Você está na página 1de 185

Rheum Dis Clin N Am 33 (2007) xi–xiii


Sailing the Vasculitis Sea

John H. Stone, MD, MPH

Guest Editor

On the topic of learning medicine, Sir William Osler penned a two-part

quote, the first half of which reads as follows: ‘‘To study the phenomena of
disease without books is to sail an uncharted sea. . .’’ Osler would have agreed
that learning how to take care of patients who have systemic vasculitis re-
quires reading (and re-reading) works written by expert cliniciansddoctors
who have developed their skills through experience with many patients and
many complications of both the diseases and their therapies. Harnessing
such clinical expertise into one issue is no trifling task. But for me, recruiting
authors for the articles in this issue of Rheumatic Disease Clinics of North
America and editing their work has been a labor of love.
Although recent years have seen the beginning of evidence-based medicine
for several forms of systemic vasculitis, much of the knowledge base related
to these disorders remains rooted on Pearls and Mythsdclinic nuggets of
wisdom and common fallacies that are important to identify and understand
in order to provide good care for patients. This issue collects a large number
of Pearls and Myths in the first article, entitled ‘‘Vasculitis: A Collection of
Pearls and Myths.’’
The rational management of vasculitis patients requires that rheumatol-
ogists collaborate closely with colleagues in other disciplines. The other

0889-857X/07/$ - see front matter Ó 2007 Elsevier Inc. All rights reserved.
doi:10.1016/j.rdc.2007.09.003 rheumatic.theclinics.com

articles in this issue of Rheumatic Disease Clinics of North America reflect

this fact abundantly:
 Dr. Ulrich Specks, a pulmonologist, serves as the senior author on two
articles: ‘‘B cell depletion in the treatment of ANCA-associated vasculi-
tis’’ and ‘‘Large airway disease in Wegener’s granulomatosis: Diagnosis
and management.’’ Dr. Specks has recruited colleagues in otolaryngol-
ogy for the second of these articles. Understanding both of these articles
will be essential to state-of-the-art management of Wegener’s patients,
both now and over the next decade.
 Dr. Jeffrey Callen, who has long been one of the premier medical derma-
tologists in the field, serves as the senior author of the article on ‘‘Pyo-
derma gangrenosum,’’ a clinical entity that poses a perpetual challenge
for rheumatologists. The illustrations of pyoderma gangrenosum pro-
vided in this article are the best I have ever seen.
 Dr. Maria Cid is an internist whose career focus has been devoted to
caring for patients with giant cell arteritis (GCA) and to understanding
this disease in the laboratory. She and her colleagues have written an
insightful description of ‘‘Five clinical conundrums’’ related to GCA.
The article is earmarked by their expertise with this disorder.
 Dr. Jennifer Thorne, an ophthalmologist, comes from a long line of
outstanding clinicians at Johns Hopkins who focus on ocular immunol-
ogy. In this issue of Rheumatic Disease Clinics, she is the senior author on
an article devoted to ‘‘Scleritis and peripheral ulcerative keratitis,’’ two
potential ophthalmologic emergencies in rheumatology.
In addition to these contributions from subspecialty colleagues, several
leading clinicians and investigators in rheumatology have contributed to this
issue. Dr. Gary Hoffman is, in many ways, the father of clinical vasculitis
investigation in the United States. Dr. Eric Matteson is a scholar of both
vasculitis history and state-of-the-art vasculitis management. Drs. Hoffman
and Matteson serve as senior authors on articles related to ‘‘Takayasu’s ar-
teritis’’ and ‘‘Cogan’s syndrome,’’ respectively. Finally, Dr. Carlo Salvarani
and colleagues have written a magnificent article on ‘‘Retroperitoneal fibro-
sis,’’ informing me (and, I suspect, you) about a number of new thoughts on
this perplexing illness.
In addition to the depth of experience of their authors, the articles in this
issue of Rheumatic Disease Clinics of North America are linked by one other
major strength: color figures. For the first time, the Vasculitis issue is graced
by color photography, which enhances substantially the learning to be
derived from these pages.
I deeply appreciate all of the hard work that my coauthors contributed to
this issue, and am also grateful to our publishers at Elsevier, Rachel Glover
and Barton Dudlick, who kindly granted the occasional eleventh extension
and oversaw the final production of the issue.

The end of this preface must conclude with Osler’s full quote: ‘‘To study
the phenomena of disease without books is to sail an uncharted sea, while to
study books without patients is not to go to sea at all.’’
As you head off to sea in taking on the care of patients with vasculitis, we
hope that these articles serve as useful maps.

John H. Stone, MD, MