the autoantibody but, depending on the
medicine physician may be helpful for assessment of transfusion
phenotype, either removes or fails to remove alloantibody.
risks in patients with complex serologic work-ups.
For example, if the patient’s serum contains an anti-Jka anti-
• Special blood bank techniques are available to minimize the
body along with an autoantibody, both the autoantibody and risk of transfusion in patients with AIHA.
the anti-Jka antibody will be adsorbed by Jka-positive adsorb-
ing cells, but only the autoantibody will be adsorbed by Jka-
negative adsorbing cells. The presence of the anti-Jka in the
Autoimmune and consumptive
patient’s serum then can be deduced by demonstrating that
thrombocytopenias
the aliquot of the serum that was adsorbed by Jka-positive
cells is nonreactive in a standard antibody screen, whereas Transfusion of platelets in patients with immune-mediated
the aliquot of serum that was adsorbed by Jka-negative cells thrombocytopenias is problematic because of the destruc-
reacts only with Jka-positive cells in a standard antibody tion of transfused platelets by the antibody. As is the case
screen. with AIHA, the autoantibody in ITP often reacts with public
Warm-reacting autoantibodies occasionally demonstrate antigens. Transfusion of platelets in patients with ITP usu-
preferential reactivity against certain antigens, even though ally is attempted only in the setting of life-threatening hem-
the patient is positive for the antigen in question. The appar- orrhage. Administration of IVIg may improve the survival of
ent specificity demonstrated by autoantibodies is often to an transfused platelets in patients with ITP, and the administra-
antigen in the Rh blood group system, most commonly to tion of IVIg or continuous infusions of platelets have been
the e (little e) antigen. These antibodies are referred to as attempted in patients with life-threatening hemorrhage and
mimicking antibodies. In these situations, the survival of patients undergoing major surgery. Elective splenectomy
antigen-positive donor RBCs usually does not differ from typically is managed with preoperative IVIg or a pulse of
that of the patient’s own RBCs; however, in some cases, corticosteroids. Intravenous Rh(D) IgG can be administered
RBCs that do not express the target antigen may survive lon- more quickly than IVIg, but its use is limited to Rh(D)-
ger following transfusion. positive, nonsplenectomized patients.
In patients with clinically significant cold-reacting auto- Except in life-threatening bleeding situations, such as
antibodies, such as anti-I, blood lacking the antigen in ques- intracranial hemorrhage, platelet transfusions should be
tion is often not available. In this situation, blood transfused avoided in consumptive thrombocytopenias, such as TTP
through a blood warmer usually will survive adequately if and heparin-induced thrombocytopenia, because they could
the patient is kept warm while other forms of treatment, exacerbate the thrombotic process that characterizes these
such as cytotoxic chemotherapy or plasmapheresis, are insti- disorders and increase the risk of arterial thrombosis.
tuted. If requested, a blood bank workup of the cold-reacting
autoantibodies can include the performance of a thermal-
Sickle cell disease
amplitude determination in which RBC binding in vitro to
the patient’s autoantibodies is assessed as a function of tem- Patients receiving long-term transfusion therapy often
perature (eg, at 4°C, 22°C, 30°C, and 37°C). The results of become alloimmunized to multiple blood components,

including RBC antigens, leukocyte antigens, and plasma The development of alloantibodies in sickle cell patients is
proteins. This is especially true in patients with sickle cell occasionally associated with autoantibody formation, which
disease, of whom ~25% become alloimmunized to non- complicates transfusion therapy. The prevalence was 8% in a
ABO RBC antigens. Patients with sickle cell disease are par- pediatric series of sickle cell patients; about half the patients
ticularly prone to alloimmunization against RBC alloantigens with detectable autoantibody had evidence of hemolysis,
and account for more than half of the requests for rare- often associated with a positive DAT for complement.
phenotype blood received by the American Red Cross Rare Hyperhemolysis is another transfusion-related complication
Donor Registry, which collects and distributes blood from observed in sickle cell disease, often presenting with severe
donors with unusual phenotypes. The reasons for the high anemia and reticulocytopenia 7-10 days after transfusion.
rate of alloimmunization to RBC antigens among patients The hematocrit is lower than the pretransfusion value, indi-
with sickle cell disease are poorly understood. Likely factors cating destruction of autologous RBCs. The DAT is often
include repeated exposure to RBC antigens in combination negative, and new alloantibodies may or may not be detect-
with differences in the frequencies of certain RBC antigens able. It is important to recognize this syndrome because its
between the predominantly white donor population and the management consists of the judicious avoidance of addi-
predominantly black patient population. Patients with sickle tional transfusions in the face of severe anemia, corticoste-
cell disease also may have a higher intrinsic immune res roids, IVIg, and erythropoietin. There is an association
ponsiveness to blood group antigens due to underlying among DHTRs, the onset of sickle cell vaso-occlusive crises,
inflammation. and the occurrence of other sickle cell disease complications.
Indications for transfusion in sickle cell disease include A transfusion reaction should always be considered in the
stroke, acute chest syndrome, aplastic crisis, and preopera- differential diagnosis of sickle cell complications.
tive preparation to reduce the risk of postoperative respira-
tory complications and vaso-occlusive crises. Sickle cell
Key points
patients who require chronic transfusion therapy accumu-
late iron much less rapidly if the transfusion occurs in the • Patients with sickle cell disease should receive extended-match
form of exchange procedures rather than simple transfu- red cells to prevent alloimmunization to non-ABO red cell
sions, although exchange carries the risk of additional donor antigens, when possible.
• There is a high incidence DHTRs in SCD and the potential for
exposures and requires adequate vascular access.
life-threatening hyper-hemolytic transfusion reactions.
Techniques for preventing and managing alloimmuniza-
• Transfusion is standard of care for the prevention and
tion in patients with sickle cell disease are varied. In many
treatment of select complications of SCD (eg, stroke, acute chest
institutions, extended antigen typing for the most important syndrome, and aplastic crisis).
antigen systems in addition to ABO and Rh(D), including
Kell, Kidd, Duffy, MNSs, Rh(C), and Rh(E), is performed on
RBCs of patients with sickle cell disease before transfusion
Massive transfusion
therapy is initiated. DNA-based methods may be preferable,
particularly in multiply transfused patients. Extended RBC Massive transfusion is defined as the replacement of one
phenotyping facilitates future identification of antibody blood volume within 24 hours, or transfusion of a certain
specificities and prophylactic transfusion of antigen-matched number of RBC units within 4-6 hours, typically in the set-
units, which may reduce the incidence of subsequent alloim- ting of severe trauma or major surgery. Coagulopathy of
munization. Most commonly, matching is performed for the massive transfusion is multifactorial; causes include: hypo-
extended Rh antigens (C, c, E, and e) and the major Kell thermia, acidosis, the dilutional effect of blood loss, inade-
antigen (K), in part because providing RBC units that are quate coagulation factor replacement, reduced hepatic
matched for these antigens does not require use of pheno- synthesis of coagulation factors in massive hepatic injury,
typically rare blood units. It is not usually possible to match disseminated intravascular coagulopathy from hypotension
routinely for all antigens for which the patient has been and tissue injury, and consumption of coagulation factors or
typed because most blood centers that identify and store platelets.
phenotypically rare blood provide these units only to patients Unfortunately, both laboratory tests and transfusion vol-
with demonstrated alloantibodies because of the limited ume do not correlate well with the severity of bleeding. In
supplies of rare-phenotype blood. The adoption of high- the absence of hypovolemic shock and significant liver dys-
throughput blood group genotyping platforms by more function, the exchange of one circulating plasma volume
blood centers will facilitate extended blood group matching does not reduce the clotting factor activities below levels
between blood donors and sickle cell patients, particularly necessary to maintain hemostasis, that is, approximately
coupled with minority donation recruitment efforts. 50% of the normal levels. Thrombocytopenia is the most

Transfusion risks | 329
frequent abnormality associated with massive transfusion. Routine transfusion of platelets to patients who are not
When transfusions of 1.5-2.0 blood volumes are adminis- bleeding and are not severely thrombocytopenic does not
tered over 4-8 hours, the mean reduction in the peripheral appear to be justified. Thromboelastography offers whole
blood platelet count is approximately 50%. One strategy for blood-based coagulation testing that can localize the coagu-
management of massive transfusion is to use whole blood lation defect to a deficiency in platelets, coagulation factors
coagulation tests, such as thromboelastography and rota- or fibrinogen, or excessive fibrinolysis. In selected centers, its
tional thromboelastometry, or other laboratory tests, includ- use has reduced indiscriminate transfusion of plasma or
ing prothrombin time (PT), international normalized ratio platelets in nonsurgical bleeding in perioperative or postop-
(INR), activated partial thromboplastin time (aPTT), plasma erative cardiac surgery patients.
fibrinogen level and platelet counts, and use the results to
guide component replacement therapy. In practice, many
trauma centers have adopted an empiric preemptive Transfusion risks
approach to prevent coagulopathy based on military experi-
ence using early aggressive plasma transfusion and/or fixed
ratios of blood components in standardized massive transfu- Clinical case
sion protocols. Patients undergoing massive transfusion Shortly after initiation of a red cell transfusion, a 63-year-old
need to be monitored for electrolyte disturbances such as patient with melena develops pain at the infusion site followed
hypocalcemia (citrate in the anticoagulant used for all blood by dyspnea, fever, chills, and low back pain. His urine is noted
components binds free calcium), hyperkalemia or hypokale- to be red and his plasma demonstrates free hemoglobin. Repeat
mia, and metabolic alkalosis (from citrate metabolism). testing of both the red cell product and the patient reveals that
the product is type A, the patient is type O, and the crossmatch
is incompatible.
Cardiopulmonary bypass
Alterations in the laboratory parameters of hemostasis are

Acute hemolytic transfusion reactions
observed in virtually all patients undergoing open-heart sur-
gery and extracorporeal circulation. Less than 10% of these The patient in this clinical case illustrates the typical presen-
patients experience severe bleeding, however, and during the tation of an acute hemolytic transfusion reaction: pain at the
history of cardiopulmonary bypass procedures, blood usage administration site, fever, chills, back pain, dark urine, and
for surgery involving extracorporeal circulation has laboratory evidence of intravascular hemolysis. ABO iso-
decreased markedly. Dilution by priming the extracorporeal hemagglutinins are complement-fixing and lead to the intra-
circuit with nonblood solutions may reduce the platelet vascular destruction of the transfused RBCs, which can
count by as much as 50%. Platelet dysfunction results from manifest as hemoglobinemia and hemoglobinuria. Often,
platelet contact with the surfaces of extracorporeal circuits, fever is the only initial sign. Activation of complement leads
including pumps and ventricular assist devices. Preoperative to the release of cytokines, including tumor necrosis factor,
therapy with antiplatelet agents, such as aspirin clopidogrel accounting for fever and chills. The serologic hallmark of an
and GPIIb/IIIa inhibitors, exacerbates platelet dysfunction. acute hemolytic reaction is a positive DAT that demonstrates
Changes in platelet function due to exposure to the extracor- both IgG and complement on the surface of the recipient’s
poreal circuit may persist for several hours after discontinu- circulating RBCs. Disseminated intravascular coagulation
ation of bypass. Although plasma coagulation factor levels also occurs and bleeding may result.
are diluted by nonblood priming solutions, coagulation fac- Patient misidentification due to clerical error or failure to
tor levels ordinarily remain above the minimal level needed follow established hospital procedures remains the most
for hemostasis, that is, approximately 50% of the normal common cause of acute hemolytic transfusion reactions;
factor levels. The extracorporeal circuit is not thought to therefore, the importance of definitive bedside patient iden-
consume clotting factors directly. Consequently, platelet tification, both at the time that type and screen specimens
transfusion to correct quantitative or qualitative platelet are obtained and at the time that the product is ready to be
defects is the mainstay of treatment of nonsurgical bleeding administered, cannot be overemphasized. Barcode and
associated with cardiopulmonary bypass procedures. In radio-frequency chip technologies to ensure correct patient
addition, because platelet products contain significant quan- identification have been shown to reduce the risk of
tities of plasma, platelet transfusion may be effective even mistransfusion.
when the primary laboratory abnormalities appear to be Acute hemolytic reactions can occur after platelet trans
coagulation factor related, although particular attention fusions, typically involving a group A patient receiving
should also be paid to fibrinogen replacement in this setting. group O platelets that contain high-titer anti-A antibody.

Treatment of acute intravascular hemolytic reactions is sup- with chills and/or rigors with a minimal or absent febrile
portive and includes fluids and vasopressors for hypotension component, particularly in patients receiving an antipyretic.
and maintenance of urine output. Symptoms are usually self-limited and respond to symptom-
atic therapy, which includes antipyretics for fever and chills
and meperidine for rigors. While the clinician can adminis-
Delayed hemolytic transfusion reactions
ter subsequent transfusions without undue risk in most
DHTRs occur when a patient develops an alloantibody to an cases, the main concern is that an elevation in temperature
RBC antigen following pregnancy, transfusion, or HSCT, but during a transfusion, although most likely the result of this
the titer of the antibody falls to below the detectable limit, innocuous febrile transfusion reaction or the patient’s
resulting in an apparently negative antibody screen, before underlying medical condition, cannot be distinguished from
a subsequent RBC transfusion. Following the subsequent an evolving life-threatening acute hemolytic or septic trans-
transfusion, the patient develops an anamnestic immune fusion reactions in which fever can be the only clue. Com-
response to the mismatched antigen, leading to delayed pleting a blood bank evaluation to rule out hemolysis is
antibody-mediated destruction of the transfused RBCs. Clin- important in this setting. The increasing adoption of univer-
ical symptoms of hemolysis including fever, anemia, and sal leukoreduction has been associated with a significant
jaundice develop 7-10 days after the transfusion; however, the reduction in febrile nonhemolytic transfusion reactions, but
link to the preceding transfusion is not always obvious. If no change in the incidence of allergic reactions.
blood bank testing is performed at this point, the DAT is
often positive for IgG, with or without complement depend-
Allergic transfusion reactions
ing on the causative antibody. Because a positive DAT can be
nonspecific, an eluate may be performed to remove the IgG Minor allergic reactions manifested by urticaria are frequent.
antibody coating the circulating RBCs in order to identify it. Antihistamines generally alleviate symptoms of allergic reac-
The antibody screen may also demonstrate the presence of a tions, but they have not been shown to prevent them. Many
new antibody, although this may lag behind the positive DAT urticarial reactions are donor specific and thus do not recur
by a few days. Hemolysis is usually IgG mediated and thus with subsequent transfusions. If a recipient experiences
extravascular, although IgG alloantibodies to Kidd blood multiple urticarial reactions, the clinician should consider
group antigens may fix complement and cause intravascular premedication with antihistamines. Washed products resus-
hemolysis. Hemoglobinuria may occur, and occasional pended in albumin and/or saline may be considered in severe
instances of severe complications such as acute renal failure cases. Although removing plasma through washing mitigates
or disseminated intravascular coagulation have been reported. allergic reactions, washing platelets impairs platelet function
The antibodies most often implicated in DHTR are directed and leads to accelerated clearance after transfusion.
against antigens in the Rh (34%), Kidd (30%), Duffy (14%), Severely IgA-deficient patients may make anti-IgA anti-
Kell (13%), and MNSs (4%) antigen systems. bodies that can cause anaphylactic reactions, but this is a rare
occurrence. Considering that ~1 in 1,200 people is IgA defi-
cient with anti-IgA antibodies and that passively transfused
Febrile nonhemolytic transfusion reactions
anti-IgA antibodies do not cause allergic reactions, the
Multiparous women and multiply transfused patients pathophysiology of recurrent and severe allergic transfusion
develop anti-leukocyte antibodies that cause febrile nonhe- reactions in IgA deficiency is incompletely understood.
molytic reactions to RBC or platelet transfusions. In addi- Washed RBCs, washed platelets, and/or platelet and plasma
tion, during the storage of blood, clinically significant products from IgA- deficient donors should be transfused
quantities of cytokines (IL-1, IL-6, IL-8, and tumor necrosis only when a patient has severe IgA deficiency (<0.05 mg/dL)
factor) are sometimes liberated from donor-derived leuko- and a concern for anaphylactic reactions. Most IgA deficient
cytes present in platelet and RBC products. Prestorage patients, even those with anti-IgA, have no adverse reactions
leukoreduction, as opposed to poststorage bedside leukofil- to transfusion. There are also reports of patients with defi-
tration, may reduce the accumulation of these biologic ciencies of haptoglobin and various complement compo-
mediators and the incidence of febrile, hypotensive, or nents, such as C4a (Rogers antigen) or C4b (Chido antigen),
hypoxic transfusion reactions. developing anaphylactic reactions to platelets.
Febrile nonhemolytic transfusion reactions typically man-
ifest during or within 4 hours of transfusion with fever
Transfusion-related acute lung injury (TRALI)
(defined as an increase in temperature of 1°C above the
patient›s baseline, typically to >38°C) with or without chills TRALI is a potentially life-threatening reaction that in many
and/or rigors. Such reactions may also manifest primarily cases appears to be caused by passive transfusion of donor

antigranulocyte antibodies (anti-HLA or anti-HNA anti- United Kingdom in late 2003, where 60% of TRALI cases
bodies), cytokines, biologically active lipids, or other sub- previously had been caused by plasma transfusions, no
stances. The resulting clinical picture is acute lung Injury reports of TRALI deaths due to plasma occurred after 2004
with noncardiogenic pulmonary edema. Signs and symp- (six deaths occurred in 2005, none from plasma). Major
toms include dyspnea, hypoxemia, hypotension, fever, and a blood suppliers in the United States now limit the use of
chest x-ray showing bilateral infiltrates with pulmonary female plasma or screen for HLA and HNA antibodies in
edema. Aggressive pulmonary support, including intubation multiparous donors. Even with these precautions in place,
and mechanical ventilation, frequently is needed. Approxi- cases of TRALI in which HLA or other granulocyte-specific
mately 80% of patients improve within 48-96 hours, and antibodies do not appear to be responsible will not be elimi-
100% of patients require oxygen support with approximately nated. Therefore, strict transfusion criteria for plasma-rich
70% requiring mechanical ventilation. Infrequently, anti- blood products, early recognition, and prompt clinical man-
bodies in the recipient may react with donor granulocytes agement are key to dealing with this potentially fatal transfu-
that are introduced by units of RBCs or platelets. In some sion reaction. Reporting suspected cases of TRALI to the
cases of TRALI, neither recipient nor donor-derived anti- blood bank is also important in limiting potential risk to
bodies can be identified. Other mechanisms have been other patients by quarantine of any co-components from the
advanced such as the priming of neutrophils by bioactive same donation and evaluating the donor with possible exclu-
lipids that accumulate during blood storage. sion from future donation if TRALI is confirmed.
In 2007, TRALI represented approximately 65% of all
transfusion-related fatalities reported to the US FDA.
Transfusion-associated circulatory overload
Although TRALI accounted for only 37% of transfusion-
related fatalities reported to the FDA in the 5-year period Dyspnea with or without hypoxia during or after transfu-
from 2008-2012, most likely due to widespread Implementa- sion, accompanied by signs of volume overload, such as an
tion of TRALI risk reduction strategies, it remains the lead- increase in blood pressure, jugular venous distention, and
ing cause of death due to transfusion in the US. elevated pulmonary arterial wedge pressure, represents
The true incidence rate of TRALI is unknown, but it may transfusion-associated circulatory overload (TACO). At ini-
occur in as many as 1 in 5,000 transfusions of any plasma- tial presentation, TACO and TRALI may be difficult to dis-
containing blood product (ie, RBCs, platelet concentrates, tinguish from each other. Despite increased awareness of
platelet apheresis units, and plasma) with a 5%-10% fatality TACO, it remains underdiagnosed or at least under-reported
rate. TRALI can be difficult to distinguish from the manifes- to hospital blood banks as a transfusion reaction. Despite this
tations of a patient’s underlying medical problems, particu- underreporting, TACO accounted for approximately 18% of
larly those of cardiac origin, such as congestive heart failure transfusion-related fatalities reported to the FDA between
and fluid overload brought on by the transfusion. A consen- 2008 and 2012, making it the second most common cause of
sus definition of TRALI is: acute lung injury (ALI) occurring reported death due to transfusion in this time period (after
during a transfusion or within 6 hours of completing a TRALI). Risk factors for TACO include extremes of age, his-
transfusion with no other temporally associated causes of tory of cardiac disease, renal failure, and transfusion of mul-
acute lung injury (ALI). ALI is defined as a syndrome of: tiple blood components within a short period of time. An
(i) acute onset, (ii) hypoxemia (PaO2/FiO2 <300 mm Hg, elevated brain natriuretic peptide (BNP) may be helpful to
O2 saturation <90% on room air, or other clinical evidence), distinguish TACO from TRALI in some cases. Therapy con-
(iii) bilateral pulmonary infiltrates, and (iv) no evidence of sists of diuretics and decreased blood administration rate.
circulatory overload.
Clinical management is supportive with the goal of revers-
Transfusion-associated graft-versus-host
ing progressive hypoxemia. There is no universal method to
disease (TA-GVHD)
prevent TRALI. Once blood from a particular donor is
implicated in a case of TRALI, that donor is excluded from TA-GVHD is an important risk in patients undergoing
the donor pool. Preventing the first cases of TRALI by those treatment of hematologic malignancies, patients undergo-
donors, however, requires the elimination of all blood donors ing HSCT, and patients with certain congenital immunode-
whose plasmas contain anti-HLA or anti-neutrophil anti- ficiency syndromes. The pathophysiology of TA-GVHD
bodies. For plasma, this is achieved by excluding female involves engraftment of small numbers of donor-derived
donors from the plasma donor pool because multiparous passenger leukocytes in a host whose immune system is
females are the most likely among a healthy donor popula- unable to recognize these cells as “foreign” and/or unable to
tion to have anti-HLA antibodies as a result of sensitization eliminate them. Unlike HPSC transplantation-associated
during pregnancy. When this approach was adopted in the GVHD, in which the hematopoietic organ is donor derived

and thus relatively protected from immune assault by include the diversion of the first aliquot of donor blood from
donor-derived T-cells, in transfusion-associated GVHD, the the collection bag to remove the skin core that otherwise
hematopoietic organ is recipient derived. Therefore, when would be introduced by the phlebotomy needle. Efforts to
TA-GVHD develops, mortality approaches 100% as a result detect the presence of bacteria in platelet units before dis-
of complications of severe pancytopenia. Patients may pensing to a patient include incubating an aliquot of the unit
develop signs and symptoms of classic transplantation- in a culture system and using a rapid strip immunoassay for
associated GVHD, including skin rash, diarrhea, liver func- bacterial antigens. Other less sensitive methods for detection
tion test abnormalities, and other symptoms related to using a surrogate marker for evidence of bacterial metabo-
pancytopenia such as infection and bleeding. The infusion of lism, such as a low pH, in an aliquot of the platelet suspen-
any cellular blood product can theoretically cause TA-GVHD. sion have been discontinued.
Irradiation of all cellular blood products before transfusion— While platelet products are typically contaminated by
but not conventional leukoreduction—virtually eliminates gram positive cocci, such as coagulase-negative Staphylo-
the risk of TA-GVHD. cocci, sepsis associated with transfusion of RBC units is most
TA-GVHD also has been described in immunocompetent often due to gram negative organisms, particularly Yersinia
patients when the donor is homozygous for an HLA haplo- enterocolitica. Fatal reactions to RBCs caused by contamina-
type shared with the recipient. Transfusion within relatively tion with Yersinia enterocolitica have been reported. This
less HLA diverse populations, such as in Japan, appears to Gram-negative organism can survive during refrigerated
increase the risk of TA-GVHD because of the increased prev- storage and lead to bacteremia or septic shock in the transfu-
alence of donors who are homozygous for an HLA haplotype sion recipient.
shared with the recipient. This appears to set up a unidirec- Malarial transmission by transfusion is uncommon, but
tional HLA mismatch in which the recipient immune system cases are occasionally reported. Currently, no FDA-approved
is unable to recognize the donor-derived passenger leuko- test is available to screen donors for malaria, and therefore
cytes as being foreign and thus is unable to eliminate the pas- screening is accomplished by donor questioning. Donors
senger leukocytes; whereas the passenger T-lymphocytes with a history of residence in a malaria-endemic area or
recognize the nonshared HLA allele on the recipient’s cells travel associated with a risk of malarial exposure are deferred
and thereby initiate a graft-versus-host reaction. For similar for up to 3 years, depending on the exposure. With the immi-
reasons, directed-donor transfusions between blood rela- gration of individuals from South America to the United
tives, such as siblings or mother to neonate, increase the risk States, there is concern that Chagas disease may emerge as a
of TA-GVHD. Therefore, all directed donations of cellular common transfusion-transmitted infection. Trypanosoma
blood products from blood relatives must be irradiated. cruzi parasites can survive several weeks of storage in blood,
and contamination of blood products with this organism is
already a significant problem in parts of South America. An
Infectious complications FDA-approved blood donor-screening test for antibodies to
T. cruzi is available. Blood donors only need to be tested at
Bacterial and parasitic transmission by transfusion
their first donation. Transfusion-transmitted babesiosis has
Bacterial contamination of platelet products is a significant been reported in New England and the upper Midwest and
issue given that platelets are stored at room temperature. has been identified in patients receiving platelets, refriger-
Before the introduction of specific precautions to reduce ated RBCs, and even frozen-thawed RBCs. A number of
bacterial contamination of platelet products, as many as 1 in investigational tests are being evaluated for donor screening
1,000 to 1 in 2,000 platelet units were contaminated with in areas endemic for Babesia. Borrelia burgdorferi, the etio-
bacteria, resulting in clinical sepsis after 1 in 4,000 platelet logic agent of Lyme disease, has yet to be confirmed as hav-
transfusions. As bacterial contamination of platelets became ing been transmitted by blood transfusions.
recognized as the most common cause of transfusion-
associated morbidity and mortality due to an infectious
Hepatitis
source in the United States (greater than hepatitis, HIV, and
other viral sources combined), methods to limit and detect Despite the exclusive use of volunteer blood donors and
the presence of bacteria in platelet components became screening of donor blood for hepatitis B and hepatitis C
mandated. Since the introduction of bacterial screening, the viruses, posttransfusion hepatitis occasionally still develops.
risk of septic transfusion reactions for apheresis platelets has Acute transfusion-related hepatitis C virus infection is sub-
declined to approximately 1 in 75,000, and the risk of a fatal clinical and anicteric in most cases. Hepatitis C virus infec-
septic reaction has declined to approximately 1 in 500,000. tion frequently becomes chronic, however, and often results
Efforts to limit the introduction of bacteria into platelets in clinically significant liver dysfunction. Recent data suggest

that patients who have undergone HSCT are at increased West Nile virus
risk for late-onset cirrhosis after hepatitis C virus exposure.
During the 2002 West Nile virus (WNV) epidemic in the
With current antihepatitis C virus antibody tests and
United States, 23 individuals acquired WNV after blood
nucleic acid testing, it is estimated that the risk of posttrans-
transfusion, developing fever, confusion, and encephalitis
fusion hepatitis C is 1 per 1.1 million units transfused. The
characteristic of WNV infection within days to weeks of
risk of HBV transmission by transfusion decreased from
transfusion. As a result, blood centers implemented nucleic
1:220,000 to approximately 1:750,000 after implementation
acid-based testing to screen all donations for WNV. In a sur-
of HBV DNA testing. Table 12-6 summarizes the estimated
vey of 2.5 million donations in 2003, 601 donations (0.02%)
risks of various transfusion-associated infections.
were found to contain WNV. A subsequent follow-up study
Photochemical pathogen inactivation strategies have just
detected no cases of transfusion-transmitted WNV infection
become licensed in the US after extensive use elsewhere in
among recipients of tested blood; however, rare break-
the world. To date, this approach appears both efficacious
through transmissions have been reported.
and relatively sparing in terms of qualitative platelet func-
tion, although decreases in quantitative platelet recovery
have been observed in some studies. Parvovirus B19
Rare transmissions of parvovirus B19 by transfusion have

HIV and human T-cell lymphotropic viruses been recognized. A recent study documented persistence of
low levels of parvovirus B19 DNA in a high percentage of
The risk of acquiring HIV-1 or HIV-2 infection as a result of
multitransfused patients. The long-term clinical implica-
transfusion currently is estimated to be 1 in 1.5 million.
tions of this finding currently are unknown. Parvovirus (and
Nucleic acid amplification testing for HIV has reduced the
other viruses without a lipid envelope such as hepatitis A
window of serologic conversion from 16 days to about 9
virus) is not eliminated by solvent detergent treatment.
days. The availability of heat-treated concentrates, solvent
Acute parvovirus B19 infection can result in impaired eryth-
detergent-treated products, and recombinant factor concen-
ropoiesis and can cause an aplastic crisis in patients with
trates has essentially eliminated HIV as a therapy risk for
sickle cell disease and other hemolytic diseases. Infection
hemophiliacs.
with this virus can also result in significant fetal harm when
Human T-cell lymphotropic virus 1 (HTLV-1) is a retrovi-
a pregnant woman is infected during weeks 9 to 20 of preg-
rus associated with adult T-cell leukemia or lymphoma and
nancy. There is no currently available blood donor screening
tropical spastic paraparesis. Because asymptomatic blood
assay for this virus.
donors can transmit this virus, screening for HTLV-1 in
blood donors is currently performed in the US. Several cases
of neuropathy had been reported in transfusion recipients Cytomegalovirus
before the availability of testing. HTLV-2, a related virus with
Leukocytes inevitably contaminate RBC and platelet prod-
antigenic cross-reactivity to HTLV-1, is endemic in certain
ucts, and they are capable of transmitting CMV infection.
Native American populations and also has been found in a
Transfusion-transmitted CMV infection is an important
high proportion of intravenous drug users. The risk of HTLV
issue in transfusion of cellular blood products to neonates,
transmission by transfusion using current test methods is
particularly low-birth-weight infants born to seronegative
approximately 1 in 2.7 million.
mothers, HSCT recipients, and other highly immunosup-
pressed patients. The risk of acquiring CMV from transfu-
sions is particularly high when pretransplantation serologic
Table 12-6 Infectious complications of transfusion testing reveals that neither the HPSC donor nor the recipient
Infectious agent Approximate risk per transfused unit previously has been exposed to CMV. In addition, transplan-
tation recipients are at increased risk for transplantation-
Hepatitis B virus 1:750,000 after implementation of HBV
associated CMV reactivation when either the donor or the
DNA testing
recipient is seropositive for CMV before transplantation.
Hepatitis C virus 1:1.1 million
The latter consideration often affects the choice of HPSC
HIV-1, HIV-2 1:1.5 million
donors.
HTLV-1, HTLV-2 1:2.7million
Bacterial sepsis 1:75,000 (platelet transfusions); 1:250,000 to For these reasons, some institutions use blood products
1:10 million (red blood cell transfusions) obtained exclusively from CMV-seronegative donors when
providing blood products to neonatal recipients or recipi-
HIV = human immunodeficiency virus; HTLV = human T-cell
ents of HPSC transplantations. However, as noted earlier, a
lymphotropic virus.

landmark randomized comparison of leukoreduced versus preoperative anemia often can be achieved simply through
CMV-seronegative blood components in CMV-seronegative iron replacement; use of erythropoietin also may be indi-
HSCT recipients (with seronegative donors) found no sig- cated in some cases. Avoidance of iatrogenic anemia by
nificant difference in the incidence of CMV infection, and avoiding unnecessary blood draws is equally important. In
CMV disease as a composite outcome and most transplanta- the setting of ICUs, routine blood draws have been demon-
tion centers, in practice, will use prestorage leukoreduced strated to result in the loss of the equivalent of 1-2 units of
blood components for CMV prevention. Other institutions RBCs per week. Critical thinking should ensure that all blood
simply use leukoreduced blood products in all recipients, tests ordered are justified and actively contribute to patient
regardless of CMV status. The latter strategy has the addi- care. The frequency, timing, and volumes of blood draws,
tional advantage of reducing the risk of alloimmunization to including use of lower volume blood collection tubes when
HLA antigens and thus of developing refractoriness to plate- appropriate, also should be coordinated to limit the volume
let transfusions. of patient blood collected for testing.
In the past, preoperative autologous donation, where the
patient would donate blood for his or her own use in the
Blood management
weeks before surgery, used to be the most common approach
to avoid allogeneic transfusion for elective surgical cases.
Clinical case Although the use of autologous blood may eliminate trans-
fusion risks because of transfusion-transmitted infection
A 44-year-old multiparous female requires orthopedic surgery.
Pretransfusion testing reveals antibodies to three separate red
(except for bacterial contamination of the unit), the risk of
cell antigens: K (Kell system), Fya (Duffy system), and E (Rh transfusion of ABO-incompatible blood due to a clerical
system). Crossmatch-compatible blood is transfused, and the error, still exists (ie, the inadvertent transfusion of the wrong
patient does well. A second operation is needed, and at this time patient’s autologous blood). Likewise, transfusion-associated
repeat screening of the patient’s plasma detects an additional complications such as those related to fluid overload in a
antibody directed against c (Rh system). Because of the multiple patient with cardiac disease can occur. Therefore, unless the
antibodies, a large number of donor units must be screened clinical condition of the patient actually warrants transfu-
to find the required number of antigen-negative units. The sion, autologous units of blood should not be used just
hematologist advises the surgeon that a comprehensive blood because they are available and “won’t hurt.” Use of preop-
management approach should be considered to reduce the
erative autologous blood donation is now broadly discour-
need for further allogeneic transfusion in this patient.
aged, as approximately 50% of autologous units are never
transfused and patients who donate autologous units preop-
The concept of blood management has been steadily gain- eratively may present to surgery with anemia that increases
ing in popularity with the recognition of the high costs asso- their overall risk of transfusion, particularly if the interval
ciated with transfusion, high frequency of inappropriate between donation and surgery is short. Of note, directed
utilization of blood products, and an ever-broadening range donations from relatives or friends selected by the patient
of adverse effects potentially associated with transfusion. have not been shown to be any safer than regular units from
Avoidance of unnecessary allogeneic transfusion is the ulti- the blood bank in terms of transmitting infectious agents. In
mate goal of blood management, and a multidisciplinary fact, blood from first-degree relatives must be irradiated to
approach is required to achieve it. The elements of blood prevent TA-GVHD, and family donors can HLA alloimmu-
management include decreasing the need for transfusion, nize the patient through transfusion, which may preclude a
using the patient’s own blood when possible, optimizing uti- subsequent transplant. Directed donors are also more likely
lization of allogeneic blood products when transfusion is to be first time donors, who have a higher incidence of infec-
indicated, and performing utilization reviews with auditing tious disease positivity. For these reasons, directed donations
and benchmarking to induce and maintain the behavioral are now generally discouraged.
changes required for the broad application of blood manage-
ment in a hospital setting.
Intraoperative techniques
A number of surgical, anesthetic, and pharmacological

Iatrogenic and preoperative anemia
approaches can be utilized to reduce intraoperative bleeding.
The cornerstone of decreasing the need for transfusion is Use of the patient’s own blood to minimize the need for, or
appropriate medical management of anemia, particularly in entirely avoid, allogeneic transfusion may be accomplished
the preoperative setting in which anemia is the most impor- through acute normovolemic hemodilution (ANH) and red
tant predictor of perioperative transfusion. Management of cell salvage or perioperative autotransfusion. ANH involves

Bibliography | 335
removal of one or more units of whole blood in the operat- anesthesiology, surgery, internal medicine, nursing, pharmacy,
ing room immediately before surgery, with adequate fluid laboratory medicine, and hospital administration.
replacement to maintain an iso- or normovolemic state.
Blood shed during surgery is dilute in this case, theoretically
Bloodless medicine
resulting in a lower net loss of red cell mass after return of
the whole blood units to the patient toward the end of the Currently, no licensed blood substitutes are available for
case. The units collected by ANH may have the added benefit clinical use in the United States. A recent meta-analysis of
of providing additional platelets and coagulation factors. hemoglobin-based blood substitutes found excess myocar-
ANH has not been established definitively to avoid alloge- dial infarction and mortality in surgical patients who
neic transfusion. Conversely, intraoperative cell salvage can received the blood substitute compared with controls;
significantly reduce the need for allogeneic transfusion, par- patient groups included trauma, cardiac surgery, vascular
ticularly in cases associated with high-volume blood loss. In surgery, and elective orthopedic surgery.
this approach, blood is suctioned from the operative field Erythropoiesis-stimulating agents can be used in
into an anticoagulated reservoir and then washed with nor- patients who decline transfusion, either therapeutically to
mal saline. The washed salvaged RBCs are concentrated for treat anemia or prophylactically before elective surgery.
reinfusion to the patient. When using cell salvage techniques, The management of Jehovah’s Witness patients who
precautions must be taken to avoid potential hazards, such as require chemotherapy for hematologic malignancies or
air emboli and infusion of inadequately washed products. In HSCT can be challenging. A comprehensive approach is
some cases, postoperative wound drainage may be collected, required, including reduced-intensity conditioning che-
filtered, and administered with or without washing. Many motherapy, reduced phlebotomy and gastrointestinal
Jehovah Witnesses will consent to autologous transfusion blood loss, optimized pretransplantation blood counts
using cell salvage, potentially allowing for more complex sur- using iron and folate, erythropoiesis-stimulating agents,
geries to be performed in this patient population. The tech- and possibly thrombopoietin mimetic agents, as well as
nique can also be helpful in patients for whom it is difficult prophylactic use of antifibrinolytic agents during the
to find compatible blood because of the presence of multiple period of thrombocytopenia.
red cell antibodies.
Key points
Judicious transfusion
• Avoiding iatrogenic anemia can help reduce the need for
Care should always be taken to transfuse the least amount of allogeneic transfusion in all patient populations.
blood products required to achieve the desired outcome. • Transfusing only when indicated (right product to the right
Clinicians must keep in mind that it is unnecessary to correct patient at the right time and for the right reason) can help avoid
unnecessary risks associated with transfusion.
a cytopenia or a clotting factor deficiency to normal levels;
• Preoperative medical management of anemia before elective
transfusion should be directed toward restoring only func-
surgery can reduce perioperative transfusions.
tionally adequate levels. For example, many patients with
• Preoperative autologous donation generally is discouraged
chronic anemia or thrombocytopenia tolerate much lower due to wastage of collected units, the residual risks of clerical
blood counts than patients with acute cytopenias involving error, bacterial contamination, and volume overload, as well as
the same lineages, and most patients tolerate clotting factor the preoperative anemia associated with these donations.
levels below 50% without difficulty. One of the major behav-
ioral changes incorporated into most blood management
programs is a shift in practice from transfusing RBC units in
multiples to a strategy of single-unit transfusions with subse- Bibliography
quent reassessment of patient status and need for further
transfusion. Evidence-based management of sickle cell disease. NHLBI Expert
Panel Report. 2014. http://www.nhlbi.nih.gov/health-pro/
Auditing of compliance with institutional transfusion guide-
guidelines/sickle-cell-disease-guidelines. A comprehensive,
lines, internal and external benchmarking, and ongoing data-
evidence-based guideline document for the management of SCD.
driven process improvement projects all contribute to improved Although aimed at primary care providers, the document is a
blood product utilization and systematic application of blood thorough source for the evidence of current standards of care
management concepts. The institutional oversight for such in SCD.
activities usually is provided by hospital transfusion committees, Anstee DJ. Red cell genotyping and the future of pretransfusion
or their equivalents, which typically include broad multidisci- testing. Blood. 2009;114:248-256. An excellent summary of recent
plinary representation from transfusion medicine, hematology, advances in blood group genotyping and its applications.

Anstee DJ. The functional importance of blood group-active Klein HG, Spahn DR, Carson JL. Red blood cell transfusion in
molecules in human red blood cells. Vox Sang. 2011;100:140- clinical practice. Lancet. 2007;370:415-426. A review of RBC
149. An updated review of the structure–function relationship transfusion, balancing the principles of oxygen transport with the
of RBC antigens, such as the Duffy glycoprotein as a chemokine uncertainties of the transfusion trigger, and the practical aspects of
receptor, and the transporter functions for the Rh proteins, band 3 component modification and transfusion reactions.
glycoprotein (Diego blood group), and aquaporin-I (Colton blood Lacroix J, Hebert PC, Fergusson DA, et al. Age of transfused blood
group). in critically ill adults. N. Engl J Med. 2015;372:1410-1418.
Avent ND. Large-scale blood group genotyping-clinical A large, multicenter RCT designed to look at morbidity and
implications. Br J Haematol. 2008;144:3-13. Another excellent mortality differences attributable to older (mean 22 day) versus fresh
summary of recent advances in blood group genotyping and its (mean 6 day). No differences were found.
applications. Lund N, Olsson ML, Ramkumar S, et al. The human Pk histo-
Bell EF. When to transfuse preterm babies. Arch Dis Child blood group antigen provides protection against HIV-1
Fetal Neonatal Ed. 2008;93:F469-F473. A succinct review of infection. Blood. 2009;113:4980-4991. Fascinating study
transfusions in the neonatal ICU, from the Iowa group that demonstrates that the presence of the Pk antigen on mononuclear
conducted one of the two RCTs comparing liberal versus restrictive cells confers resistance to HIV-1 infection, whereas cells completely
transfusion thresholds in preterm babies. lacking Pk (blood group “p” phenotype) were 10 to 1,000 times
Carson JL, Grossman BJ, Kleinman S, et al. Red blood cell more susceptible to HIV-1 infection.
transfusion: a clinical practice guideline from the AABB. Ann Nichols WG, Price TH, Gooley T, et al. Transfusion-transmitted
Intern Med. 2012;157:49-58. Clinical practice guideline for cytomegalovirus infection after receipt of leukoreduced blood
RBC transfusion developed by the AABB based on a systematic products. Blood. 2003;101:4195-4200. This paper compares the
review of randomized clinical trials evaluating transfusion incidence of transfusion-transmitted CMV infection with two
thresholds for RBC transfusions. The guideline includes a strong commonly used prevention strategies.
recommendation with high-quality evidence for adhering to a Rebulla P, Finazzi G, Marangoni F, et al. The threshold for
restrictive transfusion strategy (7 to 8 g/dL) in hospitalized, stable prophylactic platelet transfusions in adults with acute myeloid
patients. leukemia. N Engl J Med. 1997;337:1870-1875. A randomized
Carson JL, Terrin ML, Noveck H, et al. Liberal or restrictive clinical trial examining the consequences of using a prophylactic
transfusion in high-risk patients after hip surgery. New platelet transfusion threshold of 10,000/mL versus 20,000/mL. The
Engl J Med. 2011;365:2453-2462. No difference in mortality use of the lower threshold appeared to be associated with minimal
or functional (walking) capacity in elderly patients at high increased risk, but the statistical power of the study to detect
cardiovascular risk randomized to transfusion trigger of clinically meaningful increases in the risk of major hemorrhage
hemoglobin 10 g/dL or 8 g/dL. was limited.
Fergusson DA, Hebert P, Hogan DL, et al. Effect of fresh red Roback JD, Caldwell S, Carson J, et al. Evidence-based practice
blood cell transfusions on clinical outcomes in premature, very guidelines for plasma transfusion. Transfusion. 2010;50:1227-
low-birthweight infants: the ARIPI randomized trial. JAMA. 1239. Evidence-based guidance, developed by the AABB, to
2012;308:1443-1451. The first multicenter RCT designed to look inform plasma transfusion decisions in common clinical scenarios
at morbidity and mortality outcomes in premature infants. No including massive transfusion and warfarin therapy.
difference were found. Schwartz J, Winters JL, Padmanabhan A, et al. Guidelines on
Gaydos LA, Freireich EI, Mantel N. The quantitative relationship the use of therapeutic apheresis in clinical practice-evidence-
between platelet count and hemorrhage in patients with based approach from the Writing Committee of the American
acute leukemia. N Engl J Med. 1962;266:905-909. One of the Society for Apheresis: the sixth special issue. J Clin Apher.
few studies to address the relationship between the degree of 2013;28:145-284. Comprehensive analysis of the indications for
thrombocytopenia and the risk of hemorrhage. apheresis using a rigorous evidence-based approach.
Goodnough LT, Shander A. Patient blood management. Seftel MD, Growe GH, Petraszko T, et al. Universal prestorage
Anesthesiology. 2012;116:1367-1376. Excellent overview of patient leukoreduction in Canada decreases platelet alloimmunization
blood management, including rationale and practical strategies for and refractoriness. Blood. 2004;103:333-339. A large study
implementation. documenting that prestorage leukodepletion of blood products
Hébert PC, Wells G, Blajchman MA, et al. A multicenter, markedly reduces the incidence of platelet transfusion
randomized, controlled clinical trial of transfusion refractoriness.
requirements in critical care. N Engl J Med. 1999;340:409-417. Silliman CC, Boshkov LK, Mehdizadehkashi Z, et al. Transfusion-
An excellent multicenter clinical trial showing that a conservative related acute lung injury: epidemiology and a prospective
transfusion protocol for the administration of RBCs to critically ill analysis of etiologic factors. Blood. 2003;101:454-462. A large
patients did not significantly change morbidity or mortality. prospective study documenting both the incidence of TRALI and
Heddle NM, Klama L, Singer J, et al. The role of the plasma from the apparent contribution of granulocyte antibodies, cytokines, and
platelet concentrates in transfusion reactions. N Engl J Med. biologically active lipids.
1994;331:625-628. One of the first reports documenting the role of Slichter SJ, Kaufman RM, Assmann SF, et al. Dose of prophylactic
cytokines in transfusion reactions. platelet transfusions and prevention of hemorrhage. New

Bibliography | 337
Engl J Med. 2010;362:600-613. Multicenter RCT showing no Vamvakas EC, Blajchman MA. Transfusion-related mortality:
difference in bleeding episodes when patients were randomized to the ongoing risks of allogeneic blood transfusion and the
prophylactic transfusions at low, medium, or high doses. The low- available strategies for their prevention. Blood. 2009;113:
dose arm received fewer platelets but more transfusion episodes. 3406-3417. An up-to-date review of transfusion-related
Stramer SL, Glynn SA, Kleinman SH, et al. National Heart, Lung, morbidity, including hemolysis, bacterial contamination,
and Blood Institute Nucleic Acid Test Study Group. Detection and TRALI but also immunomodulation and the benefits of
of HIV-1 and HCV infections among antibody-negative blood leukoreduction.
donors by nucleic acid amplification testing. N Engl J Med. Vaughn JI, Manning M, Warwick RM, et al. Inhibition of
2004;351:760-768. Reports on the effectiveness of NAT to detect erythroid progenitor cells by anti-Kell antibodies in fetal
individuals infected with HIV-1 and hepatitis C virus who are in alloimmune anemia. N Engl J Med. 1998;338:798-803. This
the seronegative period of infection. paper was the first to describe the hypoproliferative effect of
Stanworth SJ, Estcourt LJ, Powter G, et al. A no-prophylaxis anti-Kell alloantibodies on fetal erythroid progenitor cells, thus
platelet-transfusion strategy for hematologic cancers. N. Engl explaining how the anemia in anti–Kell-induced hemolytic disease
J Med. 2013;368:1771-1780. The largest prophylactic vs. no of the newborn results from a combination of decreased production
prophylactic platelet transfusion RCT. as well as increased destruction.
Stroncek DF, Rebulla P. Platelet transfusions. Lancet. Villanueva C, Colomo A, Bosch A, et al. Transfusion strategies
2007;370:427-438. An excellent review of platelet transfusion for acute upper gastrointestinal bleeding. N Engl J Med.
practice, including the transfusion trigger and platelet 2013;368:11-21. The only multicenter RCT that showed higher
refractoriness. mortality with a liberal transfusion strategy (9 g/dL hb trigger),
Trial to Reduce Alloimmunization to Platelets Study Group. as compared to a restrictive strategy (7 g/dL).
Leukocyte reduction and ultraviolet B irradiation of platelets Yazdanbakhsh K, Ware RE, Noizat-Pirenne F. Red blood cell
to prevent alloimmunization and refractoriness to platelet alloimmunization in sickle cell disease: pathophysiology, risk
transfusions. N Engl J Med. 1997;337:1861-1869. A landmark factors, and transfusion management. Blood. 2012;120:528-537.
study comparing four major options for the reduction of A comprehensive discussion of red cell alloimmunization in sickle
alloimmunization to HLA antigens with consequent reduction in cell patients with emphasis on possible mechanisms that can
the incidence of platelet transfusion refractoriness. trigger delayed hemolytic transfusion reactions.

CHAPTER
13
Cellular basis of hematopoiesis
and stem cell transplantation
Paresh Vyas and Aly Karsan
Introduction and historical Hematopoiesis through development Clinical transplantation of HSPCs, 350
perspective, 339 and into adult life, 339 Bibliography, 353
Introduction and historical (called spleen colony-forming unit [CFU-S] cells) of

perspective myeloid, erythroid, and megakaryocytic cells in the spleens
of the recipients. These colonies arise from a single implanted
Humans produce approximately 300 billion blood cells per cell and were the first demonstration of the existence of a
day. Hematopoiesis is the process by which blood cells are repopulating hematopoietic cell that could also differentiate.
made by transiting through a hierarchy of hematopoietic This chapter will summarize the development, differentia-
stem and progenitor cells (HSPCs). Hematopoietic stem tion, and localization of HSCs and the hematopoietic niche,
cells (HSCs) are defined by their ability to self-renew as well as well as aspects of HSC purification and expansion for
as differentiate into the progenitors that replenish the entire transplantation.
blood system. Residing at the top of this hierarchy, HSCs are
located in a number of embryonic niches, settling in the
bone marrow (BM) in adult life. Contrary to previous Hematopoiesis through development
assumptions, HSCs are a heterogeneous cell population, and and into adult life
likely number tens of thousands of cells in adult life, giving
rise to hundred of millions of hierarchically organised highly Hematopoiesis occurs in waves and at multiple discrete ana-
heterogeneous progenitor cells which in turn differentiate tomical sites that change through development (Figure 13-1).
into precursor cells and eventually mature effector cells. In humans, like other vertebrates, the initial wave of hemato-
Although the field of stem/progenitor cell biology has grown poiesis occurs in the extraembryonic yolk sac (YS) blood
dramatically over the past decades, bone marrow transplan- islands from weeks 3-6 of gestation in human (Figure 13-2).
tation (BMT) has been in routine clinical practice for The yolk sac primarily produces primitive erythroid cells
>50 years and is the only routine widely used example of (termed primitive erythropoiesis) expressing embryonic glo-
stem/progenitor cell therapy. Identification of bone marrow bins that deliver oxygen to tissues in the rapidly growing
(BM) HSPCs emerged after it was recognized that survivors embryo. Additional studies suggest that primitive hematopoi-
from atomic bomb explosions in 1945 died of hematopoietic esis produces myeloid and lymphoid cells (eg, macrophages
failure from radiation damage. In the early 1960s, a series of and natural killer cells). Interestingly, the developmental
seminal experiments by Till and McCulloch showed that the potential of embryonic hemopoiesis closely resembles hema-
transfer of BM cells from donor mice into lethally irradiated topoietic cells derived from human embryonic stem cells
host mice resulted in the formation of macroscopic colonies (hESCs) and induced pluripotent stem cells (iPSCs).
Primitive hematopoiesis is transient and is replaced by
adult or definitive hematopoiesis that sustains blood pro-
Conflict-of-Interest disclosure: Dr. Vyas declares no competing duction throughout development and postnatal life.
financial interest. Dr. Karsan declares no competing financial interest.
Off-label drug use: Dr. Vyas: not applicable. Dr. Karsan: not ap- Embryonic hematopoietic activity is detected at 4-5 weeks
plicable. of gestation in a region around the ventral wall of the dorsal
| 339

340 | Cellular basis of hematopoiesis and stem cell transplantation
Weeks 2.5–4 Weeks 4–10 Weeks 5 of gestation Mainly

of gestation of gestation until term postnatal
Yolk sac AGM Liver, spleen Bone marrow
Figure 13-1 Changing anatomical locations of hematopoiesis through development. Hematopoiesis is initially detected in the extra embryonic
yolk sac, then in the embryo in a region known as the aorta-gonad-mesonephros (AGM), the placenta, the umbilical arteries, and vitelline
vessels. It then shifts to the fetal liver and finally to the bone marrow.
aorta called the aorta-gonad mesonephros (AGM) (Figure still a matter of much debate whether HSCs arise from the
13-3). In early development, blood cells arise in close con- embryo proper, from the AGM, or by colonization from the
nection with the vascular structures (both in the yolk sac yolk sac.
and the dorsal aorta) giving rise to the notion that there HSCs are then detected in the developing fetal liver, spleen,
may be a common precursor cell population that gives rise and thymus, from 6-22 weeks in humans, where they expand
to both blood and blood vessel cells called a hemangioblast and differentiate into committed progenitor cells. Expansion
or that hematopoiesis arises directly from specialized and differentiation of HSCs allows for development of defini-
“hemogenic” endothelial cells, such as those lining the ven- tive red cells, myeloid cells, and lymphoid cells (T-cells that
tral aspect of the dorsal aorta. In mice and other animals, develop in the thymus and B-cells in the marrow). HSCs are
studies have shown that definitive hematopoietic stem cells then later detected in the bone marrow. It is also unclear if
(HSCs) with serially transplantable activity together with HSCs from the AGM migrate and colonize the other embry-
long-term engraftment capacity are found in the AGM. It is onic sites or whether they arise de novo at these other sites. It
Figure 13-2 Yolk sac blood

islands in a human fetus.
(A) Transverse section in a 3
somite human embryo (21
days) at the truncal level
stained with anti-CD34
antibody. Paired dorsal aorta
(da) ventrolateral to the neural
tube (nt) and above the yolk
sac (ys) and blood islands (bi).
(B) Higher magnification of a
solid hemangioblastic
mesodermal cluster of CD34
expressing cells in a blood
island of the yolk sac (brown).
Adapted from Tavian M, et al.
Development. 1999;126:793-
803, Figure 1A and B
(dev.biologists.org).

Hematopoiesis through development and into adult life | 341
epigenetic regulators are common with aging, and are seen

recurrently in hematopoietic diseases (eg, myeloid malig-
nancies), further demonstrates the importance of epigenetic
control of normal hemopoiesis.
Hematopoiesis: a hierarchical
differentiation cascade
Hemopoietic stem and progenitor cell populations give rise

to all hematopoietic cells through a cascade of differentia-
tion. They are both highly heterogeneous populations. They
are defined by two key properties: variable ability to self-
renew and a variable ability to differentiate and generate
mature blood cells. At the top of the hierarchy are HSCs.
They have extensive self-renewal capability and differentiate
into all blood cell types (Figure 13-4). During normal steady-
state hematopoiesis, adult HSCs cycle slowly and are rela-
Figure 13-3 Transverse section through the human fetal dorsal aorta tively resistant to cytokine stimulation. Within the HSC
at embryonic day 32 showing hemopoietic CD34+ cells clusters
compartment, long-term stem cells are usually quiescent
(arrowheads) associated with ventral wall. CD34+ cells (hemopoietic
and rarely divide. Short-term stem cells are more prolifera-
and endothelial cells) are stained brown. Adapted from Tavian M,
et al. Development. 1999;126:793-803, Figure 5c (dev.biologists.org).
tive. The remarkable ability of HSCs, at the single-cell level,
to reconstitute and maintain a functional hematopoietic sys-
tem over extended periods of time in vivo demonstrates
is not clear whether HSCs must first reside in the fetal liver these key properties. Self-renewal allows HSCs to be trans-
before seeding the BM. A large transient pool of HSCs has planted between individuals, and the surviving HSCs engraft,
been identified in the placenta of mice around the time of proliferate, and differentiate for the life of the recipient.
AGM HSC development. It remains to be determined whether HSCs can be serially transplanted for many generations
an equivalent population of HSCs exists in developing human between recipients. The most primitive HSCs are rare, repre-
placenta. Given that HSCs isolated from different locations or senting approximately 1 in 104-106 BM cells.
cell sources (eg, BM, fetal liver, placenta, hESCs/iPSCs) and When HSCs are recruited into active hematopoiesis, they
from organisms of different ages have been shown to have dis- exit the G0 phase of the cell cycle, and their daughter cells
tinct gene expression patterns and phenotypic features, this may either be a replicate of the parent cells (self-renewal) or
may have implications regarding choice of stem cell sources enter into a differentiation program. This distinctive, asym-
for human transplantation therapies. metric division process is the basis for long-term preserva-
Murine HSCs are heterogeneous with respect to lineage tion of stem cells while enabling continued production of
output generating either megakaryocyte or myeloid or lym- mature cells. The daughter cells that undergo differentiation
phoid progenitors or balanced progenitor output. In the proceed through a series of maturational cell divisions, cul-
mouse, fetal HSCs show extensive proliferation and tend minating in the generation of progenitor cells.
towards greater lymphoid output. During adult life, HSCs Progenitor cells are also hierarchically arranged. As they
are more quiescent. In contrast, functional HSCs are reduced differentiate from stem cells and through the progenitor
with age, and lymphocyte-producing HSCs diminish relative hierarchy, they progressively lose self-renewal and become
to myeloid-biased HSCs, potentially contributing to immune restricted in their differentiation potential such that more
alterations with age. These studies also demonstrate that multi-potent progenitors give rise to oligo-potent and finally
changes in functional potential of HSCs during development mono-potent progenitors. Progenitors are highly prolifera-
and aging correlate with changes in gene expression and tive and very cytokine responsive. New studies of murine
growth factor requirements. Additionally, HSCs demon- hemopoiesis suggest that in contrast to transplantation, the
strate epigenetic changes over time. Similar detailed studies source of most of the cells produced daily by the blood sys-
in humans still need to be performed, but the murine studies tem during normal steady-state blood production is main-
raise the prospect that there will be similar changes in func- tained by the continuing expansion of thousands of
tional potential of human HSPCs and that this may in part hematopoietic progenitor cells, each with a minimal contri-
be the cause of the relative lymphoid deficiency in the elderly. bution to mature progeny. Single cell transplant studies in
In humans, the discovery that mutations in genes encoding mice have also revealed a bypass pathway that produces

A B Lin–CD34+CD38+CD45RA–
HSC LT-HSC HSC CD90+CD49f+Rholow
ST-HSC Lin–CD34+CD38–CD45RA–
MPP MPP
CD90–CD133+
Lin–CD34+CD38–CD45RA+ Lin–CD34+
CMP CLP CD90–CD133+CD10–
LMPP ST-HSC EMP CD133low
MEP GMP MLP GMP MLP GMP

B cell
T cell
Lin–CD34+ Lin–CD34+ CD34+CD133low/– CD34+CD133low/–
NK cell CD38lowCD45RA+ CD38+CD45RA+
CD133+CD10+ CD133+CD10–CD7–
Megakaryocyte Neutrophil
Erythroid cell Monocyte/macrophage
Eosinophil
Basophil Lymphocyte Neutrophil Eosinophil Megakaryocyte
Monocyte/ Monocyte/ Basophil Erythrocyte
macrophage macrophage
Figure 13-4 Alternative hierarchal maps of human hematopoiesis. (A) In the classical model of human hematopoiesis, a self-renewing HSC
gives rise to a multipotent progenitor (MPP) that bifurcates into a common myeloid progenitor (CMP) and a common lymphoid progenitor
(CLP), which then eventually give rise to mature myeloid and lymphoid cells respectively. (B) Based on a compilation of more recent data the
MPP appears to give rise to a lymphoid-primed MPP (LMPP), both of which may have short-term self-renewal capacity, and an erythromyeloid
progenitor (EMP), differentiated by CD133 expression. The hierarchy following the LMPP further bifurcates into a multi-lymphoid progenitor
(MLP) and a granulocyte macrophage precursor (GMP). The EMP gives rise to a precursor that produces eosinophils and basophils (EoBP) and
a megakaryocyte/erythroid progenitor (MEP). It is highly likely that as our ability to functionally and molecularly interrogate HSPCs at a single
cell level improves our understanding of the hierarchy will be further refined. Surface antigen phenotypes of human hematopoietic stem and
progenitor cells are indicated. The immunophenotypes of the EMP, EoBP and MEP are not as well characterized as the other progenitors.
LT-HSC, long-term HSC; ST-HSC, short-term HSC.
long-term repopulating myeloid progenitors. This pathway and cell surface marker expression of HSPCs. Early work on
may be operative under stress, as progenitor populations murine BM revealed that the transplantable HSCs copurified
most readily respond to stress conditions in order to up- and with lymphocytes and led to the idea that HSCs are morpho-
down-modulate production of specific blood cell types. Pro- logically indistinguishable from lymphocytes. Density gradi-
genitors differentiate into lineage-restricted precursor cells ent separation, such as Ficoll and Percoll gradient, are
and eventually mature effector cells of the hematopoietic commonly used as a preenrichment step in stem cell purifica-
system. These mature lineages include erythroid cells for tion protocols. Progenitor cells cycle actively, whereas HSCs
oxygen transport, myeloid and lymphoid cells that provide are relatively quiescent. This difference has been exploited in
immune defense, and megakaryocytes and platelets essential techniques for HSC enrichment in mouse and human sys-
for hemostasis. tems. Treatment of mice with the antimetabolite agent fluoro-
Finally, our current view of the hematopoietic cellular uracil markedly reduces progenitor cells while relatively
hierarchy will change as we accrue more information. Our sparing populations enriched in HSC activity.
understanding of the hierarchical relationships between More recently, considerable progress has been made
populations, the plasticity of commitment and the nature of prospectively isolating HSPCs using flow cytometry and

the functional potential of populations will increase as we cell surface markers. However, it is difficult to compare dif-
better purify HSPC populations. ferent immunophenotyping strategies with respect to quan-
tifying the purity of the HSC population, as various factors
such as source of HSCs (umbilical cord versus marrow),
Phenotypic characterization and
route of transplant (intravenous versus intrafemoral), the
isolation of HSPCs
type of immunodeficient mouse used for xenografting, and
Attempts to purify stem cell populations have used a combina- pretransplant manipulation of the cells can all affect the
tion of approaches based on physical and biologic properties outcome.

Figure 13-5 schematically illustrates how HSCs (and pro- capable of fully reconstituting the lymphohematopoietic sys-
genitors) are isolated and tested for function. Hematopoietic tem in humans after myeloablative chemotherapy and radia-
tissues are isolated, cells disassociated, and then labelled with tion therapy. Although it is clear that the CD34-expressing
panels of fluorescently conjugated antibodies. The cell popu- population contains a long-term repopulating HSC, there is
lations can then be analysed and separated on a FACS sorter. some evidence of an upstream deeply quiescent CD34- HSC
Early studies to isolate human HSCs found that approxi- that gives rise to the CD34+ HSC. Approximately 5%-25% of
mately 1% of human BM cells express CD34. Isolation of CD34+ cells also express low to moderate levels of CD90.
CD34+ cells enriches for HSPCs and hematopoietic engraft- CD90 expression by human hematopoietic cells decreases
ment when transplanted into irradiated nonhuman primates. with differentiation, and most lineage-restricted progenitors
Similarly, human CD34+ selected cells contain stem cells are CD34+CD90+/low cells. Additional studies demonstrate
that human HSCs do not express mature cell lineage markers
(Lin−) or CD45RA or CD38. Isolation of Lin−CD34+CD38−
CD45RA−CD90+ cells provides a relatively easy method to
YS sort for putative HSCs based on in vitro and in vivo studies.
AGM
SSC However, this remains a heterogeneous population. Sorting
FL
BM for the integrin CD49f further enriches for HSCs. Others
FSC
Disassociate have combined some of these markers with ability of HSCs
cells
to efflux dyes (eg, the mitochondrial dye rhodamine-123).
HSCs, but not progenitor cells, express high levels of the
CD34 verapamil-sensitive multidrug-resistance membrane efflux
pump (P-glycoprotein), which confers resistance to multiple
CD38 chemotherapeutic agents. This pump also excludes certain
fluorescent dyes, such as rhodamine-123 or Hoechst-33342.
By using these dyes in combination with flow cytometry, it
CD34 CD34
has been possible to identify a population of hematopoietic
cells with low dye retention, so-called side population (SP)
CD38 CD38 cells. Although this population is markedly enriched for
In vivo
engraftment HSCs, SP cells still represent a heterogeneous mix and are not
assay
equivalent to pure HSCs. Although the SP phenotype has
NOD-SCID mouse
In vitro
been useful in characterizing HSCs (and possibly other non-
clonogenic HSCs) isolated from mice, this characteristic has not trans-
assay
lated as easily into the human system. Downstream of the
Long-term culture-initiating cells
Cobblestone area–forming assay long-term repopulating HSCs, one study has shown that
Colony assay CD133 expression differentiates different progenitor popula-
Liquid culture assay
tions, such that high CD133 defines the lymphoid-primed
Figure 13-5 Isolation of HSCs. This figure shows how HSCs can be
multipotent progenitor (LMPP) that matures into lymphoid,
isolated from different sources. Cells are initially disassociated and monocyte/macrophage, and neutrophil lines. In contrast,
stained with multiple antibodies. They are then analyzed and then CD133low/− progenitors differentiate into megakaryocytes,
sorted with a fluorescence-activated cell sorter. Here mononuclear live erythrocytes, eosinophils, and basophils (Figure 13-4).
cells are separated in gate 1. These live cells are then analysed for In mice, the immunophenotype of c-Kit+, Thy-1+, Lin−
CD34 and CD38 expression. Those live cells that are CD34+CD38- are (a cocktail of surface markers found on mature cells of dis-
enriched for stem cell potential. Further purification can be tinct lineages), and Sca-1+ (so-called KTLS cells) enriches
undertaken on the basis of additional cell surface markers such as for cells with HSC activity. Flk2 expression can be used to
CD45RA, CD90, CD49f and efflux of dyes for example Rhodamine. distinguish LT-HSCs (Flk2−) from short-term repopulating
To test the functionality of isolated (sorted) cells, cells can be tested in
HSCs (ST-HSCs; Flk2+). Other protocols have used the sig-
in vivo assays (transplanted into immuno-deficient mice such as the
nal lymphocyte activation molecule (SLAM) family recep-
NOD-SCID mouse model) and in vitro in long-term culture (long
tors CD150, CD244, and CD48 to isolate murine HSCs that
term culture initiating cell culture assay and cobblestone-area forming
assay), clonogenic colony assays and liquid culture assays. FL, fetal
are highly purified as CD150+CD244−CD48−. The number
liver; FSC, forward side scatter; SSC, side scatter. Redrawn with of murine HSCs estimated using either the KTLS/FLK2- or
permission from Hoffbrand AV, Pettit JE, Vyas P. Color Atlas of Clinical the SLAM immunophenotype is ~10 000 HSCs/mouse.
Hematology. 4th ed. Philadelphia, PA: Mosby, Ltd; 2010. © John Finally, the SLAM phenotype does not translate for isola-
Wiley & Sons Ltd. tion of human HSCs.

Though current HSPC populations are still impure, they

have allowed isolation of cell populations of defined func-
tionality and are useful to identify genes and signaling
pathways that mediate human HSPC differentiation. Isola-
tion of distinct HSPC populations is also is beginning to
permit careful dissection of the hierarchical relationships
between different blood cell populations. This is essential
in describing the cellular basis of normal hematopoiesis. In
turn, this is critical when trying to understand (i) the nor-
mal cellular compartments where genetic and epigenetic
changes initially occur in hematological diseases (ie, the
disease initiating cell populations); (ii) the cell compart-
ments where subsequent mutations/epigenetic change is
acquired during disease evolution and how this changes the
hemopoietic hierarchy; and (iii) the cell populations that
propagate hemopoietic disease. Advances in cell sorting,
genetic analysis, and other technologies are making analysis
of HSCs increasingly precise. This progress certainly will
provide additional insight into HSC biology and
heterogeneity.
In routine clinical transplantation practice, isolation of
HSCs to high purity is not necessary. Safe transplantation is
routinely possible with both unfractionated mononuclear
cells and CD34+ purified grafts (usually by magnetic beads
that enrich CD34+ cells to ~60-85% purity). Indeed, trans-
plantation of highly purified HSCs may delay engraftment,
as initial engraftment in patients is most likely from pro-
genitors rather than HSCs. Nevertheless, there may be
merit in transplanting purified HSPCs in some clinical sit-
uations—for example, to reduce transplanting contaminat- Figure 13-6 Spleen colony-forming unit (CFU-S) assay. Macroscopic
ing tumour cells in autologous transplantation. In this splenic hematopoietic colonies arising from the CFU-S stem/
regard, there are encouraging clinical studies transplanting progenitor cell 14 days after injection of murine BM into lethally
CD34+Thy1(CD90)+ cells. However, this requires a flow irradiated mice. Reproduced with permission from Williams DA.
cytometry-based isolation procedure that is difficult to Stem cell model of hematopoiesis. In: Hoffman R, Benz EJ Jr, Shattil
implement on a widespread basis. SJ, Furie B, Cohen HJ, eds. Hematology: Basic Principles and Practice.
New York, NY: Churchill Livingstone, 1995.
Stem/progenitor cell assays
A number of in vitro and in vivo assays have been devel- formed and included cells in multiple stages of differen-
oped to test HSPC function. It is useful to have back- tiation (Figure 13-7). In line with the properties observed
ground knowledge of them as the assays have defined for CFU-S, it subsequently was established that colonies
HSPC populations. generated in vitro could be initiated by the proliferation
of a single colony-forming cell (CFC). In contrast to the
self-renewal potential of most CFU-S, colonies grown in
Colony-forming assays
vitro showed more limited ability to proliferate in second-
The identification of a cell capable of in vivo clonal dif- ary cultures. Therefore, CFCs were suggested to define a
ferentiation by Till and McCulloch (1961) prompted population of committed progenitors. That is why today
other groups to develop a simple quantitative assay for the we define progenitors as those cells that can form a colony
growth and differentiation of single-cell suspensions of in an in vitro colony-forming assay. HSCs can also form
mouse BM in vitro (Figure 13-6). When hematopoietic colonies. alternatively, one cannot exclude the possibility
cells were cultured in a semisolid medium (typically soft that our current HSC populations are impure and contain
agar or methylcellulose based), discrete colonies were progenitor populations.

Figure 13-7 Examples of colony-forming assays of

human hematopoietic progenitor cells. These include
CFU erythroid (CFU-E), burst-forming unit erythroid
(BFU-E), CFU granulocyte/macrophage (CFU-GM),
and CFU granulocyte/erythroid/ macrophage/
megakaryocyte (CFU-GEMM). Reprinted with
permission from Stem Cell Technologies, Inc.
BFU-E CFU-E
CFU-GM CFU-GEMM
Long-term bone marrow culture lineages of myeloablated host mice. Human HSCs cannot be
similarly identified, however, except in a clinical study.
Attempts to develop procedures that mimic the marrow
Therefore, xenograft models commonly have been used as
microenvironment resulted in the development of long-term
another surrogate assay for human HSCs. This work origi-
BM cultures. In these assays, formation of an adherent stro-
nally involved transplanting human hematopoietic cells into
mal cell layer, which produces and deposits an extracellular
severe combined immune-deficient (SCID) mice. However,
matrix meshwork, is a prerequisite for the development and
more immunodeficient and/or radioresistant mouse strains
maintenance of hematopoietic cells. In association with the
such as NOD-SCID/IL-2Rγ−/− (commonly termed NSG or
feeder layer, hematopoietic cells proliferate and differentiate
NOG mice) or NOD/Rag1−/−/IL-2Rγ−/− (NRG) or further
over several months in culture-releasing clonogenic and
engineered to express human cytokine or other molecules
mature cells. The ongoing production of these cells is the
that aid cell survival (for example IL-3, Steel factor, GM-CSF
result of differentiation and proliferation of primitive cells.
termed NSGS mice; human thrombopoietin or human
In recognition of their method of detection, these cells have
SIRP1α) are now used for these analyses. Human cells giving
been called long-term culture-initiating cells (LTC-ICs).
long-term engraftment in these immunodeficient models
They represent primitive immature hematopoietic cells that
are considered more primitive and clearly distinct from
can be assayed in vitro. The presence of LTC-ICs can be
prior multipotent primitive human hematopoietic cell pop-
detected by assaying for the presence of CFUs in cultures
ulations identified using in vitro methodology. Other trans-
maintained for a minimum of 5 weeks. Beyond this point,
plantation models have been developed, for example, using
any CFCs (progenitor cells with shorter survival time) ini-
fetal sheep for xenografts, or nonhuman primates for autolo-
tially present in the culture should have disappeared through
gous transplantation studies (often using gene transfer into
differentiation or death, and those detected will be the result
putative HSCs). Additionally, zebrafish (where HSCs are
of differentiation by LTC-IC. LTC-ICs are not necessarily
located in the kidney and not BM) have become a well-
true HSCs, and limits of these assays make it difficult to
utilized model system of hematopoiesis. Zebrafish are ame-
know whether these cells are capable of definitive long-term
nable to medium- and high-throughput analyses and have
reconstitution and maintenance of hematopoiesis in vivo. In
been used to identify many genes and soluble factors that
vivo studies can be expensive and cumbersome, however,
regulate hematopoietic development.
and they come with their own caveats. Therefore, LTC-IC
Single-cell transplantation studies can identify clonal
studies provide a reasonable in vitro surrogate assay for early
mouse and human hematopoietic cells with the ability to
human hematopoietic cells with similarity to HSCs.
mediate long-term multilineage engraftment. Several recent
research approaches, however, have demonstrated that
Transplantation assays
HSCs are a heterogeneous cell population. Earlier studies
The definitive assay for mouse HSC activity is the ability to have used phenotypic cell surface antigens to distinguish
provide long-term (>4 months) repopulation of all blood HSCs with long-term engraftment ability (LT-HSCs) and

those that mediate just short-term engraftment (ST-HSCs). siderable efforts to identify strategies to improve develop-
More recent studies (primarily in mice), however, have ment of HSCs from hESCs are ongoing.
demonstrated that some HSCs have more myeloid engraft- iPSCs are another important cell population. Briefly, iPSCs
ment ability, and some are more lymphoid biased. These can be derived from various somatic cell populations typically
subpopulations are maintained through serial transplanta- by expression of a limited number of “reprogramming genes”
tion in the mouse. that are able to convert the somatic cell population into cells
that look and essentially act like embryonic stem cells. These
studies were first done from mouse cells in 2006 and then from
Pluripotent stem cells and hematopoiesis
human cells in 2007. Like their ESC counterparts, iPSCs have
Mouse embryonic stem cells (mESCs) were first isolated in been used to derive diverse hematopoietic cell lineages. Again, to
1981. mESCs have been proven invaluable for studies of date, transplantable HSCs have not been derived from iPSCs.
basic mammalian developmental biology, including hema- This field will continue to mature, and there is considerable
topoietic development. Unlike adult stem cells (such as interest in deriving iPSCs from individuals with different genetic
HSCs), ESCs are able to undergo self-renewal indefinitely in deficiencies to use this system as a human model of genetic dis-
culture, yet maintain the ability to form all somatic cell lin- ease. Using iPSCs, gene correction strategies or other means to
eages (including hematopoietic cells). Studies with mESCs overcome the genetic defect can be analyzed. This may lead to
have been invaluable to identify genes that regulate hemato- effective therapies based on using iPSCs as a screening resource
poietic development through gene deletion and/or manipu- and would not require direct transplantation of iPSC-derived
lation. Additionally, diverse hematopoietic cell populations cells. Additionally, future developments may allow for deriva-
can be derived from mESCs in vitro and allow for interroga- tion of iPSCs from individuals with hematologic or others dis-
tion of specific genetic and cell signaling pathways that regu- eases and use of these cells to produce essentially autologous
late development of specific hematopoietic cell lineages. replacement cell populations.
Notably, attempts to derive HSCs capable of long-term mul-
tilineage engraftment largely have failed using mESCs that
The hematopoietic niche
have not been manipulated genetically. However, overex-
pression of certain transcription factors, most notably Hematopoietic cell development from HSPCs is regulated by
HoxB4, has been shown to produce hematopoietic cells signals provided by the BM microenvironment. The specific
capable of engraftment in syngeneic recipients. Some addi- constituents of the microenvironment that influence blood
tional factors have had similar effects. cell development are being elucidated, but they can be catego-
hESCs were first described in 1998. Like mESCs, hESCs rized broadly as heterologous cells, such as mesenchymal cells;
can be maintained indefinitely as a self-renewing population endothelial and neural cells; hematopoietic cells; and extracel-
in culture, yet maintain the ability to form all somatic cell lular matrix. Mesenchymal cells include a number of cell
populations. hESCs also have been used to investigate types. Some are part of the continuum of cells that produce
human hematopoiesis. Indeed, key areas of human hemato- bone, and some are perivascular without a clear role in skeletal
poiesis are distinct from the murine system. For example, biology. Both bone-related and perivascular mesenchymal
human globin genes undergo two switching events during cells have been shown to influence hematopoiesis. For exam-
embryonic-fetal development, whereas the mouse undergoes ple, mature osteoblasts are important in stem cell mobiliza-
only one switching event. hESCs also have raised consider- tion, nestin-positive mesenchymal stem cells are important for
able interest because of the potential to use these cells to pro- HSC persistence, and adipocytes have been implicated as neg-
duce large amounts of human cells and tissues suitable for ative regulators of HSC number (Figure 13.8).
studying disease mechanism, transplantation, or transfusion The heterogeneous mesenchymal stromal cell (MSC)
medicine. For example, there has been considerable interest population plays a significant role in the hematopoietic
in using hESCs to produce red blood cells (RBCs) or platelets niche. In human studies, cord blood cocultured with MSCs
as an adjunct to the blood supply. Additionally, the potential underwent a median 30-fold expansion of CD34+ cells and
to produce HSCs from hESCs is of great interest. To date, resulted in significantly improved engraftment. In mice,
however, although most mature blood cell populations have Leptin receptor expressing (LepR+) cells represent the major
been produced from hESCs, it has not been possible to dem- proportion of MSCs. LepR+ cells appear to be the main
onstrate effective engraftment of HSCs by transplantation source of new osteoblasts and adipocytes in adult bone mar-
into immunodeficient mice to any reasonable extent. Even row and form bony ossicles supportive of hematopoiesis in
genetic manipulation and overexpression of transcription vivo. LepR+ MSCs are the major source of the cytokines stem
factors, such as HoxB4, that is effective in the murine system cell factor (SCF) and chemokine (CXC motif) ligand 12
has not been similarly effective in the human system. Con- (CXCL12) in mouse bone marrow. Conditional deletion of

Adipocytes Endothelial cells

(negatively (critical for SC localization)
regulate SC
number)
Stem cell
+
Nestin MSCs Kit ligand/kit
(regulate SC CXCL12/CXCR4
number and Ang1/Tie2
localization) Vessel VCAM/α4β1
Wnts/LRP-Frz
Jagged1/Notch1 (with PTH)
N-cadherin (?)
Osteoblastic cells
(regulate SC
number and
Leptin R+ mesenchymal cells
localization (source of Kit ligand regulating SC number)
when activated)
CXCL12+ adventitial reticular cells
(regulate SC number and localization?)
Macrophages, osteoclasts(?)
Bone matrix (regulate SC localization)
(osteopontin limits
SC number; Nonmyelinating Schwann cells
calcium receptor (regulate SC quiescence)
participates in
localization)
Sympathetic neurons
(regulate SC localization)
Figure 13-8 Bone marrow niche. HSCs localize to perivascular spaces some of which are near the endosteal surface. A number of mouse models
have been used to define specific cell types and gene products that, when manipulated, result in a change in HSC location or number. The cell
types are indicated in the figure with the HSC function that their activity appears to modulate. The molecules involved are collected in the gray
box, but which cells express these molecules is still being investigated. Some of the molecules are well defined (eg, kit ligand and CXCL12) while
others are less well defined and some (N-cadherin) quite controversial.
stem cell factor gene (Scf ) in LepR+ cells leads to depletion of process that requires expression of molecules, such as SCF
quiescent HSCs and conditional deletion of the gene encod- and members of the WNT family. The second is control of
ing CXCL12 (Cxcl12 also called Sdf1) in LepR+ cells leads to the number of stem cells, a parameter that is regulated in
HSC mobilization. part by specific extracellular matrix proteins, such as osteo-
Other cell types, such as neural cells of the sympathetic pontin, a negative regulator of HSC number. The third is the
nervous system and nonmyelinated Schwann cells, also play coordinated regulation of proliferation and differentiation
a role in HSC support or localization. The sympathetic ner- of HSCs; a process that some mouse models have indicated
vous system mediates circadian modulation in the number can go awry by changes in the niche and cause myeloprolif-
of HSCs moving from BM to bloodstream on a daily basis. erative or myelodysplastic phenotypes. The fourth is cell
Mature hematopoietic cells are also thought to influence localization, a process that is important in the context of
HSC function in the BM. Specifically, macrophages help either harvesting stem cells by mobilization into the blood or
regulate HSC mobilization into blood and T-cells are delivery of transplanted HSCs to enable engraftment.
thought to influence HSC engraftment and provide relative Thus, the HSC niche is a critical aspect of the regulated
protection from immune attack. Megakaryocytes have been production of blood cells throughout life. It is a complex tis-
shown to be important for maintaining HSC quiescence, but sue in which multiple cell types and extracellular matrix pro-
different secreted factors, including CXCL4, TGFb1, and teins contribute to balance the molecular cues that govern
thrombopoietin, have been implicated in this role. There- HSC number, self-renewal, and differentiation. By unravel-
fore, a complex admixture of cells participates in what is des- ing how stem cells enter and leave the niche, methods to
ignated as the stem cell niche. mobilize stem cells for clinical harvest have been defined (see
The niche serves several functions important for hemato- section “HSPC circulation, homing, and mobilization” in
poiesis. The first is the regulation of stem cell self-renewal, a this chapter). Ongoing efforts to improve stem cell

engraftment into the niche and to discern how the niche these TFs is shown in Figure 13-9. A thorough description of
contributes to disease are contexts in which research likely the function of these proteins is not possible here. Some of
will contribute to future manipulation of the niche for clini- the key points that arise from these studies are:
cal benefit.
1. TFs are divided into families that have similar protein
domains.
Regulation of hematopoietic differentiation
2. They often bind DNA and interact with other proteins
A complex network of transcription factor and growth factor (other TFs and proteins that control transcription) via
signaling pathways regulates HSPC self-renewal, lineage specific domains.
commitment, and differentiation. Amongst transcription 3. TFs work in combinations to both activate and repress the
factors (TFs), those that are expressed either exclusively in expression of a large number of genes.
blood cells, or have restricted tissue-specific patterns of 4. TFs are required at discrete stages of hematopoiesis, and
expression, play important roles in regulating blood produc- any one TF often functions at multiple stages within one
tion. Furthermore, acquired driver mutations of these TFs lineage and can function in more than one lineage.
are pathogenic in hematological diseases such as lymphoma 5. Ultimately, TFs work in complicated networks that can
and leukemia. The importance of these TFs is also under- be modeled much like semiconductor/computing net-
scored by the conserved role they play in hematopoiesis works. TFs work in negative feedback loops, feed-forward
through evolution. Over the last two decades, this attribute loops, and cross-antagonistic loops, to mention just three
has allowed the function of these TFs to be extensively inves- such types of interaction.
tigated in animal models. In these models, genes encoding 6. TFs help to regulate the cell’s potential to make blood cells
critical TFs have been deleted, modified, overexpressed and of different lineages, proliferate, undergo apoptosis, and
misexpressed. A summary of the site of action of some of self-renew.
Pluripotential Committed
Multipotent progenitors Mature cells
stem cells precursors
e Eosinophil
αε
n Neutrophil
Aml 1 C/EBP
tal-1/SCL Pu.1
Rbtnw/Lmo2 m Monocyte/macrophage
tel
b Basophil/mast cell
NF-E2
GATA-2 GATA-3 M
FOG GATA-1 GATA-1 Platelet
E Erythrocyte
Ikaros
Pu.1 FOG
E2A
EBF
T T lymphocyte
Pax-5
B B lymphocyte
Figure 13-9 A schematic representation of hematopoiesis and where key hematopoietic-specific transcription factors have nonredundant
functions as revealed by gene deletion studies in mice. Thus, for example, the transcription factors GATA2, AML1/RUNX1, TAL-1/SCL,
LMO2/RBTN2 and TEL are all critically required in HSCs and loss of function of these genes causes a block (as indicated by the red bar) in
hematopoietic differentiation at the HSC level. Similarly, deletions of the other transcription factors cause blocks later in hematopoiesis as
indicated by the red bars.

More specifically, the TFs SCL/TAL1, LMO2 are required Summary

to specify HSCs from mesoderm. The TFs RUNX1 (AML1),
Hematopoiesis involves a regulated set of developmental
TEL1, MLL, GATA2 are required to maintain stem cells once
stages from HSCs that produce hematopoietic progenitor
they have been specified. In myelopoiesis, the TFs Pu.1, the
cells that then differentiate into more mature hematopoietic
C/EBP family (C/EBPα and C/EBPε) GFI-1, EGR-1, and
lineages, which provide all the key functions of the hemato-
NAB2 all promote the granulocyte-macrophage lineage pro-
poietic system. Hematopoietic reconstitution during BMT is
grams. GATA2 is required in stem/early progenitor cells but
mediated by a succession of cells at various stages of develop-
is also required for mast cell differentiation and in the early
ment. Immediately following transplantation, more mature
phases of megakaryocyte-erythroid lineage maturation.
cells contribute to repopulation. With time, cells at progres-
Working with GATA2 to promote erythropoiesis and mega-
sively earlier stages of development contribute; with the final
karyopoiesis are GATA1, FOG1, SCL, EKLF, p45NF-E2, and
stable repopulation being provided by long-lived, multipo-
FlI-1. In early lymphopoiesis, the TF Ikaros is required. In B
tent HSCs. Long-term hematopoiesis is sustained by a rela-
lymphopoiesis, the TFs E2A (and its family members), EBF,
tively small number of HSCs.
and PAX5 are required and finally the TF BLIMP1 is neces-
sary for plasma cell formation. In T-cell maturation, Notch
signaling activates the TF CSL, which works with the Key points
TFs GATA3, T-BET, NFATc, and FOXP3. Of note, the TF Development
SCL/TAL1, MLL, RUNX1, LMO2, PU.1, C/EBPa, PAX5, E2A, • Hematopoiesis develops in distinct waves during development.
and GATA1 are all implicated in the pathogenesis of human • Definitive HSCs first develop within the embryo in specialized
leukemia. regions of the dorsal aorta and umbilical arteries and then seed
In addition to TFs, proteins that modulate epigenetic pro- the fetal liver and BM.
file of cells (eg, regulate DNA methylation and histone modi- • HSC characteristics differ based on their site of development
fications) and regulate splicing of RNA are also commonly and age of the organism.
mutated either through loss-of-function or gain-of-function Key features of HSCs
mutations. This also suggests that these proteins play critical • Ability, at the single-cell level, to reconstitute and maintain a
functional hematopoietic system over extended periods of time
roles in normal HSPC differentiation. Examples of the pro-
in vivo.
teins are given in Figure 13.10.
P P P
JAK2V617F PRMT5
H3Y41-P H2A/4Rme
IDH1/2
TET2
Hydroxymethylcytosine
Methylcytosine H3k4me3 H3/H4ac

HDACs
EVI1 DNMT3A HATs
H3k27me3 RUNX1 EVI1
Trithorax
EZH2 (MLL proteins)
miR-29b
Polycomb repressive
complex 2 (PRC2)
ASXL1
Figure 13-10 Epigenetic modifiers that are altered in sequence or expression in hematopoietic malignancies. Genes outlined in red represent
loss-of-function mutants, whereas those outlined in green represent gain-of-function or overexpressed genes. The genes outlined in blue
represent mutants that have acquired novel (neomorphic) function. These epigenetic modifiers may impact on DNA methylation (DNMT3A,
TET2, IDH1, IDH2), histone (H3 or H4) methylation, or histone acetylation, thereby modifying gene expression across a wide range of targets.
HAT, histone acetyltransferase; HDAC, histone deacetylase; me, methyl; ac, acetyl.

as a source of HSPCs. In allogeneic transplants, there are

Key points (continued) minimal short–medium term toxicities for healthy sibling
• Self-renewal capacity for life of organism or even after donors following PBSC mobilization. Infusion of PBSCs
transplantation. results in more rapid engraftment (because more committed
• Multipotency: the ability to make all types of blood cells. progenitors are collected) and therefore fewer early compli-
• Relative quiescence: the ability to serve as a deep reserve of cations. In the allogeneic setting, there was evidence of more
cells to replenish short-lived, rapid proliferation progenitors. frequent chronic graft-versus-host disease (GVHD) from PB
• In vivo transplantation models are currently the only reliable (due to increased number of collected T-cells) and more
assays of HSC activity.
graft failure from BM (due to lower numbers of collected
Key features of hematopoietic progenitor cells
donor T-cells and possibly lower numbers of infused
• Inability to maintain long-term hematopoiesis in vivo due to
HSPCs). In a retrospective study of children, poorer out-
limited or absent self-renewal.
• More rapid proliferation and cytokine responsiveness enabling
come was seen in patients receiving mobilized PB grafts.
increased blood cell production under conditions of stress. Prospective studies in children comparing PB and BM
• Lineage commitment and thereby limited cell-type production. HSPCs are needed urgently. In addition, in the pediatric
Key features of the HSC niche population, the donor is frequently a minor, for whom the
• Anatomically and functionally defined regulatory environment risks of granulocyte colony-stimulating factor (G-CSF) and
for stem cells. leukapheresis (the smaller the size of the patient, the more
• Modulates self-renewal, differentiation, and proliferative technically difficult the collection due to the fluid shifts
activity of stem cells, thereby regulating stem cell number. involved) must be weighed against the risk of a marrow har-
• Niche function is important in maintaining hematopoietic vest. In adult and pediatric patients receiving a transplant for
integrity and niche dysfunction may contribute to hematopoietic
nonmalignant conditions (eg, aplastic anemia) or where
disease.
rapid engraftment is not a critical issue, BM may possibly be
• Niches for HSCs are dynamic, changing during development
a superior source of stem cells due to the lower rates of acute
and with physiologic stress.
• HSCs naturally traffic into and out of the niche, a feature that
and chronic GVHD.
can be exploited for stem cell transplantation or harvesting,
respectively. HLA matching and the impact on
different HSPC sources
Clinical transplantation of HSPCs When selecting PBSCs for allogeneic transplantation, HSPCs
have to be matched to avoid the alloimmune response of
Sources of HSPCs in clinical transplantation
donor immune cells against host (GVHD) and, conversely,
This section focuses on use of different donor HSPC sources, the allo-immune reaction of the host against donor cells
key differences in the donor products, and provides a glimpse leading to graft rejection. The major genetically encoded loci
of where future advances may come from. It does not focus mediating alloimmune responses are the cell surface human
on the clinical results of transplantation. leukocyte antigens (HLAs) encoded on chromosome 6.
The requirement of all donor HSPC sources is to eradicate These encode major histocompatibility complex (MHC)
disease (when transplants are performed for hematopoietic class I (HLA-A, -B, and -C) and class II (DR, DQ and DP,
diseases) and provide life-long hematopoiesis and immu- DM and DO) antigens. MHC class I antigens are present on
nity. HSPCs can be isolated from the patient themselves (for all cells; class II antigens are also present on immune antigen
autologous transplantation) and for other humans (alloge- presenting cells. A detailed review of the biology of MHC
neic transplantation). Over the last few decades, allogeneic class I and II antigens is not possible here, and readers are
donor sources have expanded and now include HLA- referred elsewhere for this (see Bibliography). Aside from the
matched related donors, HLA-matched unrelated donors, HLA genes, there are a large number of other genes encoding
haploidentical donors, and HSPCs from umbilical cord cell surface proteins that collectively are termed minor histo-
blood (UCB). HSPCs from adults and children were initially compatibility antigens. As individual proteins, they play a
harvested from BM. Increasingly, HSPCs used in both the more modest role in an alloimmune response but collec-
autologous and allogeneic settings are isolated from periph- tively they are likely to direct both GVHD and graft versus
eral blood (PB) after administration of chemotherapy disease responses that are not completely understood.
and/or cytokines that promote circulation of HSPCs, and To identify potential HSPC donors and inform donor
they have been termed mobilized peripheral blood stem cells choice, high resolution molecular typing is replacing/has
(PBSCs). It is important to mention here that there are no replaced serotyping. Detailed national guidelines exist to guide
significant survival differences between using PB versus BM donor choice based on molecular typing (United States:

Clinical transplantation of HSPCs | 351
https://bethematchclinical.org/Transplant-Therapy-and- haploidentical donor. Parents are always haploidentical with

Donor-Matching/HLA-Typing-and-Matching/; United King- their children, and siblings have a 50% chance of being hap-
dom: http://bsbmt.org/wp-content/uploads/2012/05/Stem- loidentical. Where parents share HLA alleles or even a com-
Cell-Donor-Selection-Guidelines-V2012_13.pdf). A complete plete haplotype, the potential haplo-donor and patient are
HLA-matched sibling donor is almost always the first choice of even more closely matched.
allogeneic donor. HLA-matched sibling donor cells usually UCB cells harvested from the placenta and umbilical cord
cause less GVHD and therefore less morbidity and mortality. of newborns are cryopreserved in cord blood stem cell banks.
Furthermore, matched sibling donors often are easier logistic The advantages of UCB as a source of HSPCs are: (i) they
ally to coordinate for timing of transplantation. require less stringent matching, as UCB donation contain
Using Mendelian laws of inheritance, the likelihood that a relatively fewer T-cells and the T-cells are more immuno-
sibling pair is HLA identical would be exactly 25%. Cross- logically naïve. Thus, typically, UCB only needs to be 4 6 HLA
over phenomena during meiosis explain unusual cases of matched to the recipient. (ii) UCB can also be advantageous
aberrant recombination of HLA antigens. For this reason, when a transplant needs to be performed quickly, for exam-
the chance that two siblings are HLA matched is slightly less ple, due to aggressive disease in a patient. UCB are available
then 25%. Given family sizes and the increasing age of on average in 13.5 days in contrast to a matched unrelated
patients, medically fit sibling donors are often not available. donor HSPC, which can take 3-4 months. One of the great-
Thus, many allogeneic transplants rely on matched unre- est disadvantages of UCB as an HSPC source is the slow rate
lated donors. Such donors are matched at HLA-A, -B, C, of engraftment, prolonged myelosuppression (especially
DRB1 and DQB1 loci. Ten out of ten matches are recom- platelet recovery), and delayed immune reconstitution
mended; where this is not possible, a single mismatch at (mainly in adults), which results in higher rates of death
HLA-A, -B, -C, DRB1 and DRQ1 is acceptable (see Chapter from infection. These limitations are related to the relatively
14). Despite matching for HLAs, unrelated donors are much low HSPC dose in UCB donations. To help circumvent this
more likely to be a mismatch at minor histocompatibility problem two approaches have been tried: (i) use of two dif-
antigens. Thus, there is increased risk of GVHD and thus ferent cord blood units to provide an adequate cell dose for
increased immunosuppression is required. Conversely, the engraftment; (ii) methods to expand HSPC numbers within
increased likelihood of mismatch at minor histocompatibil- one UCB donation (see below).
ity antigens increases the chance that the donor immune sys-
tem will mount an alloimmune response against recipient
HSPC circulation, homing, and mobilization
disease cells (termed graft versus disease or graft versus leuke-
mia in patients with leukemia). HSPCs migrate from one site of blood cell production, cir-
One of the major advances in unrelated hematopoietic cell culate, home, and enter other supportive sites. This cer-
transplantation has been establishment of donor marrow tainly is the case in development before the existence of
registries. More than 20 million potential donors have been bones in the fetus. HSPCs eventually transit from fetal liver
HLA typed on donor registries. The utility of these registries to nascent BM to establish hematopoiesis via the blood-
is greatest for patients with common haplotypes that are stream. Even after BM is the site of hematopoiesis, HSPCs
found frequently within the registry. Thus, for patients with traffic into and out of the BM regularly. Experiments using
common HLA types, donors are found on a routine basis for parabiotic mice, in which the circulations of two separate
~80% of patients. However, it is still difficult to find a donor mice are joined surgically, have indicated that murine HSCs
for patients with infrequent haplotypes or with polymorphic exit the BM and transit the peripheral blood (PB) system at
HLA backgrounds, such as African Americans (probability surprisingly high flux rates (estimated to be ~104-105 long-
around 15%). term repopulating HSCs [LT-HSCs] per day in a mouse).
Because of the inability to identify a fully matched related Other mouse studies have identified macrophages and
or unrelated donor in a timely fashion, additional sources of osteoblasts as cells critical for G-CSF-mediated effects on
stem cells continue to be explored. Two sources have been HSPC trafficking through regulation of the SDF1-CXCR4
used in increasing numbers, namely haploidentical donors axis (see below).
and HSPCs from UCB. Haploidentical-related donors share This ability of HSPCs to move from the BM to the PB
one haplotype with the recipient. Advances in conditioning stream is exploited for collection of stem cells for clinical
regimes (use of cyclophosphamide post-HSPC infusion) has hematopoietic cell transplantation. The rate, timing, and
allowed safe use of haploidentical HSPCs (see section destination of the HSPCs that circulate from the BM to
“Sources of HsPCs in Clinical Transplantation” in this periphery involves chemokines and their receptors, espe-
chapter). The major advantage of haploidentical HSPCs as a cially CXCL12 (or SDF-1) and its receptor CXCR4; integrins,
cell source is that ~90% of patients will have an available particularly very late antigen 4 (VLA-4, also termed integrin

α4b1); selectins, such as P-selectin glycoprotein ligand 1 demonstrated modest increase in neutrophil recovery but
(PSGL1, also termed CD162); and the calcium- sensing did so without long-term engraftment of expanded HSCs.
receptor and the intracellular-signaling molecules, Gas and Other endothelial expressed proteins, such as angiopoietin-
Rac1/Rac2. CXCL12/SDF-1 is produced by several marrow like 5 and insulin-like growth factor binding protein 2, have
stromal cell types. These include immature osteoblasts in the been shown to expand HSCs cultured from mice. These
endosteal region adjacent to stem cell niches and the endo- agents, however, have not yet been translated into clinical
thelium. The SDF-1 receptor, CXCR4, is expressed on a wide studies.
variety of hematopoietic cells and is critical in stem cell self- Recently, considerable progress on this front has been
renewal and retention of maturing hematopoietic cells made, with at least five different novel strategies for ex vivo
within the marrow as shown by targeting of the gene in mice. expansion of HSCs, and very early progenitor cells have been
Mobilization of stem cells from the marrow to the peripheral reported. This is of relevance to expand UCB-derived
blood by G-CSF is thought to be secondary to reduction of HSC/early progenitors, as there are concerns about cell dose/
SDF-1 transcripts and proteolytic cleavage of the protein. single UCB unit, time to engraftment, and speed and depth
Finally, high cell surface expression of CD47 (CD47 is a of platelet and immune reconstitution. Agents suggested to
“don’t eat me” signal) is also likely to be important, allowing expand HSC/early progenitors and possibly promote earlier
circulating HSPCs to evade phagocytosis. engraftment include the following.
Instead of needing to collect BM as a source of HSPCs,
“mobilized” PB HSPCs can be collected by apheresis. Differ- 1. Nicotinamide. This has been used in 11 patients and there
ent preparative regimens and growth factors (eg, G-CSF) can was significantly faster engraftment with longer-term
be used to increase the number of HSPCs collected, with engraftment in 9/11 patients from the UCB unit exposed
enumeration of CD34+ cells used as a surrogate marker of to nicotinamide, suggesting that long-term HSC contri-
HSPCs. Within the CD34+ population, most cells are pro- bution was not compromised.
genitors and the frequency of true HSCs is exceedingly rare. 2. Immobilized Notch ligand Delta1. Clinical studies using a
Plerixafor, a drug that blocks the chemokine receptor double-UCB transplant with one unit expanded by cul-
CXCR4 from binding to SDF-1, is now used clinically when ture with Delta1 and the other unit unmanipulated lead
needed to increase the numbers of circulating HSPCs (given to more rapid engraftment as measured by time to abso-
with G-CSF). Ways to improve homing to the marrow lute neutrophil count. Though early engraftment was
and/or engraftment are also being studied, particularly with often provided by the manipulated unit, long-term
the use of umbilical cord blood transplants. Potential ways engraftment came from the unmanipulated UCB unit.
include ex vivo fucosylation of cells to enhance selectin These trials suggest that Notch ligand-based ex vivo
interactions; inhibition of the extracellular peptidase expansion strategies may be valuable to hasten recovery
(CD26), which cleaves CXCL12/SDF-1; or pretreatment of of peripheral cell recovery, but HSCs may be lost in the
donor cells with modified prostaglandin E2 to expand HSCs, expanded cell population.
but also to improve homing to the marrow. Thus, studies of 3. Prostaglandin E2 (PGE2). Originally identified in a
stem cell niche interactions may ultimately affect clinical zebrafish screen for agents that increased HSCs, PGE2 (or
medicine, reducing the numbers of stem cells needed for more biologically stable derivatives of PGE2) increase
transplantation through more efficient mobilization, hom- CD34+ cells and expand HSCs as measured using in vivo
ing, and engraftment. transplantation assays. Clinical trials using double-UCB
units are ongoing. Much of the increase in HSC engraft-
ment due to PGE2 treatment may be from increase
Ex vivo expansion of HSCs
expression of adhesion molecules on the HSCs leading to
Because of the limited number of HSPCs available from improved trafficking to the hematopoietic niche.
some sources, there has been great interest over the years in 4. Arylhydrocarbon receptor antagonists. They expand
defining conditions that allow ex vivo expansion of HSPCs. putative HSCs from humans and nonhuman primates
Though it has been possible to expand progenitors, a key but not mice (showing that preclinical studies for some
problem has been to increase the number of HSCs without agents may need to be done in more than one model).
sacrificing “stemness” including long-lived self-renewal and This agent is now in clinical trials.
multipotentiality. Previous studies demonstrated that com- 5. UM171. A pyrimidoindole identified in a screen for
binations of hematopoietic cytokines (such as SCF, Flt chemicals that expand HSC numbers without loss of HSC
ligand, thrombopoietin, or G-CSF), with or without stromal function. Clinical studies will begin shortly. Future clini-
support cells, could provide modest expansion of putative cal translation may benefit from a combination of these
HSCs in culture. Clinical trials using these strategies typically and other approaches.

Bibliography | 353
Number of stem cells required a malignant condition). The failure to provide adequate
for engraftment myeloid and red cell production is not the major problem
(other than for cord blood transplants). In the allogeneic set-
Normal hematopoiesis is characterized by contribution from
ting, a variety of donor sources are now available. Typically,
many HSPCs, each contributing a small proportion of the
matched sibling donors are preferred, but, given the small
blood cells present at any given time. In contrast, the hemato-
family sizes and older age of patients, alternative donor sources
poietic system regenerated in animals that have undergone
are increasingly being used. The advantages and disadvantages
transplantation is, after an initial period of clonal disequilib-
of these donor sources are summarized in the Key Points.
rium, characterized by the contribution of a relatively few stem
cells. Monoclonal hematopoiesis also has been documented in
humans after allogeneic hematopoietic cell transplantation. Key points
Experiments in which mice were injected with varying num- Advantages of a sibling donor
bers of unseparated marrow cells or purified stem cells, how- • Better match at HLA and minor histocompatibility loci;
ever, have shown that the number of clones contributing to reduced chance of GVHD compared to unrelated donor cell
hematopoiesis increases in proportion to the number of stem sources.
cells injected. Limiting numbers of repopulating stem cells in a • Donor cells can potentially be obtained quickly.
graft may force such cells to exhibit a larger proliferative • Potentially, donor can be easily approached again for
response than stem cells in a transplantation that exceeds min- additional HSPCs to improve graft function or for T-cells for
donor lymphocyte infusion to augment alloimmune response
imal requirements. Although it may be possible to regenerate
against recipient disease cells.
a functioning immunohematopoietic system after transplan-
Disadvantages of a sibling donor
tation of relatively few stem cells, it is less likely that these cells
• Potentially less graft-versus-disease immune response.
are capable of sustaining hematopoiesis for the life span of the Advantages of an unrelated donor
patient without periodic replacement. • Increased graft versus disease compared to sibling donor.
Therefore, although in clinical transplantation the adage • Donors available for common haplotypes.
appears to be “the more stem/early progenitor cells, the Disadvantages of an unrelated donor
better,” the practical questions are, “How many stem/early • Increased graft versus host response compared to sibling
progenitor cells are enough?” and “How are such cells mea- donor.
sured?” In practice, mobilization of stem/early progenitor • Potentially more difficult to obtain additional HSPCs and
cells can be monitored by the total number of nucleated cells, donor lymphocytes for graft failure, and increased alloimmune
CD34+ cells, and in some centers by CFU-granulocyte mac- response against recipient disease cells.
Advantages of a haploidentical donor
rophages (CFU-GM) to optimize harvesting time and
• Donor cells commonly available.
achieve maximum yield. Higher CD34+ cell doses from BM
• Increased graft versus disease.
or mobilized PB cells (typically >2 × 106 CD34+ cells/kg of
• Potentially, donor can be easily approached again for
recipient body weight) are associated with improved survival additional HSPCs to improve graft function or for T-cells for
especially in unrelated transplants. CD34+ cell doses of >8 × donor lymphocyte infusion to augment alloimmune response
106/kg of body weight are associated with increased GVHD against recipient disease cells.
where PB HSPCs are used. For UCB transplantation, the Disadvantages of a haploidentical donor
minimum cell dose required to reliably achieve engraftment • More immunosuppression required given the greater HLA
is >2.5 × 107 total nucleated cells/kg or >1.7 × 105 CD34+ mismatch.
cells/kg. However, the greater the degree of HLA mismatch- Advantages of cord blood
ing, the higher the cell dose needed, such that for a 5/6 HLA • Donor cells available when other cell sources not available.
match many transplant centers use >4 × 107 nucleated • Donor cells quickly available.
Disadvantages of cord blood
cells/kg body weight, and >5 × 107 nucleated cells/kg body
• Longer time to engraftment.
weight for a 4/6 HLA match.
• Longer time for immune reconstitution.
Summary
Bibliography
Donor HSPCs provide both daily blood cell production and
long-lasting immunity. The major issues facing transplanta- Cellular basis of hematopoiesis
tion are mainly to better understand and harness the immune Becker, AJ, McCulloch EA, Till JE. Cytological demonstration of
response to provide long-lasting protective immunity against the clonal nature of spleen colonies derived from transplanted
infection and against malignant disease (in those patients with mouse marrow cells. Nature. 1963;197:452-454.

Bonig H, Papayannopoulou T. Mobilization of hematopoietic granulocyte subtypes derive from distinct hematopoietic
stem/progenitor cells: general principles and molecular lineages. Cell Rep. 2013;3:1539-1552.
mechanisms. Methods Mol Biol. 2012;904:1-14. An excellent Kiel MJ, Yilmaz OH, Iwashita T, Yilmaz OH, Terhorst C,
review of the basis for stem cell mobilization. Morrison SJ. SLAM family receptors distinguish hematopoietic
Méndez-Ferrer S, Lucas D, Battista M, Frenette PS. Hematopoietic stem and progenitor cells and reveal endothelial niches for stem
stem cell release is governed by circadian oscillations. Nature. cells. Cell. 2005;121:1109-1121.
2008; 452:442-447. Demonstration of how niche function involves Notta F, Doulatov S, Laurenti E, Poeppl A, Jurisica I, Dick JE.
multiple cell types, including neurons of the sympathetic nervous Isolation of single human hematopoietic stem cells capable of
system. long-term multilineage engraftment. Science. 2011;333:218-221.
Orkin SH, Zon LI. Hematopoiesis: an evolving paradigm of stem Spangrude GJ, Heimfeld S, Weissman IL. Purification and
cell biology. Cell. 2008;132:631-644. An excellent review of characterization of mouse hematopoietic stem cells. Science.
concepts in hematopoiesis. 1988;241:58-62.
Park D, Sykes DB, Scadden DT. The hematopoietic stem cell Wagers AJ, Sherwood RI, Christensen JL, Weissman IL. Little
niche. Front Biosci. 2012:17:30-39. A recent update on the stem evidence for developmental plasticity of adult hematopoietic
cell niche. stem cells. Science. 2002;297:2256-2259.
Purton LE, Scadden DT. Limiting factors in murine hematopoietic
stem cell assays. Cell Stem Cell. 2007;1:263-270.
Sutherland HJ, Lansdorp PM, Henkelman DH, Eaves AC, Stem cell sources, ex vivo expansion of HSCs,
Eaves CJ. Functional characterization of individual human pluripotent stem cells, and hematopoiesis
hematopoietic stem cells cultured at limiting dilution on
Anasetti C, Logan BR, Lee SJ, et al., and the Blood and Bone
supportive marrow stromal layers. Proc Natl Acad Sci USA.
Marrow Transplant Clinical Trials Network. Peripheral-blood
1990;87:3584-3588.
stem cells versus bone marrow from unrelated donors. N Engl J
Till JE, McCulloch EA. A direct measurement of the radiation
Med. 2012;364:1487-1496.
sensitivity of normal mouse bone marrow cells. Radiat Res.
Barker JN, Byam C, Scaradavou A. How I treat: the selection
1961;14:213-222. The first example of a quantitative assay for
and acquisition of unrelated cord blood grafts. Blood.
hematopoietic stem cells; beginning of stem cell research.
2011;117:2332-2339.
Boitano AE, Wang J, Romeo R, et al. Aryl hydrocarbon receptor
Ontogeny of hematopoiesis antagonists promote the expansion of human hematopoietic
stem cells. Science. 2010;329:1345-1348.
Bertrand JY, Chi NC, Santoso B, Teng S, Stainier DY, Traver Brunstein CG, Fuchs EJ, Carter SL, et al. Alternative donor
D. Hematopoietic stem cells derive directly from aortic transplantation after reduced intensity conditioning: results of
endothelium during development. Nature. 2010;464:108-111. parallel phase 2 trials using partially HLA-mismatched related
Chen MJ, Yokomizo T, Zeigler BM, Dzierzak E, Speck NA. Runx1 bone marrow or unrelated double umbilical cord blood grafts.
is required for the endothelial to hematopoietic cell transition Blood. 2011;118:282-288.
but not thereafter. Nature. 2009;457:887-891. Brunstein CG, Gutman JA, Weisdorf DJ, et al. Allogeneic
Dzierzak E, Speck NA. Of lineage and legacy: the development hematopoietic cell transplantation for hematologic malignancy:
of mammalian hematopoietic stem cells. Nat Immunol. relative risks and benefits of double umbilical cord blood. Blood.
2008;9:129-136. 2010;116:4693-4699.
North TE, Zon LI. Modeling human hematopoietic and Delaney C, Heimfeld S, Brashem-Stein C, Voorhies H, Manger RL,
cardiovascular diseases in zebrafish. Dev Dyn. 2003;228:568-583. Bernstein ID. Notch-mediated expansion of human cord blood
progenitor cells capable of rapid myeloid reconstitution. Nat
Med. 2010;16:232-236.
Stem cell characteristics
Evans MJ, Kaufman MH. Establishment in culture of
Baum CM, Weissman IL, Tsukamoto AS, Buckle AM, Peault pluripotential cells from mouse embryos. Nature. 1981;292:
B. Isolation of a candidate human hematopoietic stem-cell 154-156.
population. Proc Natl Acad Sci USA. 1992;89:2804-2808. Goodell MA, Brose K, Paradis G, Conner AS, Mulligan RC.
Copley MR, Beer PA, Eaves CJ. Hematopoietic stem cell Isolation and functional properties of murine hematopoietic stem
heterogeneity takes center stage. Cell Stem Cell. 2012; cells that are replicating in vivo. J Exp Med. 1996;183:1797-1806.
10:690-697. Kaufman DS. Toward clinical therapies using hematopoietic
Goodell MA, Brose K, Paradis G, Conner AS, Mulligan RC. cells derived from human pluripotent stem cells. Blood.
Isolation and functional properties of murine hematopoietic 2009;114:3513-3523.
stem cells that are replicating in vivo. J Exp Med. 1996;183: Kaufman DS, Hanson ET, Lewis RL, Auerbach R, Thomson JA.
1797-1806. Hematopoietic colony-forming cells derived from human
Görgens A, Radtke S, Möllmann M, Cross M, Dürig J, Horn embryonic stem cells. Proc Natl Acad Sci USA. 2001;98:
PA, Giebel B. Revision of the human hematopoietic tree: 10716-10721.

Bibliography | 355
Kyba M, Perlingeiro RC, Daley GQ. HoxB4 confers definitive Takahashi K, Tanabe K, Ohnuki M, et al. Induction of pluripotent
lymphoid-myeloid engraftment potential on embryonic stem stem cells from adult human fibroblasts by defined factors. Cell.
cell and yolk sac hematopoietic progenitors. Cell. 2002;109:29-37. 2007;131:861-872.
Martin GR. Isolation of a pluripotent cell line from early mouse Takahashi K, Yamanaka S. Induction of pluripotent stem cells
embryos cultured in medium conditioned by teratocarcinoma from mouse embryonic and adult fibroblast cultures by defined
stem cells. Proc Natl Acad Sci USA. 1981;78:7634–7638. factors. Cell. 2006;126:663-676.
North TE, Goessling W, Walkley CR, et al. Prostaglandin E2 Thomson JA, Itskovitz-Eldor J, Shapiro SS, et al. Embryonic
regulates vertebrate hematopoietic stem cell homeostasis. stem cell lines derived from human blastocysts. Science.
Nature. 2007;447:1007-1011. 1998;282:1145-1147.
Rocha V, Labopin M, Sanz G, Arcese W, Schwerdtfeger R, Bosi A, Wagner JE, Barker JN, DeFor TE, et al. Transplantation of
Jacobsen N, Ruutu T, de Lima M, Finke J, Frassoni F, Gluckman unrelated donor umbilical cord blood in 102 patients with
E; Acute Leukemia Working Party of European Blood and malignant and nonmalignant diseases: influence of CD34 cell
Marrow Transplant Group; Eurocord-Netcord Registry. N Engl dose and HLA disparity on treatment-related mortality and
J Med. 2004;351:2276-2285. survival. Blood. 2002;100:1611-1618.
Stadtfeld M, Hochedlinger K. Induced pluripotency: history, Yu J, Vodyanik MA, Smuga-Otto K, et al. Induced pluripotent
mechanisms, and applications. Genes & Development. stem cell lines derived from human somatic cells. Science.
2010;24:2239–2263. 2007;318:1917–1920.
Shultz LD, Ishikawa F, Greiner DL. Humanized mice in Zanjani ED. The human sheep xenograft model for the study of
translational biomedical research. Nat Rev Immunol. the in vivo potential of human HSC and in utero gene transfer.
2007;7:118-130. Stem Cells. 2000;18:151.

CHAPTER
14
Clinical hematopoietic cell
transplantation
Matthew Wieduwilt and Sergio A. Giralt
Historical perspective, 357 Conditioning regimens, 362 Late effects, 372
The hematopoietic stem cell, 357 Supportive care post-HCT, 364 Hematopoietic cell transplantation for
The hematopoietic cell transplant HCT complications, 364 specific diseases, 374
process, 358 Specific organ toxicities, 367 Summary, 388
Stem cell sources and procurement, 361 Graft-versus-host disease, 369 Bibliography, 388
Historical perspective (BMT) could be a reality for sufficient numbers of patients

to warrant its large-scale study. The landmark paper from
The advent of the atomic era and the potential for large-scale Thomas et al. (1979), demonstrating that long-term disease
human exposure to ionizing radiation either accidentally or control could be achieved in patients with refractory acute
intentionally resulted in a dramatic increase in basic and leukemias with the use of high-dose chemo-radiotherapy
preclinical research in hematopoiesis and hematopoietic cell followed by infusion of HLA identical sibling marrow, marks
transplantation (HCT) as a therapeutic strategy against the beginning of HCT.
exposure to lethal radiation. The following seminal observa- The rationale for high-dose cytotoxic chemotherapy stems
tions were required to develop the field: from the steep dose–response curve of alkylating agents and
radiotherapy and tumor cell response in human tumors.
1. Safety and feasibility of human bone marrow infusion Doubling the dose of alkylating agents increases tumor cell
2. Ability of normal stem cells to reconstitute a lethally radi- kill by a log or more, and increasing the dose of alkylating
ated host agents by five- to tenfold overcomes the resistance of tumor
3. Recognition of a potential graft-versus-tumor (GVT) cells against lower doses. High-dose chemotherapy aims to
effect operative in animal models and humans destroy the tumor cells in an expediently timely manner to
4. Safety and feasibility of cryopreserved autologous bone prevent the emergence of resistant clones. In 1978, investiga-
marrow in reconstituting lethally radiated hosts tors from the National Cancer Institute were the first to
Notwithstanding these initial observations, the initial report the use of high-dose chemotherapy followed by autol-
clinical experience with HCT was dismal, with most patients ogous BMT for patients with relapsed lymphoma. These
succumbing to transplant-related complications. It was not encouraging results were the initial clinical evidence that led
until the discovery and identification of human leukocyte to the widespread application of autologous HCT. Drs.
antigens (HLAs) as well as improvements in supportive care Mcelwain and Powles demonstrated a similar dose response
with the development of antibiotics and antifungals that suc- curve for melphalan in patients with myeloma, which led to
cessful HCT in the form of bone marrow transplantation the beginning of high-dose therapy for myeloma that is now
the most common indications for autologous HCT.
Conflict-of-interest disclosure: Dr. Giralt: Consultancy: Celgene;

Research Funding: Celgene; Sanofi-Aventis. Honoraria: Celgene; Mil- The hematopoietic stem cell
lennium; Sanofi-Aventis; Speakers Bureau: Sanofi-Aventis. Dr. Wiedu-
wilt: Research funding: Sigma Tau.
Off-label drug use: Dr. Giralt and Dr. Wieduwilt: Drug therapy for Hematopoietic progenitor cells (HPCs) are capable of reconstitu
hematologic malignancies and hematopoietic cell transplantation. ting and maintaining a complete and functional hematopoietic
| 357

358 | Clinical hematopoietic cell transplantation
system over extended periods of time. They are characterized these processes currently is not well understood but appears
by three intrinsic properties: extensive proliferative capacity, to involve lectins, integrin adhesion molecules, chemokines,
pluripotency (the ability to differentiate into all blood cell and their receptors. The ability to alter these interactions
types), and self-renewal capacity (the ability to replace the with such agents as granulocyte colony-stimulating factor
cells that became progressively committed to differentia- (G-CSF) or CXCR4 antagonists allows for “mobilization” of
tion). The HPC has been characterized by the following: HSCs into the peripheral blood system and their collection
by apheresis for HPC transplantation.
1. Ability to form multilineage colonies in semisolid soft
agar medium
2. Ability to form colony-forming units (CFUs) after being
The hematopoietic cell transplant
maintained in culture for a minimum of 5 weeks
process
3. Ability to provide long-term (>4 months) repopulation
of all blood lineages of myeloablated host mice
The HCT process is a complex procedure in which an indi-
HPCs account for 1 in 10,000 bone marrow cells, and during vidual receives a combination of chemical and physical
normal steady-state hematopoiesis, are in the G0 phase. agents to eliminate a malignant disorder or a poorly func-
Through a series of chemical signals, they are recruited into tioning bone marrow supported by reinfusion of HSC from
active hematopoiesis and undergo a series of maturational the patient or a donor. Intense medical therapy is required as
cell divisions that culminate in the generation of progenitor patients recover from the effects of the conditioning regimen
cells that have progressively limited self-renewal, are prolif- and throughout the period of profound immune-suppression
erative, and have the potential to differentiate into different that occurs while the transplanted HPC mature and recover
cell types. normal function. The components of the HCT process are
Hematopoietic cells develop in vivo in intimate associa- listed below.
tion with a heterogeneous population of stromal cells and an
extracellular matrix that constitute the microenvironment of
HCT recipient
the bone marrow. Fibroblasts, smooth muscle cells, adipo-
cytes, osteogenic cells, and macrophages compose the stro- HCT is performed not only for patients with a variety of
mal cell compartment. Extracellular matrix molecules of hematologic malignancies but also for patients with nonma-
seven distinct families have been identified, including colla- lignant hematologic disorders. The most common indica-
gens, proteoglycans, fibronectin, tenascin, thrombospondin, tions for HCT are summarized in Table 14-1.
laminin, and hemonectin. Within the marrow microenvi- Having a condition that could be amenable to treatment
ronment, two types of niches have been described that favor with HCT is not enough to deem a patient transplant eligible.
HPC self-renewal versus differentiation. The osteoblastic Transplant eligibility is determined by an extensive pre-
niche is located in the periosteal region of the bone cavity; transplant evaluation of organ function and psychosocial
and the vascular niche involves vascular sinusoids within the evaluation to assess the risk-benefit ratio of HCT compared
bone marrow. A complex network of transcription factor with alternative treatment approaches. Table 14-2 summa-
and growth factor signaling pathways tightly regulates HPC rizes the most commonly used criteria to determine HCT
recruitment, lineage commitment, and differentiation. The eligibility.
advent of flow cytometric techniques has allowed cell surface
markers to be used to prospectively isolate cell populations
Types of hematopoietic cell transplants
with selective potentials.
Stem cells migrate from one site of blood cell production, HCT traditionally has been classified according to the source
circulate, home, and enter other supportive sites. Control of of stem cells as either autologous or allogeneic.
Table 14-1 Most common indications for HCT
Autologous HCT Allogeneic HCT

No. of procedures performed No. of procedures performed
Diagnosis in the United States in 2009 Diagnosis in the United States in 2009
Myeloma 4,700 Acute myeloid leukemia 2,500

Non-Hodgkin lymphoma 2,800 Acute lymphoblastic leukemia 1,100
Hodgkin lymphoma 1,250 Myelodysplastic syndrome 1,000

The hematopoietic cell transplant process | 359
Table 14-2 Commonly used eligibility criteria for HCT
Eligibility criteria Test Transplant eligible Comments
Patient History and physical exam Usually ECOG 2 or less, Transplant related mortality is related to pre-HCT
performance Karnofsky performance status performance status. Patients with poor performance
status of >70% status generally are not considered candidates for
HCT unless chances for long-term cure are high if
HCT is successful.
Disease and Multiple Depending on disease and Patients with advanced refractory disease are generally
disease status disease status. Patients with not considered transplant eligible. Armand et al.
high-risk disease and high- (2012) recently proposed a disease and disease status
risk disease status have <10% risk classification for stem cell transplantation.
chance of 2-year survival
Infectious disease Serologies for hepatitis A, Generally, patients should not Guidelines may be changing with the advent of effective
markers B, and C. PCR for viral have documentation of active antiviral therapy (entecavir and highly active
copies HIV, HTLV-1 viral replication antiretroviral therapy). Prior hepatitis exposure does
CMV, EBV, toxoplasmosis not affect transplant outcomes.
Cardiac function Echocardiogram Ejection fraction >40% Patients with cardiac history may require more
Nuclear medicine testing No uncontrolled cardiac disease extensive pretransplant evaluation, including referral
to cardiology for stress testing or Holter monitoring.
Pulmonary Pulmonary function DLCO >40% In some series, the most important predictor of
function testing outcome is DLCO <40%.
Renal function Creatinine and creatinine Usually creatinine clearance >40 Patients with poor renal function (including
clearance cc/min patients with ESRD) can be considered for HCT on
a case-by-case basis. Autologous HCT is routinely
performed in patients with multiple myeloma on
dialysis.
Hepatic function Liver function tests Bilirubin <2-3 × ULN unless Elevated liver function tests predict liver toxicity.
(transaminases and history of Gilbert disease
total bilirubin)
Comorbidity Hematopoietic specific No cutoff determined Comorbidity scoring more useful to decide regimen
scoring comorbidity index intensity. HCT-CI most commonly used scoring
system.
Psychosocial Various Varies with institutional Essential to determine risk of noncompliance as well
evaluation guidelines as caregiver availability and social support needed
throughout the transplant process.
ECOG = Eastern Cooperative Oncology Group; HCT = stem cell transplantation; PCR = polymerase chain reaction; HIV = human
immunodeficiency virus; HTLV-1 = human T-lymphotropic virus 1; CMV = cytomegalovirus; EBV = Epstein-Barr virus; DLCO = diffusion
lung capacity; ESRD = end-stage renal disease; ULN = upper limit of normal; HCT-CI = hematopoietic cellular therapy-comorbidity index.
Autologous HCT that affect the marrow or are difficult to cure with chemo-
therapy alone (eg, severe aplastic anemia, acute and chronic
Autologous HCT involves using HPCs obtained from the
leukemia, genetic disorders) will require a third-party stem
same recipient. The stem cells can be obtained directly from
cell source to rescue the patient from the effects of the con-
the recipient’s marrow or through mobilization of stem cells
ditioning regimen and provide the vehicle for immunother-
from the marrow into the peripheral blood using G-CSF, often
apy (donor lymphocytes). In diseases in which a steep dose
with chemotherapy, or the CXCR4 antagonist plerixafor.
response to alkylating agents is observed and the role of a
GVT effect is less certain (eg, lymphoma, myeloma, and
germ cell tumor) the use of autologous stem cells is gener-
Allogeneic HCT
ally preferred, at least for first transplant. The advent of
Allogeneic HCT involves using HPCs obtained from a third gene therapy for some hemoglobinopathies and the use of
party who can be a related or unrelated donor. chimeric antigen receptor gene modified T-cells may change
The choice between performing an autologous versus an some of these traditional uses of allogeneic HCT over autol-
allogeneic HCTs can be a difficult one. In general, diseases ogous HCT.

Human leukocyte antigen typing For the majority of patients who lack a matched related
donor, an HLA-identical unrelated donor represents an
Determination of HLA types has become much more accu-
alternative stem cell source. Millions of potential donors
rate as typing has become molecularly based, replacing the
have been HLA typed and are listed with national and inter-
earlier serologic or cellular techniques. Modern HLA typing
national registries. The utility of these registries has been
relies on molecular techniques, such as polymerase chain
greatest for patients with common haplotypes for whom,
reaction (PCR) amplification of the test DNA followed by
because of linkage disequilibrium, the haplotype is found
probing with labeled short sequence-specific oligonucleotide
frequently within the registry. Thus, for patients with com-
probes or, more recently, sequencing of the major histocom-
mon HLA types, it is now possible to find donors on a rou-
patibility complex (MHC) class I and class II alleles. By con-
tine basis. It is still difficult, however, to find a donor for
vention, differences recognized by serologic typing are called
patients with infrequent haplotypes or for patients with
antigen mismatches, and differences recognized only by
polymorphic HLA backgrounds, such as African Americans
molecular techniques are called allele mismatches.
or those of mixed race.
The ideal donor is identified according to HLA compati-
bility as determined by HLA typing. The MHC refers to the
entire genetic region containing the genes encoding tissue Cord blood transplantation and haploidentical
antigens. In humans, the MHC region lies on the short arm donor transplantation
of chromosome 6 and is designated the HLA region. The
Because of the inability to identify a fully matched related or
HLA region is a relatively large section of chromosome 6
unrelated donor in a timely fashion, additional sources of
with many genes, not all of which are involved in immune
stem cells have been explored. It is estimated that a quarter
responses. The HLA region has been divided into class I,
to a third of patients in need of an unrelated donor will not
class II, and class III regions, each containing numerous gene
be able to find a match. Therefore, umbilical cord blood
loci that may encode a large number of polymorphic alleles.
(UCB) cells harvested from the umbilical cord of newborns
Class I antigens are composed of two chains: a heavy chain
represent a rapidly growing source of HSCs. UCB contains
containing the polymorphic region that combines with the
hematopoietic progenitors capable of hematopoietic recon-
nonpolymorphic light chain, b2-microglobulin, to form the
stitution, can be obtained within a short time span (available
final molecule. The class I HLA antigens include HLA-A, -B,
on average in 13.5 days in contrast to a matched unrelated
and -C antigens and are expressed on almost all cells of the
donor search of 3-4 months), and demonstrates less alloge-
body at varying densities. Class II antigens are composed of
neic reactivity responsible for graft-versus-host disease
two polymorphic chains: an α-chain and a β-chain (both
(GVHD) compared with marrow or peripheral blood grafts.
encoded on chromosome 6). Both chains of class II antigens
Because of the relative immaturity of the newborn immune
are encoded in the MHC. The class II antigens are further
system, cord blood transplantation can be performed with a
divided into DR, DQ, and DP antigens. Class II antigens are
relative low incidence of GVHD even with two and three
expressed on B-cells and monocytes and can be induced on
HLA antigen mismatches.
many other cell types following inflammation or injury. The
The greatest limitations of UCB transplantation are slow
DQ and DP antigens each have polymorphic α- and β-chains,
engraftment with prolonged cytopenias and delayed immune
which can dimerize in various combinations.
reconstitution that results in higher rates of death from
infection. All of these limitations are related to the relatively
low progenitor cell dose in cord blood units. This low pro-
Matched related and unrelated donors
genitor cell dose has hampered the ability to obtain rapid
Inheritance of HLA antigens is determined by Mendelian engraftment in patients who weigh >50 kg.
genetics with coexpression of the maternal and paternal alleles; A number of trials have been performed to determine the
the likelihood of siblings sharing both HLA haplotypes (ie, a feasibility of UCB transplantation for adults; engraftment
particular sequence of HLA-A, -B, -C, -DR, -DQ, and -DP on generally is slow with a median time to neutrophil engraft-
chromosome 6) is approximately 25%, and the chances of ment of 3-4 weeks and up to 10% of patients failing to
finding a sibling donor increases with the number of siblings engraft. A more recent analysis from the International Bone
in the family. Parents share one HLA haplotype with their off- Marrow Transplant Registry (IBMTR) showed that the
spring and are considered haploidentical. Certain HLA anti- results with UCB transplantation were equivalent to those of
gens commonly occur in association with one another, a mismatched unrelated-donor transplantation but were infe-
phenomenon called linkage disequilibrium. This limits the rior to matched unrelated–donor transplantation. Attempts
number of potential HLA haplotypes that occur and allows for to speed engraftment have included the use of two cord
the development of large volunteer donor registries. products in a single patient, supplementation with CD34

Stem cell sources and procurement | 361
selected cells from a related donor, and the use of ex vivo important predictor of outcome with patients receiving
partially expanded UCB products. Of particular interest is larger stem cell dose having more rapid engraftment, reduced
the observation that cord blood transplantation may be nonrelapse mortality (NRM), and improved survival.
associated with lower rates of relapse than other products,
potentially because of the co-infusion of maternal cells.
Peripheral blood
Another potential source of stem cells for patients without
an HLA identical donor within their families or from volunteer The observation that peripheral blood contained low levels
donor registries are mismatched family members. Donors that of circulating hematopoietic pluripotent progenitor cells was
share one haplotype with the recipient are called haploidenti- made initially in the 1970s, but it was not until the cloning
cal. Parents are always haploidentical with their children, and and development of colony stimulating factors that large
siblings have a 50% chance of being haploidentical with each numbers of circulating stem cells could be mobilized into
other. It is estimated that approximately 90% of patients will peripheral blood and peripheral blood HCT was made fea-
have an available haploidentical donor. The major challenges sible for large-scale study and use.
associated with haploidentical transplantation are the high rate Under steady-state conditions, most HSCs reside in the
of severe acute GVHD and delayed immune reconstitution. marrow, and various strategies have been developed to mobi-
Strategies to ameliorate GVHD include T-cell depletion both lize them. This includes single-agent cytokine, cytokine com-
through in vivo and ex vivo means, novel immunosuppressive binations, and combinations of chemotherapy with cytokines
combinations, and posttransplantation chemotherapy with followed by collection of peripheral blood leukocytes with
either cyclophosphamide or bortezomib. Graft failure has been leukapheresis. HSC concentration in the bloodstream usually
reduced with the use of large doses of stem cells and intensified peaks 4-6 days after initiation of therapy with cytokines
conditioning regimens. alone. When chemotherapy with cytokines is given, maxi-
Retrospective comparisons of transplant outcomes have mum recovery of stem cells in the blood occurs at the time of
shown similar results for recipients of haploidentical and marrow recovery. Collection usually is initiated when the
cord blood transplants. Trials currently are under way to white blood cell (WBC) count recovers to >1 × 109L WBC/L.
determine whether there is an optimal alternative stem cell To improve the accuracy and efficacy of stem cell collections,
source for patients lacking an HLA-compatible donor within daily measurement of peripheral blood CD34+ cell content
their family or the unrelated donor registries. has been used, and many centers initiate HSC collection
when CD34+ cell count exceeds 5-10 cells/mL.
Peripheral blood progenitor cells have almost completely
Stem cell sources and procurement replaced bone marrow as the HSC source for patients under-
going autologous HCT because of the less-invasive collec-
HPCs reside primarily in the bone marrow but circulate in tion method and more rapid blood count recovery. The
the peripheral blood at low levels. Chemotherapy, colony- more rapid recovery is thought to be due to higher stem cell
stimulating factors, and the CXCR4 inhibitor plerixafor can doses infused with PBSCs. In the autograft, increasing stem
mobilize large quantities of HSCs into the peripheral blood cell dose is associated with more rapid platelet and neutro-
that subsequently can be collected. Initially, allogeneic mar- phil recovery when stem cell doses of between 2 and 10 mil-
row always was infused fresh after collection with minimal lion CD34+ cells/kg are analyzed.
manipulation (filtering of fat globules and bone particles, Despite the use of chemotherapy–cytokine combination
plasma or red cell reduction depending on ABO incompati- regimens, mobilization failure still occurs in some patients.
bility). With the advent of the cryopreservation agent Prior treatment is the single most important factor affecting
dimethyl-sulfoxide (DMSO), cryopreservation of autolo- stem cells yields. Prior treatment with stem cell toxins, short
gous marrow became feasible and was adopted rapidly for interval since last chemotherapy, previous radiation, hypo-
autologous marrow and peripheral blood stem cell (PBSC) cellular marrow, malignancies involving the bone marrow,
harvesting and to a lesser degree for cryopreservation of allo- and refractory disease have been associated with poor mobi-
geneic marrow or PBSC. lization. This underscores the importance of referring a
potential transplantation candidate early for transplantation
evaluation before repeated salvage attempts that may
Bone marrow
adversely affect stem cell collections.
HSCs initially were obtained exclusively from the marrow A novel chemokine, plerixafor (Mozobil; Genzyme, Cam-
cavity under anesthesia with multiple aspirations by a proce- bridge, MA), was approved in 2009 as a mobilization agent
dure initially described in the 1950s. In the setting of marrow in combination with G-CSF. Plerixafor, formerly known as
transplantation, stem cell dose has been identified as an AMD3100, a bicyclam derivative, is a specific antagonist of

CXCR4, a coreceptor for the entry of HIV into host cells and progenitor cells (CFU-GM) assay. They were also the first to
initially was developed as a potential therapeutic agent for find that procedures to remove erythrocytes or granulocytes
HIV. In a phase 1 study, it induced modest leukocytosis when before freezing and washing of thawed cells before plating
administered intravenously to HIV-1-infected patients. On entailed large losses of progenitor cells. These findings laid
the basis of this observation, plerixafor was tested for its abil- the foundation for current cord blood banking. This experi-
ity to mobilize CD34+ and hematopoietic progenitor cells ence led to the first successful cord blood transplant in a
from marrow to peripheral blood. In a pilot study that young patient with Fanconi anemia.
included patients with myeloma or lymphoma, plerixafor
caused a rapid and statistically significant increase in the
total WBC and peripheral blood CD34+ counts at 4 and 6 Conditioning regimens
hours after a single injection. The results of two phase 3 ran-
domized studies, one involving lymphoma patients and the The combination of chemical and physical agents given
other for myeloma patients, were reported. Patients were before HCT is known as the conditioning or preparative
randomized to receive G-CSF alone or G-CSF in combina- regimen. The purpose of this conditioning regimen in both
tion with plerixafor. In both studies, a significantly higher the autograft and allograft setting is to eradicate the malig-
proportion of patients in the G-CSF-plus-plerixafor arm nancy exploiting the dose response phenomena that most
reached the primary end-point compared with the G-CSF- cancer cells exhibit. In the setting of allogeneic HCT, the
alone arm. Plerixafor was well tolerated, and the most com- conditioning regimen serves a second purpose, which is to
mon adverse events were gastrointestinal disorders and suppress the host immune system. The more immunosup-
injection site reactions. These results led to approval by the pressive the conditioning regimen, the better the chance for
US Food and Drug Administration (FDA) of this agent in engraftment. Conditioning regimen intensity has been clas-
2009. In the autologous transplant setting, peripheral blood sified according to their myelosuppressive effects into the
HCT has almost totally replaced bone marrow as a source of categories of myeloablative, reduced intensity, and nonmy-
stem cells because of the ability to collect a large number of eloablative. Table 14-3 lists the most commonly used condi-
stem cells that result in a more rapid engraftment and reduc- tioning regimens currently in use.
tion in complications.
In the setting of allogeneic HCT, peripheral blood HCT is Table 14-3 Commonly used conditioning regimens
associated with faster neutrophil engraftment but also with Allogeneic stem cell transplantation
higher rates of chronic GVHD. Nine randomized trials have Myeloablative conditioning
been performed comparing peripheral blood versus bone CyTBI Cyclophosphamide 120 mg/kg + TBI 8-12 Gy*
marrow in the setting of matched related donor transplanta- BuCy Cyclophosphamide 120 mg/kg + busulfan
tion. In a meta-analysis of individual data of these trials, 16 mg/kg PO or IV equivalent
peripheral blood led to faster neutrophil and platelet engraft- FluBu Fludarabine 120-150 mg/m2 + busulfan
ment and was associated with a significant increase in the 16 mg/kg PO or IV equivalent
development of grade 3-4 acute GVHD and extensive Reduced intensity/nonmyeloablative conditioning
chronic GVHD at 3 years. Peripheral blood also was associ- Flu TBI Fludarabine + TBI 2 Gy
ated with a decrease in relapse (21% vs. 27% at 3 years) both Flu Mel Fludarabine + melphalan 140 mg/m2
for advanced and early stage hematologic malignancies. Flu Cy Fludarabine + cyclophosphamide 60 mg/kg
Peripheral blood was not associated with lower rates of Flu Bu Fludarabine + busulfan 8 mg/kg PO or IV
NRM; however, in patients with advanced disease, it was equivalent
associated with improvements in overall and disease-free Cy ATG Cyclophosphamide 4 g/m2 + ATG†
survival (DFS). Predominantly autologous stem cell transplantation
In children, the increased risks of chronic GVHD has led Lymphoma
to the use of bone marrow as the preferred source of stem BEAM BCNU + etoposide + cytarabine + melphalan
cells. Use of T-cell depletion with CD34 selection has been BEAC BCNU + etoposide + cytarabine +
used to reduce the increased risk of chronic GVHD but fur- cyclophosphamide
ther prospective trials are needed. CBV Cyclophosphamide + BCNU + etoposide
Myeloma
High-dose melphalan Melphalan 200 mg/m2
Cord blood
ATG = antithymocyte globulin; TBI = total body irradiation.
Broxmeyer et al. (1989) were the first to report the presence *Various fractionation schedules in use.
of HPCs in cord blood using the granulocyte-macrophage †Mainly for conditioning in severe aplastic anemia.

Conditioning regimens | 363
Myeloablative regimens following doses of commonly administered agents: mel-

phalan <150 mg/m2; busulfan <9 mg/kg of the oral equiv-
The first conditioning regimen that achieved widespread
alent; thiotepa <10 mg/kg; and TBI <500 cGy single
application consisted of the combination of cyclophos-
fraction or 800 cGy fractionated. These definitions are
phamide and total body irradiation (CyTBI). High doses
somewhat arbitrary but are important for retrospective
of cyclophosphamide, typically 120-200 mg/kg are com-
studies.
bined with radiation in a dose of 8-12 Gy (depending on
Randomized trials have been performed comparing
the fractionation). This regimen is myeloablative and pro-
transplant outcomes in patients with hematologic malig-
foundly immunosuppressive. High-dose busulfan and
nancies using different regimen intensities. To date, the
cyclophosphamide (BuCy) conditioning was developed as
results are conflicting, but the trial sizes have been rela-
an alternative to CyTBI. For treatment of patients with
tively small. The largest trial to date performed by the
ALL, TBI-containing regimens may be slightly superior.
Blood and Marrow Transplant Clinical Trials Network
Treatment-related morbidity and mortality rates are simi-
(BMT-CTN) was stopped early due to a survival benefit in
lar after both regimens, although the patterns of toxicity
the myeloablative arm. Final report of this study is eagerly
are slightly different. TBI is associated with more pulmo-
awaited. Retrospective comparisons suggest similar out-
nary toxicity, cataract formation, and thyroid dysfunction.
comes with reduced-intensity conditioning regimens
BuCy is associated with a higher incidence of veno-
when compared with myeloablative regimens. One Center
occlusive disease (VOD) and irreversible alopecia. More
for International Blood and Marrow Transplant Research
recently, fludarabine/busulfan (FluBu) combinations have
(CIBMTR) analysis, however, did report inferior out-
become increasingly utilized because cyclophosphamide
comes for patients with AML receiving low doses of TBI
and its metabolites have been implicated in the develop-
when compared with other reduced-intensity or myeloab-
ment of VOD. Busulfan-based protocols generally are rec-
lative conditioning regimens.
ommended in very young children because of the
long-term effects of TBI.
Regimens for autologous HCT
Some conditioning regimens are used nearly exclusively

Nonmyeloablative and reduced-intensity
for autologous HCT; they include: (i) carmustine-based
conditioning
regimens, such as carmustine, etoposide, cytarabine, and
Myeloablative conditioning regimens were long consid- melphalan (BEAM) or cyclophosphamide, carmustine,
ered necessary for engraftment of allografts, but their and etoposide (CBV) for large-cell lymphoma or Hodgkin
considerable extramedullary toxicity typically limited lymphoma; (ii) the high-dose melphalan regimens used
their use to patients <50-60 years of age who had a good for myeloma; and (iii) the ifosfamide, carboplatin, and
performance status and no comorbidities. The demon- etoposide (ICE) regimens for lymphoma and germ cell
stration that engraftment can be achieved without mye- tumors (GCTs).
loablation led to the investigation of so-called
nonmyeloablative or reduced-intensity conditioning (RIC)
Conditioning for benign hematologic
regimens (also called minitransplantation regimens).
disorders
These regimens often use lower doses of busulfan, mel-
phalan, cyclophosphamide, or TBI (typically 2 Gy), often Patients with aplastic anemia, metabolic disorders, or hemo-
in combination with fludarabine. Nonmyeloablative regi- globinopathies represent a special category. There is no
mens more frequently have been used in older patients underlying malignancy that requires eradication. There is a
and in patients with comorbidities. These regimens rely higher risk of graft rejection, in part because of the nature of
more heavily on immunologic (graft-versus-leukemia the underlying disease, the lack of previous immunosup-
[GVL]) effects to induce tumor regression and contain pressive chemotherapy, and, in many cases, exposure to
lower doses of drugs with cytoreductive activity. GVHD prior transfusions with HLA sensitization. Thus, the condi-
and infections remain the major causes of NRM. tioning regimens for such patients traditionally have empha-
The characteristics of a reduced-intensity conditioning sized more immunosuppression and less myelosuppression.
regimen are that they should be associated with low non- A combination of high-dose cyclophosphamide with anti-
hematologic toxicities and some degree of mixed chime- thymocyte globulin (ATG) has emerged as the standard
rism early posttransplant, and, theoretically, they could be conditioning regimen for aplastic anemia, although it may
given without stem cell support. Operationally reduced- need modification in patients with underlying Fanconi
intensity conditioning regimens have been defined by the syndrome.

Supportive care post-HCT Phase V: late convalescence
This final phase is characterized by the almost full recovery

Successful HCT requires the patient to tolerate the side effects
of the immune system and by the potential of late complica-
of the conditioning regimen that the stem cells proliferate
tions, such as organ dysfunction or recurrence of the original
and mature adequately, and appropriate as treatment and
malignancy.
prevention of infectious complications that can occur during
the severely immune-compromised state of the patients dur-
ing the first months after HCT. Most HCTs in North America
are performed by specialized teams of physicians, nurses, and HCT complications
other personnel. Outcomes are improved when HCTs are Myelosuppression
performed by specialized transplant units that perform a
minimum of at least 10 transplants a year. Severe myelosuppresion is a universal complication of mye-
The HCT procedure can be divided in five phases as loablative conditioning regimens, regardless of the stem cell
detailed below and summarized in Table 14-4. source. The duration of the myelosuppression depends on
various factors, including stem cell dose, use of methotrexate
as GVHD prophylaxis, extent of prior therapy, and stem cell
Phase I: chemotherapy phase source. Engraftment is defined as sustained recovery of an
During this phase, chemotherapy (usually at high doses) absolute neutrophil count (ANC) of 500 neutrophils per
with or without radiation is given to the patient to elimi- liter or more for 3 consecutive days. In the context of an allo-
nate any residual malignant cells and provide space for the geneic HCT, this also implies evidence of donor cell engraft-
donor stem cells. Phase I finishes with the infusion of the ment; in the context of an autologous HCT, neutrophil
HPCs provided either by the patient in the case of an recovery is synonymous with engraftment. Filgrastim has
autologous transplant or by a donor in the case of an allo- been shown to reduce the time to neutrophil recovery in
geneic HCT. both the autologous and allogeneic setting but with no
definitive improvement in HCT outcomes.
Phase II: cytopenic phase

Graft failure
The most obvious effects of the high doses of chemoradio-
therapy are felt during this phase. Severe myelosuppression Graft failure is an unusual but often fatal complication of
and disruption of the gastrointestinal mucosa manifested as HCT. Mechanisms include immunologic rejection, abnor-
stomatitis and diarrhea during this period can last 10-28 malities in the marrow microenvironment or stroma, inad-
days. During this period, serious infections and organ toxici- equate dose or composition of the graft, viral infections
ties can occur. (in particular cytomegalovirus [CMV]), or drug-induced
myelosuppression. It often is impossible to determine the
exact cause of graft failure in an individual patient, but the
Phase III: early recovery phase
risk for graft failure is increased with increasing HLA dis-
In this initial phase of neutrophil recovery, patients can parity between the graft and host, with T-cell depletion of
develop a syndrome characterized by fever, rash, and pulmo- the graft, and in transplantation for certain diseases, such as
nary infiltrates known as the “engraftment syndrome,” severe aplastic anemia or hemoglobinopathies. The risk for
which, when identified, should be treated promptly with graft rejection can be decreased by infusing larger numbers
corticosteroids. This period also marks the most common of HSCs and by increasing the intensity of the conditioning
time when acute GVHD can begin to manifest in the allograft regimen. Successful treatment of graft failure usually
setting. involves reinfusing more stem cells either from the original
stem cell donor or another source if the original donor is
unavailable. Graft failure after autologous HCT is rare but
Phase IV: early convalescence phase
can happen because of infections or toxic drug exposure.
This phase is characterized by persistent immune deficiency Heavily pretreated patients receiving suboptimal doses of
despite normal peripheral blood cell counts. Patients remain stem cells (<2 million CD34+ cells/kg) frequently have poor
at risk of serious life-threatening opportunistic infections graft function after autologous HCT and have a higher rate
that require antibiotic, antiviral, and antifungal prophylaxis of developing secondary myelodysplastic syndrome (MDS)
as well as close monitoring by the transplant team. or AML.

HCT complications | 365
Table 14-4 HCT complications according to transplant phase
I: Chemotherapy V: Late
Phase phase II: Cytopenic phase III: Early recovery IV: Early convalescence convalescence
Infections Catheter-related GPC GNR from GI Resistant GNR or GPC Viral reactivation Viral reactivation
GPC toxicity Fungal infections CMV Pneumocystis (if active GVHD)
HSV reactivation EBV Encapsulated GPC encapsulated
Fungal infections reactivation Other GPC
viruses
Gastrointestinal Nausea and Mucositis Protracted nausea and/or
vomiting Diarrhea anorexia can be sign of
Diarrhea Nausea upper GI GVHD
Anorexia
Hepatic Transaminitis Transaminitis Transaminitis Hepatitis reactivation Cirrhosis
Sinusoidal Sinusoidal obstruction
obstruction syndrome
syndrome Liver GVHD
Cardiac Arrhythmias Hypertension from Hypertension from CNI Congestive heart
(rare) CNI failure
Fluid overload Premature coronary
vascular disease
Pulmonary Pneumonitis Pneumonia Idiopathic pneumonia Cryptogenic organizing Bronchiolitis
(rare) Fluid overload syndrome pneumonia obliterans
Hyperactive airway
disease
Neurologic Seizures from Cognitive
busulfan dysfunction-
(rare with short-term
prophylaxis) memory loss
Impaired
concentration
Endocrine Hyperglycemia Hyperglycemia from Hyperglycemia from CNI Hyperglycemia Metabolic syndrome
CNI Hypothyroidism
Renal Increased Increased creatinine Increased creatinine Chronic renal failure
creatinine due to drugs Electrolyte disturbances
Electrolyte (antibiotics,
abnormalities antifungals, CNI)
Electrolyte
disturbances
Acute graft vs. Initial presentation can be Late acute GVHD
host rash and fevers presents as acute onset
“Cytokine storm” diarrhea or rash
Chronic graft Usually presents in
vs. host the context of
immune suppression
withdrawal
Other Cataracts
Secondary
malignancies
Timing D−10 to D 0 D0 to engraftment Engraftment +5-7 days D+30 to 6-12 months 12 months post-
usually depending on HCT onward
D+15-30 immune reconsti-
tution
GPC = gram-positive cocci; GNR = gram-negative rods; GI = gastrointestinal; HSV = herpes simplex virus; CMV = cytomegalovirus; EBV =
Epstein-Barr virus; GVHD = graft-versus-host disease; CNI = calcineurin inhibitor; d = day.

inhibitors warrant the need for careful monitoring and often

Infections
dose reduction of tacrolimus and cyclosporine. Also, because
Infections are a major cause of life-threatening complica- Aspergillus is treated more successfully, cases of mucormycosis
tions in HCT. Their prevention, diagnosis, and treatment are increasingly are reported and necessitate treatment with
important components of the care of the HCT patient. Major amphotericin derivatives or posaconazole.
advances in this area have decreased NRM. Although this is Viral infections are common after HCT. Cytomegalovirus
an ever-changing field, the Centers for Disease Control and (CMV) infection used to be a major cause of pneumonia and
Prevention (CDC) recommendations published in 2000 and death in HCT recipients. CMV infection post-HCT usually
updated in 2009 provide an essential framework for treat- occurs as a consequence of CMV reactivation in patients
ment and prevention. previously exposed to CMV as indicated by positive anti-
Bacterial infections occur with high frequency during the body titers (CMV+ patients). The incidence of reactivation
neutropenic period after transplantation, and guidelines for ranges from 40% to 60% in the allogeneic setting and <5%
their prevention and management are similar to those in in the autologous setting depending on the technology used
other neutropenic patients. Many centers use prophylactic for screening, the target tissue evaluated (eg, blood, urine,
fluoroquinolone antibiotics for gut decontamination in bronchoalveolar lavage [BAL]), the conditioning regimen,
patients older than 12 years of age during the neutropenic and the method of GVHD prophylaxis. Detection of CMV in
period. Their use in younger children is controversial sec- the blood (CMV viremia), either by PCR or rapid antigen
ondary to older safety data in an animal model that restricted screening, indicates a high risk for CMV disease, usually
the use of fluoroquinolones in this age group. The American CMV pneumonia but occasionally (especially at later time
Academy of Pediatrics currently recommends that fluoro- points after transplantation) CMV hepatitis, retinitis, or gas-
quinolones be limited in children to a number of circum- troenteritis. Patients who have not been exposed before
stances that include Gram-negative bacteremia in the transplantation (CMV−) are still at risk for CMV infection
immunocompromised host in which an oral agent is desired. either by transmission from a CMV+ stem cell donor or via
As the experience with fluoroquinolones in young children transfusion of blood products from a CMV++ blood donor.
grows, there is likely to be an analysis of their benefit in this To avoid risk of CMV infection in CMV−/− donor/recipient
age group. Patients with chronic GVHD are immunosup- pairs, CMV− blood products formerly were recommended
pressed by therapy for GVHD as well as GVHD itself. They but often were not readily available. Fortunately, leukocyte
are at particular risk for fulminant infections with encapsu- filtration of blood products efficiently reduces the risk of
lated Gram-positive organisms, particularly Pneumococcus. CMV transmission, and most centers no longer require use
They should receive prophylaxis with penicillin V potassium of CMV− blood products.
or trimethoprim-sulfamethoxazole. Frequent screening for CMV viremia is mandatory in the
HCT patients are at high risk for Pneumocystis jirovecii first 3 months after allogeneic but not autologous HCT. Gan-
(formerly known as Pneumocystis carinii), and prophylaxis is ciclovir, oral valganciclovir, high-dose acyclovir, and valacy-
recommended. For those allergic to trimethoprim-sulfa- clovir have all been used for prophylaxis of CMV reactivation
methoxazole, alternatives such as pentamidine, atovaquone, or in patients at high risk. Each of these approaches has poten-
dapsone are routine. Trimethoprim-sulfamethoxazole prophy- tial problems, including cost, inconvenience, and adverse
laxis also may prevent toxoplasmosis, which occasionally has effects. Myelosuppression, especially neutropenia, is the
been reported in recipients of allogeneic transplantation. most serious and common toxicity associated with ganciclo-
Fungal infections remain a major problem in allogeneic trans- vir and valganciclovir.
plantation patients and are associated with prolonged neutro For patients who develop CMV viremia, preemptive treat-
penia and also with immunosuppression and GVHD. Yeast ment with ganciclovir or valganciclovir is initiated immedi-
(Candida) infections are rare with fluconazole prophylaxis. When ately. This strategy of preemptive treatment has significantly
such infections occur, they frequently are caused by fluconazole- decreased the occurrence of CMV disease in the early months
resistant organisms. Airborne molds, particularly Aspergillus, after transplantation. Oral valganciclovir (but not oral gan-
remain a major hazard for patients undergoing allogeneic tran ciclovir) is a convenient and effective oral alternative for pre-
splantation, despite the use of high-efficiency particulate air emptive and prophylactic treatment. Alternative medications
filtration. The azoles (voriconazole and posaconazole) and echi- for preemptive treatment include foscarnet (equally effica-
nocandins (caspofungin, micafungin, and anidulafungin) with cious but more nephrotoxic) and cidofovir (requires only
potent activity against molds have improved the outcome for once-weekly administration but is less extensively tested and
such patients. Concerns with new azoles include their toxicity much more nephrotoxic and myelosuppressive). Acyclovir
profile (neurologic and hepatic toxicity), which can be life and valacyclovir, although moderately active for CMV pre-
threatening. Interactions with the metabolism of calcineurin vention, have no role in preemptive treatment.

Specific organ toxicities | 367
Other important herpes viruses include herpes simplex regimen. As a rapidly dividing tissue that is being constantly
virus (HSV), varicella-zoster virus (VZV), Epstein-Barr virus regenerated, the toxic effects of the conditioning regimen are
(EBV), and human herpesvirus 6 (HHV-6). HSV used to be felt all throughout the gastrointestinal tract. The most com-
a major cause of mucositis and pneumonia occurring during mon manifestations of gastrointestinal toxicity are nausea,
the neutropenic phase after transplantation and is prevented vomiting, oral lesions, esophagitis, and diarrhea.
by acyclovir. VZV can cause zoster, a frequent problem after Carmustine, TBI, and cyclophosphamide are highly emeto-
transplantation with patients at risk for dissemination when genic agents, whereas melphalan and busulfan are classified as
profoundly immunosuppressed. In a single-institution moderately so. Thus, adequate control of nausea and vomiting
double-blind controlled trial, patients after an allogeneic requires prophylaxis as well as frequent use of breakthrough
transplantation who were at risk for VZV reactivation were medications. Acute emesis usually involves combination
randomized to acyclovir 800 mg twice daily or placebo given therapy with corticosteroids and 5-
hydroxytryptamine-3
from 1 to 2 months until 1 year after transplantation. Acyclo- receptor antagonists. Despite this, complete control of nausea
vir significantly reduced VZV infections at 1 year after trans- and vomiting (no nausea, no emesis, and no need for break-
plantation (hazard ratio, 0.16; p = 0.006). EBV can cause through medications) is achieved in <20% of the population.
posttransplantation lymphoproliferative disease, particularly Destruction of the oral and gastrointestinal mucosa is a signifi-
in patients who are extremely immunosuppressed because of cant dose-limiting complication of high-dose therapy
mismatched or T-cell–depleted transplantation. Treatment regimens.
with rituximab is typically first-line therapy. HHV-6 is a Stomatitis refers to the painful ulcerations and sores that
cause of posttransplantation encephalitis and aplasia. Others occur on the mouths, lips, gums, and throats of patients usu-
have postulated a link with interstitial pneumonia. ally 5-7 days after conditioning and can be seen in up to 90%
Adenovirus has been the cause of fatal hepatitis, gastroen- of HCT recipients. The most important risk factor for devel-
teritis, and pneumonitis in transplantation patients. The epi- oping severe oral mucositis is the intensity of the condition-
demiology and value of screening remains a matter of ing regimen; other factors that predict development of severe
ongoing study. Respiratory viruses, such as respiratory syn- oral mucositis are poor oral hygiene, extensive prior therapy,
cytial virus (RSV) and influenza virus, can lead to fatal pneu- and concurrent chemo-radiation. Oral mucositis is a signifi-
monias. Some centers have recommended screening of all cant cause of morbidity post HCT. Studies have shown that
patients during RSV season and treatment with ribavirin the incidence of severe oral mucositis post high-dose therapy
and immunoglobulin in patients who become infected. This can be reduced by palifermin in the setting of TBI and by
is, however, a controversial issue. BK virus and adenovirus using ice chips during chemotherapy infusion in the setting
have been associated with severe hemorrhagic cystitis. The of high-dose melphalan and by amifostine. Once oral muco-
frequency of infection, treatment, and value of screening are sitis occurs, treatment is primarily supportive with intrave-
not determined. nous hydration and alimentation if needed as well as
parenanalgesics and antibiotics to prevent infections.
Diarrhea occurs in more than half of all patients receiving
Specific organ toxicities high-dose chemotherapy and also depends on the intensity
of the conditioning. Other treatable causes of diarrhea need
Integument toxicity
to be considered, particularly Clostridium difficile infection,
TBI frequently is associated with generalized erythema fol- antibiotic-induced diarrhea, and GVHD. Persistent diarrhea
lowed by hyperpigmentation. Thiotepa is metabolized and after engraftment should be investigated thoroughly with
excreted through the sweat glands around skin folds and endoscopic evaluation for tissue procurement to rule out
dressings. Failure to take frequent showers and change dress- GVHD and other treatable causes. Treatment is supportive
ings can lead to serious thiotepa skin toxicity. Likewise, and symptomatic.
patients receiving thiotepa should not use moisturizing
cream during the days immediately after receiving the drug.
Hepatic complications
High-dose alkylator therapy as well as radiation is associated
with alopecia, usually reversible with the occasional excep- VOD or sinusoidal obstruction syndrome (SOS) is one of
tion of busulfan. the most common and lethal toxicities of HCT; it occurs in
10%-60% of patients receiving transplantations, depending
on both the risk factors for the patients and the vigor with
Gastrointestinal toxicity
which the diagnosis is pursued. VOD is caused by prepara-
After hematopoietic tissue, the gastrointestinal tract is the tive regimen toxicity and is thought to be caused by damage
single most commonly affected organ by the conditioning to endothelial cells, sinusoids, and hepatocytes in the area

surrounding terminal hepatic venules. Endothelial cells are decreased diffusing capacity of the lung for carbon monox-
directly sensitive to chemotherapy and radiation therapy, ide and elevated alveolar-arterial partial pressure of oxygen
and cytokines released during endothelial injury also may be were predictors for increased mortality. Most transplanta-
implicated. For instance, elevated levels of tumor necrosis tion centers, however, do not exclude a patient from trans-
factor a (TNFa) predict development of VOD. plantation based solely on an abnormal pre-HCT PFT.
VOD is more common in patients with evidence of prior Similarly, baseline reduced left-ventricular ejection fraction
hepatocellular damage at the time of transplantation, heavy predicted for cardiac toxicity after HCT but failed to predict
pretreatment before HCT, prolonged and elevated busulfan life-threatening events.
levels, or >10-12 Gy TBI. Other drugs, such as nitrosoureas During the early posttransplantation period (days 0-30),
(carmustine), also have been implicated in VOD/SOS. Stud- regimen-related toxicity and infectious etiologies account
ies suggest that prior exposure to gemtuzumab significantly for most of the pulmonary events. Although most focal infil-
increases the risk of VOD, especially in those who receive the trates are infectious in origin, diffuse infiltrates related to
drug shortly before transplantation. Low-dose heparin and regimen-related toxicity also should be considered. The dif-
ursodiol have been used for prevention of VOD/SOS but ferential diagnosis of diffuse infiltrates during early post-
remain controversial. HCT includes iatrogenic fluid overload, pulmonary edema
The clinical diagnosis of VOD usually requires weight gain (cardiogenic and non-cardiogenic), idiopathic pneumonia
or ascites, tender hepatomegaly, and jaundice. Ideally, the syndrome (IPS), adult respiratory distress syndrome from
diagnosis should be confirmed by liver biopsy, but liver chemoradiotherapy injury or sepsis, and diffuse alveolar
biopsy is not always possible because of the risks in critically hemorrhage (DAH). Cardiogenic pulmonary edema and
ill patients. Treatment generally has been supportive care septicemia in particular need to be excluded. After engraft-
with judicious fluid management, salt restriction, and elimi- ment, the risk of fungal and viral infection increases. Histori-
nation of any potential hepatotoxic agent. cally, CMV pneumonitis was the most common cause of
Defibrotide is an adenosine receptor agonist that increases diffuse infiltrates during days 30-150, but its incidence has
levels of endogenous prostaglandins (PGI2 and PGE2), decreased dramatically with the use of preemptive treatment
reduces levels of leukotriene B4, stimulates expression of strategies for the prevention of CMV disease. During this
thrombomodulin in endothelial cells, modulates platelet period, opportunistic and idiopathic pneumonias dominate
activity, and stimulates fibrinolysis by increasing endoge- the pulmonary complications. It takes approximately 3-6
nous tissue plasminogen activator function and decreasing months for the immune function of patients undergoing
the activity of plasminogen activator inhibitor 1. Defib- HCT to return to normal and longer for patients who suffer
rotide has little systemic anticoagulant activity, which is an from chronic GVHD. Infectious etiologies during this phase
advantage in patients with multiorgan failure. In the latest include bacteria, fungi, viruses, Nocardia, mycobacteria, and
published update, 88 patients with severe VOD were treated P. jirovecii. Furthermore, approximately 10% of patients
with defibrotide. At treatment, median bilirubin was with chronic GVHD develop bronchiolitis obliterans, a
12.6 mg/dL, and multiorgan failure was present in 97%. No severe obstructive airflow disease.
severe hemorrhage or other serious toxicity was reported.
Complete resolution of VOD was seen in 36%. Younger
Idiopathic pneumonia syndrome
patients, those receiving autologous HCT, and those with
abnormal portal flow had the highest response rates. Defib- IPS is a condition characterized by diffuse alveolar injury
rotide is approved in Europe, and the results of a recently with fever, cough, dyspnea, hypoxemia, and restrictive
completed phase 3 trial have been submitted to the FDA for physiology. Chest x-ray usually demonstrates multilobar
evaluation. pulmonary infiltrates. This is a diagnosis of exclusion.
Bronchoalveolar lavage (BAL) must be negative for infec-
tious etiologies, including bacteria, fungi, CMV, and other
Pulmonary toxicities
viral infections. The incidence of IPS is approximately 7%,
Pulmonary complications occur frequently after HCT and with a median time to onset of 21 days and hospital mortal-
are associated with a high mortality. To stratify transplanta- ity, ranging from 30% to 70%. The risk factors for IPS
tion patients into different risk categories, pretransplanta- include the use of TBI or c armustine-based conditioning
tion evaluation often includes a detailed pulmonary function regimens and previous exposure to bleomycin. Treatment
test (PFT) and two-dimensional echocardiogram or radio- of IPS is mostly supportive, but high-dose corticosteroids
nuclide ventriculography. The utility of these tests, however, often are given. They may be beneficial in patients in whom
is doubtful. In a retrospective study from the Fred Hutchin- pulmonary damage is due to carmustine or in those with
son Cancer Research Center that included 1,297 patients, the closely related syndrome of DAH.

Graft-versus-host disease | 369
Diffuse alveolar hemorrhage transplantation-related TTP do not respond to plasma

exchange. TTP outside the transplantation setting has been
DAH occurs most commonly in the first weeks after HCT
associated with immunoglobulin G (IgG) antibodies that
and presents as idiopathic pneumonia with or without
inhibit the cleaving protease of von Willebrand factor in the
hemoptysis. The classic finding on BAL is increasingly
plasma. No such mechanism was found in cases of post-
bloody returns during BAL washings. Analysis of BAL fluid
transplantation TTP, pointing to another as-yet-unidentified
usually demonstrates red blood cells, hemosiderin-laden
cause for this syndrome.
macrophages if blood has been present for more than 2-3
days, and negative microbiologic studies. Treatment of DAH
is largely supportive, but retrospective studies suggest that Bleeding
high-dose corticosteroids with a starting dose in the range of
Although all patients with thrombocytopenia are at risk for
1 g/d are often beneficial.
bleeding, several hemorrhagic syndromes are peculiar to
transplantation. Hemorrhagic cystitis early after transplanta-
Transplantation-related obstructive tion usually is attributed to toxicity to the bladder from
airway disease cyclophosphamide metabolites. Late-onset hemorrhagic cys-
titis often is associated with viral infection with BK virus and
Approximately 6%-10% of patients with chronic GVHD occasionally with adenovirus. Hemorrhagic cystitis can be
develop chronic airway obstruction. The most common his- severe and can require continuous bladder irrigation, divert-
tologic finding is constrictive bronchiolitis obliterans. Bron- ing nephrostomy tubes, and occasionally formalin instilla-
chiolitis obliterans typically presents 3-12 months after an tion until the bladder heals. As mentioned, pulmonary
allogeneic HCT with gradual onset of dyspnea, dry cough hemorrhage can be a serious complication of transplantation
associated with occasional wheezing, and inspiratory crack- and most often is attributed to preparative regimen toxicity.
les. PFTs demonstrate an obstructive pattern that does not
respond to bronchodilator therapy and reduced diffusing
capacity of the lung for carbon monoxide. Thin-section Iron overload
computed tomographic scans reveal bronchial dilation, Iron overload has been identified to be an adverse prognostic
mosaic pattern attenuation, and evidence of air trapping on factor for children with thalassemia undergoing HCT, and
expiration. The diagnosis often is based on clinical, imaging, there is an increasing evidence that iron overload also may
and spirometric findings without a tissue biopsy. There is no have deleterious effects for patients with hematologic malig-
effective treatment of patients with bronchiolitis obliterans, nancies who undergo HCT. This particular patient popula-
and current treatment mostly is directed at the underlying tion often is transfused heavily before HCT and continues to
chronic GVHD with immunosuppressive therapy. Lung require transfusions in the peritransplantation period. One
transplantation offers some promise. red blood cell unit contains 200-250 mg of iron, and signifi-
cant iron accumulation can occur after 10-20 RBC transfu-
sions. Iron overload increases the risk of infections, VOD,
Thrombotic microangiopathy
and hepatic dysfunction.
Posttransplantation thrombotic microangiopathy (TMA)
presents as a spectrum of disease, ranging from mild micro-
angiopathic anemia to thrombotic thrombocytopenic pur- Graft-versus-host disease
pura (TTP) or hemolytic uremic syndrome and occurs more
commonly after allogeneic and unrelated donor HCT. TTP Acute and chronic GVHD traditionally was defined by the
frequently presents with fever, neurologic symptoms, micro- time of onset. Acute GVHD was defined as any GVHD
angiopathic hemolytic anemia, thrombocytopenia, and occurring before day 100 after transplantation, and chronic
renal impairment. In children, TMA more closely resembles GVHD was defined as any GVHD occurring after day 100. It
hemolytic uremic syndrome. In some patients, TMA appears is now recognized that typical features of chronic GVHD can
to be related to cyclosporine nephrotoxicity and responds to occur before day 100 and that typical features of acute
discontinuing the cyclosporine. But other patients have a GVHD can occur after day 100. Acute and chronic GVHD
fulminant course and a very high mortality rate. Autopsy are no longer defined by their time of onset but rather by
findings include arteriolar thrombosis in the kidneys. In their clinical features. Two subcategories of acute (classic
many patients, fungal infection, sepsis, or severe acute and persistent or late onset or recurrent) and two subcatego-
GVHD appear to be underlying the microangiopathic pro- ries of chronic GVHD (classic and overlap acute or chronic)
cesses. Unlike patients with idiopathic TTP, patients with are recognized.

mofetil now commonly are used as alternatives to metho-

Acute GVHD
trexate to decrease the toxicity of GVHD prophylaxis.
Acute GVHD is characterized in its mildest forms by skin rash. Other methods to prevent GVHD include depleting the
As the disease worsens, the confluent rash may progress to graft of donor T-cells either by an in vitro procedure after
blistering of the skin similar to a severe burn, profound diar- procurement of the stem cells or by exposure to T-cell-
rhea with crampy abdominal pain, and hepatic dysfunction depleting antibodies, such as ATG or alemtuzumab. These
with marked hyperbilirubinemia. Acute GVHD is graded by strategies result in a significant reduction in acute GVHD
the extent of skin rash, the amount of diarrhea, and the degree but can result in higher infection rates because of delayed
of bilirubin elevation. There are several methods of grading immune reconstitution. A large randomized study compar-
GVHD, but all rely on the same features, and most continue to ing different GVHD prevention strategies is now underway
use the original Glucksberg criteria or the modified Keystone in the United States and in Europe comparing CD34 selec-
criteria. Patients with stage I disease have skin disease and a tion to posttransplant cyclophosphamide to standard GHVD
mild course. Those with stage II to IV disease have multiorgan prevention and may establish a new standard of care.
disease, and patients with stage III or IV disease have a grave Therapy for acute GVHD consists of high-dose corticoste-
prognosis, with mortality rates >90%. roids, typically 2 mg/kg/d, which are tapered upon obtaining
Acute GVHD was first considered a “pure” T-cell-mediated a response. Calcineurin inhibitors will be continued or
disease with cellular injury thought to be the result of infil- restarted. Patients not responding to or experiencing recur-
tration of T-effector cells into target tissues. Recent immu- rence on high doses of corticosteroids (considered steroid
nohistochemical studies, however, demonstrate that some refractory) have a poor prognosis from continued acute
infiltrating cells are natural killer (NK) cells rather than GVHD, infection, and chronic GVHD. Other agents added
mature T-cells. This observation has led many investigators in the steroid-refractory setting include mycophenolate
to consider acute GVHD as a “cytokine storm.” This model mofetil, pentostatin, ATG, and monoclonal antibodies, such
accounts for many of the observations made in GVHD. It as infliximab, etanercept, and rituximab. The responses are
proposes that damage to host tissues during chemotherapy low, however, and patients typically succumb to opportunis-
and infection results in the release of inflammatory cyto- tic infection in the setting of profound immunosuppression.
kines such as TNF and interleukin-1 (IL-1). These cytokines All patients with GVHD should be encouraged to participate
provoke increased MHC expression and upregulate other in clinical trials.
adhesion molecules that, in turn, amplify recognition of allo-
geneic minor HLA differences by T-cells in the donor graft.
Chronic GVHD
The reactive donor T-cells proliferate and secrete more cyto-
kines that further activate additional donor T-cells and other Chronic GVHD affects from 40% to 80% of long-term-
inflammatory cells, including macrophages induced to survivors of allogeneic HCT. Although chronic GVHD once
secrete more IL-1 and TNF. This cascade eventually produces was designated arbitrarily as any GVHD occurring after day
the clinical manifestations of GVHD. Factors such as gut 100, it is now recognized as a distinct disorder in which the
decontamination, sterile environment, lower dose prepara- manifestations often resemble those seen in spontaneously
tive regimens, and ex vivo lymphocyte depletion of a marrow occurring autoimmune disorders. The diversity of the mani-
graft may decrease GVHD by interrupting this cascade. Of festations has proven a great hindrance to clinical study of
particular interest will be the further elucidation of the role chronic GVHD. A recent National Institutes of Health con-
of CD4+ subpopulations in GVHD because in experimental sensus conference produced working definitions for clinical
models, the T-helper cell type 2 (TH2) subpopulation that and pathologic diagnosis, staging, and response criteria, as
produces IL-4 and IL-10 (in contrast to TH1 cells, which well as suggestions for supportive care, clinical trial design,
secrete IL-2 and interferon) inhibits GVHD. Recent data and biomarkers.
have shown that allogeneic PBSC transplantation is relatively Many of the features of acute GVHD also can be found in
enriched for the TH2 population, which may account for the patients with chronic GVHD. But patients with chronic
relatively moderate rate of acute GVHD seen after the large GVHD always have, in addition, other diagnostic or distinc-
T-cell load given with the peripheral blood. tive features. Diagnostic features of chronic GVHD are fea-
Prophylaxis of GVHD has been more successful than tures that are sufficient to establish the diagnosis. They are
treatment. The most commonly used prophylaxis regimens summarized in Table 14-5. Diagnostic features of chronic
combine a calcineurin inhibitor (cyclosporine or tacroli- GVHD typically involve the skin and mucosa. They include
mus) with methotrexate. Because of the renal and mucosal poikiloderma, lichen planus–like features, sclerotic features,
toxicities seen with these regimens, alternative prophylactic and morphea-like features of the skin. Lichen-type features
regimens are being explored. Sirolimus and mycophenolate and hyperkeratotic plaques of the mouth also are diagnostic,

Graft-versus-host disease | 371
Table 14-5 Signs and symptoms of chronic GVHD
Distinctive (seen in chronic

Diagnostic (sufficient to GVHD, but insufficient alone to
establish the diagnosis of establish a diagnosis of chronic Common (seen with both
Organ of site chronic GVHD) GVHD) Other features* acute and chronic GVHD)
Skin Poikiloderma Depigmentation Sweat impairment Erythema, maculopapular
Lichen planus–like features Ichthyosis keratosis pilaris rash, pruritus
Sclerotic features hypopigmentation
Morphea-like features Hyperpigmentation
Lichen sclerosus–like
features
Nails Dystrophy
Longitudinal ridging, splitting,
or brittle features onycholysis
Pterygium unguis
Nail loss (usually symmetric;
affects most nails)†
Scalp and New onset of scarring or Thinning scalp hair, typically
body hair nonscarring scalp alopecia patchy, scarce, or dull (not
(after recover from explained by endocrine or
chemoradiotherapy) other causes), premature
Scaling, papulosquamous gray hair
lesions
Mouth Lichen-type features Xerostomia, mucocele, mucosal Gingivitis, mucositis,
Hyperkeratotic plaques atrophy, pseudomembranes† erythema, pain
Restriction of mouth Ulcers†
opening from sclerosis
Eyes New-onset dry, gritty, or painful Photophobia periorbital
eyes† hyperpigmentation
Cicatricial conjunctivitis Blepharitis (erythema of
Keratoconjunctivitis sicca† the eyelids with edema)
Confluent areas of punctate
keratopathy
Genitalia Lichen planus-like features Erosions†
Vaginal scarring or stenosis Fissures†
Ulcers†
GI tract Esophageal web Excrine pancreatic Anorexia
Strictures or stenosis in the insufficiency Nausea, vomiting, diarrhea,
upper to mid-third of weight loss
the esophagus† Failure to thrive (infants and
children)
Liver Total bilirubin, alkaline
phosphatase >2-3X upper
limit of normal†
ALT or AST >2-3X upper
limit of normal†
Lung BO diagnosed with lung BO diagnosed with PFTs and BOOP
biopsy‡ radiology†
Muscles, Fasciitis Myositis or polymyositis† Edema
fascia, joints Joint stiffness or Muscle cramps
contractures secondary Arthralgia or arthritis
to sclerosis
(continued)

Table 14-5 Signs and symptoms of chronic GVHD (continued)
Distinctive (seen in chronic

Diagnostic (sufficient to GVHD, but insufficient alone to
establish the diagnosis of establish a diagnosis of chronic Common (seen with both
Organ of site chronic GVHD) GVHD) Other features* acute and chronic GVHD)
Hematopoietic Thrombocytopenia
and immune Eosinophilia
Lymphopenia
Hypo- or
hypergammaglobulinemia
Autoantibodies (AIHA and
ITP)
Other Pericardial or pleural effusions
Ascites
Peripheral neuropathy
Nephrotic syndrome
Myasthenia gravis
cardiac conduction abnormality
or cardiomyopathy
From Filipovich AH, Weisdorf D, Pavletic S, et al. Biol Blood Marrow Transplant. 2005;11:945-956.
*Can be acknowledged as part of the chronic GVHD symptomatology if the diagnosis is confirmed.
†In all cases, infection, drug effects, malignancy, or other causes must be excluded.
‡Diagnosis of chronic GVHD requires biopsy or radiology confirmation (or Schirmer test for eyes).
as is vaginal scarring. Other diagnostic features of chronic Therapy in patients with chronic GVHD has relied on cor-
GVHD are the development of an esophageal web and stric- ticosteroids after a report by the Seattle transplantation
tures, fasciitis, and joint contractures. Finally, bronchiolitis group that corticosteroids are more effective than corticoste-
obliterans is a diagnostic feature of chronic GVHD if con- roids plus azathioprine. A study comparing cyclosporine
firmed by biopsy. plus prednisone therapy with prednisone was unable to
Distinctive signs are also typical for chronic GVHD but show a benefit to combination therapy other than a steroid-
are not by themselves considered sufficient for a diagnosis. sparing effect and less bone damage compared with the
They include depigmentation, nail loss, alopecia, xerosto- prednisone-alone group. Other therapies currently under
mia, and myositis. Features such as thrombocytopenia, evaluation include psoralen plus ultraviolet A, extracorpo-
eosinophilia, lymphopenia, hypo- or hypergammaglobu- real photopheresis, pentostatin, imatinib, and rituximab.
linemia, exocrine pancreatic insufficiency, myasthenia gra- The major cause of death in patients with chronic GVHD
vis, cardiac conduction abnormalities, and nephrotic is infection from the profound immunodeficiency associated
syndrome can occur in chronic GVHD but are not sufficient with chronic GVHD and its therapy. Careful monitoring
for diagnosis. with antibiotic prophylaxis for encapsulated organisms is
Chronic GVHD used to be scored as limited or extensive warranted in all patients. Patients with frequent infections
on the basis of the need for treatment. In the new proposal, and low immunoglobulin levels may benefit from intrave-
chronic GVHD is classified as mild, moderate, or severe nous immunoglobulin replacement. Patients should remain
based on the number of organs involved and the extent of on prophylaxis for viruses, P. jirovecii pneumonia, and fun-
involvement within each organ. gal infections (yeast and mold).
The incidence of chronic GVHD is increasing because of
the older age of patients being transplanted, the predomi-
nant use of PBSCs, and the use of mismatched and unrelated Late effects
donors. The greatest risk factor for development of chronic
GVHD is prior acute GVHD. Chronic GVHD has been As the number of long-term survivors following a transplan-
poorly studied compared with acute GVHD because most tation increases and many of these individuals have returned
patients have returned to their home institutions by the time to their referring physicians, the need for understanding and
this complication develops. These same factors also have continued follow-up is essential both for the care of the survi-
hindered studies of the pathophysiology of this disorder. vors and to anticipate the needs of the group as a whole. The

Late effects | 373
joint recommendations of the European Bone Marrow Trans- Psychosocial considerations

plantation (EBMT) Group, the CIBMTR, and the American
The long-term cognitive effects of prior therapy and HCT
Society of Blood and Marrow Transplantation were recently
continue to be evaluated. Significant central nervous system
published. These recommendations are based in part on pub-
(CNS) toxicity has been seen, especially in young patients
lished data and in part on common practice. Careful evalua-
receiving intensive intrathecal chemotherapy or CNS radio-
tion on an annual basis is recommended with close monitoring
therapy before transplantation. Previous evaluations involv-
and preventive screening, especially for the problems dis-
ing quality-of-life (QOL) assessments completed by parents
cussed in this chapter. Many of the late complications seen
appear to underestimate the child’s QOL and functioning.
after HCT are especially profound for younger patients.
Newer methods of neuropsychiatric testing have begun to
reveal subtle problems that greatly affect school perfor-
Endocrine adverse effects mance. Identification of these deficiencies and adaptive
measures help to improve school functioning. Use of neuro-
Endocrine sequelae of myeloablative transplantation have
psychiatric testing should be considered on a regular basis
been well documented but may be underappreciated. Chil-
for children as well as adults who are finding tasks at home
dren should be followed to ensure that adequate growth is
and work more difficult after transplantation. The loss of
obtained through adolescence. After conditioning with
executive function after therapy has not been evaluated. For
CyTBI, 20%-70% of children develop growth hormone defi-
patients who receive transplantation as adults, changes in
ciency. Some children have benefited from growth hormone
executive function, attention, and memory have been
therapy. In addition, many patients have thyroid dysfunc-
reported and may affect the ability to return to a particular
tion, often compounded by the effects of therapy before the
job or to continue the previous role of the individual in his
transplantation.
or her family life.
The prevalence of damage to the gonadal tissue is high and
may result in delay or absence of development of secondary
sexual characteristics and the need for hormonal replace-
Second malignancies and posttransplantation
ment. The risk for gonadal damage appears to depend on
lymphoproliferative disorders
multiple factors, including age, sex, type of transplantation,
previous therapy, and conditioning regimen. For many young Survivors of allogeneic transplantation are at increased risk for
adults, the high risk of infertility after HCT is a major issue, a variety of second malignancies, including a two- to threefold
and counseling for sperm or egg banking should be consid- increased risk of solid tumors compared with their age-
ered for such young patients before HCT. Recent evaluation matched controls. The risk increases over time after transplan-
of 39 male patients 2 years following HCT at a single institu- tation, with the greater risk among younger patients. In a
tion demonstrated spermatogenesis in 28% of the patients. retrospective multicenter study that included approximately
Those more likely to have sperm included men >25 years of 20,000 patients who had received either allogeneic or synge-
age at transplantation, men with a longer interval from trans- neic transplantations, the cumulative incidence rates for the
plantation, and men who had not had chronic GVHD. development of a new solid cancer were 2.2% and 6.7% at 10
Unfortunately, although sperm banking is readily available, and 15 years, respectively. The risk was significantly elevated
currently, only fertilized eggs are readily stored. Research is for cancers of the buccal cavity, liver, brain, bone, and connec-
ongoing into the cryopreservation of unfertilized eggs or ova- tive tissue as well as malignant melanoma. Higher doses of TBI
ries. For many patients, the course of their disease does not were associated with a higher risk of solid cancers. Chronic
allow this luxury; however, counseling with fertility specialists GVHD and male sex also were associated with increased risk
after the procedure, in the future, may allow new options. of squamous cell cancers of the buccal cavity and skin. Mela-
noma and basal cell carcinoma are common in patients, espe-
cially those with chronic GVHD. Patients should be instructed
Musculoskeletal complications
to avoid ultraviolet exposures and to use sunscreens and pro-
Patients receiving high-dose corticosteroids for their disease tective clothing.
or for GVHD have an increased risk of developing avascular Posttransplantation lymphoproliferative disorders after
joint necrosis and myopathies. In addition, loss of range of allogeneic transplantation are usually related to EBV. They
motion may be seen in patients with a history of chronic occur more commonly after T-cell-depleted transplantation
GVHD even if the disease is controlled. Osteoporosis result- or in other profoundly immunosuppressed states. Treat-
ing from steroid use and therapy-induced menopause is ment consists of decreasing immunosuppression, monoclo-
common. All patients should obtain a bone densitometry at nal antibody therapy (in particular rituximab), donor
1 year after transplantation. leukocyte infusions (DLIs), and sometimes chemotherapy.

Long-term survivors of autologous transplantation are at Treatment and prevention of relapse after allogeneic HCT
considerable risk for therapy-related leukemia. In some remains a major challenge. Only recently, the use of hypo-
series, the cumulative incidence exceeds 10%. The risk is methylating agents has been shown to potentially reduce the
increased with high-dose TBI used for conditioning, is risk of relapse in patients with AML and MDS, and a phase 3
related to the type and intensity of chemotherapy received trial is under way.
before transplantation, and possibly is related to the chemo- Preemptive or prophylactic DLI has been attempted, but
therapy agents used for stem cell mobilization (high-dose no large prospective trials have been performed. The appli-
etoposide is thought to confer an increased risk). In some cation of DLI is not without toxicity and can carry a mortal-
cases, cytogenetic abnormalities were detected in the mar- ity rate of 3%-10%, with acute GVHD and marrow aplasia
row or stem cell product of patients destined to develop being the leading causes of death. The incidence of both
therapy-related MDS, further implicating pretransplanta- acute and chronic GVHD after DLI is ~40%-60%, with more
tion chemotherapy. than half of the patients who develop chronic GVHD having
extensive disease. The onset of acute GVHD typically occurs
32-42 days after DLI.
Relapse and the graft-versus-
Posttransplant cellular therapies as a strategy for relapse
malignancy effect
prevention continues to be explored. Novel technologies,
Relapse remains the most important cause of treatment fail- such as chimeric antigen receptors or antigen-specific cyto-
ure in both the autologous and the allogeneic transplant set- toxic T-lymphocytes, are promising technologies that are in
ting. The mechanisms underlying disease recurrence are early clinical trials.
poorly understood. In the setting of autologous HCT, the
existence of cancer stem cells that may be quiescent and
therefore impervious to the effects of high-dose chemother- Hematopoietic cell transplantation
apy and radiation has been postulated. The potential benefit for specific diseases
of prolonged posttransplant therapy (such as lenalidomide)
supports this concept. The task of mastering the roles and results for hematopoietic
In contrast to autologous HCT, allogeneic HCT is associ- cell transplantation (HCT) is daunting: there are various
ated with a GVT effect mediated by alloreactive donor T- stem cell sources (autologous, allogeneic), many diseases,
and B-cells that provide an inherent posttransplant immune and, for each disease, different states of aggression. Things
surveillance mechanism. The importance of GVT initially can be simplified, however, into some basic rules of thumb:
was studied by comparing relapse rates between syngeneic
and allogeneic transplantation recipients, by considering the 1. Prognosis can be estimated based on disease risk (low,
relation between GVHD and relapse, and by examining the intermediate, high) and stage (low risk for cases in remis-
effect of T-cell depletion of the graft on risk of disease recur- sion vs. high risk for relapsed or refractory disease). Thus,
rence. Patients with AML in first complete remission (CR1) patients with low disease risk and low stage have a low
and CML in chronic phase had an increased rate of recur- overall risk (OS >60%), patients with low risk and high
rence after syngeneic transplantation. Relapse rates after stage disease or intermediate risk and low stage disease
syngeneic transplantation for lymphoma or for acute lym- have an intermediate overall risk (OS approximately
phoblastic leukemia (ALL) in CR1 are not increased com- 40%), patients with intermediate risk and high stage dis-
pared with those after allogeneic transplantation. Definitive ease or high risk and low stage disease have a high overall
evidence for a GVT effect comes from the use of DLIs. DLI risk (OS 25%), and patients with high risk and high stage
confers a direct graft-versus-malignancy effect by infusion of disease have a very high overall risk (OS <10%).
alloreactive donor lymphocytes. Purposes of DLI include 2. In diseases amenable to a reduced-intensity conditioning
conversion of mixed-donor chimerism to full-donor chime- (RIC) approach, typically those with significant graft-
rism after HCT as preemptive therapy to prevent relapse or versus-leukemia effect and/or high treatment-related
for the treatment of relapse. mortality with myeloablative conditioning, results from
The mechanisms of relapse after allogeneic HCT have not “full” myeloablative and RIC conditioning appear simi-
been well studied. These malignancies have not only escaped lar. Relapses are more common in the RIC setting but this
the effects of high dose alkylating agents but have also is offset by less treatment-related mortality.
evolved mechanisms to overcome the immune mediated 3. In the vast majority of cases, different allogeneic hemato-
graft vs. tumor effects. Clonal evolution, loss of specific sur- poietic cell sources yield similar results. Thus, outcomes
face antigens, or development of local immune suppressive after matched related, matched unrelated, cord blood, and
mechanisms have all been postulated. haploidentical transplants appear similar. The d ifferences

Hematopoietic cell transplantation for specific diseases | 375
in these approaches are from the causes of failure Outcomes of allogeneic HCT in CR1 are predictably bet-
(eg, relapse in matched related donor [MRD], GVHD in ter than in CR2 or higher, with active relapse, or with refrac-
matched unrelated donor [MUD], infections in cord tory AML. Survival after allogeneic HCT in CR1 is
blood and haploidentical donor transplants). approximately 40%-60%, whereas CR2 rates are ~25%-30%
4. In the course of some diseases, high-dose chemotherapy and refractory rates are ~10%. This data should not be inter-
with autologous HCT is preferred over allogeneic HCT, preted to suggest that all patients should receive chemother-
as it can have high cure rates while sparing patients the apy alone and if not cured expect the same outcome as
treatment-related morbidity and mortality associated upfront transplantation by delaying until CR2. Only a frac-
with allogeneic HCT. tion of patients relapsing after CR1 manage to regain remis-
sion and survive treatment toxicities to get to allogeneic
HCT. On the contrary, proceeding to allogeneic HCT with
Acute myelogenous leukemia (AML)
its high treatment-related mortality and morbidity without
The success of chemotherapy alone in curing AML is largely definitive evidence of superiority over chemotherapy alone
dictated by leukemia genetics (cytogenetic and molecular does a disservice to favorable-risk patients. Notably, alloge-
abnormalities) and age. Chemotherapy alone has high cure neic HCT for AML refractory to intensive induction or sal-
rates for the favorable-risk AMLs, acute promyelocytic leuke- vage chemotherapy can be curative. A CIBMTR study
mia (APL) and core-binding factor AML (t(8;21), inv16, evaluated outcomes of AML transplanted with persistent
t(16;16)). Allogeneic HCT significantly improves overall sur- disease at transplant established a predictive score for sur-
vival for patients with poor-risk cytogenetics in first complete vival based on CR1 duration <6 months, presence of circu-
remission (CR1) and potentially intermediate-risk AML lating blasts, donor other than HLA-matched sibling, KPS
patients. A meta-analysis of 24 trials comprising 6,007 patients <90, and poor-risk cytogenetics. The authors found a 3-year
suggested a survival benefit of allogeneic HCT compared with overall survival (OS) of 19% in the entire AML cohort but a
contemporaneous chemotherapy in both poor-risk and 3-year OS of 42% in those with a risk score of 0 (favorable
intermediate-risk AML. Although intensification of anthracy- findings in all 5 categories of risk).
cline during induction improves outcomes for younger AML The optimal myeloablative conditioning regimen for
patients, it has no benefit for patients with poor-risk cytogenet- AML has not been determined in a randomized fashion but
ics or those over the age of 65. Ongoing advances in initial current studies support the use of an intravenous busulfan
therapy for AML such as daunorubicin intensification, the (Bu)-based conditioning (Bu/cyclophosphamide(Cy), Bu/
addition of cladribine, and the incorporation of targeted agents fludarabine(Flu) rather than total-body irradiation (TBI)
continue to change the role of allogeneic HCT fro AML in CR1. based conditioning. An early, randomized study demon-
Mutations in specific genes can affect prognosis within a strated the superiority of TBI/Cy over oral busulfan, a drug
defined cytogenetic group. In AML with normal karyotype, a with unreliable absorption and significant GI toxicity. Intra-
historically intermediate-risk group, mutations in specific venous busulfan has more reliable pharmacokinetics, and
genes such as nucleophosmin 1 (NPM1), fms-like tyrosine two recent studies demonstrated the superiority of IV Bu-
kinase 3 (FLT-3), and CEBPA significantly alter prognosis. based conditioning over TBI-based regimens. A retrospec-
Normal karyotype AML with biallelic loss of CEBPA or tive CIBMTR retrospective study of 1,230 AML patients in
NPM1 mutation and no FLT3-ITD mutation behave like CR1 from 2000-2006 comparing IV Bu/Cy conditioning to
favorable-risk cases, but patients with a FLT3-ITD mutation oral Bu/Cy or TBI/Cy conditioning demonstrated superior-
have very high relapse rates with chemotherapy alone and ity in multivariate analysis of IV but not oral busulfan over
have relatively good outcomes with allogeneic HCT in CR1. TBI for leukemia-free survival (LFS), overall survival (OS),
A recent study suggests that, although considered favorable- relapse, and nonrelapse mortality (NRM). A subsequent
risk, NPM1 mutant AML may derive a relapse-free but not CIBMTR prospective cohort study enrolling 1,483 patients
overall survival benefit from matched-related donor alloge- from 2009-2011 compared outcomes of IV Bu-based condi-
neic HCT over chemotherapy alone. In addition, core bind- tioning to TBI-based conditioning in AML, CML, and MDS
ing factor (CBF) cases with activating mutations in the found significantly improved OS and progression-free sur-
tyrosine kinase c-Kit have intermediate-risk of relapse and vival (PFS) with the use of IV Bu-based conditioning relative
death and should be considered for allogeneic HCT in CR1. to TBI-based conditioning with a significant 2-year OS ben-
Older AML patients (often defined as >60 years of age) gen- efit in AML specifically (57% vs 46%, p = 0.003). As such, IV
erally have poor outcomes with chemotherapy alone and busulfan-based conditioning has become standard at most
thus are candidates for an RIC allogeneic HCT in CR1 if they transplant centers. As noted, the encouraging results of RIC
have limited comorbidities and intermediate- or poor-risk regimens has brought this from a treatment for older patients
cytogenetic or molecular profiles. and those with comorbidities to being more widely used.

Indeed, a nationwide randomized Phase 3 trial comparing the lack of a GVL effect. Either way, the procedure is
myeloablative vs. RIC transplantation in younger patients uncommonly used, given that now almost all patients who
with MDS and AML (BMT CTN 0901) closed early for need allogeneic HCT have hematopoietic cells available
apparent superiority of myeloablative arm and awaits results. from a matched-related, unrelated, or haploidentical donor
In cases that have a high likelihood of relapse, however, con- or from UCB units.
ventional wisdom suggests as potent a preparative regimen In acute myeloid leukemia (AML), multiparameter flow
be offered as clinically feasible. cytometry techniques can measure minimal residual disease
Although a matched sibling donor is still a preferred (MRD) as low as 1 AML cell in a background of 10,000 nor-
hematopoietic cell source, principally due to reducing time mal cells in some patients who are in morphologic complete
from CR to transplant and potentially better match of remission. More sensitive and broadly applicable genetic
minor histocompatibility antigens, equivalent outcomes are techniques are also being developed. Patients with detectable
observed using a fully allele-level HLA-matched unrelated residual disease are at a higher risk of relapse compared with
donor. Moreover, it appears that umbilical cord blood unit those without detectable residual disease. Thus, it is tempt-
and haploidentical donor transplants result in similar out- ing to use MRD as a guide to suggest which patients should
comes as MRD or MUD transplants. Because most umbilical undergo transplant in remission. Studies have shown that
cord blood (UCB) units have slow engraftment in adults due patients with MRD at the time of transplant do far worse
to the relatively small stem cell dose, many centers now use than those without MRD, with posttransplant relapse rates
two cord units, which speeds engraftment considerably. A of 65% versus 18%, respectively, although the utility of MRD
major advantage to a UCB is reducing the time from CR to in clinical decision-making is still not clear for AML.
allogeneic HCT. Whereas a typical unrelated donor search Newly diagnosed acute promyelocytic leukemia (APL,
may take 2-3 months, cord units generally are available AML FAB M3) has cure rates in excess of 80% with the all-
within a week. In addition, allogeneic HCT using related trans-retinoic acid (ATRA) in combination with chemother-
haploidentical donors is becoming standard at most centers, apy or arsenic trioxide (ATO). For patients who relapse,
often within the context of a clinical trial. The use of post- salvage therapy is based on prior therapies and duration of
transplant high-dose cyclophosphamide has made the tech- remission (less than or >6 months). Consolidation with
nique accessible to all transplant centers. Bashey et al. autologous or allogeneic HCT is required to maximize cure
performed a retrospective study comparing 53 patients (32% rates in these patients. For patients in molecular remission
AML) undergoing T-cell replete haploidentical HCTs with (lacking detectable PML-RAR1 fusion transcript), autolo-
posttransplant cyclophosphamide with contemporaneous gous HCT offers similar DFS to allogeneic HCT but with
MRD and MUD HCTs. No significant differences were seen significantly superior 5- or 7-year OS of 60%-75% for autol-
in NRM, relapse, 6-month aGVHD incidence, DFS, or OS ogous HCT versus 50%-52% with allogeneic HCT. For those
between the three groups, although significantly less exten- patients with detectable disease after salvage chemotherapy
sive chronic GVHD was seen with haplodientical HCT (38% or relapsing after autologous HCT, allogeneic HCT offers the
haploidentical vs. 54% MRD vs. 54% MUD, p < 0.05 for best opportunity for long-term survival.
both MRD and MUD). Larger studies are needed to better
define the outcomes of haploidentical HCT relative to other
Acute lymphoblastic leukemia (ALL)
stem cell sources as haploidentical donors are often readily
available preventing delays in proceeding to allogeneic HCT, The role of allogeneic HCT in ALL differs greatly between
as seen with MUDs. When umbilical cord blood units or the pediatric and adult population. The prognosis of pediat-
haploidentical donors are used, opportunistic infections ric ALL is excellent with chemotherapy alone resulting in
including rapidly progressive invasive fungal and viral infec- 5-year OS in excess of 80%. Allogeneic HCT in first remis-
tions (eg, CMV, adenovirus) stemming from delayed sion is thus limited to the very high-risk pediatric ALL
immune reconstitution represent a major cause of NRM. populations including children with t(9;22), hypodiploid
Autologous transplantation for AML has been explored karyotype, MLL rearrangement (eg, t(4;11)), a slow response
as consolidation for patients in CR1 or CR2. Patients to therapy, and primary refractory disease. In a retrospective
undergoing autologous transplantation in CR1 may have a study, children with t(9;22) translocation achieved a 65%
decreased rate of recurrence compared with those receiving long-term event-free survival (EFS) after HCT from an
standard chemotherapy but have a higher relapse rate than HLA-identical sibling compared with an approximately 25%
patients undergoing allogeneic transplantation. There is EFS for patients treated with standard chemotherapy regi-
controversy whether the higher relapse rates in autologous mens. Several reports of infants with MLL rearrangements
compared with allogeneic transplantation are from con- treated with allogeneic HCT in first remission have docu-
taminated AML cells in the stem cell product or are from mented EFS ranging between 64% and 76%. This compares

favorably to an EFS of approximately 33% attained with the The role of HCT in later relapses is debatable due to rela-
most aggressive chemotherapy regimens in this setting. tively good results with standard chemotherapy alone.
The superiority of chemotherapy or allogeneic HCT as Despite optimal multiagent chemotherapy for adult ALL,
consolidation for Ph− ALL in CR1 has never been studied in 10% of patients fail to achieve remission after induction and
a randomized fashion, rather donor vs. no donor compari- 40%-60% of patients relapse, historically with very poor
sons have traditionally been performed. In adults, the role of long-term survival outcomes of 5%-10% due in part to a
allogeneic HCT in CR1 is evolving. Some providers have lack of effective salvage options. However, there has recently
reserved allogeneic HCT for those with high-risk ALL. High been a revolution in the treatment of relapsed and refractory
risk often is defined by the following characteristics: positive ALL with the advent of effective novel therapies allowing
minimal residual disease (MRD) at end of induction; poor- more relapsed and refractory patients to proceed to alloge-
risk cytogenetics, such as t(9;22), t(4;11), t(1;19), or complex neic HCT beyond CR1. In addition to repeating initial
karyotype (>4-5 abnormalities); WBC >30,000/mL with the induction in patients with late relapse, several novel treat-
B-cell phenotype; WBC >50,000/mL with the T-cell pheno- ment options for relapsed/refractory ALL are either FDA
type; requiring >4 weeks to achieve complete remission approved or in clinical trials: (i) the bifunctional T-cell
(CR); or age >30-35 years. Approximately 50% of patients engager blinatumomab contains variable regions of anti-
who receive transplantation in first remission become long- CD3 and anti-CD19 monoclonal antibodies joined by a pep-
term survivors. For patients with standard-risk ALL, several tide linker and has CR/CRh rate of 43% in B-lineage ALL in
studies have suggested that allogeneic transplant may yield a 189 patient Phase 2 study; (ii) inotuzumab-ozagamycin is
superior results compared with chemotherapy or autologous an antibody-drug conjugate linking calicheamicin to an
transplantation. The largest prospective trial to date address- anti-CD22 monoclonal antibody with a CR/CRp rate in
ing this question is the MRC UKALLXII-ECOG2993 trial, B-ALL of 49% in a phase 2 study; (iii) liposomal vincristine
which accrued nearly 2,000 newly diagnosed ALL patients had a CR/CRi rate of 20% as monotherapy in a nonrandom-
from 1993 to 2006. Of the 1,031 Philadelphia chromosome– ized phase 2; (iv) nelarabine yielded a CR rate of 36% in
negative patients who obtained CR with frontline therapy, relapsed/refractory T-ALL/LBL; (v) clofarabine-based regi-
the 5-year OS among patients who had a donor undergoing mens yield CR/CRp rates of 40%-60% in small studies; and
allogeneic HCT versus no donor receiving either autologous (v) autologous T-cells expressing chimeric-antigen receptors
HCT or chemotherapy was 53% versus 45%, respectively. A targeting CD19 have shown CR rates of about 90% in phase
recent meta-analysis of 2,962 Ph- ALL patients from 13 stud- 1 studies. All of these therapies may act as bridge to myeloab-
ies comparing chemotherapy with or without autologous lative allogeneic HCT which can yield long-term survival in
HCT to allogeneic HCT showed superior OS in patients about 20%-30% of relapsed/refractory patients. Similar to
under the age of 35 with a matched-sibling donor compared AML, for those with primary refractory ALL or persistent
with the no-donor group due to increased nonrelapse mor- disease in relapse, myeloabaltive allogeneic HCT can be lead
tality in older patients undergoing allogeneic HCT. Autolo- to long-term survival with multiple relapses, CMV+ donor,
gous HCT appeared inferior to chemotherapy alone, bone marrow blast percentage >25%, and older age being
although this result is complicated by the fact that a large risk factors for inferior survival.
percentage of patients randomized to autologous HCT did The role of transplantation in Ph+/BCR-ABL1+ ALL
not undergo the procedure. Notably, there was no difference deserves a special note. Philadelphia-chromosome positivity
in OS when analyzing studies that did not include autolo- leading to the BCR-ABL1 fusion protein has long been con-
gous HCT as part of comparator with allogeneic HCT. In sidered a “high-risk” feature requiring allogeneic HCT, as
addition, improved survival outcomes are being observed few patients were cured with chemotherapy alone. The addi-
with the application of more intensive pediatric-inspired tion of the BCR-ABL1 targeted tyrosine kinase inhibitor
chemotherapy regimens to younger adult Ph-negative ALL (TKI) imatinib to chemotherapy failed to significantly
patients, the very group that had a survival benefit in the improve survival in adults with a 5-year survival of approxi-
meta-analysis. mately 20%. The application of second-generation TKIs
The prognosis of patients with relapsed childhood ALL against BCR-ABL1 (dasatinib, nilotinib), especially when
depends on the site and timing of relapse. Among patients combined with intensive chemotherapy regimens, is chang-
with early marrow relapse (during chemotherapy or within 6 ing outcomes, with recent studies showing survival out-
months of stopping), only 10% achieve long-term EFS with comes comparable to allogeneic HCT although long-term
standard chemotherapy. A retrospective review found that follow up data is needed. Because long-term follow up data
children with relapsed ALL had better EFS with allogeneic with second-generation TKIs are not yet available, the con-
HCT than with chemotherapy alone for early relapse and servative approach is to consider patients with Ph+ ALL for
that, in this population, a TBI-based regimen was superior. allogeneic HCT in CR1.

Relapse after allogeneic HCT for ALL is typically incur- outcomes following a myeloablative transplant with EBMT
able. Donor leukocyte infusion has very low response rates data showing 70% survival for the best score and 20% sur-
in ALL, and its use is controversial. Second allogeneic HCT vival for the worst score.
can be successful in only small percentage of cases due to The advent of TKIs targeting BCR-ABL1 (imatinib, dasat-
high relapse rates (40%-50%) and nonrelapse mortality. inib, nilotinib, bosutinub, ponatinib) revolutionized the
In both pediatric and adult ALL, the detection of MRD by treatment of chronic phase CML and was the first great suc-
flow cytometry or by PCR of clonal TCR or IgH gene rear- cess of targeted chemotherapy. To summarize the salient
rangements is highly predictive of subsequent relapse. points of TKI therapy: (i) Primary therapy for chronic phase
Emerging data suggest that MRD is the strongest predictor is highly effective, and can produce a CCyR in ~80% of cases
of relapse even in ALL with high-risk features. An analysis of with superior response rates but not overall survival for
the GRAALL2003/2005 studies demonstrated that allogeneic second-generation TKIs (dasatinib, nilotinib) compared

HCT in CR1 only benefited Ph-negative ALL patients with imatinib. For these patients, survival at 7+ years is nearly
with positive MRD (≥10−3) at end of induction. In Ph+ ALL, 90%. Approximately 20%-30% of cases will fail primary TKI
similar results are emerging as the addition of the second- therapy; however, either from intolerance, relapse, or pro-
generation TKIs dasatinib and nilotinib to chemotherapy gression to accelerated phase or blast crisis. (ii) For patients
dramatically improve survival compared to similar combi- receiving secondary therapy for resistant disease, approxi-
nations with imatinib. As such, allogeneic HCT should be mately 50% will achieve a CCyR. The survival for these
strongly considered in CR1 for all patients with high levels of patients is ~80% at 3 years. Those who do not achieve and
MRD (generally >10−3 or 10−4 depending on study). It also maintain a CCyR often relapse with resistance mutations in
is clear that patients with detectable residual disease at the BCR-ABL1. (iii) Patients with accelerated phase or blast cri-
time of transplant or after transplant have inferior outcomes sis can achieve a CCyR with TKI therapy, but this does not
due to high relapse rate compared with patients free of appear to be associated with long-term progression-free sur-
detectable disease, although these patients typically do very vival (PFS). (iv) The third generation TKI ponatinib is active
well without allogeneic HCT. against the CML with BCR-ABL1 T315I mutation with a
Engineering of autologous T-cells to express chimeric- CCyR in 66% of chronic-phase CML patients with the T315I
antigen receptors (CARs) targeting CD19 on the surface of mutation. (v) Omecetaxine is active against CML resistant to
B-lineage lymphoblasts is a novel and powerful strategy cur- TKIs, but duration of response is typically short.
rently in clinical trials. Patient T-cells are manipulated in So, which CML patients should be considered for allogeneic
vitro to express a chimeric T-cell receptor containing an HCT? For chronic phase patients, the initial therapy should be
extracellular single-chain variable fragment (scFv) targeting a first- or second-generation TKI. Both the National Compre-
CD19, a transmembrane domain, an intracellular costimula- hensive Cancer Network and the European Leukemia Net-
tory domain(s), and a signaling domain (eg, CD3). The engi- work (ELN) have similar guidelines for monitoring response.
neered cells are then reinfused into the patient where the Using these criteria, roughly 20% of cases will become resis-
cells expand and kill cells expressing CD19, including nor- tant to primary therapy. For those cases that become resistant
mal B-lymphocytes. Complete response rates in relapsed/ to imatinib, roughly 40% will achieve a CCyR with a second-
refractory ALL are >80% with a potentially fatal cytokine generation TKI, and some of these cases eventually will relapse.
release syndrome and persistent B-cell aplasia being con- Transplantation for chronic-phase CML patients can be con-
cerning side effects. sidered in the rare cases of intolerance or resistance to all TKIs.
For patients with accelerated phase disease, allogeneic HCT
should be considered for poor responders to TKI therapy. All
Chronic myelogenous leukemia (CML)
CML patients with blast crisis should proceed to allogeneic
CML is driven by the BCR-ABL1 kinase fusion protein gen- HCT if possible, ideally after successful treatment to CR with a
erated by the t(9;22) translocation/Philadelphia chromo- TKI with or without induction chemotherapy appropriate for
some. Before the development of TKIs targeting BCR-ABL1, lymphoid or myeloid blast crisis.
allogeneic HCT was a standard therapy for CML. Allogeneic In the pre-TKI era, several studies showed that prior ther-
HCT outcomes differ by CML stage with chronic-phase apy with busulfan or interferon before transplant was associ-
CML showing 10-year OS rates of 70%-80%, accelerated- ated with poorer outcomes. This does not appear to be the
phase 30%-40%, and blast crisis ~10%. Pretransplant vari- case with TKIs. Several studies on the effect of prior imatinib
ables define a prognostic scoring system for transplantation and transplant outcomes have failed to show a deleterious
in CML. The system devised by Gratwohl using HLA match, effect of pretransplant imatinib. In addition, there is no evi-
stage, age, sex of donor and recipient, and time from diagno- dence that TKI resistant patients with ABL1 mutations have
sis to transplant is effective in defining posttransplant a poorer outcome following transplantation.

Lastly, CML was one of the earliest uses of a molecular test of cases transplanted in CR have remained in CR after 5
(quantitative RT-PCR of BCR-ABL1 mRNA) to predict sub- years.
sequent relapse following transplantation. For patients with Although high-dose chemotherapy with autologous HCT
detectable molecular, cytogenetic, or morphologic disease or confers high response rates in CLL and reported remission
relapse after allogeneic HCT, treatment generally consists of durations lasting up to 5-6 years, it is not a recommended
withdrawal of immunosuppression and a TKI, omacetaxine, modality for CLL. A randomized study from the European
and/or donor leukocyte infusion. Given the strong graft- intergroup compared autografting with observation in
versus-leukemia effect in CML, these strategies can be very responding patients after first- or second-line therapy.
effective although run a high risk of stimulating graft-versus- Autologous transplant was associated with reduced relapse
host disease. Several studies have used TKI therapy (mostly rates compared with observation (54% vs. 76%, respec-
imatinib) to treat “molecular relapse” posttransplant, with tively), but the OS at 5 years were nearly identical. Moreover,
remarkable effect. In addition, published and ongoing stud- autologous HCT has been associated with posttransplanta-
ies have used TKI prophylactically posttransplant in those tion MDS/AML with an incidence as high as 12%.
cases at very high risk of relapse, such as accelerated and blast As in ALL, chimeric antigen receptor expressing T-cells
phase CML and Ph+ ALL. targeting CD19 have also been used in CLL with success,
although with much lower response rates than those seen in
ALL. An ongoing Phase 2 study in relapsed/refractory CLL
Chronic lymphocytic leukemia (CLL)
has yielded an overall response rate of 35% with 22% of
CLL is the most prevelant adult leukemia in North America patients achieving a CR.
and Europe. Although this disease usually follows an indo-
lent course, it remains incurable with standard therapy.
Myelodysplasia
Allogeneic HCT has high cure rates with low nonrelapse
mortality with the use of reduced-intensity conditioning. Allogeneic transplantation is the only curative therapy for
Although newer therapies such as ibrutinib, idelalisib, MDS. Through the evolution of transplantation regimens, a
obinotuzumab, and bendamustine extend treatment options few general rules apply: (i) results are better for early rather
for relapsed/refractory CLL, allogeneic HCT remains appro- than late-stage disease, (ii) outcomes are worse with poor-
priate for poor-risk or advanced CLL broadly defined as risk cytogenetics or if the patient has a therapy-related
cases with primary resistance to purine analogue-containing myeloid neoplasm that arises subsequent to prior chemo-
therapy or relapse within 24 months of initial therapy. Given therapy or radiation therapy, (iii) matched-related and
the aggressive nature of the disease and intrinsic chemother- unrelated donor transplantation yield similar results, and
apy resistance leading to short remission durations in (iv) fully ablative and RIC regimens offer similar survival
responders, CLL with deletion of 17p14 (with associated p53 outcomes. The major caveat with the last statement, how-
loss) and cases with Richter’s transformation of CLL to ever, is that RIC is more effective in patients with low aber-
DLBCL or Hodgkin lymphoma ideally should undergo allo- rant blast counts, so the comparison is valid only in a
geneic HCT in first remission with curative intent. relatively early stage MDS or MDS in remission.
The rise of RIC transplantation has greatly broadened the Like many diseases, outcomes of transplantation are bet-
use of allogeneic transplantation in CLL, largely supplanting ter in cases of early disease. Thus, cases with refractory ane-
myeloablative approaches, which had a very high NRM mia (RA) have a disease-free survival exceeding 50% (indeed,
(>50%), likely because of the cumulative effects of chemo- this may exceed 70% for International Prognostic Symptom
therapy as well as the older age of the CLL population. A Score [IPSS] 0, and 60% for IPSS 0.5-1). In contrast, patients
number of studies of RIC transplants for advanced CLL have with advanced MDS or secondary MDS have overall survival
been reported and show similar outcomes. The preparative closer to 25%. The differences in DFS are mostly accounted
regimens differ, but generally they are based on fludarabine- for by relapse risk: low in early stage disease (~10% or less)
containing regimens, some with low dose TBI. In general, 4- and common in advanced disease (>40%). As in AML, cyto-
and 5-year data suggest NRM of ~20%, PFS of ~40%, and genetic risk groups largely map to outcome, again princi-
OS of 40%-60%. Of note, two studies have shown that unlike pally dictated by relapse rates after transplantation.
chemotherapy, patients with the p53 mutations or ZAP70 The optimal timing of allogeneic transplant for MDS
expression fare no differently than other risk groups follow- depends on the stage of disease as well as response to sup-
ing RIC allogeneic HCT. Not surprisingly, patients who are portive therapy and hypomethylating agents. A Markov
transplanted with chemosensitive disease and those with model examined three approaches to treatment: transplanta-
low-bulk disease or in remission do considerably better after tion right away, transplantation at leukemic progression, and
transplantation. Indeed, the Seattle group reports that >80% transplantation at a fixed time point (eg, 1 year after

diagnosis). Under this model, the transplant-first option was For patients with relapsed disease, only one randomized
associated with a longer life expectancy in IPSS Int-2 and trial from Europe, known as the CUP (Conventional Che-
high-risk disease, but delaying allogeneic HCT was the opti- motherapy, Unpurged Autograft, Purged Autograft) trial,
mal strategy for low- and intermediate-1 (int-1)-risk disease. has been conducted but closed early due to slow accrual. A
A subsequent study of nearly 400 patients with myeloabla- total of 89 patients with relapsed disease were randomized to
tive and RIC transplants showed that increasing age and a either three cycles of salvage chemotherapy versus high-dose
time from diagnosis of >12 months were associated with an therapy with autologous HCT using in vitro purged or
inferior result. The IPSS risk group classification does a fine unpurged autograft. After a median follow-up of 69 months,
job in risk classification, but the newer five-group classifica- the hazard ratio for PFS was significantly better for both
tion system, the revised IPSS (R-IPSS), better defines prog- autologous HCT arms compared with the salvage chemo-
nosis in the low- and int-1-risk groups an adds a very-high therapy arms, and there was a trend for a superior OS favor-
risk group for whom median survival is 9.6 months in the ing the high-dose therapy arms. No difference was seen in
absence of therapy. An ongoing decision analysis suggests outcomes between purged and unpurged autografts.
that R-IPSS intermediate-, high-, and very-high risk benefit There appears to be a strong graft-versus-lymphoma
from early allogeneic HCT whereas very-low- and low-risk effect in FL, and thus first ablative, and more recently, RIC
disease should delay allogeneic HCT. Based on currently approaches have been used in relapsed disease. Several non-
available data, allogeneic HCT is indicated for all transplant- randomized studies have showed lower relapse rates after
eligible patients with IPSS int-2- or high-risk or R-IPSS allografting compared with autologous HCT, but this gain
intermediate-, high-, or very-high-risk disease at diagnosis, was offset by the considerably higher cost of nonrelapse
those progressing to advanced MDS, and those patients with deaths with the ablative procedure. RIC allogeneic trans-
lower-risk disease but failing supportive care with hemato- plants have been used in FL, including cases failing an autol-
poietic growth factors, immune-suppressive therapy when ogous transplant. Two prospective studies have used
indicated, and/or hypomethylating agents. fludarabine-based RIC conditioning regimens. An MD
In lower-risk MDS, the risk of delaying transplant is pro- Anderson trial reported 6-year PFS and OS rates of 83% and
gression to more advanced MDS or AML leading to: (i) 85%, respectively, with an NRM of 15%. The Cancer and
higher relapse rates after transplantation; (ii) the potential Leukemia Group B (CALGB) trial reported 2-year PFS
that the disease will not be well-controlled at transplanta- and OS rates of 71% and 76%, respectively, with an NRM
tion; (iii) the development of complications related to cyto- of only 7%. Patients with chemotherapy-sensitive disease
penias, transfusions, or progression to AML that may delay before transplantation fared better than patients with
or preclude transplantation; and (iv) the need for induction chemotherapy-resistant disease.
therapy to eliminate blasts prior to transplantation. To be Given encouraging results for both autologous and allo
practical, HLA typing should be initiated after the diagnosis geneic HCT in relapsed follicular lymphoma, which strategy
of advanced MDS and considered in lower-risk patients. If a best serves patients? The EBMT performed a retrospective
matched-related donor is not available, an unrelated donor analysis comparing autologous (n = 726) to RIC allogeneic
search should be started with consideration of hypomethyl- HCT (n = 149) as first transplant strategy in relapsed follicu-
ating therapy as bridge to transplant. lar lymphoma. Relative to autologous HCT, RIC allogeneic
HCT yielded significantly reduced relapse rates and longer
PFS at the expense of increased nonrelapse mortality leading
The lymphomas to equivalent 5-year OS (72% autoHCT vs 69% alloHCT,
p = NS). Patients undergoing RIC allogeneic HCT had
Follicular lymphoma
increased early death compared with autologous HCT and
Follicular lymphoma (FL) typically runs a relatively indolent 2-year cumulative incidence of acute GVHD and chronic
course, but it is incurable with conventional chemotherapy. GVHD of 47% and 52%. For 292 patients relapsing after
The addition of rituximab to frontline and salvage therapies autologous HCT, 56 underwent RIC allogeneic HCT with
has made a significant impact on survival. Thus, transplanta- 3-year PFS and OS of 39% and 50%, respectively.
tion (either autologous or allogeneic) is reserved for salvage The incidence of transformation from FL to diffuse large
therapy. In the pre-rituximab area, the results of three large B-cell lymphoma is ~3% per year, with several studies
randomized trials from Europe suggested improved DFS but reporting a risk of 30% by 10 years of follow-up. Chemo-
no benefit in OS for early remission patients randomized to therapy alone is unlikely to be curative. Autologous trans-
the transplantation arm compared with conventional che- plant has been associated with a 5-year OS of ~40%-60%,
motherapy arm with an incidence of therapy-related MDS with EFS or PFS ranging 25% to 50%. Ablative allogeneic
ranging from 3.5% to 12% in the transplantation arms. transplants have not done better because of high NRM. RIC

approaches being studied, with PFS and OS ranging from for patients who respond to salvage therapy after demon-
20% to >60%, likely owing to the differences in study popu- strating resistant disease to frontline therapy. Salvage che-
lations (particularly chemotherapy responsiveness). motherapy regimens and their results are discussed in the
relevant section of this book.
The introduction of rituximab has improved the progno-
Diffuse large B-cell lymphoma
sis of DLBCL, and rituximab has a growing role in the peri-
The majority of patients with aggressive and very aggressive transplantation management of DLBCL. Rituximab is part
B-cell NHLs can be cured with frontline combination che- of front-line therapy and is typically added to salvage regi-
motherapy with or without consolidative radiotherapy. For mens (eg, R-ICE, R-DHAP) for added cytoreduction and an
patients relapsing after initial chemotherapy, autologous or in vivo purging effect that may reduce the incidence of tumor
allogeneic HCT can be curative, but securing a remission contamination in the autograft. The addition of rituximab to
durable enough to proceed to HCT can be difficult leading upfront therapy appears to have diminished the efficacy of
to numerous studies of autologous HCT in the frontline set- high-dose therapy with autologous HCT, at least in patients
ting. In the pre-rituximab era, the GELA LNH87-2 study with refractory disease or early relapse. The international
randomized diffuse large B-cell lymphoma (DLBCL) patients phase 3 CORAL study randomized refractory or first relapse
in CR1 to consolidation with standard dose chemotherapy DLBCL patients to salvage with R-ICE or R-DHAP followed
or high-dose chemotherapy with autologous HCT. For by high-dose therapy with BEAM (BCNU, etoposide, cytara-
patients with aaIPI high/intermediate or high-risk disease bine, melphalan) with autologous HCT with a second ran-
8-year DFS and OS were significantly improved with up- domization to rituximab maintenance or no maintenance
front autologous HCT. Several subsequent randomized after transplant. Response rates were nearly identical for
studies in the rituximab era have failed to show a consistent R-ICE and R-DHAP with about 50% of patients in each arm
PFS or OS benefit to upfront autologous HCT when com- proceeding to autologous HCT. 3-year PFS was 53% in
pared with standard chemoimmunotherapy except, perhaps, patients undergoing autologous HCT but dropped to only
SWOG 9704 that in a retrospective cohort analysis demon- 23% in patients with prior rituximab exposure who relapsed
strated superior 2-year PFS and OS with autologous HCT in within 12 months of initial therapy. PFS was similar if
IPI high-risk DLBCL. As it stands, upfront autologous HCT patients went to transplant in CR or PR. Notably, rituximab
for DLBCL in CR1 should be reserved for rare high-risk maintenance had no effect on EFS or OS compared with
patients, ideally within the context of a clinical trial. observation alone.
For patients failing to achieve CR with initial therapy or Several prognostic factors are associated with outcome
for patients with relapsed disease, standard-dose salvage after autologous HCT. The International Prognostic Index
therapy with chemotherapy alone is not curative. High- (IPI) is the validated scoring system designed to predict
dose chemotherapy with autologous HCT offers curative survival of patients with newly diagnosed aggressive NHLs.
potential and is the treatment of choice for patients with The IPI at relapse (second-line IPI), however, also has been
relapsed, chemotherapy-sensitive DLBCL. Autologous shown to correlate with prognosis after autologous HCT.
HCT is the standard of care for most patients with In addition, positron emission tomography (PET) scan-
chemotherapy-sensitive relapsed or refractory DLBCL. The ning appears to have predictive value. Several studies have
international, multicenter, prospective PARMA trial estab- shown that PET positivity after salvage therapy is associ-
lished the role of autologous HCT for patients with relapsed, ated with an inferior failure-free survival independent of
chemotherapy-sensitive DLBCL. In this trial, 109 of 215 age-adjusted IPI. Other poor prognostic features include
patients who had relapsed DLBCL and who responded to relapse within 12 months of diagnosis, advanced stage,
platinum-based salvage chemotherapy were assigned ran- poor performance status, and failure to achieve CR after
domly to four more courses of conventional chemotherapy transplantation.
or autologous HCT. The 5-year EFS and OS were 46% and Allogeneic HCT is not offered routinely to patients with
53%, respectively, for the transplantation arm and 12% DLBCL. Exceptions include select young patients with
and 32%, respectively, for the chemotherapy arm. Patients advanced disease, patients who failed to mobilize adequate
with relapsed DLBCL who were chemotherapy-sensitive CD34+ hematopoietic cells, or patients who failed a previ-
unequivocally fared better compared with patients with ous autologous HCT. In a review of 101 patients with
chemotherapy-resistant disease. DLBCL who failed an autologous transplant, 3-year NRM
Patients with DLBCL who demonstrate primary refrac- was 28% (higher in myeloablative vs. RIC), relapse was
tory disease or relapsed disease that is not responsive to sal- 30%, and OS 52%. Time to relapse of <12 months and che-
vage chemotherapy have poor outcomes even after high-dose motherapy refractory disease at transplant portended a
therapy with autologous HCT. Autologous HCT is indicated worse outcome.

Peripheral T-cell lymphomas potential yielding EFS rates ranging from 50% to 85% even in
patients who failed a prior autologous HCT.
Peripheral T-cell lymphomas (PTCLs) account for 10% of
non-Hodgkin lymphomas and are generally aggressive lym-
phomas. With the exception of ALK+ anaplastic large cell Classical Hodgkin lymphoma
lymphoma, which has 5-year OS rates of 60%-70% with
Front-line therapy for classical Hodgkin lymphoma has high
CHOP or CHOP-like chemotherapy alone, PTCLs tend to
cure rates. For the unfortunate patients with relapsed or
have poorer response to and shorter survival after chemo-
refractory disease, high-dose therapy with autologous HCT is
therapy alone compared with diffuse large B-cell lymphoma,
the standard of care and confers cure rates of 40%-60% in
The 5-year OS of these PTCLs is approximately 40%. Out-
patients with relapsed, chemotherapy-sensitive disease and
comes with autologous HCT are good but not clearly supe-
25%-40% in patients with chemotherapy-refractory disease.
rior to chemotherapy, especially in patients achieving a CR.
Maintenance therapy after autologous HCT has also been
For relapsed or refractory disease, autologous HCT yields
evaluated for classical Hodgkin lymphoma. Brentuximab-
5-year survival rates of about 40% in chemosensitive disease
vedotin (BV) is an antibody-drug conjugate targeting CD30
with similar outcomes observed for allogeneic HCT with
on the cell surface and is FDA approved for classical Hodgkin
myeloabaltive or RIC conditioning.
lymphoma relapsing after autologous HCT and relapsed/
refractory CD30+ anaplastic large cell lymphoma. The
AETHERA trial, a recent randomized, double-blind, placebo-
Mantle cell lymphoma
controlled phase 3 study comparing 16 BV treatments after
Mantle cell lymphoma is an uncommon lymphoma (5%- autologous HCT to placebo for classical HL at high risk for
10% of lymphomas) with a male predominance that gener- relapse or progression, demonstrated superior PFS (HR 0.57,
ally presents with advanced disease. Traditional, less P = 0.001) with BV but at the expense of increased sensory
aggressive chemotherapy alone (eg, CHOP) does not offer and motor neuropathy, grade 3-4 neutropenia, and a low rate
durable disease control in most cases and results in a median (2 cases) of fatal acute respiratory distress syndrome attribut-
OS of approximately 3 years. Most younger, newly-diagnosed able to BV. Like other maintenance studies in lymphoma,
patients receive aggressive therapy with regimens using however, no overall survival benefit has been observed to
rituximab, cyclophosphamide, vincristine, doxorubicin, and date, perhaps in part due to effective salvage with BV. The lack
prednisone with the addition of high-dose cytarabine con- of an OS benefit suggests a limited if any role for posttrans-
taining courses (eg, CALGB 59909, R-HyperCVAD, Nordic plant brentuximab vedotin given the overtreatment of a large
MCL-2) which yield overall response rates of ~90%. High- number of patients who would never have needed the drug
dose therapy with autologous HCT is typically part of con- with resultant toxicities and the high expense of the drug.
solidation therapy yielding 5-year PFS of 50%-70% and The role of allogeneic HCT in classical Hodgkin lym-
5-year OS 60%-70% with low rates of relapse after 5 years, phoma is less established and generally is pursued only in
although it is still not clear if autologous HCT is curative in patients who have marrow involvement, refractory disease,
a portion of patients. Higher proliferation fraction as mea- or relapsed or progressive disease after an autologous HCT.
sured by Ki67 immunohistochemistry is associated with In general, RIC transplant regimen is preferred to a fully
shorter EFS. Rituximab maintenance after front-line autolo- ablative regimen, as RIC is associated with fewer regimen-
gous HCT for mantle cell lymphoma is currently under related deaths and better survival in a population which has
study. The phase 3 LyMa study randomized 299 mantle cell typically previously undergone a myeloablative autologous
lymphoma patients to rituximab or observation. A recent HCT. The Gruppo Italiano retrospectively compared nearly
planned interim analysis showed improved 3-year PFS with 200 Hodgkin cases following a failed autologous transplant,
rituximab maintenance (93% vs 82%; P = 0.015) but no dif- and divided the patients into those with a donor (sibling,
ference in 2-year OS (93% vs 94%, NS), suggesting no role unrelated, or haploidentical) versus those that could not
for rituximab maintenance after upfront autologous HCT at secure a donor, with the intent that those with donor would
this time. have a RIC allogeneic HCT. The 2-year PFS and OS were
For patients with relapsed or refractory disease, allogeneic superior in the donor group (39% vs. 14% and 66% vs. 42%,
HCT is indicated as it offers the only chance for long-term sur- respectively). The Seattle group has compared the outcome
vival. Myeloablative allogeneic HCT can induce durable remis- for HLA-matched, unrelated matched, and haploidentical
sions in mantle cell lymphoma, even in heavily pretreated donors in RIC allogeneic HCTs and has found survival to be
patients, but is associated with significant treatment-related similar in all approaches, with overall survival of ~60%, and
mortality and morbidity. Reduced-intensity conditioning regi- PFS of ~40%. Chemosensitivity before RIC alloHCT pre-
mens reduce toxicity without significantly sacrificing curative dicts reduced risk of relapse.

The focus of first-line therapy for pediatric patients has at relapse with stem cells collected at diagnosis. Early trans-
been to decrease the toxicity of the current therapy while still plantation significantly improved PFS, but there was no dif-
achieving high cure rates. For children and adolescents who ference in OS. Early transplantation, however, was associated
fail conventional chemotherapy, autologous HCT is offered with a shorter period of chemotherapy and hence improved
as a curative option. For those with bone marrow involve- quality of life. Given the advent of new drugs like lenalido-
ment or with relapse following an autologous transplanta- mide and bortezomib, it is no longer clear that there is a ben-
tion, allogeneic transplantation has been performed and efit to autologous HCT in first response or what groups of
appears to be better tolerated than in the adult population. myeloma patients might benefit. An ongoing randomized
phase 3 is comparing initial therapy with lenalidomide, bort-
ezomib, and dexamethasone with or without high-dose mel-
Plasma cell dyscrasias
phalan with autologous HCT in untreated multiple myeloma
Multiple myeloma is the most common indication for to determine if the survival benefit of autologous HCT as
autologous HCT. Compared with chemotherapy, high-dose first consolidation remains since the advent imids and pro-
therapy with autologous HCT is associated with higher teasome inhibitors.
response rates and improved DFS and OS. When autologous Currently, the utility of tandem transplantation, either
HCT is given as part of the planned frontline treatment, auto-auto or auto-allo, as part of primary treatment remains
~22%-44% patients achieve CR, with median time to pro- controversial. In a randomized study from the French group,
gression and OS of 18-24 months and 4-6 years, respectively. both response rates and survival favored tandem autologous
High-dose melphalan alone at a dose of 200 mg/m2 is the transplantation, in particular for patients with significant
most commonly used preparative regimen for patients with residual disease (the lack of at least a VGPR) after their first
multiple myeloma undergoing autologous HCT. The proce- transplant. Event-free, relapse-free, and overall survival were
dure is well tolerated, with a treatment-related mortality of 10%, 13%, and 21%, in the single-transplant group, com-
~2%. The PFS and OS benefit of upfront autologous HCT pared with 20%, 23%, and 42% in the tandem transplant
relative to chemotherapy alone appears to continue in the group. A recently published meta-analysis of six randomized
era of thalidomide and lenalidomide. Palumbo et al. ran- trials with ~1,000 patients concluded that tandem autolo-
domized newly-diagnosed multiple myeloma patients to gous HCT confers higher response rates compared with sin-
200-mg/m2 melphalan with autologous HCT or chemother- gle autologous HCT, but it did not find conclusive evidence
apy alone with melphalan, prednisone, and lenalidomide for improvement in PFS or OS. A registry analysis from the
(MPR) followed by a second randomization in each arm to EBMT group, however, demonstrated that when a second
maintenance lenalidomide or no maintenance. Two major transplantation is performed within 3-6 months after the
results of the study were a significantly improved OS with first transplantation, survival is improved. An ongoing NIH/
autologous HCT relative to MPR alone (HR 0.55 P = 0.02) Intergroup study is studying single versus tandem autolo-
and improved PFS but not OS with lenalidomide mainte- gous HCT in a randomized fashion.
nance versus no maintenance. Relapse is the overwhelming cause of autologous HCT
The advent of the immunomodulators, imids such as tha- failure. Allografting potentially provides a stem cell source
lidomide and lenalidomide, and proteasome inhibitors, such free of myeloma cells and a graft-versus-myeloma effect, but
as bortezomib, as part of frontline treatment for myeloma myeloablative allogeneic HCT has been associated with
has changed the treatment paradigm. The newer agents unacceptably high treatment-related mortality. Thus, there
result in more patients attaining CR, near CR, and very good has been interest in using RIC transplants after an autolo-
partial response with frontline therapy. Current guidelines gous transplant. Numerous studies have compared tandem
state that high-dose chemotherapy with autologous SCT autologous HCT to autologous HCT followed by RIC alloge-
should be offered as initial consolidation therapy in patients neic HCT as part of upfront therapy for multiple myeloma
with newly diagnosed myeloma who are <65 years old and with mixed results. Armeson et al. conducted a meta-analysis
have a good performance status. It appears, however, that of 6 trials comprising 1,192 subjects undergoing tandem
patients with adverse prognostic features at diagnosis, such autologous HCT and 630 subjects undergoing autologous
as high-serum β2-microglobulin or an unfavorable karyo- HCT followed by RIC allogeneic HCT. T reatment-
type, such as del13 and del17p, still have a poor outcomes related mortality was significantly higher in the allogeneic
even after tandem (double) autologous HCT. HCT group without any benefit seen for PFS or OS with the
Because of concerns over the potential toxicities associ- autologous-allogeneic HCT strategy. At this time, single
ated with HCT, a strategy of delayed transplantation is autologous HCT after response to primary therapy remains
undergoing continued study. In a French randomized study, the standard at most institutions, and other approa ches
upfront transplantation was compared with transplantation are best performed in the setting of a clinical trial.

Allogeneic HCT remains an option for patients relapsing patients had a severe, rapidly progressive sensorimotor poly-
after autologous HCT. neuropathy, and 9 patients were wheelchair dependent. All
Multiple myeloma was the first disease to demonstrate a 14 evaluable patients achieved neurologic improvement or
clear disease progression benefit with maintenance therapy stabilization. Other symptoms also improved substantially.
after autologous HCT. In two large randomized trials, post- Autologous HCT results in significant clinical improvement
transplant lenalidomide therapy resulted in significant in a majority of patients and should be considered a thera-
improvements in DFS. In a randomized placebo-controlled peutic option in these patients.
North American trial reported by McCarthy et al., lenalido-
mide maintenance led to significantly improved time to pro-
Aplastic anemia and other autoimmune
gression (46 months for lenalidomide versus 27 months for
diseases
placebo, P < .001) and OS (two-sided p = 0.03, 3-year OS
88% vs 80%). The IFM2005-02 study randomized myeloma Therapy for aplastic anemia depends on the severity of the
patients to lenalidomide or placebo after autologous HCT aplasia, the availability of a matched-related donor, and the
and found a similar PFS benefit but no difference in OS with age of the patient. The standard first-line therapy for newly-
lenalidomide maintenance. In both studies, lenalidomide diagnosed severe aplastic anemia is allogeneic HCT if a
had more hematologic adverse events, and secondary pri- matched-related donor is available. If a matched-related
mary cancers occurred more frequently in lenalidomide donor is not available, immunosuppressive therapy (IST)
maintenance patients compared with placebo. with anti-thymocyte globulin (ATG) and cyclosporine is
AL (light chain amyloid) amyloidosis is a clonal plasma used for initial therapy with unrelated donor transplant
cell disorder characterized by tissue deposition of amor- reserved for patients who do not adequately respond to IST.
phous extracellular material composed in part of immuno- Long-term survival following a matched-related donor
globulin light- or heavy-chain fragments in many vital transplant exceeds 80%. Inferior survival is associated with
organs, such as the heart, lung, kidney, liver, and CNS. This older age, use of an unrelated donor, and prior transfusion.
infiltrative process ultimately leads to organ failure and The main complication of transplant is related to chronic
death. The prognosis of patients with AL amyloidosis is GVHD which, unlike in hematologic malignancies, has no
poor, with median survival of ~1-2 years. Although conven- benefit in terms of reduced risk of relapse. Thus, bone mar-
tional chemotherapy has limited utility in patients with AL row rather than peripheral blood is the highly preferred
amyloid, autologous HCT can reverse the disease process for source of stem cells. In regards to preparative regimen, most
selected patients. Nonimmunoglobulin forms of amyloido- use high-dose cyclophosphamide (50 mg/kg × 4 doses) with
sis, however, will not benefit from cytotoxic therapy, includ- ATG, although regimens incorporating fludarabine with
ing transplantation. Because of the preexisting organ ATG and lower doses of cyclophosphamide are highly effica-
dysfunction in patients with AL amyloidosis, the NRM of cious with less toxicity.
autologous HCT is still 4-8 times (NRM ~25%) higher com- For younger patients (often defined as <40 years of age)
pared with that of autologous transplantation for multiple with newly diagnosed idiopathic severe aplastic anemia and
myeloma (NRM ~2%). The causes of NRM include gastro- an HLA-identical sibling, many centers recommend imme-
intestinal bleeding, cardiac arrhythmias, and the develop- diate transplantation to minimize alloantigen sensitization
ment of intractable hypotension and multiorgan failure. with transfusions, which historically has resulted in an
Several studies have suggested a 2-3 year survival of ~70%, increased risk of graft rejection and poorer outcomes.
following autologous transplant, although patients with Although the use of cyclosporine, as well as the use of leu-
multiorgan involvement have a distinctly worse survival. A kodepleted blood products, has reduced the problem of
phase 3 trial in which AL amyloid patients were randomized rejection, sensitization should be minimized through strict
to receive autologous SCT versus oral melphalan and high- avoidance of transfusions when possible and avoidance of
dose dexamethasone suggested a benefit of the conventional directed family donations of blood products. Indeed, one
chemotherapy arm because of the high NRM of 24% in the large study showed the hazard ratio for mortality was 1.7 for
transplant arm. After a median follow-up of 3 years, the OS patients who received IST before transplant, compared with
was significantly longer in the conventional-dose group (57 those patients who underwent front-line transplantation.
vs. 22 months; P = 0.04). Secondary malignancies occur after transplantation for
POEMS syndrome a rare condition characterized by poly- severe aplastic anemia in as many of 10% of cases 15 years
neuropathy, organomegaly, endocrinopathy, monoclonal from transplant. Risk factors include age >15 years, use of
gammopathy, and skin changes, as well as a clonal plasma cyclosporine (CSP) in an IST regimen before transplant, and
cell disorder. Investigators from the Mayo Clinic performed perhaps radiation therapy as part of the transplant regimen
transplantation in 16 patients with POEMS syndrome; 15 (no longer preferred, as noted).

It is important to assess for Fanconi anemia and dyskera- Hemoglobinopathies

tosis congenita as congenital causes of bone marrow failure
Thalassemia major
in newly diagnosed aplastic anemia patients to help select the
appropriate treatment course and avoid the futile adminis- The Pesaro team has pioneered transplantation for thalas-
tration of immune suppressive therapy that is standard in semia and has reported high cure rates. Three factors predict
idiopathic aplastic anemia. Patients with Fanconi anemia adverse transplantation outcomes: hepatomegaly (>2 cm
frequently do not have all of the stigmata of the disease, and below the costal margin), hepatic fibrosis, and irregular che-
the diagnosis is overlooked easily. The sensitivity of patients lation. Quality chelation therapy is defined as deferoxamine
with Fanconi anemia to alkylating agents is well known, and therapy initiated <18 months after the first transfusion and
transplantation can be done successfully using only reduced- given for >5 days each week. Class I patients have none of
intensity regimens. Recent trials have focused on reducing these factors, class II patients have one or two factors, and
radiation exposure in addition to reducing doses of alkylat- class III patients have all three factors. For class I patients
ing agents in these patients. Patients with Fanconi anemia <17 years of age, survival, thalassemia-free survival, NRM,
are at high risk for solid tumors, especially following radia- and recurrence of thalassemia were 94%, 87%, 6%, and 7%,
tion exposure. Dyskeratosis congenita is a congenital bone respectively. The rates of survival, thalassemia-free survival,
marrow failure syndrome caused by mutations in telomerase NRM, and recurrence of thalassemia were 84%, 81%, 15%,
or telomerase-associated genes. The vast majority of patients and 4%, respectively, for class II patients. Patients with class
develop abnormal skin pigmentation, nail dystrophy, and III disease have more complications and a higher rate of graft
oral leukoplakia. Bone marrow failure occurs early in life rejection. The probability of thalassemia-free survival for
and is the leading cause of early mortality necessitating allo- young patients who are in class III is 62%, and the risk of
geneic HCT to prolong life in most patients. RIC allogeneic dying is 35%. RIC regimens have been investigated in these
HCT minimizing the use of alkylators and radiation in con- patients. Class III adults receiving reduced-dose condition-
ditioning is preferred, as patients are prone to pulmonary ing appear to have a lower rate of rejection. The Pesaro team
fibrosis, hepatic cirrhosis, and secondary malignancies. As noted a 24% chance of rejection if the individual has received
for other inherited disorders, siblings should be screened for >100 transfusions, compared with a 53% chance in patients
the recipient’s bone marrow failure syndrome. who have received fewer transfusions.
Given its immunosuppressive properties, autologous The optimal source of stem cells for patients with hemo-
transplantation has been studied as treatment for life- globinopathies is still under investigation. To avoid chronic
threatening autoimmune disorders. Autologous HCT has GVHD, the use of bone marrow rather than PBSCs has been
been used in multiple sclerosis, systemic sclerosis, rheuma- advocated. For those lacking sibling donors, unrelated and
toid arthritis, juvenile idiopathic arthritis, systemic lupus cord blood donor transplantations have shown promising
erythromatosis, dermatomyositis/polymyositis, Crohn dis- results in both pediatric and adult patients, provided donor
ease, and autoimmune cytopenias. The therapeutic rationale compatibility is stringent.
for these transplantations is that high-dose chemotherapy Cord blood transplantation has been used in cases with-
may eradicate or modulate clones of autoreactive T-cells. out a matched sibling or unrelated donor. The 2-year prob-
Although the integration of this approach into treatment of ability of survival after cord blood transplantation for
each disease will depend on the results of ongoing trials, some children with thalassemia was 79% in 33 patients who
general observations are now possible. First, allogeneic HCT received transplantation. Unfortunately, nearly a quarter
has considerable treatment-related morbidity and mortality rejected the graft.
in this population and is not typically used outside of a clini- Patients with thalassemia major frequently develop mixed
cal trial. Second, the underlying organ dysfunction often pro- chimerism following transplantation, which often leads to
gresses acutely during transplantation even if there is stability marked improvement in their transfusion requirements.
or improvement later. Third, durability of response and the The patients remain at risk for graft rejection, however, espe-
need for continued therapy after HCT remain to be defined. cially those whose percentage of host cells remains >25%.
The waxing and waning course of autoimmune disorders
make it difficult to define end points in these diseases. Results
Sickle cell disease
that have been considered encouraging in the transplantation
literature have been considered disappointing (both regard- Allogeneic transplantation is a promising therapy for sickle
ing the rates of response and toxicity) in the rheumatology cell disease. Results from children who have received trans-
literature. Nonetheless, patients with aggressive autoimmune plantation from HLA-identical siblings have a >90% survival
disorders should consider clinical trials and examine this rate, and 85% are disease free. Moreover, successful allogeneic
approach as one of their treatment options. HCT appears to prevent further sickle cell complications.

A study from Belgium demonstrated that patients who vitro function. HLA typing should be undertaken as soon as
received transplantation early in the course of their disease a diagnosis of SCIDS or other combined deficiency poten-
(less than four blood transfusions) had a 100% survival rate tially correctable by HCT is established. Allogeneic HCT
and 93% DFS rate compared with an 88% overall survival approaches are modified based on the exact diagnosis. The
rate and 80% DFS rate in patients who received transplanta- need for a preparative regimen and its intensity of condi-
tion later in the course of their disease. tioning are determined in part by the function of the lym-
Despite these successes, many recommend reserving phocytes and NK cells.
transplantation for children at high risk from their sickle cell Allogeneic HCT is undertaken in these disorders to pro-
disease because of the toxicities and risks. Frequently, how- vide a stable source of immunologically competent cells. The
ever, children at significant risk are not identified until they major complications are rejection of the marrow graft and
have suffered end-organ damage, including stroke or severe GVHD. Graft rejection occurs when sufficient immune
lung injury. In addition, the clinical course for a patient may function remains for the recipient to mount a cellular
vary over time. Attempts to identify risk factors of severe dis- immune response against donor HLA molecules. In some
ease have suggested high WBC count, severe anemia, and forms of SCIDS with absent T-cell function, such as X-linked
early dactylitis as surrogate markers. But the ability to pre- SCIDS, Janus kinase-3 (Jak3) deficiency, and complete
dict the clinical course for each individual remains elusive. recombination activation gene-1 (RAG-1) and recombina-
In addition, finding suitable, unaffected sibling donors has tion activation gene-2 (RAG-2) deficiencies, the patient is
been difficult. In one study, only 14% of patients with sib- unable to reject the hematopoietic cells. In these patients,
lings had a suitable HLA-matched donor. simple infusion of hematopoietic cells is usually all that is
Nonmyeloablative allografting has been studied in adults. required, without a preceding preparative regimen. Many of
A preparative regimen including pretransplant alemtu- the recipients who received hematopoietic cells without a
zumab (an antibody therapy to CD52, which reduces host preparative regimen fail to develop normal B-cell function
B- and T-cells), 300 cGy of TBI, and posttransplant siroli- and required ongoing IgG replacement with intravenous
mus following HLA-matching sibling CD34+ PBSC infusion immunoglobulin. This has led many centers to tailor the
has been used to remarkable effect. All 10 patients were alive preparative regimen to include some chemotherapy, most
at 30 months of follow-up, and 9 of 10 patients had stable- recently fludarabine, to attempt to ensure full immune
donor chimerism. Remarkably, there were no cases of acute reconstitution. Patients with adenosine deaminase defi-
or chronic GVHD. ciency, the largest subset of this group, require a preparative
regimen despite the absence of detectable T-cell function
because the donor lymphocytes may rescue the host cell
Immune deficiency disorders
function, thus allowing for ultimate graft rejection. Patients
Many immune deficiency disorders become evident in with normal NK cell activity (including some X-linked, Jak3,
infancy secondary to an increased rate of infections or to the and RAG defects) also often require preparative regimens,
presence of opportunistic infections. In such cases, the pos- again emphasizing the need to determine the exact defect
sibility of HIV infection must be ruled out. For patients sus- before initiating therapy.
pected of having a primary immune deficiency, definitive Results of transplantation are best for children receiving
diagnosis of the exact molecular defect is important to pre- HLA-identical sibling transplantations, with survival rang-
dict the course of the disease and to be able to tailor therapy ing from 70% to 100%. For patients lacking a sibling donor,
appropriately. The most common diseases for which alloge- results have ranged from 30% to 50%. In the past, many
neic transplantation is indicated include severe combined patients lacking a sibling donor have received haploidentical
immunodeficiency syndrome (SCIDS), adenosine deami- grafts from a parent, although the increasing availability of
nase deficiency, Wiskott-Aldrich syndrome, Nezelof syn- cord blood stem cells provides another option. Cord blood
drome, Omenn syndrome, MHC antigen deficiency, stem cells are particularly appealing because they can be
leukocyte adhesion defect, Chédiak-Higashi syndrome, accessed readily and are not infection carriers, decreasing the
chronic granulomatous syndrome, and DiGeorge anomaly. risk of CMV disease and EBV lymphoproliferative disorders
Newborns known to have or to be at high risk for severe after transplantation.
SCIDS should be isolated at birth because infection increases
the risk for complications of allogeneic HCT. Evaluation of
Inherited metabolic disorders
early complete blood counts may suggest a neutrophil (neu-
trophil adhesion disorder or Kostmann syndrome) or lym- A number of inborn errors of metabolism have been corrected
phocyte disorder, such as SCIDS. Cord blood, when with allogeneic HCT. One of the most important steps is the
available, should be studied for lymphocyte numbers and in early identification of the disorder before the development of

end-organ damage. The role of transplantation varies accord- disease. In an EBMT group prospective study, patients were
ing to the disorder identified. For instance, certain storage dis- randomized between four cycles of etoposide, ifosfamide,
orders such as Niemann-Pick type IA are not treatable by and cisplatin (VIP) versus three cycles of VIP plus a single
transplantation. Other disorders such as globoid cell leuko- cycle of high-dose therapy followed by autologous HCT. The
dystrophy, metachromatic leukodystrophy, adrenoleukodys- 3-year EFS for patients who received VIP only was 35% ver-
trophy, mannosidosis, fucosidosis, aspartylglucosaminuria, sus 42% for patients randomized to transplantation, with no
Hurler, Hunter, Maroteaux-Lamy, and Sly syndromes, and difference in OS. A US intergroup phase 3 randomized study
Gaucher disease type III have been treated successfully with failed to demonstrate any benefits in high-dose therapy for
allogeneic HCT. Siblings and parents should be HLA typed as patients with newly diagnosed intermediate- or poor-risk
soon as possible. For some of these disorders, transplantation germ cell cancer.
using marrow from a donor heterozygous for the trait will not
cure the disease. For those lacking a suitable related donor, the
Pediatric solid tumors
best donor source is unclear. The pace of the disease may make
the time required for the typical search for a matched unre- Many pediatric solid tumors demonstrate exquisite chemo-
lated donor unrealistic, thus making cord blood cells more sensitivity, leading to the exploration of HCT as a method of
attractive in these cases. GVHD in some of these disorders (eg, dose intensification for children presenting with high-risk or
adrenoleukodystrophy) may accelerate their disease process recurrent disease.
and increase the risk of rapid decline. The timing of the trans-
plantation may be difficult because not all patients with the
Neuroblastoma
same apparent diagnosis have the same course of disease.
Thus, in adrenoleukodystrophy, some patients have rapid neu- In 1999, the Children’s Cancer Group reported a study of
rologic decline at an early age, whereas others may not mani- >500 patients with high-risk neuroblastoma (defined as age
fest symptoms until later in childhood, adolescence, or >1 year, metastatic disease, amplification of MYCN onco-
adulthood, if at all. In a number of these disorders, HCT will gene, and histologic findings). All patients were treated with
halt the disease progression, but the patient may not regain the same initial regimen of chemotherapy, and those with
lost milestones or function and may actual show more rapid progression of disease were assigned randomly to more che-
deterioration. motherapy or HCT using purged autologous bone marrow.
Patients still without disease progression were then random-
ized to differentiation therapy with 13-cis-retinoic acid or no
HCT for solid tumors further therapy. The 3-year EFS was superior for the HCT
group (34% vs. 22%). Among patients assigned to receive
Germ cell cancer
cis-retinoic acid and HCT, EFS was 55% versus 18% in those
Germ cell cancer is highly curable, even in patients with dis- assigned to chemotherapy and no cis-retinoic acid. More
seminated disease. Although conventional-dose cisplatin- recently, the use of purged mobilized PBSCs has replaced
based chemotherapy cures the majority of patients, patients purged bone marrow at many centers and is associated with
presenting with advanced disease have a somewhat higher decreased HCT-related mortality. Ongoing studies are inves-
rate of recurrence. Some patients at first relapse can achieve tigating additional HCT-related strategies to further improve
a durable remission with salvage chemotherapy, but most of the outcome of high-risk patients, such as the use of seq
the patients who failed salvage chemotherapy or had uential autologous transplantations and combination thera-
cisplatin-refractory disease ultimately died of the disease. pies with high-dose radiopharmaceutical agents such as
Approximately 15%-20% of patients with multiply relapsed iodine-131 meta-iodobenzylguanidine.
or overtly cisplatin-refractory germ cell cancer, however, can
be cured with high-dose carboplatin and etoposide followed
Ewing sarcoma
by autologous HCT. In a large retrospective study, progres-
sive disease before transplantation, primary mediastinal Like neuroblastoma, the Ewing sarcoma and primitive neu-
tumor, refractoriness to conventional-dose cisplatin, and roectodermal tumor family includes chemotherapy-sensitive
human chorionic gonadotropin levels >1,000 IU/L before and radiotherapy-sensitive tumors. High-risk features of
transplantation predicted transplantation failure. The esti- Ewing sarcoma include a large primary tumor 0.8 cm in
mated 2-year survival rates were 51% and 5% for patients diameter, pelvic location of the primary tumor, and presence
with no risk factors and multiple risk factors, respectively. of overt metastatic disease at diagnosis. Patients with meta-
Transplantation has been investigated as consolidation static Ewing tumors have a DFS rate of 20% when treated
therapy after initial treatment of patients with advanced with conventional therapy. Dose intensification with stem

cell support has been tried in Ewing sarcoma patients, but Thomas ED, Buckner CD, Clift RA, et al. Marrow transplantation
several large retrospective studies have failed to show a clear for acute nonlymphoblastic leukemia in first remission. N Engl
benefit from autologous HCT compared with conventional J Med. 1979;301:597-599. Landmark article in the field that led
therapies. In a study from the National Cancer Institute, 91 to the increased use of marrow transplantation in patients with
leukemia and early phases of their disease.
patients were enrolled on a series of three protocols consist-
Thomas ED, Lochte HL Jr, Lu WC, Ferrebee JW. Intravenous
ing of induction chemotherapy, radiation to the primary site,
infusion of bone marrow in patients receiving radiation and
consolidation with TBI (8 Gy), and autologous HCT. In this
chemotherapy. N Engl J Med. 1957;257:491-496.
group, 79% of the patients achieved a CR with surgery, local Thomas ED, Lochte HL Jr, Cannon JH, Sahler OD, Ferrebee JW.
radiation, and chemotherapy; 90% of eligible patients pro- Supralethal whole body irradiation and isologous marrow
ceeded to transplantation; and 30% survived long term with- transplantation in man. J Clin Invest. 1959;38:1709-1716. These
out progression of disease. Although this proportion is two seminal papers by Dr. E. D. Thomas mark the dawn of the era
higher than expected for a poor-prognosis group of patients, of modern stem cell transplantation.
this may represent selection of a chemotherapy-sensitive
better risk group because only patients who did not progress
The hematopoietic stem cell
after chemotherapy were eligible for autologous HCT.
Bryder D, Rossi DJ, Weissman IL. Hematopoietic stem cells:
the paradigmatic tissue-specific stem cell. Am J Pathol.
Summary 2006;169:338-346. A nice review on hematopoietic stem cell
properties.
HCT is a rapidly evolving field. Results have improved over Lengerke C, McKinney-Freeman S, Naveiras O, et al. The cdx-hox
the past decades, and indications for and timing of HCT pathway in hematopoietic stem cell formation from embryonic
have changed. Transplantation is more widely applicable stem cells. Ann N Y Acad Sci. 2007;1106:197-208. A description
because of improvements in supportive care and donor of the challenges of deriving HSCs by in vitro differentiation of
embryonic stem cells and the role of cdx-hox gene pathways in
selection and the advent of RIC regimens. For patients with
enhancing efficiency.
malignant diseases, especially the lymphoid malignancies,
Orkin SH, Zon LI. Hematopoiesis: an evolving paradigm of stem
the chance for a better outcome is significantly improved if cell biology. Cell. 2008;132:631-644. An up-to-date review of
patients are referred when their disease still demonstrates concepts in hematopoiesis.
chemotherapy sensitivity. For most of the other indications, Raaijmakers MH, Scadden DT. Evolving concepts on the
it is important to identify high-risk features or poor prog- microenvironmental niche for hematopoietic stem cells. Curr Opin
nostic factors at the time of diagnosis to help determine the Hematol. 2008;15:301-306. A current perspective article incorporating
optimal timing for HCT. recent new findings in the field of hematopoietic niche biology. This
includes a summary of current knowledge about key components of
the niche, as well as a discussion regarding the potential role of the
Bibliography niche in hematologic disease and as a therapeutic target.
Rossi DJ, Bryder D, Zahn JM, et al. Cell intrinsic alterations
Historical perspective underlie hematopoietic stem cell aging. Proc Natl Acad Sci
Appelbaum F. Hematopoietic stem cell transplantation at 50. USA. 2005;102:9194-9199. Provides data regarding age-related
N Engl J Med. 2007;357:1472-1475. Excellent review on the differences in HSCs.
historical evolution of hematopoietic stem cell transplantation. Tavian M, Peault B. Embryonic development of the human
Barnes DWH, Corp MJ, Loutit JF, Neal FE. Treatment of murine hematopoietic system. Int J Dev Biol. 2005;49:243-250. A
leukaemia with X-rays and homologous bone marrow. BMJ. comprehensive review of the ontogeny of blood cell development in
1956;ii:626-627. humans, including comparison to murine and avian models.
Mathé G, Amiel JL, Schwarzenberg L, Catton A, Schneider M. Till JE, McCulloch EA. A direct measurement of the radiation
Adoptive immunotherapy of acute leukemia: experimental sensitivity of normal mouse bone marrow cells. Radiat Res.
and clinical results. Cancer Res. 1965;25:1525-1531. These two 1961;14:213-222. The first example of a quantitative assay for
seminal papers are the first descriptions of experimental and hematopoietic stem cells; beginning of stem cell research.
clinical graft versus tumor effects mediated by donor lymphocytes.
McGovern JJ Jr, Russel PS, Atkins L, Webster EW. Treatment
Stem cell transplant recipient
of terminal leukemic relapse by total-body irradiation and
intravenous infusion of stored autologous bone marrow obtained Armand P, Gibson C, Cutler C, et al. A disease risk index for
during remission. N Engl J Med. 1959;260:675-683. Landmark patients undergoing allogeneic stem cell transplantation. Blood.
article demonstrating that autologous marrow could rescue patients 2012;120:905-913. First modern validated model that looks at the
from the effects of supralethal radiation, thus opening the field of impact of diagnosis and stage on stem cell transplant outcomes,
autologous transplant for hematologic malignancies. applicable for all types of conditioning regimens.

Bibliography | 389
Pasquini MC, Wang Z. Current use and outcome of hematopoietic Flomenberg N, Baxter-Lowe LA, Confer D, et al. Impact of HLA
stem cell transplantation: CIBMTR Summary Slides, 2011. class I and class II high-resolution matching on outcomes
Available at http://www.cibmtr.org. Best online resource to of unrelated donor bone marrow transplantation: HLA-C
obtain up-to-date information regarding stem cell transplant mismatching is associated with a strong adverse effect on
activity in North America and the world. transplantation outcome. Blood. 2004;104:1923-1930. Impact of
Sorror ML, Maris MB, Storb R, et al. Hematopoietic cell HLA matching by molecular techniques, demonstration that allele
transplantation (HCT)-specific comorbidity index: a new mismatching affects outcomes.
tool for risk assessment before allogeneic HCT. Blood. Koreth J, Stevenson KE, Kim HT, et al. Bortezomib, tacrolimus,
2005;106:2912-2919. Most commonly used and validated and methotrexate for prophylaxis of graft-versus-host disease
comorbidity index used in HCT. Allows for tailoring intensity of after reduced-intensity conditioning allogeneic stem cell
conditioning to transplant recipient. Predictive value in autologous transplantation from HLA-mismatched unrelated donors.
stem cell transplantation is less certain. Together with other Blood. 2009;114:3956-3959. Novel GVHD prevention strategy for
assessments allows the reader to determine risk benefit ratio of mismatches stem cell transplantation.
stem cell transplantation in individual patients. Kurtzberg J. Update on umbilical cord blood transplantation.
Curr Opin Pediatr. 2009;21:22-29.
Luznik L, O’Donnell PV, Symons HJ, et al. HLA-haploidentical
Types of stem cell transplants bone marrow transplantation for hematologic malignancies
using nonmyeloablative conditioning and high-dose,
Aversa F, Terenzi A, Tabilio A, et al. Full haplotype-
posttransplantation cyclophosphamide. Biol Blood Marrow
mismatched hematopoietic stem-cell transplantation: a
Transplant. 2008;14:641-650. Seminal observation leading to an
phase II study in patients with acute leukemia at high risk of
effective and easily translatable strategy of GVHD prevention in
relapse. J Clin Oncol. 2005;23:3447-3454. Largest experience
the mismatch setting.
to date with haploidentical stem cell transplantation using
Petersdorf EW, Malkki M. Human leukocyte antigen matching
CD34 selection as GVHD prophylaxis. This experience led to
in unrelated-donor hematopoietic cell transplantation
the identification of the role of NK cells in preventing AML
[review]. Semin Hematol. 2005;42:76-84. Excellent review on
relapse.
HLA typing.
Ballen KK, Koreth J, Chen YB, Dey BR, Spitzer TR. Selection of
Rocha V, Labopin M, Sanz G, et al. Acute leukemia working party
optimal alternative graft source: mismatched unrelated donor,
of European blood and marrow transplant group; eurocord-
umbilical cord blood, or haploidentical transplant. Blood.
netcord registry. Transplants of umbilical-cord blood or bone
2012;119:1972-1980. Excellent review on the process of selecting
marrow from unrelated donors in adults with acute leukemia.
alternative donors.
N Engl J Med. 2004;351:2276-2285.
Barker J, Byam C, Scaradavou A. How I treat: the selection
van Rood JJ, Scaradavou A, Stevens CE. Indirect evidence that
and acquisition of unrelated cord blood grafts. Blood.
maternal microchimerism in cord blood mediates a graft-
2011;117:2332-2339. An excellent review on cord blood
versus-leukemia effect in cord blood transplantation. Proc Natl
transplantation.
Acad Sci USA. 2012;109:2509-2514. Evidence that maternal cells
Barker JN, Rocha V, Scaradavou A. Optimizing unrelated donor
may play a role in the graft-versus-leukemia effects mediated
cord blood transplantation. Biol Blood Marrow Transplant.
through cord blood transplantation.
2008;15(1)(suppl):154-161.
Brunstein CG, Eapen M, Ahn KW, et al. Reduced-intensity
conditioning transplantation in acute leukemia: the effect
Stem cell sources and procurement
of source of unrelated donor stem cells on outcomes. Blood.
2012;119:5591-5598. Bortin MM, Gale RP, Kay HE, Rimm AA. Bone marrow
Brunstein CG, Fuchs EJ, Carter SL. Alternative donor transplantation for acute myelogenous leukemia. Factors
transplantation after reduced intensity conditioning: results of associated with early mortality. JAMA. 1983;249:1166-1175.
parallel phase 2 trials using partially HLA-mismatched related A paper demonstrating the impact on cell dose on bone marrow
bone marrow or unrelated double umbilical cord blood grafts. transplant outcomes.
Blood. 2011;118:282-288. Largest retrospective analysis of cord Broxmeyer HE, Douglas GW, Hangoc G, et al. Human
blood transplantation outcomes compared with other sources of umbilical cord blood as a potential source of transplantable
stem cell transplantation. hematopoietic stem/progenitor cells. Proc Natl Acad Sci USA.
Eapen M, Rocha V, Sanz G, et al. Effect of graft source on unrelated 1989;86:3828-2832. Landmark paper that showed the potential
donor haemopoietic stem-cell transplantation in adults of cord blood as a source of stem cells and opened the field of cord
with acute leukaemia: a retrospective analysis. Lancet Oncol. blood transplantation.
2010;11:653-660. Buckner CD. Autologous bone marrow transplants to
Eapen M, Rubinstein P, Zhang MJ, et al. Comparison of outcomes hematopoietic stem cell support with peripheral blood stem
after transplantation of unrelated donor umbilical cord blood cells: a historical perspective. J Hematother. 1999;8:233-236.
and bone marrow in children with acute leukemia. Lancet. Jantunen E, Fruehauf S. Importance of blood graft characteristics
2007;369:1947-1954. in auto-HCT: implications for optimizing mobilization

regimens. Bone Marrow Transplant. 2011;5:627-635. An Ho VT, Cutler C, Carter S, et al. Blood and marrow transplant
excellent review on autologous stem cell graft characteristics. clinical trials network toxicity committee consensus summary:
Sheppard D, Bredeson C, Allan D, Tay J. Systematic review thrombotic microangiopathy after hematopoietic stem cell
of randomized controlled trials of hematopoietic stem cell transplantation. Biol Blood Marrow Transplant. 2005;11:
mobilization strategies for autologous transplantation for 571-575.
hematologic malignancies. Biol Blood Marrow Transplant. McDonald GB. Hepatobiliary complications of hematopoietic cell
2012;18:1191-1203. An excellent review. transplantation, 40 years on. Hepatology. 2010;51:1450-1460. A
Siena S, Bregni M, Brando B, et al. Circulation of CD34+ must-read for anybody wanting to learn from one of the foremost
hematopoietic stem cells in the peripheral blood of high- experts in hepatobiliary complications of HCT.
dose cyclophosphamide-treated patients: enhancement by Pihusch M. Bleeding complications after hematopoietic stem cell
intravenous recombinant human granulocyte-macrophage transplantation [review]. Semin Hematol. 2004;41(suppl 1):
colony-stimulating factor. Blood. 1989;74:1905-1914. 93-100.
This classic paper led to the current standard of care for Tichelli A, Rovó A, Gratwohl A. Late pulmonary, cardiovascular,
chemomobilization. and renal complications after hematopoietic stem cell
Sierra J, Storer B, Hansen JA, et al. Transplantation of marrow transplantation and recommended screening practices.
cells from unrelated donors for treatment of high-risk acute Hematology Am Soc Hematol Educ Program. 2008;2008:125-133.
leukemia: the effect of leukemic burden, donor HLA-matching, Tomblyn M, Chiller T, Einsele H, et al. Guidelines for preventing
and marrow cell dose. Blood. 1997;89:4226-4235. A paper infectious complications among hematopoietic cell transplant
demonstrating the impact on cell dose on bone marrow transplant recipients: a global perspective. Biol Blood Marrow Transplant.
outcomes. 2009;15:1143-1238. Important consensus statement.
Stem Cell Transplant Trialist Group. Allogeneic peripheral blood Tuncer HD, Rana N, Milani C, Darko A, Al-Hasmi S.
stem-cell compared with bone marrow transplantation in Gastrointestinal and hepatic complications of hematopoietic
the management of hematologic malignancies: an individual stem cell transplantation. World J Gastroenterol. 2012;18:1851-
patient data meta-analysis of nine randomized trials. J Clin 1860. Excellent up-to-date review.
Oncol. 2005;23:5074-5087. Meta-analysis of allogeneic peripheral
blood versus bone marrow as a stem cell source for transplantation.
Storb R, Prentice RL, Thomas ED. Marrow transplantation for Graft-versus-host disease
treatment of aplastic anemia. An analysis of factors associated Carpenter P. How I conduct a comprehensive chronic graft versus
with graft rejection. N Engl J Med. 1977;296:61-66. A paper host disease assessment Blood. 2011;118:2679-2687. A must-read
demonstrating the impact on cell dose on bone marrow transplant for anybody taking care of long-term allogeneic HCT survivors.
outcomes. Excellent resource for health care providers.
Thomas ED, Storb R. Technique for human marrow grafting. Ferrara JL, Levine JE, Reddy P, Holler E. Graft-vs-host-disease.
Blood. 1970;36:507-515. Classic paper describing the still used Lancet. 2009;373:1550-1561.
technique of marrow harvesting. Filipovich AH. Diagnosis and manifestations of chronic graft-
versus-host disease. Best Pract Res Clin Haematol. 2008;21:
251-257.
Conditioning regimens
Joseph RW, Couriel DR, Komanduri KV. Chronic graft-versus-
Bacigalupo A, Ballen K, Rizzo D, et al. Defining the intensity of host disease after allogeneic stem cell transplantation: challenges
conditioning regimens: working definitions. Biol Blood Marrow in prevention, science, and supportive care. J Support Oncol.
Transplant. 2009;15:1628-1633. Current consensus definition of 2008;6:361-372. Comprehensive review of current knowledge.
what constitutes a reduced intensity regimen. Martin PJ, Rizzo JD, Wingard JR, et al. First- and second-
Luger S, Ringden O, Zhang MJ, et al. Similar outcomes using line systemic treatment of acute graft-versus-host disease:
myeloablative versus reduced intensity regimens for allogeneic recommendations of the American Society of Blood and
transplants for AML or MDS. Bone Marrow Transplant. Marrow Transplantation. Biol Blood Marrow Transplant.
2012;47:203-211. Large retrospective comparison of transplant 2012;18:1150-1163. Excellent reviews and state-of-the-art
outcomes demonstrating that at least in AML and MDS dose may management of acute GVHD.
be important.
Late effects
Stem cell transplant complications
Burns LJ. Late effects after autologous hematopoietic cell
Bacigalupo A, Chien J, Barisione G, Pavletic S. Late pulmonary transplantation. Biol Blood Marrow Transplant. 2008;15
complications after allogeneic hematopoietic stem cell (suppl 1):21-24.
transplantation: diagnosis, monitoring, prevention, and Deeg HJ, Socie G. Malignancies after hematopoietic stem cell
treatment. Semin Hematol. 2012;49:15-24. Excellent and timely transplantation: many questions, some answers. Blood.
reviews on late HCT pulmonary complications. 1998;91:1833-1844.

Bibliography | 391
Rizzo JD, Wingard JR, Tichelli A, et al. Recommended screening Duval M, Klein JP, He W, et al. Hematopoietic stem-cell
and preventive practices for long-term survivors after transplantation for acute leukemia in relapse or primary
hematopoietic cell transplantation: joint recommendations of induction failure. J Clin Oncol. 2010;28:3730-3738.
the European Group for Blood and Marrow Transplantation, the Fernandez HF, Sun Z, Yao X, et al. Anthracycline dose
Center for International Blood and Marrow Transplant Research, intensification in acute myeloid leukemia. N Engl J Med.
and the American Society of Blood and Marrow Transplantation. 2009;361:1249-1259.
Biol Blood Marrow Transplant. 2006;12:138-151. Gupta V, Tallman MS, He W, et al. Comparable survival
Tichelli A, Rovó A, Gratwohl A. Late pulmonary, cardiovascular, after HLA well-matched unrelated or matched sibling
and renal complications after hematopoietic stem cell donor transplantation for acute myeloid leukemia in first
transplantation and recommended screening practices. remission with unfavorable cytogenetic at diagnosis. Blood.
Hematology Am Soc Hematol Educ Program. 2008;2008:125-133. 2010;116:1839-1848.
Holter Chakrabarty JL, Rubinger M, Le-Rademacher J, et
al. Autologous is superior to allogeneic hematopoietic
Relapse cell transplantation for acute promyelocytic leukemia in
second complete remission. Biol Blood Marrow Transplant.
Horowitz MM, Gale RP, Sondel PM, et al. Graft-versus-
2014;20:1021-1025.
leukemia reactions after bone marrow transplantation. Blood.
Koreth J, Schlenk R, Kopecky KJ, et al. Allogeneic stem cell
1990;75:555-562.
transplantation for acute myeloid leukemia in first complete
Schmid C, Labopin M, Nagler A, et al. Donor lymphocyte infusion
remission. JAMA. 2009;301:2349-2361. Excellent meta-analysis
in the treatment of first hematological relapse after allogeneic
of all donor versus no donor trials.
stem-cell transplantation in adults with acute myeloid leukemia:
Löwenberg B, Ossenkoppele GJ, van Putten W, et al. High-dose
a retrospective risk factors analysis and comparison with other
daunorubicin in older patients with acute myeloid leukemia.
strategies by the EBMT Acute Leukemia Working Party. J Clin
N Engl J Med. 2009;361:1235-1248.
Oncol. 2007;25:4938-4945.
Röllig C, Bornhäuser M, Kramer M, et al. Allogeneic stem-cell
transplantation in patients with NPM1-mutated acute myeloid
leukemia: results from a prospective donor versus no-donor
Transplantation for specific diseases
analysis of patients after upfront HLA typing within the SAL-
Armand P, Gibson CJ, Cutler C, et al. A disease risk index for AML 2003 trial. J Clin Oncol. 2015;33:403-410.
patients undergoing allogeneic stem cell transplantation. Blood. Schlenk RF, Döhner K, Krauter J, et al. Mutations and treatment
2012;120:905-913. outcome in cytogenetically normal acute myeloid leukemia.
N Engl J Med. 2008;358:1909-1918. Essential reading to
understand the new molecular classification of AML.
Acute myelogenous leukemia
Bashey A, Zhang X, Sizemore CA, et al. vT-cell-replete HLA-
haploidentical hematopoietic transplantation for hematologic
malignancies using post-transplantation cyclophosphamide Bassan R, Spinelli O, Oldani E, et al. Improved risk classification
results in outcomes equivalent to those of contemporaneous for risk-specific therapy based on the molecular study of
HLA-matched related and unrelated donor transplantation. minimal residual disease (MRD) in adult acute lymphoblastic
J Clin Oncol. 2013;31:1310-1316. leukemia (ALL). Blood. 2009;113:4153-4162.
Brunet S, Labopin M, Esteve J, et al. Impact of FLT3 internal Cornelissen JJ, van der Holt B, Verhoef GE, et al. Myeloablative
tandem duplication on the outcome of related and unrelated allogeneic versus autologous stem cell transplantation in adult
hematopoietic transplantation for adult acute myeloid leukemia patients with acute lymphoblastic leukemia in first remission: a
in first remission: a retrospective analysis. J Clin Oncol. prospective sibling donor versus no-donor comparison. Blood.
2012;30:735-741. 2009;113:1375-1382.
Brunstein CG, Gutman JA, Weisdor DJ, et al. Allogeneic Dhédin N, Huynh A, Maury S, et al. Role of allogeneic stem
hematopoietic cell transplantation for hematologic malignancy: cell transplantation in adult patients with Ph-negative acute
relative risks and benefits of double umbilical cord blood. Blood. lymphoblastic leukemia. Blood. 2015;125:2486-2496.
2010;116:4693-4699. Role of alternative donor transplantation Eapen M, Raetz E, Zhang MJ, et al. Outcomes after HLA-
in AML. matched sibling transplantation or chemotherapy in children
de Botton S, Fawaz A, Chevret S, et al. Autologous and allogeneic with B-precursor acute lymphoblastic leukemia in a second
stem-cell transplantation as salvage treatment of acute remission: a collaborative study of the Children’s Oncology
promyelocytic leukemia initially treated with all-trans- Group and the Center for International Blood and Marrow
retinoic acid: a retrospective analysis of the European acute Transplant Research. Blood. 2006;107:4961-4967.
promyelocytic leukemia group. J Clin Oncol. 2005;23: Gaynon PS, Harris RE, Altman AJ, et al. Bone marrow
120-126. transplantation versus prolonged intensive chemotherapy for

children with acute lymphoblastic leukemia and an initial bone Chronic myelogenous leukemia
marrow relapse within 12 months of the completion of primary
therapy: Children’s Oncology Group Study CCG-1941. J Clin Cortes J, Lipton JH, Rea D, et al. Phase 2 study of subcutaneous
Oncol. 2006;24:3150-3156. Large retrospective and prospective omacetaxine mepesuccinate after TKI failure in patients
analysis looking at the role of HCT in pediatric ALL. with chronic-phase CML with T315I mutation. Blood.
Gökbuget N, Basara N, Baurmann H, et al. High single-drug 2012;120:2573-2580.
activity of nelarabine in relapsed T-lymphoblastic leukemia/ Cortes JE, Kim DW, Pinilla-Ibarz J, et al. A phase 2 trial of
lymphoma offers curative option with subsequent stem cell ponatinib in Philadelphia chromosome-positive leukemias.
transplantation. Blood. 2011;118:3504-3511. N Engl J Med. 2013;369:1783-1796.
Goldstone AH, Richards SM, Lazarus HM, et al. In adults with Gratwohl A, Brand R, Apperley J, et al. Allogeneic hematopoietic
standard-risk acute lymphoblastic leukemia, the greatest benefit stem cell transplantation for chronic myeloid leukemia in
is achieved from a matched sibling allogeneic transplantation Europe 2006: transplant activity, long-term data and current
in first complete remission, and an autologous transplantation results. An analysis by the Chronic Leukemia Working Party of
is less effective than conventional consolidation/maintenance the European Group for Blood and Marrow Transplantation
chemotherapy in all patients: final results of the International (EBMT). Haematologica. 2006;91:513-521.
ALL Trial (MRC UKALL XII/ECOG E2993). Blood. Kantarjian H, Shah NP, Hochhaus A, et al. Dasatinib versus
2008;111:1827-1833. imatinib in newly diagnosed chronic-phase chronic myeloid
Huguet F, Leguay T, Raffoux E, et al. Pediatric-inspired therapy leukemia. N Engl J Med. 2010;362:2260-2270.
in adults with Philadelphia chromosome-negative acute Lee SJ, Kukreja M, Wang T, et al. Impact of prior imatinib
lymphoblastic leukemia: the GRAALL-2003 study. J Clin Oncol. mesylate on the outcome of hematopoietic cell transplantation
2009;27:911-918. for chronic myeloid leukemia. Blood. 2008;112:3500-3507.
Kantarjian H, Thomas D, Jorgensen J, et al. Results of inotuzumab Radich JP, Gooley T, Bensinger W, et al. HLA-matched related
ozogamicin, a CD22 monoclonal antibody, in refractory and hematopoietic cell transplantation for chronic-phase CML
relapsed acute lymphocytic leukemia. Cancer. 2013;119: using a targeted busulfan and cyclophosphamide preparative
2728-2736. regimen. Blood. 2003;102:31-35.
Lee DW, Kochenderfer JN, Stetler-Stevenson M, et al. T cells Saglio G, Kim DW, Issaragrisil S, et al. Nilotinib versus imatinib
expressing CD19 chimeric antigen receptors for acute for newly diagnosed chronic myeloid leukemia. N Engl J Med.
lymphoblastic leukaemia in children and young adults: a phase 2010;362:2251-2259.
1 dose-escalation trial. Lancet. 2015;385:517-528. Saussele S, Lauseker M, Gratwohl A, et al. Allogeneic
Maude SL, Frey N, Shaw PA, et al. Chimeric antigen receptor hematopoietic stem cell transplantation (allo HCT) for chronic
T cells for sustained remissions in leukemia. N Engl J Med. myeloid leukemia in the imatinib era: evaluation of its impact
2014;371:1507-1517. within a subgroup of the randomized German CML Study IV.
O’Brien S, Schiller G, Lister J, et al. High-dose vincristine sulfate Blood. 2010;115:1880-1885.
liposome injection for advanced, relapsed, and refractory
adult Philadelphia chromosome-negative acute lymphoblastic
Chronic lymphocytic leukemia
leukemia. J Clin Oncol. 2013;31:676-683.
Pidala J, Djulbegovic B, Anasetti C, Kharfan-Dabaja M, Kumar A. Dreger P, Corradini P, Kimby E, et al. Indications for allogeneic
Allogeneic hematopoietic cell transplantation for adult acute stem cell transplantation in chronic lymphocytic leukemia: the
lymphoblastic leukemia (ALL) in first complete remission. EBMT transplant consensus. Leukemia. 2007;21:12-17.
Cochrane Database Syst Rev. 2011;5(10):CD008818. Largest Dreger P, Döhner H, Ritgen M, et al. Allogeneic stem cell
prospective trials and meta-analysis of donor versus no donor transplantation provides durable disease control in poor-risk
outcomes in adult ALL. chronic lymphocytic leukemia: long-term clinical and MRD
Seftel, MD, Neuberg, D, Zhang, M, et al. Superiority of results of the German CLL Study Group CLL3X trial. Blood.
pediatric chemotherapy over allogeneic hematopoietic cell 2010;116:2438-2447.
transplantation for Philadelphia chromosome negative adult Michallet M, Dreger P, Sutton L, et al. Autologous hematopoietic
ALL in first complete remission: a combined analysis of Dana- stem cell transplantation in chronic lymphocytic leukemia:
Farber ALL Consortium and CIBMTR cohorts. ASH Annual results of European intergroup randomized trial comparing
Meeting 2014, Abstract 319. autografting versus observation. Blood. 2011;117:
Stock, W, Luger, S, Advani, A, et al. Favorable outcomes for older 1516-1521.
adolescents and young adults (AYA) with acute lymphoblastic Schetelig J, van Biezen A, Brand R, et al. Allogeneic hematopoietic
leukemia (ALL): early results of U.S. Intergroup Trial C10403. stem-cell transplantation for chronic lymphocytic leukemia
ASH Annual Meeting 2014, Abstract 796. with 17p deletion: a retrospective European Group for
Topp MS, Gökbuget N, Stein AS, et al. Safety and activity of Blood and Marrow Transplantation analysis. J Clin Oncol.
blinatumomab for adult patients with relapsed or refractory 2008;26:5094-5100.
B-precursor acute lymphoblastic leukaemia: a multicentre, Sorror ML, Storer BE, Sandmaier BM, et al. Five-year follow-
single-arm, phase 2 study. Lancet Oncol. 2015;16:57-66. up of patients with advanced chronic lymphocytic leukemia

Bibliography | 393
treated with allogeneic hematopoietic cell transplantation after Gisselbrecht C, Schmitz N, Mounier N, et al. Rituximab
nonmyeloablative conditioning. J Clin Oncol. 2008;26:4912-4920. maintenance therapy after autologous stem-cell transplantation
in patients with relapsed CD20+ diffuse large B-cell lymphoma:
final analysis of the collaborative trial in relapsed aggressive
Myelodysplasia lymphoma. J Clin Oncol. 2012;30:4462-4469.
Cutler CS, Lee SJ, Greenberg P, et al. A decision analysis of Haioun C, Lepage E, Gisselbrecht C, et al. Survival benefit of high-
allogeneic bone marrow transplantation for the myelodysplastic dose therapy in poor-risk aggressive non-Hodgkin’s lymphoma:
syndromes: delayed transplantation for low-risk myelodysplasia final analysis of the prospective LNH87-2 protocol—a groupe
is associated with improved outcome. Blood. 2004;104:579-585. d’Etude des lymphomes de l’Adulte study. J Clin Oncol.
Deeg JH, Scott BL, Fang M, et al. Five-group cytogenetic risk 2000;18:3025-3030.
classification, monosomal karyotype and outcome after Le Gouill S, Milpied NJ, Lamy T, et al. First-line rituximab (R)
hematopoeitic cell tranplantation for MDS or acute leukemia high-dose therapy (R-HDT) versus R-CHOP14 for young
evolving from MDS. Blood. 2012;120:1398-1408. adults with diffuse large B-cell lymphoma: preliminary results
Deeg HJ, Storer B, Slattery JT, et al. Conditioning with targeted of the GOELAMS 075 prospective multicenter randomized trial
busulfan and cyclophosphamide for hemopoetic stem cell [abstract]. J Clin Oncol. 2011;29:Abstract 8003.
transplantation from related and unrelated donors in patients Le Gouill, S,, Thieblemont, C, Oberic, L, et al. Rituximab
with myelodysplastic syndrome. Blood. 2002;100:1201-1207. maintenance versus wait and watch after four courses of
Della Porta, MG, Alessandrino, EP, Jackson, CH, et al. Decision R-DHAP followed by autologous stem cell transplantation in
Analysis of Allogeneic Stem Cell Transplantation in Patients previously untreated young patients with mantle cell
with Myelodysplastic Syndrome Stratified According to the lymphoma: first interim analysis of the phase III prospective
Revised International Prognostic Scoring System (IPSS-R). Lyma trial, a Lysa Study. ASH Annual Meeting 2014,
ASH Annual Meeting 2014, Abstract 531. Abstract 146.
De Witte T, Brand R, van Biezen A, et al. Allogeneic stem cell Robinson SP, Canals C, Luang JJ, et al. The outcome of reduced
transplantation of patients with refractory anaemia with intensity allogeneic stem cell transplantation and autologous
matched related and unrelated donors: delay of the transplant is stem cell transplantation when performed as a first transplant
associated with inferior survival. Br J Haematol. 2009;146: strategy in relapsed follicular lymphoma: an analysis from
627-636. the Lymphoma Working Party of the EBMT. Bone Marrow
Martino R, Iacobelli S, Brand R, et al. Retrospective comparison Transplant. 2013;48:1409-1414
of reduced-intensity conditioning and conventional high- Romaguera JE, Fayad L, Rodriguez MA, et al. High rate of durable
dose conditioning for allogeneic hematopoetic stem cell remissions after treatment of newly diagnosed aggressive
transplantation using HLA-identical sibling donors in mantle-cell lymphoma with rituximab plus hyper-CVAD
myelodysplastic syndromes. Blood. 2006;108:836-846. alternating with rituximab plus high-dose methotrexate and
cytarabine. J Clin Oncol. 2005;23:7013-7023.
Sandlund JT, Bowman L, Heslop HE, et al. Intensive
chemotherapy with hematopoietic stem-cell support for
Non-Hodgkin lymphoma
children with recurrent or refractory NHL. Cytotherapy.
Al-Tourah AJ, Gill KK, Chhanabhai M, et al. Population-based 2002;4:253-258.
analysis of incidence and outcome of transformed non- Schmitz N, Nickelsen M, Ziepert M, et al. Conventional
Hodgkin’s lymphoma. J Clin Oncol. 2008:26;5165-5169. chemotherapy (CHOEP-14) with rituximab or high-dose
Barr PM, Lazarus HM. Follicular non-Hodgkin lymphoma: long- chemotherapy (MegaCHOEP) with rituximab for young, high-
term results of stem-cell transplantation. Curr Opin Oncol. risk patients with aggressive B-cell lymphoma: an open-label,
2008;20:502-508. randomised, phase 3 trial (DSHNHL 2002-1). Lancet Oncol.
Damon LE, Johnson JL, Niedzwiecki D, et al. 2012;13:1250-1259.
Immunochemotherapy and autologous stem-cell Schouten HC, Qian W, Kvaloy S, et al. High-dose therapy improves
transplantation for untreated patients with mantle-cell progression-free survival and survival in relapsed follicular non-
lymphoma: CALGB 59909. J Clin Oncol. 2009;27:6101-6108. Hodgkin’s lymphoma: results from the randomized European
Geisler CH, Kolstad A, Laurell A, et al. Long-term progression- CUP trial. J Clin Oncol. 2003;21:3918-3927.
free survival of mantle cell lymphoma after intensive front-line Seshadri T, Kuruvilla J, Crump M, Keating A. Salvage therapy for
immunochemotherapy with in vivo-purged stem cell rescue: relapsed/refractory diffuse large B cell lymphoma. Biol Blood
a nonrandomized phase 2 multicenter study by the Nordic Marrow Transplant. 2008;14:259-267.
Lymphoma Group. Blood. 2008;112:2687-2693. Stiff PJ, Unger JM, Cook J, et al. Randomized phase III U.S./
Ghielmini M, Zucca E. How I treat mantle cell lymphoma. Blood. Canadian intergroup trial (SWOG S9704) comparing CHOP
2009;114:1469-1476. {+/-} R for eight cycles to CHOP {+/-} R for six cycles followed
Gisselbrecht C, Glass B, Mounier N, et al. Salvage regimens with by autotransplant for patients with high-intermediate (H-Int)
autologous transplantation for relapsed large B-cell lymphoma or high IPI grade diffuse aggressive non-Hodgkin lymphoma
in the rituximab era. J Clin Oncol. 2010;28:4184-4190. (NHL) [abstract 8001]. J Clin Oncol. 2011;29:Abstract 8001.

Tam CS, Khouri IF. Autologous and allogeneic stem cell in patients with refractory and/or recurrent Hodgkin
transplantation: rising therapeutic promise for mantle cell lymphoma: a multi-center study of the Intergruppo Italiano
lymphoma. Leuk Lymphoma. 2009;26:1-10. Linfomi showing prolonged disease free survival in patients
Thomson KJ, Morris EC, Bloor A, et al. Favorable long-term treated at first recurrence. Cancer. 2003;97:2748-2759.
survival after reduced-intensity allogeneic transplantation for
multiple-relapse aggressive non-Hodgkin’s lymphoma. J Clin
Oncol. 2009;27:426-432. Plasma cell dyscrasias
Van Kampen RJW, Canals C, Schouten HC, et al. Allogeneic Armeson KE, Hill EG, Costa LJ. Tandem autologous vs
stem-cell transplantation as salvage therapy for patients with autologous plus reduced intensity allogeneic transplantation
diffuse large B-cell non-Hodgkin’s lymphoma relapsing after in the upfront management of multiple myeloma: meta-
an autologous stem-cell transplantation: an analysis of the analysis of trials with biological assignment. Bone Marrow
European Group for Blood and Marrow Transplantation Transplant. 2013;48:562-567.
Registry. J Clin Oncol. 2011;29:1342-1348. Attal M, Haroussseau JL, Facon T, et al. Single versus double
Vitolo U, Chiappella A, Brusamolino E, et al. Rituximab autologous stem-cell transplantation for multiple myeloma.
dose-dense chemotherapy followed by intensified high-dose N Engl J Med. 2003;349:2494-2502.
chemotherapy and autologous stem cell transplantation Attal M, Lauwers-Cances V, Marit G, et al. Lenalidomide
(hdc+asct) significantly reduces the risk of progression Maintenance after Stem-Cell Transplantation for Multiple
compared to standard rituximab dose-dense chemotherapy Myeloma. N Engl J Med. 2012;366:1782-1791.
as first line treatment in young patients with high-risk Bruno B, Rotta M, Patiarca F, et al. Nonmyeloablative allografting
(aa-ipi 2-3) diffuse large b-cell lymphoma (dlbcl): for newly diagnosed multiple myeloma: the experience of the
final results of phase iii randomized trial DLCL04 of the Gruppo Italiano Trapianti di Midollo. Blood. 2009;113:
Fondazione Italiana Linfomi (FIL). ASH Annual Meeting 3375-3382.
Abstracts 2014, Abstract 688. D’Souza A, Lacy M, Gertz M, et al. Long-term outcomes after
autologous stem cell transplantation for patients with POEMS
syndrome (osteosclerotic myeloma): a single-center experience.
Hodgkin disease
Blood. 2012;120:56-62.
Burroughs LM, O’Donnell PV, Sandmaier BM, et al. Comparison Harousseau JL, Moreau P. Autologous hematopoietic stem
of outcomes of HLA-matched related, unrelated, or HLA- cell transplantation for multiple myeloma. N Engl J Med.
haploidentical related hematopoietic cell transplantation 2009;360:2645-2654.
following a myeloablative conditioning for relapse or Krishnan A, Pasquini MC, Logan B, et al and Blood Marrow
refractory Hodgkin lymphoma. Biol Blood Marrow Transplant. Transplant Clinical Trials Network (BMT CTN). Autologous
2008;14:1279-1287. haemopoietic stem-cell transplantation followed by allogeneic
Moskowitz CH, Nademanee A, Masszi T, et al. The AETHERA or autologous haemopoietic stem-cell transplantation in
trial: results of a randomized, double-blind, placebo-controlled patients with multiple myeloma (BMT CTN 0102): a phase 3
phase III study of brentuximab vedotin in the treatment of biological assignment trial. Lancet Oncol. 2011;12:1195-1203.
patients at risk of Hodgkin lymphoma progression following McCarthy PL, Owzar K, Hofmeister CC, et al. Lenalidomide after
autologous stem cell transplant. ASH Annual Meeting 2014, stem-cell transplantation of multiple myeloma. N Engl J Med.
Abstract 673. 2012;366:1770-1781.
Sarina B, Castagna L, Farina L, et al. Allogeneic transplantation Palumbo A, Cavallo F, Gay F, et al. Autologous transplantation
improves the overall and progression-free survival or Hodgkin and maintenance therapy in multiple myeloma. N Engl J Med.
lymphoma patients relapsing after autologous transplantation: a 2014;371:895-905. 
retrospective study based on the time of HLA typing and donor Rotta M, Storer BE, Sahebi F, et al. Long-term outcome of patients
availability. Blood. 2010;115:3671-3677. with multiple myeloma after autologous hematopoietic cell
Schmitz N, Pfistner B, Sextro M, et al. Aggressive conventional transplantation and nonmyeloablative allografting. Blood.
chemotherapy compared with high-dose chemotherapy with 2009;113:3383-3391.
autologous haemopoietic stem-cell transplantation for relapsed
chemo-sensitive Hodgkin’s disease: a randomised trial. Lancet.
Autoimmune diseases and aplastic anemia
2002; 359:2065-2071.
Sureda A, Robinson S, Canals C, et al. Reduced-intensity Ades L, Mary JV, Robin M, et al. Long-term outcome after bone
conditioning compared with conventional allogeneic stem-cell marrow transplantation for severe aplastic anemia. Blood.
transplantation in relapsed or refractory Hodgkin’s lymphoma: 2004;103:2490-2497.
an analysis from the Lymphoma Working Party of the Ayas M, Nassar A, Hamidieh AA, et al. Reduced intensity
European Group for Blood and Marrow Transplantation. J Clin conditioning is effective for hematopoietic SCT in dyskeratosis
Oncol. 2008;26:455-462. congenita-related BM failure. Bone Marrow Transplant.
Tarella C, Cuttica A, Vitolo U, et al. High-dose sequential 2013;48:1168-1172.
chemotherapy and peripheral blood progenitor cell autografting Brodsky RA, Jones RJ. Aplastic anaemia. Lancet. 2005;365:1647-1656.

Bibliography | 395
Griffith LM, Pavletic SZ, Tyndall A, et al. Feasibility of allogeneic Veys P, Rao K, Amrolia P. Stem cell transplantation for congenital
hematopoietic stem cell transplantation for autoimmune immunodeficiencies using reduced-intensity conditioning. Bone
disease: position statement from a National Institute of Marrow Transplant. 2005;35(suppl 1):S45-S47.
Allergy and Infectious Diseases and National Cancer Institute-
Sponsored International Workshop, Bethesda, MD, March 12
and 13, 2005. Biol Blood Marrow Transplant. 2005;11:862-870. Metabolic disorders
Passweg JR, Marsh JCW. Aplastic anemia: first-line treatment Peters C, Steward CG. Hematopoietic cell transplantation
by immunosuppression and sibling marrow transplant. for inherited metabolic diseases: an overview of outcomes
Hematology. 2010;2010:36-42. and practice guidelines. Bone Marrow Transplant.
Townsley DM, Dumitriu B, Young NS. Bone marrow failure and 2003;31:229-239.
the telomeropathies. Blood. 2014;124:2775-2783.
Adult solid tumors

Hemoglobinopathies/thalassemia
Einhorn LH, Williams SD, Chamness A, et al. High-dose
Bolanos-Meade J, Brodsky RA. Blood and marrow transplantation chemotherapy and stem-cell rescue for metastatic germ-cell
for sickle cell disease: overcoming barriers to success. Curr Opin tumors. N Engl J Med. 2007;357:340-348.
Oncol. 2009;21:158-161.
Gaziev J, Sodani P, Polchi P, et al. Bone marrow transplantation
in adults with thalassemia: treatment and long-term follow-up. Pediatric solid tumors
Ann N Y Acad Sci. 2005;1054:196-205.
Fish JD, Grupp SA. Stem cell transplantation for neuroblastoma.
Hsieh MM, Kang EM, Fitzhugh CD, et al. Allogeneic
Bone Marrow Transplant. 2008;41:159-165.
hematopoietic stem-cell transplantation for sickle cell disease.
Gardner SL, Carreras J, Boudreau C, et al. Myeloablative therapy
N Engl J Med. 2009;361:2309-2317.
with autologous stem cell rescue for patients with Ewing
sarcoma. Bone Marrow Transplant. 2008;41:867-872.
Hale GA. Autologous hematopoietic stem cell transplantation
Immune deficiencies
for pediatric solid tumors. Expert Rev Anticancer Ther. 2005;5:
Bielorai B, Trakhtenbrot L, Amariglio N, et al. Multilineage 835-846.
hematopoietic engraftment after allogeneic peripheral blood Ladenstein R, Pötschger U, Hartman O, et al. 28 years of high-
stem cell transplantation without conditioning in SCID dose therapy and HCT for neuroblastoma in Europe: lessons
patients. Bone Marrow Transplant. 2004;34:317-320. from more than 4,000 procedures. Bone Marrow Transplant.
Buckley RH, Schiff SE, Schiff RI, et al. Hematopoietic stem- 2008;41:S118-S127.
cell transplantation for the treatment of severe combined Matthay KK, Villablanca JG, Seeger RC, et al. Treatment of high-
immunodeficiency. N Engl J Med. 1999;340:508-516. risk neuroblastoma with intensive chemotherapy, radiotherapy,
Grunebaum E, Mazzolari E, Porta F, et al. Bone marrow autologous bone marrow transplantation, and 13-cos-retinoic
transplantation for severe combined immune deficiency. JAMA. acid. Children’s Cancer Group. N Engl J Med. 1999;341:
2006;295:508-518. 1165-1173.

CHAPTER
15
Myeloid disorders and inherited
marrow failure syndromes
Alison A. Bertuch and Inderjeet Dokal
Granulocytes: neutrophils, eosinophils, Acquired neutropenia, 416 Disorders of histiocytes and dendritic
and basophils, 397 Inherited disorders of neutrophil cells, 420
Monocytes and tissue histiocytes, 401 function, 418 Lysosomal storage diseases, 424
Inherited marrow failure syndromes, 402 Bibliography, 426
The term myeloid derives from the Greek myelos, meaning hematopoietic growth factor receptor–kinase signaling
“marrow,” and in its broadest sense is used to describe hema- (congenital amegakaryocytic thrombocytopenia [CAMT]).
tologic conditions or diseases originating in the bone mar- In some inherited marrow failure syndromes, the mecha-
row. Myeloid is also used more narrowly to describe disorders nism of hematopoietic failure is currently unclear.
primarily involving granulocytes (neutrophils, eosinophils,
or basophils) and monocytes, as opposed to other cell lin-
eages such as lymphoid cells. Tissue macrophages (histio- Granulocytes: neutrophils, eosinophils,
cytes), Langerhans cells, and interdigitating dendritic cells and basophils
also arise from monocytic progenitors or precursors. In con-
trast, other important immunoregulatory dendritic cell The term granulocytes refers to circulating neutrophils,
types derive from marrow progenitors that are distinct from eosinophils, and basophils, although, because of their
myeloid and monocytic progenitors. predominance in the blood, the terms neutrophil and
Bone marrow (BM) failure refers to the inability of hema- granulocyte sometimes are used synonymously. Normal
topoiesis to meet physiologic demands for production of values for the differential count for leukocytes in the blood
healthy blood cells. Pancytopenia may result from marrow are shown in Table 15-1. Neutrophils are a critical component
failure, or cytopenias involving a single myeloid lineage may of the innate immune response, and persistent neutropenia
dominate; usually lymphopoiesis is relatively preserved. BM is associated with a marked susceptibility to infection.
failure syndromes can be classified into idiopathic, inherited, Evidence also is increasing that neutrophils are a major
and secondary. The range of molecular mechanisms respon- contributor to tissue damage in inflammatory diseases.
sible for inherited marrow failure states (discussed in the Neutrophil homeostasis in the blood is regulated at three
section “Inherited marrow failure syndromes” in this chap- levels: neutrophil production in the BM (granulopoiesis),
ter) is broad, including abnormal DNA-damage response neutrophil release from the BM to blood, and neutrophil
(Fanconi anemia [FA]), defective ribosome biogenesis clearance from the blood (Figure 15-1).
(Diamond-Blackfan anemia [DBA]), defective telomere

maintenance (dyskeratosis congenita [DC]), and altered
Granulopoiesis
Conflict-of-Interest disclosure: Dr. Bertuch: Research funding:
Cancer Prevention Research Institute of Texas, National Institutes Under normal conditions, neutrophils are produced exclu-
of Health, American Society of Hematology. Dr. Dokal: Research sively in the BM, where it is estimated that 1012 are gener-
funding: MRC, Children with Cancer, Leukaemia & Lymphoma
ated on a daily basis. Granulocytic differentiation of
Research.
Off-label drug use: Dr. Bertuch: not applicable. Dr. Dokal: not hematopoietic stem cells (HSCs) is regulated by the coordi-
applicable. nated expression of a number of key myeloid transcription
| 397

398 | Myeloid disorders and inherited marrow failure syndromes
Table 15-1 Normal leukocyte values in peripheral blood in adults
Conventional units SI units

Leukocyte counts 4,000-10,000/mm3 4.0-10 × 109 /L
Afro-Caribbean: 2,800-10,000/mm3 2.8-10 × 109 /L
Leukocyte differential (% of total) (Absolute [× 103/mm3])
Neutrophils 40-70 1.4-7.5
Bands 0-15 0-1.5
Lymphocytes 20-50 0.8-5.0
Monocytes 3-8 0.1-1.0
Eosinophils 0-4 0-0.4
Reference ranges reported by specific laboratories may differ slightly from these values.
factors, including PU.1, CCAAT enhancer–binding protein transmigration. Indeed, deficiency of selectin ligands or
α (C/EBPα), C/EBPε, and GFI-1. A number of hematopoi- β2-integrins causes leukocyte adhesion deficiency, a rare
etic growth factors provide extrinsic signals that regulate syndrome manifested by normal neutrophil production but
granulopoiesis. The most important of these is granulocyte impaired emigration to sites of inflammation.
colony-stimulating factor (G-CSF), which stimulates the Once emigrated to the inflammatory site, neutrophils
proliferation of granulocytic precursors, reduces the average function primarily as tissue phagocytes. This vital function
transit time through the granulocytic compartment, and depends on several critical features of these cells. Surface
stimulates neutrophil release from the BM. receptors for immunoglobulins and complement enhance
ingestion and killing of microorganisms. Within the cell,
the reduced nicotinamide adenine dinucleotide phosphate
Neutrophil release
(NADPH) oxidase system, enzymes found in the cell’s
Neutrophils are released from the marrow into the blood in primary and secondary granules, and cytoplasmic glycogen
a regulated fashion to maintain homeostatic levels of circu- are involved in the intracellular oxidative burst that
lating neutrophils. The BM provides a large reservoir of accompanies phagocytosis and in the killing and digesting
mature neutrophils that can be mobilized readily in response of microorganisms. Maintenance of both the supply of cells
to infection. A broad range of substances has been shown to and the integrity of all of these functions is critical for
induce neutrophil release from the BM, including chemo- normal host defense mechanisms. Diseases affecting each of
kines, cytokines, microbial products, and various other these features and functions of neutrophils result in
inflammatory mediators (eg, C5a). The chemokine stromal enhanced susceptibility to infection.
derived factor-1 (SDF1, CXCL12) and chemokine receptor
CXCR4 play a key role in regulating neutrophil trafficking in Eosinophils and basophils
the BM by providing retention signals, whereas the chemo-
kine receptor CXCR2 and its ligands relay a release signal. Normally, the marrow contains a small proportion of eosinophils
and basophils. The granules of eosinophils contain histamine and
proteins important for the killing of parasites. Eosinophil
Neutrophil clearance
production is increased and eosinophilia (ie, >0.7 × 109/L) occurs
Neutrophil homeostasis in the blood is determined, in part, by in allergic disorders (eg, asthma, allergic rhinitis, dermatitis),
the rate of clearance from the circulation. Once released into parasitic infections, collagen vascular diseases, and drug reactions.
the circulation, neutrophils are cleared rapidly with a mean Eosinophilia also frequently occurs with myeloproliferative
half-life of only 6.3 hours (range 4-10 hours). Neutrophils are neoplasms and is the hallmark feature of chronic eosinophilic
cleared primarily in the liver, spleen, or BM, where apoptotic leukemia and the hypereosinophilic syndrome, a disorder
or aged neutrophils are phagocytosed by macrophages. characterized by autonomous eosinophil production with end-
organ or tissue complications.
Basophils are the least numerous blood leukocytes. Baso-
Neutrophil extravasation philic granules contain histamine, glycosaminoglycans,
Neutrophils in the circulation loosely attach and subse- major basic protein, proteases, and a variety of other vasoac-
quently adhere to vascular endothelium in response to the tive inflammatory mediators. Basophils primarily function
local production of inflammatory cytokines and chemo- to activate the immediate (type 1) hypersensitivity responses.
kines (Figure 15-1). Selectins mediate neutrophil rolling Increases in basophils are associated with hypersensitivity
and β2-integrins mediate firm adherence and vascular reactions, including drug and food allergies. Basophilia is a

Granulocytes: neutrophils, eosinophils, and basophils | 399
deviations above the mean. Neutrophilia is associated with

Bone marrow
Blood many different types of stress, including recent exercise, infec-
Stem cell tion, and inflammatory diseases (Table 15-2). A prompt increase
in the blood neutrophil count, as well as the circulating levels of
CFU-GM other leukocytes, occurs with acute stress, exercise, anxiety, and
CFU-G epinephrine administration. Only rarely does this response
more than double the count. It is attributable to demargination
Myeloblast of cells, not to the release of cells from the marrow reserve. Nor-
mally, approximately one half of the neutrophils in the circula-
Release
tion are loosely adherent to blood vessels and not counted in a
Promyelocyte G-CSF
routine blood count. These cells are described as being in the
marginal pool. The other half of the neutrophils are circulating
freely with the red cells and platelets. They are described as being
Myelocyte in the circulating pool.
Neutrophilia associated with infections and inflammatory
disorders occurs by two general mechanisms. First, during
Metamyelocyte infection, a number of inflammatory cytokines are released
into the circulation that induces the release of mature neu-
trophils from the BM. Second, the sustained cytokine
Segmented response associated with infections may stimulate neutro-
neutrophil
phil production in the BM. G-CSF, C/EBPβ and signal trans-
ducer and activator of transcription 3 (STAT3) are key
Extravasation Removal of aged
regulators of this form of stress or emergency granulopoiesis.
into tissue neutrophils In contrast to neutrophil demargination, neutrophilia asso-
• Selectins • Spleen ciated with infections and inflammatory disorders is marked
• Integrins • Liver by the presence of an increase in immature granulocytes in
• Chemokines • Bone marrow
the blood, mostly including band forms as well as metamy-
Figure 15-1 Neutrophil homeostasis. The bone marrow is the primary elocytes and occasionally early granulocytic precursors. In
site of granulopoiesis in humans. Under basal conditions, the bone addition, there often is a change in the morphology of neu-
marrow contains a large reservoir of mature neutrophils. Neutrophils are trophils with the appearance of vacuoles and “toxic granula-
released into the circulation in a regulated fashion. The principal cytokine tion,” because of more intense staining of the primary
regulating both granulopoiesis and neutrophil release into the blood is granules of neutrophils. Cells released prematurely also may
granulocyte colony-stimulating factor (G-CSF). Neutrophils in the contain bits of endoplasmic reticulum that stain as blue bod-
circulation have two general fates. In response to local infection or
ies in the cytoplasm, called Döhle bodies (Figure 15-2).
inflammation, neutrophils can emigrate into tissue. Neutrophil
Neutrophilia is also a feature of the myeloproliferative
emigration is a highly orchestrated process that includes sensing of
neoplasms, particularly CML, atypical CML, and the rare
chemokine gradients, selectin-mediated rolling on inflamed endothelium,
and integrin-mediated adhesion and diapedesis through the endothelium.
chronic neutrophilic leukemia, and it can sometimes be seen
Alternatively, senescent (aged) neutrophils are cleared from the circulation in the myeloproliferative-myelodysplastic overlap syndrome
primarily by the macrophages in the spleen, liver, and bone marrow. chronic myelomonocytic leukemia. In most cases of reactive
neutrophilia, the inciting infection (or other stress) is usually
clinically obvious, and the neutrophilia is self-limited. In
common feature of myeloproliferative neoplasms, particu- patients without demonstration of a clonal marker by either
larly chronic myeloid leukemia (CML), and can aid in diag- cytogenetic or molecular testing, clinical features such as the
nosis of these disorders. Basophilia also can be associated presence of splenomegaly, leukoerythroblastic features on
with other chronic inflammatory diseases, such as tubercu- the blood smear (teardrop and nucleated red blood cells),
losis, ulcerative colitis, and rheumatoid arthritis, but is rarely basophilia, or circulating promyelocytes or blasts are highly
seen as an isolated finding. suggestive of an underlying myeloproliferative neoplasm.
In rare cases, neutrophilia may be due to intrinsic defects
of neutrophil function. Leukocyte adhesion deficiency and
Neutrophilia
familial Mediterranean fever are associated with neutro-
Neutrophilia is an excess of circulating neutrophils and is typi- philia; they are discussed in detail in the section “Inherited
cally defined as an absolute neutrophil count (ANC) >2 standard disorders of neutrophil function” in this chapter.

Table 15-2 Causes of neutrophilia
Acute neutrophilia Chronic neutrophilia
Acute infections Chronic infections

Many localized and systemic acute bacterial, mycotic, rickettsial,
spirochetal, and certain viral infections
Inflammation or tissue necrosis Inflammation

Burns, electric shock, trauma, myocardial infarction, gout, Continuation of most acute inflammatory reactions, such as rheumatoid
vasculitis, antigen-antibody complexes, complement activation arthritis, gout, chronic vasculitis, myositis, nephritis, colitis,
pancreatitis, dermatitis, thyroiditis, drug-sensitivity reactions,
periodontitis, Sweet syndrome, familial periodic fever syndromes
Physical or emotional stimuli Tumors

Cold, heat, exercise, convulsions, pain, labor, anesthesia, surgery, Any but especially gastric, bronchogenic, breast, renal, hepatic,
severe stress pancreatic, uterine, and squamous cell cancers
Drugs, hormones, and toxins Drugs, hormones, and toxins

Epinephrine, etiocholanolone, endotoxin, glucocorticoids, Cigarette smoking, continued exposure to many substances that produce
venoms, vaccines, colony-stimulating factors acute neutrophilia; lithium; rarely, as a reaction to other drugs
Metabolic and endocrinologic disorders

Pregnancy and lactation, eclampsia, thyroid storm, Cushing disease

Hematologic disorders
Rebound from agranulocytosis or therapy of megaloblastic anemia,
chronic hemolysis or hemorrhage, asplenia, myeloproliferative
disorders
Hereditary and congenital disorders

Down syndrome, familial Mediterranean fever, leukocyte adhesion
deficiency, hereditary neutrophilia
Chronic idiopathic neutrophilia
Neutropenia
Neutropenia is commonly defined as an ANC of <1,500 that neutropenia is also more prevalent in males and in chil-
cells/µL. Neutrophil levels can be lower in health in persons dren <5 years of age.
of some ethnic and racial groups (eg, Africans, African Neutropenia is classified based on the ANC as severe
Americans, and Yemenite Jews). For example, in adult Afri- (<500/µL), moderate (500-1,000/µL), or mild (1,000-1,500/
can Americans, the 95% confidence limits for the ANC are µL). The risk of infection begins to increase with an ANC
1,300-7,400 cells/µL, which is slightly lower than for the pop- <1,000/µL. Patients with neutropenia are prone to develop
ulation of European descent. A survey of 25,222 participants bacterial infections, typically caused by endogenous flora and
in the National Health and Nutrition Examination Survey involving mucous membranes, including gingivitis, stomati-
showed that 4.5% of black participants had an ANC of tis, perirectal abscesses, cellulitis, and pneumonia. Fungal
<1,500/µL, compared with 0.79% and 0.35% of whites and infections are less common but are a major cause of mortal-
Mexican Americans, respectively. The survey also showed ity in patients with chronic severe neutropenia. There is no
Figure 15-2 Photomicrographs
of blood smears showing typical
neutrophil morphology from
(a) a healthy individual; (b) a
patient with sepsis showing Döhle
bodies (arrow) and toxic
granulations; (c) a patient with
Chédiak-Higashi syndrome showing
large cytoplasmic inclusions
(arrowhead). ASH Image Bank
images 3780 (a), 3778 (b), 2979 (c).

Monocytes and tissue histiocytes | 401
increase in susceptibility to viral or parasitic infections. The Table 15-3 Causes of neutropenia
differential diagnosis of neutropenia is broad (Table 15-3).
Congenital neutropenia syndromes*
Chemotherapy-induced neutropenia is a common compli-
Severe congenital neutropenia, with and without extrahematopoietic
cation of the treatment of cancer with myelotoxic drugs. manifestations
Neutropenia is a frequent manifestation of myelodysplastic Cyclic neutropenia
syndromes (MDS), acute leukemia, and marrow-infiltrative Shwachman-Diamond syndrome
processes, such as myelofibrosis, or metastatic carcinoma; WHIM syndrome (myelokathexis)
these are discussed in detail in their respective chapters. Poikiloderma with neutropenia, Clericuzio type
Fanconi anemia
Dyskeratosis congenita
Disorders of vesicular transport
Monocytes and tissue histiocytes Chédiak-Higashi syndrome
Griscelli syndrome, type II
Monocytes share their origin and functions with neutrophils. Hermansky-Pudlak syndrome, type II
They are phagocytes, ingesting and killing microorganisms. p14 deficiency
Monocytes also may become the fixed phagocytic cells that Cohen syndrome
line portions of the circulation, particularly in the spleen and Barth syndrome
Cartilage-hair hypoplasia syndrome
liver. In these tissues, their role is to clear particulate matter,
Pearson’s syndrome
including microorganisms, and aged or damaged blood cells Glycogen storage disease type 1b
from the circulation. They are also an important source of Neutropenia associated with immunodeficiency syndromes
inflammatory cytokines (eg, tumor necrosis factor, interleu- GATA2 deficiency
kin-1, interferon-g) that cause fever and many of the symp-
toms associated with infectious and inflammatory diseases. Acquired neutropenia
Neonatal alloimmune neutropenia
Monocytes can differentiate to tissue histiocytes, and these
Primary autoimmune neutropenia
cells may then fuse to form giant cells, as occurs in tuberculo- Secondary autoimmune neutropenia
sis and sarcoidosis. Alveolar macrophages in the lung, Kupffer Systemic lupus erythematosus
cells in the liver, and osteoclasts and phagocytic cells that Felty syndrome
comprise reticuloendothelial tissue are all forms of tissue his- Nutritional deficiencies
tiocytes that are thought to derive from blood monocytes or, Vitamin B12, folic acid, copper
in some cases, from circulating monocyte progenitors. These Myelodysplastic syndrome
Acute leukemia
phagocytic cells not only serve antimicrobial, scavenger, and
Myelophthisis (bone marrow infiltration
secretory functions but also participate in wound repair and tumor, fibrosis, granulomas)
antigen processing and presentation. Inappropriate over- Large granular lymphocytic leukemia
stimulation of tissue macrophages, usually in the setting of Neutropenia associated with infectious disease
infectious stimuli with or without underlying acquired or Sepsis
inherited immune dysregulation, may contribute to the sep- Rickettsial: human granulocytic ehrlichiosis
sis-associated systemic inflammatory response syndrome. Viral: mononucleosis, HIV
Drug-induced neutropenia
Hypersplenism
Dendritic cells
* Not restricted to disorders in which neutropenia is the only
Dendritic cells are hematopoietic-derived cells that partici- manifestation.
pate in the innate and adaptive immune responses. These
cells are distributed widely throughout virtually all tissue Mature dendritic cells also produce cytokines, which can
types and, in particular, lymphoid organs and organs with prime T-cells toward either a TH1 versus TH2 response. In
barrier function, such as the skin and mucosal surfaces. A addition, dendritic cells possess other immunomodulatory
major function of dendritic cells is to process and present functions, including the activation of other lymphocyte sub-
antigen to the immune system. Immature dendritic cells in sets and induction of immune tolerance. Specific dendritic
the peripheral tissue express surface receptors that allow cell subsets have been identified and characterized based on
them to recognize and take up extracellular antigens in their their location, cell surface phenotype, function, or develop-
environment. Upon encountering antigen, immature den- mental stage. The ability of dendritic cells to present tumor
dritic cells undergo activation and maturation and migrate antigens has led to their use in vaccine immunotherapy trials.
to secondary lymphoid organs where they bind and present An important pathologic condition of dendritic cells is Lang-
antigen in the context of major histocompatibility complex erhans cell histiocytosis, an acquired clonal disorder that can
(MHC) class I and II molecules to stimulate naive T-cells. lead to local or systemic tissue infiltration and organ failure.

Monocytosis Epidemiology
Monocytes normally account for approximately 1%-9% of Although the inherited BM failure syndromes are rare disor-
peripheral blood leukocytes, with absolute monocyte counts ders, collectively affecting just several dozen new patients in
ranging from 0.3 × 109 to 0.7 × 109/L. An increase in circulating the United States each year, a diagnosis with one of these
monocytes may be observed in chronic inflammatory conditions syndromes has profound implications for medical manage-
and infectious diseases, such as tuberculosis, endocarditis, and ment and treatment. FA is probably the most common and
syphilis. Monocytosis is a hallmark of chronic myelomonocytic is one of the best defined of these rare inherited conditions
leukemia and of the pediatric disorder juvenile myelomonocytic (common is a relative term here, as the incidence of FA in the
leukemia, and it also may be observed in other malignancies, United States has been estimated at approximately 1 in
including lymphomas and acute monocytic (monoblastic) leu- 130,000 live births).
kemias. In the inflammatory conditions, monocytosis is a reac- Several inherited marrow failure syndromes are compared
tive process resulting from the peripheral production of in Table 15-4. BM failure usually is not the only feature of the
cytokines, which stimulate monocyte production. Malignant inherited marrow failure disorders, and marrow failure may
monocytosis is presumed to be due to specific molecular defects even be absent in some patients. Alternatively, isolated mar-
affecting monocyte proliferation, differentiation, and survival. row failure or development of a malignancy may be the first
clinical manifestation of FA in the absence of other clinical
Monocytopenia stigmata, and occasionally, patients with FA may first come
to clinical attention as young adults.
Transient monocytopenia occurs with stress, infections,
including overwhelming sepsis, and as the result of cytotoxic
chemotherapy. A decreased absolute monocyte count can be Pathophysiology
encountered in BM failure states such as aplastic anemia (AA)
and, less commonly, MDS. Monocyte numbers are suppressed, A hallmark of cells from patients with FA is hypersensitivity
but circulating monocyte counts and function are maintained to genetic damage induced by DNA-damaging and cross-
in many other conditions that cause neutropenia. Monocyto- linking agents, such as DEB and mitomycin C (MMC). The
penia in association with natural killer (NK) cell deficiency and underlying molecular defects resulting in the FA phenotype
B cell lymphopenia typifies GATA2-associated diseases such as involve components of a critical pathway that regulates cel-
MonoMAC syndrome (discussed further in the section lular DNA-damage recognition and response.
“GATA2 deficiency” in this chapter.). Monocytopenia, along FA is a heterogeneous disease at the molecular level, with
with neutropenia, is characteristic of hairy cell leukemia. at least 16 distinct complementation groups (ie, distinct
genes; complementation in molecular biology means that
two different gene loci encode proteins of distinct function
Inherited marrow failure syndromes that can each provide something the other lacks,
facilitating identification in vitro) identified to date. More
Fanconi anemia than 75% of patients fall into complementation groups A or
C. Molecular cloning of the genes corresponding to each of
Clinical case these complementation groups led to the identification of a
complex DNA repair pathway composed of distinct FA
A 12-year-old boy presents to his primary care physician with proteins (Figure 15-3). (Confusingly, there is no Fanconi H
pallor and bruising. His past medical history is remarkable only
protein; it turned out to be the same complementation
for an orchiopexy during the first year of life to correct an unde-
group as Fanconi A. There is also no Fanconi K protein.)
scended testis. Pancytopenia is now noted. The patient and his
Each of these genes, when biallelically mutated, can cause
parents do not report any medication or toxin exposures. There
are no siblings. On initial examination, the boy appears to be a
FA, except for FANCB, which causes X-linked recessive
normal prepubescent male. On closer examination, however, disease.
his thumbs appear underdeveloped, and patches of cutane- DNA-damage activates an eight-protein core complex
ous hyperpigmentation are noted on his trunk. BM aspiration consisting of Fanconi A, B, C, E, F, G, L, and M proteins,
and biopsy are performed. The marrow cellularity is only 10%; which results in monoubiquitylation of the Fanconi I
the marrow aspirate shows hypocellular fragments and rare (FANCI) protein and the Fanconi D2 (FANCD2) protein, by
megakaryocytes, most of which are abnormal uninucleate forms. a ubiquitin ligase domain of FANCL. The modified FANCD2
Cytogenetic studies are normal. Exposure of peripheral blood protein then translocates to chromatin nuclear foci and
mononuclear cells to diepoxybutane (DEB) results in numerous localizes to sites of DNA repair, where it interacts with the
chromosomal breakages, confirming a diagnosis of FA.
breast cancer susceptibility protein BRCA1, as well as with

Table 15-4 Comparison of the inherited marrow failure syndromes
Male:
female Median age % diagnosed Nonhematological
Syndrome ratio at diagnosis >15 years Somatic features Hematological features cancers Screening test Genetics (genes)
BK-ASH-SAP_6E-150511-Chp15.indd 403
Fanconi anemia (FA) 1.2:1 6.6 9 Skin hyperpigmentation and cafe-au Pancytopenia, hypocellular Solid tumors Increased chromosome Autosomal recessive
lait spots, short stature, triangular bone marrow, MDS, (head, neck, breakage in cells (15), X-linked
face, abnormal thumbs/radii, leukemia gynecologic, cultured with DNA recessive (1)
microcephaly, abnormal kidneys, liver, CNS) cross-linkers (DEB and
decreased fertility MMC)
Dyskeratosis 2:1 15 46 Nail dystrophy, abnormal skin Pancytopenia, hypocellular Solid tumors (head Very short telomere X-linked recessive (1),
congenita (DC) pigmentation, leukoplakia, lacrimal marrow, MDS, leukemia and neck) length autosomal dominant
duct stenosis, pulmonary fibrosis, (4), autosomal
liver fibrosis, esophageal strictures, recessive (8)
early gray hair, osteoporosis,
cerebellar hypoplasia, retinopathy,
hypogonadism, urethral stricture
Diamond-Blackfan 1.1:1 0.25 1 Short stature, abnormal thumbs, Macrocytic anemia, Solid tumors Elevated red cell Autosomal dominant
anemia (DBA) hypertelorism, cardiac septal defect, erythroid hypoplasia in (osteosarcoma) adenosine deaminase (13), X-linked
cleft lip or palate, short neck marrow, MDS, leukemia (ADA) recessive (1)
Shwachman- 1.5:1 1 5 Short stature, exocrine pancreatic Neutropenia, anemia, None Decreased pancreatic Autosomal recessive (1)
Diamond insufficiency with malabsorption thrombocytopenia, AA, isoamylase and
syndrome (SDS) MDS, leukemia trypsinogen
Severe congenital 1.2:1 3 13 None Neutropenia, MDS, None Bone marrow exam for Autosomal dominant
neutropenia (SCN) leukemia promyelocyte arrest and recessive (6)
Congenital 0.8:1 0.1 0 Usually none Thrombocytopenia; None Bone marrow exam for Autosomal recessive (1)
amegakaryocytic decreased megakaryocytes megakaryocytes
thrombocytopenia initially, later AA; MDS,
(CAMT) leukemia
Thrombocytopenia 0.7:1 0.6 0 Absent radii, abnormal ulnae or Thrombocytopenia, MDS, None Bone marrow exam for Autosomal recessive (1)
absent radii humeri (phocomelia), thumbs leukemia megakaryocytes
syndrome (TAR) present, occasional cryptorchidism,
hypertelorism, horseshoe kidney,
hemangiomas, micrognathia, cow’s
milk allergy, cardiac anomalies
AA = aplastic anemia; CNS = central nervous system; DEB = diepoxybutane; MDS = myelodysplasia; MMC = mitomycin C.
Inherited marrow failure syndromes | 403
4/26/2016 10:13:07 PM
Ub Ub
DNA damage
AGBL
CFME I D2
Stalled replication
fork
Ub Ub
I D2
D2 I
RAD51
BRCA2 PALBN2(N)
ATR ATM (D1)
ATR activation SLX4 BRCA1 ERCC4
P I D2 P BRIP1(J)
Checkpoint DNA repair

response
Genomic
stability
Figure 15-3 A model of the Fanconi anemia (FA) pathway. The FA core complex consists of eight FA proteins (A, B, C, E, F, G, L, and M) and
this together with ATR (ataxia-telangiectasia and RAD3 related protein) is essential for the ubiquitination-activation of I-D2 complex after DNA
damage. Activated I-D2-Ub translocates to DNA repair foci where it associates with other DNA damage response proteins, including BRCA2
and RAD51 and participates in DNA repair. The proteins mutated in different FA subtypes are shaded yellow.
effector proteins FANCJ (a structure-specific helicase that cumulative DNA damage. Recent studies suggest an exacer-
binds to BRCA1) and FANCN. FANCD2 is phosphorylated bated p53/p21 DNA-damage response impairs hematopoi-
by the ataxia-telangiectasia-mutated (ATM) protein kinase etic stem and progenitor cells in patients with FA. There is
in response to ionizing radiation, linking the Fanconi also emerging evidence for dysregulation of the TGFβ
pathway to the ATM/ATR/Chk1 DNA damage–sensing and pathway.
checkpoint response pathway. Cells from patients with FA have been shown to have
Biallelic mutations in the BRCA2 gene, encoding a DNA increased sensitivity to apoptosis, and elevated levels of
repair enzyme for which constitutional heterozygous muta- apoptosis may contribute to the hematopoietic failure
tions had been linked to breast, ovarian, and prostate cancer phenotype. Additionally, FA hematopoietic progenitor
susceptibility, were shown to be the underlying abnormality cells are hypersensitive to interferon-g, a known inhibitor
in patients with FA who previously were assigned to comple- of hematopoiesis.
mentation group D1; thus, BRCA2 and FANCD1 turned out
to be identical. FANCD1/BRCA2 is essential for recruiting Clinical features and diagnosis of FA
the homologous recombination-promoting protein RAD51
into DNA-damage-inducible nuclear foci. These data clearly FA is an autosomal recessive (usually) or X-linked
establish a defect in the ability to repair certain types of DNA recessive (rarely) disorder that is characterized by
damage appropriately as the underlying abnormality in FA, pancytopenia and congenital defects in the cutaneous,
although the precise molecular mechanisms are still being musculoskeletal, and urogenital systems. Patients with
elucidated. It has been established the FA pathway has a role FA were recognized historically by characteristic physical
in protecting cells from endogenous aldehyde induced DNA findings, including short stature, microcephaly, intense
damage. patchy brown pigmentation of the skin (café au lait
Despite progress in identifying the genetic and biochemi- spots), gonadal abnormalities, and malformations of the
cal defects in FA, it is not known precisely why affected indi- thumb and kidney. Hemoglobin F levels are increased in
viduals develop BM failure or are at risk for development of FA, and 80% of patients develop signs of BM failure by
clonal hematopoietic neoplasms, including MDS and acute age 20 years. It is now clear that many patients
myeloid leukemia (AML). The cause of the progressive AA in (approximately 30%) with FA lack any abnormal physical
FA has been thought to be due to the loss of HSCs because of findings, and some may not demonstrate overt marrow

failure, so a high index of suspicion is required. cell carcinomas, with head and neck, in particular, esoph-
Chromosome breakage testing now secures the diagnosis ageal, and vulvar/vaginal tumors much more common in
in most patients. patients with FA than in the general population. In addi-
Contemporary diagnosis of FA is based on the analysis tion to hepatocellular carcinoma, peliosis hepatis and
for chromosome breaks in phytohemagglutinin- hepatic adenomas occur with increased frequency, espe-
stimulated peripheral blood lymphocytes cultured with cially in patients treated with androgens. The risk of AML
and without clastogenic agents (eg, DEB or MMC). is 700-fold in patients with FA compared with the general
Results usually are reported as percentage of cells with population but plateaus after the second decade of life,
chromosome aberrations; the percentage of such cells whereas the risk of solid tumors increases with age; in a
inducible in samples from healthy individuals depends competing risk analysis, 30% of patients with FA will
on the specific laboratory protocol but is increased develop a solid tumor by age 48 years.
dramatically in FA. The finding of an increased percentage The clinical significance of an abnormal marrow cyto-
of cells arrested at the G2/M phase (4N DNA) of the cell genetic clone (eg, monosomy 7) in the absence of mor-
cycle after treatment with a clastogenic agent, ascertained phologic dysplasia is not always clear, because these clones
by flow cytometry, also is consistent with a diagnosis of may be stable or even regress with time. The exquisite
FA. Cell cycle analysis may be used in conjunction with, sensitivity of patients with FA to the DNA-damaging
but should not be used in lieu of, chromosome breakage effects of chemotherapy and radiation poses a formidable
testing. BM cells should not be used for chromosome obstacle to the treatment of malignancies in these patients.
breakage studies because false-negative results are more The most successful treatment of solid tumors in FA
likely. These in vitro diagnostic tests have revealed that results from early detection and complete surgical exci-
~30% of patients with FA lack typical physical findings, sion. For this reason, regular tumor surveillance is an
and ~10% of patients first present at age >16 years. important aspect of medical management beginning in
Diagnosis of FA may be complicated by the develop- the late teenage years.
ment of somatic mosaicism in the lymphocytes. Somatic
mosaicism results from a genetic reversion to normal
Treatment
(non-FA), such that a subset of lymphocytes is no longer
susceptible to chromosomal breakage in response to DEB The only potentially curative option for marrow failure in
or MMC. The molecular mechanism underlying the patients with FA is allogeneic HSC transplantation (HSCT).
reversion varies, with evidence for intragenic recombina- Because of the increased sensitivity of FA cells to DNA
tion, mitotic gene conversion, or second site suppressor in damage-inducing agents, it is necessary to use modified
the small number of studied cases. Because reversion to transplantation conditioning regimens, such as attenuated
wild-type confers a growth advantage over the nonre- alkylating agent and radiation doses, or alternative non–
verted FA cells, the diagnosis of FA may be missed. In DNA-damaging agents, such as fludarabine or antithymocyte
patients for whom there is a strong suspicion for FA, the globulin (ATG). For this reason, it is critical to identify
diagnosis may be made by testing for chromosomal break- patients with FA as having the condition before proceeding
age in response to DEB or MMC using cultured skin to stem cell transplantation. Patients with FA who present
fibroblasts obtained from a punch biopsy. with MDS or AML without an observed BM failure phase
Although spontaneous chromosomal alterations can be may go unrecognized until the use of standard transplantation
observed in Bloom syndrome and ataxia telangiectasia, DEB- conditioning results in a disaster.
induced chromosomal alterations are not seen in those dis- Stem cell transplantation in FA is associated with an
orders. Patients with Nijmegen breakage syndrome (NBS) increased risk of subsequent solid tumors, particularly in the
may exhibit increased chromosomal breakage with MMC, setting of chronic graft-versus-host disease (GVHD). Stem
and this condition must be distinguished from FA; immuno- cell transplantation corrects only the hematopoietic defect,
logic abnormalities are characteristic of NBS but not typi- and the patient remains at risk for FA-related complications
cally in FA, and in doubtful cases, testing of the NBS1 gene in other tissues, such as solid tumors. Despite these limita-
can help diagnostically. tions, stem cell transplantation from a matched (unaffected)
Complications of marrow failure are the most common sibling may be considered as the initial treatment of choice
causes of death in FA, but FA is also characterized by an for patients with FA who present with marrow failure. Unre-
increased incidence of malignancies. Approximately 10%- lated donor stem cell transplantation is complicated in these
15% of patients with FA develop MDS or AML, often in patients because the intensive conditioning regimens
the context of a hypoplastic marrow and monosomy 7. required for engraftment can exacerbate cellular damage and
Patients with FA also are at increased risk for squamous increase the risk for GVHD. Transplantation outcomes are

better if transplantation occurs before the development of Dyskeratosis congenita

leukemia, so regular surveillance of the peripheral blood
counts and BM is recommended.
Clinical case
Androgens (eg, oxymetholone with a starting dose of
0.5 mg/kg/d) may elevate the blood counts in a subset of A 16-year-old boy presents to his doctor with a history of skin
patients with FA. Although the response to androgen changes, nail abnormalities, and bruising. Following referral to
therapy typically is most pronounced in the red blood cell the hematologist, examination shows he has significant nail dys-
lineage, improvements in platelet and neutrophil counts trophy and reticulate skin pigmentation around the neck. Blood
count reveals a nonsevere pancytopenia, and the BM cellularity
also have been seen. Responses may be delayed, particularly
is found to be markedly reduced. Peripheral blood chromosomal
for platelets, where first responses have been reported as far
breakage analysis following exposure to DEB is normal. Subse-
as 6 months out from initiation of treatment. The neutro
quent tests, however, show he has very short telomeres and a
phil count may also respond to G-CSF or granulocyte- missense mutation in the DKC1 gene, confirming a diagnosis of
macrophage colony-stimulating factor (GM-CSF). Some X-linked dyskeratosis congenita.
patients who initially respond to androgens may become
refractory over time. Supportive therapy with transfusions
can be considered, but in the patient who is a candidate for Clinical features
allogeneic stem cell transplantation, the use of transfusions
The spectrum of diseases encompassed by the term dyskera
should be minimized to prevent alloimmunization, and
tosis congenita (DC, sometimes abbreviated as DKC) has
transfusions should never be from family members. Iron
expanded considerably since its initial description in 1910. In
overload may develop in patients receiving chronic red
its classic form, DC is usually characterized by the mucocu-
blood cell transfusions.
taneous triad of abnormal skin pigmentation, nail dystrophy,
Because of the risk of neoplasia, patients with FA should
and leukoplakia. A wide spectrum of features affecting every
undergo regular screening for cancer. Although there is no con-
system (cutaneous, dental, gastrointestinal, neurological,
sensus on optimal frequency of such screening evaluations,
ophthalmic, pulmonary, and skeletal) in the body, particu-
annual gynecologic examination for female patients is recom-
larly the BM, have been associated with DC. Three modes of
mended after menarche, and regular dental care is also impor-
inheritance have been recognized: X-linked recessive, auto-
tant, with careful examination for head and neck cancer.
somal dominant, and autosomal recessive. The clinical phe-
Surveillance with liver ultrasound at least once yearly is recom-
notype associated with each of these genetic forms can vary
mended for patients undergoing treatment with androgens.
widely. The main causes of mortality in DC are BM failure
(~60%-70%), pulmonary disease (~10%-15%), and malig-
Key points nancy (~10%).
Clinical features of classic DC often appear in childhood.
• FA is usually an autosomal recessive and rarely X-linked
The abnormal skin pigmentation and nail changes usually
recessive cause of BM failure that is due to a germ line defect in
appear first and become more pronounced over time. The
DNA repair.
• Approximately 80% of patients with FA develop signs of
onset is usually prior to age 20 years, in many cases below the
marrow failure, but the absence of marrow failure does not rule age of 10 years. BM failure develops frequently below the age
out FA if typical physical stigmata are present. of 20 years with up to 80% of patients showing signs of BM
• The diagnostic test for FA is a DEB or MMC chromosome failure by the age of 30 years. But there is considerable varia-
breakage study. tion between patients with respect to age of onset and disease
• FA can present in adulthood and without classic features other severity even within the same family. This causes difficulty in
than BM failure. These patients remain at risk for marrow failure, making a diagnosis. Equally, it is not uncommon for the BM
leukemia, and solid tumors. It is important to screen for FA in failure or an abnormality in another system to present before
patients with idiopathic marrow failure syndromes or myelodys- the more classic mucocutaneous features, and this is being
plasia who are under 40 years of age.
recognized increasingly since the advances in its genetics.
• Chemotherapeutic agents and radiation are poorly tolerated;
The diagnostic criteria for DC vary among experts. In the
attenuated conditioning regimens are necessary for stem cell
absence of molecular testing, the minimal clinical criteria for
transplantation.
• Stem cell transplantation is the only curative option for
diagnosis of DC used by some includes the presence of at least
FA-associated hematologic manifestations. two out of the four major features (abnormal skin pigmenta-
• Careful monitoring for malignancies allows early institution of tion, nail dystrophy, leukoplakia, and BM failure) and two or
treatment, with attention to minimizing exposure to chemo- more of the other somatic features known to occur in DC. The
therapy and radiation. detection of a pathogenic variant in a DC-associated gene or

the presence of very short telomeres, may allow for a diagnosis disorder and that other syndromes with overlapping presen-
in patients prior to meeting clinical criteria alone. Short telo- tation can share the same genetic mutations. The first such
meres have also been associated with idiopathic pulmonary example was the Hoyeraal-Hreidarsson (HH) syndrome, a
fibrosis and with cryptogenic cirrhosis. severe multisystem disorder, which is characterized by growth
retardation of prenatal onset, microcephaly, cerebellar hypo-
plasia, AA, and immunodeficiency. Because of the overlap in
Pathophysiology and link to other diseases
features, it was suggested and subsequently proven that HH is
Since 1998, more than 11 DC genes have been identified, and a severe variant of DC because of the presence of DKC1 muta-
these account for ~65% of DC cases. Figure 15-4 shows the tions in males with the classical presentation of HH. DKC1
different components of the telomerase and shelterin com- mutations are not the only cause of HH, however; mutations
plexes as well as other factors important in telomere mainte- in other genes also can lead to a phenotype of HH (see sec-
nance. Mutations in the genes encoding RTEL1, TIN2, TPP1, tions on Autosomal dominant and Autosomal recessive DC).
TERT, TERC, dyskerin, NOP10, NHP2, USB1, CTC1 and
TCAB1 have been associated with DC and related disorders.
Autosomal dominant DC and its link to
telomeres and other diseases
X-linked DC and the Hoyeraal-Hreidarsson syndrome
Autosomal dominant DC is heterogeneous; to date, heterozy-
The gene responsible for X-linked DC was mapped to Xq28 gous mutations in four genes (TERC, TERT, TINF2, and
in 1986 and identified as DKC1 through positional cloning in ACD) have been characterized. The identification of heterozy-
1998. The DKC1 gene is highly conserved and encodes the gous mutations in TERC (telomerase RNA component) in
protein dyskerin. With the identification of mutations in 2001 was a major advance in the DC field, as it provided a
DKC1, the first diagnostic tests became available. It also pro- direct link between DC and telomerase. Telomerase is a ribo-
vided the first firm evidence that DC was not a homogenous nucleoprotein composed of two core components:a catalytic
Shelterin
Telomerase
Tankyrase Dyskerin
NOP10
TERT NHP2
Helicase
TIN2 TPP1 GAR1
3’
TRF1 TRF2 POT1
RTEL1 TERC
5’
3’ 5’
RAP1
TEN1
STN1
USB1 CTC1 TCAB1
Capping
snRNA processing Cajal body
Figure 15-4 Complexes important in telomere maintenance. A schematic representation of the telomerase complex (dyskerin, GAR1, NHP2,
NOP10, TERC, and TERT), the shelterin complex, and their association with different categories of dyskeratosis congenita and related diseases.
The telomerase complex is a RNA-protein complex because TERC is a 451b RNA molecule that is never translated. The other molecules
(dyskerin, GAR1, NHP2, NOP10, and TERT) are proteins. Recent studies suggest that the minimal active telomerase enzyme is composed of two
molecules each of TERT, TERC, and dyskerin. Dyskerin, GAR1, NHP2, and NOP10 are believed to be important for the stability of the
telomerase complex. The shelterin complex is made up of six proteins (TIN2, POT1, TPP1, TRF1, TRF2, and RAP1) and is important in
protecting the telomere. Mutations in components of the telomerase complex, the shelterin complex and related molecules, as occurs in different
subtypes of DC in related disorders, result in telomere shortening. CTC1 = conserved telomere maintenance component 1; RTEL1 = regulator of
telomere length 1; TCAB1 = telomere Cajal body protein 1; USB1 = U6 small nuclear RNA (snRNA) biogenesis 1.

component, which adds the repeats; telomerase reverse tran- composition and protein interaction of the components of
scriptase (TERT), and an RNA subunit, TERC, which includes shelterin complex appears to be highly ordered with TIN2
the template. It functions as a specialized polymerase that adds playing a pivotal role. In a subset of patients with DC, HH,
the telomeric repeat (TTAGGG) to the 3′ end of the DNA AA, and Revesz syndrome heterozygous mutations in the
strands after replication. Because of the semiconservative TIN2 component of shelterin have been identified. This dis-
nature of DNA replication, telomerase is essential to maintain covery extends the range of the DC spectrum of diseases
telomere length in rapidly dividing cells, such as cells of the even further. Revesz syndrome is characterized by bilateral
hematopoietic system, including activated T-cells and mono- exudative retinopathy, BM hypoplasia, nail dystrophy, fine
cytes. Telomerase is also expressed in germ cells, stem cells, and hair, cerebellar hypoplasia, and growth retardation. Patients
their immediate progeny. Without telomerase, the telomeres with TIN2 mutations tend to have severe disease, and this is
shorten with each successive round of replication, and when associated with very short telomere lengths. Interestingly,
they reach a critical length, the cells enter senescence. In cells nearly all the patients have de novo TIN2 mutations, which
in which telomerase is not present, telomere shortening is part gives rise to a different mechanism that causes the disease. In
of the normal process of cellular aging. patients with heterozygous TERC and TERT mutations,
Mutations in patients with autosomal dominant DC were studies have shown that the phenomenon of genetic antici-
identified in TERC initially, and it was through the identification pation frequently is involved; a parent of an affected child
of mutations within this molecule that led to significant expan- has the same telomerase mutation but usually no overt signs
sion of the DC phenotype to include other hematological and of disease. In the child with the same heterozygous telomer-
nonhematological disorders. First, heterozygous mutations in ase mutation, however, the disease manifests itself at a much
TERC were identified in patients with AA and soon after in younger age and is usually more severe.
patients with MDS. This started to push the original clinical
diagnosis away from the classical mucocutaneous manifesta-
Autosomal recessive DC
tions to BM failure being the initial presenting feature.
The identification of mutations in DKC1 and TERC estab- Since 2007, progress has been made in understanding the
lished the pathology of defective telomere maintenance as genetic basis of autosomal recessive (AR)-DC. A large
being the principal underlying cause of DC; both dyskerin linkage study of 16 consanguineous families, including 25
(encoded by DKC1) and TERC are now recognized to be core affected individuals, did not identify a single common
components of telomerase, and patients with DKC1 and locus, suggesting there is genetic heterogeneity within this
TERC mutations have very short telomeres compared with subtype of DC. Since this observation, mutations in eight
age-matched controls. This finding led to further study of genes have been identified as causing AR-DC. The first
the telomerase complex to determine the genetic basis of the AR-DC gene to be identified was NOP10. The homozy-
remaining uncharacterized patients. The next gene to have gous NOP10 mutation identified in a large family affected
mutations identified was TERT, which encodes the enzy- a highly conserved residue. As a result of this mutation, all
matic component of the telomerase complex. The clinical the affected individuals had reduced telomere length and
presentation in patients with TERT or TERC mutations is reduced TERC levels. To date, no additional NOP10 muta-
highly variable. Although some patients will manifest suffi- tions have been described. In a subset of AR-DC, biallelic
cient features for a clinical diagnosis of DC, many will not; mutations have been identified in TERT. These mutations
instead appearing to have idiopathic aplastic anemia or give a different profile regarding telomerase activity and
MDS. In addition, heterozygous mutations in TERT and telomere length with both being greatly reduced com-
TERC have been identified in 15% of familial pulmonary pared with heterozygous TERT mutations. Biallelic muta-
fibrosis cases (and in some patients with pulmonary fibrosis tions in NHP2 have been identified in a third subset of
who do not have a family history), as well as in some patients AR-DC patients. Again telomere lengths and TERC levels
with cryptogenic cirrhosis and leukemia. are reduced in patients compared with normal controls.
In 2008, mutations in a component (TIN2) of the shel- Both NOP10 and NHP2 are components of H/ACA ribo-
terin complex were identified in one subtype of autosomal nucleoprotein complex (H/ACA RNP). This complex is
dominant DC. In 2014, a heterozygous mutation in TPP1 composed of a RNA molecule and four proteins, dyskerin,
(also a component of the shelterin complex) was identified GAR1, NOP10, and NHP2. These four proteins are highly
in a family with AA and MDS. The shelterin complex (com- conserved and have been shown to be involved in ribo-
posed of six proteins) has at least three effects on telomeres. some biogenesis, pre-mRNA splicing, and telomere main-
It determines the structure of the telomeric terminus, it has tenance. Mutations have been identified in all components
been implicated in the generation of t-loops, and it controls of this H/ACA RNP complex in patients with DC except
the synthesis of telomeric DNA by telomerase. The for GAR1.

In 2011, biallelic mutations in the TCAB1 gene were iden- Approximately two-thirds of patients with DC respond to
tified in two patients with AR-DC. TCAB1 is a telomerase oxymetholone or danazol; in some cases, the response can
holoenzyme protein that facilitates trafficking of telomerase last several years and involve all lineages. Patients with DC
to Cajal bodies, the nuclear sites of nucleoprotein complex can respond to a dose as low as 0.25 mg oxymetholone/kg/
modification and assembly. Compound heterozygous muta- day and this can be increased, if necessary, to 2-5 mg/kg/day.
tions in TCAB1 disrupt telomere localization to Cajal bodies, It is important to monitor for side effects (eg, liver toxicity).
resulting in misdirection of telomerase RNA to nucleoli. This Powdered forms of some androgens are available for those
in turn prevents telomerase from elongating telomeres, patients with esophageal strictures who have difficulty
thereby resulting in short telomeres. swallowing. It is possible to maintain reasonable blood
In another subgroup of AR-DC, biallelic mutations counts for a prolonged period with anabolic steroids in
have been found in the USB1 gene. Mutations in this gene many patients. The concurrent use of androgen and G-CSF
have unified this subgroup of DC with patients classified is not recommended due to reports of splenic peliosis with
as having poikiloderma with neutropenia and Rothmund- or without splenic rupture in patients with DC receiving
Thomson syndrome. This subgroup of AR-DC patients these treatments simultaneously.
appears to have normal-length telomeres and therefore The only long-term cure for the hematopoietic abnor-
represent a biologically different subtype to the other malities is allogeneic HSC transplantation, but this is not
characterized DC subgroups. USB1 is responsible for U6 without risk. Significant mortality is associated with BM
snRNA 3′ end processing. Clinically, patients with USB1 transplants for patients with DC than with other BM fail-
mutations can be identical or similar to those with genetic ure syndromes. One of the main reasons for this is the high
defects in telomere length maintenance. level of pulmonary and vascular complications that present
In 2012, biallelic mutations in CTC1, encoding conserved in these patients probably as a result of the underlying telo-
telomere maintenance component 1, were identified in a rare mere defect. The conditioning regimen appears to have an
subgroup of DC patients. As CTC1 mutations initially were impact on patient survival. The standard myeloablative
identified in the pleotropic syndrome Coats plus (character- conditioning regimens are associated with frequent and
ized by retinopathy, intracranial calcifications and cysts, severe adverse effects, such as pulmonary complications
osteopenia, and gastrointestinal abnormalities), this obser- and veno-occlusive disease. Recently, the adoption of non-
vation expands the complexity of phenotypes associated myeloablative fludarabine-based protocols has allowed for
with the “telomeropathies.” Patients with biallelic CTC1 successful engraftment in some patients with fewer com-
mutations usually tend to have non-severe pancytopenia. plications and lower toxicity. The long-term survival, how-
In 2013, biallelic RTEL1 mutations were identified in a ever, is unknown at present, but the initial response is
subset of patients with HH. RTEL1 is a helicase essential for encouraging as more effective treatment for DC. As with
telomere maintenance and regulation of homologous recom- FA, patients with DC need to be followed up long term for
bination. Patients harbouring biallelic RTEL1 mutations nonhematological complications.
have very short telomeres due to incorrect resolution of
t-loops but no significant defects in homologous recombina-
tion (normal chromosomal breakage score following Key points
exposure to DEB or MMC). These patients have a very
• DC is a marrow failure syndrome classically characterized by
tight clinical phenotype, which includes global bone mar-
the triad of dystrophic nails, reticulated skin pigmentation, and
row failure, cerebellar hypoplasia, and significant immune
oral leukoplakia.
deficiency. • Nonhematologic clinical features usually develop later in life
In 2014, an individual with biallelic mutations in the ACD and may be absent in young children.
gene (encoding the shelterin component TPP1) associated • Nonhematologic features of DC may be mistaken for chronic
with a HH-like phenotype was described. This patient also GVHD in patients who received marrow transplantations for AA.
had very short telomeres. Identification of further families • DC is associated with an increased risk for MDS, AML, and
with biallelic ACD mutations will help to clarify on the pre- squamous cell carcinomas.
cise contribution of ACD as a disease gene. • DC is associated with genetic defects in telomere mainte-
nance. Very short telomere lengths are seen in these patients.
• Eleven DC genes (DKC1, TERC, TERT, NOP10, NHP2, TINF2,
Treatment USB1, TCAB1, CTC1, RTEL1, and ACD) have been identified, and
these account for ~65% of patients with DC.
BM failure is the main cause of premature mortality in DC.
• The clinical presentation can range from AA alone to severe
Use of anabolic steroids (oxymetholone and danazol)
forms, such as HH and Revesz syndromes.
can produce improvement in the hematopoietic function.

Shwachman-Diamond syndrome Ribosomal DNA
Clinical features
Shwachman and Bodian and their colleagues reported this

45S rRNA
disease independently in 1964. It is now recognized as an
autosomal recessive disorder characterized by exocrine pan-
creatic insufficiency (100%), BM dysfunction (100%), and 30S 32S
other somatic abnormalities (particularly involving the skel-
etal system). Signs of pancreatic insufficiency (malabsorp-
tion, failure to thrive) are apparent early in infancy
(pancreatic function can improve in a subset of patients with 18S 5.8S 28S 5S
Shwachman-Diamond syndrome [SDS] by 5 years of age). Nucleus
Other common somatic abnormalities include short stature Cytoplasm

(~70%), protuberant abdomen, and an ichthyotic skin rash
(~60%). Metaphyseal dysostosis is seen on radiographs with
the localization and severity varying with age. Other abnor-
malities include hepatomegaly, rib or thoracic cage abnor- 40S subunit 80S ribosome 60S subunit
malities, hypertelorism, syndactyly, cleft palate, dental
dysplasia, ptosis, and skin pigmentation. DBA: RPS7, RP10, RPS17, RPS19, RPS24, RPS26, RPS28, RPS29
The spectrum of hematological abnormalities includes neu- DBA: RPL5, RPL11, RPL26, RPL35A
5q-syndrome: RPS14
tropenia (~60%), other cytopenias (~20% have pancytopenia),
SDS: SBDS
myelodysplasia, and leukemic transformation (~25%). The age
at which leukemia develops varies widely from 1 to 43 years.
Figure 15-5 Ribosome biogenesis. Schematic showing scheme of
AML is the commonest category, and there is an unexplained
rRNA processing in human cells and the points at which this possibly
preponderance of cases of leukemia in males (M:F ratio ~3:1). is disrupted in the different bone marrow failure syndromes. The
Exocrine pancreatic insufficiency and hematological abnor- ribosomal RNAs (rRNAs) are transcribed by RNA polymerase I as a
malities are also seen in Pearson syndrome (PS), and this is single precursor transcript (45S rRNA). The 45S rRNA is then
therefore an important differential diagnosis. PS is discussed processed to 18S, 5.8S, and 28S rRNAs. The 18S is a component of the
further below. Other differential diagnoses to be excluded are 40S ribosomal subunit. The 5.8S and 28S together with 5S
cartilage hair hypoplasia syndrome and cystic fibrosis. (synthesized independently) are components of the 60S ribosomal
subunit. The 40S and 60S subunits are assembled to form the 80S
ribosomes. The processing steps affected in Shwachman-Diamond
Pathophysiology syndrome (biallelic mutations in SBDS), Diamond-Blackfan anemia
(due to heterozygous mutations in RPS7, RPS10, RPS17, RPS19,
The SDS gene (SBDS) on 7q11 was identified in 2003. The
RPS24, RPS26, RPS28, RPS29, RPL5, RPL11, RPL15, RPL26, and
majority (95%) of patients with SDS have been found to
RPL35A) and 5q- syndrome (haploinsufficiency of RPS14) are
have biallelic mutations in this gene. Recent data from differ-
indicated by the different colored stars. DBA = Diamond-Blackfan
ent sources provides compelling evidence that the SBDS pro- anemia; SDS = Shwachman-Diamond syndrome.
tein has an important role in the maturation of the 60S
ribosomal subunit (Figure 15-5). SDS therefore can be Recent analysis of patients with SDS has showed that the
regarded as a disorder of ribosome biogenesis, with hemato- incidence of myelodysplasia and transformation to AML
poietic failure similar in some respects to Diamond-Blackfan (~15%-25%) is higher than reported previously. Isochromo-
anemia and myelodysplastic syndrome with chromosome 5q some 7q and del(20q) are very frequent in SDS, but do not
deletion and haploinsufficiency for RPS14. imply a poor prognosis by themselves, as they may be stable
or decline over time and not progress to MDS. The develop-
ment of leukemia, often with features of myelodysplasia,
Treatment
usually has a poor prognosis. SDS patients with leukemia
The malabsorption in SDS responds to treatment with oral treated with conventional courses of chemotherapy usually
pancreatic enzymes. For those with neutropenia, G-CSF may fail to regenerate normal hematopoiesis. As this is a constitu-
produce an improvement in the neutrophil count. As in tional disorder, all somatic cells, including HSCs, are abnor-
other cases of BM failure, supportive treatment with red cell mal. In addition, the hematopoietic stem cells may have
and platelet transfusions and antibiotics is very important. accumulated secondary abnormalities as suggested by com-
The main causes of death are infection or bleeding. plex karyotypes (especially involving chromosome 7) often

observed in the BM from such patients. Therefore, for those cells contain many mitochondria, normal and mutant mito-
who develop leukemia, the only approach likely to be suc- chondria can coexist in the same cell; the proportion of
cessful is allogeneic SCT using low-intensity conditioning which can vary from tissue to tissue (known as hetero-
regimens that include fludarabine. The similarities between plasmy). The extent to which different tissues are affected
SDS and the other common inherited BM failure syndromes depends in part on this proportion. This explains why differ-
emphasize that SDS should be regarded as a disorder with ent tissues may be affected to varying degrees in the same
high propensity to develop both AA and leukemic transfor- patient.
mation, particularly AML with erythroid differentiation
(formerly known as AML-M6). As these complications may
Key points
not develop until adult life, it is important to continue close
hematological follow-up throughout life. Nonhematological • PS is a multiorgan mitochondrial DNA disorder characteristi-
malignancies have not been observed in patients with SDS. cally associated with marrow and pancreas dysfunction.
Treatment is largely supportive, and many patients die in
infancy.
Key points
• SDS is a rare autosomal recessive disorder characterized by BM
failure and exocrine pancreatic insufficiency. Diamond-Blackfan anemia
• The majority of patients with SDS have biallelic mutations in Clinical features
the SBDS gene, which has an important role in ribosome
DBA classically presents within the first year of life with a
biogenesis.
• Like other BM failure syndromes, patients with SDS have a
hypoproliferative, macrocytic anemia. BM examination typ-
high risk of developing MDS and leukemia. ically reveals a paucity of erythroid precursors. Approxi-
• Patients with isolated neutropenia can be treated with G-CSF; mately 50% of patients have associated nonhematologic
those developing more global BM failure can be treated with physical findings, which may include thumb anomalies, mid-
HSCT. line craniofacial defects or cleft palate, urogenital abnormali-
ties, or cardiac defects. Red blood cell adenosine deaminase
levels and fetal hemoglobin levels usually are elevated, and
Pearson syndrome
this can help in diagnosis. An increased risk of AML has been
PS, sometimes called Pearson marrow-pancreas syndrome, is reported in patients with DBA, but the risk is less than in
a multisystem mitochondrial disease originally described as a patients with FA.
disorder with sideroblastic anemia, vacuolization of hemato-
poietic precursors, and exocrine pancreas dysfunction. Other
Pathophysiology
features include hypoplastic macrocytic anemia alone or asso-
ciated with thrombocytopenia or granulocytopenia, proximal Heterozygous germ line mutations in the RPS19 gene, which
tubular insufficiency, failure to thrive, endocrine pancreatic encodes a ribosomal protein, have been reported in 25% of
insufficiency with insulin-dependent diabetes, lactic acidosis, patients with DBA. The clinical manifestations among family
hyperlipidemia with liver steatosis, muscle and neurologic members with identical RPS19 gene mutations are highly
impairment, and, frequently, early death (usually in infancy). variable, and anemia may be absent. It is not clear how a
Treatment is supportive. The few patients who survive into defect in one allele of a gene encoding a ribosomal protein
adulthood often develop neurological symptoms of Kearns- leads to red blood cell hypoplasia and not to other dramatic
Sayre syndrome. phenotypic manifestations, because ribosomes are essential
The BM cellularity can be variable and all cell lineages may for cellular protein synthesis broadly. Intact ribosomes
be affected; the anemia is usually macrocytic with prominent appear to be particularly important for normal erythropoie-
vacuoles in cells of both erythroid and myeloid lineages. sis; in patients with MDS associated with the deletion of
Ring sideroblasts can be identified in a subset of patients chromosome 5q, acquired haploinsufficiency of RPS14
with PS. Hematopoietic dysfunction associated with periph- (a gene at 5q31 that encodes another ribosomal component)
eral cytopenias is a major cause of morbidity and mortality. contributes to disease-associated anemia.
This syndrome was first described by Pearson and col- Germ line mutations in 12 other genes encoding ribo-
leagues in 1979, and the mitochondrial DNA deletions caus- some-associated proteins (RPS7, RPS10, RPS17, RPS24,
ing it were discovered a decade later in 1989. The PS-associated RPS26, RPS28, RPS29, RPL5, RPL11, RPL15, RPL26, and
mitochondrial DNA deletions can vary in size from patient RPL35A,) have been described in families with DBA who
to patient. These lead to loss of mitochondrial function have wild-type RPS19. Although the overall inheritance pat-
(defective oxidative phosphorylation) to varying degrees. As tern is autosomal dominant, many cases of DBA appear to be

sporadic, presumably resulting from new germ line muta- conditions, such as hemoglobinopathies, hereditary sidero-
tions. Recently, germ line mutations in the X-linked GATA1 blastic anemias, GATA1 mutations, and MDS, and these
gene have been identified in some male patients with “DBA- should be ruled out.
like” disease.
CDA type I
Treatment
CDA I is a rare autosomal recessive disorder (~80 pedigrees
In the majority of patients with DBA (~80%), the hemoglo- are listed in the pan-European registry) that usually presents
bin level improves with corticosteroid treatment. It is vital to in childhood or adolescence. CDA is characterized by hemo-
use the minimal dose of steroids required to support erythro- lytic anemia (usually moderate, with hemoglobin in the
poiesis to minimize adverse effects of chronic steroid use. range of 9-10 g/dL), anisopoikilocytosis with reduced levels
Patients who fail to respond to steroids, or who require high of erythrocyte membrane protein 4.1R, normal or elevated
steroid doses, may be supported with red blood cell transfu- reticulocyte count, macrocytosis, and high serum iron lev-
sions instead. Careful attention to iron overload and timely els. Serum bilirubin levels often are elevated, and some
initiation of iron chelation therapy are important for patients patients develop pigment gallstones or jaundice. Spleno-
undergoing chronic transfusions. Currently, the only curative megaly is a frequent feature. Some patients have dysmor-
treatment of marrow failure in DBA is stem cell transplanta- phologic features, including syndactyly, absent or
tion, but the risks of transplantation must be weighed against hypoplastic distal phalanges or nails, skin pigmentation
the benefits for each patient. Some patients with DBA will abnormalities, hypoplastic right third rib, and sensorineural
undergo spontaneous remission and maintain adequate deafness. BM examination shows erythroid hyperplasia,
hemoglobin levels independent of steroids. These outcomes, binucleated erythroblasts, and a distinctive pattern of inter-
however, may not be permanent, with some patients in remis- nuclear chromatin bridging.
sion experiencing relapse of their anemia and some patients CDA I has been linked to defects in the gene codanin
maintained stably on steroids becoming steroid refractory. 1 (CDAN1) at chromosome 15q15, a gene encoding a protein of
unknown function and without well-defined domains. Recently,
the second disease gene mutated in CDA I was identified as
Key points C15orf41. For unclear reasons, the anemia in type I CDA typi-
• DBA typically presents with macrocytic anemia and red blood cally responds to recombinant interferon-α. Folate supplemen-
cell hypoplasia in infancy. tation is helpful, given the chronic hemolysis. Most patients with
• Approximately 50% of patients with DBA have physical signs, type I CDA typically do not require transfusions, and transfu-
including radial ray and craniofacial abnormalities. sions can exacerbate the tendency to iron overload.
• Autosomal dominant mutations in 13 different ribosomal
genes account for ~65% of DBA patients; rarely due to hemizy-
CDA type II
gous mutations in GATA1.
• Treatment options for DBA include corticosteroids, red blood CDA II is more common than CDA I (~350 patients have been
cell transfusion support, and hematopoietic stem cell collected in European registries) and usually presents during
transplantation.
childhood with anemia of variable severity. Transfusion depen-
• Spontaneous remissions may occur in a subset of patients.
dence is uncommon. The reticulocyte count is low, and the BM
typically shows multinucleated erythroid precursors, karyor-
Congenital dyserythropoietic anemia rhexis, and pseudo-Gaucher cells. The red blood cell membrane
in patients with this disorder demonstrates abnormal glycosyl-
General clinical features
ation, apparently because of a defect in Golgi processing in
The congenital dyserythropoietic anemias (CDAs) are a het- erythroblasts. Abnormal migration of band 3 and band 4.5 on
erogeneous group of conditions characterized by ineffective sodium dodecyl sulfate gels may be useful diagnostically. CDA II
erythropoiesis and anemia, multinucleated erythroid pre- is an autosomal recessive disorder and, in 2009, was found to be
cursors in the marrow, and excess iron even in the absence of due to biallelic mutations in the gene encoding the secretory
blood transfusions. Beyond these similarities, the subtypes COPII component SEC23B.
of CDA have differing clinical features and modes of inheri- CDA II formerly was known as HEMPAS (hereditary eryth-
tance. Two types of CDA (CDA I and the more common roblastic multinuclearity with a positive acidified serum test).
CDA II) are fairly well defined, CDA III is rare, and the other A characteristic feature of HEMPAS is the ability of some
forms of CDA are very rare and poorly characterized. The group-compatible sera to lyse the patient’s erythrocytes, result-
differential diagnosis of dyserythropoiesis includes other ing in a positive acid hemolysis test (Ham test). The sucrose

hemolysis test is negative, however, and the cells are not lysed myelopoiesis that first present in infancy, usually without
by autologous serum, unlike the situation in paroxysmal noc- other physical abnormalities, which was first described in
turnal hemoglobinuria (PNH). The cell lysis is secondary to 1954. Although the original description by Kostmann was of
increased immunoglobulin M (IgM) binding to erythrocytes an autosomal recessive disorder, other congenital neutrope-
and not to increased sensitivity to complement. nia subtypes (both sporadic and autosomal dominant) have
Like other patients with congenital dyserythropoiesis, been included subsequently in this category. The neutrope-
patients with CDA II can have problems with iron overload. nia usually is recognized at birth, and the neutrophil count is
Phlebotomy and iron chelation have been used to treat iron often <0.2 × 109/L. The hemoglobin (Hb) and platelet counts
overload in these patients. Because the osmotic fragility test are usually normal, and the BM shows maturation arrest of
is usually abnormal in CDA II, some patients are misdiag- myelopoiesis at the level of the promyelocyte/myelocyte
nosed as having hereditary spherocytosis and undergo sple- state (with abundant promyelocytes but with a selective
nectomy. Splenectomy may be useful in treating anemia in reduction in myelocytes, metamyelocytes and neutrophils).
some patients, but results are variable. Severe congenital neutropenias are associated with severe
infections and early death. No patient has developed AA, but
myeloid leukemias (~25% by 25 years) can occur. The avail-
Other CDA types
ability of G-CSF has revolutionized the outcomes of these
CDA III is a rare autosomal dominant disorder characterized children. Somatic mutations in the gene that encodes the
by the presence of multinucleated erythroid precursors in G-CSF receptor have been documented during the evolution
the marrow (gigantoblasts) in addition to mild anemia and to leukemia in patients receiving G-CSF. Leukemic transfor-
low reticulocyte counts. Heterozygous mutations in KIF23 mation occurred in patients with congenital neutropenia
have been identified in patients with CDA III. The peripheral prior to the availability of G-CSF, and the precise contribu-
smear shows marked anisopoikilocytosis and basophilic tion of G-CSF therapy to the development of G-CSF recep-
stippling of the red blood cells, a picture similar to tor mutations remains unclear. For patients who become
β-thalassemia major. Additionally, exceptionally rare types refractory to G-CSF or who develop leukemia, SCT may be
of CDA (eg, CDA IV, V, VI, and VII) with variable modes of appropriate and curative.
inheritance have been proposed. Recently, a unique muta- Cyclical neutropenia is characterized by a neutrophil
tion in KLF1, which encodes the erythroid transcription fac- count that usually reaches a nadir with a 21-day periodicity.
tor KLF1, was identified; this causes major ultrastructural Around the nadir, patients may develop fever and mouth
abnormalities (a hitherto unnamed CDA), the persistence of ulcers. In cyclical neutropenia, the pattern of inheritance is
embryonic and fetal hemoglobins, and the absence of some usually autosomal dominant. Linkage analysis in affected
red cell membrane proteins. CDA is a feature of the autoin- families resulted in the localization of the disease gene to
flammatory disease Majeed syndrome, which is further in 19p13.3. Subsequent studies identified mutations in the
the section “Autoinflammatory Diseases” in this chapter. gene ELANE encoding neutrophil elastase (NE). An
extraordinary twist was the identification of ELANE muta-
tions in many patients who also had severe congenital neu-
Key points tropenia (SCN). In cyclical neutropenia, the mutations
• CDA I is characterized by moderate hemolytic anemia,
usually are clustered around the active site of the molecule,
internuclear chromatin bridging, iron overload, germ line whereas the opposite face of the molecule tends to be
biallelic mutations in CDAN1 or c15orf41, and responsiveness to mutated in congenital neutropenia patients. NE is a serine
interferon-α therapy. protease that is synthesized predominantly at the promy-
• CDA II is the most common form of CDA and can be elocytic stage and can be expected to be important in neu-
misdiagnosed as hereditary spherocytosis. Patients typically have trophil development. Recent studies suggest that ELANE
a positive acid hemolysis (Ham) test, multinucleated giant cells, mutations lead to accumulation of the nonfunctional pro-
and a low reticulocyte count. The genetic defect in this subtype tein, which in turn triggers an unfolded protein response
is in the SEC23B gene. (UPR) leading to maturational arrest. The precise mecha-
• Several rare forms of CDA have also been genetically charac-
nism leading to maturation arrest of promyelocytes remains
terized including those with heterozygous mutations in KIF23.
unclear. The original family described by Kostmann had
autosomal recessive SCN and recently has been shown to be
associated with biallelic mutations in the HAX1 gene pre-
Congenital and cyclical neutropenia
dicted to lead to defects in cell death. Biallelic mutations in
Congenital neutropenia is a heterogeneous disorder. It HAX1 account for ~10% of congenital neutropenia. The
includes Kostmann syndrome, a group of disorders of HAX1 protein is a critical regulator of the mitochondrial

membrane potential and cellular viability. Although data receptor for SDF1 (CXCL12). Evidence is convincing that
suggest HAX1 is important in controlling apoptosis, it is SDF1/CXCR4 signaling is a key regulator of neutrophil
unclear why premature death of neutrophils consistently is release from the BM. Specifically, SDF1/CXCR4 provides a
associated with HAX1 deficiency. Mutations in JAGN1 and signal to retain neutrophils in the BM. The mutations of
other genes (GFI1, G6PC3, WASP) are also associated with CXCR4 in WHIM syndrome truncate the carboxyl-termi-
SCN, demonstrating genetic heterogeneity. Patients with nal (cytoplasmic) tail, resulting in enhanced CXCR4 sig-
biallelic G6PC3 mutations may present nonsyndromically naling. These observations support the current model in
or syndromically, with additional extrahematologic and which enhanced signaling by the CXCR4 mutants in WHIM
nonmyeloid cell features, including congenital heart and syndrome leads to abnormal neutrophil retention in the
urogenital defects, prominent superficial veins, primary BM. Given that many of the clinical features affecting
pulmonary hypertension, severe lymphopenia, and thymic patients with WHIM are a consequence of hyperfunction
hypoplasia. of CXCR4, inhibitors of CXCR4 function, such as plerixa-
for, are being investigated in clinical trials with some thera-
peutic promise.
Key points
• SCN is a heterogeneous disorder, and several disease genes
(ELANE, HAX1, GFI1, G6PC3, JAGN1) have been identified to Key points
date. • WHIM syndrome is an inherited immune deficiency character-
• G-CSF has become an important therapeutic agent for these ized by susceptibility to HPV infection-induced warts, neutrope-
patients. Patients who become refractory to G-CSF or progress nia (associated with BM myelokathexis), B-cell lymphopenia, and
to leukemia can be treated by HSCT. hypogammaglobulinemia-related recurrent infections.
• The syndrome is due to heterozygous mutations in the CXCR4
gene, resulting in functional overactivity of CXCR4.
WHIM syndrome • Clinical management includes therapy with G-CSF, IVIg,
prophylactic antibiotics, and surveillance or surgical removal of
WHIM (warts, hypogammaglobulinemia, infections, and
dysplastic skin or mucosal HPV-related lesions.
myelokathexis) syndrome is a rare autosomal dominant dis-
order characterized by neutropenia, B-cell lymphopenia,
hypogammaglobulinemia, and extensive human papilloma-
Thrombocytopenia with absent radii
virus (HPV) infection. Affected individuals typically present
with recurrent bacterial (eg, Haemophilus influenzae, Strep Thrombocytopenia with absent radii (TAR) is an autosomal-
tococcus pneumoniae, Klebsiella pneumoniae, and Staphylo recessive disorder characterized by hypomegakaryocytic
coccus aureus) infections from birth with ANCs of <1,000/µL. thrombocytopenia and bilateral radial aplasia. The presence
Recurrent pneumonias may in some cases lead to severe of normal thumbs distinguishes TAR from other syndromes
bronchiectasis. Despite the peripheral neutropenia, the BM that may manifest with blood count abnormalities and radial
of affected patients is generally hypercellular with increased aplasia, such as FA. Babies with TAR often have hemorrhagic
numbers of mature neutrophils (a condition termed myelo manifestations at birth when the diagnosis usually is made,
kathexis). Patients commonly have B-cell lymphopenia, yet owing to the characteristic physical appearance combined
immunity to most viral pathogens is normal. The major with thrombocytopenia. Additional skeletal (absent ulnae,
exception is HPV, which is the cause of warts in patients with absent humeri, clinodactyly) and other somatic (microceph-
WHIM syndrome. Although some patients have few, if any, aly, hypertelorism, strabismus, heart defects) abnormalities
warts, the majority of patients suffer from extensive verruco- may be seen in some patients.
sis. They typically appear in the first or second decades of life The platelet count is usually <50 × 109/L. The leukocyte
and can involve any mucocutaneous surface. Hypogamma- count can be normal or raised sometimes up to 100 × 109/L
globulinemia is variable, ranging from normal to modestly (leukemoid reaction). BM cellularity is normal, and
decreased serum immunoglobulin (Ig) G, IgM, and IgA. myeloid and erythroid lineages are normal or increased.
Treatment with G-CSF or GM-CSF is effective in correcting Megakaryocytes are absent or decreased. Most patients
the neutropenia; those with significant hypogammaglobu- bleed in infancy and then improve after the first year. The
linemia benefit from intravenous immunoglobulin (IVIg) mainstay of management is prophylactic and therapeutic use
therapy every 6 weeks. of platelet transfusions. Patients with TAR have a good
The majority of patients with WHIM syndrome have prognosis after infancy. There have been no reports of AA,
heterozygous mutations of the CXCR4 gene. CXCR4 is a G but four of leukemia. The leukemia cases were described
protein-coupled heptahelical receptor that is the major prior to molecular testing for TAR and included both

myeloid and lymphoid leukemias. Whether TAR is associated

with an increased risk of hematologic malignancy remains Key points
unknown. • CAMT is caused by biallelic mutations in the MPL gene
In patients with TAR, thrombopoietin levels are usually encoding the TPO receptor.
elevated and thrombopoietin receptor expression on the sur- • CAMT presents in the neonatal period with bruising or
face of TAR platelets is normal. Therefore, defective mega- bleeding and severe thrombocytopenia; pancytopenia may
karyocytopoiesis or thrombocytopoiesis does not appear to develop in later childhood.
be caused by a defect in thrombopoietin production. There is • CAMT may be treated with HSCT.
some evidence that it may be due to a lack of response to

thrombopoietin in the signal transduction pathway of the GATA2 deficiency
thrombopoietin receptor (MPL). Recently, it has been deter-
mined that compound inheritance of low-frequency regula- Germ line heterozygous mutations in GATA2, which encodes a
tory SNP and a rare null mutation in the RBM8A gene zinc finger transcription factor that regulates hematopoiesis and
(which encodes a subunit of the exon-junction complex) lymphatic development, explain four MDS-predisposing condi-
reduces protein Y14 production and by unclear mechanisms tions: monocytopenia and mycobacterial infection (MonoMAC
causes TAR. syndrome); dendritic cell, monocyte, B and NK cell lymphoid
deficiency (DCML); primary lymphedema with myelodysplasia
progressing to AML (Emberger syndrome); and a subset of
Key points familial MDS. Some patients with GATA2 mutations present
with congenital neutropenia or AA. GATA2 deficiency should
• TAR is characterized by isolated thrombocytopenia and
bilateral radial aplasia.
be suspected in patients with peripheral blood monocytopenia
• The mainstay of management is platelet transfusions, and and B and NK lymphopenia in the context of less pronounced
patients usually have a very good prognosis. or absent neutropenia, anemia, and thrombocytopenia. Similar
• It was recently established that TAR is due to biallelic muta- to other inherited BM failure syndromes, the BM in GATA2
tions in the RBM8A gene, which encodes a subunit of the deficient patients is typically hypocellular for age. Additional
exon-junction complex. findings specifically suggestive of GATA2 deficiency include
increased megakaryocytes with atypical/dysplastic features,
including micromegakaryocytes, nuclear hypolobation, and
Congenital amegakaryocytic large megakaryocytes with separated nuclear lobes localized to
thrombocytopenia the periphery, and abnormal cytogenetics, most often mono-
Congenital amegakaryocytic thrombocytopenia (CAMT) is somy 7 and trisomy 8. HPV and nontuberculous Mycobacte
a rare autosomal recessive disorder characterized by amega- rium infections are common. Pulmonary alveolar proteinosis,
karyocytic thrombocytopenia. Megakaryocytes are absent or recurrent upper respiratory tract infection, panniculitis, lymph-
greatly diminished in the BM. Patients typically present edema, thrombosis, and hearing loss can be observed in 20% or
shortly after birth with petechiae, bruising or bleeding, and a more patients with GATA2 disease. AML or MDS is present
very low platelet count. Patients with CAMT frequently in 30%-50% of patients at presentation and the lifetime risk
progress to pancytopenia associated with BM aplasia (AA). is 90%.
CAMT is not associated with specific congenital anomalies; In addition to monitoring for infections and other nonhe-
however, patients with CAMT and common somatic mal- matological disease complications, serial blood count and
formations have been reported. There are also case reports of annual BM evaluations are recommended. Development of
patients with CAMT developing MDS and leukemia. MDS or AML, which often presents with high-risk cytoge-
CAMT is caused by biallelic mutations in the MPL gene, netic features, is an indication for HSC transplantation.
which encodes the thrombopoietin (TPO) receptor. Devel- Genetic counseling should be provided to family members,
opment of AA in patients with CAMT is consistent with and this should be conducted prior to family screening for
findings that TPO signaling plays an important role in the HSC donors.
maintenance and expansion of HSCs and multipotent pro-
genitors. TPO levels are typically high in CAMT. Chédiak-Higashi syndrome
Supportive care consists largely of platelet transfu-
sions. Antifibrinolytic agents may be useful to help treat Chédiak-Higashi syndrome (CHS) is a rare inherited syndrome
bleeding. Many patients progress to aplasia, and this can characterized by severe immunodeficiency, partial albinism, a
be cured with HSCT. The platelet count is not responsive mild bleeding diathesis, and progressive neurologic defects.
to TPO. The pathognomonic feature of CHS is the presence of giant

inclusion bodies in virtually all granulated cells, particularly are infrequent and may reflect the ability of patients to
neutrophils (Figure 15-2). Neutropenia is common, and the increase their neutrophil counts during acute illnesses. BM
residual neutrophils display functional defects. Approximately examination rarely is indicated. When performed, it reveals
85% of patients will progress to an accelerated phase minimal abnormalities or only a deficit of mature neutro-
characterized by a nonclonal lymphohistiocytic infiltration of phils. Spontaneous remission occurs in most patients. In
multiple organs, leading to multiorgan system failure. CHS is the largest series, 95% of 240 patients with autoimmune
inherited in an autosomal recessive fashion and is due to neutropenia had a spontaneous remission, usually within
mutations of the LYST gene. The loss of LYST protein disrupts 24 months of diagnosis. The pathogenesis is thought to be
vesicular trafficking, leading to impaired formation of secretory immune-mediated neutrophil clearance. Indeed, in the
lysozymes and resulting in hypopigmentation and dysre great majority of published cases, antibodies to neutrophil
gulated immune cell function. Of note, there are several other antigens can be detected, although multiple blood samples
rare neutropenic disorders in which impaired vesicular may need to be analyzed. Antibodies directed against
trafficking appears to be the primary mechanisms of disease FCgRIII or CDllb/CD18 (also termed the type 3 comple-
pathogenesis (Tables 15-3 and 15-4). Other than allogeneic ment receptor) are the most common antibodies detected.
BM transplantation, treatment of CHS and related syndromes Many patients remain free of infections, and no specific
is largely supportive. therapy is required. For patients with recurrent infections,
prophylactic antibiotics or intermittent treatment with
G-CSF may be indicated. Although rarely needed, high-
Acquired neutropenia
dose IVIg or corticosteroid therapy have been reported to
be effective.
Neonatal alloimmune neutropenia
Physiologic and acquired neutropenia in premature infants,

Secondary autoimmune neutropenia
including neutropenia as a result of idiopathic or immune
causes, is much more common than inherited or congenital Neutropenia occasionally is associated with autoimmune
neutropenia. In alloimmune neonatal neutropenia, a trans disease, most commonly rheumatoid arthritis (RA), systemic
placental transfer of maternal IgG is directed against paternal lupus erythematosus (SLE), and Sjögren syndrome. More-
antigens expressed on neonatal neutrophils. This is analogous over, there is a strong association of neutropenia with large
to neonatal erythroblastosis secondary to Rh incompatibility. granular lymphocytic leukemia (LGL) often in association
Rarely, the maternal antibody is secondary to autoimmune with RA. In SLE, neutropenia occurs in ~50% of patients.
neutropenia in the mother. The incidence has been esti The neutropenia is generally mild, has little impact on dis-
mated at 2 per 1,000 live births. The BM usually shows ease, and requires no specific treatment. The pathogenesis of
normal cellularity with a late myeloid arrest. Maternal neutropenia in SLE is thought to be related to accelerated
alloimmunization probably occurs during the first trimester apoptosis of mature neutrophils. Although neutrophil anti-
of pregnancy. Neutrophil-specific antibodies to HNA-1a, bodies have been implicated but are present in some patients
HNA1b, and HNA-2a can be detected in more than half of with SLE without neutropenia, the clinical utility of measur-
cases. These children may develop omphalitis or skin ing antineutrophil antibodies in SLE is questionable. As with
infections; however, they are also at risk of severe, life- SLE, the differential diagnosis for neutropenia in RA is wide,
threatening infections. Aggressive antibiotic therapy must be and drug-induced neutropenia must be considered. Felty
given for documented infection, and recombinant G-CSF syndrome is the triad of unexplained neutropenia, long-
should be considered, although the response is variable and standing RA, and variable splenomegaly. There is an
unpredictable. With supportive care, the neutropenia usually increased risk of infections in these patients. Treatment usu-
spontaneously resolves within 3-28 weeks (average of 11 ally is directed at the underlying RA.
weeks). Overall, the prognosis for infants with alloimmune
neutropenia syndrome is favorable.
Nonimmune chronic idiopathic neutropenia
In a subset of patients with chronic neutropenia, there is no

Primary autoimmune neutropenia
evidence of immune-mediated disease. The diagnosis of
Primary autoimmune neutropenia (also termed chronic nonimmune chronic idiopathic neutropenia in adults
benign neutropenia of infancy and childhood) typically (NI-CINA) is based on the presence of chronic acquired
occurs in children between the ages of 5 and 15 months but neutropenia in the absence of underlying autoimmune dis-
can be present from 1 month through adulthood. The ANC ease, cytogenetic abnormality, antineutrophil antibodies, or
is typically between 500 and 1,000/µL. Serious infections other obvious explanation for neutropenia. In addition to

Acquired neutropenia | 417
neutropenia, lymphopenia, monocytopenia, anemia, and Table 15-5 Selected drugs associated with neutropenia
thrombocytopenia occasionally are seen. BM findings are Anti-inflammatory agents Antimicrobial agents
highly variable, with both hyperplastic and hypoplastic BM Aminopyrine* Ampicillin*
cellularity reported. The pathogenesis of NI-CINA is poorly Diclofenac* Cefotazime*
understood, although it has been suggested that chronic low- Diflunisal* Cefuroxime*
grade inflammation may contribute. It is most common in Dipyrone* Flucytosine*
Ibuprofen* Fusidic acid*
young and middle-aged women. Fortunately, the clinical
Gold salts Imipenem-cilastatin*
course usually is benign, infections are infrequent, and spe- Penicillamine Nafcillin*
cific treatment is not required. Phenylbutazone Oxacillin*
Sulfasalazine Quinine*
Ticarcillin*
Drug-induced neutropenia Cardiovascular agents Chloramphenicol
Clopidogrel* Sulfonamides
Clinical case Disopyramide* Amodiaquine
Methyldopa* Dapsone
A 50-year-old teacher with a history of ulcerative colitis for 1 year Procainamide* Terbinafine
presents to the emergency room with fever, chills, and sore throat Quinidine* Vancomycin
for 24 hours. On examination, the temperature is 39.6°C, blood Ramipril*
pressure 90/60, pulse 105, and respiratory rate 28. The patient is Spironolactone* Antithyroid agents
confused and reports that her throat is very sore. The abdomen is Dipyridamole Propylthiouracil*
slightly tender and bowel sounds are absent. On the basis of the Captopril Carbimazole
patient’s presentation, therapeutic measures for septic shock are Ticlopidine Methimazole
initiated. Within a few minutes, a complete blood count (CBC)
Anticonvulsants Other agents
reveals a white blood cell (WBC) count of 1.5 × 109/L with an
Phenytoin* Amygdalin*
ANC of 0. On questioning the patient’s husband, you learn that
Carbamazepine Calcium dobesilate*
she had been in her usual state of health until a few days ago.
Cimetidine*
She has had long-standing complaints of chronic diarrhea with
Psychotropic agents Infliximab*
intermittent blood and mucous in the stool. Her only medication
Chlorprmazine* Levamisole*
is sulfasalazine begun about 3 months ago. Clozapine* Metoclopramide*
Fluoxetine* Mebhydrolin*
Mianserin Rituximab
Drug-induced neutropenia and agranulocytosis are serious Ranitidine
medical problems, with an estimated frequency of 1 per Hypoglycemic agents Famotidine
1.6-7 million and a case fatality rate of approximately 5%. Chlorpropamide Metiamide
Agranulocytosis refers to the complete absence of neutro- Tolbutamide
phils in the blood. Although certain medications carry a Glyburide*
higher risk of neutropenia (Table 15-5), it is probably *Level I evidence based on Andersohn F et al. Ann Intern Med.
wise to consider most medications as potential offenders, 2007;146:657-665.
thus emphasizing the need for a careful drug history in
all patients who present with acquired neutropenia. A
recent systematic review of the literature identified 10 mechanism often is unclear. In some well-studied cases, the
drugs that accounted for ~50% of cases of definite or prob- offending drug serves as a hapten in association with an endog-
able reports of drug-induced neutropenia: carbimazole, enous protein, probably an antigen expressed on the neutrophil
clozapine, dapsone, dipyrone, methimazole, penicillin G, surface. The immune response to this complex results in neu-
procainamide, propylthiouracil, rituximab, sulfasalazine, trophil destruction, severe neutropenia, and susceptibility to
and ticlopidine. infection. Other drugs may impair production of neutrophils by
In most cases, agranulocytosis presents within 6 months, and a direct, toxic effect on myeloid precursors.
usually within 3 months, after starting the offending drug. The Drug-induced agranulocytosis is difficult to anticipate.
clinical presentation of drug-induced agranulocytosis is often Serial blood counts are now recommended for patients on
less dramatic than in this case, but patients often have fever and some drugs (ie, sulfasalazine, clozapine, phenothiazines, and
pharyngitis as their first symptoms. Sepsis or pneumonia may antithyroid drugs) because of the relatively high frequency of
occur in 10%-30% of patients. Usually the prognosis is good drug-induced neutropenia associated with these agents.
because neutrophil counts recover within approximately Practices are not standardized, and the benefit of frequent
1 week if the offending medication is withdrawn. The disease blood counts is not established.

Management includes prompt withdrawal of all potentially Because recurrent fevers, otitis media, and sinopulmonary
offending drugs and administration of broad-spectrum anti- infections are common in young children, it may be difficult
biotics, usually with inpatient management. The mean time to to assess when a child has had “too many” infections and
recovery is ~10 days, but the duration of neutropenia is highly requires a careful workup. Certain conditions, however,
variable. Therapy with hematopoietic growth factors, particu- should raise concern for an underlying immunodeficiency
larly G-CSF, is controversial. A number of nonrandomized tri- syndrome, including neutrophil function disorders, and
als have reported a shortened duration of neutropenia, less may merit further evaluation. These include the following:
antibiotic use, and reduced hospital stay with the use of G-CSF. (i) recurrent systemic bacterial infections (eg, sepsis,
BM examination usually is not necessary in cases with other- osteomyelitis, meningitis); (ii) infections at unusual sites (eg,
wise-normal hemoglobin, platelet count, and red blood cell hepatic or brain abscess); (iii) recurrent bacterial infections
morphology. The time to hematologic recovery may be pro- (eg, pneumonia, sinusitis, cellulitis, lymphadenitis, draining
portional to the severity of the marrow defect; that is, if no otitis media); (iv) infections caused by unusual pathogens
cells at the myelocyte stage are seen on an aspirate sample, it (eg, Aspergillus pneumonia, disseminated candidiasis, Serr
probably will be several days before recovery occurs. Overall atia marcescens, Nocardia species, Burkholderia cepacia); and
survival is ~95%. A neutrophil count ≤0.1 × 109/L and the (v) chronic gingivitis or recurrent aphthous ulcers. In the
presence of sepsis or severe infection are associated with previous clinical case, the history of recurrent abscesses in
delayed neutrophil recovery and increased mortality. the setting of a normal ANC would merit further evaluation
Agranulocytosis has been associated with both cocaine for a neutrophil function disorder. The most widely known
and heroin use. In reported cases, it was caused by the adul- disorders with abnormalities of neutrophil function are
teration of the drug with levamisole, an antihelminthic drug described in the following sections.
used in veterinary medicine that is known to be associated
with agranulocytosis.
Myeloperoxidase deficiency
Myeloperoxidase (MPO) deficiency is the most common

Key points
disorder of phagocytes, with 1 in 4,000 individuals having a
• Transient neutropenia is commonly seen in infants and complete deficiency of MPO. It is inherited in an autosomal
children and may be due to infection, auto- or alloimmune recessive fashion and is due to mutations of the MPO gene.
mechanisms, unidentified causes (ie, idiopathic), or, less MPO is a primary granule enzyme that catalyzes the conver-
commonly, genetic disorders of granulopoiesis. sion of H2O2 to hypochlorous acid and other toxic interme-
• The genetic basis for many congenital neutropenia syndromes
diates that greatly enhance polymorphonuclear neutrophil
has been identified and genetic testing is becoming an impor-
microbicidal activity. The diagnosis can be made with histo-
tant diagnostic tool in the evaluation of patients with chronic
chemical assays for MPO on neutrophils. Of note, most
neutropenia.
• Neutropenia in adults is frequently due to drugs, both as a
patients (95%) with MPO deficiency are asymptomatic. An
predictable response to myelotoxic agents and as an idiosyn- increase in mucocutaneous infections with Candida strains
cratic reaction to almost any drug. Less commonly, neutropenia has been reported, particularly in patients with concurrent
in adults is due to infection, acquired hematopoietic disease, diabetes mellitus. There is no specific treatment.
autoimmune disorder, or a clonal proliferation of large granular
lymphocytes.
Leukocyte adhesion deficiency
Leukocyte adhesion deficiency (LAD) is a rare disorder man-

Inherited disorders of neutrophil ifested by delayed wound healing, recurrent bacterial infec-
function tions, and neutrophilia. There are three distinct forms of
LAD. In LAD-I, mutations of ITGB2, encoding the
Clinical case β2-integrin (CD18) chain, disrupt β2-integrin function. In
LAD-II, mutations of SLC35C1, encoding GDP-fucose
A 2-year-old boy has had recurrent furuncles and deep abscesses transporter-1, disrupt the generation of ligands on neutro-
since the first few months of life. On examination, there is no phils required for selectin binding. Patients with LAD-II also
active infection, but there are scars from drainage of previous
display short stature, abnormal facies, and severe cognitive
abscesses. CBC shows a hematocrit of 32%, WBC is 12 × 109/L,
impairment. Finally, in LAD-III, mutations of FERMT3
and the platelet count is 400 × 109/L. The differential count is
(KINDLIN3) or RASGRP2 lead to impaired β-integrin func-
normal, and the morphology of the leukocytes is normal. The
IgG level is increased; the levels of IgM and IgA are normal.
tion. In addition to immunodeficiency, patients with LAD-
III also have a bleeding diathesis due to a defect in β3-integrin

Inherited disorders of neutrophil function | 419
function on platelets. All mutations result in severely mutations of autosomal genes, including CYBA, which
impaired neutrophil chemotaxis and emigration from the encodes p22phox (the second membrane component of
blood to sites of infection. A history of consanguinity may be cytochrome b558) (5% of CGD cases); NCF-1, which
an important clue in evaluating children for LAD. Definitive encodes p47phox (20% of cases); NCF-2, which encodes
treatment of LAD requires allogeneic HSCT, with a recent p67phox (6% of cases); and NCF-4, which encodes p40phox
study reporting a 5-year survival of 75%. (<1%). The incidence is approximately 1 in 200,000 live
births.
The diagnosis of CGD is established by a typical clinical
Hyperimmunoglobulin E syndrome
history and a laboratory test demonstrating an abnormal
Hyperimmunoglobulin E syndrome (previously known as neutrophil oxidative burst. In the nitroblue tetrazolium
Job syndrome) is manifested by defective neutrophil chemo- (NBT) test, neutrophils are incubated with NBT. Patients
taxis, recurrent bacterial infections (typically involving the with CGD are unable to generate superoxide and fail to
skin, sinuses, or lung), mucocutaneous infections with Can reduce NBT to a blue formazan precipitate. In the dihy-
dida albicans, and elevated serum IgE levels. Patients may rorhodamine assay, oxidation of dihydrorhodamine to rho-
present with pruritic dermatitis in the first few weeks of life. damine by hydrogen peroxide produced by activated
Associated features that might aid in the diagnosis include neutrophils is measured by flow cytometry. Genetic testing
coarse facial features, recurrent fractures, and short stat- for both X-linked and autosomal recessive variations is also
ure. Heterozygous STAT3 mutations are identified in the available.
majority (60%-70%) of cases of hyperimmunoglobulin Treatment of CGD consists of prophylactic antibiotics,
E syndrome. In addition, mutations of DOCK8 (encoding antifungal agents, and the prompt administration of antibi-
dedicator of cytokinesis 8) are present in many cases of otics for specific infections. Chronic treatment with
the autosomal recessive form of hyperimmunoglobulin interferon-γ reduces the incidence of bacterial and fungal
E syndrome. Recently, homozygous mutations in PGM3 infections by ~70%. HSCT although curative, generally is
(encoding phosphoglucomutase 3), resulting in defective reserved for patients in whom the clinical course or specific
glycosylation, were identified in families with a hyperimmu- mutation portends a poor outcome.
noglobulin E syndrome-like phenotype. Developmental
delay, psychomotor retardation, and hypotonia were addi-
Autoinflammatory diseases
tional features of this novel hyperimmunoglobulin E syn-
drome subtype. Autoinflammatory diseases, also called periodic fever syn-
dromes, are a group of rare hereditary disorders characterized
by recurrent episodes of unprovoked inflammation in the
Chronic granulomatous disease
absence of infection. The most common and prototypical auto-
Chronic granulomatous disease (CGD) is a primary immu- inflammatory disease is familial Mediterranean fever (FMF).
nodeficiency syndrome caused by a defect in the NADPH FMF is characterized by sporadic paroxysmal attacks of fever,
oxidase system. In CGD, neutrophils and monocytes are serosal inflammation, and neutrophilia. The condition usually
unable to generate the respiratory burst that generates super- presents in early childhood with recurrent attacks of acute peri-
oxide, the precursor to hydrogen peroxide and other reactive tonitis, although pleuritis and synovitis also are frequent. These
oxygen derivatives with microbicidal activity. The disorder is attacks generally last 1-3 days and then resolve spontaneously.
characterized by recurrent bacterial and fungal infections FMF is inherited as an autosomal recessive disorder and mainly
affecting the skin, lungs, and bones with the development of occurs in populations from the Mediterranean basin, such as
granulomatous inflammatory responses in lymph nodes and Sephardic Jews, Arabs, Turks, and Armenians. Sporadic cases,
other tissue. The majority of cases are diagnosed in early however, also have been reported in individuals of French, Ital-
childhood. Infections are caused by catalase-positive organ- ian, Greek, Belgian, and Northern European heritage. Mutations
isms, with the majority of infections being caused by Asper in the MEFV gene, which encodes the protein, pyrin, appear to
gillus species, Burkholderia species, S. aureus, Nocardia cause dysregulation of inflammation control that leads to
species, and Mycobacteria species. The mutations responsible unpredictable episodes of neutrophil overactivity and tissue
for CGD occur in one of the components of the NADPH infiltration by activated neutrophils. Because the chronic, recur-
oxidase system and can be inherited in either an X-linked or rent inflammatory attacks also cause persistent elevations of
autosomal recessive manner. About two-thirds of CGD cases serum amyloid A protein, patients with FMF are at high risk of
are due to mutations affecting the X-linked gene CYBB, developing complications of amyloid A amyloidosis, especially
which encodes the gp91phox component of the membrane in the kidneys. In this regard, FMF is the leading cause of sec-
cytochrome b558 protein complex. The other cases involve ondary amyloidosis in Turkey. The diagnosis of FMF usually is

made based on clinical criteria, including unexplained episodes

that persist over many months to years in the absence of other Key points (continued)
etiologies of inflammation. Most of the common MEFV muta- • CGD is characterized by recurrent bacterial and fungal
tions are well characterized. Thus, the diagnosis now can be infections and is due to mutations that impair the ability of
confirmed and family studies can be carried out using molecu- phagocytes to generate reactive oxygen intermediates.
lar studies done by a reference research laboratory. Colchicine • FMF should be considered in children or adults with unex-
prevents clinical attacks and tissue amyloid deposition in most plained recurrent fever and inflammation and the appropriate
patients with FMF. Rare patients with exceptional, refractory ethnic background.
• Majeed syndrome should be considered in children with CDA
disease have undergone successful HSCT.
accompanied by periodic fever and chronic recurrent multifocal
Hyper-IgD syndrome is another rare autosomal reces-
osteomyelitis.
sive autoinflammatory disease found in Dutch and French
families that is associated with mutations in the mevalon-
ate kinase gene, MVK. The tumor necrosis factor (TNF)
receptor-associated periodic syndrome (TRAPS; previ-
Disorders of histiocytes and dendritic cells
ously known as familial Hibernian fever) is an autosomal
dominant disorder affecting Scottish and Irish individuals
Hemophagocytic lymphohistiocytosis
and associated with mutations in the gene-encoding TNF
receptor 1, TNFRSF1A. Cryopyrin-associated periodic syn-
dromes (CAPS) are a group of autosomal dominant inher- Clinical case
ited disorders that are caused by mutations of a pyrin-like A 9-month-old girl is admitted to the hospital after presenting
protein called NALP3, encoded by the CIAS1 gene. The with fever of 40.5°C, sore throat, and lethargy. Over the course
type of CIAS1 mutation determines the clinical severity. of the next 48 hours, the child continues to have high fevers
Familial cold autoinflammatory syndrome is the most despite broad-spectrum antibiotics and develops progressive
severe form of CAPS, followed by Muckle-Wells syndrome splenomegaly and pancytopenia. Laboratory data is also notable
and familial cold autoinflammatory syndrome. Although for a markedly elevated ferritin level of 24,000 ng/mL (normal
neutrophils are not the primary mediators of pathogenesis 4-76 ng/mL) and hypofibrinogenemia of 68 mg/dL (normal
in these non-FMF disorders, they share many clinical fea- 150-450 mg/dL). A BM biopsy reveals marked histiocyte
hyperplasia with hemophagocytosis. She begins treatment with
tures with FMF and should be considered in the differential
dexamethasone, cyclosporine, and etoposide. Mutational testing
diagnosis of unexplained recurrent fever with noninfec-
reveals the presence of a homozygous mutation in the PRF1 gene.
tious autoinflammation.
Majeed syndrome is a rare autoinflammatory syndrome
characterized by sterile, chronic recurrent multifocal osteomy- Hemophagocytosis is the histologic finding of activated
elitis, with pain and swelling around a joint, recurrent febrile macrophages engulfing leukocytes, erythrocytes, platelets,
episodes, and CDA. Inflammatory dermatoses and hepato- and their precursor cells. Hemophagocytosis may be
splenomegaly may be present. Patients present during the first observed in a variety of conditions, including hemolytic ane-
two years of life. The anemia is hypochromic and microcytic mias, infections, and malignancies. It also is a principal fea-
and may be mild or transfusion dependent. Majeed syndrome is ture of hemophagocytic lymphohistiocytosis (HLH), a
an autosomal recessive disorder caused by mutations in LPIN2, clinical syndrome characterized by fever, pancytopenia, and
which encodes lipin-2, a phosphatidic acid phosphatase. The splenomegaly that results from the abnormal activation and
role of lipin-2 in the control of chronic inflammation remains to proliferation of cytotoxic T-lymphocytes and tissue macro-
be elucidated. Nonsteroidal anti-inflammatory drugs are mod- phages (Figure 15-6). The major pathophysiologic abnor-
erately helpful. Clinical improvement was reported in two mality in HLH is the high production of inflammatory
brothers with Majeed syndrome with interleukin 1 inhibition. cytokines with abnormal T-cell activation. Severe impair-
ment in NK cell activity and cytotoxic T-cell function are
also characteristic of the disease.
Key points HLH may occur either as an inherited or acquired disor-
der (Table 15-6). Familial hemophagocytic lymphohistiocy-
• Genetic disorders affecting neutrophil function are rare causes tosis (FHL) is an autosomal recessive disease that classically
of recurrent infections, unexplained fever, and inflammation.
presents in infancy and early childhood with an estimated
• Disorders affecting neutrophil adhesion, chemotaxis, and
incidence of approximately 1 in 50,000. FHL is caused by
killing usually are diagnosed in young children with recurrent
mutations in genes that encode critical components of the
infections.
granule exocytosis pathway, which enables NK cells and

Disorders of histiocytes and dendritic cells | 421
are released to initiate cell death. FHL-3, FHL-4, and FHL-5

are caused by mutations in the Munc13-4 (UNC13D), syn-
taxin 11 (STX11), and syntaxin binding protein 2 (STXBP2)
genes, respectively. The mutation in FHL-1 is known to map
to 9q21.3-q22, but the gene remains unknown. In addition
to FHL, HLH also occurs in the context of inherited immune
deficiency syndromes, including CHS (LYST ), Griscelli syn-
drome type II (RAB27A), X-linked proliferative disease
(XLP, caused by mutations in SH2DIA), X-linked inhibitor
of apoptosis deficiency syndrome (XIAP), Hermansky-
Pudlak type II (AP3B1), and GATA2 deficiency.
Acquired HLH syndrome, also known as reactive hemo-
phagocytic syndrome or secondary HLH (sHLH), can affect
both adults and children and usually is associated with an
underlying infection, malignancy, or other immunocompro-
Figure 15-6 Hemophagocytic lymphohistiocytosis. Bone marrow
mised state. Predisposing conditions associated with sHLH
aspirate demonstrating phagocytic histiocytes with ingested platelets
include bacterial and viral infections, autoimmune or rheuma-
and RBC precursors. From ASH Image Bank, #3502.
tologic disorders, hematologic and (less commonly) nonhe-
cytotoxic T-lymphocytes to induce apoptosis in target cells. matologic malignancies, AIDS (with or without opportunistic
A genetic diagnosis can be established in 50%-80% of infections), and posttransplantation immunosuppression. The
patients with HLH, depending on the race and ethnicity of pathophysiology of sHLH appears to be similar to that of FHL
the cohort. There are 5 FHL subtypes. FHL-2 is caused by except that in these patients, the underlying predisposing dis-
mutations in the PRF1 gene, which encodes perforin. Perfo- order, and not a genetic defect, is primarily responsible for the
rin is the major component of the cytolytic granules and dysregulation of T-cells and NK cells that leads to histiocyte
forms the pore at the synapse between the effector lympho- activation, although patients with sHLH commonly have poly-
cyte and the target cells through which the cytolytic contents morphisms in genes associated with FHL that make them
more susceptible to develop this complication.
The clinical presentation, laboratory features, and histopa-
Table 15-6 Hereditary and acquired causes of HLH
thology of inherited and acquired HLH are similar. HLH
Primary HLH should be considered in the differential diagnosis in patients
Familial HLH
who develop sepsis or multiorgan dysfunction in the setting of
Chédiak-Higashi syndrome
fever, unexplained progressive pancytopenia, and hepato-
Griscelli syndrome
splenomegaly. Laboratory findings include elevated ferritin
X-linked lymphoproliferative (XLP) disease
and triglycerides with low fibrinogen. Central nervous system
X-linked inhibitor of apoptosis (XIAP)
(CNS) involvement may range from irritability, bulging fonta-
Hermansky-Pudlak syndrome type II
GATA2 deficiency
nel, and neck stiffness to seizures, cranial nerve palsies, ataxia,
Secondary HLH or coma. Lymphadenopathy, rash, and liver disease also may be
Infections present. A BM biopsy is crucial to identify histiocytic hyper-
Herpesvirus infection, particularly EBV, CMV, HHV-8, HSV plasia and hemophagocytosis of nucleated cells. It is important
HIV to note that hemophagocytosis is highly variable and may not
Parvovirus, adenovirus, hepatitis virus be observed early in the clinical course and that the absence of
Bacterial, rickettsial, fungal, and spirochete-associated infections hemophagocytosis does not exclude the diagnosis of HLH.
Malignancy Lumbar puncture and magnetic resonance imaging (MRI)
AML, MDS, lymphomas, multiple myeloma of the brain should be performed in those with suspected
Metastatic carcinoma CNS involvement.
Autoimmune diseases (macrophage activation syndrome) Although evaluation for an underlying infection should
Other immunodeficiency states be performed, both familial and secondary forms of HLH
Posttransplant are frequently triggered by an infection. Diagnostic criteria
Cytotoxic or immunosuppressive therapy for HLH have been established by the Histiocyte Society
Postsplenectomy
(Table 15-7). In addition to the original criteria proposed in
EBV = Epstein-Barr virus; CMV = cytomegalovirus; HHV-8 = human 1991 of fever, splenomegaly, cytopenias, hypertriglyceride-
herpesvirus 8; HSV = herpes simplex virus. mia or hypofibrinogenemia, and hemophagocytosis, three

Table 15-7 2004 Revised diagnostic criteria for hemophagocytic lymphohistiocytosis

The diagnosis of HLH can be established if one of either 1 or 2 is fulfilled:
1. A molecular diagnosis consistent with HLH
2. Diagnostic criteria for HLH fulfilled (five out of the following eight criteria)
A. Initial diagnostic criteria (to be evaluated in all patients with HLH)
Fever
Splenomegaly
Cytopenias (affecting ≥2 of 3 lineages in the peripheral blood):
Hemoglobin <90 g/L (in infants <4 weeks: hemoglobin <100 g/L)
Platelets <100 × 109/L
Neutrophils <1.0 × 109/L
Hypertriglyceridemia and/or hypofibrinogenemia:
Fasting triglycerides ≥3.0 mmol/L (ie, ≥265 mg/dL)
Fibrinogen ≤1.5 g/L
Hemophagocytosis in bone marrow or spleen or lymph nodes
No evidence of malignancy
B. New diagnostic criteria
Low or absent natural killer cell activity (according to local laboratory reference)
Ferritin ≥500 µg/L
Soluble CD25 (ie, soluble IL-2 receptor) ≥2,400 U/mL
If hemophagocytic activity is not proven at the time of presentation, further search for hemophagocytic activity is encouraged but not mandatory
for diagnosis if other markers are consistent with the disease. If the bone marrow specimen is not conclusive, material may be obtained from
other organs. Serial marrow aspirates over time may also be helpful. The following findings may provide strong supportive evidence for the
diagnosis: (i) spinal fluid pleocytosis (mononuclear cells) or elevated spinal fluid protein, (ii) histological picture in the liver resembling chronic
persistent hepatitis (biopsy). Other abnormal clinical and laboratory findings consistent with the diagnosis are cerebromeningeal symptoms,
lymph node enlargement, jaundice, edema, skin rash, hepatic enzyme abnormalities, hypoproteinemia, hyponatremia, increased very-low-
density lipoprotein (VLDL), and decreased high-density lipoprotein (HDL).
additional criteria were introduced in 2004; these are low or therapy 82% short term complete response rates. The Hybrid
absent NK cell activity, hyperferritinemia, and high levels of Immunotherapy for Hemophagocytic LymphoHistiocytosis
sIL-2R. At least five of eight clinical criteria or the presence and EURO-HIT-HLH phase 2 trials, which are evaluating regi-
of either familial disease or one of the known genetic abnor- mens of ATG, dexamethasone, and etoposide, are currently
malities is required for diagnosis of HLH. accruing patients in the United States and Europe, respectively.
Although sHLH may resolve after treatment of the underly- Allogeneic HSCT is recommended in patients with FHL and
ing condition or with a short course of immunosuppression, in patients with relapsed or refractory sHLH. NI-0501, a fully
untreated FHL is uniformly fatal within 1-2 months. Treat- human monoclonal antibody that targets interferon gamma, is
ment of FHL consists of chemoimmunotherapy followed by active in FHL and currently undergoing clinical trials.
allogeneic stem cell transplantation. The HLH-94 protocol of Macrophage activation syndrome (MAS) is considered to
the Histiocyte Society for FHL consists of an initial 8 weeks of be a variation of sHLH, which occurs in individuals with
dexamethasone and etoposide followed by maintenance cyclo- autoimmune disorders. The disorder most frequently seen is
sporine with pulses of etoposide and dexamethasone. Intra- systemic juvenile idiopathic arthritis (SJIA) but also can be
thecal therapy with methotrexate and corticosteroids is observed in other rheumatologic conditions, including SLE
administered in individuals with evidence of CNS involve- and Kawasaki disease. Like other forms of HLH, MAS is
ment. Results of HLH-94 demonstrate a 3-year survival rate of characterized by fever, hepatosplenomegaly, cytopenias, and
51%. The subsequent protocol, HLH-2004, is similar to the coagulopathy with the expansion of macrophages and T-cells
HLH-94 protocol and includes etoposide, dexamethasone, as well as decreased cytotoxic T-cell and NK function.
and cyclosporine with an earlier introduction of cyclosporine Approximately 10% of individuals with SJIA can develop
to reduce the risk of relapse while corticosteroids are being life-threatening MAS, although it is believed that a much
tapered. Outcomes of HLH-2004, which was closed to enroll- higher percentage may have a milder or subclinical form.
ment in 2011, remain under evaluation. A single center retro- Although MAS resembles HLH, diagnostic criteria for HLH
spective analysis of FHL patients treated with antithymocyte may not apply as some features, such as hyperferritinemia,
globulin (ATG), prednisone, maintenance cyclosporine, and lymphadenopathy, and splenomegaly, often are present dur-
intrathecal methotrexate and corticosteroids reported first line ing a flare of the underlying disease. MAS generally responds

Disorders of histiocytes and dendritic cells | 423
to high dose corticosteroids alone or in combination with Diabetes insipidus may result from intracranial extension of
cyclosporine. craniofacial bone lesions and is the most common CNS
manifestation, occurring in up to 30% of patients. LCH of
the lungs in adults frequently is associated with smoking.
Langerhans cell histiocytosis
Tissue biopsy is required to confirm the diagnosis of LCH.
Langerhans cells are specialized dendritic cells that are found Histologically, the lesions contain a mix of characteristic
in the skin and mucosa. Langerhans cell histiocytosis (LCH) Langerhans cells in a background of eosinophils, neutro-
is a clonal disorder of dendritic cells associated with polymor- phils, and lymphocytes (Figure 15-7). Langerhans cells are
phic cellular infiltration and damage at either unifocal tissue positive for CD1a and S-100, and, by ultrastructural exami-
sites or in multiple organs and tissues. Although the dendritic nation, contain the hallmark Birbeck granules. Birbeck gran-
cells in LCH expresses similar antigens, including CD1a and ules are tennis racket-shaped cytoplasmic granules ~200-400
CD207 as skin Langerhans cells, they are believed to originate nm in length and 33 nm in width with a zipper-like appear-
from a distinct myeloid dendritic cell precursor. LCH is rare, ance. Because expression of langerin (CD207) confirms the
with an annual incidence of approximately five per million. presence of Birbeck granules, electron microscopy is now
Historically, localized LCH was referred to as eosinophilic rarely done for diagnosis. Some tumors contain an abun-
granuloma, whereas clinical variants of multisystem disease dance of eosinophils and neutrophils with central necrosis,
were referred to as histiocytosis X, Letterer-Siwe disease, and whereas fibrosis and foamy macrophages are found in more
Hand-Schüller-Christian syndrome. Recently, somatic muta- long-standing lesions.
tions in BRAF (V600E) have been identified in >50% of Treatment of LCH is based on the extent and activity of
patients LCH. In addition, mutations in MAP2K1, ERBB3, or the disease. SS-LCH generally confers a good prognosis and
ARAF or have been found in patients with wildtype BRAF, frequently requires minimal or no treatment. Bony or soft
revealing activation of extracellular signal-regulated kinase tissue SS-LCH can be treated with surgical resection or bony
(ERK) signaling as major driver of LCH pathogenesis. curettage, local irradiation, or injection of steroids. Limited
Patients with LCH are categorized as having either uni- or skin disease often responds to topical steroids, nitrogen mus-
multifocal involvement of a single organ system (SS-LCH) or tard, or psoralen and ultraviolet A (PUVA) light therapy.
multisystem LCH (MS-LCH). SS-LCH most commonly Management of lung disease includes discontinuation of
involves the bone (particularly the skull, femur, pelvis, and smoking; treatment with prednisone, vinblastine, and meth-
ribs) and less commonly involves the skin, lymph nodes, and otrexate; and immunosuppressive agents. Disease-free sur-
lung. Usual presentations of limited disease include persis- vival with limited or local LCH exceeds 95%; however,
tent or recurrent and progressive bony pain or swelling, recurrences are common, and some patients require multiple
chronic skin rash, chronic ear drainage, dyspnea, cough, and courses of treatment to be cured. Therefore, patients must be
pneumothorax (isolated bony involvement resulting in ear monitored closely for evolution to multisystem disease, sec-
drainage is more common in children, and pulmonary dis- ondary malignancies, and, in the case of lung involvement,
ease occurs predominantly in adults). MS-LCH most com- progressive pulmonary compromise.
monly occurs in young children and may present with MS-LCH and SS-LCH with progressive multifocal
various combinations of bony or soft tissue masses with involvement or involvement of critical anatomic sites are
symptoms including fever, eczematoid rash, gingival swell- treated with systemic therapy. Induction therapy with
ing, cough or dyspnea, tooth loss, hepatosplenomegaly, vinblastine and prednisone commonly is used as initial
lymphadenopathy, abnormal chest x-ray, and cytopenias. therapy. Involvement of the hematopoietic system, spleen,
Figure 15-7 Langerhans cell histiocytosis.

(a) Hematoxylin-eosin stain demonstrating
Langerhans cell infiltrate. Cells have abundant
eosinophilic cytoplasm with variably shaped nuclei
ranging from cleaved, grooved, folded, indented, and
even lobated. Clusters of eosinophils surround the
infiltrate. (b) CD1a immunohistochemistry staining
Langerhans cells. From ASH Image Bank, images
3461 (a) and 3465 (b).

liver, and lung is considered high risk, with a mortality of These disorders lead to an accumulation of undigested mate-
~20% compared with <5% for patients without high-risk rial inside the lysosome, leading to cell degeneration and
features. Disease recurrence and progression are most accumulation of macromolecules in various tissues and
common in patients with extensive visceral disease and a organs of the body and resulting in organ dysfunction.
suboptimal initial response. In addition, long-term neuro Gaucher disease and Niemann-Pick disease represent a
logic complications increasingly are being recognized in subtype of lysosomal storage diseases, also known as sphin-
patients with LCH, particularly those with MS-LCH or CNS golipidoses or lipid storage disorders, in which undigested
involvement. Neurodegenerative changes may be seen on lipids accumulate in the lysosome-rich cells of the monocyte
MRI and can be accompanied by symptoms, including ataxia, or macrophage system. These conditions are of particular
dysarthria, dysmetria, and learning and behavior difficulties. importance to hematologists because they frequently present
with cytopenias and hepatosplenomegaly.
Non–Langerhans cell histiocytoses
Gaucher disease
A number of other rare histiocytic disorders that are pheno-
typically distinct from Langerhans cells have been character-
ized. Juvenile xanthogranuloma (JXG), the most common of Clinical case
these disorders, is a proliferative disorder of young children A 23-year-old man from Ukraine presents with a several-month
that generally appears as a solitary or multiple red, yellow, or history of easy bruising, worsening fatigue, and hip pain. On
brown papular skin lesions. The condition generally follows physical examination, the patient is noted to be pancytopenic
a benign clinical course and usually resolves spontaneously, with an Hb level of 8.0 and platelet count of 40 × 109/L and has
although extracutaneous lesions do rarely occur. marked hepatosplenomegaly with a spleen measuring 12 cm
Sinus histiocytosis with massive lymphadenopathy, also below the costal margin. A BM biopsy reveals the presence of
known as Rosai-Dorfman disease, is a nonmalignant prolif- lipid-laden macrophages consistent with Gaucher cells infiltrat-
ing the marrow. Measurement of leukocyte glucocerebrosidase
eration of histiocytes within lymph node sinuses and lym-
is reduced markedly measuring at <10% of normal levels.
phatics in extranodal sites. The condition most commonly
occurs in children and young adults and presents as massive,
painless, bilateral lymph node enlargement in the neck with Gaucher disease is a lipid storage disease that results from the
fever. Other nodal and extranodal sites may sometimes be deficiency of glucocerebrosidase (acid β-glucosidase), which
involved. Although spontaneous resolution is observed in hydrolyzes glucocerebroside to glucose and ceramide. Defi-
most cases, relapses can occur, and the condition occasion- ciency of the enzyme causes glucocerebroside accumulation
ally can be fatal. There is no standard treatment approach, in the cytoplasm of tissue macrophages, known as Gaucher
and therapies employed have included corticosteroids, radia- cells, resulting in a characteristic wrinkled-paper appearance
tion, thalidomide, or cytotoxic agents including vinca alka- (Figure 15-8).
loids and purine nucleoside analogues. Gaucher disease is the most common lysosomal storage dis-
ease, with an incidence of approximately 1 in 75,000 births, and
is more common in Ashkenazi Jewish populations, with an
Lysosomal storage diseases incidence of 1 in 1,000. Gaucher disease is inherited as an auto-
somal recessive disorder, with over 300 mutations having been
Lysosomal storage diseases are a collection of approximately described. The disease is divided into three clinical subtypes
50 genetically inherited disorders characterized by a defi- based on pattern and severity of neurologic involvement. Type
ciency or defect in one or more specific lysosomal enzymes. I (nonneuropathic) is most common (90% of all patients), has
Figure 15-8 Gaucher disease. (a) Proliferation

of benign-appearing macrophages with
interspersed normal hematopoietic elements.
(b) High-power view of bone marrow aspirate
demonstrating a Gaucher cell, an abnormal
macrophage with the characteristic “wrinkled-
paper” cytoplasm. From ASH Image Bank,
images 2711 (a) and 2709 (b).

Lysosomal storage diseases | 425
the most variable clinical presentation, and is associated with Oral substrate reduction therapy has also been developed for
the highest residual enzyme activity. Symptoms consist of hep- Gaucher disase. Miglustat, the first in its class, inhibits
atosplenomegaly, cytopenias, and bone disease. Skeletal mani- glucosylceramide synthase, a key enzyme in glycosphingolipid
festations include osteopenia, pain crises, and osteolytic lesions, synthesis, thereby reducing substrate accumulation. Miglustat is
with radiographs showing flaring of the ends of the long bones available for use in the United States for patients unable to
(Erlenmeyer flask deformity) and cortical thinning. Although receive ERT and in Europe for adult patients with mild to
the clinical severity may not be predicted by the genotype, early moderate disease. In clinical studies, miglustat decreased liver
onset is associated with more rapidly progressive and severe and spleen volumes by 12% and 19%, respectively, with modest
disease. Type II (acute neuronopathic) is rarest, with the lowest improvements in hemoglobin and platelet counts. Eliglustat is
enzyme activity. Disease onset occurs during infancy and an additional glucosylceramide synthase inhibitor approved in
results in progressive neurologic deterioration that includes the United States and Europe. It is approved for use first-line
generalized seizures, hypertonia, profound mental retardation, treatment of Gaucher disease type I. A randomized, double-
and death during infancy. Type III (subacute neuronopathic) blind, placebo-controlled trial of 40 adult patients with untreated
falls between types I and II in incidence, enzyme activity, and Gaucher disease type 1 demonstrated a decrease in spleen
clinical severity. Onset occurs at any time during childhood, volume by 27% (the primary study endpoint), and an increase
and manifestations include progressive dementia and ataxia, in platelet count by 41% and modest improvements in liver
bone and visceral involvement, and supranuclear palsies. volume and hemoglobin level.
The diagnosis of Gaucher disease can be established by
enzyme assay for glucocerebrosidase activity in leukocytes,
Niemann-Pick disease
fibroblasts, or urine, which is between 0% and 30% of normal
values. In addition, mutational analysis of the four most com- Type A and type B Niemann-Pick disease (NPD) are caused
mon mutations of the glucocerebrosidase gene, (p.N409S by mutations in the sphingomyelin phosphodiesterase-1
[known as N370S], IVS2+1G>A, p.L483P [known as L444P] (SMPD1) gene, which result in deficient sphingomyelinase
and c.84dupG [known as 84GG]) can detect 90%-95% of the activity and accumulation of sphingomyelin (ceramide
mutations associated with Gaucher disease in the Ashkenazi phosphorylcholine). Type C NPD is an unrelated defect
Jewish population and 50%-75% of the associated mutations caused by mutations of the NPC1 and NPC2 genes, which
in the general population. The onset of symptoms ranges result in impaired cellular processing and transport of low-
from 2 years of age to late adulthood, and it has been esti- density lipoprotein cholesterol.
mated that up to 60% of individuals harboring the most com- NPD is inherited as an autosomal recessive disorder. Type
mon N370S mutation never present to medical attention. A patients have <5% of normal sphingomyelinase activity,
Enzyme replacement therapy (ERT) is the mainstay of and the disease is characterized by hepatosplenomegaly, fail-
treatment of individuals with nonneuropathic manifesta- ure to thrive, and rapidly progressive neurodegeneration,
tions of Gaucher disease. Imiglucerase (Cerezyme) is a with death occurring by age 2-4 years. Examination reveals
recombinant modified placental glucocerebrosidase in which cherry-red maculae in approximately half of affected infants.
the glycosylation sites of the enzyme are processed to termi- Type B patients have sphingomyelinase activity within
nate in mannose sugars to improve uptake and trafficking to 5%-10% of normal, often have minimal to no neurologic
the lysozymes of macrophages via the mannose receptor. In involvement, and can survive into adulthood. Cytopenias
clinical studies, ERT at 30-60 U/kg administered every and hepatosplenomegaly are typical, and patients can
2 weeks normalized cytopenias and reduced organomegaly develop progressive pulmonary infiltrates.
within 6-12 months, whereas skeletal symptoms improve The histologic hallmark of NPD is the pathologic foam cell or
more slowly. Studies using low-dose ERT at 15-30 U/kg/ Niemann-Pick cell; these cells are histiocytes filled with lipid
month administered more frequently at three times a week droplets or particles that are uniform in size, giving the cells a
have shown that this regimen appears to produce similar “mulberry-like” or “honeycomb-like” appearance, and are
effects on cytopenias and organomegaly at significantly found in involved organs. Types A and B NPD may be readily
reduced cost. Given the high cost of ERT, published guide- diagnosed by assays for sphingomyelinase in leukocytes or cul-
lines advocate treatment only for symptomatic children and tured fibroblasts, which demonstrate reduced activity
adults with severe disease (eg, platelet counts <60,000/µL, (1%-10%) in the disease. Genetic testing can detect the most
marked splenomegaly, skeletal disease). Because glucocere- common mutations in type A, which account for approximately
brosidase does not cross the blood–brain barrier, it has lim- 90% of the mutant alleles in the Ashkenazi Jewish population.
ited utility in neuropathic forms of the disease. Velaglucerase Currently, no specific treatment exists for NPD types A
and taliglucerase are alternative recombinant glucocerebro- and B. Miglustat is approved for the treatment of progressive
sidase preparations available for ERT. neurological manifestations of NPD type C.

for clinical management in children and adults. Clin Genet.

Key points 2015;88:13-24.
• HLH is a pathologic activation and proliferation of tissue
Soulier J. Fanconi anemia. Hematology Am Soc Hematol Educ
histiocytes leading to severe multisystem clinical consequences.
Program. 2011;2011:492-497.
HLH may present in young children with an inherited predisposi-
tion (eg, due to perforin gene mutations) or in children and
adults with acquired disorders of immune regulation due to
infection, autoimmune disorder, malignancy, or acquired Dokal I. Dyskeratosis congenita. Hematology Am Soc Hematol Educ
immunodeficiency state. Program. 2011;2011:480-486.
• LCH is a clonal dendritic cell disorder that can present with Gramatges MM, Bertuch AA. Short telomeres: from dyskeratosis
involvement of a single tissue (usually the bone) or multiple congenita to sporadic aplastic anemia and malignancy.
tissues and organs, including the pituitary and hypothalamus Transl Res. 2013;162:353-363.
(with diabetes insipidus). The clinical course may be variable,
with periods of disease inactivity or chronic progression, and
treatment must be individualized. Shwachman-Diamond syndrome
• Gaucher disease is a lysosomal storage disorder caused by Dror Y, Donadieu J, Koglmeier J, Dodge J, Toiviainen-Salo S,
mutations in glucocerebrosidase, leading to abnormal accumula- Makitie O, Kerr E, Zeidler C, Shimamura A, Shah N, Cipolli
tion of glucocerebroside in tissue macrophages and resulting in M, Kuijpers T, Durie P, Rommens J, Siderius L, Liu JM. Draft
hepatosplenomegaly, cytopenias, and skeletal disorders. ERT and consensus guidelines for diagnosis and treatment of Shwachman
substrate reduction therapy can reverse both nonhematologic Diamond Syndrome. Ann N Y Acad Sci. 2011;1242:40-55.
and hematologic manifestations. Myers KC, Davies SM, Shimamura A. Cinical and molecular
pathophysiology of Shwachman-Diamond syndrome: an
update. Hematol Oncol Clin North Am. 2013;27:117-128.
Bibliography
General Pearson syndrome
Alvarez-Errico D, Vento-Tormo R, Sieweke M, Ballestar E. Manea EM, Leverger G, Bellmann F, Stanescu PA, Mircea A, Lebre
Epigenetic control of myeloid cell differentiation, identity and AS, Rotig A, Munnich A. Pearson syndrome in the neonatal
function. Nat Rev Immunol. 2015;15:7-17. period: two cases and a review of the literature. J Pediatr
Friedman AD. Transcriptional control of granulocyte and monocyte Hematol Oncol. 2009;31:947-951.
development. Oncogene. 2007;26:6816-6828. Review of transcription
factors that control neutrophil and monocyte development. Diamond-Blackfan anemia
Hsieh MM, Everhart JE, Byrd-Holt DD, Tisdale JF, Rodgers
GP. Prevalence of neutropenia in the U.S. population: age, Ball S. Diamond-Blackfan anemia. Hematology Am Soc Hematol
sex, smoking status, and ethnic differences. Ann Intern Med. Educ Program. 2011;2011:487-491.
2007;146:486-492. Population-based study of ethnic differences in Vlachos A, Blanc L, Lipton JM. Diamond Blackfan anemia: a
blood neutrophil counts. model for the translational approach to understanding human
Nauseef WM, Borregaard N. Neutrophils at work. Nat Immunol. disease. Expert Rev Hematol. 2014;7:359-372.
2014;15:602-611. Current review of neutrophil homeostasis and
emigration.
Spinner MA, Sanchez LA, Hsu AP, et al. GATA2 deficiency: a
Congenital dyserythropoietic anemia
protean disorder of hematopoiesis, lymphatics, and immunity. Iolascon A, Heimpel H, Wahlin A, Tamary H. Congenital
Blood. 2014;123:809-821. A comprehensive examination of the dyserythropoietic anemias: molecular insights and diagnostic
clinical features of GATA2 deficiency. approach. Blood. 2013;122:2162-2166.
Sullivan BM, Locksley RM. Basophils: a nonredundant contributor to Iolascon A, Russo R, Delaunay J. Congenital dyserythropoietic
host immunity. Immunity. 2009;30:12-20. Review of basophil biology. anemia. Curr Opin Hematol. 2011;18:146-151.
von Vietinghoff S, Ley K. Homeostatic regulation of blood
neutrophil counts. J Immunol. 2008;181:5183-5188.
Zarbock A, Ley K. Mechanisms and consequences of neutrophil Congenital and cyclical neutropenia
interaction with the endothelium. Am J Pathol. 2008;172:1-7. Donadieu J, Fenneteau O, Beaupain B, Mahlaoui N, Chantelot CB.
Congenital neutropenia: diagnosis, molecular bases and patient
management. Orphanet J Rare Dis. 2011;6:26.
Fanconi anemia
Hauck F, Klein C. Pathogenic mechanisms and clinical
Schneider M, Chandler K, Tischkowitz M, Meyer S. Fanconi implications of congenital neutropenia syndromes. Curr Opin
anaemia: genetics, molecular biology, and cancer—implications Allergy Clin Immunol. 2013;13:596-606.

Bibliography | 427
Klein C. Genetic defects in severe congenital neutropenia: Papadaki HA, Palmblad J, Eliopoulos GD. Non-immune chronic
emerging insights into life and death of human neutrophil idiopathic neutropenia of adult: an overview. Eur J Haematol.
granulocytes. Annu Rev Immunol. 2011;29:399-413. 2001;67:35-44. A review of nonimmune chronic idiopathic
neutropenia.
Starkebaum G. Chronic neutropenia associated with autoimmune
WHIM syndrome disease. Semin Hematol. 2002;39:121-127. A nice review of
Al Ustwani O, Kurzrock R, Wetzler M. Genetics on a WHIM. neutropenia associated with autoimmune disorders.
Br J Haematol. 2014;164:15-23.
Kawai T, Malech HL. WHIM syndrome: Congenital immune Drug-induced neutropenia
deficiency disease. Curr Opin Hematol. 2009;16:20-26.
Andersohn F, Konzen C, Garbe E. Systematic review: agranulocytosis
induced by nonchemotherapy drugs. Ann Intern Med.
Thrombocytopenia with absent radii 2007;146:657-665. Systematic review of drug-induced neutropenia.
Andres E, Maloisel F. Idiosyncratic drug-induced agranulocytosis
Albers CA, Newbury-Ecob R, Ouwehand WH, Ghevaert C. New
or acute neutropenia. Curr Opin Hematol. 2008;15:15-21.
insights into the genetic basis of TAR (thrombocytopenia-
Systematic review of drug-induced neutropenia.
absent radii) syndrome. Curr Opin Genet Dev. 2013;23:316-23.
Curtis BR. Drug-induced immune neutropenia/agranulocytosis.
Toriello HV. Thrombocytopenia-absent radius syndrome. Semin
Immunohematology. 2014;30:95-101.
Thromb Hemost. 2011;37:707-712.
Congenital amegakaryocytic Disorders of neutrophil function

thrombocytopenia Boxer LA. Neutrophil abnormalities. Pediatr Rev. 2003;24:52-62.
Excellent review of congenital disorders of neutrophil function.
Ballmaier M, Germeshausen M. Congenital amegakaryocytic
Ferguson PJ, Chen S, Tayeh MK, et al. Homozygous
thrombocytopenia: clinical presentation, diagnosis and
mutations in LPIN2 are responsible for the syndrome of
management. Semin Thromb Hemost. 2011;37:673-681.
chronic recurrent multifocal osteomyelitis and congenital
dyserythropoietic anaemia (Majeed syndrome). J Med Genet.
Neutrophilia 2005;42:551-557. Original report of LPIN2 gene mutations in
Majeed syndrome.
Reding MT, Hibbs JR, Morrison VA, Swaim WR, Filice GA. Marciano BE, Spalding C, Fitzgerald A, et al. Common severe
Diagnosis and outcome of 100 consecutive patients with infections in chronic granulomatous disease. Clin Infect Dis.
extreme granulocytic leukocytosis. Am J Med. 1998;104:12-16. 2015;60:1176-1183. Survey of infections in 268 patients with
Classic paper on neutrophilia. chronic granulomatous disease followed over 4 decades.
Seebach JD, Morant R, Ruegg R, et al. The diagnostic value of the Martinon F, Aksentijevich I. New players driving inflammation
neutrophil left shift in predicting inflammatory and infectious in monogenic autoinflammatory diseases. Nat Rev Rheumatol.
disease. Am J Clin Pathol. 1997;107:582-591. Classic paper on 2015;11:11-20. Comprehensive review of the genetics and biology
neutrophilia. of the autoinflammatory diseases.
Qasim W, Cavazzana-Calvo M, Davies EG, et al. Allogeneic
hematopoietic stem-cell transplantation for leukocyte adhesion
Immune neutropenia and neutropenia
deficiency. Pediatrics. 2009;123:836-840. Update on HSCT for
associated with autoimmunity
leukocyte adhesion deficiencies.
Berliner N, Horwitz M, Loughran TP. Congenital and acquired Sassi A, Lazaroski S, Wu G, et al. Hypomorphic homozygous
neutropenia. Hematology Am Soc Hematol Educ Program. mutations in phosphoglucomutase 3 (PGM3) impair immunity
2004:63-69. and increase serum IgE levels. J Allergy Clin Immunol.
Bux J, Behrens G, Jaeger G, Welte K. Diagnosis and clinical 2014;133:1410-1419.e13. Original report of gene mutations in
course of autoimmune neutropenia in infancy: analysis of 240 PGM3 in hyperimmunoglobulin IgE syndrome.
cases. Blood. 1998;91:181-186. Classic paper on autoimmune Seger RA. Modern management of chronic granulomatous
neutropenia in infants and young children. disease. Br J Haematol. 2008;140:255-266. Review of the
Gramatges MM, Fani P, Nadeau K, Pereira S, Jeng MR. Neonatal molecular pathogenesis, clinical features, and treatment of chronic
alloimmune thrombocytopenia and neutropenia associated granulomatous disease.
with maternal human leukocyte antigen antibodies. Pediatr Sowerwine KJ, Holland SM, Freeman AF. Hyper-IgE syndrome
Blood Cancer. 2009;53:97-99. Review of neonatal alloimmune update. Ann N Y Acad Sci. 2012;1250:25-32.
neutropenia. van de Vijver E, van den Berg TK, Kuijpers TW. Leukocyte
O’Malley DP. T-cell large granular leukemia and related adhesion deficiencies. Hematol Oncol Clin North Am.
proliferations. Am J Clin Pathol. 2007;127:850-859. Excellent 2013;27:101-116, viii. Review on the molecular pathogenesis,
review of neutropenia, particularly LGL and Felty’s syndrome. clinical features, and treatment of leukocyte adhesion defiency.

Vitale A, Rigante D, Lucherini OM, et al. Biological the Histiocyte Society. Eur J Cancer. 2003;39:2341-2348. Large
treatments: new weapons in the management of monogenic registry study of 274 adults with LCH.
autoinflammatory disorders. Mediators Inflamm. Badalian-Very G, Vergilio JA, Degar BA, et al. Recurrent
2013;2013:939847. BRAF mutations in Langerhans cell histiocytosis. Blood.
2010;116:1919-1923.
Gadner H, Minkov M, Grois N, et al. Therapy prolongation
Hemophagocytic lymphohistiocytosis improves outcome in multisystem Langerhans cell histiocytosis.
Arico M, Janka G, Fischer A, et al. Hemophagocytic Blood. 2013;121:5006-5014.
lymphohistiocytosis. Report of 122 children from the Grois N, Potschger U, Prosch H, et al. Risk factors for diabetes
International Registry FHL Study Group of the Histiocyte insipidus in Langerhans cell histiocytosis. Pediatr Blood Cancer.
Society. Leukemia. 1996;10:197-203. Registry study of FHL. 2006;46:228-233. Analysis of 1,741 patients treated on clinical
Feldmann J, Callebaut I, Raposo G, et al. Munc13-4 is essential for studies for LCH.
cytolytic granules fusion and is mutated in a form of familial Histiocyte Society. Langerhans cell histiocytosis evaluation and
hemophagocytic lymphohistiocytosis (FHL3). Cell. 2003;115: treatment guidelines. http://www.histiocytesociety.org. Accessed
461-473. Initial report describing Munc13-4 mutations in FHL. March 4, 2015. Clinical guidelines published by the Histiocyte
Henter JI, Elinder G, Soder O, Hansson M, Andersson B, Society, April 2009.
Andersson U. Hypercytokinemia in familial hemophagocytic Titgemeyer C, Grois N, Minkov M, Flucher-Wolfram B, Gatterer-
lymphohistiocytosis. Blood. 1991;78:2918-2922. Report Menz I, Gadner H. Pattern and course of single-system disease
describing the presence of elevated inflammatory cytokines in FHL. in Langerhans cell histiocytosis data from the DAL-HX 83- and
Henter JI, Horne A, Arico M, et al. HLH-2004: diagnostic and 90-study. Med Pediatr Oncol. 2001;37:108-114. Describes the
therapeutic guidelines for hemophagocytic lymphohistiocytosis. outcome of 170 patients with SS-LCH treated on German LCH
Pediatr Blood Cancer. 2007;48:124-131. Updated guidelines from studies.
the Histiocyte Society on the diagnosis and treatment of HLH. Vassallo R, Ryu JH, Schroeder DR, Decker PA, Limper AH. Clinical
Henter JI, Samuelsson-Horne A, Arico M, et al. Treatment outcomes of pulmonary Langerhans’-cell histiocytosis in
of hemophagocytic lymphohistiocytosis with HLH–94 adults. N Engl J Med. 2002;346:484-490. Retrospective review of
immunochemotherapy and bone marrow transplantation. pulmonary LCH outcomes.
Blood. 2002;100:2367-2373. Large series reporting the outcome of Weitzman S, Jaffe R. Uncommon histiocytic disorders: the
children treated with stem cell transplantation for HLH. non-Langerhans cell histiocytoses. Pediatr Blood Cancer.
Jordan MB, Allen CE, Weitzman S, Filipovich AH, McClain 2005;45:256-264. Review of non-LCH disorders.
KL. How I treat hemophagocytic lymphohistiocytosis. Blood. Willis B, Ablin A, Weinberg V, et al. Disease course and late
2011;118:4041-4052. sequelae of Langerhans’ cell histiocytosis: 25-year experience
Katano H, Cohen JI. Perforin and lymphohistiocytic proliferative at the University of California, San Francisco. J Clin Oncol.
disorders. Br J Hematol. 2005;128:739-750. Excellent review 1996;14:2073-2082. Classic paper describing the natural history
describing spectrum of perforin mutations in HLH. of LCH.
Mahlaoui N, Ouachee-Chardin M, de Saint Basile G, Willman CL, Busque L, Griffith BB, et al. Langerhans’-cell
et al. Immunotherapy of familial hemophagocytic histiocytosis (histiocytosis X)—a clonal proliferative disease.
lymphohistiocytosis with antithymocyte globulins: a N Engl J Med. 1994;331:154-160. Classic paper describing the
single-center retrospective report of 38 patients. Pediatrics. clonal nature of LCH.
2007;120:e622-628. Study that prompted the current US and
European trials that combine immunotherapy with chemotherapy
Lysosomal storage diseases
for HLH treatment.
Schneider EM, Lorenz I, Muller-Rosenberger M, Steinbach Barton NW, Brady RO, Dambrosia JM, et al. Replacement
G, Kron M, Janka-Schaub GE. Hemophagocytic therapy for inherited enzyme deficiency–macrophage-targeted
lymphohistiocytosis is associated with deficiencies of cellular glucocerebrosidase for Gaucher’s disease. N Engl J Med.
cytolysis but normal expression of transcripts relevant to 1991;324:1464-1470. Classic paper describing effectiveness of
killer-cell-induced apoptosis. Blood. 2002;100:2891-2898. Report enzyme replacement therapy in Gaucher disease.
describing NK cell defects found in HLH. Grabowski GA. Gaucher disease and other storage disorders.
Stepp SE, Dufourcq-Lagelouse R, Le Deist F, et al. Perforin Hematology Am Soc Hematol Educ Program. 2012;2012:13-18.
gene defects in familial hemophagocytic lymphohistiocytosis. Review describing clinical manifestations, molecular genetics,
Science. 1999;286:1957-1959. Classic paper describing identifying and treatment of Gaucher disease and other lysosomal storage
mutations in perforin as the cause for some cases of FHL. disease.
Mistry PK, Lukina E, Ben Turkia H, et al. Effect of oral eliglustat
on splenomegaly in patients with Gaucher disease type 1: the
Langerhans cell histiocytosis
ENGAGE randomized clinical trial. JAMA. 2015;313:695-706.
Arico M, Girschikofsky M, Genereau T, et al. Langerhans cell Report describing first line treatment with substrate reduction
histiocytosis in adults. Report from the International Registry of therapy in Gaucher disease.

Bibliography | 429
Vanier MT. Niemann-Pick diseases. Handb Clin Neurol. Gaucher disease after 2 to 5 years of treatment: a report from
2013;113:1717-1721. Review of Niemann-Pick disease. the Gaucher Registry. Am J Med. 2002;113:112-119. Paper
Weinreb NJ, Charrow J, Andersson HC, et al. Effectiveness of describing long-term results of ERT therapy in Gaucher disease.
enzyme replacement therapy in 1,028 patients with type 1 Zimran A. How I treat Gaucher disease. Blood. 2011;118:1463-1471.

CHAPTER
16
Chronic myeloid leukemia and
myeloproliferative neoplasms
Ruben A. Mesa and Brady L. Stein
Introduction, 431 Polycythemia vera, 448 Special management situations across
Chronic myelogenous leukemia Essential thrombocythemia, 457 MPNs, 470
(BCR-ABL1 positive), 432 Myelofibrosis (primary and Other MPNs, 471
Classical BCR-ABL1–negative MPNs, 443 post-ET/PV), 462 Bibliography, 483

primary myelofibrosis (PMF), polycythemia vera (PV), and
multimedia component. essential thrombocythemia (ET) were grouped together.
Subsequently, CML was found to be associated in all cases
with a chromosome translocation resulting in a pathognomonic
Introduction fusion of the BCR gene on chromosome 22 and the ABL1 gene
on chromosome 9, so CML is often considered separately from
The myeloproliferative neoplasms (MPNs) are a phenotypi- the other MPN. At present, CML is the only one of the four
cally diverse group of stem cell-derived clonal disorders char- major MPNs characterized by a disease-defining genetic abnor-
acterized by proliferation of one or more of the components mality. In 2005, the discovery of the activating Janus kinase
of the myeloid lineage (ie, erythroid, granulocytic, mega- (JAK2) V617F tyrosine kinase (TK) mutation in most cases of
karyocytic, or mast cell). MPNs share several clinical and labo- ET, PV, and myelofibrosis (MF) ushered in a new era of discovery
ratory features, including a thrombotic tendency; frequent and understanding. From that point on, classification of MPNs
organomegaly (hepatomegaly or splenomegaly) caused by has transitioned to include not only traditional morphologic
sequestration of excess blood cells or abnormal proliferation assessments but also disease-associated molecular genetic classi-
of hematopoietic cells; increased metabolic rate, leading to fication. Accordingly, in 2008, the World Health Organization
constitutional symptoms; hypercellularity of the bone mar- (WHO) revised the classification of what were then known as
row due to clonal marrow hyperplasia, associated with MPDs and renamed this group of disorders as MPNs to under-
increased numbers of circulating granulocytes, red blood cells, score their clonal, neoplastic nature (Table 16-1).
or platelets; and absence of marked cellular dysplasia. Each Genetic studies have now identified clonal genetic abnor-
MPN has the potential to undergo a stepwise progression that malities involving cytoplasmic or receptor TKs in the majority
terminates in marrow failure due to ineffective hematopoiesis of MPN patients. These genetic abnormalities, most com-
caused by fibrosis or in transformation to “blast phase,” which monly translocations or point mutations, result in abnormal,
is clinically similar to acute leukemia. In recognition of the constitutively active TK signaling that leads to pathologic pro-
shared clinical, laboratory, and histological features of these liferation of myeloid precursors.
conditions, William Dameshek first used the term myeloprolif- In addition to BCR-ABL1 in CML, activating point muta-
erative disorders (MPDs) in 1951. Later, chronic myelogenous tions of JAK2 are observed in almost all patients with PV and
leukemia (now often called chronic myeloid leukemia [CML]), in a significant proportion of patients with ET, PMF, and
other myeloid disorders. Calreticulin (CALR) mutations
have been recently reported in substantial proportions of
Conflict-of-interest disclosure: Dr. Mesa: Consulting: Novartis; JAK2 V617F–negative ET and MF patients. Somatic activat-
research: Incyte, Gilead, Genentech, CTI, Promedior, NS Pharma.
Dr. Stein: Advisory board/consulting for Incyte Corporation. ing JAK2 exon 12 mutations and myeloproliferative leukemia
Off-label drug use: Not applicable. (MPL) mutations are less frequently identified mutations in
| 431

432 | Chronic myeloid leukemia and myeloproliferative neoplasms
Table 16-1 2008 WHO classification of MPNs

Myeloproliferative neoplasms (MPNs)
Chronic myelogenous leukemia, BCR-ABL1 positive easuring 11 cm below the left subcostal margin. Routine
m
Chronic neutrophilic leukemia complete blood count (CBC) showed leukocytosis (white blood
Polycythemia vera cells [WBCs] = 40 × 109/L) with predominance of neutrophils
Primary myelofibrosis and neutrophil precursors (8% myelocytes, 4% metamyelocytes),
Essential thrombocythemia normocytic anemia (hemoglobin [Hb] = 10.2 g/dL, hematocrit
Chronic eosinophilic leukemia, not otherwise specified = 35%, mean corpuscular hemoglobin = 85 fL), and a normal
Systemic mastocytosis platelet counts (platelets = 335 × 109/L). Also noted on labora-
MPN, unclassifiable tory examination were basophilia (6%) and eosinophilia (3%).
Myeloid (and lymphoid) neoplasms associated with eosinophilia A bone marrow biopsy was performed and showed a hypercel-
and abnormalities of PDGFRA, PDGFRB, or FGFR1 lular marrow (90% cellularity) with granulocytic proliferation.
Myeloid and lymphoid neoplasms associated with PDGFRA Metaphase cytogenetics showed t(9;22) (q34;q11) [20] and
rearrangement fluorescent in situ hybridization (FISH) showed the BCR-ABL1
fusion gene.
Myeloid neoplasms associated with PDGFRB rearrangement
Myeloid and lymphoid neoplasms associated with FGFR1
abnormalities
CML is a pluripotent hematopoietic stem cell neoplasm
characterized by the BCR-ABL1 fusion gene, which is usually
JAK2 V617F–negative PV and ET/MF patients, respectively. derived from a balanced translocation between the long
A spectrum of somatic mutations in genes involved in vari- arms of chromosomes 9 and 22, t(9;22)(q34;q11), resulting
ous cellullar processes have also been identified recurrently in a derivative chromosome known as the Philadelphia (Ph)
in MPNs, including in genes that regulate DNA methylation chromosome. CML accounts for 15%-20% of adult leuke-
(TET2, DNMT3A, IDH1/IDH2), histone modification mia cases. The worldwide annual incidence of CML is 1-2
(ASXL1, EZH2), RNA splicing (SF3B1, U2AF1 ZRSR2, cases per 100,000 persons, with a slight male predominance
SRSF2), signal transduction (LNK, CBL, NRAS), and tran- (male-to-female ratio, 1.3:1). However, with the advent of
scription (RUNX1, TP53). Mutations in CSF3R, which successful therapies that are enabling patients to live much
encode the granulocyte colony-stimulating factor receptor, longer than in the past, the prevalence of CML continues to
have been recently described in patients with chronic neu- increase and is projected to reach 150,000 cases by 2040. The
trophilic leukemia. In addition, systemic mastocytosis (SM) median age at diagnosis is 67 years, but CML has been diag-
frequently is associated with somatic mutations in KIT (eg, nosed in patients of all age groups. In children, CML is most
KIT D816V) and TET2 and less frequently with mutations in commonly seen in the 10- to 14-year-old age group and
DNMT3A, CBL, SF3B1, and ASXL1. accounts for only 2% of childhood leukemia. Radiation
Recognition of MPN-associated mutations is of diagnos exposure has been implicated as a risk factor; however,
tic importance and complements traditional morphological unlike other myeloid leukemias, there has been no evidence
assessments in distinguishing MPN-subtypes. Additionally, for a causal association between CML and exposure to
molecular discoveries have aided in prognostic assessment and organic solvents, industrial chemicals, or alkylating agents.
have led to development of novel therapeutics. In this chapter,
the diagnosis, pathobiology, clinical features, treatment, and
prognosis of classical and atypical MPNs are reviewed. Pathobiology
The Ph chromosome [der(22q)] initially was identified in

patients with CML at what is now the Fox Chase Cancer
Chronic myelogenous leukemia Center in Philadelphia in 1960. The t(9;22)(q34;q11) trans-
(BCR-ABL1 positive) location in CML juxtaposes the 3′ segment of the c-ABL
oncogene (normally encoding the Abelson TK) from the
long arm of chromosome 9 to the 5′ part of the breakpoint
Clinical case
cluster region (BCR) gene on the long arm of chromosome
A 60-year-old male construction worker with a history of coro- 22. The resultant hybrid oncogene is transcribed as a chime-
nary artery disease and hyperlipidemia came to see a physician ric BCR-ABL1 mRNA, which, in turn, is translated into a
for persistent fatigue of 2 months duration. He complained of functional abnormal protein. At diagnosis, characteristic
intermittent episodes of palpitations, dizziness, weight loss, and t(9;22)(q34;q11) is present in approximately 95% of CML
discomfort in the left upper quadrant of the abdomen. Physi- cases. The remaining cases have either variant translocations
cal examination was remarkable for palpable splenomegaly
involving a third and sometimes a fourth chromosome or

Chronic myelogenous leukemia (BCR-ABL1 positive) | 433
karyotypically cryptic translocations. In these cases, routine

cytogenetic analysis is unable to detect the Ph chromosome, Ph chromosome
ch 9 9q+ (22q–) ch 22
and the diagnosis relies on demonstration of either gene
fusion by FISH or the fusion transcript by reverse transcrip-
tase (RT) polymerase chain reaction (PCR).
Three separate breakpoint regions in the BCR gene are BCR-ABL BCR
associated with distinct disease phenotypes. In typical CML,

the BCR gene is interrupted at either its e13 or e14 loci (exons
ABL
12-16). Collectively, this region is referred to as the major
breakpoint cluster region (M-BCR). In a rearrangement
involving M-BCR, the 5′ BCR segments on chromosome 22 5'
are joined with the sequences from c-ABL that are 3′ from 5' M-BCR (near e1 locus)
the a2 breakpoint (a breakpoint near the 5′ end of c-ABL).
This union gives rise to hybrid transcripts called e13a2 a2 SH1
NLS
(b2a2) and e14a2 (b3a2). These transcripts are translated Actin binding M-BCR (near e13
or e14 locus)
into 210-kDa proteins, collectively known as p210 BCR-
3' m-BCR (near e19 locus)
ABL1. Importantly, the rearranged c-ABL segment here 3'
includes sequences necessary for TK activity. As a result, the ABL gene BCR gene
p210 BCR-ABL1 oncoprotein functions as a constitutively
active TK that can phosphorylate a number of cytoplasmic BCR-ABL rearrangements
substrates with other activities of the chimeric protein, lead-
ing to alterations in cell proliferation, differentiation, adhe-
sion, and survival.
Two alternative translocations involving BCR and ABL
also have been implicated in the pathogenesis of hematologic
malignancies (Figure 16-1). In one of these, a similar seg-
ment of c-ABL is transposed onto a locus of BCR that is
downstream (39) from the M-BCR locus, a region referred
to as µBCR (exons 17-20). Translocations involving µ-BCR
yield a larger fusion gene than those involving M-BCR, and
this larger fusion gives rise to a 230-kDa p230 BCR-ABL1 e1a2 e13a2 or e14a2 e19a2
protein. The p230 BCR-ABL1 product has been found in transcript transcript transcript
uncommon CML variant cases that are characterized by
chronic neutrophilia with or without thrombocytosis; these p190BCR-ABL p210BCR-ABL p230BCR-ABL
cases tend to have a more indolent disease course than CML
associated with p210 BCR-ABL1. Distinction between p230 Figure 16-1 Schematic of molecular pathogenesis of t(9;22)(q34;q11) in
BCR-ABL1 CML and chronic neutrophilic leukemia can be chronic myelogenous leukemia (CML). The 39 portion of the ABL gene
challenging. on the telomeric region of the long arm of chromosome 9 is translocated
The third type of BCR-ABL1 rearrangement juxtaposes the to the BCR gene on chromosome 22 to form the characteristic
same c-ABL segment to the minor BCR breakpoint region 22q– abnormality referred to as the Philadelphia (Ph) chromosome.
(m-BCR), which is located upstream (59) from the M-BCR Breakpoints within the ABL gene occur within intron 1b or 2, both of
(exons 1-2). The resultant smaller chimeric oncogene gener- which are 59 (upstream) to the a2 exon. The a2 and downstream exons of
ated by this rearrangement gives rise to a 190-kDa p190 BCR- ABL encode the Src homology (SH) domains of the ABL kinase,
ABL1 protein product. The p190 BCR-ABL1 transforming including the SH1/tyrosine kinase domain, DNA binding domain,
nuclear localization signal (NLS), and actin binding site. The breakpoints
protein most often is found in a portion of de novo acute lym-
on chromosome 22 occur at 1 of 3 different locations within BCR,
phoblastic leukemia (ALL) cases referred to as Ph-positive ALL
yielding hybrid oncogenes of varying length consisting of 59 BCR
(see Chapter 18). Sometimes, the p190 BCR-ABL1 product can sequences and 39 ABL sequences. Each hybrid oncogene gives rise to a
be detected in CML, either coexpressed with p210 BCR-ABL1 chimeric transcript, which encodes a fusion protein with oncogenic
(5%-10% of cases) or detected alone in atypical cases that are activity. These include p190 BCR-ABL1 (resulting from fusion at the
often associated with monocytosis. Coexpression of p190BCR- minor breakpoint or m-BCR site), p210 BCR-ABL1 gene product
ABL1 and p210BCR-ABL1 is attributed to alternative splicing (resulting from fusion at the major breakpoint or M-BCR site), and p230
of the transcript arising from the M-BCR chimeric oncogene. BCR-ABL1 (resulting from fusion at the micro breakpoint or µ-BCR site).

The leukemic clone in CML has a tendency to acquire In the peripheral blood, neutrophilia and immature circu-
additional oncogenic mutations over time. Clinically, the lating myeloid cells are hallmark features of chronic-phase
acquisition of additional cytogenetic or molecular abnor- CML (CP CML). More than 50% of patients present with a
malities is associated with progression to accelerated and WBC count of >100 × 109/L, with blasts usually accounting
blast phases of disease or resistance to TK inhibitors (TKIs). for <2% of the WBCs. Absolute basophilia is usually present,
At the chromosomal level, additional mutations are identi- and eosinophilia is common. Anemia may be present in up to
fied in 50%-80% of advanced-disease cases. These changes half of patients. Roughly 15%-35% of patients present with
include monosomy 7, t(3;21), amplification of t(9;22), tri- platelet counts >700 × 109/L, although extreme thrombocy-
somy 8, trisomy 19, and abnormalities of chromosome 17. tosis (ie, >1,500 × 109/L) is uncommon. The high cell turn-
At the molecular level, mutations in the kinase domain of over and hypercatabolic state of CML are associated with
BCR-ABL1 can emerge and represent a prevalent mecha- elevated lactate dehydrogenase (LDH) and uric acid levels.
nism of TKI resistance and subsequent treatment failure and The marrow in CP CML typically shows myeloid hyper-
disease progression. To date, greater than 80 amino acid sub- plasia and an elevated myeloid-to-erythroid ratio (often
stitutions have been identified in patients with resistance to >10:1). Bone marrow blasts are <10%. Maturation of pre-
imatinib. Moreover, the risk of acquisition of imatinib resis- cursors is normal in CML, and dysplastic features are not
tance mutations is associated with the phase of the disease. routinely found. Megakaryocytes are often smaller than nor-
In an analysis of 256 patients with various stages of CML, the mal, in contrast to large megakaryocytes that can be seen in
overall incidence of Tyrosine Kinase domain (TKD) muta- other MPNs. Marrow basophilia is noted in one-fourth of
tions was 26% in chronic phase, 44% in accelerated phase, cases. Increased reticulin fibrosis is found in 30% of the cases
73% in myeloid blast crisis, and 81% in lymphoid blast crisis. and may have negative prognostic impact on outcomes.
Importantly, identification of a BCR-ABL1 kinase domain Pseudo-Gaucher cells and sea blue histiocytes, secondary to
mutation can influence treatment selection after an inade- increased cell turnover, frequently are found. A progressive
quate response and/or TKI resistance is encountered. For symptom burden and change in laboratory characteristics
example, the T315I mutation is resistant to all TKIs, except mark progression to AP or BP CML; these abnormalities are
ponatinib, as discussed below. In addition, patients with summarized in Table 16-2.
F317L/V/I/C and V299L mutations appear resistant to dasat- A suspected case of CML can be confirmed by assays of the
inib, while patients with Y253H, E255K/V, and F359V/C/I peripheral blood to detect either the BCR-ABL1 fusion gene
are resistant to nilotinib. The V299L mutation appears to at the chromosome level or its chimeric transcripts. At diag-
confer resistance to bosutinib. nosis, the sensitivity of FISH or RT-PCR of peripheral blood
is equal to that of bone marrow.
FISH allows for identification and quantitation of the chi-
Diagnosis
meric oncogene among interphase nuclei on a peripheral
The majority of CML patients present in the chronic phase blood smear; usually, 200-500 nuclei are screened. RT-PCR
of disease, most commonly with an insidious onset, and are is carried out on peripheral blood-derived RNA and is an
diagnosed based on abnormalities observed on CBC. Com- extremely sensitive technique; RT-PCR can detect the BCR-
mon symptoms at presentation can include fatigue, night ABL1 transcript in <1 of 105 cells. Both methods can detect
sweats, weight loss, and gout attacks. Many patients also “masked” or cryptic chromosomal translocations that are
present with splenomegaly (50%-90%) at diagnosis, which missed by conventional cytogenetics in ~5% of cases. FISH
may be symptomatic, and painless hepatomegaly may be has the advantage of identifying unusual variant rearrange-
present in up to half of the patients. Thrombotic and ments that are outside the regions amplified by the RT-PCR
hemorrhagic complications are relatively infrequent (<5%), primers. The RT-PCR method, unlike FISH, can differenti-
although purpura is a common complaint. Hyperleukocyto- ate between the fusion genes encoding the p210 BCR-ABL1
sis alone does not routinely cause symptoms because of the product and the p190 BCR-ABL1 product. Because of the
relative maturity of the leukemic cells and their smaller size lower cost and ability to discriminate the breakpoints, RT-
compared with the immature, large, poorly deformable blast PCR is becoming the preferred molecular assay for CML
cells seen in acute leukemia; however, in rare cases, males diagnosis. Quantitation of BCR-ABL1, by RT-PCR, is not
with very high WBC counts can present with leukostasis- particularly helpful at the time of diagnosis but becomes
related priapism. A progressively severe symptom burden, important for clinical decision making and monitoring of
marked by constitutional symptoms including fever, night minimal residual disease in patients on therapy. Because the
sweats, weight loss, bleeding, and worsening splenomegaly false-positive rate of FISH for BCR-ABL1 fusion gene is ~3%
may herald onset of accelerated-phase or blast-phase CML, (due to random overlay of chromsomes in examined cells,
defined below. resulting in an occasional probe fusion signal suggestive of

Table 16-2 Clinicopathological features in chronic-, accelerated-, and blast-phase CML
Symptom Blood and bone marrow findings (WHO classification)

Chronic phase
Fatigue Neutrophilic leukocytosis with immaturity
Weight loss Peripheral blasts <10%
Nocturnal sweats Thrombocytosis
Left upper-quadrant abdominal pain Basophilia and/or eosinophilia
Early satiety Normocytic anemia
Palpitations and/or dyspnea BCR-ABL1 rearrangement (usually p210 BCR-ABL1, may be b3a2 or b2a2 variants or both)
Bleeding/bruising High LDH
Priapism Hyperuricemia
Marrow myeloid and megakaryocytic hyperplasia, mild/moderate fibrosis, <10% blasts,
minimal dysplasia, t(9;22) +/- other abnormalities
Accelerated phase (several definitions exist)
Unexplained fever or bone pain, Increasing WBC count unresponsive to therapy
progressive weight loss, and sweats
Increasing spleen size (can also result in Peripheral blood basophils ≥20%
splenic infarction)
Persistent thrombocytopenia (<100 × 109/L) unrelated to therapy, or persistent thrombocytosis
(>1,000 × 109/L) unresponsive to therapy
Blasts 10%-19% of WBCs in peripheral blood and/or nucleated bone marrow cells
Cytogenetic evidence of clonal evolution
Blast phase
Bleeding, bruising Blasts ≥20% of peripheral blood white cells or of nucleated bone marrow cells
Infections Extramedullary blast proliferation
Prominent constitutional symptoms
Large foci or clusters of blasts in the bone marrow biopsy
Massive splenomegaly
Tissue manifestations of extramedullary
disease
BCR-ABL1 fusion), quantitative RT-PCR is the preferred to accelerated-phase or blast crisis CML. Before the develop-
technique for monitoring disease response. ment of TKIs, patients with CML who did not undergo stem
Although a positive RT-PCR or FISH assay confirms the cell transplantation (SCT) had a median survival of roughly 5-7
diagnosis of CML, a complete staging of the disease still years, and 30% of patients survived beyond 10 years. This has
requires a marrow evaluation (Table 16-2). Most investigators dramatically changed in the era of the TKIs. With 8 years of
believe a marrow sample is still necessary at diagnosis in order follow-up of patients randomized to imatinib in the study that
to assess the percentage of undifferentiated blasts and to eval- led to its regulatory approval in 2001, the overall survival was
uate for the presence of additional cytogenetic abnormalities, 93% if only CML-related deaths were considered.
such as trisomy 8, a second Ph, isochrome 17q, or trisomy 19, Before the development of TKIs, multivariate prognostic
which adversely influence prognosis. Conventional cytoge- models (eg, the 1984 Sokal score including age, spleen size,
netic studies identify a Ph chromosome in 90%-95% of cases; platelet counts, and % blasts) and the 1998 Hasford (Euro)
more than half of the karyotypically negative cases will have a score (added eosinophil and basophil percent to Sokal score)
detectable BCR-ABL1 rearrangement by molecular assay. The were useful to help identify high-risk patients. Recent studies
clinical course of the BCR-ABL1-positive, Ph-negative patient suggest that these older prognostic models still can predict
is identical to that of patients with Ph-positive CML. the probability of achieving a cytogenetic remission in
patients treated with imatinib mesylate and second-generation
Disease course and prognosis TKIs. As opposed to the Sokal and Hasford score, the EUTOS
score was developed in imatinib-treated patients, and
Chronic-phase CML (CP CML)
includes spleen size and basophil percent in its estimation of
Historically, patients diagnosed with the chronic phase of CML the likelihood of achieving complete cytogenetic response in
remained stable for an average of 3-5 years before progressing imatinib-treated patients.

While these scoring systems are useful, in the current era The median survival is 3-6 months for older patients and
of TKIs, the most important prognostic indicators are phase roughly 8 months for younger adults; patients with lymphoid
of disease at diagnosis and the speed and depth of response blast crisis survive 4-5 months longer on average than those
to TKI therapy. Appropriate monitoring strategies continue with myeloid blasts. In blast phase, the presence of >50%
to evolve, and are discussed subsequently. blast cells in the blood and cytogenetic progression have
been identified as independent predictors of worse survival.
Accelerated-phase CML (AP CML) Deaths usually are due to metabolic derangements, infection,
bleeding, and extramedullary leukemic infiltration.
The AP CML is accompanied by the acquisition of additional
molecular lesions, genomic instability, and progressive
impairment of myeloid cell differentiation. This latter fea- Management
ture leads to the accumulation of immature precursors and Chronic phase
undifferentiated blasts in the marrow, blood, and extramed- Imatinib mesylate
ullary tissue. The clinical symptoms associated with the
accelerated phase (Table 16-2) may be minor, delayed, or The promise of targeted therapy for CML was realized with
completely absent. the regulatory approval of the first small-molecule TKI for
The median survival from the onset of accelerated phase, cancer, imatinib mesylate, in May 2001. Imatinib binds the
without an SCT or TKIs, is only 12-18 months. Death occurs adenosine triphosphate (ATP) binding site in the catalytic
predominantly because of transformation to blast phase domain of the BCR-ABL1 oncoprotein and inhibits the BCR-
with the associated life-threatening complications of marked ABL1 TK activity. This interaction prevents the transfer of
leukocytosis and complete failure of normal hematopoiesis. phosphate groups to tyrosine residues on substrate molecules
Various clinical and laboratory criteria have been proposed involved in downstream signal transduction pathways.
for defining entry into accelerated phase, some of which The drug also interferes with the TK activities of normal
have been identified by multivariate analyses as prognosti- ABL and with the kinase activity of the ARG, PDGFRA,
cally useful. Patients with karyotypic evolution alone, except PDGFRB, and KIT TKs. These actions are useful for the
for the acquisition of chromosome 17 abnormalities, with- treatment of other hematopoietic (eg, SM without KIT
out accompanying clinical and other laboratory changes, mutations, chronic eosinophilic leukemia [CEL]) and non-
may follow a more indolent course to blast transformation. hematopoietic (eg, gastrointestinal stromal tumor) disor-
ders; the first two are discussed later in this chapter.
The pivotal phase 3 study comparing imatinib to the com-
Blast-phase (blast crisis) CML (BP CML)
bination of interferon α (IFNα) and cytarabine (the Interna-
Progression of CML to acute leukemia, synonymous with tional Randomized Interferon and STI571 [IRIS] trial)
“blast phase” and “blast crisis,” evolves most commonly demonstrated the superiority of imatinib compared with
from a preceding accelerated phase and is reached when the IFNα plus cytarabine, with higher rates of complete hemato-
proportion of blasts in the blood or marrow is >20% (Table logic response (CHR), major cytogenetic response (MCyR),
16-2). Sudden transformation to blast phase in patients tak- and complete cytogenetic response (CCyR); freedom from
ing imatinib is observed with an annual rate of 1%-2%. progression to AP or blast crisis CML; and better tolerance of
Myeloid lineage markers (eg, CD33, CD13, CD14, and therapy. The majority of patients who were at high risk
CD15) are expressed by the blast cells in more than half of according to current prognostic models achieved MCyR at a
the cases of BP CML. Up to one-third express B-cell-precur- rate of 69%-78.9% in the imatinib arm. An 8-year follow-up
sor lymphoid markers (eg, CD10, CD19, and CD20). Undif- report provided long-term efficacy and safety data on 553
ferentiated acute leukemia and cases displaying both myeloid patients who were randomized to the first-line imatinib arm
and lymphoid cell surface markers account for the remain- of the IRIS study. The rate of major molecular response
der. Most CML cases express the p210 BCR-ABL1 gene (MMR) was 86%. None of the patients who achieved MMR
product, and only rare cases are associated with p190 BCR- at 12 months progressed to AP or BP CML. The estimated
ABL1 alone. Thus, a case of Ph-positive ALL that subse- overall survival and event-free survival at 8 years were 85%
quently is found to be associated with p210 BCR-ABL1 and 81%, respectively. The estimated freedom from progres-
might actually represent previously unrecognized CML pre- sion to AP or BP CML was 92%. There were low yearly rates
senting in lymphoid blast crisis. The clinical and laboratory of progression to AP or BP CML in years 4-8 after starting
features of BP CML are summarized in Table 16-2. Cytoge- imatinnib treatment (0.9%, 0.5%, 0%, 0%, and 0.4%). Only
netic abnormalities in addition to t(9;22) are found in 3% of patients who achieved CCyR progressed to AP or BP
65%-80% of cases of blast phase disease. CML. More patients (55%) remained on imatinib therapy

while the remainderhave discontinued treatment due to randomized phase 3 study, 4 treatment arms were compared,
insufficient therapeutic outcomes (16%), side effects (6%), including imatinib 400 mg daily, imatinib 600 mg daily, ima-
death (3%), and other reasons (17%). tinib 400 mg daily with pegylated IFN, and imatinib 400 mg
Despite impressive results with imatinib, several attempts with subcutaneus cytarabine. While rates of CCyR were sim-
have been made to improve response rates and decrease resis- ilar between these treatment arms, higher rates of superior
tance in newly diagnosed patients through the use of higher molecular responses (4 log reduction) were noted in the
doses of imatinib (600-800 mg/d). The rationale for use of pegylated-IFN group compared to imatinib alone (30% vs
higher-dose imatinib is based in part on interpatient variabil- 14%; P = .001); however, the long-term impact is not known
ity of drug uptake into target hematopoietic cells, which and rash and depression were more common consequences.
depends on human organic cation transport -1 (OCT-1) Adding pegylated IFN alfa-2b also resulted in higher rates of
activity. Accordingly, prior studies have shown superior long- MMR compared to imatinib alone (82% vs 54%, P = .002),
term outcomes, including higher rates of MMR, overall sur- but IFN discontinuation rates were 61% in the combination
vival, and event-free survival and lower rates of kinase domain group. Another large study of newly diagnosed CML patients
resistance mutations in those patients with higher (than could not confirm the benefit of adding IFN alfa to imatinib
median) OCT-1 activity compared to low OCT-1 activity. 400 mg daily. Currently, neither high-dose imatinib nor
Therefore, measurement of OCT-1 activity represents a preimatinib in combination with IFN are recommended front-
therapy assessment of imatinib responsiveness; those with low line treatments and would be considered investigational.
activity appear at risk for suboptimal response with standard
dose imatinib. Phase 2 studies demonstrated that higher-dose Toxicity Adverse effects include myelosuppression, fatigue,
imatinib yields higher rates of CCyR and MMRs at earlier gastrointestinal disturbances such as nausea and diarrhea,
time points for low- or intermediate-Sokal-risk newly diag- edema (periorbital and peripheral), and muscle cramps.
nosed CML patients. A phase 3 clinical trial of high-Sokal-risk Long-term consquences may rarely include hypophosphate-
CML conducted by the ELN however, showed no significant mia and decrease in bone mineral density. Cardiotoxicity,
difference in CCyR and MMR rates at 12 months between including congestive heart failure, is rare, unlike other TKIs
patients treated with 400 mg/d versus 800 mg/d of imatinib. described below. In many patients who experience unac-
The phase 3 Tyrosine Kinase Inhibitor Optimization and ceptable adverse effects, transient dose reduction or treat-
Selectivity (TOPS) prospective randomized trial, which comment interruption with supportive care allows for patients to
pared high-dose (800 mg/d of imatinib) and standard-dose resolve adverse effects and resume full-dose or modified
imatinib (400 mg/d of imatinib), showed a higher rate of therapy.
CCyR and MMR at 6 months but did not differ at 12 months.
Half of the patients in the high-dose imatinib had to undergo
Dasatinib
dose reduction to less than 600 mg/d of imatinib, and there
was a higher rate of grade 3 and 4 hematologic adverse events Dasatinib, which lacks structural similarity to imatinib, has
in the highe- dose imatinib arm. No significant differences in activity against Src family kinases in addition to ABL kinases.
CCyR or MMR were noted between the two treatment arms Dasatinib does not rely on a conformational change of ABL
based on Sokal risk scores. Responses were achieved sooner in for binding and thus appears to be less susceptible to the
patients who received the 800 mg/d of imatinib, but longer development of resistant kinase domain mutations that alter
follow-up will be necessary to determine the significance of ABL conformation. Dasatinib is approved for the treatment
this finding. Finally, the TIDEL-II study evaluated higher dose of adults with newly-diagnosed CP CML and CP CML with
imatinib (600 mg daily), followed dose escalation to 800 mg resistance or intolerance to prior therapy.
daily if the plasma trough at day 22 was reduced. Thereafter, if
patients failed to meet molecular targets (BCR-ABL1 ≤10%, Frontline therapy The phase 3 randomized open-label
≤1%, and ≤0.1% at 3, 6, and 12 months), they were either trial, Dasatinib versus Imatinib study in Treatment-Naive
increased to imatinib 800 mg daily and later nilotinib if failing CML-Chronic Phase (DASISION) showed that CML
the same target (cohort 1), or directly to nilotinib (cohort 2). patients on 100 mg/d of dasatinib achieved higher CCyR
At 2 years, 55% and 30% remained on imatinib and nilotinib, (77% vs. 66%) at 12 months and MMR (46% vs. 28%) com-
respectively, only 12% failed a target at 3 months, and MMR pared with patients treated with imatinib 400 mg/d. Further-
was 73% (4.5 log reduction 34%). This novel study suggests more, the rates of accelerated- and blast-phase progression
that an imatinib-based initial therapy approach with an early were less in the dasatinib compared with the imatinib treat-
switch is practical and effective. ment arm (1.9% vs. 3.5%). Similar response rates and
Another attempt to improve imatinib response rates decreased toxicity have been demonstrated with dasatinib
involved the addition of pegylated IFN to imatinib. In one 100 mg orally once daily. Data from the 3-year follow-up of

patients enrolled in the DASISION trial showed that CCyR been approved for the treatment of newly diagnosed CP
rates between dasatinib- and imatinib-treated patients were CML and CP CML in adult patients resistant or intolerant to
87% vs. 83%, but the median time to CCyR was shorter in prior therapy.
dasatinib-treated patients (3.1 months vs. 5.8 months). In a
landmark analysis, dasatinib was associated with a higher Frontline therapy In the phase 3 randomized open-label
percentage of patients achieving BCR-ABL1 levels ≤ 10% at 3 trial, Evaluating Nilotinib Efficacy and Safety in Clinical
months (84% vs. 64%) and associated with improved pro- Trials—Newly Diagnosed Patients (ENESTnd), nilotinib
gression-free survival. The cumulative MMR rates favored (300 mg twice daily or 400 mg twice daily) was compared
dasatinib (69% vs. 55%) and deeper responses by 4.5 log with 400 mg/d of imatinib. CML patients on 300 mg or 400 mg
reduction of BCR-ABL1 were also higher (22% vs. 12%). twice daily of nilotinib had superior CCyR in 12 months
There were also fewer transformations to AP or BP CML in compared with patients treated with imatinib 400 mg/d
the dasatinib- versus imatinib-treated group (3% vs. 5%). (80% and 78% vs. 65%). The time to progression to AP or
The 3 year OS was 93.7% and 93.2%, respectively, in dasat- BP CML was better with the nilotinib-treated patients. Data
inib and imatinib-treated patients. A very high response rate from the 36-month follow-up showed superiority of nilo-
(36%) of 4.5-log major molecular remission by 12 months tinib 300 mg or 400 mg orally twice daily compared with 400
was observed when combining dasatinib with IFN, com- mg orally once daily of imatinib in terms of rates of MMR
pared with <11% with imatinib, nilotinib, or dasatinib alone defined as a ≤0.1% of BCR-ABL1:ABL1 ratio on the interna-
at that time point. tional scale (IS) (73% and 70% vs. 53%), deeper molecular
response 4 log reduction (MR4.0) (50% and 44% vs. 26%),
Second-line therapy (after imatinib resistance or rates of AP or BP CML progression (2 patients [0.7%] and 3
intolerance) Dasatinib was first investigated in CP CML patients [1.1%] vs. 12 patients [4.2%]). The estimated 3-year
patients with resistance or intolerance to imatinib. In one of overall survival were not statistically significantly different
the clinical trials (START-C), 90% of CP CML patients with among the three groups (95%, 97%, and 94%), but the
resistance or intolerance that were switched to dasatinib authors reported better overall survival for those treated
therapy (70 mg orally twice daily, average daily dose approx- with nilotinib compared with those treated with imatinib, if
imately 100 mg) achieved CHR (median follow-up, 15 only CML-related deaths were considered (98.1% vs. 98.5%
months), and MCyRs and CCyRs were noted in 59% and vs. 95.2%, HR = 0.35, P = .0356). In another analysis of the
49% of patients, respectively. Although no responses were three-year data, treatment-emergent mutations were less
seen in patients with the T315I mutation, disease control was frequently observed in nilotinib-treated patients (11 patients
noted across all other TK mutations. The MMR rate at 12 in each nilotinib-dosing arm), compared to imatinib-treated
months was 25%. Progression-free survival at 15 months patients (21 patients).
was 90%, and overall survival was 96%. Long-term follow-
up (6 years) from a dose optimization study of dasatinib in Second-line therapy (after imatinib resistance or
imatinib-resistant or intolerant CP-CML patients demon- intolerance) Like dasatinib, nilotinib has also demon-
strated sustained benefit, including a 43% MMR rate and strated significant clinical activity and an acceptable safety
estimated OS of 71% in those treated with 100 mg daily. and tolerability profile in patients with imatinib-resistant or
Finally, dasatinib was also shown to be superior to high dose intolerant CP CML, except in those who carry the T315I
imatinib (800 mg) in those with prior imatinib resistance mutation. Four year follow-up from an international phase
(400 mg and 600 mg doses) regarding all endpoints (CHR, 2 study of CP CML resistant/intolerant patients treated with
MCyR, CCyR, MMR, and progression-free survival). nilotinib revealed that 59% achieved a major cytogenetic
response, 45% achieved a CCyR, and OS was estimated at
Toxicity Myelosuppression is common compared to ima- 78%. Deeper responses at 3 and 6 months correlated with
tinib. Unique toxicities include pleural effusion, suggesting later outcomes, including OS. In an expanded access, open-
that those with lung disease, congestive heart failure, and label study of 1,422 patients who failed prior imatinib,
hypertension may not tolerate this agent. Other unique but CCyR was attained in 34% of nilotinib-treated patients. In
uncommon toxicities include pulmonary hypertension and another study of patients in CCyR, but with detectable BCR-
platelet dysfunction. ABL1 after more than 2 years on imatinib, patients random-
ized to nilotinib had higher rates of undectable BCR-ABL1
compared to those randomized to imatinib at 2 years (22.1%
Nilotinib
vs 8.7%, P = 0.0087); deeper responses (MR4.5) at 2 years
Nilotinib is a structural derivative of imatinib that is a were also more commonly observed in nilotinib-treated
30-fold more potent inhibitor of BCR-ABL1 activity and has patients.

Toxicity Unique toxicities include hyperglycemia, hyperli- follow-up in those with imatinib-resistance/intolerance,
pasemia, hyperbilirubinemia, and QT prolongation. Vascu- 59% achieved and maintained MCyR, and 35% achieved
lar events including peripheral arterial occlusive disease, MMR; estimated OS was 91%. The major cyogenetic
coronary syndromes, and cerebrovascular accidents have response rate was 26.9% for those who received prior ima-
also been reported in patients with and without cardiovascu- tinib followed by nilotinib or dasatinib; with a minimum of
lar risk factors. 28.5 months follow-up of 118 patients, MCyR and CCyR
were achieved in 32% and 24% and 2 year estimated OS was
Bosutinib 83%. Transformation to AP/BP CML occurred during the
treatment course in five patients.
Frontline therapy Bosutinib, a dual Src/Abl kinase inhib-
itor, is not currently FDA approved for the frontline treat-
Toxicity Unique toxcities are primarily gastrointestinal,
ment of CP CML because the phase 3 Bosutinib Efficacy and
including diarrhea, nausea, vomiting, and transaminitis;
Safety in Newly Diagnosed Chronic Myeloid Leukemia
rash has also been reported.
(BELA) trial, which compared bosutinib with imatinib in
newly diagnosed CP CML, did not achieve its primary end-
point rate of CCyR at 12 months. This study included 502 Ponatinib
newly diagnosed CML patients who were randomly assigned
The T315I mutation results in resistance to TKI therapy
(1:1) to bosutinib 500 mg orally once daily or imatinib 400
because it causes a structural change that closes the ATP-
mg orally once daily. There was no difference in the rate of
binding pocket of BCR-ABL1 so that drugs cannot enter and
CCyR at 12 months between bosutinib compared to ima-
inhibit the kinase domain. Typically, patients with TKI resis-
tinib (70% vs. 68%, P = .601). The BELA trial did show that
tance due to an acquired T315I mutation should be consid-
the MMR rate at 12 months was higher with bosutinib ver-
ered for alternative therapies, including investigational
sus imatinib (41% vs. 27%, P = 0.001). Bosutinib treatment
agents, allogeneic SCT, or IFN therapy depending on the
also resulted in a faster time to CCyR and MMR compared
patient’s age, comorbidity profile, and donor availability.
with imatinib. There were also fewer on-treatment transfor-
Recently, a third generation oral pan-BCR-ABL1 TKI, pona-
mations to AP or BP CML and fewer CML-related deaths
tinib, has shown activity in patients with CML with T315I
with bosutinib. Bosutinib was associated with more frequent
mutations. In the phase 2 PACE trial, refractory CP, AP, and
gastrointestinal and liver-related side effects, whereas neu-
BP CML or Ph-positive ALL resistant or intolerant to dasat-
tropenia, edema, and musculoskeletal problems were more
inib or nilotinib, or with the resistant T315I mutation, were
common with imatinib. Recently, 24 month follow-up was
treated with ponatinib (45 mg orally once daily). A total of
presented, showing similar cumulative rates of CCyR in
88% of the patients in the cohort had resistance to either
bosutinib and imatinib-treated patients (79% and 80%,
dasatinib or nilotinib. Of 267 CP CML patients, 56% attained
respectively). Cumulative rates of MMR were 59% and 49%,
MCyR, (51% with resistance or intolerance to dasatinib or
respectively. Early response (BCR-ABL1 ≤ 10%) predicted
nilotinib and 70% with the T315I mutation) 46% achieved
better long-term outcomes in both treatment groups,
CCyR (40% and 66% of those with resistance/intolerance
responses were durable, and discontinuations were more
and T315I mutation, respectively), and 34% attained MMR
commonly noted with bosutinib (transaminitis).
(27% and 56%). The median time to MCyR was 2.8 months,
and the rate of sustained MCyR for 12 months was 91%.
Second-line therapy (after imatinib-resistance or
intolerance) Bosutinib was FDA approved for the treat-
ment of adult patients with CP, AP, or BP CML who are Toxicity In the PACE study, the most common adverse events
resistant or intolerant to imatinib, based on a single-arm included thrombocytopenia, rash, dry skin, and abdominal
open-label multicenter study of CP, AP, and BP CML pain. Serious arterial thrombotic events were noted in 9%, of
patients who received at least one prior TKI (either imatinib which 3% were thought to be treatment-related. Continued
or imatinib followed by nilotinib or dasatinib). A total of 546 follow-up revealed a significant increase in the frequency
patients were enrolled of which 73% were imatinib-resistant (>27%) of arterial and venous thrombosis, as well as occlusive
and 27% were imatinib-intolerant. The efficacy endpoint of events leading to brief suspension of marketing/sales. Ponatinib
the study was MCyR at week 24 for CP CML patients and was released from this suspension in January 2014, but is now
confirmed CHR and overall hematologic response (OHR) part of risk evaluation and mitigation (REMS) strategy with
by week 48 for AP or BP CML patients. In CP CML patients, restricted use and careful monitoring recommended. It is not
MCyR at week 24 was 33.8% for those who only received yet clear if the thrombotic risk is dose dependent.
prior imatinib therapy. With a minimum of 2 years of A schema for treating CP CML is shown in Figure 16-2.

identified using the international scale (IS), or ≥3 log

Diagnosis of chronic-phase CML: reduction of the BCR-ABL1 mRNA if IS is not used. Com-
frontline therapy
plete molecular responses are achieved if there is no dete
ctable BCR-ABL1 by quantitative PCR (qPCR) using a
Decision making: highly sensitive assay (≥4.5 log reduction from standard-
CML risk category, comorbidities, concurrent medications, ized baseline).
patient preference, cost
Several studies have suggested that achievement of early
molecular response (EMR, defined as <10% BCR-ABL1
Imatinib Dasatinib Nilotinib transcript level reduction at 3 months of first-line TKI ther-
apy) is associated with durable long-term responses and sur-
vival in CP CML patients. One study involving 282 patients
Intolerance, *resistance, or loss of response showed that those with >9.84% BCR-ABL1 transcript levels
Yes? Yes? Yes? at 3 months have significantly inferior 8-year probabilities of
overall survival, progression-free survival, and cumulative
incidence of CMR and CCyR compared with those who have
Switch to another TKI: lower transcript levels. Similarly, another study reported that
**mutational profile, comorbidities
>10% BCR-ABL1 transcript level by IS at 3 months after
imatinib treatment correlated with worse 5-year overall sur-
Nilotinib or Nilotinib Dasatinib vival compared with patients with BCR-ABL1 levels of
dasatinib Bosutinib Bosutinib 1%-10% and <1%.
Bosutinib Ponatinib Ponatinib Guidelines regarding milestones for response and recom-
Ponatinib (T315I mutation) (T315I mutation)
(T315I mutation) mendations for monitoring have been created by the Euro-
pean LeukemiaNet (ELN) and the National Comprehensive
Cancer Network (NCCN). ELN response criteria for first-
Figure 16-2 Failure per ELN criteria includes absence of CHR at
3 months, or Ph+ >95%; BCR-ABL1 >10% and/or Ph+ >35% at line TKI therapy are shown in Table 16-3. Per the NCCN,
6 months; and BCR-ABL1 >1% and/or Ph+ >0 at 12 months. The BCR-ABL1 ≤10% is a response milestone at 3 months, but
NCCN recommendations suggest that BCR-ABL1 > 10% at failure to achieve EMR does not mandate therapy change;
3 months is an indicator to adjust therapy. BCR-ABL1 kinase the same is true for the 6 month evaluation (pCyR if molecu-
domain testing is indicated in cases of resistance or when a lar monitoring is not available). Complete cytogenetic
confirmed loss of response occurs. Results from mutational testing response is expected milestone at 12 and 18 months; the
can inform choice of the next TKI, as the T315I is resistant to all NCCN panel did not consider absence of MMR as a failure if
TKI’s except ponatinib. CCyR was attained.
ELN 2013 guidelines currently recommend monitoring
qPCR from the peripheral blood every 3 months until a
Response criteria and monitoring of TKI therapy
patient achieves MMR, then each 3-6 months. If cytogenet-
The development of TKIs has completely changed the stan- ics are used to assess response to therapy, these should be
dard approach for all phases of CML, and response to these repeated at 3, 6, and 12 months until a CCyR is attained.
therapies has a substantial impact on prognosis. As such, NCCN guidelines also recommend monitoring with qPCR
response to therapy and many clinical trial endpoints are each 3 months and after CCyR has been attained, each 3
measured by meeting certain milestones. Criteria for com- months for 3 years, then each 3-6 months indefinitely. If
plete hematological response (CHR) includes resolution of patients do not achieve these responses, they should undergo
symptoms and signs of the disease, including palpable sple- full restaging; and if their clinical situation worsens, muta-
nomegaly, normalization of peripheral counts, with leuko- tional testing and modification of therapy should be consid-
cytes <10 × 109/L and absence of immaturity (myelocytes, ered. Per the NCCN guidelines, BCR-ABL1 kinase domain
promyelocytes, blasts, etc), and platelets <450 × 109/L. mutation testing is recommended for CP CML patients who
Complete cytogenetic response (CCyR) is achieved if there failed to achieve PCyR or BCR-ABL1/ABL1 ≤10% by IS at 3
are no Ph-positive metaphases, whereas partial CyR is months, CCyR at 12 and 18 months, or any loss of response
characterized by 1%-35% Ph-positive metaphases, major (hematologic relapse, cytogenetic relapse or 1 log increase in
CyR (MCyR) by 0%-35% Ph-positive metaphases (com- BCR-ABL1 transcript levels and MMR loss). Also, another
plete + partial), and minor responses by >35% Ph-positive important aspect of monitoring includes assessments of
metaphases. A major molecular response (MMR) is patient compliance to therapy, which may uncover the etiol-
attained if BCR-ABL1 transcriptions of 0.1% or less are ogy for an apparent treatment failure.

Table 16-3 Expected milestones and response to first-line TKI therapy per ELN 2013 recommendations
Time Optimal Warning Failure

Diagnosis Not applicable High-risk CML, or with additional NA
cytogenetic abnormalities
3 months MCyR (Ph-positive metaphases Less than MCyR(Ph-positive metaphases No CHR and/or Ph-positive
≤35%) and/or BCR-ABL1 ≤10% 36-95%) and/or BCR-ABL1 >10% metaphases > 95%
6 months CCyR (Ph-positive metaphases: Less than CCyR (Ph-positive metaphases Less than MCyR (Ph-positive metaphases
0%) and/or BCR-ABL1 <1% 1%-35%) and/or BCR-ABL1 1%-10% >35%) and/or BCR-ABL1 >10%
12 months MMR (BCR-ABL1 ≤0.1%) Less than MMR (BCR-ABL1 >0.1%-1%) Less than CCyR (Ph-positive metaphases >
0%) and/or BCR-ABL1 >1%
After 12 months BCR-ABL1 ≤0.1% Additional cytogenetic abnormalities in Loss or CHR, CCyR, or MMR, or
Ph-negative cells (abnormalities of acquisition of kinase domain
chromosome 7) mutations or additional cytogenetic
abnormalities
Optimal responses correlate with favorable long-term outcomes, and no treatment change recommended. Warning suggests a need for more
frequent monitoring to identify any need to change therapeutic strategies. Failure requires a change in therapeutic strategy.
Adapted from Table 5 in Baccarani M, et al. Blood. 2013:122;872-884.
ELN = European LeukemiaNet; CCyR = complete cytogenetic response; CyR = cytogenetic response; CHR = complete hematologic response;
MMR = major molecular responses; NA = not applicable; PCyR = partial cytogenetic response; Ph = Philadelphia chromosome.
Discontinuation of TKIs in responding patients Finally, a multicenter prospective study (A-STIM) of 80

patients that discontinued imatinib after prolonged CMR
Currently, TKI therapy is continued indefinitely as long as
(median 41 months) observed that 29 (36%) lost MMR after
tolerated and as long as desired responses are achieved and
a median of 4 months discontinuation. The investigators
sustained because disease relapse can occur if imatinib ther-
suggested loss of MMR was a practical and safe criterion
apy is discontinued. There are ongoing studies that attempt
indicating a need to restart treatment. In this study, treat-
to answer the question of whether TKI therapy can be dis-
ment free remission was estimated at 61% at 36 months,
continued safely in some CML patients. The study by Ross,
and in those that required retreatment, second CMR was
et al. (TWISTER) showed that at 24 months, the estimate
obtained in approximately 7 months. Reports that piogli-
of treatment-free remission was 47%; relapses typically
tazone, an inhibitor of PPARγ that influences STAT3 signal-
occurred within 4 months of imatinib discontinuation. Nei-
ing, can induce complete molecular remissions in patients
ther disease progression nor acquisition of kinase domain
with persistent low-level residual disease, and that this drug
mutations were observed in those with molecular relapse,
or immunotherapy approaches can potentially eradicate
and relapsed patients were sensitive to reinitation of ima-
quiescent leukemia stem cells, has prompted development of
tinib. In a prospective, multicenter, nonrandomized study
trials to see if combining agents with TKIs may allow TKI
called Stop Imatinib (STIM), imatinib therapy was stopped
discontinuation in a broader range of patients. Currently,
in patients who previously were treated with imatinib for >2
discontinuation of TKI therapy is still considered investiga-
years, who were ≥18 years of age, and who achieved CMR
tional, and additional studies are needed.
defined as a >5-log reduction in BCR-ABL1 and ABL1 levels
and undetectable transcripts by quantitative RT-PCR. The
TKIs in accelerated and blast phase CML
median follow-up of the study was 17 months. Of the 69
patients who had at least a 12-month follow-up (median = In a published phase 2 study, HRs occurred in approximately
24 months), 39% remained in CMR, while 58% had disease 80% of imatinib-treated patients with AP CML, while CHR,
relapse within 6 months after stopping imatinib therapy. All MCyR, and CCyR occurred in 53%, 24%, and 17% of
patients who had a molecular relapse, however, responded patients, respectively. Overall survival and disease progres-
when retreated with imatinib. Unusually, many patients in sion rates at 12 months were optimal among patients receiv-
STIM had previously been treated with IFN, which may have ing 600 mg/d (78% and 44%, respectively). Toxicity was
influenced their response. Prognostic factors predictive of acceptable. Furthermore, imatinib therapy in accelerated-
maintenance of CMR after imatinib discontinuation phase disease can serve as a bridge to SCT.
included low Sokal risk, longer duration of imatinib therapy Imatinib can transiently control CML blast crisis in a pro-
(≥50 months), and male sex. portion of patients. Both lymphoid and myeloid phenotypes

respond, and optimal results are achieved with a dose of Additional treatment strategies
600 mg/d, as for accelerated-phase disease. Imatinib induced
Other TKIs
OHRs in ~50% of study subjects, 8%-21% achieved CHRs,
and ~30% achieved stable or sustained hematologic res Several novel TKIs are under development, including
ponses (lasting ~4 weeks). MCyRs occurred in 16% of ABL001, which binds to a different part of the ABL kinase
patients, and CCyRs occurred in 7% of patients. The median (myristoyl site) than the five approved ATPase domain–
overall survival for patients who achieved a sustained hema- binding TKIs. Patterns of resistance to this molecule are
tologic response was 19 months. Myelosuppression was expected to be distinct from those to the other TKIs.
common, and nonhematologic toxicities were mild to mod-
erate and seldom required discontinuation of therapy. The
Omacetaxine
remarkable and encouraging results for blast crisis with this
well-tolerated single agent have led to ongoing studies com- Omacetaxine, a protein translation inhibitor formerly
bining imatinib with conventional acute leukemia chemo- known as homoharringtonine, was approved by the FDA in
therapy regimens. 2012 for patients with CP or AP CML and with resistance or
Dasatinib, at a dose of 140 mg/d, led to CHR, MCyR, and intolerance to ≥2 TKIs. This approval was based on a trial
CCyR in 45%, 39%, and 32% of patients with AP CML, with MCyR rates of 20% in CP CML, and MHR of 27% in
respectively. Responses were achieved irrespective of imatinib AP CML. The final analysis, with 24 months follow-up,
status (resistant or intolerant). The 12-month progression- reported a MCyR and median OS of 18% and 40.3 months,
free survival and overall survival rates were 66% and 82%, respectively in those with CP CML; 14% of patients with AP
respectively. In another study, a subgroup of patients with AP CML achieved MHR, for a median of 4.7 months. The most
CML randomized to 140 mg once daily or 70 mg twice daily common toxicities were hematological, with ≥ grade 3
experienced comparable rates of major hematologic response adverse events in 79% and 73% of CP and AP CML patients.
(MHR) (66% vs 68%) and MCyR (39% vs 43%), but once- The requirement for frequent subcutaneous administration
daily dosing associated with a more favorable safety profile. has limited use of this agent.
Two-year follow-up from a study of patients with BP CML
treated with either 140 mg daily or 70 mg twice daily suggested
Stem cell transplantation
that once-daily dosing had comparable efficacy and better tol-
erability. In those with myeloid-BP CML treated with once With the development of TKIs, rates of allogeneic SCT dra-
daily dasatinib, the MHR was 28%, MCyR was 25%, and OS at matically declined for those with CP CML. Currently, alloge-
24 months was 24%. For those with lymphoid BP CML, cor- neic stem cell transplant is typically reserved for those who
responding rates were 42%, 50%, and 21%, respectively. fail all available TKIs and those with advanced phase disease.
Dasatinib is approved for AP and myeloid or lymphoid BP The phase of disease has a significant impact on transplant
CML with resistance or intolerance to other therapy. outcome, which is best for those with CP CML. Prior use of
With 2 years of follow-up, nilotinib, at a dose of 400 mg TKIs does not appear to influence transplant outcome.
orally twice daily in patients with AP CML, led to CHR, Although myeloablative allogeneic SCT can cure up to
MCyR, CCyR and MMR in 31%, 32%, 21%, and 11% of 40% of adult patients with AP CML, the salvage rate of
patients, respectively. The 24-month overall survival rate patients who receive transplantation in the setting of blast
was 70%. Nilotinib is approved for use in AP CML, but not crisis is dismal. Rather, patients usually are treated with
approved for use in BP CML. induction chemotherapy similar to that used for acute leuke-
Bosutinib has also been approved for use advanced phase mia, often with TKIs, in hopes of achieving a second CP
disease in those with intolerance or resistance to prior ther- CML that may improve the success of an SCT. Transplanta-
apy. In the AP or BP CML patients, the responses included tion in the second chronic phase yields outcomes compara-
confirmed CHR at week 48 (prior imatinib = 30.4%) and ble to those for transplantation in the accelerated phase
OHR at week 48 (prior imatinib = 55.1%). For the BP CML (ie, 20%-40% long-term disease-free survival).
patients, 15% achieved CHR and 28.3% achieved OHR by Across all ages, the incidences of acute graft-versus-host
week 48. Finally, ponatinib was studied in AP CML and is an disease (GVHD) range from 8%-63%, with severe and fatal
option for those with advanced disease and intolerance/ GVHD affecting up to 20% and 13% of patients, respec-
resistance to prior therapy. Among 82 patients with AP tively. The use of alternative donors could expand access to
CML, 55% had a MHR, 39% had an MCyR, 24% attained those in need of transplantation without a matched donor.
CCyR, and 16% had an MMR. Finally, among patients with Conditioning commonly involves a myeloablative approach,
BP CML, 31% had MHR, 23% had an MCyR, and 18% had a and nonmyeloablative approaches are considered investiga-
CCyR. tional. However, there is rationale for nonmyeloablative and

Classical BCR-ABL1–negative MPNs | 443
reduced-intensity conditioning approaches, based on the Pathobiology

fact that CML cells are highly susceptible to the graft-versus-
MPNs are clonal neoplasms
leukemia (GVL) effect of an allograft. The overall leukemia
relapse rate after matched unrelated-donor SCT is somewhat PV was among the first disorders demonstrated to have a
lower than after matched-related transplantations, suggest- clonal origin. In the 1960s and 1970s, studies of blood cells
ing that minor antigen disparity enhances a GVL effect. In from women with PV who were germ line heterozygous for
addition, relapse rates are higher after transplantation with glucose-6-phosphate dehydrogenase (G6PD) isoenzyme
T-cell-depleted stem cells compared with unmanipulated types demonstrated that cells expressed either one G6PD
stem cells, implicating that donor graft immune function is type or the other type, but not a balanced mix of variants
important in clearing residual disease. cells, as would be expected in a polyclonal cell population.
The potency of the GVL effect is further illustrated by the Subsequently, molecular analyses of other polymorphic
success of donor lymphocyte infusion (DLI) for relapsed dis- X-linked genes yielded similar findings.
ease after SCT. CML is the disease that responds best to DLI. Circulating red blood cells, platelets, granulocytes, and
DLI induces remission in 54%-93% of patients with early sometimes B lymphocytes in patients with PV are all prog-
hematologic or cytogenetic relapse after allografting. In eny of the malignant clone. By comparison, the majority of T
addition to DLI, TKI therapy is often effective in the setting lymphocytes and natural killer cells are polyclonal and do
of posttransplant relapse. Accordingly, molecular monitor- not exhibit the karyotypic changes or other genetic abnor-
ing in the posttransplant setting is important to identify malities that characterize the abnormal clone.
those at higher risk for relapse. Similarly, a clonal hematopoietic population can be dem-
onstrated in most, but not all, women with ET when evaluat-
Key points ing for G6PD isoenzyme expression or for inactivation
patterns of polymorphic alleles of X-linked genes in myeloid
• CML is a pluripotent hematopoietic stem cell neoplasm populations. Skewing of X chromosome inactivation pat-
characterized by the BCR-ABL1 fusion gene, which is derived
terns is noted in up to 20%-25% of normal women >50 years
from a balanced translocation between the long arms of
of age, and age-related skewing is present in >50% of women
chromosomes 9 and 22, t(9;22)(q34;q11), also known as the
>75 years. Thus, these analyses of clonality are specific only
Philadelphia (Ph) chromosome.
• Typical blood findings include a left-shifted leukocytosis, with
for selected younger women, but they illustrate generally the
basophilia and often thrombocytosis. clonal nature of disease. In some younger women with sus-
• Prognosis has been remarkably improved by the development tained thrombocytosis, clonality assays do not show a clonal
of TKIs, and is dependent on the phase at presentation (CP, AP, population. The inability to detect a clonal cell population in
or BP) and depth of response to therapy. a portion of younger women with ET may relate to the infre-
• There are now five TKIs available for use in CP-CML. Imatinib, quency of the affected clone as a proportion of total myeloid
dasatinib, and nilotinib are frontline options for CP CML. cells, restriction of clonal maturation to the megakaryocytic
Dasatinib, nilotinib, bosutinib, and ponatinib (T315I) can be lineage, or alternative pathogenic mechanisms that lead to
used in those with intolerance or resistance to prior TKI polyclonal platelet overproduction.
therapy.
• Meeting treatment milestones strongly influences prognosis,
and identifies those with resistance or loss of response, who JAK2 mutations
require a switch to another TKI. Consensus guidelines are
A watershed moment arose in the understanding of the
available to direct appropriate assessments during months 3, 6,
12, and beyond and aid in management decisions.
MPNs when in 2005 the laboratory of William Vainchenker
first identified the point mutation of JAK2 V617F and both
its prevalence and relevance in MPN pathogenesis (Table
Classical BCR-ABL1–negative MPNs 16-4). JAK2 is an intracellular signaling molecule that is cou-
pled to several cell surface hematopoietic growth factor
As discussed above, the classical BCR-ABL1–negative MPNs receptors that lack intrinsic kinase domains, including the
include the three most prevalent illnesses in this category, erythropoietin (EPO) receptor and the thrombopoietin
PV, ET, and MF (which itself includes primary myelofibrosis (TPO) receptor. The specific JAK2 point mutation most
[PMF], post–polycythemia vera myelofibrosis [post-PVMF], closely associated with MPN, V617F, causes constitutive
and post–essential thrombocythemia myelofibrosis [post- activation of the JAK2 kinase domain, which results in eryth-
ET/MF]). This category also includes rarer disorders such as ropoiesis losing its dependence on EPO signaling and
chronic neutrophilic leukemia and chronic eosinophilic becoming virtually autonomous. JAK2 V617F can be found
leukemia. in ~95% of patients with PV, and diagnostic mutation assays

Table 16-4 Somatic mutations seen in patients with MPNs CALR, MPL, and other MPN-associated mutations
PV ET MF MPL
Gene name Mutation effect (%) (%) (%)
JAK2 (V617F) JAK/STAT signaling 95-97 50-60 50-60 After discovery of JAK2 V617F, much effort was expended in
JAK2 exon 12 JAK/STAT signaling 1-2 0 0 identifying other mutations important in diagnosis and
CALR JAK/STAT signaling 0 25 30 pathogenesis in MPNs (Table 16-4). Some of these muta-
MPL JAK/STAT signaling 0 3-5 5-10 tions are primarily found in MPNs, while others are found
CBL JAK/STAT signaling – 0-2 5-10 across the spectrum of both chronic and acute myeloid
SH2B3/LNK JAK/STAT signaling 2-20 2-6 3-6 neoplasms.
ASXL1 Epigenetic modification 2 5-10 10-35 The first recurrent non-JAK2 MPN mutations described
EZH2 Epigenetic modification 1-2 1-2 7-10 were somatic activating mutations in the gene encoding the
IDH1/2 Epigenetic modification 2 1 5 TPO receptor (MPL) in 3%-5% of ET patients and 5%-10%
DNMT3A Epigenetic modification 5-10 2-5 8-12 of those with PMF. These latter mutations have not been
TET2 Epigenetic modification 10-20 4-5 10-20 found in patients with PV. By comparison, germ line MPL
SF3B1 MRNA splicing 2 2 4 mutations that lead to constitutive overexpression of the
SRSF2 MRNA splicing – – 4-17 gene product have been identified in several kindreds with
U2AF1 MRNA splicing <1 <1 1-8
hereditary thrombocytosis. Hereditary thrombocytosis is
ZRSR2 MRNA splicing <1 <1 <1
rare. There were also non-V617F germ line JAK2 variants
TP53 DNA repair 1-2 1-2 1-2
(JAK2 V617I and JAK2 R564Q) identified in two families
with hereditary thrombocytosis.
are readily available for routine clinical testing. This muta- CALR
tion is homozygous in at least one-third of PV cases, a situa-
In late 2013, the laboratories of Robert Kralovics and
tion that arises by acquired uniparental disomy of the region,
Anthony Green identified a prevalent mutation in the calre-
including the mutated gene on chromosome 9p24.
ticulin gene (CALR) with important diagnostic and poten-
Analysis of JAK2 V617F–negative PV patients led to the
tially pathogenetic implications in ET and MF patients.
identification of acquired-activating mutations in exon
CALR mutations are present in 25%-30% ET and MF and
12 of JAK2 in most, but not all, JAK2 V617F–negative PV
are present in the majority of JAK2 wild-type cases (Table
patients. Of note, unlike the more pleiotropic JAK2 V617F
16-4). These mutations are mutually exclusive from JAK2
allele, which is seen in a spectrum of myeloid malignan-
V617F and MPL mutations, and absent in patients with PV.
cies, JAK2 exon 12 mutations are found in JAK2 V617F–
Mutations in CALR have been found infrequently in other
negative PV and most often are identified in patients
myeloid neoplasms, including myelodysplastic syndrome
with erythrocytosis without associated thrombocytosis or
(MDS), and MPN/MDS overlap syndromes. The mutations
leukocytosis.
in CALR occur in the terminal exon 9 of the gene and the two
Despite the high frequency of the JAK2 V617F mutation
most common mutations are a 52-base frameshifting dele-
in PV, this molecular mutation is not specific to this disease
tion, which has been referred to in the literature as a type 1
type, as it is also frequently identified in about 50% of ET
deletion (most common mutation) or a 5-base-pair inser-
and PMF (also about 50%). Patients with post-ET/-PV MF
tion, which has been referred to as a type 2 mutation. There
have JAK2 V617F as prevalent as the preceding MPN of ET/
are prognostic implications for the presence of the CALR
PV. The presence of the JAK2 V617F mutation across all
mutation, which marks individuals with a more indolent
subtypes of MPN raises the question of what other factors
disease course compared to JAK2 or MPL mutations; how-
contribute to the phenotypic heterogeneity within different
ever, data are conflicting about whether the type of CALR
MPNs that share the same JAK2 V617F mutation. Differ-
mutation (type 1 or 2) is prognostic.
ences in allele burden, downstream intracellular signaling,
host genetic background, acquisition of other molecular
Additional MPN mutations
mutations, and the hematopoietic progenitor tissue type tar-
geted by the JAK2 V617F mutation can all influence pheno- In addition to the main mutations which have been identi-
type. A germ line haplotype (46/1, GGCC) at the 3′ region of fied in MPNs (JAK2, MPL, CALR), a variety of additional
JAK2 also has been associated with a three- to fourfold mutations have been identified, which are likely relevant in
increased risk of developing a JAK2 V617F mutant or MPL MPN pathogenesis but may have roles in other myeloid neo-
mutant MPN, although a weak association with JAK2 V617F plasms as well as a role in the pathogenesis of disease pro-
wild-type MPN also was observed. gression to MPN blast phase (MPN-BP). These additional

mutations include several epigenetic modifiers, including The number of circulating hematopoietic progenitors
ASXL1, DNMT3A, TET2, EZH2, IDH1/IDH2. Additionally, (CD34+ cells) is increased significantly in PMF compared to
the presence of mutations in genes encoding factors impor- other myeloid neoplasms and normal controls. The number
tant in mRNA splicing, including SF3B1 and U2AF2 (Table of circulating CD34+ cells in PMF patients can be 50-fold
16-4), has been reported. Increasing literature regarding the higher than in PV or ET. Higher levels of circulating CD34+
diagnostic and prognostic implication of these mutations are cells in PMF are associated with more advanced bone mar-
discussed in the respective disease-associated chapters. row fibrosis and other disease characteristics. One study
observed shorter survival and earlier transformation to acute
myeloid leukemia (AML) among PMF patients with circu-
Karyotype and cytogenetic abnormalities
lating CD34+ cell counts >300/mL. A “hostile” marrow
Clonal karyotypic abnormalities can be frequently seen in microenvironment alone does not appear to account for
patients with MPN and are most commonly identified in elevated CD34+ counts, because extensive myelofibrosis sec-
individuals with PMF (Table 16-5). Increasing chromosomal ondary to marrow involvement with carcinoma or lymphoid
mutations and greater karyotypic complexity have been malignancies is associated only mild increases in the num-
associated with the potential progression toward MPN-BP. bers of circulating progenitor cells.
A recent analysis of over 800 patients with PMF identified a The marked reactive mesenchymal cell proliferation in
karyotypic abnormality in approximately 43% of patients, PMF has been linked to inflammatory response cytokines
with 68.2% of these individuals having a sole karyotypic and megakaryocyte- and monocyte-derived growth factors.
abnormality, 18.2% having 2 mutations, and 13.6% having 3 Platelet-derived growth factor (PDGF) and basic fibroblast
or more karyotypic abnormalities. Prognostic assessment of growth factor (bFGF) are constitutively overproduced by
the karyotype will be discussed in the setting of all prognostic clonal megakaryocytes in PMF. Platelets and monocytes or
factors in the section on myelofibrosis. macrophages in PMF marrow release increased amounts of
transforming growth factor beta (TGFβ), which also con-
tributes to the stromal reaction. In PMF patients, elevated
The bone marrow microenvironment
levels of IL-1 and tumor necrosis factor alpha (TNFα) are
Although histologic changes can occur in the development associated with augmented production or release of PDGF,
of ET/PV (eg, changes in marrow cellularity, expanded bFGF, angiogenic factors such as vascular endothelial growth
megakaryocyte numbers, and megakaryocyte clustering), factor (VEGF), and osteogenic cytokines, in addition to
PMF is most commonly characterized by marrow fibrosis direct or indirect effects of TGFβ. Neoangiogenesis is more
and extramedullary hematopoiesis (EMH). EMH commonly significant in PMF compared with other MPNs, and high
affects the liver and spleen. These processes result from the marrow microvascular density is associated with advanced
proliferation and emigration of neoplastic hematopoietic splenomegaly and shorter survival.
cells and production of cytokines within the marrow micro-
environment, leading to reactive proliferation of fibroblasts Additional biological abnormalities
and other mesenchymal cells. Throughout all stages of dis- Endogenous myeloid colony growth
ease, the circulating populations of red blood cells, granulo-
cytes, and platelets are clonal. Before the discovery of JAK2 V617F, a number of pathophys-
iologic growth and survival abnormalities were identified in
progenitor cells from PV patients. One such characteristic
Table 16-5 Individual karyotypic abnormalities identified in patients
biologic feature, which was used as a minor diagnostic crite-
with PMF
rion for PV by several groups, is the ability of erythroid pro-
Karyotype % genitors to form colonies in serum-containing cultures in
20q− 23.3 the absence of EPO, a phenomenon often called endogenous
13q− 18.2 erythroid colony (EEC) growth. EEC growth is not specific
Trisomy 8 11.1 for PV; it also can be observed in some cases of ET and PMF
Trisomy 9 9.9 and in rare cases of congenital polycythemia. JAK2 V617F
Duplication of chromosome 1q 9.7 provides a mechanistic explanation for this phenomenon in
−7/7q 7.1 acquired MPNs.
Monosomal karyotype 5.7 Biologic and functional abnormalities frequently are
i(17q) 1.4 found in hematopoietic progenitor cells and platelets from
12p− 1.1
ET patients also. Myeloid progenitors demonstrate increased
inv(3) 0.6
in vitro sensitivity to cytokines, with excessive colony growth

in the majority of cases and endogenous megakaryocyte col- MPN symptom assessment form (MPN-SAF). Evidence sup-
ony formation in a significant proportion of ET cases. A sub- porting the role of symptoms in the diminution of both
stantial proportion of ET patients demonstrate EEC growth, quality of life and life expectancy is substantial. Patient-
megakaryocytic progenitor sensitivity to TPO (the major reported outcome tools have identified the following symp-
megakaryocytic growth and differentiation factor), or both. toms in decreasing order of prevalence: fatigue, concentration
problems, early satiety, inactivity, night sweats, itching,
Thrombopoietin abdominal discomfort, bone pain, weight loss, and fevers.
Additional symptoms have been identified related to the
Additional phenotypic features of unclear pathogenic signifi- presence of increased peripheral blood counts in PV and ET
cance include decreased expression of c-Mpl (the TPO recep- including dyspnea, dizziness, visual changes, weight loss, epi-
tor) in PV megakaryocytes and platelets and overexpression gastric pain, excessive sweating, or painful paresthesias
of messenger RNA (mRNA) encoding polycythemia rubra of the hands and feet associated with erythema (erythromel
vera 1 (PRV1) in mature granulocytes but not in myeloid algia). These symptoms are believed to relate to hypervis
progenitors. Neither of these findings is specific for PV, cosity, hypercatabolism, elevated histamine release, and
although some groups use them as supportive evidence for a microvascular or vasomotor instability. Vasomotor manifes-
PV diagnosis in challenging cases. Normal to elevated plasma tations, which appear to be due to platelet-endothelial inter-
levels of TPO are observed in most patients with ET, as well actions and inflammation in small arterioles, usually
as in many patients with PV. Some earlier studies attributed manifest as symptoms in the central nervous system or acral
the higher TPO levels to decreased expression of the c-Mpl extremities.
receptor on megakaryocytes and platelets, with decreased Close assessment of MPN symptom burden has been
platelet-mediated clearance of bound ligand. More recent found to be of prognostic significance in PMF (See section
studies have not confirmed those earlier findings, suggesting on MF prognosis). Additionally, improvement in MPN-
that low c-Mpl expression levels and high plasma TPO levels related symptoms has been included in the response criteria
are not specific or consistent abnormalities in ET. Increased for ET/PV (ELN 2013) as well as MF (IWG-MRT 2013) (see
mean platelet volume and qualitative platelet aggregation respective disease sections). Finally, the improvement/reso-
defects are also found in up to 90% of patients with ET; how- lution of MPN-related symptoms has proven to be an
ever, they are not useful clinical predictors of the risk of important aspect of efficacy for new therapies including JAK
thrombosis or bleeding. Qualitative platelet defects can also inhibitors.
be observed in other MPNs.
Physical findings
Diagnosis
Physical findings in MPNs include (i) manifestations of
The clinical presentation of MPN is variable, but clinical sus- prior thrombotic events; (ii) the presence of plethora and
picion is usually driven by abnormal laboratory features rubor in the setting of erythrocytosis; (iii) the development
such as erythrocytosis, thrombocytosis, and leukocytosis in of pallor or cachexia in the setting of anemia, (iv) muscle
the absence of a clear explanation, sometimes accompanied wasting and other hypercatabolic signs, especially in PMF;
by splenomegaly, constitutional symptoms, or a vascular (v) splenomegaly (present in 60%-80% of patients with
event. Additional consideration of a MPN can occur with myelofibrosis at presentation, less common in PV and ET);
any of the signs and symptoms described below, as outlined and (vi) hepatomegaly, although this is most common in the
in Table 16-6. setting of patients with MF who previously have been sple-
nectomized. Extramedullary hematopoiesis also can cause
not only hepatomegaly but also lymphadenopathy and asci-
MPN-related symptoms
tes (through peritoneal seeding; can also occur via portal
All three of the Ph-negative MPNs can range from being hypertension from thrombosis) and form extramedullary
asymptomatic to highly symptomatic at disease presentation masses anywhere in the body.
with, in general, patients with MF being the most symptom-
atic, followed by those with PV, and then ET the least symp-
Laboratory features
tomatic (Table 16-6). The constellation of MPN-related
symptoms has been quantified and can be measured and An elevation in the red cell count, white cell count, or platelet
tracked to the development of now validated MPN-specific count are common features across MPNs with erythrocyto-
patient-reported outcome tools including the myelofibrosis sis being pathognomonic for PV. Erythrocytosis in PV may
symptom assessment form (MFSAF) and the more general be absent in the setting of either disease progression for

Table 16-6 Laboratory, physical findings, and symptoms at presentation for BCR-ABL1–negative MPNs
Feature PV ET MF (PMF/post–ET/PV MF)

Laboratory features
Erythrocytosis +++ + Absent
Leukocytosis Variable Variable Variable
Thrombocytosis Variable +++ Variable
Leukoerythroblastosis Absent Absent +++
Decreased serum erythropoietin +++ Variable Absent
Elevated lactate dehydrogenase Absent Absent Common
Hyperuricemia Uncommon Uncommon Variable
JAK2V617F mutation 90%-95% 50% 50%
CALR mutation Absent 25% 30%
MPL mutation Absent 3%-5% 5%-10%
JAK2 exon 12 mutation 2%-5% Absent Absent
Physical findings
Splenomegaly + + +++
Hepatomegaly Absent Absent +
Plethora ++ Absent Absent
Pallor Absent Absent Variable
Disease-related symptoms (MPN-10)*
Fatigue 84% 85% 94%
Early satiety 56% 60% 74%
Abdominal discomfort 48% 48% 65%
Inactivity 54% 60% 76%
Problems with concentration 58% 62% 68%
Night sweats 47% 52% 63%
Pruritus 46% 62% 52%
Bone pain 45% 48% 53%
Fever 17% 19% 24%
Weight loss 28% 33% 47%
Additional presenting features
Erythromelalgia + ++ Absent
Thrombosis† Variable Variable Variable
Hemorrhage Variable Variable Variable
Portal hypertension Variable Variable Variable
*See Geyer HL, Mesa RA. Blood 2014;124:3529-3537.
+ to +++: Occasional to very common.
†Venous thrombosis: any site is possible, noting that MPNs frequently underlie splanchnic vein thrombosis and, less commonly, dural venous
sinus thrombosis. Arterial thrombosis: any site is possible, though cardiovascular and cerebrovascular events prevail.
post-PV fibrosis, or restrained hematopoiesis in the setting cause of anemia in PMF is often multifactorial, including
of iron deficiency or other nutritional deficiencies. Throm- impaired erythropoiesis, hematopoietic failure, hemolysis,
bocytosis can be seen in MPNs, as well as other myeloid neo- hemorrhage (usually gastrointestinal), and hypersplenism.
plasms (eg, del5q MDS) and MDS/MPN overlap syndromes. Symptoms due to anemia are common among patients with
Leukocytosis, while characteristic of CML, again can be seen PMF. Approximately 50%-70% of patients are anemic at
across the spectrum of MPNs. A leukoerythroblastic periph- presentation, and 25% of patients have hemoglobin levels of
eral blood pattern can be seen, which includes the presence <8 g/dL. Progressive thrombocytopenia, usually in the setting
of nucleated red blood cells and left-shifted white cell differ- of hematopoietic failure and hypersplenism (ie, sequestra-
ential; this is suggestive of fibrosis. tion and destruction of circulating platelets in the spleen),
Anemia is the major hematologic complication of PMF significantly increases the risk of bleeding.
and usually occurs in ET and PV only in the setting of pro- Additional laboratory findings include a normal or decreased
gression or excess myelosuppression by medical therapy. The serum EPO level in the setting of erythrocytosis in PV,

elevations in LDH, and hyperuricemia. Low-grade dissemi- bleeding. Bleeding is more common in older individuals and
nated intravascular coagulopathy may arise in some patients in patients with platelet counts ≥1,500 × 109/L.
and can exacerbate disease-related thrombotic and hemor- Portal hypertension with ascites and varices develops in
rhagic risk. Secondary iron overload may develop because of up to 7% of patients with PMF. This complication arises
inappropriate iron loading by the gut and iatrogenically, due from thrombotic vasculopathy involving the portal circula-
to red blood cell transfusion dependency. Autoimmune tion and extramedullary hematopoiesis. Portal hypertension
complications have been described in association with PMF, and massive splenomegaly predispose to splenic infarction.
including hemolytic anemia and vasculitis. Splenic infarction should be suspected in a patient with PMF
who presents with acute or subacute left upper-quadrant
Additional MPN manifestations pain radiating to the shoulder, with or without associated
nausea and fever. Splenic infarcts do not have clear prognos-
Vascular events
tic value. Pulmonary hypertension can occur as a complica-
Roughly 15% of patients diagnosed with PV have suffered tion of PMF and is underrecognized.
an arterial or, less commonly, a venous thrombotic event
within the previous 2 years. One-fifth of patients with PV
Nonhepatosplenic extramedullary hematopoiesis
present with a large-vessel thrombotic complication, such as
a transient ischemic attack, cerebrovascular accident, myo- Rare patients (<10%) with MF develop nonhepatosplenic
cardial infarction, deep venous thrombosis, or hepatic vein extramedullary hematopoiesis in locations such as the verte-
thrombosis. It is important to consider the diagnosis of PV bral column (paraspinal or intraspinal lesions, which can
in people who develop an apparently unprovoked thrombo- lead to cord compression), lung, pleura, retroperitoneum,
embolic event, in particular at unusual sites (ie, dural sinuses eye, kidney, bladder, mesentery, and skin. Lung extramedul-
or splanchnic veins). In particular, a substantial proportion lary hematopoiesis is associated with pulmonary hyperten-
of patients who present with Budd-Chiari syndrome and sion and marked reduction in overall survival and impaired
portal vein thrombosis are found to be JAK2 V617F positive, quality of life.
yet the erythrocytosis of PV is masked by bleeding or
hypersplenism.
Blastic transformation
Approximately 10%-25% of patients with ET have suf-
fered a thrombotic or bleeding complication prior to diag- Ph-negative MPN confers a risk of blast phase, just as in
nosis. As illustrated by the previous clinical case, young CML. The majority of individuals who progress to blast
women with ET have a significantly increased risk of first- phase do so from the MF phase of the disease (PMF or post–
trimester abortions (up to 36% of documented pregnancies) ET/PV MF). About 1/3 of deaths linked to MF arise from
and may suffer recurrent fetal loss; studies have not identi- complications of leukemic progression. Individuals have
fied clinical or laboratory risk factors for pregnancy- progressed to blast phase directly from ET/PV, although
associated complications in ET with the exception of poorly usually in the setting of having received clearly leukemogenic
managed platelet counts. In ET, thrombosis in the arterial therapy such as radioactive phosphorus (P-32) or alkylator
system, including the cerebral, coronary, and peripheral therapy, such as melphalan.
arteries, is roughly three times more common than throm-
boembolic complications involving the veins.
Epistaxis is reported by 15%-20% of patients at diagnosis, Polycythemia vera
and gastrointestinal bleeding is reported by approximately
5%. In some cases, bleeding can be attributed to the develop-
ment of acquired type 2 von Willebrand disease (vWD) which Clinical case
may occur with extreme thrombocytosis (usually ≥1,000 × A 60-year-old male violinist presented with intractable gener-
109/L), The mechanism of acquired vWD in thrombocytosis alized pruritus. The patient’s general practitioner immediately
appears to be related to platelet binding, cleavage, and ulti- noticed multiple skin excoriations. The patient was prescribed
mately clearance of large von Willebrand factor (vWF) multi- antihistamines and steroid cream and was sent home. A week
mers. There are also recent data suggesting that endothelial later, he returned, complaining of the same problem. He also
cells with the JAK2 V617F mutation may play a role in acquired started to experience facial flushing and painful erythematous
vWD. Normal vWF activity is restored after the platelet count swelling of all his fingers, impairing his ability to play the
is normalized by cytoreductive drugs or platelet-pheresis. In violin. Physical exam revealed erythematous swelling of both
hands, multiple skin excoriations, and palpable spleen 3 cm
ET, hemorrhage is reported in only 6% of patients at diag
below the left subcostal margin. Vital signs including oxygen
nosis, most commonly as gastrointestinal or oral mucosal

Polycythemia vera | 449
resultant liver dysfunction, which masks erythrocytosis and

Clinical case (continued) can confound the diagnosis of PV.
saturation were within normal limits. The doctor ordered a
CBC, which showed the following: WBCs = 12 × 109/L,
Question 2: Is erythrocytosis primary or secondary?
Hb = 19 g/dL, mean corpuscular hemoglobin: 85 fL, platelet
count = 530 × 109/L. Additional blood tests showed a serum In an adult with erythrocytosis, additional studies are
erythropoietin level of 2 U/L (normal, 7-20 U/L) and the pres- required to differentiate PV from secondary causes of eryth-
ence of a JAK2 V617F mutation. The patient was phleboto-
rocytosis (Table 16-7). Rare patients may have both PV and
mized a unit of blood (450 ml) as well as started on aspirin
a secondary cause of erythrocytosis. Most acquired second-
(81 mg by mouth once daily) with improvement in pruritus
ary polycythemic states are associated with elevated or high-
and also erythematous swelling of both hands. He has no
history of bleeding or blood clots.
normal serum EPO levels, which may be appropriately
elevated (ie,in the setting of chronic tissue hypoxemia) or
inappropriately elevated (ie, due to exogenous administra-
Introduction tion of recombinant EPO or endogenous EPO overproduc-
tion by the kidney or liver or by a tumor). A smoking history
PV is defined by an elevated hemoglobin or hematocrit or (including exposure to secondhand smoke) should be
red cell mass (RCM) in the absence of conditions that induce obtained; arterial blood gas and carboxyhemoglobin deter-
secondary erythrocytosis, such as hypoxia or inappropriate minations should be obtained in smokers or those with
EPO production by a tumor. Excessive red blood cell pro- occupational exposure to hydrocarbon fumes. An oxygen
duction by the bone marrow in PV often is associated with saturation <92% suggests the possibility of EPO-driven
concomitant increases in circulating platelets and granulo- polycythemia. Evidence for underlying lung or cardiac dis-
cytes because the acquired defect underlying PV involves a ease should be sought, and evaluation for sleep apnea
multipotential hematopoietic progenitor cell. through monitoring of nocturnal oxygen saturations should
PV is the most common MPN in the United States, with be considered. An abdominal ultrasound is useful to assess
an annual incidence rate of roughly 1.1 cases per 100,000 for renal or hepatic cysts or tumors or for splenomegaly, if it
persons per year. There is a slight male predominance. The is not obvious from physical examination. In the absence of
median age at diagnosis is 60-65 years, with roughly 5% of other causes of elevated EPO and with a normal oxygen satu-
cases occurring in those <40 years old. Radiation exposure ration, an oxygen dissociation curve (partial pressure of oxy-
and, rarely, environmental or toxic factors have been linked gen at which hemoglobin is half saturated) should be
to the disease. determined to evaluate for a high oxygen affinity hemoglo-
binopathy, if there is sufficient clinical suspicion. The family
Diagnosis history is often suggestive in these cases. Alternative causes
of secondary polycythemia also must be considered, includ-
Question 1: Is there a polycythemic state?
ing EPO-producing renal or liver tumors; in addition, cere-
Absolute polycythemia versus relative polycythemia
bellar hemangiomas, uterine myomas, ovarian tumors,
An elevated hematocrit may result from either an increase in parotid tumors, lymphomas, and adrenal tumors can be
the total RCM (absolute polycythemia) or a decrease in the associated with pathologic EPO production. Budd-Chiari
total plasma volume (relative polycythemia). The latter con- syndrome can be a presenting sign of PV; however, in the
dition usually is due to moderate to severe intravascular acute setting, hepatic necrosis can be associated with tran-
dehydration, such as that due to diarrhea or loss of fluid into siently elevated EPO levels, misleadingly suggesting a
third spaces (effusions or edema), sometimes exacerbated by secondary, EPO-driven erythrocytosis. Therefore, it is
diuretic use. Because laboratory normal ranges are based on important to assess patients after they recover from the acute
statistical distributions, 2.5% of healthy people will have a complications of Budd-Chiari syndrome. Androgen replace-
hematocrit value above the normal laboratory reference ment therapy or androgen abuse by aspiring athletes and
range. In contrast, in some cases, a normal hemoglobin and bodybuilders may induce polycythemia. Absolute polycy-
hematocrit can be a sign of disease; examination of old blood themia may occur in up to 15% of patients after renal trans-
counts may demonstrate that the patient’s hemoglobin has plantation. This complication usually develops 8-24 months
increased substantially from his or her personal baseline, after transplantation and remits spontaneously in one-
although the level is still within the laboratory normal range. fourth of patients; when treatment is required, therapy with
In addition, patients with fluid overload can present with a angiotensin-converting enzyme inhibitors or angiotensin
normal hematocrit in the setting of an elevated RCM; this is receptor blockers often is effective. JAK2 V617F mutations
most common in patients with Budd-Chiari syndrome and have not been identified in patients with secondary

Table 16-7 Causes of secondary erythrocytosis polycythemia without a concurrent diagnosis of a MPN;
Neonatal therefore, the presence of the JAK2 V617F allele supports a
Normal intrauterine environment (Hb F) diagnosis of PV even in patients with concomitant second-
Twin-twin transfusion syndrome or maternal-fetal bleeds ary polycythemia.
Infants of diabetic mothers Differentiation between PV and secondary polycythemia
Intrauterine growth retardation
is important both for prognosis and treatment because sec-
Adrenal hyperplasia
Thyrotoxicosis
ondary polycythemia does not carry a risk of leukemic or
Congenital fibrotic transformation and has a lower risk of thrombosis.
Trisomies of 13, 18, or 21 In either case, however, symptoms may result from increased
Mutant high oxygen-affinity hemoglobin viscosity of the blood. In secondary polycythemia, phlebot-
Congenital low 2,3-bisphosphoglycerate omy may occasionally be indicated to decrease blood viscos-
Autonomous high-EPO production (including Chuvash-type ity and improve oxygenation when symptoms occur or
polycythemia associated with VHL mutations)
prophylactically, especially when hematocrit values exceed
Autosomal dominant polycythemia (including truncating EPO
receptor mutations)
60%. With reduction in blood viscosity, symptoms such as
HIF2A (EPAS1) mutation headaches, plethora, and dizziness quickly abate. Caution
Proline hydroxylase (EGLN1) mutation must be exercised, however, when the cause of secondary
Other congenital polycythemic states polycythemia is physiologically appropriate, such as in cya-
Acquired notic congenital heart disease, chronic hypoxia, or high-
Arterial hypoxemia affinity hemoglobins. Tissue hypoxemia may be worsened by
High altitude
phlebotomy when the physiologically appropriate compen-
Cyanotic congenital heart disease
Chronic lung disease satory increase in hematocrit is reversed. If underlying severe
Sleep apnea and hypoventilation syndromes cardiac disease exists, 0.9% normal saline should be given
Other causes of impaired tissue oxygen delivery concurrently with phlebotomy to compensate for the blood
Smoking volume that is removed.
Carbon monoxide poisoning
Renal lesions
Renal tumors Familial polycythemic states
Renal cysts
Diffuse parenchymal disease Primary familial and congenital erythrocytosis is usually
Hydronephrosis autosomal dominant and most commonly associated with
Wilms tumor low-serum EPO levels. Approximately 10% of such cases
Renal artery stenosis have been linked to germ line truncating mutations of the
Renal transplantation
EPO receptor that abrogate an important inhibitory domain
Miscellaneous tumors
Parotid tumors
and lead to constitutive EPO receptor signaling. In contrast,
Cerebellar hemangiomas normal or high EPO levels are found in patients with Chu-
Lymphomas vash-type congenital polycythemia due to abnormalities in
Uterine myomata cellular oxygen sensing. This autosomal-recessive disorder
Cutaneous leiomyomata was first recognized among the population of the Chuvash
Bronchial carcinoma
region in the center of the European part of Russia and is
Ovarian tumors
associated with a high risk of thrombotic and hemorrhagic
Adrenal tumors
Meningiomas complications. Sporadic cases of Chuvash-type polycythe-
Pheochromocytomas mia with homozygous or compound heterozygous inheri-
Drugs and chemicals tance patterns subsequently have been identified among
Androgens other ethnic groups. These patients have mutations involv-
Erythropoiesis-stimulating agents (eg, epoetin alfa or darbepoetin alfa) ing a region of the von Hippel-Lindau (VHL) gene that is
Novel erythropoietic agents
distinct from the autosomal-dominant VHL mutations asso-
Nickel
Cobalt ciated with the von Hippel-Lindau syndrome. The Chuvash-
Hepatic lesions type VHL mutations impair the function of the VHL gene
Hepatomas product to facilitate degradation of hypoxia-inducible factor
Cirrhosis 1 (HIF1), an oxygen-responsive transcriptional factor that
Hepatitis upregulates EPO expression. More recent studies of families
Modified from Pearson TC, Messinezy M. Pathol Biol (Paris). with autosomal-dominant heritable erythrocytosis have
2001;49:170-177. identified germ line mutations in the HIF2A gene that lead to

defective oxygen sensing and resultant polycythemia; of diagnostic criteria for PV and without an apparent cause for
note, these mutations are heterozygous and result in dysreg- secondary polycythemia. Typical features of PV will develop
ulation of the HIF transcriptional complex. Another autoso- in 10%-40% of such patients over time, but most cases will
mal-dominant familial polycythemia is caused by germ line resolve spontaneously, or a cause of secondary polycythemia
mutations in proline hydroxylase domain 2 (PHD2). PHD2 is ultimately will become clear. JAK2 testing is helpful to distin-
an Fe(II)- and 2-oxoglutarate–dependent oxygenase that guish these groups. Some younger patients have been
hydroxylates HIF2A to allow it to be targeted for ubiquitina- described with chronic “idiopathic erythrocytosis” that per-
tion and degradation by VHL. sists for many years without evolution to PV or development
of thrombohemorrhagic complications. The etiology of this
Diagnostic criteria for PV syndrome is unclear; however, a subset of patients with idio-
pathic erythrocytosis who do not meet the criteria for a diag-
From PVSG to WHO
nosis of PV are JAK2 exon 12 mutation positive.
In the absence of causes for secondary polycythemia and a A low-serum or inappropriately normal-serum EPO level
family history of erythrocytosis and in the presence of a low is common with PV. Serum EPO levels within the normal
serum EPO level, PV is the most likely diagnosis. The Poly- range also occur in PV, especially when EPO levels are not
cythemia Vera Study Group (PVSG) originally established measured until after the patient has undergone initial thera-
criteria in the 1970s to standardize the diagnosis of PV to peutic phlebotomy. Normal EPO levels also may be seen in
follow a uniform group of patients on treatment protocols. hypoxemia-induced secondary polycythemia after the ele-
In light of the discovery of JAK2 V617F and other MPN- vated RCM has restored adequate renal oxygen delivery.
associated molecular lesions, the WHO has published a set of Elevated EPO levels strongly suggest hypoxemia or another
diagnostic criteria that includes genetic testing for JAK2 form of secondary erythrocytosis but rarely may be seen in
V617F as a major criterion (Table 16-8). JAK2 V617F muta- PV if there is also a concomitant secondary polycythemic
tion testing by PCR-based assays is >90% sensitive for PV state. Assays for EEC growth, although not readily available
and appears to be 100% specific for clonal myeloprolifera- to most clinicians, are minor criteria for the diagnosis of PV
tion. JAK2 mutation testing is useful as a first-intention diag- and are helpful in the evaluation of JAK2 V617F–negative
nostic test for evaluation of an elevated hematocrit, before patients who present with polycythemia. Platelets show
any other investigation, but it cannot distinguish inapparent abnormal in vitro aggregation in up to 80% of cases. Addi-
PV from other clonal myeloid disorders such as ET or MF. tional laboratory abnormalities that are present in at least
Therefore, JAK2 V617F mutations are not specific for PV half of PV patients include an elevated LAP score, elevated
given their occurrence in ET, PMF, and other myeloid malig- LDH, and hyperuricemia (Table 16-6).
nancies; however, the presence of the JAK2 V617F allele and Bone marrow evaluation is not critical for the confirma-
erythrocytosis does provide two major criteria for the diag- tion of suspected PV when other diagnostic criteria are pres-
nosis of PV using the new WHO criteria. For patients who ent, including JAK2 V617F, but is useful in supporting the
are JAK2 V617F negative, testing for JAK2 exon 12 is of value diagnosis, in establishing prognosis, or in the diagnostic eval-
in JAK2 V617F type patients with suspected PV. uation of JAK2 V617F-negative patients. Common marrow
Some patients present with an elevated RCM and a nor- findings with PV include erythroid and megakaryocytic
mal or near-normal serum EPO level in the absence of other hyperplasia, with abnormal megakaryocytes (typically large
Table 16-8 WHO 2008 diagnostic criteria for PV
Major 1. Hemoglobin >18.5 g/dL in men,>16.5 g/dL in women, or other evidence of increased red blood cell volume*
2. Presence of JAK2 V617F or other functionally similar mutation such as JAK2 exon 12 mutation
Minor 1. Bone marrow biopsy showing hypercellularity for age with trilineage growth (panmyelosis) with prominent erythroid,
granulocytic, and megakaryocytic proliferation
2. Serum erythropoietin level below the reference range for normal
3. Endogenous erythroid colony formation in vitro
Diagnosis Requires the presence of both major criteria and one minor criterion or the presence of the first major criterion together with
two minor criteria
Adapted from Swerdlow SH, et al., eds. World Health Organization Classification of Tumours of Haematopoietic and Lymphoid Tissues. Lyon,
France: IARC Press; 2008.
* Hemoglobin or hematocrit ≥99th percentile of method-specific reference range for age, sex, altitude of residence or hemoglobin >17 g/dL
in men, >15 g/dL in women if associated with a documented and sustained increase of at least 2 g/dL from an individual’s baseline value that
cannot be attributed to correction of iron deficiency or elevated red blood cell mass >25% above mean normal predicted value.

cells with hyperlobated nuclei) organized in clusters (Figure British Committee for Standards in Hematology (BCSH)
16-3). Advanced reticulin fibrosis (ie, grades 3 or 4) is found which have two different cohorts, JAK2-positive polycythe-
in ~5% of cases at diagnosis, but in at least 20% after 10-15 mia vera, which requires both the mutation in JAK2 (JAK2
years and >50% after 20 years of disease. Karyotypic analysis V617F or JAK2 exon 12) as well as a high hematocrit (>0.52%
may offer diagnostic utility in ambiguous cases that are JAK2 in men or >0.48% in women) or raised red cell mass (>25%
wild type. At diagnosis, ~15% of karyotypes from PV patients above predicted).
contain nonrandom chromosomal abnormalities, including
trisomy 8, trisomy 9, del(13q), and del(20q). Monosomy 7,
Disease course and prognosis
del(5q), and other abnormalities seen with AML and myelo-
dysplasia are not common in PV. The frequency of a clonal In the short term, patients with PV have a risk of thrombotic
karyotype in late-stage (>10 years) PV increases to 80%. Of events, either microvascular or macrovascular, as well as risk
note, there is no clear association between specific karyotypic of hemorrhagic events. Additionally, constitutional symp-
abnormalities and progression to either AML or the pretermi- tomatic burden and splenomegaly may be bothersome. Pro-
nal stage of high-grade marrow fibrosis. gression to fibrosis or leukemia can result in an increase in
As of the writing of this chapter, a committee of the World symptoms. Table 16-9 lists risk factors for clinical complica-
Health Organization is reconsidering the current diagnostic tions in PV.
criteria for PV (Barbui, et al, Leukemia, 2014). The rationale
Table 16-9 Risk factors for clinical complications of PV
behind these changes are several, including the concern that
the hematocrit target of the 2008 criteria has lost sensitivity Thrombosis
in striving for extreme specificity. The final format of the Age >60 years old
revised criteria have not been finalized, but likely will include History of previous thrombosis (venous or arterial)
High rate of phlebotomy
a decrease in the hemoglobin threshold for diagnosis to
16.5 g/dL, likely will remove endogenous erythroid colony Hemorrhage
formation in vitro as a minor criteria given the lack of broad Postoperative state with uncontrolled hematocrit or platelet count
availability of this assay and standardization, and likely will Platelet antiaggregating therapy (particularly in elderly patients)
Thrombocytopenia during late-stage postpolycythemic myelofibrosis
mandate a bone marrow aspirate and biopsy for confirma-
tion of morphologic features. Until these have been ratified Acute leukemia
by the WHO and formally published, one should consider Previous therapy with chlorambucil, busulfan, or 32P
Evolution to post-PVMF
the 2008 criteria by the WHO as the standard. It is worthy of
Possibly pipobroman use
note that there are competing diagnostic criteria by the
a b

Figure 16-3 Megakaryocyte morphologic abnormalities observed in myeloproliferative disorders. (a) Megakaryocyte hyperplasia, clustering, and
nuclear hyperlobation in a bone marrow core biopsy sample from a patient with PV (hematoxylin-eosin stain; original magnification, ×170);
(b) a marrow aspirate from a patient with essential thrombocythemia (Wright-Giemsa stain; magnification, ×170). Photos courtesy of Steven J.
Kussick (University of Washington, Seattle, WA).

Vascular events in PV Disease progression

Thrombosis and bleeding are the major causes of morbidity Post-PV MF
and death with PV. Data from the European Collaboration
Patients with PV can progress to postpolycythemic MF
on Low-dose Aspirin in Polycythemia Vera (ECLAP) study
(post-PV MF) or to overt AML (ie, MPN-BP) (Table 16-10).
revealed a thrombotic complication rate of 5.5 events per
Post-PV MF typically develops in individuals who have had
100 patients per year at a median follow-up time of 2.7 years.
PV for at least 10 years. Changing phenotypic features of a
Two-thirds of those events were arterial, and one-third were
PV patient moving towards MF include decreasing erythro-
venous. The risk of a thrombotic complication in the ECLAP
cytosis (ie, requiring less phlebotomy in the absence of iron
cohort was increased in PV patients >65 years old (hazard
deficiency, or anemia despite cessation of cytoreductive ther-
ratio [HR], 8.6), with a history of prior thrombosis (HR,
apy), decreasing thrombocytosis, increasing leukocytosis,
4.85), or >65 years old and with thrombosis (HR, 17.3);
increasing splenomegaly, increased symptom profile (includ-
these represent the major factors used to assess thrombotic
ing increasing weight loss, night sweats, bone pain, fever),
risk in PV patients. In addition, cardiovascular morbidity
and finally increasing intramedullary fibrosis. The criteria
and mortality in PV were linked significantly to smoking,
for post-PV MF as agreed by the IWG-MRT are listed in
diabetes, and congestive heart failure.
Table 16-10 and include (i) being certain the disease was
The etiology of hypercoagulability in PV is not well under-
originally PV, (ii) a clear change in medullary fibrosis, and
stood, but hyperviscosity, due to uncontrolled erythrocytosis
(iii) a clear change in clinical phenotype. At least 25%-50%
qualitative platelet dysfunction, and granulocyte activation
of patients with post-PV MF develop AML, although not all
likely contribute to the pathogenesis of thrombosis in PV.
PV patients who transform to AML progress through a post-
Recent studies in patients treated with aspirin and hydroxy-
PV MF phase.
urea suggest that WBC count is an important predictor of
thrombotic risk in patients with controlled erythrocytosis,
MPN blast phase
suggesting that granulocyte activation contributes to throm-
bosis in patients treated with standard therapies. It is likely PV can transform to MPN-BP; criteria agreed upon by the
that patients with known genetic hyper-coagulable states IWG-MRT include persistently increased blasts in excess of
such as factor V mutation and concomitant PV are at higher 20% in the marrow or blood (Table 16-10). A baseline rate of
risk of thrombosis, although this has not been confirmed in MPN-BP from PV is considered to be modest, as MPN-BP
epidemiologic studies. The risk of postoperative thrombosis occurs in roughly 1%-3% of patients treated with phlebot-
or bleeding is increased in patients with PV. Up to 80% of omy alone (felt not to be leukemogenic). In contrast, certain
patients with PV who undergo surgery in the context of a therapies, which have been utilized in PV, are felt to increase
hematocrit or platelet count markedly above the normal the risk of MPN-BP, include Phosphorus-32 (32P) treat-
range will suffer a thrombohemorrhagic event. By contrast, ment, chlorambucil, busulfan, and alkylating agent combi-
only 5%-10% of patients suffer these perioperative compli- nations and are associated with increased risk of
cations when their counts are normalized for at least several transformation to MPN-BP (up to 15-fold increased risk in
weeks preoperatively. randomized PVSG trials). The ECLAP study noted a higher
Table 16-10 Criteria of progression of PV
Type of progression Criteria Details

Post-PV myelofibrosis Major criteria • Documentation of a previous diagnosis of polycythemia vera as defined by the WHO
(both major criteria criteria
and 2 minor criteria • Bone marrow fibrosis grade 2-3 (on 0-3 scale) or grade 3-4 (on 0-4 scale)
required) Minor criteria • Anemia or sustained loss of requirement of either phlebotomy (in the absence of
cytoreductive therapy) or cytoreductive treatment for erythrocytosis
• A leukoerythroblastic peripheral blood picture
• Increasing splenomegaly defined as either an increase in palpable splenomegaly of by
5 cm (distance of the tip of the spleen from the left costal margin) or the appearance of
a newly palpable splenomegaly
• Development of 1 of three constitutional symptoms: >10% weight loss in 6 months,
night sweats, unexplained fevers >37.5°C
MPN-BP (either criterion) Bone marrow ≥20% Blasts
Peripheral blood ≥20% Blasts with an absolute blast count of ≥1 × 109/L that lasts for at least 2 weeks

rate of AML/MDS transformation with pipobroman use;

Diagnosis of PV
this agent is no longer available in the United States but is
still available for use in Europe and elsewhere. Leukemias
that occur after chemotherapy or radiation rarely may be
Assess PV risk score All PV patients
associated with abnormalities of chromosome 5 or 7. and assess MPN Control of hematocrit (<45%)
Although early observational studies suggested that MPN- symptoms (MPN 10) Low-dose aspirin
BP might be increased in patients receiving hydroxyurea, the

largest prospective PV study to date, the ECLAP study,
Decide on need for concurrent cytoreduction based
enrolled 1,638 patients and noted no increase in MPN-BP in on PV risk and symptoms
patients treated with hydroxyurea, with median a follow-up No Yes
time of 8.4 years after PV diagnosis and 2.5 years after study
enrollment. There is no evidence that IFNα or anagrelide is
leukemogenic. Monitor for symptom Frontline cytoreduction
burden, vascular events, HU or HU-vs-IFN
progression clinical trial
Prognosis
Worsening
symptom burden
PV is a chronic disease that is incurable without stem cell Worsening
Vascular event,
symptom burden
transplant. Although early studies suggested that untreated Vascular event,
progression
HU resistance or
erythrocytosis and thrombocytosis lead to a median survival progression
intolerance
Phlebotomy
of only 18 months because of a high incidence of fatal throm- intolerance
boembolic events, more recent observational studies, includ-
ing younger individuals with PV and patients with chronic Consider ruxolitinib (PV)
or IFN (trial) or HU if
idiopathic erythrocytosis suggest that the clinical diagnosis of not previously received
PV often is preceded by an asymptomatic prodrome of many
years. Once the diagnosis is secure, treatment with phlebot-
Figure 16-4 A treatment algorithm for PV on the basis of vascular
omy, with aspirin, and with or without cytoreductive agent’s risk and symptomatic burden.
results in a near-normal life span for the average patient who
is diagnosed at a median of 60-65 years of age (Table 16-11).
Although it exceeds 10 years, however, the median survival of Assessing response in PV
individuals diagnosed with PV before 40 years of age is con-
siderably shorter than the life expectancy for a healthy person The European LeukemiaNet has published guidelines on the
of similar age, even with active management. expectation of therapy for PV (Barbui, Blood. 2013 Jun
6;121(23):4778-4781.) which are broken into categories of
complete remission, partial remission, no response or pro-
Management
gressive disease. Complete remission (CR) requires a durable
A stepwise approach to therapy of PV incorporates elements resolution of disease signs (symptoms and splenomegaly),
of hematocrit control, antiplatelet therapy, selective control of hematocrit without phlebotomy, control of leuko-
cytoreduction with hydroxyurea (hydroxycarbamide) as cytosis/thrombocytosis, absence of vascular events, absence
frontline therapy and role of ruxolitinib as second-line of progression, bone marrow histologic normalization and
therapy (Figure 16-4). absence of >1+ reticulin fibrosis. Partial remission (PR)
requires all features of complete remission except bone mar-
Table 16-11 PV risk (3 groups) survival row requirements. No response is any response falling short
Age, years >67 (5 points) of CR or PR. Molecular response remains experimental and
57-66 (2 points) not assessed as part of CR or PR.
<57 (0 points)
Leukocytes >15 × 109/L (1 point) vs. <15 × 109/L
Hematocrit control
Prior thrombosis Yes (1 point) vs no (0 points)
Risk group point Sum above points. Median survival: The mainstay of treatment of PV remains phlebotomy to
cutoffs/survival Low risk (0 points): 27.8 years maintain the hematocrit closer to a normal physiologic
Intermediate risk (1-2 points): 18.9 years range. Generally accepted goals include a hematocrit of
High risk (≥3 points): 10.9 years <45% for men. Although frequent phlebotomy is often
Adapted from Tefferi A, et al. Leukemia. 2013;27:1874-1881. required shortly after diagnosis, once the hematocrit is

within the desired range, the interval between phlebotomies Cytoreduction

may often be extended to every 3-6 months. Phlebotomy to
Hydroxyurea
the point of iron deficiency without concomitant cytoreduc-
tive therapy may be associated with reactive elevation in the Hydroxyurea, a ribonucleotide reductase inhibitor, is typi-
platelet count, although the thrombohemorrhagic risk of cally the first choice for most patients with PV requiring cyto-
reactive thrombocytosis in this clinical setting has not been reductive therapy. Hydroxyurea emerged as the cytoreductive
delineated. Iron supplementation should be avoided to pre- of choice based on PVSG studies showing a lower rate of
vent undue elevation in hemoglobin and hematocrit levels thrombosis compared to phlebotomy alone and lower risks of
and provide better phlebotomy control. secondary leukemia compared to chlorambucil and 32P. The
The target goals for phlebotomy have recently been evalu- mutagenic and leukemogenic potential of hydroxyurea has
ated by the CYTO-PV study conducted in Europe. This study been a subject of concern; but overall, the AML/MDS risk
began with the preconception that any hematocrit under a with chronic hydroxyurea therapy appears lower in magni-
value of 50% likely was equivalent in terms of prophylaxis tude compared to other cytoreductive agents, such as chlo-
against vascular events, but perhaps higher Hct targets would rambucil, 32P, pipobroman, and busulfan. Nevertheless,
lead to a decreased burden on patients by requiring fewer because of uncertainty regarding these concerns, hydroxyurea
phlebotomy procedures as well as hopefully a decrease in often is avoided in younger adults, and it should be used only
toxicities related to iron deficiency in this subgroup. Patients after a thorough discussion of the potential risks and benefits.
were randomized between a target hematocrit of 45%-50% Additional adverse effects of hydroxyurea include cytopenias,
or less than 45%. Individuals were allowed to have been on gastrointestinal disturbances, and, less commonly, chronic
cytoreductive therapy. The achieved median hemoglobin mucocutaneous ulcers. Because of the observations from the
was about 48% in the higher threshold group and 43%-44% PVSG regarding early thrombotic risk, hydroxyurea should
in the lower threshold group. Tthere was a significant be used in high-risk patients at the start of aggressive phle-
increased risk of vascular events in individuals in the cohort botomy. In general, patients with PV should avoid practices
managing hematocrit above 45%, with a hazard ratio of that augment their hypercoagulability or risk of vascular com-
about 3.8. This difference was statistically significant in all plications, such as smoking and the use of oral contraceptives
patient subgroups. This study has a significant impact on or estrogen hormone replacement therapy.
practice, and a hematocrit of 45% should now be considered
the ceiling of target hematocrit for PV.
Interferons
IFNs have demonstrated activity against many chronic

Antiplatelet therapy
myeloid neoplasms including chronic myeloid leukemia.
The ECLAP study, a double-blind randomized trial, com- Amongst patients with PV, both short-acting and longer-
pared low-dose aspirin with placebo among 518 patients acting IFNs have demonstrated ability to control erythrocy-
who had no indication for anticoagulation and no preexist- tosis, leukocytosis, and or thrombocytosis. Recent studies
ing clear indication or contraindication to aspirin therapy with pegylated IFNα have demonstrated significant clinical
and demonstrated that low-dose aspirin (ie, 100 mg/d) efficacy, including clinical and molecular remissions in a sub-
reduces the rate of thrombosis and cardiovascular deaths in stantial proportion of patients with improved tolerability;
patients with PV receiving standard phlebotomy and sup- current trials are aimed at assessing the efficacy and safety of
portive care. The aspirin-treated group experienced 60% pegylated IFNα in a larger cohort of PV patients. Two large
fewer major thromboses and cardiovascular deaths (3.2% vs global phase 3 clinical trial programs of IFN use in patients
7.9% absolute incidence) after roughly 3 years of follow-up. with PV continue to be ongoing, with their data eagerly
Low-dose aspirin also can effectively control erythromelalgia anticipated. The first, from the MPD Research Consortium
and other vasomotor symptoms in most patients. PVSG tri- (www.mpd-rc.org) is a randomized trial of newly diagnosed,
als showed that higher doses of aspirin (ie, 500-900 mg/d) high-risk patients with PV, randomizing them between
offer no added benefit but increase the risk of bleeding com- pegylated IFNα 2a (Pegasys) versus hydroxyurea. In parallel,
plications, especially when combined with dipyridamole. there is a randomized phase 3 for newly diagnosed, high risk
Elderly PV patients on high-dose aspirin are at highest risk PV patients with pegylated IFN proline, pegylated IFNα 2b
of bleeding. The ECLAP trial, however, observed only a mod- (AOP2014). IFNα therapy is safe during pregnancy, in con-
est increase in epistaxis and no increase in major bleeding on trast to hydroxyurea, which may be teratogenic (although
low-dose aspirin. The role of clopidogrel and other antico- experience from sickle cell anemia populations suggests that
agulants in thrombosis prevention in PV is not well defined, hydroxyurea is a low-risk agent, so abortion is not justified
and these agents are not considered standard of care. solely based on inadvertent fetal hydroxyurea exposure).

JAK inhibition considered acceptable. Duration of optimal time (ie, namely

extension beyond 6 months) of anticoagulation continues to
JAK2 inhibitor therapy was first tested in myelofibrosis
be unclear.
truly on the basis of unmet medical need and in concern for
Pruritus may be a particularly disturbing symptom that
safety until the safety signal of JAK inhibitors was fully iden-
often is unresponsive to phlebotomy or cytoreductive ther-
tified. The first trials of JAK inhibition in non-myelofibrosis
apy therapy. JAK inhibition with ruxolitinib has been helpful
patients was a phase 2 trial conducted between the US and
for some patients; additionally antihistamines, psoralen and
Europe in individuals with ET and PV who had previously
ultraviolet A phototherapy, cholestyramine, or selective
failed hydroxyurea. In PV, this study demonstrated the abil-
serotonin reuptake inhibitors (ie, paroxetine) may provide
ity to control PV-related MPN symptoms, control of hema-
symptomatic relief. Cytoreductive therapy with hydroxyurea
tocrit in individuals who had previously failed hydroxyurea,
or IFN (IFNα or pegylated IFNα) may help in refractory
and good tolerability. The dose-limiting toxicity of ruxoli-
cases. Painful splenomegaly and unacceptable hypercata-
tinib, which traditionally has been anemia or thrombocyto-
bolic symptoms usually require treatment with hydroxyurea
penia, was not a major factor in individuals with PV and ET.
or IFN (IFNα or pegylated IFNα). Splenectomy or splenic
A subsequent phase 3 trial was undertaken, called the
irradiation may be necessary for palliation in selected
RESPONSE trial, which administered ruxolitinib vs. best
patients who are intolerant of or unresponsive to cytoreduc-
alternative therapy in a cohort of PV patients who had
tive agents.
problematic disease defined by having been resistant or
intolerant to hydroxyurea, having palpable splenomegaly,
and having significant symptoms. This trial demonstrated Key points
clearly statistically superior control of hematocrit in the
ruxolitinib arm, control of reduction in splenomegaly by • PV is a clonal hematopoietic disorder characterized by
increased progenitor cell sensitivity to growth-promoting
volumetric studies, improvement in PV-related symptoms.
cytokines and is associated most commonly with activating
There was a trend for a reduced number of thrombotic
somatic mutations in the JAK2 tyrosine kinase, either JAK2 V617F
event in the ruxolitinib arm. Additionally, a phase 3 double-
or exon 12 mutations.
blind, double-dummy study of ruxolitinib vs hydroxyurea • PV (primary, absolute polycythemia) must be differentiated
in patients with inadequately controlled symptoms demon- from relative and secondary polycythemias. Secondary polycy-
strated improved symptomatic control of PV in ruxolitinib- themias may be either physiologically appropriate (elevated EPO
treated patients. induced by tissue hypoxemia) or physiologically inappropriate
The FDA approved ruxolitinib for patients with PV who (renal or ectopic overproduction of EPO or response to other
had had an inadequate response to hydroxyurea or could not erythropoietic stimuli). Familial polycythemic states often are
tolerate hydroxyurea in the fall of 2014. Ruxolitinib may be associated with truncating EPO receptor mutations (low
considered in other patients with PV and a very difficult endogenous EPO level) or VHL/HIF2A mutations (high endog-
symptomatic burden, such as increasing splenomegaly, severe enous EPO level).
• Unless erythrocytosis is an incidental finding, at diagnosis,
constitutional symptoms, difficult hematocrit control, or
patients with PV may exhibit a range of disease related MPN
potentially recurring vascular events despite hydroxyurea.
symptoms related to hyperviscosity, hypercatabolism, microvas-
cular events, or thromboembolic complications. These symp-
Additional therapeutic considerations toms include fatigue, pruritus, night sweats, complex headaches,
and problems with cognition.
Acute thrombotic events in patients with PV are managed • The median survival time of PV seems to be less than age
with therapeutic systemic anticoagulation in a similar man- matched controls but can be quite lengthy when patients are
ner to other patients who present with acute thrombosis. It is managed appropriately; thrombosis and bleeding are the major
also important to control the hematocrit and platelet count causes of morbidity and mortality in the first 10 years of disease.
to minimize progression or recurrence; phlebotomy to nor- Progression to post-PV MF or AML is associated with poor
malize the hematocrit quickly should be initiated if patients survival.
• Major risk factors for thrombosis with PV include increasing
have a hematocrit >45%. The utility of platelet-pheresis for
age, leukocytosis, and prior vascular events.
thrombocythemic patients with acute thrombosis and the
• The hematocrit and platelet count should be normalized for
optimal target platelet count after depletion is unknown.
several weeks before elective surgery to minimize the risk of
Antiplatelet therapy in addition to warfarin may be useful in perioperative thrombosis and bleeding in patients with PV.
selected cases of PV-associated arterial thrombosis, but only Postoperative thrombosis prophylaxis should be given whenever
after the acute event is stabilized with full anticoagulation possible.
and only if the potential additive risk of bleeding is

Essential thrombocythemia | 457
Diagnosis
Is there sustained primary thrombocytosis?
• All patients with PV should be phlebotomized to maintain the
hematocrit closer to a physiologic range (ie, <45% for men, The first key step in the diagnosis of ET is clinical suspicion,
<42% for women, and <37% during late pregnancy). Aspirin at and that suspicion should be based on the presence of sus-
a dose of ~100 mg/d should be routinely used for thrombosis tained thrombocytosis unexplained by other causes. Throm-
prophylaxis, unless a clear contraindication exists. bocytosis is defined as above the normal range for your
• Cytoreductive therapy for symptomatic and higher-risk PV
laboratory or, according to current diagnostic criteria, >450
patients should be considered with either hydroxyurea or IFN
× 109/L. In the past, the threshold for ET thrombocytosis was
(pegylated IFN). Phase 3 trials between IFN products and
at a level of 600 × 109/L but this value was reduced in WHO
hydroxyurea are undergoing randomized trials in newly
diagnosed PV patients.
2008 diagnostic criteria with the realization that there are
• Ruxolitinib is approved now for PV patients with inadequate patients with clear ET with associated molecular confirma-
(resistant or intolerant) response to hydroxyurea. Compared to tion (ie, JAK2 V617F, CALR, or MPL) with platelet counts
best alternative care, JAK inhibition demonstrated better control between 450 × 109 and 600 × 109/L.
of hematocrit, splenomegaly, and PV related symptoms. Secondary causes of thrombocytosis include many reac-
tive states including iron deficiency, infection, inflammation,
surgery, trauma, tissue injury or infarction, malignancy, and
Essential thrombocythemia the postsplenectomy state. Increased levels of inflammatory
mediators, including interleukin-1b (IL-1b) and IL-6 in
addition to IL-11 (a direct thrombopoietic stimulator), have
Clinical case been associated with reactive thrombocytosis. An elevated
A 40-year-old previously healthy female landscaper complained
C-reactive protein measurement is a surrogate marker for
of increased fatigue and numbness on her face and legs for 3 increased levels of IL-6, which can suggest an occult inflam-
weeks. She reported these symptoms to her internist, who per- matory process. The absolute value of the platelet count usu-
formed a CBC, which revealed a platelet count of 1,062 × 109/L. ally does not help distinguish reactive thrombocytosis from
She was then referred to a hematologist, whose evaluation ET, although reactive conditions infrequently cause eleva-
included iron studies and inflammatory markers (erythrocyte tions in platelet counts of >2,000 × 109/L.
sedimentation rate levels and C-reactive protein levels), which Finally distinguishing ET from other primary causes of
were within normal limits. FISH for BCR-ABL1 was negative. JAK2 thrombocytosis relies on the exclusion of other MPNs (espe-
was wild type, but she did have a mutation in calreticulin (CALR). cially occult PV, PMF, and CML) and MDS [especially
Marrow biopsy showed a hypercellular bone marrow (70%
5q– syndrome, chromosome 3(q21;q26) abnormalities, or
cellularity) with megakaryocytic hyperplasia. There was no mild
refractory anemia with ring sideroblasts associated with
reticulin fibrosis.
marked thrombocytosis (RARS-T)]. Primary thrombocyto-
sis can also be seen with a variety of congential syndromes,
Introduction and needs to be distinguished, in particular with children
and young adults with unexplained thrombocytosis. Idenit-
ET is about as prevalent as PV, with approximately 150,000
fying whether the platelet counts had been normal earlier in
cases in the United States (estimated from claims databases)
life and that thrombcytosis represents a change from base-
with an annual incidence rate of approximately 0.5 to 1.5
line can be very helpful. Additionally, there are familial cases
cases per 100,000 persons per year. The median age at diag-
of MPNs, but also familial cases of mutations in MPL dis-
nosis is approximately in the mid-seventh decade of life;
tinct from ET that can be drivers of thrombocytosis. Identi-
however, the distribution of cases is quite broad and include
fication of ET-related molecular markers can be quite helpful
many individuals who are young adults discovered at time
in equivocal cases.
of routine blood test analysis, pregnancy, or evaluation of
ET-related symptoms. Women with ET outnumber men
Diagnostic criteria for ET
with ET 1.5- to 2-fold, particularly among those with ET
diagnosed in the third to fifth decade of life. Morbidity and The WHO revised the diagnostic criteria for ET in 2008
mortality from ET in the first decade of disease are predom- (Table 16-12). The first criterion was a sustained thrombocy-
inantly related to thrombotic, vasomotor, and, less com- tosis in excess of 450 × 109/L as previously discussed. The
monly, hemorrhagic complications. Long-standing ET, second required that appearance of megakaryocytes were
similar to PV, can progress to either post-ET, MF, or even distinct from those seen in PMF (Figure 16-5). The third was
blast phase. that alternative myeloid neoplasm which can lead to

Table 16-12 WHO 2008 Diagnostic criteria for ET

Major 1. Sustained platelet count >450 × 109/L
2. Bone marrow biopsy specimen showing proliferation mainly of the megakaryocytic lineage with increased numbers of
enlarged, mature megakaryocytes; no significant increase or left-shift of neutrophil granulopoiesis or erythropoiesis
3. Not meeting World Health Organization criteria for polycythemia vera, primary myelofibrosis, chronic myelogenous
leukemia, myelodysplastic syndrome, or other myeloid neoplasm
4. Demonstration of JAK2 V617F or other clonal marker; or in the absence of a clonal marker, no evidence for reactive
thrombocytosis
Diagnosis Requires meeting all four major criteria
Adapted from Swerdlow SH, et al., eds. World Health Organization Classification of Tumours of Haematopoietic and Lymphoid Tissues. Lyon,
France: IARC Press; 2008.
thrombocytosis was not present. Finally, the last criterion is rarely used in the 21st century. Moderate leukocytosis may
was the demonstration of a molecular marker associated be found in up to one-half of ET patients. The peripheral
with ET. At the time of the WHO 2008, these included JAK2 blood smear often is notable for large or giant platelets
V617F (about 50%), or MPL (about 3-5%) and if neither (Figure 16-5) with occasional eosinophils, basophils, or cir-
was present than exclusion of other reactive states was a culating megakaryocyte fragments. Howell-Jolly bodies sug-
requirement (iron deficiency etc.). CALR was found to be gest hyposplenism; if the cause of this is not obvious (ie,
diagnostic in most of the other cases that are JAK2 and MPL prior splenectomy), then celiac disease, amyloidosis, or a
wild-type. A small minority of ET patients (<10-15%) are hemoglobinopathy resulting in functional hyposplenism
“triple-negative” for JAK2/CALR/MPL mutations. should be considered.
Specific assessment of BCR-ABL1 should be undertaken The significance of blood leukoerythroblastic features and
because CML can present with isolated thrombocytosis and their possible association with PMF rather than ET has been
a normal WBC count. Patients with isolated thrombocytosis investigated. One study suggested that patients with more
in association with BCR-ABL1 should be considered to have prominent circulating immature cells, teardrop erythrocytes,
atypical CML and should be managed accordingly. mild splenomegaly, and stainable marrow reticulin fibrosis
more commonly progressed to overt MF and had a signifi-
Additional laboratory features of ET cantly shorter survival, suggesting that their natural histories
were more consistent with PMF and that they might have
The leukocyte alkaline phosphatase (LAP) score typically is been classified as having “prefibrotic” or “early fibrotic” PMF
elevated in ET, as opposed to being reduced in CML; this test rather than ET. This area is complex and requires an experi-
enced morphologist and clinical correlation. Moreover,
although patients with evidence of leukoerythroblastosis
may be at higher risk for progression to post-ET myelofibro-
sis, the diagnostic and therapeutic strategy for patients with
clonal thrombocytosis does not differ based on the likeli-
hood for progression to overt MF.
Bone marrow and cytogenetic findings
Marrow evaluation is important in suspected ET cases to

assess for histopathologic features that are characteristic of
ET to confirm the presence of adequate iron stores and to
assess for fibrosis or other atypical features that might sug-
gest an alternative diagnosis. Increased numbers and clusters
of large megakaryocytes with hyperploid nuclei are seen in
~90% of marrow samples from patients with ET, and the
Figure 16-5 Increased numbers of platelets, including abnormally
bone marrow may be normocellular or only mildly hypercel-
large platelets (arrow), are characteristic of essential thrombocythemia. lular (Figure 16-5). Trilineage dysplasia and significant retic-
Similar platelet abnormalities may also be seen in other ulin and collagen fibrosis are commonly minimal or absent.
myeloproliferative disorders. From the American Society of If >15% ring sideroblasts are present, the rare provisional
Hematology Image Bank, #100445. entity of RARS-T rather than ET must be considered,

although the prognostic importance of this distinction is JAK2 mutational status, and the presence of additional
unclear. The presence of trilineage dysplastic morphologic cardiovascular risk factors (Table 16-13). The relative
features should raise the suspicion of MDS. importance of associated cardiovascular risk factors with
arterial thromboembolism has not been clearly demon-
strated; however, ET patients with a history of cigarette
Disease course and prognosis
smoking, diabetes mellitus, or hypercholesterolemia are at
The disease course of ET can range from a near normal life higher risk of cerebral or cardiac thrombotic events.
expectancy, to significant morbidity and mortality and death Although limited data exist, most experts consider oral
from progressive disease. ET is associated with vascular contraceptive use, estrogen replacement therapy, and other
events, constitutional symptoms, microvascular compromise acquired and congenital prothrombotic conditions to be
and vasomotor instability, and can potential progress along modifiers of thrombotic risk with ET. Although some
similar pathways of PV to either post-ET MF or to MPN-BP. studies have found an increased risk of bleeding in ET
Finally, it is crucial when discussing disease course to be cer- patients with a platelet count of >1,500 × 109/L, no data
tain the patient indeed has WHO 2008 defined ET, as opposed suggest that absolute platelet number is clearly predictive
to an early case of PMF (which can present before overt retic- of thromboembolic complications in the absence of other
ulin fibrosis is observed) and carries a worse prognosis. clinical risk factors. Nevertheless, controlling the platelet
count may be protective, and studies with hydroxyurea
(targeted at maintaining the platelet count at <600 × 109/L)
Vascular events in ET
and anagrelide (targeted at maintaining the platelet count
Between 10% and 50% of patients with ET will have a throm- at <400 × 109/L) have shown that these drugs decrease the
botic episode during the first decade after diagnosis, and 4% risk of primary or recurrent thrombotic events (from
will suffer a hemorrhagic complication. Older age and previ- 24%-36% during the 27-month observation period),
ous history of thrombohemorrhagic complications are the though whether this benefit is solely due to platelet control
major risk factors for vascular events in most studies. Com- or other drug effects is unclear.
plications of either type are uncommon in patients <40 years
of age. One prospective observational study found no signifi-
Risk in ET
cant increase in the risk of thrombosis (1.91 cases per 100
patient-years, with 4.1 years of median follow-up) among In a study of 800 WHO-defined ET patients, multivariate
untreated ET patients <60 years of age with a negative prior analysis found age, leukocytosis, and prior vascular events as
thrombohemorrhagic event history and a platelet count most prognostic for survival (Table 16-14). ET patients can
<1,500 × 109/L when compared with age-matched healthy be divided into low-, intermediate-, and high-risk groups
controls. Older age and a platelet count >1,500 × 109/L have (IPSET). With a median follow-up of 6.2 years, the 10-year
been cited as predictors of bleeding risk in ET; however, the estimated risk of MPN-BP was <1%, and thrombosis-free
influence of comorbid factors must be considered (Table survival at 10 years of 89%, 84%, and 69% for low-, interme-
16-13). Some ET patients with uncontrolled thrombocytosis diate-, and high-risk patients, respectively. Of note, this study
and clinical bleeding are found to have vWD, as described did not find a link between these risk scores and risk of pro-
previously for PV. gression to post-ET MF. Risk for thrombosis was based on
The overall risk of thrombotic complications in ET can advanced age, JAK2 mutational status, CV risk factors, and
be stratified based on age, history of prior thrombosis, prior thrombosis.
Table 16-14 ET risk for survival and thrombosis in WHO-defined ET

Table 16-13 Historical risk factors for vascular events in ET (3 groups)
Low risk (all of the following) Parameter 0 Points 1 Point 2 Points
Age <60 years old Age, years <60 — ≥60
No history of thromboembolism Leukocytes (× 109/L) <11 ≥11 —
No cardiovascular risk factors (smoking, hypercholesterolemia) History of thrombosis No Yes
Indeterminate risk
Neither low- nor high-risk disease Low risk 0 Points: Median survival not reached
High risk (one or both) (>30 years)
Age >60 years old 1-2 Points: Median survival 24.5 years
History of thromboembolism 3-4 Points: Median survival 13.8 Years
Adapted and modified from Tefferi A. Leuk Res. 2001;25:369-377. Adapted from Passamonti F et al. Blood. 2012 Aug 9;120(6):1197-1201.

Symptom burden in ET post-ET MF phase, is uncommon. Amongst WHO 2008

defined ET cases, rates of progression to MPN-BP are <5% at
A common misperception is that ET patients are asymptom-
10 years.
atic, whereas recent studies of ET patients utilizing the MPN-
SAF demonstrate that patients with ET can indeed be highly
symptomatic. Overall ET is less symptomatic, taking all Management
comers, than PV or PMF. However, well over 50% of ET
The management of ET is a stepwise process which begins
patients are symptomatic.
with appropriate use of antiplatelet therapy, and then decid-
ing on the need and value of cytoreductive therapy. Cytore-
Disease progression in ET ductive therapy currently includes significant data with use
Post-ET myelofibrosis of hydroxyurea or anagrelide. IFNα has also been found to
be a valid consideration as cytoreduction in ET, and JAK
If the patient was diagnosed previously with ET without the inhibitor approaches remain experimental in this MPN indi-
careful assessment of bone marrow aspirate and biopsy, it is cation (Figure 16-6).
possible the patient had early PMF even if they did not have
significant splenomegaly or anemia at diagnosis. Amongst
patients with originally confirmed ET, the confirmation of Antiplatelet therapy
progression to post-ET MF, the IWG-MRT, have consensus
Low-dose aspirin (ie, <325 mg/d) as primary prophylaxis for
criteria to define this progression (criteria listed in Table
ET patients is recommended in view of the beneficial results
16-15) and highlight these clinical features of progression.
with aspirin in PV in the ECLAP study and the favorable
Changes (ie, increasing allele burden) in current molecular
outcomes of patients treated with aspirin and hydroxyurea
markers (JAK2 V617F, CALR, or MPL) are not a feature of
in the PT-1 trial. Additionally, vasomotor symptoms and
progression, although karyotypic abnormalities are more
erythromelalgia are often responsive to therapy with aspirin.
common in post-ETMF than in ET. Exact rates of progres-
As with PV, the risk of bleeding in ET appears to be elevated
sion are difficult to accurately estimate given that past ET
with high doses (>325 mg/d) of aspirin. Therefore, lower
patient datasets are enriched for those with early PMF. For
aspirin doses, in the range of 75-300 mg/d, normally are
WHO 2008 defined ET, the rates of progression to post-ET
advised. As for PV, the role of clopidogrel or other antiplate-
MF have been reported to be <10% at 10 years.
let agents in ET is unknown. Whether aspirin is necessary in
asymptomatic low-risk patients with ET is unknown and
MPN blast phase
remains a point of clinical judgment.
Diagnostic criteria for MPN-BP from ET have the same cri-
teria independent of whether a patient originally had ET, PV,
Cytoreduction
or PMF (Table 16-15). However, in the absence of leukemo-
genic therapy for ET (radiophosphorus, alkylators), progres- Cytoreduction in ET in the frontline setting is considered in
sion to MPN-BP directly from ET, without an intervening individuals who either (i) are high risk by IPSET; (ii) have
Table 16-15 Criteria of progression of ET
Type of
progression Criteria Details
Post-ET Major criteria • Documentation of a previous diagnosis of essential thrombocythemia defined by the WHO criteria
myelofibrosis • Bone marrow fibrosis grade 2–3 (on 0–3 scale) or grade 3–4 (on 0–4 scale)
(both major, Minor criteria • Anemia and a ≥2g/dL decrease from baseline hemoglobin level
2 minor) • A leukoerythroblastic peripheral blood picture
• Increasing splenomegaly defined as either an increase in palpable splenomegaly of by 5 cm
(distance of the tip of the spleen from the left costal margin) or the appearance of a newly palpable
splenomegaly
• Increased LDH (above reference level)
• Development of 1 of three constitutional symptoms: >10% weight loss in 6 months, night sweats,
unexplained fevers >37.5°C
MPN-BP (either Bone marrow ≥20% Blasts
criterion) Peripheral blood ≥20% Blasts with an absolute blast count of ≥1 × 109/L that lasts for at least 2 weeks

aspirin (75-100 mg daily) and were randomized to receive

Diagnosis of ET either hydroxyurea or anagrelide, with a goal platelet count
of <400 × 109/L. After a median follow-up of 39 months,
patients in the anagrelide group were significantly more
Assess ET risk score
and assess MPN likely than those in the hydroxyurea group to have reached
symptoms (MPN 10) the adverse primary endpoint of thrombosis or serious hem-
Low-dose aspirin
orrhage. Specifically, compared with hydroxyurea plus aspi-
rin, patients receiving anagrelide plus aspirin had increased
Decide on need for concurrent cytoreduction based rates of arterial thrombosis, serious hemorrhage, and devel-
on ET risk and symptoms opment of marrow fibrosis but a decreased rate of venous
No Yes
thromboembolism. Patients receiving anagrelide were more
likely to withdraw from their assigned treatment because of
toxicity or treatment failure. Taken together, these two stud-
Monitor for symptom Frontline cytoreduction
burden, vascular events, HU, anagrelide, ies suggest that patients with ET at high risk of thrombosis
progression or HU-vs-IFN should be treated with low-dose aspirin and hydroxyurea
clinical trial
with a goal platelet count of 400 × 109 to 600 × 109/L.
Worsening A recent study compared anagrelide to hydroxyurea in
Worsening symptom burden
symptom burden Vascular event,
those with WHO-defined ET; this noninferiority study
Vascular event, progression (excluding aspirin in the anagrelide arm with concerns of
progression HU resistance or
intolerance
doubling up on the antiplatelet properties of anagrelide)
demonstrated no significant difference between hydroxy-
urea and anagrelide with regard to rate of platelet control
Consider IFN (trial) or
HU or anagrelide if and vascular events. Practice patterns of which agent to use
not previously received first varies, and the ELN guidelines present hydroxyurea as
frontline therapy and anagrelide as second-line therapy. In
Figure 16-6 A treatment algorithm for ET on the basis of vascular risk the United States and Japan, and in younger patients,
and symptomatic burden. anagrelide has been considered frequently as frontline ther-
apy by some investigators and hydroxyurea by others.
suffered prior vascular events; (iii) are intermediate-risk

Interferons
patients, especially if cardiovascular risk factors are present;
or (iv) have problematic ET-related symptoms unresponsive Among patients with ET, both short-acting and longer-act-
to aspirin therapy. Amongst frontline cytoreductive options ing IFN have demonstrated ability to control thrombocyto-
in ET, hydroxyurea and anagrelide are considerations. Three sis and/or leukocytosis. Recent studies with pegylated IFNα
important randomized trials of high-risk patients with ET have demonstrated significant clinical efficacy, including
inform regarding management. clinical and molecular remissions in a substantial proportion
The first study from Italy, a randomized trial of 114 of patients with improved tolerability; current trials are
patients including a placebo arm, convincingly showed a aimed at assessing the efficacy and safety of pegylated IFNα
role for hydroxyurea in decreasing thromboembolic events in a larger cohort of ET patients. One large global phase 3
in high-risk ET patients who either are >60 years of age, clinical trial program of IFN use in patients with ET contin-
have a previous history of thromboembolism, or both. A ues from the MPD Research Consortium (www.mpd-rc.org)
follow-up report of this study (median treatment time, 73 and is a randomized trial of newly diagnosed, high-risk
months) revealed a continued benefit for hydroxyurea: patients with ET, randomizing them between pegylated IFNα
45% of patients in the control group suffered a thrombotic 2a (Pegasys) versus hydroxyurea. IFNα therapy is consid-
event versus 9% of patients in the hydroxyurea group. Of ered reasonably safe during pregnancy, in contrast to
note, 1.7% of control patients and 3.9% of the group receiv- hydroxyurea and anagrelide, which may be teratogenic
ing hydroxyurea developed secondary myeloid malignan- (although experience from sickle cell anemia populations
cies (AML/MDS), a difference that was not statistically suggests that hydroxyurea is a low-risk agent, so abortion is
significant. not justified solely based on inadvertent fetal hydroxyurea
A second important randomized study in ET was the PT-1 exposure). Until phase 3 trials are complete, IFNs should be
trial. A total of 809 patients with ET at high risk of thrombo- considered second- or third-line therapy in ET, except in
sis were enrolled in the PT-1 study. They received low-dose pregnancy.

JAK inhibition
JAK inhibition has only been investigated in a limited way in
• The life expectancy of patients with ET is longer than for those
ET. Ruxolitinib was active in a phase 2 study of those with
with other MPNs, but patients are at risk for ET-related morbidity
hydroxyurea failure and did decrease thrombocytosis, ET
and mortality over time. Vascular events (bleeding, thromboem-
related symptoms, and splenomegaly. The reduction in bolic events [arterial >venous]), ET related symptoms, and
thrombocytosis was partial in this latter trial, and on this vasomotor complications affect most individuals eventually and
basis has not yet been moved forward to larger studies. At are the major causes of morbidity and mortality during the first
this juncture, JAK inhibition would be considered experi- decade of disease.
mental in ET. • Risk factors for bleeding and thrombosis should be identified
and considered in a risk-based treatment approach for patients
with ET.
Other drugs in trials for ET
• Cytoreductive agents and antiplatelet therapy (low-dose aspirin)
Other agents are in clinical trials for different situations in ET, are indicated for patients with ET, high-risk features, or poor
including long-acting formulations of anagrelide, histone disease control. Hydroxyurea is considered the first choice by most
deactylase inhibitors, and the telomerase inhibitor imetelstat. investigators because it was superior to anagrelide at preventing
most thrombohemorrhagic events in a randomized trial. IFNαα is
being investigated as a frontline therapy option for ET. Anagrelide
Additional therapeutic considerations can be a useful second-line agent, and some groups use it as a
Managing acute vascular events first-line therapy. Low-dose aspirin alone should be used for
low-risk patients and may effectively treat vasomotor symptoms.
In the setting of acute arterial or venous events, emergency • ET is the most common MPN detected in young women and
platelet-pheresis occasionally may be indicated to reduce the is associated with recurrent fetal loss. Aspirin and, if cytoreduc-
platelet count if it is extremely high, but this is not a data- tion is required, IFNαα can be used to strive for a favorable
driven practice. Standard heparin and warfarin therapy are pregnancy outcome, but data are limited.
indicated for venous thrombosis as for non-ET patients who
present with thrombosis. Large-vessel arterial thrombotic Myelofibrosis (primary and
events may require acute intervention with a heparin or post-ET/PV)
thrombolysis. Indefinite warfarin is reserved for patients
with recurrent events or if the primary event was catastrophic
Clinical case
and unprovoked in a patient with a normal platelet count on
cytoreductive therapy. In either circumstance, the patient A 71-year-old man with a remote history of prostate cancer diag-
should be monitored closely for bleeding while receiving nosed status post (s/p) radical prostatectomy, gout, and cholecys-
anticoagulation. Platelet-lowering agents should be used to titis presented with early satiety, discomfort over the left upper
maintain the platelet count in a safe range, ideally <400 × quadrant of the abdomen, night sweats, intermittent fevers, and
chills. Physical examination revealed an enlarged spleen (15 cm
109/L. Aspirin is indicated following an arterial thromboem-
below the left subcostal margin) and an enlarged liver. A CBC
bolic event if the patient is not already taking it. The useful-
showed leukocytosis (WBCs = 21 × 109/L), normocytic anemia
ness of aspirin after a venous thrombotic event is more
(Hb = 9.4 g/dL, mean corpuscular volume [MCV] = 88 fL), and
questionable, and aspirin treatment should be deferred until a normal platelet count (292 × 109/L). The WBC differential
warfarin therapy is discontinued. includes neutrophils = 67%, lymphocytes = 11%, monocytes =
1%, basophils = 2%. Review of the peripheral blood smear shows
Key points some circulating blasts, teardrop cells, nucleated red blood cells
(RBCs), and immature WBCs (myelocytes and metamyelocytes).
• ET diagnosis requires first excluding reactive thrombocytosis JAK2-V617F was detected with an allele burden of 34%. Meta-
and then excluding other myeloid neoplasms. Reactive or phase cytogenetics showed 46,XY,del(13)q in all 20 metaphases
secondary thrombocytosis may be due to iron deficiency, examined. A bone marrow biopsy showed a hypercellular marrow
infection, inflammation, tissue injury, or trauma, malignancy, with megakaryoyctic hyperplasia and atypia, marrow blasts were
and post-splenectomy state or functional hyposplenism. The 4%, grade 2 of 3 reticulin fibrosis with special stains.
platelet count alone does not distinguish reactive thrombocyto-
sis from ET.
• JAK2, CALR, or MPL mutations are present in 80%-90% of ET
patients; their presence proves the existence of a clonal myeloid Introduction
disorder but these mutations are not specific for ET, and their
MF is the most severe of the MPNs and includes three differ-
absence does not exclude a diagnosis of ET.
ent groups of patients: there are those with primary MF, and

Myelofibrosis (primary and post-ET/PV) | 463
MF that evolved from ET or PV (post-ETMF and post-PVMF) histopathologic, clinical, and laboratory features that distin-
(see the respective sections), and those who are said to have guish PMF from other myeloid and non myeloid diseases. In
post-ET or post-PV MF. Although there are likely differences most cases, the diagnosis of PMF is satisfied by finding
in terms of biological underpinnings between these three sub- increased bone marrow reticulin or collagen fibrosis, leuko-
sets of MF, management has been largely identical. PMF as an erythroblastic peripheral blood findings, and splenomegaly,
initial diagnosis is clearly less common than ET and PV. The in the absence of a secondary cause for these findings or of
annual incidence of PMF has been reported at between 0.2 to features better fitting ET or PV. CML should always be ruled
0.5 cases per 100,000 persons per year. The median survival out by molecular studies for the BCR-ABL1 rearrangement.
time is 3.5-5.5 years, although the natural history of PMF is JAK2 V617F, CALR, and MPL mutations demonstrate the
quite variable depending on the presence or absence of poor presence of clonal hematopoiesis, but their presence is not
prognostic features. A subset of low-risk patients will live lon- specific for PMF, and their absence does not exclude a diag-
ger than 10 years. The median age at diagnosis of PMF is ~65 nosis of PMF, although patients lacking all three (ie, triple
years, with 70% of cases diagnosed after 60 years of age and negative) are a small minority of PMF patients. Acute myelo-
approximately 10% of cases diagnosed at <45 years of age. fibrosis resulting from AML with megakaryocytic differen-
tiation (ie, M7 AML by French-American-British [FAB]
criteria and acute megakaryoblastic leukemia by WHO clas-
Diagnosis
sification criteria) or other primary myeloid disorders may
Suspecting PMF in a patient without a prior known hemato- be confused with PMF. In most cases, acute megakaryoblas-
logical neoplasm can be challenging in that the clinical fea- tic leukemia is distinguished by a rapid disease onset, pancy-
tures individually each have a broad differential diagnosis, topenia, mild splenomegaly, and a high frequency of marrow
including fatigue, splenomegaly (the differential diagnosis is myeloblasts with a megakaryocytic immunophenotype (ie,
more narrow for massive splenomegaly), anemia, leuko- positive for CD61). Many children who are diagnosed with
erythroblastosis, cachexia, vascular events, and intramedul- PMF probably have acute megakaryoblastic leukemia. Other
lary fibrosis noted at time of bone marrow done for other cases of acute myelofibrosis have a similar marrow histopa-
reasons (ie, workup of cytopenias, lymphoma staging, etc). thology and fulminant disease course akin to M7 AML but
Presenting with more than one of these features raises suspi- lack the 20% blast count required for the diagnosis of AML.
cion of PMF. Reasons to suspect Post ET/PV MF in patients Patients with MDS may present with fibrosis in their bone
who initially had established PV or ET are described above. marrow but usually lack the prominent splenomegaly and
peripheral blood leukoerythroblastosis typical of PMF. The
marrow in MDS with fibrosis should reveal trilineage dyspla-
Diagnostic criteria
sia without osteosclerosis. Clonal abnormalities of chromo-
The diagnostic criteria for Post ET/PV MF have been dis- some 5 or 7 may be detected in either MDS or PMF and thus
cussed in the context of PV and ET sections (Tables 16-10 do not exclude a diagnosis of PMF. Late-stage PV or ET,
and 16-15). Diagnostic features of PMF have been published where marrow fibrosis and myeloid metaplasia have devel-
by the WHO (Table 16-16) and incorporate molecular, oped, may be indistinguishable from PMF by morphologic
Table 16-16 Diagnostic criteria for PMF
Disease Criteria
Major 1. Presence of megakaryocyte proliferation and atypia, usually accompanied by either reticulin and/or collagen fibrosis, or, in
the absence of significant reticulin fibrosis, the megakaryocyte changes must be accompanied by an increased bone marrow
cellularity characterized by granulocytic proliferation and often decreased erythropoiesis (ie, prefibrotic cellular-phase disease)
2. Not meeting World Health Organization criteria for polycythemia vera, chronic myelogenous leukemia, myelodysplastic
syndrome, or other myeloid neoplasm
3. Demonstration of JAK2 V617F or other clonal marker (ie, MPL W515L/K), or, in the absence of a clonal marker, no evidence
of bone marrow fibrosis due to underlying inflammatory or other neoplastic diseases
Minor 1. Leukoerythroblastosis
2. Increase in serum lactate dehydrogenase level
3. Anemia
4. Palpable splenomegaly
Diagnosis Diagnosis requires meeting all three major criteria and two minor criteria.
Adapted from Swerdlow SH, et al., eds. World Health Organization Classification of Tumours of Haematopoietic and Lymphoid Tissues.
Lyon, France: IARC Press; 2008.

and clinical criteria. The biology, prognosis, and treatment of picture of peripheral blood leukoerythroblastosis. The LAP
PMF and postpolycythemic myelofibrosis or ET are identi- score in patients with PMF usually is elevated, although it
cal. Other malignant and nonmalignant causes of marrow may be normal or low in up to one-fourth of patients. Because
fibrosis are listed in Table 16-17. Of note, secondary marrow of the high marrow cell turnover, LDH, bilirubin, and uric
changes of increased reticulin fibrosis with megakaryocytic acid levels commonly are increased as well. Haptoglobin lev-
hyperplasia resembling PMF have been associated with sys- els may be decreased, and there may be other clinical and
temic lupus erythematous, Sjögren’s syndrome, psoriatic laboratory indicators of low-grade idiopathic hemolysis.
arthritis, and other chronic autoimmune diseases, most
commonly in the absence of splenomegaly or significant leu- Bone marrow histological features
koerythroblastic peripheral blood findings. These patients
frequently have constitutional symptoms and anemia with Approximately 20%-30% of patients with PMF are believed
positive antinuclear antibody and direct antiglobulin titers. to present in the prefibrotic stage. The early prefibrotic stage
Clinical, hematologic, and marrow fibrotic abnormalities in of PMF, also referred to as the “cellular” or “proliferative”
these conditions often respond to corticosteroids. stage, may be associated with thrombocytosis and modest
leukoerythroblastosis but may be difficult to distinguish
from ET. This condition may cause diagnostic difficulties in
Laboratory features
the absence of clear evidence of megakaryocyte atypia on
In addition to a high frequency of anemia, patients with PMF bone marrow biopsy. Splenomegaly may be present, and the
present with leukocytosis, leukopenia, thrombocytosis, or marrow reveals hypercellularity, left-shifted myeloid matu-
thrombocytopenia in 50%, 7%, 28%, and 37% of cases, ration, and increased megakaryocyte numbers with cluster-
respectively. Egress of immature cells from the marrow into ing and nuclear dysplasia. The fibrotic stage of PMF is
the blood is characteristic of this disorder. Erythroid precur- associated with reticulin or collagen fibrosis in addition to
sors may account for up to 20% of the circulating nucleated more characteristic clinical and peripheral blood changes
cells, and circulating blast cells are found in up to 30% and more prominent megakaryocyte atypia (Figure 16-6).
of cases at diagnosis (blasts of >20% indicate AML). In most With increasing degrees of fibrosis, a diagnostic marrow
cases, the peripheral blood smear reveals the typical leuko- aspirate often is unobtainable, yielding a “dry tap.” Progres-
erythroblastic features of immature myeloid cells, nucleated sive medullary fibrosis is characterized by accumulation of
red blood cells, teardrop erythrocytes, and large platelets. extracellular reticulin fibers (revealed by silver staining) and
Although these findings are sensitive for the diagnosis of collagen (revealed by trichrome staining). In advanced stages
PMF, they are not highly specific. Secondary causes of mar- of PMF, the hematopoietic space is completely replaced by
row fibrosis (Table 16-17) also may give rise to a similar fibroblasts and extracellular matrix material (Figure 16-7).
Osteosclerosis may develop in some cases.
Table 16-17 Differential diagnosis of PMF
Acute myelofibrosis (acute megakaryoblastic leukemia, AML-M7) Disease course and prognosis
Myelodysplasia with fibrosis
The course MF is variable and includes risk of vascular
Late-stage PV, ET, or CML with evolution to myelofibrosis
Malignant causes of secondary myelofibrosis
events, although overall the rate of arterial and venous vas-
Hairy cell leukemia cular events are probably lower than those in the untreated
Hodgkin lymphoma counterparts with PV and ET. Splenic infarctions, portal
Non-Hodgkin lymphoma vein thrombosis, as well as events in other vascular distribu-
Plasma cell dyscrasias tions, including deep vein thrombosis, pulmonary emboli,
Acute lymphoblastic leukemia and arterial thrombosis occur.
Metastatic carcinoma Overall, patients with MF are more symptomatic than
Multiple myeloma those with ET or PV. The profile of symptoms also tends to
Chronic myelomonocytic leukemia be different in myelofibrosis than with ET and PV. There
Systemic mastocytosis tend to be fewer symptoms, which are associated with hyper-
Eosinophilic leukemia viscosity, namely issues such as erythromelalgia, dizziness,
Nonmalignant causes of secondary myelofibrosis difficulties with concentration, but more difficulties with
Granulomatous infections (tuberculosis, histoplasmosis) hypercatabolic symptoms, such as night sweats, and fatigue
Paget disease is more pronounced. Symptoms directly relating to the
Autoimmune disorders (eg, systemic lupus, Sjögren syndrome, spleen are clearly increased including pain/abdominal dis-
psoriatic arthritis, primary autoimmune myelofibrosis)
comfort, early satiety, undesired weight loss, and fevers, the

a b

c d

Figure 16-7 Bone marrow histopathologic findings in various stages of chronic idiopathic myelofibrosis. (a) The early fibrotic, cellular stage is
characterized by hypercellularity and myeloid hyperplasia (hematoxylin-eosin [H&E] stain; original magnification, ×85); with (b) variable
reticulin fibrosis (reticulin stain; magnification, ×85). (c) The later fibrotic stage is associated with extensive collagen deposition and loss of
hematopoietic precursors, except for occasional residual megakaryocytes (H&E stain; magnification, ×85); (d) with more extensive reticulin
fibrosis (reticulin stain; magnification, ×85). Photos courtesy of Steven J. Kussick (University of Washington, Seattle, WA).
latter being almost exclusively associated with myelofibrosis patients, nearly 75% will have a hemoglobin value less than
compared to ET or PV. normal, with approximately 50% having hemoglobin value
of <10g/dL and 25% being red cell transfusion dependent.
Cytopenias
Splenomegaly
Cytopenias are most common in MF among the MPN, with
anemia being by far the most common. Anemia in the setting Splenomegaly appears to largely originate from splenic
of myelofibrosis is complex and potentially multifactorial, sequestration of immature myeloid cells, and this leads to
including underproduction from the marrow, splenic seques- partially ineffective EMH in that there is little evidence that
tration, and myelosuppression from medication. In post-PV the spleen is actually a useful source of circulating hemato-
MF, there can be anemia as carryover from iron deficiency poietic elements in the patient with myelofibrosis. Splen
from earlier stages of disease in which phlebotomy was ectomized patients with myelofibrosis have typically seen
employed. Additionally, bleeding from variceal portal hyper- cytopenias improve, and although there is histological evi-
tension or other sources should always be considered in the dence of EMH, no evidence exists that the spleen is a produc-
differential diagnosis of worsening anemia. Among PMF tive source of hematopoietic elements in MF. The exact

relationship between the spleen and the pathogenesis of worsening cytopenias rather than consequences of myelo-
myelofibrosis is incompletely understood, but it is clearly proliferation (splenomegaly as an example).
known that the splenomegaly can be massive; more than
70% of patients with primary myelofibrosis have palpable
Prognosis in MF
splenomegaly and many will have enlargement to a degree
that can lead to pain, splenic infarction, abdominal discom- The outlook for PMF is more variable than for other MPNs.
fort, early satiety, and masking of weight loss. A number of studies have identified clinical and laboratory
features that predict a more aggressive disease course and
shorter survival in patients with PMF (Table 16-18). In gen-
Disease progression in MF
eral, morbidity and mortality are related to hematopoietic
Most patients die while the disease remains in the MF phase failure, thrombosis, hypersplenism, advanced age, and evo-
of the disease. Causes of death include cardiovascular com- lution to AML. AML is the cause of death in 5%-30% of
plications, infection, thrombosis, and further debilitation of patients. The risk of developing AML is increased among
the illness exacerbating underlying comorbidities. Progres- patients with severe anemia and a high number of circulat-
sion to MPN-BP can be seen as a cause of death in up to 30% ing immature myeloid cells.
of patients with myelofibrosis with the same diagnostic crite- A hemoglobin level of <10 g/dL is the most consistent
ria outlined in the sections on PV and ET; namely >20% adverse prognostic indicator for patients with PMF. Age also
myeloid blasts in the bone marrow or >20% circulating appears to be important; patients <55 years of age at diagno-
myeloid blasts in the peripheral blood with >1.0 × 109/L sis have a median survival of 8-10 years, whereas the median
absolute myeloid blast count in the peripheral blood (Tables survival of older patients (variably defined as >55, >60, or
16-10 and 16-15). Some patients with myelofibrosis progress >65 years of age) is 3-5 years in most studies. The presence of
to MDS, in some cases because of prior therapy they had constitutional symptoms, increased circulating peripheral
received, such as alkylator therapy, which can lead to ther- blasts, and elevated WBC (>25 × 109/L) also contribute to
apy-associated MDS. Others that likely were MPN/MDS poor outcomes in MF. These clinico-pathologic factors
overlap syndromes progress along an MDS phenotype with became the basis for the development of the International
Table 16-18 The three prognostic scoring systems currently used in PMF
IPSS DIPSS DIPSS-Plus

Risk factor (No. of points) HR (95% CI) (no. of points) HR (95% CI) (no. of points) HR (95% CI)
Age >65 years 1 1.95 (1.61-2.36) 1 1.98 (1.52-2.60) DIPSS low = 0 –
Constitutional symptoms* 1 1.97 (1.62-2.40) 1 2.06 (1.61-2.65) DIPSS Int-1 = 1 1.9 (1.2-3.1)
Hemoglobin <10 g/dL 1 2.89 (2.46-3.61) 2 4.18 (3.03-5.78) DIPSS Int-2 = 2 3.6 (2.1-6)
WBC count >25 × 109/L 1 2.40 (1.83-3.14) 1 1.33 (1.33-2.29)
Blood blasts >1% 1 1.80 (1.50-2.17) 1 1.82 (1.39-2.4) DIPSS-high = 3 7.3 (4-13.3)
RBC transfusion
dependence – – – – 1 1.4 (1.1-2)
Thrombocytopenia
(<100 × 109/L) – – – – 1 1.6 (1.2-2.2)
Unfavorable karyotype† – – – – 1 2.4 (1.7-3.4)
Median survival Median survival Median survival
Risk group No. of factors (years) No. of factors (years) No. of factors (years)
Low 0 11.3 0 NR 0 15.4
Intermediate-1 1 7.9 1-2 14.2 1 6.5
Intermediate-2 2 44 3-4 4 2-3 2.9
High >=3 2.3 5-6 1.5 4-6 1.3
Data from Cervantes F, et al. Blood. 2009;113:2895-2901; Passamonti F, et al. Blood. 2010;115:1703-1708; and Gangat N, et al. J Clin Oncol.
2011;29:392-397.
NR = not reached; DIPSS = Dynamic International Prognostic Scoring System; CI = confidence interval; HR = hazard ratio; IPSS =
International Prognostic Scoring System.
* Constitutional symptoms include fever, night sweats, weight loss >10% from baseline on the year prior to diagnosis.
† Unfavorable karyotype includes complex karyotype, one or two abnormalities that includes +8, -7/7q-, i(17q), -5/-5q, 12p-, inv(3), 11q23
rearrangement.

Prognostic Scoring System (IPSS). The IPSS, however, was Management

designed primarily to evaluate prognosis only at the time of
The management of myelofibrosis requires assessment of
original diagnosis. A subsequent prognostic scoring system
both the patient’s prognostic expectations, utilizing either
that takes into account changes in the risk profile brought
the DIPSS or perhaps in an evolving pattern weaving in
about by acquisition of other disease factors during the
molecular data using the newly minted MIPSS scoring sys-
course of the disease was established and designated the
tem. Additionally, a thorough understanding of the diffi-
Dynamic International Prognostic Scoring System (DIPSS).
culty the patient is experiencing in terms of symptom
The same factors were considered in both scoring systems
burden using the MPN-SAF or the MPN10 is helpful
except that hemoglobin <10 g/dL was given a higher score (2
(Figure 16-8).
points) compared with other risk factors in the DIPSS. An
updated version of this prognostic scoring system called
DIPSS-Plus takes into account transfusion dependence,
Observation
platelet counts <100 × 109/L, and the presence of certain
chromosomal defects. Among 116 younger PMF patients, a Given the variable prognosis and presentation in MF, the
hemoglobin <10 g/dL, the presence of constitutional symp- initial diagnosis does not necessarily lead to a therapeutic
toms, and circulating blasts >1% were independent adverse intervention. In the setting of post-ET/PV MF, the pace of
predictors and should be used to determine whether younger disease and burden the patient experiences is variable, and in
patients undergo evaluation for allogeneic SCT. In that the absence of worsening disease burden (ie, worsening
cohort, patients with 0 or 1 adverse factors had a median symptoms, problematic splenomegaly, increasing blasts,
survival of 176 months, whereas patients with 2 or 3 adverse problematic cytopenias) patients might be best remaining
factors had a median survival of only 33 months. The pres- on their preexisting ET/PV therapy until the MF declares
ence of more adverse factors predicts for worse overall sur- itself to be more problematic. In those with asymptomatic,
vival and increased risk for AML transformation. low-risk PMF and likely low-risk molecular features, a case
Prognostic scores in MF continue to evolve, and new can be made for observation. Experimental trials in low-risk
scores require validation before being accepted into medical PMF and slowly progressive post-ET/PV MF with a goal of
practice. In the future, combining all of these factors of prog- delaying disease progression would be reasonable and are
nosis, clinical, mutational, karyotypic, and potentially being considered with pegylated IFN or other agents, which
comorbidities will further help to refine the prognosis. might slow disease progression.
Diagnosis of MF
(primary, post-ET, or post-PV MF)
Calculate DIPSS MF score and

assess MPN symptoms (MPN 10)
Low risk Low risk Intermediate to high risk

Median survival = Median survival < Assess role and timing of allogeneic SCT (donor, risk, candidate)
185 months 185 months Allogeneic SCT: urgent, delayed, or never
Low symptoms High symptoms
Delayed or not
Urgent allogeneic SCT
allogeneic SCT
Observation Possible role of Proceed to JAK2 inhibitor JAK2 single-agent failure

vs IFN trial JAK2 inhibitor allogeneic SCT (unless anemia or Refractory cytopenias
(trial) or IFN (possible prior JAK2 cytopenias are
(trial) inhibitor) (trial) main problem)
Clinical trials:
Ruxolitinib combination
Nonruxolitinib JAK2 inhibitor
New targeted approaches
Figure 16-8 A treatment algorithm for MF on the basis of prognosis and symptomatic burden.

Stem cell transplantation JAK1 and JAK2 and is the first pharmacologic agent to be
FDA approved in MF. The pivotal phase 3, multicenter, dou-
Initially, it was a concern that the marrow fibrosis character-
ble-blind, placebo-controlled, randomized controlled trial,
istic of MF would lead to graft failure and that SCT might be
Controlled Myelofibrosis Study with Oral JAK Inhibitor
contraindicated. Studies using standard ablative condition-
Treatment (COMFORT-I), showed at least a 35% spleen size
ing regimens demonstrated that a subset of patients achieve
reduction assessed by radiologic imaging (magnetic reso-
long-term clinical and molecular remission after SCT and
nance imaging or computed tomography [CT] scan) at 24
that engraftment failure is uncommon. Results using unre-
weeks in 41.9% of patients in the ruxolitinib arm compared
lated donors were equivalent to those with HLA-matched
with just 0.7% in the placebo. Furthermore, decreases in the
sibling transplantations. An early multi-institutional experi-
total symptom score by >50% at 24 weeks were better in
ence, including patients who were predominantly <55 years
the ruxolitinib-treated patients (45.9%) compared with those
old and received a marrow or peripheral blood stem cells
receiving placebo (5.3%). A European counterpart also was
predominantly from related donors, demonstrated a 47%
conducted called COMFORT-II, which compared ruxoli-
5-year probability of survival and a 40% rate of histologic
tinib versus best available therapy (BAT). In COMFORT-II,
and hematologic remission.
a spleen volume reduction of at least 35% was noted in 28%
In a recent study of 21 patients (age 27-68 years), reduced-
of patients on ruxolitinib compared with 0% in patients on
intensity conditioning resulted in long-term disease-free
BAT at 48 weeks. Similarly, quality-of-life measures and dis-
survival in the majority of patients with acceptable toxicities.
ease-related symptoms were better in the ruxolitinib treated
Moreover, patients treated with reduced-intensity condi-
patients. Emerging data show that patients with MF treated
tioning regimens can achieve stable donor chimerism, sig-
with ruxolitinib who have high-risk disease and have at least
nificant improvement in blood counts, and significant
a >10% reduction in spleen size have better overall survival
decrease in marrow fibrosis.
compared with the control arm of the trials and matched
historical controls. Of note, patients who are JAK2 V617F
When to consider SCT in MF positive or negative respond equally to ruxolitinib, such that
mutational testing is not required to guide the use of JAK
The decision to undertake stem cell transplantation is a com-
inhibitor therapy. The main side effects of ruxolitinib are
plex one. Transplant is primarily a consideration for indi-
reversible, dose-dependent anemia and thrombocytopenia,
viduals with intermediate- to high-risk MF; the risk of
and care must be taken in using these agents in patients with
transplantation probably outweighs the benefits in those
preexisting anemia or thrombocytopenia. Now, with over 3
with low-risk disease. In individuals with intermediate-one
years of follow-up data, responses seem durable, and the
risk disease, there is the greatest lack of clarity and other new
favorable impact on survival remains present. The impact of
molecular markers, such as triple negativity for JAK2 V617F,
ruxolitinib on allele burden has been minimal (of unknown
CALR, and MPL, as well as the presence of ASXL1, may help
significance), and impact on progression to AML seems
further refine that choice. Additional factors that are consid-
favorable for ruxolitinib-treated patients; however, the stud-
ered are an individual’s age, the presence of a donor, the
ies were not powered for this as an endpoint.
quality of match, and an individual’s own risk philosophy.
Two JAK inhibitors are currently in phase 3 clinical testing
Many approaches are being investigated prior to stem cell
and may be available outside of the setting of the clinical trial
transplantation, including the use of JAK inhibition in trying
in the near future. The first, pacritinib, is a JAK2 and FLT3
to optimize the course; if JAK inhibition is used pretrans-
inhibitor which in phase 2 studies demonstrated the ability
plant, questions regarding duration and best approach for
to improve splenomegaly and symptoms but did not seem to
weaning JAK inhibition prior to transplant or overlapping it
cause significant anemia or thrombocytopenia. Phase 3 clin-
with a conditioning regimen remain. Pretransplant treat-
ical trials utilizing this agent in individuals with underlying
ments may also include iron chelation as well as cytoreduc-
significant thrombocytopenia are ongoing. The second
tion in individuals who have an increased blast percentage in
agent, momelotinib, is a JAK1 and JAK2 inhibitor in which
the blood or bone marrow.
early trials demonstrated abilities to improve MF-related
splenomegaly and symptoms while not exacerbating anemia
Medical therapy or in some individuals improving anemia. Ongoing Phase 3
studies are enrolling comparing randomizing patients
JAK inhibition
between momelotinib and ruxolitinib to assess for possible
The high prevalence of both JAK2 mutations became the frontline therapy and potential superiority around the aspect
basis for the development of novel JAK inhibitors. The first of anemia. Multiple combination strategies for use with rux-
of its class, ruxolitinib is a potent and selective inhibitor of olitinib are undergoing in clinical trials using agents ranging

from PI3 kinase inhibitors, SMO inhibitors, and immuno- their cytogenetic profile. Pomalidomide has had activity in
modulatory drugs, and androgens such as danazol. MF-associated anemia and thrombocytopenia. Randomized
clinical trials show that this response rate was inadequate to
justify FDA registration of this medication.
Cytoreduction
Splenectomy is associated with a mortality rate of 6%-8%,
The cytoreductive medications used in ET and PV still have with even a much higher rate of morbidity attributed to
occasional utility in the setting of myelofibrosis, specifically abdominal thrombotic events, postoperative hemorrhage,
hydroxyurea, anagrelide, and IFN. All of these agents can postsplenectomy leukocytosis and thrombocytosis, and its
reduce thrombocytosis; hydroxyurea and IFN having poten- use was limited due to the need for highly skilled surgical care
tial benefits when leukocytosis is problematic. Leukocytosis as well as the risks involved. The advent of JAK inhibitor ther-
requiring intervention is present in MF primarily in the set- apy and the significant impact on splenomegaly have
ting of postsplenectomy or in individuals that are having sig- decreased the need and consideration for splenectomy. How-
nificantly progressive disease and a leukocyte count in excess ever, splenectomy can be successfully employed in individuals
of 50 × 109-75 × 109/L are seen. Clinical trials have helped to who have failed JAK inhibitor therapy, but would be consid-
demonstrate that these cytoreductive therapies are not, how- ered a salvage option. Given the risk involved, most would
ever, overly impactful in improving MF-associated spleno- not recommend splenectomy in JAK inhibitor naïve patients.
megaly or symptoms as they have all been used in the control Splenectomy pre–stem cell transplantation had been consid-
arms or best alternative therapy arms of clinical trials, such ered in the past as potentially a resource for expediting count
as COMFORT-II with minimal impact. Their major utiliza- recovery poststem cell transplantation, but given the risks
tion is in control of counts in the postsplenectomy setting in and the morbidity of the procedure this is not recommended.
patients with myelofibrosis or in individuals who have resid- Finally, despite the massive splenomegaly which can be char-
ual thrombocytosis that has not normalized in the setting of acteristic of myelofibrosis, splenic rupture fortunately
other MF-directed therapy such as JAK inhibitor therapy. remains a rare complication of myelofibrosis, and if it were to
There have been patients who have benefited from concur- occur, splenectomy would need to be considered.
rent cytoreduction in addition to JAK inhibitor therapy
when JAK inhibitor therapy has been inadequate to also con- Key points
trol thrombocytosis.
• MF includes both post-ET/PV MF patients as well as those with
PMF. Management of all subtypes of MF is similar.
Cytopenia-directed therapy • More than two-thirds of patients with PMF are diagnosed at
age 60 years or older. Although the median survival time is only
Anemia in MF has responds to a variety of different 3.5-5.5 years, it is quite variable and clinical features and
approaches. ESAs are most helpful in patients with serum molecular findings help refine prognostic expectations.
EPO levels <100 IU/L who have not yet progressed to trans- • PMF must be differentiated from advanced PV, ET, CML, acute
fusion dependence. Splenomegaly may worsen during ESA megakaryoblastic leukemia, myelodysplasia with fibrosis,
therapy. Androgens, including danazol, oxymetholone, nan- infiltrative malignant processes, and infectious or autoimmune
drolone, and testosterone, can lead to anemia or platelet diseases associated with reactive marrow fibrosis.
responses in 10-35% of patients. A small subset of patients • The majority of patients with PMF develop anemia, spleno-
with myelofibrosis who have evidence of hemolytic anemias megaly, and significant symptoms during the course of their
disease.
can respond to corticosteroids. Immunomodulatory drugs
• Therapeutic approaches to PMF are guided by assessment of
also can have an impact on myelofibrosis-associated anemia,
prognosis, MF symptom burden, and degrees of cytopenias.
perhaps from impact on the intramedullary cytokine milieu
Consideration of curative hematopoietic stem cell transplant is
which may be inhibiting hematopoiesis. Thalidomide and an important consideration in higher-risk patients who are
prednisone are usually dosed at 50 mg of thalidomide a day excellent candidates.
with a corticosteroid taper of predisone (starting dose, • Therapy with JAK inhibition has been very impactful by
0.5-1 mg/kg/d) reduced over 3 months. Thrombocytopenia decreasing disease associated splenomegaly, improving
responses can also be seen with thalidomide and prednisone. MF-related symptoms, and decreasing disease associated
Although responses have occurred in less than 50% of morbidity and mortality.
patients, there have been anecdotal reports of patients having • Splenectomy can be considered for palliation of symptomatic
years of transfusions independent response to thalidomide massive splenomegaly in individuals who have failed JAK
therapy. Toxicities can include sedation and neuropathy. inhibition; however, postsplenectomy complications include
perioperative mortality, rebound thrombocytosis, and massive
Lenalidomide can have activity in MF-associated anemia in
hepatomegaly.
those individuals that have the deletion 5q chromosome in

Special management situations surgical procedures should proceed as necessary, but with
across MPNs awareness of a higher risk of vascular complications, particu-
larly in those with uncontrolled thrombocytosis, erythrocy-
Pregnancy tosis, or leukocytosis in the postoperative period and that
The presence of an MPN increases the risk of miscarriage as appropriate postoperative vascular prophylaxis should be
well as abruptio placentae, preeclampsia, and intrauterine considered. In general, the estimation of surgical risk is
growth retardation. Based on the age distribution of MPN dependent upon the procedure itself. For example, there is a
patients, the MPNs of ET and PV are by far the most likely to high risk of deep vein thrombosis in orthopedic procedures.
be an issue in the woman of childbearing potential. Patients The underlying risk of the individual and their prior history,
may either have a preexisting MPN prior to conception, or including whether they have required prior anticoagulation,
there are individuals in which the blood tests obtained after have had prior thrombotic events, and have required prior
the time of a known pregnancy can uncover the underlying cytoreduction is also considered. Even amongst individuals
MPN. It has been demonstrated that fetal loss in women who previously have not been on cytoreduction, if they
with ET is significantly higher than the general population undergo an elective surgical procedure of high risk, in the
with potential risk factors having been identified, including ideal world, preoperative cytoreduction to normalization of
previous pregnancy complications as well as the presence of the counts as well as DVT prophylaxis postoperatively
the JAK2 V617F. It is felt that the risk factors that have been should be considered. In general, as a theme, MPN patients
associated, specifically with ET and PV in the arena of undergoing elective surgical procedures despite all prophy-
thrombosis, remain very relevant in complications during lactic measures do have a higher risk of events and should
pregnancy. Venous thrombosis is increased in general in have active concurrent care by the hematologist with the sur-
woman during pregnancy, and this risk seems to be higher in gical team.
individuals with MPNs, including in the postpartum period
and those with a higher risk of preexisting vascular events. Splanchnic vein thrombosis
The European LeukemiaNet has published guidelines for
treatment strategy for MPNs during pregnancy. In low-risk Abdominal vein thrombosis, including Budd-Chiari syn-
pregnancies, it is recommended to control the hematocrit in drome, portal vein occlusion, and mesenteric vein thrombo-
patients with preexisting PV to <45% or a mid-gestation- sis, are all more frequently encountered in patients with
specific range, whichever is lower. The use of low-dose aspirin MPNs; and, indeed, screening of individuals, even in the
or prophylactic low-dose molecular-weight heparin may be absence of thrombocytosis or other overt features of MPNs,
recommended after delivery until six weeks postpartum. High- do find a significant prevalence of either JAK2 V617F or
risk pregnancies are defined as having previous venous or arte- other MPN-related mutations in this group. The exact natu-
riothrombotic events, whether pregnant or not during that ral history of individuals not meeting WHO diagnostic crite-
time, previous hemorrhage attributed to MPN, previous preg- ria for MPNs but having a splanchnic vein thrombosis and
nancy complications that may have been caused by a MPN, an MPN-defining lesion remain unclear. The European Leu-
and marked sustained increase in a platelet count of >1,500 × kemiaNet has recommendations for low-molecular-weight
109/L. High-risk patients may require low-molecular-weight heparin followed by lifelong anticoagulation with an inter-
heparin throughout pregnancy, monitoring for bleeding com- national normalized ratio in the range of 2.0 to 3.0 if warfa-
plications. If the platelet count is >1,500 × 109/L, the hematolo- rin is used. For patients with thrombocytosis, hydroxyurea
gist may utilize IFNα. If there has been previous major bleeding, should be used to normalize platelets as soon as possible.
avoid aspirin and consider IFNα to reduce thrombocytosis. Currently, ongoing phase 2 studies are investigating the
Finally, if individuals were on preexisting cytoreductive ther- potential benefits of IFNα in this group of patients even with
apy for risk of their disease, amongst the major MPN therapies, normal or even low blood counts at the time of their event.
only IFNα is felt to be safe during pregnancy. No drug is actu-
ally approved and licensed for use during pregnancy, but the MPN blast phase
risk profile in high-risk patients is typically felt to be acceptable
with the use of IFNα, whereas hydroxyurea, anagrelide, and The therapy of MPN-BP is challenging in that the life expec-
ruxolitinib are either known or suspected teratogens. tancy of individuals with MPN-BP is exceedingly short
and the MPN-BP is frequently refractory to induction che-
motherapy. It has been found that approximately 40%-50%
Elective surgery in MPN patients
of patients with MPN-BP treated with AML-like induction
Elective surgical procedures likely have an increased risk of chemotherapy will return to a more chronic phase of an
having postoperative vascular complications. Emergency MPN, but usually only for a short amount of time. In this

Other MPNs | 471
setting, if stem cell transplant can be performed, it should encoding for the receptor of colony-stimulating factor 3
occur in a rapid fashion. In general, it is viewed that MPN- (CSF3R) in a cohort of CNL patients. Maxson, et al hypoth-
BP is a setting most appropriate for clinical trials and that esized that patients with CNL (and atypical CML) would
induction should be only utilized in individuals who are harbor oncogenes that would be sensitive to kinase inhibi-
considered to be stem cell transplant candidates, and most tion. Using a deep sequencing approach with coverage of
other patients should receive palliative approaches. The dif- 1,862 genes, they found that 16 of 27 (59%) harbored CSF3R
ficulty of therapy of MPN-BP highlights the importance of mutations, including 8 of 9 with CNL. Two types of muta-
consideration of stem cell transplant earlier in the course of tions were observed: membrane proximal mutations and
the disease in individuals that have high-risk features. Cyto- frameshift or nonsense mutations that truncated the cyto-
reductive therapy with hypomethylating agents has been uti- plasmic tail of CSF3R. The influence upon downstream sig-
lized in individuals with MPNs. naling pathways, and subsequently, sensitivity to kinase
inhibiton, differed depending on the type of mutation: trun-
cation mutations activated the SRC family-TNK2 kinase sig-
Other MPNs naling and showed sensitivity to dasatinib, whereas proximal
mutations activated the JAK-STAT pathway. As proof of
Chronic neutrophilic leukemia
concept, a patient carrying a JAK-STAT activating CSF3R
mutation experienced clinical improvement in neutrophilic
Clinical case leukocytosis and thrombocytopenia when treated with rux-
A 64-year-old previously healthy executive noticed a change in olitinib. In a murine model, transplantation with the predo-
her abdominal girth for about 3 months. This was accompanied moninant mutation linked with diagnosis of CNL, CSF3R
by a feeling of bloatedness, early satiety, occasional nausea, and T618I, recapitulated a fatal MPN characterized by granulo-
intermittent episodes of itching. She decided to have a routine cytic proliferation and infiltration of the liver and spleen;
blood test at a local clinic and was found to have the following JAK inhibition with ruxolitinib reduced the leukocyte count
CBC results: WBCs = 27 × 109/L, Hb = 12.9 g/dL, hematocrit = and spleen weight. Subsequently, in another study, 10 of 12
40%, MCV = 8 fL, platelet count = 315 × 109/L, absolute neu- (83%) WHO-defined CNL cases were found to carry CSF3R
trophil count (ANC) = 25 × 109/L; occasional metamyelocytes
mutations; 33% coexpressed SETBP1 mutations.
and myelocytes were noted but accounting for 5% of WBCs, no
myeloblasts were seen. She was referred to a hematologist who
noted hepatosplenomegaly and mild cervical lymphadenopathy Diagnosis
by physical examination. A bone marrow aspiration and biopsy
was subsequently performed, which showed increased numbers While some patients have an incidental discovery of leuko-
of neutrophilic granulocytes, a hypercellular marrow (95%), no cytosis, others present with fatigue and constitutional symp-
dysplastic changes, and 3% myeloblasts. Metaphase cytogenet- toms, such weight loss and night sweats. Splenomegaly is the
ics showed 46, XX [20]. Molecular testing for BCR-ABL1, PDGFRA, most frequently found clinical feature in patients with CNL.
PDGRB, FGFR1, and JAK2 V617F were all unremarkable. Subse- Some patients will present with gastrointestinal tract bleed-
quently, her physician sent peripheral blood for CSF3R mutation ing, thrombocytopenia, pruritus, and gout. Transformation
testing, which returned positive for T618I. to acute leukemia has been reported.
CNL is defined by the WHO as having a sustained, nonre-
Chronic neutrophilic leukemia (CNL) is a very rare active leukocytosis >25 × 109/L, with >80% segmented/band
chronic MPN recognized as distinct entity by the 2008 WHO neutrophils, <10% immature granulocytes, and <1% periph-
classification. CNL has been a historically challenging diag- eral blasts. Bone marrow biopsy demonstrates hypercellular-
nosis to make, requiring exclusion of reactive neutrophilia ity with a striking neutrophil proliferation with myeloid-
and other MPNs, such as typical and atypical CML and to-erythroid ratio reaching up to 20:1. Blasts or promyelo-
chronic myelomonocytic leukemia (CMML). The incidence cytes are not increased in number; dysplasia and reticulin
and prevalence of CNL is difficult to estimate and males and fibrosis are not evident. Cytogenetic studies are normal in
females appear to be equally affected. CNL occurs more ~90% of cases of CNL, but in the remaining cases, clonal
commonly in older patients (often in the seventh decade), karyotypic anomalies may include +8, +9, +21, del(20q),
though adolescent cases have been described. del(11q), and del(12p). Other MPNs should be excluded,
and there should be no evidence of BCR-ABL1, PDGFRA/
PDGFRB, or FGFR1 mutations. Given the recent discovery
Pathobiology
of CSF3R mutations in CNL patients, it is expected that this
In 2013, understanding of the genetic basis of CNL was molecular genetic marker will be incorporated into revised
improved with the demonstration of mutations in the gene diagnostic criteria, aiding diagnostic capability.

Course and prognosis

The clinical course of CNL is heterogeneous. Disease accelera-
primary care doctor who noticed palpable lymph nodes in the
tion often manifests with the development of progressive neu-
neck and axillary regions and a palpable spleen tip by physical
trophilia with resistance to previously effective therapy,
examination. Routine blood work showed some normocytic
progressive splenomegaly, or worsening thrombocytopenia, anemia (Hb = 10.1 g/dL, MCV = 92 fL); leukocytosis (WBCs =
or with cytogenetic clonal evolution. Blastic transformation 25 × 109/L) with increased lymphocytes (40%), monocytes
was reported to occur in a significant proportion of patients at (28%), and eosinophils (12%); and mild thrombocytopenia
a median of 21 months from diagnosis. Progressive neutro- (platelets = 97 × 109/L). Review of blood work 6 months prior
philia associated with anemia and thrombocytopenia have showed similar CBC findings. The patient saw a hematologist
been reported, as has transformation to myelodysplasia and and underwent a bone marrow aspiration and biopsy, which
AML. Although CNL is regarded as a relatively slowly progres- showed dysplastic changes in the erythroid and megakaryo-
sive disease with survival ranging from 6 months to >20 years, cytic lineages with 5% blasts in the bone marrow. The biopsy
one retrospective analysis of 40 patients with CNL reported a showed a hypercellular marrow with spindle-shaped mast cell
infiltration grade of 50%. Flow cytometry of the bone marrow
median survival time of 23.5 months. Most common causes of
aspirate showed increased CD25 expression on mast cells. A
death included intracranial hemorrhage (n = 9), progressive
KITD816V mutation also was identified. Metaphase cytogenet-
disease (n = 5), blastic transformation (n = 4), infection (n = 1),
ics showed 46, XX [20]. Total tryptase level was 450 ng/mL.
and treatment-related complications (n = 1). The patient was diagnosed with systemic mastocytosis with
associated hematologic nonmast cell disease, specifically chronic
myelomonocytic leukemia.
Management
Optimal treatment for patients with CNL remains to be

defined. Splenectomy has resulted in worsening of neutro- Mastocytosis encompasses a heterogeneous spectrum of
philic leukocytosis and cannot be recommended. Treatment disorders characterized by mast cell proliferation and accu-
of CNL to date has consisted largely of cytoreductive agents, mulation (Figure 16-2). Clinical manifestations of mast cell
such as hydroxyurea, where clinical responses occur but lack disorders are caused by uncontrolled proliferation of tissue
durability. Similar to other chronic MPNs, IFNα has been mast cells and the release of mast cell-derived mediators.
used. Allogeneic sibling SCT can be curative, but is usually While cutaneous mastocytosis (CM) is usually a chronic,
is reserved for patients with accelerated or blastic transfor- indolent disease, systemic mastocytosis (SM) can be either
mation. Given the potential for blastic transformation and indolent or more aggressive and even life-threatening. Given
progressive refractory neutrophilia, however, allogeneic that SM has a spectrum of clinicopathologic features in com-
SCT may be appropriate for younger patients. The role of mon with MPNs, the revised 2008 WHO classification
tyrosine kinase inhibitors in the treatment of CNL is included SM under the broader umbrella of MPNs. Masto-
intriguing, but not yet defined. In the first report of CSF3R cytosis can be classified according to clinocopathologic
mutations in CNL, Maxson, et al described a patient with and laboratory findings (Table 16-19). The incidence of
a membrane proximal mutation (CSF3R T618I) and mastocytic disorders is poorly defined; SM is felt to be a very
improvement in neutrophilic leukocytosis and thrombocy- rare disease. Although mastocytosis can be diagnosed at any
topenia when treated with ruxolitinib. In another report, a age, CM is more common in children, whereas SM occurs
patient with a membrane proximal mutation (also CSF3R predominantly in adults. These disorders appear to have a
T618I) and a SETBP1 mutation was refractory to ruxoli- slight male predominance.
tinib and hydroxyurea. The safety and efficacy of ruxolitinib
in CNL (and atypical CML) are currently under investiga-
tion in a multicenter study (clinical trials identifier: Pathobiology
NCT02092324). Mast cells are long-lived hematopoietic cells with unique
biologic properties and a unique spectrum of mediators and
Systemic mastocytosis cell surface antigens. Mature mast cells are best known for
their involvement in allergic inflammation mediated by
Clinical case allergen-specific immunoglobulin E (IgE) and tend to reside
in diverse organs, often in close vicinity to smaller or larger
A 67-year-old retired woman has been experiencing fever, blood vessels. Mast cell survival depends largely on stem cell
chills, diarrhea, a persistent urticaria-like rash, flushing, and factor (SCF); KIT is the protein TK receptor for SCF, and is
palpitations for the past 5 months. She decided to see a
expressed by mast cells, and accordingly, most cases of

Other MPNs | 473
a b
Figure 16-9 Bone marrow involvement with systemic

mastocytosis. (a) Marrow biopsy shows areas of scattered
infiltration or complete replacement by elongated, spindle-
shaped cells (hematoxylin-eosin stain; original magnification,
×85). (b) Tryptase staining of the core marrow biopsy
identifies the infiltrating cells as mast cells (magnification, ×85).
(c) Abnormal mast cells also may be identified in a marrow
aspirate specimen (arrows); however, this is not a reliable assay
for diagnosis (Wright-Giemsa stain; magnification, ×425).
Photos courtesy of Steven J. Kussick (University of Washington,
Seattle, WA).
Table 16-19 The 2008 WHO classification of mastocytosis mastocytosis are associated with somatic-activating point
Cutaneous mastocytosis mutations of KIT. The most common point mutations
Urticaria pigmentosa/maculopapular cutaneous mastocytosis result from a Val for Asp substitution at codon 816
Diffuse cutaneous mastocytosis (D816V), which is found in ~80% of SM patients (skin,
Solitary mastocytoma of skin peripheral blood, and bone marrow) and results in ligand-
Systemic mastocytosis independent activation of KIT, promoting mast cell pro-
Indolent systemic mastocytosis liferation and survival. A KIT juxtamembrane mutation
Bone marrow mastocytosis in codon 560 also has been described in a human mast cell
Smoldering systemic mastocytosis line called HMC-1 and rarely is found in SM. Rare juxta-
Aggressive systemic mastocytosis membrane and transmembrane variants of KIT point
Systemic mastocytosis with associated clonal hematologic non–mast mutations also have been described, as well as alternative
cell lineage disease KIT D816 codon mutations such as D816Y/H/F/I.
Mast cell leukemia Other somatic mutations, including TET2, DNMT3A,
Mast cell sarcoma ASXL1, SF3B1, and CBL mutations also have been identi-
Extracutaneous mastocytoma fied in a subset of mastocytosis patients, particularly in
Adapted from Horny H-P, et al. In: Swerdlow SH, Campo E, et al., eds. those with an associated hematological non–mast cell dis-
World Health Organization Classification of Tumours of Haematopoietic ease (AHNMD). In a study of 39 KIT D816V-mutated SM
and Lymphoid Tissues. Lyon, France: IARC Press; 2008. patients, Schwaab et al reported that the presence of

additional somatic mutations (most frequently TET2, SRSF2, osteopenia, fractures). Anaphylaxis after a bee sting can also
ASXL1, CBL, and RUNX1) were more common in those with indicate underlying mastocytosis.
advanced SM and contributed to inferior survival. When SM The diagnosis of CM is confirmed by the demonstration
is diagnosed in conjunction with another hematologic neo- of pathologic mast cell infiltration of the skin. SM requires
plasm (~30%-40% of cases), the underlying neoplasm is involvement of at least one extracutaneous tissue by clonal
typically of myeloid rather than lymphoid origin. Mutations mast cells (bone marrow is the most commonly involved
in KIT D816V have been identified in both the mast cell and organ). Diagnostic criteria for CM, SM, and variant presen-
AHNMD compartment, which potentially may indicate a tations of SM are summarized in Table 16-20.
shared pathogenetic origin in a hematopoietic progenitor.
Patients with ISM appear to have more of a pure KIT D816V-
driven disease. Apart from organ infiltration, the conse-
quences of mastocytosis stem from mediator release as mast Life expectancy can be quite variable, ranging from only a
cells are activated and degranulate. Mast cells contain a vari- few months in aggressive SM variants to nearly normal life
ety of mediators, including histamine, inflammatory cyto- spans in more indolent disease. CM in children tends to have
kines such as IL-3 and IL-16, and tumor necrosis factor. In an indolent course and often is associated with spontaneous
addition, mast cells contain secretory granule proteases, regression. Adults with CM rarely may evolve to SM. The
most commonly tryptase which is increased in mast cell dis- presence of cutaneous involvement in SM appears to confer
eases. An increase in tryptase levels serve as a minor criterion an indolent behavior, whereas lack of skin involvement is
for diagnosis (unless AHNMD is present), and while the level associated with aggressive behavior. Predictive factors of
itself cannot distinguish SM subgroups, marked increases are poor prognosis in SM include older age at onset of symp-
seen in more advanced/aggressive subtypes. Additionally, toms, absence of CM, low platelets, hypoalbuminemia, hepa
measurement serves as a practical means of assessing mast tosplenomegaly, anemia, and elevated LDH.
cell burden and monitoring response to therapy.
Management
Diagnosis
Treatment of CM includes H1 and H2 antihistamines, cro-
The presenting clinical features for patients with mastocyto- molyn and other mast cell stabilizers, topical or intralesional
sis depend on the extent of organ infiltration, consequences glucocorticoids, and psoralen and ultraviolet A photother-
of mediator release, and whether or not a nonmast cell dis apy. Adults with chronic CM may require long-term con-
order is also present. Cutaneous manifestations of masto tinuous or intermittent symptomatic treatment.
cytosis typically include a reddish-brown maculopapular For adult patients with indolent variants of SM, treatment
eruption (urticaria pigmentosa) or, less often, a diffuse ery- of mediator related symptoms with combinations of H1 and
thema, plaques, or nodules. The classic description of urti- H2 antihistamines, anticholinergic drugs, proton pump
caria following stroking of the skin is coined as the Darier inhibitors, cromolyn and other mast cell stabilizers, may be
sign. Telangiectasia macularis eruptiva perstans, character- sufficient to alleviate symptoms. Patients with SM should
ized by red-brown macules with irregular borders and a tel- carry epinephrine in an injectable form available at all times
angiectasia-like appearance, is a less common cutaneous for managing anaphylaxis. Aspirin and nonsteroidal anti-
manifestation. Cutaneous manifestations may be the only inflammatory drugs have been helpful for some patients
consequence of mast cell disease in children. Blistering can with flushing and syncope, but hypersensitivity to these
occur in pediatric patients and represents an aggressive form drugs is relatively common and must be excluded. Accord-
of urticaria pigmentosa. ingly, a major goal in the management of mastocytosis is the
Mastocytosis is typically systemic in adults and includes avoidance known triggers, which can include opioid analge-
bone marrow infiltration; major and minor criteria for bone sics, such as morphine and codeine, which are known mast
marrow involvement are described in Table 16-5. Other cell degranulators, as well as anesthesia, stress, and infection.
organs commonly involved include the liver, spleen, lymph IFNα can be helpful for patients with painful skeletal lesions
nodes, and gastrointestinal mucosa. Clinical features of SM or mast cell tumors that threaten bony integrity. Those with
are categorized in four distinct groups: (i) constitutional osteoporosis can be treated with calcium and/or bisphos-
symptoms (eg, fatigue, fever, weight loss); (ii) cutaneous phonate therapy when indicated.
manifestations, as described above; (iii) systemic mediator- The aggressive variants of SM may progress to end-stage
related symptoms (eg, abdominal pain or bloating, dyspepsia, organ fibrosis or failure and may be complicated by patho-
diarrhea, flushing, headache, hypotension, anaphylaxis); and logic fractures, severe cytopenias, or both. Patients with evi-
(iv) musculoskeletal complaints (eg, bone pain and myalgias, dence of end-organ damage without major bony complications

Other MPNs | 475
Table 16-20 WHO criteria for diagnosis of cutaneous and systemic mastocytosis
Cutaneous mastocytosis
Skin lesions demonstrating the typical clinical findings and typical infiltrates of mast cells in a multifocal or diffuse pattern in an adequate skin
biopsy. Absence of features/criteria for the diagnosis of SM.
Systemic mastocytosis
The diagnosis of SM may be made if one major criterion and one minor criterion are present or if three minor criteria are fulfilled.
Major criterion
Multifocal, dense infiltrates of mast cells (≥15 mast cells in aggregates) detected in sections of bone marrow and/or other extracutaneous
organ(s).
Minor criteria
a. In biopsy sections of bone marrow or other extracutaneous organs, >25% of the mast cells in the infiltrate are spindle shaped or have
atypical morphology or, of all mast cells in bone marrow aspirate smears, >25% are immature or atypical mast cells.
b. Detection of KIT point mutation at codon 816 in bone marrow, blood, or other extracutaneous organ(s).
c. Mast cells in bone marrow, blood, or other extracutaneous organs that coexpress CD117 with CD2 and/or CD25.
d. Serum total tryptase persistently >20 ng/mL (unless there is an associated clonal myeloid disorder, in which case this parameter is
not valid).
Indolent systemic mastocytosis
Meets criteria for SM.
No evidence of an associated clonal hematologic non–mast cell lineage disease.
No “C” findings.
Mast cell burden is low, and skin lesions are almost invariably present.
* Bone marrow mastocytosis: bone marrow involvement, but no skin lesions.
* Smoldering systemic mastocytosis: with two or more “B” findings but no “C” findings.
Aggressive systemic mastocytosis
One or more “C” findings.
No evidence of mast cell leukemia.
* Lymphadenopathic mastocytosis with eosinophilia (provisional subvariant): progressive lymphadenopathy with peripheral blood
eosinophilia, often with extensive bony involvement and hepatosplenomegaly, but usually without skin lesions. Exclude cases with
rearranged PDGFRA.
Systemic mastocytosis with associated clonal hematologic non–mast cell lineage disease
Associated clonal hematologic non–mast cell lineage disorder (MDS, MPN, AML, lymphoma, or other hematologic neoplasm that meets the
criteria for a distinct entity in the WHO classification).
Mast cell leukemia
Diffuse bone marrow infiltration by atypical immature mast cells. Bone marrow aspirate contains >20% mast cells. Usually >10% circulating
mast cells on peripheral blood.
“B” findings
1. Bone marrow biopsy showing >30% infiltration by mast cells (focal, dense aggregates) and/or serum total tryptase level >20 ng/mL.
2. Signs of dysplasia or myeloproliferation in non–mast cell lineage, but insufficient criteria for definitive diagnosis of hematopoietic neoplasm
by WHO, with normal or only slightly abnormal blood counts.
3. Hepatomegaly without impairment of liver function, and/or palpable splenomegaly without hypersplenism, and/or palpable or visceral
lymphadenopathy.
“C” findings
1. Bone marrow dysfunction manifested by one or more cytopenia (ANC <1 × 109/L, Hb <10 g/dL, or platelets <100 × 109/L), but no frank
non–mast cell hematopoietic malignancy.
2. Palpable hepatomegaly with impairment of liver function, ascites, and/or portal hypertension.
3. Skeletal involvement with large-sized osteolysis and/or pathologic fractures.
4. Palpable splenomegaly with hypersplenism.
5. Malabsorption with weight loss due to gastrointestinal mast cell infiltrates.
Adapted from Horny H-P, et al. In: Swerdlow SH, et al., eds. World Health Organization Classification of Tumours of Haematopoietic and
Lymphoid Tissues. Lyon, France: IARC Press; 2008.

may benefit from IFNα with or without corticosteroids, 29 months and was 44 months in responders. Responses
although most responses are only partial. Single-agent were also noted in those with mast cell leukemia (44%
cladribine or 2-chlorodeoxyadenosine, given at 5 mg/m2/d or major response). The median change in mast cell burden
0.13-0.17 mg/kg/d as a 5-day treatment cycle every 4-6 weeks was 59% among 72 evaluable, and the decrease in tryptase
induced clinical and laboratory responses (ie, decreased neared 60%. Grade 3/4 toxicities encountered with the use
serum tryptase and urinary histamine metabolites) in patients of midostaurin included neutropenia (5%), leukopenia
with symptomatic SM. Patients who fail to respond to these (4%), febrile neutropenia (3%), anemia and thrombocyto-
interventions and those who progress to mast cell leukemia penia (each 3%), nausea (6%), vomiting (6%), fatigue
should receive multiagent antileukemic or investigational (4%), and increased amylase/lipsase (3% and 4%). Masi-
chemotherapy. tinib is another tyrosine kinase inhibitor (wild-type KIT)
Allogeneic SCT should be considered for younger patients that could represent an emerging treatment for indolent
with aggressive SM who achieve a remission with chemo- mastocytosis. In a phase 2 study of 25 patients with symp-
therapy. In a retrospective analysis of 57 patients, 28% tomatic SM or CM, an overall clinical response of 56% was
achieved CR after transplant; for all subtypes, the OS was noted, with improvements in pruritus and flushing. Adverse
57% at 3 years, including 74% for SM-AHNMD, 43% with events included edema, muscle spasm, and rash; one patient
ASM, and 17% for mast cell leukemia. Symptomatic SM in developed reversible agranulocytosis.
the presence of a nonmast cell clonal hematologic disease
should be treated as indicated both for the hematologic
Myeloproliferative neoplasm, unclassifiable
malignancy and for the SM complications. Generally, the
underlying non-SM malignancy determines the overall clini- The term MPN, unclassifiable (MPN-U) should be used to
cal course, although in cases in which aggressive forms of SM describe only those patients who meet clinical, laboratory,
like aggressive systemic mastocytosis (ASM) coexist with and morphologic criteria of MPNs but who fail to present
low-grade myeloid neoplasms, the aggressive SM may take features of any single MPN entity or patients who present
precedence. with overlapping features of two or more MPN entities. The
The crucial role of KIT in normal mast cell development demonstration of pathognomonic molecular abnormalities,
and the evidence that KIT mutations may be important in such as BCR-ABL1 fusion or the PDGFRA, PDGFRB, or
SM pathogenesis prompted treatment of mastocytosis FGFR1 rearrangements, excludes the diagnosis of MPN-U.
patients with TKIs. Imatinib mesylate, because of its inhibi- In the era of newly discovered mutations, it is expected that
tory properties against KIT, was the first to enter the clinical the number of MPN-U will likely decrease. The exact inci-
arena. The presence of KIT D816V mutation confers resis- dence, median age at onset, and sex distribution of MPN-U
tance to imatinib mesylate by affecting the catalytic pocket are not truly known.
of the KIT protein, preventing imatinib from binding and
exerting its inhibitory activity. Therefore, KIT mutation
Clinical features
analysis is important in therapeutic decision making in SM.
A trial of imatinib is a consideration in those with aggres- The clinical features of patients with MPN-U is variable, as
sive SM who lack the D816V mutation. Other TKIs also this is a heterogeneous group of disorders. Patients can pres-
appear to have promising effects in patients with ent with minimal to no organomegaly and well-preserved
KIT-D816V-positive SM. Midostaurin is a novel TKI that peripheral blood counts in the very early stages of the disease
has displayed potent activity (inhibitory concentration at or massive organomegaly, extensive myelofibrosis, and
50% of 30-40 nM) against wild-type and KIT-816 mutants. severe cytopenias in advanced cases. Unexplained portal or
On the basis of these data, in one study, midostaurin 100 mg splanchnic vein thrombosis may be the initial presenting fea-
twice a day was administered to patients with aggressive SM ture in these patients.
in continuous 28-day cycles until progression or intolerable
toxicity. In a preliminary report of the fully accrued trial,
the overall response rate was 60%, including a major
response rate of 75%, good partial response rate of 21%, The clinical course and prognosis for patients with MPN-U
and a minor partial response of 4%. Major response was can be extremely heterogeneous. Patients with early stage
defined as normalization of albumin levels, resolution of disease can safely be followed every 6 months and gener-
liver transaminitis and other liver function tests, relief of ally will develop features of unique MPN entities. Patients
ascites and pleural effusion, improvement of hemoglobin in whom unique MPN entities are no longer recognizable
and platelet levels, and reversal of weight loss. After a tend to have aggressive clinical courses and very poor
median of 26 months, overall survival was estimated near prognosis.

Other MPNs | 477
Chronic eosinophilic leukemia, not cough, pruritus, diarrhea, angioedema, and muscle pain.
otherwise specified End-organ damage can be a manifestation of a direct eosino-
phil leukemic infiltrate or secondary to the release of cyto-
kines or contents of toxic granules. The most serious clinical
Clinical case
findings relate to endomyocardial fibrosis due to eosino-
A 35-year-old male graduate student came to the university philic infiltration of the heart, leading to constrictive peri-
health clinic because of nonproductive cough, diarrhea, fatigue, carditis, fibroplastic endocarditis, myocarditis, or intramural
intermittent fevers (102°F [38.9°C]), and muscle aches. He ini- thrombus formation (due to scarring of the mitral or tricus-
tially attributed these symptoms to stress, but presented due to pid valves). Peripheral and central nervous system findings
persistence over a 2-month period. A CBC showed the following:
can include mononeuritis multiplex, peripheral neuropathy,
WBCs = 19 × 109/L, Hb = 11.5 g/dL, MCV = 83 fL, platelets =
and paraparesis, as well as cerebellar involvement, epilepsy,
188 × 109/L, ANC = 12 × 109/L, and AEC = 3.4 × 109/L. There
dementia, cerebral infarction, and eosinophilic meningitis.
were 3% circulating blasts in the peripheral blood. Workup for
connective tissue diseases, parasitic infections, and allergies was
Pulmonary involvement includes idiopathic infiltrates,
unremarkable. His subsequent evaluation by a hematologist fibrosis, pulmonary effusions, and pulmonary emboli. Skin
CBC confirmed an eosinophilic leukocytosis and a bone marrow manifestations are common and can take many forms,
aspiration and biopsy showed 6% bone marrow blasts with no including angioedema, urticaria, papulonodular lesions, and
dysplastic changes. Metaphase cytogenetics were normal (46, erythematous plaques. Gastrointestinal involvement by
XY [20]). He had no abnormalities in PDGFRA, PDGFRB, FGFR1, eosinophilia can result in ascites, diarrhea, gastritis, colitis,
and BCR-ABL1 was absent. pancreatitis, cholangitis, or hepatitis.
The WHO criteria for diagnosis of CEL-NOS require the
CEL is characterized by an autonomous, clonal proliferation presence of eosinophilia (1.5 × 109/L); a clonal cytogenetic or
of eosinophil precursors resulting in persistent elevation of molecular abnormality or blasts cells >2% in the peripheral
eosinophils in the peripheral blood, bone marrow, and blood or >5% in the bone marrow; lack of BCR-ABL1, PDG-
peripheral tissues. By definition, in CEL, not otherwise speci- FRA/PDGFRB, or FGFR1 rearrangements; bone marrow
fied (CEL-NOS), the absolute peripheral blood eosinophil blasts <20%; and the absence of inv(16)(p13.1q22). Consid-
count has to exceed 1.5 × 109/L, and patients must not have eration for idiopathic HES requires exclusion of patients
the Ph chromosome (BCR-ABL1 fusion gene); rearrange- with infectious, allergic, autoimmune, or collagen vascular
ments of PDGFRA/PDGFRB or FGFR1, MDS, MPN; or disorders or pulmonary or neoplastic conditions (including
MPN/MDS overlap syndromes. CEL-NOS requires demon- clonal lymphoid disorders), which are known to be asso
stration of eosinophil clonality and must be differentiated ciated with secondary eosinophilia. Idiopathic HES is
from idiopathic hypereosinophilic syndrome (HES). Idio- therefore classified in patients who have the following char
pathic HES has been historically defined by a persistent (>6 acteristics: (i) persistent eosinophilia (>1.5 × 109/L) lasting
months) peripheral blood eosinophilia (1.5 × 109/L) without for at least 6 months (though this is in evolution as treatment
a clear underlying secondary cause and without evidence of needs can be urgent, and waiting 6 months inappropriate;
eosinophil clonality but associated with end-organ damage (ii) no reactive causes of eosinophilia; (iii) no associated
or dysfunction. Although CEL-NOS is a rare MPN, the true clonal myeloid neoplasm like AML, MDS, MDS/MPN,
incidence of these neoplasms is unknown. Complicating MPN, and SM; (iv) no cytokine-producing immunopheno-
matters further, the revised 2008 WHO classification sepa- typically aberrant T-cell population; (v) no increased myelo-
rated previously classified patients with CEL into myeloid blasts in the peripheral blood or bone marrow; and (vi) no
(and lymphoid) neoplasms associated with eosinophilia and evidence of eosinophil clonality and with end-organ dam-
abnormalities of PDGFRA, PDGFRB, or FGFR1 or CEL- age. If the previous six criteria were fulfilled except that there
NOS depending on the presence or absence of rearrange- is no end-organ damage, then its best classified as idiopathic
ments of PDGFRA/PDGFRB or FGFR1, respectively. hypereosinophilia. Lymphocyte variant hypersinophilia also
Nonetheless, myeloproliferative eosinophilic syndromes needs to be excluded; this entity currently lacks standard
seem to occur much more often in men than in women. The diagnostic criteria, but patients have skin manifestations,
peak incidence is in the fourth decade, but CEL-NOS can increased IgE, and clonal T-cell gene rearrangement with
occur at any age, including childhood. excessive IL-5 production.
Diagnostic criteria Course and prognosis

A minority of CEL-NOS patients are identified incidentally, CEL-NOS typically carries a poor prognosis. Blast transfor-
and more commonly, patients present with fever, fatigue, mation can occur, and poor prognostic features include

marked splenomegaly, cytogenetic abnormalities, and dys- maintained at the tolerated dose 3 times per week for a total
plastic myeloid features in the bone marrow. Idiopathic HES of 12 doses. This resulted in a 91% complete hematologic
can have variable course and overall survival, and tends to be response after a median of 2 weeks. The median duration of
a chronic disorder. In one series, including patients with response was 3 months. A second retrospective study of 12
idiopathic HES and eosinophilic leukemia, 80% of patients patients with HES or CEL treated with alemtuzumab reported
were alive at 5 years after diagnosis, and 42% were alive at 15 a complete hematological response in 10 (83%) for a median
years. of 66 weeks. Eleven relapses were reported, and 5 achieved a
second CHR with retreatment. Infusion reactions and viral
infections including CMV, zoster, and EBV were reported.
Management
Despite these results, the data on alemtuzumab remain lim-
Treatment is indicated for patients with evidence of end- ited and the drug is best considered an investigational ther-
organ damage. Therapy for CEL-NOS and idiopathic HES is apy for this condition at this time.
aimed primarily at decreasing the eosinophil count, improv-
ing symptoms, and preventing end-organ damage or throm-
boembolic complications. Inadequate data exist to support Myeloid and lymphoid neoplasms associated
initiation of therapy based on a specific eosinophil count in with eosinophilia and abnormalities of
the absence of organ disease. Corticosteroids (prednisone 1 PDGFRA, PDGFRB, or FGFR1
mg/kg/d) have typically been the treatment of choice in HES
The WHO recognizes three rare conditions classified as
to reduce eosinophil numbers and minimize the cytotoxic
myeloid–lymphoid neoplasms associated with marked and
effects of the eosinophilic granules. Steroid-resistant
persistent eosinophilia and chromosomal rearrangements,
patients traditionally have been treated with hydroxyurea.
leading to constitutive activation of the PDGFRA/PDGFRB
IFNα can elicit sustained hematologic and cytogenetic
or FGFR1 genes. These are separate entities from CEL and
remissions in idiopathic HES and CEL-NOS patients refrac-
from HES, which are subcategories of MPNs. Although the
tory to other therapies, including prednisone and hydroxy-
partner gene involved heavily influences the clinical features,
urea. Lack of steroid responsiveness, or failure of HU or IFN
separate consideration needs to be given to PDGFRA- and
may warrant consideration of cytotoxic chemotherapeutic
PDGFRB-rearranged eosinophilic disorders because they
agents, such as vincristine and etoposide; these agents also
carry major therapeutic relevance due to the exquisite sensi-
have been used in patients with HES resistant to other thera-
tivity to imatinib mesylate therapy.
pies and in patients with aggressive CEL. Imatinib is also a
consideration, but expectations regarding response are
lower compared to those patients with PDGFRA or PDG- Myeloid and lymphoid neoplasms associated with
FRB rearrangements. PDGFRA rearrangement
Anti–IL-5 antibody approaches (eg, mepolizumab) have
been undertaken in HES based on the cytokine’s role as a
Clinical case
differentiation, activation, and survival factor for eosino-
phils. Long-term results have been presented in 78 patients A 52-year-old mechanic suffered a stroke of unclear etiology,
who had a median exposure of 251 weeks. Ninety-seven per- followed by recurrent headache, rhinorrhea, wheezing, weight
cent of patients experienced adverse effects, but approxi- loss of 15 lb (~7 kg), diarrhea, night sweats, pruritus, and lower-
extremity edema. He underwent a routine blood test, including
mately one-third were considered drug-related; cough,
a CBC, which showed the following: WBCs = 15 × 109/L, Hb
fatigue, headache, upper respiratory infections, and sinusitis
= 10.3 g/dL, MCV = 89 fL, platelets = 224 × 109/L. The most
were most commonly observed. Suppression of eosinophilia
notable feature on the WBC differential was an eosinophilia with
was noted, and in the first 4 months, the median prednisone an AEC of 2.7 × 109/L. There were no circulating blasts in the
dose decreased from 20 to 0 mg (1.8 mg was the median dose peripheral blood. Workup for an underlying connective tissue
over the course of the study. Despite these findings, mepoli- disease, other neoplastic process, and parasitic infection were
zumab has not yet achieved approval by the FDA for CEL- negative. A CT scan of the sinus revealed thickening of the right
NOS at the time of this writing, but is available on a sphenoid sinus. Total IgE was elevated (IgE = 283 kU/L). CT of
compassionate use protocol. the chest showed patchy opacities consistent with bronchiolitis
Use of alemtuzumab (anti-CD52 monoclonal antibody) in or vasculitis. CT scan of the abdomen and pelvis confirmed sple-
refractory HES based on the expression of the CD52 antigen nomegaly. Transthoracic echocardiography showed a dimin-
on eosinophils has also been reported. A study that included ished ejection fraction of 30% and the presence of restrictive
cardiomyopathy. FISH for the CHIC2 deletion, a surrogate for the
11 patients with HES and CEL used alemtuzumab in escalat-
FIP1L1-PDGFRA fusion, was positive.
ing doses of 5, 10, 30 mg intravenously from days 1-3, then

Other MPNs | 479
Although the true incidence of PDGFRA-related neoplasms and precursor eosinophils. It is important to note that
is not known, it is clear these are rare hematologic disorders. FIPL1-PDGRFA rearrangements are not exclusively associ-
These neoplasms are considerably more common in men ated with an MPN phenotype, since these rearrangements
than in women (male-to-female ratio, 9:1 to 17:1) and usu- have also associated with presentations of acute myeloid and
ally are diagnosed between the ages of 25 and 55 years acute lymphoid leukemia. Fibrosis can be present. Immuno-
(median age of onset is late 40s). Approximately 5%-10% of phenotyping is typical for activated eosinophils with expres-
patients in industrial countries who present with idiopathic sion of CD23, CD25, and CD69. Mast cells usually are double
hypereosinophilia can be found to have the FIP1L1-PDGFRA negative for CD2 and CD25. The gold standard for the diag-
fusion. nosis of these neoplasms is demonstration of the fusion gene.
As mentioned, most cases of CEL present with normal
karyotype; thus, FISH and RT-PCR are preferred methods of
Pathobiology
testing. FISH testing relies on the probe for the CHIC2 gene,
PDGFRA is a member of the family of class III receptor TKs, which is deleted uniformly in patients with the FIP1L1-
which also includes PDGFRB, KIT, and FLT3. The PDGFRA PDGFRA fusion gene. RT-PCR can be used both for diagno-
gene is located on the long arm of chromosome 4 (4q12) and sis and monitoring of disease response and for minimal
has been implicated in the chronic eosinophilic syndromes residual disease monitoring.
as a result of a cryptic interstitial deletion at 4q12, leading to
the juxtaposition and in-frame fusion of FIP1L1 and PDGFRA.
This deletion evades standard cytogenetic banding tech-
niques, and explains why most cases of CEL apparently have In the preimatinib era, the prognosis of patients with HES
a normal karyotype. Expression of FIP1L1-PDGFR trans- was poor; the median survival time was 9 months, and the
formed a murine hematopoietic cell line and was constitu- 3-year survival was only 12%. Patients generally had
tively active in these cells and led to increased STAT5 advanced disease, with congestive heart failure accounting
phosphorylation. Similar transforming properties were for 65% of the identified causes of death. However, imatinib
noted when STRN-PDGFRA or ETV6-PDGFRA fusion genes has had a positive impact on the natural history. More
were transfected into murine hematopoietic cell lines. Sev- recently, an observed 5-year survival rate of 80%, decreasing
eral other partner genes have been implicated in the patho- to 42% at 15 years, was noted. In another series reported
genesis of PDGFRA-related neoplasms, including BCR, from the Mayo Clinic, with long-term follow-up, the median
ETV6, KIF5B, and CDK5RAP2. survival was not reached, and 18/22 (82%) were alive, though
2 leukemic transformations were reported.
Diagnosis
Management
PDGFRA-related neoplasms are multisystem disorders asso-
ciated with bone marrow and peripheral blood eosinophilia. The mainstay of therapy for patients with PDGFRA-related
The most common presenting signs and symptoms are neoplasms is the use of imatinib mesylate. Initially reported
weakness, fatigue, cardiopulmonary symptoms, myalgias, as case studies in 2001, one of the first of the pivotal reports
rash, and fever. Splenomegaly is a common finding with a identifying FIPL1-PDGRA as a therapeutic target of ima-
minority of patients also presenting with hepatomegaly. tinib was reported by Cools et al in 2003. Following this
Organ damage occurs as a result of release of cytokines or report, investigators from the NIH reported on improved
direct organ infiltration by eosinophils and possibly mast hematological parameters, including reversal of bone mar-
cells. The most serious complication of PDGFRA-related row fibrosis, along with molecular remissions in 5 of 6
neoplasms is cardiac in nature, including endomyocardial patients. Subsequently, several single- and multi-institution
fibrosis with ensuing restrictive cardiomyopathy. studies have looked at the efficacy of low to conventional
The 2008 WHO classification defines PDGFRA-related doses of imatinib for the treatment of PDGFRA-related
neoplasms as MPNs with prominent eosinophilia and the neoplasms. These studies report remarkably similar results,
presence of the FIP1L1-PDGFRA fusion gene. Thus, the where patients found to have PDGFRA gene rearrange-
most prominent diagnostic feature of patients with PDG- ments have rapid, deep, and durable responses to low to
FRA-related neoplasms is the presence of peripheral blood conventional doses of imatinib mesylate (100-400 mg/d).
mature eosinophilia. Anemia and thrombocytopenia Along these lines, the ELN reported the results of 11
occasionally are present. An elevated serum tryptase of patients treated for at least 12 months with imatinib. Over-
>12 ng/mL is usually also present. Bone marrow biopsy all, 11 of 11 evaluable patients achieved at least a 3-log
demonstrates marked hypercellularity with increased mature reduction in FIP1L1-PDGFRA fusion transcripts, and 9 of

11 patients achieved a complete molecular remission. Simi- PDGFRB-related neoplasms are a distinct group of
larly, an Italian multicenter study demonstrated high levels myeloid neoplasms with histopathological presentations
of durable (median, 251 months) complete molecular including chronic myelomonocytic leukemia (CMML),
remissions in 27 patients with PDGFRA-related neoplasms. juvenile myelomonocytic leukemia, atypical CML, and
In those with known eosinophilic heart disease, steroids are MDS/MPN-unclassified, associated with rearrangement of
recommended concurrently with imatinib during the first the PDGFRB gene, located on the long arm of chromosome
1-2 weeks of therapy given prior reports of cardiogenic 5 (5q31-33). In 1994, Golub et al. (1994) were the first to
shock. characterize the t(5;12)(q31-q33;p13) translocation involv-
Withdrawal of imatinib therapy is typically followed by an ing ETV6 (12p13) and PDGFRB (5q33). Since then, 25 part-
increase in FIP1L1-PDGFRA transcript levels, and presently, ner genes have been identified to collaborate in the
indefinite therapy is recommended. Interestingly, in a retro- development of PDGFRB-related neoplasms. Similarly to its
spective report of 44 patients by the French Eosinophil Net- counterpart, PDGFRB-related neoplasms are extremely
work, CHR and CMR was obtained in 44/44 and 43/44, uncommon disorders, and the true incidence of it is not
respectively. Among 11 patients in whom imatinib was dis- completely known. In fact, among >56,000 cytogenetically
continued, 5 remained in CHR or CMR (range 9-88 months). defined cases from the Mayo Clinic, only 0.04% exhibited
This strategy requires confirmation in a prospective setting. the t(5;12) translocation. In another prospective study, of
Although less common than in BCR-ABL1–positive CML, 556 with MPN, only 10 cases with PDGFRB rearrangements
kinase domain mutations that confer resistance to imatinib were noted. PDGFRB-related neoplasms are more common
therapy can occur in FIP1L1-PDGRFA rearrangement-posi- in men than in women (male-to-female ratio, 2:1), with a
tive myeloid or lymphoid neoplasms, and includes T674I median age of onset in late 40s.
and D842V. Other TKIs have been used in this setting with
only modest and transient benefit.
Diagnosis
Myeloid neoplasms associated with PDGFRB Extramedullary manifestations, such as skin infiltration and
rearrangement splenomegaly, are common. Patients with PDGFRB-related
neoplasms tend to present with anemia and thrombocyto
penia, along with leukocytosis; features characteristic of
Clinical case CMML including a monocytic leukocytosis with associated
A 66 year-old retired computer software analyst underwent a eosinophilia are often seen. Confirmation of diagnosis for
general check-up, which is part of his yearly routine since retir- PDGFRB-related neoplasms requires demonstration of
ing 4 years prior. He is feeling well and has no complaints. His MPN with prominent eosinophilia and occasional neutro-
workup, which included a lipid profile and complete metabolic philia or monocytosis and the presence of the ETV6-
panel, were unremarkable. He was notified, however, of the PDGFRB fusion gene or an alternative PDGFRB gene
presence of abnormalities in his CBC. His CBC showed WBCs = re
arrangement. The classic t(5;12)(q31-q33;p13) can be
24 × 109/L, Hb = 11.0 g/dL, MCV = 81 fL, platelets = 324 × detected easily by conventional metaphase analysis, so FISH
109/L, ANC = 16 × 109/L, absolute lymphocyte count = 2.5 × or RT-PCR usually is used for the confirmation of diagnosis
109/L, AMC = 3 × 109/L, and AEC = 2.7 × 109/L. There was
and determination of fusion gene. Peripheral blood usually
1% circulating peripheral blood blasts. Review of a prior CBC
presents with leukocytosis (neutrophilia or monocytosis)
dating back to 8 months prior showed leukocytosis (WBCs =
along with anemia and thrombocytopenia. The bone mar-
19 × 109/L) with neutrophilia (ANC = 14 × 109/L), absolute
monocytosis (AMC = 1.5 × 109/L), and absolute eosino-
row is hypercellular, with increased fibrosis and mast cells
philia (AEC = 2.6 × 109/L). He was referred to a hematolo- can be increased in number.
gist who noted the prior findings and reviewed a peripheral
blood smear, which showed dysplastic changes in some of Course and prognosis
the neutrophils. A bone marrow aspiration and biopsy was
recommended, which he underwent 3 days later. The bone Prognosis in the preimatinib era was very poor for patients
marrow showed 2% bone marrow blasts in the aspirate with with PDGFRB-related neoplasms; the median survival time
increased numbers of monocytes. Dysplastic-looking neutro- did not exceed 2 years. As above, with imatinib, the 10 year-
philic precursors and monocytes also were appreciated. The survival appears favorable, and based on this small cohort of
core biopsy showed a hypercellular bone marrow (80%) with 26 patients treated with imatinib and mentioned above, it
mild reticulin fibrosis. Metaphase cytogenetics showed t(5;12)
also appears that TKIs positively impact the outcome of
(q31~33;p12).
patients with PDGFRB-related neoplasms.

Other MPNs | 481
Management with lymphomatous presentation. Hypercatabolism and

splenomegaly are common features of AML and MPN
Imatinib mesylate is also the mainstay of therapy for patients
patients.
with PDGFRB-related neoplasms. In 2002, Apperley et al
Diagnostic criteria established by the 2008 WHO classi-
reported four patients treated with imatinib 400mg daily, and
fication include the presence of an MPN with prominent
normalization of blood accounts occurred within 4 weeks, the
eosinophilia and occasional neutrophilia or monocytosis
t(5;12) translocation was undetectable by 12-36 weeks, and
or the presence of AML or precursor B- or T-cell lympho-
transcript levels decreased in those with the ETV6-PDGFRB
blastic leukemia and the presence of FGFR1 rearrange-
fusion. A report on 12 patients with PDGFRB-related neo-
ment. The most common chromosomal translocation
plasms who received imatinib therapy for a median of 47
associated with FGFR1-related neoplasms is t(8;13)
months revealed normalization of peripheral blood cell
(p11;q12), which results in expression of the ZNF198-
counts and disappearance of eosinophilia in 11; 10 had com-
FGFR1 fusion TK. Fifteen fusion gene partners have been
plete resolution of cytogenetic abnormalities and decrease or
described in FGFR1 rearrangement neoplasms, including
disappearance of fusion transcripts as measured by RT-PCR.
CEP110, FGFR1OP1, FGFR1OP2, TRIM 24, MYO18A,
A retrospective report of an expanded cohort of 26 patients
HERVK, and BCR.
with a median follow-up of 10.2 years (imatinib duration 6.6
years) reported a 90% 10-year survival, a 96% response rate,
and that no patient with complete cytogenetic (N = 13) or
molecular (N = 8) response lost their response. Course and prognosis
The prognosis for patients with FGFR1-related neoplasms is

very poor, with evolution to AML typically occurring within
Myeloid and lymphoid neoplasms associated
1-2 years. The clinical aggressiveness and diminished aware-
with FGFR1 abnormalities
ness about the features of this entity and the lack of appro
This uncommon and heterogeneous group of neoplasms ved TKIs make the management of these patients very
arise from pluripotent hematopoietic stem cells and are challenging.
associated with rearrangements in the FGFR1 gene and
eosinophilia. Formerly known as 8p11 myeloproliferative
syndrome or 8p11 stem cell syndrome, FGFR1-related neo- Management
plasms can present as classic MPNs, precursor B- or T-cell
Early intensive chemotherapy followed by allogeneic SCT
lymphoblastic leukemia, or AML. FGFR1-related neoplasms
remains the only potential curative therapy for patients with
have been reported across a wide age range (3-84 years), and
FGFR1-related neoplasms. Interestingly, midostaurin has
the median age of diagnosis is 32 years. Females constitute
demonstrated in vitro activity against one subtype of the
approximately 40% of the cases.
FGFR1 fusion gene and in one patient, resulted in improved
leukocytosis, splenomegaly, and lymphadenopathy and 6
Pathobiology months of clinical stability prior to transplantation. Addi-
tional TKIs with anti-FGFR1 activity are being evaluated,
The molecular consequences of FGFR1 rearrangements are
including ponatinib.
remarkably well described for such an unusual disorder. In
all FGFR1-related neoplasms, the N-terminal partner con-
taining self-association motif is fused to the C-terminal TKD Key points (see also Figures 16-10, 16-11,
of FGFR1. These fusion genes (ZNF198-FGFR1), when and 16-12)
expressed in primary murine hematopoietic cells, cause an • CNL is characterized by a sustained mature, neutrophilic
MPN that recapitulates the human MPN phenotype. Fur- leukocytosis, often with splenomegaly, and CSF3R mutations
thermore, these constitutively active FGFR1 fusion genes identified in substantial proportions of patients. CNL is a
activate downstream effector molecules, such as PLC-g, progressive MPN, with hydroxyurea and allogeneic SCT as
STAT5, and PI3K/AKT. treatment options, though TKI may have a future role
depending on the type of CSF3R mutation that is
present.
Diagnosis • Mastocytosis includes cutaneous and systemic mastocytosis;
the latter is more commonly identified in adults and includes
Clinical manifestations include fever, weight loss, and
indolent and aggressive subtypes. Mastocytosis can be
night sweats. Lymphadenopathy is common in patients

Key points (see also Figures 16-10, 16-11, Fatigue, constitutional symptoms
and 16-12) (continued) Splenomegaly
Mature sustained, nonreactive
Clinical and leukocytosis (leukocytes >25 × 109/L,
associated with an underlying hematological malignancy or laboratory with >80% segmented/band neutrophils,
can manifest as mast cell leukemia. Consequences stem from features <10% immature granulocytes, and
<1% peripheral blasts)
mediator-release or organ infiltration, and most patients have
Bone marrow hypercellularity with
KIT D816V mutations. Treatments are supportive, including neutrophil predominance
H1/H2 blockade, and mast cell stabilizers. For patients with
bone involvement, IFN is an option. Patients with KIT 816V Molecular
CSF3R mutations
mutations are resistant to imatinib, but midostaurin is markers
under investigation and has preliminarily shown promising

Hydroxyurea
results. Treatment
Transplantation
options
• Myeloproliferative neoplasms with eosinophilia include CEL, TKIs?
and those with PDGFRA/PDGFRB, and FGFR1 rearrangements.
Cardiac involvement can be a source of morbidity and Figure 16-10 CNL is a rare MPN, and its presenting features are
morality, especially in those with CEL, and PDGFRA-rearranged similar to those of other MPNs, including fatigue, constitutional
disease. Characteristic findings include sustained eosinophilia, symptoms, and symptoms from splenomegaly. CNL is characterized
often with monocytosis in those with PDGFRB-rearranged by a sustained, nonreactive, mostly mature leukocytosis with marrow
disease. Those with FGFR1 rearrangements may also have hypercellularity. It is expected that most CNL patients will harbor
splenomegaly and lymphadenopathy. Patients with PDGFRA/ CSF3R mutations, either membrane-proximal mutations or
PDGFRB rearranged disease are uniquely senstive to imatinib, frameshift or nonsense mutations that will ultimately be incorporated
which has had a very positive impact on prognosis for these into revised diagnostic criteria, aiding diagnostic capability. Optimal
rare neoplasms. treatment for patients with CNL remains to be defined. The role of
TKIs in the treatment of CNL is intriguing, but not yet defined.
Skin lesions: urticaria pigmentosa and Darier sign (urticaria after stroking skin)
Constitutional and mediator release symptoms
Clinical and
Anaphylaxis
laboratory
Osteopenia, fracture
features
Hepatosplenomegaly
*Increased tryptase (>20 ng/mL is a minor criterion)
Bone marrow/ **Major criterion: multifocal dense infiltrates (≥15 MCs in aggregates)
EC organ MC *Minor criterion: >25% with atypia, immaturity, or spindle-shaped appearance
finding *Minor criterion: MC expression of CD117 with CD2 and/or CD25
Molecular
*KIT D816V (minor criterion)
markers
Subtypes Indolent SM Aggressive SM SM with AHNMD MC leukemia

• No AHNMD • No MC leukemia SM, in addition to Diffuse infiltration with
• No C findings • ≥1 “C” findings (malabsorption, WHO criteria for AML, atypical, immature MCs
skeletal lesions, HSM with MDS, MPN, lymphoma and increased circulating
impaired liver function or immature MCs
hypersplenism, cytopenias)
Trigger avoidance
H1/H2 blockers, proton pump inhibitors, cromolyn
Treatment IFNα
options Cladribine, chemotherapy
Allogeneic SCT
Midostaurin? Masitinib?
Figure 16-11 Systemic mastocytosis is diagnosed in the presence of 1 major criterion (**) and 1 minor criterion (*) or in the presence of 3
minor criteria; also see Table 16-20. EC = extracutaneous; MC = mast cell; HSM = hepatosplenomegaly.

Bibliography | 483
Cardiac, pulmonary, neurological, and dermatological involvement

Splenomegaly
Clinical and
Lymphadenopathy (FGFR1)
laboratory
Sustained eosinophilia (>1.5 × 109/L)
features
Anemia, thrombocytopenia
Bone marrow fibrosis
PDGFRA/PDGFRB- and
Subtypes HES CEL, NOS
FGFR1-rearranged neoplasms
Additional Persistent, primary PDGFRA PDGFRB FGFR1 Clonal abnormality;

diagnostic eosinophilia with May have features blast cells >2% in the blood or
features end-organ damage; in common with <20% in the bone marrow;
no increased CMML, JMML, lack of BCR-ABL1, PDGFRA/PDGFRB,
blood/bone MDS/MPN-U, or FGFR1 rearrangements;
marrow blasts; atypical CML absence of inv(16)(p13.1q22)
no clonal disease
or aberrant T-cell
population
First-line FGFR1:
treatment Steroids PDGFRA/PDGFRB: imatinib Chemotherapy, HU, IFNα
options allogeneic SCT, clinical trial
Figure 16-12 The hypereosinophilias discussed in this chapter are typically characterized by sustained eosinophilia and end-organ consequences, the
most severe of which can be cardiac in nature. HES is distinguished by absence of a clonal marker, and steroids are considered a frontline therapeutic
option. It is critical to recognize PDGFRA/PDGFRB rearranged neoplasms given their remarkable sensitivity to low dose imatinib. On the other
hand, FGFR1 rearranged neoplasms carry an aggressive clinical course and poor prognosis, with an unmet treatment need. CEL-NOS is characterized
by clonal eosinophilia and often peripheral or marrow blasts, but not by BCR-ABL1 or other molecular genetic abnormalities such as PDGFRA/
PDGFRB/FGFR1. Typical first-line treatments for CEL-NOS can include either hydroxyurea or IFNα. *See text for diagnostic criteria (Table 16-5).
Bibliography Baccarani M, Deininger MW, Rosti G, et al. European

LeukemiaNet recommendations for the management of chronic
Introduction myeloid leukemia: 2013. Blood. 2013;122:872-884.
Baccarani M, Rosti G, Castagnetti F, et al. Comparison of imatinib
Delhommeau F, Dupont S, Della Valle V, et al. Mutation in TET2
400 mg and 800 mg daily in the front-line treatment of
in myeloid cancers. N Engl J Med. 2009;360:2289-2301.
high-risk, Philadelphia-positive chronic myeloid leukemia: a
Johansson P. Epidemiology of the myeloproliferative disorders
European LeukemiaNet Study. Blood. 2009;113:4497-4504.
polycythemia vera and essential thrombocythemia. Semin
Bonifazi F, de Vivo A, Rosti G, et al. Chronic myeloid leukemia and
Thromb Hemost. 2006;32:171-173.
interferon-alpha: a study of complete cytogenetic responders.
Mehta J, Wang H, Iqbal SU, Mesa R. Epidemiology of
Blood. 2001;98:3074-3081.
myeloproliferative neoplasms in the United States. Leuk
Branford S, Rudzki Z, Parkinson I, et al. Real-time quantitative
Lymphoma. 2014;55:595-600.
PCR analysis can be used as a primary screen to identify patients
Swerdlow SH, Campo E, Harris NL, et al., eds. World Health
with CML treated with imatinib who have BCR-ABL kinase
Organization Classification of Tumors of Hematopoietic and
domain mutations. Blood. 2004;104:2926-2932.
Brummendorf TH, Cortes JE, de Souza CA, et al. Bosutinib versus
Tefferi A. The history of myeloproliferative disorders: before and
imatinib in newly diagnosed chronic-phase chronic myeloid
after Dameshek. Leukemia. 2008;22:3-13.
leukaemia: results from the 24-month follow-up of the BELA
Zhang SJ, Rampal R, Manshouri T, et al. Genetic analysis of
trial. Br J Haematol. 2015;168:69-81.
patients with leukemic transformation of myeloproliferative
Cortes JE, Baccarani M, Guilhot F, et al. Phase III, randomized,
neoplasms shows recurrent SRSF2 mutations that are associated
open-label study of daily imatinib mesylate 400 mg versus 800 mg
with adverse outcome. Blood. 2012;119:4480-4485.
in patients with newly diagnosed, previously untreated chronic
myeloid leukemia in chronic phase using molecular end points:
Chronic myelogenous leukemia, BCR-ABL1 tyrosine kinase inhibitor optimization and selectivity study. J
Clin Oncol. 2010;28:424-430.
positive
Cortes JE, Kantarjian HM, Brümmendorf TH, et al. Safety and
Apperley JF, Cortes JE, Kim DW, et al. Dasatinib in the treatment efficacy of bosutinib (SKI-606) in chronic phase Philadelphia
of chronic myeloid leukemia in accelerated phase after imatinib chromosome-positive chronic myeloid leukemia patients with
failure: the START A trial. J Clin Oncol. 2009;27:3472-3479. resistance or intolerance to imatinib. Blood. 2011;118:4567-4576.

Cortes JE, Kantarjian HM, Rea D, et al. Final analysis of the to imatinib at a dose of 400 to 600 milligrams daily: two-year
efficacy and safety of omacetaxine mepesuccinate in patients follow-up of a randomized phase II study (START-R). Cancer.
with chronic- or accelerated-phase chronic myeloid leukemia: 2009;115:4136-4147.
results with 24 months of follow-up. Cancer, 2015 May Kantarjian H, Sawyers C, Hochhaus A, et al. Hematologic
15;121(10):1637-1644. and cytogenetic responses to imatinib mesylate in chronic
Cortes JE, Kantarjian H, Shah NP, et al. Ponatinib in refractory myelogenous leukemia. N Engl J Med. 2002;346:645-652.
Philadelphia chromosome-positive leukemias. N Engl J Med. Kantarjian HM, Shah NP, Cortes JE, et al. Dasatinib or
Cortes JE, Kim DW, Kantarjian HM, et al. Bosutinib versus leukemia: 2-year follow-up from a randomized phase III trial
imatinib in newly diagnosed chronic-phase chronic (DASISION). Blood. 2012;119:1123-1129.
myeloid leukemia: results from the BELA trial. J Clin Oncol. Kantarjian H, Shah NP, Hochhaus A, et al. Dasatinib versus
Deininger M, O’Brien SG, Guilhot F, et al. International leukemia. N Engl J Med. 2010;362:2260-2270.
Randomized Study of Interferon vs STI571 (IRIS) 8-year follow- Kantarjian H, Talpaz M, O’Brien S, et al. High-dose imatinib
up: sustained survival and low risk for progression or events mesylate therapy in newly diagnosed Philadelphia
in patients with newly diagnosed chronic myeloid leukemia chromosome-positive chronic phase chronic myeloid leukemia.
in chronic phase (CML-CP) treated with imatinib [abstract]. Blood. 2004;103:2873-2878.
Blood. 2009;114(22):1126. Larson RA, Hochhaus A, Hughes TP, et al. Nilotinib vs imatinib
Druker BJ, Sawyers CL, Kantarjian H, et al. Activity of a specific in patients with newly diagnosed Philadelphia chromosome-
inhibitor of the BCR-ABL TK in the blast crisis of chronic positive chronic myeloid leukemia in chronic phase: ENESTnd
myeloid leukemia and acute lymphoblastic leukemia with 3-year follow-up. Leukemia. 2012;26:2197-2203.
the Philadelphia chromosome. N Engl J Med. 2001;344: Mahon FX, Réa D, Guilhot J, et al. Discontinuation of imatinib
1038-1042. in patients with chronic myeloid leukaemia who have
Druker BJ, Talpaz M, Resta DJ, et al. Efficacy and safety of a maintained complete molecular remission for at least 2 years:
specific inhibitor of the BCR-ABL tyrosine kinase in chronic the prospective, multicentre Stop Imatinib (STIM) trial. Lancet
myeloid leukemia. N Engl J Med. 2001;344:1031-1037. Oncol. 2010;11:1029-1035.
Guilhot F, Chastang C, Michallet M, et al. French Chronic Marin D, Ibrahim AR, Lucas C, et al. Assessment of BCR-
Myeloid Leukemia Study Group. Interferon-2b combined with ABL1 transcript levels at 3 months is the only requirement
cytarabine versus interferon alone in chronic myelogenous for predicting outcome for patients with chronic myeloid
leukemia. N Engl J Med. 1997;337:223-229. leukemia treated with tyrosine kinase inhibitors. J Clin Oncol.
Hansen JA, Gooley TA, Martin PJ, et al. Bone marrow transplants 2012;30:232-238.
from unrelated donors for patients with chronic myeloid Mathisen MS, Kantarjian HM, Cortes J, Jabbour EJ. Practical
leukemia. N Engl J Med. 1998;338:962-968. issues surrounding the explosion of tyrosine kinase inhibitors
Hehlmann R, Hanfstein B, Muller MC, et al. The prognostic for the management of chronic myeloid leukemia. Blood Rev.
significance of early molecular and cytogenetic response for 2014;28:179-87.
long-term progression-free and overall survival in imatinib- McGlave PB, Shu XO, Wen W, et al. Unrelated donor marrow
treated chronic myeloid leukemia (CML). J Clin Oncol. transplantation for chronic myelogenous leukemia: 9 years’
2012;30:6510. experience of the national marrow donor program. Blood.
Hughes TP, Kaeda J, Branford S, et al. International Randomized 2000;95:2219-2225.
Study of Interferon Versus STI571 (IRIS) Study Group. Michor F, Hughes TP, Iwasa Y, et al. Dynamics of chronic myeloid
Frequency of major molecular responses to imatinib or leukemia. Nature. 2005;435:1267-1270.
interferon alfa plus cytarabine in newly diagnosed chronic NCCN Guidelines, version 1.2015, Chronic Myelogenous
myeloid leukemia. N Engl J Med. 2003;349:1423-1432. Leukemia 8/28/14, accessed 2/7/15 (www.nccn.org).
Hughes TP, Lipton JH, Spector N, et al. Deep molecular responses O’Brien SG, Guilhot F, Larson RA, et al. Imatinib compared
achieved in patients with CML-CP who are switched to with interferon and low-dose cytarabine for newly diagnosed
nilotinib after long-term imatinib. Blood. 2014;124:729-736. chronic-phase chronic myeloid leukemia. IRIS investigators.
Jabbour E, Kantarjian HM, Saglio G, Steegmann JL, Shah NP, N Engl J Med. 2003;348:994-1004.
Boque C, et al. Early response with dasatinib or imatinib in O’Dwyer ME, Mauro MJ, Blasdel C, et al. Clonal evolution and
chronic myeloid leukemia: 3-year follow-up from a randomized lack of cytogenetic response are adverse prognostic factors for
phase 3 trial (DASISION). Blood. 2014 Jan 23;123(4):494-500. hematologic relapse of chronic phase CML patients treated with
Kantarjian HM, Cortes J, O’Brien S, et al. Imatinib mesylate imatinib mesylate. Blood. 2004;103:451-455.
(STI571) therapy for Philadelphia chromosome-positive Olavarria E, Ottmann OG, Deininger M, et al. Chronic Leukemia
chronic myelogenous leukemia in blast phase. Blood. Working Party of the European Group of Bone and Marrow
2002;99:3547-3553. Transplantation (EBMT). Response to imatinib in patients who
Kantarjian H, Pasquini R, Lévy V, et al. Dasatinib or high-dose relapse after allogeneic stem cell transplantation for chronic
imatinib for chronic-phase chronic myeloid leukemia resistant myeloid leukemia. Leukemia. 2003;17:1707-1712.

Bibliography | 485
Or R, Shapira MY, Resnick I, et al. Nonmyeloablative allogeneic Lasho TL, Mims A, Elliott MA, Finke C, Pardanani A, Tefferi A.
stem cell transplantation for the treatment of chronic myeloid Chronic neutrophilic leukemia with concurrent CSF3R and
leukemia in first chronic phase. Blood. 2003;101:441-445. SETBP1 mutations: single colony clonality studies, in vitro
Preudhomme C, Guilhot J, Nicolini FE, et al. Imatinib plus sensitivity to JAK inhibitors and lack of treatment response to
peginterferon alfa-2a in chronic myeloid leukemia. N Engl J ruxolitinib. Leukemia. 2014;28:1363-1365.
Med. 2010;363:2511-2521. Maxson JE, Gotlib J, Pollyea DA, et al. Oncogenic CSF3R
Radich JP, Gooley T, Bryant E, et al. The significance of BCR- mutations in chronic neutrophilic leukemia and atypical CML.
ABL molecular detection in chronic myeloid leukemia patients N Engl J Med. 2013;368:1781-1790.
“late,” 18 months or more after transplantation. Blood. Pardanani A, Lasho TL, Laborde RR, et al. CSF3R T618I is a
2001;98:1701-1707. highly prevalent and specific mutation in chronic neutrophilic
Ross DM, Branford S, Seymour JF, et al. Safety and efficacy of leukemia. Leukemia. 2013;27:1870-1873.
imatinib cessation for CML patients with stable undetectable
minimal residual disease: results from the TWISTER study.
Blood. 2013;122:515-522. Systemic mastocytosis
Rousselot P, Charbonnier A, Cony-Makhoul P, et al. Loss of
Gotlib J, Akin C. Mast cells and eosinophils in mastocytosis,
major molecular response as a trigger for restarting tyrosine
chronic eosinophilic leukemia, and non-clonal disorders. Semin
kinase inhibitor therapy in patients with chronic-phase
Hematol. 2012;49:128-137.
chronic myelogenous leukemia who have stopped imatinib
Gotlib J, Kluin-Nelemans HC, George TI, et al. Midostaurin
after durable undetectable disease. J Clin Oncol. 2014;32:
(PKC412) demonstrates a high rate of durable responses in
424-430.
patients with advanced systemic mastocytosis: results from the
Saglio G, Jim DW, Issaragrisil S, et al. Nilotinib versus imatinib
fully accrued global phase 2 CPKC412D2201 trial [abstract].
for newly diagnosed chronic myeloid leukemia. N Engl J Med.
Blood. 2014;124(21):636.
2010;362:2251-2259.
Pardanani A. Systemic mastocytosis in adults: 2015 update on
Shah NP, Guilhot F, Cortes JE, et al. Long-term outcome with
diagnosis, risk stratification, and management. Am J Hematol.
dasatinib after imatinib failure in chronic-phase chronic
2015;90:250-262.
myeloid leukemia: follow-up of a phase 3 study. Blood.
Schwaab J, Schnittger S, Sotlar K, Walz C, Fabarius A, Pfirrmann
2014;123:2317-2324.
M, et al. Comprehensive mutational profiling in advanced
Soverini S, Martinelli G, Rosti G, et al. ABL mutations in late
systemic mastocytosis. Blood. 2013;122:2460-2466.
chronic phase chronic myeloid leukemia patients with up-front
Soucie E, Hanssens K, Mercher T, et al. In aggressive forms of
cytogenetic resistance to imatinib are associated with a greater
mastocytosis, TET2 loss cooperates with c-KITD816V to
likelihood of progression to blast crisis and shorter survival:
transform mast cells. Blood. 2012;120:4846-4849.
a study by the GIMEMA Working Party on Chronic Myeloid
Swerdlow SH, Campo E, Harris NL, et al., eds. World Health
Leukemia. J Clin Oncol. 2005;23:4100-4109.
Organization Classification of Tumors of Hematopoietic and
Tiu R, Kalaycio M. Targeted therapy for patients with chronic
myeloid leukemia: clinical trial experience and challenges in
Tefferi A, Levine RL, Lim KH, et al. Frequent TET2 mutations
inter-trial comparisons. Leuk Lymphoma. 2012;53:1263-1272.
in systemic mastocytosis: clinical, KITD816V and FIP1L1-
Valent P, Hadzijusufovic E, Schernthaner GH, Wolf D, Rea D, le
PDGFRA correlates. Leukemia. 2009;23:900-904.
Coutre P. Vascular safety issues in CML patients treated with
Traina F, Visconte V, Jankowska A, et al. Single nucleotide
BCR/ABL1 kinase inhibitors. Blood. 2015;125:901-906.
polymorphism array lesions, TET2, DNMT3A, ASXL1 and
Wadhwa J, Szydlo RM, Apperley JF, et al. Factors affecting
CBL mutations are present in systemic mastocytosis. PLoS One.
duration of survival after onset of blastic transformation of
2012;7:e43090.
chronic myeloid leukemia. Blood. 2002;99:2304-2309.
Ustun C, Reiter A, Scott BL, et al. Hematopoietic stem-cell
Yeung DT, Osborn MP, White DL, et al. TIDEL-II: first-line
transplantation for advanced systemic mastocytosis. J Clin
use of imatinib in CML with early switch to nilotinib for
Oncol. 2014;32:3264-3274.
failure to achieve time-dependent molecular targets. Blood.
Valent P, Sperr WR, Akin C. How I treat patients with advanced
2015;125:915-923.
systemic mastocytosis. Blood. 2010;116:5812-5817.
Chronic neutrophilic leukemia

Chronic eosinophilic leukemia, not otherwise
Elliott MA, Hanson CA, Dewald GW, Smoley SA, Lasho TL,
specified
Tefferi A. WHO-defined chronic neutrophilic leukemia: a long-
term analysis of 12 cases and a critical review of the literature. Gotlib J. World Health Organization-defined eosinophilic
Leukemia. 2005;19:313-317. disorders: 2014 update on diagnosis, risk stratification, and
Elliott MA, Tefferi A. Chronic neutrophilic leukemia 2014: Update management. Am J Hematol. 2014;89:325-337.
on diagnosis, molecular genetics, and management. Am J Noel P. Eosinophilic myeloid disorders. Semin Hematol.
Hematol. 2014;89:651-658. 2012;49:120-127.

Pitini V, Teti D, Arrigo C, et al. Alemtuzumab therapy for Carbuccia N, Murati A, Trouplin V, et al. Mutations of ASXL1
refractory idiopathic hypereosinophilic syndrome. Br J gene in myeloproliferative neoplasms. Mutations of ASXL1 gene
Haematol. 2004;127:477. in myeloproliferative neoplasms. Leukemia. 2009;23:2183-2186.
Swerdlow SH, Campo E, Harris NL, et al., eds. World Health Caramazza D, Maffioli M, Cazzola M, et al. Post-polycythemia and
Organization Classification of Tumors of Hematopoietic and post-thrombocythemia myelofibrosis have distinctive clinical
Lymphoid Tissues. Lyon, France: IARC Press; 2008. phenotypes: an international multicenter study on 718 patients
Rothenberg ME, Klion AD, Roufosse FE, et al. Mepolizumab HES [abstract]. Blood. 2014;124(21):1824.
Study Group. Treatment of patients with the hypereosinophilic Cervantes F, Dupriez B, Pereira A, et al. New prognostic scoring
syndrome with mepolizumab. N Engl J Med. 2008;358:1215-1228. system for primary myelofibrosis based on a study of the
Roufosse FE, Kahn JE, Gleich GJ, et al. Long-term safety International Working Group for Myelofibrosis Research and
of mepolizumab for the treatment of hypereosinophilic Treatment. Blood. 2009;113:2895-2901.
syndromes. J Allergy Clin Immunol. 2013;131:461-467. Cortelazzo S, Finazzi G, Ruggeri M, et al. Hydroxyurea for patients
Verstovsek S, Tefferi A, Kantarjian H, et al. Alemtuzumab therapy with essential thrombocythemia and a high risk of thrombosis.
for hypereosinophilic syndrome and chronic eosinophilic N Engl J Med. 1995;332:1132-1136.
leukemia. Clin Cancer Res. 2009;15:368-373. Dameshek W. Some speculations on the myeloproliferative
syndromes. Blood. 1951;6:372-375.
Polycythemia vera, essential thrombocytosis, Delhommeau F, Dupont S, Della Valle V, et al. Mutation in TET2
in myeloid cancers. N Engl J Med. 2009;360:2289-2301.
and primary myelofibrosis
Ding J, Komatsu H, Wakita A, et al. Familial essential
Adamson JW, Fialkow PJ, Murphy S, Prchal JF, Steinman L. thrombocythemia associated with a dominant-positive
Polycythemia vera; stem cell and probable clonal origin of the activating mutation of the c-Mpl gene, which encodes for the
disease. N Engl J Med. 1976;295:913-916. receptor for thrombopoietin. Blood. 2004;103:4198-4200.
Ang SO, Chen H, Hirota K, et al. Disruption of oxygen El Kassar N, Hetet G, Li Y, Briere J, Grandchamp B. Clonal
homeostasis underlies congenital Chuvash polycythemia. Nat analysis of haemopoietic cells in essential thrombocythaemia.
Genet. 2002;32:614-621. Br J Haematol. 1995;90:131-137.
Arcasoy MO, Degar BA, Harris KW, Forget BG. Familial Emanuel RM, Dueck AC, Geyer HL, et al. Myeloproliferative
erythrocytosis associated with a short deletion in the neoplasm (MPN) symptom assessment form total symptom
erythropoietin receptor gene. Blood. 1997;89:4628-4635. score: prospective international assessment of an abbreviated
Argetsinger LS, Campbell GS, Yang X, et al. Identification of JAK2 symptom burden scoring system among patients with MPNs.
as a growth hormone receptor-associated tyrosine kinase. Cell. J Clin Oncol. 2012;30:4098-4103.
1993;74:237-244. Emanuel R, Dueck AC, Geyer HL, et al. Myeloproliferative (MPN)
Barbui T, Barosi G, Birgegard G, Cervantes F, Finazzi G, symptom burden response thresholds: assessment of MPN-
Griesshammer M, et al. Philadelphia-negative classical SAF TSS quartiles as potential markers of symptom response
myeloproliferative neoplasms: critical concepts and [abstract]. Blood. 2013:122(21):122.
management recommendations from European LeukemiaNet. Fruchtman SM, Mack K, Kaplan ME, Peterson P, Berk PD,
J Clin Oncol. 2011;29:761-770. Wasserman LR. From efficacy to safety: a polycythemia
Barbui T, Thiele J, Vannucchi AM, Tefferi A. Rethinking vera study group report on hydroxyurea in patients with
the diagnostic criteria of polycythemia vera. Leukemia. polycythemia vera. Semin Hematol. 1997;34:17-23.
2014;28:1191-1195. Gangat N, Caramazza D, Vaidya R, et al. DIPSS plus: a refined
Barosi G, Mesa RA, Thiele J, Cervantes F, Campbell PJ, Verstovsek Dynamic International Prognostic Scoring System for primary
S, et al. Proposed criteria for the diagnosis of post-polycythemia myelofibrosis that incorporates prognostic information from
vera and post-essential thrombocythemia myelofibrosis: a karyotype, platelet count, and transfusion status. J Clin Oncol.
consensus statement from the International Working Group for 2011;29(4):392-397.
Myelofibrosis Research and Treatment. Leukemia. 2008;22: Gilliland DG, Blanchard KL, Levy J, Perrin S, Bunn HF.
437-438. Clonality in myeloproliferative disorders: analysis by means
Barosi G, Mesa R, Finazzi G, Harrison C, Kiladjian JJ, Lengfelder of the polymerase chain reaction. Proc Natl Acad Sci USA.
E, et al. Revised response criteria for polycythemia vera and 1991;88:6848-6852.
essential thrombocythemia: an ELN and IWG-MRT consensus Harrison CN, Campbell PJ, Buck G, et al. Hydroxyurea compared
project. Blood. 2013;121:4778-4781. with anagrelide in high-risk essential thrombocythemia. N Engl
Baxter EJ, Scott LM, Campbell PJ, et al. Acquired mutation of the J Med. 2005;353:33-45.
tyrosine kinase JAK2 in human myeloproliferative disorders. Harrison C, Kiladjian JJ, Al-Ali HK. JAK inhibition with
Lancet. 2005;365:1054-1061. ruxolitinib versus best available therapy for myelofibrosis. N
Berk PD, Goldberg JD, Silverstein MN, et al. Increased incidence Engl J Med. 2012;366:787-798.
of acute leukemia in polycythemia vera associated with Heuck G. Two cases of leukemia with peculiar blood and bone
chlorambucil therapy. N Engl J Med. 1981;304:441-447. marrow findings, respectively. Arch Pathol Anat. 1879;78:475-496.

Bibliography | 487
James C, Ugo V, Le Couedic JP, et al. A unique clonal Nangalia J, Massie CE, Baxter EJ, et al. Somatic CALR mutations
JAK2 mutation leading to constitutive signalling causes in myeloproliferative neoplasms with nonmutated JAK2. N Engl
polycythaemia vera. Nature. 2005;434:1144-1148. J Med. 2013;369:2391-2405.
Jones AV, Chase A, Silver RT, et al. JAK2 haplotype is a major risk Osler W. Chronic cyanosis with polycythaemis and enlarged
factor for the development of myeloproliferative neoplasms. Nat spleen: a new entity. 1903. Am J Med Sci. 2008;335:411-417.
Genet. 2009;41:446-449. Pardanani AD, Levine RL, Lasho T, et al. MPL515 mutations in
Kiladjian JJ, Cassinat B, Chevret S, Turlure P, Cambier N, myeloproliferative and other myeloid disorders: a study of 1182
Roussel M, et al. Pegylated interferon-alfa-2a induces complete patients. Blood. 2006;108:3472-3476.
hematologic and molecular responses with low toxicity in Pardanani A, Laborde RR, Lasho TL, et al. Safety and efficacy
polycythemia vera. Blood. 2008;112:3065-3072. of CYT387, a JAK1 and JAK2 inhibitor, in myelofibrosis.
Kilpivaara O, Mukherjee S, Schram AM, et al. A germline JAK2 Leukemia. 2013;27:1322-1327.
SNP is associated with predisposition to the development of Passamonti F, Thiele J, Girodon F, et al. A prognostic model to
JAK2(V617F)-positive myeloproliferative neoplasms. Nat predict survival in 867 World Health Organization-defined
Genet. 2009;41:455-459. essential thrombocythemia at diagnosis: a study by the
Klampfl T, Gisslinger H, Harutyunyan AS, et al. Somatic International Working Group on Myelofibrosis Research and
mutations of calreticulin in myeloproliferative neoplasms. Treatment. Blood. 2012;120:1197-1201.
N Engl J Med. 2013;369:2379-2390. Percy MJ, Furlow PW, Lucas GS, et al. A gain-of-function
Kralovics R, Passamonti F, Buser AS, et al. A gain-of-function mutation in the HIF2A gene in familial erythrocytosis. N Engl J
mutation of JAK2 in myeloproliferative disorders. N Engl J Med. Med. 2008;358:162-168.
2005;352:1779-1790. Pikman Y, Lee BH, Mercher T, et al. MPLW515L is a novel
Kroger N, Zabelina T, Alchalby H, et al. Dynamic of bone somatic activating mutation in myelofibrosis with myeloid
marrow fibrosis regression predicts survival after allogeneic metaplasia. PLoS Med. 2006;3:e270.
stem cell transplantation for myelofibrosis. Biol Blood Marrow Prchal JF, Axelrad AA. Bone-marrow responses in polycythemia
Transplant. 2014;20:812-815. vera [letter]. N Engl J Med. 1974;290:1382.
Kröger N, Giorgino T, Scott BL, et al. Survival of allogeneic Rondelli D, Barosi G, Bacigalupo A, et al.; Myeloproliferative
stem cell transplantation vs conventional therapies per DIPSS Diseases-Research Consortium. Allogeneic hematopoietic
stratification in patients with primary myelofibrosis younger stem-cell transplantation with reduced-intensity conditioning
than 65 years: a retrospective analysis on 673 patients [abstract]. in intermediate- or high-risk patients with myelofibrosis with
Blood. 2014;124(21):633. myeloid metaplasia. Blood. 2005;105:4115-4119.
Landolfi R, Marchioli R, Kutti J, et al. Efficacy and safety of low-dose Rosti V, Massa M, Vannucchi AM, et al. The expression of
aspirin in polycythemia vera. N Engl J Med. 2004;350:114-124. CXCR4 is down-regulated on the CD341 cells of patients with
Lasho TL, Finke CM, Hanson CA, et al. SF3B1 mutations in myelofibrosis with myeloid metaplasia. Blood Cells Mol Dis.
primary myelofibrosis: clinical, histopathology and genetic 2007;38:280-286.
correlates among 155 patients. Leukemia. 2012;26:1135-1137. Scott LM, Tong W, Levine RL, et al. JAK2 exon 12 mutations in
Levine RL, Wadleigh M, Cools J, et al. Activating mutation polycythemia vera and idiopathic erythrocytosis. N Engl J Med.
in the tyrosine kinase JAK2 in polycythemia vera, essential 2007;356:459-468.
thrombocythemia, and myeloid metaplasia with myelofibrosis. Scherber R, Dueck AC, Johansson P, et al. The Myeloproliferative
Cancer Cell. 2005;7:387-397. Neoplasm Symptom Assessment Form (MPN-SAF):
Lussana F, Rambaldi A, Finazzi MC, et al. Allogeneic international prospective validation and reliability trial in 402
hematopoietic stem cell transplantation in patients with patients. Blood. 2011;118:401-408.
polycythemia vera or essential thrombocythemia transformed Stein BL, Williams DM, O’Keefe C, et al. Disruption of the ASXL1
to myelofibrosis or acute myeloid leukemia: a report from gene is frequent in primary, post-essential thrombocytosis
the MPN Subcommittee of the Chronic Malignancies and post-polycythemia vera myelofibrosis, but not essential
Working Party of the European Group for Blood and Marrow thrombocytosis or polycythemia vera: analysis of molecular
Transplantation. Haematologica. 2014;99:916-921. genetics and clinical phenotypes. Haematologica. 2011;96:
Marchioli R, Finazzi G, Specchia G, et al. Cardiovascular events 1462-1469.
and intensity of treatment in polycythemia vera. N Engl J Med. Tefferi A, Cervantes F, Mesa R, et al. Revised response criteria for
2013;368:22-33. myelofibrosis: International Working Group-Myeloproliferative
Mead AJ, Rugless MJ, Jacobsen SE, et al. Germline JAK2 mutation Neoplasms Research and Treatment (IWG-MRT) and European
in a family with hereditary thrombocytosis. N Engl J Med. LeukemiaNet (ELN) consensus report. Blood. 2013;122:1395-
2012;366:967-969. 1398.
Mesa RA, Niblack J, Wadleigh M, et al. The burden of fatigue Tefferi A, Guglielmelli P, Finke C, et al. Integration of mutations
and quality of life in myeloproliferative disorders (MPDs): an and karyotype towards a genetics-based prognostic scoring
international internet-based survey of 1179 MPD patients. system (GPSS) for primary myelofibrosis [abstract]. Blood.
Cancer. 2007;109:68-76. 2014;124(21):406.

Tefferi A, Rumi E, Finazzi G, et al. Survival and prognosis among Myeloid (and lymphoid) neoplasms associated
1545 patients with contemporary polycythemia vera: an with eosinophilia and abnormalities of
international study. Leukemia. 2013;27:1874-1881. PDGFRA, PDGFRB, or FGFR1
Tefferi A, Vardiman JW. Classification and diagnosis of
myeloproliferative neoplasms: the 2008 World Health Baccarani M, Cilloni D, Rondoni M, et al. The efficacy of
Organization criteria and point-of-care diagnostic algorithms. imatinib mesylate in patients with FIP1L1-PDGFRalpha-
Leukemia. 2008;22:14-22. positive hypereosinophilic syndrome. Results of a multicenter
Tefferi A, Verstovsek S, Barosi G, et al. Pomalidomide is active in prospective study. Haematologica. 2007;92:1173-1179.
the treatment of anemia associated with myelofibrosis. J Clin Cheah CY, Burbury K, Apperley JF, et al. Patients with myeloid
Oncol. 2009;27:4563-4569. malignancies bearing PDGFRB fusion genes achieve durable
Vannucchi AM, Guglielmelli P, Rotunno G, et al. Mutation- long-term remissions with imatinib. Blood 2014;123:3574-3577.
enhanced International Prognostic Scoring System (MIPSS) Chen J, Deangelo DJ, Kutok JL, et al. PKC412 inhibits the zinc
for primary myelofibrosis: an AGIMM & IWG-MRT project finger 198-fibroblast growth factor receptor 1 fusion tyrosine
[abstract]. Blood. 2014;124(21):405. kinase and is active in treatment of stem cell myeloproliferative
Vannucchi AM, Kiladjian JJ, Griesshammer M, et al. Ruxolitinib disorder. Proc Natl Acad Sci U S A. 2004;101:14479-14484.
versus standard therapy for the treatment of polycythemia vera. Cools J, DeAngelo DJ, Gotlib J, et al. A tyrosine kinase created by
N Engl J Med. 2015;372:426-435. fusion of the PDGFRA and FIP1L1 genes as a therapeutic target
Vaquez H. On a special form of cyanosis accompanied by of imatinib in idiopathic hypereosinophilic syndrome. N Engl J
excessive and persistent erythrocytosis. Comp Rend Soc Biol. Med. 2003;348:1201-1214.
1892;12: 384-388. David M, Cross NC, Burgstaller S, et al. Durable responses to
Verstovsek S, Mesa RA, Gotlib J, et al. A double-blind, placebo- imatinib in patients with PDGFRB fusion gene-positive and
controlled trial of ruxolitinib for myelofibrosis. N Engl J Med. BCR-ABL-negative chronic myeloproliferative disorders. Blood.
2012;366:799-807. 2007;109:61-64.
Visconte V, Makishima H, Jankowska A, et al. SF3B1, a splicing Golub TR, Barker GF, Lovett M, et al. Fusion of PDGF receptor
factor is frequently mutated in refractory anemia with ring beta to a novel ets-like gene, tel, in chronic myelomonocytic
sideroblasts. Leukemia. 2012;26:542-545. leukemia with t(5;12) chromosomal translocation. Cell.
Wernig G, Kharas MG, Okabe R, et al. Efficacy of TG101348, 1994;77:307-316.
a selective JAK2 inhibitor, in treatment of a murine model Gotlib J, Cools J, Malone JM III, et al. The FIP1L1-PDGFRalpha
of JAK2V617F-induced polycythemia vera. Cancer Cell. fusion tyrosine kinase in hypereosinophilic syndrome and
2008;13:311-320. chronic eosinophilic leukemia: implications for diagnosis,
Zhang SJ, Rampal R, Manshouri T, et al. Genetic analysis of classification, and management. Blood. 2004;103:2879-2891.
patients with leukemic transformation of myeloproliferative Jovanovic JV, Score J, Waghorn K, et al. Low-dose imatinib
neoplasms shows recurrent SRSF2 mutations that are associated mesylate leads to rapid induction of major molecular responses
with adverse outcome. Blood. 2012;119:4480-4485. and achievement of complete molecular remission in FIP1L1-
PDGFRA-positive chronic eosinophilic leukemia. Blood.
2007;109:4635-4640.
Myeloproliferative neoplasm, unclassifiable
Metzgeroth G, Walz C, Erben P, et al. Safety and efficacy
Swerdlow SH, Campo E, Harris NL, et al., eds. World Health of imatinib in chronic eosinophilic leukaemia and
Organization Classification of Tumors of Hematopoietic and hypereosinophilic syndrome: a phase-II study. Br J Haematol.
Lymphoid Tissues. Lyon, France: IARC Press; 2008. 2008;143:707-715.
Thiele J, Kvasnicka HM, Orazi A. Bone marrow histopathology Stover EH, Chen J, Folens C, et al. Activation of FIP1L1-
in myeloproliferative disorders—current diagnostic approach. PDGFRalpha requires disruption of the juxtamembrane
Semin Hematol. 2005;42:184-195. domain of PDGFRalpha and is FIP1L1-independent. Proc Natl
Thiele J, Kvasnicka HM, Vardiman J. Bone marrow Acad Sci U S A. 2006;103:8078-8083.
histopathology in the diagnosis of chronic myeloproliferative Swerdlow SH, Campo E, Harris NL, et al., eds. World Health
disorders: a forgotten pearl. Best Pract Res Clin Haematol. Organization Classification of Tumors of Hematopoietic and
2006;19:413-437. Lymphoid Tissues. Lyon, France: IARC Press; 2008.

CHAPTER
17
Acquired marrow failure
syndromes: aplastic anemia,
paroxysmal nocturnal
hemoglobinuria, and
myelodysplastic syndromes
Phillip Scheinberg, Amy E. DeZern, and David P. Steensma
Introduction, 489 Paroxysmal nocturnal Myelodysplastic syndromes, 503
Aplastic anemia, 489 hemoglobinuria, 499 Bibliography, 519
Introduction altered hematopoietic growth factor receptor/kinase signal-

ing (congenital amegakaryocytic thrombocytopenia). Simi-
The term bone marrow failure refers to the inability of hema- lar mechanisms may underlie some acquired marrow failure
topoiesis to meet physiologic demands for the production of syndromes, such as acquired haploinsufficiency for ribo-
healthy blood cells. Pancytopenia may result from marrow somal protein RPS14 in MDS associated with chromosome
failure, or cytopenias involving a single myeloid lineage (eg, 5q deletion, which parallels heterozygous ribosomal protein
anemia) may dominate; usually lymphopoiesis is relatively mutations observed in DBA.
preserved. This chapter reviews acquired marrow failure syndromes,
The causes of marrow failure are diverse and may either including aplastic anemia (AA), paroxysmal nocturnal
be extrinsic to the marrow, as in the disordered immune hemoglobinuria (PNH), and MDS. For discussion of inher-
response that characterizes aplastic anemia, or intrinsic, as in ited marrow failure syndromes such as Fanconi anemia, dys-
the hematopoietic progenitor or stem cell defects that under- keratosis congenita, and Diamond-Blackfan anemia, please
lie the myelodysplastic syndromes (MDS). Bone marrow fail- refer to Chapter 15.
ure syndromes can be acquired or, more rarely, congenital.
The range of molecular mechanisms responsible for con-
genital marrow failure states is broad, including abnormal Aplastic anemia
DNA damage response (Fanconi anemia [FA]), defective
ribogenesis (Diamond-Blackfan anemia [DBA]), abnormal Definition
telomere dynamics (dyskeratosis congenita [DC]), and
AA is a hematopoietic stem cell (HSC) disorder associated
Conflict-of-interest disclosure: Dr. Scheinberg: speaker: Novartis. with markedly reduced marrow cellularity and deficient
Dr. DeZern: no competing financial interest. Dr. Steensma: advisory blood cell production. Classification and prognosis in AA
board: Incyte, Novartis, Celgene, MEI Pharma, Takeda, and Genoptix.
Off-label drug use: Dr. Scheinberg: cyclosporine, rabbit antithy- are related to the severity in the depression of peripheral
mocyte globulin, alemtuzumab, cyclophosphamide; Dr. DeZern: not blood counts. Severe AA (SAA) is defined by depression of
applicable. Dr. Steensma: eltrombopag, romiplostim, lenalidomide blood counts involving at least two hematopoietic lineages
for non-del5q, darbepoetin, epoetin.
| 489

490 | Acquired marrow failure syndromes
Table 17-1 Classification of aplastic anemia by severity
Severe aplastic anemia* Moderate aplastic anemia

Bone marrow cellularity <30% Decreased bone marrow cellularity
Depression of at least two of the following three hematopoietic lineages: Depression of at least two of three hematopoietic lineages not
Absolute neutrophil count <0.5 × 109/L fulfilling the severity criteria as specified in the left column
Transfusion dependence, with absolute reticulocyte count, <60 × 109/L,
platelet count <20 × 109/L
* Very severe aplastic anemia is reserved for patients who fulfill criteria for severe AA but with an absolute neutrophil count <0.2 × 109/L.
(ie, absolute reticulocyte count <60 × 109/L, absolute neu- certain antibiotics such as chloramphenicol and gold salts
trophil count <0.5 × 109/L, or platelet count <20 × 109/L) are more clearly associated with development of AA. Envi-
and bone marrow hypocellularity (<30%, excluding lym- ronmental exposure to benzene has also been strongly linked
phocytes). Very severe AA has an absolute neutrophil count to marrow failure.
of <0.2 × 109/L, whereas moderate AA is characterized by Hepatitis-associated AA accounts for 2%-5% of cases of
depression of blood counts not fulfilling the definition of AA in Europe and 4%-10% of cases in East Asia. AA has been
severe disease (Table 17-1). AA is associated with normal reported to occur in 28%-33% of patients requiring ortho-
cytogenetics. An abnormal karyotype in a patient with a topic liver transplantation for fulminant non-A, non-B, and
hypocellular marrow is more consistent with a diagnosis of non-C hepatitis. This seronegative hepatitis in patients with
MDS, although some investigators believe that certain chro- posthepatitis AA does not appear to be caused by any of the
mosomal abnormalities such as trisomy 8 or deletion 13q known hepatitis viruses and often is referred to as hepatitis/
can still be consistent with an AA diagnosis. AA syndrome. An immune pathogenesis following a puta-
AA may be acquired and idiopathic, or it can arise in the con- tive trigger is suspected, but the precise mechanism of this
text of an inherited marrow failure syndrome. This distinction form of marrow failure is unknown. AA evolves with a typi-
carries profound implications for management and treatment. cal delay of several weeks to months after the episode of hep-
For example, immunosuppression is a therapeutic option in atitis, usually after improvement of liver enzymes.
acquired AA, although this treatment modality is ineffective in Regardless of the etiology, hematopoiesis is reduced mark-
inherited forms of marrow failure. AA is a diagnosis of exclu- edly in all patients with AA, as reflected by marrow histology,
sion and systemic causes for pancytopenia should be ruled out. low numbers of circulating or marrow CD34 cells, diminished
The diagnosis of AA usually is reserved for naturally occurring numbers of long-term culture-initiating cells (a surrogate
conditions and excludes those patients with a recent history of measure of HSCs), and poor hematopoietic colony format
cytotoxic chemotherapy or exposure to ionizing radiation. ion in cells obtained from an aplastic marrow. Obligatory
Epidemiology, etiology, and pathogenesis Table 17-2 Classification of aplastic anemia by etiology
Acquired aplastic anemia
AA is rare in Western Europe and the United States (2 cases
Primary
per million population per year), but the incidence in China,
Idiopathic aplastic anemia
Southeast Asia, and Mexico is estimated to be three to four Pregnancy-associated aplastic anemia
times higher. AA is primarily a disease of children and Aplastic anemia/paroxysmal nocturnal hemoglobinuria syndrome
younger adults, with another peak in incidence in patients 60 (AA/PNH)
years and older; in the latter group, some cases may reflect Secondary
diagnostic overlap with hypoplastic MDS. Drug associated
AA can be acquired or constitutional. Idiopathic acquired Iatrogenic/cytotoxic
AA is perceived as a T-cell-mediated autoimmune process Idiosyncratic
that accounts for most cases in North America. Idiopathic Radiation associated
AA is more common than AA associated with medication Iatrogenic
exposure, pregnancy, or hepatitis (Table 17-2). The associa- Accidental
tion of AA with drug exposure has been of great interest for Viruses
decades. However, the level of evidence linking AA to specific Epstein-Barr virus
drugs is variable. The nonsteroidal anti-inflammatory agents Cytomegalovirus
indomethacin, diclofenac, and butazones (such as phenylbu- Hepatitis/aplastic anemia syndrome (seronegative for hepatitis
viruses)
tazone), antithyroid medication (propylthiouracil), certain
Pancytopenia of autoimmune diseases
anticonvulsants (such as hydantoins, carbamazepine), and

Aplastic anemia | 491
involvement of multiple lineages points toward HSCs or very telomerase, TERC and TERT, have been described in about
early hematopoietic progenitor cells as main targets of the 5%-10% of AA patients who lack the overt clinical stigmata
pathophysiologic mechanism in AA (Figure 17-1). of DC (a congenital disorder associated with germ line
Clinical response to immunosuppressive therapy targeting TERC and TERT mutations and, more commonly, muta-
T cells (eg, antithymocyte globulin [ATG]), described fur- tions in DKC1 [encoding dyskerin]). Telomeres progres-
ther below in the “Immunosuppressive therapy” section, sively shorten over successive cell divisions, and the
supports an immune-mediated pathogenesis of AA. This telomerase complex repairs the naturally occurring erosion
notion is corroborated by laboratory data where: (i) acti- of chromosome ends, avoiding replicative senescence. This
vated cytotoxic T-cells produce interferon-γ and tumor complex consists of an RNA template (encoded by TERC), a
necrosis factor α (TNFα), which suppress hematopoiesis; catalytic subunit (encoded by TERT), and associated pro-
(ii) Fas receptor (CD95) expression is increased by hemato- teins, which add short nucleotide repeats (TTAGGG) to the
poietic progenitor and stem cells are increased in patients ends of chromosomes (telomeres) during each cellular divi-
with AA likely induced by interferon-γ and TNFα triggering sion. When a critically short telomere length is reached, cell
apoptosis upon ligand binding; (iii) T-cell associated cyto- death ensues. Although telomerase expression is generally
kines exert a direct inhibitory effect on hematopoietic pro- low in most somatic cells, expression in lymphocytes and
genitors; (iv) T-bet, a transcription factor that binds to the HSCs is high, in view of their life-long high replicative
interferon-gamma promoter region, is upregulated in aplas- capacity. Telomerase dysfunction is associated with acceler-
tic anemia; and (v) a decrease in regulatory T cells (Tregs) and ated telomere shortening, and may result in premature
an increase in interleukin-17-producing T-helper cells (Th17 death of rapidly proliferating cells (eg, HSCs). Thus, it is
cells) have been described at diagnosis in AA, with the ratio possible that the inability to properly counter telomeric
between these two T-cell subsets normalizing in responding attrition may contribute to AA pathogenesis in some
patients to immunosuppression. Many of these in vitro patients because of a lower capacity to restore hematopoie-
observations have been confirmed in a murine AA model, sis. It is unclear whether a TERC or TERT genetic defect is
corroborating with an autoimmune pathogenesis. Although sufficient to define the pathogenesis in some cases of appar-
the T-cell-mediated process is intrinsically polyclonal, oligo- ent acquired AA or whether it is a contributor to other
clonal expansion of CD8+ T cells has been observed in some pathophysiologic mechanisms of marrow destruction, such
AA patients, raising the possibility that these might represent as an aberrant immune system. Some mutants appear to
immunodominant autoreactive clones. Human leukocyte respond to immunosuppressive therapy suggesting that the
antigen (HLA)-DR15, a split of HLA-DR2, is overrepre- interaction between these genetic defects and the immune
sented in AA (40%-50% of patients, compared with an anti- system is more complex. More recently, mutations in
gen frequency of ~20% in the general population), which GATA2, an important transcription factor for HSC mainte-
also suggests an immune pathogenesis of idiopathic AA. nance, have been described in some AA cases associated
Although diverse triggers, such as viruses or chemical hap- with distinct marrow findings of severe reduction in mono-
tens, may serve as inciting events in individual cases, the final cytes, B and NK cells.
autoimmune pathway appears to be uniform.
Intrinsic HSC defects may also play a pathogenic role in
some AA cases. Mutations in genes encoding components of
Symptoms of AA are a consequence of lack of production of
Indirect
blood cells. Patients present with pallor and fatigue due to
Immunologically
mediated
effects anemia, with mucocutaneous bleeding due to thrombocyto-
Autoantigens penia, or with infection due to neutropenia. More severe
Altered proteins hemorrhage in the central nervous system or gastrointestinal
tract is not typical on disease presentation. AA usually arises
in a previously healthy patient who has no history of malig-
Chemicals (eg, benzene) Direct toxic
Medications nancy and no exposure to cytotoxic drugs or history of radi-
effects Hematopoietic stem
Viruses ation exposure. A family history of marrow failure or
(eg, EBV, CMV, cell and progenitor
Ionizing cell compartment dysmorphology may help identify inherited causes of pancy-
hepatitis virus) radiation
topenia, such as FA or DC. Drug and chemical exposures
Figure 17-1 Types of stem cell injury in aplastic anemia. Aplastic should be queried in the interview, but these are notoriously
anemia can result both from direct toxic effects on hematopoietic stem difficult to evaluate quantitatively as the history is subject to
cells and progenitors and from an aberrant T-cell-driven immune recall bias. Confirmation of a causal relationship is difficult
response. CMV = cytomegalovirus; EBV = Epstein-Barr virus. to ascertain in practice, and management is not likely

to differ from those cases without a putative trigger. Discon- Marrow biopsy is the gold standard for assessing marrow
tinuation of a drug strongly suspected to be associated with cellularity. Because residual marrow cellularity in AA may be
the onset of pancytopenia is reasonable for a few weeks; patchy and variable, results should be interpreted within the
however, a prolonged observation period of several weeks to clinical context of the patient. Cytogenetic results are typi-
months before initiation of therapy is not recommended, cally normal in AA, and a finding of an abnormal karyotype
especially when pancytopenia remains severe. is suggestive of a dysplastic process. The identification of
Splenomegaly and hepatomegaly are not typical features monosomy 7 in the karyotype at diagnosis strongly suggests
of AA and should point toward another diagnosis. Short MDS, as most of these cases are refractory to immunosup-
stature, musculoskeletal abnormalities (particularly radial pression and carry a higher rate of progression to high-risk
ray anomalies), dysplastic nails, skin rashes, oral leukoplakia, MDS and leukemia.
exocrine pancreatic insufficiency, or other congenital anom-
alies may suggest an inherited bone marrow failure state (see
Differential diagnosis
Chapter 15). The absence of characteristic physical findings
or a suggestive family history does not rule out an inherited When evaluating a patient with pancytopenia and a hypocel-
marrow failure syndrome, which can manifest in adulthood lular marrow, the physician must exclude a number of other
with apparent acquired AA or MDS and no physical stig- conditions before a diagnosis of AA can be made (Table 17-3;
mata. The detection of genetic defects associated with FA or Figure 17-2). The most common disorders include MDS,
DC in some adults with AA but without dysmorphology has acute leukemia, PNH, myelofibrosis, hairy cell leukemia, cer-
blurred the distinction between inherited and acquired tain infections (tuberculosis), nutritional deficiency (eg,
forms of marrow failure. It is important to query in the med- anorexia nervosa), T-cell large granular lymphocyte (T-LGL)
ical interview about earlier blood count abnormalities, mac- disease (T-LGL can coexist with AA), and HIV infection. The
rocytosis, or relevant pulmonary (fibrosis) or liver disease diagnostic approach to the patient with pancytopenia
(cirrhosis) in the patient or the patient’s family, which are includes the following: history including medications, previ-
frequent in telomere disorders. ous chemotherapy or radiation exposure, occupational toxic
Malignant or markedly dysplastic cells are not seen in the exposures, HIV risk factors, family history; physical exami-
marrow in AA, although mild dysplastic features sometimes nation, paying particular attention to presence of organo-
are noted, which can cause diagnostic confusion with hypo- megaly, lymphadenopathy, or congenital abnormalities
plastic MDS. In the AA marrow, all myeloid hematopoietic (short stature, nail dystrophy, abnormalities in skin, arms,
cell lines are diminished, whereas residual lymphocytes head, eyes, mucosa, or skeletal); complete blood count,
and plasma cells frequently are observed. The marrow is including reticulocyte count and peripheral smear examina-
characteristically hypocellular with an expansion of the fat tion; liver function tests, vitamin B12 and folate levels, lactate
cells and no increase in reticulin. dehydrogenase (LDH), haptoglobin, and flow cytometry for
Table 17-3 Differential diagnosis of

Pancytopenia
idiopathic aplastic anemia and
Hypocellular bone marrow Hypercellular bone marrow
pancytopenia
Primary marrow disorders Direct marrow involvement
Acquired aplastic anemia Myelodysplastic syndrome
Constitutional forms of marrow failure (FA, DC) Paroxysmal nocturnal hemoglobinuria
Hypocellular MDS Primary myelofibrosis
Hairy cell leukemia Lymphoma
Aleukemic leukemia Metastatic carcinomas and sarcomas
Rarely Systemic illnesses
Lymphoma Systemic lupus erythematosus
Myeloma Hypersplenism
Primary myelofibrosis Sepsis
Systemic illnesses Alcoholism
Hypothyroidism Brucellosis
Anorexia nervosa Ehrlichiosis
Infections Vitamin deficiencies (vitamin B12, folate)
Tuberculosis
Q fever

Myelophthisis, lymphoma
Metastatic carcinoma, sarcoma
Pancytopenia Nonhematopoietic infiltrate
Bone marrow aspirate/biopsy Cellularity >30%
Cellularity <30% Dry tap Blasts in Hepatosplenomegaly

Splenomegaly buffy coat Leukoerythroblastosis
Consistent flow and/or spicules Bone marrow fibrosis
+TRAP stain
Absent constitutional
symptoms, lymphadenomegaly,
and hepatosplenomegaly
Hairy cell Acute leukemia Myelofibrosis
leukemia
Abnormal cytogenetics Normal cytogenetics ± modest Positive chromosome

Frank dysplasia (excess dysplastic changes fragility test
blasts, atypical mega-
karyoctes, severe
dyserythropoiesis) Family history of AA or pulmonary
or hepatic fibrosis
Acquired aplastic anemia Characteristic phenotypic abnormalities Fanconi anemia
Very short telomere length
Hypocellular MDS
Flow cytometry + for PNH
Consider underlying telomeropathy
Aplastic anemia/PNH
Figure 17-2 Diagnostic algorithm of primary marrow causes of pancytopenia. In patients with a hypocellular marrow, the main differential is
between hypocellular MDS and AA. Normal cytogenetics and only modest dysplastic changes favors AA, whereas more pronounced dysplasia
(micromegakaryocytes, left shift myelopoiesis with increase in blasts, significant dyserythropoiesis) and an abnormal cytogenetics favors MDS.
Patients with AA and a paroxysmal nocturnal hemoglobinuria (PNH) clone, are classified as AA/PNH. In those with normal or increased marrow
cellularity, differential includes a myelophthisic process (lymphoma, metastatic carcinoma, or sarcomas), rare presentations of leukemia,
myelofibrosis, and hairy cell leukemia. Close attention should be paid to the spicules, which may harbor a blast population indicating a leukemic
process. Hepatosplenomegaly, increase in marrow reticulin (fibrosis), and leukoerythroblastic peripheral smear favors myelofibrosis. In hairy cell
leukemia, marrow cellularity is often elevated, but in 10%-20% of cases, cellularity can be low. A dry tap and splenomegaly in a pancytopenic
patient hints toward the possibility of hairy cell leukemia. The diagnosis is confirmed by lymphoid immunophenotyping (with markers such as
CD123, CD103, CD11c, and CD25) and a marrow that stains positive for tartrate resistant acid phosphatase (TRAP) and annexin A1 by
immunohistochemistry.
PNH evaluation; bone marrow aspirate and biopsy with micronutrient deficiencies. Pancytopenia in this setting is
cytogenetic studies; and chromosome fragility tests, particu- associated with a hypocellular marrow with serous fat atro-
larly in children and younger adults for FA testing. phy. Vitamin B12 and folate levels should be determined in all
The presence of dysplastic immature hematopoietic cells patients, although the marrow in vitamin B12 or folate defi-
or blast cells should lead to a diagnosis of hypoplastic MDS or ciency is typically hypercellular and megaloblastic rather than
acute leukemia. Similarly, marrow cytogenetic analysis may hypocellular. HIV infection or AIDS is associated with cyto-
detect a cytogenetic abnormality diagnostic of lymphoid or penia, morphologic dysplasia, and marrow hypocellularity in
myeloid leukemic disorders. Hairy cell leukemia frequently ~10% of cases. A careful inquiry into HIV risk factors and an
presents as pancytopenia with difficulty in aspirating the HIV test are prudent.
marrow, or a “dry tap,” along with splenomegaly. Pancytope- T-LGL disease is a rare condition characterized by circu-
nia can arise in the setting of anorexia nervosa as an epiphe- lating T-cells bearing the CD57 marker of effector or cyto-
nomenon of the eating disorder, possibly because of multiple toxic T-cells. T-LGL disease, like PNH, can coexist with AA

or MDS. T-LGL disease should be considered if increased

LGL are noted on the peripheral blood smear or if the patient
has rheumatoid arthritis, which is known to be associated • Chromosome abnormalities favor an MDS diagnosis over AA,
with T-LGL disease. Single-lineage cytopenia is more com- but both karyotypic changes and clonal somatic mutations can
mon in T-LGL with clinical presentation of isolated anemia sometimes be seen in patients with AA.
or neutropenia most typical. Flow cytometry and testing for • Chromosome fragility tests are important to exclude FA in
a clonal T-cell receptor gene rearrangement is appropriate children and younger adults <40 years of age presenting with
when T-LGL is suspected. idiopathic marrow failure.
A possible underlying cause of AA is FA, which can pres-

ent with apparent acquired AA in adulthood without other
Therapy
classic features of this disease. Therefore, diepoxybutane
(DEB) or mitomycin C testing to exclude chromosome fra- Without treatment, almost all patients with severe or very
gility should be considered in patients with AA <40 years of severe AA eventually will succumb to infection or to hemor-
age, even in the absence of musculoskeletal abnormalities. rhagic complications. Therefore, such patients require
FA is discussed further in Chapter 15. urgent therapy once a diagnosis is confirmed. The standard
The distinction between AA and hypoplastic MDS may be of care for moderate AA, in contrast, is not established.
difficult to make, and increasing evidence suggests that Except for cases in which there is transfusion dependence,
immune-mediated mechanisms similar to those postulated to therapy is optional because survival is not affected by treat-
cause AA may contribute to the cytopenias associated with ment. Rarely, patients with AA can spontaneously recover
some cases of hypoplastic MDS and also normocellular or normal hematopoiesis. Spontaneous remission is most often
hypercellular MDS, even in the absence of a preceding diag- seen with drug-induced AA and usually occurs within 1-2
nosis of AA. Such evidence includes the identification of months of discontinuing the offending drug.
clonal-activated cytotoxic T-cell populations in both AA and Therapy of severe AA consists of allogeneic hematopoietic
MDS, the coexistence of PNH and T-LGL clones in both AA stem cell transplantation (HSCT) or immunosuppressive
and MDS, and improved blood counts in a subset of MDS therapy. At the time of diagnosis, all potential transplanta-
patients treated with ATG or cyclosporine immunosuppres- tion candidates should be HLA typed to identify a sibling
sion (see next section on myelodysplastic syndromes). donor or a potential histocompatible unrelated donor.
Hypolobated neutrophils, dysplastic megakaryocytes, or Because registries frequently take several months to identify
abnormally localized and increased immature precursors a donor, this process should be initiated if a sibling donor
favor a diagnosis of hypoplastic MDS rather than AA. Some- cannot be identified, especially in younger patients.
times the only way to make the distinction between AA and
MDS is by detection of an abnormal cytogenetic clonal popu-
Supportive care, transfusions, and
lation, but even this may not be diagnostic of MDS because
hematopoietic growth factors
some cytogenetically abnormal clones can be observed tran-
siently in AA. Supportive care is instituted to sustain blood counts and
In acquired AA, PNH clones can be detected by flow alleviate symptoms and risks associated with pancytopenia
cytometry in 40%-50% of cases, but these are usually small and consist of transfusion of irradiated, leukocyte-depleted
(<10% of cells). A PNH clone can expand later in the course blood products (blood and/or platelets) due to the risk for
of disease leading to frank hemolysis; this occurs most com- alloimmunization from chronic transfusions. If the patient
monly in patients with larger preexisting PNH clones at is cytomegalovirus (CMV) negative, it is best to use CMV-
diagnosis. PNH clones can remain stable over time or reduce negative blood products or leukocyte-depleted products.
in size having no clinical consequence. Indicators of the Transfusions should not be withheld from symptomatic
presence of a PNH clone include elevated LDH, absent hap- patients. In patients who are candidates for allogeneic HSCT,
toglobin, increased reticulocytes, and erythroid predomi- the use of leuko-depleted blood products is critical to
nance in the marrow. decrease the risk of alloimmunization, and transfusions
should never be given from family members because doing
Key points so could increase the risk of subsequent graft rejection.
Prophylactic transfusions to maintain platelets >10 × 109/L
• AA is a diagnosis of exclusion and can result from intrinsic
generally prevents major hemorrhagic complications. The role
stem cell defects, immunologic impairment of hematopoiesis, or
of preventive antibiotics in neutropenic patients is not well
toxic effect of an exogenous exposure.
defined. Prophylaxis for Pneumocystis jirovecii should be
• PNH clones are frequently seen in patients with AA.
used for 6 months, and prophylaxis for herpesviruses should

be considered in the first 1-2 months following ATG + cyclo- from mobilized peripheral blood when compared with a
sporine (CsA) in seropositive individuals. The role for anti- bone marrow source. In contrast to allogeneic HSCT under-
bacterial and antifungal prophylaxis is less well defined and taken for malignant disorders, where GVHD offers potential
could be considered on an individual case. graft-versus-tumor benefits, GVHD is to be avoided at all
Most patients with AA have an elevated serum erythro- costs in the AA setting, because its occurrence is associated
poietin level and do not respond to recombinant erythropoi- with decreased survival and long-term quality of life. Thus,
etin. Although typical AA will not respond to myeloid bone marrow is the preferred source of HSCs in AA patients
growth factors either (ie, granulocyte colony-stimulating undergoing HSCT.
factor [G-CSF] or granulocyte-macrophage colony-stimu- The standard conditioning regimen in matched-sibling
lating factor [GM-CSF]), some patients do improve neutro- transplantation includes ATG and cyclophosphamide, which
phil counts, and these growth factors may have a role in results in a long-term survival rate of about 75%-80% in
decreasing infectious morbidity while awaiting definitive patients of all ages and is superior to conditioning regimens
treatment with immunosuppression or HSCT. In several based on total lymphoid irradiation. The latter conditioning
randomized trials, the addition of G-CSF to standard ATG regimen has been shown to result in a higher incidence of
and cyclosporine therapy did not improve the rates of hema- secondary malignancies, infertility, and, in children, retarda-
tologic response rate or survival. tion of growth and development.
Corticosteroids are ineffective, increase the risk of infec- A matched-sibling HSCT is the preferred first therapy in
tion, and should not be used as therapy in AA. The role of children and young adults diagnosed with SAA (up to the age
corticosteroids in severe AA is limited to serum sickness proof 40), as transplant-related mortality and GVHD rates are
phylaxis with concurrent ATG administration. Androgens lower in this younger age group. GVHD and infection remain
may have a supportive role in some patients throughout the limiting factors to the success of transplantation, especially in
treatment course of AA. Androgens, however, should not be older patients with severe AA. GVHD increases in frequency
used as primary upfront therapy. and severity in recipients >20 years of age, and more so >40
years of age. The increased risk of GVHD contributes to the
poor survival in older patients, especially in those >40 years
Key points
of age. The historically high rejection rate associated with AA
• If transfusions are needed in a patient with AA, use irradiated, patients appears to have improved to rates similar to those
leukocyte-depleted blood products. observed in patients transplanted for other conditions, likely
• Transfusions should not be from family members (especially in because of less frequent use of immunogenic blood products
transplant candidates). and better conditioning regimens. Standard prophylactic
• Transfusions should be used with caution but should not be
therapy for GVHD includes a calcineurin inhibitor (cyclo-
withheld in symptomatic anemic patients or in those at higher
sporine or tacrolimus) and methotrexate.
risk for bleeding (eg platelets <10 × 109/L).
In general, marrow transplantation using unrelated
• AA does not usually respond to G-CSF or erythropoietin, but
myeloid growth factors may have a role in decreasing infectious
donors has resulted in higher morbidity and mortality than
morbidity while the patient awaits definitive treatment with transplantations using matched sibling donors in AA and
immunosuppression or stem cell transplantation for severe AA. generally has been reserved for patients who lack a matched
• Corticosteroids should not be used as therapy in AA except as sibling donor and fail to respond to one or more rounds of
prevention of serum sickness in patients receiving ATG. ATG and cyclosporine therapy. Transplantation results with
unrelated donors are typically best in younger patients (ie,
children and young adults) who have not had significant
Hematopoietic stem cell transplantation
infections or developed alloimmunization.
In AA, the pretransplantation conditioning regimen primar- For older patients, reduced-intensity transplantation condi-
ily is administered to provide immunosuppression, which tioning regimens using low doses of total-body irradiation or
enables the donor stem cells to engraft and also eliminate fludarabine have shown promise in reducing rejection rates.
activated immune cells that may be causing the marrow In recent years, however, outcomes with matched unrelated-
aplasia. Bone marrow has been a traditional source for the donor HSCT have improved likely because of more stringent
stem cell graft, but the use of peripheral blood stem cells has donor selection with high-resolution-molecular tissue typing,
gained in popularity in the past 10-15 years. This practice has less toxic and more effective conditioning regimens, and higher
resulted in an untoward consequence in transplanted AA quality transfusion and antimicrobial supportive care. In some
patients, where several reports in the recent years from reports in children, outcomes with a matched-unrelated
Europe and the United States showing an increase rate of HSCT have compared favorably to those observed with sib-
graft-versus-host disease (GVHD), with stem cells derived ling donors, and this treatment modality is becoming the

preferred salvage treatment modality in younger patients who The usual time to response to h-ATG/CsA therapy in SAA
fail an initial course of immunosuppression when a matched- is approximately 10-12 weeks. In most studies, responses are
unrelated histocompatible donor is available. defined as achieving blood counts that no longer fulfill crite-
Outcomes with mismatched-unrelated, haploidentical, or ria for severe disease, as well as transfusion independence.
umbilical cord donors are not as favorable, with higher rates Total restoration of blood counts will occur in a minority of
of graft rejection, infectious complications, acute and patients, and recovery can be protracted.
chronic GVHD, and transplant-related mortality. These The overall response rate at 3 months in patients receiving
higher risk transplants usually are undertaken when immu- h-ATG/CsA is between 60% and 80%. Hepatitis-associated
nosuppression and other non-immunosuppression strate- and drug-related AA appears to be equally responsive to
gies to improve blood counts have failed and no related- or immunosuppressive therapy as idiopathic AA. Although
unrelated-histocompatible donor is available. most patients who will respond to immunosuppressive ther-
apy will do so by 6 months, in a small minority of patients,
time to recovery may be longer. Achieving hematologic
Key points response (partial or complete) to immunosuppression is
• Outcomes with HSCT are better in younger patients (espe- very important in SAA because it strongly associates with
cially patients <20 years old); in patients >40 years old, long-term survival.
transplantation-related mortality and morbidity increase. Both horse- and rabbit-derived ATG have activity in SAA,
• Bone marrow is the preferred source of stem cells in AA, not with most of the experience with horse occurring in the
peripheral blood stem cells, unlike the situation with hematologi- upfront setting and experience with rabbit in the salvage set-
cal neoplasms. ting. A repeat course of r-ATG and CsA may be given to
• Matched unrelated-donor transplantation should be reserved h-ATG refractory patients, which results in additional
for patients for whom an initial course of immunosuppression responses in approximately 35% of patients. In responders
has failed, especially in children and young adults.
to h-ATG/CsA, relapse has been reported in 35% of patients
by 5 years. Relapses can be related temporally to the discon-
tinuation of CsA or to the reduction of its dose. Cyclospo-
Immunosuppressive therapy
rine should be continued for at least 6 months. The benefit of
The principal immunosuppressive agent used in severe AA is a taper in abrogating or reducing relapse rates has not been
ATG, which is manufactured by delivering human T-cells to confirmed in prospective studies; however, most practicing
a horse or rabbit. The immunized animal then produces hematologists institute a slow CsA taper after 6 months in an
antibodies against antigens expressed on the surface of a attempt to prevent hematologic relapses. Relapsed patients
T-cell, which subsequently are harvested and purified. The may respond to an increased dose or reintroduction of CsA
resulting polyclonal animal serum has lymphocytotoxic or a second course of r-ATG, which results in hematologic
properties, and administration to humans leads to varying responses in about 60%-70% of patients. Approximately
degrees of lymphocyte depletion. Several in vitro and in vivo 25% of patients remain chronically dependent on CsA to
differences are observed between the two types of ATG maintain adequate blood counts.
despite a similar manufacturing process. Rabbit ATG The greater lymphocytotoxicty of r-ATG and its effective-
(r-ATG) has a longer half-life and results in a more durable ness in salvaging refractory and relapsed SAA patients
lymphocyte depletion compared to horse ATG (h-ATG). A prompted its use as initial therapy with the anticipation that it
difference in T-cell binding affinity, cytokine release and would be superior to h-ATG. In a randomized study, how-
T-cell subset depletion and reconstitution has also been ever, results with r-ATG were disappointing. The hematologic
shown distinct between the ATGs. response rate with r-ATG was 37% compared with 68% for
Initial investigations using h-ATG or CsA alone in AA were horse ATG at 6 months, and survival was inferior in the r-ATG
succeeded by studies of h-ATG and CsA in combination, with arm. These results suggest that h-ATG/CsA remains the pre-
improved response rates over monotherapy, becoming the ferred first-line immunosuppressive therapy in severe AA.
standard regimen. Multiple efforts to improve outcomes As alternative therapy to h-ATG/CsA, high-dose cyclo-
beyond h-ATG/CsA have been disappointing. Addition of phosphamide has been used, with response rates compar
mycophenolate mofetil, G-CSF or sirolimus did not improve able to that of h-ATG/CsA but with early reports suggesting
hematologic responses or decrease the relapse and clonal evo- fewer rates of relapse and clonal evolution. This initial favor-
lution rates. The use of more lymphocytotoxic agents such as able experience with cyclophosphamide from a single center
rabbit ATG, alemtuzumab, or cyclophosphamide led to was not reproduced in a randomized study; however, infec-
worse outcomes than with h-ATG/CsA in randomized stud- tious complications in the cyclophosphamide arm were
ies, due to a lower response rate and/or excess toxicities. substantial, leading to prolonged severe neutropenia,


hospitalizations, intensive care unit admissions, fungal patients, and a matched-unrelated HSCT is a viable option in
infections, and deaths. In addition, relapses and clonal evo- younger patients. Despite significant progress with matched
lutions have been observed with cyclophosphamide therapy unrelated-donor marrow transplantation, data from large reg-
at rates similar to that of h-ATG/CsA. istries suggest that survival rates remain inferior when com-
More recently, alemtuzumab has shown activity in the sal- pared with matched related-donor transplantation. Therefore,
vage setting, producing a hematologic response in 30%-40% patients of all ages who do not have a matched-sibling donor
of those patients with AA refractory to initial h-ATG/CsA should receive h-ATG/CsA as a first-line therapy.
and about 50%-60% in patients with relapsed SAA. These
results are similar to those reported with r-ATG in similar
Late clonal complications
settings, which prompted a trial of alemtuzumab as initial
therapy in SAA. The results, however, were disappointing Although 40%-50% of patients with AA will have PNH
because of a low response rate with alemtuzumab (~20%) clones at presentation, most are small and evolution to frank
compared with that of h-ATG/CsA (60%-70%). PNH is relatively infrequent. PNH that occurs after treat-
More recently, a nonimmunosuppressive approach has ment, however, frequently is subclinical and rarely is associ-
demonstrated significant activity in refractory AA. The ated with overt hemolysis or thrombosis. More concerning is
thrombopoietin receptor agonist eltrombopag produced evolution to MDS, which most frequently is associated with
improvement in blood counts in patients with severe AA either monosomy 7 or a trisomy 8 karyotype. Evolution to
who were refractory to at least one course of immunosup- MDS can occur in up to 15%-20% of patients in the first 20
pression. A hematologic response rate of approximately 40% years after diagnosis, an event usually associated with a
has been reported with this single agent with multilineage decrease in blood counts or refractoriness to immunosup-
responses observed. This outpatient oral therapy was well pression. The prognosis of patients with chromosome 7
tolerated. Based on this data, eltrombopag was FDA abnormalities is generally poor, whereas those with trisomy
approved in 2014 in SAA after insufficient response to initial 8 can respond to immunosuppressive therapy.
immunosuppressive therapy. Other cytogenetic abnormalities can be identified in fol-
The use of eltrombopag continues to be investigated as low-up of AA, which may not necessarily signify progression
salvage therapy, as well as in the upfront setting along with to MDS. Some of these abnormalities may be transient and
immunosuppression, to better define its role in AA. Clonal may not be associated with dysplastic marrow findings,
evolution, an initial concern with the use of eltrombopag, worsening in blood counts, or refractoriness to further thera-
has been observed in clinical trials in AA but at rates similar pies. The exception is the appearance of monosomy 7, which
to that expected in the population studied. However, the commonly is associated with frank dysplasia, with the only
effect of eltrombopag on late complications of relapse and curative approach being an HSCT from a related or alterna-
clonal evolution will need to be defined in larger AA cohorts tive donor. Recurrent genetic abnormalities in ASXL1,
with longer follow-up. DNMT3A, TET2, and BCOR genes have been recently
described in AA but their relevance is not clearly defined, and
clones bearing these markers may disappear with time.
Immunosuppression versus hematopoietic
stem cell transplantation
Key points
Although an incomplete response, relapse, and clonal evolu-
• Allogeneic stem cell transplantation from a matched sibling
tion to a clonal disease limit the success of h-ATG/CsA, the
donor is the treatment of choice for patients with severe AA in
morbidity and mortality of patients who receive HSCT can
children and young adults.
be substantial, primarily from infections and GVHD. Trans-
• For older patients, those without sibling donors, and those
plantation outcomes are age dependent. Currently, matched- who refuse transplantation or have significant comorbidities that
sibling allogeneic marrow transplantation is the treatment of precludes HSCT, immunosuppression with horse ATG plus
choice for children and adolescents, and it may be consid- cyclosporine combination should be initiated as soon as possible
ered as a first treatment option in adults up to the age of 40 once the diagnostic workup is completed.
years with severe AA for whom transplantation-related mor- • In patients without matched sibling donors, regardless of age,
bidity and mortality are relatively low and potential remain- horse ATG plus cyclosporine should be the preferred initial
ing life span is relatively long (Figure 17-3). For patients treatment. Horse ATG is superior to rabbit ATG as a first-line therapy.
without a matched-sibling donor or patients >40 years, • Outcomes with matched unrelated-donor transplantation have
immunosuppressive therapy is the initial treatment of choice. been improving and may be considered as the preferred salvage
treatment in children and young adults who fail an initial course of
In immunosuppression-refractory patients, sibling-donor
immunosuppression and have a histocompatible unrelated donor.
HSCT may constitute a second-line therapy in selected older

<40 with matched SAA diagnosed

sibling
>40 or no matched sibling

Sibling HSCT
>40 with histocompatible sibling,

Response at 6 months Horse ATG plus cyclosporine consider HSCT if suitable
Stop cyclosporine No histocompatible donor or

No response at 6 months
Long-term follow-up not suitable for HSCT
Children and young adults with

matched unrelated donor
Repeat immunosuppression
• Rabbit ATG plus cyclosporine
• Alemtuzumab
Consider matched unrelated
donor HSCT
No response at 6 months Response at 6 months
Non-HSCT options
• Eltrombopag HSCT options Long-term follow-up
• Androgens (12-week trial) • Mismatched unrelated
• G-CSF + Epo (12-week trial) • Haploidentical
• Supportive care (transfusions) • Umbilical cord
• Experimental protocols (alternative
immunosuppressants)
Figure 17-3 Algorithm for initial management of severe aplastic anemia (SAA). In patients who are not candidates for a matched related
hematopoietic stem cell transplantation (HSCT), immunosuppression with horse antithymocyte globulin (ATG) plus cyclosporine should be the
initial therapy. We assess for response at 3 and 6 months, but usually wait 6 months before deciding on further interventions in case of
nonresponders. In patients who are doing poorly clinically with persistent neutrophil count <0.2 × 109/L, we proceed to salvage therapies earlier
between 3 and 6 months. Transplant options are reassessed at 6 months and donor availability, age, comorbidities, and neutrophil count become
important considerations. We favor a matched unrelated HSCT in younger patients with a histocompatible donor, and repeat immunosuppression
for all other patients. In patients with a persistently low neutrophil count in the very severe range, we may consider a matched unrelated donor HSCT
in older patients. In patients who remain refractory after two cycles of immunosuppression, further management is then individualized, taking into
consideration suitability for a higher-risk HSCT (mismatched unrelated, haploidentical, or umbilical cord donor), age, comorbidities, neutrophil
count, and overall clinical status. *Non-HSCT options can be considered as salvage after insufficient response to 1 or 2 courses of immunosuppressive
therapy, which includes oral eltrombopag (FDA approved for this indication), androgens or combination growth factors. In those who are not
suitable for transplantation and a repeat course of immunosuppression due to advanced age, comorbidities, lack of donor, poor performance, or
personal preference, non-HSCT options can be considered earlier after refractoriness to initial course of therapy. Some authorities in SAA consider 50
years of age as the cutoff for sibling HSCT as first-line therapy. Adapted from Scheinberg P, Young NS. Blood. 2012;120:1185-1196.
Key points (continued) Key points (continued)

• The combination of ATG and cyclosporine is more effective an option for those who are not eligible for HSCT due to lack of
than single-agent immunosuppression in severe AA. a histocompatible donor, age comorbidities, or personal
• Relapses occur in about one-third of responders to horse ATG preference.
plus cyclosporine but often respond well to reinstitution of • Clonal evolution to MDS can occur in 10%-20% of patients
immunosuppressive therapy. long term.
• Repeat courses of rabbit ATG and cyclosporine may be given to • Higher risk transplant modalities from mismatched-unrelated,
refractory patients, resulting in a salvage rate of approximately 35%. haploidentical, or umbilical cord donors should be reserved for
• Eltrombopag is effective in 40% of SAA patients who have an patients refractory to two or more courses of immunosuppres-
insufficient response to initial immunosuppressive therapy and is sion who remain severely pancytopenic.

Paroxysmal nocturnal hemoglobinuria | 499
Paroxysmal nocturnal hemoglobinuria complement activation on the surface of erythrocytes.

Depending on the type of mutation in the PIGA gene, vari-
Definition ous degrees of CD55 and CD59 deficiency can occur. Patients
with PNH may have in their circulation an admixture of nor-
PNH is a rare, clonal HSC disorder that manifests with a mal complement-resistant red blood cells (so-called PNH I
chronic intravascular hemolytic anemia from uncontrolled cells), as well as mildly (PNH II) or markedly (PNH III)
complement activation, a propensity for thrombosis, and abnormal complement-sensitive cells. The difference in the
bone marrow failure. The hemolysis is largely mediated by proportion of these red blood cell populations contributes to
the alternative pathway of complement. These clinical mani- the variability in intravascular hemolysis seen in patients.
festations result from the lack of specific cell surface pro- PNH patients have a propensity for thrombosis. Several
teins, CD55 and CD59, on PNH cells resulting from a theories have been postulated to account for this hypercoag-
somatic mutation in the PIGA gene in HSCs, which results in ulability, but the mechanism has not been clearly defined. It
failure to synthesize the glycosylphosphatidylinositol (GPI) is believed that thrombophilia in PNH is related to the degree
anchor. of hemolysis and thereby indirectly related to the size of PNH
clone. Possible prothrombotic pathways include platelet
activation by complement components, procoagulable mic-
Pathophysiology
roparticles derived from GPI-deficient erythrocytes, or slow-
Hemolysis in PNH is complement mediated and is a direct ing of the microcirculation because of vasoconstriction
result of the mutated PNH cells acquiring a deficiency of induced by products of hemolysis. It also has been suggested
complement regulatory proteins. In PNH, because of the that intravascular hemolysis exposes red blood cell phospho-
defect of the enzyme encoded by the mutant PIGA gene lipids that may serve to initiate coagulation.
(Figure 17-4A), the first step in biosynthesis of the GPI anchor PNH is also a disorder of marrow failure. PNH clones
protein (AP) cannot be completed normally (Figure 17-4B), expand only in the context of immune-mediated bone mar-
and all GPI-anchored proteins are absent on the surface of row failure, explaining the close association between AA and
progeny cells of all hematopoietic lineages derived from the PNH. According to the most predominant hypothesis, PNH
affected stem cell with increased susceptibility to hemolysis stem cells, which can be found in very low frequencies in
(Figures 17-4C and D). healthy individuals, have a selective advantage in certain cir-
The intravascular hemolysis in PNH is due to the lack of cumstances of immune dysregulation. Under conditions of
GPI-anchored proteins (CD55 and CD59) that attenuate T-cell-mediated immune attack on HSCs, GPI-deficient stem
A C
Protein Protein
1 2 3 4 5 6
B Protein
Plasma membrane
PIGA
PIGC D
PIGH
PIGQ
PIGY Endoplasmic
DPM2 reticulum
Mutant PIGA Hemolyzing

stem cell PNH cells cells
Figure 17-4 Pathogenesis of PNH. (A) In hematopoietic stem cells, acquired somatic mutations of the PIGA gene may occur. This controls the
key step in the biosynthesis of GPI anchor proteins. (B) GPI anchor biosynthesis takes place in the endoplasmic reticulum. PIGA is 1 of 7
subunits involved in the first step of GPI anchor biosynthesis. (C) After multiple steps including protein attachment to the GPI anchor and fatty
acid remodeling, the GPI anchored protein should be transported to the plasma membrane. This cannot occur in PNH patients. (D) These
mutations (in panel A) can decrease the function or totally inactivate the enzyme encoded by PIGA. As a consequence, all proteins using this
type of anchor are deficient from the membrane of affected progeny derived from the mutant stem cells and cause the PNH phenotype and
hemolysis.

cells appear to thrive due to selective survival advantage com- iron losses may have microcytic red blood cell indices. Ele-
pared with healthy stem cells, which facilitate their expan- vated LDH and absent haptoglobin together with urine
sion. This close association between immune-mediated hemosiderin indicate the presence of intravascular hemoly-
depletion of normal stem and progenitor cells explains the sis. Patients with PNH who do not receive transfusions
coexistence of hematopoietic failure and frequent cytopenias develop various degrees of iron deficiency anemia over time.
related to impaired blood cell production (Figure 17-5). Various degrees of thrombocytopenia and neutropenia also
may be present in patients with PNH associated with AA. In
the absence of AA, the bone marrow shows relative expan-
Laboratory findings and diagnosis sion of erythroid series and most often is hypercellular.
The diagnosis of PNH is both a laboratory and a clinical diag-
nosis. The laboratory diagnosis of PNH formerly relied on Key points
the demonstration of abnormally complement-sensitive
• PNH is an acquired clonal HSC disorder characterized by
erythrocyte populations such as the Ham test or sucrose deficiency of GPI-linked proteins in blood and bone marrow cells
lysis test. These two tests are primarily of historical interest. due to a somatic mutation in the PIGA gene.
Currently, the diagnosis of PNH is secured by abnormal lab- • Patients with PNH experience chronic hemolytic anemia
oratory measures including a reticulocyte count, lactate (intravascular) from uncontrolled complement activation. They
dehydrogenase levels, complete blood count indicative of may also suffer with a propensity for thrombosis and bone
hemolysis, and peripheral blood flow cytometry to detect the marrow failure (indicated by leukopenia and/or thrombocytope-
deficiency of the GPI-AP. This absence of GPI-APs is detected nia in addition to anemia).
after staining cells with monoclonal antibodies (eg, CD55, • Flow cytometric techniques to identify cell populations lacking
CD59) and/or a reagent known as fluorescein-tagged pro- GPI-linked proteins, such as CD55 and CD59, confirm the
diagnosis of PNH and are used to estimate the size of PNH clone.
aerolysin variant (FLAER) that binds a portion of the GPI
anchor. The erythrocytes may be classified as type I, II, or III
PNH cells, as noted above. It should be noted that testing of
a PNH clone solely in erythrocytes is not adequate for evalu-
ation of PNH, since hemolysis and transfusions may greatly Chronic hemolytic anemia of various degrees is the most
underestimate the size of the clone. For these reasons, granu- common manifestation of PNH. Despite the name of the
locyte and monocyte clones are frequently detected when disease, hemoglobinuria with darker-stained urine at a par-
erythrocyte clones are not. Hematopathologists have recently ticular time of the day is reported by only a minority of
published guidelines for diagnosis of PNH using flow cytom- patients. Symptoms related to hemolysis include back and
etry. In patients with brisk hemolysis associated with PNH, abdominal pain; headache; smooth muscle dystonias, such
macrocytic anemia due to compensatory reticulocytosis typi- as esophageal spasm and erectile dysfunction (due to scav-
cally is present (if hematopoiesis is not suppressed), but some enging of nitric oxide by free plasma hemoglobin); and
PNH patients with iron deficiency due to chronic urinary severe fatigue often out or proportion to the degree of
Figure 17-5 Mechanisms of anemia in

Classical PNH Anemia Aplastic anemia/PNH PNH. Anemia in PNH can be a result of
in PNH
increased red blood cell (RBC) destruction
due to intravascular hemolysis of
Destruction due to underlying Production defect due to GPI-deficient RBCs, decreased production
hemolytic process underlying marrow failure of RBCs due to immune-mediated bone
marrow failure, or a combination of these
• Reticulocytopenia
• Leukopenia two mechanisms. Hemolysis can be
• Thrombocytopenia compensated for by increased production
(patients with increased reticulocytes), or
compensation may be inadequate (patients
with low reticulocyte counts).
Extravascular hemolysis Intravascular hemolysis

• Elevated LDH (if on eculizumab)
• Reticulocytisis • Direct Coombs + with C3
• Schistocytes • Spherocytes

Paroxysmal nocturnal hemoglobinuria | 501
anemia. Exacerbations of hemolysis can occur with infec- failure-related anemia. Chronic hemolysis should be treated
tions, surgery, or transfusions and manifest as acute worsen- with supportive measures, such as transfusions, supplemen-
ing of anemia. If severe, hemolysis can result in acute renal tation of folate and iron, and, in the context of renal failure,
failure because of pigment nephropathy. Icterus often is recombinant erythropoietin administration.
present intermittently and typically worsens during hemo- A humanized monoclonal antibody to the C5 terminal
lytic exacerbations. complement component, eculizumab, has shown efficacy in
The most concerning complication of PNH is thrombosis. decreasing intravascular hemolysis, decreasing the need for
It is the leading cause of death in the disease. Thrombosis transfusions, and improving the quality of life in patients
may occur at any site in PNH: venous or arterial. Common with PNH. Eculizumab effectively stops hemolysis and alle-
sites include intra-abdominal (hepatic, portal, splenic, or viates the need for transfusions in the majority of patients. It
mesenteric) and cerebral (cavernous or sagittal sinus) veins, is the only FDA-approved therapy for PNH. Treatment with
with hepatic vein thrombosis (also known as Budd-Chiari eculizumab is associated with few complications, but because
syndrome) being the most common. Deep venous thrombo- the terminal components of complement are important to
sis, pulmonary emboli, and dermal thrombosis are also prev- protect from Neisseria meningitidis, vaccination against this
alent. For unclear reasons, thrombotic complications are less microorganism is important before initiation of eculizumab
common in PNH patients of Asian descent. The thrombotic therapy (at least 2 weeks in advance). The decision about
propensity is particularly enhanced during pregnancy. Clini- when to start eculizumab needs to take into consideration
cally, the complication of thrombosis is more prevalent in the degree of chronic hemolysis, frequency of acute hemo-
patients as the PNH clone increases in size. Thrombosis may lytic attacks, severity of constitutional symptoms, throm-
occur in any PNH patient, but those with a large percentage botic history, and frequency of transfusions—parameters
of PNH cells (>50% granulocytes) are at greatest risk. Com- that should be balanced against the need for chronic lifelong
plement inhibition with eculizumab is the most effective biweekly infusions and the high cost of the drug.
means to stop thrombosis in PNH. If a diagnosis of a thrombosis is made in a PNH patient,
Patients with PNH suffer from anemia but may also have aggressive treatment is warranted. Anticoagulation and ecu-
other cytopenias depending on the degree of the associated lizumab are indicated for acute thrombotic events; however,
marrow failure. The marrow failure component of PNH can primary prophylactic anticoagulation has not been well
vary from subclinical disease to severe aplastic anemia (AA) established to be beneficial in PNH. Anticoagulation after
and may be categorized as an overlap syndrome of AA/PNH. the acute event in a PNH patient well maintained on eculi-
The disease presentations of PNH and AA do have consider- zumab may not be necessarily lifelong.
able overlap, as they may represent different spectrums of The majority of classical PNH patients will respond to
the same disorder. The PNH clone is often considered a eculizumab; however, the hemoglobin response is highly
marker of an immune form of marrow failure, as it may pre- variable and may depend on underlying bone marrow fail-
dict response to immunosuppressive therapy in AA; thera- ure, concurrent inflammatory conditions, genetic factors,
pies directed at PNH hemolysis will not improve the patients’ and the size of the PNH red cell clone following therapy.
component of underlying marrow failure. Patients do require close monitoring whilst on eculizumab
The classification of PNH has been proposed by the Inter- treatment. Unfortunately, not all patients have their disease-
national PNH Interest Group (IPIG) and includes 3 sub- specific needs met by eculizumab. Eculizumab does not
types: (i) classical PNH, which includes hemolytic and improve underlying bone marrow failure. There are also
thrombotic patients who have evidence of PNH in the reports of patients who have a coexistent autoimmune dis-
absence of another clinical bone marrow failure disorder; (ii) ease with ongoing activation of complement from their
PNH in the context of other primary bone marrow disor- underlying disease which lead to suboptimal responses from
ders, such as aplastic anemia or myelodysplastic syndrome; eculizumab. Transient breakthrough intravascular hemolysis
and (iii) subclinical PNH, in which patients have small PNH can be observed following viral or bacterial infections. Preg-
clones but no clinical or laboratory evidence of hemolysis or nancy can be another limitation on the efficacy of eculi-
thrombosis. zumab. Pregnancy is a hypercoagulable state itself, and there
have been concerns both about the potential for increased
maternal and fetal morbidity in a pregnant patient as well as
Treatment
the safety of eculizumab therapy in pregnancy. There are
The variability in the clinical manifestations of PNH makes multiple case reports and case series reporting successful
it necessary to individualize the treatment plan. Anemia is pregnancies in patients on eculizumab. However, what has
often the dominant issue to be addressed. Anemia resulting been observed is the tendency for breakthrough hemolysis at
from hemolysis should be distinguished from bone marrow later stages of pregnancy which requires reduced dosing

interval by the third trimester. Japanese patients can be asymptomatic, but others have symptomatic anemia and
another group of suboptimal responders to eculizumab. remain dependent on transfusions. Thus, there is a need for
They may carry a single missense C5 heterozygous mutation, complement inhibition that reduces C3 accumulation on
c.2654G→A, which prevents binding and blockade by eculi- PNH erythrocytes to address the shortcomings of eculi-
zumab while retaining the functional capacity to cause zumab in PNH.
hemolysis. The polymorphism accounts for the poor Life-threatening and fatal meningococcal infections have
response to eculizumab in patients carrying the mutation. occurred in patients treated with eculizumab due to the
Lastly, eculizumab only compensates for the CD59 defi- complement blockade and inability to fight encapsulated
ciency on PNH erythrocytes, but not the CD55 deficiency. pathogens. As these infections can be life threatening or fatal,
Thus, PNH patients on eculizumab accumulate C3 frag- the recommendation is for meningococcal vaccination at
ments on their CD55 deficient red cells leading to extravas- least 2 weeks prior to administering the first dose of eculi-
cular hemolysis through the accumulation of opsonins that zumab. In patients where the risks of delaying eculizumab
are recognized by the reticuloendothelial system (Figure therapy outweigh the risk of developing a meningococcal
17-6). Laboratory evidence of extravascular hemolysis in infection, a fluoroquinolone (ciprofloxacin) can be given as
eculizumab-containing patients includes increased reticulo- a bridge. Furthermore, any infection can increase comple-
cytes, persistent anemia, and often direct antiglobulin testing ment and increase hemolysis, even in patients well managed
that is positive for C3 deposition. These patients may remain with stable eculizumab dosing. Attention and increased vigi-
lance at the time of infection is imperative in these patients.
Instructions to notify providers for fevers, headaches, or
Classical pathway other symptoms should be provided to all patients so that
and lectin pathway
prompt medical attention is available.
C1q
The approach to severe bone marrow failure associated
Alternative with PNH should be similar to that taken for severe AA.
C1r/C1s Immunosuppressive therapy with horse ATG and cyclospo-
C3
rine can be effective in improving blood counts and may
Alternative
C4a pathway
allow for better compensation of hemolysis. Immunosup-
(tickover) pressive drugs, however, are mostly ineffective in patients
with purely hemolytic forms of PNH who have adequate
C2b C4a, 2b C3bB
marrow reserve.
Hematopoietic stem cell transplantation (HSCT) is the only
C3 convertase C3b, Bb curative therapy for PNH. However, it is not recommended as
upfront therapy in the eculizumab era given the risks of trans-
Eculizumab plant-related morbidity and mortality. HSCT is a reasonable
Extravascular hemolysis
due to C3 deposition therapeutic option in patients who do not respond to therapy
C5 convertase with eculizumab or those patients who have severe pancytope-
nia due to underlying bone marrow failure. The transplant
Membrane attack Intravascular paradigm pursued is often with reduced intensity condition-
complex hemolysis ing regimens, as myeloablation is not required to eradicate the
PNH clone. The use of HSCT may be revisited in the future as
Figure 17-6 The complement cascade, paroxysmal nocturnal patients and healthcare providers weigh the cost-benefit ratio
hemoglobinuria and eculizumab. PNH cells have a deficiency in of HSCT versus lifetime of eculizumab therapy.
GPI-anchored proteins on their cell surface. Absence of CD55 and
CD59 leads to uncontrolled complement activation on the surface of
PNH cells. Deficiency of CD59 increases MAC formation and induces Prognosis
intravascular hemolysis, which is central to the pathophysiology of
Thrombotic events, progression of the marrow failure com-
PNH. Deficiency of CD55 leads to increased C3 convertase activity
and C3d-associated extravascular hemolysis. Eculizumab therapy for
ponent, and age >55 years at diagnosis have been correlated
PNH is a humanized monoclonal antibody that targets C5. By with a poorer prognosis for PNH patients. The clonal evolu-
preventing C5 activation, eculizumab prevents the formation of the tion of PNH to MDS or acute leukemia markedly shortens
MAC leading to a significant reduction in intravascular hemolysis of survival. Patients diagnosed with classical PNH without leu-
PNH cells. Use of eculizumab can lead to increased extravascular kopenia, thrombocytopenia, or other complications main-
hemolysis in some patients. tained on therapy can anticipate long-term survival.

Myelodysplastic syndromes | 503
or the presence of certain AML-defining karyotypes such as

Key points t(15;17); thus, all patients with MDS have <20% marrow
• Eculizumab, a monoclonal antibody against the C5 terminal blasts, by definition.
complement component, effectively blocks hemolysis in patients MDS may arise de novo—80% to 85% of cases are idio-
with symptomatic PNH and alleviates the need for transfusions in pathic—or may be secondary to a recognized exposure to a
most cases. Eculizumab also appears to reduce thrombotic events. DNA-damaging agent. Secondary or therapy-related MDS
There are limitations to this treatment in some patients including (t-MDS) can be induced by drugs that alkylate DNA bases
breakthrough hemolysis and risk of meningococcal infections. (eg, chlorambucil, cyclophosphamide, melphalan), inhibi-
• Prompt evaluation of PNH patients is indicated when
tors of topoisomerase II (eg, topotecan, etoposide, anthracy-
symptoms are suggestive of thrombosis because the risk of
clines), therapeutic or accidental exposure to ionizing
clotting is high.
radiation, or environmental or occupational exposure to
• Treatment of bone marrow aplasia with immunosuppressive
therapy will not eliminate the PNH clone and is generally
other DNA toxins, such as hydrocarbons. Proving a connec-
ineffective in primary hemolytic PNH. Immunosuppression, tion between a suspect exposure and subsequent develop-
however, may be helpful in patients with AA/PNH syndrome. ment of MDS can be difficult, but the presence of a complex
• Allogeneic HSCT has curative potential but is indicated only in karyotype (defined as ≥3 acquired chromosome abnormali-
patients with severe cytopenias and severe thrombotic complica- ties), abnormalities of chromosomes 5 and 7, or somatic
tions refractory to medical therapy. TP53 mutation suggests the possibility of t-MDS.
Among the potential peripheral blood cytopenias, anemia
(often macrocytic) is the most commonly observed cytopenia
Myelodysplastic syndromes in MDS, present in >90% of cases at diagnosis. So-called dys-
plastic cell morphology (Figure 17-7; discussed in “Diagnostic
Clinical case Evaluation” later in this chapter) is diagnostically important,
reflects failure of cells to differentiate and mature normally,
A 79-year-old retired man with past medical history of type 2
and often is accompanied by cellular dysfunction that exacer-
diabetes mellitus and hypertension develops fatigue and exertional
bates the signs or symptoms of cytopenias. For example,
dyspnea. Physical examination demonstrates generalized pal-
lor; hepatosplenomegaly and lymphadenopathy are absent. A
hypogranular neutrophils with impaired bactericidal activity
complete blood count reveals a hemoglobin level of 8.3 g/dL, mean compound the infection risk associated with neutropenia,
cell volume (MCV) of 102 fL, white blood cell count of 2.9 × 109/L whereas platelets that lack intracellular granules or express
with 33% neutrophils (absolute neutrophil count 0.96 × 109/L) abnormally low levels of procoagulant cell surface markers
and no circulating blasts, and a platelet count of 68 × 109/L. may be ineffective in achieving hemostasis, even when large
Serum vitamin B12 and red cell folate levels are normal, and a numbers of these “dud” cells are present. As a result, the infec-
chemistry group is unremarkable; the ferritin level is 348 ng/mL tion and bleeding risks in MDS correlate poorly with the cir-
and reticulocyte proportion is 0.6%. Peripheral smear shows oval culating neutrophil and platelet count, and some MDS
macrocytes and hypogranular and hypolobated neutrophils. The patients with severe cytopenias are less symptomatic than
patient undergoes marrow aspiration and biopsy, which reveals
other patients who have more modest cytopenias. The bone
a hypercellular marrow for age (80% cellularity) with erythroid
marrow in MDS usually is normocellular or hypercellular for
hyperplasia and megaloblastoid erythroid maturation, reduced
age, but 10%-20% of cases are accompanied by a hypocellular
granulocyte progenitors with dysplastic maturation, and scattered
abnormally hypolobated megakaryocytes. Ring sideroblasts are
marrow, and such cases of “hypoplastic MDS” or “hypocel-
absent, and reticulin is graded at 1+. There are 8% blasts in the lular MDS” may be difficult to distinguish from AA.
bone marrow that express CD34+ and various myeloid markers; a
subset of these cells aberrantly co-express CD7. The karyotype is
Classification
47, XY, +8 [12], 46, XY, del(20)(q11q13) [8]. Mutational profiling
demonstrates clonal mutations in DNMT3A, ASXL1, and TET2. The current prevailing classification of MDS is the fourth
edition of the WHO Classification of Tumors of Hematopoi-
etic and Lymphoid Tissues, published in 2008 (Table 17-4).
Introduction
The 2008 WHO MDS classification is a minor modifica-
Myelodysplastic syndromes (MDS) include a heterogeneous tion of the third edition WHO classification, formally pub-
group of clonal, acquired disorders characterized by ineffec- lished in 2001, which in turn was built on the 1982 MDS
tive hematopoiesis, resulting in peripheral blood cytopenias. classification of the French-American-British (FAB)
MDS carry a variable risk of progression to acute myeloid Cooperative Group, the first formal MDS classification. A
leukemia (AML). AML is defined by the World Health Orga- revised WHO classification is planned for publication in
nization (WHO) as ≥20% blast cells in the marrow or blood, either 2016 or 2017.

Table 17-4 2008 World Health Organization (WHO) classification of myelodysplastic syndromes and neoplasms
WHO-estimated
Peripheral blood: patients with
Name Abbreviation key features Bone marrow: key features MDS (%)
Refractory cytopenias with
unilineage dysplasia (RCUD):
Refractory anemia RA Anemia; <1% peripheral Unilineage erythroid dysplasia (in 10%-20%
blood blasts >10% of cells); <5% blasts
Refractory neutropenia RN Neutropenia; <1% Unilineage granulocytic dysplasia; <1%
peripheral blood blasts <5% blasts
Refractory thrombocytopenia RT Thrombocytopenia; <1% Unilineage megakaryocytic dysplasia; <1%
blasts <5% blasts
Refractory anemia with ring RARS Anemia; no blasts Unilineage erythroid dysplasia; ≥15% 3%-11%
sideroblasts of erythroid precursors are ring
sideroblasts; 5% blasts
Refractory cytopenias with RCMD Cytopenia(s); <1% blasts; Multilineage dysplasia ± ring 30%
multilineage dysplasia no Auer rods sideroblasts; <5% blasts; no Auer
rods
Refractory anemia with excess RAEB-1 Cytopenia(s); <5% blasts; Unilineage or multilineage dysplasia; 40%*
blasts, type 1 no Auer rods 5%-9% blasts; no Auer rods
Refractory anemia with excess RAEB-2 Cytopenia(s); 5%-19% Unilineage or multilineage dysplasia;
blasts, type 2 blasts; ± Auer rods 10%-19% blasts; ±Auer rods
Myelodysplastic syndrome Del(5q) Anemia; normal or high Isolated 5q31 chromosome <5%
(MDS) associated with platelet count; deletion; anemia, hypolobated
isolated del(5q) ≤1% blasts megakaryocytes
Childhood MDS, including RCC Pancytopenia <5% marrow blasts for RCC; marrow ~1%
refractory cytopenia of usually hypocellular
childhood (provisional)
MDS, unclassifiable MDS-U Cytopenias; ≤1% blasts Does not fit other categories; dysplasia; ?
<5% blasts; if no dysplasia, MDS-
associated karyotype
If peripheral blood blasts are 2%-4%, the diagnosis is RAEB-1 even if marrow blasts are <5%. If Auer rods are present, the WHO considers the
diagnosis RAEB-2 if the blast proportion is <20% (even if <10%) or acute myeloid leukemia (AML) if ≥20% blasts. Cases of RCUD, RARS, or
RCMD where the peripheral blood blasts are exactly 1% should be considered MDS-U, according to the WHO. For all subtypes, peripheral
blood monocytes must be <1 × 109/L. Bicytopenia may be observed in RCUD subtypes, but pancytopenia with unilineage marrow dysplasia
should be classified as MDS-U. Therapy-related MDS (t-MDS), whether due to alkylating agents, topoisomerase II inhibitors, or radiation,
is classified together with therapy-related AML (t-MDS/t-AML) in the WHO classification of AML and precursor lesions. The listing in this
table excludes MDS/MPN overlap categories, such as chronic myelomonocytic leukemia, juvenile myelomonocytic leukemia, and RARS with
thrombocytosis.
From Swerdlow SH, Campo E, Harris NL, et al., eds. WHO Classification of Tumours of Haematopoietic and Lymphoid Tissues. 4th ed. Lyon,
France: IARC Press; 2008:87-107.
*This 40% figure represents the proportion of patients with RAEB-1 or RAEB-2, collectively.
Important classification factors in the current WHO MDS blast count. Cases with both MDS and myeloproliferative fea-
schema include the specific myeloid cell lineages in which tures, such as leukocytosis or thrombocytosis, are classified in
>10% of cells are dysplastic, the marrow and peripheral blood a separate “overlap” category of MDS/myeloproliferative neo-
blast proportion; whether or not ≥15% of erythroid precursor plasms (MPNs), which includes chronic myelomonocytic leu-
cells in the marrow are ring sideroblasts (an arbitrary thresh- kemia (CMML, defined by >1 × 109/L blood monocytes) and
old); whether or not Auer rods are present; and, to a limited refractory anemia with ring sideroblasts and thrombocytosis
extent, the presence of cytogenetic abnormalities, such as (RARS-T, which requires a platelet count >450 × 109/L). While
interstitial deletion of the long arm of chromosome 5. The the WHO classification is useful diagnostically, it has only lim-
WHO has grouped t-MDS with therapy-related AML (t-AML) ited prognostic value, and other tools (described below) are
because the outcome in such patients is poor regardless of the more useful for risk stratification.

The observation that alkylating agents, topoisomerase may also have a biological basis related to a protective effect
inhibitors, and ionizing radiation predispose patients to of having two X chromosomes. However, one specific MDS
both MDS and AML; evolution of MDS to AML in some subtype, MDS associated with isolated deletion of the long
patients over time; the existence of shared cytogenetic abnor- arm of chromosome 5 and a marrow morphology that
malities, such as deletions or gains in all or parts of chromo- includes hypolobated megakaryocytes and erythroid hypo-
somes 5, 7, 8, or 20; and shared common somatic mutations, plasia (ie, 5q– syndrome), is more common in women than
such as TET2 and ASXL1, imply a biologic continuum in men.
between MDS and AML. Whereas loss or gain of chromo- Accurate estimates of the incidence of MDS have been dif-
somal material is common in MDS, chromosomal transloca- ficult to obtain since MDS cases have not historically been
tions are less common in MDS than in AML, and certain captured by cancer registries and many elderly patients with
point mutations (eg, FLT3) common in AML are rarely seen mild cytopenias are incompletely evaluated. However, cur-
in MDS. The so-called “good-risk” recurrent AML-associated rent registry and claims-based algorithms suggest there are
translocations, t(8;21), t(15;17), and inv(16), are rare in 30,000-40,000 new cases of MDS diagnosed per year in the
patients with dysplasia, and the WHO classifies patients with United States. Most patients have lower-risk disease at the
these abnormalities as having AML regardless of the blast time of initial diagnosis.
count or marrow dysplasia. MDS are rare in the pediatric age group and represent
The natural history of MDS includes a risk of progres- ~5% of hematologic malignancies in patients <18 years of
sion to treatment-refractory AML (~25%-30% likelihood age. When MDS do arise in children, they are frequently
overall, with some subtypes of MDS such as RAEB-2 at associated with Down syndrome, congenital marrow failure
much greater risk), but most patients with MDS do not syndromes, or germ-line defects of DNA repair, such as Li-
develop AML. Instead, the majority of patients who are Fraumeni syndrome or Bloom syndrome. Children with
diagnosed with MDS will die from complications of cyto- Shwachman-Diamond syndrome, congenital neutropenia,
penias, most commonly infections resulting from absolute or Fanconi anemia are at markedly increased risk of develop-
neutropenia and neutrophil dysfunction, and less frequently ing MDS. In all of these inherited conditions, MDS arise in
thrombocytopenia-associated bleeding or anemia-exacer- the context of hematopoietic deficits and typically present in
bated cardiovascular events. Because MDS are primarily dis- late childhood or in adolescence. Children who develop
eases of older persons, some patients succumb to unrelated MDS without excess blasts but who appear to lack a predis-
conditions that are common in the elderly; they die with posing congenital syndrome are provisionally classified by
MDS, rather than from MDS. the WHO as having refractory cytopenia of childhood
(MDS-RCC).
Refractory anemia with excess blasts is also relatively com-
Epidemiology
mon in children, and the bone marrow is often hypocellular
Aging is the most important risk factor for development of rather than the hypercellular marrow characteristic of adults;
MDS, in part because of the progressive accumulation of there is also a high incidence of unfavorable biologic fea-
somatic mutations of HSCs across the human life span. Even- tures, such as monosomy 7. Refractory anemia with ring sid-
tually, a mutation or combination of mutations can occur in eroblasts (RARS) and 5q– syndrome are rare in children,
such a way in a hematopoietic cell that its progeny acquire a although a number of forms of congenital sideroblastic ane-
growth and survival advantage and clonal hematopoiesis mia can be confused with MDS, such as sideroblastic anemia
emerges. The expanded clone of cells is then at risk for acquir- due to germ line mutations of ALAS2, which does not carry
ing additional mutations that increase its malignant poten- a risk of progression to AML.
tial. Somatic mutations associated with clonal hematopoiesis Familial MDS or AML with monosomy 7 has been
such as DNMT3A, TET2, or ASXL1 can be detected in >10% reported in at least 10 families in the absence of either pheno-
of individuals over age 70 years with normal blood counts, typic abnormalities or any history of hematologic disorders.
and patients with clonal hematopoiesis have an increased risk Germ line mutations in RUNX1 and GATA2 transcription
of subsequent diagnosis of MDS and other hematological factors also predispose to MDS. Germ line RUNX1 muta-
neoplasms, as well as increased all-cause mortality. tions are associated with a prodrome of thrombocytopenia.
The median age at diagnosis of MDS in the United States GATA2 mutations are sometimes nonsyndromic but can be
and Europe is ~70 years. In China and Eastern Europe, the associated with mycobacterial infections, lymphedema, and
median age at diagnosis is more than a decade younger than monocytopenia (MonoMAC syndrome).
in the West, possibly due to environmental factors. Overall, In most adult patients with MDS, the etiology is unknown,
there is a slight male predominance in MDS that may be and there is no specific predisposing factor identifiable other
related in part to occupational exposures, but this imbalance than advanced age.

Pelger-Huët anomaly, and are referred to as Pelgeroid or

Key points pseudo–Pelger-Huët cells. Peripheral blood smears may be
• MDS are characterized by ineffective hematopoiesis, leading highly suggestive of the diagnosis, but are never conclusive by
to peripheral blood cytopenias. The marrow is often hypercel- themselves. A marrow aspirate is essential to establish defini-
lular for age. tively a diagnosis of MDS, and the bone marrow core biopsy
• Macrocytic anemia is the most common cytopenia associated provides complementary information on cellularity and
with MDS. Functional defects in neutrophils and platelets can architecture, megakaryocyte morphology, and the presence
exacerbate the risk of infection from neutropenia or bleeding of fibrosis—useful information that may inform therapeutic
from thrombocytopenia.
decisions.
• Aging and exposure to alkylating agents, topoisomerase II
The bone marrow biopsy in MDS usually demonstrates
inhibitors, or ionizing radiation are all risk factors for develop-
hypercellularity, which, in the setting of cytopenias in the
ing MDS.
• MDS are rare in children, and when they occur are often
peripheral blood, indicates ineffective hematopoiesis. On the
associated with congenital marrow failure syndromes. marrow aspirate, megaloblastoid red blood cell precursors
• Germ line mutations in RUNX1 and GATA2 are associated with a with asynchronous maturation of the nucleus and the cyto-
subsequent risk for MDS development. RUNX1 mutations are also plasm are usually evident, and multinucleated erythroid pre-
associated with thrombocytopenia and are often mistaken for ITP cursors are common (Figure 17-7). Ring sideroblasts, which
until MDS develops or additional family members are diagnosed. are erythroid precursors with iron-stuffed mitochondria
• The 2008 WHO classification of MDS is the current standard, (stored as mitochondrial ferritin, a unique type of ferritin)
but it should be used in conjunction with risk stratification tools surrounding at least one-third of the nucleus, may be identi-
to assess prognosis. fied via the Prussian blue reaction, and often there is pre-
dominance of immature myeloid cells and dysplastic
Diagnostic evaluation granulocytic precursors. Dysplastic megakaryocytes may be
smaller or larger than normal and may be hypolobated or
After a medical history and physical examination, the diagno- hyperlobated. Dysplastic features in all lineages can include
sis of MDS is readily established in most patients by a complete nuclear and cytoplasmic blebs and misshapen nuclei.
blood count, careful review of the blood smear, bone marrow Cytogenetic studies can further support a diagnosis of
examination, and basic laboratory tests to rule out other disor- MDS, and specific aberrations correlate with prognosis and
ders that mimic MDS. Vitamin B12 and folate deficiency, HIV response to treatment (see Tables 17-5, 17-6 and 17-7). In a
infection, copper deficiency, alcohol abuse, and adverse effects small percentage of cases, fluorescence in situ hybridization
of medication (eg, antimetabolites such as methotrexate or (FISH) analysis with probes directed towards chromosomes
azathioprine) need to be excluded. The diagnosis of MDS is frequently rearranged in MDS (eg, 5, 7, 8, 20) reveals specific
currently based primarily on morphologic criteria demon- chromosomal translocations and losses or gains of DNA seg-
strating dysplastic features in the peripheral blood and >10% ments that were not detected with standard cytogenetic
of bone marrow precursor cells in one or more lineages— methods. The clinical relevance of small clones detectable
erythroid, myeloid, megakaryocytic (Figure 17-7). only by FISH is uncertain. The yield of FISH is very low if
In one large study, the median hemoglobin of patients karyotyping is successful.
diagnosed with MDS was 9.5 g/dL, and 75% of patients had Flow cytometric analysis of the bone marrow, which is
a level <11 g/dL. Only 20% of patients had both a platelet now a standard procedure for diagnosing and subclassifying
count >100 × 109/L and an absolute neutrophil count >1.0 × 109/L patients with acute leukemia, is being used increasingly to
at diagnosis, indicating that a presentation with anemia evaluate patients suspected of having MDS. A number of
alone in MDS is relatively uncommon. Although patients investigative groups have described abnormal cell popula-
with MDS often seek medical attention because of symp- tions and inappropriate antigen expression detected by flow
toms related to cytopenias, especially fatigue or poor exercise cytometry, and these investigators continue to study the
tolerance, many patients are asymptomatic at diagnosis and diagnostic specificity and prognostic importance of specific
are discovered to have MDS only when a complete blood flow findings. Because accurate classification according to
count is performed as a screening test or to evaluate another WHO criteria is based, at least in part, on bone marrow mor-
condition. phology, flow cytometry should be viewed as a complemen-
Oval macrocytic red blood cells, hypogranular and hypol- tary test that is best interpreted in the context of the
obulated granulocytes, and giant or hypogranular platelets appearance of the marrow morphology. Specifically, flow
can be identified in the peripheral blood of many patients with cytometric enumeration of marrow blasts should not replace
MDS. Bilobated hyposegmented neutrophils in MDS resem- a manual differential from the marrow aspirate, since it is
ble those seen in the clinically inconsequential congenital subject to technical artifacts.

A B
C D
E F
G H
Figure 17-7 Typical blood and marrow cell morphology in patients with MDS. (A and B) Multinucleated erythroid precursors (arrows); the cells
in panel A have a visible cytoplasmic bridge, which is uncommonly observed. Wright-Giemsa stained marrow aspirate. Source: ASH Image Bank
(imagebank.hematology.org), #00030315. (C) Megaloblastoid erythroid cell maturation (nuclear-cytoplasmic dys-synchrony). The chromatin
pattern of these cells is fine, suggesting relative immaturity, whereas the lightening of the cytoplasm indicative of early hemoglobinization is an
event typically associated with later stages of maturation. Source: ASH Image Bank #00002571. (D) Hypolobated neutrophil (pseudo–Pelger-
Huët cell) found in the peripheral blood of a patient with refractory cytopenias with multilineage dysplasia. The cell vaguely resembles a
pince-nez, a style of eyeglasses popular in the 19th century supported without earpieces. Source: ASH Image Bank #00002117. (E and F)
Hypogranular neutrophils (arrows). These would be expected to have poor bactericidal activity. The double arrow in panel F indicates a small,
dysplastic megakaryocyte. Source: ASH Image Bank #00001435. (G and H) Micromegakaryocytes in a Wright-Giemsa stained aspirate (G) and
hematoxylin-eosin-stained core trephine biopsy specimen (H). These may have an eccentric, hypolobulated or round nucleus. These images are
from a patient with 5q- syndrome. Source: ASH Image Bank #00001446 (H) and #00001448 (G). (I) Ring sideroblasts (a Prussian blue reaction
on a marrow aspirate, seen at low power magnification and counterstained with neutral red). Source: ASH Image Bank #00001157.

Table 17-5 The 1997 Revised International Prognostic Scoring System (IPSS) for myelodysplastic syndromes
Category score (sum all three subscores for overall IPSS score)
Prognostic factor 0 (best) 0.5 1 1.5 2.0 (worst)
Marrow blasts (%) <5 5-10 – 11-20 21-30*
Karyotype Good: normal, isolated Intermediate: all karyotypes Poor: abnormal chromosome – –
-Y, isolated del(5q), or not defined as good or poor 7 or a complex karyotype
isolated del(20q) (≥3 anomalies)
Peripheral blood 0 or 1 2 or 3 – – –
cytopenias†
Scoring system: A point value from 0 to 2.0 is determined for each of the three prognostic factors in Table 17-5, and the three values are
summed to obtain the total IPSS score (see Table 17-6).
While replaced by the IPSS-R in 2012 (see below), numerous clinical trial protocols still use the original IPSS for determination of eligibility.
From Greenberg P et al. Blood. 1997;89:2079-2088.
*No longer considered myelodysplastic syndrome (redefined as acute myeloid leukemia by World Health Organization in 2001).
†IPSS definition of peripheral blood cytopenias: hemoglobin <10 g/dL; absolute neutrophil count <1.8 × 109/L; and platelet count <100 × 109/L
Increasingly, molecular profiling is playing an important patients with ICUS are, by definition, not known to have a
role in evaluation of patients suspected of having MDS, espe- clonal disorder.
cially in ambiguous cases with bland morphology but no A related situation occurs when patients have cytopenias
other explanation for cytopenias. Almost all patients with and a nondiagnostic bone marrow, yet a clonal cytogenetic
MDS have a somatic mutation detectable in one of the 25-40 abnormality typical for MDS (eg, del[5q]) is present. The
most commonly mutated MDS-associated genes, so the neg- clinical behavior of such patients is typical for MDS, with a
ative predictive value of a normal result on an MDS muta- risk of death from cytopenias and progression to AML. In
tion panel is high, and another cause for cytopenias should the 2008 version of the WHO classification of MDS, such
be carefully sought in such cases. However, because clonal patients with unremarkable morphology but an abnormal
hematopoiesis is common in healthy older people, detection karyotype were considered as having MDS, unclassifiable
of a mutation in patients with a normal karyotype and with- subtype. Certain karyotypes (eg, trisomy 8, del[20q], and
out morphological changes of dysplasia should be inter- loss of the Y chromosome) are not specific enough to define
preted with caution. Molecular profiling can also aid in a case as MDS in the absence of dysplastic morphology.
prognostic assessment and decisions about stem cell trans- Finally, some patients with normal blood counts who
plant. Detection of an SF3B1 mutation, for instance, would undergo molecular genetic analysis of blood and marrow—
support a diagnosis of RARS rather than a congenital sidero- for instance, during marrow aspiration performed as part of
blastic anemia or reactive cause of sideroblastic anemia, a staging evaluation for non-Hodgkin lymphoma or whole
while the finding TP53 mutation makes stem cell transplant exome sequencing to evaluate a non-hematological condi-
less likely to be successful. tion—will be found to have an acquired mutation in an
Some patients present with cytopenias and nondiagnostic MDS-associated gene such as DNMT3A, TET2, or ASXL1.
bone marrow findings (ie, minimal or no dysplastic changes This state has been termed clonal hematopoiesis of indetermi-
and no increase in marrow myeloblasts) and without any nate potential (CHIP), and may represent a myeloid progen-
cytogenetic abnormalities—a situation in which MDS is itor or stem cell equivalent of MGUS or monoclonal B-cell
possible and an alternative diagnosis is not apparent, yet lymphocytosis. As described above, such mutations are
WHO-defined diagnostic criteria for MDS are not met. present in >10% of the population >70, and patients with
Such patients have been termed as having idiopathic CHIP have approximately 1% per year risk of developing a
cytopenia(s) of undetermined significance (ICUS). Close hematological malignancy as well as an increased all cause
observation of patients with ICUS is recommended. Some mortality.
patients with ICUS will develop overt MDS or AML over Overall, the diagnosis of MDS is evolving toward the
time, whereas others eventually will be found to have an approach used in AML, in which morphologic, cytogenetic,
alternative diagnosis; many will be stable for a prolonged and flow cytometric data are assessed together to make an
period, with or without complications from the cytopenias. accurate diagnosis and determine the optimal treatment.
Unlike other undetermined-significance hematologic con- This strategy will become increasingly important as biologi-
ditions (eg, monoclonal gammopathy of undetermined sig- cally distinct subsets of MDS patients who respond to spe-
nificance [MGUS], monoclonal B-cell lymphocytosis), cific therapies are defined.

valid only for patients with de novo disease treated with sup-
Key points portive care and is not useful during the course of the disease
• Complete blood counts, blood morphology, marrow aspirate or in previously treated patients; within each IPSS risk
and core biopsy, and cytogenetic testing are key to establishing group, there are wide variations in patient outcomes. Despite
a diagnosis of MDS. these shortcomings, the IPSS has greater prognostic value
• Flow cytometry may provide complementary information but than the 2008 WHO classification system for individual
cannot be used to establish a diagnosis of MDS in the absence of patients, and it remains widely used for clinical trial enroll-
marrow morphology. Blast counts should be based primarily on ment purposes.
a manual assessment of the marrow aspirate by an experienced
Several newer MDS prognostic systems have been intro-
morphologist.
duced since 2007 to try to overcome limitations of the IPSS.
• Vitamin B12 and folate deficiency, HIV infection, copper
These newer risk stratification models include the WHO-
deficiency, alcohol abuse, and medication effects (eg,
antimetabolites such as methotrexate) can cause cytopenias
based Prognostic Scoring System (WPSS), which integrates
and dysplastic changes in blood cells and need to be the WHO classification with karyotyping data and the
excluded. degree of anemia; a modified form of the WPSS includes the
• Molecular abnormalities are present in most cases of MDS, presence or absence of marrow fibrosis. A general risk
and molecular testing can be used as a supplemental diagnostic model proposed by investigators at the M.D. Anderson
tool and as an aid in prognostic assessment. Cancer Center in Houston, Texas, is valid across a broad
spectrum of MDS patients, including those with exposure-
related MDS and those who previously have been treated
(eg, with a hypomethylating agent). A risk model specific to
Prognosis
lower-risk MDS also was developed at the M.D. Anderson
In 1997, the International Prognostic Scoring System (IPSS) Cancer Center and has been independently validated by
(Tables 17-5 and 17-6) was developed to help stratify patients other groups.
with MDS by their risk of disease progression to acute leuke- In 2012, a revised version of the IPSS (IPSS-R) was pub-
mia and death. The overall IPSS score is based on the sum of lished, based on analysis of >7,000 patients from more than
three subscores—scores for the karyotype, percentage of 10 countries (Tables 17-7, 17-8, and 17-9). The primary
bone marrow blasts, and number of qualifying cytopenias. changes in the IPSS-R are that it includes a broader range of
Patients >60 years of age with a low IPSS score have a median cytogenetic abnormalities than the small list of MDS-associ-
survival of 4.8 years, whereas patients in this age group with ated karyotypes that were included in the 1997 IPSS version,
a high IPSS score have a median survival of only <6 months, and the IPSS-R also weighs cytogenetic findings more heavily
if treated with supportive care alone. For each IPSS risk than other variables. In addition, degree of cytopenias is
group, outcomes tend to be better for younger patients than given more weight in the IPSS-R than in IPSS, and blast cut-
for older patients. offs are different. Like the original IPSS, however, the IPSS-R
A major limitation of the 1997 IPSS is that it does not is only valid in patients with de novo MDS and only at the
distinguish between patients with severe and modest degrees time of diagnosis. In addition, other prognostically impor-
of cytopenias, which may influence outcome. For example, a tant variables, such as the presence of comorbid conditions
platelet count of 9 × 109/L is not weighted any differently by and the patient’s performance score, molecular genetic find-
the IPSS than a count of 90 × 109/L, although several studies ings, and the kinetics of clonal evolution and disease progres-
have shown that severe thrombocytopenia is an important sion, are not accounted for by the IPSS-R or any of the other
risk factor for disease progression and death. The IPSS is major prognostic tools.
Table 17-6 Risk stratification of International Prognostic Scoring System (IPSS)
Median Median survival Median survival (years) Time until 25% of surviving
Total survival (years) for patients for patients ≥60 years old patients in category developed
Risk category score (years) <60 years old (n = 205) (n = 611) leukemia (years)
Low risk 0 5.7 11.8 4.8 9.4
Intermediate-1 (INT-1) 0.5 or 1.0 3.5 5.2 2.7 3.3
Intermediate-2 (INT-2) 1.5 or 2.0 1.2 1.8 1.1 1.1
High ≥2.5 0.4 0.3 0.5 0.2
From Greenberg P et al. Blood. 1997;89:2079-2088.

Table 17-7 MDS cytogenetic risk stratification system used in the revised International Prognostic Scoring System (IPSS-R)
Updated cytogenetic classification for use in IPSS-R (n = 7012)
Median 25% of patients Proportion of patients

Risk group Included karyotypes survival, years to AML, years in this group
Very good del(11q), -Y 5.4 N/R 4%
Good Normal, del(20q), del(5q) alone or with 1 other anomaly, 4.8 9.4 72%
del(12p)
Intermediate +8, del(7q), i17q, +19, any other single or double 2.7 2.5 13%
abnormality not listed, 2 or more independent clones
Poor Abnormal 3q, -7, double abnormality include -7/del(7q), 1.5 1.7 4%
complex with 3 abnormalities
Very poor Complex with >3 abnormalities 0.7 0.7 7%
N/R = not reached.
Adapted from Greenberg PL et al. Blood. 2012;120:2454-2465.
of ribosomal components contribute to DBA (see section

Key points
on DBA in Chapter 15). As originally described, the 5q–
• The IPSS is the most widely used risk stratification system in syndrome is associated with erythropoietin-refractory
MDS, but was revised in 2012 (IPSS-R) to include a broader macrocytic anemia, normal or increased platelet count,
range of karyotypes and other modifications. giant platelets, dyserythropoiesis, hypolobated megakaryo-
• Factors associated with poorer outcomes in MDS include cytes, variable neutropenia, female predominance, pro-
advanced age, comorbid conditions and poor performance longed survival, and a low rate of leukemic transformation.
score, increased marrow and blood blasts, more severe
It is important to differentiate the 5q– syndrome from
cytopenias and transfusion dependence, higher-risk karyotypes
other myeloid disorders in which chromosome 5q deletions
(eg, a complex karyotype or monosomy 7), and the presence of
are found. Patients with the del(5q) without the character-
certain mutations (eg, TP53 or RUNX1).
istic clinical and morphologic features of 5q– syndrome
may have a more aggressive clinical course and shorter sur-
Biology vival than those with the classic syndrome, although they
still may respond to lenalidomide treatment. The extent of
Chromosome and molecular biology
the chromosome 5q deletion in MDS also has prognostic
Approximately one-half of patients with de novo MDS and value, with small interstitial deletions associated with bet-
most patients with t-MDS have cytogenetic abnormalities ter outcomes than larger deletions.
detectable on routine G-banded metaphase karyotyping. The clinical and genetic heterogeneity found in MDS and
Cytogenetic results have independent prognostic signifi- the typical advanced age at disease onset support the idea that
cance (Table 17-7). New clonal cytogenetic aberrations multiple cooperating genetic lesions contribute to leukemo-
emerge in >25% of patients with MDS during the course of genesis. Unlike AML and MPNs, which frequently demon-
their disease, which suggests genomic instability of some strate chromosomal translocations, gains and losses of entire
form, although microsatellite instability is not common. In chromosomes (eg, monosomy 5 and 7 or trisomy 8) or of
patients with MDS who have a normal karyotype, more sen- large DNA segments (eg, many megabase pairs of chromo-
sitive analytical techniques, such as single-nucleotide poly- somes 5q, 7q, or 20q) are more common in MDS, which has
morphism arrays and array-based comparative genomic made pinpointing individual genes that contribute to the
hybridization, frequently detect areas of loss of heterozygos- development or progression of MDS via a candidate-gene
ity and uniparental disomy, which often are clonally restricted approach a formidable challenge. In recent years, high-
(ie, not present in germ line tissue). throughput resequencing techniques revealed recurrent
One particular clonal abnormality involving interstitial point mutations in more than 30 different genes, some of
or terminal deletion of part of the long arm of chromo- which are shared with AML and other neoplasms (Figure
some 5 (5q–) has received a great deal of attention in recent 17-8). These techniques also demonstrate that the majority of
years because patients with deletions of chromosome 5q cells in the marrow are clonal even in lower-risk MDS with
preferentially respond to lenalidomide therapy (see section <5% blasts.
“Treatment of MDS” later in this chapter). Haploinsuffi- Activating mutations in proto-oncogenes such as NRAS,
ciency of a 5q-encoded ribosomal protein, RPS14, contrib- FLT3, and JAK2 are detected in many cases of AML or MPN
utes to defective erythropoiesis, just as germ line mutations but are uncommon in MDS. Although Ras/Raf pathway

Table 17-8 Revised International

IPSS-R
Prognostic Scoring System for MDS
(IPSS-R) (2012 version) Parameter Categories and associated scores
Cytogenetic risk group Very good Good Intermediate Poor Very poor
0 1 2 3 4
Marrow blast proportion <2% 2%-<5% 5%-10% >10%
0 1 2 3
Hemoglobin ≥10 g/dL 8-<10 g/dL <8 g/dL
0 1 1.5
Absolute neutrophil ≥0.8 × 109/L <0.8 × 109/L
count 0 0.5
Platelet count ≥100 × 109/L 50-100 × 109/L <50 × 109/L
0 0.5 1
Possible range of summed scores: 0-10.
mutations are common in the MDS-MPN overlap syn- recurrent mutations are also found in EZH2, IDH1 and
dromes of CMML and juvenile myelomonocytic leukemia, IDH2, and ASXL1. Another class of recurrent mutations in
these mutations are rare in MDS without MPN features and MDS are those in genes that encode components of the spli-
usually are found only after progression to acute leukemia. ceosome and alter RNA splicing, especially SF3B1, which is
These data suggest that aberrant activation of signal trans- present in the majority of patients with RARS. Other com-
duction pathways may not be a major mechanism of aber- mon mutations in spliceosome components include SRSF2
rant cell growth and clonal dominance in early MDS, and U2AF1. Mutations in genes such as STAG2 or RAD21
which distinguishes these diseases from other myeloid that encode components of the cohesin protein complex,
malignancies. which regulates the separation of sister chromatids during
The TP53 tumor suppressor gene, which regulates cell- cell division, are found in up to 20% of MDS.
cycle progression, DNA repair, and apoptosis, is mutated in More than 80% of patients with MDS have at least one
5%-10% of MDS cases overall and in a higher proportion of somatic mutation detectable in hematopoietic cells. Several
t-MDS. TP53 mutation is often associated with a complex of these mutations have IPSS-independent prognostic sig-
karyotype and has a strong negative prognostic significance. nificance. For instance, patients with mutations in TP53,
RUNX1 point mutations also are relatively common in DNMT3A, NRAS, RUNX1, ASXL1, or EZH2 have a greater
patients with t-MDS. risk of leukemia progression or death than would be pre-
Mutations in genes altering DNA methylation and chro- dicted by the IPSS, and patients with IPSS low-risk disease
matin remodeling are common in MDS. TET2 mutations, who harbor one of these mutations have an outcome more
for example, are present in 20%-25% of patients, and similar to IPSS intermediate-1–risk disease.
IPSS-R (see: http://www.mds-foundation.org/ipss-r-calculator/)

Table 17-9 Survival and AML progression
risk with the 2012 Revised International
Prognostic Scoring System for MDS
Time until
%patients Median (IPSS-R).
Median 25% of
(n=7,012; survival for
Risk group Points survival, patients
AML data on pts under 60
years develop
6,485) years
AML, years
Very low 0-1.5 19% 8.8 Not reached Not reached
Low 2.0-3.0 38% 5.3 8.8 10.8
Intermediate 3.5-4.5 20% 3.0 5.2 3.2
High 5.0-6.0 13% 1.5 2.1 1.4
Very high >6.0 10% 0.8 0.9 0.7

MDS mutation landscape

Proliferation Impaired differentiation
BRAF GNAS KRAS CDKN2A PTPN11 PTEN
<1% <1% 1% <1% <1% <1%
RUNX1 ETV6 SETBP1
9% 3% 7%
JAK2 CBL NRAS Other

3% 2% 4%
STAG2
NPM1 TP53 and other
2% 8% cohesins
Epigenetic regulation
5 to 10%
EZH2 DNMT3A
6% 8% Pre-mRNA splicing SF1 SF3A1
ASXL1
14% 1% 1%
U2AF1
8%
IDH1/2 ZRSR2
PRPF40B
2% 5%
1%
TET2 SF3B1
21% UTX ATRX 22%
1% <1% SRSF2 U2AF65
11% <1%
IPSS-independent good prognosis No clear independent effect IPSS-independent poor prognosis
Figure 17-8 Recurrent somatic mutations in MDS, including approximate frequency of the most common recurrent somatic mutations in MDS
and their prognostic significance. Some mutations influence the phenotype and are therefore more common in specific subtypes of MDS; for
instance, SF3B1 mutations are found in >50% of patients with RARS, ATRX mutations are most common in MDS associated with acquired
alpha thalassemia, and SRSF2 mutations are more common in MPN/MDS overlap syndromes such as chronic myelomonocytic leukemia. The
prognostic value of SF3B1 mutations is unclear with some series suggesting no independent value and others suggesting a more favorable
outcome even when accounting for other clinicopathological parameters. The frequency of TET2, EZH2, ASXL1, IDH2, NRAS, KRAS, JAK2,
BRAF, GNAS, CBL, CDKN2A, PTPN11, RUNX1, ETV6, NPM1, PTEN, and TP53 mutations and their prognostic significance are derived from
Bejar R, et al. N Engl J Med. 2011;364:2496-2506. Frequency of SF3B1, ZRSR2, SRSF2, SF3A1, PRPF40B, USAF65, U2AF1 (=U2AF35), and SF1
mutations is from Yoshida K et al. Nature 2011;478:64-69, with SF3B1, ZRSR2, SRSF2, and U2AF1 data frequency combined with Thol F, et al.
Blood. 2012;119:3578-3584. The negative prognostic impact with SRSF2 mutations is from Thol F, et al. Blood. 2012;119:3578-3584.
Patients who develop t-MDS secondary to exposure to the MDS1-EVI1 (MECOM) genes; such patients often have a
mutagenic or carcinogenic agents almost always have chro- normal or elevated platelet count at the time of diagnosis
mosomal abnormalities. t-MDS is most commonly associ- and have a grim prognosis.
ated with previous treatment with alkylating agents or
exposure to ionizing radiation, and these cases frequently Key points
demonstrate losses involving chromosomes 5 or 7. The
latency period for t-MDS arising after alkylating agent ther- • One-half of patients with de novo MDS and most patients
apy is typically 3-7 years. Patients treated with epipodophyl- with secondary, therapy-related MDS have a clonal cytogenetic
abnormality.
lotoxins (eg, etoposide) can develop specific translocations
• 5q– syndrome has a relatively benign prognosis, but not all
involving the breakpoint at 11q23; the latency period
patients with del(5q) have 5q– syndrome. Deletion of RPS14, a
between exposure and MDS/AML development is typically
gene on chromosome 5q that encodes a ribosomal subunit,
1-3 years. These 11q23 translocations lead to transcription of contributes to the erythropoietic defect in del(5q) MDS and links
a fusion protein involving the mixed-lineage leukemia del(5q) MDS to DBA, which is due to heterozygous germ line
(MLL) gene. Translocations and inversions of 3q21/3q26 can mutations in genes, such as RPS19 encoding ribosomal proteins.
arise after etoposide treatment and involve rearrangement of

of TNFα and its receptors. In marrow cultures, strategies that

Key points (continued) block TNFα-mediated signals or TGFβ family members, such
• Patients with t-MDS who have been exposed to alkylating as the use of anti-TNFα antibodies or activin IIA/B receptor
agents or ionizing radiation usually have abnormalities of ligand traps that block erythropoiesis-inhibiting growth and
chromosomes 5 and 7, whereas those who have been exposed differentiation factor 11 (GDF11), significantly increase the
to epipodophyllotoxins usually have abnormalities of chromo- numbers of hematopoietic colonies compared with untreated
some 11q23. cells. Increased apoptosis has been identified in both mature
• More than 30 genes are known to harbor somatic mutations cells and immature CD34+ cells from patients with lower-risk
in patients with MDS. Those with IPSS-independent prognostic
MDS, compared with healthy controls and patients with
value include DNMT3A, EZH2, TP53, RUNX1, ASXL1, and NRAS.
higher risk MDS or de novo AML. In patients with higher-
risk MDS or AML, cell survival signals dominate.
Several studies have suggested that the bone marrow
Cell biology
microenvironment is abnormal in MDS. The growth of stro-
A major challenge in unraveling the complex pathogenesis of mal progenitors is defective, with reduced colony growth
MDS is distinguishing primary events from secondary effects and failure of cultures to grow to confluence. Furthermore,
of specific initiating mutations within HSCs and progenitor stromal support of the growth and maturation of normal
cells or the marrow microenvironment. As described above, hematopoietic progenitors also is impaired, consistent with
MDS arises from DNA mutation-driven clonal expansion of a functional defect. Stromal cells may play an important role
multipotent or pluripotent HSCs or progenitor cells. Most in the development and maintenance of abnormal signaling
studies of adults with MDS have shown that ineffective networks mediated by TNFα, Fas, and other soluble factors.
hematopoiesis, as opposed to the lack of hematopoietic Conditional knockout of the Dicer gene in osteoprogenitor
activity that characterizes AA, is the major factor contribut- cells in mice induced development of an MDS-like syn-
ing to pancytopenia in MDS. Abnormal responses to cyto- drome, and an activating mutation of β-catenin promoted
kine growth factors, impaired cell survival, and defects in the myeloid transformation in another murine model. The rel-
bone marrow microenvironment are all implicated in the evance of these models for the clinical condition is unclear.
pathogenesis of MDS.
Analysis of X-linked polymorphisms and newer molecu-
Key points
lar techniques indicate that the malignant clone in MDS
includes both CD34+ cells and more differentiated myeloid, • MDS is clonal disorder that arises in hematopoietic stem and
erythroid, and megakaryocytic cells. B cells are sometimes progenitor cells and affects the entire myeloid compartment.
part of the clonal process, but T cells are rarely involved, The heterogeneous nature of MDS and the advanced age at
although distinct T-cell clones akin to those seen in large disease onset infer the existence of multiple cooperating genetic
granular lymphocyte disorders may be detected in associa- lesions.
tion with MDS. Cell culture studies with primary MDS sam- • Whole genome sequencing shows that most patients with
MDS have somatic nonsense or missense mutations.
ples have shown reduced growth of multilineage colony-
• Both the “soil” (microenvironment) and the “seed” (hemato-
forming unit (CFU)–granulocyte-erythroid-monocyte-
poietic progenitor cells) may be abnormal in MDS, contributing
megakaryocyte progenitors (GEMM) and of lineage-
to failed hematopoiesis.
restricted burst-forming unit–erythroid, CFU–erythroid, • Abnormal responses to cytokine growth factors, impaired
CFU–granulocyte-macrophage, and CFU–megakaryocyte hematopoietic progenitor cell survival and excessive intramedul-
progenitors. These abnormalities in the progenitor compart- lary apoptosis, and defects in the marrow microenvironment
ment likely contribute to the development of peripheral have all been implicated in the pathogenesis of MDS.
blood cytopenias and might underlie the responses of some
patients to pharmacologic doses of hematopoietic growth
Treatment of MDS
factors.
Experimental evidence also implicates inhibitory cytokines With the exception of allogeneic HSCT, no therapeutic
and increased intramedullary apoptosis as contributors to options in MDS have demonstrated curative potential. How-
ineffective hematopoiesis in early MDS. Death receptor ever, three medications have specific US Food and Drug
ligand binding may contribute to excessive apoptosis of Administration (FDA) approval for MDS-related indications
hematopoietic precursors, resulting in ineffective hematopoi- (azacitidine, decitabine, and lenalidomide), and these drugs
esis. For example, in several studies, bone marrow cells from offer benefit to a subset of patients. Advanced age, the pres-
patients with MDS demonstrated increased expression of Fas ence of comorbidities, and a lack of a suitable donor limit the
and Fas ligand, TGFβ family members and their receptors, or availability of allogeneic HSCT, but use of reduced-intensity

conditioning approaches and alternative stem cell sources gastrointestinal symptoms. Deferiprone (L1) is widely used
(eg, umbilical cord blood and mismatched donors, including for chelation therapy in thalassemia, but a risk of agranulo-
haploidentical donors) are expanding the roster of poten- cytosis limits its use in MDS. Platelet transfusions also may
tially eligible patients. Therefore, patients with MDS who are be necessary in some patients with MDS, but development of
potentially candidates for transplantation should be evalu- alloimmunization is problematic.
ated early in the disease course by a physician with expertise
in stem cell transplantation. In many centers, reduced-
Hematopoietic growth factors
intensity stem cell transplantation is now routinely per-
formed for patients in their late 60s and early 70s. Hematopoietic growth factors are an integral part of the
Goals of MDS therapy for individual patients depend in treatment of MDS, despite the lack of a specific FDA-
part on the stage of disease, and include symptom control, approved indication for any of the available agents. Erythro-
reduction of transfusion needs, delay of disease progression, cyte growth factors in particular may reduce transfusion
and extension of survival. Prognostic systems such as the requirements by improving hemoglobin levels, and these
IPSS-R, supplemented by molecular testing, allow clinicians agents are generally well tolerated.
to incorporate clinicopathological risk factors for death and Studies with recombinant erythropoiesis-stimulating
disease progression into therapeutic decisions. agents (ESAs; epoetin and darbepoetin) demonstrated ery-
throid response rates in the range of 20% to 40%. The com-
bination of ESA and G-CSF may be more effective in
Supportive care: transfusions and iron chelation
improving anemia than treatment with ESA alone, especially
Despite the availability of several active treatments for MDS, in patients with RARS. No prospective studies have shown
transfusion support remains a mainstay of therapy for many an alteration in survival with ESAs in MDS, although several
patients. Patients receiving red blood cell transfusions at retrospective studies suggest that ESAs may improve life
least once every 8 weeks have a poorer survival than those expectancy and there is no increase in AML progression with
who do not require regular transfusions, probably because a ESA use.
need for transfusions is a marker of more advanced hemato- An 8- to 12-week trial of an ESA is appropriate for anemic
poietic failure and higher risk disease. In a number of stud- patients with serum erythropoietin levels <500 U/L. Patients
ies, lower-risk patients with MDS who have a ferritin >1000 with serum erythropoietin levels ≥500 U/L respond only
ng/mL have been shown to experience poorer survival than rarely to ESA therapy, and patients who have heavy transfu-
lower-risk MDS patients with a ferritin ≤1000 ng/mL, sug- sion needs are less likely to respond than those who do not
gesting that transfusion-related iron overload also might be require transfusions.
a contributing factor to poorer outcomes in transfusion- Both G-CSF (filgrastim, tbo-filgrastim) and GM-CSF (sar-
dependent patients. gramostim, molgramostim) have been evaluated in patients
Because the correlation between serum ferritin and iron with MDS and increase the neutrophil count in up to 60%-
burden is relatively poor and patients receiving transfusions 90% of patients, which may help some patients who have
develop iron overload at different rates, newer techniques for recurrent infections. Some patients treated with G-CSF, how-
noninvasively measuring hepatic iron concentration, such as ever, just produce more functionally defective neutrophils
quantitative (R2*/T2*) magnetic resonance imaging (MRI), and derive no benefit from myeloid growth factors, although
may be useful in determining which patients are the best can- the increased number of white cells may provide false reassur-
didates for iron chelation. Cardiac T2* MRI results may also ance to the patient and physician. Concerns regarding use of
be clinically helpful in determining patients’ risk from iron G-CSF and risk of leukemic transformation were addressed
overload, but T2* signals in the heart are rarely abnormal in a randomized controlled trial of 102 patients with high-risk
until patients have received at least 80-100 units of blood. MDS who were treated with either G-CSF or supportive care.
Consideration should be given to initiation of iron chela- No differences in frequency or time to progression to AML
tion therapy with parenteral deferoxamine or oral defera- were seen between the two groups overall, but survival was
sirox in patients who have a reasonable life expectancy, are shorter in patients with 5%-19% blasts who received G-CSF.
red blood cell transfusion dependent, and have evidence of Pegfilgrastim has been associated with splenic rupture and
tissue iron overload. No controlled prospective data, how- leukemoid reactions in MDS and, if used, should be adminis-
ever, support a survival benefit from iron chelation in MDS, tered only with caution and started at low doses (eg, 1-3 mg,
and such therapy is costly and can have adverse effects. In rather than the standard 6 mg vial).
elderly patients with MDS, a dose of deferasirox high enough Thrombopoietin (TPO)-receptor agonists approved for
to cause a negative iron balance (ie, at least 20-30 mg/kg/day) use in immune thrombocytopenia, romiplostim and eltrom-
often results in elevated creatinine or intolerable bopag, have been evaluated in patients with MDS and can

improve the platelet count in many patients and reduce transcription rates. Methylated cytidine residues cluster in
bleeding events. Patients who are not heavily platelet trans- so-called cytosine-phosphate-guanine (CpG) islands, which
fusion dependent and who have an endogenous TPO level are located near the promoter regions of many genes. When
<500 pg/mL are most likely to benefit. However, some these regions are hypermethylated, expression of nearby
patients experience an increase in blood or marrow blast genes is silenced and this represents a mechanism for regu-
proportion during romiplostim or eltrombopag therapy, lating gene transcription. DNA methyltransferase 1
which may be because some myeloblasts have functional (DNMT1) is the enzyme responsible for maintenance of
TPO receptors. In one placebo-controlled study of romip- cytidine methylation patterns, and the aza-substituted cyto-
lostim monotherapy, progression to AML was observed in sine nucleoside analogs azacitidine and decitabine can
6% of patients treated with romiplostim, compared with inhibit DNMT1 by incorporating into RNA or DNA and
2.4% with placebo, with the majority of progressions seen irreversibly binding to this enzyme, resulting in generalized
among patients who already had excess blasts before treat- hypomethylation of DNA and reversal of gene silencing.
ment. When romiplostim was used in pilot studies in com- Although these so-called epigenetic changes occur in vitro in
bination with azacitidine, decitabine, or lenalidomide, cells exposed to DNMT1 inhibitors, it is not clear whether
however, an increased rate of progression to AML was not these epigenetic effects are responsible for the clinical activ-
observed. Another concern with TPO agonists is the possi- ity of azacitidine or decitabine in MDS or whether other bio-
bility of development of marrow fibrosis with long-term use, logic effects (eg, DNA damage) also play a role.
because mice engineered to overexpress TPO develop a Azacitidine is the first and, as of this writing, only medica-
myelofibrosis-like picture, but the clinical relevance of this is tion that has been shown in a randomized trial to improve
unclear and, to date, fibrosis in TPO-agonist treated patients survival in higher-risk MDS patients. In a multicenter trial
with MDS has been rare. Rebound thrombocytopenia can (AZA-001), 358 patients with IPSS intermediate-2 or high-
occur with discontinuation of TPO agonists. Thrombocyto- risk MDS were randomized to receive either azacitidine
penic patients who have bleeding from mucosal surfaces (eg, 75 mg/m2 subcutaneously for 7 consecutive days every
urinary bladder or gut) may benefit from topical therapy or 28 days or conventional care (ie, best supportive care, either
careful use of the antifibrinolytic agent epsilon aminoca- alone or with low-dose cytarabine or AML-like induction
proic acid. chemotherapy using infusional cytarabine and an anthracy-
cline). The median survival time was 24 months in patients
receiving azacitidine, versus 15 months in patients receiving
Key points conventional care. Although the complete response rate in
• Transfusion support with leukocyte-depleted blood products is
the azacitidine treated group was a modest 17%, subsequent
an integral part of supportive care for most patients with MDS. analysis demonstrated that a complete response was not nec-
Iron chelation may become necessary in carefully selected essary for patients to achieve a survival benefit; however, it is
low-risk patients who are receiving regular red cell transfusions. unclear whether stable disease alone or minor hematologic
• Data are insufficient to determine whether treating MDS improvements are beneficial. Azacitidine is approved for
patients with hematopoietic growth factors alters disease intravenous administration and subcutaneous dosing. Intra-
progression or survival. venous administration avoids injection-site reactions, but
• ESAs lead to a red blood cell response in ~20%-40% of requires either central or peripheral venous access.
patients; adding G-CSF to ESAs can lead to red blood cell Decitabine is also active in MDS, but a European multi-
response in ~40% of patients, and responses to combined
center study designed to show a survival benefit with
therapy may be more common among patients with RARS.
decitabine in MDS was negative. The overall survival of the
• ESAs are less effective in patients with high pretreatment
control arm in that study (8 months) suggests that a different
serum erythropoietin levels (≥500 U/L).
• TPO receptor agonists (thrombopoiesis-stimulating agents)
population was enrolled compared to AZA-001.
can raise the platelet count in some patients with MDS and Clinical response to hypomethylating agents may be delayed,
decrease platelet transfusions and clinically significant bleeding and an adequate therapeutic trial of either agent requires at least
events, but they have been associated with increased blast four to six treatment cycles. Although the initial FDA approval
proportion in some cases and are not FDA approved for MDS. of decitabine was for a regimen of 15 mg/m2 administered every
8 hours for 9 doses intravenously (in a hospital-based setting),
the most commonly used regimen in clinical practice is 20 mg/
Hypomethylating agents (DNA methyltransferase
m2 intravenously once daily for 5 consecutive days, repeated
inhibitors)
every 4-6 weeks. In a multicenter study of this 5-day decitabine
Cytidine residues in mammalian DNA can be methylated, regimen, 17% of patients achieved a complete response, 15%
and DNA methylation is a dynamic process that affects achieved a marrow response, and 18% experienced hematologic

improvement, similar to the response rates observed with Immunomodulatory drugs

azacitidine therapy.
The drug thalidomide has multiple biological effects, includ-
The most common adverse events associated with both
ing alteration of immune cell subsets, inhibition of TNFα
hypomethylating agents are neutropenia and thrombocyto-
and other cytokines, and inhibition of neoangiogenesis in
penia, which often improve over time with continued treat-
the marrow. These effects are mediated by modulation of the
ment as the MDS clones are suppressed and normal
activity of an E3 ubiquitin ligase complex that includes the
hematopoiesis recovers. The optimal maintenance dosing
protein cereblon. When thalidomide was used for MDS in
once patients achieve a response is unknown, but some
the 1990s, responses were seen in ~20% of patients, but the
maintenance therapy appears to be required to maintain
drug was difficult to tolerate (especially for elderly patients)
responses. Thus far, no therapy has been demonstrated to
due to sedation, constipation, peripheral neuropathy, and
improve survival for patients with lower risk MDS, though
other adverse events.
DNA hypomethylating agents can improve peripheral counts
Lenalidomide was then generated by chemical modifica-
and reduce transfusion needs in a minority of such patients.
tion of thalidomide, and has an improved safety profile
Given the frequency of mutations in pathways that alter
without the neurologic toxicity seen with thalidomide.
DNA methylation in MDS, it is reasonable to hypothesize
Lenalidomide has more potent immunomodulatory, anti-
that such mutations might serve as biomarkers for therapeu-
TNFα, and anti-vascular endothelial growth factor effects
tic response to hypomethylating agents. Indeed, in both a
than thalidomide. Its primary mechanism is via cereblon-
French study and in a trial run by the defunct Bone Marrow
mediated alteration in the degradation rate of casein kinase
Failure Consortium, the presence of mutations in TET2 or
1, a serine-threonine kinase which is encoded on chromo-
DNMT3A predicted a modestly higher likelihood of response
some 5q and modulates Wnt/β-catenin signaling.
to azacitidine therapy. The response rate was high enough in
After phase 1 testing suggested a high response rate in
the wild-type group, however, that this mutation signature
del(5q) MDS, lenalidomide was tested in a phase 2 trial in
cannot be used to select therapy in the clinic.
patients with IPSS low-risk or intermediate-1-risk disease
Deacetylase (DAC) inhibitors are another therapeutic
who were red blood cell transfusion dependent and had a
strategy for MDS that also is based on the principle of epi-
deletion of chromosome 5q31, either alone or in associa-
genetic modification. DAC inhibitors are agents that main-
tion with other chromosomal abnormalities. Of 148
tain chromatin in a transcriptionally active state by inhibiting
patients enrolled in this phase 2 study, 67% achieved trans-
deacetylation of histone tails on chromatin; the acetylation
fusion independence, with a median time to response of 4.6
state of cytoplasmic proteins also is modified during DAC
weeks. The median increase in hemoglobin was 5.4 g/dL
inhibitor therapy, but their significance is unclear. In vitro,
and the median duration of response was >2 years. A major
these agents lead to reversal of transcription repression and
cytogenetic response (ie, elimination of the del(5q) clonal
gene silencing and are synergistic in reactivating silenced
abnormality) occurred in 44% of patients. The major
genes when combined with hypomethylating agents. Several
adverse effect was myelosuppression, with grade 3 to 4 neu-
DAC inhibitors currently are under clinical investigation in
tropenia and thrombocytopenia seen in up to 55% of
MDS and AML, but in three cooperative group randomized
patients; treatment-emergent cytopenias are associated
trials, E1905, S1117, and AZA-Plus, the combination of
with a moderately higher likelihood of response. These
azacitidine plus a DAC inhibitor (entinostat [MS-275] in
results led to the approval of lenalidomide by the FDA in
E1905, vorinostat in S1117, valproic acid in Aza-Plus) was
2005 for patients with del(5q) with IPSS low-risk or inter-
not superior to azacitidine monotherapy, and combination
mediate-1–risk disease who are red blood cell transfusion
therapy was associated with more adverse effects, such as
dependent.
fatigue and thrombocytopenia. Similarly, in a fourth ran-
A second phase 2 trial of lenalidomide was conducted in
domized study funded by an industry sponsor, addition of
patients with the same eligibility who did not have del(5q). In
the DAC inhibitor pracinostat to azacitidine did not improve
this patient population, responses were less frequent and of
responses compared to azacitidine monotherapy. Therefore,
shorter duration compared with those in patients with del(5q);
the future of DAC inhibitors in MDS is unclear.
26% of patients became red blood cell transfusion indepen-
Once hypomethylating agents fail the patient, the prognosis
dent, with a median response duration of 41 weeks. A third
is grim, with a median survival <6 months. Switching from
trial was conducted comparing a starting dose of 5 mg daily to
one failed hypomethylating agent to the other agent or adding
10 mg for 21 out of 28 days in patients with del(5q) MDS,
lenalidomide or a deacetylase inhibitor is usually not helpful.
because many patients starting at the 10-mg dose require dose
Such patients should instead be referred for HSCT or enrolled
reduction due to treatment-emergent cytopenias. Complete
in clinical trials whenever feasible. Responses are seen in some
response rates and cytogenetic response rates were superior in
patients with low-dose cytarabine or clofarabine.

the 10-mg arm. A placebo-controlled trial did not show an

increase in disease progression with lenalidomide use. Clinical case
Lenalidomide at high doses (>10 mg/day) has some clini- The patient described above has an IPSS score of 1.5: 0.5 point
cal activity in patients with high-risk disease (eg, excess blasts for being pancytopenic, 0.5 point for an intermediate-risk
or complex karyotype) or AML, but is not FDA approved for karyotype, and 0.5 point for having a blast count between 5%
these indications. Patients with a low platelet count are much and 10%. This places him in the IPSS intermediate-2 risk group,
less likely to achieve benefit from lenalidomide than those with an expected median survival of just over 1 year. His IPSS-R
with a platelet count >50 × 109/L. score is 5.5 (high)—2 points for an intermediate karyotype, 2
points for blast proportion, 1 point for hemoglobin between
8 and 10 g/dL, and 0.5 points for platelet count between 50
Immunotherapy × 109 and 100 × 109/L—with an expected median survival of
1.5 years. The AZA-001 randomized trial demonstrated that
An autoreactive T-cell-mediated process suppressing hema- IPSS intermediate-2-risk and high-risk patients gain a median
topoiesis may contribute to pancytopenia in some patients survival benefit of 9 months from azacitidine, which would
with MDS. Several studies have demonstrated that treatment be an appropriate therapy in this situation. Unfortunately, the
approaches analogous to immunosuppressive therapy of AA patient is too old to consider allogeneic HSCT using currently
may be beneficial in MDS. Therapy with anti-thymocyte available approaches, but if he were younger, HSCT should be
globulin (ATG) or a calcineurin inhibitor (ie, cyclosporine considered.
or tacrolimus) benefits some patients with lower risk disease
(<10% blasts), especially those who are <60 years of age, lack Allogeneic HSCT
transfusion dependence, and have either a normal karyotype
or trisomy 8. Selection of patients most likely to respond to Even though allogeneic HSCT is the only potentially curative
ATG or cyclosporine therapy remains challenging, and mar- therapy in MDS, <10% of patients with MDS currently
row hypocellularity and HLA-DR15 positivity have predicted undergo HSCT, due to older age, comorbid conditions, lack
response to immunosuppressive therapy in some studies but of a suitable donor, high cost, patient concern over the risk
not in others. of transplantation, and failure of clinicians to refer patients
who might be transplant candidates to transplant centers.
Younger patients (ie, ≤40 years) without excess blasts at the
Key points time of transplant may have a long-term disease-free-sur-
• Azacitidine has been demonstrated to improve survival by a vival rate exceeding 50% after an HLA-matched HSCT.
median of 9 months in patients with higher-risk MDS. Patients with high IPSS score or treatment-resistant disease
Decitabine, another hypomethylating cytosine analog, also have survival rates <30% after HSCT. Patients with a com-
produces responses in MDS and delays AML progression. Both plex monosomal karyotype, defined as two or more autoso-
drugs are approved by the FDA for the treatment of MDS. mal monosomies or one monosomy plus additional
• Azacitidine and decitabine induce DNA hypomethylation structural chromosomal abnormalities, are at particularly
through the inhibition of DNMT1, but it is not clear whether this high risk for poor outcome. When patients with complex
mechanism is responsible for the clinical effects. karyotype also have a TP53 mutation, the long-term disease-
• Lenalidomide led to transfusion independence in 67% of
free survival even with HSCT is <10%.
lower-risk MDS patients with deletions of chromosome 5q, and
Allogeneic HSCT should be seriously considered for
some patients also achieved a cytogenetic remission. Lenalido-
patients with higher-risk MDS who have a matched donor
mide also has some effectiveness in anemic patients with
lower-risk MDS who lack deletion of chromosome 5q, but the
and a good performance status. The use of reduced-intensity
response rate is only 20%-25%, and responses are not durable. conditioning regimens may permit allogeneic HSCT in older
• Lenalidomide’s mechanism of action in MDS is via binding to individuals.
cereblon and modulation of ubiquitination of casein kinase 1 Given the risks of allogeneic HSCT, defining the optimal
and alteration of casein kinase’s clearance rate. Casein kinase is time to refer patients for transplantation is an important
encoded on chromosome 5q. consideration. One analysis indicated that performing trans-
• Some patients with MDS respond to ATG or cyclosporine/ plantation in patients with lower-risk disease (ie, IPSS low
tacrolimus immunotherapy, but selecting the most appropriate and intermediate-1 risk) only at the time of progression of
patients for this therapy remains challenging. Younger patients, disease resulted in greater life expectancy than when HSCT
those with lower risk disease (ie, those who are not yet transfu-
was performed earlier in the disease course. In contrast,
sion dependent or have required transfusions for only a short
patients with higher-risk disease (IPSS intermediate-2 and
time) with normal karyotype or trisomy 8 seem most likely to
high risk) benefited from HSCT shortly after diagnosis.
benefit.
Unfortunately, disease relapse occurs in the majority of

high-risk patients after HSCT and thus represents a continu- General therapeutic approach
ing challenge. Strategies to reduce relapse rates are being
An approach to MDS therapy is outlined in Figure 17-9. All
studied and include pre- and posttransplantation interven-
patients should receive supportive care with transfusions
tions with novel therapies. No clear benefit has been shown
and antimicrobial agents as needed. Iron chelation therapy
for the administration of one or more courses of cytotoxic
can be considered for selected RBC transfusion-requiring
chemotherapy or hypomethylating agent therapy before
lower risk patients with an expected long life expectancy and
HSCT, although pretransplantation therapy may be useful
evidence of transfusional hemosiderosis.
to reduce the burden of marrow blasts to ≤10% before
For lower-risk patients (ie, those without excess blasts or
the HSCT.
an adverse karyotype) in whom the clinical picture is domi-
HSCT is the treatment of choice for children and young
nated by anemia, the initial therapeutic choice depends on
adults with MDS. It is imperative to perform a DEB test to
the karyotype and the serum EPO level. For patients with
exclude FA before performing a transplantation in a child or
del(5q), lenalidomide is an appropriate first choice and is
young adult with apparently de novo MDS, because patients
FDA approved for this indication. For patients without
with FA suffer severe toxicity with conventional condition-
del(5q) but with serum EPO <500 U/L, epoetin or darbepo-
ing regimens and also require close monitoring for nonhe-
etin are recommended.
matologic tumors after transplantation. Although most
The most appropriate therapy for lower-risk patients with
patients with FA have dysmorphic features such as short
either anemia with serum EPO >500 U/L and without
stature and radial ray anomalies, many do not. A bone mar-
del(5q), pancytopenia or a clinical picture dominated by
row examination and cytogenetic testing should be per-
individual cytopenias other than anemia (ie, neutropenia or
formed on any related donor when the recipient is a child or
thrombocytopenia) is unclear. Hypomethylating agents can
young adult with bone marrow monosomy 7, because there
be beneficial in some patients with lower-risk disease, though
have been instances in which an unsuspected clonal disorder
their effect on survival in this group is unclear. Patients with
has been detected in the prospective donor. Particular care
isolated thrombocytopenia may overlap with immune
must be exercised in determining the proper conditioning
thrombocytopenia and may benefit from corticosteroids,
regimen and best time to perform HSCT in infants and
romiplostim, intravenous gamma globulin, or other immune
young children because of the toxic effects of radiation on
thrombocytopenia-directed therapies. Immunosuppressive
the developing central nervous system and because of differ-
therapy, lenalidomide, supportive care alone, or HSCT are
ences in drug metabolism compared with older children and
all reasonable choices in the other patient groups, depending
adults.
on patient-specific factors. Many of the patients in these
groups do not truly have “lower risk” disease—for instance,
the population with pancytopenia is enriched for those with
Key points
EZH2 mutations, which confers increased risk —and, in the
• Allogeneic HSCT remains the only routinely curative approach future, molecular profiling may help assign them to a higher-
in MDS and is an important consideration if the patient is risk group, likely resulting in increased therapy with hypo-
young, otherwise healthy, and has an HLA-identical sibling or a methylating agents or other potentially disease-modifying
closely matched unrelated donor. Cure rates overall are approaches.
~30%-40%.
For higher-risk patients, the treatment approach differs
• Reduced-intensity (nonmyeloablative) conditioning regimens
depending on whether the patient is a transplant candidate.
are associated with a lower transplantation-related mortality but
Higher-risk patients who are HSCT candidates should pro-
higher relapse rate in MDS; overall survival is similar with
reduced-intensity and conventional myeloablative conditioning.
ceed with definitive HSCT therapy as soon as feasible.
Reduced-intensity conditioning regimens may permit HSCT to HSCT may be preceded by a few treatment cycles of a hypo-
be performed in older and sicker patients who would not methylating agent as a “bridging” therapy to try to cytore-
tolerate myeloablative conditioning. duce or at least keep the disease stable until a donor is
• Transplantation at the time of progression for patients with identified, insurance approval is obtained, and pretrans-
lower risk disease, and as soon as feasible after the time of plant screening tests are completed. Patients who are not
diagnosis for patients with higher risk disease, yields the greatest HSCT candidates can be treated with a hypomethylating
life expectancy. agent; some investigators prefer azacitidine over decitabine
• HSCT is the treatment of choice for pediatric MDS; because of the demonstrated survival advantage in this
however, donors and recipients must be screened carefully to
setting.
exclude familial disorders such as FA that would alter the
Once initial therapy fails, no optimal second-line therapy
management.
is defined, and the choice depends on clinical circumstances.

Bibliography | 519
MDS initial diagnosis (using WHO 2008 diagnostic criteria,

supplemented by novel genomics approaches)
Is there a need for No

Clinical monitoring
treatment now?
Development of a
Yes need for therapy
Lower risk Individualized risk assessment, Higher risk
using IPSS-R or other tools
No, other important
Is anemia isolated, cytopenias are also present Is the patient a
or the major problem? transplant candidate?
Yes No Yes
Is del5q No Serum No Azacitidine or decitabine Allo-SCT, perhaps

present? EPO <500 U/L? (ie, HMA) until disease with HMA or
progression, relapse, or chemotherapy as
Yes Yes drug intolerance bridging therapy
Lenalidomide; if sEPO ESA ± G-CSF Optimal approach is Relapse

<500 U/L, ESA trial unclear; consider
before or after G-CSF or TPO agonist,
HMA, IST, clinical trial Donor lymphocyte
Failure infusion or second
Failure Failure
allo-SCT, possibly after
Failure cytoreductive therapy
Optimal therapy unclear; consider HMA, IST, androgens,

Failure
lenalidomide (if not already used), or clinical trial
Allo-SCT, enrollment in a clinical Enrollment in a clinical trial or

trial, or palliative/supportive care supportive/palliative care
Figure 17-9 A general approach to MDS therapy, as described in the accompanying text. All patients should receive supportive care. Low-
intensity therapies are most suited for lower risk MDS, whereas patients with higher-risk disease should move to allogeneic stem cell transplant,
if feasible, or otherwise be considered for azacitidine or decitabine therapy. WHO = World Health Organization; ESA = erythropoiesis-
stimulating agent; G-CSF = granulocyte colony stimulating factor; IST = immunosuppressive therapy; TPO = thrombopoietin, IPSS-R = 2012
Revised International Prognostic Scoring System; MDS = myelodysplastic syndromes; HMA = hypomethylating agent (azacitidine or decitabine);
alloSCT = stem cell transplantation; sEPO = serum erythropoietin level. Partly based on National Comprehensive Cancer Network (NCCN)
guidelines; see http://www.nccn.org. Source: Bejar R, Steensma D. Blood. 2014;124:2793-2803.
Supportive care is the default, and clinical trial enrollment is management, and therapy of AA. Summary of results of
always appropriate, if a well-designed study is available for unrelated-donor HSCT in AA in recent years.
which the patient is eligible.
Paroxysmal nocturnal hemoglobinuria
Bibliography Brodsky, RA. Paroxysmal nocturnal hemoglobinuria. Blood.

2014;124:2804-2811.
Aplastic anemia Kelly RJ, Hill A, Arnold LM, et al. Long-term treatment with
Scheinberg P, Nunez O, Weinstein B, et al. Horse versus rabbit eculizumab in paroxysmal nocturnal hemoglobinuria: sustained
antithymocyte globulin in acquired aplastic anemia. N Engl J efficacy and improved survival. Blood. 2011;117:6786-6792.
Med. 2011;365:430-438. Prospective randomized study comparing
horse ATG, rabbit ATG, and alemtuzumab as first-line therapy in Myelodysplastic syndromes
severe AA.
Scheinberg P, Young NS. How I treat acquired aplastic anemia. Bejar R, Steensma DP. Recent developments in myelodysplastic
Blood. 2012;120:1185-1196. Practical review of diagnosis, syndromes. Blood. 2014;124:2793-2803. This article discusses

new insights from genomic discoveries that are illuminating Koreth J, Pidala J, Perez WS, et al. Role of reduced-intensity
MDS pathogenesis, increasing diagnostic accuracy, and refining conditioning allogeneic hematopoietic stem-cell transplantation
prognostic assessment, and which will one day contribute to more in older patients with de novo myelodysplastic syndromes:
effective treatments and improved patient outcomes. an international collaborative decision analysis. J Clin Oncol.
Fenaux P, Mufti GJ, Hellstrom-Lindberg E, et al. Efficacy of 2013;31:2662-2670. Decision analysis of transplant decisions for
azacitidine compared with that of conventional care regimens patients aged 60-70 based on IPSS score. For low/intermediate-1
in the treatment of higher-risk myelodysplastic syndromes: IPSS, nontransplantation approaches are preferred. For
a randomised, open-label, phase III study. Lancet Oncol. intermediate-2/high IPSS, RIC transplantation offers overall and
2009;10:223-232. Azacitidine was the first agent shown to improve quality-adjusted survival benefit.
survival in patients with higher risk MDS. This report describes the List A, Dewald G, Bennett J, et al. Lenalidomide in the
AZA-001 clinical trial comparing azacitidine to conventional care. myelodysplastic syndrome with chromosome 5q deletion. N
Greenberg PL, Tuechler H, Schanz J, et al. Revised International Engl J Med. 2006;355:1456-1465. Lenalidomide is very effective in
Prognostic Scoring System for myelodysplastic syndromes. patients with lower risk MDS and deletions of chromosome 5q31.
Blood. 2012;120:2454-2465. The revised version of the IPSS, This paper describes a 148-patient clinical trial of lenalidomide
which includes a more comprehensive list of karyotypic in 5q deletion MDS in which 67% of patients responded and no
abnormalities than the original IPSS but it is still only valid at the longer required transfusions.
time of diagnosis.

CHAPTER
18 © Dennis Kunkel Microscopy, Inc
Acute myeloid leukemia

B. Douglas Smith and Eytan M. Stein
Definition and epidemiology, 521 Treatment, 524 Acute promyelocytic leukemia, 527
Clinical manifestations, 521 Monitoring residual disease, 525 Pediatric AML, including Down
Subtype classification, 522 AML relapse, 526 syndrome, 528
Prognostic factors, 522 Older patients with AML, 526 Bibliography, 529

therapy have increased incidence with age, typically have a
multimedia component. 5- to 10-year latency period, and frequently are associated
with an antecedent therapy-related myelodysplastic syndrome
(MDS) and unbalanced loss of genetic material involving
Definition and epidemiology chromosomes 5 or 7. t-AML associated with exposure to
topoisomerase II inhibitors is less common and encompasses
Acute myeloid leukemia (AML) is a heterogeneous clonal 20%-30% of t-AML patients, has a shorter latency period of
hematopoietic progenitor or stem cell malignancy in which 1-5 years, is less often preceded by a myelodysplastic phase,
immature hematopoietic cells proliferate and accumulate in and may be associated with balanced recurrent chromo-
bone marrow, peripheral blood, and other tissues. This process somal translocations involving 11q23 (MLL gene) or 21q22
results in inhibition of normal hematopoiesis, characterized by (RUNX1). Other environmental risk factors include exposure
neutropenia, anemia, thrombocytopenia, and the clinical fea- to benzene and ionizing radiation. Patients with inherited
tures of bone marrow failure. AML accounts for 90% of all bone marrow failure syndromes (eg, Fanconi anemia, Shwach-
acute leukemias in adults, with an estimated 20,800 new cases man-Diamond syndrome, severe congenital neutropenia),
and 10,500 deaths expected in the United States in 2015. The genetic disorders (eg, Down syndrome), and MDS and myelo-
annual incidence is approximately 3.5 per 100,000 and increases proliferative neoplasms are also at increased risk of developing
with age, with approximately a tenfold increased risk between AML. Patients with these conditions who develop AML have
ages 30 (1 case per 100,000) and 65 years (1 case per 10,000). poorer treatment outcomes than patients with de novo AML.
The median age at diagnosis is approximately 67 years, with
~6% of patients <20 years of age and 34% of patients 75 years
or older. Overall survival in adults remains poor, with <50%
5-year survival in patients <45 years of age that continues to fall
to <10% in patients >60 years of age at diagnosis. In children,
Patients with AML generally present with nonspecific signs and
overall survival has improved to ~60%.
symptoms related to infiltration of the bone marrow and other
Most cases of AML have no apparent cause. The most com-
organs with leukemic blasts, including pallor, fatigue, bone pain,
mon known risk factor is previous exposure to radiation or
hepatosplenomegaly, fever, bruising, and bleeding. Tissue infil-
chemotherapy, particularly topoisomerase II inhibitors and
tration of the skin, gingiva, and central nervous system (CNS) is
alkylating agents, which results in therapy-related AML
more common with monocytic subtypes. CD56 expression, in
(t-AML), and accounts for ~10%-20% of all AML cases. AML
addition to monocytic subtypes, increases extramedullary risk
that arises after exposure to alkylating agents or radiation
at presentation. Patients with leukocytosis and leukemic blasts
>50,000/µL are at increased risk of pulmonary and CNS com-
Conflict-of-interest disclosure: Dr. Smith declares no competing
financial interest. Dr. Stein declares no competing financial interest. plications from leukostasis. Pathologically, this process shows a
Off-label drug use: Dr. Smith: not applicable. Dr. Stein: not applicable. combination of microinfarction and hemorrhage.
| 521

522 | Acute myeloid leukemia
AML may be associated with a variety of laboratory status. An absolute blast count >50,000/µL at diagnosis is asso-
derangements in addition to abnormal blood counts. Coag- ciated with increased risk of vascular complications.
ulation abnormalities are particularly common and severe in Originally, the morphologic characteristics of the newly
patients with acute promyelocytic leukemia (APL), but they diagnosed disease were considered important determinants
may be seen in all subtypes. Metabolic abnormalities related in prognostication. It now is evident, however, that chromo-
to tumor lysis syndrome also may be present, including somal (cytogenetic) and molecular abnormalities are the pri-
hyperuricemia, hyperkalemia, hyperphosphatemia, and mary tools and far better than morphology in assigning
hypocalcemia. In patients with monocytic leukemia, severe prognosis for patients with newly diagnosed AML. It is
hypokalemia may be present. imperative that the initial diagnostic workup of suspected
AML include testing for the commonly described abnormal-
ities, typically determined by bone marrow aspirate studies.
Subtype classification
Acquired, nonrandom, clonal chromosomal abnormali-
ties, including balanced translocations, inversions, deletions,
In the 1970s, AML was subclassified according to the
French-American-British (FAB) classification system using
Table 18-1 World Health Organization 2008 classification of acute
morphologic and cytochemical criteria to define eight major
myeloid leukemia (AML) and related myeloid neoplasms
AML subtypes (M0-M7) on the basis of greater than or equal
to 30% blasts, lineage commitment, and the degree of blast 1. AML with recurrent genetic abnormalities
cell differentiation. The FAB system has been replaced by the a. AML with balanced translocations/inversions
World Health Organization (WHO) classification, which i. AML with t(8;21)(q22;q22); RUNX1-RUNX1T1
was developed to incorporate epidemiology, clinical fea- ii. AML with inv(16)(p13.1q22) or t(16;16)(p13.1;q22);
tures, biology, immunophenotype, and genetics into the CBFB-MYH11
iii. Acute promyelocytic leukemia with t(15;17)(q22;q12);
diagnostic criteria. The WHO has identified a number of
PML-RARa
genetically defined subgroups of AML (Table 18-1). A new
iv. AML with t(9;11)(p22;q23); MLLT3-MLL
revision of the WHO classification, incorporating point
v. AML with t(6;9)(p23;q34); DEK-NUP214
mutations, will debut in 2016-2017.
vi. AML with inv(3)(q21q26.2) or t(3;3)(q21;q26.2);
AML is now defined as greater than or equal to 20% RPN1-EVI1
myeloblasts, monoblasts or promonocytes, erythroblasts, or vii. AML (megakaryoblastic) with t(1;22)(p13;q13);
megakaryoblasts in the peripheral blood or bone marrow, RBM15-MLK1
except in patients with the following cytogenetic abnorm b. AML with gene mutations
alities, who are classified as having AML irrespective of i. Provisional entity: AML with mutated NPM1
blast count: t(8;21)(q22;q22), inv(16)(p13q22), t(16;16) (nucleophosmin)
(p13;q22), and t(15;17)(q22;q12). Immunophenotypic char- ii. Provisional entity: AML with mutated CEBPA (CCAAT/
acterization using surface antigens remains important in enhancer binding protein A)
AML and may include progenitor-associated antigens (eg, 2. AML with myelodysplasia-related changes
human leukocyte antigen-DR [HLA-DR] [except in APL], 3. Therapy-related myeloid neoplasms
CD34, CD117) and myeloid antigens (eg, CD13, CD33). 4. AML, not otherwise specified
Complex composite immunophenotypes, including expres- a. AML with minimal differentiation
sion of lymphoid markers, also may be seen. b. AML without maturation
c. AML with maturation
d. Acute myelomonocytic leukemia
Prognostic factors e. Acute monoblastic/monocytic leukemia
f. Acute erythroid leukemia
AML is a clinically and biologically heterogeneous disease. i. Pure erythroid
Adverse clinical prognostic features include advanced age at ii. Erythroleukemia, erythroid/myeloid
diagnosis, extramedullary disease (including CNS leukemia), g. Acute megakaryoblastic leukemia
disease related to previous chemotherapy or radiation treat- h. Acute basophilic leukemia
ment (t-AML), and the presence of an antecedent hematologic i. Acute panmyelosis with myelofibrosis
disorder (typically, MDS or myeloproliferative disorders). 5. Myeloid sarcoma
6. Myeloid proliferations related to Down syndrome
Patients >60 years and especially those >75 years of age have
a. Transient abnormal myelopoiesis
poor long-term survival because of both disease- and host-
b. Myeloid leukemia associated with Down syndrome
related factors, including increased expression of multidrug
7. Blastic plasmacytoid dendritic cell neoplasm
resistance genes, medical comorbidities, and poor performance

Prognostic factors | 523
monosomies, and trisomies may be found in up to 50% of NPM1, CEBPA, MLL, RUNX1,
patients with AML. The karyotype is considered complex FLT3, RAS, WT1
when there are more than three abnormalities, and this is 45% normal karyotype
seen in 10%-20% of patients, often in association with TP53

gene mutation or deletion. Cytogenetic findings remain one
of the most important prognostic tools and often are classi-
11%
fied into favorable, intermediate, and unfavorable risk complex
5% t(8;21) karyotype
groups, but clinical study groups have not always been con- KIT
sistent in assigning the Individual abnormalities. It is uni RAS TP53
FLT3 6% inv(16)
versally agreed, however, that patients with the t(15;17)
(q22;q12-21) found in APL have excellent outcomes. 2% t(9;11)
Balanced abnormalities of t(8;21)(q22;q22), inv(16)(p13.1 2% t(11q23) 23% various
q22), and t(16;16)(p13.1;q22) involve the heterodimeric NPM1, CEBPA,
FLT3 1% t(6;9) MLL, RUNX1
components of core-binding factor (CBF) and are associated
1% inv(3)/t(3;3)
with a relatively favorable prognosis. Complex karyotype, NRAS
inv(3)(q21q26)/t(3;3)(q21;q26), and monosomal karyotype
(at least two autosomal monosomies or one single-autosomal Figure 18-1 Major cytogenetic subgroups of AML (excluding acute
monosomy combined with at least one structural abnormal- promyelocytic leukemia) and associated gene mutations.
ity) are associated with particularly poor outcomes.
Molecular alterations also provide important prognostic Cooperative Oncology Group genetically profiled all of the
information for many patients with AML, particularly those patients treated on protocol E1900, a randomized trial of
with normal karyotype disease (Figure 18-1). This is the 90 mg/m2 daily × 3 vs. 45 mg/m2 daily × 3 of daunorubicin
largest cytogenetic subset of AML, and without further abil- with both trial arms receiving 7 days of infusional cytara-
ity to classify these patients, most generally fall into an inter- bine. Analysis demonstrated that certain mutations co-occur
mediate-risk group. Yet, these intermediate-risk patients frequently—like NPM1 and FLT3-ITD while others, such as
have variable outcomes with conventional treatment strate- IDH mutations and TET2 mutations, appear to be mutually
gies, which may be explained by the underlying molecular exclusive, leading to insights into pathways of leukemogene-
heterogeneity associated with their disease. For example, sis and hierarchies of clonal evolution. In addition, combin-
20%-25% of patients with AML have fms-like tyrosine ing clinical outcomes with genetic profiling has started to
kinase 3 (FLT3) length mutations (inclusive of ITDs, inser- define new groups of patients with a favorable prognosis. For
tions, deletions), which are associated with inferior progno- example, patients with co-occurring IDH2 R140Q and
sis. In addition, heterozygous mutations in exon 12 of the NPM1 mutations appear to have outcomes as good as or bet-
nucleophosmin member 1 (NPM1) gene have been found in ter than patients with core binding factor leukemias. Finally,
40%-60% of AML patients with a normal karyotype, and identification of these mutations has served as a platform for
mutated NPM1, in conjunction with wild-type FLT3, is asso- the development of novel therapeutic inhibitors of the
ciated with a favorable prognosis. Finally, mutations of the mutated proteins and led to the ongoing efforts to develop
CCAAT–enhancer binding protein A (CEBPA), a gene clinical trials combining novel agents to target each genetic
encoding a myeloid transcription factor important for nor- alteration.
mal granulopoiesis, also appear to be associated with favor- Recent efforts to combine the information from cytoge-
able clinical outcomes. Certain mutations such as IDH1/ netics and molecular changes have been set forth by the
IDH2, KIT and FLT3 may have therapeutic implications, as European Leukemia Net (ELN) and have culminated in the
specific inhibitors are available or in development. evolution of the traditional risk groups into favorable, inter-
While evaluating for mutations in NPM1, FLT3, and mediate-1, intermediate-2, and adverse categories, which is a
CEBPA have become part of routine testing to aid in risk reminder that refining prognosis will continue to evolve as
stratification for patients with AML associated with a normal the impact of more targets is recognized (Table 18-2).
karyotype, a host of other molecular alterations including
mutations in genes defining epigenetic pathways such as Key points
DNMT3A, IDH1, IDH2, TET2, and others have been
described in many patients with AML. In addition, genetic • The most important prognostic indicators in AML are patient
profiling of patients with a normal karyotype has started to age, cytogenetics, and molecular genetics.
yield insights into distinct prognostic subgroups of patients • Complex cytogenetic abnormalities and monosomal karyo-
types are associated with poor clinical outcomes.
with various co-occurring mutations. The Eastern

cin and 7 days of cytarabine were more effective than 2 and

Key points (Continued ) 5 days, respectively, and that 10 days of cytarabine was not
• t(15;17), t(8;21), and inv(16) are cytogenetic abnormalities better than 7 days. Also, 100 mg/m2 of cytarabine for 7 days
associated with favorable outcomes. was as effective as 200 mg/m2 for the same duration. Dauno-
• Patients with cytogenetically normal AML and FLT3-ITD rubicin at a dose of 30 mg/m2 was inferior to 45 mg/m2, and
mutations have an unfavorable prognosis, whereas those with recently, daunorubicin 90 mg/m2 has been shown, in large
wild-type FLT3 and mutations of NMP1 or CEBPA have a more cooperative group trials, to be superior to 45 mg/m2 even in
favorable prognosis. selected patients >60 years of age. In a UK study, daunorubi-
cin 60 mg/m2 was equivalent to 90 mg/m2, establishing
Treatment 60 mg/m2 as an appropriate standard dose and 90 mg/m2 as
a reasonable alternative in patients with adequate cardiac
Treatment for AML generally is divided into remission status.
induction and postremission therapy. Standard remission Despite many attempts to improve outcomes from “7+3”
induction regimens in the United States for all AML sub- since it was first developed in the 1970s, no treatment regi-
types, excluding APL (see section “Acute promyelocytic leu- men has succeeded in replacing it as standard induction,
kemia” later in this chapter), almost always include 7 days of though this may be about to change. Recent efforts have
infusional cytarabine and 3 days of an anthracycline, com- focused on adding therapies targeting molecular mutations
monly known as the “7+3” or “3&7” strategy. This strategy to traditional induction and include FLT3 inhibitors, polo-
results in complete remission (CR) in 70%-80% of adults like kinase inhibitors, and agents blocking multidrug resis-
<60 years of age and 30%-50% of selected adults >60 years of tance gene overexpression without definitive evidence of
age with a good performance status. The Cancer and Leuke- bettering outcomes. Results for the targeted agent gemtu-
mia Group B (CALGB) established that 3 days of daunorubi- zumab ozogamicin (GO) have been conflicting to date, but a
recent meta-analysis exploring its potential benefit when
added to chemotherapy suggested that it can be safely added
Table 18-2 ELN standardized reporting system for correlation of
to chemotherapy and that it provides a survival advantage to
cytogenetic and molecular genetic data in AML with clinical data
patients without adverse prognostic factors. Unfortunately,
Genetic group Subsets GO remains unavailable in the United States as of 2015. Two
Favorable t(8;21)(q22;q22);RUNX1-RUNX1T1 studies have shown superiority for adding a kinase inhibitor
Inv(16)(p13.1q22) or t(16;16(p13.1q22; to 3&7: sorafenib extended relapse-free survival regardless of
CBFB-MYH11 FLT3 status, and in a large cooperative group study of
Mutated NPM1 without FLT3/ITD midostaurin plus 3&7 for FLT3 mutant AML, both event-
(normal karyotype)
free and overall survival were improved.
Mutated CEBPA (normal karyotype)
Once remission has been achieved, further therapy is
Intermediate-1 Mutated NPM1 and FLT3/ITD
required to prevent relapse. Options include repeated
(normal karyotype)
Wild-type NPM1 and FLT3/ITD
courses of consolidation chemotherapy or allogeneic hema-
(normal karyotype) topoietic stem cell transplantation (HSCT). Allogeneic
Wild-type NPM1 without FLT3/ITD HSCT allows combination of myeloablative or nonmyeloab-
(normal karyotype) lative chemotherapy with a graft-versus-leukemia effect
Intermediate-2 t(9;11)(p22;q23); MLLT3-MLL from the donor cells. Autologous HSCT is routinely not done
Cytogenetic abnormalities not otherwise for patients with AML. Several studies have prospectively
classified as favorable or adverse evaluated the role of intensive consolidation with HiDAC.
Adverse inv(3)(q21q26.2) or t(3;3)(q21;q26.2); The CALGB randomized patients in first remission to four
RPN1-EVI1 courses of cytarabine using either a continuous infusion of
t(6;9)(p22;q23); DEK-NUP214
100 mg/m2 for 5 days or a 3-hour infusion of 400 mg/m2
t(v;11)(v;q23); MLL rearrangement
or 3 g/m2 twice daily on days 1, 3, and 5. Significant CNS
-5 or del(5q)
toxicity was observed in patients >60 years old randomized
-7
Abnormal 17p
to the high-dose arm, and thus, this regimen is not recom-
Complex karyotype* mended for older patients. In patients <60 years old, there
*Defined as 3 or more chromosome abnormalities in the absence of was a significant improvement in disease-free survival asso-
one of the WHO designated recurring translocations of inversions: ciated with the high-dose regimen, and this was most pro-
t(8;21), inv(16) or t(16;16), t(15;17), t(9;11), t(v;11)(v;q23), t(6;9), nounced in patients with favorable cytogenetics, including
inv(3) or t(3;3). t(8;21) and inv(16).

Monitoring residual disease | 525
Although it has become standard to offer at least two cycles of alternative donor sources of stem cells to allow allogeneic
of HiDAC at 1 to 3 g/m2 to younger patients with AML, there transplant options for patients without fully matched sibling
are no clear data defining the optimal number or intensity of or unrelated donors. Trials utilizing partially matched-
HiDAC cycles. Randomized trials from the United Kingdom related donors, including haploidentical donors, as well as
Medical Research Council failed to demonstrate that three cord blood as sources of stem cells are under way by national
cycles of HiDAC consolidation were better than two cycles. cooperative transplant groups. Finally, using nonmyeloabla-
Although it is clear that patients with core binding factor tive or reduced-intensity conditioning regimens is another
(CBF) leukemias specifically benefit from HiDAC, some of way to broaden the application of allogeneic SCT toward
these patients also have mutations in KIT, which are associ- patients who may not be medically fit to undergo a full pre-
ated with an inferior outcome; clinical trials of chemother- parative regimen.
apy combined with tyrosine kinase inhibitors including
dasatinib (which inhibits KIT) are ongoing in these patients.
Consolidation chemotherapy, in general, has not been Key points
proven to be of benefit for patients >60 years old, but older
• Treatment of AML generally involves remission induction
patients able to tolerate additional treatment often are followed by postremission therapy.
offered modified dosing of bolus cytarabine or additional • The standard of care for induction for all AML subtypes in
courses of “7+3.” Maintenance therapy outside of APL has adults, excluding APL (FAB-M3), remains 3 days of an anthracy-
not been adopted. Two pediatric randomized trials from the cline combined with 7 days of cytarabine.
Leucémies Aiguës Myéloblastiques de l’Enfant (LAME) and • Consolidation chemotherapy with two to four cycles of HiDAC
the Children’s Cancer Group (CCG) failed to demonstrate is of particular benefit for patients <60 years old with favorable
that maintenance therapy improves outcomes. prognosis cytogenetics involving CBF [t(8;21) and inv(16)]; it is
Several studies of postremission therapy in AML have not routinely recommended for patients >60 years old.
compared intensive chemotherapy consolidation to HSCT • Allogeneic stem cell transplantation appears to be of benefit
for AML patients in first remission who have intermediate- or
by assigning younger patients with a human leukocyte anti-
poor-risk cytogenetics.
gen (HLA)-matched sibling donor to allogeneic HSCT and
• Retrospective data suggest that allogeneic transplantation may
randomizing other patients to chemotherapy or autologous
be of benefit for patients with FLT3-ITD+.
HSCT. Meta-analyses have shown that autologous HSCT
decreases relapse risk but increases treatment-related mor-
tality compared with chemotherapy consolidation, thus
resulting in similar overall survival rates of approximately Monitoring residual disease
40%-45% at 3-5 years. There is no specific indication for its
use in any prognostic subgroup, but it continues to be used Although morphologic methods remain the gold standard
in some settings, especially in Europe. for determining the status of disease in AML, more sensi-
Allogeneic HSCT is probably the most effective antileuke- tive immunologic and molecular methods for detecting the
mic therapy currently available and offers a combination of presence of minimal residual disease (MRD) are available.
the therapeutic efficacy of the conditioning regimen and the Leukemia-associated immunophenotypes can be identified
graft-versus-leukemia effect from the donor cells. It is, how- as “signatures” for some patients with AML, and the pres-
ever, associated with significant morbidity and mortality. A ence of MRD as measured by immunophenotype has been
recent comprehensive meta-analysis by Koreth et al. (2009) shown to predict for disease relapse in some studies. The
of prospective clinical trials of allogeneic HSCT in AML genetic abnormalities associated with a significant propor-
patients in first CR evaluated 24 trials and more than 6,000 tion of AML cases provide unique markers that can be used
patients. In this analysis, allogeneic HSCT resulted in signifi- to monitor MRD. Polymerase chain reaction (PCR) offers a
cantly improved 5-year overall survival, from 45% to 52% qualitative detection of abnormal gene rearrangements or
for patients with intermediate-risk cytogenetics and from fusion genes, whereas real-time PCR offers both the advan-
20% to 31% in patients with poor-risk cytogenetics. There tage of higher sensitivity and the possibility of quantifica-
was no benefit of allogeneic HSCT for patients with good- tion. MRD is now being used to risk stratify patients in
risk cytogenetics. Retrospective analyses of uniformly treated current clinical trials. The optimal sensitivity for detection
patients have shown that allogeneic HSCT was also benefi- methods remains to be determined, and prospective clini-
cial for cytogenetically normal AML patients with FLT3- cal trials are required to assess whether additional pos-
ITD+, FLT3-ITD–/NPM1–, and FLT3-ITD–/CEBPA–, and tremission treatment with chemotherapy, allogeneic
prospective trials using molecular and cytogenetic risk strat- transplantation, or other agents will improve outcomes for
ification are under way. Other efforts are focusing on the use patients with persistent MRD.

AML relapse a high frequency of poor prognostic features, including

antecedent hematologic disorders, unfavorable cytogenet-
The majority of adult patients with AML experience relapse ics, and multidrug resistance (MDR1) phenotypes. Also,
despite initially attaining CR. The prognosis of relapsed dis- older patients are often less able to tolerate intensive che-
ease is poor, and these patients should be considered for motherapy because of medical comorbidities, polyphar-
investigational trials. Most AML relapses occur within macy, poor performance status, and limited social supports.
2 years of diagnosis. The duration of first remission is of There is no universally accepted standard of care for the
critical prognostic importance, and patients with an initial treatment of older patients, but they generally are offered
CR of <6 months are unlikely to respond to standard chemo- either conventional “7+3” induction, hypomethylators,
therapeutic agents. Patients whose initial CR duration was repeated cycles of low-dose subcutaneous cytarabine, sup-
>12 months may have up to a 50% chance of responding to portive care with antibiotics and transfusions, hospice care,
a HiDAC-containing regimen, even if they had previous or an investigational trial. Although remission can be
exposure to this agent. Examples of reinduction regimens attained in ~50% of selected older patients with a good
include cytarabine, etoposide, and mitoxantrone (MEC), performance status using 7+3, these responses are offset by
fludarabine, high-dose cytarabine and granulcoyte colony- mortality rates of 5%-20% and their short duration, with
stimulating factor priming (FLAG), clorfarabine and cytara- <10% of elderly patients being long-term survivors.
bine, and cladribine, cytarabine, and growth factor (CLAG). The use of hypomethylating drugs (azacitidine and
No combination has proven more effective in the few ran- decitabine) has become a common treatment strategy for
domized trials attempted. Likewise, moving the noted com- elderly patients or patients who are deemed unfit for tradi-
binations into the frontline setting has not shown superior tional cytotoxic chemotherapy base on the findings that
outcomes to “7+3” induction. Patients who achieve a second these agents can result in bone marrow stabilization, reduc-
remission should be considered for standard or reduced- tion in transfusion needs, and even complete remissions in
intensity allogeneic transplantation if possible because the between 10%-20% of patients. Decitabine is now approved
duration of second remission with chemotherapy alone is for this indication in Europe and is often used off-label in
generally short and almost always shorter than CR1. The the United States; azacitidine is approved in the US for
prognosis for patients who relapse after allogeneic transplan- patients with AML who have 20-30% blasts. There is con-
tation is dismal. troversy about how to define “unfit” patients; previous
There are many categories of novel agents for AML as well studies used physician judgment, and more recent studies
as new combination strategies that are under investigation, have employed specific criteria for patients unlikely to ben-
including chemotherapeutics (eg, topoisomerase II inhibi- efit from intensive induction.
tors, purine nucleoside analogs), signal transduction inhibi- Major cooperative group or multicenter trials, which
tors (FLT3-ITD, PIM), hypomethylating agents and other generally have focused on patients <75 years old with de
epigenetic modifiers (HDAC, ASXL), cell cycle modulators novo AML and those having a good performance status,
(CDK, polo-like kinase), and agents targeting cellular show 3- to 5-year overall survival rates of only 10%-20%;
metabolism (IDH1/2). In general, with the possible excep- however, many of these patients are not offered any treat-
tion of the IDH inhibitors, targeted agents have had limited ment for AML despite randomized data clearly demon-
single-agent activity in relapsed AML and may be more strating a survival benefit favoring treatment with
effective in combination with chemotherapy. chemotherapy over supportive care in this population.
Clinical experience suggests that quality of life is better for
those who achieve CR, but data are sparse. Although there
Older patients with AML are clearly frail and debilitated older patients who cannot
tolerate any treatment, emerging data suggest that age
Clinical case alone should not be used as the major determinant of
treatment because several intensive options, including
An 82-year-old woman with a history of myocardial infarction, intensified doses of daunorubicin and reduced-intensity
diabetes, and peripheral vascular disease presents with shortness
stem cell transplantations, are both feasible and effective in
of breath and is found to be pancytopenic. Bone marrow biopsy
selected patients >60 years old. Many, if not most, older
shows 40% myeloblasts with monosomy 7.
patients with AML fail to benefit from therapy due to its
lack of therapeutic efficacy, not due to intolerable toxici-
Most patients with AML are >60 years old, and their prog- ties. Novel therapies are clearly needed for this population,
nosis is dismal, with median survival times of only 8-12 and there are many ongoing clinical trials with both single
months among the most “fit” patients. Older patients have agent and combinations using cytotoxics, antibodies,

Acute promyelocytic leukemia | 527
hypomethylating agents, and nonmyeloablative transplan-

tations. Older AML patients should be encouraged to par-
ticipate in clinical trials whenever possible.
Acute promyelocytic leukemia
Clinical case
A 23-year-old Hispanic female presents with 2 weeks of dyspnea,
bruising, and menorrhagia. Laboratory evaluation shows pancy-
topenia with elevated prothrombin and partial thromboplastin
times and markedly decreased fibrinogen. Bone marrow aspira-
tion shows intensely myeloperoxidase-positive promyelocytes
and t(15;17). Figure 18-2 A faggot cell. Source: ASH Image Bank/Peter Maslak.
APL (FAB-M3) is a clinically, cytogenetically, and prog- STAT5B-RARα fusions, are resistant to ATRA. It is crucial
nostically distinct subtype of AML that accounts for ~5%- that ATRA is started as soon as the diagnosis of APL is sus-
15% of all adult AML cases, with a higher incidence in pected, even before pathologic confirmation, as this agent
younger patients and among Hispanics. It is the most cur- reverses the coagulopathy common at presentation and is the
able form of AML in adults. Almost all leukemic cells from cornerstone of all treatment plans for APL.
patients with APL have a balanced reciprocal translocation Combination regimens with ATRA and an anthracycline
between chromosomes 15 and 17, which results in the (with or without cytarabine) induces remission in >90% of
fusion of the promyelocytic leukemia (PML) and retinoic patients, and long-term cures are achieved in >70%-80% of
acid receptor-α (RARα) genes, a PML-RARα fusion gene patients in many series. Primary resistance to chemotherapy
product, and disruption of normal differentiation. APL is virtually nonexistent. Arsenic trioxide may be the single
blasts contain granules with proteolytic enzymes, the most active agent in APL, and efforts to incorporate this
release of which induces severe coagulopathy and fibrino- agent early into treatment have confirmed its importance.
lysis, predisposing patients to both hemorrhage and In addition to offering a survival benefit when given as
thrombosis. consolidation for newly diagnosed patients (as opposed to
APL exists in hypergranular (typical) and microgranular cytotoxic chemotherapy-based consolidations), arsenic
forms. In hypergranular APL, the promyelocytes are strongly combined with ATRA produces high rates of durable CR in
myeloperoxidase positive and have bi-lobed or kidney- newly diagnosed patients with low-risk disease (less than
shaped nuclei. The cytoplasm has densely packed, large gran- 10,000 peripheral white cell count at presentation) with low
ules, and characteristic cells containing bundles of Auer rods rates of hematologic toxicity as compared to ATRA plus an
(faggot cells, named after a chiefly British English term for anthracycline. The ATRA/arsenic combination led to an
the bundle of sticks that these Auer rods resemble) may be enviable 100% complete remission rate, a 97% event-free
found in most cases (Figure 18-2). Cases of microgranular survival, and a 99% overall survival at 2 years. ATRA/
APL have predominantly bi-lobed nuclei, are strongly myelo- Arsenic is now the considered the standard of care for
peroxidase positive, and often have a very high leukocyte patients with low risk APL. Finally, early use of arsenic has
count and doubling time. APL is characterized by low expres- also been recognized as contributing to the elimination of
sion or absence of HLA-DR, CD34, CD117, and CD11b. The maintenance ATRA and chemotherapy in most lower-risk
diagnosis is confirmed with cytogenetics, reverse transcrip- APL patients.
tase PCR (RT-PCR) for the PML-RAR fusion transcript, and Higher-risk APL patients continue to receive combina-
fluorescence in situ hybridization with probes for PML and tion therapy with ATRA + chemotherapy, typically an
RARα, or anti-PML antibodies. anthracycline. There is some controversy regarding the best
APL promyelocytes have the unique ability to undergo chemotherapy to include with ATRA during induction, but
differentiation with exposure to all-trans retinoic acid an anthracycline alone appears to be sufficient, and either
(ATRA). Detection of t(15;17) or the underlying PML-RARα daunorubicin 60 mg/m2 for 3 days or idarubicin 12 mg/m2
rearrangement is predictive of response to ATRA in virtually on days 2, 4, 6, and 8 can be used. Consolidation
100% of cases. Some infrequent APL variants, such as protocols differ between the United States and European
t(11;17)(q23;q21) with ZBTB16-RARα and cases with cooperative groups but generally include several cycles of

anthracycline-based chemotherapy. The role of infusional Pediatric AML, including Down

cytarabine during induction is not clear, but patients pre- syndrome
senting with a white blood cell (WBC) count of ≥10,000/µL
may benefit from intermediate-dose cytarabine or HiDAC
Clinical case
during either induction or consolidation. Some protocols
for high-risk patients have also incorporated prophylactic A 6-year-old boy presents with a 4-week history of fatigue and
intrathecal chemotherapy. The role of maintenance therapy fever and a 1-week history of bruising and pallor. Laboratory
is also debated in APL, but ATRA with or without 6-mer- evaluation shows pancytopenia. Bone marrow aspiration shows
myeloblasts with granules and an occasional Auer rod. Cytoge-
captopurine and oral methotrexate frequently is offered to
netic studies reveal t(8;21).
patients in CR, particularly in patients not treated with
arsenic as part of their regimen. The optimal combination
and duration of maintenance have not been defined. Pediatric AML has unique clinical features, risk stratification
Despite the success of ATRA-based regimens, there is schemas, and therapeutic approaches. Cutaneous involve-
still an early mortality of ~10% in APL patients, primarily ment is more common in children, particularly in infants
because of hemorrhagic complications. Predictors of early diagnosed at <1 year of age. Poor-prognosis cytogenetics are
death resulting from hemorrhage include WBC count at less frequent in children, and within the pediatric spectrum,
presentation, abnormal creatinine, peripheral blast count, age is not a critical prognostic indicator, except for children
presence of coagulopathy, and age. Also, patients must be with Down syndrome. Children may tolerate intensive che-
closely monitored for the development of APL differentia- motherapy better than adults, and this may affect the opti-
tion syndrome, a potentially fatal constellation of findings, mal therapeutic approach. Standard induction chemotherapy
including interstitial pulmonary infiltrates, hypoxemia, in pediatrics typically includes cytarabine and an anthracy-
respiratory distress, fluid retention, weight gain, pleural or cline with the addition of a third agent, such as etoposide.
pericardial effusions, and, sometimes, renal failure. Rapid Most current pediatric AML protocols use at least four cycles
administration of dexamethasone 10 mg twice daily for at of chemotherapy with HiDAC-based consolidation. Autolo-
least 3 days at the earliest manifestations of the syndrome gous HSCT has been abandoned by most pediatric groups,
can be lifesaving. whereas the role of allogeneic HSCT is highly variable. In
The persistence or reappearance of PML-RARα fusion North America, most children with favorable features are
gene transcripts in patients with APL is highly predictive treated with chemotherapy alone, whereas most children
of clinical relapse, and frequent monitoring by RT-PCR with poor-risk features are offered allogeneic HSCT from
has been integrated into most clinical trials. The ideal either a related or unrelated donor. Children with favorable
monitoring approach is not clear, as most patients achiev- cytogenetics have an overall survival rate of ~70% irrespec-
ing a molecular remission will not relapse. However, tive of response to the first cycle of induction, whereas chil-
relapsed disease can be treated effectively with arsenic tri- dren with adverse cytogenetics or poor response to the first
oxide, which causes differentiation and apoptosis of APL cycle of induction therapy with >15% residual blasts have an
cells, alone or in combination with ATRA. In addition, overall survival rate of only 15%.
autologous stem cell transplantation can be considered for Children with Down syndrome have a 46- to 83-fold
patients in second remission if the stem cells are negative increased risk of AML and are generally younger than other
for PML-RARα. Allogeneic stem cell transplantation gen- pediatric AML patients. AML associated with Down syn-
erally is not recommended for patients with APL but may drome tends to be classified as FAB-M7 (acute megakaryo-
be considered for patients beyond first relapse. blastic leukemia [AMKL]), and acquired GATA1 mutations
have been described in the leukemic blasts. AMKL in Down
syndrome may be preceded by transient myeloproliferative
Key points disorder (TMD), a condition unique to children with Down
• APL is a unique subtype of AML that is exquisitely sensitive to syndrome. TMD is a clonal disorder characterized by circu-
ATRA, anthracyclines, and arsenic trioxide. lating blasts and dysplastic features and usually is diagnosed
• ATRA should be started immediately if the diagnosis of APL is in the first few weeks after birth. Although TMD typically
suspected. resolves spontaneously within the first 3 months, intensive
• Cure rates are high in APL. supportive care may be required, and early death has been
• APL may be complicated by a life-threatening coagulopathy or reported in as many as 15%-20% of cases. For those who
differentiation syndrome. survive, ~20%-30% will later develop AMKL. Children with
• APL differentiation syndrome should be treated promptly with Down syndrome and AML who are <2-4 years of age have
dexamethasone 10 mg twice daily for at least 3 days.
better prognosis compared with both non-Down syndrome

Bibliography | 529
AML and Down syndrome AML patients >4 years of age at groups have predictive value in secondary AML and in patients
diagnosis. This superior prognosis may be related to undergoing transplantation.
enhanced sensitivity of the leukemic blast to cytarabine. Koreth J, Schlenk R, Kopecky KJ, et al. Allogeneic stem cell
Children with Down syndrome have greater toxicities with transplantation for acute myeloid leukemia in first complete
remission. JAMA. 2009;301:2349-2361.
treatment and usually are not offered HSCT in first
Kottaridis PD, Gale RE, Frew ME, et al. The presence of a FLT3
remission.
internal tandem duplication in patients with acute myeloid
leukemia (AML) adds important prognostic information
to cytogenetic risk group and response to the first cycle of
Bibliography chemotherapy: analysis of 854 patients from the United
Kingdom Medical Research Council AML 10 and 12 trials.
Byrd JC, Dodge RK, Carroll A, et al. Patients with t(8;21)(q22;q22) Blood. 2001;98:1752-1759. Describes important prognostic value
and acute myeloid leukemia have superior failure-free and of FLT3-ITD in AML.
overall survival when repetitive cycles of high-dose cytarabine Lo-Coco F, Avvisati G, Vignetti M, et al. Retinoic acid and arsenic
are administered. J Clin Oncol. 1999;17:3767-3775. Shows that trioxide for acute promyelocytic leukemia. N Engl J Med.
failure-free survival and overall survival are improved in patients 2013;369:111-121. Randomized trial of nonchemotherapy based
with t(8;21) AML with repetitive cycles of HiDAC. treatment (ATRA + arsenic trioxide) vs traditional ATRA +
Döhner H, Estey EH, Amadori S, et al.; European LeukemiaNet. chemotherapy for newly diagnosed, low-risk APL showing similar
Diagnosis and management of acute myeloid leukemia outcomes and less hematologic toxicity. Longer follow-up shows
in adults: recommendations from an international expert outcomes better with ATRA + arsenic and this combination is now
panel, on behalf of the European LeukemiaNet. Blood. the standard treatment approach for patients with low risk APL.
2010;115:453-474. Mayer RJ, Davis RB, Schiffer CA, et al. Intensive post-remission
Elonen E, Almqvist A, Hanninen A, et al. Comparison between chemotherapy in adults with acute myeloid leukemia. N Engl
four and eight cycles of intensive chemotherapy in adult acute J Med. 1994;331:896-903. Randomized study of three different
myeloid leukemia: a randomized trial of the Finnish Leukemia cytarabine regimens in postremission therapy of AML. Benefit of
Group. Leukemia. 1998;12:1041-1048. Once in remission after HiDAC is limited to younger patients. HiDAC is suggested to be
two cycles of induction, six additional cycles were not better than harmful to older patients.
two cycles, with no difference in either relapse-free survival or Mrozek K, Marcucci G, Nicolet D, et al. Prognostic significance
overall survival. of the European LeukemiaNet standardization system for
Fenaux P, Chastang C, Chevret S, et al. A randomized comparison reporting cytogenetic and molecular alterations in adults with
of all trans-retinoic acid (ATRA) followed by chemotherapy acute myeloid leukemia. J Clin Oncol. 2012;20:4515-4523.
and ATRA plus chemotherapy and the role of maintenance Validation of the prognostic significance of the ELN’s combing
therapy in newly diagnosed acute promyelocytic leukemia. The both cytogenetic and molecular markers, now a standard tool to
European APL Group. Blood. 1999;94:1192-1200. Randomized prognosticate newly diagnosed AML.
study showing that concurrent ATRA plus chemotherapy is Patel JP, Gönen M, Figueroa ME, et, al. Prognostic relevance of
associated with lower relapse rate and better overall survival than integrated genetic profiling in acute myeloid leukemia. N Engl J
sequential ATRA until remission followed by chemotherapy. Also Med. 2012;366:1079-1089.
shows benefit of maintenance in APL. San Miguel JF, Vidriales MB, Lopez-Berges C, et al. Early
Fernandez HF, Sun Z, Yao X, et al. Anthracycline dose immunophenotypical evaluation of minimal residual disease in
intensification in acute myeloid leukemia. N Engl J Med. acute myeloid leukemia identifies different patient risk groups
2009;361:1249-1259. and may contribute to postinduction treatment stratification.
Grimwade D, Walker H, Oliver F, et al. The importance of Blood. 2001;98:1746-1751. Concludes that immunophenotypical
diagnostic cytogenetics on outcome in AML: analysis of evaluations to detect MRD in the CR marrow after induction have
1,612 patients entered into the MRC AML 10 trial. The important prognostic value.
Medical Research Council Adult and Children’s Leukaemia Swerdlow SH, Campo E, Harris NL, et al., eds. World Health
Working Parties. Blood. 1998;92:2322-2333. Provides large Organization Classification of Tumours of Haematopoietic and
data set correlating cytogenetics with outcome in patients with Lymphoid Tissues. Lyon, France: IARC; 2008.
AML. Provides evidence that cytogenetically defined prognostic

CHAPTER
19
and lymphoblastic lymphoma
Mark R. Litzow and Mats Heyman
Introduction, 531 Cytogenetics, 533 Treatment of ALL, 537
Classification and diagnosis of ALL, 531 Molecular genetics, 534 Late complications of therapy, 545
Immunophenotyping, 532 Prognostic factors, 535 Bibliography, 546
Introduction higher frequency of poor-risk prognostic factors based on

disease biology, comorbidities associated with older age
Acute lymphoblastic leukemia (ALL) is the most common that impair the ability to tolerate the intensive multiagent
leukemia in children (representing 23% of all pediatric can- chemotherapeutic regimens that have been used success-
cer diagnoses and 76% of leukemias among children <15 fully in children, differences in the treatment regimens used
years of age) but accounts for only 20% of adult acute leuke- by pediatric versus adult hematologists and medical oncol-
mia. Lymphoblastic lymphoma is rarer, representing 2% of ogists, and treatment adherence. The treatment and prog-
adult and 30% of pediatric non-Hodgkin lymphomas. These nosis of lymphoblastic lymphoma mirror those of its
entities are closely related biologically and clinically and may leukemic counterpart, but the distribution of immunophe-
share presenting features, although symptoms of bone mar- notypes and genetic aberrations as well as clinical prognos-
row failure are much more common in ALL. Otherwise typi- tic markers differs.
cal cases of lymphoblastic lymphoma with bone marrow
involvement exceeding 25% are classified as ALL.
The prognosis for both adult and especially childhood Classification and diagnosis of ALL
ALL has improved substantially in recent years with the use
of risk-directed combination induction-consolidation- The French-American-British (FAB) morphologic classifica-
continuation (maintenance) regimens that include central tion of ALL was based largely on morphology and contained
nervous system (CNS) prophylaxis. In children, treatment little prognostic or therapeutic information that might help
now results in complete remission (CR) rates of 97%-99%, to guide treatment choice. The World Health Organization
5-year event-free survival rates of 75%-87%, and 5-year (WHO) classification was revised in 2008 and has changed
survival rates of 90%-94%. The use of similar treatment the classification to reflect increased understanding of the
regimens in adults with ALL also has improved the progno- biology and molecular pathogenesis of the diseases. In addi-
sis, with CR rates of 65%-95% and 5-year survival rates of tion to discarding the FAB terms, the WHO classification
25%-60%, with more favorable results in younger adults divides these heterogeneous lymphoid diseases into two
than in older adults. The less favorable prognosis for adults major categories: precursor lymphoid neoplasms and mature
with ALL is related to several factors, including a much lymphoid neoplasms. The precursor lymphoid diseases
include both B-lymphoblastic leukemia/lymphoma and
T-lymphoblastic leukemia/lymphoma (Table 19-1). The new
classification further subdivides the precursor B-cell ALL
Conflict-of-interest disclosure: Dr. Litzow: research funding from cases by recurring molecular-cytogenetic abnormalities to
Amgen. Dr. Heyman: no competing financial interest.
Off-label drug use: Drs. Litzow and Heyman: blinatumomab, provide prognostic and therapeutic information and to facil-
dasatinib, ponatinib. itate the implementation of specific molecularly targeted
| 531

532 | Acute lymphoblastic leukemia and lymphoblastic lymphoma
Table 19-1 WHO classification of precursor lymphoid neoplasms between T-cell ALL, B-cell precursor (BCP) lymphoblastic
(B and T lymphoblastic leukemia/lymphoma) ALL, and mature B-cell ALL. Additionally, early T-cell pre-
B lymphoblastic leukemia/lymphoma, NOS cursor (ETP) and late T-cell maturation-stages have in some
B lymphoblastic leukemia/lymphoma with recurrent genetic therapeutic contexts been associated with adverse prognosis,
abnormalities but this has not so far been used for stratification by most
B lymphoblastic leukemia/lymphoma with t(9:22)(q34;q11.2); study groups.
BCR-ABL1 Typically, BCP ALL cases are terminal deoxynucleotidyl
B lymphoblastic leukemia/lymphoma with t(v;11q23); MLL transferase (TdT) positive, human leukocyte antigen (HLA)-
rearranged
DR positive, and almost always positive for CD19 and
B lymphoblastic leukemia/lymphoma with t(12;21)(p13;q22);
CD79a; CD10 and CD22 are positive in most cases. Infants
TEL-AML1 (ETV6-RUNX1)
with ALL, typically with genetic rearrangement of the MLL-
B lymphoblastic leukemia/lymphoma with hyperdiploidy
B lymphoblastic leukemia/lymphoma with hypodiploidy
gene usually lack CD10 expression. The lymphoblasts in
(hypodiploid ALL) T-cell ALL are TdT positive and most often express CD7 and
B lymphoblastic leukemia/lymphoma with t(5;14)(q31;q32); cytoplasmic CD3. ETP ALL has a unique immunologic
IL3-IGH marker and gene expression profile reminiscent of a double-
B lymphoblastic leukemia/lymphoma with t(1;19) (q23;p13.3); negative thymocyte that retains the ability to differentiate
E2A-PBX1(TCF3-PBX1) into T-cell and myeloid, but not B-cell, lineages. These cases
T lymphoblastic leukemia/lymphoma were initially associated with a dismal treatment outcome
with chemotherapy, but recent reports suggest that the
adverse outcome may be limited to a higher incidence of
therapies. Burkitt lymphoma/leukemia is the one subset of induction failure, whereas postinduction outcome may be
ALL that is classified as a mature B-lymphoid neoplasm. similar to non-ETP cases with intensive chemotherapy strat-
Examination of a bone marrow aspirate is important in ified according to MRD. Mature B-cell ALL, Burkitt ALL,
the diagnostic evaluation of suspected ALL because as many has a unique immunophenotype with expression of surface
as 10% of patients with ALL lack circulating blasts at the immunoglobulin, has a strong expression of CD20, is nega-
time of diagnosis and because bone marrow cells tend to be tive for TdT expression, and also has distinctive morpho-
better than blood cells for genetic studies. Fibrosis or tightly logic and cytogenetic features. These ALLs are associated
packed marrow occasionally can lead to difficulties with with chromosome translocations involving the MYC proto-
marrow aspiration that necessitate a biopsy to make the oncogene on chromosome 8, which encodes a transcription
diagnosis. In patients with marrow necrosis (<2% of cases), factor.
multiple marrow aspirations are sometimes needed to obtain The distribution of the immunophenotypic subsets differs
diagnostic tissue. slightly between adult and pediatric ALL. T-cell ALL accounts
for less than 10% of children below 10 and increases with age
during adolescence and constitutes approximately 25% of
Immunophenotyping adult ALL, though its incidence decreases again with increas-
ing age. Mature B-cell/Burkitt ALL accounts for ~2%-5% of
Because leukemic lymphoblasts lack specific morphologic adult and pediatric ALL cases, and BCP ALL accounts for the
and cytochemical features, immunophenotyping by flow remaining cases. There are also racial or ethnic differences in
cytometry is essential for diagnosis. Genetic analyses may the distribution, with T-cell ALL accounting for 10%-12% of
contribute additional information. A panel of antibodies is white and 25% of black children with ALL.
needed to establish the diagnosis and to distinguish among Myeloid-associated antigens may be expressed on otherwise
the different immunologic subclasses of leukemic cells. In typical lymphoblasts. The pattern of myeloid-associated anti-
general, the panel includes antibodies to at least one very gen expression is correlated with certain genetic features of blast
sensitive marker for each hematopoietic and lymphoid lin- cells. CD15, and CD65 are frequently expressed in ALL patients
eage (CD19 for B-lineage cells, CD7 for T-lineage cells, and with a rearranged MLL gene, and CD13 and CD33 are expressed
CD13 or CD33 for myeloid cells) and antibodies to a rela- in patients with the ETV6-RUNX1 (also known as TEL-AML1)
tively specific marker (cytoplasmic CD79a and CD22 for fusion as well as in Philadelphia chromosome-positive (Ph+)
B-lineage cells, cytoplasmic CD3 for T-lineage cells, and ALL. The presence of myeloid-associated antigens lacks prog-
CD20 and surface immunoglobulin for mature B cells). nostic significance but can be useful in immunologic monitor-
Although ALL can be classified according to the normal ing of patients for minimal residual leukemia. A summary of
sequential stages of normal T-cell and B-cell development, CD markers and specific immunophenotypic techniques and
most groups find it therapeutically useful only to distinguish findings in ALL is found in Chapter 10.

Cytogenetics | 533
Table 19-2 Clinical and biologic characteristics of selected genetic subtypes of ALL
Frequency (%) Estimated event-free survival (%)
Genetic abnormality Adult Pediatric Adult Pediatric Therapeutics
B-cell
Hyperdiploidy >50 chromosomes 6-7 23-29 30-50 at 5 years 80-90 at 5 years Antimetabolites
Hypodiploid <44 chromosomes 2 1 10-20 at 3 years 30-40 at 3 years
t(8;14); t(2;8); t(8;22); c-MYC 5 2 50-80 at 3 years 75-85 at 3 years Short-term intensive
overexpression multiagent chemotherapy
with rituximab
t(12;21)(p13;q22)/ETV6-RUNX1 fusion 0-3 20-25 Unknown 85-95 at 5 years Intensive asparaginase
t(1;19)(q23;p13.3)/TCF3-PBX1 fusion 2-3 4-5 40-70 at 3 years 85-90 at 5 years High-dose methotrexate
t(9;22)(q34;q11)/ 25-30 2-3 40-60 at 2 years 70 at 5 years (DFS) ABL1 tyrosine kinase
BCR-ABL1 fusion inhibitors (imatinib/
dasatinib)
t(4;11)(q21;q23)/MLL-AF4 fusion 3-7 2 10-20 at 3 years 30-40 at 5 years FLT3 inhibitors, DOT1L-
Inhibitors
BCR-ABL1–like/IKZF1 alterations Unknown 15-20 Unknown 40-50 at 5 years Tyrosine kinase/JAK2
inhibitors in some cases
iAMP21 Unknown 2 Unknown 60-70 at 5 years Intensive glucocorticoid,
asparaginase and
vincristine
T-cell
NOTCH1 mutations 60-70 50 ~50 at 4 years 90 at 5 years γ-secretase or other Notch
signaling pathway
inhibitors
HOX11 overexpression 30 7 70-80 at 3 years 90 at 5 years
HOX11L2 13 20 ~20 at 2 years ~45 at 5 years
t(9;9)(q34;q34)/NUP214-ABL1 fusion 5 4 Unknown Unknown ABL kinase inhibitors
(imatinib/dasatinib)
ALL = acute lymphoblastic leukemia; iAMP21 = intrachromosomal amplification of chromosome 21.
Cytogenetics According to the modal chromosomal number, ALL can

be classified into several ploidy subgroups. Hyperdiploidy
ALL arises from a lymphoid progenitor cell that has sus- (also known as “high hyperdiploid”), defined as 51-67 chro-
tained multiple specific genetic injuries that lead to malig- mosomes, is seen in approximately 25% of childhood cases
nant transformation and proliferation. Thus, genetic and in 6%-7% of adult cases and is associated with a favor-
classification of blast cells is expected to yield more relevant able prognosis in childhood ALL and in some studies of
biologic information than that obtained by other means. adult ALL. High hyperdiploid karyotype may be associated
More than 75% of adult and childhood cases can be readily with an increased cellular accumulation of methotrexate and
classified into prognostically or therapeutically relevant sub- its polyglutamates, an increased sensitivity to antimetabo-
groups based on the modal chromosome number (or DNA lites, and a marked propensity of these cells to undergo
content estimated by flow cytometry), and specific chromo- apoptosis. The outcome of hyperdiploid cases in adults is not
somal rearrangements. Table 19-2 lists selected cytogenetic comparable to the excellent outcome of childhood cases, and
and molecular genetic abnormalities that can be identified hyperdiploidy has lacked favorable prognosis in some adult
by conventional cytogenetic analysis and/or fluorescent in studies. By contrast, hypodiploidy with <44 chromosomes,
situ hybridization (FISH) with prognostic and therapeutic especially near haploidy (24-31 chromosomes) and low
relevance. Increasingly, as described in later sections, molec- hypodiploidy (32-39 chromosomes), is consistently associ-
ular techniques are Identifying new genetic abnormalities ated with an adverse prognosis in both children and adults
that have prognostic and therapeutic implications and that with ALL.
should be integrated with cytogenetic abnormalities in Flow cytometric determination of cellular DNA content is
patient management so that treatment strategies can be spe- a useful adjunct to cytogenetic analysis because it is auto-
cifically tailored to the different genetic subsets of ALL. mated, rapid, and inexpensive, and its measurements are not

affected by the mitotic index of the cell population; results targeted exome capture, and whole-genome sequencing) and
can be obtained in almost all cases. Flow cytometric studies epigenetic approaches, has identified many specific genetic
can sometimes identify a small but drug-resistant subpopu- alterations. As a result, virtually all patients with ALL can be
lation of near-haploid or low-hypodiploid cells that may classified according to specific genetic abnormality to date.
have been missed by standard cytogenetic analysis. These studies also provided insight into the complex interac-
Specific reciprocal translocations have important biologic tions of multiple genetic alterations in leukemogenesis and
and clinical significance. Some translocations can mobilize response to therapy. T-ALL cases can be classified into sev-
the promoter-enhancer element of the immunoglobulin eral distinct genetic subgroups that correspond to discrete
heavy- or light-chain gene or the T-cell antigen receptor β/γ T-cell development stages: HOX11L2, LYL1 plus LMO2,
or α/δ gene to sites adjacent to a variety of transcription fac- TAL1 plus LMO1 or LMO2, HOX11, and MLL-ENL. In addi-
tor genes and can result in the overexpression of the tran- tion, mutations in the IL7R, JAK1, and JAK3 genes activate
scription factor. An example of this type of translocation the IL7R/JAK-STAT pathway; and loss of the PTEN tumor
occurs in Burkitt ALL, in which the transcription factor MYC suppressor gene drives aberrant PI3K-AKT signaling. The
is translocated to the promoter-enhancer element of the genetic landscape of T-ALL also includes loss of transcrip-
immunoglobulin heavy- or light-chain and, consequently, is tion factors (eg, WT1, LEF1, RUNX1, ETV6, BCL11B), epi-
expressed aberrantly. More often, the genetic rearrange- genetic (eg, EZH2, SUZ12, and PHF6) tumor suppressors,
ments result from the fusion of two genes encoding different cell cycle inhibitors (eg, CDKN2A, RB, and CDKN1B), gains
transcription factors. These chimeric transcription factors of oncogenes (eg, MYB) and chromosomal rearrangements
may regulate genes involved in the differentiation, self- that can result in fusion products, including CALM-AF10,
renewal, proliferation, and drug resistance of hematopoietic MLL1-ENL, and NUP214-ABL1. Whereas HOX11L2 gener-
stem cells. Included in this group of translocations are those ally confers a poor outcome, HOX11 and MLL-ENL are asso-
involving the MLL gene on chromosome 11q23, the most ciated with a favorable outcome. However, the favorable
common of which is t(4;11), which results in the creation of benefit of the MLL-ENL fusion is mainly confined to patients
the MLL-AF4 (alias AFF1) fusion gene. Other fusion genes with T-cell ALL or who are 1 to 9 years of age. BCP-ALL cases
result in the aberrant activation of tyrosine kinases, which with the MLL-ENL fusion have a poorer prognosis.
play a critical role in pathogenesis of these diseases. An Among many other mutations in T-cell ALL, NOTCH1 or
important example of this type of translocation is the Phila- FBXW7 mutations are associated with a favorable prognosis
delphia chromosome, where the t(9;22) results in the BCR- in adult and childhood ALL; K-RAS, N-RAS, and PTEN
ABL1 fusion gene and causes constitutive activation of the mutations are associated with a poor prognosis; and the
ABL tyrosine kinase, which is directly linked to disease NUP214-ABL1 fusion is responsive to tyrosine kinase
pathogenesis. inhibition.
In B-lymphoblastic ALL with the Philadelphia chromo-
some and BCR-ABL1 fusion, IKZF1 is deleted in both pediat-
Molecular genetics ric and adult cases (75%-85%). A subgroup of Philadelphia
chromosome-negative B-cell precursor ALL with a gene
In addition to cytogenetic analysis, there are compelling rea- expression pattern similar to BCR-ABL1 fusion cases (BCR-
sons to perform molecular genetic studies in ALL. First, ABL1-like ALL) occurs in as many as 12%-14 % of children
molecular analyses can identify several important submicro- but in up to 27% of young adults with ALL. In a large pedi-
scopic genetic alterations not routinely visible by standard atric Dutch and German cohort 16% and 17% of B-cell pre-
karyotyping procedures, such as the ETV6-RUNX1 (also cursor ALL cases belonged to the Ph+-like group and the
known as TEL-AML1) fusion, intrachromosomal amplifica- IKZF1-deleted group respectively (14% and 15% of all
tion of chromosome 21, deletions of tumor suppressor cases). There is a large overlap: 40% of the patients in the
genes, and mutations of proto-oncogenes. Second, cases with BCR-ABL1-like group had IKZF1 deletions. In the same
clinically important genetic rearrangements can be missed Dutch/German study, about 10% of B-cell precursor patients
because of technical errors (eg, karyotyping residual normal had overexpression of the cytokine receptor–like factor 2
metaphase cells rather than leukemic metaphase cells). (CRLF2) gene resulting in supernormal JAK/STAT signaling.
Hence, FISH and reverse-transcriptase polymerase chain A variety of genetic lesions result in CRLF2 overexpression.
reaction (RT-PCR) assays are used frequently. Another recent detailed genomic analysis of 154 precursor
More recently, the application of microarray-based B-cell ALL cases (with whole genome sequencing study in 42
genomewide analysis of gene expression and DNA copy cases) identified structural alterations and mutations acti-
number, complemented by high-throughput sequencing vating kinase and cytokine receptor signaling in all cases
technologies (transcriptome sequencing [mRNA-seq], studied. Importantly, fusion transcripts that would be

Prognostic factors | 535
responsive to tyrosine kinase inhibitors (TKIs) were identi- also associated with prognosis but also genetic subgrouping
fied in many cases, including fusions involving ABL1, ABL2, has seen the same development over time, with generally
CSF1R, and PDGFR. In this study, increased CRLF2 expres- decreasing impact when risk-adapted therapy has been
sion was seen in 6%-15% of children and young adults with implemented. Because of the impact of therapy and the large
precursor B-cell ALL and, remarkably, in ~50% of the cases number of prognostic factors, which sometimes covary and
with Down syndrome. Rearrangements of EPOR and JAK2 interact with each other, all current protocols have more or
may be sensitive to ruxolitinib. less complicated algorithms for stratification taking several
Among children with hypodiploid ALL, near-haploid ALL of these factors into account.
cases frequently have alterations targeting receptor tyrosine Of the many variables that influence prognosis, genetic
kinase signaling and Ras signaling (71%), and low-hypodip- subtype, initial white blood cell (WBC) count, immunophe-
loid cases are characterized by alterations in TP53 (91%) that notype, age at diagnosis, and evidence of early treatment
commonly present in normal cells and may be inherited. response are the most important. Table 19-3 lists some prog-
Both of these two hypodiploid subtypes are responsive to nostic factors that may be used for risk stratification and/or
PI3K inhibitors in preclinical models. risk-adapted therapy in current clinical protocols.
Recent genome-wide association studies have identified
germ line single-nucleotide polymorphisms of several genes
Clinical prognostic factors
that are strongly associated with ALL susceptibility, racial
disparities in the incidence, and treatment outcome, includ- Children aged 1-9 years have a better outcome than either
ing the incidence of bone osteonecrosis due to corticoste- infants or adolescents, who, in turn, fare significantly better
roids. Hence, inherited genetic variations not only affect the than adults with ALL. How much of this age-dependent dif-
development of ALL but also the response to treatment. ference in outcome between infants, children, and adults
that depends on leukemia biology and how much depends
on differences in administered therapy is unclear. Among
Prognostic factors adults, the outcome of therapy worsens with increasing age.
Leukocyte count is a continuous variable, with increasing
The most important prognostic factor in ALL is the therapy counts conferring a poorer outcome. In childhood ALL,
administered, which means that most risk factors as we know there is a general agreement to use a presenting age between
them are valid only in the context of a particular therapy. 1 and 9 years and a leukocyte count of <50 × 109/L as mini-
Many clinical prognostic factors were identified early on, mal criteria for low-risk B-lymphoblastic ALL; age and
when patients were first cured of ALL with considerably less leukocyte count have little prognostic value in T-cell ALL. In
intensive therapy than currently used. Some of these factors adult ALL, age <35 years and leukocyte count <30 × 109/L are
have lost their independent prognostic significance with considered favorable prognostic indicators, and leukocyte
more intensive therapy. Genetic subgroups have further count >100 × 109/L is considered a poor prognostic feature
refined the stratification systems, since many of these are for T-cell ALL in some protocols.
Table 19-3 Prognostic factors used for risk stratification

Prognostic factors Favorable Adverse
Adult
Age (years) <35 >60
Leukocyte count (× 109/L) <30 for B-cell >100 for T-cell
Immunophenotype Thymic T-ALL Early T-cell precursor (in some studies)
Genotype High hyperdiploid (in some studies) BCR-ABL1; MLL-rearrangement
Hypodiploidy <44
Minimal residual disease after induction Low/absent High
Pediatric
Age (years) 1 to 9 <1 or >10
Leukocyte count (× 109/L) <50 >50
Immunophenotype B-lymphoblastic T-cell
Genotype Hyperdiploidy >50; ETV6-RUNX1 Hypodiploidy <44; MLL-AF4, IKZF1
deletions or mutations
Minimal residual disease after induction <0.01% >1%

Cytogenetic and molecular genetic arrays showing similar expression-patterns to Ph+ cases but
prognostic factors lacking the t(9;22) translocation (BCR-ABL1-like ALL) has
been shown to have a poor prognosis. This is also true for
The prognostic impact of age and, to a lesser extent, leukocyte
patients only defined by a deletion/mutation of the IKZF1-
count can be explained by their association with specific
gene coding for the developmentally important transcrip-
genetic abnormalities. For example, there is a preponderance
tion factor IKAROS in leukemic cells. The prognostic
of cases with favorable genetic abnormalities of hyperdip-
importance of CRLF-2 overexpression (as well as the under-
loidy >50 chromosomes or ETV6-RUNX1 in patients aged
lying genetic lesions) is less clear than for the Ph+-like and
1-9 years. ETV6-RUNX1 mutations are very rare in adults,
IKZF1-deleted groups, although cases with rearrangements
whereas the adverse genetic abnormality MLL rearrangement
of CRLF-2 have been associated with a worse prognosis. Nev-
occurs in 70%-80% of infant cases. The t(9;22) translocation
ertheless, it has attracted attention due to the potential thera-
resulting in a Philadelphia chromosome (Ph+) occurs in
peutic target offered by JAK/STAT overactivation. All these
25%-30% of adult patients, and the incidence of Ph+ ALL
groups are enriched in the group of patients of B-cell precur-
increases with increasing age. Approximately 50% of cases of
sor immunophenotype and lacking the typical genetic
B-lymphoblastic ALL in patients >60 years of age are Ph+.
lesions of prognostic importance (ie, patients with “other”
Although many genetic abnormalities are associated with
clonal abnormalities or normal karyotype and negative
clinical outcome (Table 19-3), there is no consensus on the
specific genotypes used for treatment stratification. The directed genetic analyses). These are sometimes referred to as
Children’s Oncology Group uses trisomy of chromosomes “B-other.”
4, 10, and 17 (triple trisomy) as a favorable prognostic factor
in its clinical trials. Some groups, but not others, use the Minimal residual disease detection
t(1;19) translocation and the recently discovered iamp(21)
aberration for stratification to higher risk therapy. There is The response to initial therapy is a very important prognos-
also clinical heterogeneity within each specific genetic sub- tic marker. This measure accounts for the combination of
type. Some groups have shown that combining genetic drug sensitivity or resistance of leukemic cells and the phar-
alterations with cytogenetics and minimal residual disease macodynamics of the drugs, which are affected by the phar-
detection (see below) are powerful tools to risk stratify macogenetics of the host. Initially, this was assessed as the
patients and are supplanting traditional risk factors. In rate of clearance of leukemic cells from the blood or bone
patients with MLL-AF4 fusion, infants and adults have a marrow on morphological examination Flow cytometric
worse prognosis than children. The basis of these differences profiling of aberrant immunophenotypes and PCR amplifi-
may be related to some combination of secondary genetic cation of fusion transcripts or immunoglobulin and T-cell
events, the developmental stage of the target cell undergoing receptor gene rearrangements genes, which are at least 100-
malignant transformation, and the pharmacogenetic or fold more sensitive than conventional morphologic determi-
pharmacokinetic features of the patient. In addition to nations, have replaced morphologic assessment and allows
t(4;11), the MLL gene is involved in a number of other minimal residual disease (MRD) to be detected at very low
translocations in ALL (eg, t[11;19], t[9;11]). Because of levels (0.01% or lower), providing a useful means to identify
their poor prognosis, allogeneic transplantation in first CR patients at very low or high risk of relapse. The main limita-
(CR1) is recommended for adult patients with translocation of flow cytometry is the requirement for data
tions involving the MLL gene. interpretation.
The clinical significance of many prognostic factors The choice of MRD assessment methodology can also
changes with improvements in treatment. For example, the depend on the facilities and expertise that are available. The
outcome for patients with Ph+ ALL has improved substan- vast majority of ALL patients have leukemic blasts with immu-
tially with the addition of TKIs to treatment, which is nophenotypic characteristics distinct enough to allow for their
described in detail later. In fact, children with this genotype reliable recognition by flow cytometry when present at very
who are treated with intensive chemotherapy and imatinib low level among an overwhelming number of normal cells.
have a 5-year disease-free survival in the order of 70%, and This high degree of applicability of flow cytometry is an
the outcome seems to be similar for patients treated with important factor in the selection of MRD assessment tech-
chemotherapy and allogeneic stem-cell transplantation. nique, especially since many facilities have expertise in flow
Another study confirmed that imatinib in combination with cytometry. Allele-specific oligonucleotide (ASO) real-time
intensive chemotherapy is well tolerated and beneficial for quantitative PCR analysis of rearranged immunoglobulin (Ig)
children with Ph+ ALL. and T-cell receptor (TCR) genes requires the development of
As mentioned above, a group of patients originally identi- patient-specific probes, which is laborious and expensive and/
fied in unsupervised cluster experiments from expression or the presence of fusion transcripts. Furthermore, ASO-PCR

Treatment of ALL | 537
is limited to Ig/TCR gene rearrangements with junctional encountered at diagnosis, sometimes even before chemo-
regions that yield sufficient sensitivity. MRD detection based therapy is initiated, especially in patients with high leukemic
on PCR amplification of fusion transcripts is in most cases cell burden and those with T-cell or mature B-cell ALL.
both highly sensitive and specific but is limited to the relatively Patients should be given hyper-hydration with intravenous
few cases with an identified gene fusion. Both MLL-rearrange- fluids (3,000-4,500 mL/m2/day) to maintain diuresis and
ments and Philadelphia-positive ALL may routinely be moni- dilute harmful metabolites. If impaired kidney function is
tored by this technique. manifest rasburicase (recombinant urate oxidase) should be
MRD can be measured by these two techniques in most given to patients at high risk of tumor lysis syndrome to treat
patients and has become a crucial factor for risk stratification or prevent hyperuricemia, whereas allopurinol may be suffi-
in childhood ALL. It is mainly risk stratification according to cient if urate is moderately elevated and the risk of tumor
MRD (most often at the end of induction) that has led to the lysis is lower. With rasburicase available, there is no indica-
reduced prognostic impact of clinical and genetic factors. tion for alkalinization of urine. A phosphate binder, such as
A new deep-sequencing based method has been developed, aluminum hydroxide, lanthanum carbonate, or sevelamer,
which is potentially even more sensitive and more robustly should be given to treat or prevent hyperphosphatemia. Cal-
reproducible. It is also likely to be less sensitive to multi- cium acetate or carbonate may be used if the serum calcium
clonality and can probably be used in almost all cases. This concentration is low, but is seldom necessary if no alkali is
method may replace flow cytometry and PCR eventually, administered.
particularly in cases where the other methods cannot be Infections are common in febrile patients with newly
used, but it remains to be investigated whether the increased diagnosed ALL as well as in those who are already receiving
sensitivity translates into grounds for altered stratification. therapy. Therefore, any patient with ALL who presents with
Patients who achieve molecular or immunologic remis- fever, especially those with neutropenia, should be given
sion (<.01% of aberrant cells) after conventional remission broad-spectrum antibiotics until infection is excluded. Usu-
induction have excellent outcomes and can continue treat- ally, all patients with ALL are given either trimethoprim-sul-
ment for standard (or low-risk) ALL. Patients with high lev- famethoxazole, atovaquone, dapsone, or inhaled pentamidine
els of MRD after remission induction or those who fail to as prophylactic therapy for Pneumocystis jirovecii pneumo-
achieve clinical remission (>5% leukemic cells in bone mar- nia. Some pediatric and many adult trials also recommend
row) may become candidates for allo-HSCT, as are patients some form of antibacterial, antiviral, and antifungal prophy-
with persistent levels of MRD at later time-points. For exam- laxis in patients with severe leukopenia during the intensive
ple, in some protocols, patients with MRD at the end of con- phases of treatment, but practices are not uniform.
solidation therapy are selected for allo-HSCT. MRD detection The use of hematopoietic growth factors for adults with
after achievement of morphologic remission in adults with ALL has been found to be safe and in some studies has
ALL has been associated with a significantly worse prognosis. reduced the number of induction deaths. All blood products
Measurements of the level of fusion transcripts, usually in should be irradiated prior to SCT to prevent alloimmuniza-
blood, may be more sensitive than the alternative methods tion. Other important supportive care measures include the
for MRD measurement. use of indwelling catheters, amelioration of nausea and vom-
MRD measurement is now used to improve risk stratifica- iting, pain control, and continuous psychosocial support for
tion and to allocate patients to allo-HSCT in most pediatric the patient and family.
trials and in some adult clinical trials. MRD detection, either
before or after allo-HSCT, in adults with Ph+ ALL has been
Treatment of Burkitt lymphoma/leukemia
associated with a lower disease-free survival (DFS). Finally,
in children and adults
MRD level is, at least in some protocols, a strong predictor of
treatment outcome at the time of second remission and The outcome for both children and adults with Burkitt lym-
before allo-HSCT for relapsed leukemia in both pediatric phoma/leukemia has improved dramatically during the past
and adult ALL. decades. The improved outcomes have resulted from the use
of fractionated high doses of alkylating agents, such as cyclo-
phosphamide or ifosfamide, with high-dose methotrexate.
Treatment of ALL These agents are combined with vincristine, an anthracycline
(doxorubicin or daunorubicin), and high-dose cytarabine
Supportive care
and administered in rapid succession over 4-6 months.
Optimal management of patients with ALL requires careful To reduce the large tumor bulk often present at diagnosis
attention to supportive care. Hyperuricemia and hyperphos- and to limit the severity of tumor lysis syndrome, a “reduc-
phatemia with secondary hypocalcemia are frequently tion” phase consisting of a week of glucocorticoid treatment

and a dose of vincristine and cyclophosphamide before inserted before maintenance, at least for medium risk
intensive chemotherapy has often been incorporated into patients. Higher-risk patients are in some protocols sub-
treatment regimens. Because of an extremely high predispo- jected to intensive block-therapy after induction before con-
sition to CNS involvement in Burkitt lymphoma/leukemia, tinued therapy with either continuous more standard
intensive CNS-directed therapy is given with high doses of elements or allogeneic stem-cell transplant for selected sub-
systemically administered cytarabine and methotrexate as groups. CNS-directed therapy is started early and is given for
well as intrathecal administration with these agents in com- different lengths of time, depending on the patient’s risk of
bination with hydrocortisone. CNS irradiation is typically relapse and the intensity of the primary systemic treatment.
omitted and reserved for adult patients with overt CNS dis-
ease. Recurrence after the first year rarely, if ever, occurs;
Remission induction
therefore, maintenance (continuation) therapy has not been
shown to be beneficial and is not recommended. Rates of CR range from 97% to 99% with contemporary
Using this aggressive approach, the survival for these chemotherapy. The induction regimen usually contains
patients has ranged from 50% to 60% in many adult series to three or four drugs, typically a glucocorticoid (prednisone,
>80% in pediatric series. Because the lymphoblasts in mature prednisolone, or dexamethasone), vincristine, and either
B-cell ALL exhibit strong expression of CD20, ongoing stud- asparaginase or an anthracycline. Four-drug inductions
ies have incorporated the anti-CD20 monoclonal antibody commonly include all these drugs from the beginning,
rituximab into frontline regimens in an attempt to further sometimes with the addition of cyclophosphamide for
improve outcome. The data from these trials appear very higher-risk patients. The intensive chemotherapy is in some
promising, with survival rates of approximately 80% in protocols preceded by a prephase of single corticosteroid to
adults. The addition of rituximab to frontline therapies for reduce the leukemic cell burden. The response to this pre-
Burkitt lymphoma/leukemia has also is been tested in chil- phase is also used for stratification. The efficacy of predni-
dren with some promise of further Improvement. Prospec- sone and dexamethasone is dose dependent. Although both
tive studies with randomized questions concerning the use drugs yielded comparable results when given in equivalent
of rituximab in children are planned. Diagnosis and treat- doses, dexamethasone still appears to yield improved CNS
ment of Burkitt lymphoma are also discussed in Chapter 21, control and is used preferentially in postremission therapy
in the context of aggressive non-Hodgkin lymphomas. in current clinical trials. However, if dexamethasone is used
at higher doses (10 mg/m2/d) in induction, this intensifica-
tion has in some studies offset the reduced relapse-rate by
Treatment of B-precursor ALL and T-ALL in
an increase in induction deaths and deaths in remission. Of
children
the various anthracyclines given to patients with ALL, none
Although risk-directed therapy is a standard therapeutic has proved superior to any other; however, daunorubicin is
strategy for childhood ALL, there is no consensus on the risk used most commonly.
criteria and the terminology for defining prognostic sub- The pharmacodynamics of asparaginase differ by formu-
groups. Because some of the prognostic factors are present at lation, and in terms of leukemic control, the dose-intensity
diagnosis or are the result of genetic investigations, the and duration of asparaginase treatment (ie, the amount of
results of which become known during the first weeks of asparagine depletion) are far more important than the type
therapy, whereas other important stratifying factors (early of asparaginase used. Because of the lower immunogenic-
response and MRD) are only known after evaluation of ini- ity, less frequent dosing, and feasibility in intravenous
tial therapy, all current protocols include a more or less com- administration of PEG-asparaginase (a polyethylene glycol
plex stratification system. The final risk-groups are frequently form of the Escherichia coli asparaginase), compared with
identified a few months into the therapy. Usually, childhood the native product, PEG-asparaginase has replaced native
ALL cases are divided into low- (standard-) risk, high- E. coli asparaginase as the first-line treatment in most pro-
(intermediate- or average-) risk, and very-high-risk groups, tocols, but availability of the pegylated product is a limiting
although the US Children’s Oncology Group advocates four factor in some countries. As a result, native E. coli asparagi-
categories, including a very-low-risk group. Infants are often nase has not been commercially available in the United
treated with separate regimens as are children with Ph+ ALL States after 2012.
after the introduction of tyrosine-kinase inhibitors. Treat- Immunoreactivity against asparaginase is a significant
ment for lower-risk-groups typically consists of a remission problem and may cause allergic reactions as well as silent
induction phase, an intensification (consolidation) phase, inactivation of the drug. Most major allergic reactions to
and prolonged continuation (maintenance) therapy to erad- both native and pegylated asparaginase seem to be associated
icate residual disease. A delayed intensification phase is often with inactivation, but not all antibody formation causes

inactivation of asparaginase activity. However, given the prognosis in the past. With improved outcome, both boys
potential for inactivation, all patients with significant allergic and girls are now treated with the same duration of 2 to 2.5
reactions should be treated with the alternative product years of continuation therapy in most, but not all, clinical
derived from Erwinia chrysanthemi. Monitoring of asparagi- trials. The administration of methotrexate and mercaptopu-
nase activity with subsequent possibility of dose-adjustment rine to the limits of tolerance (as indicated by low leukocyte
is used in some protocols, both to avoid over-treatment and counts) has been associated with improved clinical outcome.
to detect silent inactivation, which would also indicate a Many investigators advocate that the drug dosage be adjusted
change in product used. Such individualized dosing of to maintain leukocyte counts <3 × 109/L and neutrophil
asparaginase based on serum asparaginase activity may counts between 0.5 × 109 and 1.5 × 109/L to ensure adequate
improve outcome, but these assays are not used universally. dose-intensity during the continuation treatment in child-
In one randomized study of rituximab in CD20+ ALL, an hood ALL, yet not induce excessive myelosuppression. In
unexpected finding was reduced adverse effects from aspara- one study of maintenance therapy, the dose intensity of mer-
ginase therapy in the patients receiving rituximab, probably captopurine was the most important pharmacologic factor
because of reduced synthesis of antibodies against the drug. influencing treatment outcome. Overzealous use of mercap-
topurine is counterproductive, however, resulting in inter-
ruption of chemotherapy because of neutropenia and
Consolidation and delayed Intensification therapy
reduction of overall dose intensity. Furthermore, the exten-
At the end of induction, most protocols have a point of sion of the maintenance phase has been associated with the
response evaluation and stratification. In some protocols, development of secondary MDS and AML, which may be the
risk-adapted therapy diverges, whereas some protocols have true limiting factor for optimizing the use of this element.
a common start of post induction therapy to allow for MRD- There is no strong evidence for a difference in outcome if
evaluation. Post induction therapy starts with consolidation the mercaptopurine is taken in the evening or the morning,
therapy. Although there is no dispute on the importance of but it is recommended to be taken daily at a fixed time-point
this treatment, there is no consensus on the best regimen and to facilitate compliance and should for pharmacokinetic rea-
duration of treatment. Many protocols continue with a ther- sons not be taken with milk or milk products. The coadmin-
apy element developed by the Berlin-Frankfurt-Münster istration of food does not seem to influence the outcome if
consortium (BFM) with cyclophosphamide, 6-mercaptopu- the therapy is titrated to adequate myelosuppression.
rine, and repeated 4-day blocks of injections of cytarabine, Although methotrexate is used orally in most clinical trials,
whereas other regimens include high-dose methotrexate parenteral administration could circumvent problems of
with mercaptopurine or regimens based on a lower dose of decreased bioavailability and poor treatment adherence,
methotrexate. Patients with poor response to therapy are in especially in adolescents. Antimetabolite treatment should
some protocols shifted to more intensive, block-based ther- not be withheld because of isolated increases of liver
apy. Delayed intensification (or reinduction), also first intro- enzymes; such liver toxicity is tolerable and reversible.
duced by the BFM, is a widely used approach consisting of a A few patients (1 in 300) have an inherited homozygous
repetition of therapy similar to the first remission induction deficiency of thiopurine S-methyltransferase, the enzyme
therapy 3 months after the end of remission induction. This that catalyzes the S-methylation (inactivation) of mercapto-
element has been repeated (double delayed intensification) purine. Mercaptopurine should be reduced markedly (eg,
in studies with somewhat conflicting results, probably tenfold reduction) in these patients to avoid potentially fatal
reflecting the treatment stratification and the context of the hematologic toxicity. Approximately 10% of patients are
therapy. Extended asparaginase therapy starting early during heterozygous for the enzyme deficiency and have intermedi-
postinduction therapy or in continuation from induction ate levels of thiopurine methyltransferase. This subgroup
and used throughout or during parts these therapy-elements can be treated safely with only moderate reductions in mer-
has received increasing attention and is under study in a ran- captopurine dosage and appears to have better clinical out-
domized fashion in several protocols. comes than patients with the homozygous wild-type
phenotype. Importantly, patients with this enzyme defi-
ciency are at higher risk for therapy-related leukemia.
Maintenance (continuation) therapy
Whether dose reduction of mercaptopurine can reduce the
A combination of methotrexate administered weekly and risk of therapy-related leukemia in these patients is unknown,
mercaptopurine administered daily constitutes the usual since it seems that the duration of the therapy is of impor-
continuation regimen for ALL. In some protocols, boys have tance. Although thioguanine is more potent than mercapto-
been treated with a longer duration of continuation therapy purine, leads to higher concentrations of thioguanine
than girls because male sex has been associated with a poorer nucleotides in cells and cytotoxic concentrations in

cerebrospinal fluid, and produces superior antileukemic A recent meta-analysis of T-cell ALL showed no conclusive
response, its prolonged use has been associated with pro- evidence to suggest that treatment strategies including CNS-
found thrombocytopenia, an increased risk of death, and an irradiation (either prophylactic for all or only for risk-groups
unacceptable rate of hepatic veno-occlusive disease. There- or patients with frank CNS-involvement) had better out-
fore, mercaptopurine remains the drug of choice for ALL. comes than therapies completely omitting irradiation, which
The Clinical Pharmacogenetics Implementation Consor- emphasizes the importance of systemic and intrathecal ther-
tium has developed guidelines for thiopurine therapy apy also for patients at highest baseline risk.
(updates at http://www.pharmgkb.org) based on the associa- Systemic treatment, including high-dose methotrexate,
tion between clinical effects and phenotype or genotype of intensive asparaginase, dexamethasone, and optimal intra-
the thiopurine methyltransferase. thecal therapy, is important to control CNS leukemia. Triple
Intermittent pulses of vincristine and a glucocorticoid intrathecal therapy with methotrexate, cytarabine, and
have improved the efficacy of antimetabolite-based continu- hydrocortisone is more effective than intrathecal methotrex-
ation regimens and have been adopted widely in the treat- ate alone in preventing CNS relapse but may not improve
ment of childhood ALL. In a large intergroup randomized the overall outcome. A meta-analysis showed that adding
trial featuring intensive reinduction, the addition of six intravenous methotrexate for patients treated with triple
pulses of vincristine and dexamethasone during early con- intrathecal therapy improves outcome by reducing both
tinuation treatment failed to improve the outcome for chil- CNS and non-CNS relapses. Because the presence of ALL
dren with intermediate-risk ALL but has also been shown to blasts in the cerebrospinal fluid, even from traumatic lumbar
be of benefit in some recent studies. Thus, whether this pulse puncture, has been associated with an increased risk of CNS
therapy is necessary in contemporary regimens featuring relapse and poor event-free survival, special precaution
early intensification of therapy is unclear. should be taken to decrease the rate of traumatic lumbar
puncture (eg, transfusion to increase platelet count to ≥50 ×
109/L for initial intrathecal treatment, having the most expe-
CNS-directed treatment
rienced clinician perform the procedure with the patient
Prophylactic cranial irradiation, once a standard treatment, under deep sedation or general anesthesia), and intrathecal
is being replaced by intrathecal and systemic chemotherapy therapy should be intensified in patients with blasts in the
to reduce radiation-associated late complications. Two early CSF even if this is due to a traumatic lumbar puncture.
clinical trials tested the feasibility of complete omission of Patients should remain in a prone position for at least
prophylactic cranial irradiation from treatment. Although 30 minutes after the procedure to enhance the distribution
the cumulative risks of an isolated CNS relapse were rela- of the chemotherapy within the CSF and to avoid post-spinal
tively low (4% and 3%), the event-free survival rates were headache.
only 68.4% and 60.7%. Since then, systemic therapy has
improved markedly, and several studies have published
Stem cell transplantation
comparable results for patients treated without cranial irra-
diation. Other protocols have omitted irradiation for The indications for hematopoietic stem cell transplantation
increasing fractions of the patients but still prescribe irradia- (HSCT) during first remission should be reviewed continu-
tion for higher-risk patients. A radiation dose of 12 Gy ously as treatment improves and new agents become avail-
appeared to provide adequate protection against CNS able. These are detailed in Chapter 14. At this time, very poor
relapse, even in high-risk patients (eg, those with T-cell ALL early response to remission induction treatment, possibly
and leukocyte counts >100 × 109/L). Two recent studies, with the exception of patients aged 1-6 years with favorable
however, tested the feasibility of total omission of prophy- leukemic cell genetics, is the most frequent indication for
lactic cranial irradiation, even in patients with T-cell ALL, transplantation, followed by patients with remaining detect-
hyperleukocytosis, or overt CNS leukemia at diagnosis. In able MRD at high level after consolidation. Few convincing
these two studies, the 5-year survival rates were 85.6% and results indicate that cytogenetic changes only (without tak-
81%, and the cumulative risks of an isolated CNS relapse ing response to therapy into consideration) should indicate
were only 2.7% and 2.6%, respectively. Importantly, all 11 HSCT in first remission. Except in some small studies, trans-
patients with isolated CNS relapse in the first study remained plantation failed to improve the outcome of infant patients
in second remission for 0.4 to 5.5 years. with MLL rearrangement. Hypodiploid cases did not appear
These promising results suggest that prophylactic cranial to benefit from transplantation, but the number of patients
irradiation can be omitted safely in all patients in the context treated with this modality was very small.
of effective intrathecal and systemic chemotherapy and sev- The use of imatinib has dramatically improved early treat-
eral protocols without irradiation are currently recruiting. ment results in children with BCR-ABL1 positive ALL,

including those with poor early response to chemotherapy, MRD in the marrow. CNS relapses are associated with a
raising the question of whether transplantation should be higher level of MRD in the bone marrow than testicular
performed in first remission even in children with this sub- relapses. Importantly, submicroscopic bone marrow involve-
type of ALL. An ongoing international study of BCR-ABL1 ment at a level of 0.01% (10−4) or higher by PCR at the time
positive ALL is testing whether the intensity of chemother- of overt extramedullary relapse confers a very poor outcome.
apy can be reduced and cranial irradiation can be limited Hence, patients with extramedullary relapse and MRD in
only to patients with overt CNS leukemia at diagnosis by bone marrow require more intensive treatment to prevent
substitution of imatinib with a second-generation inhibitor subsequent hematologic relapse than truly isolated extra-
(dasatinib) that readily penetrates into the CNS. medullary relapses. The efficacy of retrieval therapy in chil-
Whether transplantation would improve the poor out- dren with an isolated CNS relapse depends partly on duration
comes in patients with early T-cell ALL remains to be deter- of CR1 and partly on whether CNS irradiation was previ-
mined. A recent large study based on chemotherapy indicates ously performed. The strategy of delaying cranial or cranio-
that MRD-stratified therapy with chemotherapy may be suf- spinal irradiation for 6-12 months to allow initial
ficient to also cure this subgroup without HSCT. intensification of systemic chemotherapy has yielded long-
term second event-free survival rates of 70%-80% in chil-
dren with isolated CNS relapse. In one study, 12 months of
Treatment for relapse in children
intensive systemic chemotherapy and reduced-dose cranial
Although an isolated hematologic relapse (blood and mar- irradiation (18 Gy) resulted in an excellent 4-year event-free
row) remains most common, relapses may involve extra- survival rate among children with B-cell ALL who had not
medullary sites, such as the testes, CNS, lymph nodes, skin, received cranial irradiation during initial treatment and had
liver, spleen, and other organs. The strongest predictor of a an initial remission duration of >18 months. One-third of
fatal outcome after relapse is a short duration of first remis- patients with early testicular relapse and two-thirds of
sion. Relapse on therapy has an especially poor prognosis. patients with late testicular recurrence became long-term
Additional risk factors for therapy failure include T-cell survivors after salvage chemotherapy and testicular irradia-
immunophenotype and an isolated hematologic relapse. tion. For patients with bilateral testicular relapse, local irra-
Prolonged second remissions (>3 years) can be achieved diation (22-26 Gy) usually is recommended, but the optimal
with chemotherapy in as many as half of patients with late dose of irradiation is unclear. In patients with unilateral tes-
relapses (ie, >6 months after cessation of initial maintenance ticular relapse, some leukemia therapists advocate unilateral
chemotherapy) but in only ~10% of patients with early orchiectomy with reduced irradiation (15 Gy) to the “unin-
relapse. The presence of MRD after reinduction treatment volved testicle,” but others would rely on intensive chemo-
also indicates a very poor prognosis, but is not as useful dis- therapy alone to spare the testicular function. Indeed,
criminator as in frontline therapy. In patients who experi- successful treatment in some patients with testicular relapse
ence hematologic relapse while on therapy or shortly has been achieved without the use of any testicular
thereafter and in patients with high levels of MRD after irradiation.
remission induction for relapse, allogeneic hematopoietic
SCT is the only acceptable treatment.
Targeted therapies
For patients without histocompatible-related donors,
transplantation of stem cells from cord blood or marrow The best example of targeted therapy is the use of the TKI,
from matched-unrelated donors has yielded encouraging imatinib, in BCR-ABL1 positive ALL. Second- and third-
results. Outcome may be further improved by a new strategy generation TKIs (eg, dasatinib, nilotinib, ponatinib) with
using a reduced-intensity conditioning regimen and selection different specificities and pharmacokinetic properties have
of donor-derived alloreactive natural killer cells or selective been developed to address the problem of resistance to
depletion of α/β T cells from the graft used in haploidentical imatinib and are currently tested in cases of rising MRD
transplantation. Among various chemotherapeutic regimens after analysis of possible fusion-gene mutations, resulting in
tested in relapsed ALL, the addition of bortezomib, a protea- a fusion-protein unresponsive to the previously used TKI.
some inhibitor, to a standard four-drug induction appears to Currently, it is standard practice to test for resistance in
be promising and warrants additional study. In one random- case of rising MRD/molecular relapse including analysis of
ized trial, patients received mitoxantrone during induction possible fusion protein mutation. Change of drug therapy is
had superior progression-free survival than those treated recommended in the case of an identified mutation accord-
with idarubicin. ing to the resistance profiles of available compounds. Other
Although extramedullary relapse can occur without obvi- novel agents include inhibitors of FLT3, JAK-enzymes, Abl-
ous marrow disease, many occurrences are associated with family kinases, mTOR, farnesyltransferase, proteasome,

DNA methylation, and histone deacetylase. Some of these Special subgroups of ALL in children
principles are currently tested in clinical trials; others seem
less promising based on initial studies. Immunotherapeutic Patients with Down syndrome have a 10- to 20-fold higher
options are progressively emerging. Rituximab (anti-CD20) relative risk for leukemia, and they constitute ~2% of pedi-
showed benefit in a large randomized study of patients with atric ALL. They have the same age range as the general pedi-
>20% CD20-expressing blasts. Alemtuzumab (anti-CD52) atric population with the exception of a lack of cases in the
and epratuzumab (anti-CD22) and the recombinant immu- infant age group. ALL patients with Down syndrome have a
notoxin inotuzumab ozogamicin (anti-CD22) are active in much lower incidence of T-cell and mature B-cell ALL and
ALL, and their use is being explored in various settings. Blin- have a low frequency of other specific genetic subtypes of
atumomab has showed promising results in relapsed and precursor B-cell ALL but have a high frequency of activating
refractory cases of CD19 positive B-precursor ALL and is somatic JAK2 mutations, affecting approximately 20% of the
approved for that indication, and testing is currently moving cases. A recent compilation of data from several study groups
to trials using this agent in first high-risk relapse to eliminate showed that up to 69% of Down syndrome cases have CRLF2
MRD before SCT. rearrangements, some of which co-occurred (about 21% of
Other promising drugs include the antimetabolite, nelara- all cases) with activating JAK2 mutations. Although the out-
bine, which was incorporated in a large frontline study of come has improved with modern treatment, these patients
intensive chemotherapy for T-cell ALL with results indicat- still fared significantly worse than other children with ALL.
ing a partial reversal of the adverse outcome of the early The reason for this is likely a combination of reduced toler-
T-cell precursor immunophenotype with use of the intensive ance to chemotherapy, such as dexamethasone and metho-
regimen, though data on the nelarabine randomization have trexate resulting in reduced compliance to protocol
not yet been presented. Although the ETP immunopheno- treatment but also to excessive treatment-related deaths.
type was associated with an increased risk of induction Another possible contributing factor is the paucity of the
death, the overall EFS and OS of the ETP-group was not sig- genetic changes associated with better prognosis in this
nificantly lower than the non-ETP patients (87% and 93% vs patient population. The JAK2/CRLF2 alterations in them-
87% and 92%, respectively) with this therapy. selves do not seem to confer an adverse prognosis compared
A recent development of targeted immunotherapy in to other children with DS.
combination with gene therapy is the use of autologous Infant ALL accounts for 2%-3% of childhood ALL and is
T-cells transduced with chimeric antigen receptors characterized by a high frequency of 11q23 chromosomal
(CARs). The transduced construct usually contains the abnormalities and rearrangements of the MLL gene (70%-
variable region of an antibody against CD19 linked with 80%), a CD10-negative pro-B immunophenotype, a ten-
various costimulatory domains and anchoring part of the dency towards hyperleukocytosis, CNS-involvement, and an
T-cell receptor, but CARs directed against other antigens inferior outcome. Large collaborative studies are necessary
have also been described. Such cells have been successfully to study this rare subset of patients, but despite very large
used to treat patients with multiple relapses of B-lineage consortia efforts, progress has been very modest, if detect-
(CD19-positive) ALL and relapse after SCT. In some of able at all, over the last 15 years with overall survival hover-
these cases, the therapy has been used to get patients to a ing between 50% and 60%.
second SCT and, in some studies, persistent CAR T-cells Several studies have shown that adolescents and young
uphold what appears to be durable remission, although adults (ages 15-39 years) treated on pediatric trials fare sig-
the follow-up time is so far relatively short. There has been nificantly better than the same age groups treated on adult
significant toxicity, most notably cytokine-release syn- protocols. The superior outcome with pediatric regimens has
drome (CRS) with very high levels of inflammatory cyto- been attributed to more effective treatment, including use of
kines, that have severely affected patients and even fatalities higher doses of nonmyelosuppressive drugs (eg, asparaginase,
have been described. The CRS has been associated with corticosteroids, and vincristine) and to the better treatment
bulky disease prior to therapy and has in most cases has adherence by patients, parents, and clinicians, although this
been possible to control with the anti-IL-6 receptor anti- remains an important topic of investigation. Several com-
body, tocilizumab. Patients with persistent CAR T-cells bined adult and pediatric consortia are using common regi-
remain B-cell depleted and require immunoglobulin sub- mens to treat children and young adults to understand the
stitution. Relapses after CAR therapy have either been basis for this difference. Uniform therapy and treatment
associated with the loss of persisting circulating CAR stratification make it possible to analyze what accounts for
T-cells in the patient or loss of CD19 on the leukemic cells. the higher-risk profile of this patient population, and rigor-
Ongoing studies seek to define the role of this new tech- ously recorded studies including both pediatric and adult
nology for more widespread clinical use. departments may address some of the compliance issues.

Treatment of B-precursor ALL and T-ALL in Center and uses hyperfractionated cyclophosphamide (simi-
adults lar to the approach used for Burkitt lymphoma/leukemia),
dexamethasone, vincristine, and doxorubicin without aspar-
Tailoring treatment to assessed risk has resulted in improved aginase during induction and high-dose cytarabine and
outcomes in pediatric ALL. In adults with ALL, although risk methotrexate during consolidation. In their trials, over 90%
stratification has been used with success in the treatment of of patients achieve CR with 3-year survival rates of 50%.
Burkitt ALL and, more recently, BCR-ABL1 positive ALL Addition of rituximab for CD20 positive patients and nelar-
(described later), over the past two decades, the majority of abine for T-ALL patients to Hyper-CVAD in phase 2 studies
patients with B- and T-cell ALL have been treated without has suggested improved outcomes, but the results of ran-
specific consideration of biologic risk. Treatment for these domized trials with these agents are awaited.
adults, in general, has followed the same basic strategy of Pioneered in the treatment of pediatric ALL, the contribu-
induction, consolidation-intensification, CNS prophylaxis, tion of asparaginase to response rates and duration of
and maintenance therapy that has been used so successfully response in adults is not clear, as there are no randomized
in pediatric ALL. The relative contribution of each of these studies supporting its use in adult patients. The toxicities of
phases toward improved prognosis and disease curability has asparaginase in adults include pancreatitis, hepatotoxicity,
not been determined rigorously in adult ALL. Nevertheless, and coagulopathy. A study by the Cancer and Leukemia
the use of regimens patterned after those used in childhood Group B (CALGB), now known as the Alliance, 9511, with
ALL has resulted in the achievement of remission in the the long-acting asparaginase, pegaspargase, showed that
majority (75%-90%) of adults with ALL, although cure rates patients who achieved effective asparagine depletion had a
historically were only ~30%-40% overall. The general treat- superior outcome to patients who did not achieve asparagine
ment strategy for adults with ALL is described in the follow- depletion. Ongoing trials by the German Multicenter ALL
ing sections. Because of the lower survival rates in adults (GMALL) group of pegaspargase suggest a potential survival
with ALL treated with aggressive combination chemotherapy benefit in older adults with ALL when the drug is adminis-
approaches, the use of allo-SCT in CR1 has been explored, tered at slightly lower doses than have been used by the
and results of these studies are reviewed in this chapter. Cur- pediatricians.
rent clinical research efforts are focused on better risk strati- The goal of using granulocyte colony-stimulating factor
fication with implementation of biologically-directed (G-CSF) is to shorten the period of neutropenia to prevent
therapies tailored to disease subset, as described in the fol- possibly fatal infections, and previous studies demonstrate the
lowing sections, for adolescents and young adults with ALL utility of this drug with induction regimens for ALL. In the
and those with Ph+ ALL. Leucémie Aigüe Lymphoblastique de l’Adulte (LALA)-94 trial,
patients were randomized to receive G-CSF, granulocyte-
macrophage colony-stimulating factor (GM-CSF), or no col-
Induction phase
ony-stimulating factor (CSF). When given on day 4 of induction
Over the past 20 years, intensification of the induction regi- until return of absolute neutrophil count of 1,000/mL,
men for adults with ALL has resulted in significant improve- patients receiving G-CSF had significantly shorter hospital
ment in CR rates, with >80% of patients achieving remission stays, less time to neutrophil recovery, and fewer severe infec-
in many current multicenter studies. Building on a backbone tions compared with patients who did not receive G-CSF.
of vincristine, a glucocorticoid (prednisone or dexametha- The CALGB 9111 trial highlighted the benefit of using this
sone), and often asparaginase, the addition of an anthracy- drug in patients prone to difficulty with hematologic recov-
cline (daunorubicin or doxorubicin) has resulted in ery, specifically older patients. The study observed a trend
improved CR rates ranging from 72% to 92%. Given the toward increased CR rates in patients 60 years of age or older
high CR rate observed with these 4-drug induction regi- in the G-CSF arm compared with the placebo arm. Although
mens, it has been difficult to demonstrate further improve- G-CSF does not affect DFS or overall survival (OS), it appears
ments in overall CR rates with the addition of other drugs to be safe and also enables patients to proceed with pos-
such as cyclophosphamide or cytarabine during induction. tremission therapy.
The Italian Gruppo Italiano Malattie Ematologiche Maligne
dell’Adulto (GIMEMA) reported that, similar to childhood
Consolidation therapy
ALL, a good response to 1 week of pretreatment prednisone
before chemotherapy, defined as a decrease in circulating Traditionally, agents similar to the four or five drugs used
blasts to 1,000/mL, was predictive of a longer CR duration during remission induction, with the addition of antime-
and survival. An alternative treatment regimen known as tabolites, such as methotrexate, mercaptopurine, or thiogua-
Hyper-CVAD was developed at the M.D. Anderson Cancer nine, are used for postremission treatment. The postremission

treatment modules in adult series typically have been mod- when more than 5 leukocytes per microliter of cerebrospinal
eled after the pediatric regimens. Cyclophosphamide, high- fluid are seen along with the presence of lymphoblasts in the
dose cytarabine, and etoposide also have been incorporated cerebrospinal fluid. Symptoms may include headache, men-
into many postremission strategies, although it has been dif- ingismus, fever, or cranial nerve palsies. Some patients, how-
ficult to analyze critically the contribution of each drug or ever, have no symptoms. Risk factors for CNS involvement in
schedule to outcome in adult ALL series. adults include mature B-cell ALL, high serum lactate dehy-
Although induction chemotherapy leads to CR rates that drogenase levels (>600 U/L), and the presence of a high pro-
are >90% in many series, the relapse rate in adult ALL liferative index at diagnosis (>14% of lymphoblasts in the S
patients with adult intensive regimens has been 50%-75%, and G2/M phase of the cell cycle). If symptomatic CNS dis-
leading to many variations of postremission consolidation ease is present at diagnosis, such as focal cranial nerve palsies,
treatment in an attempt to eradicate MRD and improve concurrent radiation therapy and intrathecal chemotherapy
DFS. Adult consolidation regimens have evolved from pedi- are used.
atric schedules that have been shown to be successful. Pos- Initially, CNS-directed therapy included the use of intra-
tremission therapy in ALL can include a wide range of drugs, thecal methotrexate and 24 Gy of cranial radiation in the
including cytarabine, etoposide, teniposide, methotrexate, pediatric population. This strategy was incorporated into an
mercaptopurine, and thioguanine. In addition, the use of early adult trial that compared CNS prophylaxis with no
autologous HSCT (auto-HSCT) and allo-HSCT has been CNS treatment, resulting in an improved CNS relapse rate of
incorporated into ALL treatment, as will be discussed in a 19% versus 42% at 24 months. Although in children it is
separate section. known that combination treatment can result in toxicities
The CALGB compared a more intensive consolidation that include seizures, early dementia, cognitive dysfunction,
regimen that included both early and late intensification and slow growth, the long-term effects on adults are less
using eight drugs with previous CALGB trials in a phase 2 clear. It is known that combined radiation and intrathecal
study. The results showed that median remission duration chemotherapy in adults can cause substantial acute toxicities
improved to 29 months, whereas median survival extended that may delay postremission consolidation treatment. An
to 36 months. Likewise, the M.D. Anderson group used alternative strategy that combines intrathecal chemotherapy
extended consolidation within its Hyper-CVAD regimen by without radiation has been investigated. This treatment regi-
alternating cyclophosphamide, doxorubicin, vincristine, and men includes so-called triple therapy that uses intrathecal
dexamethasone (cycles 1, 3, 5, and 7) with high doses of methotrexate, cytarabine, and corticosteroids without
methotrexate and cytarabine (cycles 2, 4, 6, and 8) in a sin- radiation.
gle-arm trial. In this study of 204 patients treated between CNS relapse rates as low as 5% have been achieved with-
1992 and 1998, median survival time was 35 months, and the out radiation by using combination intrathecal treatment in
5-year survival rate was 39%. The Italian GIMEMA group conjunction with high-dose systemic treatment that can
conducted a study that included randomization of 388 penetrate the cerebrospinal fluid. The German GMALL
patients to postremission intensification followed by main- investigators have reported higher CNS relapse rates of 9%
tenance chemotherapy versus early maintenance therapy versus 5% when CNS-directed radiation was postponed.
without intensification. Therefore, although CNS-directed prophylactic therapy is
In summary, all of these regimens result in similar DFS required in ALL treatment, there is no single modality or
rates of ~30%-40% in adult patients with ALL who are combination that has been proven to be superior.
entered into cooperative group trials. Outcomes vary con-
siderably, however. Younger patients with favorable-risk
Maintenance therapy
cytogenetics can have DFS rates of ~60%; in contrast, older
adults defined as >60 years old still have a dismal prognosis, The rationale behind the use of maintenance treatment is the
with <10%-15% achieving long-term survival. elimination of slowly-growing subclones that persist after
induction and consolidation treatments by exposing them to
antimetabolite drugs over long periods of time, ranging from
CNS prophylaxis in adults
18 months up to 3 years after initial diagnosis. Commonly
Although <10% of adults with ALL will present with CNS used components of maintenance therapy include daily
involvement, CNS relapse will occur in 35%-75% of patients mercaptopurine and oral weekly methotrexate, supple-
at 1 year if prophylactic CNS-directed therapy is not incor- mented by monthly pulses of vincristine, corticosteroids,
porated into treatment. A lumbar puncture at the time of and periodic intrathecal methotrexate.
ALL diagnosis always is performed in pediatric studies, but is Despite the lack of randomized trials investigating the
variably timed in adult ALL regimens. CNS disease is present importance of maintenance treatment in adults with ALL,

Late complications of therapy | 545
two older trials showed inferior results compared with his- imatinib), DFS rates of 60%-75% have been reported in
torical controls when maintenance therapy is not included. these studies. For patients >60 years old with B-cell ALL for
Thus, on the basis of these data and the clear success of pro- whom the incidence of the BCR-ABL1 positive disease
longed maintenance therapy in pediatric studies, mainte- approaches 50%, the addition of TKIs to standard regimens
nance regimens mimicking those used in pediatric protocols appears to be improving remission duration, even without
routinely are incorporated into the treatment regimens of the addition of allo-HSCT. Longer follow-up will be needed,
adult B- and T-cell ALL. however, to confirm these promising results. The emergence
of TKI resistance remains the largest obstacle to long-term
survival. Thus, new trials that incorporate the third-
Risk-directed treatment of adult ALL
generation TKI, ponatinib, which has an advantage over cur-
The focus of current treatment studies for adults with ALL is rent TKIs as it maintains activity in cases with a variety of
to begin to adapt treatment according to biologic risk. Simi- ABL kinase domain mutations that result in TKI resistance,
lar to the progress that has been made in Burkitt lymphoma/ including the T315I, have been undertaken, but the use of
leukemia when a specific therapeutic approach is applied ponatinib may be limited by its vascular toxicity.
based on the underlying biology of the disease, targeted Other novel agents such as inotuzumab, blinatumomab,
treatment of Ph+ ALL that incorporates the TKIs, imatinib and CAR-T cells discussed above for pediatric ALL have also
or dasatinib, into frontline therapy and the application of a been used in adults.
pediatric-inspired regimen for older adolescents and young Transplantation of ALL is discussed further in Chapter 14.
adults may be changing the treatment paradigm for adults
with ALL, resulting in improvements in survival. The role of
allo-HSCT in first remission based on risk group also is Late complications of therapy
reviewed.
Emphasis on the intensive use of methotrexate and gluco-
corticoids has led to an increased frequency of neurotoxic-
BCR-ABL1 positive ALL
ity and osteonecrosis, underscoring the need for judicious
Treatment and outcome of patients with BCR-ABL1 positive use of even seemingly benign agents. Many long-term sur-
ALL has changed dramatically during the past decade with vivors of childhood ALL, especially those who received high
the addition of imatinib, a targeted ABL TKI, to frontline cumulative doses of glucocorticoid, methotrexate, or cra-
therapy. Previously, the standard approach for adult patients nial irradiation, have developed severe osteoporosis. Such
in whom the presence of the Ph chromosome portended a development highlights the need for early identification of
very poor prognosis with standard therapy alone (median bone lesions and therapeutic intervention to prevent frac-
survival of 1 year) was to recommend, whenever possible, tures. Treatment with anthracyclines can produce severe
allo-HSCT in first remission. Allo-HSCT in first remission cardiomyopathy, especially when they are given in high
resulted in improved DFS rates ranging from 30% to 65% in cumulative and peak doses to young girls. Cardiac abnor-
the preimatinib era. Recent studies have demonstrated that malities are persistent and progressive years after anthracy-
the addition of the TKI, imatinib or dasatinib, into frontline cline therapy. In one study, dexrazoxane prevented or
chemotherapy is feasible, does not add to systemic toxicities, reduced anthracycline-induced cardiotoxicity without
and significantly increases remission rates. Several studies interfering with antileukemic activity. In current clinical
have reported CR rates >90% for these high-risk patients. trials, only limited doses of anthracyclines are used, even
Results from these studies also suggest that the addition of for high-risk cases, to decrease the risk of subsequent
TKIs to standard therapy can rapidly reduce MRD. In older cardiomyopathy.
adults, studies are emerging that report good results with Cranial irradiation has been implicated as the cause of
addition of a TKI to vincristine and corticosteroids in induc- numerous late sequelae in children, including second can-
tion in combination with CNS prophylaxis followed by sub- cer, neurocognitive deficits, and endocrine abnormalities
sequent consolidation chemotherapy or allo-HSCT with that can lead to obesity, short stature, precocious puberty,
reductions in toxicity. Because it has been demonstrated that and osteoporosis. In general, these complications are seen in
patients with BCR-ABL1 positive ALL have improved DFS girls more often than in boys, and in young children more
when allo-HSCT is performed without evidence of MRD, often than in older children. A long-term follow-up study of
the addition of TKIs before or after allogeneic transplanta- survivors of childhood ALL revealed a >10% cumulative
tion appears to be improving OS in this high-risk group. risk of second neoplasms at 30 years and a higher than aver-
When imatinib is added to frontline therapy, followed by age mortality rate among patients who had received cranial
allo-HSCT (and sometimes followed by post-HSCT irradiation. The most devastating complication is the

development of malignant brain tumors. The median time Maude SL, Frey N, Shaw PA, et al. Chimeric antigen receptor
to the diagnosis of secondary high-grade brain tumor is 9 T cells for sustained remissions in Leukemia. N Engl J Med.
years, and the median time to diagnosis of meningioma is 2014;371:1507-1517.
20 years. Although neurocognitive problems clearly are Parikh SA, Litzow MR. Philadelphia chromosome-negative acute
lymphoblastic leukemia: therapies under development. Future
linked to cranial irradiation, they also can be caused by
Oncol. 2014;10:2201-2212.
systemic and intrathecal therapy. Knowledge of potential
Portell, CA, Sweetenham, JW. Adult lymphoblastic lymphoma.
treatment sequelae to modify treatment strategy and of
Cancer J. 2012;18:432-438.
appropriate screening measures to permit early detection Pui C-H, Evans WE. A 50-year journey to cure childhood acute
of complications should greatly improve the quality of life lymphoblastic leukemia. Semin Hematol. 2013;50:185-196.
of survivors of ALL. Roberts KG, Li Y, Payne-Turner D, et al. Targetable kinase-
activating lesions in Ph-like acute lymphoblastic leukemia. N
Engl J Med. 2014;371:1005-1015.
Bibliography Schultz KR, Carroll A, Heerema NA, Bowman WP, et al.
Long-term follow-up of imatinib in pediatric Philadelphia
Beldjord, K, Chevret, S, Asnafi, V, et al. Oncogenetics and minimal chromosome-positive acute lymphoblastic leukemia: Children’s
residual disease are independent outcome predictors in adult Oncology Group Study AALL0031. Leukemia. 2014;28:1467–
patients with acute lymphoblastic leukemia. Blood. 2014;123: 1471.
3739-3749. Topp, MS, Gokbuget, N, Stein, AS, et al. Safety and activity of
Bhojwani D, Yang JJ, Pui CH. Biology of childhood acute blinatumomab for adult patients with relapsed or refractory
lymphoblastic leukemia. Pediatr Clin North Am. 2015;62:47-60. B-precursor acute lymphoblastic leukaemia: a multicentre,
Faham M, Zheng J, Moorhead M, Carlton VEH, et al. Deep- single-arm, phase 2 study. Lancet Oncol. 2015;16:57-66.
sequencing approach for minimal residual disease detection in
acute lymphoblastic leukemia. Blood. 2012;120:5173-5180.

CHAPTER
20
Hodgkin lymphoma
Ann S. LaCasce and Andreas Engert
Introduction, 547 Therapy for relapsed or refractory Nodular lymphocyte-predominant
Frontline therapy for early-stage HL, 550 HL, 558 HL, 564
Frontline therapy for advanced-stage Follow-up of patients with HL, 562 Bibliography, 566
HL, 554 Pediatric HL, 564
Introduction variants of HRS cells, including the: (i) mononuclear variant

“Hodgkin cell”; (ii) “Reed-Sternberg cells,” which have mul-
Hodgkin lymphoma (HL) represents approximately 10% of tilobulated nuclei with small nucleoli; and (iii) the “lympho-
all cases of malignant lymphoma. This group of diseases cyte predominant” (LP) or “popcorn cells,” which are seen
most commonly affects young adults and usually presents in NLPHL and have lobulated, vesicular nuclei with multiple
with painless lymphadenopathy involving the neck and small nucleoli that are located peripherally.
chest. Systemic symptoms of fevers, night sweats, and unex- For many years, the cell of origin of the HRS cell was
plained weight loss are common. Today, the vast majority of unknown. Using single-cell microdissection techniques,
patients with HL are cured with combination chemotherapy HRS cells were isolated from histological sections of HL.
with or without radiation. Given the long-term survival of Analysis by PCR identified clonal immunoglobulin genes
patients with Hodgkin lymphoma, current efforts are with evidence of somatic mutation, thus proving that HRS
focused on reducing late, treatment-related toxicities. cells were derived from germinal center B lymphocytes.
Unusually for human cancers, HRS cells account for the
minority of cells in affected lymph nodes, from as little as 1%
Epidemiology
in HL, and are surrounded by a background of mixed inflam-
Approximately 9,000 patients per year in the United States matory cells, which varies according to the histologic sub-
are diagnosed with HL. HL is divided into two distinct enti- type and includes B- and T-cells, plasma cells, eosinophils,
ties, classical Hodgkin lymphoma (cHL) (95% of cases) and neutrophils, macrophages, and fibroblasts.
nodular lymphocyte-predominant Hodgkin lymphoma To make a definite diagnosis of HL, an adequate tissue
(NLPHL). The disease has a bimodal age distribution with biopsy is critical. Fine-needle aspirate is not adequate to
one peak in the early 20s and the second in the mid-60s. evaluate architecture and establish the histologic subtype.
There is a slight male predominance. Incisional or excisional biopsy is preferred, although image-
guided core-needle biopsy in patients without peripheral
lymphadenopathy may yield sufficient tissue. Within classi-
Pathology cal cHL, there are four histologic subtypes: nodular sclerosis
The malignant cell in HL is the Hodgkin Reed-Sternberg (NS), mixed cellularity, lymphocyte rich (LR), and lympho-
(HRS) cell, a large, bi-lobed cell with two or more nuclei cyte depleted (LD). In cHL, the HRS express CD30 in nearly
with eosinophilic nucleoli. There are several morphologic all cases and CD15 in ~85%. CD20 is positive in ~20% of
cases, but other B- and T-cells markers, including CD45,
typically are absent. The B-cell transcription factors, OCT-2
Conflict-of-interest disclosure: Dr. LaCasce: research funding: and BOB1 usually are decreased. The transcription factor
Seattle Genetics. Dr. Engert: Consultancy, presentation and research B-cell lineage specific activator protein (BSAP, also known
funding: Takeda, Millennium.
Off-label drug use: Dr. LaCasce: Rituximab for the treatment of as PAX-5) also is expressed in HRS cells in most cases, and
lymphocyte-predominant Hodgkin lymphoma. its presence can be helpful in distinguishing cHL from
| 547

548 | Hodgkin lymphoma
anaplastic large-cell lymphoma, which also expresses CD30 Risk factors

and exhibits large, atypical cells, but those cells do not
In the developed parts of the Western world, the risk of cHL,
express BSAP/PAX-5. Epstein-Barr virus (EBV), as evi-
in particular the NS subtype, is associated with factors indica-
denced by LMP-1 or EBV small nuclear transcripts (EBER),
tive of a high standard of living, including small family size,
is found in a subset of cHL, including more than half of cases
and has been postulated to be related to a delayed exposure to
of mixed cellularity (75%), and nearly all cases of LD HL.
common childhood illnesses or other environmental factors.
Nodular sclerosis (NS) HL is composed of nodular areas
A diagnosis of infectious mononucleosis confers an increased
with fibrous bands. The HRS cells may be rare in NS but also
risk for the subsequent development of cHL. In the develop-
may be found in sheets (the so-called syncytial variant of
ing world and areas of lower socioeconomic status, however,
NS). In the mixed-cellularity variant, HRS cells are more
the majority of cases of cHL are of the mixed cellularity and
abundant and are surrounded by neutrophils, eosinophils,
LD subtypes, which are more commonly associated with EBV.
macrophages, and plasma cells without areas of fibrosis. The
Patients who are immunocompromised, from either
nodal appearance is most commonly diffuse. LR HL typically
human immunodeficiency virus (HIV) infection, immuno-
appears nodular but also can be diffuse. Typical HRS are
suppression due to solid organ or hematopoetic stem cell
present in LR HL, but mononuclear cells and cells that are
transplant, or who are treated with immunosuppressive
similar to LP cells in NLPHL are seen. The background is
medications for autoimmune or inflammatory disease, are at
composed predominately of small lymphocytes. The least
higher risk for the development of cHL, which is typically
common subtype, LD HL, has a diffuse histologic appear-
associated with EBV. The risk of HIV-associated cHL has
ance with a large number of HRS cells, which may appear to
risen in the era of highly active antiretrovirals. In addition,
be atypical, in a background of fibrosis and necrosis with few
the risk of cHL is increased in patients with autoimmune dis-
inflammatory cells.
eases, including rheumatoid arthritis, lupus, and sarcoidosis,
even in the absence of immunosuppressive therapy.
Pathogenesis The risk of developing cHL is higher among relatives of
patients with cHL, and specific HLA haplotypes (most nota-
Although HRS cells are derived from germinal center B-cells,
bly, HLA-A1) are associated with a higher risk. In identical
HRS cells do not express the majority of germinal center cell
twins, the risk is increased approximately 100-fold.
markers and do not transcribe RNA for the production of
immunoglobulins or show evidence of somatic hypermu
tation. As such, the B-cell transcription factors OCT-2, Key points
BOB-1, and PU.1 are not expressed and signaling through • 9,000 new cases of HL are diagnosed per year in the United States.
the B-cell receptor is downregulated. The nuclear factor • cHL represents 95% of cases with the remainder being NLPHL.
kappa B (NF-κB) and Janus kinase–signal transducer and • There are four histologic subtypes of cHL: NS, mixed
activation of transcription signaling (JAK-STAT) pathways cellularity, LR, and LD.
have been implicated as key components of the growth and • The malignant cell in HL is the Hodgkin Reed-Sternberg (HRS)
survival of HRS cells. Genetic analyses revealed frequent cell, which is a germinal center B cell. It is accompanied by a
9p24.1 amplification, resulting in the upregulation of pro- mixed infiltrate of neutrophils, eosinophils, macrophages, and
grammed death 1 ligands (PD-1 ligands) and JAK2. other reactive cells in the nodal tissue.
The microenvironment in cHL is critical to the survival of

the HRS cell. The HRS cells secrete a variety of cytokines and
chemokines, including tumor necrosis factor alpha (TNFα),
interleukin-4 (IL-4), IL-5, IL-6, chemokine ligand-5 (CCL5), Patients with cHL typically present with nontender lymph-
CCL17, CCL22, and fibroblast growth factor (FGF), that adenopathy, with the neck being the most commonly
influence the surrounding inflammatory background. In involved site of disease. B symptoms, defined as fevers
addition, the surrounding cells, including T- and B-cells, >100.4°F (38.0°C), drenching night sweats, and involuntary
macrophages, neutrophils, plasma cells, eosinophils, and weight loss of >10% of body weight in the preceding 6 months
macrophages, also express multiple cytokines and chemo- occur in many patients with advanced-stage disease but are
kines, including IFN (interferon) gamma, IL-2, IL-10, and present in <20% of patients with early-stage disease. Pruritus,
IL-13, which provide signaling that supports the growth and which may be intense and typically is not associated with a
survival of the HRS cells. In addition, the interaction of PD-1 rash (although patients may develop secondary excoriations),
ligands on the surface of the HRS cell and PD-1 on sur- is seen in 10%-15% of patients. Although it occurs rarely
rounding T-cells results in down regulation of T-cell activity (<10% of cases), patients may experience intense pain in the
and an ineffective immune response against HRS cells. sites of disease upon alcohol ingestion.

Introduction | 549
The clinical presentation also varies according to the his- tomography (CT) and functional imaging accurately identify
tologic subtype of cHL. NS cHL accounts for 70% of cases in splenic involvement by disease. Patients are staged with CT
the Western world. Males and females are affected in equal scans and positron emission tomography (PET), as HL is
proportion and, at diagnosis, most patients are between the strongly fluorodeoxyglucose (FDG) avid.
ages of 15 and 35 years. Mediastinal involvement, which may Recent studies have demonstrated a very high sensitivity
be bulky, is common in NS cHL and patients may present of PET for bony involvement. In addition, given that HL
with respiratory symptoms. Mixed-cellularity cHL is the typically involves nodal stations in a contiguous manner,
second-most-common subtype in the industrial world, rep- patients with asymptomatic disease above the diaphragm are
resenting 20% of cHL. The median age of presentation is 38 exceedingly unlikely to have occult marrow involvement,
years, and males are affected more commonly. Patients typi- likely obviating the need for marrow sampling in this subset
cally present with peripheral lymphadenopathy. The medi- of patients. In addition, the recently proposed Lugano clas-
astinum is less commonly involved. Splenic involvement sification for the initial staging of HL suggests that bone
occurs in 30% and bone marrow occurs in 10%. LR cHL marrow biopsies are no longer required for any patients
accounts for 5% of all cases. Patients typically present with given the high sensitivity of PET-CT.
early stage disease affecting peripheral nodes. LD cHL is the Patients should be evaluated with a complete blood count
least common subtype, at 1% of cases in the Western world. (CBC) with differential and assessment of renal and hepatic
The median age of onset is in the 30s, and males are more function, including albumin, before initiating chemotherapy.
often affected. It is more common in the industrial world HIV testing should be considered. Erythrocyte sedimentation
and in HIV-infected individuals. Extranodal and intra- rate (ESR) commonly is elevated and is prognostic in early
abdominal disease commonly is seen, and advanced-stage stage disease. Lactate dehydrogenase (LDH) is rarely elevated
disease and systemic symptoms are common. except in patients with extensive, advanced-stage disease. Pul-
monary function testing and assessment of cardiac function
should be obtained before the initiation chemotherapy when-
Staging and workup
ever possible but should not delay the initiation of therapy in
The Ann Arbor staging system has been employed in HL for a young patient without comorbidities.
more than 40 years and currently is used with the Cotswold
modification (Table 20-1). X is assigned to bulky disease,
defined as any mass >10 cm in transverse dimension or to an Key points
intrathoracic mass occupying greater than one-third of the • Patients with cHL typically present with painless lymphade-
maximal intrathoracic diameter, as measured at T5/6. The nopathy in the neck and chest with or without B symptoms.
S designation refers to splenic involvement by disease. • Staging evaluation is performed with CT and PET scans with
Given that nearly all patients, including those with early bone marrow biopsy under selective circumstances. Marrow
stage disease, receive chemotherapy that treats microscopic biopsy is likely not necessary in early stage cases.
disease, the use of staging laparotomy is no longer employed. • Baseline evaluation includes an assessment of left-ventricular
In addition, modern imaging techniques, including computed cardiac function and pulmonary function tests.
Table 20-1 Staging of HL
Stage I. Involvement of one lymph node region

Stage II. Involvement of two or more lymph node regions or lymph node structures on the same side of the diaphragm. Hilar nodes should
be considered to be “lateralized” and when involved on both sides, constitute stage II disease. For the purpose of defining the number
of anatomic regions, all nodal disease within the mediastinum is considered to be a single lymph node region, and hilar involvement
constitutes an additional site of involvement.
Stage III. Involvement of lymph node regions or lymphoid structures on both sides of the diaphragm.
Stage IV. Diffuse or disseminated involvement of one or more extranodal organs or tissue beyond that designated E, with or without associated
lymph node involvement.
All cases are subclassified to indicate the absence (A) or presence (B) of the systemic symptoms of significant unexplained fever, night sweats,
or unexplained weight loss exceeding 10 percent of body weight during the 6 months before diagnosis.
The designation “E” refers to extranodal contiguous extension (ie, proximal or contiguous extranodal disease) that can be encompassed within
an irradiation field appropriate for nodal disease of the same anatomic extent. More extensive extranodal disease is designated stage IV.
The subscript “X” is used if bulky disease is present. This is defined as a mediastinal mass with a maximum width that is equal to or greater
than one-third of the internal transverse diameter of the thorax at the level of T5/6 interspace or >10 cm maximum dimension of a nodal
mass. No subscripts are used in the absence of bulk.

Frontline therapy for early-stage HL malignancies and cardiac dysfunction, and the improve-
ment in outcomes with the addition of effective chemother-
apy. By definition, EFRT, also known as subtotal nodal
Clinical case
radiotherapy (STNRT), includes both the involved lymph
A 24-year-old woman presents with a persistent dry cough nodes and the grossly normal adjacent lymph nodes. Typical
for 2 months duration. She has no weight loss, fever, or night extended fields were the mantle field and the inverted Y field.
sweats. On exam, she looks well and has bilateral cervical lym- IFRT, which encompassed only the clinically involved lymph
phadenopathy. Chest x-ray reveals a widened mediastium and
nodes, replaced EFRT, in part based on two randomized
subsequent chest CT is notable for 3.5 × 3 cm anterior medias
studies. In the European Organization for Research and
tinal mass. Mediastinoscopy and biopsy are performed and
Treatment of Cancer (EORTC) H8 study, more than 1,500
reveal classical HL, NS subtype with neoplastic HRS cells express-
ing CD30, CD15, and negative for CD20. EBER is negative. PET
patients were stratified according to risk (see detailed discus-
and CT scans demonstrate disease localized to the mediastinum. sion in the following section). Chemotherapy consisted of
Laboratory studies show a mild leukocytosis at 12.5 with 80% MOPP/ABV. Favorable patients received three cycles of
neutrophils and 10% lymphocytes with an otherwise-normal chemotherapy plus IFRT versus STNRT. Unfavorable risk
CBC. ESR is 25. PET and CT scans after two cycles of therapy patients received four or six cycles of chemotherapy with
show mediastinal uptake less than blood pool, and she com- IRFT versus four cycles of chemotherapy with STNRT.
pletes four cycles of chemotherapy followed by involved-field There was no difference in 5-year or 10-year event-free sur-
radiotherapy (IFRT) to 30 Gy. vival (EFS) or OS. Bonadonna et al (2004) randomized
patients to receive either STNRT versus IFRT following four
Before the 1980s, patients with HL underwent laparotomy to cycles of ABVD chemotherapy. The freedom from progres-
fully stage the extent of disease. Patients without splenic sion (FFP) and OS at 12 years were equivalent in both arms
involvement and disease confined to the neck and chest were at 94% and 96%, respectively. Current studies are under way
treated with extended-field radiotherapy (EFRT), consisting to evaluate the use of involved-node radiotherapy, in which
of a mantle field and the para-aortic region. Approximately the initially involved lymph nodes plus an additional margin
95% of patients achieved a complete remission (CR), with of ≤5 cm of surrounding, radiographically-uninvolved tis-
75% of patients remaining disease free in the long term. With sue is treated.
the introduction of chemotherapy, multiple clinical trials In terms of the dose of the radiotherapy employed, most
subsequently demonstrated the superiority of chemotherapy studies have used 20-40 Gy administered in 1.8-2 Gy frac-
plus radiation compared with radiation alone in terms of tions. The German Hodgkin Study Group (GHSG) ana-
progression-free survival (PFS) and overall survival (OS). lyzed two studies in which EFRT at 20 Gy, 30 Gy, or 40 Gy
Overall, the prognosis of early stage cHL using currently was administered following COPP/ABVD chemotherapy
available therapies is excellent, with >85% of patients being and demonstrated no difference in OS. Current studies
cured of disease with initial therapy and 95% of patients alive typically employ 20-30 Gy of IFRT for nonbulky disease
at 5 years. Currently, the majority of patients with early stage and 30-36 Gy of IRFT in the presence of bulk. More
cHL are treated with combined modality therapy given its recently, the International Radiation Oncology Group
superior freedom from treatment failure (FFTF). Chemo- (ILROG) recommended the use of even more limited fields
therapy alone is associated with a higher risk of relapse but based on data demonstrating equivalent efficacy compared
has less long-term toxicity compared with combined modal- with involved field and the lower likelihood of late toxicity.
ity treatment, and the OS is likely not different given the This technique, involved-site radiotherapy (ISRT) incor-
availability of effective second-line therapies. porates the pre- and postchemotherapy tumor volume plus
ABVD (doxorubicin [Adriamycin], bleomycin, vinblastine, a margin of normal surrounding tissue. This is particularly
dacarbazine) is the favored chemotherapy in most centers for relevant when combined modalities are used and reflects
HL in terms of efficacy and toxicity, including risk of myelo- the tumor volume reduction during the chemotherapy
dysplastic syndromes (MDS) or acute myeloid leukemia phase. In addition, involved node radiotherapy (INRT)
(AML) and infertility based on a randomized study comparing also uses pre- and posttreatment imaging but with a much
MOPP, MOPP/ABVD, and ABVD (see full discussion in more limited margin. Given that INRT requires high qual-
“Frontline therapy for advanced-stage HL”). ity pretreatment imaging fused with end of treatment scans,
most patients receive ISRT. Although the use of more lim-
ited radiotherapy fields and lower radiation dose will likely
Radiotherapy
reduce late toxicity compared with that of larger fields,
Over time, the extent and dose of radiotherapy has decreased long-term follow-up will be required to confirm this
given associated long-term toxcities, particularly secondary hypothesis.

Frontline therapy for early-stage HL | 551
Risk stratification at ~91%-93% and 97% at 5 years and 86%-90% and 94% at
8 years, respectively. Toxicity was comparable between all the
A number of prognostic indicators have been identified in
arms. Overall, there have been 55 (4.6%) secondary malig-
early-stage cHL and are employed in clinical trials to risk
nancies, including 38 solid tumors, 15 cases of non-HL, and
stratify patients (Table 20-2). The GHSG scale includes five
two cases of AML. In the follow-up HD13, the GHSG exam-
risk factors, including bulky mediastinal disease as defined
ined the relevance of bleomycin and dacarbazine in the ABVD
by more than one-third of the maximal intrathoracic cavity,
regimen in patients with favorable risk early-stage disease.
ESR of ≥30 in the presence of B symptoms or ≥50 without
One thousand, five hundred, and two patients were assigned
B symptoms, extranodal extension of disease, and three or
to two cycles of ABVD, ABV, AVD or AV followed by 30 Gy
more lymph node sites of involvement. The EORTC scale
of IFRT. The AV and ABV arms were closed early due to
includes age ≥50 years, bulky mediastinal disease, ESR of
excess events. The primary objective was noninferiority of the
≥30 in the presence of B symptoms or ≥50 without B symp-
experimental arms compared with ABVD by excluding a
toms, and four or more nodal sites of involvement. In Cana-
5 year difference in FFTF of 6%. The 5 year FFTF of the regi-
dian and some US cooperative group studies, patients with
mens was 93%, 81%, 89%, and 77%, respectively, for AVBD,
stage IIB disease are considered to have advanced-stage dis-
ABV, AVD, and AV. The authors concluded that ABVD
ease. The presence of bulky mediastinal disease is considered
remains the standard therapy in this patient population.
to be unfavorable by all groups.
In the National Cancer Institute of Canada (NCIC)/ECOG
HD.6 trial, the use of chemotherapy alone was compared
Early favorable disease with STNRT without chemotherapy in favorable risk patients
defined as age <40, fewer than four lymph node sites involved,
In the EORTC H8F study, patients were randomized to three
and absence of mixed cellularity or LD histology, ESR <50—
cycles of MOPP/ABV (n = 270) plus IFRT versus STNRT
patients in the experimental arm received ABVD for four to
alone (n = 272). Both the 5-year EFS of 98% versus 74%
six cycles (n = 59) depending on response. Patients who
(P < 0.001) and the 10-year overall 97% versus 92% (P =
achieved CR following two cycles of ABVD received a total of
0.001) favored the combined modality arms. The GHSG
four cycles, whereas all others received six cycles. The control
subsequently examined the role of fewer cycles of chemo-
arm for favorable patients consisted of STNRT (n = 64). The
therapy and lower dose radiation in the HD10 study: 1,370
study was closed early after the EORTC study demonstrated
patients without risk factors (fewer than three nodal sites,
the superiority of combined modality therapy using IFRT
nonbulky disease without extranodal extension, ESR <30
compared with STNRT. At a median follow-up of 11.3 years,
without B symptoms or ESR <50 with B symptoms) were
there was no difference in the freedom from progressive dis-
randomized in a two-by-two design to four versus two cycles
ease at 89% and 87%, respectively, or OS at 98% in both
of ABVD and 30 Gy versus 20 Gy of IFRT. With a median
arms. The outcome of patients who achieved CR after two
follow-up of 7.5 years, there was no difference in FFTF or OS
cycles of ABVD was favorable with freedom from disease
progression and OS of 94% and 98%, respectively.
Table 20-2 Risk factors in early-stage Hodgkin lymphoma
Organization Risk factors Early unfavorable disease

EORTC Age <50
In patients with unfavorable risk Hodgkin lymphoma, the
No LMA (less than one-third maximum
intrathoracic diameter) EORTC H8U study randomized 996 patients to three arms:
ESR <50 without B sx four cycles of MOPP/ABV plus IFRT versus six cycles of
ESR <30 with B sx MOPP/ABV chemotherapy followed by IFRT or four cycles
<4 lymph node groups of MOPP/ABV followed by STNRT. There was no difference
GHSG No LMA (less than one-third maximum in 5-year EFS (84%-88%) or 10-year OS (84%-88%). The
intrathoracic diameter) GHSG HD8 study also demonstrated preserved efficacy with
ESR <50 without B sx reduced toxicity using IFRT compared with extended-field
ESR <30 with B sx radiotherapy (EFRT) in unfavorable patients (stage IA-IIA
No extranodal extension with risk factors as defined in the HD10 study, or stage IIB
<3 lymph node groups disease with at least three lymph node sites of involvement or
B sx = fevers, drenching night sweats, unexplained weight loss; an elevated ESR). Patients received COPP (cyclophospha-
EORTC = European Organization for Research and Treatment of mide, vincristine, procarbazine, and prednisone) and ABVD
Cancer; ESR = erythrocyte sedimentation rate; GHSG = German for two cycles each and were randomized to either 30 Gy of
Hodgkin Study Group; LMA = large mediastinal mass. EFRT (n = 532) or Gy IFRT (n = 532). All patients received

an additional 10 Gy of radiotherapy to bulky sites of disease 12 weeks of chemotherapy followed by 36 Gy to bulky sites.
defined as a single lymph node or conglomerate mass mea- At a median follow-up of 8 years, the 8-year FFP was 96% in
suring > 5cm. At a median follow-up of 54 months, the FFTF patients without bulky disease and 92% in patients with
and OS were 83% and 91%, respectively, and not different bulky disease. In patients with bulky mediastinal disease
between the two arms. Toxicity, in terms of myelosuppres- treated in the ECOG 2496 study, 136 received ABVD for six
sion and gastrointestinal side effects, were more prominent cycles plus 36 Gy, and 131 patients were treated with 12 weeks
in the EFRT arm. In the GHSG follow-up study, HD11, a of Stanford V with 30 Gy of IFRT to sites >5 cm and 36 Gy to
total of 1,395 patients were included in order to compare the mediastinum. At a median follow-up of 5.5 years, there
four cycles of ABVD with four cycles of BEACOPP (bleomy- was no difference in the 5-year failure-free survival (FFS) at
cin, etoposide, doxorubicin, cyclophosphamide, vincristine, 85% versus 77% (P = 0.13) or OS 95% versus 92% (P = 0.31).
procarbazine, and prednisone) and IFRT of 20 or 30 Gy. Toxicity was similar between arms.
There were no differences between the chemotherapy arms,
so that the less toxic ABVD combined with 30 Gy should be
Chemotherapy alone
used in this setting.
The subsequent HD14 study examined the role of intensi- Given the late effects of radiotherapy, including secondary
fied chemotherapy with the incorporation of the escalated malignancies, especially breast cancer in women <30 years of
BEACOPP regimen in this patient population. This regimen age and cardiovascular disease, a number of studies have
had been superior to COPP/ABVD in patients with evaluated the use of chemotherapy only in Hodgkin lym-
advanced-stage disease (see full discussion in the section phoma, including the Canadian/ECOG HD.6, discussed
“Frontline therapy for advanced-stage HL”). In the HD14 above. For patients in this study, patients with both favor-
study, 1,528 patients were randomized to standard therapy able and unfavorable disease who achieved a complete
with four cycles of ABVD versus two cycles of escalated BEA- remission after 2 cycles of ABVD had excellent outcomes
COPP followed by two cycles of ABVD. All patients received with FFTF and OS at 12 years of 94% and 98% respectively.
30 Gy of IFRT. The intensified arm resulted in improvement In contrast, those patients who were not in CR after two
in the primary endpoint of 5-year FFTF at 95% versus 88% cycles of ABVD had a significantly poorer outcome (Hay
(P < 0.001) in the standard arm. The OS, however, in both et al., 2013). There is only a single randomized trial compar-
arms was excellent at 97%, highlighting the ability to salvage ing chemotherapy to combined modality treatment in which
patients initially treated with ABVD. Grade 3 toxicity was a proportion of the patients received IFRT. Straus et al
significantly more prominent in the BEACOPP arm in terms (2004) randomized 152 patients with stage IA-IIIA HL to six
of leukopenia, thrombocytopenia, and infection. Second cycles of ABVD chemotherapy alone or ABVD with IFRT or
malignancies were similar in both arms with two cases of modified EFRT. There was no difference in FFP (86% vs.
MDS or AML in the BEACOPP group, although the median 81%; P = 0.61) or OS (97% vs. 90%; P = 0.08) at 5 years,
follow-up remains relatively short at 43 months. although the study was closed early for poor accrual; in addi-
In terms of the role of chemotherapy only in unfavorable tion, the study included patients with stage III disease and
patients without bulky disease, the Canadian/ECOG HD.6 was underpowered to detect a difference.
study randomized patients to ABVD alone (n = 137) versus
ABVD for two cycles followed by STNRT (n = 139). At a
Risk-adapted strategies
median follow-up of 11.3 years, the freedom from disease
progressive favored the combined modality arm at 94% Recent evidence suggests that interim PET scans (ie, con-
versus 86% (P = .006). OS, however, was superior in the ducted after the first two or three cycles of therapy) are
chemotherapy-only arm at 92% compared with 81% (P = highly predictive of outcome in HL, especially in advanced
.04) in the radiation-containing arm because of late deaths in disease. For example, Gallamini et al (2007) evaluated 260
patients receiving radiotherapy. patients with stage IIB-IV HL, the majority of whom were
treated with ABVD chemotherapy with or without radia-
tion. Patients underwent PET scans after two cycles of ther-
Stanford V
apy. Approximately 20% of patients were PET positive. At a
The Stanford V regimen, which includes meclorethamine, median follow-up of 2 years, the PFS in PET-negative
doxorubicin, vinblastine, prednisone, vincristine, bleomy- patients was 95%, whereas only 12.8% of patients with a
cin, and etoposide, is an alternative approach used in early- positive PET scan were free from disease (P < 0.0001). The
stage cHL. In patients with stage I or IIA nonbulky cHL, predictive value in patients with early-stage disease is less
87 patients received 8 weeks of Stanford V plus 30 Gy of clear. In a study of 77 patients, 5 of 16 interim PET-positive
IFRT, and 61 patients with bulky, early-stage disease received patients had early-stage disease and only one patient

Frontline therapy for early-stage HL | 553
relapsed. In addition, a retrospective series of 96 patients Toxicity of chemotherapy

with nonbulky, early-stage disease, end-of-treatment, not
In general, ABVD and Stanford V are well tolerated with nau-
interim PET was predictive of PFS at 94% in the PET-
sea, vomiting, constipation, alopecia, and peripheral neuropa-
negative group compared with 54% in the PET-positive
thy being common. Anthracycline-related cardiotoxicity in the
group (P < 0.0001).
absence of mediastinal radiotherapy is rare in this patient pop-
Given the prognostic value of interim PET scans in pre-
ulation, as the total cumulative dose of doxorubicin adminis-
dicting FFS, a number of ongoing clinical trials are using
tered is 300 mg/m2 or less. An evaluation of left-ventricular
PET-directed risk-adapted therapy. In the United Kingdom
function is typically obtained before the initiation of chemo-
(UK) RAPID trial, patients with early-stage HL underwent
therapy, although asymptomatic cardiac dysfunction is uncom-
imaging with PET scans after three cycles of therapy. PET-
mon in this patient population, especially younger patients.
negative patients were randomized to receive no further
The majority of patients receiving ABVD will develop
therapy versus IFRT. PET-positive patients received con-
significant granulocytopenia with anemia and thrombocyto-
solidation with IFRT. The primary endpoint was a <7% dif-
penia being uncommon. When patients present for chemo-
ference between the RT and observation and RT arms. In
therapy, many will have low absolute neutrophil counts, and
the preliminary analysis, the study was positive and sug-
some patients will be frankly neutropenic. Despite this, retro-
gested that this strategy would allow some patients to avoid
spective data suggests the risk of febrile neutropenia is very
RT. The US Intergroup has completed a trial of a similar
low, <1% per cycle. Patients who receive white blood cell
risk-adapted strategy in patients with early-stage, nonbulky
growth factors may have an increased risk of bleomycin lung
disease. Patients who are PET negative after two cycles of
toxicity. The majority of patients may be treated safely with
ABVD receive two additional cycles without radiotherapy.
full-dose therapy, on time, without growth factors. For
PET-positive patients receive two cycles of escalated
patients who develop febrile neutropenia, granulocyte colony-
BEACOPP plus radiotherapy. The results of this study are
stimulating factor (G-CSF) for the minimal number of days to
pending. The EORTC evaluated similar intensified
support the white blood cell count should be administered.
approaches in interim PET-positive patients with both
Bleomycin-associated pneumonitis is common and pres-
favorable and unfavorable disease. The study was also
ents in ~20% of patients receiving a full course of ABVD
designed as a noninferiority study, and the chemotherapy
chemotherapy. The discontinuation of bleomycin does not
only arms were closed early given the number of events in
appear to adversely affect the efficacy of ABVD chemother-
both the favorable and unfavorable arms, which precluded
apy. There are not well-studied guidelines for following
the possibility of a positive study. Therefore, risk-adapted
patients who are receiving bleomycin. Baseline pulmonary
strategies remain experimental at this time.
function tests may be obtained before chemotherapy. A low
baseline diffusing capacity (DLCO) should be corrected for
Summary of frontline therapy
baseline hemoglobin levels and interpreted carefully in
For patients with favorable disease, current options include patients with extensive disease in the chest. A high index of
three to four cycles of ABVD plus IFRT, typically 30 Gy, suspicion is critical for the early recognition of bleomycin
except for patients meeting the criteria for the GHSG HD 10 lung toxicity. Patients who develop cough and or dyspnea on
study, where two cycles of ABVD plus 20 Gy IFRT is an exertion with or without fevers should be evaluated promptly
appropriate option. The Stanford V regimen for 8 weeks plus by physical examination for the presence of basilar crackles
30 Gy of IFRT to sites >5 cm is an alternative approach. For and oxygen desaturation with ambulation and or at rest.
patients with unfavorable, nonbulky disease, options include Chest x-ray may reveal an interstitial pattern of abnormality,
four cycles of ABVD plus 30 Gy of IFRT. Two cycles of esca- and a decline in the DLCO on pulmonary function testing is
lated BEACOPP followed by two cycles of ABVD, followed typical. Bleomycin should be discontinued promptly and
by 30 Gy of IFRT in patients fitting the criteria for the GSHG steroids should be administered for patients with significant
HD14 study results in improved disease control without an symptoms or hypoxemia. The value of serial pulmonary
OS benefit at the expense of increased toxicity. For patients function testing has not been demonstrated clearly but may
with bulky disease, options include four to six cycles of show asymptomatic decreases in DLCO.
ABVD for 12 weeks or Stanford V followed by 36 Gy of IFRT. With regard to fertility, the risk of premature ovarian fail-
Chemotherapy alone with six cycles of ABVD for patients ure with ABVD is very low. Studies evaluating the prophy-
without bulky disease is an alternative, especially in young lactic use of gonadotropin-releasing hormone (GNRH)
women with a high risk of radiotherapy-related breast can- agonists have been equivocal. For patients diagnosed in their
cer, and existing data suggests a 5%-8% lower PFS without 30s who desire to retain fertility, referral to a reproductive
demonstrable difference in OS. endocrinologist should be considered.

Key points Clinical case (continued)

• Patients with early-stage cHL are risk stratified according to a CD30 and negative for ALK-1, CD3, CD20, and CD45, and was
number of factors, including the presence of bulky disease, ESR, consistent with cHL, NS subtype. PET and CT scans demonstrated
and number of nodal sites of involvement. extensive bilateral cervical, supraclavicular, axillary, mediastinal,
• More than 90% of patients with favorable disease and 85% of hilar, and retroperitoneal adenopathy with SUVs of 7.3-18.5, and
patients with unfavorable disease are cured with initial therapy. small bilateral pulmonary nodules. The patient had no B symp-
• Therapeutic options include combined modality therapy: two toms, ESR was elevated at 82, and he had six adverse prognostic
to four cycles of ABVD (favorable), four to six cycles of ABVD features by the International Prognostic Score (IPS), including male
(unfavorable), or the Stanford V regimen plus IFRT. gender, age >45 years, white blood cells (WBC) 15.5, hemoglobin
• Chemotherapy alone with ABVD in patients without bulky (Hb) 8.6 mg/dL, albumin 3.0 g/dL, and stage IV disease. Ejec-
disease is an alternative in selected cases. tion fraction (EF) was normal at 55% on pretreatment multigated
acquisition scan. Treatment was given with six cycles of ABVD
without complication until 2 days after completion of cycle 6 when
Frontline therapy for advanced- the patient noticed progressive dyspnea with exercise. Chest x-ray
stage HL and a CT scan of the chest demonstrated no pulmonary infiltrates
or nodules, but an echocardiogram demonstrated EF of 20%.
Clinical case
A 68-year-old man with a history of hypertension and asthma ABVD
presented with firm, fixed 3-4 cm right-sided submandibular and
Since the early 1990s, the treatment of patients with advanced-
cervical adenopathy. Biopsy of a right axillary lymph node demon-
stage HL has relied on combination chemotherapy with ABVD
strated large, pleomorphic lymphoma cells positive for CD15 and
(Table 20-3). ABVD was first introduced by Bonadonna and
Table 20-3 Frontline chemotherapy regimens in HL
Method of
Regimen Drugs administration When administered Cycle
ABVD Doxorubicin 25 mg/m2 IV Days 1 and 15 Q28 days
Bleomycin 10 units/m2 IV Days 1 and 15
Vinblastine 6 mg/m2 IV Days 1 and 15
Dacarbazine 375 mg/m2 IV Days 1 and 15
BEACOPP Bleomycin 10 mg/m2 IV Day 8 Q21 days
(baseline) Etoposide 100 mg/m2 IV Days 1-3
Doxorubicin 25 mg/m2 IV Days 1
Cyclophosphamide 650 mg/m2 IV Day 1
Vincristine 1.4 mg/m2 (capped at 2.0 mg) IV Day 8
Procarbazine 100 mg/m2 IV Days 1-7
Prednisone 40 mg/m2 IV Days 1-14
BEACOPP Bleomycin 10 mg/m2 IV Day 8 Q21 days
(escalated) Etoposide 200 mg/m2 IV Days 1-3
Doxorubicin 35 mg/m2 IV Days 1
Cyclophosphamide 1,250 mg/ m2 IV Day 1
Vincristine 1.4 mg/m2 (capped at 2.0 mg) IV Day 8
Procarbazine 100 mg/m2 IV Days 1-7
Prednisone 40 mg/m2 IV Days 1-14
Stanford V Doxorubicin 25 mg/m2 IV Weeks 1, 3, 5, 7, 9, 11
Vinblastine 6 mg/m2 IV Weeks 1, 3, 5, 7, 9, 11
Vincristine 1.4 mg/m2 (capped at 2.0 mg) IV Weeks 2, 4, 6, 8, 10, 12
Bleomycin 5 U/m2 IV Weeks 2, 4, 6, 8, 10, 12
Mustard 6 mg/m2 IV Weeks 1, 5, 9
Etoposide 60 mg/m2 IV Weeks 3, 7, 11
Prednisone 40 mg/m2 PO QOD Weeks 1-9; taper by 10 mg
QOD weeks 10 and 11
IV= intravenous; PO = per os (by mouth); Q = every; QOD = every other day.

Frontline therapy for advanced-stage HL | 555
coworkers in the late 1970s as an effective, noncross-resistant 69%, 64%, and 67%, respectively; the 5-year OS was 82%
chemotherapy option for patients with relapsed HL. Unlike using ABVD; and ABVD resulted in fewer pulmonary and
MOPP, the first successful combination chemotherapy regimen hematologic toxicities, treatment-related deaths, and second
used in the treatment of HL, ABVD was associated with little malignancies, including acute leukemia, than observed with
sterility and minimal secondary myelodysplasia or leukemia. MOPP/ABV. In a UK study comparing ABVD to other com-
The success of ABVD in the second-line setting eventually bination hybrid regimens (ChlVPP/PABIOE and ChlVPP/
led to efforts to combine MOPP and ABVD in 12 alternating EVA), the 3-year EFS and OS with ABVD were 75% and
treatment cycles (MOPP/ABVD alternating regimen). As 90%, respectively, once again similar to multidrug regimens,
frontline therapy in patients with stage IV HL, MOPP/ABVD with less infectious and neurologic toxicity than observed
proved superior to MOPP alone with a FFP at 8 years of with the hybrid regimens.
64.6% versus 35.9% (P < 0.005) and 8-year OS rates of 83.9% As a result of these trials, ABVD became the standard of
versus 63.9% (P < 0.06). In 1982, the Cancer and Leukemia care for initial therapy of advanced-stage HL; however,
Group B (CALGB) in the United States initiated a random- recently, two other combination chemotherapy regimens,
ized multicenter trial in 361 patients with stage III-IV previ- Stanford V and BEACOPP (Table 20-3) have challenged the
ously untreated HL, comparing MOPP for six to eight cycles role of ABVD as the standard frontline regimen in this
(n = 123), ABVD for six to eight cycles (n = 115), and alter- patient population.
nating MOPP/ABVD for 12 cycles (n = 123). In this trial,
complete response (CR) rate (67% MOPP, 82% ABVD, and
Stanford V
83% MOPP/ABVD, P = 0.006) and 5-year FFS (50% MOPP,
61% ABVD, and 65% MOPP/ABVD, P = 0.02) were superior In a phase 2 trial with Stanford V administered weekly over
in the ABVD and MOPP/ABVD arms, with less neutropenia, 12 weeks (Table 20-3) followed by 36 Gy consolidative
thrombocytopenia, and infectious toxicity in the ABVD arm radiotherapy to bulky mediastinal disease, tumor masses, or
(18%, 2%, and 2% ABVD; 47%, 36%, and 11% MOPP; and lymphadenopathy ≥5 cm, the 5-year FFP was 89% in pati
53%, 28%, and 12% MOPP/ABVD, respectively). ents with bulky stage I-II and advanced-stage HL (n = 142).
A subsequent randomized phase 3 US Intergroup trial In a randomized Italian trial in 355 patients with stage IIB,
compared ABVD and a MOPP/ABV hybrid (combined III, and IV HL comparing six cycles of ABVD (n = 122),
MOPP and ABV drugs in each cycle rather than alternating 12 weeks of Stanford V (n = 107), and six cycles of
cycles of MOPP and ABVD) for 8-10 cycles in 856 patients MOPPEBVCAD (mechlorethamine, vincristine, procarba-
with stage III-IV HL. CR rates (76% vs. 80%, P = 0.16) and zine, prednisone, epirubicin, bleomycin, vinblastine, lomus-
5-year FFS rates (63% vs. 66%, P = 0.42) were similar in tine [CCNU], doxorubicin, and vindesine), 5-year FFS were
patients receiving ABVD or MOPP/ABV, respectively. In 78%, 54%, and 81%, respectively (P < 0.01). In this trial,
this study, the 5-year FFS for patients with 0, 1, 2, 3, 4, and however, radiation was given only to sites of initial bulky
≥5 risk factors by the IPS (Table 20-4), were 76%, 72%, 81%, disease or sites of residual disease after chemotherapy, and
only 66% of patients receiving Stanford V also received
Table 20-4 International Prognostic Score in advanced-stage HL radiation in the Italian study, compared with 91% in the
Number of risk factors* 5-year FFP (%) 5-year OS (%)
prior single-institution trial. At 10 years of follow-up,
10-year FFS were 75%, 74%, and 49% (P < 0.001) and
0 84 ± 4 89 ± 2
10-year OS were 87%, 80%, and 78% (P = 0.4) for the ABVD,
1 77 ± 3 90 ± 2
MOPPEBVCAD, and Stanford V arms, respectively; how-
2 67 ± 2 81 ± 2
ever, in looking at Stanford V patients treated with and
3 60 ± 3 78 ± 3
without radiotherapy, 10-year disease-free survival (DFS)
4 51 ± 4 61 ± 4
was statistically in favor of those patients receiving radiation
>5 42 ± 5 56 ± 5
(76% vs. 33%, P = 0.004).
From Hasenclever D, Diehl V. N Engl J Med. 1998;339:1506-1514.
In a subsequent multicenter UK study in 520 patients with
FFP = freedom from progression, OS = overall survival.
stage IIB, III, IV, or bulky stage I-II HL, there were no differ-
* The IPS is derived from a retrospective analysis of 5,141 patients
ences in 5-year FFS and OS in patients receiving six to eight
treated at 25 centers from 1983-1992 with advanced-stage HL.
Risk factors identified in this retrospective study included age >45
cycles of ABVD (53% patients received radiation to sites of
years, male gender, WBC >15,000/mm3, Hb <10.5 g/dL, absolute residual disease) or 12 weeks of Stanford V (73% patients
lymphocyte count <600/mm3 or <8% of WBC, albumin <4.0 g/dL, received radiation to disease sites >5 cm or splenic nodules).
and stage IV disease. More recent data on the value of IPS suggest More pulmonary toxicity was observed in the ABVD arm in
that the impact might have narrowed in the modern treatment era this trial, although myelosuppression and neuropathy was
(Moccia et al., 2012). slightly worse with Stanford V.

ECOG 2496, a randomized phase 3 US Intergroup trial, follow-up trial examined eight cycles of BEACOPP escalated
demonstrated no significant difference in 5-year FFS (73% versus four cycles of BEACOPP escalated plus four cycles of
vs. 71%, P = 0.29) or 5-year OS (88% vs. 87%, P = 0.87) in BEACOPP baseline. Five-year FFTF and OS were 86.4% and
812 patients with bulky stage I-II, III, or IV HL receiving six 92%, respectively, with eight cycles of BEACOPP escalated
to eight cycles of ABVD with radiation to bulky mediastinal compared with 84.8% and 90.3% with the 4+4 arm, and tox-
disease or 12 weeks of Stanford V chemotherapy with radio- icities were not significantly reduced with the 4+4 approach.
therapy to disease >5 cm or splenic nodules, respectively. In a randomized Italian study comparing six cycles of
With median follow-up of 5.25 years, 26 second malignan- ABVD with four cycles of BEACOPP escalated plus two
cies were observed (14 ABVD and 12 Stanford V). Therefore, cycles BEACOPP baseline, the 5-year PFS was superior for
as a result of these trials, six cycles of ABVD has remained the BEACOPP (81%) compared with ABVD (68%) (Table
standard treatment for patients with advanced-stage HL; 20-5). The trial included a total of 197 patients for this com-
Stanford V may be acceptable in selected patients for whom parison and was not powered to detect differences in OS.
a shortened treatment duration or reduction in cumulative In a second Italian cooperative group study, the 5-year rate
doses of bleomycin or doxorubicin is desirable. of freedom from first progression was 85% in patients
receiving four cycles of BEACOPP escalated plus four cycles
BEACOPP baseline compared with 73% in patients receiv-
BEACOPP
ing six to eight cycles of ABVD (P = 0.004). In this trial, the
The GHSG HD9 trial randomized patients aged 15-65 years overall survival at 5 years was 84% with ABVD and 89% with
with stage IIB, III, and IV HL to either eight cycles of COPP/ BEACOPP. But again, this trial was not powered to detect
ABVD (cyclophosphamide, vincristine, procarbazine, and differences in OS. More recently, two French multicenter tri-
prednisone alternating with ABVD), BEACOPP baseline als also compared BEACOPP 4+4 with ABVD demonstrat-
(bleomycin, etoposide, doxorubicin, cyclophosphamide, ing significant differences in terms of PFS with 69% vs 84%
vincristine, procarbazine, and prednisone), or BEACOPP (P = 0.0003) and 75% vs 93% (P = 0.008), respectively.
escalated (Table 20-3). With 10 years of follow-up, FFTF was Again, there were nonsignificant differences in OS.
64%, 70%, and 82% with OS of 75%, 80%, and 86% in the Despite the superior efficacy of BEACOPP escalated com-
COPP/ABVD, BEACOPP baseline, and BEACOPP escalated pared with hybrid regimens and ABVD in patients <60 years
arms, respectively. By IPS, FFTF at 5 years was 92%, 87%, of age, there have been concerns of increased acute toxicity,
and 82% with BEACOPP escalated for patients with 0-1, 2-3, a higher incidence of secondary AML or MDS as well as
and 4-7 risk factors, respectively. Only 13% of patients infertility with BEACOPP escalated. In addition, there is
receiving BEACOPP escalated had 4-7 risk factors on this clearly more acute toxicity including hematologic and infec-
trial, however, and although differences in FFTF were sig- tious complications. To reduce the toxicity of BEACOPP
nificantly in favor of BEACOPP escalated among all three escalated without compromising on its efficacy, the GHSG
risk groups, statistically significant improvements in OS performed a three-armed prospectively randomized pan-
were observed only in patients with IPS scores of 2-3 European trial in which a total of 2,182 patients with newly
(P = 0.27 for IPS 0-1, P < 0.0027 for IPS 2-3, and P = 0.16 for diagnosed advanced-stage HL aged 18-60 years were ran-
IPS 4-7). When analyzed by age, there was not a significant domly assigned to receive either 8 cycles of BEACOPP esca-
improvement in FFTF and OS for patients aged 60-65 years, lated, 6 cycles of BEACOPP escalated or 8 cycles of
and increased toxicity occurred with BEACOPP escalated. BEACOPP baseline given in 14 day interval (BEACOPP 14).
Therefore, BEACOPP escalated is not recommended for When 8 and 6 cycles of BEACOPP escalated were directly
patients >60 years of age. To reduce toxicity, the HD12 compared in the sequential analysis at 5 years of follow-up,
Table 20-5 Direct comparison of

Treatment 5-y PFS Difference (%) 5-y OS Difference (%) Reference
ABVD and BEACOPP variants
ABVD 68 13 (P = .038) 84 8 (P = NS) Federico et al. (2009)
4+2 81 92
ABVD 73 12 (P = .004) 84 5 (P = 0.38) Viviani et al. (2011)
4+4 85 89
ABVD 73 @ 4 years 15 87 4 (P = 0.28) Carde et al. (2012)
4+4 83 90
ABVD 75 18 (P = .007) 92 7 (P = 0.06) Mounier et al. (2014)
4+4 93 99
NS = not significant.

Frontline therapy for advanced-stage HL | 557
the difference between these two regimens was significant in Three hundred patients (43%) received IFRT, and although
terms of FFTF with 89.3% for 6× BEACOPP escalated and nonradiotherapy patients tended to be more often in CR,
85.4% for 8× BEACOPP escalated (P = 0.009). Overall sur- 5-year EFS was 86% with radiotherapy and 71% without
vival in the intention-to-treat population was also signifi- radiotherapy (P < 0.001) with a 5-year OS of 93% and 87%,
cantly higher in the 6× BEACOPP escalated group than respectively (P < 0.001).
in the 8× BEACOPP escalated group. T reatment-related In contrast, a randomized study of 739 patients with
mortality was 0.8% for 6× BEACOPP escalated, compared advanced-stage HL assigned patients with a CR after six to
with 2.1% for 8×. In addition, the number of sAML-MDS eight cycles of MOPP/ABV hybrid to observation or IFRT,
cases was 19 with 8× and only 2 with 6× BEACOPP escalated and demonstrated no difference in 5-year OS (P = 0.07) or
(P = 0.0001). EFS (P = 0.35) in the radiotherapy (n = 172, OS of 85% and
Thus, 6 cycles of BEACOPP escalated are clearly better EFS of 79%) group compared with the observation group
tolerated and more effective, resulting in an improved tumor (n = 161, OS of 91% and EFS of 84%). The HD12 trial ran-
control of 89.3% and an overall survival of 95.3% at 5 years, domized responding patients after BEACOPP with stage IIB,
and thus constitute the new standard of care in many coun- III, and IV HL and with bulk or residual tumor on CT imag-
tries. However, it should be noted that this regimen is associ- ing to either additional consolidative radiotherapy or no
ated with mandatory G-CSF support, dose adaptation upon radiotherapy. In this trial, 730 patients were randomized to
toxicity as well as treating higher-risk patients within the radiotherapy or no radiotherapy and 5-year FFTF was 87%
hospital particularly for the first course. An earlier analysis in those patients who did not receive radiotherapy, com-
had demonstrated that among 3,402 patients treated with pared to 90.4% in the radiotherapy arm (P = 0.08). However,
BEACOPP escalated, that patients ≥40 years with poor per- high-risk patients received radiation irrespective of their
formance status had a significantly higher risk of treatment randomization.
related mortality. In contrast, the majority of patients (2,164) In the GHSG HD15 trial, patients who had residual dis-
who were younger than 40 years and had a good ECOG per- ease ≥2.5 cm after 6-8 cycles of BEACOPP were evaluated by
formance status had an overall treatment related mortality PET, and those patients who were PET positive received
rate of 0.7%. 30 Gy radiotherapy given in 1.8-2.0 Gy five times weekly.
From the total of 2,182 patients, 739 had residual disease
≥2.5 cm; PET was positive in 191 patients (26%). PFS at
Radiation therapy as consolidation in
48 months was 92.6% in PET− and 86.2% in PET-positive
stage III-IV HL
patients. Although the radiation of PET-positive patients
Several studies have examined the role of consolidative with residual mass was not performed in a randomized fash-
radiotherapy in patients with advanced-stage HL, and to ion and there was no biopsy-proven active disease, the high
date, no study has demonstrated a clear OS advantage with tumor control rate suggests that radiating PET-positive dis-
combined modality therapy in patients achieving a CR or ease after BEACOPP is a pragmatic and feasible approach
partial remission (PR) with chemotherapy alone. The H89 that might also be applicable to other patients.
Groupe d’Etude des Lymphoma de l’Adulte (GELA) study
randomized 533 patients with stage III-IV HL to six cycles of
Autologous transplant as consolidation in
MOPP/ABV hybrid or six cycles of ABVPP followed by
stage III-IV HL
either two more cycles of chemotherapy or STNRT for
patients achieving a CR or PR after six cycles. Ten-year OS Several trials have examined the role of autologous trans-
was superior in the chemotherapy-alone arms (90% for plant to improve outcomes in patients with high-risk,
ABVPP × 8, 78% for MOPP/ABV × 8, 82% for MOPP/ABV advanced-stage HL, and to date none have demonstrated a
× 6 with radiation, and 77% for ABVPP × 6 with radiation, role for this following ABVD chemotherapy. A European
P = 0.03). Using an ABVD backbone, Laskar et al (2004) intergroup trial randomized 163 patients with stage III-IV
demonstrated an improvement in 8-year EFS and OS with HL and two risk factors (elevated LDH, bulky disease, stage
chemotherapy and IFRT in patients achieving a CR; how- IV with two or more extranodal sites, anemia, or inguinal
ever, this trial was small with only 179 patients randomized involvement) achieving a CR or PR after four cycles of
to observation versus radiotherapy and included all stages ABVD or doxorubicin containing induction (MOPP/ABVD,
(n = 80 with stage III-IV disease). A larger, although non- MOPP/ABV, CVPP/ABV) to either four additional cycles of
randomized, UK study restricted to advanced-stage HL the same induction chemotherapy or autologous stem cell
patients analyzed outcomes in 807 patients treated with transplantation after BEAM (carmustine, etoposide, cytara-
six cycles of ABVD, CHLVPP/PABlOE, or CHLVPP/EVA bine, and melphalan) or CBV (cyclophosphamide, carmus-
and IFRT to sites of residual masses or bulky disease. tine, and etoposide) conditioning regimens. With continued

chemotherapy, the 5-year FFS with continued chemother- traditional chemotherapy backbones to further improve
apy was 82% compared with 75% (P = 0.4) with consolida- outcomes in advanced-stage HL. Brentuximab vedotin (BV)
tive autologous transplantation, and the 5-year OS were is an anti-CD30 antibody-drug conjugate, which has been
88% and 88% (P = 0.99), demonstrating no clear benefit approved by the US Food and Drug Administration (FDA)
from early high-dose consolidation in high-risk advanced- for the treatment of relapsed and refractory HL. This drug is
stage HL. Two other studies using hybrid induction chemo- further described in the section “Brentuximab vendontin”
therapy regimens versus chemotherapy followed by and recently has been added to the ABVD regimen. In a
myeloablative transplant in 126-158 high-risk patients also phase 1 trial, patients with stage III-IV cHL received ABVD
similarly demonstrated no difference in OS with frontline with escalating doses of BV ranging from 0.6-1.2 mg/kg. Side
transplantation. effects included significant pulmonary toxicity, leading to
removal of bleomycin from the regimen and treatment of an
expanded cohort of patients with AVD plus BV. This combi-
Future directions and upcoming studies in
nation was found to be both safe and effective. A prospec-
frontline therapy for advanced-stage HL
tively randomized study in patients with stage III and IV cHL
Most cooperative groups today also include bulky stage II as comparing the AVD plus BV regimen to traditional ABVD is
well as stage IIB patients as advanced-stage disease. In the currently ongoing. In addition, there is also a new substan-
future, therapy for advanced-stage HL may be tailored based tially modified BEACOPP variant that incorporates BV;
on a patient’s pretreatment risk factors (clinical or selected based on promising phase 2 data, this variant (“BRECADD”)
biologic markers) or on results of interim PET or CT. In pre- is currently being compared to 6 cycles of BEACOPP esca-
vious studies, 2-year PFS for patients with a positive PET or lated. Long-term follow-up will be needed from these ran-
CT after two cycles of ABVD were 12.8% compared with domized studies; however, these approaches may offer an
95% for interim PET- or CT-negative patients (P < 0.0001), alternative strategy for improving therapy in high-risk
and interim PET or CT results were more predictive of out- advanced-stage patients using a novel targeted agents.
come than IPS score in multivariable analysis. Although the
prognostic potential of PET or CT may be dependent on ini- Key points
tial treatment, timing of scans, and definitions of PET nega-
tivity, PET or CT does represent an important tool that may • ABVD results in superior CR rates and 5-year FFS when
direct future therapy. To date, altering therapy based on compared with MOPP with less toxicity than MOPP/ABV hybrid
interim PET or CT results remains investigational; however, or MOPP/ABVD in patients with advanced-stage HL.
• Stanford V (combined chemotherapy over 12 weeks followed
a number of trials are ongoing examining this question in
by radiotherapy) results in similar 5-year FFS and OS as ABVD in
patients with advanced-stage HL. Risk-adapted therapy was
patients with advanced-stage HL.
first described by Dann et al (2007), who prospectively
• Escalated BEACOPP is associated with superior PFS and FFTF in
assigned 108 patients with unfavorable HL (stages I-II with patients with advanced-stage HL and may be considered as
risk factors, B symptoms, bulky disease, or stages III-IV) to frontline therapy for younger patients. Compared to 8 cycles,
either two cycles of BEACOPP escalated if the IPS score was 6 cycles of BEACOPP escalated were shown to be more effective
≥3 or BEACOPP baseline if the IPS score was <3 followed by and better tolerated. The benefit with respect to OS formally
interim PET or CT or gallium scan. Patients with a positive remains unclear, however, as patients failing ABVD often can
interim scan received four more cycles of BEACOPP esca- undergo effective salvage therapy and stem cell transplant.
lated, and those with a negative interim scan received four • Consolidative radiotherapy following chemotherapy is
cycles of BEACOPP baseline. Using this IPS and imaging- controversial in patients with advanced-stage HL treated with
directed therapeutic approach resulted in 5-year EFS and OS ABVD. In patients treated with BEACOPP, only PET-positive
residual disease ≥2.5 cm should be radiated.
of 85% and 90%, respectively. Currently, UK and US inter-
group studies are exploring intensification of therapy in
patients who are interim PET or CT positive after two cycles Therapy for relapsed or refractory HL
of initial ABVD therapy, whereas the GHSG HD18 trial is
reducing therapy after two cycles of initial BEACOPP esca- Clinical case
lated in patients who are interim PET or CT negative. Cen-
tral review of interim PET or CT is critical in these studies A 32-year-old man presented with stage IVB cHL in 2009
involving the bone marrow, liver, lungs, spleen, and multiple
using PET or CT to alter treatment approaches.
vertebrae. He received six cycles of ABVD with a negative PET or
In addition to efforts to risk-stratify therapy based on ini-
CT after cycles 2 and 4. PET or CT 1 month after cycle 6 demon-
tial high-risk features or interim PET or CT, efforts also are
strated a new liver lesion and biopsy confirmed HL. He received
being made to incorporate promising new therapies into

Therapy for relapsed or refractory HL | 559
high-dose cyclophosphamide, methotrexate, and etopo-

Clinical case (continued) side followed by BEAM and then transplant. In this trial,
three cycles of ICE (ifosfamide, carboplatin, etoposide), achieved 284 patients were included; there was higher toxicity in the
a second CR on PET or CT, and underwent autologous stem cell more aggressively treated group and no difference in terms
transplantation in 2010. One year following transplant, he devel- of tumor control and OS between the two groups. Thus,
oped progressive mediastinal and intra-abdominal adenopathy more chemotherapy does not seem to further improve the
and new pulmonary nodules, and biopsy of a retroperitoneal outcome of these patients and new drugs with different
lymph node by endoscopic ultrasound confirmed recurrent HL. mechanism of action are needed.
He then received brentuximab vedotin for 10 cycles, achieving
For patients who are refractory to frontline or salvage
a PR. Brentuximab vedotin initially was given every 3 weeks,
therapy, PFS of 25%-38% with high-dose therapy and stem
but due to neuropathy and neutropenia, the cycle length was
cell transplantation has been reported, suggesting that these
increased to 4 weeks, and he remained on it for 16 cycles. He
has one brother who is not an HLA match, but he does have
patients also benefit from HDCT and ASCT. A retrospective
several donor options through the National Marrow Donor study from Vancouver also demonstrated 15-year OS of 39%
Program registry. in patients with refractory disease to initial induction ther-
apy, compared with 67% in chemosensitive patients. In a
retrospective analysis of 175 patients with stable disease (SD)
Salvage therapy and autologous stem
or progressive disease (PD) to initial induction and who also
cell transplant
failed to respond according to CT findings to salvage ther-
Salvage chemotherapy followed by autologous stem cell apy, 5-year OS and PFS were 36% and 32%. Therefore, in
transplant is the standard of care for patients with relapsed patients with primary refractory disease defined either as
or refractory HL. With respect to salvage regimens before failure to achieve a CR on CT (not PET/CT) with frontline
autologous transplant, patients typically receive two to chemotherapy or a short remission duration who respond to
three cycles and then proceed to transplant. There are no salvage therapy, high-dose therapy with ASCT may lead to
randomized data on optimal salvage regimens, however, prolonged, durable remissions.
and numerous options exist. Regimens include ICE, GVD PET/CT may be helpful in guiding treatment approaches
(gemcitabine, vinorelbine, liposomal doxorubicin), DHAP for these patients. A number of studies recently have dem-
(dexamethasone, cytarabine, cisplatin), ESHAP (etoposide, onstrated the prognostic value of FDG PET/CT in this set-
methylprednisolone, cytarabine, cisplatin), GDP (gem- ting with EFS/PFS of 10%-31% in patients who are PET
citabine, dexamethasone, and cisplatin), IGEV (ifosfamide, positive compared with 68%-93% for patients with a nega-
gemcitabine, vinorelbine, prednisolone), mini-BEAM (car- tive PET/CT before stem cell transplantation. Although
mustine, etoposide, cytarabine, melphalan), and Dexa- using PET/CT to determine eligibility for stem cell trans-
BEAM (dexamethasone, carmustine, etoposide, cytarabine, plantation remains investigational, it is reasonable to rec-
melphalan) (Table 20-6) with responses ranging from 70% ommend two to three cycles of salvage chemotherapy,
to 90%. Ideally, the salvage regimen chosen should result in confirmation of response by PET/CT, and then ASCT in
a high overall response rate with acceptable toxicity and responding patients. For those with clearly progressive dis-
not impair stem cell mobilization if transplantation is ease on PET/CT, alternative salvage regimens should be
planned. offered and, if patients respond, ASCT could be considered.
There are two prospectively randomized trials comparing For those patients with improving disease on CT scan, but
standard dose with high dose chemotherapy (HDCT) in persistent PET positivity, ASCT is still recommended based
patients with relapsed HL: The British National Lymphoma on previous retrospective data and the possibility of false-
Group randomized 40 patients with relapsed disease to positive PET/CT. Further prospective study is needed in this
either BEAM followed by autologous stem cell transplant patient population that incorporates centrally reviewed
(ASCT) or mini-BEAM alone, demonstrating a significant PET/CT into the response assessment and transplant
PFS benefit (P = 0.005) with HDCT. A larger trial of 161 determination.
chemosensitive patients randomized to two cycles of Tandem transplantation has been evaluated with poor
Dexa-BEAM and ASCT or two more cycles of Dexa-BEAM risk or refractory disease and remains investigational. In the
demonstrating a 3-year FFTF of 55% with transplantation largest multicenter trial of 247 patients, 105 patients with
compared with 34% without transplant. Neither trial, how- primary refractory (defined as less than a PR or progression
ever, demonstrated an OS benefit perhaps because of limited within the first 90 days from induction doxorubicin contain-
follow-up or small patient numbers. A pan European trial ing chemotherapy) or at least two risk factors (time to relapse
then compared two cycles of DHAP followed by HDCT and <12 months, stage III-IV at relapse, or relapse within irradi-
ASCT with a variant that included additional single agent ated sites) underwent tandem ASCT, while 95 patients with

Table 20-6 Salvage combination chemotherapy regimens utilized for relapsed or refractory Hodgkin lymphoma
Method of
Regimen Drugs administration When administered Cycle
GVD (not previously Gemcitabine 1,000mg/m2 IV Days 1 and 8 Q21 days
transplanted) Vinorelbine 20 mg/m2 IV Days 1 and 8
Liposomal doxorubicin 15 mg/m2 IV Days 1 and 8
GVD (previously Gemcitabine 800 mg/m2 IV Days 1 and 8 Q21 days
transplanted) Vinorelbine 15 mg/m2 IV Days 1 and 8
Liposomal doxorubicin 10 mg/m2 IV Days 1 and 8
ICE Ifosfamide 5,000 mg/m2 IV over 24 h Day 2 Q14 days
Mesna 5,000 mg/m2 IV over 24 h Day 2
Etoposide 100 mg/m2 IV Days 1-3
Carboplatin AUC = 5 (maximum dose of 800 mg) IV Day 2
DHAP Dexamethasone 40 mg IV/PO Days 1-4 Q21 days
Cisplatin 100 mg/m2 IV over 24 h Day 1
Cytarabine 2,000 mg/m2 IV every 12 h Day 2
ESHAP Etoposide 40 mg/m2 IV Days 1-4 Q21 days
Methylprednisolone 500 mg IV Days 1-5
Cytarabine 2,000 mg/m2 IV Day 5
Cisplatin 25 mg/m2 CIV Days 1-4
Mini-BEAM BCNU (carmustine) 60 mg/m2 IV Day 1 Q21-28 days
Etoposide 75 mg/m2 IV Days 2-5
Cytarabine 100 mg/m2 IV every 12 h Days 2-5
Melphalan 30 mg/m2 (maximum of 50 mg) IV Day 5
Dexa-BEAM Dexamethasone 24 mg PO Days 1-10 Q28 days
BCNU (carmustine) 60 mg/m2 IV Day 2
Melphalan 20 mg/m2 IV Day 3
Etoposide 200 mg/m2 IV every 12 h Days 4-7
Cytarabine 100 mg/m2 IV every 12 h Days 4-7
G-CSF 300-480 mg SQ Day 9 until WBC > 2,500/µL
IGEV Ifosfamide 2,000 mg/m2 IV Days 1-4 Q21 days
Gemcitabine 800 mg/m2 IV Days 1 and 4
Vinorelbine 20 mg/m2 IV Day 1
Prednisolone 100 mg PO Days 1-4
GDP Gemcitabine 1,000 mg/m2 IV Days 1 and 8 Q21 days
Cisplatin 75 mg/m2 IV Days 1 and 8
Dexamethasone 40 mg PO Days 1-4
ChlVPP Chlorambucil 6 mg/m2 PO Days 1-14 Q28 days
Vinblastine 6 mg/m2 IV Days 1 and 8
Procarbazine 100 mg/m2 PO Days 1-14
Prednisone 40 mg PO Days 1-14
Brentuximab vedotin 1.8 mg/kg (capped at maximum of 100 kg) IV Day 1 Q 21 days
Source for GVD: Bartlett NL et al. Ann Oncol. 2007;18:1071-1079. Source for ICE: Moskowitz CH et al. Blood. 2001;97:616-623. Source for
DHAP: Josting A et al. Ann Oncol. 2002;13:1628-1635. Source for Mini-BEAM: Kuruvilla J et al. Cancer. 2006;106:353-360. Source for Dexa-
Beam: Josting A et al. Ann Oncol. 1998;9:289-295. Source for IGEV: Santoro A et al. Haematologica. 2007;92:35-41. Source for GDP: Kuruvilla J
et al. Cancer. 2006;106:353-360. Source for Ch1VPP: Vose JM et al. J Clin Oncol. 1991;9:1421-1425. Source for brentuximab vedotin: Chen R
et al. J Clin Oncol. 2011;29:8031 [abstract].
AUC = area under the concentration-time curve; CIV = continuous intravenous; IV= intravenous; PO = per os (by mouth); Q = every;
SQ = subcutaneous.
intermediate risk disease (only one risk factor) received a FFP was 46% and OS 57%, respectively, which compares
single transplant. In this study, 5-year FFP and OS were 73% favorably to historically observed 3- to 5-year PFS of 25%-
and 85%, in the intermediate-risk group, for those in the 39% in primary refractory patients. Transplantation for HL
high-risk group who underwent tandem transplant, 5-year is discussed further in Chapter 14.

Therapy for relapsed or refractory HL | 561
Therapeutic options for patients relapsing of patients with relapsed or refractory HL after previous
after autologous stem cell transplantation ASCT. BV is composed of a CD30 antibody conjugated by a
plasma-stable linker to the antimicrotubule agent, mono-
Although many of the previously discussed combination sal-
methyl auristatin E (MMAE). In phase I testing, the maxi-
vage regimens, including ICE, GVD, DHAP, ESHAP, GDP,
mum tolerated dose was 1.8 mg/kg every 3 weeks with
and IGEV, have significant activity in patients with HL that
an overall response rate of 38% in 45 patients, including
has progressed after autologous stem cell transplant, the goal
11 complete responses. In a pivotal phase II study with
in this patient population often has been palliation, with
102 relapsed (29%) or refractory (71%) cHL patients who
minimization of symptoms as well as treatment-related tox-
previously had received a median of 3.5 prior therapies
icity. Therefore, single-agent regimens often are preferred,
(range 1-13), the overall response rate (ORR) was 75% with
and combination regimens are reserved for patients with
a 33% achieving CR. The median duration of response was
organ involvement or significant disease-related symptoms.
9.0 months for patients in CR or PR and 2.8 months for
A number of single-agent regimens can be utilized in the pal-
those with SD/PD; the OS was 40.5 months in this trial and
liative setting and include vinblastine, etoposide, gem-
grade 3-4 toxicities consisted of sensory neuropathy (8%),
citabine, and vinorelbine. With vinblastine, 4-6 mg/m2
neutropenia (20%), and thrombocytopenia (8%). There-
weekly or every 2 weeks until disease progression or toxicity,
fore, with its significant single-agent activity and tolerability,
response rates as high as 59% and median EFS of 14 months
BV should be considered as initial therapy for patients with
have been reported. Lomustine at a dose of 100-200mg/m2
relapsed HL post–autologous transplantation. BV may be
orally every 4-8 weeks could also be helpful in a palliative
administered for up to 16 cycles, with dose reductions or
setting. Gemcitabine and vinorelbine both have single-agent
delays if needed for myelosuppression or neuropathy.
activity in 39%-50% of patients. The histone deacetylase
More recently, the prospectively randomized phase III
inhibitor panobinostat also has activity in this population,
AETHERA trial randomized a total of 322 cHL patients after
including multiply relapsed disease. Selected patients with
treatment with high dose chemotherapy and autologous
nonbulky lymphadenopathy and no organ involvement who
stem cells between consolidation treatment with brentux-
are otherwise asymptomatic also could be observed in this
imab vedotin or placebo. Treatment was given at 1.8 mg/kg
setting.
in 3-week intervals for up to 16 cycles. In the primary end-
Radiotherapy should also be considered in the setting of
point, PFS per independent review facility, the hazard ratio
relapsed HL. In a retrospective analysis of salvage radiother-
of this trial was 0.57 (p = 0.001) with a median PFS in the
apy used in 100 patients at first treatment failure, typically
brentuximab vedotin arm of 42.9 months and 24.1 months
after COPP/ABVD initial therapy, 5-year FFTF and OS were
in the group treated with placebo. The results were even
28% and 51% with radiotherapy alone. Advanced stage at
more obvious in the per investigator analysis with a hazard
relapse and B symptoms adversely affected OS in multivari-
ratio of 0.50 and a 24-months PFS rate of 65% for brentux-
ate analysis. Therefore, in highly selected patients with lim-
imab vedotin and 45% for placebo. Although there was more
ited stage disease at relapse who may not be eligible for
toxicity with brentuximab vedotin, the overall feasibility was
autologous transplantation due to age and comorbid condi-
good with a median of 15 cycles of brentuximab vedotin
tions, IFRT may lead to prolonged remissions. For younger
received. Thus, consolidation treatment with brentuximab
patients with relapsed HL, because of potential risks of sec-
vedotin might become an option for patients with higher
ond malignancies within the radiation field and improved
risk of relapse.
survival with autologous stem cell transplantation, radio-
therapy alone is not recommended at first relapse. IFRT,
however, should be considered in these patients as consoli- Allogeneic transplant
dation postautologous transplant to bulky, nonirradiated
Allogeneic transplant has been used for patients with relapsed
sites or to sites of relapsed limited stage disease in previously
HL after prior autologous transplant, although the presence
nonirradiated fields. For those patients with limited stage
of a graft-versus-Hodgkin lymphoma effect remains contro-
relapse after transplantation, radiotherapy may lead to pro-
versial. Most trials of allogeneic transplant in HL demonstrate
longed remissions and may delay the need for palliative
2-year PFS rates of 30% and OS of 35%-60%. A European
chemotherapy.
Blood and Marrow Transplantation trial compared reduced
intensity (n = 89) to myeloablative (n = 79) allogeneic stem
cell transplant. In this trial, with reduced-intensity condition-
Brentuximab vedotin
ing, 1-year treatment-related mortality (TRM) was 23% and
The FDA approved brentuximab vedotin (BV) in 2011, a 5-year OS was 28%, compared with 46% 1-year TRM and
novel anti-CD30 drug-antibody conjugate for the treatment 5-year OS of 22% with myeloablative conditioning. Five-year

PFS in the reduced intensity group was 18%. Other prospec- and cardiovascular disease continually increased beginning
tive and retrospective studies have reported 2- to 3-year PFS 10-15 years from the start of treatment. Within the first
of 25%-32% and OS of 43%-64% with reduced-intensity 5 years from diagnosis, patients typically are monitored for
conditioning. Overall, these studies and others demonstrate HL recurrence with history and symptom-directed evalua-
that in selected patients with available donors who are at least tion, physical examinations, and laboratory testing (CBC,
a good PR, reduced-intensity allogeneic stem cell transplanta- platelets, chemistries, and ESR if elevated at initial diagnosis)
tion is an option for patients with relapsed or refractory HL every 2-3 months for the first 2 years and every 3-6 months
after prior autologous transplantation and may lead to pro- during years 3-5. Limited data exist, however, regarding the
longed DFS in 18%-32% of patients. utility of routine blood work in detecting relapsed disease,
and in one series, relapse was detected in 55% of patients by
history, 23% by chest x-ray, and 1% by laboratory findings.
Key points
In a second study of 107 patients, 22 patients relapsed and
• Salvage chemotherapy followed by autologous transplant 64% were detected clinically, 9% detected by laboratory test-
offers superior PFS compared with chemotherapy alone in ing, and 9% by CT imaging.
patients with relapsed, chemosensitive HL. With respect to imaging studies, although the convention
• Selected patients with chemorefractory HL may benefit from
has been to recommend CT of initially involved sites every
autologous transplantation; particularly if they respond to
6-12 months during the first 5 years from diagnosis, several
salvage therapy or achieve a negative PET or CT before
previous studies have demonstrated no survival benefit with
transplantation.
• Brentuximab vedotin leads to overall response rates of 75% in
routine CT surveillance, and it does not appear to be cost
patients with relapsed HL following autologous transplant. effective. A more recent examination of follow-up PET/CT
Additional options for patients relapsing after autologous demonstrated a high false-positive rate, with an overall posi-
transplant include close observation if asymptomatic and no organ tive predictive value of only 28%, limiting its utility as a
compromise; combination regimens in Table 20-6; single agents follow-up tool for HL. In this study, 161 patients had 299
like gemcitabine, vinorelbine, or vinblastine; or radiotherapy. routine or clinically indicated follow-up PET/CTs (defined
as PET/CT performed based on clinical suspicion), and in
Follow-up of patients with HL this setting the true positive rates were only 5% and 13%,
respectively. Therefore, with the low risk for relapse in most
patients with HL, and no demonstrated survival benefit with
Clinical case
routine surveillance CT or PET/CT, follow-up should con-
An 18-year-old nonsmoking man with no history of cardiac dis- sist of history and physical exam with only symptom-
ease, diabetes, or elevated cholesterol presented with bulky stage directed imaging during the first 5 years after HL diagnosis.
IIB cHL involving the mediastinum and bilateral supraclavicular The risks of routine laboratory and imaging studies, includ-
nodes. He received six cycles of ABVD with mantle field irradia- ing false-positive findings, should be discussed with the
tion. He was followed every 6 months with CT scans for 2 years
patients, and the use of routine imaging should be limited
and then annually with CT scans until year 5 with no recurrence.
without clinical suspicion. Routine surveillance PET/CT
After his fifth year, he relocated for a new job opportunity and
currently is not recommended based on low-positive predic-
followed only as needed with a primary care physician (PCP).
About 9 years after diagnosis, he acutely developed nausea
tive value and associated costs of this testing and workup of
and chest discomfort and was seen in a local emergency room. false-positive results.
Because of lack of cardiac risk factors and initially normal elec- Secondary, late therapy-related effects in HL survivors
trocardiogram and troponin, he was admitted to a nonteaching include hypothyroidism, fertility issues, secondary cancers,
service for observation with the thought that this was gastro- and cardiovascular disease. The risks of second malignancies
intestinal discomfort. Subsequent troponins continued to rise and cardiovascular disease continue even beyond 30 years
and the patient was urgently taken to cardiac catheterization, from diagnosis. Therefore, monitoring of late complications
demonstrating a 90% occluded left-anterior descending artery. is a lifelong endeavor for HL survivors. In a meta-analysis,
second cancers were more commonly encountered in
Follow-up of patients with HL must address both the risk of patients receiving combined modality treatment or radiation
relapse as well as potential late complications of therapy. In a alone compared with chemotherapy alone, and no signifi-
study of 1,261 patients treated for HL before the age of 41 cant differences in the second malignancy rate were observed
from 1965-1987, 534 patients died, including 54% due to with IFRT versus EFRT so far. Therefore, any patient receiv-
HL, 22% from second malignancies, and 9% from cardio- ing previous radiotherapy should be monitored for second
vascular disease. The likelihood of HL recurrence declined malignancy and cardiovascular disease. Specifically, annual
after 5 years, whereas the incidence of second malignancies breast screening (typically mammography) is recommended

Follow-up of patients with HL | 563
8-10 years after completion of treatment or at the age of 40, with nontransplanted patients with HL. Second malignan-
whichever comes first, in women who received chest or axil- cies occurred in 15 patients, at a median of 9 years posttrans-
lary radiation. The risk of secondary breast cancers is associ- plantation and consisted of AML or MDS (n = 6), non-HL
ated with young age at the time of radiation, and women <30 (n = 3), non–small cell lung cancer (n = 2), colon cancer (n =
years of age are particularly at risk. Screening MRI, in addi- 2), gastric cancer (n = 1), and adenocarcinoma of unknown
tion to mammography, may be of benefit in this patient primary (n = 1). All patients with AML or MDS had MOPP
population, though the risk of false positive findings on MRI as part of their initial treatment regimen. In 100 patients
is substantial. Lung cancer risk is increased in patients receiv- treated with autologous stem cell transplantation in Van-
ing mediastinal radiation, particularly patients with a smok- couver, second malignancies occurred in seven patients at a
ing history, and chest imaging annually should be considered median time of 4.2 years from transplantation, and consisted
for these patients at greatest risk. Cardiovascular disease, of AML or MDS, glioblastoma, renal cell carcinoma, colon
including increased risk of coronary artery disease, pericar- carcinoma, non-HL, and breast cancer. In this series, five
dial disease, cardiomyopathy, and valvular disease, also is patients developed cardiovascular disease, including myo-
observed in HL survivors, particularly after mediastinal radi- cardial infarctions 4-11 years posttransplant, arrhythmia,
ation or anthracycline-based chemotherapy, starting about 5 and aortic stenosis.
years after treatment and also is associated with age at treat- In the largest retrospective series of outcomes postautolo-
ment. Although optimal screening strategies are unclear, gous transplantation, 16 of 494 patients who underwent
monitoring and aggressive management of cardiovascular autologous stem cell transplantation for relapsed or refrac-
risk factors, including smoking, hypertension, diabetes, and tory disease developed second malignancies, and second
hyperlipidemia, is recommended along with consideration malignancy was associated with use of total body irradiation
of a baseline stress test or echocardiogram. (TBI) as part of conditioning, age ≥40 years, or use of radia-
Other late toxicities associated with radiotherapy include tion before transplantation. Therefore, all survivors of HL
hypothyroidism, which can occur in up to 50% of patients, who are transplanted should be monitored lifelong for sec-
and radiation pneumonitis or lung fibrosis, which is fairly ond malignancies, with close attention to those treated with
uncommon occurring in 3%-10% of patients. Annual thy- radiation either pre- or posttransplant or with TBI during
roid function tests are recommended for patients with radia- transplant conditioning. In addition, patients typically expe-
tion to the neck or upper mediastinum, and evaluation for rience hypogonadism posttransplantation, and monitoring
pulmonary fibrosis should be considered in symptomatic for consequences of hormonal deficiency is recommended,
patients. including monitoring for bone mineral density reduction
With respect to secondary MDS and leukemia, these con- using DEXA scanning.
ditions affect up to 1% of patients receiving ABVD chemo- Immunity typically wanes postautologous transplanta-
therapy and have been observed in up to 3.2% of patients tion, and it is recommended that patients receive pneumo-
treated with eight cycles of BEACOPP escalated. In contrast, coccal, tetanus, Haemophilus influenzae type b, hepatitis B,
with 6 cycles of BEACOPP escalated and radiation only to and annual influenza vaccinations. Measles, mumps, and
PET-positive residual disease ≥2.5 cm, the AML/MDS rate rubella (MMR) and varicella vaccinations can be considered
was only 0.2%. With respect to fertility, patients treated with in immunocompetent patients no sooner than 24 months
MOPP or BEACOPP have a high risk of infertility depending posttransplantation.
on the age at treatment and the number of cycles received.
All patients receiving chemotherapy should be counseled
about this risk and referred for sperm banking or reproduc- Key points
tive endocrinology evaluation. The impact of ABVD on
• Routine follow-up for HL survivors consists of history and
gonadal function seems more modest, but robust data from
directed physical examination with symptom directed laboratory
larger cohorts are lacking. testing or imaging. Surveillance CT, PET or CT, and laboratory
Finally, in addition to these risks, patients who undergo testing have not been shown to improve survival or increase
autologous stem cell transplant for relapsed disease should detection of relapsed disease.
be monitored for risks of secondary leukemia, other second- • Monitoring for secondary malignancies and cardiovascular
ary malignancies, hypogonadism and its complications, disease is a lifelong endeavor for HL survivors. Annual mammog-
including declines in bone mineral density, and also should raphy is recommended starting 8-10 years after completion of
be considered for revaccination. In a retrospective study of treatment in women treated with chest or axillary radiation.
153 patients treated with autologous stem cell transplant for Smoking cessation, cardiovascular risk assessment, and monitor-
relapsed HL, the relative risk of second malignancies was ing for hypothyroidism are recommended, particularly in
patients receiving mediastinal or neck radiation.
6.5 compared with the general population and 2.4 compared

Pediatric HL
HL represents 7% of childhood cancers and is rare in children mass failed to improve despite 6 months of intermittent steroids
under the age of 10 but is the most common malignancy in the and antibiotics, and subsequent biopsy demonstrated atypi-
late teens. NS accounts for the majority of cases at approxima cal large cells with large nuclei that were CD20, PAX-5, BCL-2,
tely 70%. Mixed cellularity accounts for 30% and NLPHL and CD45 positive and CD15 and CD30 negative, consistent
with NLPHL. CTs of the C/A/P demonstrated bilateral cervical
accounts for 1%-15%. LD is rare, except in association with HIV.
adenopathy, but no other sites of disease and bone marrow
The vast majority of pediatric patients with HL are cured of
biopsy was negative.
disease. Most children in the US with HL are treated in large
referral centers, often in the context of clinical trials. As with
adults, limiting exposure to radiation, avoiding alkylator- NLPHL is an uncommon subtype of HL, representing about
based regimens, such as MOPP, and reducing anthracycline 5% of cases, with unique pathologic features distinguishing
exposure has been employed to reduce secondary malignan- it from cHL. The neoplastic cell is a large cell, the LP cell,
cies, infertility, and other late toxicities given the long life otherwise known as popcorn cell due to the large-folded, or
expectancy of these patients. Patients are typically risk strati- multilobulated nucleus (Figure 20-1). Unlike the classic
fied with early-stage disease defined as stage I and IIA; HRS cell, these cells are typically CD30 and CD15 negative,
advanced stage includes patients with stage III and IV disease, with CD19, CD20, CD45, and CD79a positivity (Figure
bulky mediastinal disease, and all patients with B symptoms. 20-2) and are also PAX-5 and OCT-2 positive. The sur-
In the recently published COG AHOD0031 study of 1,712 rounding background lymphocytes are predominantly small
children with intermediate risk HL, patients were stratified CD20+ B-cells, with rare eosinophils, neutrophils, and
according to rapid early response (RER) defined by at least 60% plasma cells (Figure 20-3). Surrounding the LP cells, CD4+
reduction in lymph node diameter by CT following 2 cycles T-cell rosettes are found as well as CD21-positive follicular
of ABVE-PC (doxorubicin, bleomycin, vinblastine, etoposide, dendritric cells, consistent with the germinal center deriva-
prednisone, cyclophosphamide). RERs received 2 additional tion of this malignancy.
cycles of ABVE-PC followed by PET/CT. Patients who were Because of the rare occurrence of this malignancy, presen-
PET negative were randomized to observation versus IFRT. tation, treatment, and patient outcomes are not well
PET positive patients received IFRT. Slow early responders described. In a retrospective analysis of 8,298 patients
(SERs) who were PET positive were treated with 2 additional enrolled on clinical trials for HL through the GHSG, 394
cycles of ABVE-PC with or without 2 cycles of DECA (dexa- patients had NLPHL. In this series, the median age at diag-
methasone, etoposide, cisplatin and cytarabine). All SERs recei nosis was 37 years, 75% of patients were male, and 79% had
ved IFRT. Overall, the 4 year EFS and OS were 85% and 98%; early-stage disease. The presence of B symptoms or bulky
87%/99% for RERs and 77%/95% for SERs (P < 0.001). No dif- disease is unusual, observed in <10% of patients. Unlike
ference in outcome was seen in RERs with or without the inclu- cHL, patients with NLPHL typically have peripheral adenop-
sion of IFRT with EFS at 4 years at 87% (P = 0.87). Slow early athy (axillary or inguinal) at diagnosis rather than central or
responders who were PET positive benefitted from the addition
of DECA with EFS 71% vs 55% (P = 0.05).
Other approaches in children include OEPA (vincristine,
etoposide, prednisone and doxorubicin) for males and
OPPA (vincristine, procarbazine, prednisone, and doxoru-
bicin) for females, the Stanford V regimen, COPP/ABV and
escalated BEACOPP with or without low dose radiotherapy.
Additional risk adapted strategies to reduce the exposure to
chemotherapy and RT are under investigation.
Nodular lymphocyte-predominant HL
Clinical case
A 19-year-old college lacrosse player presented with left-sided
cervical adenopathy and large parotid mass of 6 cm, initially
thought to be secondary to acute infectious mononucleosis. The
Figure 20-1 LP or popcorn cells in NLPHL with typical folded,
multilobulated nucleus.

Nodular lymphocyte-predominant HL | 565
NLPHL. With T-cell-rich B-cell lymphoma, large atypical

B-cells that are CD20 positive are surrounded by an abun-
dant background of T-cells and histiocytes. Because ~5% of
NLPHL eventually develop NHL, including T-cell-rich B-cell
NHL or diffuse large cell lymphoma, biopsy of recurrent
lymph nodes is necessary to determine therapy at relapse. In
a series of 22 patients treated with rituximab for NLPHL,
nine patients relapsed, including five who underwent biopsy
at recurrence and, of these, two had diffuse large cell NHL
within 13 months of follow-up.
With respect to treatment for NLPHL, no standard front-
line or relapsed therapy exists, although a number of options
are available. In the large GHSG series, outcomes for these
patients appear to be excellent, with ORR of 85%-91% in
patients with early-stage NLPHL compared with ORR of
Figure 20-2 CD20 staining on large LP cells in NLPHL.
83%-86% in patients with early-stage cHL treated with the
same regimens. For advanced-stage patients, outcomes are
mediastinal involvement; nodal involvement is not contigu-
similarly good with ORR of 78% in patients with NLPHL
ous, and extranodal involvement is uncommon.
compared with 78% in patients with stage III-IV classical
An association exists with this subtype of lymphoma and
HL. FFTF at 50 months was 88% in patients with NLPHL
a benign condition, progressive transformation of germinal
and 82% for patients with cHL. Interestingly, late relapses
centers, as well as with NHL, particularly T-cell-rich B-cell
>1 year after therapy are observed more commonly in
lymphoma and diffuse large B-cell lymphoma. Progressive
patients with LPHL (7.4%) compared with cHL (4.7%).
transformation of germinal centers is described as lymph
Adverse prognostic factors in LPHL include advanced stage,
nodes with large, well-defined nodules with an excess of
hemoglobin <10.5 g/dL, age >45 years, and lymphopenia
B-cells, or germinal centers overrun by lymphocytes. Pro-
(<8% of total white cell count).
gressive transformation of germinal centers may be observed
For early-stage NLPHL, typically IFRT alone is recom-
before, simultaneous with, or following a diagnosis of
mended. Given the excellent long-term survivals of these
NLPHL. This entity is thought to be a benign condition, but
patients, however, the risks of late secondary malignancies
as it occurs concurrently or following a diagnosis of NLPHL,
and cardiovascular disease must be considered. In a single
biopsy of recurrent adenopathy always is required in this dis-
institution study of 113 patients with stage I-II NLPHL,
ease to confirm relapse.
93 patients were treated with radiation alone, 13 patients
Likewise, T-cell-rich B-cell lymphoma and NLPHL can
received combined modality therapy, and seven received
occur simultaneously or in succession, and frequently T-cell-
chemotherapy alone. Twenty of 106 patients receiving radio-
rich B-cell lymphoma can be confused pathologically for
therapy relapsed (19%), compared with six of seven patients
treated with chemotherapy alone, and 10-year PFS for stage
I and II patients treated with radiotherapy were 89% and
72%, respectively. No differences were observed in patients
treated with extended-field, regional, or limited-field radio-
therapy, and the use of combined modality therapy did
not improve PFS or OS over radiotherapy alone. However,
12 patients receiving radiotherapy developed second cancers,
five of which were fatal. The GHSG evaluated 131 patients
with stage IA LPHL treated with extended-field (n = 45),
involved-field (n = 45), and combined modality treatment
(n = 41), and found an FFTF rate of 95% and OS of 99% at
43 months, with no differences with respect to FFTF or OS
among the three treatment arms. In this series, only one sec-
ondary cancer was observed—that is, gastric carcinoma in
the extended-field group. In a retrospective evaluation from
the Australasian Radiation Oncology Group of 202 patients
Figure 20-3 Low-power view of NLPHL. with stage I-II NLPHL treated with radiotherapy alone,

the 15-year FFP and OS were 83% and 82%, respectively. or CVP as part of the initial treatment with those for stage
Second malignancies included non-HL (n = 9), acute leuke- III-IV disease. Limited data exist comparing ABVD to alkyl-
mia (n = 1), and 18 carcinomas, including breast, lung, gas- ator therapy either alone or in combination with rituximab.
trointestinal, melanoma, and unknown primary carcinomas. Therefore, these regimens frequently are utilized as frontline
Deaths resulting from cardiac and respiratory complications and salvage therapy for relapsed-stage III-IV NLPHL.
after radiotherapy occurred in nine patients. In contrast to
these three trials demonstrating excellent PFS and OS with
Key points
radiotherapy alone in early-stage NLPHL, a retrospective
comparison of 32 patients treated with radiotherapy alone • NLPHL cells typically lack CD30 and CD15 staining and are
versus 56 patients receiving combined modality therapy with CD19, CD20, CD45, and CD79a positive.
ABVD for two cycles and radiotherapy demonstrated • NLPHL is associated with progressive transformation of
improved PFS survival (65% vs. 91%, P = 0.0024) with com- germinal centers (a benign condition) and also transformation to
bined modality therapy as compared with radiation alone. diffuse large B-cell or T-cell-rich B-cell NHL, so biopsy at relapse
is necessary.
Therefore, most series support favorable outcomes with
• Unlike HL, NLPHL is associated with noncontiguous nodal
radiotherapy alone in early stage IA NLPHL, and at least one
spread and late relapses. No standard therapy exists for NLPHL
series advocates treating these patients similar to current
and includes IFRT, combined modality therapy, or observation
cHL treatment approaches with combined modality therapy. for early-stage disease and combination chemotherapy with
Because of the risks of second malignancies and the excellent rituximab for advanced-stage disease as well as single-agent
long-term outcomes observed in patients with LPHL, in rituximab in the relapsed setting.
selected patients for whom the disease is completely resected,
observation could be discussed as alternative to IFRT.
Chemotherapy alone is used for non stage IA patients or Bibliography
for those with very high risk of late complications of radio-
therapy due to field size of radiotherapy required. In early- Epidemiology, pathology, pathogenesis, and
stage patients, for whom the risks of secondary malignancies risk factors
with radiotherapy are of concern, chemotherapy alone Biggar RJ, Jaffe ES, Goedert JJ, et al. Hodgkin lymphoma
occasionally is utilized. For example, a study conducted in and immunodeficiency in persons with HIV/AIDS. Blood.
the United Kingdom and France prospectively examined 2006;108:3786-3791.
the results of three cycles of cyclophosphamide, vinblas- Correa P, O’Conor GT. Epidemiologic patterns of Hodgkin’s
tine, and prednisolone (CVP) in children and adolescents disease. Int J Cancer. 1971;8:192-201.
with stage I-II NLPHL. In this trial, the ORR was 100% and Diepstra A, Niens M, Vellenga E, et al. Association with HLA
the 40-month FFTF and OS were 75% and 100%, respec- class I in Epstein-Barr-virus-positive and with HLA class III
tively. In a study of single-agent rituximab as frontline in Epstein-Barr-virus-negative Hodgkin’s lymphoma. Lancet.
2005;365:2216-2224.
therapy in 28 patients with stage IA NLPHL, the ORR was
Gutensohn N, Cole P. Childhood social environment and
100% and at 36 months the PFS was 81%. Therefore,
Hodgkin’s disease. N Engl J Med. 1981;304:135-140.
although the FFTF may be slightly lower with chemother- Hehn ST, Grogan TM, Miller TP. Utility of fine-needle aspiration
apy alone compared with radiotherapy or combined as a diagnostic technique in lymphoma. J Clin Oncol.
modality therapy, these early-stage patients who relapse 2004;22:3046-3052.
can be effectively salvaged with additional chemotherapy Hummel M, Anagnostopoulos I, Dallenbach F, et al. EBV
and radiotherapy, and such an approach may reduce the infection patterns in Hodgkin’s disease and normal lymphoid
rates of second malignancies. tissue: expression and cellular localization of EBV gene
In the advanced-stage setting, chemotherapy options products. Br J Haematol. 1992;82:689-694.
include six cycles of ABVD or BEACOPP, alkylator regimens Kuppers R, Rajewsky K, Zhao M, et al. Hodgkin disease: Hodgkin
(CVP or CHOP), or rituximab. Because of the CD20 expres- and Reed-Sternberg cells picked from histological sections show
sion on LP cells, rituximab increasingly is utilized in the clonal immunoglobulin gene rearrangements and appear to be
derived from B cells at various stages of development. Proc Natl
frontline treatment of advanced-stage NLPHL. In one of the
Acad Sci USA. 1994;91:10962-10966.
initial series examining the activity of this agent, Ekstrand
Landgren O, Engels EA, Pfeiffer RM, et al. Autoimmunity and
et al. (2003) reported an ORR of 96% in 22 patients, 12 with susceptibility to Hodgkin lymphoma: a population-based
previously untreated NLPHL; however, response duration is case-control study in Scandinavia. J Natl Cancer Inst.
limited, with a median FFP of 9.2 months. With the notable 2006;98:1321-1330.
activity of single-agent rituximab in relapsed and previously Landgren O, Pfeiffer RM, Kristinsson SY, et al. Survival patterns in
untreated HL, it is frequently combined with ABVD, CHOP patients with Hodgkin’s lymphoma with a pre-existing hospital

Bibliography | 567
discharge diagnosis of autoimmune disease. J Clin Oncol. Frontline therapy for early-stage HL
2010;28:5081-5087.
Loddenkemper C, Anagnostopoulos I, Hummel M, et al. Advani R, Hong F, Fisher RI, et al. Randomized phase III
Differential Emu enhancer activity and expression of BOB.1/ trial comparing ABVD + radiotherapy and the Stanford V
OBF.1, Oct2, PU.1, and immunoglobulin in reactive B-cell regimen in patients with stage I/II bulky mediastinal hodgkin
populations, B-cell non-Hodgkin lymphomas, and Hodgkin lymphoma: a subset analysis of the US intergroup trial E2496
lymphomas. J Pathol. 2004;202:60-69. [abstract]. Blood. 2010;116:abstract 416.
Mack TM, Cozen W, Shibata DK, et al. Concordance for Barnes JA, LaCasce AS, Zukotynski K, et al. End-of-treatment
Hodgkin’s disease in identical twins suggesting genetic but not interim PET scan predicts outcome in nonbulky
susceptibility to the young-adult form of the disease. N Engl J limited-stage Hodgkin’s lymphoma. Ann Oncol. 2011;22:
Med. 1995;332:413-418. 910-915.
Marafioti T, Hummel M, Foss HD, et al. Hodgkin and Reed- Behringer K, Goergen H, Hitz F, et al. Omission of dacarbazine or
Sternberg cells represent an expansion of a single clone bleomycin, or both, from the ABVD regimen in treatment of
originating from a germinal center B-cell with functional early-stage favourable Hodgkin’s lymphoma (GHSG HD13):
immunoglobulin gene rearrangements but defective an open-label, randomised, non-inferiority trial. Lancet.
immunoglobulin transcription. Blood. 2000;95:1443-1450. 2015;385:1418-1427.
Mueller N, Evans A, Harris NL, et al. Hodgkin’s disease and Boleti E, Mead GM. ABVD for Hodgkin’s lymphoma: full-dose
Epstein-Barr virus. Altered antibody pattern before diagnosis. chemotherapy without dose reductions or growth factors. Ann
N Engl J Med. 1989;320:689-695. Oncol. 2007;18:376-380.
Siegel R, Naishadham D, Jemal A. Cancer statistics, 2012. CA Bonadonna G, Bonfante V, Viviani S, et al. ABVD plus
Cancer J Clin. 2012;62:10-29. subtotal nodal versus involved-field radiotherapy in early-
Steidl C, Connors JM, Gascoyne RD. Molecular pathogenesis of stage Hodgkin’s disease: long-term results. J Clin Oncol.
Hodgkin’s lymphoma: increasing evidence of the importance of 2004;22:2835-2841.
the microenvironment. J Clin Oncol. 2011;29:1812-1826. Cosset JM, Henry-Amar M, Meerwaldt JH, et al. The EORTC trials
Swerdlow SH, Campo E, Harris N. World Health Organization for limited stage Hodgkin’s disease. The EORTC Lymphoma
Classification of Tumors of Haematopoeitic and Lymphoid Cooperative Group. Eur J Cancer. 1992;28A:1847-1850.
Tissues. Lyon: IARC Press; 2008. Engert A, Plutschow A, Eich HT, et al. Reduced treatment
von Wasielewski R, Werner M, Fischer R, et al. Lymphocyte- intensity in patients with early-stage Hodgkin’s lymphoma.
predominant Hodgkin’s disease. An immunohistochemical N Engl J Med. 2010;363:640-652.
analysis of 208 reviewed Hodgkin’s disease cases from the Engert A, Schiller P, Josting A, et al. Involved-field radiotherapy
German Hodgkin Study Group. Am J Pathol. 1997;150:793-803. is equally effective and less toxic compared with extended-field
radiotherapy after four cycles of chemotherapy in patients with
early-stage unfavorable Hodgkin’s lymphoma: results of the
HD8 trial of the German Hodgkin’s Lymphoma Study Group.
Clinical presentation, staging, and workup
J Clin Oncol. 2003;21:3601-3608.
Atkinson K, Austin DE, McElwain TJ, et al. Alcohol pain in Evens AM, Cilley J, Ortiz T, et al. G-CSF is not necessary to
Hodgkin’s disease. Cancer. 1976;37:895-899. maintain over 99% dose-intensity with ABVD in the treatment
Gobbi PG, Cavalli C, Gendarini A, et al. Reevaluation of of Hodgkin lymphoma: low toxicity and excellent outcomes in a
prognostic significance of symptoms in Hodgkin’s disease. 10-year analysis. Br J Haematol. 2007;137:545-552.
Cancer. 1985;56:2874-2880. Ferme C, Eghbali H, Meerwaldt JH, et al. Chemotherapy plus
Lister TA, Crowther D, Sutcliffe SB, et al. Report of a committee involved-field radiation in early-stage Hodgkin’s disease. N Engl
convened to discuss the evaluation and staging of patients J Med. 2007;357:1916-1927.
with Hodgkin’s disease: Cotswolds meeting. J Clin Oncol. Gallamini A, Hutchings M, Rigacci L, et al. Early interim
1989;7:1630-1636. 2-[18F]fluoro-2-deoxy-D-glucose positron emission tomography
Mauch PM, Kalish LA, Kadin M, et al. Patterns of presentation of is prognostically superior to international prognostic score in
Hodgkin disease. Implications for etiology and pathogenesis. advanced-stage Hodgkin’s lymphoma: a report from a joint
Cancer. 1993;71:2062-2071. Italian-Danish study. J Clin Oncol. 2007;25:3746-3752.
Ng AK, Bernardo MP, Weller E, et al. Long-term survival Hay AE, Klimm B, Chen BE, et al; NCIC Clinical Trials Group
and competing causes of death in patients with early-stage (Canada) and German Hodgkin Study Group (GHSG). An
Hodgkin’s disease treated at age 50 or younger. J Clin Oncol. individual patient-data comparison of combined modality
2002;20:2101-2108. therapy and ABVD alone for patients with limited-stage
Purz S, Mauz-Korholz C, Korholz D, et al. [18F] Hodgkin lymphoma. Ann Oncol. 2013;24:3065-3069.
Fluorodeoxyglucose positron emission tomography for Ho P, Sherman P, Grigg A. Intermittent granulocyte colony-
detection of bone marrow involvement in children and stimulating factor maintains dose intensity after ABVD therapy
adolescents with Hodgkin’s lymphoma. J Clin Oncol. complicated by neutropenia. Eur J Haematol. 2012; 88:
2011;29:3523-3528. 416-421.

Horning SJ, Hoppe RT, Advani R. Efficacy and late effects of chemotherapy with autologous stem cell transplantation: five-
Stanford V chemotherapy and radiotherapy in untreated year results of a randomized trial on behalf of the GOELAMS
Hodgkin’s diseaseL mature data in early and advanced stage Group. Cancer. 2008;113: 3323-3330.
patients. Blood. 2004;104:308. Bonadonna G, Valagussa P, Santoro A. Alternating non-cross-
Horning SJ, Hoppe RT, Breslin S, et al. Stanford V and resistant combination chemotherapy or MOPP in stage IV
radiotherapy for locally extensive and advanced Hodgkin’s Hodgkin’s disease. A report of 8-year results. Ann Intern Med.
disease: mature results of a prospective clinical trial. J Clin 1986;104:739-746.
Oncol. 2002;20:630-637. Bonadonna G, Zucali R, Monfardini S, et al. Combination
Hutchings M, Mikhaeel NG, Fields PA, et al. Prognostic value of chemotherapy of Hodgkin’s disease with adriamycin,
interim FDG-PET after two or three cycles of chemotherapy in bleomycin, vinblastine, and imidazole carboxamide versus
Hodgkin lymphoma. Ann Oncol. 2005;16:1160-1168. MOPP. Cancer. 1975;36:252-259.
Loeffler M, Diehl V, Pfreundschuh M, et al. Dose-response Borchmann P, Haverkamp H, Diehl V, et al. Eight cycles of
relationship of complementary radiotherapy following four escalated-dose BEACOPP compared with four cycles of
cycles of combination chemotherapy in intermediate-stage escalated-dose BEACOPP followed by four cycles of baseline-
Hodgkin’s disease. J Clin Oncol. 1997;15:2275-2287. dose BEACOPP with or without radiotherapy in patients with
Longo DL, Glatstein E, Duffey PL, et al. Radiation therapy advanced-stage hodgkin’s lymphoma: final analysis of the
versus combination chemotherapy in the treatment of early- HD12 trial of the German Hodgkin Study Group. J Clin Oncol.
stage Hodgkin’s disease: seven-year results of a prospective 2011;29:4234-4242.
randomized trial. J Clin Oncol. 1991;9:906-917. Canellos GP, Anderson JR, Propert KJ, et al. Chemotherapy of
Martin WG, Ristow KM, Habermann TM, et al. Bleomycin Advanced Hodgkin’s Disease with MOPP, ABVD, or MOPP
pulmonary toxicity has a negative impact on the outcome of Alternating with ABVD. N Engl J Med. 1992;327:1478-1484.
patients with Hodgkin’s lymphoma. J Clin Oncol. 2005;23: Carde P, Mounier N, Fortpied C, et al. ABVD (8 cycles) versus
7614-7620. BEACOPP (4 escalated cycles => 4 baseline) in stage III-IV
Meyer RM, Gospodarowicz MK, Connors JM, et al. ABVD alone high-risk Hodgkin lymphoma (HL): First results of EORTC
versus radiation-based therapy in limited-stage Hodgkin’s 20012 Intergroup randomized phase III clinical trial [abstract].
lymphoma. N Engl J Med. 2012;366:399-408. J Clin Oncol. 2012;30(suppl):abstract 8002.
Santoro A, Bonadonna G, Valagussa P, et al. Long-term results of Cisesi T, Bellei M, Luminari S, et al. Long-term follow-up analysis
combined chemotherapy-radiotherapy approach in Hodgkin’s of HD9601 trial comparing ABVD versus Stanford V versus
disease: superiority of ABVD plus radiotherapy versus MOPP MOPP/EBV/CAD in patients with newly diagnosed advanced-
plus radiotherapy. J Clin Oncol. 1987;5:27-37. stage Hodgkin’s lymphoma: a study from the Intergruppo
Straus DJ, Portlock CS, Qin J, et al. Results of a prospective Italiano Linfomi. J Clin Oncol. 2011;29:4227-4233.
randomized clinical trial of doxorubicin, bleomycin, Dann EJ, Bar-Shalom R, Tamir A, et al. Risk-adapted BEACOPP
vinblastine, and dacarbazine (ABVD) followed by radiation regimen can reduce the cumulative dose of chemotherapy
therapy (RT) versus ABVD alone for stages I, II, and IIIA for standard and high-risk Hodgkin lymphoma with no
nonbulky Hodgkin disease. Blood. 2004;104:3483-3489. impairment of outcome. Blood. 2007;109:905-909.
van der Kaaij MA, Heutte N, Meijnders P, et al. Premature Diehl V, Franklin J, Pfreundschuh M, et al. Standard and
ovarian failure and fertility in long-term survivors of Hodgkin’s Increased-Dose BEACOPP Chemotherapy Compared with
lymphoma: a European Organisation for Research and COPP-ABVD for Advanced Hodgkin’s Disease. N Engl J Med.
Treatment of Cancer Lymphoma Group and Groupe d’Etude 2003;348:2386-2395.
des Lymphomes de l’Adulte Cohort Study. J Clin Oncol. Duggan DB, Petroni GR, Johnson JL, et al. Randomized
2012;30:291-299. comparison of ABVD and MOPP/ABV hybrid for the treatment
von Tresckow B, Plutschow A, Fuchs M, et al. Dose-intensification of advanced Hodgkin’s disease: report of an intergroup trial.
in early unfavorable Hodgkin’s lymphoma: final analysis of J Clin Oncol. 2003;21:607-614.
the German Hodgkin Study Group HD14 trial. J Clin Oncol. Engert A, Diehl V, Franklin J, et al. Escalated-dose BEACOPP
2012;30:907-913. in the treatment of patients with advanced-stage Hodgkin’s
lymphoma: 10 years of follow-up of the GHSG HD9 study.
J Clin Oncol. 2009;27:4548-4554.
Frontline therapy for advanced-stage HL
Engert A, Haverkamp H, Kobe C, et al. Reduced-intensity
Aleman BM, Raemaekers JM, Tirelli U, et al. Involved-field chemotherapy and PET-guided radiotherapy in patients
radiotherapy for advanced Hodgkin’s lymphoma. N Engl J Med. with advanced stage Hodgkin’s lymphoma (HD15 trial): a
2003;348:2396-2406. randomised, open-label, phase 3 non-inferiority trial. Lancet.
Arakelyan N, Berthou C, Desablens B, et al. Early versus late 2012;379:1791-1799.
intensification for patients with high-risk Hodgkin lymphoma-3 Federico M, Bellei M, Brice P, et al. High-dose therapy and
cycles of intensive chemotherapy plus low-dose lymph node autologous stem-cell transplantation versus conventional therapy
radiation therapy versus 4 cycles of combined doxorubicin, for patients with advanced Hodgkin’s lymphoma responding to
bleomycin, vinblastine, and dacarbazine plus myeloablative front-line therapy. J Clin Oncol. 2003;21:2320-2325.

Bibliography | 569
Federico M, Luminari S, Iannitto E, et al. ABVD compared with Laskar S, Gupta T, Vimal S, et al. Consolidation radiation after
BEACOPP compared with CEC for the initial treatment of complete remission in Hodgkin’s disease following six cycles
patients with advanced Hodgkin’s lymphoma: results from the of doxorubicin, bleomycin, vinblastine, and dacarbazine
HD2000 Gruppo Italiano per lo Studio dei Linfomi Trial. J Clin chemotherapy: is there a need? J Clin Oncol. 2004;22:62-68.
Oncol. 2009;27:805-811. Moccia AA, Donaldson J, Chhanabhai M, et al. International
Ferme C, Mounier N, Casasnovas O, et al. Long-term results Prognostic Score in advanced-stage Hodgkin’s lymphoma: altered
and competing risk analysis of the H89 trial in patients utility in the modern era. J Clin Oncol. 2012;30:3383-3388.
with advanced-stage Hodgkin lymphoma: a study by the Mounier N, Brice P, Bologna S, et al. ABVD (8 cycles) versus
Groupe d’Etude des Lymphomes de l’Adulte (GELA). Blood. BEACOPP (4 escalated cycles ≥4 baseline): final results in stage
2006;107:4636-4642. III-IV low-risk Hodgkin lymphoma (IPS 0-2) of the LYSA H34
Gallamini A, Hutchings M, Rigacci L, et al. Early interim 2-[18F] randomized trial. Ann Oncol. 2014;25:1622-1628.
fluoro-2-deoxy-D-glucose positron emission tomography is Proctor SJ, Mackie M, Dawson A, et al. A population-based
prognostically superior to International Prognostic Score in study of intensive multi-agent chemotherapy with or
advanced-stage Hodgkin’s lymphoma: a report from a Joint without autotransplant for the highest risk Hodgkin’s disease
Italian-Danish Study. J Clin Oncol. 2007;25:3746-3752. patients identified by the Scotland and Newcastle Lymphoma
Gobbi PG, Levis A, Chisesi T, et al. ABVD versus modified Group (SNLG) prognostic index. A Scotland and Newcastle
Stanford V versus MOPPEBVCAD with optional and limited Lymphoma Group Study (SNLG HD III). Eur J Cancer.
radiotherapy in intermediate- and advanced-stage Hodgkin’s 2002;38:795-806.
lymphoma: final results of a multicenter randomized trial by Sieniawski M, Reineke T, Nogova L, et al. Fertility in male patients
the Intergruppo Italiano Linfomi. J Clin Oncol. 2005;23:9198- with advanced Hodgkin lymphoma treated with BEACOPP: a
9207. report of the German Hodgkin Study Group (GHSG). Blood.
Gordon LI, Hong F, Fisher RI, et al. A randomized phase III trial of 2008;111:71-76.
ABVD vs. Stanford V +/- radiation therapy in locally extensive Wongso D, Fuchs M, Plütschow A, et al. Treatment-related
and advanced stage Hodgkin’s lymphoma: an Intergroup Study mortality in patients with advanced-stage Hodgkin lymphoma:
Coordinated by the Eastern Cooperatve Oncology Group an analysis of the German Hodgkin Study Group. J Clin Oncol.
(E2496) [abstract]. Blood. 2010;116:abstract 415. 2013;31:2819-2824.
Hasenclever D, Diehl V. A prognostic score for advanced Vaagussa P, Santoro A, Fossati-Bellani F, et al. Second acute
Hodgkin’s disease. International prognostic factors project on leukemia and other malignancies following treatment for
advanced Hodgkin’s disease. N Engl J Med. 1998;339:1506-1514. Hodgkin’s disease. J Clin Oncol. 1986;4:830-837.
Horning SJ, Hoppe RT, Breslin S, et al. Stanford V and Viviani S, Santoro A, Ragni G, et al. Gonadal toxicity after
radiotherapy for locally extensive and advanced Hodgkin’s combination chemotherapy for Hodgkin’s disease. Comparative
disease: mature results of a prospective clinical trial. J Clin results of MOPP vs ABVD. Eur J Cancer Clin Oncol.
Oncol. 2002;20:630-637. 1985;21:601-605.
Horning SJ, Williams J, Bartlett NL, et al. Assessment of the Viviani S, Zinzani PL, Rambaldi A, et al. ABVD versus BEACOPP
Stanford V regimen and consolidative radiotherapy for bulky for Hodgkin’s lymphoma when high-dose salvage is planned.
and advanced Hodgkin’s disease: Eastern Cooperative Oncology N Engl J Med. 2011;365:203-212.
Group Pilot Study E1492. J Clin Oncol. 2000;18:972-980. Younes A, Connors JM, Park SI, et al. Frontline therapy with
Hoskin PJ, Lowry L, Horwich A, et al. Randomized comparison brentuximab vedotin combined with ABVD or AVD in patients
of the Stanford V regimen and ABVD in the treatment of with newly diagnosed advanced stage Hodgkin lymphoma
advanced Hodgkin’s lymphoma: United Kingdom National [abstract]. Blood. 2011;118:abstract 955.
Cancer Research Institute Lymphoma Group Study ISRCTN
64141244. J Clin Oncol. 2009;27:5390-5396.
Therapy for relapsed or refractory HL
Johnson PWM, Radford JA, Cullen MH, et al. Comparison
of ABVD and alternating or hybrid multidrug regimens for Aleman BMP, van den Belt-Dusebout AW, Klokman WJ, et al.
the treatment of advanced Hodgkin’s lymphoma: results Long-term cause-specific mortality of patients treated for
of the United Kingdom Lymphoma Group LY09 Trial Hodgkin’s disease. J Clin Oncol. 2003;21:3431-3439.
(ISRCTN97144519). J Clin Oncol. 2005;23:9208-9218. Anderlini P, Saliba R, Acholonu S, et al. Fludarabine-melphalan
Johnson PW, Sydes MR, Hancock BW, et al. Consolidation as a preparative regimen for reduced intensity conditioning
radiotherapy in patients with advanced Hodgkin’s lymphoma: allogeneic stem cell transplantation in relapsed and refractory
survival data from the UKLG LY09 randomized controlled trial Hodgkin’s lymphoma: the updated MD Anderson Cancer
(ISRCTN97144519). J Clin Oncol. 2010;28:3352-3359. Center Experience. Haematologica. 2008;93:257-264.
Kobe C, Dietlein M, Franklin J, et al. Positron emission Andre M, Henry-Amar M, Pico JL, et al. Comparison of high-
tomography has a high negative predictive value for progression dose therapy and autologous stem-cell transplantation with
or early relapse for patients with residual disease after first-line conventional therapy for Hodgkin’s disease induction failure: a
chemotherapy in advanced-stage Hodgkin lymphoma. Blood. case-control study. Societe Francaise de Greffe de Moelle. J Clin
2008;112:3989-3994. Oncol. 1999;17:222-2229.

Bartlett NL, Niedzwiecki D, Johnson JL, et al. Gemcitabine, with carmustine, etoposide, cytarabine, and melphalan
vinorelbine, and pegylated liposomal doxorubicin (GVD), a salvage therapy prior to autologous stem cell transplantation
salvage regimen in relapsed Hodgkin’s lymphoma: CALGB for recurrent or refractory Hodgkin lymphoma. Cancer.
59804. Ann Oncol. 2007;18:1071-1079. 2006;106:353-360.
Castagna L, Bramanti S, Balzarotti M, et al. Predictive value of Lazarus HM, Rowlings PA, Zhang MJ, et al. Autotransplants
early 18F-fluorodeoxyglucose positron emission tomography for Hodgkin’s disease in patients never achieving remission:
(FDG-PET) during salvage chemotherapy in relapsing/ a report from the Autologous Blood and Marrow Transplant
refractory Hodgkin lymphoma (HL) treated with high-dose Registry. J Clin Oncol. 1999;17:534-545.
chemotherapy. Br J Haematol. 2009;145:369-372. Lavoie JC, Connors JM, Phillips GL, et al. High-dose
Chen R, Gopal A, Smith S, et al. Results from a pivotal phase chemotherapy and autologous stem cell transplantation for
II study of brentuximab vedotin (SGN-35) in patients with primary refractory or relapsed Hodgkin lymphoma: long-term
relapsed or refractory Hodgkin lymphoma (HL) [abstract]. outcome in the first 100 patients treated in Vancouver. Blood.
J Clin Oncol. 2011;29:abstract 8031. 2005;106:1473-1478.
Colwill R, Crump M, Couture F, et al. Mini-BEAM as salvage Linch DC, Winfield D, Goldstone AH, et al. Dose intensification
therapy for relapsed or refractory Hodgkin’s disease before with autologous bone-marrow transplantation in relapsed and
intensive therapy and autologous bone marrow transplantation. resistant Hodgkin’s disease: results of a BNLI randomised trial.
J Clin Oncol. 1995;13:396-402. Lancet. 1993;341:1051-1054.
Corradini P, Sarina B, Farina L. Allogeneic transplantation for Little R, Wittes RE, Longo DL, et al. Vinblastine for recurrent
Hodgkin’s lymphoma. Br J Haematol. 2011;152:261-272. Hodgkin’s disease following autologous bone marrow
Crump M, Smith AM, Brandwein J, et al. High-dose etoposide transplant. J Clin Oncol. 1998;16:584-588.
and melphalan, and autologous bone marrow transplantation Moskowitz CH, Nademanee A, Masszi T, et al. Brentuximab
for patients with advanced Hodgkin’s disease: importance of vedotin as consolidation therapy after autologous stem-cell
disease status at transplant. J Clin Oncol. 1993;11:704-711. transplantation in patients with Hodgkin’s lymphoma at risk
Devizzi L, Santoro A, Bonfante V, et al. Vinorelbine: an active of relapse or progression (AETHERA): a randomised, double-
drug for the management of patients with heavily pretreated blind, placebo-controlled, phase 3 trial. Lancet. 2015;385:
Hodgkin’s disease. Ann Oncol. 1994;5:817-820. 1853-1862.
Engert A. Hodgkin’s lymphoma: who needs consolidation Moskowitz CH, Nimer SD, Zelenetz AD, et al. A 2-step
treatment? Lancet. 2015;385:1810-1812. comprehensive high-dose chemoradiotherapy second-line
Forero-Torres A, Berryman B, Advani R, et al. Prolonged treatment program for relapsed and refractory Hodgkin disease: analysis
with brentuximab vedotin (SGN-35) in patients with relapsed by intent to treat and development of a prognostic model.
or refractory Hodgkin lymphoma (HL) or systemic anaplastic Blood. 2001;97:616-623.
large cell lymphoma [abstract]. Blood. 2011;118:abstract 3711. Moskowitz AJ, Yahalom J, Kewalramani T, et al. Pretransplantation
Jabbour E, Hosing C, Ayers G, et al. Pretransplant positive functional imaging predicts outcome following autologous
positron emission tomography/gallium scans predict poor stem cell transplantation for relapsed and refractory Hodgkin
outcome in patients with recurrent/refractory Hodgkin lymphoma. Blood. 2010;116:4934-4937.
lymphoma. Cancer. 2007;109:2481-2489. Morschhauser F, Brice P, Ferme C, et al. Risk-adapted
Josting A, Katay I, Rueffer U, et al. Favorable outcome of patients salvage treatment with single or tandem autologous stem-
with relapsed or refractory Hodgkin’s disease treated with high- cell transplantation for first relapse/refractory Hodgkin’s
dose chemotherapy and stem cell rescue at the time of maximal lymphoma: results of the prospective multicenter H96 trial by
response to conventional salvage therapy (Dexa-BEAM). Annals the GELA/SFGM study group. J Clin Oncol. 2008;26: 5980-5987.
of Oncology. 1998;9:289-295. Robinson S, Sureda A, Canals C, et al. Reduced intensity
Josting A, Muller H, Borchmann P, et al. Dose intensity of conditioning allogeneic stem cell transplantation for Hodgkin’s
chemotherapy in patients with relapsed Hodgkin’s lymphoma. lymphoma: identification of prognostic factors predicting
J Clin Oncol. 2010;28:5074-5080. outcome. Hematologica. 2009;94:230-238.
Josting A, Nogova L, Franklin J, et al. Salvage radiotherapy in Santoro A, Bredenfeld H, Devizzi L, et al. Gemcitabine in
patients with relapsed and refractory Hodgkin’s lymphoma: a the treatment of refractory Hodgkin’s disease: results of a
retrospective analysis from the German Hodgkin Lymphoma multicenter phase II study. J Clin Oncol. 2000;18:2615-2619.
Study Group. J Clin Oncol. 2005;23:1522-1529. Santoro A, Magagnoli M, Spina M, et al. Ifosfamide, gemcitabine,
Josting A, Rudolph C, Reiser M, et al. Time-intensified and vinorelbine: a new induction regimen for refractory and
dexamethasone/cisplatin/cytarabine: an effective salvage therapy relapsed Hodgkin’s lymphoma. Haematologica. 2007;92:35-41.
with low toxicity in patients with relapsed and refractory Schmitz N, Pfistner B, Sextro M, et al. Aggressive conventional
Hodgkin’s disease. Ann Oncol. 2002;13:1628-1635. chemotherapy compared with high-dose chemotherapy with
Kuruvilla J, Nagy T, Pintilie M, et al. Similar response rates and autologous haemopoietic stem-cell transplantation for relapsed
superior early progression-free survival with gemcitabine, chemosensitive Hodgkin’s disease: a randomised trial. Lancet.
dexamethasone, and cisplatin salvage therapy compared 2002;359:2065-2071.

Bibliography | 571
Sureda A, Robinson S, Canals C, et al. Reduced intensity Torrey M, Poen J, Hoppe RT. Detection of relapse in early-stage
conditioning compared with conventional allogeneic stem-cell Hodgkin’s disease: role of routine follow-up studies. J Clin
transplantation in relapsed or refractory Hodgkin’s lymphoma: Oncol. 1997;15:1123-1130.
an analysis from the lymphoma working party of the European
Group for Blood and Marrow Transplantation. J Clin Oncol.
2008;26:455-462. Pediatric HL
Sweetenham JW, Carella AM, Taghipour G, et al. High-dose
Friedman DL, Chen L, Wolden S, et al. Dose-intensive response-
therapy and autologous stem-cell transplantation for adult
based chemotherapy and radiation therapy for children and
patients with Hodgkin’s disease who do not enter remission
adolescents with newly diagnosed intermediate-risk Hodgkin
after induction chemotherapy: results in 175 patients reported
lymphoma: a report from the Children’s Oncology Group Study
to the European Group for Blood and Marrow Transplantation.
AHOD0031. J Clin Oncol. 2014;32:3651.
Lymphoma Working Party. J Clin Oncol. 1999;17:3101-3109.
Linabery AM, Ross JA. Trends in childhood cancer incidence in
Vose JM, Bierman PJ, Anderson JR, et al. CHLVPP chemotherapy
the U.S. (1992-2004). Cancer. 2008;112:416-432.
with involved-field irradiation for Hodgkin’s disease: favorable
Metzger ML, Weinstein HJ, Hudson MM, et al. Association
results with acceptable toxicity. J Clin Oncol. 1991;9:1421-1425.
between radiotherapy vs no radiotherapy based on early
Younes A, Bartlett NL, Leonard JP, et al. Brentuximab vedotin
response to VAMP chemotherapy and survival among children
(SGN-35) for relapsed CD30-positive lymphomas. N Engl J
with favorable-risk Hodgkin lymphoma. JAMA. 2012;307:
Med. 2010;363:1812-1821.
2609-2616.
Yuen AR, Rosenberg SA, Hoppe RT, et al. Comparison between
Perry CL, Smith MA, Linet M, et al. Lymphomas and
conventional salvage therapy and high-dose therapy with
reticuloendothelial neoplasms. In: Cancer Incidence and
autografting for recurrent or refractory Hodgkin’s disease.
Survival among Children and Adolescents: United States SEER
Blood. 1997;89:814-822.
Program, 1975-1995. National Cancer Institute, Bethesda, MD;
1999:35.
Follow-up of patients with HL
Dryver ET, Jernstrom H, Tompkins K, et al. Follow-up of patients
Nodular lymphocyte-predominant HL
with Hodgkin’s disease following curative treatment: the
routine CT scan is of little value. Br J Cancer. 2003;89:482-486. Chen RC, Chin MS, Ng AK, et al. Early-stage, lymphocyte-
Guadagnolo BA, Punglia RS, Kuntz KM, et al. Cost-effectiveness predominant Hodgkin’s lymphoma: patient outcomes from a
analysis of computerized tomography in the routine follow-up large, single-institution series with long follow-up. J Clin Oncol.
of patients after primary treatment for Hodgkin’s disease. J Clin 2010;28:136-141.
Oncol. 2006;24:4116-4122. Eichenauer DA, Fuchs M, Pluetschow A, et al. Phase 2 study of
El-Galaly T, Mylam KJ, Brown P, et al. PET/CT surveillance in rituximab in newly diagnosed stage IA nodular lymphocyte
patients with Hodgkin lymphoma in first remission is associated predominant Hodgkin lymphoma: a report from the German
with low positive predictive value and high costs. Haematologica Hodgkin Study Group. Blood. 2011;118:4363-4365.
2012;97:931-936. Ekstrand B, Lucas J, Horowitz S, et al. Rituximab in lymphocyte
Franklin J, Pluetschow A, Paus M, et al. Second malignancy predominant Hodgkin disease: results of a phase 2 trial. Blood.
risk associated with treatment of Hodgkin’s lymphoma: 2003;101:4285-4289.
meta-analysis of the randomised trials. Annals of Oncology. Lee A, LaCasce AS. Nodular lymphocyte predominant Hodgkin
2006;17:1749-1760. lymphoma. Oncologist. 2009;14:739-751.
Goodman K, Riedel E, Serrano V, et al. Long-term effects of high Nogova L, Reineke T, Brillant C, et al. Lymphocyte-predominant
dose chemotherapy and radiation for relapsed and refractory and classical Hodgkin’s lymphoma: a comprehensive analysis
Hodgkin’s lymphoma. J Clin Oncol. 2008;26:5240-5247. from the German Hodgkin Study Group. J Clin Oncol.
Ng AK, Bernardo MVP, Weller E, et al. Second malignancy 2008;26:434-439.
after Hodgkin disease treated with radiation therapy with or Nogova L, Reineke T, Eich HT, et al. Extended field
without chemotherapy: long-term risks and risk factors. Blood. radiotherapy, combined modality treatment or involved
2002;100:1989-1996. field radiotherapy for patients with stage IA lymphocyte-
Ng A, Constine LS, Advani R, et al. ACR appropriateness criteria: predominant Hodgkin’s lymphoma: a retrospective analysis
follow-up of Hodgkin’s lymphoma. Curr Probl Cancer. from the German Hodgkin Study Group (GHSG). Ann Oncol.
2010;34:211-227. 2005;16:1683-1687.
Sureda A, Arranz A, Iriondo A, et al. Autologous stem cell Pellegrino B, Terrier-Lacombe MJ, Oberlin O, et al. Lymphocyte-
transplantation for Hodgkin’s disease: results and prognostic predominant Hodgkin’s lymphoma in children: therapeutic
factors in 494 patients from the Grupo Espanol de Linfomas/ abstention after initial lymph node resection: a study of
Transplante Autologo de Medula Osea Spanish Cooperative the French Society of Pediatric Oncology. J Clin Oncol.
Group. J Clin Oncol. 2001;19:1395-1404. 2003;21:2948-2952.

Savage K, Skinnider BF, Al-Mansour M, Sehn LH, Gascoyne adolescents with early stage nodular lymphocyte predominant
RD, Connors JM. Treating limited stage nodular lymphocyte Hodgkin’s lymphoma: an Anglo-French collaborative report.
predominant Hodgkin lymphoma similarly to classical Eur J Cancer. 2011;48:1700-1706.
Hodgkin lymphoma with ABVD may improve outcomes. Blood. Wirth A, Yuen K, Barton M, et al. Long-term outcome after
2011;118:4585-4590. radiotherapy alone for lymphocyte-predominant Hodgkin
Shankar A, Hall G, Gorde-Grosjean S, et al. Treatment outcome lymphoma. Cancer. 2005;104:1221-1229.
after low intensity chemotherapy (CVP) in children and

CHAPTER
21
Non-Hodgkin lymphoma
Brad Kahl, Grzegorz S. Nowakowski, and David Yang
Overview of lymphocyte development and classification of Aggressive B-cell lymphomas, 590
lymphoid malignancies, 573 Bibliography, 611
Indolent B-cell NHL, 583
Overview of lymphocyte development signaling has also been found to be necessary for lymphoma
and classification of lymphoid development and evolution with many mature B-cell malig-
malignancies nancies showing sensitivity to kinase inhibitors, such as ibru-
tinib and idelalisib, which disrupt BCR signaling.
The lymphoid system forms the backbone of the human Collectively, the primary function of B-cells is to generate
immune system, contributing to both the innate (non-specific) a vast diversity of immunoglobulins. Generating this diver-
immune response through natural killer (NK) cells and the sity begins with the combinatorial diversity produced from
adaptive (specific) immune response through B- and T-cells. random V, D, and J rearrangements. Combinatorial diver-
B- and T-cell lymphomas account for the vast majority of non- sity is amplified by junctional diversity produced by the
Hodgkin lymphomas, neoplasms that originate in these cells. action of terminal deoxynucleotidyl transferase (TdT) where
Knowledge of B- and T-cell development is important in nucleotides are randomly added or deleted at the sites of V,
understanding the biology and, in turn, providing insight into D, and J fusion. Successful rearrangement of the heavy and
the behavior of the numerous subtypes of these lymphomas light immunoglobulin chains (either kappa or lambda)
that are derived from their normal B- and T-cell counterparts. results in expression of functional IgM and IgD on the sur-
face mature B-cells that exit the marrow. These mature, but
antigen naïve, B-cells then gain additional diversity when
B-cell development and the biology of
exposed to antigen in the germinal centers of secondary lym-
B-cell lymphomas
phoid organs such as lymph nodes, mucosa associated lym-
Common lymphoid progenitors in the bone marrow derived phoid tissue, or the spleen. Here, somatic hypermutation
from hematopoietic stem cells are the source of B- and occurs in the V genes of the heavy and light chains, fine tun-
T-cells. Unlike T-cells, full B-cell maturation occurs in the ing their affinity to their cognate antigens. B-cells expressing
bone marrow and begins with recombination of the V, D, immunoglobulin with just the right amount of antigen affin-
and J gene segments of the immunoglobulin heavy chain ity differentiate to memory B-cells or plasma cells while all
(IgH) followed by the light chain genes in order to generate a the others undergo apoptosis. Finally, class switching also
functional immunoglobulin that is expressed on the cell sur- occurs in the germinal center and involves changing the
face as B-cell receptor (BCR). The survival and maturation of heavy chain that is expressed to produce IgG, IgA, or IgE.
B-cells in the bone marrow, as well as the differentiation of The classification of B-cell lymphomas is based on the
mature B-cells that have exited the bone marrow, is depen- observation that malignant B-cells appear to be frozen at par-
dent on operative BCR signaling. Importantly, BCR ticular differentiation stages that reflect their origin and, in
part, dictate their biology (Figure 21-1). Distinct stages of
Conflict-of-interest disclosure: Dr. Kahl: Research support: B-cell development and differentiation are characterized by
Genentech, AbbVie; Consultancy: Celgene, Roche, Seattle Genetics, distinct cytologic features, expression patterns of differentia-
Millennium, Infinity, Cell Therapeutics. Dr. Nowakowski: Research tion markers, and BCR. These characteristics form the basis
support: Celgene, Bayer; Consultancy: Celgene, Bayer, Seattle
Genetics. Dr. Yang: None. of pathologic diagnosis of lymphoid neoplasms. For example,
Off-label drug use: None. B-lymphoblastic leukemia/lymphoma arises from an
| 573

574 | Non-Hodgkin lymphoma
Central lymphoid tissue Peripheral lymphoid tissue

Precursor B cells Peripheral (mature) B cells
Bone marrow Interfollicular area Follicular area Perifollicular area
Long-lived plasma cell

IgG, IgA, IgM,
Mantle
IgD, IgE
M
D cell
Naive Antigen
Progenitor B cells
B cell
AG
Centro-
Pre-B cell FDC cyte
M
Extrafollicular Memory B cells
B blast Marginal zone
M
Immature
B cell
M
M
Short-lived
plasma cell
Centroblast
Apoptotic
B cell
Precursor B-cell neoplasms Pre-GC neoplasm GC neoplasms Post-GC neoplasms

B-lymphoblastic Mantle cell lymphoma Follicular lymphoma Marginal zone and MALT lymphomas
leukemia/lymphoma CLL/SLL (some) Burkitt lymphoma Lymphoplasmacytic lymphoma
DLBCL (some) CLL/SLL (some)
Hodgkin lymphoma Plasma cell myeloma
Figure 21-1 Schematic representation of B-cell differentiation (WHO 2008). CLL/SLL = chronic lymphocytic leukemia/small lymphocytic
lymphoma; DLBCL = diffuse large B-cell lymphoma; GC = germinal center; MALT = mucosa-associated lymphoid tissue. Reproduced with
permission from Harald Stein.
immature B-cell (Figure 21-1) and accordingly, diagnosis Table 21-1 Risk factors in the development of non-Hodgkin
requires the identification of immature B-cells that have mor- lymphoma
phologic characteristics of blasts; co-express B-cell markers, Viral infection EBV, HTLV-1, HHV-8,
such as CD19, with markers of immaturity, such as TdT and hepatitis C
CD10; and do not express BCR on their surface. Likewise, fol- Bacterial infection Helicobacter pylori
licular lymphoma (FL) arises from a germinal center B-cell Chlamydophila psittaci
(Figure 21-1) and has morphologic characteristics of nodular Impaired/altered immunity Ataxia-telangiectasia
growth, resembling B-cell follicles, while expressing the ger- Congenital disorders Wiskott-Aldrich syndrome
minal center marker CD10 with surface IgM, IgD, IgG or IgA. X-linked lymphoproliferative
The transformation of benign B-cells to their malignant syndrome
counterparts is closely linked to the mandate of B-cells to Severe combined
generate immunological diversity. Conditions under which immunodeficiency
malignant transformation is fostered include viral infection, Other immunodeficiency states
chronic bacterial infection, immune deficiency, and exposure Acquired conditions of AIDS (HIV infection)
to toxins (Table 21-1). Given the degree to which the immu- immunodeficiency Organ or stem cell transplantation
noglobulin genes of B-cells are subjected to DNA damage in Aging
Autoimmune and rheumatologic
the bone marrow and germinal centers, it is not surprising
disease
that reciprocal translocations involving an immunoglobulin
gene locus and a proto-oncogene form the hallmark of many Environmental or Herbicides
occupational Pesticides
types of B-cell lymphoma (Table 21-2).

Overview of lymphocyte development and classification of lymphoid malignancies | 575
Table 21-2 Phenotypic markers and chromosomal translocations in non-Hodgkin lymphomas
NHL sIg CD5 CD10 CD20 Other Cyclin D1 Cytogenetics Oncogene Function
CLL/SLL Weak + − Dim CD23+ FMC− − No diagnostic − −
abnormalities*
Follicular ++ − + + − − t(14;18) BCL2 Anti-apoptosis
Mantle cell ++ + − + CD23− FMC+ + t(11;14) Cyclin D1 Cell cycle regulator
Marginal zone/ + − − + – t(11;18) AP12−MALT Resistance to
extranodal Helicobacter
marginal zone pylori treatment
lymphoma
Lymphoplasmacytic ++ − − + CD25+/− − – MYD88 Proliferation
lymphoma CD38+/−
Hairy cell leukemia ++ − − + CD11c+, CD25+, Weak – BRAF Proliferation
CD103+
DLBCL + Rare +/− + − − t(14;18), t(3;14), BCL2 Anti-apoptosis
t(3;v) BCL6 Transcription factor
Rare t(8;X), cMYC Proliferation
EZH2‡ Histone modifier
MYD88§ Proliferation
PMBCL + − −/+ + CD30+/− − t(16;X)† CIITA MHC class II
transactivator
Burkitt lymphoma + − + + TdT− − t(8;14), t(2;8), cMYC Transcription factor
t(8;22)
ALCL, ALK-positive − − − − CD30+, CD2+/−, − t(2;5) ALK Tyrosine kinase
CD3−/+
EMA+
ALCL, ALK-negative − − − − CD30+, CD2+/−, − t(6;7)(p25.3;q32.3) DUSP22 Phosphatase
CD3−/+
EMA–
ALCL = anaplastic large-cell lymphoma; CLL = chronic lymphocytic leukemia; DLBCL = diffuse large B-cell lymphoma; PMBCL = primary
mediastinal large B-cell lymphoma; MALT = mucosa-associated lymphoid tissue; sIg = surface immunoglobulin; SLL = small lymphocytic
lymphoma; TdT = terminal deoxynucleotidyl transferase.
* A number of prognostic cytogenetic abnormalities have been identified (see Chapter 22).
† A number of partner chromosomes described.
‡ Exclusively in GCB-like DLBCL.
§ Exclusively in ABC-like DLBCL.
innate, rather than adaptive, immune system. Meanwhile, as

T-cell development and biology of the
developing T-cells that express αβ TCR mature in the thy-
T-cell lymphomas
mus, their αβ TCR is complexed with surface CD3 and
In contrast to B-cell development, T-cell progenitors derived either CD4 or CD8, which identifies helper and cytotoxic
from common lymphoid progenitors exit the marrow and T-cell subsets, respectively (Figure 21-2).
develop in the thymus. Similar to B cells, each T cell recog- The cell of origin approach that was so effective for cate-
nizes a specific antigen, but through a T-cell receptor (TCR) gorization of B-cell lymphomas has been more difficult to
rather than BCR. Similar to BCRs, diversity of TCRs is gen- apply to T-cell lymphomas, due to a combination of factors
erated through recombination of V, D, and J gene segments including the complexity of mature T- and NK-cell lineages,
of the four TCR genes, alpha (α), beta (β), gamma (γ) and with numerous functional subsets demonstrating marked
delta (δ). Mature T-cells express either αβ TCR or γδ TCR. phenotypic and morphologic diversity compounded by evi-
Of note, αβ TCRs can only recognize antigen presented in dence of plasticity. In addition, with the noticeable exception
the context of major histocompatibility complex (MHC) of ALK-positive anaplastic large cell lymphoma (ALCL), few
while γδ TCRs do not have this restriction. As such, NK-cells recurrent cytogenetic abnormalities have been associated
and γδ T-cells, which express γδ TCR, do not require antigen with mature T-cell lymphomas and accordingly, contribute
sensitization to become active and operate as part of our little to their categorization. Instead, clinical features and

Central lymphoid tissue Peripheral lymphoid tissue

Precursor T cells Peripheral (mature) T and NK cells
NK Spleen
Mucosa
Peripheral blood
Subcapsular
cortical thymocyte γδ T cell Antigen
Progenitor
Skin
T cell/pro-
thymocyte
TFH
Common CD4+ FDC

αβ Effector T cell
thymocyte CD8+ Naive
T cell T blast CD4
CD4 CD4 CD4
CD4
CD8 Medullary
thymocytes Memory T cells
CD8
Thymus
CD8 CD8
Bone marrow
CD8 Follicle
Naive T blast
T cell
Effector T cell
T-lymphoblastic lymphoma/leukemia Peripheral (mature) T-cell and NK-cell lymphomas/leukemias
Figure 21-2 Schematic representation of T-cell differentiation (WHO 2008). FDC = follicular dendritic cells; NK = natural killer; TFH =
Τ-helper follicular cells. Reproduced with permission from Harald Stein.
anatomic location of the disease have played a major role in chromatin, from mature lymphocytes, which have more
defining many of the mature T- and NK-cell neoplasms opaque and condensed chromatin. Some lymphoid malig-
included in the World Health Organization (WHO) classifi- nancies, such as chronic lymphocytic leukemia (CLL), have
cation, which can be grouped according to their presenta- characteristic patterns of chromatin condensation, with CLL
tion as disseminated (leukemic), predominantly extranodal lymphocytes typically showing a “soccer-ball” nuclear pat-
or cutaneous, or predominantly nodal disease (Table 21-3). tern. Likewise, Burkitt lymphoma cells can be recognized on
smear preparations by their finely granular chromatin and
strikingly blue, vacuolated cytoplasm.
Diagnostic testing in lymphoproliferative
Lymph node biopsies and bone marrow core biopsies
disorders
lack the cytologic detail of smear preparations because tis-
Diagnosis of lymphoproliferative disorders requires some sue specimens must be fixed in formalin and dehydrated, a
expertise and relies on a combination of morphologic find- process that shrinks the cells and obscures cytologic detail.
ings (peripheral blood, bone marrow, or lymph node), The benefit of tissue specimens is that they provide a glimpse
immunophenotyping, cytogenetics, and molecular genetics. of the underlying architecture, a critical component in dif-
ferentiating between benign from malignant lymphoid pro-
liferations and in the classification of lymphoid malignancies.
Morphology
Lymphoid malignancies typically obliterate and “efface”
Well-stained peripheral blood and bone marrow aspirate underlying normal architectural features and the pattern of
smears provide excellent cytologic detail, facilitating evalua- malignant growth, for example nodular versus diffuse,
tion of nuclear chromatin patterns and cytoplasmic color- guides subsequent classification. These patterns can be dif-
ation as well as revealing the presence of cytoplasmic ficult to recognize in small biopsy specimens and accord-
inclusions and vacuoles in lymphoid cells. The degree of ingly, needle core biopsies of suspected lymphoid
nuclear chromatin condensation is helpful in differentiating malignancies can be extremely challenging for pathologists
lymphoid blasts, which have finely granular or “open” to interpret.

Table 21-3 2008 World Health Organization classification of B-cell and T-cell neoplasms
B-cell neoplasms T-cell neoplasms

Precursor B-cell neoplasms* Precursor T-cell neoplasms*
B-lymphoblastic leukemia/lymphoma NOS T-lymphoblastic leukemia/lymphoma
B-lymphoblastic leukemia/lymphoma with recurrent genetic abnormalities
Mature B-cell neoplasms Mature T-cell neoplasms

Aggressive lymphomas Leukemic or disseminated
Diffuse large B-cell lymphoma: variants, subgroups, and subtypes/entities T-cell large granular lymphocytic leukemia†
Diffuse large B-cell lymphoma, NOS Chronic lymphoproliferative disorders of NK cells†
Common morphologic variants: centroblastic, immunoblastic, anaplastic T-cell prolymphocytic leukemia
Rare morphologic variants Aggressive NK-cell leukemia
Molecular subgroups: germinal center B-cell like (GCB) and activated Adult T-cell leukemia/lymphoma
B-cell like (ABC) Systemic EBV-positive T-cell lymphoproliferative disorders
Immunohistochemical subgroups: CD5+ DLBCL, GCB, and non-GCB of childhood
Diffuse large B-cell lymphoma subtypes
T-cell/histiocyte-rich large B-cell lymphoma Extranodal
Primary DLBCL of the CNS Extranodal NK/T-cell lymphoma, nasal type
Primary cutaneous DLBCL, leg type Enteropathy-type T-cell lymphoma
EBV-positive DLBCL of the elderly Hepatosplenic T-cell lymphoma
Other lymphomas of large B cells
Primary mediastinal large B-cell lymphoma Cutaneous
Intravascular large B-cell lymphoma Mycosis fungoides†
DLBCL associated with chronic inflammation Sézary syndrome†
Immunodeficiency-associated lymphoma Primary cutaneous CD30+ T-cell lymphoproliferative
Lymphomatoid granulomatosis disorder†
ALK-positive large B-cell lymphoma Primary cutaneous CD4+ small/medium T-cell lymphoma†
Plasmablastic lymphoma Primary cutaneous anaplastic large cell lymphoma
Large B-cell lymphoma arising in HHV-8–associated multicentric Lymphomatoid papulosis
Castleman disease Subcutaneous panniculitis-like T-cell lymphoma
Primary effusion lymphoma Primary cutaneous γδ T-cell lymphoma
B-cell lymphoma, unclassifiable, with features intermediate between DLBCL Primary cutaneous CD8+ aggressive epidermotropic
and Burkitt lymphoma cytotoxic T-cell lymphoma
B-cell lymphoma, unclassifiable, with features intermediate between DLBCL Hydroa vacciniforme-like lymphoma
and classical Hodgkin lymphoma
Burkitt lymphoma Nodal
Mantle cell lymphoma Peripheral T-cell lymphoma, NOS
Angioimmunoblastic T-cell lymphoma
Indolent lymphomas Anaplastic large-cell lymphoma, ALK positive
Follicular lymphoma Anaplastic large-cell lymphoma, ALK negative
Primary cutaneous follicle center lymphoma
Extranodal marginal zone lymphoma of mucosa-associated lymphoid tissue
(MALT)
Nodal marginal zone lymphoma
Splenic marginal zone lymphoma
Splenic B-cell lymphoma/leukemia, unclassifiable
Lymphoplasmacytic lymphoma
Heavy chain disease
Plasma cell neoplasms
CLL/SLL
B-cell prolymphocytic leukemia
Hairy cell leukemia
CLL = chronic lymphocytic leukemia; CNS = central nervous system; DLBCL = diffuse large B-cell lymphoma; HHV-8 = human herpesvirus 8;
NK = natural killer; NOS = not otherwise specified; SLL = small lymphocytic lymphoma.
* All precursor neoplasms are considered aggressive.
† Indolent T-cell neoplasms, all other T-cell neoplasms are considered aggressive.

Immunophenotyping can be useful markers of malignancy and for identifying spe-

cific lymphoma subtypes. Use of microarray technology has
Immunophenotyping can be performed by flow cytometry
defined gene expression profiles of various lymphoid malig-
on live cells from either liquid specimens or disaggregated
nancies and compared them to normal lymphoid popula-
tissue. For fixed specimens, immunophenotyping is typi-
tions. This technique has been successfully applied to a
cally performed by 3,3'-diaminobenzidine (DAB) staining
number of B-cell lymphomas, including diffuse large B-cell
of tissue on glass slides. Immunophenotyping complements
lymphoma (DLBCL), FL, CLL, and MCL to identify expres-
morphologic assessment by illuminating details of cell biol-
sion patterns that correlate with patient outcome. However,
ogy that would be otherwise imperceptible through the
technical difficulty with assessing gene expression profiles in
microscope. By determining cell lineage, maturation stage,
the clinical laboratory, especially in formalin-fixed tissues,
and the presence of any aberrant antigen expression, immu-
has hampered clinical application of these findings. Despite
nophenotyping findings can be combined with morpho-
this, pathologists and oncologists have managed to apply the
logic findings to arrive at a diagnosis. For example, mantle
DLBCL gene expression discoveries to the clinical realm by
cell lymphoma (MCL) is characterized by effacement of
utilizing surrogate immunohistochemistry based expression
normal nodal architecture by small nongerminal center
panels to differentiate the better-prognosis germinal center
(CD10 negative) B-cells (CD20 positive), with aberrant
B-cell-like DLBCL from the poor-prognosis activated B-cell-
coexpression of CD5 (typically a T-cell marker, but
like DLBCL. More recently, next-generation sequencing
expressed on a subset of B cells) and cyclin D1 (a protein
(NGS) technology has been utilized to deeply interrogate the
that is not expressed in normal lymphocytes; its expression
genomes of various lymphoid malignancies. While many
results from the translocation that underlies MCL). Other
such studies are still ongoing, landmark discoveries of single
characteristic immunophenotypic profiles of lymphoid
causative mutations of BRAF V600E in hairy cell leukemia
malignancies can be found on Table 21-2.
(HCL) and MYD88 L265P in Waldenström macroglobulin-
For B-cell malignancies, clonality can also be identified by
emia have thus far been reported (Table 21-2).
light-chain restriction of the surface immunoglobulin.
B-cells normally express κ and λ light chains in a ratio of 2:1.
A clonal expansion can be identified by a marked predomi- Classification of non-Hodgkin lymphomas
nance of either κ- or λ-expressing B cells that would not be
The classification of lymphoproliferative disorders contin-
expected in a reactive process. The immunophenotyping of
ues to evolve as our understanding of the biology of these
T-cell neoplasms is less conclusive than for B-cell disorders
diseases progresses. The current classification system used is
because T-cells lack the equivalent of light-chain restriction.
the World Health Organization (WHO) Classification of
Several findings can be suggestive of neoplasia, including
Tumors of Hematopoietic and Lymphoid Tissues, which was
expression of CD4 or CD8 on the majority of the T-cells,
updated in 2008 (Table 21-3) (Swerdlow et al., 2008) and
lack of expression of CD4/CD8 on the majority of T cells, or
will be updated again in 2016-2017. The B- and T-cell neo-
coexpression of CD4 and CD8 on the majority of T-cells.
plasms are separated into precursor (lymphoblastic) neo-
Often, however, molecular techniques to look at TCR gene
plasms and mature B- or T-cell neoplasms. Overall, ~90% of
rearrangements are necessary to differentiate reactive from
all non-Hodgkin lymphomas (NHLs) in Western countries
clonal T-cell processes.
are of mature B-cell origin, with DLBCL and FL being the
most common subtypes. In children, Hodgkin lymphoma
(HL) is more predominant, and the aggressive NHLs of lym-
Molecular genetics and cytogenetics
phoblastic lymphoma and Burkitt lymphoma (BL) are much
Molecular genetic techniques can be helpful in assessing clon- more commonly encountered than indolent neoplasms. The
ality when morphology and immunophenotyping are incon- incidence of NHL is lower among Asian populations, in
clusive. These involve isolating the DNA from a sample and whom T-/NK-cell neoplasms are more frequent.
subjecting it to polymerase chain reaction (PCR) to detect While the premise of the WHO classification is to separate
rearrangements of immunoglobulin or TCR genes. The dem- lymphoid malignancies into distinct, nonoverlapping enti-
onstration of a dominant rearrangement of the immunoglob- ties, it also recognizes that the biology of particular tumors
ulin or TCR genes is indicative of a clonal process. cross the boundaries between current categories. These gray
Chromosomal translocations are common in lympho zone malignancies are recognized as B-cell lymphoma,
proliferative disorders and may contribute to the transfor- unclassifiable, with features intermediate between (i) DLBCL
mation process or cellular proliferation (Table 21-2). and classical Hodgkin lymphoma (cHL) and (ii) DLBCL and
Commercial probes are available for detection of most trans- Burkitt lymphoma. Common gene expression and epigene-
locations by fluorescent in-situ hybridization (FISH) and tic profiles between primary mediastinal large B-cell

lymphoma and cHL indicate a true biologic grey zone hepatitis C seropositivity were associated with increased risk
between these two entities exists. Likewise, certain cases of for certain B-cell lymphomas (Morton et al., 2014).
DLBCL have been found to have expression profiles of Immunosuppression associated with HIV infection or iat-
Burkitt lymphoma, although they differed clinically and rogenically induced immune suppression in the organ trans-
genetically from classic Burkitt lymphoma and vice versa. plantation setting is associated with an increased incidence of
Biologically, many of these cases may lie in the gray zone aggressive B-cell lymphomas, likely due to dysregulated B-cell
because they have rearrangements in both BCL2 and cMYC proliferation and susceptibility to viruses such as Epstein-
genes (“double-hit” lymphomas) and are more clinically Barr virus (EBV) (Table 21-1). In children, the incidence of
aggressive than standard DLBCLs. NHL is increased in several disorders that have in common
For clinical purposes, the NHLs can be broadly separated immunodeficiency from primary immune disorders, includ-
into indolent or aggressive categories (Table 21-3). Indolent ing ataxia-telangiectasia, Wiskott-Aldrich syndrome, com-
lymphomas generally are incurable with most standard ther- mon variable or severe combined immunodeficiency, and
apeutic approaches and are typified by a chronic course with X-linked lymphoproliferative disorder.
repeated relapses and progression with standard therapy. Infection with the bacterium Helicobacter pylori is strongly
Some of these patients, however, survive many years with associated with gastric MALT lymphoma (Table 21-1).
remarkably stable disease even in the absence of specific Patients with MALT limited to the stomach often achieve CR
therapy. Median survival is usually 8-10 years but not following successful therapy to eradicate H pylori, indicating
uncommonly may exceed 15-20 years. Most, but not all, that the lymphoma remains dependent in part on continued
aggressive lymphomas are potentially curable with combina- antigenic drive from the microorganism. Associations have
tion chemotherapy. Aggressive subtypes usually have a more also been made between orbital infection by Chlamydophila
acute presentation often with B-symptoms and a more rapid psittaci and orbital adnexal MALT lymphoma, infection with
progression than the indolent entities. In the event of failure Campylobacter jejuni and immunoproliferative small intesti-
to achieve complete remission (CR) following treatment or nal disease, and Borrelia burgdorferi and cutaneous MALT
with relapse after an initial therapeutic response, survival lymphoma. These intriguing associations need to be firmly
usually is measured in months rather than years. Some of established by additional investigation, with variable
these patients, however, are cured by second-line chemo- responses observed with antimicrobial agents.
therapy and stem cell transplantation approaches, as Certain viral infections have been linked with specific sub-
described later in this chapter. types of NHL. EBV has a clear pathogenic role in endemic as
well as some cases of sporadic BL and in many cases of HIV-
related aggressive B-cell lymphoma. EBV positive DLBCL of
Epidemiology, pathogenesis, and
the elderly is thought to be associated with age related immu-
molecular characterization
nosuppression. EBV is strongly associated with extranodal
Data from cancer registries show that the incidence of NHL T-/NK-cell lymphoma, nasal type, which is seen most com-
has been increasing steadily in North America and other monly in Asia and in Central and South America. It is also
industrial countries with a doubling of cases between 1970 detected in 70%-80% of cases of angioimmunoblastic T-cell
and 1990 and stabilization thereafter. In 2015, there will be lymphoma (AITL). The gamma herpesvirus human herpes-
an estimated 71,850 new cases of NHL, representing 4.3% of virus 8 (HHV-8, also called Kaposi sarcoma–associated her-
all cancer diagnoses and 3.4% of all cancer deaths. The rea- pesvirus [KSHV]), first described in Kaposi sarcoma but also
sons for this increasing incidence are unknown but are the associated with an unusual primary body cavity lymphoma
subject of ongoing epidemiologic investigations. Associa- (primary effusion lymphoma), is most commonly seen in
tions have been made with occupational exposure to certain patients with AIDS. HHV-8 also has been described in asso-
pesticides and herbicides (Table 21-1). Agricultural workers ciation with multicentric Castleman disease. The retrovirus
with cutaneous exposure to these agents have an approxi- human T-cell lymphotropic virus 1 (HTLV-1) is associated
mately two- to six fold increased incidence of NHL, possibly with adult T-cell leukemia/lymphoma endemic to Japan,
contributing to the relatively greater frequency of lymphoma central Africa, and the Caribbean. Chronic hepatitis C virus
in rural versus urban populations. Risks may differ between infection has been linked to the development of B-cell NHL,
B- and T-cell lymphoma. A large epidemiologic study from particularly marginal zone lymphoma and DLBCL, possibly
the International Lymphoma Epidemiology Consortium via chronic BCR stimulation through direct binding of a
(InterLymph) identified eczema, T-cell activating autoim- viral envelope protein.
mune diseases, a family history of myeloma, and occupation Specific chromosomal translocations are strongly associ-
as a painter as increasing the risk for T-cell lymphoma. A ated with individual subtypes of B-cell NHL (Table 21-2).
history of B-cell-activating autoimmune disease and The majority of these arise early in B-cell differentiation,

during the process of immunoglobulin gene rearrangement, Staging and prognostic factors
when errant fusion of immunoglobulin promoter and
Staging procedures generally include careful physical exami-
enhancer elements with other genes leads to dysregulated
nation for lymphadenopathy and organomegaly; computed
oncogene expression. Careful study of such transloca-
tomography (CT) scans of the neck, chest, abdomen, and pel-
tions has provided important insights into pathogenic
vis; fluorodeoxyglucose positron emission tomography (FDG-
mechanisms in lymphoma. The most frequent of these
PET) imaging; and bone marrow biopsy. CT or magnetic
translocations are: (i) t(14;18), with resultant overexpression
resonance imaging (MRI) of the brain and evaluation of the
of the anti-apoptotic gene BCL2, which is present in ∼85% of
cerebrospinal fluid are indicated in patients with Burkitt or
FLs; (ii) t(11;14) with cyclin D1 overexpression, which is
lymphoblastic lymphomas and also should be considered in
present in virtually all MCLs; and (iii) t(8;14), t(2;8), and
patients with aggressive histology lymphoma involving high-
t(8;22) of BL, which fuse an immunoglobulin heavy- or
risk sites, including the sinuses or testis. The Ann Arbor stag-
light-chain gene promoter to the cMYC transcription factor.
ing system, identifying patients as having stage I (localized) to
BCL6, a chromosome 3 transcription factor gene capable of
stage IV (extensive) disease, originally was devised for use in
promiscuous rearrangement with multiple translocation
HL but was later adopted for use in NHL. Patients are further
partners, is most commonly identified in DLBCL. The t(2;5)
stratified as to the absence (A) or presence (B) of symptoms,
(p23;q35) fuses the ALK gene with nucleophosmin and is
namely, fevers, drenching night sweats, or weight loss of 10%
found in a subset of CD30-positive ALCL. Several other
or more within 6 months of diagnosis. Several limitations
translocation partners with the ALK gene also have been
become apparent when the Ann Arbor classification is applied
described in this disease. This translocation and ALK expres-
to NHL and as a result, a revised staging system, called the
sion are associated with a more favorable prognosis in ALCL
Lugano classification, was proposed in 2014 (Table 21-4).
(see also the section “Peripheral T-cell lymphomas” in this
Patients with Ann Arbor stage I or II disease can be grouped
chapter). ALK+ lymphoma can be found in the breast in
and considered as having “limited stage” disease while patients
association with certain types of breast implants.
with Ann Arbor stage III or IV disease can be grouped in con-
Gene expression profiling has defined molecular signa-
sidered as having “advanced stage” disease. Other recommen-
tures in lymphoma that have been utilized to identify prog-
dations from the Lugano classification include the following:
nostically significant disease subsets in DLBCL, FL, MCL,
(i) FDG-PET/CT be considered standard imaging for FDG
CLL, and T-cell ALCL as well as illuminating the existence of
avid lymphomas while CT is indicated for non-avid histolo-
gray zone lymphomas that lie between DLBCL and Burkitt
gies; (ii) reserving the suffix A or B only for HL; (iii) eliminat-
lymphoma as well as DLBCL and cHL. More recently, next-
ing the X designation for bulky disease (since there is no
generation sequencing has provided some early insight into
universal definition for bulk) and replacing it with a recording
the mutational landscape of several lymphomas including
of the largest nodal diameter; and (iv) elimination of the need
the previously mentioned single causative mutations of
for staging bone marrow biopsies in HL if a PET-CT was used
BRAF V600E in HCL and MYD88 L265P in Waldenstrom
for staging.
macroglobulinemia. Additionally, the mutational landscape
Lymphoma staging has only limited prognostic usefulness.
of GCB-like DLBCL has been found to be distinct from ABC-
To more fully incorporate additional relevant prognostic fea-
like DLBCL, with GCB-like DLBCL harboring an activating
tures, models have been developed in the most common
EZH2 mutation while ABC-like DLBCL harbors activating
NHLs, DLBCL and FL, and, more recently, MCL. The most
MYD88 mutations similar to Waldenström macroglobulin-
widely used clinical prognostic model to stratify patients
emia. These discoveries continue to refine lymphoma classi-
with aggressive NHLs is the International Prognostic
fication and elucidate novel therapeutic targets.
Table 21-4 Lugano staging system for NHL
Stage
Lugano Ann Arbor Involvement Extranodal (E) Status
Limited I One node or a group of adjacent nodes Single extranodal lesions without nodal involvement
Limited II Two or more lymph node regions on the same side of Stage I or II by nodal extent with limited contiguous
the diaphragm extranodal involvement
Advanced III Involvement of lymph node regions on both sides of Not applicable
the diaphragm, nodes above the diaphragm with or
spleen involvement
Advanced IV Additional noncontiguous extralymphatic involvement Not applicable

Table 21-5 The IPI in DLBCL in the rituximab era
Risk group Risk factors (no.)* Distribution of cases (%) 4 year PFS (%) 4-year OS (%)
BCCA DSHNHL
n = 365 n = 1062 BCCA DSHNHL† BCCA DSHNHL†
Low (L) 0, 1 28% 52% 85% 87% 82% 91%
Low-intermediate (LI) 2 27% 21% 80% 75% 81% 81%
High-intermediate (HI) 3 21% 17% 57% 59% 49% 65%
High (H) 4, 5 24% 10% 51% 56% 59% 59%
Estimates are rounded off.
BCCA = British Columbia Cancer Agency; DLBCL = diffuse large B-cell lymphoma; DSHNHL = German High-Grade Non-Hodgkin
Lymphoma Study Group; OS = overall survival; PFS = progression-free survival.
*IPI risk factors are age ≥60 years, abnormal LDH, PS ≥2, stage III or IV, and >1 extranodal sites.
†3-year PFS and OS.
Index (IPI; Shipp et al., 1993). The purpose was to identify chapter.). The IPI is useful in comparing studies and also in
pretreatment variables that predict relapse-free and overall the investigation of new prognostic factors to determine the
survival (OS) in patients treated with doxorubicin-contain- independent effect on outcome.
ing combination chemotherapy. The following five risk fac- The IPI score is predictive of survival in indolent lympho-
tors were found to be independently associated with clinical mas, namely, FL, although using the IPI, the majority of
outcome and often are referred to by the mnemonic APLES: these patients fall into the low-risk or low-intermediate-risk
(i) age >60 years, (ii) PS >2, (iii) elevated serum lactate dehy- categories. As such, a new index was developed specifically
drogenase (LDH), (iv) number of extranodal sites of disease for FL called the Follicular-Lymphoma International Prognos-
>1, and (v) stage III or IV. The IPI score is derived as a simple tic Index (FLIPI) in hopes of better stratifying patients (Table
additive score from 0-5 and has been widely adopted to esti- 21-6). This index can be remembered by the mnemonic No-
mate prognosis in patients with NHL and is useful in some of LASH. The five clinical factors that are the strongest predic-
the other lymphoma subtypes. Of note, these survival esti- tors of outcome in multivariate analysis were: (i) number
mates established before the use of rituximab diffuse large (no.) of nodal sites of disease (>4), (ii) elevated LDH, (iii)
B-cell lymphoma. age >60 years, (iv) stage III or IV disease, and (v) hemoglo-
Limited studies support that the IPI is still prognostic in bin <12 g/L. Compared with the IPI, the FLIPI provides a
the rituximab treatment era (Table 21-5). A revised IPI better distribution of patients across the risk categories of
(R-IPI) may define new risk groups in rituximab-treated low risk (0 to 1 factor), intermediate risk (2 factors), or high
patients: very good risk (0 risk factors, 4-year PFS 90%); risk (>3 factors). The 10-year OS rates were 71% (low risk),
good risk (1, 2 risk factors, 4-year PFS 70%); and poor risk 51% (intermediate risk), and 36% (high risk), respectively
(>3 risk factors, 4-year PFS 50%). The DSHNHL group also (Table 21-6). Similarly, an international prognostic index
evaluated the usefulness of the IPI in patients enrolled on for MCL (the Mantle Cell Lymphoma International Prog-
prospective clinical trials, with a predominance of low-risk nostic Index [MIPI]) also has been developed, incorporating
patients, and found that it did effectively separate patients age, performance status (PS), LDH, and white blood cell
into the previously established risk categories, although the (WBC) level (Table 21-7).
difference between the high-intermediate and high-risk
groups was small (Ziepert et al., 2010) (Table 21-5).
Although the IPI scoring system provides useful prognos- Table 21-6 Follicular Lymphoma International Prognostic Index
(FLIPI)
tic information, there is no definitive evidence that outcome
is altered by using intensive regimens in high-risk patients. Risk model No. of Distribution 5-year 10-year
Numerous studies have been reported, and others are still in and group factors of cases (%) OS (%) OS (%)
progress, assessing the utility of the IPI and “risk-adjusted” FLIPI*
or “risk-adapted” therapeutic strategies. These include trials Low 0-1 36 91 71
of high-dose therapy (HDT) and autologous stem cell trans- Intermediate 2 37 78 51
plantation (ASCT) for aggressive lymphoma patients with High ≥3 27 53 36
high IPI scores; however, such strategies currently are not *FLIPI risk factors: No-Lash = number of nodal sites of disease (>4);
established as standard approaches and remain experimental elevated LDH, age >60 years, stage III or IV disease, and hemoglobin
(see the section “Diffuse large B-cell lymphoma” later in this ≤12 g/L.

Table 21-7 The Mantle Cell Lymphoma International Prognostic NCCN guidelines at http://www.nccn.org/about/nhl.pdf,
Index (MIPI) which is an outstanding resource for the treating clinician.
Points Age, Years ECOG PS LDH/ULN WBC, cells/mm3 Patient surveillance following treatment of lymphoma
0 <50 0-1 ≤0.67 <6,700
should address both long-term complications of therapy and
1 50-59 — 0.67-0.99 6,700-9,999 disease recurrence. Long-term effects of therapy depend on
2 60-69 2-4 1.00-1.49 10,000-14,999 the type of treatment and whether radiotherapy was also
3 ≥70 — ≥1.50 ≥5000 administered. For example, radiotherapy to the head and
neck region leads to decreased salivation with dental caries,
MIPI risk factors are age, PS, LDH, WBC level.
and if the thyroid was included in the radiation field, a large
Formula for MIPI: [0.03535 × age (years)] + 0.6978 (if ECOG >1) +
[1.367 × log10(LDH/ULN)] + [log10(WBC count)].
proportion of patients eventually may become hypothyroid.
Simplified MIPI: low risk = 0-3 points; intermediate risk = 4-5 Thyroid-stimulating hormone (TSH) level should moni-
points; high risk = 6-11 points. tored with each follow-up visit. Hypothyroidism may raise
ECOG PS = Eastern Cooperative Oncology Group performance the LDH and prompt a fruitless search for disease recur-
status; LDH = [lactate] dehydrogenase; PS = performance status; rence. Women who have had mantle radiation should
ULN = upper limit of normal; WBC = white blood cell. receive a mammogram 10 years after radiation or at age 40
years. In younger women, MRI breast imaging also can be
considered given reduced sensitivity of mammogram.
Role of FDG-PET imaging
Long-term survivors are at risk of second malignancies,
FDG-PET scanning is useful both for staging and assessing which is dependent on the treatment administered. For
response to lymphoma therapy and is generally recom- example, radiated patients are at risk for carcinomas and sar-
mended as part of routine staging and end-of-treatment comas in the radiated field, while those who have had alkyl-
response assessment in FDG-avid lymphomas. The 5-point ating agents are at risk for therapy-related myelodysplastic
scale (Deauville criteria, Table 21-8) should be used for PET syndrome or acute myeloid leukemia. Once primary therapy
interpretation and scores of 1-3 at completion of therapy are has been completed and remission documented, patients
considered consistent with complete remission, regardless of typically are followed every 3 months for the first 2 years,
the size of any residual masses. Some studies indicate that then every 6 months until 5 years, and then annually
interim PET scanning, performed midtreatment, can iden- thereafter.
tify patients at higher risk for treatment failure; however, it is Most recurrences of aggressive lymphoma occur in the
unknown whether therapy should be altered based upon the first 2 years after treatment, although late relapses beyond 5
results of a midtreatment PET scan. False-positive results years do occur in a small minority of patients. Patients with
can occur in the setting of inflammation, granulomatous indolent lymphoma have a lifelong risk of relapse and typi-
disease, and infection, and a biopsy should be performed in cally are seen every 3 months for the first 2 years and then
a PET-positive patient in a remission by CT scan if high-dose every 6-12 months indefinitely. There is no evidence that
chemotherapy and stem cell transplant (HDC/SCT) are routine CT or PET imaging affects outcome of patients, and
under consideration. newer guidelines recommend minimizing surveillance imag-
ing in indolent lymphomas and discourage any surveillance
imaging in aggressive lymphoma.
Patient management and follow-up
With over 60 lymphoma subtypes, detailed management

guidelines for each subtype and disease stage are beyond the Key points
scope of this chapter. The reader is encouraged to refer to the
• NHLs are biologically and clinically heterogeneous; accurate
diagnosis by a hematopathologist using the WHO classification is
essential for optimal management.
Table 21-8 Deauville 5-point scale for PET interpretation in • The majority of NHLs are of B-cell origin and are categorized
lymphoma broadly as indolent versus aggressive subtypes.
• The incidence of NHL is increasing in Western countries.
Score Visual description
• Specific chromosomal translocations are associated with
1 No uptake specific subtypes of lymphoma and are pathogenetically
2 Uptake ≤ mediastinum involved in malignant transformation and progression.
3 Uptake > mediastinum but ≤ liver • The IPI score provides important prognostic information for
4 Update moderately higher than liver outcome and survival in aggressive lymphomas. The FLIPI has
5 Update markedly higher than liver been developed specifically for FL.

Indolent B-cell NHL | 583
Indolent B-cell NHL derived from follicle center cells and can have a follicular,
follicular and diffuse, or diffuse growth pattern. Unlike
The indolent B-cell lymphomas include the cell types shown nodal FL, the neoplastic cells are usually BCL-2 negative. It
in Table 21-3, and the most commonly encountered subtype typically occurs as solitary or localized skin lesions on the
is FL, which accounts for 20%-30% of all lymphomas. Other scalp, forehead, or trunk, and only 15% present with multi-
subtypes include marginal zone lymphomas (nodal, splenic, focal lesions. The clinical course is usually very indolent and
and extranodal [MALT] types) and lymphoplasmacytic lym- can be managed with low-dose radiation and other site
phoma. This category also includes CLL/SLL, which is dis- directed approaches.
cussed in Chapter 22. Gene expression profiling has been explored in FL. In the
largest study, molecular signatures divided patients into four
Clinical case quartiles with widely disparate median survival times (3.9,
10.8, 11.1, and 13.6 years; Dave et al., 2004). Interestingly,
A 53-year-old man is diagnosed with stage IV FL after noticing a the signatures largely consisted of nonmalignant cells from
lump on his neck while shaving. A biopsy reveals a lymph node the microenvironment. One signature was termed immune
with enlarged, closely packed follicles with distorted architec-
response-1 and was associated with a more favorable progno-
ture. Inside the follicles are small lymphocytes with irregular
sis and had high expression of genes expressed in T-cells. In
nuclei. The cells stain positive for CD20 and CD10. The staging
contrast, immune response-2 had high expression of genes
evaluation reveals widespread lymphadenopathy, involving five
nodal groups, with the largest node measuring just over 3 cm
expressed in monocytes or dendritic cells. Whether these sig-
and 10% marrow involvement. The hemoglobin and LDH are natures can be used to develop targeted therapies or are still
normal. He has no disease-related symptoms and his Eastern relevant in rituximab-treated patients is unknown.
Cooperative Oncology Group (ECOG) PS is 0. The FLIPI score is
2 and he has low tumor burden by Groupe d’Etude des Lympho-
Management of localized follicular lymphoma
mes Folliculaires (GELF) criteria.
Limited-stage (Ann Arbor I or II) FL is relatively uncommon
and as a result, there are no randomized studies indicating
Follicular lymphoma
the optimal management strategy. Rather, most of the data
FL is the prototypical and most common indolent lym- are observational, derived from single-institution databases.
phoma, with about 15,000 new cases diagnosed each year in Older studies suggested a proportion of patients might be
the United States. Although incurable, the prognosis is rela- cured with an external beam radiation. MacManus and
tively good (median OS >10 years) and appears to be improv- Hoppe (1996) found that ~40% of limited-stage patients
ing in the immunochemotherapy era. with FL remained disease free at 10 years after radiation
FLs are derived from GCB and are graded based on treatment; late relapses beyond 10 years were unusual. Other
the number of centroblasts per high-power field: grade 1 studies also reported a 10-year disease-free survival (DFS)
(0-5), grade 2 (6-15), and grade 3 (>15). Grade 3 is further rate of ~40%-50%, suggesting that cure is possible in a pro-
classified into Grade 3A (centrocytes present) and grade 3B portion of patients with this approach (Wilder et al., 2001).
(solid sheets of centroblasts). For purposes of classification Given the excellent long-term outcomes for patients with
and clinical management, grades 1 and 2 can be considered localized FL, there is concern for late-onset radiation-
one entity, and the WHO classification uses the term induced complications, including second primary cancers.
grade 1-2. Grade 3 FL is relatively uncommon (<20% of all Recent data indicate that radiation fields can be reduced
FLs), and the natural history of this entity is less clear. Most without adversely impacting disease control (Campbell
contemporary clinical trials will allow grade 3A to be et al., 2011). As a result, contemporary strategies tend to uti-
included with grade 1-2 cases, whereas grade 3B is excluded. lize an involved field approach. Studies evaluating chemo-
It is believed that grade 3B is best managed like therapy plus radiation (combined modality therapy [CMT])
DLBCL, but whether or not it is curable remains unclear. have demonstrated improved progression-free survival
Immunophenotypically, FL cells are CD20+, CD10+, BCL6+, (PFS) without an obvious effect on OS (Seymour et al.,
BCL2+, and CD5-. Up to 90% of cases have a t(14;18) with a 2003). Therefore, the CMT approach is likely best reserved
higher frequency observed in grade 1-2 FLs. for the rare patient who presents with bulky (node >5 cm)
In the 2008 WHO classification, there are a number of limited stage FL. Finally, an alternative management strategy
identified variants of FL. These include primary intestinal for this patient population is watch and wait. A Stanford
FL, extranodal FLs, and pediatric FL. Primary cutaneous fol- report of stage I and II patients, who received no initial ther-
licular center lymphoma, a provisional entity in the updated apy, showed that more than half of the 43 patients did not
WHO classification, should be distinguished from FL. It is require therapy at a median of 6 years and 85% of patients

were alive at 10 years (Advani et al., 2004). A report from a reasonable. From a decision-making standpoint, patients
large observational database found the following treatment with symptomatic, high-tumor-burden FL are the most
approaches were utilized for 471 stage I FL patients: R-chemo straightforward. They require treatment, although there is
28%, XRT 27%, observation 17%, CMT 13%, rituximab little consensus on which treatment is best.
12%, and other 3% (Friedberg et al., 2012). Approaches uti-
lizing systemic therapy produced better PFS outcomes than
Management of asymptomatic, low-tumor-burden
XRT alone. There were no OS differences between any of the
follicular lymphoma
approaches.
Asymptomatic patients may be candidates for a strategy of
watch and wait. To determine whether watch and wait is an
Approach to patients with advanced-stage
option, one should make an assessment of the tumor bur-
follicular lymphoma
den. The GELF criteria (Table 21-10) are the most com-
Patients with advanced stage FL generally are considered monly used criteria to assess tumor burden and to assess
“incurable” with standard chemotherapy. The disease gener- eligibility for clinical trials. The watch-and-wait strategy was
ally is responsive to treatment, however, and there are first advocated at Stanford University when two retrospec-
numerous effective treatment options. As a result, the prog- tive studies suggested no detriment in patient outcome.
nosis is excellent relative to other cancers. A typical patient Three randomized clinical trials later confirmed the Stanford
will undergo a number of different treatments, often sepa- observations. Low-tumor-burden FL patients assigned to
rated by several years. Advanced-stage FL can be thought of watch and wait experienced the same OS compared with
as a chronic disease that requires long-term management, patients assigned immediately to treatment. The median
and the management is largely a matter of how to sequence time to first chemotherapy in all studies was 2.3-3 years. All
the different therapies. of these studies, however, were conducted in the preritux-
The approach to a newly diagnosed patient needs to be imab era.
individualized, factoring in the presence or absence of symp- To date, there are no studies comparing rituximab plus
toms, the tumor burden, the patient age and comorbidities, chemotherapy to watch and wait, and there is only one ran-
and the goals of therapy. A 2 × 2 table can be constructed to domized clinical trial comparing single-agent rituximab to
help with the initial approach of separating patients by watch and wait in patients with previously untreated, asymp-
symptoms and tumor burden (Table 21-9). Using this tomatic, low-tumor-burden FL (Ardeshna et al., 2014).
approach, four patient categories are generated: (i) asymp- Patients were assigned to watch and wait (Arm A), rituximab
tomatic, low tumor burden; (ii) asymptomatic, high tumor at 4 weekly doses (Arm B), or rituximab at 4 weekly doses
burden; (iii) symptomatic, low tumor burden; and (iv) plus a single dose every 2 months for 2 years (Arm C). A
symptomatic, high tumor burden. Patients with asymptom- significant prolongation in PFS and prolongation in the time
atic, low-tumor-burden FL can be considered for watch and to first chemotherapy was observed for the patients random-
wait or single-agent rituximab. Patients with asymptomatic, ized to rituximab. With a median follow-up of 32 months,
high-tumor-burden FL generally start therapy soon after the proportion of patients progression free at 3 years was
diagnosis, although selected patients may be observed ini- 33%, 60%, and 81% in Arms A, B, and C, respectively. The
tially, such as the very elderly or those who just meet the proportion of patients free of chemotherapy or radiation at
high-tumor-burden criteria (eg, three nodes in the 3-4 cm 3 years was 48%, 80%, and 91% in Arms A, B, and C,
range). Patients with symptomatic, low tumor burden are
uncommon and care should be taken to look for alternative Table 21-10 GELF criteria for high tumor burden
causes of the patient-reported symptoms. If no alternative
Any nodal or extranodal mass >7 cm
explanations are uncovered, then initiation of treatment is
Three or more nodal sites with diameter of >3 cm
Elevated LDH
Table 21-9 Algorithm for the approach to the newly diagnosed FL Hb <10 g/dL, ANC <1.5 × 109/L, Plts <100 × 109
patient Spleen >16 cm by CT scan
Risk or organ compression or compromise
Low tumor burden High tumor burden Significant serous effusions
Symptoms Watch and wait vs R-chemotherapy +/– MR
Meeting any one criterion qualifies as high tumor burden. All must
absent single-agent rituximab vs watch and wait
be absent to qualify as low tumor burden.
Symptoms Single-agent rituximab R-chemotherapy +/– MR
ANC = absolute neutrophil count; GELF = Groupe d’Etude des
present vs R-chemotherapy
Lymphomes Folliculaires; Hb = hemoglobin; LDH = lactate
R = rituximab; MR = maintenance rituximab. dehydrogenase; Plts = platelets.

respectively. There is no difference, however, in the OS at 3 Therapy of symptomatic, high-tumor-burden

years (95% in all arms). The study also evaluated quality of follicular lymphoma
life (QOL). Given that these patients are symptom free, the
The addition of rituximab to conventional chemotherapy,
main QOL issues tend to be anxiety, depression, and adjust-
has improved outcomes in FL, including response rates, PFS,
ment to illness. The study found that anxiety and depression
event-free survival (EFS), and OS. Table 21-11 summarizes
were more common in patients with low-tumor-burden FL
major studies combining rituximab with chemotherapy.
than in the general population, but still relatively infrequent
Clearly, rituximab added to chemotherapy is a therapeutic
at 13% and 3%, respectively. Patients in all treatment arms
advance in FL; however, the optimal chemotherapy backbone
adapted to their illness over time. The patients identified as
remains unsettled. Data generated prior to the introduction
“anxious” adapted more readily when assigned to rituximab
of bendamustine in the US indicated the most commonly
treatments. The interpretation of this study vary. It is rea-
used regimens in the United States were R-CHOP (rituximab,
sonable to conclude that: (i) given no OS difference observed
cyclophosphamide, vincristine, prednisone) (60%), R-CVP
to date, watch and wait remains a reasonable standard for
(rituximab, cyclophosphamide, prednisone) (27%), and R-
the asymptomatic, low-tumor-burden FL population; (ii)
fludarabine-based (13%) (Friedberg et al., 2009). A random-
some benefits are associated with immediate rituximab ther-
ized comparison of these regimens indicated R-CHOP had the
apy, such as improved PFS and a longer time to first chemo-
best risk-benefit profile of the three, as it was more active than
therapy (these benefits should be discussed with patients);
R-CVP and less toxic than R-FM (Federico et al., 2013).
and (iii) a subset of patients (perhaps 15%) with particular
However, bendamustine, an alkylating agent, has gained
difficulty adjusting to their diagnosis may experience a QOL
widespread adoption as the chemotherapy platform of
benefit from single-agent rituximab.
choice in FL. A phase 3 trial comparing BR to R-CHOP
If administering single-agent rituximab to a patient with
demonstrated better efficacy and reduced toxicity with BR
low-tumor-burden FL, should one utilize a maintenance
(Rummel et al., 2013). In this multicenter phase 3 study, 549
strategy or simply retreat at progression? This dosing ques-
patients with high-tumor-burden indolent NHL and MCL
tion was addressed in the RESORT study (Kahl et al., 2014).
(median age 64 years) were randomized to receive benda-
After induction therapy with single-agent rituximab, patients
mustine 90 mg/m2 on days 1 and 2, with rituximab 375 mg/m2
with low-tumor-burden FL were randomized to receive
on day 1, every 28 days (the BR group) or to receive stan-
maintenance rituximab until treatment failure or to be peri-
dard R-CHOP chemotherapy every 21 days. The overall
odically retreated with rituximab (retreated with 4 weekly
response rates (ORR) were similar in the BR versus R-CHOP
doses at each progression) until treatment failure. The trial
groups (92.7% vs 91.3%, respectively), but the CR rate was
revealed no difference in the time to treatment failure
significantly higher in the BR group (39.8%) compared with
between the two dosing strategies. Patients on the mainte-
the R-CHOP group (30.0%) (P = 0.03). When evaluating
nance arm, however, utilized four times as much rituximab.
just the FL patients, with a median follow-up of 45 months,
There was no difference in quality of life, depression, or anx-
the median PFS was significantly longer after BR compared
iety between the two strategies. On the basis of these results,
with R-CHOP (median PFS, not reached vs 40.9 months, P =
a re-treatment strategy is preferred if opting for single-agent
0.007). OS did not differ between both groups. There was
rituximab in this patient population.
less hematologic toxicity, alopecia, infections, peripheral
Table 21-11 Randomized trials of chemotherapy versus R-chemotherapy in high tumor burden, advanced-stage follicular lymphoma
Study Treatment N Median follow-up ORR Time to event OS

Hiddemann, et al. R-CHOP vs CHOP 223 vs 205 1.5 years 96% vs 90% 88% vs 70% (2-year 95% vs 90%
Blood. 2005 DoR) (2-year OS)
Marcus, et al. R-CVP vs CVP 162 vs 159 4.5 years 81% vs 57% 38 months vs 14 months 83% vs 77%
J Clin Oncol. 2008 (median DoR) (4-year OS)
Herold, et al. R-MCP vs MCP 181 vs 177 6 years 92% vs 75% 57% vs 25% (6-year 80% vs 65%
Ann Onc. 2011 PFS) (6-year OS)
Bachy, et al. R-CHVP-IFN(6) vs 175 vs 183 8 years 81% vs 72% 44% vs 28% (8-year 70% vs 79%*
Ann Onc. 2011 CHVP-IFN(12) EFS) (8-year OS)
CVP = cyclophosphamide, vincristine, prednisone; DoR = duration of response; DFS = disease-free survival; EFS = event-free survival; MCP =
mitoxantrone, chlorambucil, prednisone; R-CVP = rituximab, cyclophosphamide, vincristine, prednisone; R-MCP = rituximab, mitoxantrone,
chlorambucil, prednisone.
* P value not significant.

neuropathy, and stomatitis with BR. Drug-associated ery- use depends on a number of factors, including the prior
thematous skin reactions were seen more frequently in the treatment utilized, duration of prior response, patient age,
BR group. These data suggest that BR is a better option for comorbid illnesses, and goals of therapy. Options range from
untreated high-tumor-burden FL. low-risk strategies, such as single-agent rituximab, to high-
A confirmatory randomized phase 3 trial (BRIGHT study) risk/high-reward strategies, such as allogeneic stem cell
was conducted in North America (Flinn et al., 2014). Previ- transplantation (alloSCT), with many options in between. A
ously untreated indolent NHL patients with high tumor bur- recent report from a large observational database reveals a
den were randomized to BR or R-CHOP/R-CVP. Control substantially worse overall survival for patients who relapse
arm patients were identified as an R-CHOP or R-CVP candi- within 2 years of initial therapy with R-CHOP compared to
date prior to randomization. The primary endpoint was to patients who remain in remission longer than 2 years (Casulo
show noninferiority of BR in the CR rate. Seventy percent of et al., 2015). These high-risk patients should be considered
the 447 enrolled patients had FL, and in these patients BR for more aggressive strategies (such as SCT) or investiga-
therapy was found to be noninferior to the R-CHOP/R-CVP tional approaches.
control arm for CR rate (30% vs 25%) and the overall response Bendamustine is approved in the United States for use in
rate (99% vs 94%). Time-to-event data was not reported. Side patients with rituximab-refractory indolent B-cell lym-
effect profiles were distinct, with more GI toxicity and rash phoma. A pivotal trial in 100 patients reported an ORR of
with BR and more neuropathy and alopecia with R-CHOP/R- 75% with a median PFS of 9.3 months (Kahl et al., 2010).
CVP. Although, the BRIGHT data do not exactly replicate the The FDA-approved dose of single-agent bendamustine is
StIL data for BR, they do suggest that BR remains a very rea- 120 mg/m2 given intravenously on days 1 and 2 of 21-day
sonable alternative to R-CHOP or R-CVP in FL. cycles. Approximately two-thirds of patients required dose
The question of whether to administer maintenance ritux- modifications or delays, mainly due to cumulative myelo-
imab after frontline R-chemotherapy was addressed in the suppression. In addition, most practitioners prefer to admin-
phase 3 PRIMA trial (Salles et al., 2010). The study evaluated ister bendamustine with rituximab. An expert panel has
the efficacy and safety profile of maintenance rituximab in published guidelines on bendamustine dosing when com-
newly diagnosed FL patients who responded to initial treat- bined with rituximab, and a recommend 90 mg/m2 on days
ment with rituximab plus chemotherapy. Induction treat- 1 and 2 repeated every 28 days (Cheson et al., 2010). Lower
ment was selected by center; R-CHOP (75%), R-CVP (22%), doses (such at 70 mg/m2) may be more appropriate in the
or R-FCM (3%). Patients were randomized to either obser- elderly.
vation or a single dose of rituximab every 2 months for 2 A new option for relapsed FL is the PI3Kδ inhibitor ide-
years. At a median follow-up of 36 months from randomiza- lalisib. Idelalisib targets the delta isoform of PI3K, an
tion, the 2-year PFS in the maintenance rituximab arm was enzyme downstream from the B-cell receptor which even-
75% versus 58% in the observation arm (P < 0.0001). The tually signals through AKT and mTOR. In a phase 2 study
beneficial effect of maintenance rituximab was seen irrespec- of 125 patients with indolent NHL who were considered
tive of the induction chemotherapy backbone and in both refractory to both rituximab and an alkylating agent, idelal-
CR and partial remission (PR) patients. Grade 3-4 adverse isib was administered at a dose of 150 mg BID until PD or
events were slightly higher in the maintenance rituximab patient withdrawal (Gopal et al., 2014). The response rate
arm (24% vs 17%). No difference in OS was observed. Given was 57% with a median duration of 12.5 months. Grade 3
the lack of OS benefit, the decision regarding the use of or higher toxicities included neutropenia (27%), transami-
maintenance rituximab can be individualized. Rituximab nase elevations (13%), diarrhea (13%), and pneumonia
administration does carry low risk for neutropenia and (7%). Based upon this data, idelalisib received accelerated
rarely more serious toxicities such as progressive multifocal approval by the FDA in 2014. Another option for patients
leukoencephalopathy. As maintenance, rituximab generally who relapse after an alkylator-based therapy are fludara-
is well tolerated; it has become a commonly utilized strategy bine-based regimens. They should be used with caution in
in the United States. Regarding the duration of maintenance heavily pretreated or elderly patients, however, due to
rituximab, available data suggest that 2 years is safe, but data immunosuppression. Patients are at increased risk of Pneu-
beyond 2 years is lacking. mocystis jirovecii and reactivation of herpes zoster. Prophy-
laxis with trimethoprim/sulfamethoxazole and acyclovir
should be considered, particularly in elderly patients. Fur-
Therapy for relapsed and refractory follicular lymphoma
thermore, if ASCT is considered as a future treatment
Multiple options exist for the treatment of patients who have option, the number of cycles with fludarabine should be
failed first-line therapy, and the decision of which therapy to minimized to avoid stem cell toxicity.

RIT is also a viable option for patients with indolent B-cell Marginal zone lymphomas
NHL if the bone marrow is minimally involved and the dis-
The WHO classification separates the marginal zone B-cell
ease is not bulky. With Y90 ibritumomab tiuxetan, response
lymphomas (MZL) into extranodal MZL of MALT, nodal
rates are ~70% and response duration is, on average, 11-15
MZL, and splenic MZL (SMZL). The morphology of these
months. Single-agent rituximab can be used in relapsed lym-
disorders is characterized by an infiltrate of centrocyte-like
phoma, although now that most patients have received it with
small cleaved cells, monocytoid B-cells, or small lympho-
their primary therapy and possibly as maintenance, more and
cytes; they may exhibit an expanded marginal zone sur-
more patients are becoming rituximab refractory. For patients
rounding lymphoid follicles. The immunophenotype is
who are still rituximab sensitive, it is an attractive option for
characterized by expression of CD20 but lack of CD5 or
elderly patients who will not tolerate cytotoxic agents well.
CD10 expression (Table 21-2); this marker profile is useful
in distinguishing MZL from SLL, MCL, and FL. A feature
Stem cell transplantation common to many cases of MZL is association with chronic
antigenic stimulation by microbial pathogens or autoanti-
HDC with autologous stem cell transplant (ASCT) and
gens, as described above. Examples include gastric MALT
alloSCT are both useful strategies in the management of FL,
(H pylori), cutaneous MALT (B burgdorferi), ocular adenxal
particularly for younger patients with high-risk features,
MALT (C psittaci), nodal MZL (hepatitis C), SMZL (hepa-
such as a brief remission to previous therapy. A review of 904
titis C), parotid MALT (Sjögren syndrome), and thyroid
patients in the International Bone Marrow Transplant Reg-
MALT (Hashimoto thyroiditis). There is significant geo-
istry who underwent autologous or allogeneic transplanta-
graphic variation in the association with certain microbial
tion for FL revealed that durable remissions could be induced
pathogens. For example, the prevalence of C psittaci in
with either technique (van Besien et al., 2003). A lower
patients with ocular adnexal MALT appears to be 50%-80%
5-year recurrence rate with allogeneic transplantations was
in Italy, Austria, Germany, and Korea, whereas this organ-
offset by a higher treatment-related mortality (TRM) com-
ism is observed infrequently in Japan, China, and the
pared with autologous transplantation, leading to similar
United States.
5-year survival rates of 51%-62%. To reduce the TRM of
alloSCT, most centers now favor a nonmyeloablative strat-
egy in FL. Results utilizing a nonmyeloablative allogeneic
MALT lymphomas
SCT strategy vary widely in the literature. For example, a
series of 62 patients treated at the Fred Hutchinson Cancer Extranodal MZLs or MALT lymphomas constitute ~70% of
Center (FHCC) demonstrated a 3-year OS and PFS of 67% all MZLs. They occur in mucosal sites, predominantly gas-
and 54%, respectively (Rezvani et al., 2008). Alternatively, a tric or intestinal, and some nonmucosal extranodal sites,
highly selected group (n = 47) treated at the M.D. Anderson including the lung, salivary gland, ocular adnexa, skin, and
Cancer Center achieved an 11-year OS and PFS of 78% and thyroid. These sites often are affected by chronic infection
72%, respectively (Khouri et al., 2012). or inflammation, such as Sjögren syndrome or Hashimoto
There is one small, randomized clinical trial (the CUP thyroiditis. The typical presentation of MALT lymphoma is
trial), examining ASCT versus standard therapy in patients an isolated mass in any of these extranodal sites or an ulcer-
with relapsed follicular lymphoma (Schouten et al., 2003). ative lesion in the stomach. Clinically, they are typically
The study, conducted in the prerituximab era, found indolent lymphomas, with 10-year OS rates of 90% in many
improved PFS and a trend toward improved OS. An interest- series. MALT lymphomas can be characterized as either gas-
ing long-term analysis of patients receiving myeloablative tric (30%-40%) or nongastric (60%-70%), and the approach
chemotherapy followed by ASCT comes from investigators to disease management is site specific. Approximately 90%
at St. Bartholomew’s Hospital (London) and the Dana- of gastric MALT lymphomas are associated with H pylori
Farber Cancer Institute (Boston) (Rohatiner et al., 2007). A infection. Newly diagnosed patients typically report dys-
cohort of 121 patients, with a median follow-up of 13.5 pepsia, pain, reflux symptoms, or weight loss. Upper endos-
years, were noted to have a plateau in the remission duration copy can reveal erythema, erosions, ulcers, or masses.
curve beginning around year 8. Nearly half the patients were A consistent observation has been that 70%-80% of gastric
still in remission at 10-15 years, suggesting some patients MALT lymphomas durably regress following effective
may be cured. Results were substantially better for patients H pylori antibiotic therapy (Nakamura et al., 2012). The
treated in second remission as opposed to later in the disease most widely used antibiotic regimen is a combination of
course, suggesting there may be an optimal window to con- amoxicillin, omeprazole, and clarithromycin. Metronida-
sider ASCT in FL. zole is an effective alternative antibiotic in patients with a

penicillin allergy. Lymphoma responses can be slow, taking Nodal MZL

up to 6 months to 1 year. Repeat assessment of H pylori
Nodal MZL, previously known as monocytoid B-cell lym-
either by histologic examination or a urea breath test is nec-
phoma, also arises from marginal zone B-cells. Whenever
essary to ensure that the bacteria have been eradicated. The
nodal MZL is diagnosed, a careful history and physical
strongest predictor for lymphoma nonresponse to antibi-
examination should be pursued for a possible coexisting
otic therapy is the presence of the t(11;18) translocation,
extranodal MALT lymphoma component, which may be
which is present in 20%-30% of cases. In the series reported
identified in up to one-third of cases. It more commonly
by Nakamura et al. (2012), only 3 out of 30 patients with
presents with advanced-stage than with MALT-type MZL.
t(11:18) experienced lymphoma regression following
The t(11;18) karyotypic changes identified in MALT are
H pylori eradication therapy. In patients who do not respond
absent in nodal MZL, and no specific or recurring karyotypic
to antibiotics or in H pylori-negative cases, involved-field
anomaly has been described. IgM monoclonal gammopathy
radiotherapy (IFRT) has been highly effective with DFS or
can occur in ~10% of cases. HCV infection is reported in up
PFS rates of >90% at 10 years (Goda et al., 2010). The prog-
to 25% of patients. Across reported series, the 5-year OS is
nosis for early stage gastric MALT is excellent, with most
60%-70%; however, the EFS is only 30%, which likely reflects
series reporting 10-year OS rates in excess of 90%. For
more commonly encountered advanced-stage disease. Man-
patients with advanced stage disease, regimens similar to
agement is similar to the approach recommended in follicu-
those used in FL, including rituximab-based combinations,
lar lymphoma.
can be used. Transformation to DLBCL is possible, but a
In the updated WHO classification, a new category was
remarkable observation has been the regression of early
introduced—pediatric nodal MZL—which has distinctive
stage H pylori-positive gastric diffuse large B-cell lympho-
clinical and morphologic characteristics. There is a male pre-
mas with H pylori eradication therapy. This observation was
dominance (20:1), and patients usually present with local-
noted in DLBCL clearly arising from gastric MALT (trans-
ized asymptomatic adenopathy in the head and neck region.
formation) and in de novo DLBCL (no apparent underlying
Morphologically, the infiltrate is similar to that seen in
MALT) (Kuo et al., 2012).
adults, except that progressively transformed germinal cen-
Nongastric MALT lymphomas usually have an indolent
ters often are seen.
course, including the one-third of patients who present
with stage IV disease (Thieblemont et al., 2000). OS at 10
years exceeds 90% in many series. The most common loca-
Splenic MZL
tions are the salivary glands (26%), ocular adnexa (17%),
skin (12%), lung (8%), upper airways (7%), thyroid (6%), SMZLs are uncommon. The median age at diagnosis is 68
and intestinal tract (5%) (Zucca et al., 2003). Treatment years, and it is more common in females. Patients usually
approaches depend on both stage and site of primary present with symptomatic splenomegaly. Generalized
involvement and may include surgery, radiation therapy, lymphadenopathy is uncommon, but patients may have
or chemotherapy. Radiation therapy produces excellent associated splenic hilar nodal or hepatic involvement. The
results in limited-stage disease (Goda et al., 2010). Many bone marrow and blood typically are involved and villous
patients can be managed with a watch-and-wait approach. lymphocytes may be seen. Diagnosis usually is based on
Patients with advanced-stage disease typically can be man- spleen histology following splenectomy or after bone mar-
aged using the same principles used for FL. Rituximab row examination. Clinically, it can be confused with CLL,
added to chemotherapy was shown to improve EFS in a MCL, FL, HCL, or WM. Unlike CLL and MCL, it is typically
RCT (Zucca et al., 2013). Recurrences tend to occur in a CD5 negative. Unlike FL, it is CD10 negative. Unlike HCL,
site-specific fashion (ie, pulmonary MALT tends to recur in which is CD103 positive and replaces the splenic red pulp,
the lung), and so monitoring can be primarily (although SMZL is CD103 negative and replaces the splenic white pulp.
not exclusively) site directed. An emerging story is the asso- There is no reliable immunophenotypic method to distin-
ciation of C psittaci and ocular adenexal MALT. Several guish SMZL from WM, so morphologic and clinical features
regions of Europe have reported associations in >50% of must be used. Clinically, SMZL is more likely to have signifi-
cases. A recent European multicenter phase 2 study, testing cant splenomegaly, whereas WM typically has a larger
the efficacy of doxycycline, detected evidence of C psittaci monoclonal protein (Arcaini et al., 2009). A prognostic
in 39 out of 44 patients (Ferreri et al., 2012). Lymphoma model, using hemoglobin <12 g/dL, elevated LDH, and albu-
regression was observed in 65% of the doxycycline-treated min <3.5 g/dL, has identified three distinctive risk groups
patients and the responses tended to be durable. Nonre- (Arcaini et al., 2006). OS at 5 years was 88%, 73%, and 50%
sponse was associated with failure to eradicate the C psittaci for patients with 0, 1, and 2-3 risk factors, respectively. Sple-
organism. nectomy has long been considered the optimal first-line

therapy in symptomatic patients or in the case of cytopenias is used more commonly because of the short duration of
due to splenomegaly. However, single-agent rituximab has therapy required, and it also is available as a subcutaneous
been reported to be remarkably active, with an ORR of 100% injection. In one large series of 233 patients with long-term
in a series of 16 patients. In an observational retrospective follow-up, the ORR and CR rate with either of these agents
study, rituximab produced more durable remissions than was 97% and 80%, respectively (Else et al., 2009). The
did splenectomy (Kalpadakis et al., 2013). Regimens active median recurrence-free survival was 16 years, and many of
in other indolent lymphomas are appropriate for patients the relapses were observed 5-15 years after treatment, high-
requiring systemic therapy. Some cases have been associated lighting the unique natural history of this disease. It cur-
with hepatitis C infection, and responses have been reported rently is recommended that assessment of response should
with clearance of the virus with pegylated interferon and be determined 4-6 months following the end of treatment
ribavirin (Hermine et al., 2002). and if only a PR is attained then a second course can be given
(Jones et al., 2012).
HCL-variant is categorized separately in the WHO classi-
Lymphoplasmacytic lymphoma and
fication and, despite its name, it is considered unrelated to
Waldenström macroglobulinemia
HCL. It does not harbor the BRAF-V600E mutation. It dif-
Lymphoplasmacytic lymphoma (LPL) is defined in the fers from HCL in the lack of monocytopenia and by the pres-
WHO classification as an indolent neoplasm of small B- ence of an elevated white blood cell count. The bone marrow
lymphocytes, plasmacytoid lymphocytes, and plasma cells. is easier to aspirate as the reticulin fiber content is low. The
The lymphoma cells may express B-cell markers CD19 and immunophenotype of HCL-variant also differs in that the
CD20 and are CD5 and CD10 negative, much like the MZLs cells are CD25 negative. CD103 is expressed infrequently
(Table 21-3). These entities are described further in Chapter 23. and CD11c is usually positive. Unlike HCL, HCL-variant
responds poorly to purine analogs. Splenectomy can result
in good PR in two-thirds of patients. Case reports have
Hairy cell leukemia
shown CRs after rituximab.
Hairy cell leukemia (HCL) is an indolent B-cell lymphopro-
liferative disorder characterized by neoplastic lymphocytes
Transformation to aggressive lymphoma in
with cytoplasmic “hairy” projections on the cell surface, a
indolent lymphomas
positive tartrate-resistant acid phosphatase stain, and an
immunophenotype positive for surface immunoglobulin, Histologic transformation (HT) is the development of
CD19, CD20, CD22, CD11c, CD25, and CD103 (Table 21-2). aggressive NHL in patients with an antecedent history of
Marrow biopsy demonstrates a mononuclear cell infiltrate indolent lymphomas. It most commonly occurs in follicular
with a “fried egg” appearance of a halo around the nuclei and lymphoma but can occur in any of the indolent lymphomas.
increased reticulin and collagen fibrosis. Nearly 100% of The British Columbia Cancer Agency reported on the inci-
cases will harbor the BRAF V600E mutation, abnormally dence and outcome of 600 patients with FL who subse-
activating the BRAF-MEK-ERK pathway, and BRAF inhibi- quently developed transformed lymphoma. Diagnoses were
tors have demonstrated single agent activity in relapsed HCL either made clinically (sudden increase in LDH >2× the
(Pettirosi et al., 2015). upper limit of normal, discordant nodal growth, or unusual
The typical presentation is that of a middle-age man extranodal sites of involvement) (37%) or pathologically
(median age, 50-55 years) with pancytopenia, monocytope- (63%). In this series, the annual risk of transformation was
nia, splenomegaly, and an inaspirable bone marrow (dry 3% per year, with a 10- and 15-year risk of 30% and 45%,
tap). HCL is rare, representing 2% of all leukemias. Making respectively. Overall, the median posttransformation sur-
the proper diagnosis is crucial because of its generally favor- vival time was 1.7 years, with superior outcomes observed in
able prognosis, with a 10-year OS exceeding 90%, and its limited-stage patients. Similar results were observed in a
excellent treatment response to nucleoside analogs (Grever, series from St. Bartholomew’s, where histologic transforma-
2010). Most patients with HCL require therapy to correct tion was observed in 28% of patients with FL by 10 years.
cytopenias and associated complications in addition to the However, outcomes for patients with HT appear to be mark-
presence of symptomatic splenomegaly. If a patient is asymp- edly better in the rituximab era with median OS of 50 months
tomatic and cytopenias are minimal, patients may be ini- in one series (Link et al., 2013). FDG-PET imaging can be
tially observed. HCL has a unique sensitivity to purine helpful to direct the site to biopsy when establishing HT.
analogs. The nucleoside analogs cladribine or pentostatin Histologically, DLBCL is the most frequently observed sub-
are the treatments of choice in HCL in view of the high type. One should assay for MYC and BCL-2 by FISH and by
response rates and durable remissions achieved. Cladribine immunohistochemistry. The treatment is directed at the

aggressive lymphoma and depends on a variety of factors, DLBCL constitutes 25%-30% of all NHLs and can present
including age, comorbidities, and extent of prior treatment with nodal or extranodal disease. Bone marrow involvement
for the FL. Anthracyclines should be included in the treat- with large cell lymphoma occurs in fewer than 10% of cases.
ment if not previously used and if there are no contraindica- Another 10%-20% of patients have discordant marrow
tions. Consideration for stem cell transplant consolidation is involvement with a low-grade B-cell lymphoma, despite a
warranted in eligible patients (Villa et al., 2013). nodal biopsy consistent with DLBCL.
In addition to the B-cell markers CD20 and CD19, the
Key points neoplastic cells may also express CD10 (30%-60%), BCL6
(60%-90%), and IRF4/MUM1 (35%-65%). Rare cases may
• Follicular NHL is the most common indolent NHL. express CD5 (10%) and must be distinguished from the
• Patients with asymptomatic, advanced-stage indolent NHL blastoid variant of MCL, which is cyclin D1 positive. As
may be followed without specific therapy to assess the pace of
described, two molecularly distinct subtypes of DLBCL are
disease, or single-agent rituximab may be used to delay the use
recognized: GCB, which has a gene expression profile similar
of systemic chemotherapy.
to germinal center B-cells (CD10+ and BCL6+); and ABC,
• Rituximab plus chemotherapy is recommended in patients with
symptomatic disease or high tumor burden by the GELF criteria.
which has a profile similar to activated peripheral B cells
• Maintenance rituximab improves PFS. To date, there is no (IRF4/MUM+) with a prominent NFkB gene signature.
proven OS benefit.
• There are a multitude of therapeutic options for relapsed
Clinical prognostic factors in DLBCL
indolent lymphoma, including stem cell transplantation.
Approximately 60% of patients diagnosed with DLBCL can
be cured with rituximab-based chemotherapy; however, low-
Aggressive B-cell lymphomas and high-risk groups can be further defined by clinical and
biological factors. Although the IPI is robust and relevant
The most prevalent of the aggressive lymphomas is DLBCL. in the modern rituximab treatment era, it does not capture
Other histologies in this category include MCL, BL, lympho- all prognostic information. Bulky disease (often defined at
blastic lymphoma, and most of the T- and NK-cell lympho- >10 cm, although there is not universal agreement on the
mas (Table 21-3). These neoplasms are characterized by a definition of bulk in DLBCL) appears to increase the risk for
more acute presentation and, although often curable (except local relapse. Particular sites of extranodal disease also are
for MCL), are associated with relatively short survival in the associated with prognosis. Concordant involvement of the
absence of therapy-induced remission. This chapter focuses bone marrow with DLBCL, but not discordant involvement
on the mature B- and T-/NK-cell neoplasms. with a low-grade lymphoma, is associated with an inferior
outcome (Sehn et al., 2011). The patient described earlier has
Clinical case an IPI score of 3 (advanced stage, poor PS, elevated LDH),
placing him in a high-intermediate risk group with an
A 52-year-old man is diagnosed with stage IVB DLBCL. On CT
expected 5-year probability of survival with R-CHOP of
imaging, the largest nodal mass was 6 cm in the retroperitoneal
50%-60%, and the bone marrow involvement does not
region, bone marrow biopsy shows involvement with a low-
affect outcome because it is low-grade lymphoma. Testicular
grade lymphoma. Laboratory studies show a normal complete
blood count (CBC), and chemistries aside from an LDH elevated
involvement with DLBCL also is associated with an aggressive
1.5 times normal. His ECOG PS is 2. Immunophenotypic stains course with a propensity for late relapses, CNS relapse (lepto-
of the lymphoma cells reveal them to express CD19, CD20, meningeal and parenchymal), and recurrence in the contra-
κ-light chains, BCL2, and MUM1/IRF4. They are negative for lateral testicle. Even patients with localized disease should be
CD10 and BCL6 expression. treated with six cycles of R-CHOP with radiotherapy to the
contralateral testicle. Strong consideration also is given for
CNS prophylaxis as outlined in the following section.
DLBCL is composed of large B-cells with a diffuse growth pat-

Biological prognostic factors in DLBCL
tern. The new WHO classification recognizes several subcate-
gories of DLBCL, including molecular subtypes (GBC and Although the IPI is easy to apply and remains valid in the
ABC; see later sections); pathologic subtypes, including T-cell- current treatment era, it fails to capture underlying biologi-
rich B-cell lymphoma; and defined disease entities, including cal heterogeneity. As described above, DLBCL can be divided
PMBCL. Other than primary CNS lymphoma (PCNSL) treat- molecularly by GEP into the GCB and ABC subtypes, which
ment approaches are similar for the DLBCL subtypes. also have a distinct signature from PMBCL. ABC DLBCL has

Aggressive B-cell lymphomas | 591
an inferior prognosis, independent of the IPI. Furthermore, lymphomas that can occur after a proceeding FL or de novo
it is clear that prognosis also is affected by the tumor micro- as the first manifestation of disease presentation. The combi-
environment in which a signature rich in extracellular matrix nation of MYC driving cellular proliferation and BCL2 pre-
and histocytes (stromal 1) is associated with a favorable venting apoptosis has proven to be very difficult to cure.
prognosis and a signature rich in angiogenesis markers (stro- Double-hit lymphomas (DHL) can occur with DLBCL or
mal 2) is associated with an aggressive course (Lenz et al., more commonly, B-cell lymphoma, unclassifiable with fea-
2008). The use of GEP has had limited clinical utility due to tures intermediate between BL and diffuse large B-cell lym-
long turnaround time, the need to use fresh frozen tissue, phoma (Johnson et al., 2009). The outcome is extremely
and technical complexity. poor in lymphomas with dual BCL2 and MYC translocation
Immunohistochemical (IHC) algorithms have been used B-cell lymphomas, and these patients are considered for
in an attempt to capture the cell-of-origin (COO) phenotype investigational therapies. With the recent availability of a
using a methodology that can be applied routinely in clinical MYC antibody for IHC analysis, two large-scale studies have
practice. Hans et al. (2004) first reported IHC algorithm to evaluated the prognostic importance of MYC and BCL2 pro-
distinguish the GCB versus non-GCB subgroups using tein expression (double expressers) in DLBCL patients
CD10, BCL6, and IRF4/MUM1. Using the cDNA microarray treated with R-CHOP chemotherapy (Green et al., 2012;
as the gold standard, the sensitivity of the IHC COO sub- Johnson et al., 2012). MYC protein expression was found in
grouping was 71% for the GCB group and 88% for the non- approximately one-third of cases, far more than that cap-
GCB group. Other algorithms have been proposed that also tured by fluorescence in situ hybridization (FISH) analysis
have a lower sensitivity than gene expression profiling. These (11%) or high MYC mRNA expression, suggesting the mul-
results, however, have been inconsistent as to whether the tiple roads of MYC deregulation exist. Importantly, the dou-
COO distinction by IHC can be applied to rituximab-treated ble expressers exhibited a poor outcome (5-year OS 39% vs
patients. One study found that none of the applied five dif- 70%, P < 0.001; Green et al., 2102: 3-year OS 43% vs 86%,
ferent IHC algorithms could distinguish COO subgroups P < 0.001; Johnson et al). The significance of this double-hit
with prognostic significance. In contrast, another study score is independent of the IPI and patients with a high IPI
found that the Tally algorithm, which uses CD10, GCET, and dual-protein expression have an extremely aggressive
IRF4/MUM1, and FOXP1, showed the best concordance course with a median PFS of only 6 months.
with microarray data and maintained prognostic signifi-
cance. Given these inconsistencies and lack of data suggest-
ing that alternate therapies may affect outcome, the COO Treatment of newly diagnosed DLBCL
information, whether by molecular profiling or immunohis-
Advanced-stage DLBCL
tochemistry, should not be used to direct treatment deci-
sions outside of clinical trials. The backbone of treatment of all subtypes of DLBCL is
Recent technological advances in GEP, allows real time anthracycline-based treatment with R-CHOP chemother-
COO determination from formalin-fixed paraffin-embedded apy. With this approach, ~60% of patients are cured.
tissue (FFPET). The Lymphoma/Leukemia Molecular Profil- Rituximab has several mechanisms of action, including
ing Project developed Lymph2Cx assay, a parsimonious digital the ability to sensitize otherwise-resistant lymphoma cells to
gene expression (NanoString)-based test for COO assignment chemotherapy agents in vitro, perhaps in part via downregu-
in FFPET. A 20-gene assay was trained using 51 FFPET biop- lation of the BCL-2 protein. The Groupe d’Etude des Lym-
sies and the locked assay was subsequently validated using an phomes de l’Adulte (GELA) published a landmark phase 3
independent cohort of 68 FFPET biopsies (Scott et al., 2014). clinical trial in which 399 patients 60-80 years of age with
Comparisons were made with COO assignment using the previously untreated advanced-stage CD20 + DLBCL were
original COO model on matched frozen tissue. The assay was randomized to receive CHOP for eight cycles or R-CHOP on
highly accurate, with only 1 case with definitive COO being a standard 21-day schedule. An improvement in all end-
incorrectly assigned with >95% concordance of COO assign- points, including CR rate, EFS, and OS, favoring R-CHOP
ment between two independent laboratories. The test turnover CHOP was demonstrated. With longer follow-up, the
around time is several days making Lymph2Cx attractive for results held, and R-CHOP quickly became the standard of
implementation in clinical trials and practice. care for advanced-stage DLBCL around the world (Coiffier
MYC is translocated in ~5%-10% of DLBCLs, and early et al., 2002) (Table 21-12). More recently, the outcome of
studies have suggested that it is associated with an aggressive patients in this study followed for a median 10 years was
course in the pre- and postrituximab treatment eras (Bar- reported demonstrating a 10-year PFS for R-CHOP treated
rans et al., 2010; Savage et al., 2009). In some cases, there is patients of 35% (vs 20% with CHOP alone) and 10-year OS
also a t(14;18) involving BCL2, the so-called double-hit of 43.5% (vs 27.6%) (Table 21-12). A similar phase 3 study

was carried out by the US ECOG intergroup (E4494) study CHOP-14 for six or eight cycles, with or without rituximab
comparing six to eight cycles of CHOP versus R-CHOP in in elderly patients and also demonstrated a significant
elderly patients with aggressive lymphoma, which included a improvement in all endpoints with the rituximab combina-
second randomization in CR patients comparing observations (Pfreundschuh et al., 2008). Of note, the latter study
tion and rituximab maintenance therapy every 6 months for also established that six cycles of R-CHOP-14 was associated
2 years. Unlike the GELA study, there was no response rate with the best outcome. Although, six versus eight cycles of
or OS difference detected, although there was a benefit in R-CHOP-21 have not been compared, six should be pre-
TTF for the R-CHOP arm. The analysis was confounded to ferred in most circumstances to minimize toxicities.
some extent by the secondary randomization to mainte- Two randomized studies (GELA LNH-03-6B and the Brit-
nance versus no-maintenance rituximab. Maintenance ther- ish National Cancer Research Institute [NCRI]) compared
apy was beneficial for the TTF only in the CHOP-induction R-CHOP-21 (ie, repeated every 21 days) with R-CHOP-14
subset. As such, interpretation of these results supports the (every 14 days), and there was no improvement of FFS or OS
use of R-CHOP induction without subsequent maintenance using the of the shortened cycle interval, thus confirming
rituximab therapy. that R-CHOP-21 remains the standard (Table 21-12). Based
Two other randomized controlled studies have been pub- upon the observation that elderly females fare better with
lished supporting the benefit of the addition of rituximab to R-CHOP than do elderly males, and that elderly males clear
anthracycline-based chemotherapy in DLBCL. The Mab- rituximab more rapidly, dose-dense rituximab regimens are
Thera International Study Group (MInT) study included being tested in elderly males. A trial where elderly males were
young (<60 years), low-risk (aaIPI 0 or 1) patients with treated with higher dose of rituximab given at 500 mg/m2
DLBCL (including PMBCL) who primarily received either while females received standard dose of 375 mg, showed that
CHOP or CHOP plus etoposide (CHOEP) with or without outcomes of male patients treated with high dose rituximab
rituximab. The rituximab-containing regimens demon- was equivalent to outcomes of females. However, this was
strated an improvement in EFS and OS (Pfreundschuh et al., not a randomized study of different rituximab doses, and a
2006) (Table 21-12). The Rituximab With CHOP Over Age confirmatory phase 3 trial would need to be conducted
60 Years (RICOVER-60) trial by the same group evaluated before the standard practice is changed. This observation,
Table 21-12 Key trials of diffuse large B-cell lymphoma using rituximab-containing regimens
Author (trial/phase) N Treatment Patient selection PFS/EFS OS

Coiffier et al. (GELA/III) 202 R-CHOP × 8 vs × Age 60-80 y 57% vs 38% (2 y) 70% vs 57% (2 y)
197 CHOP × 8 Stage II-IV 35% vs 20% (10 y) 43.5% vs 28% (10 y)
Pfreundschuh et al. 413 R-CHOP-like* × 6 vs Age 18-60 y 79% vs 58% (3 y) 93% vs 84% (3 y)
(MInT/III) 410 CHOP like* × 6 aaIPI 0 or 1 74% vs 56% (6 y) 90% vs 80% (6 y)
Stage I(+bulk or II-IV)
Pfreundschuh et al. 306 R-CHOP-14 × 6 Age 61-80 y 66.5% (3 y) 78% (3 y)
(RiCOVER-60/III)† 304 R-CHOP-14 × 8 Stage I-IV 63% (3 y) 72.5%(3 y)
209 CHOP-14 × 6 47% (3 y) 68% (3 y)
219 CHOP-14 × 8 53% (3 y) 66% (3 y)
Cunningham et al. (ASCO 2011) 540 R-CHOP-21 × 8 Age 61-80 y 81% vs 83% (2 y) 81% vs 83% (2 y)
(NCRI/III) 540 R-CHOP-14 × 6 + G-CSF
Delarue et al. (ASCO 2012) 296 R-CHOP-21 × 8 Age 60-80 y 60% vs 56% (3 y) 72% vs 69% (3 y)
(LNH03-6B/III) 304 R-CHOP-14 × 6 aaIPI >1
Recher et al. (LNH03-2B/III) 196 R-ACVBP Age 18-59 y 87% vs 73% (3 y) 92% vs 89% (3 y)
183 R-CHOP aaIPI 1
Wilson et al. (NCI/II) 72 DA-EPOCH-R Age >18 y 79% (5 y) 80% (5 y)
Stage II-IV
Survival estimates shown for rituximab-containing regimens only and are rounded off where applicable to the nearest whole number.
EFS = event free survival; G-CSF = granulocyte colony-stimulating factor; GELA = Groupe d’Etude des Lymphomes de l’Adulte; PFS =
progression-free survival; MInT = MabThera International Study Group; NCRI = British National Cancer Research Institute Study; OS =
overall survival; R = rituximab; RiCOVER-60 = rituximab with CHOP Over Age 60 Years; y = year.
*87% DLBCL; CHOP-like = CHOP-21 or CHOEP-21 in 92%; radiotherapy given to sites of bulk, extranodal disease (physician’s discretion).
†80% DLBCL.

however, highlights a need for comparing equivalent doses of Novel strategies to improve cure rates in DLBCL
monoclonal antibodies in randomized clinical trials.
Although the outcome of DLBCL has improved with
R-CHOP chemotherapy, ~40% of patients fail after primary
Treatment of limited-stage DLBCL
therapy. The majority of patents with relapsed DLBCL suc-
Approximately 25% of cases of DLBCL are limited stage, cumb to the disease. There are several ongoing approaches
which typically includes patients with stage I disease and to improve outcome in DLBCL, including the use of a new
nonbulky (<10 cm) stage II disease. Some groups or studies combination chemotherapy backbone (eg, ACVBP, DA-
also will include patients with bulky stage I disease and EPOCH), novel enhanced anti-CD20 monoclonal antibody
exclude patients with B-symptoms. A large randomized (eg, obinutuzumab [GA101], ofatumumab), maintenance
SWOG trial (SWOG-8736) established that CMT was supe- therapy, and adding novel targeted agents to R-CHOP (so
rior to CHOP alone for the treatment of localized (stage I(E), called XR-CHOP, where X represents novel agent).
non-bulky stage II(E)) aggressive lymphoma (Miller et al., A phase 3 randomized controlled trial (LNH03-2B) com-
1998). In this study, the 5-year PFS (77% vs 65%, P = 0.03) paring the dose-intensive regimen R-ACVBP (doxorubicin
and OS (82% vs 72%, P = 0.02) for three cycles of CHOP [Adriamycin], cyclophosphamide, vindesine, bleomycin,
followed by IFRT was superior to that of eight cycles of and prednisone) with sequential consolidation (methotrex-
CHOP alone. An update of the study with longer follow-up, ate, rituximab/etoposide/ifosfamide, cytarabine) with stan-
however, showed that the treatment advantage for the CMT dard R-CHOP in patients with DLBCL 18-59 y and only one
was not sustained because of an excess of late relapses, which adverse prognostic factor by the aaIPI (Recher et al., 2011).
was offset by increased toxicity in the chemotherapy-alone The dose-intensive regimen was associated with a more
arm. A stage-adjusted IPI has been proposed for limited- favorable PFS and OS but with higher toxicity.
staged disease that includes stage II disease, age >60 years, PS In North America, dose-adjusted (DA)-EPOCH-R (eto-
>2, and elevated LDH as risk factors. The 5-year OS rates poside, prednisone, vincristine, cyclophosphamide, doxoru-
reported in the updated follow-up for patients with 0, 1 or 2, bicin, rituximab) currently is under evaluation in the
and 3 risk factors were 94%, 71%, and 50%, respectively. management of DLBCL. A phase 3 trial in newly diagnosed
The benefit of rituximab has not been specifically analyzed DLBCL, with biomarker analysis to evaluate outcomes in the
in a randomized controlled trial in localized DLBCL. The ABC and GCB subtypes has been completed and the analysis
MInT study did include some patients with localized disease by of the results is ongoing. Although generally well tolerated,
nature of the inclusion criteria. The SWOG completed a phase DA-EPOCH-R is associated with greater toxicity than
2 study evaluating three cycles of R-CHOP, with four doses of R-CHOP, with hospitalization for febrile neutropenia occur-
rituximab, followed by IFRT (40-46 Gy if CR and 50-55 Gy if ring in 19% of cycles, despite mandatory granulocyte col-
PR) in patients with localized aggressive B-cell lymphoma, the ony-stimulating factor (G-CSF) support.
majority of who had DLBCL. Patients had to have at least one New anti-CD20 antibodies also are being studied in a vari-
risk factor by the stage-modified IPI (Persky et al., 2008). The ety of B-cell lymphomas, including GA-101. In contrast to
study population was similar to the SWOG study described rituximab, obinutuzumab (GA-101) has enhanced anti-
earlier, enabling a historical comparison to determine the body-dependent cell-mediated cytotoxicity (ADCC) and
impact of the addition of rituximab to CMT. The 2-year PFS increased direct cell death induction but low-complement
was superior in the R-CHOP patients (95% vs 83%). dependent cytotoxicity (CDC). A phase 3 study comparing
With potential acute and more concerning long-term side R-CHOP to G-CHOP (GOYA) for the first-line treatment of
effects of radiotherapy, there has been interest in whether a advanced stage DLBCL has been completed and is maturing
subgroup of patients with limited-stage DLBCL can be for analysis.
selected to receive chemotherapy alone. Mature, conclusive Several novel, targeted agents have been shown to have
data is lacking. However, a preliminary report of a random- significant activity in relapsed and refractory DLBCL,
ized clinical trial does suggest that RT can be safely omitted including bortezomib, lenalidomide, and ibrutinib. These
in selected patients (Lamy et al., 2014). Patients with limited agents were combined with R-CHOP (Bor-R-CHOP,
stage, nonbulky (<7 cm) DLBCL were randomized to 4-6 R2-CHOP and IR-CHOP respectively). Due to mechanism
cycles of R-CHOP or 4-6 cycles of R-CHOP followed by of action of these compounds on either BCR signaling or
40Gy RT. Patients in CR by PET imaging after four cycles downstream, these combinations are particularly promising
(84%) did not receive cycles five and six of R-CHOP. The in ABC subtype of DLBCL. All 3 “XR-CHOPs” are currently
patients assigned to no RT had EFS and OS that was not dif- being evaluated in large randomized studies. While some of
ferent compared to patients receiving RT, suggesting RT may the studies allow patients with all subtypes of DLBCL, other
be unnecessary in selected patients. limit eligibility to patients with non-GCB DLBCL defined by

IHC or novel, real time GEP techniques, like Lymph2CX (salvage) combination chemotherapy regimen should be
using Nanostring platform. given such as ICE (ifosfamide, carboplatin, etoposide),
DHAP (dexamethasone, AraC, cisplatin), or GDP (gem-
citabine, dexamethaxone, cisplatin) followed by HDC/ASCT
Primary transplant in advanced-stage DLBCL
if chemotherapy-sensitive disease is demonstrated. The evi-
It has been proposed that patients in IPI poor-risk groups dence supporting the use of HDC/ASCT in relapsed DLBCL
might benefit from a more aggressive treatment approach, is based on the PARMA study. Patients who relapsed with
such as HDC and ASCT or alloSCT. A recent European aggressive lymphoma (excluding CNS or bone marrow
phase 3 randomized trial in aggressive NHL patients tested involvement) following an initial CR to primary therapy,
eight cycles of CHOP versus two cycles of cyclophospha- received two cycles of DHAP salvage chemotherapy. If che-
mide, epirubicin, vindesine, and prednisone followed by mosensitivity (ie, a PR or CR to salvage chemotherapy) was
HDC and ASCT and found higher 5-year EFS for patients in demonstrated, patients were then randomized to receive
the transplantation arm. There was an OS benefit for trans- either further chemotherapy with DHAP or HDC with
plantation in a subgroup analysis confined to the high-inter- BEAC (carmustine, etoposide, cytarabine, and cyclophos-
mediate IPI risk patients. The role of ASCT in frontline phamide) and ASCT. The transplant arm resulted in an
therapy for DLBCL, however, remains unclear in the era of improvement in both the 5-year EFS (46% vs 12%, P =
rituximab. The SWOG group conducted a randomized 0.001) and OS (53% vs 32%, P = 0.038).
phase 3 study investigating the benefit of HDC/ASCT in first The optimal salvage therapy recently has been investigated
remission in patients with advanced-stage diffuse large-cell in two phase 3 randomized controlled trials. The Collabora-
NHL with an aaIPI 2 or 3. This study was initiated in 1997 tive Trial in Relapsed Aggressive Lymphoma (CORAL) study
and did allow patients with a T-cell phenotype although by randomized patients with relapsed DLBCL (or those who
disease frequency, the majority of patients had DLBCL. The had not achieved a CR) to receive R-DHAP or R-ICE for
initial induction regimen was five cycles of CHOP (n = 215); three cycles followed by HDC (with BEAM)/ASCT if a
however, this was amended to R-CHOP (n = 182) in DLBCL response was demonstrated. There was also a second ran-
patients in 2003 to align with the new standard of care. domization following transplant to either rituximab or
Patients in at least a PR after five cycles of R-CHOP were observation to evaluate the role of maintenance therapy
randomized to receive either three further cycles or auto- (Gisselbrecht et al., 2010). At diagnosis, 62% of the patients
transplant using total body irradiation (TBI) or BCNU had been treated with a CHOP-like regimen with rituximab.
(carmustine)-based regimens. In the intention to treat (ITT) The ORR was similar between R-DHAP and R-ICE (63% vs
analysis, the 2-year PFS favored transplant (69% vs 56%, P = 63.5%), and there was no difference in either EFS or OS and
0.005), however, there was no difference in the 2-year OS maintenance rituximab did not affect outcome. Patients
(74% vs 71%, P = 0.32). Of note, 18% of patients in the stan- who previously had received rituximab with their primary
dard arm subsequently underwent treatment with salvage therapy had an inferior response rate (51% vs 83%,
therapy and ASCT at relapse. In exploratory analyses, there P < 0.001) and 3-year EFS (21% vs 47%), suggesting that this
was no differential treatment interaction by phenotype (B- represents a very chemoresistant group. Additional poor
vs T-cell) or induction regimen (CHOP vs R-CHOP). prognostic factors that emerged from this study were early
Patients with high-risk aaIPI, however, had a superior PFS relapse <1 year and an aaIPI of 2 or 3. Interestingly, a subse-
and OS (2-year 82% vs 64%) in the transplant group. Given quent correlative study suggested that patients with GCB
the lack of OS for the group as a whole, primary transplant is DLBCL had an improved outcome to R-DHAP compared
still considered experimental in the primary therapy setting with R-ICE (3-year PFS 52% vs 32%, P = 0.018), which was
even in high-risk subgroups. even more striking if cases were defined by gene expression
profiling (GEP) (3-year PFS 100 % vs 27%), but the
numbers were small. These results suggest selective drug sen-
Management of relapsed and refractory DLBCL
sitivity in molecular DLBCL subtypes but await confirma-
Given the implications of recurrent DLBCL, efforts should tion in other data sets. A second phase 3 trial was conducted
be undertaken to obtain a diagnostic biopsy unless there is by the NCIC (National Cancer Institute of Canada) compar-
unequivocal progression on imaging. In addition, some ing DHAP to the outpatient salvage regimen GDP (gem-
patients may relapse with indolent lymphoma, which would citabine, dexamethasone, cisplatin) in aggressive lymphomas
be managed very differently. Following confirmation of using a noninferiority design. In 2005, the protocol was
recurrence, patients should undergo full restaging investiga- amended for aggressive B-cell lymphomas to include ritux-
tions. If the patient does not have significant comorbidities imab with each salvage regimen. The ORR, EFS, and OS was
and is <70 years of age (<75 in some centers), second-line similar between the treatment arms, but (R)-GDP

was associated with less grade 3 or 4 toxicity (P = 0.0003), who present with a bulky anterior mediastinal disease that
including, febrile neutropenia (9% vs 23%, P < 0.0001) and can be locally invasive into the lung and chest wall occasion-
patients had superior QOL scores. ally with symptoms of superior vena cava syndrome. Distant
There is very little information regarding the effectiveness spread is uncommon at diagnosis. At relapse, involvement of
of salvage therapy in patients with refractory disease, which unusual extranodal sites can occur in PMBCL, including the
typically is defined as either progressive disease (PD) on pri- kidneys, adrenals, ovaries, liver, spleen, and CNS.
mary therapy or relapse within 3 months. In the CORAL Histologically, sclerosis is typically present, and pheno-
study, only 11% of patients had PD on primary therapy. typically, the cells may lack surface immunoglobulin expres-
Limited studies suggest that these cases are rarely responsive sion but express B-cell markers, such as CD19 and CD20.
to salvage therapy and most often are unable to undergone CD30 is present in 80% of cases; however, it is usually weak
HDC/ASCT. Further patients with chemo-resistance to sec- and heterogeneous. Interestingly, recent gene expression
ond-line therapy also have a very poor prognosis, and these analysis has shown that PMBCL is molecularly distinct from
high-risk groups should be considered for investigational typical DLBCL and shares many components of the molecu-
therapies. lar signature with cHL. It had long been speculated that there
may be a pathogenic overlap between the nodular sclerosis
Management of nontransplant-eligible patients with subtype of cHL based on shared clinical features, including a
relapsed or refractory DLBCL, including novel therapies young age of onset and mediastinal predominance, as well as
pathologic features, including predominant fibrosis and
Many patients are not eligible for curative intent treatment tumor cells that are CD30+. In addition, composite and
with salvage chemotherapy and HDC/ASCT due to advanced sequential lymphomas have been reported, and a GZL with
age or comorbidities. Given that the goal of treatment in this overlapping features of both malignancies is now defined in
setting is typically palliative, single-agent chemotherapy the WHO (see the section “B-cell lymphoma, unclassifiable,
often is used because it is less toxic than combination regi- with features intermediate between DLBCL and cHL”), fur-
mens. Use of the salvage regimens outlined previously in this ther highlighting the biological continuum between these
typically older group of patients usually is quite toxic, and it diseases.
is unknown whether these regimens prolong OS. The excep- A novel recurrent translocation involving CIITA (MHC
tion is select cases with late relapses or low-secondary IPI class II transactivator) was found to be recurrent in PMBCL,
risk score, which rarely may be cured with a salvage combi- occurring in 38% of patients and also found in 15% of cHL
nation regimen. Gemcitabine and etoposide have been used (Steidl et al., 2011) (Table 21-2). Cases with these chromo-
in the palliative setting in addition to oral low-dose chemo- somal breaks had an inferior disease-specific survival. Prior
therapy (metronomic) with good tolerance and modest studies also found reduced expression of MHC class II genes,
response rates. Rituximab often is added to chemotherapy, and this also is linked to an inferior outcome.
although there is no firm data showing improved outcome. The outcome of patients with PMBCL is favorable and
Palliative radiotherapy can be effective for localized, symp- was so even in the prerituximab treatment era (5-year OS,
tomatic disease. A number of novel, targeted agents show 70%; Savage et al., 2006), although patients have primary
significant clinical activity in DLBCL. Interestingly, due to refractory disease and very low cure rates. Rituximab appears
targeted nature of these, often the activity appears to be to improve outcomes further. The MiNT study included 87
molecular subtype specific. For example, lenalidomide and patients with PMBCL, which confirmed an improved CR(u)
ibrutinib showed significant activity in relapsed and refrac- rate, a reduction of PD, and an improvement in 3-year EFS
tory DLBCL and appears to be particularly active in ABC (78% vs 52%, P = 0.001) with the addition of rituximab to
(non-GCB) DLBCL. There are many more agents under CHOP-like chemotherapy, but the OS was similar (89% vs
investigations for this patient population and entry into a 78%, P = 0.158) (Rieger et al., 2011). Consolidative involved
clinical trial should be strongly encouraged if the patient’s PS field radiation is typically administered after R-CHOP ther-
is still good. apy. DA-EPOCH-R (without radiation) also has been evalu-
ated in a phase 2 study showing encouraging results with a
Special situations: management of specific median follow-up of 5 years (EFS 93% and OS 97%) (Dun-
clinicopathologic entities of DLBCL leavy et al., 2013). DA-EPOCH-R without radiation and
R-CHOP plus IFRT are considered reasonable standards in
Primary mediastinal (thymic) large B-cell lymphoma
PMBCL. Given that many patients are young females, the
PMBCL was recognized as specific entity in the WHO clas- avoidance of radiation is an attractive notion in PMBCL.
sification based on unique clinicopathologic presentation. Studies are ongoing investigating whether PET scanning
Patients are typically females with a median age of 35 years can be used to select patients who may benefit from

consolidative radiotherapy and conversely spared if they are including CNS. Due to inadequacy of R-CHOP, various
in a CR. The role of RT in this setting is unclear and the intensified chemoimmunotherapy strategies have been pro-
International Extranodal Lymphoma Study Group (IELSG) posed, largely based on experience in Burkitt lymphoma;
has initiated a randomized trial (IELSG37), whereby patients however, advanced age of patients and often poor perfor-
who are PET negative after a rituximab-containing regimen mance status limits the use of intensive chemotherapy. Due
are randomized to observation versus radiotherapy. to its rarity, data largely come from retrospective reviews,
and as such comparison between regimens is difficult. The
B-cell lymphoma, unclassifiable, with features intermediate intensified including HyperCVAD and CODOXM/IVAC
between DLBCL and cHL appear to compare favorably with R-CHOP historical con-
trols; however, one must bear in mind that patients who are
In the updated WHO 2008, a new category was defined that candidates for such intensive therapy are frequently younger
shows overlapping clinical, morphological, or immunophe- and have better PS, as such results may not be generalizable
notypic features between cHL and DLBCL, particularly to majority of patients with DHL. A study presented at ASH
PMBCL. These cases of so-called GZL usually occur in young 2014 by Dunleavy et al utilizing R-EPOCH in a multicenter
men between 20 and 40 years old who present with an ante- phase 2 study which included DHL demonstrated a promis-
rior mediastinal mass and may have supraclavicular lymph ing early PFS (87% at 14 months). R-EPOCH is usually bet-
node involvement. A broad spectrum of cytological appear- ter tolerated than other intensified regimens, therefore might
ance can occur within the same tumor. The immunopheno- be offer an option for patients who are not a candidates for
type often is transitional between PMBCL and cHL (see
intensive chemoimmunotherapy or ASCT. Novel agents,
Chapter 20) with the tumor cells CD45+, CD20+, CD30+,
particularly those targeting BCL2, may offer a promise in
and CD15+. Cases of morphologically nodular sclerosis cHL
DHL and are under investigation.
with strong and uniform expression of CD20 and CD15-
would favor a diagnosis of GZL. In contrast, cases resem-
bling PMBCL but that are CD20- and CD15+ or EBV+, also Key points
would support GZL. Given disease rarity, there is little infor- • Diffuse large B-cell lymphoma (DLBCL) is the most common
mation regarding clinical outcome or optimal therapy. Small histologic subtype of NHL.
series suggest these are best managed like other aggressive • The IPI and cell of origin phenotype remain prognostic in the
NHLs rather than like HL. rituximab treatment era in DLBCL. Studies are ongoing whether
patients classified as high risk by the IPI or ABC phenotype
T-cell-rich DLBCL should be treated with an alternate therapy other than R-CHOP.
• Treatment with R-CHOP-21 (ie, repeated every 21 days) for six
T-cell rich DLBCL is an uncommon variant of DLBCL. Typi- cycles is a standard of care in advanced disease; the role of
cally, the neoplastic cells comprise <10% of cellular popula- consolidative radiation in advanced disease is not well defined.
tion and are outnumbered by a background of abundant • In limited stage disease, abbreviated chemotherapy with 3-4
T-cells and histiocytes. Histologically, it can resemble nodu- cycles of R-CHOP plus involved-field radiotherapy (IFRT) can be
lar lymphocyte predominant HL. Typical presentation used.
• Presence of relapsed disease should be documented by biopsy
includes middle-aged men often with an advanced disease at
whenever possible.
diagnosis and liver and spleen involvement. The treatment
• Transplant eligible patients with relapsed DLBCL are usually
with R-CHOP results in similar results to these seen in other
treated with salvage chemotherapy (RDHAP, RICE, and RGDP
subtypes of DLBCL. appear to have similar efficacy) followed by high dose chemo-
therapy and stem cell transplantation.
“Double-hit” DLBCL (DHL) • PMLBCL patients are treated with DAEPOCH-R without RT or
R-CHOP plus IFRT. There are no randomized studies directly
Five to 10% of DLBCL patients have “double-hit” lymphoma comparing these approaches.
(DHL), defined as presence of c-myc and either bcl2 or bcl6 • Double-hit lymphoma (DHL) represents particularly poor
translocations (detected by FISH). These high-risk patients prognostic category when treated with R-CHOP; the role of
have lower OS when treated with R-CHOP. As such, more intensive regimes remains to be established.
R-CHOP is considered an inadequate therapy for DHL, with
majority of patients dying within 2 years of the diagnosis
Primary CNS lymphoma
(Petrich et al., Blood 2014).
The majority of patients present with poor prognostic PCNSL can occur in the brain parenchyma, spinal cord, eye
features, including an advanced age, elevated LDH, and (ocular) (Figure 21-3), cranial nerves, or meninges. Of note,
an advanced stage often with extranodal involvement, although 95% of cases of PCNSL are DLBCLs, rare cases of

within the deep regions of the brain. Patients with 0, 1 to 4,

or 5 of these factors have a 2-year OS rate of 80%, 48%, or
15%, respectively.
The median survival after surgery alone is ~1-4 months.
Whole-brain radiation is associated with a high response
rate of 90%, but the median survival is only 12 months.
CHOP has poor CNS penetration and should not be used in
PCNSL. The exception is intravascular large B-cell lym-
phoma with CNS involvement as the mechanism of spread is
likely different. Although there have been no randomized
controlled studies to establish the best therapy, in retrospec-
tive analyses, outcomes are superior when high-dose metho-
trexate (HD-MTX) (3-8 g/m2) is incorporated into first-line
Figure 21-3 Intraocular large B-cell lymphoma on slit lamp regimens. With this approach, the 5-year OS is approximately
examination. 30%-40%. Some studies have added other CNS-penetrant
chemotherapy drugs, such as cytarabine (ara-C). Rituximab
PTCL, low-grade lymphoma, and BL also have been reported. has poor penetration across the blood–brain
In addition to B-cell markers, CD10 expression is observed barrier but is currently being tested in clinical trials given
in only 10%-20%, but BCL6 expression is common (60%- synergy with chemotherapy. A phase 3 trial randomizing
80%). PCNSLs are rare and may occur in immunocompe- younger patients in a CR following HD-MTX to either
tent patients or in association with immunosuppression WBRT (45 Gy) versus observation demonstrated an
related to HIV infection or organ and marrow transplanta- improvement in median PFS (18 months vs 12 months) but
tion. With the introduction of highly active retroviral ther- OS was similar (reviewed in Ferreri, 2012). For older patients
apy (HAART), the incidence of PCNSL has decreased in >60 years, the risk of neurotoxicity is considerable and mani-
HIV-infected persons. It appears, however, to be increasing fests as dementia, ataxia, and incontinence, with a median
in incidence in immunocompetent patients. In the latter time to onset of approximately 1 year. With concerns of neu-
group, the median age is 60 years, and it is discovered based rotoxicity, even in younger patients, numerous studies are
on focal neurologic symptoms, personality changes, or evaluating chemotherapy alone with CNS-penetrant drugs.
symptoms of increased intracranial pressure. Ocular involve- The CALGB evaluated the combination of HD-MTX, temo-
ment can occur in 10%-20% of patients and may be the sole zolomide, and rituximab with consolidative HDC using
site of disease at presentation (intraocular lymphoma). Con- ara-C and etoposide without WBRT and the 3-year PFS and
current leptomeningeal disease is found in 16% through CSF OS were 50% and 67%, respectively. Studies are also investi-
analysis but occurs as the sole site in <5%. B symptoms are gating lower doses of radiation in patients in a CR after che-
extremely uncommon and should raise suspicions of sys- motherapy. If chemotherapy is contraindicated because of
temic involvement (reviewed by Ferreri, 2012). age or comorbidities, WBRT (40-50 Gy) is recommended.
Stereotactic-guided biopsy is the optimal method to diag- HDC/ASCT has been evaluated in the upfront and salvage
nosis CNS lymphoma, and gross total resection should be setting. Several small phase 2 studies have evaluated upfront
avoided. Steroids can interfere with pathologic diagnosis, transplant with cure rates ranging from 40% to 77% using a
and if they are started for neurologic symptoms, they should variety of lead-in chemotherapy and HDC regimens. There
be withheld in patients with a presumptive radiologic diag- are two ongoing randomized trials comparing WBRT and
nosis of CNS lymphoma to increase diagnostic biopsy yield. HDC/ASCT. In patients with relapsed or refractory primary
A contrast-enhanced MRI should be performed, along with CNS, HDC/ASCT is associated with a 2-year OS of 45%, a
lumbar puncture with CSF analysis. A slit-lamp examination TRM of 16%, and severe neurotoxicity in 12%. Currently, it
should be performed to rule out concurrent ocular involve- should be considered experimental but may be appropriate
ment. Staging should include CT imaging, bone marrow in select young patients in the relapsed setting. Temozolo-
aspirate and biopsy, and, in men, testicular ultrasound, as mide either alone or in combination with rituximab has
4%-12% of patients can have extraneural disease. shown an ORR of 26% and 53%, respectively, in relapsed
A prognostic scoring system has been developed in PCNSL and refractory patients. The combination of high-dose
given the limitations of the Ann Arbor staging system and methotrexate, rituximab, and temozolomide (MRT) is well
the IPI in this disease. The following five factors are associ- tolerated and associated with significant activity in a recently
ated with a poor prognosis: age >60; PS >2; elevated LDH; reported small phase 2 study. CR was achieved in 14/18
high CSF fluid protein concentration; and tumor location (78%) patients at a median of 4 months. Three of 18 patients

achieved partial response (PR). At a median follow-up of orbit), and reported a low rate of CNS relapse (3%)
15.5 months from treatment initiation, 10/18 patients (Abramson et al., 2010). Use of HD-MTX, however, is lim-
remain in CR and median PFS has not been reached. ited in elderly patients, particularly with poor renal function.
No evidence suggests that intrathecal chemotherapy Similar strategy of systemic methotrexate prophylaxis is cur-
improves outcome if HD-MTX is being used; however, the rently under evaluation in treatment of primary testicular
CSF should be reanalyzed on treatment to ensure clearance DLBCL, a subset of DLBCL associated with particularly high
of the malignant cells. risk of CNS relapse in the study conducted by IELSG. Of
note, patients with testicular lymphoma, in addition to CNS,
have a high risk of relapse in contralateral testicle, and pro-
Secondary CNS lymphoma
phylactic radiation therapy to remaining testicle is usually
The rate of secondary involvement of CNS in aggressive lym- recommended.
phoma varies by histology, occurring in up to 30% of BL (see Despite the limitations and lack of evidence-based data to
section “Burkitt lymphoma” in this chapter) and lympho- direct treatment, patients considered high risk either by the
blastic lymphoma. In these highly aggressive lymphomas, extranodal site involved or by the Hollender risk model
CNS prophylaxis is routinely incorporated using intrathecal should be evaluated for occult CSF involvement using cytol-
(IT) and systemic chemotherapy with or without cranial ogy. Flow cytometry also has been shown to be a more sensi-
irradiation and has been shown to reduce the rate of CNS tive tool for the detection of CNS involvement and should be
relapse and prolong survival. Secondary CNS lymphoma employed where possible to rule out CNS disease at the time
also is seen in DLBCL occurring in the brain parenchyma, of diagnosis. Those with positive findings should undergo
leptomeningeal compartment, or both, as an isolated event, further staging with a MRI and be treated aggressively for
or with systemic relapse. The overall risk of CNS relapse and CNS disease. Cases negative by CSF can be considered for
progression in DLBCL is only ~5% but can be up to 25%- prophylactic strategies and where possible evaluated in a
30% in specific high-risk subgroups. A number of extrano- prospective clinical trial. A management algorithm has been
dal sites have been associated with a higher risk of CNS proposed in a recent comprehensive review (Siegal and
relapse, including testis, breast, kidney, and bone marrow Goldschmidt, 2012).
(concordant).
Although these and other studies can effectively identify
Burkitt lymphoma
subgroups with a high risk for CNS disease, demonstrating a
benefit for CNS prophylaxis has proven to be much more BL is among the most aggressive of all human malignancies,
difficult in DLBCL. Furthermore, many of the studies evalu- with a rapid doubling time, acute onset, and progression of
ating CNS prophylaxis were published before the routine use symptoms. Histologically, BL has a diffuse growth pattern of
of rituximab, which does appear to reduce risk, albeit to a medium-size cells and a high mitotic rate, as depicted by
modest degree. The RiCOVER-60 study evaluated 1,217 nearly 100% of cells being Ki-67 positive due to deregulated
patients with aggressive lymphoma (81% DLBCL) and high-level expression of cMYC arising from reciprocal trans-
reported that 58 patients (4.8%) developed CNS relapse or location with immunoglobulin-heavy (t8;14) or variant
progression with a median time of 8 months (1-39 months) light-chain gene loci (t2;8 or t8;22) (Table 21-2). There is
with a median survival from CNS relapse of only 3 months. also a high rate of cell death or apoptosis, and the dead cells
Those patients who received rituximab had a lower risk of are phagocytosed by histiocytes, which gives a “starry-sky”
CNS relapse; however, the magnitude of difference was very appearance at low power. The B-cells are positive for CD19,
small (3.6% vs 5.9%, P = 0.043). Other studies have con- CD20, BCL6, and CD10. BCL2 is usually negative, but rare
firmed that rituximab appears to reduce the risk of relapse, weakly positive cases may be seen. Lack of TdT is critical to
particularly in patients in a CR, suggesting the benefit in part rule out ALL/lymphoma. Gene expression profiling studies
may be due to better systemic disease control (Villa et al., show that BL has a distinct molecular signature distinguish-
2009). The risk is not altogether eliminated, however, given ing it from DLBCL.
the poor CNS penetration of rituximab. Modeled after BL Originally described in its endemic form in African chil-
and lymphoblastic lymphoma, intrathecal CNS prophylaxis dren presenting with jaw or facial masses, BL also occurs in
often is administered to high-risk DLBCL patients, but the sporadic form in the Western world, predominantly in chil-
protective benefit is unknown. Prophylactic use of HD-MTX dren and young adults. It also is seen in HIV-infected
(3.0-3.5 g/m2) with R-CHOP was evaluated retrospectively patients. Most endemic and some sporadic cases show evi-
in 65 patients with high-risk DLBCL (elevated LDH, involve- dence of EBV infection and presence of the EBV genome.
ment of >1 extranodal sites, 4-5 Hollender criteria, high-risk Clinically, patients with BL frequently present with a bulky
location: bone marrow, testes, epidural, liver, adrenal, renal, abdominal mass, B-symptoms, and extranodal disease,

including bone marrow involvement, is common (up to with a modified Magrath regimen. Other therapeutic
70%). A leukemic phase can be seen, but pure acute leukemia approaches have included the hyperfractionated cyclophos-
is extremely rare. CNS dissemination, usually in the form of phamide, vincristine, doxorubicin, and dexamethasone
leptomeningeal involvement, may be present at diagnosis in (HyperCVAD)/methotrexate-cytarabine regimen and ALL-
up to 40% of patients; as a result, HD-MTX and intrathecal type regimens. It also is unknown whether consolidative
chemoprophylaxis are integrated into the therapy for all BL ASCT improves outcome in BL.
patients. Given the limited data in older patients, the results from
Therapy for BL must be instituted quickly because of the 12 large treatment series (10 prospective and 2 retrospective)
rapid clinical progression of the disease. Admission to hos- were combined to better determine outcome in patients with
pital and tumor lysis precautions are essential and include BL in patients >40 years of age (Friedberg et al., 2005). In
vigorous hydration and allopurinol with close monitoring of total, 470 patients were identified, 183 of whom were >40
laboratory studies, including electrolytes and renal function. years. The median OS at 2-years with intensive short-duration
Early dialysis is indicated at the first signs of decreasing renal chemotherapy in older patients was only 39% compared
function, hyperkalemia, or hyperphosphatemia. Recently, with 71% when all patients were considered, suggesting an
recombinant uric acid oxidase (rasburicase) has been shown unmet need in older BL patients. The DA-EPOCH-R regi-
to be very effective in preventing uric acid nephropathy and men was evaluated in 19 immunocompetent patients with
its secondary metabolic complications. Multiple studies BL and demonstrated an EFS rate of 95% (Dunleavy et al.,
have shown that CHOP chemotherapy is inadequate for the 2013). This may be the preferred approach to older BL and is
treatment of BL, and intensified therapies result in higher the subject on ongoing prospective trials. Several studies
cure rates. Multiagent combination chemotherapy that have evaluated the impact of rituximab to intensive therapy
includes high doses of alkylating agents and CNS prophy- and suggest a beneficial effect.
laxis have improved the outcome for adults and children
with the disease. Given disease rarity, there are no random-
B-cell lymphoma, unclassifiable, with features
ized controlled treatment trials in adults comparing these
between DLBCL and BL
approaches. Magrath et al. (1996) at the National Cancer
Institute demonstrated a risk-adapted strategy that is useful B-cell lymphomas with features intermediate between
for treatment stratification in both adults and children. DLBCL and BL have morphologic and genetic features of
Low-risk patients were those with a single extra-abdominal both DLBCL and BL and typically a very aggressive course.
mass or completely resected abdominal disease and a nor- Morphologically, it appears intermediate between DLBCL
mal LDH, and all other patients were considered high and BL with cells usually smaller than typical DLBCL but
risk. Low-risk patients received three cycles of cyclophos- larger than typical BL with a high proliferation rate, starry-
phamide, vincristine, doxorubicin, and methotrexate sky appearance, and an immunophentype consistent with
(CODOX-M) only, and high-risk patients received CODOX- BL. Other cases may be morphologically similar to BL but
M alternating with ifosfamide, etoposide, and cytarabine have an atypical immunophenotype or genetics. So-called
(IVAC) for a total of four cycles (ie, two cycles each of double-hit lymphomas with dual translocation of cMYC and
CODOX-M and IVAC). All patients received intrathecal BCL2 are included in this category unless they are morpho-
chemoprophylaxis with each cycle, and those with CNS dis- logically identical to DLBCL. Unlike BL, the MYC partner
ease at presentation received additional intrathecal therapy chromosome is often non-Ig. Overall, the prognosis is poor,
during the first two cycles. Approximately half of the patients and for the double-hit lymphomas, the risk of secondary
were adults, and the 2-year EFS for all patients was 92%. CNS involvement is high and they rarely are cured with standard
Two other phase 2 studies have used the Magrath regimen therapy. Dose-intensive therapies, including CNS-penetrant
with modifications. In a United Kingdom study, adult (age drugs, are under investigation; please see discussion in the
range, 16-60 years; median, 26.5 years), non-HIV patients section “‘Double-hit’ DLBCL (DHL)” above.
were treated with dose-modified CODOX-M (3 g/m2) for
three cycles if determined to be low risk (ie, normal LDH, PS
Immunodeficiency-associated
of 0 or 1, Ann Arbor stage I or II, and no tumor mass >10 cm),
lymphoproliferative disorders
and all other patients were considered high risk and treated
with alternating dose-modified CODOX-M/IVAC. Congenital or acquired immunodeficiency states are asso
The 2-year PFS for the patients with BL was 64%. At the ciated with an increased incidence of lymphoproliferative
Dana-Farber Cancer Institute, an older population (median disorders. The WHO classification identifies four such cate-
age, 47 years) of patients was treated with a modified gories: (i) primary immunodeficiency disorders, including
Magrath regimen, and the reported 2-year EFS was 71% Wiskott-Aldrich syndrome, ataxia-telangiectasia, common

variable or severe combined immunodeficiency, X-linked increase in life-threatening infections, which also may
lymphoproliferative disorder, Nijmegen breakage syndrome, reflect the exclusion of poor-prognosis patients because
hyper-IgM syndrome, and autoimmune lymphoproliferative patients could have no more than one of the following: CD4
syndrome; (ii) HIV infection; (iii) post–solid organ or mar- <100, PS >2, or prior AIDS. Thus, rituximab should be given
row transplantation with iatrogenic immunosuppression; to HIV patients if the CD4 count is >50, particularly given
and (iv) methotrexate- or other iatrogenic-related immuno- the strong evidence in the HIV-negative setting. Concurrent
suppression for autoimmune disease. The lymphomas seen administration of G-CSF is advised given the high rate of
in these settings are heterogeneous and may include HL or, infection in this population. DA-EPOCH has been tested in
more commonly, aggressive NHL. Chédiak-Higashi syn- HIV-aggressive lymphoma, the majority of which had
drome also has been associated with an increased incidence DLBCL but with suspension of HAART to avoid any drug
of pseudolymphoma and true NHL. interactions. At 53 months, the PFS and OS were 60% and
Lymphoproliferative disorders associated with primary 73%, respectively. The AMC also tested EPOCH-R (AMC
immune deficiencies (PIDs) most commonly are seen in 034) in patients with HIV-positive aggressive B-cell lym-
pediatric patients and frequently are associated with EBV phomas with rituximab given either concurrently or sequen-
infection. Extranodal disease including the CNS is common. tially, and the 2-year OS rate was 63% and 66%, respectively.
Lymphomas occurring in patients with PID do not differ HAART use was at the discretion of the treating physician
morphologically compared with immunocompetent hosts. but was used in the majority of patients. There was no
DLBCL is the most frequent histologic type, although T-cell greater risk of infection except in patients with a CD4 <50
lymphomas are more common in ataxia-telangiectasia. (Sparano et al., 2010). More recently, the NCI piloted a sec-
EBV-related lymphomatoid granulomatosis is associated ond-generation regimen short course SC-EPOCH-RR
with Wiskott-Aldrich syndrome. These malignancies (dose-dense rituximab), with G-CSF support, in HIV-
respond poorly to standard therapy. Therapy depends on positive DLBCL patients in the hope of improving efficacy
both the underlying disorder and the specific lymphoma but reducing toxicity. Dose-dense rituximab was intended
subtype; allogeneic transplantation has been used success- to enhance the chemotherapy and minimize the number of
fully in some patients. Novel immunotherapeutic or phar- treatment cycles. HAART was suspended during treatment.
macologic strategies targeting EBV are being explored. A PET scan was performed after two cycles: if negative, only
one further cycle was given; and if positive, two to three
cycles were given. The 5-year PFS and OS were 84% and
HIV-associated lymphomas
64%, respectively. A pooled analysis of these two AMC trials
HIV-associated lymphomas are typically DLBCL or BL. with patients either treated with R-CHOP or R-EPOCH
Approximately two-thirds of cases are EBV associated, and suggested that patients receiving R-EPOCH had an
many carry a cMYC oncogene translocation. CNS involve- improved EFS and OS after adjusting for the aaIPI and CD4
ment is frequent. Outcomes for HIV associated lymphomas count. The TRM was greater in patients with CD4 counts
were historically poor. However, the era of highly active <50 (37% vs 6%, P = 0.01) regardless of the regimen used. It
antiretroviral therapy (HAART), outcomes are similar to remains unclear whether HAART is mandatory during che-
non-HIV lymphoma, as long as the HIV is under good con- motherapy; however, with the possible exception of the SC-
trol and the CD4 count is over 200 cells/uL. Given the EPOCH-RR, studies support continuing HAART in patients
importance of optimal HIV control, HAART is usually given treated with a variety of regimens, including Hyper-CVAD,
concurrently with chemotherapy and in communication CDE (infusional cyclophosphamide, doxorubicin, etopo-
with the HIV specialist to avoid antiretrovirals that can exac- side), EPOCH-R, and R-CHOP, particularly because newer
erbate chemotherapy toxicity. antiretrovirals have fewer drug interactions than in the past.
The optimal chemotherapy and the role of rituximab Use of zidovudine is avoided because of increased risk of
with anthracycline combinations in HIV-associated DLBCL myelosuppression and the potential for deleterious drug
have been the subject of debate. One small randomized interactions.
study conducted by the AIDS Malignancy Consortium Management of BL has historically been extremely chal-
(AMC 010) demonstrated no improvement in outcome lenging in the HIV population due the comorbidities com-
comparing R-CHOP with CHOP and an increase in treat- bined with the required intensity of lymphoma therapy.
ment-related infectious deaths. A subsequent analysis, how- AMC 048 reported encouraging results using a modified
ever, indicated that the toxicity was higher in patients with a CODOX-M/IVAC for HIV-positive BL. The major modifi-
CD4 count <50. Furthermore, a phase 2 French study using cation was a reduction in the methotrexate dose to 3 gm/m2.
R-CHOP in HIV-positive aggressive lymphomas (85% Use of HAART therapy was optional. The 2-year OS was
DLBCL) demonstrated a 2-year OS of 75% without an 69% and tolerability was acceptable.

Posttransplant lymphoproliferative disorders patients failing a reduction of immunosuppression were

given four weekly cycles of rituximab followed by CHOP as
Posttransplant lymphoproliferative disorders (PTLDs) occur
sequential treatment (ST). In 70 patients, the ORR was 60%
as a consequence of immunosuppression in recipients of
following rituximab, which increased to 90% with ST with
solid organ, bone marrow, or stem cell allograft. The risk is
CHOP. With a median follow-up of 5 years, the median TTP
higher in solid organ transplants that warrant a higher degree
and median PFS were 77 months and 48 months, respec-
of immunosuppression (10%-25% in heart and lung trans-
tively. The 5-year PFS was 66% and 5-year OS was 57%
plant) than those that require a lower immune suppression
(Trappe et al., 2011). The TRM with CHOP was 11%.
dosing (1%-5% kidney and transplant). PTLDs are com-
Reduced immunosuppression and single-agent rituximab
posed of a spectrum of disorders, ranging from EBV-positive
are reasonable first-line treatments in the majority of patients
infectious mononucleosis (early lesions) to polymorphic
with close surveillance and sequential therapy with R-CHOP
PTLDs, which most often are clonal to full-blown mono-
in those who do not achieve a CR. For patients who present
morphic PTLDs that can be either EBV positive (common)
with very high-risk aggressive disease, R-CHOP can be con-
or EBV negative and are further subdivided into B-cell lym-
sidered frontline treatment with G-CSF support with strong
phomas (common) and T-cell lymphomas (rare), and are
consideration also for PCP prophylaxis.
indistinguishable from their counterparts in immunocom-
petent hosts. HL-type PTLDs also can occur; however, indo-
Mantle cell lymphoma
lent B-cell lymphomas arising in transplantation recipients
are not among the PTLDs. EBV-negative PTLD has increased In many ways, MCL falls between the indolent and aggres-
over the last decade and typically is late onset (median time sive lymphomas, unfortunately combining the poorer attri-
from transplant to PTLD of 50-60 months vs 12 months in butes of each; namely, the lack of curability with standard
EBV positive), has a poorer response to therapy, and is more therapy and a relatively aggressive clinical course. With bet-
frequently monomorphic. ter recognition of MCL as a unique entity, and treatment
PTLDs have diverse clinical presentation depending on strategies developed specifically for MCL, the median OS of
location. Extranodal involvement is common, particularly MCL appears to be improving, now longer than 5 years
the gastrointestinal (GI) tract (~25%), lung, skin, and bone (Hermann et al., 2009).
marrow. Primary CNS lymphoma also can occur. The goal MCL has distinctive clinical features include median age
of treatment is to cure the lymphoma but also to preserve of 64, a striking male predominance, and a strong tendency
graft function. Although a minority (20-50%) of patients to present with advanced stage disease. Extranodal involve-
will respond to a reduction in intensity of immunosuppres- ment is common, including bone marrow and peripheral
sive drugs, most require additional systemic therapy par blood, plus a peculiar tendency to invade the GI tract, which
ticularly for monomorphic or late PTLDs. Tolerance to may present as a distinctive syndrome of lymphomatous
chemotherapy is poor in PTCL patients, with TRM reported polyposis of the large bowel. Even patients without overt
to be as high as 31% in older series using CHOP chemo- colonic polyposis frequently have subclinical GI epithelial
therapy. With historically poor tolerance to combination invasion, which can be demonstrated on biopsy.
chemotherapy, single-agent rituximab has been explored in Cytologically, the majority of MCLs consist of small lym-
the first-line setting in PTLD. The ORR has ranged from phocytes with notched nuclei. The architectural pattern of
40% to 75%, and it is extremely well tolerated; however, the lymph node usually is diffuse but may show a vaguely
remission duration may be short in many patients. In the nodular or mantle zone growth pattern. A spectrum of mor-
first prospective phase 2 study, 43 PTLD patients who had phologic variants has been recognized, including small cell,
failed to respond to a reduction in immunosuppression were which is composed of small round lymphocytes and clumped
treated with single-agent rituximab. The ORR was 44% at chromatin, mimicking SLL/CLL, and a blastoid variant,
day 80 (CR 21%) and the 1-year OS was 67%. An updated which has a high mitotic rate and is clinically very aggressive.
analysis from this study evaluating 60 patients demonstrated The immunophenotype of MCL is distinctive. Cases are typ-
an ORR of 59% (CR 42%), but the median PFS was only 6 ically CD5+, FMC7+, and CD43+ but CD10− and CD23–
months and the 2-year OS was 52% (Choquet et al., 2007). (Table 21-2). Some of the salient features that distinguish
Elevated LDH was predictive of disease progression as well as MCL from SLL or CLL are the expression of cyclin D1 and
a shorter time from the date of transplant. Using a PTLD- FMC7 and the lack of CD23 expression (Table 21-2). Fur-
adapted prognostic score incorporating age (>60 years), ele- thermore, MCL has a more intense IgM or IgD and CD20
vated LDH, and PS (>2), patients with a score of 0, 1, or 2/3 expression than SLL/CLL. Virtually all MCLs carry the
had 2-year OS estimates of 88%, 50%, and 0%, respectively, t(11;14)(q13;q32) on karyotypic analysis or by FISH tech-
suggesting that single-agent rituximab may be suboptimal in nique. This reciprocal translocation juxtaposes the immuno-
high-risk groups. This prompted a study in which PTLD globulin heavy-chain locus and the cyclin D1 (BCL-1) gene.

Biologic and clinical features have prognostic value in have not fared as well, with a median PFS of ~3 years and
MCL. Cellular proliferation may be the most powerful pre- substantially more toxicity. Another intensive approach that
dictor. cDNA microarray analysis has demonstrated that generated highly promising results comes from the Nordic
genes associated with cellular proliferation show striking Lymphoma Study Group. They tested an intensive induction
variability among MCL cases, ranging from low to very high immunochemotherapy with alternating cycles of “maxi”
expression. Patients in the lowest quartile of expression have R-CHOP and rituximab plus cytarabine followed by in vivo
median survival times of 6-8 years, whereas patients in the purge (with rituximab) and ASCT in a phase 2 trial. The
highest expression quartile have survivals of <1 year. For study was limited to patients age <65 years and the median
clinical practice, Ki-67 staining can provide an estimate of age was 56 years. The ORR was 96%, and the 6-year EFS and
proliferation. Three prognostic groups have been identified OS were 56% and 70%, respectively. Long-term follow-up
using cut-points of <10% (best), 10%-29% (intermediate), demonstrated a continuing pattern of relapse, suggesting
and >30% (worst). With regards to clinical factors, the IPI cures are not likely, even with this approach. The European
does not provide adequate prognostic usefulness when MCL Network has presented results of large phase 3 random-
applied to MCL, leading to the generation of an MCL-specific ized clinical trial in MCL patients <65 years (Hermine et al.,
index (Hoster et al., 2008). The MCL international prognos- 2010). This trial compared the efficacy of six courses of
tic index (MIPI) identified four clinical features: age, PS, R-CHOP followed by myeloablative radiochemotherapy and
LDH, and WBC as independently associated with OS (Table ASCT versus alternating courses of R-CHOP/R-DHAP fol-
21-7). The MIPI score can separate patients into three risk lowed by a high-dose cytarabine containing myeloablative
groups and is quite valuable for characterizing patients on a regimen and ASCT. The study was designed to test the con-
clinical trial. It is not always useful in clinical practice, as tribution of cytarabine in the management of younger MCL
older age and poor PS may classify a patient as “high risk,” patients (median age 56 years). The 3-year PFS was signifi-
but such a patient may not be a candidate for therapy cantly better in the cytarabine-containing arm (75% vs 60%)
intensification. and most intensive strategies now incorporate high dose
cytarabine into the regimen.
Until recently, there were no trials focusing on the ~50%
Management of newly diagnosed MCL
of MCL patients who are not candidates for an intensive
There is no “standard” therapy or approach to MCL. It is a therapy approach. The European MCL Network conducted
relatively uncommon lymphoma subtype (6% of new cases), a trial for patients >60 years, who were assigned randomly to
making comparative trials difficult to conduct. A small num- induction with either R-CHOP or the R-FC (rituximab,
ber of cases have a course similar to the indolent lymphomas fludarabine, cyclophosphamide) regimen (Kluin-Neilmans
and a period of observation is reasonable (Martin et al., et al., 2012). Responding patients underwent a second ran-
2009). Most patients have symptomatic disease and require domization to maintenance therapy with rituximab (MR) or
treatment. There are a variety of phase 2 studies in the litera- interferon-α (IFNα), each given until progression. The
ture, and only recently were the first randomized clinical tri- median age of the 560 study participants was 70 years.
als reported by the European MCL Network. This group has Although response rates were similar between R-CHOP
adopted a strategy of separating patients by age and design- (86%) and R-FC (79%), the OS was significantly better in
ing trials using intensive treatment strategies for younger the R-CHOP arm (62% vs 47% at 4 years, P = 0.005). The
patients (defined as age 65 or less) and nonintensive strate- inferior survival in the R-FC group was due to a combina-
gies for older patients (defined as age 60 or more). The 5-year tion of inferior disease control and increased death from
overlap is intentional to allow patients between the ages of infectious complications related to the immunosuppressive
60 and 65 to be candidates for either approach, depending effects of fludarabine. Remission duration was significantly
on comorbidities. This strategy is a useful one for clinical longer in the rituximab group than in the IFN group. At 4
practice. years, 58% of the MR group remained in remission com-
For the younger patient with MCL, several different inten- pared with 29% of the IFN group. Subgroup analysis indi-
sive strategies appear to produce comparable results. The first cated the benefit of MR was restricted to the R-CHOP-treated
intensive strategy to gain widespread application was the patients, and that R-CHOP plus MR-treated patients experi-
R-HyperCVAD with alternating R-MTX/cytarabine, pio- enced improved 4-year OS compared with R-CHOP plus
neered by investigators at the M.D. Anderson Cancer Center. IFN–treated patients (87 vs 63%, P = 0.005), respectively.
This single-institution study enrolled 99 patients with a This trial indicates that R-CHOP followed by MR is a rea-
median age of 61 years. The approach produced response sonable front-line approach for older MCL patients,
rates >95% and long-term follow-up revealed a 5-year PFS of although emerging data suggests R-CHOP is not the best
~50% (Romaguera et al., 2010). Older patients on this trial induction platform for older MCL patients.

A phase 3 trial compared R-CHOP to an R-CHOP-like ORR of 57% and median PFS of 11.1 months (Wang et al.,
regimen (VR-CAP) where bortezomib has replaced vincris- 2012). Most promising of the new agents is ibrutinib, a bru-
tine (Robak et al., 2015). The VR-CAP regimen was superior ton’s tyrosine kinase inhibitor which interferes with signal-
to R-CHOP for complete response rates (53% vs 42%), ing through the B-cell receptor pathway. In a single arm
median PFS (24.7 months vs 14.4 months), and 4-year OS phase 2 trial (n = 111) in relapsed/refractory MCL, the ORR
rate (64% vs 54%). The rates of neutropenia and thrombo- was 68% with a median PFS of 13.9 months (Wang et al.,
cytopenia were higher in the VR-CAP patients. The benda- 2013). Ibrutinib was FDA approved for patients with recur-
mustine-rituximab (BR) regimen also appears to be a rent MCL in late 2013.
preferred alternative to R-CHOP. A large randomized trial
compared BR with R-CHOP in patients with newly diag- Peripheral T-cell lymphomas
nosed indolent and MCL lymphoma (Rummel et al., 2013).
For the entire study population, the BR was better tolerated PTCLs represent 10%-15% of all NHLs in Western popula-
than the R-CHOP, with less alopecia, neutropenia, and tions and are a heterogenous group of mature T-cell neo-
infections. In the MCL patients (n = 93), median age 70, BR plasms arising from postthymic T-cells at various stages of
was superior to R-CHOP for median PFS (35 vs 22 months, differentiation. NK-cell lymphomas are included in this
P = 0.006). A similarly designed trial was conducted in North group because of the close relationship between these two
America (Flinn et al., 2014). MCL patients (n = 67) com- cell types. The importance of the T-cell phenotype and the
prised a subset of the study population. MCL patients impact on prognosis is now well established but is a relatively
assigned to BR were more likely to achieve a complete remis- recent advance. A recent large retrospective study, the Inter-
sion than patient assigned to R-CHOP or R-CVP (50% vs national Peripheral T-Cell Lymphoma Project (ITLP), col-
27%). Taken together, these studies suggest that VR-CAP lected 1,153 cases of PTCLs from 22 centers from around the
and the BR regimen are better induction platforms than world and highlighted the geographic, clinicopathologic,
R-CHOP. Trials testing maintenance rituximab after BR are and prognostic differences of this diverse group of diseases
ongoing. (Vose et al., 2008).
Given disease rarity, there are no randomized controlled
trials establishing that an alternate regimen is superior to
Management of relapsed MCL CHOP, and thus CHOP remains the standard therapy of
Younger patients relapsing after intensive therapies are can- PTCLs. There is a range of disease in this category (Table
didates for alloSCT. The literature varies widely in the effi- 21-3), and a minority have a more favorable prognosis or a
cacy of this approach, but it does appear to have curative more indolent course.
potential for a fraction of patients (25%-50%). A multi-
center experience using a reduced-intensity conditioning Indolent PTCLs
(RIC) approach demonstrated 2-year EFS and OS rates of
Mycosis fungoides and Sézary syndrome
50% and 53%, respectively. The 2-year transplant-related
mortality rate was 32%, highlighting the high-risk/high- In contrast to nodal NHLs, which are mostly B-cell derived,
reward nature of allogeneic SCT in relapsed MCL. For older ~75% of primary cutaneous lymphomas have a T-cell phe-
patients, the BR regimen is highly active in relapsed MCL, notype and two-thirds are either mycosis fungoides (MF) or
with ORR of 75%-92% reported in two small studies. The Sézary syndrome (SS). MF is an epidermotropic, primary
proteasome inhibitor bortezomib is FDA approved for cutaneous T-cell lymphoma and represents the most com-
relapsed MCL and has moderate activity, with an ORR of mon of all primary cutaneous lymphomas (50%). MF usu-
33% and a median PFS of 6 months. The mTOR inhibitor ally has an indolent course, but similar to indolent B-cell
temsirolimus is European Union approved for relapsed lymphomas, it is incurable. MF is limited to the skin in its
MCL, demonstrating on ORR of 22% and median PFS of 4.8 early phases and appears as plaques or patches; but with
months in a pivotal study. The immunomodulatory agent time, it evolves to diffuse erythroderma or cutaneous nod-
lenalidomide is FDA approved for recurrent MCL. In the ules or tumors, usually with associated adenopathy. The
EMERGE study (n = 134), lenalidomide produced response early stage lesions are characteristically in a bathing suit dis-
rates of 28%. Although the median PFS was just 4 months, tribution and are often pruritic in nature. Extracutaneous
the median duration of response of 16.6 months, indicating disease can occur in advanced stages and may indicate histo-
responder can experience durable benefit (Goy et al., 2013). logic transformation. The histology varies with stage of the
Lenalidomide, which potentiates immune effector cells, disease, but epidermotropism is seen with typical plaques
appears to be even more active when combined with ritux- and intradermal collections of so-called Pautrier microab-
imab. A phase 1/2 trial in relapsed MCL (n = 52) reported an scesses. The T-cells are CD4+/CD8–, often with aberrant loss

of the T-cell antigens CD2, CD3, CD5, and CD7. Progression acetylation, thus altering the gene expression of cell-cycle
to nodal disease, organ infiltration, and circulating clonal and apoptotic regulatory proteins. Vorinostat and romidep-
T-cells (SS) represents the advanced stage of disease. A sin both are approved for the treatment of CTCLs. Vorino-
unique clinical staging system has been proposed by the stat is orally available and has an ORR of ~30% and a median
International Society for Cutaneous Lymphomas (ISCL) and duration of response (DoR) of ~6 months. A phase 2 trial
the Cutaneous Lymphoma Task Force of the European Orga- with romidepsin demonstrated an ORR 35% (CR 6%) with
nization of Research and Treatment of Cancer (EORTC) for a median DoR of 15 months in one study and 11 months in
MF and SS. The extent of cutaneous and extracutaneous dis- another. Side effects that are common with histone deacety-
ease is the most important prognostic factor in MF, with a lase (HDAC) inhibitors are fatigue, nausea, vomiting, neu-
10-year disease-specific survival ranging from 97% to 98% tropenia, and thrombocytopenia. Prolonged QT also can
for patients with limited patch/plaque disease (<10% of skin occur, and thus electrolytes should be monitored closely and
surface; stage I) to 20% for patients for patients with lymph an electrocardiogram should be performed in high-risk
node involvement. patients during therapy. Alemtuzumab, the humanized
SS is a distinct disorder characterized by erythroderma, monoclonal antibody targeting CD52, also has been used in
generalized lymphadenopathy, and the presence of Sézary MF and SS with some success; however, patients are at high
cells in the skin, lymph nodes, and peripheral blood. It is risk of opportunistic infections. More recent studies report
associated with an aggressive course with a 5-year OS rate of single agent activity for lenalidomide (ORR 28%) and low
20%-30% with lower rates seen with high Sézary cell counts. dose pralatrexate given at 15 mg/m2 for 3 out of every 4
Because MF is incurable and the use of early therapy does weeks (ORR 45%).
not affect survival, a nonaggressive approach is recom- Allogeneic transplant has been explored in select cases of
mended (reviewed Prince et al., 2009). Patients with stage IA MF and SS. The European Group for Blood and Marrow
disease may be managed expectantly with careful surveil- Transplantation recently reported a multi-institutional
lance. If treatment is needed, topical steroids or topical restrospective study evaluating alloSCT (myeloablative and
nitrogen mustard, electron-beam radiotherapy, or cutane- RIC) in 60 patients with MF (n = 36) or SS (n = 24). Almost
ous photochemotherapy with oral psoralen plus ultraviolet half had refractory disease at the time of alloSCT and the
A (PUVA) typically are employed. Phototherapy with PUVA median number of prior regimens was four. With a median
or ultraviolet B (UVB) is recommended for more widespread follow-up of 3 years, the 3-year PFS and OS were 34% and
disease. Low-dose radiotherapy can be helpful to improve 53%, respectively, with higher survival rates observed in the
symptoms and cosmesis. Patients with progressive disease RIC group (3-year PFS 52% vs 29%, P = 0.006).
and those with systemic dissemination may be appropriately Large-cell transformation in MF is defined as large cells in
treated with methotrexate or corticosteroids, although >25% of the infiltrate or if these cells form microscopic nod-
responses are usually brief. ules. The incidence ranges from 8% to 39% and typically is
Combination chemotherapy regimens are not particularly associated with a poor prognosis, but long-term survivors
effective and provide only transient responses. Single-agent can be seen. One study evaluated 100 cases of transformed
treatments are preferred, particularly with slowly progressive MF and the median survival was 2 years with a 5-year OS and
disease, because of a high risk of myelosuppression and disease-specific survival (DSS) of 33% and 38%, respectively.
infection and only modest response durations seen with The factors associated with a poor DSS were CD30-negative
combination chemotherapy. Gemcitabine (ORR 48%-75%), status, folliculotropic MF, generalized skin lesions, and
pentostatin (ORR 28%-71%), and liposomal doxorubicin extracutaneous transformation. Those cases with zero fac-
(ORR 56%-88%) have single-agent activity. Alternatively, tors had a 2-year DSS of 83% compared with 14%-33% in
IFNα, bexarotene, vorinostat, and denileukin diftitox all patients with three or four factors. The optimal management
have efficacy in advanced-stage MF and SS. Bexarotene is an is unclear, but for young patients, systemic chemotherapy
oral retinoid and is FDA approved for cutaneous T-cell lym- should be used and consideration should be made for autol-
phoma (CTCL). In a multicenter trial of 94 patients with ogous or allogeneic transplantation particularly with high-
advanced stage MF/SS, the ORR was 45% but with only 2% risk disease. Consolidative radiation may be considered in
CRs. The common toxicities are hypertriglyceridemia (82%) local transformation.
and central hypothyroidism (29%). Denileukin diftitox is a
recombinant fusion protein that combines interleukin 2
Primary cutaneous ALCL
(IL-2) with the cytotoxic A chain of diphtheria toxin with
an ORR of 49%. It is approved by the FDA for patients with Primary cutaneous ALCL (C-ALCL) is part of a spectrum
relapsed CTCL whose tumors express the IL-2 receptor sub- of diseases in the category of primary cutaneous CD30+
unit (CD25). Histone deacetylase inhibitors prevent histone T-cell lymphoproliferative disorders that also includes

lymphomatoid papulosis and “borderline” cases that have analogs have been used in the relapsed setting. Splenectomy
overlapping features of both disorders. C-ALCL is the sec- may be useful in cases with an accompanying splenomegaly,
ond most common type of CTCL. Patients are typically refractory cytopenias, or autoimmune hemolytic anemia or
older males (median age 60 years), presenting with a soli- thrombocytopenia. The anti-CD52 monoclonal antibody
tary nodule with multifocal disease occurring in only 20% alemtuzumab can be used in select cases (ORR 50%). There
of patients. Partial or complete spontaneous regression are case reports of long-term remission from LGL leukemia
occurs in ~25% of cases. C-ALCL must be distinguished in RA patients treated with rituximab.
from systemic ALCL with secondary cutaneous involve-
ment through staging procedures. Aggressive PTCLs
The outcome is very favorable with a 10-year DSS of 95%.
Adult T-cell leukemia/lymphoma
It is notable that patients with localized C-ALCL with one
draining lymph node involved have a similarly good progno- Adult T-cell lymphoma/leukemia (ATL) is caused by infec-
sis. For localized C-ALCL, radiation is the preferred therapy tion with HTLV-1 and occurs in areas of endemic infection
as the impact of chemotherapy is unknown. Progression to (eg, the Caribbean basin and southwestern Japan). The
systemic involvement can occur in a minority of cases. For cumulative incidence of adult T-cell lymphoma/leukemia
more advanced stage cases, the best management is unclear. among HTLV-1 carriers is 2.5% in Japan. The virus can be
An argument can be made to treat conservatively in mini- transmitted in breast milk and blood products. The malig-
mally symptomatic patients with radiotherapy for a few nant cells have a distinct cloverleaf appearance and are
lesions or low-dose methotrexate, similar to the management CD7−, and most are CD4+/CD8− and CD25+. The following
of lymphomatoid papulosis. If a more aggressive behavior is clinical variants have been recognized: (i) acute type with a
observed, multiagent chemotherapy is reasonable. rapidly progressive clinical course, bone marrow and periph-
The anti-CD30 antibody drug conjugate (ADC) brentux- eral blood involvement, hypercalcemia with or without lytic
imab vedotin, described further below in the section “PTCL- bone lesions, skin rash, generalized lymphadenopathy, hepa-
NOS,” is being evaluated in C-ALCL, since these tumors tosplenomegaly, and pulmonary infiltrates; (ii) lymphoma
usually express CD30. There are a variety of case reports type with prominent adenopathy but lacking peripheral
demonstrating its efficacy. blood involvement but also associated with an aggressive
course; (iii) chronic type with lymphocytosis and occasion-
ally associated with lymphadenopathy, hepatosplenomegaly,
T-cell large granular lymphocytic leukemia
and cutaneous lesions but having an indolent course; and
T-cell large granular lymphocytic leukemia (T-LGL) is (iv) smoldering type with <5% circulating neoplastic cells,
defined by a persistent (>6 months) increase in the number skin involvement, and prolonged survival. The chronic and
of peripheral blood large granular lymphocyte cells without smoldering forms can progress to the acute form after a vari-
an identifiable cause. The lymphocytosis is usually between 2 able length of time. In the ITLP, 126 patients (9.6% of all
and 20 × 109/L. The T-cells are CD3+, CD8+, and CD57/ PTCLs) were identified with either the acute (13%) or lym-
CD16 and are expressed in most cases, but CD56 is negative. phoma-type (87%) ATL. Opportunistic infections are com-
It arises more commonly in rheumatoid arthritis or other mon, and Strongyloides serology is recommended before
autoimmune disorders. Most cases have an indolent clinical starting therapy.
course, with a median survival time of ~13 years, but rare Survival times in the acute and lymphomatous variants
cases with an aggressive course also have been described. Of are ~6 and ~10 months, respectively. The median survival
note, T-LGL should be distinguished from NK-cell leukemia, time for the chronic form is 2 years. The 4-year OS for the
which does have a fulminant aggressive course (see the sec- acute, lymphoma, chronic, and smoldering types has been
tion “Aggressive NK-cell leukemia”). In T-LGL, moderate reported to be 5%, 5.7%, 27%, and 63%, respectively.
splenomegaly is the most common clinical finding, and Asymptomatic patients with the smoldering or chronic types
lymphadenopathy is rare. Severe neutropenia with or with- can be monitored closely. For young, fit patients with the
out anemia is common, but thrombocytopenia is rare. A acute and lymphoma subtypes, the intensive regimen VCAP
variety of autoimmune disorders, including hemolytic ane- (vincristine, cyclophosphamide, doxorubicin and predniso-
mia, thrombocytopenia, and pure red blood cell aplasia, also lone/AMP [doxorubicin, ranimustine, prednisolone])/VECP
may occur. If treatment is required for cytopenias, immuno- (vindesine, etoposide, carboplatin, prednisolone) may be
modulatory agents such as low-dose methotrexate, cyclospo- considered. The Japan Clinical Oncology Group (JCOG)
rine A, cyclophosphamide, chlorambucil, or corticosteroids reported a phase 3 trial comparing the dose-intensive regimen
can be effective. Responses can take up to 4 months, and lon- VCAP/AMP/VECP versus CHOP-14 alone and showed a
ger therapy often is needed to maintain the response. Purine more favorable CR rate (40% vs 25%, P = 0.02) and 3-year

OS (24% vs 13%) that was significant after adjusting for Treatment approaches in PTCL have paralleled those
prognostic factors but only for the one-sided P-value DLBCL; as a result, CHOP is considered the standard ther-
(P = 0.028) (Tsukasaki et al., 2007). The median survival for apy, despite consistent evidence that it is rarely curative. Fur-
the intensive regimen was just over 1 year, but toxicity was thermore, because of disease rarity, most studies have
high (grade 4 neutropenia in 98% and grade 3/4 infections in combined all subtypes that could obscure benefits in select
32%). Thus, this regimen should be used only in carefully subtypes. Limited analyses suggest that the use of anthracy-
selected patients, particularly with the lymphoma subtype. clines, a key component of CHOP, may not affect outcome
Relapse rates remain high, and patients should be referred in PTCL-NOS (Vose et al., 2008). Although the limitations
for considered for transplant. of CHOP are fully recognized, there is no clear evidence that
A number of phase 2 studies have evaluated the antiretro- any other approach is superior.
viral zidovudine (AZT) and IFN with response rates up to The DSHNHL group retrospectively analyzed the out-
92% and median OS of 11 months in untreated patients. For come of PTCL patients (n = 331) that had been enrolled in
the leukemia subtype, these results are superior to what is phase 2 or phase 3 aggressive lymphoma studies and evalu-
achieved with combination chemotherapy. For the chronic ated the impact of etoposide. In patients <60 years with a
and smoldering type, a recent meta-analysis demonstrated normal LDH, EFS was extended with etoposide (P = 0.003),
100% OS after 10 years with this approach. whereas OS did not improve significantly (P = 0.176). The
Chemokine receptor 4 (CCR4) is expressed in ~90% of addition of etoposide appeared to have the greatest impact
cases of ATL. Mogamulizumab (KW-0761) is a humanized in the favorable group of patients with ALK-positive ALCL
monoclonal antibody targeting CC4, and a recently (3-year EFS 91% vs 82%, P = 0.012). In patients with PTCL-
reported phase 2 study demonstrated an ORR of 50%, NOS, ALK-negative ALCL, and AITL, there was only a
including eight CRs, in 27 treated patients. The median PFS trend to improved 3-year EFS (61% vs 48%; P = 0.057),
and OS were 5.2 months and 13.7 months, respectively with no OS difference observed; however, patient numbers
(Ishida et al., 2012). The most common side effects were were small. Given that this is not a randomized compari-
lymphopenia (96%), neutropenia (52%) and thrombocy- son, the true benefit of the addition of etoposide remains
topenia (52%), infusion reaction (89%), and skin rashes unknown.
(63%). This agent is also be explored in other CCR4+ CTCL Alemtuzumab selectively targets CD52, which is present
and PTCLs. on normal T-cells but expression across PTCLs is more vari-
able. In the initial studies of alemtuzumab in relapsed or
PTCL, not otherwise specified; systemic anaplastic large cell refractory PTCLs, the ORR was 36%, but the TRM was
lymphoma; and angioimmunoblastic T-cell lymphoma also 36% because of profound immunosuppression and
opportunistic infections that can occur. Several studies have
PTCL, not otherwise specified (PTCL-NOS); systemic ana- evaluated alemtuzumab with CHOP-21 or CHOP-14 in the
plastic large cell lymphoma (ALCL); and angioimmunoblas- management of PTCL with variable results. Toxicity has
tic T-cell lymphoma (AILT) are the most common subtypes been problematic, including opportunistic infections. Phase
of PTCL encountered in North America, representing 66% 3 studies are ongoing to determine whether the addition of
of all PTCL cases. alemtuzumab improves outcome in PTCL.
Pralatrexate is a novel folate analogue that has enhanced
PTCL-NOS PTCL-NOS is the most common subgroup of uptake and cellular retention compared with MTX. Early
PTCLs, accounting for up to 30% of cases worldwide. This is studies suggested a sensitivity of TCLs over BCLs. The phase
the default PTCL category for any mature T-cell neoplasm 2 PROPEL study evaluated pralatrexate (with vitamin B12
that does not fit into any of the specified categories in and folate) in relapsed/refractory PTCLs and demonstrated
the WHO classification. Patients typically present with an ORR 29% (CR 11%), a median PFS of 3.5 months and a
advanced-stage disease, and the 5-year OS is 20%-30% in median DoR of 10.5 months (O’Connor et al., 2011). The
most series. The morphologic spectrum of PTCL-NOS is main toxicities were mucositis, thrombocytopenia, and neu-
wide, including the histiocyte-rich lymphoepithelioid, or tropenia. These results led to FDA approval of pralatrexate
Lennert lymphoma. Typically, the neoplastic cells are CD4+/ in September 2009 for the treatment of relapsed/refractory
CD8; CD5 and CD7 frequently are downregulated, and PTCL. Studies are ongoing combining pralatrexate with
~30% are CD30. Gene expression profiling has been explored other agents in the up-front and relapsed settings.
in heterogeneous PTCL-NOS to determine whether there Romidepsin is an HDAI that has been evaluated in CTCLs
are reproducible molecular subsets and to better define and PTCLs. A phase 2B registration study evaluated
prognostic markers within PTCL-NOS. In comparison with romidepsin in 130 patients with relapsed or refractory PTCL.
B-cell lymphomas, however, large-scale studies are lacking. The ORR was 25% (CR 15%), median DoR was 17 months,

and median PFS was 4 months, leading to FDA approval in hepatosplenomegaly. It was originally believed to be a
2011. Side effects were as previously described in the CTCL form of immune dysregulation, with polyclonal gammopa-
studies. A phase 1b study is ongoing combining CHOP with thy and other hematologic abnormalities (Coombs-positive
romidepsin for the primary treatment of PTCL. hemolytic anemia) reflecting B-cell hyperactivity. Oppor
Belinostat is another HDAI that has demonstrated tunistic infections can occur because of the underlying
responses in relapsed or refractory PTCL in a phase 2 trial. immunodeficiency.
Belinostat was granted approval by the FDA for the treat- Survival is similar to PTCL-NOS (5 year ~30%); however,
ment of patients with PTCL who have received at least one a small proportion may have a more indolent course. CHOP
prior therapy. A phase 2 trial (BELIEF trial) of belinostat in typically is used for the primary therapy, and although the
120 patients with PTCL reported overall and complete response rate is high, relapse is common and infectious com-
remission rates of 26% and 11%, respectively, with a median plications are problematic. GELA evaluated AITL patients
duration of response of 13 months. enrolled on different therapeutic protocols and found no
CD30 is expressed uniformly in ALCL but also highly improvement of survival with any therapy, including HDC/
restricted, making it an attractive target in this disease. Stud- ASCT. With the presence of EBV-infected B-immunoblasts
ies with the nascent anti-CD30+ in relapsed systemic ALCL and the evidence of B-cell hyperstimulation, GELA also
were largely disappointing, however, and thus to enhance recently evaluated R-CHOP in AITL in a phase 2 study. Of
tumor activity an antibody-drug conjugate (ADC), brentux- 25 evaluable patients, the ORR was 80% (CR 44%) but with
imab vedotin (formerly known as SGN-35), was developed. a median follow-up of 2 years, the 2-year PFS was only 42%,
The ADC conjugates the CD30 monoclonal antibody to the which was similar to a prior study using CHOP alone. With
microtubulin inhibitor, monomethyl auristatin E (MMAE), poor outcomes using conventional therapy, immunomodu-
by an enzyme-cleavable dipeptide linker. Following binding latory agents also have been explored, including cyclospo-
to CD30+ and uptake into the cell, MMAE is released and rine, lenolidomide, thalidomide, and interferon. A
interferes with tubulin formation. A phase 2 study recently retrospective study evaluating cyclosporine in relapsed or
was reported in relapsed or refractory systemic ALCL that refractory AITL demonstrated an ORR 67% and a median
demonstrated an ORR 86% (CR 57%), median duration of DoR of 13 months.
response of 12.6 months, and a median PFS 13.3 months,
which also prompted FDA approval for this disease in 2011. Systemic anaplastic large-cell lymphoma ALCL is
The main side effect of brentuximab vedotin is peripheral composed of large CD30+ anaplastic cells with a predilection
neuropathy. Studies are ongoing evaluating brentuximab for a sinusoidal and cohesive growth pattern. In the WHO
vedotin in the upfront setting with CHP, omitting the vin- classification, primary systemic ALCL is separated from
cristine because of overlapping toxicity. cutaneous ALCL; and more recently, ALK-positive ALCL
With the disappointing results with CHOP, some groups has been defined as a distinct entity (Table 21-3). Cases of
are evaluating new chemotherapy combinations in PTCL. ALK-positive ALCL are associated with a characteristic chro-
Gemcitabine has shown reasonable single-agent activity in mosomal translocation, t(2;5)(p23;q35), resulting in a fusion
previously treated patients and is being explored with other gene, NPM-ALK, encoding a chimeric protein with tyrosine
agents. Bendamustine has been associated with 50% response kinase activity. With the availability of antibodies to the ALK
rate in relapsed/refractory PTCL. A number of novel agents protein, ALK expression can be demonstrated in 60%-85%
are under investigation in PTCLs. of all systemic ALCL, with higher frequencies seen in the
pediatric and young adult age-groups. In contrast, although
Angioimmunoblastic T-cell lymphoma AITL is a well- ALK-negative ALCL lacks any defining features, there is
defined, distinct PTCL subtype, with unique pathobiologic accumulating evidence that it should be separated from
features. Key morphologic findings of AITL include an other PTCLs; as a result, it is considered a provisional entity
expanded CD21+ follicular dendritic cell network and promi- in the updated WHO classification.
nent arborizing high-endothelial venules (HEV). The neo-
plastic cells in AITL are mature CD4+/CD8– T-cells, expressing ALK-positive ALCL Morphologically ALK-positive ALCL
most pan-T-cell antigens. EBV-positive B-cells are seen in has pathognomonic “hallmark cells” recognized by their
most cases, and EBV-positive DLBCL has been reported. It eccentric, horseshoe, or kidney-shaped nuclei. In addition
appears that the cell of origin is the follicular helper T-cell with to strong expression of CD30, ALK-positive ALCL is usu-
T-cells CD10+, BCL6+, and CXCL13+ and derivation also is ally positive for epithelial membrane antigen (EMA) and
supported by gene-expression profiling studies. cytotoxic markers (TIA1, granzyme B, and perforin). Sev-
Patients are typically in their sixth or seventh decade and eral studies have established that patients with ALK-
have advanced-stage disease, often with B-symptoms and positive ALCL have a more favorable prognosis with

anthracycline-based chemotherapy than patients who have ALK-negative ALCL associated with breast implants ALCL
ALK-negative ALCL and other PTCLs, as well as DLBCL, at associated with implants typically involves the capsule with-
least in the prerituximab treatment era. The improved out- out invasion of the breast tissue, or it presents as an unex-
come in part is related to the young age at presentation. The plained seroma or mass, which usually is CD30+ ALK
international T-cell lymphoma project (ITLP) confirmed negative. The neoplastic cells float in the effusion fluid or
the superior outcome of ALK-positive ALCL (5-year FFS become embedded tissue; importantly, however, breast
60%; 5-year OS 60%) compared with ALK-negative ALCL parenchyma usually is not involved and the ALCL cells are at
(5-year FFS 36%; 5-year OS 49%). If the comparison is con- distance from the breast tissue. It is associated with both sili-
fined to patient’s <40 years old, however, there was no dif- cone and saline implants. Recently, investigators at the Uni-
ference in survival. Similar findings were reported from a versity of Southern California have collected 90 cases to date
retrospective analysis of patients with ALCL enrolled on worldwide and put forth recommendations (Brody et al.,
GELA studies, which reported that in patients <40 years of 2012). A total capsulectomy should be performed, and
age, there was no impact of ALK status on PFS or OS (Sibon because bilateral cases have been reported, removal of the
et al., 2012). uninvolved breast implant should be considered. The grow-
Given the favorable outcome with anthracycline-based ing body of literature supports that ALK-negative ALCL in
chemotherapy, CHOP is considered to be the standard ther- this setting appears to have a more indolent clinical course,
apy of ALK-positive ALCL. Patients with multiple IPI factors and most patients can be observed following removal of the
have a poor outcome, however, and could be considered for implant and capsule. Recent reports suggest similar survival
clinical trials. Given the strong and uniform expression of compared with those who received chemotherapy or radia-
CD30, the ADC brentuximab vedotin has been tested in the tion, but rare aggressive cases have been reported. Cases that
relapsed or refractory setting and has significant efficacy, have identified a distinct breast mass may be better classified
prompting evaluation in the frontline setting. as a typical systemic ALK-negative ALCL and may be treated
Crizotinib is a small molecule inhibitor of the ALK tyro- accordingly (Aladily et al., 2012).
sine kinase that has demonstrated activity in a subset of
patients with ALK positive nonsmall-cell lung cancer. Sev-
Extranodal NK-/T-cell lymphoma, nasal type
eral case reports of patients with multiply relapsed ALK posi-
tive anaplastic large cell lymphoma (ALCL) reported Extranodal NK-/T-cell lymphomas, nasal type, display great
complete responses after treatment with crizotinib. variation in racial and geographic distribution, with the
majority of cases occurring in the Far East. Patients are typi-
ALK-negative ALCL Patients with ALK-negative ALCL tend cally males, 40-50 years old. The tumor cells show angioin-
to be older at presentation; the clinical presentation is simi- vasion and necrosis is prominent. The designation NK/T is
lar to ALK-positive cases, but sites of extranodal disease may used to reflect the fact that although most are NK-cell
vary. Pathologically, it is not reproducibly distinguished derived (CD2+, CD56+, CD3 [cytoplasmic]+, EBV+), rare
from ALK-positive ALCL other than lacking the ALK pro- cases with identical clinical and cytologic features exhibit an
tein. ALK-negative ALCL has been difficult to define, in part EBV-positive or CD56−, cytotoxic T-cell marker positive
due to a lack of uniformly applied diagnostic criteria across (TIA1, perforin, and granzyme B). Circulating EBV in the
studies. Recently, a recurrent balanced translocation t(6;7) peripheral blood can often be detected, providing another
(p25.3;q32.3) has been identified in ALK-negative ALCL, method of disease monitoring. The majority of cases remain
but the significance is unknown. Previously, it was argued localized with <20% presenting with advanced-stage disease.
that ALK-negative ALCL had a similar outcome to PTCL- Despite the predominant nasal location, spread to the CSF is
NOS and they should be grouped together. In recent years, uncommon. Most occur in the nasal region, but identical
there is accumulating evidence that they differ not only tumors also can occur at extranasal sites, such as the skin,
pathologically and genetically but also prognostically. The soft tissue, GI tract, and testis (ie, extranasal). It appears that
ITLP compared the outcome of ALK-negative ALCL with cases involving extranasal regions may have a more aggres-
PTCL-NOS and established that ALK-negative ALCL had a sive course. From the ITLP the 5-year OS for stage I/II
more favorable 5-year FFS (36% vs 20%, P = 0.012) and NK-/T-cell lymphomas was ~50% and 15% for nasal and
OS (49% vs 32%, P = 0.032). These data confirm that extranasal sites, respectively, and the corresponding esti-
ALK-negative ALCL should be considered distinct from both mates for stage III/IV patients were 30% and <10%. The IPI
ALK-positive ALCL and PTCL-NOS. Although the survival does not stratify patients well because most have localized
is more favorable than PTCL-NOS, it is still poor, particu- disease and often with good PS. A Korean index using B
larly with multiple IPI factors. Novel therapies, including symptoms, stage (I/II vs III/IV), regional lymph nodes,
brentuximab vedotin, are being explored. LDH, and PS appears to be more useful in prognostication,

particularly for the low and low-intermediate IPI cases and Aggressive NK-cell leukemia
may help to guide treatment decisions. Patients fall into four
Aggressive NK-cell leukemia is a rare form of leukemia that
risk groups with widely disparate outcomes: group 1: no RF,
almost always is associated with EBV infection and has a
5-year OS ~ 81%; group 2: 1 RF, 5-year OS ~64%; group 3:
median survival of only 3 months. It is seen most often in
2 RF, 5-year OS ~34%; and group 4: 3 or 4 RF, 5-year OS
Asians, and the median age of onset is 42 years. Typically, the
7%. Risk factors identified in other studies have also included
bone marrow and peripheral blood are involved in addition
local tumor invasion (tone or skin), high Ki-67, or EBV
to the liver and spleen. Patients often have fever and consti-
DNA titer >6.1 × 107 copies/mL.
tutional symptoms and multiorgan failure with coagulopa-
Accumulating evidence indicates radiotherapy is impor-
thy and hemophagocytic syndrome. It is unclear whether
tant in the management of patients with localized NK-/T-
aggressive NK-cell leukemia represents the leukemic phase
cell lymphoma with more favorable outcomes observed
of extranodal NK-/T-cell lymphoma. There is no known
using high doses of radiotherapy (50-60 Gy) early in the
curative therapy, and responses to chemotherapy are usually
frontline setting. Recently, the use of platinum as a radiosen-
brief. Some encouraging results have been seen with l-aspar-
sitizer has been explored and may allow for the use of lower,
aginase–based treatment in this disease and extranodal
less-toxic doses of radiation. Furthermore, because systemic
NK-/T-cell lymphoma but both require further study.
relapse can occur with single-modality radiotherapy, other
novel combinations are being tested. The outcome with
Rare aggressive PTCL subtypes
CHOP has been disappointing, and it has been speculated
that this may be due to overexpression of p-glycoprotein Subcutaneous panniculitis-like T-cell lymphoma Sub
expression conferring multidrug resistance. Concurrent cutaneous panniculitis-like T-cell lymphoma (SCPTCL) is
radiation (40 Gy) and cisplatin, followed by three cycles of an extremely uncommon PTCL subtype that preferentially
VIPD (etoposide, ifosfamide, cisplatin), was evaluated in infiltrates the subcutaneous tissue. It has been determined
stage IE/IIE nasal NK-/T-cell lymphoma. In this highly that tumors with gδ phenotype have a far inferior prognosis
selected population, the CR rate was 83% and the 3-year PFS to those with aβ phenotype (5-year OS, 11% for gδ vs 82%
was 85% (Kim et al., 2009). Similarly, concurrent radiother- for aβ) (Willemze et al., 2008). In the WHO classification,
apy (50 Gy) and DeVIC chemotherapy (dexamethasone, SCPTCL is confined only to the aβ, which usually have a
etoposide, ifosfamide, carboplatin) was evaluated in a phase CD4–/ CD8+, and CD5– phenotype. Cases with a gδ pheno-
1/2 trial in localized nasal NK-/T-cell lymphoma with good type are combined in a new, rare PTCL entity, termed pri-
results (CR 77%, 2-year PFS 67%) (Yamaguchi et al., 2009). mary cutaneous gδ T-cell lymphoma (see section “Primary
In the absence of a randomized trial, the most recent NCCN cutaneous PTCL, rare aggressive subtypes”) because of simi-
guidelines suggest either high-dose radiotherapy alone (>50 lar aggressive behavior. The optimal therapy for aβ SCPTCL
Gy) for stage 1 patients without risk factors (as described) or is unknown, with durable responses observed with both
concurrent chemoradiotherapy (stage 1 or 2) using either of CHOP and immunosuppressive agents.
the noted regimens for localized NK-/T-cell lymphoma.
For advanced-stage disease, l-asparaginase has emerged as Hepatosplenic T-cell lymphoma Hepatosplenic T-cell
an active agent in NK-/T-cell lymphomas with an ORR of lymphoma is a rare PTCL subtype occurring usually in
87% (CR 50%) in relapsed or refractory patients. Anti- young men (median age 34 years) presenting with hepatos
thrombin levels require close monitoring. A phase 2 study plenomegaly and bone marrow involvement. Up to 20% of
evaluating l-asparaginase in combination with MTX and hepatosplenic T-cell lymphomas occur in the setting of
dexamethasone (AspaMetDex) in previously treated immunosuppression, most commonly following solid organ
patients, demonstrates an ORR of 78% (CR 61%) and a transplantation. It also has been observed in patients treated
median DoR of 12 months (Jaccard et al., 2011). A phase 2 with azathioprine and the TNFα inhibitor, infliximab, which
study evaluating the SMILE regimen (steroid, methotrexate, is used in Crohn disease. The splenic red pulp is diffusely
ifosfamide, l-asparaginase, etoposide) in 38 patients with involved, and the liver will show a sinusoidal pattern. Most
either newly diagnosed stage IV or relapsed or refractory tumor cells are CD3+, CD4–, and CD8–, and most are associ-
NK-/T-cell lymphoma demonstrated an ORR after two ated with isochrome 7q. The majority of cases are of the gδ
cycles of 79% (CR 45%) and 19 patients subsequently under- TCR type; however, rare cases that are of the aβ TCR type
went SCT. The 1-year OS rate was 55%, but grade 4 neutro- have been reported. The prognosis is extremely poor with rare
penia occurred in 92% and the grade 3/4 infection rate was long-term survivors. The optimal therapy is unknown; how-
61%. Additional studies incorporating l-asparaginase in the ever, CHOP does not appear to cure this disease. Long-term
frontline treatment of both localized and advanced-stage survivors have been reported with high-dose chemotherapy
NK-/T-cell lymphoma are ongoing. and ASCT or alloSCT and referral at diagnosis is suggested.

Enteropathy-associated T-cell lymphoma Enteropa- tumor cells are CD3+, CD4−, CD8+, and cytotoxic marker
thy-associated T-cell lymphoma (EATL) is a rare, aggressive positive, and the clinical course is aggressive.
intestinal tumor with a male predominance that often occurs
in the setting of celiac disease. It most commonly involves
Transplant in PTCL
the jejunum or ileum. Patients often present with abdominal
pain, and intestinal perforation can occur. The prognosis is Multiple retrospective studies have been published evaluat-
extremely poor due to chemotherapy resistance and difficult ing the impact of upfront transplantation in PTCL, as has
treatment delivery related to abdominal complications that been comprehensively reviewed (Yared et al., 2012). Trial
can arise in the setting of malabsorption. In some cases, interpretation and comparisons are difficult for a number of
there is a childhood history of celiac disease, but more com- reasons, including the evaluation of heterogeneous patient
monly, the disease occurs in adulthood. Alternatively, there populations, potential for selection bias, and the dearth of
is a prodrome of refractory disease or a concomitant diagno- intention-to-treat (ITT) data. Because there are no reported
sis of celiac disease is found at the time the lymphoma is dis- prospective randomized phase 3 trials comparing HDC/
covered. In the updated WHO classification, a sporadic, ASCT with conventional-dose chemotherapy, specifically for
monomorphic variant, type II EATL, has been defined that PTCL, it remains challenging to determine the relative impact
occurs in 10%-20% of cases and has a broader geographic of patient selection versus true differences in efficacy.
distribution that includes Asia. An association with celiac GELA performed a retrospective analysis of the impact of
disease has not been definitively proven in this subtype; thus, upfront autologous transplant in T-cell lymphomas. Limit-
this may represent a distinct disease entity. In the common ing the study to patients who achieved CR, a matched-pair
subtype, the neoplastic cells are CD3+, CD7+, CD4−, CD8−/+, analysis was performed comparing dose-intensive chemo-
CD56− and contain cytotoxic proteins. The monomorphic therapy alone (ACVB or NCVB (mitoxantrone substitution)
form is CD3+, CD4−, CD8+, and CD56+. versus chemotherapy plus HDC/ASCT. No difference in
The ITLP recently reported on 62 patients with EATL, DFS or OS was found, but the ACVB is considered more
which represented 5.4% of all lymphomas worldwide, most dose-intensive than CHOP.
commonly in Europe. Type I and type II EATL represented Several phase 2 prospective studies of upfront transplant
66% and 34% of the cases, respectively. The 5-year FFS was have been published and represent more homogeneous pop-
only 4% and OS was 20%, with the majority of patients ulations of treated patients. The Nordic group completed the
treated with CHOP-type chemotherapy. Similar disappoint- largest prospective phase 2 trial of upfront transplant (NLG-
ing results are observed in other studies with CHOP-type T-01) in 160 patients with PTCL, excluding ALK-positive
therapy, which has prompted evaluation of HDC/ASCT (see ALCL. The planned treatment scheduled was CHOEP-14 for
“Transplant in PTCL”). six cycles (CHOP-14 in patients >60 years), followed by
BEAM/BEAC and ASCT in responding patients (d’Amore
Primary cutaneous PTCL, rare aggressive subtypes et al., 2012). In total 160 patients represented the ITT popula-
Primary cutaneous gd T-cell lymphoma In the updated WHO tion. Most patients had good functional status (71% with PS
classification, primary cutaneous gδ T-cell lymphoma is now scores of 0 or 1), but 72% had an IPI score of >2. The CR rate
considered a distinct entity, which also includes cases previ- pretransplant was 81% to transplant, and the overall trans-
ously known as SCPTCL with a gδ phenotype, as described plant rate was 70% with a TRM of 4%. With median follow-
earlier. Clinically, the extremities are commonly affected, up of 5-years, the 5-year PFS was 44% and 5-year OS was
and the presentation can be variable, with patch or plaque 51%. Patients with ALK-negative ALCL appeared to have a
disease or subcutaneous and deep dermal tumors that may superior 5-year PFS (61%) compared with PTCL-NOS
exhibit necrosis and ulceration. The clonal T-cells have an (38%), EATL (38%), or AILT (49%), but this was not statisti-
activated gδ cytotoxic phenotype and most are CD4−/CD8−. cally significant. The 5-year OS for patients who underwent
Prognosis is poor in this disease, particularly with subcuta- transplant was 61% compared with 28% in those who did
neous fat involvement, with a fulminant clinical course and not. These results suggest that this approach may be appro-
chemoresistance. priate in select patients but still represent level 2 evidence
given the absence of data from a phase 3 trial.
Primary cutaneous aggressive epidermotropic CD8+ T-cell In eligible patients, HDC/ASCT represents the standard of
lymphoma This provisional entity typically presents with care for relapsed or refractory PTCL. In the original PARMA
generalized cutaneous lesions appearing as eruptive papules, study in which HDC/ASCT emerged as superior to second-
nodules, and tumors with central ulceration and necrosis. line chemotherapy alone in relapsed aggressive NHL, immu-
Histologically, there is marked epidermotropism, and inva- nophenotyping was not routinely performed. A subsequent
sion into the dermis and adnexal structures is common. The report of prognostic factors did not identify a difference in

Bibliography | 611
outcome in B- versus T-cell lymphomas; however, the num-

ber of patients with PTCLs was small. There has been no Key points
similar randomized study in PTCLs, but a number of retro- • BL should be treated with dose-intensive regimens as R-CHOP
spective studies report a salvage rate in this setting ranging is inadequate. The treatment should include CNS prophylaxis.
from 18% to 60% (Yared et al., 2012). Given the overall body • Patients with congenital or acquired immunodeficiency have
of evidence, ASCT frequently is offered to patients with an increased risk of lymphoma and often respond poorly to
PTCL with relapsed, chemosensitive disease. therapy.
AlloSCT, with either myeloablative or RIC, also has been • PTCLs have an inferior outcome to DLBCL. The exception is
ALK-positive ALCL, which has a high cure rate with CHOP
reported to yield durable remission in many cases (3-year
chemotherapy unless multiple IPI factors are present at diagnosis.
EFS 23%-64%). Evidence supporting a graft-versus-PTCL
effect comes from studies with donor lymphocyte infusions.
The largest study published to date evaluated 77 previously
Bibliography
treated patients with mainly myeloablative conditioning
(74%). The 5-year PFS was 53%, but the TRM was 34% at 5 Overview of lymphocyte development
years. A phase 2 trial evaluating RIC and alloSCT in 17
Kuppers R, Klein U, Hansmann M-L, Rajewsky K. Cellular origin
patients, demonstrated a 3-year PFS of 64% with a TRM of of human B-cell lymphomas. N Engl J Med. 1999;341:1520-
6%. Allogeneic transplantation is promising in the treatment 1529. An overview of the biology of lymphocyte development and
of PTCL, but it is limited by the availability of stem cell its relation to lymphomagenesis.
donors and toxicity related to graft-versus-host disease. Lenz G, Staudt LM. Aggressive lymphomas. N Engl J Med.
2010;362:1417-1429.
Swerdlow SH, Campo E, Harris NL, et al. WHO Classification
Novel PTCL therapies of Tumors of Hematopoietic and Lymphoid Tissues. 4th ed.
A number of agents are being explored in PTCL, three of Lyon, France: IARC Press; 2008. The current state-of-the-art
which have FDA approval for use today in relapsed/refrac- classification integrating morphologic, phenotypic, and biologic
features for all hematologic neoplasms.
tory disease. Pralatrexate is a novel folate analogue that has
enhanced uptake and cellular retention compared with
MTX. Early studies suggested a sensitivity of TCLs over Lymphoma epidemiology, staging
BCLs. The phase 2 PROPEL study evaluated pralatrexate and evaluation
(with vitamin B12 and folate) in relapsed/refractory PTCLs
Barrington SF, Mikhaeel NG, Kostakoglu L, et al. Role of
and demonstrated an ORR 29% (CR 11%), a median PFS of
imaging in the staging and response assessment of lymphoma:
3.5 months and a median DoR of 10.5 months (O’Connor et
Consensus of the international conference on malignant
al., 2011). The main toxicities were mucositis, thrombocyto- lymphoma imaging working group. J Clin Oncol. 2014;32:
penia and neutropenia. These results led to FDA approval of 3048-3058.
pralatrexate in September 2009 for the treatment of relapsed/ Cheson BD, Fisher RI, Barrington SF, et al. Recommendations for
refractory PTCL. Studies are ongoing combining pralatrex- initial evaluation, staging and response assessment of Hodgkin
ate with other agents in the up-front and relapsed settings. and non-Hodgkin lymphoma: the Lugano classification. J Clin
As described previously, romidepsin is a HDAIs that has Oncol. 2014:32;3059-3066.
been evaluated in CTCLs and PTCLs. A phase 2B registra- Morton LM, Slager SL, Cerhan JR, et al. Etiologic heterogeneity
tion study was published evaluating romidepsin in 130 among non-Hodgkin lymphoma subtypes: the InterLymph NHL
patients with relapsed or refractory PTCL. The ORR was Subtypes Project. J Natl Cancer Inst Monogr. 2014;48:130-144.
25% (CR 15%), median DoR was 17 months, and median
PFS was 4 months, leading to FDA approval in 2011. Belino- Follicular lymphoma
stat, another HDAC inhibitor, was FDA approved for R/R
PTCL in 2014 and demonstrates similar activity to romidep- Advani R, Rosenberg SA, Horning SJ. Stage I and II follicular non-
Hodgkin lymphoma: long-term follow-up of no initial therapy.
sin. Alisertib is an aurora kinase inhibitor which has demon-
J Clin Oncol. 2004;22:1454-1459.
strates meaningful single agent activity in R/R PTCL and is
Ardeshna KM, Qian W, Smith P, et al. Rituximab versus a
under ongoing study. watch and wait approach in patients with advanced-stage
CD30 is expressed uniformly in ALCL but is also highly asymptomatic, non-bulky follicular lymphoma: an open label
restricted to neoplastic cells, making it an attractive target in randomised phase 3 trial. Lancet Oncol. 2014;15:424-435.
this disease. Studies are ongoing evaluating brentuximab Campbell BA, Voss N, Woods R, et al. Long-term outcomes
vedotin in the upfront setting with CHP, omitting the vin- for patients with limited stage follicular lymphoma. Cancer.
cristine because of overlapping toxicity. 2010;116:3797-3806.

Casulo C, Byrtek M, Dawson KL, et al. Early relapse of Rohatiner AZS, Nadler L, Davies AJ, et al. Myeloablative therapy
follicular lymphoma after rituximab plus cyclophosphamide, with autologous bone marrow transplantation for follicular
doxorubicin, vincristine, and prednisone defines patients at lymphoma at the time of second or subsequent remission: long
high risk for death: an analysis from the National LymphoCare term follow up. J Clin Oncol. 2007;25:2554-2559.
Study. J Clin Oncol. 2015;33:2516-2522. Rummel MJ, Niederle N, Maschmeyer G, et al. Bendamustine plus
Cheson BD, Clemens-Martin W, Pieper A, et al. Optimal use rituximab versus CHOP plus rituximab as first-line treatment
of bendamustine in chronic lymphocytic leukemia, non- for patients with indolent and mantle-cell lymphomas: an open-
Hodgkin’s lymphoma, and multiple myeloma: treatment label, multicentre, randomized, phase 3 non-inferiority trial.
recommendations from an international consensus panel. Clin Lancet. 2013;381:1203-1210.
Lymph Myeloma Leuk. 2010:10;21-27. Salles G, Seymour JF, Offner F, et al. Rituximab maintenance for 2
Dave SS, Wright G, Tan B, et al. Prediction of survival in follicular years in patients with high tumour burden follicular lymphoma
lymphoma based on molecular features of tumor-infiltrating responding to rituximab plus chemotherapy (PRIMA): a phase
immune cells. N Engl J Med. 2004;351:2159-2169. 3, randomized controlled trial. Lancet. 2010;377:42-51.
Federico M, Luminari S, Dondi A, et al. R-CVP versus R-CHOP Schouten HC, Qian W, Kvaloy S, et al. High-dose therapy
versus R-FM as first line therapy for advanced stage follicular improves progression free survival and survival in relapsed
lymphoma: final result of FOLL5 Trial conducted by the follicular non-Hodgkin’s lymphoma: results from the
Fondazione Italiana Linformi. J Clin Oncol. 2013;31:1506-1513. randomized European CUP trial. J Clin Oncol. 2003;21:
Flinn IW, van der Jagt R, Kahl, BS, et al. Randomized trial of 3918-3927.
bendamustine-rituximab or R-CHOP/R-CVP in first-line van Besien K, Loberiza FR, Bajorunaite R, et al. Comparison of
treatment of indolent NHL or MCL: the BRIGHT study. Blood. autologous and allogeneic hematopoietic stem cell transplantation
2014;123:2944-2952. for follicular lymphoma. Blood. 2003;102:3521-3529.
Friedberg JW, Taylor MD, Cerhan JR, et al. Follicular lymphoma
in the United States: first report of the national LymphoCare
Marginal zone lymphomas
study. J Clin Oncol. 2009;27:1202-1208.
Friedberg JW, Byrtek M, Link BK, et al. Effectiveness of first-line Arcaini L, Lucioni M, Boveri E, Paulli M, et al. Nodal marginal
management strategies for stage I follicular lymphoma: Analysis of zone lymphoma: current knowledge and future directions of an
the national lymphocare study. J Clin Oncol, 2012;30:3368-3375. heterogeneous disease. Eur J Haematol. 2009;83:165-174.
Gopal AK, Kahl BS, de Vos S, et al. PI3K inhibition by idelalisib in Ferreri A, Govi S, Pasini E, et al. Chlamydophila psittaci
patients with relapsed indolent lymphoma. N Engl J Med. 2014 eradication with doxycycline as first line targeted therapy for
Mar 13;370:1008-1018. ocular adnexae lymphoma: final results of an international
Hiddemann W, Kneba M, Dreyling M, et al. Frontline therapy phase II trial. J Clin Oncol. 2012;30:2988-2994.
with rituximab added to the combination of cyclophosphamide, Goda JS, Gospodarowicz M, Pintilie M, et al. Long-term outcome
doxorubicin, vincristine, and prednisone (CHOP) significantly in localized extranodal mucosa-associated lymphoid tissue
improves the outcome for patients with advanced-stage lymphomas treated with radiotherapy. Cancer. 2010;116:
follicular lymphoma compared with therapy with CHOP alone: 3815-3824.
results of a prospective randomized study of the German Low- Hermine O, Lefrère F, Bronowicki JP, et al. Regression of splenic
Grade Lymphoma Study Group. Blood. 2005;106:3725-3732. lymphoma with villous lymphocytes after treatment of hepatitis
Kahl BS, Bartlett NL, Leonard JP, et al. Bendamustine is effective C virus infection. N Engl J Med. 2002;347: 89-94.
therapy in patients with rituximab-refractory indolent B-cell Kalpadakis C, Pangalis GA, Angelopoulou MK, et al. Treatment of
non-Hodgkin lymphoma. Cancer. 2010;116:106-114. splenic marginal zone lymphoma with rituximab monotherapy:
Kahl BS, Hong F, Williams ME, et al. Rituximab Extended Progress report and comparison with splenectomy. Oncologist.
Schedule or Re-Treatment Trial (RESORT) for low tumor 2013;18:190-197.
burden follicular lymphoma: Eastern Cooperative Oncology Kuo SH, Yeh KH, Wu MS, et al. Helicobacter pylori eradication
Group Protocol E4402. J Clin Oncol. 2014;32:3096-3102. therapy is effective in the treatment of early-stage H pylori-
Khouri IF, McLaughlin P, Saliba RM, et al. Nonmyeloablative positive gastric diffuse large B-cell lymphomas. Blood.
allogeneic transplantation with our without 90 Yttrium 2012;119:4838-4844.
ibritumomab tiuxetan is potentially curative for relapsed Nakamura S, Sugiyama T, Matsumoto T, et al. Long-term clinical
follicular lymphoma: 12-year results. Blood. 2012;119:6373-6378. outcome of gastric MALT lymphoma after eradication of
Marcus R, Imrie K, Solal-Celigny P, et al. Phase III study of R-CVP Helicobacter pylori: a multicentre cohort follow-up study of 420
compared with cyclophosphamide, vincristine, and prednisone patients in Japan. Gut. 2012;61:507-513.
alone in patients with previously untreated advanced follicular Thieblemont C, Berger F, Dumontet C, et al. Mucosa-associated
lymphoma. J Clin Oncol. 2008;26:4579-4586. lymphoid tissue lymphoma is a disseminated disease in one
Rezvani AR, Storer B, Maris M, et al. Nonmyeloablative allogeneic third of 158 patients analyzed. Blood. 2000;95:802-806.
hematopoietic cell transplantation in relapsed, refractory, and Zucca E, Conconi A, Laszlo D, et al. Addition of rituximab to
transformed indolent non-Hodgkin’s lymphoma. J Clin Oncol. chlorambucil produces superior event-free survival in the
2008;26:211-217. treatment of patients with extranodal marginal zone B cell

Bibliography | 613
lymphoma: 5-year analysis of the IELSG-19 randomized study. Transformed lymphoma

J Clin Oncol. 2013;31:565-572.
Link BK, Maurer MJ, Nowakowski GS, et al. Rates and
outcomes of follicular lymphoma transformation in the
Waldenström macroglobulenemia immunochemotherapy era: a report of the University of Iowa/
Mayo Clinic SPORE Molecular Epidemiology Resourse. J Clin
Dimopoulos MA, Anagnostopoulos A, Kyrtsonis MC, et al. Onc. 2013;21:3272-3278.
Primary treatment of Waldenström macroglobulinemia with Villa D, Crump M, Panzarella T, et al. Autologous and allogeneic
dexamethasone, rituximab, and cyclophosphamide. J Clin stem cell transplantation for transformed follicular lymphoma:
Oncol. 2007;25:3344-3348. a report of the Canadian blood and marrow transplant group.
Dimopoulos MA, Kastritis E, Owen RG, et al. Treatment J Clin Oncol. 2013;31:1164-1171.
recommendations for patients with Waldenstrom
macroglobulinemia (WM) and related disorders: IWWM-&
consensus. Blood. 2014;124:1404-1411. Mantle cell lymphoma
Kyriakou C, Canals C, Cornelissen JJ, et al. Allogeneic
Delarue R, Haioun C, Ribrag V, et al. CHOP and DHAP plus
stem-cell transplantation in patients with Waldenström
rituximab followed by autologous stem cell transplantation
macroglobulinemia: report from the Lymphoma Working Party
in mantle cell lymphoma: a phase 2 study from the Groupe
of the European Group for Blood and Marrow Transplantation.
d’Etude des Lymphomes de l’Adulte. Blood. 2013;121:48-53.
J Clin Oncol. 2010;28:4926-4934.
Flinn IW, van der Jagt R, Kahl, BS, et al. Randomized trial of
Kyriakou C, Canals C, Sibon D, et al. High-dose therapy
bendamustine-rituximab or R-CHOP/R-CVP in first-line
and autologous stem-cell transplantation in Waldenstrom
treatment of indolent NHL or MCL: the BRIGHT study. Blood.
macroglobulinemia: the Lymphoma Working Party of the
2014;123:2944-2952.
European Group for Blood and Marrow Transplantation. J Clin
Giesler CH, Kolstad A, Laurell A, et al. Nordic MCL2 trial update:
Oncol. 2010;28:2227-2232.
six-year follow-up after intensive immunochemotherapy for
Treon SP, Branagan AR, Hunter Z, Santos D, et al. Paradoxical
untreated mantle cell lymphoma followed by BEAM or BEAC
increases in serum IgM and viscosity levels following
+ autologous stem-cell support: still very long survival but late
rituximab in Waldenström’s macroglobulinemia. Ann Oncol.
relapses do occur. Br J Haematol. 2012;158:355-362.
2004;15:1481-1483.
Goy A, Sinha R, Williams ME, et al. Single agent lenalidomide
Treon SP, Cao Y, Xu L, et al. Somatic mutations in MYD88
in patient with mantle cell lymphoma who relapse or progress
and CXCR4 are determinants of clinical presentation on
after or were refractory to bortezomib: phase II MCL-001
overall survival in Waldenström macroglobulinemia. Blood.
(EMERGE) study. J Clin Oncol. 2013;31:3688-3695.
2014;123:2791-2796.
Hermann A, Hoster E, Zwingers T, et al. Improvement of overall
Treon SP, Ioakimidis L, Soumerai JD, et al. Primary therapy
survival in advanced stage mantle cell lymphoma. J Clin Oncol.
of Waldenström macroglobulinemia with bortezomib,
2009;27:511-518.
dexamethasone, and rituximab: WMCTG clinical trial 05-180.
Hess G, Herbrecht R, Romaguera J, et al. Phase III study to
J Clin Oncol. 2009;27:3830-3835.
evaluate temsirolimus compared with investigator’s choice
therapy for the treatment of relapsed or refractory mantle cell
lymphoma. J Clin Oncol. 2009;27:3822-3828.
Hairy cell leukemia
Hoster E, Dreyling M, Klapper W, et al. A new prognostic index
Else M, Dearden CE, Matutes E, et al. Long-term follow-up (MIPI) for patients with advanced-stage mantle cell lymphoma.
of 233 patients with hairy cell leukaemia, treated initially Blood. 2008;111:558-565.
with pentostatin or cladribine, at a median of 16 years from Kluin-Neilmans HC, Hoster E, Hermine O, et al. Treatment of
diagnosis. Br J Haematol. 2009;145:733-740. older patients with mantle-cell lymphoma. N Engl J Med.
Grever MR. How I treat hairy cell leukemia. Blood. 2010;115:21-28. 2012;367:520-531.
Jones G, Parry-Jones N, Wilkins B, et al. Revised guidelines for the Martin P, Chadburn A, Christos P, et al. Outcome of deferred
diagnosis and management of hairy cell leukemia and hairy cell initial therapy in mantle-cell lymphoma. J Clin Oncol.
leukemia variant. Br J Haematol. 2012;156:186-195. 2009;27:209-213.
Pettirossi V, Santi A, Imperi E, et al. BRAF inhibitors reverse Robak T, Huang H, Jin J, et al. Bortezomib-based therapy for newly
the unique molecular signature and phenotype of hairy diagnosed mantle cell lymphoma. N Engl J Med. 2015;372:944-953.
cell leukemia and exert potent antileukemic activity. Blood. Romaguera JE, Fayad L, Feng L, et al. Ten-year follow-up after
2015;125:1207-1216. intense chemoimmunotherapy with rituximab-hyperCVAD with
Ravandi F, O’Brien S, Jorgensen J, et al. Phase 2 study of cladribine rituximab-high dose methotrexate/cytarabine (R-MA) and without
followed by rituximab in patients with hairy cell leukemia. stem cell transplantation in patients with untreated aggressive
Blood. 2011;118(14):3818-3823. mantle cell lymphoma. Br J Haematol. 2010;150: 200-208.
Tiacci E, Trifonov V, Schiavoni G, et al. BRAF mutations in hairy Rummel MJ, Niederle N, Maschmeyer G, et al. Bendamustine plus
cell leukemia. N Engl J Med. 2011;364:2305-2315. rituximab versus CHOP plus rituximab as first-line treatment

for patients with indolent and mantle-cell lymphomas: an open- double-hit lymphoma: a multicenter retrospective analysis.
label, multicentre, randomized, phase 3 non-inferiority trial. Blood. 2014;124: 2354-2361.
Lancet. 2013;381:1203-10. Pfreundschuh M, Schubert J, Ziepert M, et al. Six versus eight cycles
Wang M, Fayad L, Wagner-Bartak N, et al. Lenalidomide in of bi-weekly CHOP-14 with or without rituximab in elderly
combination with rituximab for patients with relapsed or patients with aggressive CD201 B-cell lymphomas: a randomised
refractory mantle-cell lymphoma: a phase 1-2 clinical trial. controlled trial (RICOVER-60). Lancet Oncol. 2008;9:105-116.
Lancet Oncol. 2012;13:716-723. Pfreundschuh M, Trumper L, Osterborg A, et al. CHOP-like
Wang L, Rule S, Martin P, et al. Targeting BTK with ibrutinib in chemotherapy plus rituximab versus CHOP-like chemotherapy
relapsed or refractory mantle cell lymphoma. N Engl J Med. alone in young patients with good-prognosis diffuse large-B-
2013;369:507-516. cell lymphoma: a randomised controlled trial by the MabThera
International Trial (MInT) Group. Lancet Oncol. 2006;7:379-391.
Pfreundschuh M, Poeschel V, Zeynalova S, et al. Optimization of
rituximab for the treatment of diffuse large B-cell lymphoma
Abramson JS, Hellmann M, Barnes JA, et al. Intravenous (II): extended rituximab exposure time in the SMARTE-
methotrexate as central nervous system (CNS) prophylaxis R-CHOP-14 trial of the german high-grade non-Hodgkin
is associated with a low risk of CNS recurrence in high- lymphoma study group. J Clin Oncol. 2014; 32: 4127-4133.
risk patients with diffuse large B-cell lymphoma. Cancer. Rieger M, Osterborg A, Pettengell R, et al. Primary mediastinal
2010;116:4283-4290. B-cell lymphoma treated with CHOP-like chemotherapy with
Barrans S, Crouch S, Smith A, et al. Rearrangement of MYC is or without rituximab: results of the Mabthera International
associated with poor prognosis in patients with diffuse large Trial Group study. Ann Oncol. 2011;22:664-670.
B-cell lymphoma treated in the era of rituximab. J Clin Oncol. Rosenwald A, Wright G, Chan WC, et al. The use of molecular
2010;28:3360-3365. profiling to predict survival after chemotherapy for diffuse
Coiffier B, Lepage E, Briere J, et al. CHOP chemotherapy plus large-B-cell lymphoma. N Engl J Med. 2002;346:1937-1947.
rituximab compared with CHOP alone in elderly patients with Scott DW, Mottok A, Ennishi D, et al. Prognostic significance of
diffuse large-B-cell lymphoma. N Engl J Med. 2002;346:235-242. diffuse large B cell lymphoma cell of origin determined by digital
Dunleavy K, Pittaluga S, Maeda LS, et al. Dose-adjusted EPOCH- gene expression in formalin-fixed paraffin-embedded tissue
rituximab therapy in primary mediastinal B-cell lymphoma. biopsies. J Clin Oncol. 2015;33:2848-2856.
N Engl J Med. 2013:368;1408-1416. Sehn LH, Scott DW, Chhanabhai M, et al. Impact of concordant
Ferreri AJ. How I treat primary CNS lymphoma. Blood. and discordant bone marrow involvement on outcome in
2012;118:510-522. diffuse large B-cell lymphoma treated with R-CHOP. J Clin
Gisselbrecht C, Glass B, Mounier N, et al. Salvage regimens with Oncol. 2011;29:1452-1457.
autologous transplantation for relapsed large B-cell lymphoma Shipp MA, Harrington DP, Anderson JR, et al. A predictive model
in the rituximab era. J Clin Oncol. 2010;28:4184-4190. for aggressive non-Hodgkin’s lymphoma. The International
Green TM, Young KH, Visco C, et al. Immunohistochemical Non-Hodgkin’s Lymphoma Prognostic Factors Project. N Engl
double-hit score is a strong predictor of outcome in patients J Med. 1993;329:987-994.
with diffuse large B-cell lymphoma treated with rituximab plus Siegal T, Goldschmidt N. CNS prophylaxis in diffuse large B-cell
cyclophosphamide, doxorubicin, vincristine, and prednisone. lymphoma: if, when, how and for whom? Blood Rev. 2012;26: 97-106.
J Clin Oncol. 2012;30:3460-3467. Villa D, Connors JM, Shenkier TN, Gascoyne RD, Sehn LH, Savage
Johnson NA, Savage KJ, Ludkovski O, et al. Lymphomas with KJ. Incidence and risk factors for central nervous system relapse
concurrent BCL2 and MYC translocations: the critical factors in patients with diffuse large B-cell lymphoma: the impact of
associated with survival. Blood. 2009;114:2273-2279. the addition of rituximab to CHOP chemotherapy. Ann Oncol.
Johnson NA, Slack G, Savage KJ, Connors JM, et al. Concurrent 2009;21:1046-1052.
expression of MYC and BCL2 in diffuse large B-cell lymphoma Ziepert M, Hasenclever D, Kuhnt E, et al. Standard International
treated with rituximab plus cyclophosphamide, doxorubicin, prognostic index remains a valid predictor of outcome for
vincristine and prednisone. J Clin Oncol. 2012;30(28): 3452-3459. patients with aggressive CD20+ B-cell lymphoma in the
Miller TP, Dahlberg S, Cassady R, et al. Chemotherapy alone rituximab era. J Clin Oncol. 2010;28:2373-2380.
compared with chemotherapy plus radiotherapy for localized
intermediate and high-grade non-Hodgkin lymphoma. N Engl J
Burkitt lymphoma
Med. 1998;339:21-26.
Persky DO, Unger JM, Spier CM, et al. Phase II study of rituximab Dunleavy K, Pittaluga S, Shovlin M, et al. Low intensity therapy in
plus three cycles of CHOP and involved-field radiotherapy adults with Burkitt’s Lymphoma. N Engl J Med. 2013:369;
for patients with limited-stage aggressive B-cell lymphoma: 1915-1925.
Southwest Oncology Group study 0014. J Clin Oncol. Magrath I, Adde M, Shad A, et al. Adults and children with small
2008;26:2258-2263. noncleaved-cell lymphoma have a similar excellent outcome
Petrich AM, Gandhi M, Jovanovic B, et al. Impact of induction when treated with the same chemotherapy regimen. J Clin
regimen and stem cell transplantation on outcomes in Oncol. 1996;14:925-934.

Bibliography | 615
HIV lymphomas in patients with refractory or relapsing extranodal NK/T-cell

lymphoma, a phase 2 study. Blood. 2011;117:1834-1839.
Dunleavy K, Little RF, Pittaluga S, et al. The role of tumor Kim SJ, Kim K, Kim BS, et al. Phase II trial of concurrent radiation
histogenesis, FDG-PET, and short-course EPOCH with dose- and weekly cisplatin followed by VIPD chemotherapy in
dense rituximab (SC-EPOCH-RR) in HIV-associated diffuse newly diagnosed, stage IE to IIE, nasal, extranodal NK/T-cell
large B-cell lymphoma. Blood. 2010;115:3017-3024. lymphoma: Consortium for Improving Survival of Lymphoma
Sparano JA, Lee JY, Kaplan LD, et al. Rituximab plus concurrent study. J Clin Oncol. 2009;27:6027-6032.
infusional EPOCH chemotherapy is highly effective in Lee J, Suh C, Park YH, et al. Extranodal natural killer T-cell
HIV-associated B-cell non-Hodgkin lymphoma. Blood. lymphoma, nasal-type: a prognostic model from a retrospective
2010;115:3008-3016. multicenter study. J Clin Oncol. 2006;24:612-618.
O’Connor OA, Pro B, Pinter-Brown L, et al. Pralatrexate
Posttransplant lymphoproliferative disorders in patients with relapsed or refractory peripheral T-cell
lymphoma: results from the pivotal PROPEL study. J Clin
Choquet S, Oertel S, LeBlond V, et al. Rituximab in the Oncol. 2011;29:1182-1189.
management of post-transplantation lymphoproliferative Prince HM, Whittaker S, Hoppe RT. How we treat mycosis
disorder after solid organ transplantation: proceed with caution. fungoides and Sezary syndrome. Blood. 2009.
Ann Hematol. 2007;86:599-607. Pro B, Advani R, Brice P, et al. Brentuximab vedotin (sgn-35)
Trappe R, Oertel S, Leblond V, et al. Sequential treatment with in patients with relapsed or refractory systemic anaplastic
rituximab followed by CHOP chemotherapy in adult B-cell large-cell lymphoma: results of a phase II study. J Clin Oncol.
post-transplant lymphoproliferative disorder (PTLD): the 2012;30:2190-2196.
prospective international multicentre phase 2 PTLD-1 trial. Sibon D, Fournier M, Briere J, et al. Long-term outcome of adults
Lancet Oncol. 2011;13:196-206. with systemic anaplastic large-cell lymphoma treated within the
Trappe R, Choquet S, Oertel SHK, et al. Sequential treatment with Groupe d’Etude des Lymphomes de l’Adulte trials. J Clin Oncol.
rituximab and CHOP chemotherapy in B-cell PTLD–moving 2012;30:3939-3946.
forward to a first standard of care: results from a prospective Sibon D, Fournier M, Briere J, et al. Prognostic factors and long
international multicenter trial. ASH Annual Meeting Abstracts. term outcome of 138 adults with systemic anaplastic large-
2009;114:100. cell lymphoma: a retrospective study by the Groupe d’Etude
Des Lymphomes De l’Adulte (GELA). ASH Annual Meeting
Abstracts. 2010;116:322-.
PTCL
Simon A, Peoch M, Casassus P, et al. Upfront VIP-reinforced-
Bazarbachi A, Suarez F, Fields P, Hermine O. How I treat adult ABVD (VIP-rABVD) is not superior to CHOP/21 in newly
T-cell leukemia/lymphoma. Blood. 2011;118:1736-1745. diagnosed peripheral T-cell lymphoma. Results of the
Bekkenk MW, Geelen FA, van Voorst Vader PC, et al. Primary and randomized phase III trial GOELAMS-LTP95. Br J Haematol.
secondary cutaneous CD30(+) lymphoproliferative disorders: 2010;151:159-166.
a report from the Dutch Cutaneous Lymphoma Group on the Swerdlow SH, Campo E, Harris NL, et al. WHO Classification of
long-term follow-up data of 219 patients and guidelines for Tumours of Haematopoietic and Lymphoid Tissue. 4th ed. Lyon,
diagnosis and treatment. Blood. 2000;95:3653-3661. France: IARC Press; 2008.
d’Amore F, Relander T, Lauritzsen GF, et al. Upfront autologous Tsukasaki K, Utsunomiya A, Fukuda H, et al. VCAP-AMP-VECP
stem-cell transplantation in peripheral T-cell lymphoma: compared with biweekly CHOP for adult T-cell leukemia-
NLG-T-01. J Clin Oncol. 2012. lymphoma: Japan Clinical Oncology Group Study JCOG9801.
Delabie J, Holte H, Vose JM, et al. Enteropathy-associated T-cell J Clin Oncol. 2007;25:5458-5464.
lymphoma: clinical and histological findings from the international Vose J, Armitage J, Weisenburger D. International T-cell Lymphoma
peripheral T-cell lymphoma project. Blood. 2011;118:148-155. Project. International peripheral T-cell lymphoma study: pathology
Delfau-Larue MH, de Leval L, Joly B, et al. Targeting findings and clinical outcomes. J Clin Oncol. 2008;25:4124-4130.
intratumoral B-cells with Rituximab in addition to CHOP in Willemze R, Jansen PM, Cerroni L, et al. Subcutaneous
angioimmunoblastic T-cell lymphoma. A clinicobiological panniculitis-like T-cell lymphoma: definition, classification, and
study of the GELA. Haematologica. 2012. prognostic factors: an EORTC Cutaneous Lymphoma Group
Di Sabatino A, Biagi F, Gobbi PG, Corazza GR. How I treat Study of 83 cases. Blood. 2008;111:838-845.
enteropathy-associated T-cell lymphoma. Blood. 2012;119: Yamaguchi M, Tobinai K, Oguchi M, et al. Phase I/II study of
2458-2468. concurrent chemoradiotherapy for localized nasal natural
Ishida T, Joh T, Uike N, et al. Defucosylated anti-CCR4 killer/T-cell lymphoma: Japan Clinical Oncology Group Study
monoclonal antibody (KW-0761) for relapsed adult T-cell JCOG0211. J Clin Oncol. 2009;27:5594-5600.
leukemia-lymphoma: a multicenter phase II study. J Clin Oncol. Yared J, Kimball A. The role of high dose chemotherapy and
2012;30:837-842. autologous stem-cell transplantation in peripheral T-cell
Jaccard A, Gachard N, Marin B, et al. Efficacy of L-asparaginase lymphoma: a review of the literature and new perspectives.
with methotrexate and dexamethasone (AspaMetDex regimen) Cancer Treat Rev. 2012.

CHAPTER
22
Chronic lymphocytic leukemia/
small lymphocytic lymphoma
Clive S. Zent and Timothy G. Call
Epidemiology, 617 Risk stratification, 621 B-cell prolymphocytic leukemia, 633
Biology, 617 Management, 624 Bibliography, 634
Diagnosis and clinical evaluation, 619 Complications of CLL, 631

expression profile, suggesting a common cell of origin for
multimedia component. CLL. Functional, immunophenotype, and gene expression
data suggest that CLL is most closely related to the CD5+ B-1
B-cell subpopulation. In human adults, B-1 cells constitu-
Epidemiology tively produce polyreactive antimicrobial (natural) antibod-
ies that are an important component of innate immunity.
Chronic lymphocytic leukemia/small lymphocytic lym-
phoma (CLL) is the most prevalent lymphoid malignancy in
North America, Europe, and some other regions of the Etiology
world. The estimated incidence of CLL in the United States is
The cause of CLL remains unknown. There is considerable
~5.3:100,000 persons per year (>15,000 new diagnoses per
evidence to suggest a genetic predisposition to the disease.
year), and the incidence of CLL increases with age. The esti-
The risk of CLL in diverse populations is highly variable with
mated prevalence of CLL in the US is 120,000-140,000 per-
the highest risk in populations with northern European
sons. The median age of diagnosis is ~72 years, with a male
genetic heritage and considerably lower incidence in popula-
to female ratio of ~2:1. CLL is a familial disease in about
tions of East Asian genetic heritage, irrespective of where they
5%-10% of patients. SLL is a variant of CLL discussed in the
live. In addition, for the 5%-10% of patients with familial
section “Diagnosis” later in this chapter.
CLL, their first-degree relatives have a significantly increased
risk (~7-fold) of developing CLL or another B-cell malig-
Biology nancy. However, the course of familial CLL in individuals
with the disease is not determined by familial status, suggest-
Cell of origin ing that the familial component pertains only to the risk of
acquiring the disease. Genomewide genetic studies in familial
CLL is an indolent malignancy of mature B cells. The cell of
CLL cohorts have implicated multiple germ line genetic poly-
origin of CLL is not fully defined. CLL cells from patients
morphisms, and it is unlikely that CLL predisposition in
with somatically hypermutated and unmutated immuno-
most familial or sporadic cases is related to a single genetic
globulin heavy-chain variable region (IGHV) (see section
defect. Although the possibility has been extensively studied,
“Pathophysiology” later in this chapter) have a similar gene
there are no validated environmental risk factors for CLL.
Conflict-of-interest disclosure: Dr. Zent: research funding from

Novartis, Biothera, GlaxoSmithKline, Genzyme, and Genentech. Pre-CLL conditions
Dr. Call: no conflicts of interest to declare.
Off-label drug use: Idelalisib, ibrutinib, ofatumumab, pentostatin, Aging is associated with major changes in humoral immu-
cladribine. nity, including decreased antibody repertoire and increased
| 617

618 | Chronic lymphocytic leukemia/small lymphocytic lymphoma
frequency of oligoclonal B-cell populations. When people outcome. Patients with mutated IGHV tend to have CLL
60 years and older with normal complete blood counts are cells that are anergic while patients with unmutated CLL
screened with high sensitivity flow cytometry, >5% have a have cells that are responsive to BCR cross linking in vitro.
small circulating monoclonal B-cell population. Most of However, the relationship between IGHV mutation status,
these monoclonal B-cells are not associated with an BCR activation, and CLL disease biology is not yet fully
increased absolute lymphocyte count (lymphocytosis), are understood.
termed screening monoclonal B cell lymphocytosis (MBL) CLL cells have apoptotic defects that contribute to
and are of unknown clinical importance. The biological increased survival in the stromal microenvironment of the
importance of this form of MBL and its relationship to the lymphoid tissues and bone marrow. Important components
development of CLL is under investigation. However, the of apoptosis resistance are upregulation of the anti-apoptotic
high prevalence of this nonclinical MBL suggests that the molecules BCL2 and MCL1. The molecular mechanisms of
development of CLL is a stepwise process affecting only a these defects are not fully understood. However, 13q14
small percentage of patients with a preexisting monoclonal deletion, the most common defect detected in CLL by inter-
population of B-cells. This screening MBL is clearly differ- phase fluorescent in situ hybridization (FISH), results in the
ent from MBL detected clinically in patients with deletion of genes coding for the inhibitory microRNAs
lymphocytosis. (mIR) mIR15 and mIR16 that downregulate expression of
the BCL2 gene. The mechanism by which BCL2 expression
is upregulated in CLL patients without 13q14 deletion is not
Pathophysiology
known.
CLL is a disease characterized by progressive accumulation Defects in the DNA damage repair pathway in CLL cells
of monoclonal B-cells that preferentially grow in the prolif- are associated with more aggressive disease, cause resistance
eration centers (pseudofollicles) of lymph nodes with an to DNA damaging chemotherapies, and increase the risk of
overall tumor cell proliferation rate of 0.1%-1% per day and disease transformation. These defects are an important but
prolonged overall cell survival (~3-6 months) because of rare event in CLL patients at diagnosis (<10%). They increase
defective apoptosis. An important driver of CLL survival and in frequency with disease progression and occur in approxi-
growth is B-cell receptor (BCR) signaling and multiple mately 50% of patients refractory to therapies containing
mechanisms of sustained activation of the BCR in CLL have DNA damaging chemotherapy. TP53 defects disrupting p53
been described. Antigen binding specificity of BCR is deter- protein function occur either because of loss of one allele of
mined by the composition of the variable regions of the TP53 by 17p13 deletion and a dysfunctional mutation in the
immunoglobulin molecule; and the stereochemistry of this remaining TP53 allele, biallellic dysfunctional mutations, or
region is remarkably similar to at least one other patient with a single dominant negative mutation. Disruption of ATM
CLL in ~30% of cases, an event that is has a very low proba- function can also result in a defective DNA damage repair
bility of occurring by chance. This stereotypy suggests a pathway in CLL cells. One allele of ATM is lost in the 11q22.3
restricted VDJ and VJ allele use or selection by common deletion, and complete loss of function of ATM in these cells
antigens. Studies have shown that CLL idiotypic antibodies can occur because of disruptive mutations in the remaining
frequently react to autoantigens including antigenic targets allele. Loss of ATM function can also occur because of bial-
on apoptotic cells, tend to be polyreactive, and in some cases lelic disruptive ATM mutations.
can even be activated by self-epitopes. These findings pro- The pathophysiological effects of CLL cells are complex
vide important insights into the biology of CLL and have and not fully understood. Accumulation of CLL cells in the
also identified the BCR and its signaling pathway as a thera- lymph nodes, spleen, and liver cause enlargement and dis-
peutic target. ruption of function of these and possibly surrounding
Antigen-responsive B lymphocytes in the germinal center organs. Bone marrow infiltration and the effects of CLL cells
can be induced to undergo antigen-driven somatic hyper- on myelopoietic cells and the bone marrow microenviron-
mutation of the immunoglobulin genes, which alters epitope ment can decrease hematopoiesis resulting in cytopenias.
affinity for antigen. Somatic hypermutation of the variable CLL cells have an early detrimental effect on normal immune
region of IGHV is defined as ≥2% sequence difference from function. This results in impaired immunological response
germ line, and occurs in >50% of patients with CLL. CLL to infection, defective immunological self-recognition, and
patients with these “mutated” IGHV generally have a less possibly defective immune surveillance for other malignan-
aggressive disease course and better overall survival. In con- cies. The mechanism of the constitutional effects of progres-
trast, patients with CLL cells that have not undergone sive CLL, including weight loss, drenching night sweats,
somatic hypermutation of IGHV (so-called “unmutated” fevers, and fatigue, are not fully understood but could be the
CLL) generally have a more aggressive disease and poorer result of dysregulated cytokine production.

Diagnosis and clinical evaluation | 619
expression usually parallels that of the light chain. Low CD20

Key points expression can be confirmed by a negative study with the low
• CLL is the most prevalent lymphoid malignancy in North affinity CD20 binding antibody FMC7. If the monoclonal B
America. cells do not have the typical CLL immunophenotype (mono-
• The incidence of CLL increases with age. clonal B cells with that are CD20 dim, light chain dim, CD5+/
• Risk of CLL is higher in populations of Northern European CD23+), a wide differential diagnosis needs to be considered
heritage, and CLL is relatively uncommon in Asia. (Table 22-1). The leukemic phase of mantle cell lymphoma
• CLL is a familial disease in 5%-10% of patients. Familial CLL is an important consideration and can be evaluated by FISH
does not increase the risk of a more aggressive disease course.
analysis for t(11;14).
The International Workshop on Chronic Lymphocytic Leu-
kemia (IWCLL) published a revised version of the previous
Diagnosis and clinical evaluation 1996 National Cancer Center-Working Group guidelines in
2008 (IWCLL 2008). The IWCLL 2008 guidelines require at
Presentation least one of the following criteria to make a diagnosis of CLL in
CLL, including the SLL variant, is usually diagnosed on eval- a patient with a documented CLL immunophenotype mono-
uation of an incidental finding of asymptomatic leukocyto- clonal B cell population: (i) peripheral blood B cell count of at
sis/lymphocytosis or lymphadenopathy/splenomegaly. Only least 5 × 109/L; (ii) lymphadenopathy or splenomegaly on
~20% of patients have symptomatic disease at diagnosis. physical examination or CT scanning; (iii) CLL associated
CLL can present with symptomatic anemia, bleeding due to cytopenia; (iv) CLL related symptoms. Patients with CLL,
thrombocytopenia (~5%), symptomatic adenopathy or sple- based on these criteria, who have a B-cell count <5 × 109/L are
nomegaly (abdominal distention or early satiety), or consti- considered to have the small lymphocytic lymphoma (SLL)
tutional symptoms. Constitutional symptoms include variant of the disease. Patients with clinical lymphocytosis and
profound fatigue, drenching night sweats, fevers, and invol- a circulating monoclonal B-cell population with CLL immu-
untary weight loss. nophenotype who do not meet these criteria are considered to
In the contemporary era, when patients are diagnosed ear- have clinical MBL. Assessing the B-cell counts in the peripheral
lier than in historical series, physical examination is often blood requires quantitative flow cytometric immunopheno-
negative at diagnosis. Possible findings include firm rubbery typing that may not be readily available in all clinical settings.
nontender lymphadenopathy, which is frequently symmetri- However, in studies of patients with early stage CLL, an abso-
cal, and palpable liver or spleen enlargement. lute lymphocyte count of over 11 × 109/L is highly predictive of
an absolute B-cell count > 5 × 109/L.
Diagnosis
Peripheral blood lymphocyte morphology
Peripheral blood flow cytometry
CLL cells have a similar appearance to normal small lympho-
In ~90% of patients the diagnosis of CLL can be made by cytes. CLL cells have increased cell membrane fragility and
immunophenotypic characterization of peripheral blood tend to break during the process of making a blood smear giv-
lymphocytes by flow cytometry. B-cell clonality is deter- ing rise to smudge cells which are characteristic but not
mined by demonstrating light chain exclusion in the B pathognomonic for CLL (Figure 22-1). A subset of circulating
(CD19+) lymphocytes. CLL cells characteristically have dim CLL cells can also have prolymphocytic morphology. Higher
CD20 and dim light-chain expression and coexpress CD5 percentages of prolymphocytes in the periphe ral blood
and CD23. CD79b is a component of the BCR and (>55%) of patients with immunophenotypically diagnosed
Table 22-1 Chronic B-cell

Disease sIg CD20 CD5 CD23 CD10 CD103
lymphoproliferative disorders:
prototypic immunophenotype Chronic lymphocytic leukemia dim dim + + - -
Lymphoplasmacytic lymphoma + + -/+ -/+ - -
Mantle cell lymphoma + + + -/dim - -
Nodal marginal zone lymphoma + + - −/+ - -
Splenic marginal zone lymphoma + + −/+ −/+ − -/+
Follicular lymphoma + + − −/+ +/− −
Hairy cell leukemia + + − − − +
B cell prolymphocytic leukemia + + −/+ - - -

adequate tissue for architectural analysis of the lymphoid tis-

sue. The pathognomonic characteristic of CLL is prolifera-
tion centers (pseudofollicles) (Figure 22-2).
Bone marrow study
Bone marrow study is rarely required for the diagnosis of

CLL. Lymphoid tissue is preferable to bone marrow for diag-
nostic purposes in patients with a nondiagnostic flow cytom-
etry immunophenotype. Bone marrow studies can be helpful
in assessing cytopenias found in conjunction with the diag-
nosis of CLL, but are otherwise of little value.
Differential diagnosis
Figure 22-1 A peripheral blood smear of a patient with CLL (Giemsa The differential diagnosis of leukemic phase B-cell malignan-
stain; magnification ×400) shows small lymphocytes and numerous cies with small- to moderate-sized circulating lymphocytes
smudge cells. with mature morphology (chronic B-cell lymphoproliferative
disorders) includes CLL, mantle cell lymphoma, splenic
CLL have previously been considered to be indicative of marginal zone lymphoma, nodal marginal zone lymphoma,
transformation to “secondary” B-cell prolymphocytic leuke- lymphoplasmacytic lymphoma, hairy cell leukemia, and
mia (PLL). However, this finding could indicate clonal evolu- B-cell prolymphocytic leukemia. These B-cell lymphoprolif-
tion of CLL with a MYC translocation or other adverse event erative disorders can have distinct immunophenotypes
rather than transformation to a distinct second disease. (Table 22-1), but a definitive diagnosis can require additional
testing (eg, FISH for t(11;14) for mantle cell lymphoma) or a
diagnostic lymph node biopsy.
Lymph node biopsy
Lymph node biopsy is required for diagnosis of CLL only in

Staging
patients with an immunophenotype that is atypical for CLL
or patients without detectable monoclonal B-cells on periph- Clinical staging using clinical evaluation and the complete
eral blood flow cytometry. Patients who require this proce- blood count (Table 22-2) is useful for categorizing patients
dure should have an excision or wide incisional biopsy and identifying the small subpopulation of patients with
because fine needle aspiration biopsy often does not provide advanced stage disease that require therapy diagnosis.
A B

Figure 22-2 Section of lymph node (hematoxylin and eosin stain) from a patient with CLL. (A) Low-magnification photomicrograph (×20)
showing proliferation centers (pseudofollicles). (B) High-magnification photomicrograph (×400) of a proliferation center showing central large
lymphocytes rimmed by small lymphocytes.

Risk stratification | 621
Table 22-2 Clinical staging

Binet classification Rai classification
Stage Definition Risk group Stage Definition
A <3 lymphoid areas Low 0 Lymphocytosis only
B >3 lymphoid areas Intermediate I Lymphadenopathy
II Hepato- or splenomegaly
Hemoglobin <10 g/dL or High III Hemoglobin <11 g/dL
C
platelets <100 × 109/L IV Platelets <100 × 109/L
However, these methods are of limited predictive value for Karyotype analysis is a useful method of detecting chro-
the majority of patients who have earlier stage disease. The mosomal defects in dividing cells. Its ability to provide
advances in the treatment of CLL over the past 25 years have genetic information for CLL patients is limited by the low
also resulted in considerable improvement in survival com- level of cell division in CLL cells especially from patients with
pared to that reported in the original staging studies. earlier stage disease. CLL cells can be induced to divide in
vitro using mitogens and Toll Like Receptor (TLR) agonists,
but these methods can cause artifacts and are not universally
Key points available. Although there has been recent interest in the use
• The IWCLL 2008 criteria for the diagnosis of CLL require an
of complex karyotype to predict the prognosis of patients
absolute B-cell count of at least 5 × 109/L. with advanced stage CLL with treatment refractory disease,
• A FISH probe for t(11:14) can help distinguish mantle cell the clinical role of these data still needs to be established.
lymphoma from CLL. FISH analysis provides an accessible method of testing
• A bone marrow biopsy is not required to diagnose CLL. CLL cells for known chromosome defects using specific
probes. The prognostic value of these data has been exten-
sively studied and a hierarchical approach to ranking risk is
clinically useful. Currently used genetic profiles use probes
Risk stratification
for 17p13 (TP53 locus), 11q22.3 (ATM locus), trisomy 12,
and 13q14 (miR15A and miR16-1 loci). The hierarchical
Patients with CLL have a highly variable clinical course and
stratification for risk of disease progression is 17p13 dele-
outcome. Although the median time from diagnosis to first
tion > 11q22.3 deletion > trisomy 12 > 13q14 deletion. Alth
treatment is 5-7 years and median survival is >10 years, the
ough this methodology is currently being modified by the
wide ranges for these parameters limits the clinical utility of
addition of data from gene sequencing, the model continues
these data to plan patient management and provide accurate
to have clinical utility. Inclusion of a probe for 14q32 (IGH
prognostic estimates. Because most CLL patients are now
locus) can be useful for discrimination between CLL and
diagnosed with earlier stage disease, there is an important
mantle cell lymphoma in leukemic phase. In addition, tra
need for better prognostic markers. The most useful prog-
nslocations involving IGH do occur in a small subpopu
nostic markers available utilize the biological characteristics
lation of patients with CLL and could have prognostic
of the patient’s CLL cells.
significance.
Data from FISH analysis are limited by the probe set and
sensitivity of the assay. Most laboratories analyze all nucle-
Genetic analysis
ated cells in the submitted sample. In early stage CLL, when
CLL is characterized by recurrent genetic abnormalities that the percentage of CLL cells in a blood specimen can be low,
can be used to predict disease biology. The most commonly subclonal populations with a specific genetic defect can be
used analysis is FISH, which is a reliable and relatively sensi- present at a percentage below the detection threshold of
tive method of detecting specific chromosomal abnormali- FISH analysis (generally ~5%). In most patients, peripheral
ties in interphase cells. This methodology has been blood is the preferred sample for analysis. Bone marrow
complemented by the clinical availability of conventional aspirates usually contain a large number of nucleated red
sequencing methods to detect abnormalities in individual blood cell precursors that decrease assay sensitivity.
genes of interest, and this will likely be further expanded by Gene sequencing has considerably improved the precision
the ability of next generation sequencing and array-based of analysis of genetic defects in the DNA damage repair path-
technologies to provide rapid and affordable gene testing in way in CLL. 17p13 deletion resulting in loss of one allele
the near future. This discussion will focus on methodologies of TP53 and 11q22.3 deletion resulting in loss of one allele
that are currently clinically available. of ATM usually only affect one chromosome, and the

consequences of these deletions depend largely on the func- Clonal evolution and selection
tional integrity of the remaining allele of TP53 or ATM
The CLL cell population frequently contains genetically
respectively. Patients with 17p13 deletion (~5% of CLL at
defined subclones with the potential to expand and alter the
diagnosis, but more common in later disease stages) have a
course of the disease. Serial analysis with FISH showed that
~80% rate of dysfunctional mutations in the remaining TP53
the apparent rate of detection of new subclones (clonal evolu-
allele leading to loss of p53 function in those cells. In addi-
tion) in an initially untreated CLL population was ~5% per
tion, disruption of p53 function in CLL can occur because of
year. Subsequent studies using considerably more sensitive
dysfunctional mutations in TP53 in the absence of 17p13
(<1% allele frequency) next generation sequencing (NGS)
deletion in ~5% of patients with CLL. These mutations can
methods have shown a high rate of small subclones of cells
result in loss of p53 function because they are biallelic, asso-
with adverse genetic defects in previously untreated CLL
ciated with uniparental disomy, or because the gene product
patients. These data suggest that progression of CLL can be
is dominant negative and thus inhibits the activity of remain-
associated with clonal expansion and clonal evolution. The
ing normal p53. Patients with 11q22.3 deletion (~10% of
role of evaluation of clonal complexity of the CLL cell popu-
CLL at diagnosis) have a ~30% rate of dysfunctional muta-
lation in clinical management is currently being investigated.
tions in the remaining ATM allele resulting in loss of ATM
function and poor prognosis. Patients with 11q22.3 deletion
that retain a wild type ATM have a better prognosis than BCR analysis
patients with loss of ATM function, but still have an inferior The BCR signaling essential for CLL cell survival and prolif-
outcome compared to most patients with a monoallelic dys- eration (Figure 22-3) can be modulated by IGHV somatic
functional ATM mutation. This suggests that the 11q22 dele- hypermutation and stereotypy. IGHV somatic hypermuta-
tion results in loss of additional genes (eg, BIRC3) that can tion and VH family usage can be determined by standard
have adverse effects on prognosis. sequencing in the clinical laboratory and does not change
Genomewide analysis of CLL cells has considerably during the course of disease in CLL. Somatic hypermutation
improved our understanding of the molecular genetics of with gene sequence having <98% identity to germ line
CLL. These studies identified several additional genes (mutated, ~55% of patients) is generally associated with less
including NOTCH1 and SF3B1 with recurrent mutations in aggressive disease and longer survival compared to patients
CLL. Activating mutations of NOTCH1 are detected in with ≥98% identity to germ line (unmutated, ~45% of
~10% of patients with CLL at diagnosis and these patients patients). Patients with 97%-98% germ line identity should
have more aggressive disease and a significantly increased be considered to have borderline mutation because of the
risk of transformation to diffuse large B-cell lymphoma potential errors in sequencing and the arbitrary nature of the
(DLBCL). Dysfunctional mutations in the gene coding for 2% cutoff that is used because of the difficulty in distinguish-
the splicing factor 3b subunit (SF3B1) of the spliceosome ing between mutations and unknown single nucleotide poly-
occur in ~10% of CLL patients at diagnosis and are associ- morphisms (SNPs). The exception to this finding is patients
ated with decreased duration of response to therapy and with IGHV utilizing VH3-21 who have a poorer prognosis
decreased overall survival. irrespective of mutation status.
Conventional sequencing for TP53 mutations is clinically
available and covers >90% of known defects in CLL. Use of
Stereotypy
this assay can increase the detection of TP53 disruption in
CLL at diagnosis. The ATM gene is very large and clinical The shape of the immunoglobulin antigen-binding site can
sequencing is currently not available but could be in the near be inferred from analysis of the IGHV sequence. Antigen
future. Conventional sequencing analysis is clinically avail- binding sites with high homology to previously described
able for analysis of NOTCH1 and SF3B1 mutations in sites occur in ~30% of patients. These quasi-identical or ste-
patients with CLL. Current limitations to clinical use of gene reotyped BCR can be classified according to their shape into
sequencing in the routine evaluation of CLL patients at diag- one of 19 major subsets, each of which have prognostic
nosis are the cost of these assays, the still limited data on the implications. The clinical implication of stereotypy on man-
utility of the more recently discovered prognostic mutations, agement of CLL patients continues to be defined.
and the absence of well validated methods of integrating the
data into a predictive model. In addition, the sensitivity of
Protein expression
standard sequencing methods is ~10% allele frequency
which limits the ability to detect small subclones of CLL cells CLL cells can be analyzed for proteins that are differentially
which could have deleterious consequences in patients with expressed in populations of patients with higher or lower
early stage disease. risk of CLL progression.

Risk stratification | 623
Antigen
BCR
CD79 Extracellular
A B A B
PIP2 PIP3
PI3Kδ AKT Intracellular
P P mTOR p70S6K
P P P P
LYN SYK LYN SYK
P P P BTK P PLCγ2
P S6
SOS DAG
Fostamatinib Idelalisib Ibrutinib

GS-9973 IPI-145 CC-292
AMG-319 ONO-4059
RAS PKCβ CARD11
ACP-196
BCL10
MALT1
MEK
IKK
ERK
NF-κB
Proliferation, survival, maturation, migration
Figure 22-3 The B-cell receptor (BCR) comprises the idiotypic immunoglobulin and accessory signalling molecules Ig (CD79A) and Igβ
(CD79B). BCR activation induces signalling via a series of molecules to activate transcription factors (eg, NF-κB) that promote cellular survival
and proliferation. Signalling requires phosphorylation (P in purple circles) by protein kinases (pink symbols) and the lipid kinase PI3Kδ (green
symbol). Sites of pathway inhibition by targeted kinase inhibitors are shown. Redrawn from Wiestner A. Hematology Am Soc Hematol Educ
Program. 2014;2014:125-134.
ZAP70 is expressed by some normal and malignant B-cells somatic hypermutation. Population studies show that
during differentiation and maturation and has a role in BCR higher levels of CD38 correlate with more aggressive disease
signaling in CLL cells. ZAP70 assays were initially proposed and poorer outcome. However, the level of CD38 is not
as a surrogate marker for IGHV mutation. However, subse- constant in patients with CLL, and there is difficulty deter-
quently studies showed a poor correlation (~70%) between mining the best cut off for this continuous variable for risk
these parameters, which is not clinically useful. However, stratification.
higher levels of expression of ZAP70 are an independent CD49d is the 4-integrin subunit that can associate with
marker of more aggressive CLL. Clinical use of this prognos- CD29 to form the 4b1-integrin (VLA-4). VLA-4, which is
tic factor has been limited because accurate quantification of expressed by B-cells including CLL, binds VCAM-1
intracellular proteins by flow cytometry in clinical laborato- (expressed by endothelial cells and bone marrow stromal
ries is technically difficult. cells) and the extracellular matrix molecule fibronectin.
CD38 is a multifunctional surface molecule expressed by VLA-4 has an important role in trafficking through the
hematopoietic cells including B-cells during maturation. endothelium required to home to the lymph nodes and bone
CD38 is ligand of CD31 (PECAM1) and also has enzymatic marrow. In CLL patients, increased CD49d expression is
activity important for calcium metabolism and can interact associated with a shorter time to first treatment and poorer
with the BCR/CD19 complex in B-cells. CD38 expression overall survival. Expression levels of CD49d are reported to
by circulating CLL cells correlates with the rate of cellular be stable over time in individual patients. The availability
turnover. CD38 is not a reliable surrogate marker for IGHV and reporting of CD49d in clinical practice is variable.

Serum measurements patients with more aggressive disease and poor outcome. As
data accumulate, accurate risk stratification using a selection
β2-Microglobulin (B2M) is a polypeptide associated with
of prognostic factors could be developed to predict time to
HLA I on the cell membrane. Serum levels can be increased
first treatment, progression free survival, and overall survival
in several hematological malignancies including CLL.
as well as the risk of transformation to diffuse large B-cell
Increased serum B2M levels that exceed 2× the upper limit
lymphoma. In clinical practice, the most useful currently
of normal are associated with increased CLL tumor burden
available prognostic factors are FISH analysis, TP53 muta-
and shorter treatment-free and overall survival. B2M is
tion analysis, and IGHV somatic hypermutation analysis.
metabolized in the kidneys and levels are increased in
Predictive risk models using these and other parameters have
patients with renal impairment. Adjustment of the B2M
been proposed. TP53 disruption analysis is currently the
level for calculated glomerular filtration rate has been
only predictive marker in patients with CLL used to deter-
reported to improve its value as a prognostic factor.
mine treatment modality. Additional prognostic factor use
depends on which assays are clinically available and depend-
CLL cell kinetics able and the practice requirements and expectations.
Lymphocyte doubling time (LDT) is an estimate of time
taken for a patient’s absolute lymphocyte count (ALC) to Key points
double. A clinically useful value requires a baseline ALC >
• TP53 disruption in CLL is associated with inferior prognosis
15 × 109/L and 2 weekly counts over a period of at least
and resistance to chemoimmunotherapy.
2 months. The LDT should then be calculated using linear
• Patients with IGHV somatic hypermutation have superior
regression. The initial studies in small patient cohorts survival compared to those without somatic hypermutation.
reported in the 1980s concluded that a LDT of <12 months • CD38 levels can be variable over the course of the disease.
was associated with poorer prognosis. However, ALC is a • B2M levels may be elevated in the presence of renal
labile parameter that is poorly predictive of the total tumor impairment.
burden in CLL (<10% of CLL cells are in the circulation at
any one time) and LDT should not be used as the sole param-
eter to predict a patient’s prognosis or initiate treatment. Management
CLL cellular kinetics can be more accurately measured in
vivo with nonradioactive deuterated (“heavy”) water. These Management of CLL is in rapid flux because of more accu-
studies have shown that CLL birth rates range from 0.1% to rate and earlier diagnosis, better risk stratification, recogni-
over 1% of the entire clone per day and that those patients tion of complications and methods to prevent them, and the
with birth rates of over 0.35% per day had a higher risk of development of highly effective targeted therapies and
disease progression. Of note, the level of clonal growth was immunotherapy.
not predictive of changes in the ALC. A recent follow up At present there is no proven benefit to early treatment of
study presented in abstract form confirmed these findings patients with CLL. However, earlier diagnosis does allow for
and suggested that early measurement of cellular kinetics implementation of an active management plan to prevent
can predict time to treatment. complications of disease, early management of complica-
tions, and appropriate timing of treatment. Patients need to
Prognosis at diagnosis be well educated about their disease, the clinical manifesta-
tions of disease progression and complications, precaution-
Developing an accurate and accessible prognostic evaluation ary measures (see section “Complications of CLL” in this
system in newly diagnosed early-intermediate stage CLL chapter), and measures to improve general fitness. The inter-
patients has been challenging because of our rapidly chang- val of patient follow up can be determined by using clinical
ing understanding of the biology of the disease, the large monitoring and risk factor analysis. The indications for ini-
number of potential prognostic factors, limitations of some tiation of treatment of progressive disease in both previously
of the published studies, and the indolent nature of the dis- untreated patients and those with relapsed/refractory disease
ease which frequently makes the results of clinical studies of are based on the IWCLL 2008 guidelines.
novel prognostic factors redundant before they are com-
pleted. In addition, factors such as TP53 disruption, which
Monoclonal B-cell lymphocytosis
occurs in <10% of patients at diagnosis, are detected at low
frequency with currently-used clinical assays but are subse- Patients with an incidental detection of a monoclonal
quently responsible for a disproportionate number of B-cell population without lymphocytosis, paraproteinemia,

Management | 625
lymphadenopathy, or visceromegaly do not need further anemia) need to be carefully excluded from this evaluation.
investigation or follow up. Patients with MBL with CLL Fit patients should have an ECOG performance score (PS) of
immunophenotype and lymphocytosis that do not meet the 0-1, no evidence of organ impairment, and no major com
criteria for diagnosis of CLL are considered to have clinical orbidity. Patients who are unfit with PS ≥3, major organ
MBL with an annual ~1-2% risk of progression to CLL failure, or limiting comorbidity should be considered for
requiring treatment. These patients should be actively moni- supportive and palliative care. Patients intermediate between
tored in same way as patients with early stage CLL. These fit and unfit (less fit) need to be considered for therapy
patients have an increased incidence of infections and skin options with lower toxicity.
cancers and the same monitoring as outlined below for CLL
patients should be followed.
CLL biological risk evaluation
Initial treatment of progressive CLL As detailed above, TP53 disruption by deletion ± mutation
(17p-/TP53mutation) predicts poor response to chemoimmu-
Indications for treatment notherapy and is an indication for alternative treatment
Patients are considered to have active progressive disease approaches. CLL patients can have long intervals between
requiring treatment if they have symptomatic disease, rapid diagnosis and treatment during which their CLL biology can
disease progression, or bone marrow failure (Table 22-3). be altered by either subclonal selection or new mutations
(clonal evolution). In patients without a previously demon-
strated TP53 disruption, FISH with a probe for 17p13 and
Evaluation of fitness TP53 sequencing should be performed within 6 months of
Choice of treatment for an individual patient depends on the initiation of treatment.
biology of the patient’s disease and their physical fitness.
Physical fitness should be determined using a minimum of Evaluation of CLL disease burden
standard evaluations of organ function (eg, estimated creati-
nine clearance) and performance status. Fitness for treatment Evaluation of CLL disease burden before initiation of treat-
should be assessed on an individual basis rather than by using ment is useful for planning therapy and evaluating response.
chronological age alone. The role of more sophisticated Patients require a clinical evaluation of disease burden based
methods of quantifying comorbidity and physical fitness on symptoms and physical exam with bidimensional mea-
such as the cumulative illness rating scale (CIRS) are still surement of the largest lymph node in the cervical, axillary,
investigational. Decreased fitness caused by potentially and inguinal/femoral regions on each side and the size of the
reversible CLL induced causes (eg, fatigue and symptomatic liver and spleen (measured as centimeters below the costal
Table 22-3 Indications for initiation of treatment in CLL (IWCLL 2008)
Indication Description Precautions

Bone marrow Anemia (Hb <11 g/dL) ± thrombocytopenia (<100 × 109/L) Require bone marrow study to confirm bone marrow
failure failure
Symptomatic Unintentional weight loss >10% during the past 6 months Exclude other causative pathologies,* eg, sleep disorder,
disease Fatigue: ECOG performance status ≥2 depression, hypothyroidism, chronic infection/
Fevers >38°C not due to infection for ≥2 weeks inflammation
Drenching night sweats
Splenomegaly Massive (>6 cm below costal margin) or symptomatic
(abdominal distention, early satiety, pain)
Lymphadenopathy Massive (>10-cm maximum diameter) or symptomatic Exclude infectious lymphadenitis and transformation to
diffuse large B-cell lymphoma
Progressive Increase in absolute lymphocyte count (ALC) of >50% in Baseline ALC for calculation of LDT must be >30 × 109/L.
lymphocytosis 2 months or lymphocyte doubling time (LDT) of <6 LDT needs to be determined by using multiple
months serial ALC counts (2 weekly ALC for >3 months) to
perform linear regression analysis. All other potential
causes of changes in ALC (eg, infection, recent use of
corticosteroids) need to be excluded. ALC alone should
not be used as an indication for treatment.
*Use of fatigue as a sole indication for treatment of patients with CLL requires a careful evaluation and exclusion of all alternative etiologies.

margin at rest and at maximal inspiration in the midclavicu- evolution of therapy for CLL has exceeded the ability to do
lar line). Patients require a bone marrow study to determine randomized controlled trials.
if they have any other bone marrow pathology and to deter- CIT combining alkylating agents and anti-CD20 mono-
mine the extent of involvement of the bone marrow by CLL, clonal antibodies has been shown to be highly effective,
which requires reporting of the percentage cellularity and and possibly less toxic, in less fit patients. Bendamustine
percentage involvement of the bone marrow cellularity by and rituximab (BR) was compared to FCR in a large ran-
CLL cells in the bone marrow biopsy. Imaging is required to domized clinical trial for initial treatment of fit patients.
determine the size of the nonpalpable lymph nodes in the Patients treated with FCR had a significantly higher com-
chest, abdomen, and pelvis. CLL cells are usually not FDG plete response (CR) rate and duration of response but also
avid, and PET scans should not be routinely used for CLL a higher risk of serious adverse events. The study authors
evaluation prior to initiation of therapy. suggested that BR could be more tolerable and equally effi-
cacious in fit patients >65 years and those with mutated
IGHV.
Pretherapy precautions
Multiple alternative regimens combining alkylating
Use of monoclonal antibody therapy and myelosuppressive agents and monoclonal antibodies can also be considered as
drugs increases the risks of reactivation of latent infections. initial therapy for less fit patients with CLL. Comparison of
CLL patients should be tested for evidence of infection with the results of the clinical trials using chlorambucil and anti-
hepatitis B and hepatitis C viruses before initiation of che- CD20 monoclonal antibodies is difficult because of the dif-
motherapy or monoclonal antibodies and patients at high ferences in the chlorambucil regimens. Although the
risk of reactivation (guidelines can be consulted to define addition of anti-CD20 monoclonal antibodies to chloram-
this population) should be have antiviral therapy to mini- bucil regimens increases the rates of neutropenia and adds
mize this risk. Patients could also benefit from antiherpesvi- the additional risk of infusion reactions, the improvement
rus and anti-Pneumocystis prophylaxis, although the value of in response suggests that chlorambucil should no longer be
these precautions is not proven. Effective therapy of CLL can used as monotherapy. Chlorambucil combined with ritux-
cause rapid cytotoxicity of CLL cells with toxic conse- imab resulted in an 84% overall response rate (ORR) with
quences, including tumor lysis syndrome. Prophylactic allo- 10% CR rate in an elderly population (median age 70 years)
purinol and hydration together with appropriate monitoring with a median progression free survival (PFS) of 24 months.
is suggested. The major adverse event was neutropenia (41% grade 3/4).
Patients requiring therapy for progressive CLL can be cat- A phase 2 trial of chlorambucil and rituximab in elderly
egorized according to CLL biology and physical fitness (Fig- patients (>65 years) had an 82% ORR with 19% CR and a
ure 22-4A). Genetic analysis identifying 17p-/TP53mutation median PFS of 34 months with use of maintenance ritux-
characterizes a subset of patients with very-high-risk CLL imab. Addition of ofatumumab to chlorambucil monother-
with a poor and short duration responses to “conventional” apy increased the ORR from 69% to 82% and CR rate from
chemotherapy containing regimens that require alternative 1% to 14% with significantly improved PFS. Chlorambucil
therapies. Patients with very poor physical fitness are unlikely monotherapy was compared to combinations with ritux-
to benefit from active treatment and should be considered imab or obinutuzumab in the German-led CLL11 clinical
for best supportive care. trial for patients with decreased fitness based on CIRS scores
and decreased renal function. The results, updated in 2015,
showed that combinations of chlorambucil with either
Chemoimmunotherapy
obinutuzumab or rituximab significantly improved PFS
Chemoimmunotherapy (CIT) combining 1 or 2 chemother- and OS compared to chlorambucil monotherapy. There
apy agents (fludarabine, pentostatin, cladribine, cyclophos- was significantly better PFS (29 vs 15 months) and time to
phamide, bendamustine, and chlorambucil) and anti-CD20 next treatment (43 vs. 33 months) for obinutuzumab/chlo-
monoclonal antibodies (rituximab, ofatumumab, and rambucil vs. rituximab/chlorambucil, but no significant dif-
obinutuzumab) is highly effective for most patients requir- ference in OS. Chlorambucil monotherapy is thus inferior
ing initial therapy for progressive CLL. The combination of to chlorambucil in combination with an anti-CD20 mono-
fludarabine, cyclophosphamide, and rituximab (FCR) uti- clonal antibody, suggesting that chlorambucil should no
lizes the synergistic effect of combining cyclophosphamide longer be used as monotherapy. However, the optimal chlo-
with fludarabine, and the addition of rituximab to this che- rambucil and anti-CD20 monoclonal antibody regimen has
motherapy combination has been shown to improve overall not yet been determined. Clinical trials of BCR signaling
survival. The optimal choice of chemotherapy agents and pathway inhibitor based regimens vs. CIT regimens are
monoclonal antibodies remains uncertain because the rapid ongoing.

Management | 627
A
Initial treatment of
progressive CLL
Unfit
Supportive care
No TP53 Disrupted
disruption TP53
Less fit Fit Less fit Fit

BR/PCR FCR/BR/PCR Ibrutinib Ibrutinib
Clb + mAb Clinical trial Idelalisib + rituximab Idelalisib + rituximab
Clinical trial Clinical trial Then consider
RIC allo-SCT/CART
B
Progressive relapsed/
refractory CLL
Unfit
Supportive care
Purine Very high

analogue sensitive risk
Less fit Fit Less fit Fit

BR FCR/BR/PCR Ibrutinib Ibrutinib
PCR Ibrutinib Idelalisib + rituximab Idelalisib + rituximab
Clb + mAb Idelalisib + rituximab Clinical trial Clinical trial
Ibrutinib Clinical trial Then consider
Idelalisib + rituximab RIC allo-SCT/CART
Clinical trial
Figure 22-4 Risk-stratified therapy for progressive CLL. (A) Initial selection of therapy for patients with progressive CLL should be based on
patient fitness and the biology of the disease. Patients with predicted defective p53 function based on FISH analysis for 17p13 deletion and
sequencing of TP53 (disrupted TP53) should be treated with target therapy. Fit patients who achieve a good response should then be considered
for immunotherapy with either RIC allo-SCT or CART therapy. Patients without disrupted TP53 should be treated with chemoimmunotherapy
(CIT) or on clinical trials. CIT regimens include fludarabine, cyclophosphamide and rituximab (FCR), bendamustine and rituximab (BR),
pentostatin, cyclosphosphamide and rituximab (PCR), and chlorambucil and anti-CD20 monoclonal antibodies (Clb + mAb). (B) Therapy for
progressive relapsed/refractory disease is selected by using data on patient fitness, response to previous purine analogue containing CIT, and
TP53 status. Patients with a progression-free survival of at least 2 years after purine analogue containing CIT are considered purine analogue
sensitive. Patients with purine analogue refractory CLL and those with disrupted TP53 are considered very high risk.
(15 vs 59 months), and median OS (42 months vs not

Therapy for very-high-risk CLL
reached at median follow-up of 70 months). Similar poor
Patients with progressive CLL and 17p-/TP53mutation had responses were achieved with other chemoimmunothera-
poor responses to drugs with a mechanism of action pies including BR. Alternative therapies with cytotoxic
requiring an intact DNA damage response pathway. In mechanisms that were not dependent on an intact DNA
patients receiving initial therapy with FCR in the German damage response pathway such as high-dose corticoste-
CLL8 trial, those with 17p-/TP53mutation had a signifi- roids and the anti-CD52 monoclonal antibody alemtu-
cantly inferior response compared to those without pre- zumab achieve higher ORR, but responses were generally
dicted TP53 dysfunction: ORR (75% vs 98%), median PFS of short duration. The development of highly effective

therapies with small molecules targeted drugs that inhibit for retreatment with similar CIT regimes. However, the dura-
signaling pathways essential for CLL cell survival and tion of response to second treatment is usually shorter than to
growth has heralded a new era in the management of these the initial therapy. Patients with a PFS of <2 years to purine
patients. analogue containing CIT and those with 17p13-/TP53mutation
should be considered to be very-high risk and require
alternative treatment regimens (Figure 22-4B). Physically fit
BCR pathway inhibition
very-high-risk patients with a good response to therapy

Ibrutinib targets Bruton tyrosine kinase (BTK) and is cur- should then be considered for immunotherapy (RIC allo-SCT
rently FDA approved for initial treatment of patients with or CART) for maintenance of remission.
CLL and 17p13 deletion. Ibrutinib is discussed further below. Patients with very-high-risk relapsed/refractory CLL pre-
Use of ibrutinib monotherapy as initial therapy for progres- viously had a very poor prognosis prior to the development
sive CLL in 33 patients with 17p-/TP53mutation showed an of targeted small molecule inhibitor therapy. These drugs
ORR of 97% with 55% partial responses (PR) and 42% par- interrupt pathways required for CLL cell survival and prolif-
tial response with persistent lymphocytosis (PR-L) at 24 eration utilizing mechanisms that are independent of the
weeks with an estimated 91% PFS at 24 months. However, DNA damage response pathway and are thus effective in
this therapy is not likely to be curative and there are concerns patients with 17p-/TP53mutation and those resistant to purine
about the risk of development of resistance and clonal evolu- analogues.
tion/selection leading to transformation to diffuse large Inhibitors of the BCR pathway (Figure 22-3) are a unique
B-cell lymphoma. Consequently, more effective therapy or class of drugs that are highly effective in CLL and have
maintenance interventions will be required to improve long- changed practice. Their introduction has required an ongo-
term outcomes. ing re-evaluation of the role of prognostic factors in predict-
Idelalisib targets the phosphatidylinositol-4,5- bisphos ing response to treatment, revision of response criteria,
phate-3-kinase catalytic subunit delta (PI3Kδ), and is app changes in duration of therapy, and the need to recognize
roved in Europe in combination with rituximab for initial and manage different adverse events. Several new classes of
treatment of patients with CLL and 17p-/TP53mutation. FDA drugs in development could add considerably to the therapy
approval is for only relapsed/refractory CLL as described of CLL. These include targeted inhibitors of BCL2, later gen-
below. eration more effective unconjugated anti-CD20 monoclonal
Several additional BTK and PI3K inhibitors are currently antibodies, CART therapies, and checkpoint inhibitors.
in development. An additional promising agent in develop-
ment is the targeted small molecule inhibitor of BCL2
BTK inhibition: ibrutinib
(venentoclax; ABT199) with impressive initial results in the
treatment of patients with 17p-/TP53mutation. Other poten- BTK is an important component of the BCR signaling path-
tial options will be multidrug regimens and the use of way (Figure 22-3) expressed in hematopoietic tissue except
immune modulation by reduced-intensity conditioning T-cells and plasma cells. Ibrutinib is an orally administered
(RIC) allogeneic hematopoietic stem cell transplant (allo- molecule that binds covalently to a cysteine residue near the
SCT) and chimeric antigen receptor T-cell (CART) therapies enzymatic site of BTK resulting in irreversible inhibition.
discussed below. Although ibrutinib has a short half life, BTK binding is irre-
versible, and cellular BTK enzymatic activity can only be
restored by synthesis of new BTK protein which extends the
Treatment of relapsed/refractory CLL
therapeutic effect and the drug can thus be given once daily.
Patients with relapsed/refractory CLL can often be safely Ibrutinib is FDA approved as monotherapy for patients with
monitored until they meet the IWCLL 2008 criteria for pro- previously treated CLL. Onset of action is rapid with resolu-
gressive disease detailed in Table 22-3. Pretreatment evaluation of symptoms and decreases in lymphadenopathy and
tion of relapsed/refractory patients is similar to that required visceromegaly within days of starting therapy followed by a
prior to initial treatment, and includes evaluation of fitness, slower but progressive recovery from cytopenia. Therapy is
CLL biological risk, assessment of CLL disease burden, and frequently (>70%) associated with exacerbation of lympho-
screening for hepatitis B and C. An additional evaluation cytosis that does not affect response to therapy, usually peaks
required for patients with relapsed disease is evaluation of the after one month of therapy, and subsequently slowly
quality of the initial response to therapy. Patients who have declines. The rate of decline of lymphocytosis is usually
previously responded to purine analogue containing CIT reg- slower in patients with less aggressive disease characterized
imens with a response of at least 2 years duration should be by mutated IGHV and 13q14 deletion and faster in patients
considered purine analogue sensitive and can be considered with unmutated IGHV. The median time to resolution of

Management | 629
lymphocytosis on ibrutinib therapy is 19 weeks, but pro- is frequently complicated by minor spontaneous petechiae
longed lymphocytosis up to 124 weeks has been seen in and bruising. Severe hemorrhages are less common and
patients with ongoing treatment responses. Treatment patients on anticoagulant and antiplatelet therapy are at the
related lymphocytosis is a class effect associated with use of highest risk of these complications. Ibrutinib therapy should
drugs that inhibit BCR pathway signaling and does not be stopped for 3-7 days before and after surgical procedures
require specific management. because of the risk of bleeding. Atrial fibrillation is a rare but
important complication of ibrutinib therapy that could be
the result of the inhibition of BTK and related kinases (eg,
Response
TEK). Patients with relapsed/refractory disease CLL have
Ibrutinib monotherapy has been highly effective in patients high rates of infections, and ibrutinib does not appear to
with relapsed/refractory CLL with limited alternative alter this risk. Ibrutinib is metabolized by cytochrome P450
options. However, direct comparisons with other previously enzyme 3A (CYP3A) and potential drug interactions need to
used therapies have been difficult because of the slow but be considered in its use. Concomitant use of moderate or
ongoing response to ibrutinib therapy in many patients and strong CYP3A inhibitors requires ibrutinib dose reductions.
the difficulty in response evaluation in patients with increas-
ing or persistent lymphocytosis. These difficulties have been
Resistance
partially overcome by the revision of the standard IWCLL
2008 response criteria to include the new category of PR-L. Patients with relapsed/refractory CLL can have primary fail-
Ibrutinib monotherapy of patients with relapsed/refractory ure to respond to treatment with ibrutinib or subsequently
CLL including many with very-high-risk disease (17p-/ acquire resistance to the drug after an initial response to
TP53mutated and purine analogue refractory) has achieved treatment. The mechanisms of failure to respond to therapy
high response rates (ORR ~90%). Although most of these are not well understood. Of the patients that acquire resis-
responses were PR or PR-L (~80%) with low CR rates tance to ibrutinib therapy, a high percentage (~30%) has
(<10%), CR rates continued to increase with ongoing ther- transformation to diffuse large B-cell lymphoma. This is
apy with a median time to CR of 21 months in one study. considered a progression of CLL based on the IWCLL 2008
The duration of response is considerably better than those criteria. Disease progression on ibrutinib therapy is most
reported for previously used therapies with an estimated 30 frequent in patients with 17p-/TP53mutation, 11q22.3 dele-
month PFS of 69% and OS of 79%. Both PFS and OS appear tion, and complex karyotypic abnormalities (>1 aberration).
to be inferior in patients with 17p13 deletion (and possible Acquired resistance to ibrutinib therapy can also occur
also in patients with 11q22.3 deletion). In contrast, initial because of mutations that prevent ibrutinib from inhibiting
studies have not shown major differences in response rates BCR signaling. Two such mutations have been described, a
and duration of response based on other prognostic factors cysteine to serine change at amino acid 481 in BTK that pre-
used to assess disease risk in CLL including IGHV mutation vents ibrutinib binding to the active enzymatic site and a
status, expression of CD38 and ZAP70, and other FISH gain of function mutations in the gene coding for PLCg2 that
determined genetic abnormalities. Ibrutinib monotherapy result in autonomous BCR signaling. These mutations have
thus provides a highly effective but noncurative option for not yet been detected in patients with CLL prior to initiation
patients with relapsed refractory CLL. The challenge is to of treatment with ibrutinib suggesting that they occur either
learn how to most effectively use this novel agent, improve occur because of new mutations or by selection of pre-
its efficacy, and determine if there is a subgroup of respond- existing subclones of ibrutinib resistant cells too small to
ing patients that will not have rapid disease progression if detect by current assays.
therapy is discontinued.
Combination therapy
Toxicity
Ibrutinib combination therapy is being tested in clinical tri-
Ibrutinib is generally well tolerated. Diarrhea and skin rashes als. Combination with anti-CD20 monoclonal antibodies is
are relatively common, often transient, and can resolve with potentially attractive because monoclonal antibodies are
no specific management. Less common but more serious most effective at killing circulating CLL cells. However,
drug specific complications include bleeding, atrial fibrilla- enthusiasm for these combinations is tempered by data sug-
tion, arthritis and arthralgia, fatigue, cytopenias, and infec- gesting that ibrutinib could decrease cell mediated monoclo-
tions. BTK signaling is important for platelet activation and nal antibody dependent cytotoxicity. Combinations of
ibrutinib decreases platelet adhesion to von Willebrand fac- ibrutinib with anti-CD20 monoclonal antibodies and CIT
tor, increasing the risk of arterial bleeding. Ibrutinib therapy are currently being tested in clinical trials.

PI3Kd Inhibition: idelalisib BCL2 inhibition: venetoclax (ABT-199)
The PI3K p110δ (PI3Kδ) enzyme is expressed primarily in Venetoclax is an oral selective targeted small-molecule
hematopoietic tissue and is especially important for B-cell inhibitor of the anti-apoptotic molecule BCL2 that is
maturation and survival. In CLL, PI3K activity is constitu- expressed at high levels in CLL cells. Venetoclax monother-
tively activated suggesting that it could be a good target for apy and in combination with rituximab for treatment of
therapy. Specific inhibitors of PI3Kδ could thus provide patients with relapsed/refractory CLL are reported to achieve
targeted therapy for CLL. Idelalisib is an orally bioavailable ORR in excess of 80% with CR rates of ~30% with some
selective inhibitor of PI3Kδ which also inhibits CXCR4 patients achieving MRD negative status. Responses were
and CXCR5 signaling. Inhibition of PI3Kδ prevents phos- similar in patients with 17p13 deletion. The major toxicity
phorylation of the serine/threonine kinases AKT and was severe tumor lysis, but the risk of this complication has
mTOR resulting in decreased BCR pathway signaling been decreased by a revised administration regimen.
(Figure 22-3). Although still in development, venetoclax could be an addi-
tional targeted small molecule therapy for treating CLL.
Response
Immunotherapy
Idelalisib is FDA approved for CLL therapy in combination
with rituximab for patients with comorbidities and for Restoring immune surveillance and cytotoxicity capable of
treatment of patients with relapsed SLL variant of CLL who preventing recurrence of CLL in patients who have minimal
have had at least two prior therapies. Idelalisib monother- residual disease after effective therapy can result in long-
apy for patients with relapsed/refractory CLL (70% refrac- term disease control and possibly even cure. The first effec-
tory to previous treatment and 24% 17p-/TP53mutation) tive modality was RIC allo-SCT and CART therapy is now
resulted in an ORR of 72% with 39% PR and 33% PR-L and being evaluated in clinical trials. In addition, new modalities
a median PFS of 15.8 months. A randomized controlled that could inhibit tumor-induced immunosuppression
study tested idelalisib and rituximab vs. rituximab alone in (checkpoint inhibitors) could potentially also have roles in
CLL patients with relapsed/refractory disease and comor- CLL management. Lenalidomide has been shown to be an
bidities that precluded the use of CIT. Addition of idelalisib effective tumor immunity modulator in CLL, but its clinical
to rituximab significantly improved RR, PFS, and OS. A role in the management of CLL remains uncertain.
subsequent update of this study presented in abstract form RIC allo-SCT is effective therapy in relapsed/refractory
showed that response rates and PFS in the patients receiv- CLL including patients with very-high-risk disease. However
ing idelalisib and rituximab was not affected by 17p-/ therapy is complicated by chronic graft-versus-host disease
TP53mutation, IGHV mutation status or levels of ZAP70 with treatment related mortality of ~20%. Optimum results
expression. are achieved in younger and fitter patients with minimal
residual CLL who have not had extensive prior therapy The
Toxicity availability of highly effective targeted small molecule thera-
pies and alternative immunotherapies has reduced the
Idelalisib inhibition of BCR signaling causes the class effect enthusiasm for RIC allo-SCT in CLL and the indications for
of lymphocytosis that is unlikely to have any clinical signifi- use of this therapy are currently uncertain. Patients with
cance. Idelalisib does have important potential toxicity that very-high-risk CLL who are possible candidates for immune
requires careful monitoring of patients. Gastrointestinal therapy should be referred for evaluation at a center specializing
complications include diarrhea that can be severe and non- in the treatment of CLL early in the course of their disease.
responsive to motility inhibiting drugs, severe colitis, and
intestinal perforation. Hepatic toxicity includes frequent
CART therapy
transaminitis and less common more severe hepatitis that
usually resolves on drug cessation. Pneumonitis requiring Ex vivo introduction of chimeric genes into autologous
drug cessation and treatment with corticosteroids has been T-cells using lentivirus vectors can induce “autologous”
reported. Additional toxicities include pyrexia, fatigue, nau- anti-CLL immunity. The chimeric genes constructs include
sea, rash, neutropenia, hypertriglyceridemia, and hypergly- code for antibody variable regions (eg, B-cell specific anti-
cemia. Idelalisib has not been shown to increase the risk of CD19 or more CLL specific anti-ROR1) together with
infection above the baseline for this patient population. Ide- immunostimulatory molecules (eg, CD3ζ, CD28, CD137).
lalisib induces CYP3A and thus the risk of adverse drug CART therapies are being evaluated in ongoing clinical trials
interactions must be considered. after promising initial results in CLL.

Complications of CLL | 631
disseminated varicella-zoster. Herpes simplex virus reactiva-

Key points tion can result in local lymphadenitis and systemic herpes
• FCR chemotherapy has a higher CR rate and duration of simplex virus infections. Cytomegalovirus reactivation can
response than other CIT regimens, but may have an increased cause serious disease in previously untreated patients, but it
risk of serious adverse events in in less fit patients and those >65 is more common in patients with advanced-stage disease
years of age. and those treated with lymphotoxic therapies. CLL patients
• Ibrutinib targets BTK and is approved by the FDA for relapsed/ with advanced-stage disease and those undergoing immuno-
refractory CLL and for initial therapy of CLL with 17p13 deletion. suppressive therapy or allogeneic hematopoietic stem cell
• Ibrutinib is metabolized via CYP3A and concomitant use with
transplant are at high risk of fungal and atypical bacterial
CYP3A inhibitors requires dose reductions.
infections.
• Idelalisib and rituximab are approved for treatment of
relapsed/refractory CLL in less fit patients.
• Venetoclax (ABT-199) has been effective in clinical trials for
Prevention
relapsed CLL, but can cause tumor lysis syndrome.
Preventative measures, education, and rapid and effective
responses to infection can decrease the risk and conse-
Complications of CLL quences of serious infections. Patients need to be trained to
recognize and seek immediate medical evaluation for serious
CLL is complicated by defective innate and acquired infections and especially fevers ≥38.5°C. Vaccination responses
immune function that develops early in the clinical course are usually suboptimal in patients with CLL. However, pneu-
of the disease. This immune dysfunction generally becomes mococcal vaccine responses can be improved by addition of
more severe with disease progression and is exacerbated by the conjugated 13-valent vaccine to the standard 23-valent
conventional therapies. Immunodeficiency increases the polysaccharide vaccine. Influenza vaccines are likely to be of
risk of infection and autoimmune disease and defective most value in patients with early stage CLL but should be
immune surveillance could contribute to the increase risk of administered to all patients and household members if pos-
second malignancy. CLL patients are also at increased risk sible. Live vaccines (eg, shingles and yellow fever) are
of clonal evolution to DLBCL. DNA damaging chemother- contraindicated.
apy can increase the risk of myeloid and nonhematological Prophylactic antimicrobial therapy is not of proven value
malignancies. in CLL. Pneumocystis and herpes virus prophylaxis are com-
monly used during and after therapies with lymphotoxic
drugs (purine analogues, alemtuzumab, and high dose corti-
Infections
costeroids). Prophylactic antiviral therapy can be useful in
Serious infections result in considerable morbidity and are a decreasing the risk of varicella-zoster and herpes simplex
major cause of death in CLL patients. Defective responses to virus reactivation in patients with recurrent infections. The
antigens by nonmalignant B-cells results in quantitative and use of intravenous immunoglobulin (IVIG) in management
qualitative defects in antibody production. Although abso- of CLL is also not well established. IVIG 0.4 mg/kg every 4
lute T-cell counts are usually increased in patients with CLL, weeks has been shown to decrease the risk of infections but
CD4/CD8 ratios are reversed with decreased T-cell receptor not overall survival. It can cause serious adverse events and is
repertoire and markedly impaired T-cell function. Innate expensive. Use should probably be limited to patients with
immunity is impaired by monocyte, dendritic and NK cell recurrent major infections (at least 2 in 6 months) and
dysfunction; decreased serum complement levels; and bone should not be based on IgG levels alone.
marrow failure associated neutropenia. Effective management of infections in patients with CLL
can be challenging. Evaluation should focus on encapsulated
bacteria, atypical, and opportunistic infections. Treatment
Clinical
should be based on the assumption that all CLL patients are
Impaired humoral immunity markedly increases the risk of immune compromised.
overwhelming bacterial infections by encapsulated organ-
isms (eg, Streptococcus pneumoniae and Staphylococcus
Autoimmune disease
aureus) at all stages of CLL. Defective T-cell immunity
increases the risk of herpesvirus reactivation. Reactivation of Approximately 5%-10% of CLL patients have autoimmune
varicella-zoster virus results in shingles, which is frequently complications, most of which are hematological (eg, auto-
complicated by postherpetic neuralgia and can also lead to immune hemolytic anemia or immune thrombocytopenia).

Hematological disease ITP

Clinical CLL patients with progressive bone marrow failure
Most (>90%) autoimmune cytopenia is caused by loss of self
usually first develop anemia and only subsequently throm-
tolerance attributed to disruption of T-cell function by CLL
bocytopenia. CLL patients presenting with thrombocytope-
cells. This causes pathological production of high-affinity
nia without anemia should be evaluated for causes of platelet
polyclonal IgG antibodies directed against blood cell anti-
sequestration. In those with insidious onset thrombocytope-
gens by nonmalignant B cells resulting in autoimmune
nia and platelet counts >50 × 109/L, hypersplenism should
hemolytic anemia (AIHA) or immune thrombocytopenia
be considered. In contrast, acute onset (<2 weeks) or severe
(ITP). In contrast, production of a self-reactive monoclonal
(platelet counts <30 × 109/L) thrombocytopenia in CLL
antibody (usually IgM) by CLL cells is rare and occurs in
patients is more likely to be caused by ITP. Anti-platelet anti-
<10% of patients with AIHA or ITP. Pure red blood cell apla-
body assays have low specificity and sensitivity and are not
sia (PRCA) can be mediated by either autoantibodies or
useful in making the diagnosis of ITP which remains one of
direct T-cell cytotoxicity. Autoimmune cytopenias occur
exclusion. A bone marrow biopsy is required to show ade-
throughout the course of CLL and cause 15%-20% of noni-
quate megakaryocytopoiesis and confirm the diagnosis of
atrogenic cytopenias in CLL patients. Patients with autoim-
ITP.
mune cytopenia should not be classified as advanced stage
disease unless they have concomitant bone marrow failure
Management Patients with no bleeding complications
demonstrated by bone marrow biopsy.
and platelet counts above ~20-30 × 109/L should be carefully
observed and educated, but do not need active treatment.
AIHA Those needing treatment usually respond to immunosup-
Clinical AIHA is usually characterized by reticulocytosis in pression with corticosteroids. Thrombopoietin agonists can
the absence of bleeding, elevated serum LDH and indirect be useful if patients have a slow or inadequate response to
bilirubin levels, and a positive direct antiglobulin test (DAT) immunosuppression. Splenectomy is considered less effec-
that detects surface bound anti-red blood cell IgG antibodies tive in CLL patients compared to primary ITP. Patients with
and the complement degradation product C3d. However, ITP and bone marrow failure can be treated with similar
patients with AIHA- and CLL-related bone marrow failure regimens as those used to manage complex AIHA. Caution is
(complex AIHA) are often not able to generate a reticulocyte advised with use of ibrutinib in the presence of severe throm-
response to anemia. In addition, although DAT tests are bocytopenia because of the increased risk or bleeding.
positive in over 90% of CLL patients with AIHA, ~15%-20%
of CLL patients have a positive DAT during the course of PRCA
their disease and only ~35% of these patients develop AIHA. Clinical Autoimmune PRCA presents with anemia, a very
Appropriate management of CLL patients with anemia and low absolute reticulocyte count, and no evidence of hemoly-
features suggestive of AIHA often requires a bone marrow sis or bleeding. A definitive diagnosis requires a bone mar-
study to determine the relative roles of hemolysis and CLL row study showing an erythroid lineage maturation arrest.
induced bone marrow failure. The differential diagnosis includes parvovirus and other
virus infections. Because patients with CLL have inadequate
Management Patients with AIHA and adequate erythro- humoral immune response to infections, detection of parvo-
poiesis (simple AIHA) can be treated with immunosuppres- virus and CMV viremia by PCR is more useful than viral
sion using corticosteroids. Patients with severe anemia or a serology.
slow response to corticosteroid therapy can benefit from
addition of IVIG. AIHA relapses are common and many Management PRCA should be treated with immunosup-
patients require long-term immunosuppression or addi- pression using prednisone and cyclosporine. Clinical
tional treatment such as anti-CD20 monoclonal antibodies. improvement is often slow because of the lag time to restora-
Patients with both AIHA and CLL related bone marrow fail- tion of erythropoiesis. Long-term immunosuppression is
ure require treatment for both conditions. Combinations frequently required to maintain adequate hemoglobin
therapy with cyclophosphamide, corticosteroids, and anti- levels.
CD20 monoclonal antibodies can be effective. Because
purine analogues are myelosuppressive and can cause auto-
Autoimmune neutropenia
immune cytopenia when used as monotherapy, these agents
should be probably be avoided. There is preliminary data This is a rare and poorly understood condition that should
that therapy with B-cell receptor pathway inhibitors could be be considered in patients with isolated neutropenia of uncer-
effective in management. Splenectomy is of limited value. tain etiology.

B-cell prolymphocytic leukemia | 633
Nonhematological disease
Key points
Patients with CLL have an increased risk of autoimmune
• CLL is associated with both significant humoral and T-cell
acquired angioedema, paraneoplastic pemphigus, and glo-
mediated immunodeficiency.
merulonephritis. A clinically important consequence of
• Five percent to 10% of CLL patients have autoimmune
immune dysregulation in CLL patients is exaggerated cuta- complications, the most common being AIHA and ITP.
neous arthropod bite reactions which can be complicated by • Diffuse large B-cell lymphoma and Hodgkin lymphoma are
cellulitis and transient painful adenopathy. both seen in increased frequency in CLL.
• CLL is associated with an increased incidence of both
Second malignancies non-melanoma and melanoma skin cancers.
Hematological malignancies
B-cell prolymphocytic leukemia
Lymphoid malignancies
B-cell prolymphocytic leukemia (B-PLL) is a very rare
DLBCL can occur at any time in the course of CLL (inci-
mature B-cell lymphoid malignancy with a median age at
dence ~0.5% per year) with the highest risk in patients with
diagnosis of 69 years and equivalent incidence in males and
NOTCH1 mutations and 17p-/TP53mutation. In ~80% of
females. The etiology and B-cell of origin of the disease is
patients with CLL who develop a DLBCL, a CLL cell under-
unknown.
goes clonal transformation to a highly aggressive DLBCL
with very poor prognosis. In contrast, ~20% of CLL patients
developing DLBCL have clonally unrelated de-novo DLBCL Clinical presentation
with a considerably more favorable prognosis. These two eti-
Patients with B-PLL usually present with very high ALC,
ologies can be distinguished by VDJ rearrangement analysis
splenomegaly that can be massive, minimal or no lymphade-
of paired CLL and DLBCL cell samples. Patients with CLL
nopathy, and anemia and thrombocytopenia.
are also at increased risk of developing Hodgkin lymphoma
and other B-cell malignancies.
Diagnosis
Management De novo DLBCL requires standard evalua- Diagnosis is suggested by a high percentage (~90%) of lym-
tion and management. There is no standard of care for clon- phocytes with prolymphocytic morphology. These cells are
ally evolved DLBCL in patients with CLL. medium sized with large condensed nuclei, a prominent
large nucleolus, and a small amount of basophilic cytoplasm
Nonhematological malignancies without cytoplasmic projections. On flow cytometric analy-
sis these cells B-PLL cells are monoclonal B-cells that have
Skin cancer
bright light chain and CD20 expression and usually do not
CLL markedly increases the risk of skin malignancies. Squa- express CD5 or CD23. These features are useful in differen-
mous cell carcinoma and basal cell carcinoma (BCC) are tiating primary B-cell PLL from CLL with high levels of
increased ~5- to 10-fold and have more aggressive biology prolymphocytes and the leukemic phase of mantle cell

with increased risk of local invasion and distant metastasis. lymphoma. The other considerations in the differential
The risk of melanoma is significantly increased with more diagnosis are marginal zone lymphoma and hairy cell

aggressive biology and poorer outcome. leukemia.
Management Patients need to be educated on limiting Genetic analysis

ultraviolet radiation exposure and require frequent skin
checks with prompt evaluation and management of suspi- FISH analysis for t(11;14) should be done to exclude the
cious lesions. diagnosis of mantle cell lymphoma. Approximately 50% of
patients have 17p-/TP53mutation that is associated with poorer
responses to chemotherapy regimens.
Other malignancies
CLL patients are at increased risk of noncutaneous second

Treatment
malignancies which are a major cause of morbidity and
mortality. Patients should minimize high-risk behavior and B-PLL is a rare disease with limited data from clinical trials and
follow and preventative screening guidelines. no standard of care therapy. Patients with this disease should

be referred to specialized lymphoid malignancy programs for Bulian P, Shanafelt TD, Fegan C, et al. CD49d is the strongest
management. Patients without 17p-/TP53mutation can respond flow cytometry-based predictor of overall survival in chronic
to CIT regimens similar to those used in the treatment of CLL. lymphocytic leukemia. J Clin Oncol. 2014;32:897-904.
Patients with 17p-/TP53mutation can respond to a combination Delgado J, Pratt G, Phillips N, et al. Beta2-microglobulin is a better
predictor of treatment-free survival in patients with chronic
of anti-CD20 monoclonal antibodies and alemtuzumab. BCR
lymphocytic leukaemia if adjusted according to glomerular
pathway inhibitors could be useful, but there is currently no
filtration rate. Br J Haematol. 2009;145:801-805.
published data on their use to treat B-PLL. RIC-alloSCT had
Dohner H, Stilgenbauer S, Benner A, et al. Genomic aberrations
been used to maintain responses in patients who achieve min- and survival in chronic lymphocytic leukemia. N Engl J Med.
imal residual disease status. 2000;343:1910-1916.
Messmer BT, Messmer D, Allen SL, et al. In vivo measurements
document the dynamic cellular kinetics of chronic lymphocytic
Bibliography leukemia B cells. J Clin Invest. 2005;115:755-764.
Rosenquist R, Cortese D, Bhoi S, Mansouri L, Gunnarsson R.
Prognostic markers and their clinical applicability in chronic
lymphocytic leukemia: where do we stand? Leuk Lymphoma.
Epidemiology
2013;54:2351-2364.
Dores GM, Anderson WF, Curtis RE, et al. Chronic lymphocytic Rossi D, Khiabanian H, Spina V, et al. Clinical impact of small
leukaemia and small lymphocytic lymphoma: overview of the TP53 mutated subclones in chronic lymphocytic leukemia.
descriptive epidemiology. Br J Haematol. 2007;139:809-819. Blood. 2014;123:2139-2147.
Rossi D, Rasi S, Spina V, et al. Integrated mutational and
cytogenetic analysis identifies new prognostic subgroups in
Biology chronic lymphocytic leukemia. Blood. 2013;121:1403-1412.
Chiorazzi N, Ferrarini M. Cellular origin(s) of chronic Zent CS, Burack WR. Mutations in chronic lymphocytic leukemia
lymphocytic leukemia: cautionary notes and additional and how they affect therapy choice: focus on NOTCH1,
considerations and possibilities. Blood. 2011;117:1781-1791. SF3B1, and TP53. Hematology Am Soc Hematol Educ Program.
Rothstein TL, Griffin DO, Holodick NE, Quach TD, Kaku 2014;2014:119-124.
H. Human B-1 cells take the stage. Ann N Y Acad Sci.
2013;1285:97-114. Management
Seifert M, Sellmann L, Bloehdorn J, et al. Cellular origin and
pathophysiology of chronic lymphocytic leukemia. J Exp Med. Borge M, Almejún MB, Podaza E, et al. Ibrutinib impairs the
2012;209:2183-2198. phagocytosis of rituximab-coated leukemic cells from chronic
lymphocytic leukemia patients by human macrophages.
Haematologica. 2015;100:e140-e142.
Diagnosis and clinical evaluation Brown JR, Byrd JC, Coutre SE, et al. Idelalisib, an inhibitor of
phosphatidylinositol 3-kinase p110delta, for relapsed/refractory
Hallek M, Cheson BD, Catovsky D, et al. Guidelines for the diagnosis
chronic lymphocytic leukemia. Blood. 2014;123:3390-3397.
and treatment of chronic lymphocytic leukemia: a report from
Byrd JC, Furman RR, Coutre SE, et al. Three-year follow-up of
the International Workshop on Chronic Lymphocytic Leukemia
treatment-naive and previously treated patients with CLL and
(IWCLL) updating the National Cancer Institute-Working Group
SLL receiving single-agent ibrutinib. Blood. 2015;125:2497-2506.
(NCI-WG) 1996 guidelines. Blood. 2008;111:5446-5456.
Byrd JC, Jones JJ, Woyach JA, Johnson AJ, Flynn JM. Entering
Huh YO, Lin KI, Vega F, et al. MYC translocation in chronic
the era of targeted therapy for chronic lymphocytic leukemia:
lymphocytic leukaemia is associated with increased
impact on the practicing clinician. J Clin Oncol. 2014;32:
prolymphocytes and a poor prognosis. Br J Haematol.
3039-3047.
2008;142:36-44.
Cheson BD, Byrd JC, Rai KR, et al. Novel targeted agents and
Put N, Van Roosbroeck K, Konings P, et al. Chronic lymphocytic
the need to refine clinical end points in chronic lymphocytic
leukemia and prolymphocytic leukemia with MYC
leukemia. J Clin Oncol. 2012;30:2820-2822.
translocations: a subgroup with an aggressive disease course.
Conte MJ, Bowen DA, Wiseman GA, et al. Use of positron
Ann Hematol. 2012;91:863-873.
emission tomography-computed tomography in the
management of patients with chronic lymphocytic leukemia/
Risk stratification small lymphocytic lymphoma. Leuk Lymphoma. 2014;55:
2079-2084.
Brachtl G, Pinon Hofbauer J, Greil R, Hartmann TN. The Da Roit F, Engelberts PJ, Taylor RP, et al. Ibrutinib interferes
pathogenic relevance of the prognostic markers CD38 and with the cell-mediated anti-tumor activities of therapeutic
CD49d in chronic lymphocytic leukemia. Ann Hematol. CD20 antibodies: implications for combination therapy.
2014;93:361-374. Haematologica. 2015;100:77-86.

Bibliography | 635
Dreger P, Corradini P, Kimby E, et al. Indications for allogeneic Herman SE, Johnson AJ. Molecular pathways: targeting
stem cell transplantation in chronic lymphocytic leukemia: the phosphoinositide 3-kinase p110-delta in chronic lymphocytic
EBMT transplant consensus. Leukemia. 2007;21:12-17. leukemia. Clin Cancer Res. 2012;18:4013-4018.
Dreger P, Dohner H, Ritgen M, et al. Allogeneic stem cell Hillmen P, Gribben JG, Follows GA, et al. Rituximab plus
transplantation provides durable disease control in poor-risk chlorambucil as first-line treatment for chronic lymphocytic
chronic lymphocytic leukemia: long-term clinical and MRD leukemia: Final analysis of an open-label phase II study. J Clin
results of the German CLL Study Group CLL3X trial. Blood. Oncol. 2014;32:1236-1241.
2010;116:2438-2447. Hillmen P, Robak T, Janssens A, et al. Chlorambucil plus
Dreger P, Schetelig J, Andersen N, et al. Managing high-risk ofatumumab versus chlorambucil alone in previously untreated
CLL during transition to a new treatment era: stem cell patients with chronic lymphocytic leukaemia (COMPLEMENT
transplantation or novel agents? Blood. 2014;124: 3841-3849. 1): a randomised, multicentre, open-label phase 3 trial. Lancet.
Eichhorst B, Fink AM, Busch R, et al. Frontline 2015;385:1873-1883.
chemoimmunotherapy with fludarabine (F), cyclophosphamide Keating MJ, O’Brien S, Albitar M, et al. Early results
(C), and rituximab (R) (FCR) shows superior efficacy in of a chemoimmunotherapy regimen of fludarabine,
comparison to bendamustine (B) and rituximab (BR) in cyclophosphamide, and rituximab as initial therapy for chronic
previously untreated and physically fit patients (pts) with lymphocytic leukemia. J Clin Oncol. 2005;23:4079-4088.
advanced chronic lymphocytic leukemia (CLL): final analysis Khouri IF, Keating MJ, Saliba RM, Champlin RE. Long-term
of an international, randomized study of the German CLL follow-up of patients with CLL treated with allogeneic
Study Group (GCLLSG) (CLL10 Study) [abstract]. Blood. hematopoietic transplantation. Cytotherapy. 2002;4:217-221.
2014;124(21):19. Levade M, David E, Garcia C, et al. Ibrutinib treatment affects
Fama R, Bomben R, Rasi S, et al. Ibrutinib-naive chronic collagen and von Willebrand factor-dependent platelet
lymphocytic leukemia lacks Bruton tyrosine kinase mutations functions. Blood. 2014;124:3991-3995.
associated with treatment resistance. Blood. 2014;124: 3831-3833. McMullen JR, Boey EJ, Ooi JY, Seymour JF, Keating MJ, Tam
Farooqui MZ, Valdez J, Martyr S, et al. Ibrutinib for previously CS. Ibrutinib increases the risk of atrial fibrillation, potentially
untreated and relapsed or refractory chronic lymphocytic through inhibition of cardiac PI3K-Akt signaling. Blood.
leukaemia with TP53 aberrations: a phase 2, single-arm trial. 2014;124:3829-3830.
Lancet Oncol. 2015;16:169-176. Moreno C, Villamor N, Colomer D, et al. Allogeneic stem-
Foa R, Del Giudice I, Cuneo A, et al. Chlorambucil plus rituximab cell transplantation may overcome the adverse prognosis of
with or without maintenance rituximab as first-line treatment unmutated VH gene in patients with chronic lymphocytic
for elderly chronic lymphocytic leukemia patients. Am J leukemia. J Clin Oncol. 2005;23:3433-3438.
Hematol. 2014;89:480-486. Nelson MH, Paulos CM. Novel immunotherapies for hematologic
Furman RR, Cheng S, Lu P, et al. A novel mutation in Bruton malignancies. Immunol Rev. 2015;263:90-105.
tyrosine kinase confers acquired resistance to ibrutinib (PCI- Porter DL, Levine BL, Kalos M, Bagg A, June CH. Chimeric
32765) in CLL [abstract]. Blood. 2014;122(21):4914. antigen receptor-modified T cells in chronic lymphoid
Furman RR, Sharman JP, Coutre SE, et al. Idelalisib and rituximab leukemia. N Engl J Med. 2011;365:725-733.
in relapsed chronic lymphocytic leukemia. N Engl J Med. Roberts AW, Ma S, Brander DM, et al. Determination of
2014;370:997-1007. recommended phase 2 dose of ABT-199 (GDC-0199) combined
Goede V, Eichhorst B, Fischer K, Wendtner CM, Hallek M. with rituximab (R) in patients with relapsed/refractory
Past, present and future role of chlorambucil in the treatment (R/R) chronic lymphocytic leukemia (CLL) [abstract]. Blood.
of chronic lymphocytic leukemia. Leuk Lymphoma. 2014;56:1-8. 2014;124(21):325.
Goede V, Fischer K, Busch R, et al. Obinutuzumab plus Schetelig J, van Biezen A, Brand R, et al. Allogeneic hematopoietic
Chlorambucil in Patients with CLL and Coexisting Conditions. stem-cell transplantation for chronic lymphocytic leukemia
N Engl J Med. 2014;370:1101-1110. with 17p deletion: a retrospective European Group for
Goede V, Fischer K, Engelke A, et al. Obinutuzumab as frontline Blood and Marrow Transplantation analysis. J Clin Oncol.
treatment of chronic lymphocytic leukemia: updated results of 2008;26:5094-5100.
the CLL11 study. Leukemia. 2015;29:1602-1604. Seymour JF, Davids MS, Pagel JM, et al. ABT-199 (GDC-0199) in
Gribben JG, Riches JC. Immunotherapeutic strategies including relapsed/refractory chronic lymphocytic leukemia and small
transplantation: eradication of disease. Hematology Am Soc lymphocytic lymphoma: High response rates among patients
Hematol Educ Program. 2013;2013:151-157. with high risk disease features including unmutated IGHV.
Gribben JG, Zahrieh D, Stephans K, et al. Autologous and Haematologica. 2014;99:249-S703.
allogeneic stem cell transplantations for poor-risk chronic Souers AJ, Leverson JD, Boghaert ER, et al. ABT-199, a potent and
lymphocytic leukemia. Blood. 2005;106:4389-4396. selective BCL-2 inhibitor, achieves antitumor activity while
Hallek M, Fischer K, Fingerle-Rowson G, et al. Addition of sparing platelets. Nat Med. 2013;19:202-208.
rituximab to fludarabine and cyclophosphamide in patients Stilgenbauer S, Schnaiter A, Paschka P, et al. Gene mutations and
with chronic lymphocytic leukaemia: a randomised, open-label, treatment outcome in chronic lymphocytic leukemia: results
phase 3 trial. Lancet. 2010;376:1164-1174. from the CLL8 trial. Blood. 2014;123:3247-3254.

Wiestner A. BCR pathway inhibition as therapy for chronic malignancy, which is the most common cause of death. Leuk
lymphocytic leukemia and lymphoplasmacytic lymphoma. Lymphoma. 2004;45:507-513.
Hematology Am Soc Hematol Educ Program. 2014;2014: Kyasa MJ, Parrish RS, Schichman SA, Zent CS. Autoimmune
125-134. cytopenia does not predict poor prognosis in chronic
Woyach JA, Furman RR, Liu TM, et al. Resistance mechanisms for lymphocytic leukemia/small lymphocytic lymphoma. Am J
the Bruton’s tyrosine kinase inhibitor ibrutinib. N Engl J Med. Hematol. 2003;74:1-8.
2014;370:2286-2294. Melchardt T, Weiss L, Greil R, Egle A. Viral infections and their
management in patients with chronic lymphocytic leukemia.
Leuk Lymphoma. 2013;54:1602-1613.
Middleton O, Cosimo E, Dobbin E, et al. Complement deficiencies
Complications
limit CD20 monoclonal antibody treatment efficacy in CLL.
Bowen DA, Call TG, Shanafelt TD, et al. Treatment of autoimmune Leukemia. 2015;29:107-114.
cytopenia complicating progressive chronic lymphocytic Moreno C, Hodgson K, Ferrer G, et al. Autoimmune cytopenia in
leukemia/small lymphocytic lymphoma with rituximab, chronic lymphocytic leukemia: prevalence, clinical associations,
cyclophosphamide, vincristine, and prednisone. Leuk and prognostic significance. Blood. 2010;116:4771-4776.
Lymphoma. 2010;51:620-627. Parikh SA, Rabe KG, Call TG, et al. Diffuse large B-cell lymphoma
Bowen DA, Rabe KG, Schwager SM, et al. Infectious lymphadenitis (Richter syndrome) in patients with chronic lymphocytic
in patients with chronic lymphocytic leukemia/small leukaemia (CLL): a cohort study of newly diagnosed patients.
lymphocytic lymphoma: a rare, but important, complication. Br J Haematol. 2013;162:774-782.
Leuk Lymphoma. 2014;56:311-314. Pasiarski M, Rolinski J, Grywalska E, et al. Antibody and
Brewer JD, Habermann TM, Shanafelt TD. Lymphoma- plasmablast response to 13-valent pneumococcal conjugate
associated skin cancer: incidence, natural history, and clinical vaccine in chronic lymphocytic leukemia patients—preliminary
management. Int J Dermatol. 2014;53:267-274. report. PLoS One. 2014;9:e114966.
Davis JE, Ritchie DS. The passive-aggressive relationship Quinquenel A, Al Nawakil C, Baran-Marszak F, et al. Old DAT
between CLL-B cells and T cell immunity. Leuk Res. and new data: Positive direct antiglobulin test identifies a
2014;38:1160-1161. subgroup with poor outcome among chronic lymphocytic
Dhalla F, Lucas M, Schuh A, et al. Antibody deficiency secondary leukemia stage A patients. Am J Hematol. 2015;90:E5-8.
to chronic lymphocytic leukemia: should patients be treated Rai KR, Sawitsky A, Cronkite EP, Chanana AD, Levy RN,
with prophylactic replacement immunoglobulin? J Clin Pasternack BS. Clinical staging of chronic lymphocytic
Immunol. 2014;34:277-282. leukemia. Blood. 1975;46:219-234.
Fabbri G, Khiabanian H, Holmes AB, et al. Genetic lesions Riches JC, Davies JK, McClanahan F, et al. T cells from CLL
associated with chronic lymphocytic leukemia transformation patients exhibit features of T-cell exhaustion but retain capacity
to Richter syndrome. J Exp Med. 2013;210:2273-2288. for cytokine production. Blood. 2013;121:1612-1621.
Francis S, Karanth M, Pratt G, et al. The effect of immunoglobulin Robak E, Robak T. Skin lesions in chronic lymphocytic leukemia.
VH gene mutation status and other prognostic factors on Leuk Lymphoma. 2007;48:855-865.
the incidence of major infections in patients with chronic Royle JA, Baade PD, Joske D, Girschik J, Fritschi L. Second cancer
lymphocytic leukemia. Cancer. 2006;107:1023-1033. incidence and cancer mortality among chronic lymphocytic
Hamblin AD, Hamblin TJ. The immunodeficiency of chronic leukaemia patients: a population-based study. Br J Cancer.
lymphocytic leukaemia. Br Med Bull. 2008;87:49-62. 2011;105:1076-1081.
Hamblin TJ. Non-hemic autoimmunity in CLL. Leuk Res. Sampalo A, Brieva JA. Humoral immunodeficiency in chronic
2009;33:366-367. lymphocytic leukemia: role of CD95/CD95L in tumoral damage
Jamroziak K, Tadmor T, Robak T, Polliack A. Richter syndrome and escape. Leuk Lymphoma. 2002;43:881-884.
in chronic lymphocytic leukemia: updates on biology, clinical Schollkopf C, Rosendahl D, Rostgaard K, Pipper C, Hjalgrim H.
features and therapy. Leuk Lymphoma. 2015;56:1-10. Risk of second cancer after chronic lymphocytic leukemia. Int J
Jurado-Camino T, Cordoba R, Esteban-Burgos L, et al. Chronic Cancer. 2007;121:151-156.
lymphocytic leukemia: a paradigm of innate immune cross- Sinisalo M, Vilpo J, Itala M, Vakevainen M, Taurio J, Aittoniemi
tolerance. J Immunol. 2015;194:719-727. J. Antibody response to 7-valent conjugated pneumococcal
Kaufman M, Limaye S, Driscoll N, et al. A combination of vaccine in patients with chronic lymphocytic leukaemia.
rituximab, cyclophosphamide and dexamethasone effectively Vaccine. 2007;26:82-87.
treats immune cytopenias of chronic lymphocytic leukemia Tsimberidou AM, Wen S, McLaughlin P, et al. Other malignancies
Leuk Lymphoma. 2009;50:892-899. in chronic lymphocytic leukemia/small lymphocytic lymphoma.
Kubica AW, Brewer JD. Melanoma in immunosuppressed J Clin Oncol. 2009;27:904-910.
patients. Mayo Clin Proc. 2012;87:991-1003. Velez NF, Karia PS, Vartanov AR, Davids MS, Brown JR, Schmults
Kyasa MJ, Hazlett L, Parrish RS, Schichman SA, Zent CS. CD. Association of advanced leukemic stage and skin cancer tumor
Veterans with chronic lymphocytic leukemia/small lymphocytic stage with poor skin cancer outcomes in patients with chronic
lymphoma (CLL/SLL) have a markedly increased rate of second lymphocytic leukemia. JAMA Dermatol. 2014;150:280-287.

Bibliography | 637
Visco C, Barcellini W, Maura F, Neri A, Cortelezzi A, Rodeghiero B-cell prolymphocytic leukemia

F. Autoimmune cytopenias in chronic lymphocytic leukemia.
Am J Hematol. 2014;89:1055-1062. Campo E, Catovsky D, Montserrat E, Muller-Hermelink
Zent CS, Ding W, Reinalda MS, et al. Autoimmune cytopenia in HK, Harris NL, Stein H. B-cell prolymphocytic leukemia.
chronic lymphocytic leukemia/small lymphocytic lymphoma: In: Swerdlow SH, Campo E, Harris NL, et al, eds. WHO
changes in clinical presentation and prognosis. Leuk Lymphoma. classification of tumours of hematopoietic and lymphoid tissues.
2009;50:1261-1268. Lyon, France: International Agency for Research on Cancer,
Zent CS, Ding W, Schwager SM, et al. The prognostic significance 2008.
of cytopenia in chronic lymphocytic leukemia/small Dearden C. How I treat prolymphocytic leukemia. Blood.
lymphocytic lymphoma (CLL) Br J Haematol. 2008;141:615-621. 2012;120:538-551.

CHAPTER
23
Plasma cell disorders
Shaji K. Kumar and Jesús San Miguel
Introduction, 639 Disease definitions, 641 Active (symptomatic) MM, 645
Normal PC development, 639 MGUS, 641 Other PC disorders, 666
Detection of PC disorders, 640 Smoldering (or asymptomatic) MM, 644 Bibliography, 673
Introduction sections, we cover the various PC disorders, including diag-

nostic criteria and individual treatment approaches.
Plasma cell (PC) disorders—sometimes still collectively
referred to by the older term dyscrasias—are clonal prolifera-
tions of PCs resulting in a spectrum of clinical conditions, Normal PC development
ranging from the asymptomatic presence of a monoclonal
protein in the blood (serum) or urine to symptomatic dis- Development of a fully functional antibody-secreting PC
ease states with destruction of healthy tissue. In symptomatic from a B lymphocyte is a multistep process that is initially
PC neoplasms, the symptoms may result from tissue infiltra- independent of antigen exposure, followed by a late antigen-
tion by the monoclonal PCs, effects of a secreted monoclonal driven phase. Normal differentiation from early B-cells to
protein, or interaction between the PCs and the marrow PCs is characterized by three B-cell specific DNA remodeling
microenvironment. mechanisms involving the immunoglobulin (Ig) genes: VDJ
Although PC disorders have been identified in patients of rearrangement, somatic hypermutation, and class switch
all ages, these diseases are most commonly seen in the older recombination. Rearrangements of the immunoglobulin
population, typically in the sixth decade of life and beyond. genes occur early in B-cell precursors in the bone marrow,
The most common and the earliest identifiable phase of PC while antigen recognition, selection, somatic hypermutation,
disorders is monoclonal gammopathy of undetermined sig- and class switch recombination take place in the germinal
nificance (MGUS), characterized by relatively low burden of center of a lymph node.
clonal PCs with no measurable impact on the individual, The earliest B-cell precursor shows spontaneous antigen-
and not requiring any therapeutic intervention. The vast independent rearrangement of the immunoglobulin heavy
majority of patients with MGUS enjoy a normal life span chain, which is then followed by light-chain rearrangement.
with no visible consequence of the monoclonal process, but Once successful light-chain rearrangement occurs, the cell
a relatively small fraction will develop features of one of the expresses a complete immunoglobulin molecule on its sur-
symptomatic forms. In contrast to asymptomatic MGUS, face (typically IgM or IgD), which identifies it as a mature
patients with symptomatic multiple myeloma (MM), light- B-cell. Next, antigen-dependent development begins when a
chain amyloidosis (AL), or Waldenström macroglobulin- naïve mature B-cell in a germinal center recognizes an anti-
emia (WM) usually require active treatment. In the ensuing gen with its membrane-bound surface antibody, which trig-
gers two separate processes: somatic hypermutation and
class switch. Somatic hypermutation is a process by which
cells introduce mutations into the variable region genes,
Conflict-of-interest disclosure: Dr. Kumar: consultancy: Celgene,
Millennium, Onyx, Janssen, Bristol-Myers Squibb, Sanofi, Skyline Diag- providing a repertoire of competent cells with varying degree
nostics; research funding: Celgene, Millennium, Onyx, Janssen, Bristol- of affinity for the antigen. Class switch involves changing the
Myers Squibb, Sanofi, Novartis. Dr. San-Miguel: consultancy: Jansen, heavy chain that is expressed to produce other antibody
Novartis, Celgene, Millennium, Onyx, Bristol-Myers Squibb, Sanofi.
Off-label drug use: Dr. Kumar: not applicable. Dr. San-Miguel: not classes besides IgM or IgD, including IgG, IgA, or IgE; this
applicable. provides tools for a multifaceted response to the exposed
| 639

640 | Plasma cell disorders
antigen. Finally, the antigen-exposed hypermutated and using antibodies specific to each of the heavy- and light-
class-switched PC migrates to the bone marrow, where it chain, or (iii) serum free light-chain (FLC) assay.
interacts with marrow stromal cells before finally differenti- The serum or urine protein electrophoresis (SPEP, UPEP)
ating into a long-lived antibody-producing PC. involves charge-based separation of the serum or urine pro-
teins on a gel, which allows detection of the presence of a
monoclonal protein. This test has relatively low sensitivity
Detection of PC disorders and can miss small monoclonal proteins and presence of
monoclonal light chain. The next step in the process of
The diagnosis of a PC disorder rests on the ability to demon- monoclonal protein assessment is an immunofixation (IFE)
strate one or more of the following: a monoclonal protein in study, performed on the serum or urine, using antibodies
the serum or urine, or the presence of monoclonal PCs in the directed against each of the heavy chains and the κ and λ
bone marrow (BM), peripheral blood, or in discrete soft tis- light chains. This allows identification of the type of mono-
sue masses (Figure 23-1). Demonstration of the monoclonal clonal protein in terms of heavy chain and light-chain iso-
protein may require (i) protein electrophoresis performed type, as well as detection of small amounts of monoclonal
on serum or urine, (ii) immunofixation of serum or urine protein otherwise not detected on protein electrophoresis.
A Serum or urine Serum free Serum or urine

protein electrophoresis light-chain assay immunofixation
100000
Normal Myeloma λ LCMM PEL
10000
G
1000 Renal
Serum λ (mg/L)
impairment
M-spike A
100
Normal M
10 sera
K
1
κ LCMM
PEL L
0.1
0.1 1 10 100 1000 10000 100000
Serum κ (mg/L)
Bone marrow Plasmacytoma Peripheral blood
Figure 23-1 Diagnostic tests for monoclonal protein and PCs.

MGUS | 641
However, unlike the SPEP or UPEP, IFE is not quantitative. bony or extramedullary), WM, light-chain amyloidosis
In a small proportion of patients, both these tests can be (AL), light-chain deposition disease (LCDD) and POEMS
negative, as the PCs may only secrete immunoglobulin FLCs syndrome (ie, polyneuropathy, organomegaly, endocrinop-
(κ or λ free light chains). However, the serum FLC assay athy, monoclonal gammopathy, and skin changes) (Table
allows quantitation of monoclonal FLCs circulating 23-1). In addition, several other conditions have been
unbound to the heavy chain, by virtue of the assay’s reactiv- described in the context of a monoclonal protein or mono-
ity against exosed FLC epitopes that are normally hidden clonal PCs; these are relatively rare and their pathophysiol-
when light chains are bound to the heavy chain. The FLC ogy is not well understood (Table 23-2).
assay signals the presence of a clonal process when the ratio
between κ and λ FLC is skewed from the normal range, and
more importantly, the FLC assay also allows quantitation of MGUS
the clonal light chain, facilitating disease monitoring. With
these three tests, over 98% of patients with PC disorders can MGUS represents the starting point or the initiating abnor-
be demonstrated to have a monoclonal protein, leaving mality for all PC disorders. There appears to be some varia-
behind a very small minority who are truly “nonsecretory” tion in the prevalence of this disease based on geography and
in that they do not secrete any monoclonal protein. race. In a study from a southeastern Minnesota community,
Recently, a novel assay that allows measurement of spe- an M protein was found in 15 of the 1,200 adults (1.25%)
cific heavy- and light-chain combinations has become avail- aged 50 years or over. Similarly, monoclonal protein was
able. This enables measurement of each isotype-specific detected in the sera of 334 persons from among 30,279
heavy and light chain (ie, IgGκ, IgGλ, IgAκ, IgAλ, IgMκ, and French adults studied, translating to a prevalence of 1.1%. In
IgMλ), and the specific measurement of monoclonal (intact) a convenience sample of community-dwelling elderly sub-
Ig, as well as the suppression of the uninvolved HLC pair, in jects aged 63-95 years seen for health screening examina-
order to test the impact of immune paresis. This is not widely tions, M protein was seen in 2.7% of Japanese compared to
available clinically yet but is likely to be available in the 10% of Americans. In contrast, African Americans had a
future. greater than twofold-higher prevalence of monoclonal gam-
The other aspect of the diagnosis of a PC disorder is dem- mopathy than whites; the higher rate is similar to that seen
onstration of monoclonal PCs; this is the hallmark of the among Africans living in Ghana.
disease class but is not always required for all stages of the The prevalence of monoclonal gammopathy increases
disease spectrum. PCs normally reside in the bone marrow with age, and is greater in men than women. The etiology of
(<5%); clonal PCs may be found in a bone marrow aspirate MGUS remains unknown, but various risk factors have been
of trephine biopsy. The bone marrow examination gives an identified. Increased risk has been reported in first-degree
estimate of the tumor cell burden and can vary anywhere relatives of patients with PC disorders, survivors of nuclear
from a nearly normal looking marrow to a marrow almost radiation exposure, and those with occupational exposure to
completely replaced by clonal PCs. Unfortunately, the mar- petroleum products and pesticides.
row involvement in MM can be patchy, resulting in sam- Recent studies have conclusively demonstrated that MGUS
pling variation during biopsy. Varying numbers of PCs can precedes myeloma in all patients and that the MGUS is likely
also be detected in circulation in the vast majority of to be present for 8-10 years or more before the diagnosis of
myeloma patients, especially with the use of multi-parameter MM. During the transition from MGUS to active MM, there
flow cytometry (MFC). Finally, a small proportion of pati are several alterations that occur in the PCs as well as the
ents will present with soft tissue masses (plasmacytomas), marrow microenvironment. During this transition, clonal
with or without associated bony destruction; biopsy of plas- PCs not only increase in the numbers, but also acquire addi-
macytomas shows sheets of monoclonal PCs. The demon- tional characteristics including increased proliferative rate of
stration of clonality in the PC depends on their exclusive the PCs, decreased dependence on the bone marrow micro-
expression of the κ or λ light chain, which can be detected by environment, increasing genetic complexity, and the ability
immunohistochemistry or immunophenotyping. to secrete numerous cytokines that can drive associated fea-
tures like bone destruction. This is paralleled by changes in
the tumor microenvironment such increased bone marrow
Disease definitions angiogenesis, a shift in the osteoclast-osteoblast balance lead-
ing to increased tumor resorption, and suppression of the
The spectrum of PC disorders consists of MGUS, smolder- immune system affecting both humoral and cellular immu-
ing (asymptomatic) MM (SMM), symptomatic/active MM, nity. However, the critical events that mediate this “malig-
plasmacytoma (which can be solitary or multiple, and also nant switch” in the PCs remain unclear (Figure 23-2).

Table 23-1 Diagnostic criteria and differential diagnosis of monoclonal gammopathies
Disorder Disease definition

Monoclonal gammopathy of Serum monoclonal protein level ≥3 g/dL, bone marrow PCs <10%, and absence of end-organ damage, such
undetermined significance as lytic bone lesions, anemia, hypercalcemia, or renal failure, that can be attributed to a PC proliferative
(MGUS) disorder
Smoldering myeloma Serum monoclonal protein (IgG or IgA) level ≥3 g/dL or urinary monoclonal protein ≥500 mg per 24 h
(SMM; also referred to as and/or clonal bone marrow plasma cells 10%–60% and absence of end-organ damage, such as lytic bone
asymptomatic MM) lesions, anemia, hypercalcemia, or renal failure, that can be attributed to a PC proliferative disorder
Multiple myeloma (MM; also Bone marrow PCs ≥10% or biopsy proven bony or extramedullary plasmacytoma, presence of serum or
referred to as active MM or urinary monoclonal protein (except in patients with true nonsecretory MM), plus evidence of lytic bone
symptomatic MM) lesions, anemia, hypercalcemia, or renal failure (CRAB) that can be attributed to the underlying PC
proliferative disorder.
An expanded definition described in the text includes those patients with a more than one bone lesion on
CT-PET or MRI, ≥60% PCs in the marrow or serum FLC ratio <0.01 or >100.
Plasma cell leukemia (PCL) ≥2,000/µL circulating clonal PCs or ≥20% of WBC
May be present at diagnosis, or develop during disease evolution in MM
Solitary plasmacytoma Biopsy-proven solitary lesion of bone or soft tissue (extramedullary) with evidence of clonal PCs, normal
skeletal survey, and MRI of spine and pelvis, and absence of end-organ damage, such as anemia,
hypercalcemia, renal failure, or additional lytic bone lesions, that can be attributed to a PC proliferative
disorder
Primary systemic amyloidosis Tissue deposition of light-chain–derived amyloid fibrils (beta pleated sheets), apple green birefringence on
(also known as light chain polarizing microscopy, and tissue deposits stain for κ or λ light chain
amyloidosis, AL) Presents as cardiomyopathy, nephrotic syndrome, malabsorption, hepatic failure, peripheral neuropathy, or
other symptoms based on organ involvement
Small clonal PC infiltration in the bone marrow (usually <10%)
Typically, lytic lesions and hypercalcemia not seen; rarely, coexist with myeloma or myeloma may develop
subsequently
POEMS syndrome Major criteria: polyneuropathy, monoclonal PC disorder, sclerotic bone lesions, elevated vascular endothelial
(polyneuropathy, growth factor, and the presence of Castleman disease
organomegaly, Minor criteria: organomegaly (spleen, liver, or lymph nodes), volume overload (peripheral edema, pleural
endocrinopathy, effusion, ascites), endocrinopathy (adrenal, thyroid, pituitary, gonadal, parathyroid, pancreatic), skin
monoclonal protein, skin changes (hyperpigmentation, hypertrichosis, plethora, hemangiomata, white nails), papilledema,
changes) thrombocytosis
The diagnosis of POEMS syndrome requires 3 of the major criteria, 2 of which must include
polyradiculoneuropathy and clonal PC disorder, and at least 1 of the minor criteria.
Waldenström IgM monoclonal gammopathy regardless of the size of the M protein
macroglobulinemia (WM)* 10% or greater bone marrow infiltration by lymphoplasmacytic cells
Typical immunophenotype (surface IgM+, CD5+/−, CD10−, CD19+, CD20+, CD22+, CD23−)
Cryoglobulinemia Cryoglobulins are either immunoglobulins or a mixture of immunoglobulins and complement components
that precipitate out of blood at temperatures lower than 37°C
Type I (5%-25%): isolated monoclonal Ig (typically IgG or IgM, less commonly IgA or free immunoglobulin
light chains) typically Waldenström macroglobulinemia or MM
Type II (essential mixed cryoglobulinemia; 40%-60%): mixture of polyclonal Ig along with a monoclonal
immunoglobulin, typically IgM or IgA, with rheumatoid factor activity, often related to infections
Type III (40%-50%): mixed cryoglobulins consisting of polyclonal immunoglobulins, often secondary to
connective tissue diseases
Light chain deposition disease Granular deposits of monoclonal light chains, lacks the fibrillar ultrastructure of amyloid deposits that can
(LCDD) affect kidneys, heart, or liver
Associated with elevated free light chains, κ more common than λ (compared with amyloidosis, where λ is
more common)
Heavy chain deposition Rare condition associated with nonamyloid deposits arising from immunoglobulin light chains as well as
disease (HCDD) fragments of heavy chains
*Patients with serum IgM concentration <3.0 g/dL, in the absence of anemia, hepatosplenomegaly, lymphadenopathy, and systemic
symptoms and minimal or no lymphoplasmacytic infiltration of the bone marrow (<10%) are considered to have an IgM MGUS rather than
Waldenström macroglobulinemia.

MGUS | 643
Table 23-2 Uncommon PC proliferative disorders related to or associated with a monoclonal protein
Diagnosis Presenting features

Scleromyxedema Generalized papular and sclerodermoid cutaneous eruption with waxy firm papules and plaques with mucin
deposition, increased fibroblast proliferation, and fibrosis on histologic examination
Systemic manifestations may involve the cardiovascular, gastrointestinal, pulmonary, musculoskeletal, renal, or
nervous systems
Monoclonal gammopathy is usually IgG with a predominance of λ light chains
Capillary leak syndrome Rare disorder characterized by episodes of severe hypotension, hypoalbuminemia, and hemoconcentration
Elevated levels of vascular endothelial growth factor (VEGF) and angiopoietin 2
Typically, a prodromal phase is followed by an extravasation phase with edema, hypotension, and
hemoconcentration
Schnitzler syndrome Chronic urticaria associated with IgM monoclonal gammopathy typically IgM κ
May also have bone pain, skeletal hyperostosis, arthralgias, lymphadenopathy, and intermittent fevers
TEMPI syndrome Rare constellation of telangiectasias, erythrocytosis with elevated erythropoietin, MGUS, perinephric fluid
collections, and intrapulmonary shunting
Favorable responses have been with the proteasome inhibitor bortezomib
The diagnosis of MGUS, by definition, is made in an indi- important in young patients (especially those <50 years)
vidual with no symptoms attributable to PC proliferation with risk factors for disease progression (Table 23-3).
itself. Most often, MGUS is identified incidentally while indi- Once it is clear that the diagnosis is MGUS rather than a
viduals are undergoing work up for another condition that more aggressive PC disorder, an assessment of the risk of
has been associated with monoclonal gammopathy or dis- progression to MM or another related disorder such as AL
ease states in which PC disorders are part of the differential amyloidosis needs to be undertaken. Patients with MGUS
diagnosis. In the remaining individuals, monoclonal protein have a fixed ~1% per year risk of progression to myeloma or
testing is performed due to an elevated total protein detected another related disorder, translating to roughly 20% pro-
on a blood chemistry group or an elevated sedimentation gression rate at 25 years from diagnosis. Various prognostic
rate or detection of rouleaux on a peripheral blood smear. factors predicting for increased risk of disease progression,
Once the monoclonal protein is detected, it is important typically to MM, have been described (Table 23-3). At the
to ensure that the monoclonal protein is not associated with time of identification of the M-protein, it is often not clear
any of the disease states described above. The degree to what the trend in the laboratory values has been, and repeat
which evaluation is carried out depends on the clinical situ- testing in 3-6 months provides an important assessment of
ation, especially the quantity of monoclonal protein and the stability of the M-protein.
presence of any symptoms, and should at the minimum Management of MGUS is essentially limited to periodic
include a complete blood count (CBC), serum calcium, observation and monitoring for progression. Patients should
serum creatinine, serum FLC assay, and a 24 hour urine for
M-protein excretion. More detailed evaluation including a
Table 23-3 Factors associated with increased risk of progression in
bone marrow examination and imaging of the bones for lytic MGUS
lesions is not always required (particularly if the M-protein
1. Higher M-protein levels at diagnosis*
is small, the patient is completely asymptomatic) but may be
2. Non-IgG monoclonal protein*
3. Abnormal FLC ratio*
Tumor microenvironment 4. Percentage of PCs in bone marrow
MGUS SMM Symptomatic 5. Suppression of uninvolved immunoglobulins
MM 6. Presence of circulating PCs or clonal B-cells
7. Bone density abnormalities
8. Advanced age
9. Bence Jones proteinuria
Clonal PCs Clonal “malignant” PCs 10. Increasing M protein concentration
FLC = free light chain; PC = plasma cell; Ig = immunoglobulin;
M = monoclonal.
*A combination of M protein ≥1.5 g/dL or 15 g/L, non-IgG M protein
and an abnormal serum free light chain ratio (kappa:lambda ratio)
Figure 23-2 The transition from MGUS to myeloma. have been shown to identify persons at the highest risk of progression.

have periodic evaluation that includes a CBC, serum cal- characterized by increasing tumor burden as demonstrated by
cium, serum creatinine and a 24-hour urine for protein elec- increased M protein levels (serum M protein ≥3 g/dL or urinary
trophoresis. Risk-based monitoring is likely appropriate, monoclonal protein ≥500 mg per 24 h) or increasing marrow
with those at lower risk (ie, those with an IgG monoclonal plasmacytosis (clonal bone marrow plasma cells 10%–60%).
protein, <10 g/dL) having evaluations every 2 to 3 years with This phase in the disease evolution is characterized by increas-
the rest of the patients being followed annually. In young ing tumor burden that still do not lead to any end organ dam-
patients with several risk factors, some physicians prefer to age. In some patients, the first diagnosis of PC disorder may be
monitor every 6 months during the first 2 years after MGUS made at the smoldering (SMM) stage. It is likely that SMM does
is discovered. The adequate frequency of evaluation remains not represent a unique biological entity, but represents a transi-
controversial with at least one study suggesting that patients tional phase between preponderance of clonal yet benign PCs
may present with advanced myeloma between visits despite (MGUS) and clonal and malignant PCs (active MM).
optimal follow-up. In contrast to MGUS, patients who fit the criteria for SMM
are at a significant risk of progression to symptomatic MM.
The risk is estimated to be 10% per year for the first 5 years
Light-chain MGUS
from the time of initial recognition, and 4% per year over the
A small proportion of patients with monoclonal PC prolifera- subsequent 5 years and 1% per year after the first 10 years—
tion may not secrete an intact immunoglobulin monoclonal the latter a rate similar to what is seen in MGUS. As with
protein, but rather a κ or λ light chain only. Light-chain MGUS, multiple risk factors for progression of SMM to MM
MGUS has been defined as an abnormal free light-chain ratio have been described over the decades, and many of the charac-
with no heavy chain expression, plus increased serum concen- teristics overlap with those described for MGUS (Table 23-4).
tration of the involved light chain. In a large study, prevalence It should be noted that SMM is not a uniform entity, since
of light-chain MGUS was 0.8% (95% confidence interval [CI] it includes indolent low-risk forms that behave as MGUS and
0.7-0.9), contributing to an overall MGUS prevalence of 4.2% “early myelomas” at high risk of developing symptomatic
(3.9-4.5). The median age of patients with light-chain MGUS diseases. This heterogeneous outcome is supported by the
was 68 (range 50.0-96.0) years, compared with 70 (50.0-96.0) identification of the above-mentioned risk factors predicting
years for those with conventional MGUS. Risk of progression progression to symptomatic MM. The Mayo Clinic risk-
to MM in patients with light-chain MGUS was 0.3 (0.1-0.8) classification proposes three different subgroups of SMM:
per 100 person-years, in contrast to 1 per 100 person years for group 1 with ≥3 g/dL of monoclonal protein and ≥10% of
conventionally defined MGUS. All progressions of light-chain PCs in bone marrow, in which the median time to progres-
MGUS were to light-chain MM. Renal disease was relatively sion (TTP) to symptomatic MM is 2 years; group 2 with
more frequent in this population, with 23% of 129 patients <3 g/dL of monoclonal protein and ≥10% bone marrow PCs
with light-chain MGUS being diagnosed with renal disease. M-protein with a median TTP of 8 years; and group 3 with
≥3 g/dL of monoclonal protein but with <10% PCs bone
marrow infiltration, translating into a median TTP of 19
Biclonal gammopathies
years. The Spanish group has proposed a risk-classification
Simultaneous presence of more than one type of M-protein
can be seen in as many as 5% of patients with monoclonal Table 23-4 Factors associated with increased risk of progression of
gammopathies. This situation likely represents the prolifera- SMM to MM
tion of two separates clones of PCs, producing M-proteins of 1. Higher M-protein levels at diagnosis
different immunoglobulin classes and often with different 2. Abnormal FLC ratio
light chain. In one study, 20 patients (2%) from among 1,034 3. Percentage of PCs in bone marrow
patients with monoclonal gammopathy had two distinct 4. Suppression of uninvolved immunoglobulins
monoclonal spikes; 3 were associated with lymphoma, 7 5. Presence of circulating PCs
with myeloma, 9 with MGUS, and 1 with an autoimmune 6. A high predominance of abnormal PCs (≥95%) (defined by
rheumatologic condition. phenotype and flow-based assessment) from the total PCs in
the marrow
7. Presence of FISH abnormalities (t(4;14), deletion 17p, gain
Smoldering (or asymptomatic) MM 1q21, and hyperdiploidy)
8. IgA isotype
Development of MM is invariably preceded by a distinct MGUS 9. Evolving M-component
phase, as has been demonstrated by recent studies. However, an FLC = free light chain; FISH = fluorescent in situ hybridization;
intermediate phase can be identified in a number of patients, M = monoclonal.

Active (symptomatic) MM | 645
based on the percentage of PC with aberrant phenotype monoclonal light chain in the urine (≥0.2 g/24 h), in the
(high risk if ≥95% of the total PC are clonal) plus immune absence of intact monoclonal immunoglobulin in the serum,
paresis (decrease in one or two of the uninvolved immuno- and no evidence of MM, light-chain amyloidosis, or other
globulins), with a median TTP of 23 months when the two related plasma-cell proliferative disorders. In a study of 101
risk factors were present, as compared with 73 months when patients with IBJP from the Mayo Clinic, 27 patients (27%)
only one risk factor was present and not reached when none developed MM and seven (7%) developed light-chain amy-
of the risk factors was present. Other adverse risk factors loidosis during 901 person-years of follow-up. The major
include an abnormal serum free light-chain (FLC); evolving risk factors for progression included the amount of urinary
levels of paraprotein; and cytogenetic abnormalities such as M protein excretion, bone marrow PC percentage, abnormal
t(4;14), gain of 1q21 or hypodiploidy. Upon using either the free-light-chain ratio (<0.01 or >100), and reduction of all
Mayo or Spanish criteria, the high-risk SMM patients have a three uninvolved immunoglobulins. An entity of light-chain
70% risk of progression at 3 years. Moreover, more sensitive smoldering MM (LC SMM) was defined as presence of
criteria can be used to identify patients at “ultra-high risk of monoclonal light-chain excretion of 0.5 g/24 h or greater or
progression” (70% transform at 2 years): presence of FLC at least 10% bone marrow PCs, or both, in the absence of
ratio ≥100, more than 60% of BMPCs, focal lesions on spinal end organ damage. The cumulative probability of progres-
magnetic resonance imaging. These patients are now consid- sion to active MM or AL amyloidosis in this cohort was
ered as early myeloma, and they should be candidates for 27.8% at 5 years, 44.6% at 10 years, and 56.5% at 15 years.
immediate treatment. Patients who meet the criteria for LC SMM clearly have a
The current standard of care for patients with SMM significant risk of progression, albeit less than conventionally
remains careful observation. However, the above mentioned defined SMM, and require close follow-up.
risk stratification classifications may allow a more individu-
alized disease management, and accordingly, there is an
increasing interest in exploring the concept of therapeutic Active (symptomatic) MM
intervention in patients with SMM, particularly in those
with high-risk factors. The long period of development of MM is characterized by the proliferation of a single clone of
MM, starting with MGUS and proceeding through a smol- PCs that produce a monoclonal protein. The PC prolifera-
dering phase provides a unique opportunity for intervention tion results in extensive skeletal involvement, with osteolytic
that may potentially diminish or eliminate the progression lesions, hypercalcemia, anemia, or soft tissue plasmacyto-
to symptomatic disease. Until recently, this had not been mas. In addition, the excessive production of nephrotoxic
pursued in clinical trials due to lack of effective therapies monoclonal immunoglobulin can result in renal failure and
with favorable side effect profile. With introduction of the an increased risk of developing potentially life-threatening
new drugs, this paradigm has shifted, and interventional infections due to the lack of functional immunoglobulins.
clinical trials are being conducted in patients with SMM who
are at the highest risk of progression to MM.
Epidemiology
Initial clinical trials using thalidomide to prevent SMM to
MM transition did not lead to any significant benefit and MM accounts for 1% of all malignancies and 10% of all
was limited by drug-associated toxicities such as peripheral hematological malignancies. The 2015 annual estimate in
neuropathy. Recently, a phase 3 trial compared lenalidomide the United States for new cases of MM is 21,700 and for
and dexamethasone with the current standard of observa- deaths is 10,710. SEER (surveillance, epidemiology, and end
tion among patients with high risk SMM demonstrated a results) data incidence age-adjusted rates from 1992 through
significantly decreased risk of progression to MM with active 1998 show an overall incidence of 4.5 per 100,000 per year,
therapy, and more importantly, an improved overall survival with the incidence among whites being 4.2 per 100,000 per
with this intervention. Additional trials examining similar year and among African Americans 9.3 per 100,000 per year.
approach in larger trials as well as more intense approaches In contrast to the higher incidence in persons of African
in smaller patient groups are ongoing. descent, the incidence of MM is lower in Asians and His-
panic populations, and there is a slight male preponderance.
The median age at diagnosis of MM is 65-70 years.
Idiopathic Bence Jones proteinuria and
light-chain SMM
Etiology
Patients may present with isolated monoclonal free light-
chains in the urine, or Bence Jones proteinuria. Idiopathic The cause of MM is unknown. Radiation may play a role
Bence Jones proteinuria (IBJP) has been defined as and an increased incidence was reported in the atomic

bomb survivors. Increased incidence has been associated Genomic abnormalities

with exposure to benzene and other organic solvents as
Nearly all MM cells harbor multiple genetic abnormalities,
well as petroleum products. An increased risk has been
and there is significant heterogeneity between the PCs in any
reported in farmers, particularly those who use herbicides
given individual highlighting the genetic instability and the
and insecticides. Chronic antigen stimulation has been
clonal evolution that accompanies disease progression.
suggested as a cause and infection with HHV-8 has been
Genetic abnormalities in MM can be broadly categorized
postulated as an etiologic agent, though the evidence
into chromosomal abnormalities including translocations
remains inadequate. A relationship between MM and pre-
mainly involving the IGH locus on chromosome 14q32 and
existing inflammatory diseases has been suggested, and PC
trisomies of multiple chromosomes, point mutations (which
dyscrasias associated with chronic antigen stimulation
can involve multiple genes), epigenetic changes involving
have been reported. Genetic factors have been implicated
methylation and histone modifications, and gene and
based on multiple family members affected by MM, and
microRNA (miRNA) dysregulation.
studies have shown a two-fold increased risk of MGUS or
myeloma among first-degree relatives affected by MM. It
remains unclear how much of this is related to the genetic IGH translocations
factors or shared environmental exposures. A range of sin-
gle nucleotide polymorphism (SNP) variants as well as cer- Translocations in the myeloma cell typically affect the
tain HLA subtypes have been linked to increased risk of immunoglobulin-heavy-chain region, likely reflecting the
developing MM. genetic instability in this chromosomal region associated
with the class switching that occurs during PC development.
These translocations lead to placement of an oncogene under
Pathogenesis the control of an immunoglobulin promoter region leading
As mentioned above, MM is always preceded by an MGUS to their activation. About 40% of MM tumors have IGH
phase which, based on recent studies, is likely to be present translocations involving five recurrent chromosomal part-
for 8-10 or more years prior to onset of symptomatic MM. ners leading activation of specific oncogenes: chromosome
The specific changes that lead to the evolution of MM from 11q13 (CCND1), 4p16 (FGFR/MMSET), 16q23 (MAF),
MGUS remain unclear but likely reflect the cumulative effect 6p21 (CCND3), and 20q11 (MAFB).
of changes within the myeloma cell as well as the tumor
microenvironment. Until recently, the pathogenesis models
Gains and losses of chromosomal material
assumed that MM develops through a multistep transforma-
tion from normal PCs to MGUS (implying PC immortaliza- Patients with MM can be grouped into two major categories
tion) and subsequent transformation into active MM, in according to ploidy status assessed by karyotyping or by flow
which clonal PCs are responsible for end-organ damage. based approaches: the hyperdiploid group (more than 46/47
However, studies based on fluorescent in situ hybridization chromosomes but <75 chromosomes) and the nonhyperdip-
(FISH) panels of MM-associated genetic abnormalities, SNP loid group, composed of hypodiploid (up to 44/45 chromo-
arrays and whole-genome sequencing have demonstrated somes); diploid or pseudodiploid (44/45 to 46/47); and near
that most genetic lesions typically observed in MM are tetraploid (more than 74 chromosomes) subgroups. Nonhy-
already present in MGUS patients, and that the progression perdiploid MM is characterized by a very high prevalence of
from MGUS to SMM and eventually to MM involving a IGH translocations involving the five recurrent partners
clonal expansion of a genetically abnormal PCs, implying a described above. Likewise, monosomy/deletion 13 and gains
complex evolutionary process with intraclonal heterogene- on 1q occur predominantly in nonhyperdiploid MM. In
ity. Three distinct patterns of genomic evolution have been contrast, the hyperdiploid group is associated with recurrent
proposed, based on data generated by new genomic trisomies involving odd-numbered chromosomes (ie, 3, 5, 7,
approaches: stable genomes, without major differences 9, 11, 15, and 19) and a low incidence of structural chromo-
between diagnosis and progression clones; linear evolution, somal abnormalities.
in which the progression clone apparently derives from the Loss of chromosome 13 is the most common monosomy
major clone at diagnosis, but continues to diversify through in MM and is strongly associated with IGH translocations,
additionally acquired lesions; and branching (nonlinear) deletion of 17p, and gains on 1q (40%-50% of newly diag-
models, where a progression clone derives from a minor nosed patients). The next common is the loss of the short
subclone that is barely present at diagnosis. Patients with arm of chromosome 17, which leads to the loss of TP53,
MGUS with high-risk cytogenetics usually follow the last among other genes. del17p is less common and can be seen
two evolutionary models. in 5%-10% of patients with newly diagnosed MM but is seen

in higher frequency in relapsed disease, especially in patients are considered late oncogenic events and are associated with
treated with alkylating agents. disease progression. Dysregulation of MYC by chromosome
Recent studies have demonstrated lesions of chromosome translocation is a common secondary event in MM. Most
1 as one of the most common abnormalities in MM; mostly karyotypic abnormalities involving MYC correspond to
they are 1q gains, as the result of tandem duplications and complex translocations and insertions that often are nonre-
jumping segmental duplications of the chromosome 1q ciprocal and frequently involve three different chromo-
band. Recently, studies suggested that 1p losses (especially somes. Activating RAS mutations are also considered
1p22 and 1p32 deletions) are also frequent in MM patients, molecular markers of disease progression. Thus, the preva-
but the key genes are not yet clear. lence of activating KRAS and NRAS mutations is over 75%
in MM cases at relapse but <25% at initial diagnosis. TP53
inactivation, via either deletion or mutation, seems to be
Mutations detected by whole-genome sequencing
more frequently associated with disease progression. Meth-
Whole-genome sequencing strategies have shown that there ylation of CDKN2B and CDKN2A is more prevalent in
are approximately 35 nonsynonymous mutations per myeloma advanced MM and extramedullary forms of the disease.
simple. A recent study including 203 patients has shown that
131 (65%) had evidence of mutations in one or more of 11
Interaction between PCs and their
recurrently mutated genes: ACTG1, RB1, CYLD, PRDM1,
microenvironment
TRAF3, BRAF, FAM46C, DIS3, TP53, NRAS, KRAS. Mutations
were often present in subclonal populations, and multiple The interaction between the MM cells and the tumor micro-
mutations within the same pathway (eg, RAS and BRAF) were environment, typically the bone marrow, is key to disease
observed in the same patient. The complex MM mutational sig- pathogenesis and progression. In the marrow, MM cells
nature is similar to other hematological malignancies, such as adhere to extracellular matrix proteins and bone marrow
acute myeloid leukemia, but is in contrast to hairy cell leukemia stromal cells through a series of adhesion molecules present
and WM, in which single unifying mutations are seen (BRAF on the MM cells or the stromal cells or endothelial cells in
and MYD88, respectively). the marrow. In addition, marrow stromal cells produce a
stromal cell-derived factor (SDF)-1 that binds to CXCR4 on
the surface of myeloma cells, inducing homing of the PCs
Epigenetic modifications and microRNA expression
into the marrow.
Information on the role of epigenetics on MM pathogenesis Adhesion of myeloma cells to the bone marrow microen-
is as yet limited. Silencing of certain tumor suppressor genes vironment induces a cell adhesion-mediated drug resis-
(GPX3, RBP1, SPARC, CDKN2A, SOCS, TGFBR2), overex- tance phenotype through different mechanisms such as
pression of the histone methyltransferase MMSET, and the induction of cell cycle arrest and apoptosis inhibition. This
presence of mutations of UTX (histone demethylase) have is mediated at least in part through the induction of secre-
been described. Furthermore, genome-wide methylation tion of cytokines, such as tumor necrosis factor α (TNF-α),
studies have shown both global DNA hypomethylation and interleukin-6 (IL-6), insulin-like growth factor (IGF)-1,
gene-specific DNA hypermethylation in MM, with certain IL-21, SDF-1α, and vascular endothelial growth factor
epigenetic signatures being associated with prognostic cyto- (VEGF), by PCs and/or bone marrow stromal cells brought
genetic groups. Various studies have shown that microRNAs about by the adhesion of MM cells to stromal cells and
(miRNAs) expression is deregulated in myeloma cells com- endothelial cells. These cytokines in turn triggers signaling
pared with normal PCs and that their expression profile is pathways such as RAF/MEK/MAPK, PI3K/AKT, NF-kB,
associated with genetic abnormalities. Moreover, several and JAK/STAT, all of which mediate cell proliferation and
miRNAs are known to be involved in MM pathogenesis. prevent apoptosis.
Indeed, a mechanism has been identified by which miRNAs Another characteristic finding at all stages of monoclonal
act on MDM2 expression to regulate p53; thus, miR-192, gammopathies is increased angiogenesis, reflected by
194, and 215 re-expression in myeloma cell lines induces increased microvascular density in the bone marrow. The
degradation of MDM2, with subsequent up regulation of MVD not only increases with dose progression, it also
p53 and inhibition of cell growth. increases the risk of progression from MGUS to MM.
Increased MVD correlates with more aggressive disease
characteristics such as increased MM cell proliferation and
Late genetic events
increasing circulating cells. Increased MVD has also been
Some genetic changes in MM, such as secondary transloca- seen in the extramedullary disease and in isolated plasmacy-
tions, mutations, deletions, and epigenetic abnormalities, tomas is associated with increase in the risk of progression to

MM. The increased angiogenesis is driven by the M cells Table 23-5 Clinical presentation of MM
through a combination of proangiogenic and antiangiogenic Symptoms and signs Mechanism
factors, the overall balance between the two determining the
Anemia Marrow infiltration, direct destruction of
findings. erythroblasts, anemia of renal failure
Easy bruising and Thrombocytopenia, acquired von
bleeding Willebrand disease or inhibition of
MM and bone disease other clotting factors by monoclonal
protein
Bone disease is a hallmark of MM and is a common disease
Bone pain Lytic bone lesion, pathologic fractures
defining event, taking the shape of diffuse osteopenia, dis-
Fatigue Anemia, hyperviscosity, renal failure
crete lytic lesions or large destructive lesions leading to path-
Recurrent infections Hypogammaglobulinemia, suppressed
ological fractures. Overall, the bone disease seen in MM cellular immunity, neutropenia
reflects an altered balance between the osteoblastic and Altered mental status, Hypercalcemia, hyperviscosity
osteoclastic activity. Interaction of MM cells with stromal confusion
cells and other cells in the microenvironment induces the Neurological deficits Cord compression due to paraspinal mass/
secretion of numerous osteoclast-activating factors such as vertebral fractures, nerve compression
RANK-L (receptor activator of NF-κB ligand), macrophage from plasmacytomas
inflammatory protein (MIP)-1α, activin A, VEGF, hepato-
cyte growth factor (HGF), IL-3, IL-7, TNF-α, IL-6, IL-1β,
severity of disease related complications, and gathering the
and MIP-3α.
data for risk stratification. The typical testing associated with
Among these, RANK-L and MIP-1α play a key role in the
the initial work up is detailed in Table 23-6.
pathogenesis of bone disease. RANK-L binds to its functional
Newer imaging techniques have greater sensitivity com-
receptor RANK (TNF receptor superfamily) on osteoclasts,
pared with radiographic bone survey for detection of MM
stimulating osteoclastogenesis and inducing bone resorp-
bone lesions. Computed tomography (CT) has the highest
tion. Osteoprotegerin (OPG), a soluble decoy receptor for
sensitivity for the detection of bone lesions and with the
RANK-L, provides a counterbalance to the activity of RANK.
whole body low-dose modality the radiation exposure is
In MM, this balance is disrupted by increased expression of
much lower than with conventional CT. Magnetic resonance
RANK-L and decreased expression of OPG on stromal cells
imaging (MRI) has the highest resolution for soft tissue and
as a result of their interaction with MM cells. MIP-1α
bone marrow infiltration, but it is inferior to CT for assess-
potently stimulates osteoclast formation through enhancing
ment of bone disease. Finally, positron emission tomogra-
the activity of RANK-L and directly stimulating osteoclast
phy (PET) allows assessment of tumor metabolism and
differentiation. Along with the osteoclast activation, there is
disease activity. CT, MRI, and PET are not yet routine in
inhibition of osteoblast formation and function, mediated
MM evaluation but are widely employed in specific clinical
both by both soluble factors and direct cellular interactions
circumstances.
of MM and stromal cells. MM cells produce numerous solu-
ble factors which inhibit osteoblast differentiation and/or
function, such as Wnt signaling antagonists like DKK1, Table 23-6 MM: Important tests in evaluation
sclerostin, and soluble frizzled related proteins (sFRP-2/3), Complete blood count, including differential to assess for circulatory
and bone morphogenetic protein (BMP) inhibitors such as PCs
activin A. Chemistry with BUN, creatinine, calcium, CRP, LDH
Serum protein electrophoresis with immunofixation
Quantitative immunoglobulins
Clinical presentation and diagnostic 24-hour urine protein electrophoresis with immunofixation
considerations Serum free light chain
Skeletal survey (plain films or whole body low dose CT)
The symptoms of MM are usually nonspecific and can Serum β2-microglobulin, albumin
include fatigue, bone pain, easy bruising and bleeding, recur- Bone marrow aspirate and biopsy
rent infections, or symptoms related to the so-called “CRAB” Cytogenetics/FISH
(calcium elevation, renal dysfunction, anemia, and bone dis- MRI,* PET scan*
ease) features. The common symptoms and the underlying BUN = blood urea nitrogen; CRP = C reactive protein; FISH =
pathology are detailed in Table 23-5. fluorescence in situ hybridization; LDH = lactate dehydrogenase.
The initial workup should be aimed at confirming the *MRI and PET scans are used in specific circumstances and are not
diagnosis, estimating the tumor burden, assessing the routinely performed in all patients.

Staging and risk stratification: prognostic factors t(4;14), t(14;16), and t(14;20) have very poor outcomes com-
pared to the rest of the patients and these patients are consid-
As with all malignancies, staging systems have been devel-
ered to have high-risk MM, while those with trisomies have
oped for MM. One of the most widely used was the Durie-
the best outcomes. Recent studies have suggested that chro-
Salmon staging system (DSS), which has been replaced to a
mosome 1 abnormalities such as 1q amplification and 1p
large extent by the International Staging System (ISS) (see
deletion may also indicate a poorer outcome.
below). The staging system is purely a prognostic classifica-
Response to frontline therapy is one of the most important
tion and does not necessarily guide therapy. In addition to
prognostic factors in most hematological malignancies—
the ISS, other risk stratification systems have been proposed,
myeloma being no exception—whereby the better the qua
incorporating other prognostic factors known to be relevant
lity of the response the longer the survival. Patients achieving
in this disease. Well-accepted prognostic factors and risk
complete response (CR) display significantly longer survival
stratification systems are detailed in Tables 23-7 and 23-8.
as compared to partial responders (PR); moreover, patients
MM is a heterogeneous disease in terms of outcomes, with
failing to achieve at least PR with an optimized induction
nearly a quarter of patients dying within the first 2-3 years
therapy should be considered as high-risk patients with a
following diagnosis and a similar fraction living >10 years. In
survival of <2 years. However, the definition of CR is far
recent times, there has been increased effort towards identi-
from optimal, and more sensitive techniques for evaluating
fication of the patients with high-risk myeloma. The main
minimal residual disease (MRD) both outside the BM (eg,
drivers of the heterogeneity in outcome are the genetic
imaging techniques such as MRI or PET) and inside the BM
abnormalities seen in myeloma. Patients with del17p,
(eg, immunophenotyping by multiparametric flow cytome-
try, MFC, or molecular analysis by allele specific oligonucle-
Table 23-7 Prognostic factors in MM otide polymerase chain reaction [ASO-PCR] or next
Tumor-related prognostic factors Host-related prognostic factors generation sequencing [NGS]) are needed.
FISH: del17p, t(4;14), t(14;16), Advanced age The Italian and Arkansas groups have shown that failure
and t(14;20) to achieve complete PET suppression after autologous trans-
High lactate dehydrogenase level Poor performance status plant is associated with shorter survival. Using MFC, the
High β2-microglobulin level Renal failure Spanish and UK groups have both shown that, in trans-
(International Staging System planted as well as in elderly MM patients, persistence of
stage III) MRD is associated with significantly poorer outcome, and
Presence of circulating PCs that this parameter is of significantly more prognostic power
High PC proliferative rate (eg, than negative immunofixation; however, further standard-
measured by labeling index) ization of MFC is still required. ASO-PCR also predicts
Presence of extramedullary disease outcome and is probably one log more sensitive than MFC,
Table 23-8 Risk stratification systems for MM
System Stage I Stage II Stage III

Durie-Salmon staging system (DSS)
Parameters Calcium normal or <12 mg/dL, hemoglobin Neither stage I nor stage III One of the following: hemoglobin value
value >10 g/dL, normal skeletal survey <8.5 g/dL, calcium >12 mg/dL, multiple
or solitary plasmacytoma, low M protein lytic lesions, high M component with
with an IgG <5g/dL or IgA <3 g/dL, an IgG >7 g/dL or IgA >5 g/dL, Bence
Bence Jones protein >4 g/24 h Jones protein >12 g/24 h
International Staging System (ISS)
Parameters β2M ≤3.5 mg/dL and albumin ≥3.5 g/dL β2M >3.5 mg/dL but <5.5 mg/dL, β2M ≥5.5 mg/dL
or albumin >3.5 g/dL
Median OS 62 months 44 months 29 months
International Myeloma Working Group (IMWG) consensus criteria
High risk Standard risk Low risk
Parameters ISS II/III and t(4;14) or 17p13 del All others ISS I/II and absence of t(4;14), 17p13 del
and +1q21 and age <55 years
Median OS 2 years 7 years >10 years
Proportion 20% 60% 20%
β2M = β2-microglobulin.

but not every patient has a rearrangement or mutation sus- The International Myeloma Working Group (IMWG) has
ceptible to ASO-PCR (50% testable by ASO-PCR vs. 95% by developed a set of uniform response criteria for disease
MFC). Preliminary data from NGS indicate high applicabil- assessment in MM (Table 23-9).
ity and sensitivity making the technique a possible alterna-
tive to MFC. It should be noted that due to the patchy pattern
Initial therapy for newly diagnosed MM
of myeloma BM infiltration, a negative MRD result might
not be indicative of disease eradication but may instead be It is important to ensure that the patient requires therapy
the result of a nonrepresentative BM sample. from the underlying PC disorder; currently, patients with
SMM should not be treated outside of a clinical trial. The
goals of the initial therapy are to control the disease process
Treatment approaches for MM
as rapidly as possible and to reverse the complications of the
The treatment paradigms for MM have changed dramatically disease while minimizing the toxicity and allowing collection
during the past decade as result of improved understanding of stem cells for autologous stem cell transplantation in
of the disease biology, better risk assessment, availability of patients eligible to undergo high dose therapy and autolo-
more effective antimyeloma agents, systematic use of autologous stem cell transplantation (HDT/ASCT). The early and
gous stem cell transplantation, and better appreciation of the rapid control of disease without significant toxicity has played
importance of supportive care. The median overall survival an important role in reducing the early mortality in MM.
in MM has improved two- to three-fold in the past decade, Despite these uniform goals, substantial differences exist
according to data from single institution studies as well as in terms of the approaches to initial therapy of myeloma,
large national databases such as SEER. The overall approach and unfortunately limited data are available from random-
and goals are summarized in Figure 23-3. A sequential ized trials to provide firm guidance. Traditionally, the initial
approach to newly diagnosed MM includes risk stratifica- therapy of myeloma has been based on whether patients
tion/ prognostication, immediate interventions for reversal would be considered eligible for HDT/ASCT, in order to
of acute disease-related complications, initiation of systemic reduce the likelihood of compromised stem cell collection
therapy with the goal of maximizing the response benefit, from drugs such as melphalan or prolonged use of other
consolidation and maintenance strategies designed to agents. Definition of HDT/ASCT eligibility varies across
improve the depth and duration of the response achieved regions and practices and is typically based on age and per-
initially and consistent use of supportive care strategies along formance status. However, many of the currently used initial
the entire course. Unfortunately, in the current era, the vast treatment regimens incorporating the newer drugs do not
majority of patients relapse after initial disease control and significantly impact the ability to collect stem cells and as a
additional therapies will need to be employed for continued result the need to classify patients based on the transplant
control of MM. eligibility has diminished over time. Many combination reg-
Assessment of response to therapy is typically performed imens have been studied during the initial treatment phase
through serial assessment of the monoclonal protein levels of MM, incorporating the old and the new drugs, and the
in the serum or urine, with bone marrow or plasmacytoma most commonly used regimens are discussed below in detail
evaluation utilized in patients with nonsecretory disease. and the individual drugs and classes are listed in Table 23-10.
Figure 23-3 Stages in the initial

risk stratification
SCT management of MM.

Diagnosis and
Induction Consolidation Maintenance

eligible
SCT
Induction followed by continuous therapy
ineligible
Tumor burden
Supportive care

Table 23-9 Response and relapse definitions (IMWG)
Response or category Definition

Response categories
Stringent complete CR as defined below, PLUS
Response (sCR) Normal FLC ratio AND
Absence of clonal cells in bone marrow by immunohistochemistry or immunofluorescence‡
Complete response (CR) Negative immunofixation on the serum and urine AND
Disappearance of any soft tissue plasmacytomas AND
<5% PCs in bone marrow
Very good partial Serum and urine M-protein detectable by immunofixation but not on electrophoresis OR
response (VGPR) ≥90% reduction in serum M-protein PLUS
urine M-protein level <100 mg per 24 hours
Partial response (PR) ≥50% reduction of serum M-protein and reduction in 24-h urinary M-protein by ≥90% or to <200 mg/24 h if
the serum and urine M-protein are not measurable, a ≥50% decrease in the difference between involved and
uninvolved FLC levels is required in place of the M-protein criteria
If serum and urine M-protein are not measurable, and serum free light assay is also unmeasurable, ≥50%
reduction in PCs is required in place of M-protein, provided baseline bone marrow PC percentage
was ≥30%
In addition to the above listed criteria, if present at baseline, a ≥50% reduction in the size of soft tissue
plasmacytomas is also required
Stable disease (SD) Not meeting criteria for CR, VGPR, PR or progressive disease
Progression categories
Progressive disease Any one or more of the following:
Increase of ≥25% from baseline in:
Serum M-component and/or (the absolute increase must be ≥0.5 g/dL)
Urine M-component and/or (the absolute increase must be ≥200 mg/24 h)
Only in patients without measurable serum and urine M-protein levels: the difference between involved and
uninvolved FLC levels. The absolute increase must be ≥10 mg/dL
Bone marrow PC percentage: the absolute percentage must be ≥10%
Definite development of new bone lesions or soft tissue plasmacytomas or definite increase in the size of
existing bone lesions or soft tissue plasmacytomas
Development of hypercalcemia (corrected serum calcium ≥11.5 mg/dL or 2.65 mmol/L) that can be attributed
solely to the PC proliferative disorder
Clinical relapse One or more of the following:
Development of new soft tissue plasmacytomas or bone lesions
Definite increase in the size of existing plasmacytomas or bone lesions. A definite increase is defined as a 50%
(and at least 1 cm) increase as measured serially by the sum of the products of the cross-diameters of the
measurable lesion
Hypercalcemia (≥11.5 mg/dL)
Decrease in hemoglobin of ≥2 g/dL
Rise in serum creatinine by ≥2 mg/dL
Relapse from CR One or more of the following:
Reappearance of serum or urine M-protein by immunofixation or electrophoresis
Development of ≥5% PCs in the bone marrow
Appearance of any other sign of progression (i.e., new plasmacytoma, lytic bone lesion, or hypercalcemia)
MRD, minimal residual disease; CR, complete response; FLC, free light chain; PR, partial response; SD, stable disease; sCR, stringent complete
response; VGPR, very good partial response.
Durie BG, et al. Leukemia. 2006;20:1467-1473.
* All response categories require two consecutive assessments made at anytime before the institution of any new therapy; all categories also
require no known evidence of progressive or new bone lesions if radiographic studies were performed.
† Confirmation of bone marrow assessment is not required.
‡ Presence/absence of clonal cells is based upon the k/l ratio. An abnormal k/l ratio by immunohistochemistry and/or immunofluorescence
requires a minimum of 100 PCs for analysis. An abnormal ratio reflecting presence of an abnormal clone is k/l of >4:1 or <1:2.

Table 23-10 Commonly used classes of drugs in myeloma and new or better observed in 32%-38%. Introduction of the novel
investigational drugs drugs allowed us to reduce the intensity of dexamethasone
Class Common drugs dose as shown by a randomized controlled trial comparing
Currently approved drugs
lenalidomide with standard dexamethasone (RD with capi-
Alkylating agents Melphalan, cyclophosphamide, tal letter D; dexamethasone on days 1-4, 9-12, and 17-20 of a
bendamustine, carmustine (BCNU) 28 day schedule) with lenalidomide with reduced intensity
Corticosteroids Prednisone, methylprednisolone, dexamethasone (Rd with lowercase letter d to indicate lower
dexamethasone dose; weekly dexamethasone). After 4 months of therapy,
Anthracyclines Doxorubicin 79% of the RD patients and 68% of the Rd patients had
IMiDs Thalidomide, lenalidomide, achieved a partial response or better; however, at 1 year, OS
pomalidomide was superior in the Rd arm as compared to the RD arm (92%
Proteasome Inhibitors Bortezomib, carfilzomib, ixazomib versus 87%). Grade 3-4 AEs and early deaths were higher in
Histone deacetylase Panobinostat the RD group with the most common serious toxicities being
inhibitors DVT, infections, and fatigue.
Monoclonal antibodies Daratumumab (anti-CD38), elotuzumab Bortezomib was also studied in combination with dexa-
(anti-CS1 or signaling lymphocytic
methasone and compared to VAD as induction therapy
activation molecule F7 [SLAMF7])
prior to SCT in a phase 3 trial, resulting in deeper responses
Representative drugs currently in clinical trials
and reduced need for tandem ASCT as well as improved PFS
Oral proteasome inhibitors Oprozomib
post-SCT. In the current era, bortezomib tends to be used
Spindle kinesin inhibitor Filanesib
more in combination with cyclophosphamide, lenalido-
Bcl2 inhibitor ABT199 (venetoclax)
mide, or thalidomide, as described below.
Akt inhibitors Afuresertib
The Italian and Spanish groups have reported that the
Cdk inhibitors Dinaciclib
triplet VTD is significantly superior to TD both before and
after transplant and this triple combination also appears
to be superior to VDd. Moreover, VTD was not only asso-
Treatment of transplant-eligible patients
ciated with higher responses but also a significant prolon-
Prior to initiation of stem cell collection, patients with newly gation in PFS, reaching a median of 56 months in the
diagnosed MM receive induction therapy for 3-6 cycles, with Spanish study. Another initial regimen that is widely used
the intent of disease control and reduction of tumor burden, is bortezomib (Velcade)-cyclophosphamide (Cytoxan)-
and improvement in end-organ damage observed at diagno- dexamethasone (VCD or CyBorD), resulting in an overall
sis. Until the introduction of newer therapies, the combina- response rate pretransplant of almost 90% but with a
tion of vincristine, doxorubicin, and dexamethasone (VAD) lower CR rate (-20%). Recent data using VRD indicate
was the gold standard as a preparatory regimen for young that most patients (>90%) achieve at least PR with over
newly diagnosed MM patients who are candidates for HDT, 20% stringent CR (sCR). A recent meta-analysis of four
with partial response (PR) rates ranging between 52% and European randomized trials (IFM, HOVON/GMMG,
63%, and CR rates from 3%-13%. However, novel drug PETHEMA/GEM, and GIMEMA), including 1,572
combinations are superior to VAD or VAD-like regimens for patients, supports the use of bortezomib containing regi-
initial tumor reduction, and VAD is therefore primarily of mens as induction therapy before HDT. Bortezomib-based
historical interest. regimens were superior in terms of response rate, EFS, and
Initial trials with the newer drugs beginning in the 1990s OS to non-bortezomib-based regimens. Thus, CR rate
evaluated combination with dexamethasone or alkylating pretransplant was 14% versus 4%, and the CR rate post-
agents. Three randomized trials compared thalidomide (T)- transplant was 26% versus 14% for bortezomib-based and
based regimens (TAD, T+VAD or cyclophosphamide non-bortezomib-based regimens, respectively. The PFS
[C]TD) versus VAD as initial therapy in transplant eligible was 35.9 months versus 28.6 months (P < 0.001) and the
patients, and have shown superiority for the triplets of tha- 3-year OS was 79% versus 74% (P = 0.04) (median not
lidomide combinations with approximately 80% with PR, reached) in favor of the bortezomib-based regimens.
including 10%-20% CR. In contrast, the efficacy of thalido- Similar efficacy, if not higher, has been reported in pilot
mide plus dexamethasone is similar to that of VAD. studies for the combination of carfilzomib (second generation
Lenalidomide and dexamethasone (Rd) was subsequently proteasome inhibitor) with lenalidomide plus dexametha-
evaluated in similar studies. In previously untreated patients, sone, with up to 50% CR rate in newly diagnosed patients.
initial therapy with Rd results in overall response rates of The combination of carfilzomib with thalidomide and dexa-
91%-95%, with IMWG very good partial response (VGPR) methasone has also been tested with promising results.

The incorporation of the novel drugs such as immuno- controversial results, HDT/ASCT is considered the standard
modulatory drugs (IMiDs) and the proteasome inhibitors of care for patients in MM <70 years and is often used in
have led to unprecedented response rates and response depth older patients with good performance status. A large study
compared to older alkylator and steroid based therapies. with long follow-up, which included studies from European
Moreover, combination regimens that include an IMiD and and United States cooperative groups and institutions and
a proteasome inhibitor has led to very high response rates pooled data from more than 10,000 MM patients diagnosed
but at the cost of higher toxicity rates compared to combina- between 1982 and 2002, showed a important significant
tions with one or the other. Considerable controversy exists reduction in risk of death among patients treated with HDT
as to whether a doublet regimen (VD, Rd, TD) or a triplet compared to those who received standard chemotherapy.
regimen (VTD, VRD, VCD, PAD) should be used for induc- In the setting of novel agents, it is important to redefine if
tion. It is clear that triplets are more effective than doublets HDT enhances the response rates obtained with the new
in inducing a VGPR/CR rate before HDT and is associated induction regimens, which are associated with higher
with improved survival. This does not necessarily mean that response rates and deeper responses than was seen in the ear-
if patients do not achieve a VGPR following induction that lier studies of HDT. Analysis of recent trials clearly demon-
the treatment should be changed before proceeding to trans- strates an increase in CR rates following HDT, suggesting
plant (with the exception of those that achieve <PR, that that induction with novel agents and HDT are complemen-
should be considered as primary refractory and may benefit tary rather than alternative treatment approaches.
from a rescue regimen prior to high dose melphalan). Melphalan, 200 mg/m2 (MEL200), is considered the gold
Four drug combinations have also been investigated. standard conditioning regimen for ASCT, based on a ran-
However, the results of the EVOLUTION study comparing domized trial demonstrating better OS with MEL200 com-
VDCR versus VDR versus VDC, a phase 2 trial testing the pared with melphalan, 140 mg/m2, in combination with
efficacy of VTD plus cyclophosphamide, and the CYCLONE TBI. Other studies using intensification of the melphalan
study investigating the combination of carfilzomib, cyclo- doses or addition of alkylating agents have not demonstrated
phosphamide, thalidomide, and dexamethasone have been significant improvement in terms of response rate or sur-
disappointing, with an increased toxicity with no significant vival. More recently, the Spanish group reported data show-
benefit in response rate. ing that the combination of oral busulfan and melphalan
In comparison to melphalan, novel drugs had less impact (BU-MEL) was equivalent to MEL200 in terms of PFS dur-
on stem cell collection; thalidomide or bortezomib combi- ing the first 2 years, but with a significant long-term benefit
nations did not affect stem cell collection or posttransplant for the BU-MEL arm (PFS: 41 vs. 31 months, P = 0.009).
blood count recovery. Several studies have demonstrated an However, this trial was stopped because BU-MEL with oral
impact of lenalidomide on the ability to collect stem cells, an busulfan resulted in an unacceptable toxicity with a high
effect that is more pronounced in older patients and in those incidence of veno-occlusive disease (VOD) and transplant
who have received more than 6 cycles of treatment. This related mortality. With the use of intravenous busulfan the
problem can be overcome by avoiding prolonged therapy incidence of VOD is drastically reduced, and a randomized
prior to stem cell collection, use of plerixafor to mobilize trial by the Spanish group is ongoing in order to confirm
stem cells, or using a cyclophosphamide based mobilization these results.
regimen (chemomobilization). Several pilot studies are investigating the addition of
bortezomib to melphalan using different doses and sched-
ules. In a phase 1 trial investigating the bortezomib dose
Role of high-dose therapy and autologous stem cell
(1-1.6 mg/m2) and sequence (24 hours before or 24 hours
transplantation (HDT/ASCT)
after melphalan administration), the combination of
Eight randomized trials have compared HDT/ASCT with MEL200 along with the aforementioned bortezomib dose
conventional chemotherapy without HDT/ASCT and in all and schedule resulted in a PR rate or better of 93%, with a
but one the HDT was associated with a significant increase in toxicity profile and engraftment kinetics similar to that
depth of response well as event free survival (EFS); however, observed in an historical control receiving MEL200 alone.
a significant prolongation in OS was seen in only three stud- The International Francophone Myeloma (IFM) group used
ies, not in all. These discrepancies likely resulted from differ- bortezomib on days −6, −3, 1, and 4 along with MEL200 on
ences in designs, such as inclusion of only chemotherapy day −2 in 35 patients with high-risk MM, and the results
sensitive patients for ASCT in some studies but not others, appear to be superior in terms of response rates when com-
allowing delayed HDT at the time of relapse, differences in pared with historical controls. Other trials are examining the
the conditioning regimens, and differences in the intensity role of adding newer proteasome inhibitors such as carfil
and duration of the chemotherapy arm. In spite of these zomib to the conditioning regimen.

Post-HDT consolidation and maintenance been reported for lenalidomide maintenance with respect to
control arms in both trials (46 months vs. 27 months in the
Consolidation is typically a short course treatment given
CALGB trial and 41 vs. 23 months in the IFM trial), and this
after HDT aiming to improve response or to induce deeper
translated into an OS advantage in the American but not in
responses; while maintenance treatment is given for a pro-
the French trial. Additionally, the tolerability was much bet-
longed period of time to maintain the response achieved.
ter than with thalidomide. A third Italian randomized trial
The Arkansas group was the first to demonstrate the efficacy
has explored the value of lenalidomide maintenance after
of consolidation and maintenance therapies through their
induction with 4 cycles of lenalidomide-dexamethasone fol-
four consecutive total therapy programs with up to 50%
lowed by intensification with either 6 courses of lenalido-
long-term survivors. More recently, Ladetto et al. have
mide, melphalan and prednisone, or melphalan 200 mg/m2
shown that the use of four consolidation cycles of VTD
(MEL200) supported with ASCT. Despite an improvement
(bortezomib, thalidomide, and dexamethasone) in patients
in PFS, the OS is not significantly different in that study.
in ≥VGPR after ASCT, increase the CR rate from 15% to
Maintenance with lenalidomide is associated with an
49%, and molecular remissions from 3% after ASCT to 18%
increased risk of secondary malignancies, and this needs to
after VTD.
be considered when discussing maintenance with patients.
An Italian randomized trial compared VTD with TD both
Bortezomib has been tested as maintenance therapy in
as induction and consolidation in patients undergoing dou-
two randomized trials. In the HOVON-65 Study, induction
ble ASCT, again confirming efficacy of consolidation. VTD
with either PAD (bortezomib, doxorubicin, dexamethasone)
improved the CR rate postconsolidation by 30% as com-
or VAD (vincristine, doxorubicin, dexamethasone) was fol-
pared with 16% with TD, with 3 years PFS of 60% vs. 48%.
lowed by MEL200 with ASCT, and maintenance with bort-
Importantly, the superior PFS with VTD over TD consolida-
ezomib in the PAD arm or thalidomide in the VAD arm.
tion was maintained across poor prognosis subgroups
This study showed a significant longer PFS and OS for
[t(4;14) and/or del(17q), del(13q), β2-M, LDH, ISS]. Two
patients treated with bortezomib than for those treated with
cycles of lenalidomide and dexamethasone was used after
thalidomide; although the results of bortezomib versus tha-
one or two HDT in the French maintenance trial, and was
lidomide maintenance in the HOVON-65 study cannot be
shown to improve the response depth. Bortezomib and
isolated from the differences in induction, the increases of
dexamethasone has also been studied as post-HDT consoli-
CR rate during maintenance with bortezomib (12%) or tha-
dation, with improvement in depth of response and progres-
lidomide (11%) were similar, and a landmark analysis from
sion free survival but with no improvement in OS.
start of maintenance for PFS did not show a significant dif-
The concept of continued therapy or maintenance to con-
ference between thalidomide and bortezomib maintenances.
trol the residual disease has been explored in MM for a long
In PETHEMA/GEM05menos65 trial, after HDT/ASCT,
time. While initial trials with steroids or interferon did sug-
patients who received maintenance with bortezomib-
gest some benefit, the toxicity was significant and long-term
thalidomide showed a significantly longer PFS (78% at 2
tolerance was very poor with these treatments. Following the
years) compared with those who received thalidomide (63%)
demonstration of its efficacy, thalidomide alone or in com-
or interferon (49%) alone, but without differences in OS.
bination was evaluated in multiple phase 3 trials as a mainte-
nance agent. Seven randomized trials have demonstrated a
Early or delayed autologous transplant
superiority of thalidomide maintenance (with or without
prednisone) versus no maintenance or prednisone or inter- The timing of HDT/ASCT is now being challenged by the
feron in terms of PFS; although only two trials showed efficacy of continuous treatment with novel agents. The
improvements in OS. A meta-analyses of these trials do con- optimal results obtained with “long-term” treatment with
firm that thalidomide prolongs PFS and OS when applied as novel agents combinations has led many investigators to
post-HDT maintenance. However, the benefit of thalido- suggest reserving HDT/ASCT for the time of relapse. In
mide maintenance for patients who are already in CR was order to clarify this debate, several groups are testing these
not clear, and results in those with poor cytogenetic profile two alternatives in randomized trials. Data derived from the
are contradictory. Additionally, the high dropout rate due to E4A03 clinical trial investigating the use of lenalidomide
intolerance to thalidomide, mainly due to polyneuropathy, with either high-dose or low-dose dexamethasone has shown
limits its applicability as a maintenance drug. that the patients who proceeded to HDT/ASCT achieved a
More recently, lenalidomide maintenance has been stud- superior OS compared to those who continued on primary
ied in two large randomized trials, one conducted by the IFM therapy (3-years OS: 94% vs. 78%). However, this was a ret-
and the other by the US Cancer and Leukemia Group B rospective subanalysis with potential bias. A recent prelimi-
(CALGB) group. A significant benefit in terms of PFS has nary report pooled data from 791 patients enrolled in two

alike prospective phase 3 randomized Italian trials for new As mentioned above, the use of tandem HDT/ASCT has
diagnostic MM patients under 65, both using the same decreased in the last decade due to the apparent similar ben-
lenalidomide-dexamethasone induction followed by MEL200/ efit that could be obtained upon using consolidation therapy
ASCT (early transplant) vs. six cycles of conventional che- with novel agents. Nevertheless, the value of tandem HDT/
motherapy (melphalan or cyclophosphamide) plus lenalido- ASCT will probably be revisited in the era of novel agents
mide and steroids and, followed by lenalidomide-based and based on the recent information derived from a meta-
regimens or placebo as maintenance. In both trials, patients analysis, including patients from the French, Italian, Ger-
assigned to nontransplant arm were allowed to receive HDT/ man, Dutch, and Spanish Cooperative Groups, showing that
ASCT at relapse (delayed transplant). Early HDT/ASCT patients with high-risk cytogenetics benefit more from tan-
improved PFS1 (3-year rate: 59% vs. 35%, P < 0.001) and dem HDT/ASCT and that this procedure may, at least in
PFS2 (3-year rate: 77% vs. 68%, P = 0.01), and marginally OS part, abrogate the adverse prognosis of t(4;14) and deletion
(4-year rate: 83% vs. 72%, P = 0.09) in comparison with 17p. In addition, a second transplant at late relapse is being
delayed HDT/ASCT at relapse. Although current data favor increasingly used, provided that the duration of the response
the use of frontline HDT/ASCT even in the novel agent era, to first transplant has lasted for at least 2 years. This consid-
the definitive answer will come from the confirmation of the eration would support the collection of sufficient cells for
above mentioned Italian trials, and other ongoing trials such multiple transplants at the time of initial harvest.
as the IFM/Dana-Farber Cancer Institute (DFCI) 2009 study
that seeks to address the question of timing of transplant Allogeneic transplantation
(early vs. late) for patients receiving lenalidomide, bortezo-
mib, and dexamethasone based induction in a prospective Six clinical trials have compared double HDT vs. HDT fol-
fashion. lowed by an allogeneic transplant using reduced intensity con-
ditioning regimens (allo-RIC). Only two out of the six studies
Tandem transplants: double autologous (GIMEMA and EBMT) showed a significant prolongation in
and autologous/allogeneic PFS and OS with the ASCT/allo-RIC combination. Differ-
ences in patient characteristics, GVHD prophylaxis, and con-
To improve the limited antitumor efficacy of available thera- ditioning regimens could contribute to these discrepant
peutic agents in the era of conventional chemotherapy, the results. Two meta-analyses and a systematic review confirm
Arkansas group investigated the capacity of tandem HDT/ that although the allo therapeutic approach improves response
ASCT to improve the outcome of myeloma patients. The rates compared to tandem HDT/ASCT, the mortality associ-
first so-called “total therapy” study (a name first used by St. ated with the procedure neutralizes the potential benefits of a
Jude Children’s Research Hospital in Memphis, TN, to higher antitumor effect. It has been argued that evaluation of
describe pediatric acute lymphoblastic leukemia studies in the efficacy of allogeneic transplant should be done at the long
the 1960s and 1970s) was based on a series of alternative term. Accordingly, a recent long-term follow-up analysis of
induction treatments and two consecutive ASCT condition- the NMAM200 EBMT trial (that did not show statistical
ing with MEL200 (tandem transplants), or MEL200-ASCT differences in survival at the initial analysis), after a median
with a second allogeneic transplant if an HLA-identical follow-up of 7 years, shows a significant survival benefit for
donor was available, followed by interferon maintenance. patients allocated to ASCT/allo-RIC arm (at 96 months: PFS
The tandem HDT/ASCT approach was associated with a 22% vs. 12%, OS 49% vs. 36%). In contrast, the analysis of the
high response rate and gained many adaptors in subsequent HOVON-50 randomized trial has failed to show improve-
years; and it also became almost standard of care in some ment of tandem ASCT/allo-RIC as compared to single HDT/
centers until the introduction of novel agents and the possi- SCT followed by thalidomide maintenance therapy even after
bility of efficient consolidation schemes. a median of follow-up of 90 months. Moreover, a high pro-
The analysis of accumulated experience tandem HDT/ portion of patients developed extramedullary relapses without
ASCT yields contradictory results, however. In a Cochrane bone marrow involvement, indicating that although the dis-
systematic review of controlled studies from 1995 to 2011, ease may be under control in the bone marrow milieu, extra-
including five randomized trials enrolling 1,406 patients in medullary spread may occur. In any event, about 10%-15% of
total, the median EFS prolongation respect to a single HDT/ patients undergoing an allogeneic transplant are likely cured.
ASCT ranged between 3 to 12 months in four of the five tri-
als, while OS was statistically significantly improved in one
Treatment of transplant-ineligible patients
trial only. A formal meta-analysis has not been possible due
to the high clinical and methodological heterogeneity The combination of melphalan and prednisone (MP) was
between the trials examined. studied extensively in the nontransplant MM population

beginning in the 1960s and was the standard therapy until the of 80% (20% of CR), and following maintenance therapy
advent of the new drugs. Overall response rates from different with VT or VP (see below) the PFS and OS were 37 and
studies varied from 40% to 60% (CRs were rare), and median 60 months, respectively.
survival was around 3 years. With the introduction of the new
drugs and initial studies showing excellent efficacy when
MP with thalidomide and bortezomib (VMPT)
combined with alkylating drugs, a series of phase 3 trials were
undertaken examining the impact of adding thalidomide, Subsequent trials have sought to build upon the VMP regi-
lenalidomide, or bortezomib to MP in this population. men by adding thalidomide to the combination (VMPT)
with or without prolonged maintenance therapy. Palumbo
and colleagues randomized patients to receive either nine
MP and thalidomide (MPT)
5-week cycles of VMP or nine 5-week cycles of VMPT and
Six randomized trials compared the combination of MP continued maintenance thalidomide along with alternate
plus thalidomide (MPT) with the standard MP. PR rate was week bortezomib. While response rates and PFS were higher
42%-76% vs. 28%-48% with MPT and MP, respectively, and in the four-drug combination with maintenance, the OS was
PFS was 14-28 vs. 10-19 months. In three out of the six trials, not different. Toxicity was significantly higher using the
the PFS advantage observed with MPT also translated into a four-drug regimen with more neutropenia, cardiac events,
significant OS advantage (37-52 vs. 28-32 months), but this and thromboembolic events. During the trial, the treatment
was not confirmed in the three other trials. A meta-analysis schedule for bortezomib was changed from twice weekly to
of pooled data from 1,682 patients from the aforementioned once weekly, allowing a comparison of the two approaches.
six MPT trials showed that the addition of thalidomide to It was found that the cumulative doses of bortezomib admin-
MP is associated with a significant improvement in PFS (5.4 istered were similar with the two approaches but with sig-
months of benefit; hazard ratio of 0.67, 95% CI 0.55-0.80) nificant reduction in severe sensory peripheral neuropathy
and a nearly significant improvement in OS (6.6 months of from 16% to 3%. As a result of this study, bortezomib is
benefit; hazard ratio of 0.82, 95% CI 0.66-1.02). These results increasingly being used once weekly as part of different
support the use of MPT as one of the standards of care for drug combinations.
elderly MM patients. With respect to its toxicity, the median
incidences of grade 3-4 peripheral neuropathy (PN) and
MP with lenalidomide (MPR)
venous thromboembolism (VTE) were 13% and 6%, respec-
tively, meaning that antithrombotic prophylaxis is required MP has also been compared to the combination of melpha-
when using MPT. lan, prednisone, and lenalidomide in a 3-arm phase 3 trial:
MP versus MP with lenalidomide (MPR) without mainte-
nance versus MPR with lenalidomide maintenance (MPR-
MP and bortezomib (VMP)
R). Four hundred and fifty-nine patients were randomized
The VISTA trial compared MP to bortezomib and MP to MP (nine 4 week cycles of melphalan 0.18 mg/kg/day and
(VMP) with patients receiving nine 6-week cycles of either prednisone 2 mg/mg/kg/day, days 1-4), MPR (nine 4 week
melphalan (at a dose of 9 mg/m2) and prednisone (60 mg/m2) cycles of MP plus lenalidomide 10 mg days 1-21), or nine
on days 1 to 4, alone or in combination with bortezomib cycles of MPR with indefinite lenalidomide maintenance (10
(1.3 mg/m2) on days 1, 4, 8, 11, 22, 25, 29, and 32 during mg days 1-21 every 4 weeks). While addition of lenalidomide
cycles 1 to 4 and on days 1, 8, 22, and 29 during cycles 5 to 9. to MP led to higher response rates, and improved PFS when
Median PFS was 24 months with VMP as compared to 17 lenalidomide maintenance was used, there was no difference
months with MP, and with a 5-year follow-up VMP showed in the OS between the arms. Toxicity was substantially higher
a persistent significant survival benefit, with a 13.3-month in the lenalidomide arms. Two recent randomized trials have
increase in the median OS (HR 0.695, P = 0.0004; median OS compared MPR-R versus MPT-T with no significant differ-
56.4 vs. 43.1 months). Grade 3-4 adverse events, however, ences in terms of PFS (approximately 20 months) and OS
were more frequent in patients receiving VMP (46% vs. (approximately 50 months), although the tolerability was
36%). Long-term follow up of this trial showed consistent better for the lenalidomide combination.
results. Subsequently, the Spanish group investigated the
VMP regimen but with weekly administration of bortezo-
MP vs. thalidomide and dexamethasone (TD)
mib; patients received the first cycle as a conventional twice-
weekly dose and the other five cycles were administered as a TD was initially studied as an induction therapy in the
weekly dose. The results showed that after 6 cycles, the inci- younger patients but then was compared with MP in a cohort
dence of grade 3 or grade 4 PN dropped to 7%, with an ORR of elderly MM patients. Although it induced higher response

rates than MP (68% vs. 50%), OS was significantly shorter months with continuous lenalidomide-dexamethasone, 20.7
with TD (41.5 vs. 49.4 months); the effect being particularly months with 18 cycles of lenalidomide-dexamethasone, and
evident in patients over 75 years of age (20 vs. 41 months). 21.2 months with MPT. Overall survival at 4 years was 59%
This paradoxical finding resulted from the higher frequency with continuous lenalidomide-dexamethasone, 56% with 18
of nondisease-related deaths observed in the TD group dur- cycles of lenalidomide-dexamethasone, and 51% with MPT.
ing the first year. Thus, TD in elderly patients is not a good Compared with MPT, continuous lenalidomide-dexametha-
option unless patients receive reduced doses of both drugs. sone was associated with fewer hematologic and neurologic
toxic events but a moderate increase in infections. Subse-
quent subgroup analyses also clearly showed consistent ben-
MP vs. cyclophosphamide, thalidomide,
efit for patients over 75, as were seen for the younger patients.
and dexamethasone (CTDa)
Cyclophosphamide, another alkylator agent with proven

Bendamustine combinations
efficacy in MM, has been evaluated in the Medical Research
Council (MRC) Myeloma IX study, a randomized trial in More recently, the alkylating agent bendamustine has been
elderly patients that compared the efficacy and safety of a incorporated into the treatment armamentarium of MM. It
regimen of cyclophosphamide, thalidomide, and attenuated has structural similarities with alkylating agents and purine
dexamethasone (CTDa) compared with MP. Although analogs and is currently approved in Europe for the treat-
CTDa produced a significant two-fold improvement in ment of newly diagnosed MM patients who are not candi-
overall response rates (64% vs. 33%), survival outcomes dates for HDT-ASCT and who cannot receive thalidomide
were not significantly different between the two regimens. or bortezomib because they have a peripheral neuropathy.
CTDa was associated with higher rates of thromboembolic The rationale for the approval was a randomized trial in
complications, peripheral neuropathy, infection, and consti- which bendamustine plus prednisone (BP) proved to be
pation than MP, indicating that adequate management of superior to MP with respect to CR rate (32% vs. 13%, P =
adverse events is required to allow patients to continue the 0.007), with a benefit in terms of time-to-treatment failure
regimen and eventually benefit from the CTDa treatment. (14 months for BP compared with 10 months for MP; P =
0.020), but without any benefit to overall survival. The toxic-
ity profile was comparable, and hematological toxicity and
MP and thalidomide (MPT) vs. bortezomib, thalidomide
nauseas and vomiting were the most frequent AEs reported
and prednisone (VMP)
with BP. Bendamustine plus prednisone in combination
The Spanish group compared VMP with VTP in a random- with bortezomib is currently being evaluated in several pilot
ized trial (previously mentioned) to identify the best partner clinical trials.
for bortezomib, an alkylating agent or an immunomodula-
tory drug. VTP resulted in slightly greater efficacy (CR rate
Second generation of proteasome inhibitors
of 28% for VTP vs. 20% for VMP) but also toxicity, espe-
cially cardiac side effects (11% with VTP and none with Second-generation proteasome inhibitors in combination
VMP) and PN (9% of grade 3-4 for VTP and 7% for VMP). with alkylators are also emerging as new therapeutic options
The long-term survival was longer with VMP than with VTP: for newly diagnosed nontransplant-eligible MM patients. In
PFS (32 vs. 23 months) and OS (63 vs. 43 months), and VMP a pilot phase 1/2 trial, carfilzomib combined with MP (CMP)
was also associated with better tolerability. is yielding promising efficacy results (ORR of 92% and rate
of very good partial response or better of 42%) with an
acceptable toxicity profile with no grade 3-4 PN, and is the
MPT vs. lenalidomide-dexamethasone (Rd)
rationale for a randomized trial comparing CMP with VMP.
The combination of lenalidomide and dexamethasone was The combination of carfilzomib plus cyclophosphamide and
initially studied in a phase 3 trial comparing different doses low-dose dexamethasone is being evaluated in a series of 53
of dexamethasone, and subgroup analysis from that trial newly diagnosed elderly MM patients, showing ORR rates of
clearly showed good outcomes with the combination in older 100%, including 53% of CR rate and 22% of stringent CR.
patients. The regimen was then studied in a large random- No grade 3-4 PN was reported and tolerability was good.
ized trial (the FIRST trial), where 1,623 patients were assigned Ixazomib (MLN9708), oral second-generation proteasome
to lenalidomide and dexamethasone in 28-day cycles until inhibitor FDA approved in 2015, plus MP in a biweekly
disease progression (535 patients), to the same combination or weekly scheme is also currently undergoing a phase
for 18 cycles (541 patients), or to MPT for 72 weeks (547 1/2 clinical trial to evaluate the efficacy and safety of this
patients). The median progression-free survival was 25.5 combination. Carfilzomib has also been combined with

lenalidomide-dexamethasone (KRd) in a pilot phase 1/2 trial benefit accrued form the maintenance versus addition of
in newly diagnosed MM patients, including young and thalidomide in the induction phase. The role of lenalidomide
elderly patients. Results of a subanalysis of 23 elderly MM maintenance in this patient population, following initial
patients showed impressive efficacy (100% of ORR, with therapy with MPR (MP plus lenalidomide) was evaluated in
stringent CR in 65% of patients), and an acceptable toxicity the MM-015 study. PFS was significantly improved in the
profile (13% of grade 1-2 PN). All patients remained free of MPR-R group compared with MPR, with no difference seen
progression and alive at the median follow up of 1 year. in OS. There was increased hematological toxicity with
These results support a phase 3 study of KRd vs. Rd in all age addition of lenalidomide, and there was also an increased
groups. lenalidomide-dexamethasone alone is being also incidence of second primary malignancy with the use of
compared with lenalidomide-dexamethasone plus ixazomib lenalidomide maintenance. The FIRST trial also analyzed the
(MLN9708) in a randomized trial in nontransplant candi- value of maintenance, Rd for 18 months vs. Rd continuous.
dates MM patients supported by the positive preliminary In the FIRST trial that compared continuous lenalidomide-
results showed with ixazomib given weekly plus lenalido- dexamethasone until progression with fixed-time
mide-dexamethasone (96% of patients had achieved at least lenalidomide-dexamethasone (18 cycles) showed a longer
PR, including a ≥VGPR rate of 44% and a CR rate of 26%) PFS for continuous treatment (25.5 vs. 20.7 months, HR:
with good tolerability. 0.70, P = 0.0001) with no differences in OS (59.4% vs. 55.7%
at 4 years). Finally, second generation proteasome inhibitors
are being evaluated as part of consolidation (carfilzomib in a
Maintenance therapy in elderly patients
modified schedule) or maintenance therapy (ixazomib
Three studies compared thalidomide maintenance following weekly until disease progression), preliminary data indicate
MPT (MPT-T) versus MP with no maintenance. While that ixazomib upgrade the depth or response.
improvement in TTP and PFS was seen in all, only one study
found a significant improvement in OS for MPT-T compared
Individualizing treatment of elderly patients
with MP alone. The MRC Myeloma IX trial of attenuated
CTD versus MP also involved thalidomide maintenance The above-mentioned novel treatment combinations offer
randomization, again with improved PFS, and no physicians the possibility of tailoring treatment approaches
improvement in OS. Thalidomide maintenance was poorly taking an individual patient’s profile and their preference
tolerated in these trials with a high proportion of patients into account. Physicians should undertake two actions
discontinuing therapy with in a year. In a Spanish study that before prescribing treatment to elderly patients: (i) assess the
compared VMP and VTP as induction, patients received patient’s biological age, comorbidities, frailty, and disability
maintenance with either VT or VP for up to 3 years. in order to select the most appropriate drug regimen, adapt-
Maintenance therapy improved the depth of response as well ing the dose if required (Table 23-11); and (ii) optimize the
as PFS but not the OS. The Italian trial comparing VMPT supportive care treatment with bisphosphonates, antibiotics,
with VMP as induction therapy also included maintenance antivirals, anticoagulants, growth factors, and pain control.
with VT following VMPT, which led to improved CR rates, For unfit elderly patients, dose adjustments are key to
PFS, as well as OS, but it remains unclear how much of the improving tolerability. Bortezomib should always be given in a
Table 23-11 Recommended

Agent Dose level 0 Dose level 1 Dose level 2
dose modifications for functional
Bortezomib 1.3 mg/m2 twice / week 1.3 mg/m2 once / week 1.0 mg/m2 once / week impairment in elderly patients
d 1, 4, 8, 11 / 3 weeks d 1, 8, 15, 22 / 5 weeks d 1, 8, 15, 22 / 5 weeks
Thalidomide 100 mg/d 50 mg/d 50 mg qod
Lenalidomide 25 mg/d 15 mg/d 10 mg/d
d 1-21 / 4 weeks d 1-21 / 4 weeks d 1-21 / 4 weeks
Dexamethasone 40 mg/d 20 mg/d 10 mg/d
d 1, 8, 15, 22 / 4 week d 1, 8, 15, 22 / 4 week d 1, 8, 15, 22 / 4 week
Melphalan 0.25 mg/kg (9 mg/m2) 0.18 mg/kg (7.5 mg/m2) 0.13 mg/kg (5 mg/m2)
d 1-4 / 4-6 weeks d 1-4 / 4-6 weeks d 1-4 / 4-6 weeks
Prednisone 50 mg qod 25 mg qod 12.5 mg qod
Cyclophosphamide 100 mg/d 50 mg/d 50 mg qod
d 1-21 / 4 weeks d 1-21 / 4 weeks d 1-21 / 4 weeks
qod = every other day.

weekly scheme and as a subcutaneous formulation, probably in failure, earlier intervention based on biochemical progression
combination with low-dose steroids (prednisone may be better may be warranted. Clinical trials should always be considered
tolerated than dexamethasone), considering a low dose of mel- in patients with relapsed disease. Most of the combinations
phalan or, as a probably better alternative, cyclophosphamide. that have been studied can also be used in the relapsed setting,
Oral drugs can be more convenient for frail elderly patients; based on the prior use of the specific drugs and the presence of
lenalidomide can be given at a standard dose with low-dose toxicities. HDT therapy should be strongly considered in
dexamethasone, while thalidomide should not be given at doses patients who elected to delay the procedure as part of their
higher than 50-100 mg daily in combination with oral cyclo- initial line of therapy. In those patients who already received
phosphamide and prednisone on alternate days. HDT, a second HDT may be appropriate if the duration of
Other factors should be considered when making treat- response to initial HDT was 18 months or more.
ment decisions. In patients who have a history of venous It is important to separate young (<65 years) from elderly
thromboembolism, a bortezomib-based combination may (>65 years) patients with relapsed MM. In young patients
be a preferred treatment choice as it is less thrombogenic. relapsing after transplantation, one can discriminate three
However, appropriate anticoagulant prophylaxis has been cohorts of patients: early relapse (<1 year), intermediate
shown to reduce VTE complications to approximately relapse (1–3 years), and late relapse (>3 years). If the relapse
3% in patients treated with lenalidomide- or thalidomide- occurs within the first year after transplantation, patients
containing regimens. In patients with preexisting neurop- should be immediately considered high risk and, in order
athy, Rd would be a good choice for upfront treatment to overcome drug resistance, should be treated with multi-
since they are not associated with neurotoxicity; carfilzo- drug combinations incorporating novel agents. Allogeneic
mib does not induce peripheral neuropathy. In patients transplantation should be considered for these patients. If
with renal failure, bortezomib, thalidomide can be admin- relapse occurs 1-3 years after autologous SCT, many inves-
istered at the full approved dose; lenalidomide requires tigators would favor rescue with novel agents used in a
adjustments of the starting dose based on creatinine sequential (not simultaneous) manner, starting with one
clearance. line of treatment (different from the one used as induction)
and shifting to the second and subsequent lines only when
disease progression occurs. Within this category of patients,
Treatment of relapsed MM
those <60 years old with an HLA-identical donor and a
With the current treatments, the vast majority of patients with suboptimal response to the first line of treatment should be
MM will eventually relapse, and therapy will have to be reinsti- considered for allogeneic SCT. Finally, if relapse occurs
tuted. The choice of regimens at the time of relapse will more than 3 years after the first autologous SCT, an attrac-
depend on a variety of factors as outlined in Table 23-12, espe- tive possibility is re-induction with the initial treatment or
cially the type of relapse, efficacy, and toxicity of previous other novel-agent combination followed by a second autol-
therapies and further treatment options. The timing of initia- ogous SCT.
tion of therapy has to be carefully considered. As with initial For elderly patients, treatment decisions at relapse must
therapy, it has to be guided by the clinical scenario. Patients, take into account the general condition of the patient. Once
especially after HDT, can have a slow biochemical progression the patient relapses, after up-front treatment, the durations
with no clear end-organ damage, a situation akin to MGUS of subsequent responses to rescue therapies are progressively
and SMM, where expectant observation may be the appropri- shortened. Therefore, the current goal in relapsed MM is to
ate step. However, in patients who presented with significant optimize the efficacy of novel drugs through their most
complications, such as neurological complications and renal appropriate combinations, to establish optimal sequences of
treatment, and to promote active clinical research on experi-
Table 23-12 Considerations for choosing treatment for relapsed MM mental agents that have already shown promising activity in
Presence of end-organ damage from the relapsed disease in vitro and animal models.
Drugs used before, the responses observed, and the time elapsed
from prior exposure
Thalidomide
Prior use of HDT in those eligible for the procedure
Residual toxicity from prior therapy Single-agent thalidomide can induce a response in 25%-58%
Bone marrow and organ function of relapsed/refractory patients and higher response rates
Duration of initial response of 41%-55% when combined with dexamethasone. It
Presence of high risk chromosome abnormalities
should be noted that these results were obtained in pati
Age and functional status
ents just relapsing after alkylating agents. In addition to cor-
Goals and preferences of the patient
ticosteroids, thalidomide has been combined with alkylators

(melphalan, cyclophosphamide, or bendamustine) achiev- response of 8 months. In a phase 3 study, Richardson et al.
ing overall response rates as high as 79%. randomly assigned 669 patients with relapsed myeloma to
receive either an intravenous bortezomib or high-dose
Lenalidomide dexamethasone; overall response rate was significantly
higher with bortezomib (38%) compared with dexametha-
With lenalidomide as a single agent, approximately 18%- sone (18%), and the median times to progression in the
25% of relapsed or refractory patients have achieved a partial bortezomib and dexamethasone groups were 6.2 months
response, and the median duration of response for respond- and 3.5 months, respectively. Bortezomib has since then
ing patients is 20 months. The combination of lenalidomide been studied in combination with alkylator drugs and
and dexamethasone has been studied in two large phase 3 doxorubicin. One of the most commonly used and effec-
trials. The results of both were comparable with 59% of tive combinations has been that of cyclophosphamide
patients responding to the combination, including a 14% added to bortezomib and dexamethasone. In an interna-
complete response rate. This was significantly better than tional phase 3 study, 646 patients were randomly assigned
what was observed with single agent dexamethasone (PR to receive either intravenous bortezomib with or without
22.5% and CR 2%). In addition, both time to progression pegylated doxorubicin (PLD). The overall response rate
(11 to 13 months versus approximately 5 months) and OS was similar between bortezomib (41%) and PLD + bort-
(not reached versus 24 months) were superior in the lenalid- ezomib (44%). Median time to progression was increased
omide-dexamethasone arm. Significant numbers of patients from 6.5 months to 9.3 months with the combination and
who initially had a partial response improved the response to the 15-month survival rate for the combination was 76%
a complete or very good partial response with continued compared with 65% for bortezomib alone.
treatment. As with thalidomide, lenalidomide has been com-
bined with alkylators (melphalan, cyclophosphamide or
Carfilzomib
bendamustine) as well as anthracyclines with excellent toler-
ability and efficacy. Carfilzomib is a next-generation selective proteasome inhi
bitor that has been approved for treatment of relapsed MM.
Pomalidomide In an open-label, single-arm phase 2 study (PX-171-003-A1),
patients received single-agent carfilzomib 20 mg/m2 intra-
Pomalidomide is another IMiD that has been approved for venously twice weekly for 3 of 4 weeks in cycle 1, followed
treatment of relapsed or refractory myeloma. In the initial by 27 mg/m2 at the same schedule. Among the enrolled
phase 2 trial from Mayo Clinic, pomalidomide 2 mg daily patients, 95% were refractory to their last therapy and 80%
in combination with dexamethasone 40 mg weekly, were refractory or intolerant to both bortezomib and
resulted in 63% overall response in patients with relapsed lenalidomide. Overall response rate was 24% with a median
myeloma who had 2-3 prior regimens. Responses were duration of response of 7.8 months and OS of 15.6 months.
seen irrespective of their prior drug exposures, with 40% Common toxicities encountered were fatigue, anemia, nau-
of lenalidomide-refractory patients, 37% of thalidomide- sea, and thrombocytopenia. Neutropenia was limited to
refractory patients, and 60% of bortezomib-refractory grades 1 and 2, occurring in 10%-15% of the patients. In a
patients achieving a response. These results have been con- phase 2 study, 129 bortezomib-naive patients with relapsed,
firmed in a phase 3 randomized trial comparing pomalid- refractory MM received intravenous carfilzomib either at 20
omide plus low-dose dexamethasone versus high-dose mg/m2 for all treatment cycles or 20 mg/m2 for cycle 1 and
dexamethasone in patients who had failed to prior bort- then 27 mg/m2 for all subsequent cycles. The clinical benefit
ezomib and lenalidomide. The combination significantly response (overall response rate + minimal response) was
improved PFS (4 vs. 1.9 months, HR = 0.50; P <0.001) and 59.3% and 64.2%; median duration of response was 13.1
OS (13.1 vs. 8.1 months, HR = 0.72; P = 0.009). Triple com- months and not reached, and median time to progression
binations of pomalidomide with cyclophosphamide-pred- was 8.3 months and not reached, respectively. A recent
nisone or with bortezomib-dexamethasone are also being phase 3 trial has compared the efficacy of the combination
investigated and the preliminary results indicate that the with lenalidomide versus the standard lenalidomide-
PFS increases to 10 months. dexamethasone regimen (CRd versus Rd). CRD was associ-
ated with a significantly longer PFS (26.3 vs. 17.6 months),
and OS (HR: 0.79), as well as higher RR including 31.8% CR
Bortezomib
vs. 9.3% in the control arm. Combinations with pomalido-
Single-agent response rates in relapsed/refractory myeloma mide, cyclophosphamide, and panobinostat are also being
range from 28% to 38% with a median duration of tested.

Panobinostat Akt inhibitor in the setting of myeloma. In a phase 2 trial of

perifosine, alone or with dexamethasone, the best response
Acetylation and deacetylation of histone proteins play an
to single agent among 48 enrolled patients was an MR in 1
important role in the regulation of gene expression. Aber-
patient and stable disease in 22 patients (46%). Addition of
rant recruitment of histone deacetylase (HDAC) may play a
dexamethasone in 37 patients with disease progression led to
role in the changes in gene expression in cancer cells. HDAC
a partial response in 13%. The most common adverse events
inhibitors are thought to affect multiple pathways in MM
included nausea, vomiting, diarrhea, fatigue, increased cre-
reversing the abnormalities of cell apoptosis and cell cycle,
atinine, and anemia. Subsequent trials have examined the
potentially sensitizing MM cells to apoptosis. HDAC inhib-
combination of perifosine with lenalidomide or bortezomib
itors have exhibited only modest activity (minor responses
with modest evidence of activity.
or disease stabilization) as single agents. The combination
The mTOR kinase downstream in the PI3K/Akt pathway
of panobinostat with bortezomib-dexamethasone produ
has been another target explored in myeloma. Both CCI779
ced 35% PR in bortezomib-refractory patients in phase
and RAD001 have been studied in phase 2 trials in patients
2 studies.
with relapsed disease with very little clinically relevant anti-
In a large phase 3 trial, the proportion of patients achiev-
myeloma activity. The combination of temsirolimus and
ing an overall response did not differ between treatment
bortezomib have been studied in a phase 1/ 2 study. In the
groups (61% for panobinostat vs. 55% for the control arm),
phase 2 study, the proportion of patients with a partial
but the proportion of patients with a complete or near com-
response or better was 33%. The most common treatment-
plete response was significantly higher in the panobinostat
related grade 3-4 adverse events were thrombocytopenia and
group. This in turn translated to an improved progression
diarrhea.
free survival (PFS: 12 vs. 8 months) for the panobinostat arm
Other novel agents under investigation include the kinase
but with similar overall survival. (The improvement in the
spindle protein inhibitor filanesib (ARRY-520), which has
outcomes appeared to be particularly of interest among
shown 22% ≥PR when combined with low-dose dexametha-
patients who had been previously exposed to proteasome
sone in patients refractory to bortezomib, lenalidomide, and
inhibitors and IMiDs (PFS: 12.5 vs. 4.7 months). The results
dexamethasone.
of this trial led to its approval in its use in patients with
relapsed myeloma. In contrast, in a phase 3 trial comparing
vorinostat with or without bortezomib, the PFS was only Monoclonal antibodies
marginally improved (7.6 vs. 6.8 months), and no difference
Monoclonal antibody therapy has seen resounding success
was seen in OS despite the overall response rate being signifi-
in B-cell malignancies, but until recently has had minimal
cantly higher with the vorinostat combination (56% vs.
success with PC disorders. Elotuzumab is a humanized
41%), Over a third of the patients had serious adverse events
monoclonal IgG1 antibody targeting human CS1 (SLAMF7),
in these trials with the addition of the HDAC inhibitor, with
a cell surface glycoprotein. SLAMF7 is highly and uniformly
hematological and gastrointestinal toxicity and fatigue being
expressed on MM cells, with limited expression on natural
the most common.
killer (NK) cells and CD8+ cells and little to no expression in
normal tissues. In a phase 1 study, escalating doses of elotu-
Ixazomib zumab (2.5, 5, 10, and 20 mg/kg IV) were administered on
days 1 and 11 in combination with bortezomib (1.3 mg/m2
Ixazomib (MLN9708) has a response rate of 15% of relapsed/
IV) administered on days 1, 4, 8, and 11 of a 21-day cycle.
refractory patients (most of whom who had previously been
Dexamethasone 20 mg PO was added for patients with dis-
exposed to bortezomib) and was tested in a phase 3 trial in
ease progression The MTD was not reached, and the most
combination with Rd, which led to its FDA approval in 2015.
frequent grade 3/4 side effects were lymphopenia, fatigue,
Other oral proteasome inhibitors, such as oprozomib, are
thrombocytopenia, neutropenia, hyperglycemia, peripheral
also in an early phase of development.
neuropathy, pneumonia, and anemia. A partial response or
better was observed in 13/27 (48%) evaluable patients,
Other novel agents for relapsed or refractory diseasey including 7% CR and 41% PR. Another study examined the
drug in combination with lenalidomide, using escalating
Signal transduction inhibitors
doses of elotuzumab with standard lenalidomide dose. No
Various signal transduction pathways have been considered DLTs were observed at the maximum dose tested (20 mg/kg).
critical for the survival and proliferation of myeloma cells. A partial response or better was seen in 82% (23/28) of
Among these, PI3K-Akt pathway has been most studied treated patients and 96% (21/22) of lenalidomide-naive
from a therapeutic standpoint. Perifosine is the best-studied patients. Richardson et al. studied two doses of elotuzumab

in combination with lenalidomide and dexamethasone in a overall response rate (ORR) of 31% in relapsing patients, the
randomized phase 2 trial: patients received elotuzumab 10 or higher response rates in combinations with corticosteroids,
20 mg/kg IV (days 1, 8, 15, and 22 every 28 days in first 2 thalidomide, or bortezomib.
cycles and days 1 and 15 of subsequent cycles), lenalidomide
25 mg PO (days 1–21), and dexamethasone 40 mg PO weekly. Anthracycline-based regimens for relapsed
Among 63 enrolled patients, 81% had a PR including 37% or refractory disease
with VGPR. The overall response rates were 90% in the 10
mg/kg group (n = 31) and 72% in the 20 mg/kg group (n = Various permutations of doxorubicin-containing chemo-
32). Median time to response was 30 days (range, 21–100). therapy regimens—doxorubicin and cyclophosphamide
The most common toxicities were hematologic and infusion (AC), doxorubicin, BCNU, cyclophosphamide, and predni-
reactions. A recent randomized trial (ELOQUENT) has ran- sone (ABC-P), CAP, VCAP, VBAP, and BAP—have been tried
domized 646 relapse/refractory patients after 1-3 prior lines in patients with relapsed and refractory disease, resulting in
of therapy to receive lenalidomide-dexamethasone plus/ response rates of 7%-28%. Response duration and survival
minus elotuzumab; the triplet combination was associated tend to be short—<6 and 12 months, respectively. Respond-
with slightly higher RR (79% vs. 66%) and longer PFS (19.4 ing patients tend to live 7-10 or even 22 months longer than
vs. 14.9 months). nonresponders. Patients who have relapsed disease, rather
Daratumumab is a monoclonal antibody that is directed than resistant or refractory disease, have good response rates.
against CD38, a cell surface marker that is highly expressed Augmenting these combinations with high-dose corticoste-
on plasma cells (approved by FDA in 2015). Results from roids has been tried with some benefit, yielding response
phase 1/2 dose escalation studies have demonstrated activity rates of 47%-59% but with a high rate of infection. In sin-
in monotherapy, with 30%-40% responses at the optimized glearm studies, there does not appear to be any advantage to
doses and only mild infusion reactions, which were well con- the addition of cyclophosphamide to VAD, VAMP, or vincris-
trolled with steroids. In the SIRIUS trial that included 112 tine, epirubicin, and dexamethasone (VED).
heavily pretreated and double refractory MM patients, sin-
gle-agent daratumumab induced 29% RR, with a DOR of 7.4 Management of high-risk myeloma
months and a PFS of 3.7 months. These positive results have and risk-adapted therapy
prompted the investigation of anti CD38 in combination
Similar to other hematologic malignancies, in MM (MM)
with lenalidomide plus dexamethasone, with RR close to
patient-specific variables have been long recognized to influ-
90% in lenalidomide sensitive patients and 50% in patients
ence prognosis. There are three main patient characteristics
previously refractory to lenalidomide. Combination with
that influence survival: age, comorbidities (renal failure, car-
bortezomib is also under investigation. Isatuximab is another
diac failure, etc.), and frailty. Disease-related risk factors are
monoclonal antibody that is directed at the same target and
mainly represented by biomarkers cytogenetics [t(4;14), 17
is currently under investigation.
deletion, t (14;20), t(14;20), +1q and complex karyotyping]
and molecular signatures that are associated with outcome.
Alkylator-based regimens for relapsed or refractory disease
In addition, there are many factors related to tumor burden,
There is cross-resistance among the alkylators, but it is not including advanced clinical stage (III), high LDH, high num-
absolute and may be circumvented by increasing dose inten- ber of circulating PCs (CPC), and extramedullary disease
sity. Without extreme dose intensification, 5%-20% of (EMD). In addition, resistance to therapy is the major cause
patients with melphalan-resistant disease respond to cyclo- of treatment failure (early relapse, refractory disease, or sub-
phosphamide or BCNU as single agents or in combination optimal response). The expected survival in these patients is
with prednisone. Response rates as high as 30%-38% can be usually inferior to 3 years, and it can be even shorter in the
obtained if prednisone is administered with the cyclophos- following subgroups: (i) patients with coexistence of two or
phamide. Higher doses of cyclophosphamide (eg, 600 mg/m2 more adverse cytogenetic features; (ii) one cytogenetic
intravenously for 4 consecutive days) result in response rates adverse feature plus either high LDH or advanced clinical
of 29%-43%. Both response duration and OS tend to be stage (ISS 3) or failure to achieve a complete response with
short, approximately 3 and 9 months, respectively. Consoli- induction, or patients who have adverse cytogenetics but fail
dating the chemotherapy into a 1-day schedule rather than a to eradicate residual disease; (iii) patients with a high num-
4-day schedule does not improve response rate but does ber of CPCs; and (iv) finally, patients who fail to respond
increase the toxicity. Similarly, administration of 3.6 g/m2 (less than partial response to an optimized induction ther-
over 2 days with prednisone appears to produce comparable apy), that includes a proteasome inhibitor and an immuno-
response. Bendamustine as a single agent produced an modulatory agent.

Choice of therapy in high-risk myeloma patients tandem autologous stem cell transplants (ASCT), with a
doubling of PFS (42 months versus 21 months, P = 0.004)
The concept of high-risk cytogenetics emerged from data
and 4-year OS (76% versus 33%, P = 0.0001), as compared
using induction therapy with conventional chemotherapy
with single ASCT. Although some data suggest that high-risk
followed by ASCT, showing a 20%-50% decrease in OS for
patients may benefit from allogeneic stem cell transplant, the
high- as compared to standard-risk patients. The first novel
data correspond to small selected series of patients.
drug tested was thalidomide used as either induction (TD,
The data with the second-generation novel agents carfil-
TAD, CTD) or maintenance therapy, and again high-risk
zomib and pomalidomide is primarily in the relapsed/refrac-
patients did significantly poorer than standard-risk. More-
tory MM (RRMM). In the pivotal phase 2 trial that led to the
over, studies derived from the UK group (MRC IX intensive,
approval of carfilzomib, patients with isolated translocation
MRC IX non intensive, MRC IX maintenance) indicate that
(4;14) had a remarkably high ORR 63.6% with a median PFS
thalidomide is not superior to conventional chemotherapy
of 4.1 months and OS of 15.8 months (similar to standard
in patients with high-risk cytogenetic abnormalities. Regard-
risk 4, 6, and 19 months), suggesting that this group did as
ing lenalidomide, a small study conducted by Kapoor et al. in
well as the standard-risk group. This reflects perhaps a class
newly diagnosed patients showed that high-risk patients dis-
effect, as bortezomib also benefits the t(4;14) MM. Carfilzo-
play significantly shorter PFS (18 vs. 26 months, for high vs.
mib, however, did not improve the poor outcome of dele-
standard risk, respectively). THE IFM group has shown that
tion17p either by itself or in combination with other
lenalidomide maintenance is of benefit in patients with dele-
abnormalities. The MM-003 phase 3 trial (56) as well as an
tion 17p (PFS: 29 vs. 14 months). Nevertheless, it should be
IFM phase 2 study (57) showed that pomalidomide–low-
noted that these PFS values are clearly inferior to those of the
dose dexamethasone might provide a comparable survival
overall series of patients (42 months). Therefore, it could be
benefit to deletion 17p RRMM patients (12.6 months vs. 14
concluded that lenalidomide maintenance improves the out-
months to no deletion 17p). Moreover, a phase 1/2 trial
come especially in patients with del (17p) but does not com-
combined carfilzomib-pomalidomide-dexamethasone for
pletely overcome the poor prognosis of high-risk
RRMM saw no differences between standard- and high-risk
cytogenetics.
cytogenetics in terms of median PFS or OS. Still, it remains
There are now strong data that demonstrate that the poor
to be seen whether these data can be validated in prospective
prognosis associated with chromosomal translocation (4; 14)
trials, especially in the newly diagnosed setting.
may be improved by the addition of bortezomib as part of
induction and consolidation in newly diagnosed transplant-
eligible patients. In a metaanalysis of 3 European trials, Cavo Supportive care
et al. present compelling data supporting benefit from a
Bone disease: assessment and treatment
bortezomib-based regimen in patients with high-risk cyto-
genetics, especially translocation (4;14) and deletion 17p, Bone involvement is the most frequent clinical complication
since this improve outcome although does not completely in patients with MM. About 70% of patients have lytic bone
overcome adverse prognosis of these abnormalities, parti lesions with or without osteoporosis, and another 20% have
cularly deletion 17p. In transplant ineligible patients severe osteoporosis without lytic lesions. This frequency cor-
with high-risk MM, the Spanish GEM-05 trial that inclu responds to conventional skeletal radiography assessment, a
ded bortezomib-melphalan-prednisone versus bortezomib- technique that is associated with low sensitivity. CT has the
thalidomide-prednisone induction revealed shorter PFS/OS highest sensitivity for the detection of bone defects and with
for HRMM. In the lenalidomide-dexamethasone (high vs. the whole body low-dose modality the radiation exposure is
low dose) trial the 2-year OS was also significantly shorter much lower than with conventional CT; the scanning time is
for high-risk patients (76% vs. 91%). An IFM study includ- short and may replace conventional X-ray in the near future.
ing 1,095 elderly patients shows that whatever the treatment MRI has the highest resolution for soft tissue and bone mar-
t(4;14) and deletion 17p were associated with shorter PFS row infiltration; it is particularly valuable for differentiation
and OS as compared to patients without these abnormalities. between benign and malignant fractures but is inferior to CT
Therefore, the first generation novel agents may have for assessment of bone disease. Finally, PET allows assess-
improved but certainly did not overcome the adverse prog- ment of tumor metabolism and disease activity (vs. inactive
nosis of high-risk MM in elderly patients not eligible for or necrosis) and may be of prognostic significance; however,
HDT. Regarding ASCT, Cavo et al., in a meta-analysis from it still requires standardization.
3 European phase 3 trials has shown that patients with high- The intravenous agents pamidronate and zoledronic acid
risk cytogenetics and who failed to achieve CR after are of clinical benefit in the treatment of bone disease in
bortezomib-based induction did significantly better with patients with MM. Pamidronate is administered at a monthly

dose of 90 mg via a 2-hour intravenous infusion. Zoledronic insufficiency, and chemotherapy with cytotoxic agents.
acid, at a monthly dose of 4 mg, is at least as effective as Severe neutropenia and thrombocytopenia at the time of
pamidronate and has the advantage that it can be adminis- diagnosis are unusual. About 10% of patients have a platelet
tered via a 15-min infusion. In patients with renal function count of <100 × 109/L, but platelet counts lower than
impairment, the dose of zoledronic acid must be reduced to 20 × 109/L with risk of severe bleeding are very unusual
a maximum of 3 mg. It was suggested that bisphosphonates unless therapy-related myelodysplastic syndrome is present.
should be used indefinitely once initiated. However, the A number of trials have shown the beneficial effect of
appearance of severe late complications, such as osteonecro- recombinant human erythropoietin and darbepoetin alfa in
sis of the jaw, related to the duration of bisphosphonate the treatment of myeloma-associated anemia. Levels above
exposure has resulted in a reconsideration of the initial rec- 12 g/dL should be avoided due to association with a higher
ommendations. Osteonecrosis of the jaw is associated with risk of thrombosis and poorer outcomes. The major cause of
the duration of bisphosphonate exposure, type of bisphos- erythropoietin failure is iron deficiency. Iron repletion
phonate (eg, higher with zoledronic acid than with pamidro- should be indicated when there is evidence of functional iron
nate), and history of recent dental procedure. The current deficiency measured by an increased soluble transferrin
recommendations for treatment with bisphosphonates in receptor. It seems that the best iron supplemental therapy is
MM patients, based on consensus panels from both the the administration of iron saccharide. Treatment with gran-
IMWG and the ASCO, do not recommend the initial use of ulocyte colony-stimulating factor (G-CSF) may be required
bisphosphonates for more than 2 years. In relapsed patients, to treat chemotherapy-induced severe granulocytopenia.
treatment with bisphosphonates can be restarted and admin- Patients treated with lenalidomide may require G-CSF ther-
istered concomitantly with active therapy. Finally, in pati apy, but dose reduction or selection of an alternate agent is
ents in whom the bone disease is a consequence of excess usually a better strategy.
RANK-L activity, newer molecules such as denosumab
might be of benefit. Renal failure
Between 15% and 20% of patients with MM have hypercal-
cemia at the time of diagnosis. A common complication of About 20% of patients with MM have a serum creatinine
hypercalcemia is renal impairment caused by interstitial higher than 2 mg/dL at diagnosis. However, in some series,
nephritis. Treatment of hypercalcemia with hydration and up to 10% of patients with newly diagnosed MM have renal
bisphosphonates is a medical emergency. Zoledronic acid is the failure severe enough to require dialysis from the time of
bisphosphonate of choice (quicker response and significantly diagnosis. The main causes of renal failure in MM are light-
longer time to recurrence compared with pamidronate). chain excretion resulting in cast nephropathy (myeloma
Some patients develop pathological fractures of long kidney) and glomerular deposition of immunoglobulin
bones and require orthopaedic surgery. In the event of (light-chain amyloidosis or immunoglobulin deposition dis-
extensive lesions, stabilization surgery can be followed by ease). The prognosis mainly depends on the reversibility of
local radiation therapy. Prophylactic orthopaedic interven- renal function. The median survival of patients with revers-
tion should be considered in patients with large lytic lesions ible renal failure is similar to that of patients with normal
in weight-bearing bones at high risk of fracture. Patients renal function, whereas patients with nonreversible renal
with severe back pain due to vertebral compression fractures failure have a median survival of less than 12 months.
can benefit from vertebroplasty or kyphoplasty. Spinal cord Until recently vincristine–doxorubicin (Adriamycin)–
compression caused by a vertebral fracture is very rare in high-dose dexamethasone (VAD) or high-dose dexametha-
patients with MM. This complication instead is usually sone was the treatment of choice for patients with renal
caused by a plasmacytoma arising from a vertebral body, and insufficiency. But this has changed with the availability of
management is described further below. novel drugs. Taking into account that the action of bortezo-
mib is very quick, it is probably the ideal agent for rapidly
decreasing light chains in order to prevent the development
Anemia and bone marrow failure
of irreversible renal failure by avoiding further tubular light-
Approximately 35% of patients with newly diagnosed MM chain damage. In a retrospective series of 24 patients with
have hemoglobin lower than 9 g/dL. In addition, severe ane- relapsed/refractory MM and dialysis-dependent renal fail-
mia is a frequent complication later in the course of the ure, the overall response rate (RR) was 75%, with 30% com-
disease due to disease progression. Anemia is associated plete remissions (CR) or near-CR. Subsequent studies have
with a significant loss in quality of life and poor prognosis. confirmed the benefit of bortezomib-based therapies (in
The main causes of anemia in MM are bone marrow replace- combination with dexamethasone +/− doxorubicin or
ment by PCs, relative erythropoietin deficiency, renal IMiDs) in patients with newly diagnosed myeloma and renal

failure. Thalidomide can also be used in patients with renal a loading dose of 100 mg followed by 25 mg every 6 hours
failure, as well as lenalidomide; however, the dose of lenalid- followed by progressive tapering. Simultaneous local radia-
omide must be adjusted to the degree of renal failure. tion therapy should be started as soon as possible. If the spi-
With regard to the use of high-dose therapy/autologous nal cord compression is caused by a vertebral collapse or by
stem cell transplantation (SCT) in patients with MM and spinal instability rather than a plasmacytoma (which is very
renal failure, the largest experience comes from the Arkan- rare), urgent surgical decompression followed by fixation
sas group, with a reversibility of renal failure of 43% but using a bone graft or methacrylate is required.
higher morbidity and mortality (6% and 13% after a single
or tandem transplant, respectively) than in patients with Infection
normal renal function. Chemoresistant disease, low serum
albumin, and older age were associated with a poorer out- Infectious complications are a major cause of morbidity and
come. In any event, the dose of melphalan must be reduced mortality in patients with MM. The highest risk of infection is
to 140 mg/m2. In patients with no overt myeloma and low observed during the first 2 months of starting therapy, in
PC mass in whom renal function impairment is due to glo- patients with severe chemotherapy-induced granulocytope-
merular light-chain deposition (light-chain deposition dis- nia, and in those with relapsed and refractory disease. The
ease), the likelihood of response is higher than that in MM main cause of infection in MM is the impaired antibody pro-
because of the low PC mass at the time of transplantation. In duction, leading to a decrease in the uninvolved immunoglob-
this situation, there is no need for tumor reduction with ulins. Other important causes include chemotherapy-induced
induction chemotherapy before stem cell mobilization and granulocytopenia, renal function impairment, and glucocorti-
high-dose therapy. coid treatment, particularly high-dose dexamethasone. Most
Theoretically, the removal of nephrotoxic light chains infections in newly diagnosed patients and during the first
with plasma exchange could avoid further renal failure and cycles of chemotherapy are caused by Streptococcus pneu-
hopefully prevent irreversible renal failure. The Mayo Clinic moniae, Staphylococcus aureus, and Haemophilus influenzae;
group, in a small controlled trial, compared chemotherapy while in patients with renal failure, as well as in those with
with chemotherapy plus plasma exchange and found only a relapsed and/or refractory advanced disease, >90% of the
trend in favor of the group including plasma exchange. Simi- infectious episodes are caused by Gram-negative bacilli or
larly, in a large randomized trial, there was no conclusive S aureus.
evidence that plasma exchange improved the outcome of An infectious episode in a patient with MM should be
patients with MM and acute renal failure. When excluding managed as a potentially serious complication requiring
the patients who die in this early period, the median survival immediate therapy. In case of suspected severe infection and
of patients with MM and nonreversible renal failure needing before the identification of the causal agent, treatment
chronic dialysis is almost 2 years, and 30% of them survive against encapsulated bacteria and gram-negative microor-
for more than 3 years. Thus, long-term dialysis is a worth- ganisms should be initiated.
while supportive measure for patients with MM and end- Although prophylaxis of infection in patients with MM is
stage renal failure. The use of high cut-off dialysis filters is a controversial issue, some general guidelines can be offered.
very promising, and hopefully prospective ongoing studies Intravenous immunoglobulin prophylaxis is not recom-
will confirm its benefit. mended. Pneumococcal vaccination is recommended, par-
ticularly in patients with IgG myeloma with high-serum
M-protein levels, which are usually associated with very low
Spinal cord compression
levels of uninvolved immunoglobulins. Antibiotic prophy-
Spinal cord compression from a plasmacytoma, which laxis is likely of benefit within the first 2 months of initiation
occurs in about 10% of patients, is the most frequent and of therapy, especially in patients at high risk of infection
serious neurological complication in MM. The dorsal spine (recent past history of serious infections, such as recurrent
is the most common site of involvement, followed by the pneumonia or renal failure). Patients treated with bortezo-
lumbar region. The clinical picture of spinal cord compres- mib should receive prophylaxis against varicella zoster
sion consists of back pain and paraparesis. Although it can infections.
evolve for several days or even a few weeks, the onset can be
abrupt, resulting in severe paraparesis or paraplegia in a few
Venous thromboembolism
hours. Spinal cord compression is an emergency requiring
immediate medical action. When suspected, urgent MRI Patients with MM have an increased risk of thrombosis, with
should be performed. If confirmed, treatment with high- a baseline risk of 3%-4% of venous thrombotic events. This
dose dexamethasone must be started immediately at risk is significantly enhanced in the face of therapy, with

higher risk associated with use of specific agents. High-dose the drugs, especially some of the oral proteasome inhibitors
dexamethasone, cytotoxic chemotherapy such as doxorubi- currently in clinical trials. These patients should be managed
cin, and immunomodulatory drugs (thalidomide and symptomatically, and dose reduction should be pursued
lenalidomide) increase this risk substantially. Other factors when feasible.
such reduced mobility due to neurological complications or
bone pain, associated fractures, concurrent use of erythro-
poietic agents, prior personal or family history of thrombotic Other PC disorders
events, all increase the risk of thromboembolic events. The
Solitary plasmacytoma of bone
current recommendations in patients with MM who are
started on the IMiDs is to use full-dose aspirin in the absence The existence of a solitary plasmacytoma has been recognized
of risk factors for thrombosis, and use full-dose anticoagula- in up to 3% of patients with a PC dyscrasia, usually on the
tion for those at higher risk. Subtherapeutic doses of antico- vertebral column. The diagnostic criteria require the exis-
agulants such as fixed, small doses of warfarin are not tence of a solitary PC tumor in which the biopsy confirms PC
recommended. histology, a negative skeletal survey, absence of PC infiltra-
tion in a random bone sample (<10%), as well as no evidence
of anemia, hypercalcemia, or renal impairment. Some groups
Management of treatment-related toxicities
suggest that patients in whom a paraprotein persists after the
In addition to VTE that has been described above, there are eradication of plasmacytoma with local treatment should
several common toxicities that are encountered with the cur- undergo a review of the diagnosis. The treatment of choice is
rently used anti-myeloma agents. Hematological toxicity local radiotherapy, but about two-thirds of patients with soli-
(myelosuppression) is the most common and is seen with tary bone plasmacytoma develop MM at 10 years’ follow-up,
nearly all the drugs with the exception of corticosteroids and with a median time to progression of 2 years.
thalidomide. Neutropenia can be seen with nearly all classes
of drugs including the traditional cytotoxic drugs as well as
Extramedullary plasmacytoma
newer agents such as lenalidomide and pomalidomide. The
mechanism of neutropenia with IMiDs is believed to be a Extramedullary plasmacytoma is a PC tumor that arises out-
maturation blockade rather than inhibition of cell division side the bone marrow, most frequently in the upper respira-
as with the traditional chemotherapy. This should be man- tory tract (nose, paranasal sinuses, nasopharynx, and
aged through a combination of dose reduction as well as use tonsils). Other sites include parathyroid gland, orbit, lung,
of growth factors. Use of growth factors can follow the gen- spleen, gastrointestinal tract, testes and skin. In most cases
eral ASCO guidelines on their use. Thrombocytopenia can the lesion is unique, although the presence of more lesions
also be seen with all these drugs, though they may be more (multiple plasmacytomas) has also been reported. Diagnosis
profound in the context of the proteasome inhibitors such as is based on the detection of the PC tumor in an extramedul-
bortezomib and carfilzomib. The thrombocytopenia associ- lary site, in the absence of bone marrow PC infiltration, bone
ated with PIs tends to be more transient and cyclic with rapid lytic lesions, and other signs of MM (end-organ damage).
recovery following the initial effect of the drug. Lymphope-
nia can be seen with many of the drugs, especially PIs, but
Nonsecretory MM
typically do not need dose modifications. There is increased
risk of infections, especially herpes zoster reactivation with This specific type of MM requires particular attention, since
PIs, and these patients should be prophylactic acyclovir. it is very difficult to diagnose. The only way to make a defini-
Peripheral neuropathy can be associated with many of tive diagnosis is to demonstrate the presence of tissue infil-
these drugs, especially bortezomib and thalidomide as well tration (usually bone marrow) by cells with PC morphology.
as vincristine, which is hardly used as part of myeloma ther- However, PC infiltration must be >10%, and clonality must
apy anymore. It is important to carefully ask the patients be assessed by immunophenotyping (demonstration of
regarding neuropathy symptoms in order to identify it early, cytoplasmic immunoglobulins with restricted light chain:
and dose reductions can be instituted. In patients with pain- positive production without excretion). In addition, the
ful neuropathy, the offending drug should be discontinued. serum free light chains are abnormal and this is a most useful
Gastrointestinal toxicity is also commonly encountered parameter for the follow-up. However, exceptional cases
with many of the drugs. Diarrhea can accompany the use of exist in which no monoclonal protein can be observed within
bortezomib, carfilzomib, panobinostat, and potentially with the PCs. In these cases, it is mandatory to demonstrate clon-
long term use of lenalidomide. Constipation is a common ality by the study of the rearrangement status of the immu-
side effect of thalidomide. Nausea can be seen with many of noglobulin genes.

Other PC disorders | 667
Plasma cell leukemia Epidemiology

Plasma cell leukemia (PCL) is a rare, aggressive form of MM AL amyloidosis typically develops from the background of
characterized by high levels of PCs circulating in the periph- MGUS but can be associated with other lymphoproliferative
eral blood. PCL can either originate de novo (primary PCL) disorders in which there is excess secretion of κ or λ free light
or as a secondary leukemic transformation of MM (second- chains, including WM or chronic lymphocytic leukemia. It
ary PCL). It was initially described by Robert Kyle in 1974 as may be diagnosed simultaneously with symptomatic MM, or
blood plasmacytosis of more than 20% of total nucleated may precede or follow a diagnosis of MM. In a series of 1,596
cells, or an absolute number of circulating PCs >2 × 109/L. patients with AL amyloidosis seen at the Mayo Clinic, only
The circulating PCs appear morphologically similar to the six (0.4 %) who did not have MM at diagnosis later showed
marrow PCs, though plasmablastic morphology is common, progression to overt myeloma. In a series of 4,319 patients
and these cells often lack CD56 expression, in contrast to with a diagnosis of MM who had at least six months of
more typical MM cells. From a cytogenetic standpoint, all follow-up, 47 patients (1.1 %) had a subsequent diagnosis of
abnormalities seen in MM can also be seen in PCL, but there AL amyloidosis.
appears to be a higher prevalence of monosomy 13, deletion
17p, and abnormalities in chromosome 1, in particular 1q21 Clinical presentation
amplification and del1p, abnormalities typically seen in
higher proportion in relapsed myeloma. The clinical presentation is dictated by the spectrum and
Despite introduction of novel agents for MM, the out- severity of the organs involved and can be varied with non-
comes of patients with PCL remains uniformly poor, with specific symptoms. The common presentations based on the
median OS of only around 1 year. In patients with secondary organ system are detailed in Table 23-13. Approximately
PCL, the survival is even shorter. Modestly improved sur- 10% of patients have coexisting MM, characterized by ≥10 %
vival has been observed in recent years, as shown by an anal- PCs on bone marrow examination or lytic bone lesions.
ysis of the SEER database of 445 patients with primary PCL Such patients may present with signs and symptoms related
diagnosed between 1973 and 2009, which reported median to the MM, such as anemia, bone pain, hypercalcemia, and
overall survival times of 5, 6, 4, and 12 months for those infection.
patients diagnosed during 1973-1995, 1996-2000, 2001-
2005, and 2006-2009, respectively. There are no specific Diagnosis and staging
treatment approaches for PCL, but multidrug combinations
including both proteasome inhibitor and an IMiD appear to Diagnosis of AL amyloidosis hinges on demonstrating the
be a logical choice, along with the use of HDT/ASCT or allo- presence of amyloid deposition in any body tissue and prov-
geneic transplant in eligible patients, followed by prolonged ing that this amyloid arises from clonal immunoglobulin
maintenance until progression. light chains. In addition, identification of a monoclonal pro-
tein in the serum or urine provides supportive information
for the diagnosis and a measure to follow disease response.
Light-chain amyloidosis The latter will allow classification of the primary disorder,
which in the majority would be classified as MGUS if not for
Systemic amyloidosis represents a spectrum of disorders the presence of amyloid formation.
characterized by deposition of insoluble beta pleated sheets
of amyloid fibrils in various organs, leading to major organ
Detecting amyloid deposition
dysfunction that can be fatal. While over 40 different pro-
teins (eg, transthyretin [TTR]) have been described as poten- The typical sites for demonstrating amyloid deposits are the
tially amyloidogenic, the most common form of amyloidosis, organs involved or the bone marrow and abdominal subcu-
and the one that is the subject of current discussion, is the taneous fat. Less invasive procedures can be employed to
immunoglobulin light chain-derived systemic amyloidosis, demonstrate amyloid using abdominal fat pad aspirate
also called AL amyloidosis or just AL. (60%-80%), rectal biopsy (50%-70%), or bone marrow
AL is rare; the incidence is approximately 6-10 cases per biopsy (50%-55%). A combination of fat aspirate and a mar-
million person-years. Like other PC dyscrasias, the age- row biopsy is preferred, as patients typically would undergo
specific incidence rates increase in each decade of life after a bone marrow examination for their underlying monoclo-
age 40 years. The median age at diagnosis is 64 years, and nal gammopathy, and a fat aspirate can be performed conve-
less than 5% of patients with AL are <40 years. There is a niently at the same time. Either or both are positive in 90%
slight male predominance with nearly 60% of patients of patients with AL amyloidosis. If these sites are absent for
being male. amyloid, biopsy directed towards the affected organ should

Table 23-13 Spectrum of organ involvement and clinical features in AL amyloidosis
Organ Clinical features

Kidney Involved in ~70% of patients
Typically presents as nephrotic range proteinuria, renal; renal failure at diagnosis uncommon
Edema, hyperlipidemia
Heart Seen in approximately 60% of patients
Typically presents with restrictive cardiomyopathy or conduction disturbances, arrhythmias, rarely with
microvascular angina due to vascular involvement
Dyspnea on exertion, orthopnea, syncope, edema, fatigue, sudden cardiac death, signs of congestive failure
Liver Hepatic involvement can be seen in up to 60% of patients
Presents with elevated liver tests, especially elevated alkaline phosphatase and bilirubin; frank hepatic failure uncommon
Hepatomegaly, weight loss, fatigue, jaundice
Nervous system Typically affects peripheral nerves, sensory more than motor or autonomic nerves
Numbness, paresthesia, and pain due to peripheral nerves; postural hypotension, bladder and bowel abnormalities,
related to autonomic dysfunction
Gastrointestinal tract Approximately 30% of patients
Bleeding, gastroparesis, constipation, bacterial overgrowth, malabsorption, and intestinal pseudo-obstruction
resulting from dysmotility
Soft tissue and muscle Seen in nearly a third of the patients
Macroglossia, proximal muscle weakness
Coagulation system Increased bleeding or skin purpura related to vascular friability, altered coagulation profile with dysfibrinogenemia
and factor X deficiency, decreased synthesis of coagulation factors in patients with advanced liver disease; and
acquired von Willebrand disease
be utilized. On hematoxylin- and eosin-stained biopsy sec- or urine protein electrophoresis with immunofixation can
tions, amyloid appears as a pink, amorphous, waxy sub- identify a monoclonal protein in nearly 90% of the patients.
stance, but it binds strongly to Congo red (ie, it is Addition of the serum free light-chain assay to the diagnostic
“Congophilic”), imparting a green birefringence under work up increases the yield to over 98% of the patients. Most
polarized light, and to thioflavine-T, producing an intense patients with AL amyloidosis have little or no intact mono-
yellow-green fluorescence. clonal immunoglobulin but are characterized by the pres-
ence of monoclonal free light chain. The monoclonal light
chain type is λ in approximately 70% of cases, κ in 25%, and
Identifying amyloid type
biclonal in 5%.
Traditionally, the identification of the protein origin of the
amyloid fibrils has utilized immunohistochemistry or immu-
Clinical evaluation and disease staging
nofluorescence. However, this method can lead to false posi-
tives and can lead to an avoidable misdiagnosis as AL amyloid Laboratory studies should include a complete blood count
in the presence of a coincidental MGUS. It is important to with differential; chemistries with liver and renal function;
realize that over 30 different proteins have been identified that electrolyte panel; coagulation screening studies including
can lead to amyloid deposits. Laser microdissection with mass prothrombin time (PT), partial thromboplastin time (PTT);
spectrometry utilizes amyloid identified in biopsy specimens serum and urine protein electrophoresis with immunofixa-
using Congo red staining to be carved out of biopsy sample tion; serum free light chain assay; 24-hour urinary protein
and subjected to tandem mass spectrometry-based proteomic estimation; assessment of creatinine clearance; brain natri-
analysis. This technique can identify the amyloid type with uretic peptide (BNP) or NT-proBNP, troponin, and thyroid
over 98% specificity and sensitivity. Given the risk of false stimulating hormone (TSH). Detailed coagulation testing
positives (and negatives) with alternate methods, the type of should be considered for those with abnormal bleeding or
amyloid present is best established by mass spectrometry. screening tests. Bone marrow aspirate and biopsy along with
a fat aspirate should be done as discussed previously for
assessment of PCs as well as amyloid detection and identifi-
Detecting and quantifying the monoclonal process
cation. ECG and echocardiogram should be performed to
The PC proliferation in AL amyloidosis is typically low bur- assess for cardiac involvement; the echocardiographer should
den, with <10% PCs in over half of the patients. Serum and be alerted to the suspicion for amyloid, since findings can be

subtle. MRI can provide helpful information if echocardio- Treatment approaches

gram is nondiagnostic and suspicion is high. Patients with
Treatment approaches in AL amyloidosis have closely paral-
neurologic symptoms should have electromyography (EMG)
leled the developments in the field of MM, and many of the
and nerve conduction studies for diagnosis as well as baseline
drugs and regimens have evolved from the findings in MM.
assessment for future response determination. Gastric emp-
While the specific drugs that are currently employed reflect
tying studies may be of benefit in patients with upper gas
their use in PC disorders in general, specific doses and com-
trointestinal symptoms. Additional evaluation should be
binations, and the anticipated toxicities widely differ and
determined based on suspected organ involvement.
careful attention needs to be paid to this. The three most
The prognosis of AL amyloidosis varies considerably
common approaches in AL amyloidosis consist of HDT and
depending on the number and extent of organ involve-
autologous stem cell transplantation, melphalan and dexa-
ment. The overall outcome, while getting slightly better
methasone, and bortezomib based combinations. Small
with time, still remains poor with over 40% of patients
studies have also evaluated the majority of the commonly
dying within 1 year of diagnosis. Though multiple prognos-
used MM regimens in AL amyloidosis as well.
tic models have been proposed for patients with amyloido-
sis, models that incorporate the markers of cardiac damage
or dysfunction have high predictive value for early deaths in
Response assessment
AL amyloidosis. The Mayo Clinic Amyloid Staging system
classifies patients into having stage I, II, III, or IV disease There are critical differences between response assessment in
based upon the identification of zero, 1, 2, or 3 of the fol- AL amyloidosis (Table 23-14) and that used in MM. Response
lowing risk factors: NT-pro-BNP ≥1,800 ng/L, cardiac tro- assessment in AL amyloidosis needs to capture both the
ponin T ≥0.025 μg/L, and a difference between involved hematologic response as well as any improvements in organ
and uninvolved serum free light chains ≥18 mg/dL. Median function, the latter being more important from a patient
overall survival from diagnosis was 94, 40, 14, and 6 months, outcome standpoint. Given that nearly half of the patients
respectively. with AL amyloidosis do not have measurable levels of intact
Table 23-14 Criteria for assessment of treatment response in AL amyloidosis
Hematologic or organ
response Description of response
Hematologic response
Complete response (CR) Normalization of the FLC levels and ratio, negative serum and urine immunofixation
Very good partial response Reduction in the difference between involved FLC and uninvolved FLC (dFLC) to <40 mg/L
(VGPR)
Partial response (PR) ≥50 percent reduction in the dFLC
No response (NR) Less than a PR
Progression Free light chain increase of 50% to ≥100 mg/L. If patient achieved a CR previously, any detectable M protein or
abnormal FLC ratio (light chain must be double). If patient achieved a PR previously, 50% increase in serum
M protein to ≥0.5 g/dL or 50% increase in urine M protein to ≥200 mg/day (a visible peak must be present).
Organ response
Cardiac Response: NT-proBNP response (≥30% and ≥300 ng/L decrease in patients with baseline NT-proBNP
≥650 ng/L) or NYHA class response (≥2 class decrease in subjects with baseline NYHA class 3 or 4).
Progression: NT-proBNP progression (≥30% and ≥300 ng/L increase) or cardiac troponin progression (≥33%
increase) or ejection fraction progression (≥10% decrease).
Kidney Response: 50% decrease (at least 0.5 g/day) of 24-hour urine protein (urine protein must be ≥0.5 g/day
pretreatment). Serum creatinine and creatinine clearance (or calculated estimated GFR) must not worsen
by 25% over baseline.
Progression: 50% increase (at least 1 g/day) of 24-hour urine protein to ≥1g/day or 25% worsening of serum
creatinine or creatinine clearance
Liver Response: 50% decrease in abnormal alkaline phosphatase value. Decrease in liver size radiologically at
least 2 cm.
Progression: 50% increase of alkaline phosphatase above the lowest value.
Peripheral nervous system Response: Improvement in electromyogram nerve conduction velocity.
Progression: Progressive neuropathy by electromyography or nerve conduction velocity.

immunoglobulin M protein, serum free light-chain assay has The most commonly used conditioning regimen in AL
become the marker of choice for following the effect of treat- amyloidosis remains melphalan, 200 mg/m2, given over two
ments on the clonal PCs. However, the ultimate goal of ther- days. In patients with renal dysfunction and in those with
apy in AL amyloidosis is to reverse organ dysfunction. poorer performance status, a risk-adapted strategy of reduc-
Multiple studies have demonstrated a close relationship ing the melphalan dose has been tried, but studies suggest
between hematological response and organ response, with that dose reduction of melphalan may be associated with
deep responses (VGPR or better) being associated with a inferior outcomes. In the Mayo series of 454 patients, 100-
higher rate of organ response. day mortality was 9%. A partial response or better was seen
in 80%, including 40% with a CR. The median overall sur-
vival was 113 months with estimated rates of survival at 1
Initial treatment of AL amyloidosis
and 5 years of 87 and 66%, respectively. Estimated 5-year
The initial approach to treatment of AL amyloidosis depends survival rates for those attaining a hematologic CR, very
to a great extent on the eligibility for HDT/ASCT. Initial good partial response, partial response (PR), and less than a
experience with HDT in amyloidosis was beset with high PR were 90%, 74%, 56%, and 35%, respectively. Similar
treatment-related mortality related primarily to cardiac results were seen in a Boston University series, with a 100-
adverse events. Incorporation of standard prognostic factors day mortality of 11% and a median overall survival of 6.3
into transplant eligibility has greatly reduced the mortality years. Complete response (CR) was seen in 34% and trans-
associated with this procedure. Currently, patients with age lated into a superior event-free (8.3 versus 2 years) and over-
≤70 years, troponin T <0.06 ng/mL, NT-proBNP <5,000 ng/L, all (13.2 versus 5.9 years) survival.
ECOG performance status ≤2, New York Heart Association Recent studies have suggested a response and risk adapted
functional status class I or II, and no more than two organs strategy of using post-HDT consolidation in patients who
significantly involved (liver, heart, kidney, or autonomic fail to achieve a VGPR with the HDT. Initial results
nerve) can be considered for HDT. The decision to proceed appear promising, and this approach needs to be studied
should follow a careful discussion with the patient with prospectively.
respect to the potential toxicities and the anticipated out-
comes. It should also be kept in mind that patients who are
Melphalan and dexamethasone
initially considered ineligible for HDT might become eligible
following initial non-HDT therapy, especially in the setting Melphalan (0.22 mg/kg per day PO on days 1 through 4
of a deep hematological response. It remains unclear how every 28 days) combined with dexamethasone (40 mg/day
SCT compares with conventional therapies using melpha PO on days 1 through 4 every 28 days) is considered a stan-
lan dexamethasone- or a bortezomib-based regimen, as no dard regimen for patients with AL amyloidosis. Most of the
phase 3 trials have directly assessed this. The support for data supporting this regimen comes from small studies,
HDT is based largely on single-institution studies demon- mostly retrospective. In the initial studies, CR and PR rates
strating an improved outcome among patients who went to of 33% and 67%, respectively, were noted at 4 cycles, with
HDT compared with those who, though eligible, could not 71% of the responding patients also having significant
proceed to HDT or chose not to proceed. In addition, single improvement of involved organs. At a median follow-up of
center data as well as data from the CIBMTR suggest that 5 years, median progression-free and overall survivals were
patients undergoing HDT have high rates of hematological 3.8 and 5.1 years, respectively. In a subsequent report of 119
responses as well as organ responses and median progression HCT-ineligible patients treated with this regimen at this cen-
free survival measured in years. In contrast, a randomized ter, hematologic and complete response rates were 76 and 31
French trial demonstrated better outcomes with oral chemo- percent with a median survival of 7.4 years.
therapy compared with HDT.
Stem cell collection for HDT in AL amyloidosis is typically
Bortezomib-based regimens
performed with G-CSF alone, and the experience with che-
momobilization or plerixafor based mobilization remains Given the efficacy of bortezomib in MM and specifically the
limited. Unlike in MM, G-CSF mobilization can be associ- impact on serum free light chain levels, there has been sig-
ated with significant complications in this disease. Increased nificant interest in examining its role in patients with AL
risk of fluid retention, cardiac arrhythmias and sudden car- amyloidosis. In a single center retrospective series of 43
diac death, development of renal failure, large pleural effu- patients with AL amyloidosis treated with cyclophospha-
sions that may require thoracentesis, increased risk of mide, bortezomib, and the overall hematologic response
thrombosis as well as bleeding, have all been reported with rate of 81% including 42% CR rate. The estimated 2-year
stem cell mobilization in AL amyloidosis. progression-free survival was 67% and 41% for newly


diagnosed and relapsed patients, respectively. Bortezomib is secondary to hyperviscosity, cryoglobulinemia related,
typically given subcutaneously at 1.3 mg/m2 once a week, bleeding disorders, autoimmune hemolytic anemia), mar-
along with oral cyclophosphamide 500 mg once a week, and row or splenic infiltration by the lymphoplasmacytic cells
oral dexamethasone 40 mg once a week. Bortezomib has (hepatosplenomegaly, lymphadenopathy), or autoimmune
been studied in combination with dexamethasone in a non- phenomenon driven by the monoclonal protein (neuropa-
randomized phase 2 trial of bortezomib administered either thy). Most patients with WM present with nonspecific con-
once-weekly (1.6 mg/m2 days 1, 8, 15, and 22 of 35-day stitutional symptoms, and some of the patients may be
cycles) or twice-weekly (1.3 mg/m2 days 1, 4, 8, and 11 of asymptomatic at diagnosis. The most common presenting
21-day cycles). Seventy patients with relapsed AL amyloido- features include weakness, fatigue, weight loss, and chronic
sis were treated, with a hematologic response rate of 69% oozing of blood from the nose or gums. Recurrent infections
including 38% CRs. Estimated median overall survival was may occur due to a decrease in other unaffected immuno-
62 months. The twice weekly regimens had a similar response globulins. Clinical evaluation should include complete blood
rate but with higher rates of adverse events. count chemistry with BUN, creatinine, calcium, and LDH,
serum and urine protein electrophoresis, quantitative immu-
noglobulins, serum free light chain, serum viscosity, bone
IMiD-based regimens
marrow aspirate, and biopsy.
Thalidomide, lenalidomide, and pomalidomide have been To make a diagnosis of WM, an IgM monoclonal protein
studied in AL amyloidosis, either with dexamethasone or in of any size must be present in the serum, with 10% or more
combination with cyclophosphamide. The hematological infiltration of the bone marrow by small lymphocytes that
response rates range from 50% to 80%, with up to half of the exhibit lymphoplasmacytic features and express a typical
patients achieving an organ response. IMiDs are not as well immunophenotype: surface IgM+, CD5+/−, CD10−, CD19+,
tolerated as they are in the setting of myeloma, and lower CD20+, CD22+, CD23−, CD25+, CD27+, FMC7+, CD103−,
doses appear to mitigate this problem to some extent. IMiD CD138−. The PC component may be CD138+, CD38+, and
therapy has been associated with more toxicity among the CD45− or CD45dim. The phenotypic pattern is of critical
patients with significant heart disease and low starting blood importance in excluding other conditions, including chronic
pressures and may be related to lowering of blood pressure lymphocytic leukemia and mantle cell lymphoma.
seen with this class of drugs. Examination of the laboratory It is important to distinguish symptomatic disease from
tests among patients receiving IMiDs clearly show an increase early or precursor forms such as IgM MGUS or smoldering
in NT-proBNP levels accompanied by worsening cardiac WM. IgM MGUS is characterized by serum IgM concentra-
function, an effect that appear related to the therapy. tion <3.0 g/dL, absence of anemia, hepatosplenomegaly,
lymphadenopathy, systemic symptoms, and minimal
(<10%) or no lymphoplasmacytic infiltration of the bone
marrow. Patients who meet criteria for WM but have no end
WM is a rare disorder with an incidence of approximately 3 organ damage in the form of clinical symptoms and lack evi-
per million people per year with 1,400 new cases diagnosed dence of anemia, hepatosplenomegaly, lymphadenopathy,
in the United States each year. The median age at diagnosis is or hyperviscosity are considered to have smoldering WM.
64 years, with a gender distribution similar to other PC dis- IgM MM is quite rare, comprising only 0.5% of the Mayo
orders with approximately 60% being males. In contrast to Clinic series. Pathological distinction based on the marrow
MM, WM is much more common in European descent than appearance can be difficult in some instances, and clinical
in other ethnic groups. presentation may be relied on to correctly classifying these
The etiology of WM is unknown, though association with patients. Presence of lytic bone lesions clearly suggests the
infections and exposure to pesticides and other organic sol- presence of MM pathology rather than typical WM. In
vents suggest environmental impact. A recurrent mutation contrast, symptoms of hyperviscosity and the presence of
of the MYD88 gene (MYD88 L265P) is present in the vast lymphadenopathy or splenomegaly favor a diagnosis of WM.
majority of patients with WM, though this finding is not Other features that may help with making the diagnosis
specific to WM and can be seen in other B-cell neoplasms. includes the presence of typical chromosomal abnormalities
The pattern of somatic mutations suggests development at such as the IgH translocations seen in MM.
an early stage of B-cell differentiation: a post-germinal cen-
ter IgM memory B-cell that has undergone somatic hyper-
Treatment approaches
mutation but has failed to undergo isotype class switching.
The clinical presentation of WM is tied to the presence of Many patients with WM are asymptomatic; these patients
the IgM monoclonal protein in the blood (symptoms can be observed until they develop symptoms without

compromising their long-term outcomes. Indications for hematologic toxicities being the most common. Based on
treatment include systemic symptoms (weakness, fever, night this study, ibrutinib has been approved as a single agent for
sweats, fatigue, weight loss), along with physical findings the treatment of patients with WM.
(symptomatic lymphadenopathy, hepatomegaly, and/or Nucleoside analogs such as fludarabine or cladribine
splenomegaly, findings related to hyperviscosity), and cyto- have significant activity in WM and have been used in
penias (anemia, thrombocytopenia, neutropenia). Hypervis- combination with rituximab in regimens such as cladribine
cosity can lead to a variety of symptoms, including mucosal and rituximab, fludarabine and rituximab, and fludara-
bleeding, blurred vision, headaches, dizziness, paresthesias, bine, cyclophosphamide and rituximab (FCR). However,
retinal vein engorgement and flame-shaped hemorrhages, use of this class of drugs have been associated with stem
papilledema, and neurologic impairment, all of which points cell toxicity with subsequent myelodysplasia diagnosis
towards the need for instituting therapy. as well as increased risk of transformation to high-grade
The initial management of patients with symptomatic lymphoma. The predominant short-term toxicities with
WM depends on the age and the potential for HDT and nucleoside-analog-containing regimens are myelosuppres-
ASCT, functional status, presence and severity of symptoms, sion and immunosuppression.
especially hyperviscosity related symptoms and presence of
other comorbidities. Patients who can be future candidates
POEMS syndrome
for autologous HCT should avoid treatment with agents that
might interfere with stem cell collection (alkylators except POEMS (polyneuropathy, organomegaly, endocrinopathy,
for cyclophosphamide, purine nucleoside analogs). monoclonal protein, skin changes) syndrome (also referred
Patients with symptoms of hyperviscosity require emer- to as osteosclerotic myeloma, Crow-Fukase syndrome, PEP
gent plasmapheresis in addition to specific systemic therapy syndrome [PC dyscrasia, endocrinopathy, polyneuropathy],
for WM. The large size of the IgM molecule allows for rapid or Takatsuki syndrome) is a disorder characterized by the
removal using plasmapheresis resulting in symptomatic presence of a monoclonal PC disorder, peripheral neuropa-
improvement in a rapid fashion. Red blood cell transfusions thy, and one or more of the following features: osteosclerotic
should be avoided, if possible, prior to plasmapheresis, since myeloma, Castleman disease (angiofollicular lymph node
they might further increase serum viscosity. Along with plas- hyperplasia), increased levels of serum vascular endothelial
mapheresis, systemic therapy should be started. growth factor (VEGF), organomegaly, endocrinopathy,
Rituximab, an anti-CD20 monoclonal antibody, has little edema, typical skin changes, and papilledema (Table 23-15).
effect in MM but is an important component of the current The cause of POEMS syndrome is unknown, although
treatment regimens for WM. Patients with mild symptoms chronic overproduction of proinflammatory and other cyto-
and no urgent requirement for intervention can be consid- kines, like vascular endothelial growth factor and IL-6, have
ered for a rituximab therapy used as a single agent. While been implicated in the symptomatology seen in this disor-
rituximab is well tolerated and can be safely combined with der. It is believed that the stromal cells in response to the
a variety of other drugs, transient increase in serum IgM clonal PC population produce these cytokines.
levels (IgM flare) and associated hyperviscosity may occur The exact incidence of this disorder is unknown. In a
after the administration of rituximab and can lead to clini- Mayo Clinic series of 99 patients with POEMS, the median
cal consequences; therefore, careful short-term follow up is age was 51 years and 63% were males. Patients may present
recommended. with a wide constellation of symptoms and often this
In the vast majority of patients, rituximab should be com- makes the diagnosis difficult. Common symptoms and
bined with other chemotherapy regimens. The most com- signs, in addition to the polyneuropathy, monoclonal pro-
monly used regimens include dexamethasone, rituximab, tein with associated PC disorder, organomegaly, osteoscle-
cyclophosphamide (DRC), bortezomib plus rituximab with rotic bone lesions, skin changes and endocrinopathy,
or without dexamethasone (BRD), and bendamustine plus include weight loss, fatigue, papilledema, edema, ascites,
rituximab (BR). Overall response rates of 60%-90% have and pleural effusion. Peripheral neuropathy typically
been observed with these regimens. begins as a bilateral, symmetric, progressive symptoms
Ibrutinib, a small-molecule inhibitor of Bruton tyrosine including tingling and paresthesias, which can also include
kinase (BTK), showed considerable efficacy in WM in a motor innervation. EMG studies show slowing of nerve
phase 2 trial. Patients with symptomatic WM who had conduction, prolonged distal latencies, and severe attenua-
received at least one prior treatment had an overall response tion of compound muscle action potentials. Overall, 88%
rate of 62% when treated with daily ibrutinib in a phase 2 of patients had a monoclonal protein in the serum and/or
trial. Responses were relatively rapid, with median time to urine. The type of light chain seen in POEMS syndrome is
response of 4 weeks. Toxicity was very manageable with almost always λ.

Bibliography | 673
Table 23-15 Mayo Clinic criteria for the diagnosis of POEMS several years to improve fully, other symptoms tend to
syndrome respond rapidly following HDT. At a median follow-up of 45
Both of the following mandatory criteria must be present: months, 5-year overall and progression-free survival rates
Polyneuropathy were 94% and 75%, respectively.
Monoclonal PC proliferative disorder
Plusat least one additional major criterion:
Osteosclerotic or mixed sclerotic/lytic lesion visualized on plain Bibliography
films or computed tomography measuring at least 0.8 cm in the
longest dimension
MGUS
Castleman disease Bianchi G, Kyle RA, Colby CL, et al. Impact of optimal follow-
Elevated serum or plasma vascular endothelial growth factor up of monoclonal gammopathy of undetermined significance
(VEGF) levels at least three to four times the upper limit of on early diagnosis and prevention of myeloma-related
normal complications. Blood. 2010;116:2019-2025.
Plusat least one minor criterion: Bowden M, Crawford J, Cohen HJ, Noyama O. A comparative
Organomegaly (splenomegaly, hepatomegaly, or
study of monoclonal gammopathies and immunoglobulin
lymphadenopathy)
levels in Japanese and United States elderly. J Am Geriatr Soc.
1993;41:11-14.
Extravascular volume overload (peripheral edema, ascites, or
Cesana C, Klersy C, Barbarano L, Nosari AM, Crugnola
pleural effusion)
M, Pungolino E, et al. Prognostic factors for malignant
Endocrinopathy (excluding diabetes mellitus or hypothyroidism)
transformation in monoclonal gammopathy of undetermined
Skin changes significance and smoldering multiple myeloma. J Clin Oncol.
Papilledema 2002;20:1625-1634.
Thrombocytosis or polycythemia Cohen HJ, Crawford J, Rao MK, Pieper CF, Currie MS. Racial
differences in the prevalence of monoclonal gammopathy
in a community-based sample of the elderly. Am J Med.
Treatment 1998;104:439-444.
Katzmann JA, Clark R, Kyle RA, Larson DR, Therneau TM, Melton
In patients with one to three isolated bone lesions and no LJ III, et al. Suppression of uninvolved immunoglobulins
evidence of bone marrow involvement, limited field radia- defined by heavy/light chain pair suppression is a risk factor for
tion at a dose of 40 to 50 Gy is the preferred treatment progression of MGUS. Leukemia. 2013;27:208-212.
modality. In a retrospective study from the Mayo Clinic, Kyle RA, Therneau TM, Rajkumar SV, et al. A long-term study
radiation therapy to the lesions resulted in complete or of prognosis of monoclonal gammopathy of undetermined
partial hematologic response, VEGF response, FDG-PET significance. N Engl J Med. 2002;346:564-569.
Kyle RA, Therneau TM, Rajkumar SV, et al. Prevalence of
response, and clinical responses were seen in 31%, 14%,
monoclonal gammopathy of undetermined significance
22%, and 47%, respectively. Improvements were seen in
(MGUS) among Olmsted County, MN residents 50 years of age
peripheral neuropathy, anasarca, organomegaly, papill-
[abstract]. Blood. 2003;102:934a (A3476).
edema, skin changes, serum M-spikes, and plasma VEGF Landgren O, Kyle RA, Hoppin JA, et al. Pesticide exposure
levels. At a median follow-up of 43 months, the estimated and risk of monoclonal gammopathy of undetermined
overall and event-free survival at 4 years was 97% and 5%, significance in the Agricultural Health Study. Blood.
respectively. 2009;113:6386-6391.
Systemic chemotherapy should be considered for patients Landgren O, Kyle RA, Pfeiffer RM, et al. Monoclonal gammopathy
with widespread osteosclerotic lesions or bone marrow of undetermined significance (MGUS) consistently precedes
involvement. Many of the drugs used for myeloma treat- multiple myeloma: a prospective study. Blood. 2009;113:
ment, including melphalan and steroid combination, bort- 5412-5417.
ezomib, thalidomide and lenalidomide, have been used for Ögmundsdóttir HM, Haraldsdóttir V, Jóhannesson GM, et al.
Monoclonal gammopathy in Iceland: a population-based
treatment with varying degrees of benefit. Anecdotal reports
registry and follow-up. Br J Haematol. 2002;118:166-173.
suggest a role for agents with anti-cytokine/anti-VEGF activ-
Rajkumar SV, Kyle RA, Therneau TM, et al. Presence of
ity may help in ameliorating some of the signs and symp-
monoclonal free light chains in the serum predicts risk of
toms of this disorder. In young patients who require systemic progression in monoclonal gammopathy of undetermined
therapy, HDT followed by autologous stem cell transplanta- significance (MGUS). Br J Haematol. 2004;127:308-310.
tion has been shown to be of benefit. In a large series reported Saleun JP, Vicariot M, Deroff P, Morin JF. Monoclonal
from the Mayo Clinic, clinical improvement was seen in gammopathies in the adult population of Finistere, France.
nearly all patients. While neurologic symptoms often took J Clin Pathol. 1982;35:63-68.

Therneau TM, Kyle RA, Melton LJ III, et al. Incidence of Avet-Loiseau H, Li C, Magrangeas F, et al. Prognostic significance
monoclonal gammopathy of undetermined significance and of copy-number alterations in multiple myeloma. J Clin Oncol.
estimation of duration before first clinical recognition. Mayo 2009;27:4585-4590.
Clin Proc. 2012;87:1071-1079. Fonseca R, Bergsagel PL, Drach J, Shaughnessy J, Gutierrez N,
Stewart AK, et al. International Myeloma Working Group
molecular classification of MM: spotlight review. Leukemia.
Smoldering (or asymptomatic) MM 2009;23:2210-2221.
Gutierrez NC, Sarasquete ME, Misiewicz-Krzeminska I, et al.
Bianchi G, Kyle RA, Larson DR, Witzig TE, Kumar S, Dispenzieri
Deregulation of microRNA expression in the different genetic
A, et al. High levels of peripheral blood circulating PCs as
subtypes of MM and correlation with gene expression profiling.
a specific risk factor for progression of smoldering MM.
Leukemia. 2010;24:629-637.
Leukemia. 2013;27:680-685.
Kumar S, Fonseca R, Ketterling RP, et al. Trisomies in multiple
Dispenzieri A, Kyle RA, Katzmann JA, et al. Immunoglobulin free
myeloma: impact on survival in patients with high-risk
light chain ratio is an independent risk factor for progression
cytogenetics. Blood. 2012;119:2100-2105.
of smoldering (asymptomatic) multiple myeloma. Blood.
Lohr JG, Stojanov P, Carter SL, et al. Widespread genetic
2008;111:785-789.
heterogeneity in MM: implications for targeted therapy. Cancer
Dispenzieri A, Stewart AK, Chanan-Khan A, et al. Smoldering
Cell. 2014;25:91-101.
multiple myeloma requiring treatment: time for a new
Lopez-Corral L, Sarasquete ME, Bea S, et al. SNP-based mapping
definition? Blood. 2013;122:4172-4181.
arrays reveal high genomic complexity in monoclonal
Kyle RA, Larson DR, Therneau TM, et al. Clinical course of light-
gammopathies, from MGUS to myeloma status. Leukemia.
chain smouldering MM (idiopathic Bence Jones proteinuria): a
2012;26:2521-2529.
retrospective cohort study. Lancet Haematol. 2014;1:e28-e36.
Morgan GJ, Walker BA, Davies FE. The genetic architecture of
Kyle RA, Remstein ED, Therneau TM, Dispenzieri A, Kurtin
MM. Nat Rev Cancer. 2012;12:335-348.
PJ, Hodnefield JM, et al. Clinical course and prognosis of
Pichiorri F, Suh SS, Rocci A, et al. Downregulation of p53-
smoldering (asymptomatic) MM. N Engl J Med.
inducible microRNAs 192, 194, and 215 impairs the p53/
2007;356:2582-2590.
MDM2 autoregulatory loop in MM development. Cancer Cell.
Larsen JT, Kumar SK, Dispenzieri A, Kyle RA, Katzmann JA,
2010;18:367-381.
Rajkumar SV. Serum free light chain ratio as a biomarker for
Walker BA, Wardell CP, Chiecchio L, et al. Aberrant global
high-risk smoldering MM. Leukemia. 2013;27:941-946.
methylation patterns affect the molecular pathogenesis and
Mateos MV, Hernandez MT, Giraldo P, et al. Lenalidomide plus
prognosis of multiple myeloma. Blood. 2011;117:553-562.
dexamethasone for high-risk smoldering MM. N Engl J Med.
2013;369:438-447.
Pérez-Persona E, Vidriales M-B, Mateo G, et al. New criteria to Clinical presentation and diagnostic considerations
identify risk of progression in monoclonal gammopathy of Kyle RA, Rajkumar SV. MM. Multiple myeloma. N Engl J Med.
uncertain significance and smoldering multiple myeloma based 2004;351:1860-1873.
on multiparameter flow cytometry analysis of bone marrow Rajkumar SV, Dimopoulos MA, Palumbo A, et al. International
PCs. Blood. 2007;110:2586-2592. Myeloma Working Group updated criteria for the diagnosis of
Rajkumar SV, Gupta V, Fonseca R, Dispenzieri A, Gonsalves MM. Lancet Oncol. 2014;15:e538-e548.
WI, Larson D, et al. Impact of primary molecular cytogenetic
abnormalities and risk of progression in smoldering MM. Staging and risk stratification: Prognostic factors
Leukemia. 2013;27:1738-1744.
Chng WJ, Dispenzieri A, Chim CS, et al. IMWG consensus
on risk stratification in multiple myeloma. Leukemia.
Active (or symptomatic) MM 2014;28:269-277.
Durie BG, Salmon SE. A clinical staging system for MM.
Rollig C, Knop S, Bornhauser M. Multiple myeloma. Lancet.
Correlation of measured myeloma cell mass with presenting
2015;385:2197-2208.
clinical features, response to treatment, and survival. Cancer.
Vincent Rajkumar S. MM: 2014 Update on diagnosis, risk-
1975;36:842-854.
stratification, and management. Am J Hematol. 2014;89:999-1009.
Greipp PR, San Miguel J, Durie BG, et al. International staging
system for multiple myeloma. J Clin Oncol. 2005;23:3412-3420.
Kumar SK, Dispenzieri A, Lacy MQ, Gertz MA, Buadi FK, Pandey
Genomic abnormalities
S, et al. Continued improvement in survival in MM: changes
Avet-Loiseau H. Role of genetics in prognostication in myeloma. in early mortality and outcomes in older patients. Leukemia.
Best Pract Res Clin Haematol. 2007;20:625-635. 2014;28:1122-1128.
Avet-Loiseau H, Attal M, Moreau P, et al. Genetic abnormalities and Martinez-Lopez J, Lahuerta JJ, Pepin F, et al. Prognostic value of
survival in multiple myeloma: the experience of the Intergroupe deep sequencing method for minimal residual disease detection
Francophone du Myélome. Blood. 2007;109:3489-95. in multiple myeloma. Blood. 2014;123:3073-3079.

Bibliography | 675
Paiva B, Gutiérrez NC, Rosiñol L, et al. High-risk cytogenetics pretransplantation therapy in multiple myeloma: a randomized
and persistent minimal residual disease by multiparameter phase 3 PETHEMA/GEM study. Blood. 2012;120:1589-1596.
flow cytometry predict unsustained complete response after Reeder CB, Libby EN, Costa LJ, et al. A phase I/II trial
autologous stem cell transplantation in multiple myeloma. of cyclophosphamide, carfilzomib, thalidomide and
Blood. 2012;119:687-691. dexamethasone (CYCLONE) in patients with newly diagnosed
Rawstron AC, Child JA, de Tute RM, et al. Minimal residual MM: final results of MTD expansion cohort [abstract]. Blood.
disease assessed by multiparameter flow cytometry in multiple 2013;122(21):3179.
myeloma: impact on outcome in the Medical Research Council Sonneveld P, Asselberg-Hacker E, Zweegman S, et al. Dose
Myeloma IX Study. J Clin Oncol. 2013;31:2540-2547. escalation phase 2 trial of Carfilzomib combined with
thalidomide and low-dose dexamethason In newly diagnosed,
transplant eligible patients with multiple myeloma. A trial of the
Treatment approaches for MM European Myeloma Network [abstract]. Blood. 2013;122(21):688.
Alexanian R, Dimopoulos M. The treatment of multiple myeloma. Sonneveld P, Goldschmidt H, Rosiñol L, et al. Bortezomib-based
N Engl J Med. 1994;330:484-489. versus nonbortezomib-based induction treatment before
Mikhael JR, Dingli D, Roy V, et al. Management of Newly autologous stem-cell transplantation in patients with previously
Diagnosed Symptomatic MM: Updated Mayo Stratification of untreated multiple myeloma: a meta-analysis of phase III
Myeloma and Risk-Adapted Therapy (mSMART) Consensus randomized, controlled trials. J Clin Oncol. 2013;31:3279-3287.
Guidelines 2013. Mayo Clin Proc. 2013;88:360-376. Zervas K, Mihou D, Katodritou E, et al. VAD-doxil versus VAD-
doxil plus thalidomide as initial treatment for MM: results of
a multicenter randomized trial of the Greek Myeloma Study
Treatment of transplant-eligible patients Group. Ann Oncol. 2007;18:1369-1375.
Cavo M, Tacchetti P, Patriarca F, et al. Bortezomib with Zonder JA, Crowley J, Hussein MA, et al. Lenalidomide and
thalidomide plus dexamethasone compared with thalidomide high-dose dexamethasone compared with dexamethasone as
plus dexamethasone as induction therapy before, and initial therapy for multiple myeloma: a randomized Southwest
consolidation therapy after, double autologous stem-cell Oncology Group trial (S0232). Blood. 2010;116:5838-5841.
transplantation in newly diagnosed MM: a randomised phase 3
study. Lancet. 2010;376:2075-2085.
Role of high-dose therapy and autologous stem cell
Harousseau JL, Mathiot C, Attal M, et al. VELCADE/
transplantation (HDT/ASCT)
dexamethasone (Vel/D) versus VAD as induction treatment
prior to autologous stem cell transplantion (ASCT) in newly Attal M, Harousseau JL, Stoppa AM, et al. A prospective,
diagnosed MM (MM): updated results of the IFM 2005/01 trial randomized trial of autologous bone marrow transplantation
[abstract]. Blood. 2007;110(11):450. and chemotherapy in MM. Intergroupe Francais du Myelome.
Jakubowiak AJ, Dytfeld D, Griffith KA, et al. A phase 1/2 study of N Engl J Med. 1996;335:91-97.
carfilzomib in combination with lenalidomide and low-dose Barlogie B, Attal M, Crowley J, et al. Long-term follow-up of
dexamethasone as a frontline treatment for multiple myeloma. autotransplantation trials for multiple myeloma: update of
Blood. 2012;120:1801-1809. protocols conducted by the Intergroupe Francophone du
Kumar S, Flinn I, Richardson PG, et al. Randomized, multicenter, Myelome, Southwest Oncology Group, and University of
phase 2 study (EVOLUTION) of combinations of bortezomib, Arkansas for Medical Sciences. J Clin Oncol. 2010;28:1209-1214.
dexamethasone, cyclophosphamide, and lenalidomide in Barlogie B, Kyle RA, Anderson KC, et al. Standard chemotherapy
previously untreated MM. Blood. 2012;119:4375-4382. compared with high-dose chemoradiotherapy for multiple
Lokhorst HM, van der Holt B, Zweegman S, et al. A randomized myeloma: final results of phase III US Intergroup Trial S9321.
phase 3 study on the effect of thalidomide combined with J Clin Oncol. 2006;24:929-936.
adriamycin, dexamethasone, and high-dose melphalan, followed Bladé J, Rosiñol L, Sureda A, et al. High-dose therapy
by thalidomide maintenance in patients with multiple myeloma. intensification compared with continued standard
Blood. 2010;115:1113-1120. chemotherapy in multiple myeloma patients responding to
Mikhael JR, Schuster SR, Jimenez-Zepeda VH, et al. the initial chemotherapy: long-term results from a prospective
Cyclophosphamide-bortezomib-dexamethasone (CyBorD) randomized trial from the Spanish cooperative group
produces rapid and complete hematologic response in patients PETHEMA. Blood. 2005;106:3755-3759.
with AL amyloidosis. Blood. 2012;119:4391-4394. Child JA, Morgan GJ, Davies FE, et al. High-dose chemotherapy
Rajkumar SV, Jacobus S, Callander NS, et al. Lenalidomide plus with hematopoietic stem-cell rescue for MM. N Engl J Med.
high-dose dexamethasone versus lenalidomide plus low-dose 2003;348:1875-1883.
dexamethasone as initial therapy for newly diagnosed MM: Fermand JP, Katsahian S, Divine M, et al. High-dose therapy and
an open-label randomised controlled trial. Lancet Oncol. autologous blood stem-cell transplantation compared with
2010;11:29-37. conventional treatment in myeloma patients aged 55 to 65 years:
Rosiñol L, Oriol A, Teruel AI, et al. Superiority of bortezomib, long-term results of a randomized control trial from the Group
thalidomide, and dexamethasone (VTD) as induction Myelome-Autogreffe. J Clin Oncol. 2005;23:9227-9233.

Lahuerta JJ, Mateos MV, Martinez-Lopez J, et al. Busulfan 12 mg/ autologous stem-cell transplantation procedure. J Clin Oncol.
kg plus melphalan 140 mg/m2 versus melphalan 200 mg/m2 as 2009;27:1788-1793.
conditioning regimens for autologous transplantation in newly Stewart AK, Trudel S, Bahlis NJ, et al. A randomized phase 3 trial
diagnosed MM patients included in the PETHEMA/GEM2000 of thalidomide and prednisone as maintenance therapy after
study. Haematologica. 2010;95:1913-1920. ASCT in patients with MM with a quality-of-life assessment:
Moreau P, Facon T, Attal M, et al. Comparison of 200 mg/m2 the National Cancer Institute of Canada Clinicals Trials Group
melphalan and 8 Gy total body irradiation plus 140 mg/m2 Myeloma 10 Trial. Blood. 2013;121:1517-1523.
melphalan as conditioning regimens for peripheral blood stem
cell transplantation in patients with newly diagnosed multiple
myeloma: final analysis of the Intergroupe Francophone du
Tandem transplants: double autologous and
Myélome 9502 randomized trial. Blood. 2002;99:731-735. autologous/allogeneic
Roussel M, Moreau P, Huynh A, et al. Bortezomib and high-dose Attal M, Harousseau J-L, Facon T, et al. Single versus double
melphalan as conditioning regimen before autologous stem cell autologous stem-cell transplantation for MM. N Engl J Med.
transplantation in patients with de novo multiple myeloma: 2003;349:2495-2502.
a phase 2 study of the Intergroupe Francophone du Myélome Cavo M, Salwender H, Rosiñol L, et al. Double vs single autologous
(IFM). Blood. 2010;115:32-37. stem cell transplantation after bortezomib-based induction
Segeren CM, Sonneveld P, van der Holt B, et al. Overall and regimens for MM: an integrated analysis of patient-level data from
event-free survival are not improved by the use of myeloablative phase European III studies [abstract]. Blood. 2013;122(21):767.
therapy following intensified chemotherapy in previously Cavo M, Tosi P, Zamagni E, et al. Prospective, randomized
untreated patients with multiple myeloma: a prospective study of single compared with double autologous stem-cell
randomized phase 3 study. Blood. 2003;101:2144-2151. transplantation for multiple myeloma: Bologna 96 clinical
study. J Clin Oncol. 2007;25:2434-2441.
Einsele H, Liebisch P, Bargou R, Meisner C, Metzner B, Wandt H.
Post-HDT consolidation and maintenance
Single high-dose chemoradiotherapy versus tandem high-dose
Attal M, Harousseau JL, Leyvraz S, et al. Maintenance therapy melphalan followed by autologous stem cell transplantation:
with thalidomide improves survival in patients with multiple preliminary analysis [Xth International Myeloma Foundation
myeloma. Blood. 2006;108:3289-3294. Workshop, Sydney]. Haematologica. 2005;90(S1):131.
Attal M, Lauwers-Cances V, Marit G, et al. Lenalidomide Goldschmidt H. Single vs. double high-dose therapy in MM:
maintenance after stem-cell transplantation for MM. N Engl J second analysis of the GMMG-HD2 trial [Xth International
Med. 2012;366:1782-1791. Myeloma Foundation Workshop, Sydney 2005]. Haematologica.
Cavo M, Pantani L, Petrucci MT, et al. Bortezomib-thalidomide- 2005;90(S1):38.
dexamethasone is superior to thalidomide-dexamethasone as Kumar A, Kharfan-Dabaja MA, Glasmacher A, Djulbegovic
consolidation therapy after autologous hematopoietic stem B. Tandem versus single autologous hematopoietic cell
cell transplantation in patients with newly diagnosed multiple transplantation for the treatment of MM: a systematic review
myeloma. Blood. 2013;120:9-19. and meta-analysis. J Natl Cancer Inst. 2009;101:100-106.
Gay F, Caravita T, Cavalli M, et al. Maintenance therapy with Naumann-Winter F, Greb A, Borchmann P, Bohlius J, Engert
lenalidomide significantly improved survival of young newly A, Schnell R. First-line tandem high-dose chemotherapy and
diagnosed MM patients [abstract]. Blood. 2013;122(21):2089. autologous stem cell transplantation versus single high-dose
Ladetto M, Pagliano G, Ferrero S, et al. Major tumor shrinking chemotherapy and autologous stem cell transplantation in MM,
and persistent molecular remissions after consolidation with a systematic review of controlled studies. Cochrane Database
bortezomib, thalidomide, and dexamethasone in patients with Syst Rev. 2012;10:CD004626.
autografted myeloma. J Clin Oncol. 2010;28:2077-2084.
Maiolino A, Hungria VT, Garnica M, et al. Thalidomide plus
Allogeneic transplantation
dexamethasone as a maintenance therapy after autologous
hematopoietic stem cell transplantation improves progression- Bruno B, Rotta M, Patriarca F, et al. A comparison of allografting
free survival in MM. Am J Hematol. 2012;87:948-952. with autografting for newly diagnosed myeloma. N Engl J Med.
McCarthy PL, Owzar K, Hofmeister CC, et al. Lenalidomide 2007;356:1110-1120.
after stem-cell transplantation for MM. N Engl J Med. Costa L, Armeson K, Hill E. Tandem autologous transplantation
2012;366:1770-1781. versus autologous plus reduced-intensity conditioning allogeneic
Sonneveld P, Scheid C, van der Holt B, et al. Bortezomib induction transplantation in the management of newly diagnosed
and maintenance treatment improves survival in patients with MM: meta-analysis of all prospective trials with biological
newly diagnosed MM: extended follow-up of the HOVON-65/ randomisation. Bone Marrow Transplant. 2013;47(S1):S45.
GMMG-HD4 trial [abstract]. Blood. 2013;122:404. Gahrton G, Iacobelli S, Bjorkstrand B, et al. Autologous/reduced-
Spencer A, Prince HM, Roberts AW, et al. Consolidation therapy intensity allogeneic stem cell transplantation vs autologous
with low-dose thalidomide and prednisolone prolongs the transplantation in multiple myeloma: long-term results of the
survival of multiple myeloma patients undergoing a single EBMT-NMAM2000 study. Blood. 2013;121:5055-5063.

Bibliography | 677
Garban F, Attal M, Michallet M, et al. Prospective comparison of and prednisone in elderly patients with untreated MM: a
autologous stem cell transplantation followed by dose-reduced randomised trial. Lancet Oncol. 2010;11:934-941.
allograft (IFM99-03 trial) with tandem autologous stem cell Mateos MV, Richardson PG, Schlag R, et al. Bortezomib plus
transplantation (IFM99-04 trial) in high-risk de novo multiple melphalan and prednisone compared with melphalan and
myeloma. Blood. 2006;107:3474-3480. prednisone in previously untreated MM: updated follow-up and
Kharfan-Dabaja M, M. H, Reljic T, Bensinger W, Djulbegovic impact of subsequent therapy in the phase III VISTA trial. J Clin
B, Kumar A. Comparative effi cacy of tandem autologous- Oncol. 2010;28:2259-2266.
autologous versus tandem autologous-reduced intensity Mateos MV, Oriol A, Martinez-Lopez J, et al. Maintenance
allogeneic haematopoietic cell transplantation in MM: results of therapy with bortezomib plus thalidomide or bortezomib plus
a systematic review and meta-analysis. Bone Marrow Transplant. prednisone in elderly multiple myeloma patients included in the
2013;47(suppl 1):S44. GEM2005MAS65 trial. Blood. 2012;120:2581-2588.
Krishnan A, Pasquini MC, Logan B, et al. Autologous Morgan GJ, Davies FE, Gregory WM, et al. Cyclophosphamide,
haemopoietic stem-cell transplantation followed by allogeneic thalidomide, and dexamethasone (CTD) as initial therapy for
or autologous haemopoietic stem-cell transplantation in patients with multiple myeloma unsuitable for autologous
patients with MM (BMT CTN 0102): a phase 3 biological transplantation. Blood. 2011;118:1231-1238.
assignment trial. Lancet Oncol. 2011;12:1195-1203. Palumbo A, Bringhen S, Caravita T, et al. Oral melphalan and
Lokhorst HM, van der Holt B, Cornelissen JJ, et al. Donor prednisone chemotherapy plus thalidomide compared with
versus no-donor comparison of newly diagnosed myeloma melphalan and prednisone alone in elderly patients with MM:
patients included in the HOVON-50 MM study. Blood. randomised controlled trial. Lancet. 2006;367:825-831.
2012;119:6219-6225. Palumbo A, Bringhen S, Rossi D, et al. Overall survival benefit for
Lokhorst H, van der Holt B, Cornelissen J, et al. No improvement bortezomib-melphalan-prednisone-thalidomide followed by
of overall survival after extended follow-up of donor versus maintenance with bortezomib-thalidomide (VMPT-VT) versus
no donor analysis of newly diagnosed myeloma patients bortezomib-melphalan-prednisone (VMP) in newly diagnosed
included in the HOVON 50/54 study [abstract]. Blood. MM patients [abstract]. Blood. 2012;120(21):200.
2013;122(21):2132. Palumbo A, Hajek R, Delforge M, et al. Continuous lenalidomide
Rosiñol L, Pérez-Simón JA, Sureda A, et al. A prospective treatment for newly diagnosed MM. N Engl J Med.
PETHEMA study of tandem autologous transplantation versus 2012;366:1759-1769.
autograft followed by reduced-intensity conditioning allogeneic Ponisch W, Mitrou PS, Merkle K, et al. Treatment of
transplantation in newly diagnosed multiple myeloma. Blood. bendamustine and prednisone in patients with newly
2008;112:3591-3593. diagnosed MM results in superior complete response rate,
Stem cell transplantation for MM: Executive summary of final prolonged time to treatment failure and improved quality of
report: N05-03C, Version 1.0. 2005-2011 September 19. Epub life compared to treatment with melphalan and prednisone—a
May 1, 2014. randomized phase III study of the East German Study Group
of Hematology and Oncology (OSHO). J Cancer Res Clin
Oncol. 2006;132:205-212.
Treatment of transplant-ineligible patients San Miguel JF, Schlag R, Khuageva NK, et al. Bortezomib plus
Facon T, Mary JY, Hulin C, et al. Melphalan and prednisone melphalan and prednisone for initial treatment of MM. N Engl J
plus thalidomide versus melphalan and prednisone alone or Med. 2008;359:906-917.
reduced-intensity autologous stem cell transplantation in Waage A, Gimsing P, Fayers P, et al. Melphalan and prednisone
elderly patients with MM (IFM 99-06): a randomised trial. plus thalidomide or placebo in elderly patients with MM. Blood.
Lancet. 2007;370:1209-1218. 2010;116:1405-1412.
Fayers PM, Palumbo A, Hulin C, et al. Thalidomide for previously Wijermans P, Schaafsma M, Termorshuizen F, et al. Phase III study
untreated elderly patients with multiple myeloma: meta-analysis of the value of thalidomide added to melphalan plus prednisone
of 1685 individual patient data from 6 randomized clinical in elderly patients with newly diagnosed MM: the HOVON 49
trials. Blood. 2011;118:1239-1247. Study. J Clin Oncol. 2010;28:3160-3166.
Hulin C, Facon T, Rodon P, et al. Efficacy of melphalan and Wildes TM, Rosko A, Tuchman SA. Multiple myeloma in the
prednisone plus thalidomide in patients older than 75 years older adult: better prospects, more challenges. J Clin Oncol.
with newly diagnosed MM: IFM 01/01 trial. J Clin Oncol. 2014;32:2531-2540.
2009;27:3664-3670.
Ludwig H, Hajek R, Tothova E, et al. Thalidomide-dexamethasone
Treatment of relapsed MM
compared with melphalan-prednisolone in elderly patients with
multiple myeloma. Blood. 2009;113:3435-3442. Dimopoulos MA, Weber D, Kantarjian H, Delasalle KB, Alexanian
Mateos MV, Oriol A, Martinez-Lopez J, et al. Bortezomib, R. HyperCVAD for VAD-resistant MM. Am J Hematol.
melphalan, and prednisone versus bortezomib, thalidomide, 1996;52:77-81.
and prednisone as induction therapy followed by maintenance Jakubowiak AJ, Benson DM, Bensinger W, et al. Phase I trial of
treatment with bortezomib and thalidomide versus bortezomib anti-CS1 monoclonal antibody elotuzumab in combination

with bortezomib in the treatment of relapsed/refractory MM. Terpos E, Morgan G, Dimopoulos MA, et al. International
J Clin Oncol. 2012;30:1960-1965. Myeloma Working Group recommendations for the treatment
Palumbo A, Boccadoro M, Bruno B, Triolo S, Pileri A. of MM-related bone disease. J Clin Oncol. 2013;31:2347-2357.
Cyclophosphamide (3.6 g/m2) therapy with G-CSF support for Terpos E, Moulopoulos LA, Dimopoulos MA. Advances in
resistant myeloma. Haematologica. 1994;79:513-518. imaging and the management of myeloma bone disease. J Clin
Richardson PG, Schlossman RL, Alsina M, et al. PANORAMA Oncol. 2011;29:1907-1915.
2: panobinostat in combination with bortezomib and Vesole DH, Oken MM, Heckler C, et al. Oral antibiotic
dexamethasone in patients with relapsed and bortezomib- prophylaxis of early infection in MM: a URCC/ECOG
refractory myeloma. Blood. 2013;122:2331-2337. randomized phase III study. Leukemia. 2012;26:2517-2520.
Richardson PG, Wolf J, Jakubowiak A, et al. Perifosine plus
bortezomib and dexamethasone in patients with relapsed/
refractory MM previously treated with bortezomib: results of a Plasmacytoma
multicenter phase I/II trial. J Clin Oncol. 2011;29:4243-4249.
Dimopoulos MA, Moulopoulos LA, Maniatis A, Alexanian R.
San Miguel J, Weisel K, Moreau P, et al. Pomalidomide plus
Solitary plasmacytoma of bone and asymptomatic MM. Blood.
low-dose dexamethasone versus high-dose dexamethasone
2000;96:2037-2044.
alone for patients with relapsed and refractory MM (MM-
Hill QA, Rawstron AC, de Tute RM, Owen RG. Outcome
003): a randomised, open-label, phase 3 trial. Lancet Oncol.
prediction in plasmacytoma of bone: a risk model utilizing
2013;14:1055-1066.
bone marrow flow cytometry and light-chain analysis. Blood.
Wilson K, Shelley W, Belch A, et al. Weekly cyclophosphamide
2014;124:1296-1299.
and alternate-day prednisone: an effective secondary therapy in
Paiva B, Chandia M, Vidriales MB, Colado E, Caballero-Velazquez
MM. Cancer Treat Rep. 1987;71:981-982.
T, Escalante F, et al. Multiparameter flow cytometry for
staging of solitary bone plasmacytoma: new criteria for risk of
Supportive care progression to myeloma. Blood. 2014;124:1300-1303.
Warsame R, Gertz MA, Lacy MQ, Kyle RA, Buadi F, Dingli D,
Chapel HM, Lee M, Hargreaves R, et al.; The UK Group for
et al. Trends and outcomes of modern staging of solitary
Immunoglobulin Replacement Therapy in MM. Randomised
plasmacytoma of bone. Am J Hematol. 2012;87:647-651.
trial of intravenous immunoglobulin as prophylaxis against
infection in plateau-phase MM. Lancet. 1994;343:1059-1063.
Dimopoulos MA, Kastritis E, Bamia C, et al. Reduction of Nonsecretory MM
osteonecrosis of the jaw (ONJ) after implementation of
preventive measures in patients with MM treated with Chawla SS, Kumar SK, Dispenzieri A, et al. Clinical course and
zoledronic acid. Ann Oncol. 2009;20:117-120. prognosis of non-secretory MM. Eur J Haematol. 2015;95:57-64.
Dispenzieri A, Kyle RA. Neurological aspects of MM and related Coriu D, Weaver K, Schell M, et al. A molecular basis for
disorders. Best Pract Res Clin Haematol. 2005;18:673-688. nonsecretory myeloma. Blood. 2004;104:829-831.
Dudeney S, Lieberman IH, Reinhardt MK, Hussein M. Drayson M, Tang LX, Drew R, Mead GP, Carr-Smith H, Bradwell
Kyphoplasty in the treatment of osteolytic vertebral AR. Serum free light-chain measurements for identifying and
compression fractures as a result of MM. J Clin Oncol. monitoring patients with nonsecretory multiple myeloma.
2002;20:2382–2387. Blood. 2001;97:2900-2902.
Gimsing P, Carlson K, Turesson I, et al. Effect of pamidronate 30 Fonseca R. Non-secretory myeloma: clinical and biologic
mg versus 90 mg on physical function in patients with newly implications. Oncology. 2013;27:932-933.
diagnosed MM (Nordic Myeloma Study Group): a double-blind, Kumar S, Perez WS, Zhang MJ, Ballen K, Bashey A, To LB, et
randomised controlled trial. Lancet Oncol. 2010;11:973-982. al. Comparable outcomes in nonsecretory and secretory MM
Kristinsson SY, Fears TR, Gridley G, et al. Deep vein thrombosis after autologous stem cell transplantation. Biol Blood Marrow
after monoclonal gammopathy of undetermined significance Transplant. 2008;14:1134-1140.
and multiple myeloma. Blood. 2008;112:3582-3586. Lonial S, Kaufman JL. Non-secretory myeloma: a clinician’s guide.
Kyle RA, Yee GC, Somerfield MR, et al. American Society of Oncology. 2013;27:924-928, 930.
Clinical Oncology 2007 clinical practice guideline update on the Rubio-Felix D, Giralt M, Giraldo MP, Martinez-Penuela
role of bisphosphonates in MM. J Clin Oncol. 2007;25: JM, Oyarzabal F, Sala F, et al. Nonsecretory MM. Cancer.
2464-2472. 1987;59:1847-1852.
Morgan GJ, Davies FE, Gregory WM, et al. First-line treatment
with zoledronic acid as compared with clodronic acid in MM
PC leukemia
(MRC Myeloma IX): a randomised controlled trial. Lancet.
2010;376:1989-1999. Avet-Loiseau H, Roussel M, Campion L, Leleu X, Marit G,
Nucci M, Anaissie E. Infections in patients with MM in the Jardel H, et al. Cytogenetic and therapeutic characterization
era of high-dose therapy and novel agents. Clin Infect Dis. of primary PC leukemia: the IFM experience. Leukemia.
2009;49:1211-1225. 2012;26:158-159.

Bibliography | 679
Egan JB, Shi C-X, Tembe W, et al. Whole-genome sequencing for primary amyloidosis: importance of achieving a complete
of multiple myeloma from diagnosis to plasma cell leukemia response. Haematologica. 2007;92:1415-1418.
reveals genomic initiating events, evolution, and clonal tides. Gertz MA, Lacy MQ, Dispenzieri A, et al. Refinement in patient
Blood. 2012;120:1060-1066. selection to reduce treatment-related mortality from autologous
Fernandez de Larrea C, Kyle RA, Durie BG, Ludwig H, Usmani stem cell transplantation in amyloidosis. Bone Marrow
S, Vesole DH, et al. PC leukemia: consensus statement on Transplant. 2013;48:557-561.
diagnostic requirements, response criteria and treatment Gertz M, Lacy M, Dispenzieri A, et al. Troponin T level as an
recommendations by the International Myeloma Working exclusion criterion for stem cell transplantation in light-chain
Group. Leukemia. 2013;27:780-791. amyloidosis. Leuk Lymphoma. 2008;49:36-41.
Gonsalves WI, Rajkumar SV, Go RS, et al. Trends in survival of Gertz MA, Zeldenrust SR. Treatment of immunoglobulin light
patients with primary plasma cell leukemia: a population-based chain amyloidosis. Curr Hematol Malig Rep. 2009;4:91-98.
analysis. Blood. 2014;124:907-912. Jaccard A, Moreau P, Leblond V, et al. High-dose melphalan
Mahindra A, Kalaycio ME, Vela-Ojeda J, Vesole DH, Zhang MJ, versus melphalan plus dexamethasone for AL amyloidosis.
Li P, et al. Hematopoietic cell transplantation for primary PC N Engl J Med. 2007;357:1083-1093.
leukemia: results from the Center for International Blood and Kourelis TV, Kumar SK, Gertz MA, et al. Coexistent multiple
Marrow Transplant Research. Leukemia. 2012;26:1091-1097. myeloma or increased bone marrow PCs define equally high-
Ramsingh G, Mehan P, Luo J, Vij R, Morgensztern D. Primary risk populations in patients with immunoglobulin light chain
PC leukemia: a Surveillance, Epidemiology, and End amyloidosis. J Clin Oncol. 2013;31:4319-4324.
Results database analysis between 1973 and 2004. Cancer. Kumar S, Dispenzieri A, Katzmann JA, et al. Serum
2009;115:5734-5739. immunoglobulin free light-chain measurement in primary
Usmani SZ, Nair B, Qu P, Hansen E, Zhang Q, Petty N, et al. amyloidosis: prognostic value and correlations with clinical
Primary PC leukemia: clinical and laboratory presentation, features. Blood. 2010;116:5126-5129.
gene-expression profiling and clinical outcome with Total Kumar S, Dispenzieri A, Lacy MQ, et al. Revised prognostic
Therapy protocols. Leukemia. 2012;26:2398-2405. staging system for light chain amyloidosis incorporating cardiac
van de Donk NWCJ, Lokhorst HM, Anderson KC, biomarkers and serum free light chain measurements. J Clin
Richardson PG. How I treat plasma cell leukemia. Blood. Oncol. 2012;30:989-995.
2012;120:2376-2389. Kumar SK, Gertz MA, Lacy MQ, et al. Recent improvements in
survival in primary systemic amyloidosis and the importance of
an early mortality risk score. Mayo Clin Proc. 2011;86:12-18.
Light-chain amyloidosis
Kumar SK, Hayman SR, Buadi FK, et al. Lenalidomide,
Dietrich S, Schonland SO, Benner A, et al. Treatment with cyclophosphamide, and dexamethasone (CRd) for light-chain
intravenous melphalan and dexamethasone is not able to amyloidosis: Long-term results from a phase 2 trial. Blood.
overcome the poor prognosis of patients with newly diagnosed 2012;119:4860-4867.
systemic light chain amyloidosis and severe cardiac involvement. Landau H, Hassoun H, Rosenzweig MA, et al. Bortezomib and
Blood. 2010;116:522-528. dexamethasone consolidation following risk-adapted melphalan
Dispenzieri A, Buadi F, Laumann K, et al. Activity of and stem cell transplantation for patients with newly diagnosed
pomalidomide in patients with immunoglobulin light-chain light-chain amyloidosis. Leukemia. 2013;27:823-828.
amyloidosis. Blood. 2012;119:5397-5404. Mhaskar R, Kumar A, Behera M, et al. Role of high-dose
Dispenzieri A, Gertz MA, Kyle RA, et al. Prognostication chemotherapy and autologous hematopoietic cell
of survival using cardiac troponins and N-terminal pro- transplantation in primary systemic amyloidosis: A systematic
brain natriuretic peptide in patients with primary systemic review. Biol Blood Marrow Transplant. 2009;15:893-902.
amyloidosis undergoing peripheral blood stem cell Moreau P, Jaccard A, Benboubker L, et al. Lenalidomide in
transplantation. Blood. 2004;104:1881-1887. combination with melphalan and dexamethasone in patients
Dispenzieri A, Lacy MQ, Rajkumar SV, et al. Poor tolerance to with newly diagnosed AL amyloidosis: a multicenter phase 1/2
high doses of thalidomide in patients with primary systemic dose-escalation study. Blood. 2010;116:4777-4782.
amyloidosis. Amyloid. 2003;10:257-261. Palladini G, Dispenzieri A, Gertz MA, et al. New criteria
Dispenzieri A, Lacy MQ, Zeldenrust SR, et al. The activity of for response to treatment in immunoglobulin light chain
lenalidomide with or without dexamethasone in patients with amyloidosis based on free light chain measurement and
primary systemic amyloidosis. Blood. 2007;109:465-470. cardiac biomarkers: Impact on survival outcomes. J Clin Oncol.
Gertz MA. I don’t know how to treat amyloidosis. Blood. 2012;30:4541-4549.
2010;116:507-508. Palladini G, Perfetti V, Obici L, et al. Association of melphalan
Gertz MA, Lacy MQ, Dispenzieri A, et al. Autologous stem cell and high-dose dexamethasone is effective and well tolerated in
transplant for immunoglobulin light chain amyloidosis: A patients with AL (primary) amyloidosis who are ineligible for
status report. Leuk Lymphoma. 2010;51:2181-2187. stem cell transplantation. Blood. 2004;103:2936-2938.
Gertz MA, Lacy MQ, Dispenzieri A, et al. Effect of hematologic Reece DE, Sanchorawala V, Hegenbart U, et al. Weekly and twice-
response on outcome of patients undergoing transplantation weekly bortezomib in patients with systemic AL amyloidosis:

results of a phase 1 dose-escalation study. Blood. 2009;114: Ghobrial IM, Fonseca R, Greipp PR, et al.; Eastern Cooperative
1489-1497. Oncology Group. Initial immunoglobulin M “flare” after
Sanchorawala V, Patel JM, Sloan JM, et al. Melphalan, rituximab therapy in patients diagnosed with Waldenström
lenalidomide and dexamethasone for the treatment of macroglobulinemia: An Eastern Cooperative Oncology Group
immunoglobulin light chain amyloidosis: results of a phase II Study. Cancer. 2004;101:2593-2598.
trial. Haematologica. 2013;98:789-792. Ghobrial IM, Gertz M, Laplant B, et al. Phase II trial of the oral
Sanchorawala V, Skinner M, Quillen K, et al. Long-term outcome of mammalian target of rapamycin inhibitor everolimus in
patients with AL amyloidosis treated with high-dose melphalan relapsed or refractory Waldenström macroglobulinemia. J Clin
and stem-cell transplantation. Blood. 2007;110:3561-3563. Oncol. 2010;28:1408-1414.
Sanchorawala V, Wright DG, Rosenzweig M, et al. Lenalidomide Ghobrial IM, Hong F, Padmanabhan S, et al. Phase II trial of
and dexamethasone in the treatment of AL amyloidosis: results weekly bortezomib in combination with rituximab in relapsed
of a phase 2 trial. Blood. 2007;109:492-496. or relapsed and refractory Waldenström macroglobulinemia.
Siragusa S, Morice W, Gertz MA, et al. Asymptomatic J Clin Oncol. 2010;28:1422-1428.
immunoglobulin light chain amyloidosis (AL) at the time of Leleu X, Soumerai J, Roccaro A, et al. Increased incidence of
diagnostic bone marrow biopsy in newly diagnosed patients transformation and myelodysplasia/acute leukemia in patients
with MM and smoldering myeloma: a series of 144 cases and a with Waldenström macroglobulinemia treated with nucleoside
review of the literature. Ann Hematol. 2011;90:101-106. analogs. J Clin Oncol. 2009;27:250-255.
Tsai SB, Seldin DC, Quillen K, et al. High-dose melphalan and Owen RG, Treon SP, Al-Katib A, et al. Clinicopathological
stem cell transplantation for patients with AL amyloidosis: definition of Waldenström’s macroglobulinemia: consensus
trends in treatment-related mortality over the past 17 years at a panel recommendations from the Second International
single referral center. Blood. 2012;120:4445-4446. Workshop on Waldenström’s Macroglobulinemia. Semin
Venner CP, Lane T, Foard D, et al. Cyclophosphamide, Oncol. 2003;30:110-115.
bortezomib, and dexamethasone therapy in AL amyloidosis Treon SP. How I treat Waldenström macroglobulinemia. Blood.
is associated with high clonal response rates and prolonged 2009;114:2375-2385.
progression-free survival. Blood. 2012;119:4387-4390. Treon SP, Branagan AR, Ioakimidis L, et al. Long-term outcomes to
Vrana JA, Gamez JD, Madden BJ, et al. Classification of fludarabine and rituximab in Waldenström macroglobulinemia.
amyloidosis by laser microdissection and mass spectrometry- Blood. 2009;113:3673-3678.
based proteomic analysis in clinical biopsy specimens. Blood. Treon SP, Ioakimidis L, Soumerai JD, et al. Primary therapy of
2009;114:4957-4959. WMwith bortezomib, dexamethasone, and rituximab: WMCTG
Wechalekar AD, Goodman HJ, Lachmann HJ, et al. Safety and clinical trial 05-180. J Clin Oncol. 2009;27:3830-3835.
efficacy of risk-adapted cyclophosphamide, thalidomide, Treon SP, Soumerai JD, Branagan AR, et al. Lenalidomide and
and dexamethasone in systemic AL amyloidosis. Blood. rituximab in Waldenström’s macroglobulinemia. Clin Cancer
2007;109:457-464. Res. 2009;15:355-360.
Wechalekar AD, Lachmann HJ, Offer M, et al. Efficacy of Xu L, Hunter ZR, Yang G, et al. MYD88 L265P in Waldenström
bortezomib in systemic AL amyloidosis with relapsed/refractory macroglobulinemia, immunoglobulin M monoclonal
clonal disease. Haematologica. 2008;93:295-298. gammopathy, and other B-cell lymphoproliferative disorders
Wechalekar AD, Schonland SO, Kastritis E, et al. A European using conventional and quantitative allele-specific polymerase
collaborative study of treatment outcomes in 346 patients with chain reaction. Blood. 2013;121:2051-2058.
cardiac stage III AL amyloidosis. Blood. 2013;121:3420-3427.
Weiss BM, Hebreo J, Cordaro DV, et al. Increased serum free light
chains precede the presentation of immunoglobulin light chain POEMS syndrome
amyloidosis. J Clin Oncol. 2014;32:2699-2704. Badros A, Porter N, Zimrin A. Bevacizumab therapy for POEMS
syndrome. Blood. 2005;106:1135.
Dispenzieri A. POEMS Syndrome. Blood Rev. 2007;21:285-299.
Dispenzieri A, Klein CJ, Mauermann ML. Lenalidomide therapy in
Ansell SM, Kyle RA, Reeder CB, et al. Diagnosis and management a patient with POEMS syndrome. Blood. 2007;110:
of Waldenström macroglobulinemia: Mayo stratification of 1075-1076.
macroglobulinemia and risk-adapted therapy (mSMART) Dispenzieri A, Kyle RA, Lacy MQ, et al. POEMS syndrome:
guidelines. Mayo Clin Proc. 2010;85:824-833. Definitions and long-term outcome. Blood. 2003;101:
Gertz MA, Reeder CB, Kyle RA, Ansell SM. Stem cell transplant 2496-2506.
for Waldenström macroglobulinemia: an underutilized Dispenzieri A, Lacy MQ, Hayman SR, et al. Peripheral blood stem
technique. Bone Marrow Transplant. 2012;47:1147-1153. cell transplant for POEMS syndrome is associated with high
Ghobrial IM, Fonseca R, Gertz MA, et al. Prognostic model rates of engraftment syndrome. Eur J Haematol. 2008;80:
for disease-specific and overall mortality in newly diagnosed 397-406.
symptomatic patients with Waldenström macroglobulinaemia. D’Souza A, Lacy M, Gertz M, et al. Long-term outcomes after
Br J Haematol. 2006;133:158-164. autologous stem cell transplantation for patients with POEMS

Bibliography | 681
syndrome (osteosclerotic myeloma): a single-center experience. Li J, Zhang W, Jiao L, et al. Combination of melphalan and
Blood. 2012;120:56-62. dexamethasone for patients with newly diagnosed POEMS
Humeniuk MS, Gertz MA, Lacy MQ, et al. Outcomes of patients syndrome. Blood. 2011;117:6445-6449.
with POEMS syndrome treated initially with radiation. Blood. Tang X, Shi X, Sun A, et al. Successful bortezomib-based
2013;122:66-73. treatment in POEMS syndrome. Eur J Haematol. 2009;83:
Ishikawa O, Nihei Y, Ishikawa H. The skin changes of POEMS 609-610.
syndrome. Br J Dermatol. 1987;117:523-526. Takatsuki K, Sanada I. PC dyscrasia with polyneuropathy and
Jaccard A, Royer B, Bordessoule D, et al. High-dose therapy endocrine disorder: Clinical and laboratory features of 109
and autologous blood stem cell transplantation in POEMS reported cases. Jpn J Clin Oncol. 1983;13:543-555.
syndrome. Blood. 2002;99:3057-3059.

Ash Sap 2016

Enviado por

Dados do documento

Direitos autorais

Formatos disponíveis

Compartilhar este documento

Compartilhar ou incorporar documento

Opções de compartilhamento

Você considera este documento útil?

Este conteúdo é inapropriado?

Direitos autorais:

Formatos disponíveis

Ash Sap 2016

Enviado por

Direitos autorais:

Formatos disponíveis

CHAPTER

BK-ASH-SAP_6E-150511-Chp01.indd 1 4/27/2016 6:35:25 PM

Processed 5’ CAP Poly (A)-3’

BK-ASH-SAP_6E-150511-Chp01.indd 2 4/27/2016 6:35:26 PM

BK-ASH-SAP_6E-150511-Chp01.indd 3 4/27/2016 6:35:26 PM

BK-ASH-SAP_6E-150511-Chp01.indd 4 4/27/2016 6:35:26 PM

(RNA-induced silencing complex), where it is unwound by a ATP

BK-ASH-SAP_6E-150511-Chp01.indd 5 4/27/2016 6:35:27 PM

BK-ASH-SAP_6E-150511-Chp01.indd 6 4/27/2016 6:35:27 PM

A Restriction enzyme digestion

Enzyme Recognition sequence Digestion products Overhang

B Ligation of “sticky ends”

can regenerate the original fragments, but

BK-ASH-SAP_6E-150511-Chp01.indd 7 4/27/2016 6:35:28 PM

treatment with a tyrosine kinase inhibitor. The cycle number

BK-ASH-SAP_6E-150511-Chp01.indd 8 4/27/2016 6:35:28 PM

BK-ASH-SAP_6E-150511-Chp01.indd 9 4/27/2016 6:35:29 PM

BK-ASH-SAP_6E-150511-Chp01.indd 10 4/27/2016 6:35:30 PM

BK-ASH-SAP_6E-150511-Chp01.indd 11 4/27/2016 6:35:30 PM

Biotinylated probes Genomic DNA

Apply magnetic field

Wash and elute

Amplify recovered library fragments and sequence

BK-ASH-SAP_6E-150511-Chp01.indd 12 4/27/2016 6:35:31 PM

BK-ASH-SAP_6E-150511-Chp01.indd 13 4/27/2016 6:35:31 PM

BK-ASH-SAP_6E-150511-Chp01.indd 14 4/27/2016 6:35:31 PM

The power of molecular biology to provide important

BK-ASH-SAP_6E-150511-Chp01.indd 15 4/27/2016 6:35:31 PM

6 VH-FR1 primers JH primer

Denaturation (94°C)/renaturation (4°C)

BK-ASH-SAP_6E-150511-Chp01.indd 16 4/27/2016 6:35:32 PM

BK-ASH-SAP_6E-150511-Chp01.indd 17 4/27/2016 6:35:32 PM

BK-ASH-SAP_6E-150511-Chp01.indd 18 4/27/2016 6:35:32 PM

Glossary complementary Sequence of the second strand of DNA

BK-ASH-SAP_6E-150511-Chp01.indd 19 4/27/2016 6:35:32 PM

BK-ASH-SAP_6E-150511-Chp01.indd 20 4/27/2016 6:35:32 PM

BK-ASH-SAP_6E-150511-Chp01.indd 21 4/27/2016 6:35:32 PM

BK-ASH-SAP_6E-150511-Chp01.indd 22 4/27/2016 6:35:32 PM

BK-ASH-SAP_6E-150511-Chp02.indd 23 4/28/2016 12:43:38 PM

Table 2-1 Principles of effective consultation and interphysician communication

BK-ASH-SAP_6E-150511-Chp02.indd 24 4/28/2016 12:43:38 PM

Stop warfarin 5 days before surgery

Low risk Moderate risk High risk

No bridging Consider bridging Bridging anticoagulation

BK-ASH-SAP_6E-150511-Chp02.indd 25 4/28/2016 12:43:39 PM

Table 2-2 Perioperative cessation and resumption of new oral anticoagulants

Type of NOAC* Dabigatran Rivaroxaban/apixaban

Bleeding risk of intervention Low High Low High Edoxaban

BK-ASH-SAP_6E-150511-Chp02.indd 26 4/28/2016 12:43:39 PM

Management of perioperative hemorrhage surgical reexploration should be considered. Some systemic

BK-ASH-SAP_6E-150511-Chp02.indd 27 4/28/2016 12:43:39 PM

BK-ASH-SAP_6E-150511-Chp02.indd 28 4/28/2016 12:43:39 PM

Adapted from ACCP guidelines.

BK-ASH-SAP_6E-150511-Chp02.indd 29 4/28/2016 12:43:39 PM

thrombocytopenia [ITP] may be caused by both platelet

Common inpatient consultations

BK-ASH-SAP_6E-150511-Chp02.indd 30 4/28/2016 12:43:39 PM

• Drug-induced immune thrombocytopenia (DITP)

• Platelet clumping: pseudothrombocytopenia

• Postoperative patient: dilutional, HIT

• <20 × 109/L: Primary ITP (including DITP), TTP/HUS/DIC

Determine the timing of • Within 5–10 days: HIT or DITP; PTP

• Bleeding signs present: ITP, DITP, PTP

BK-ASH-SAP_6E-150511-Chp02.indd 31 4/28/2016 12:43:40 PM

erson-years of OC use. This 2- to 3-fold increased risk is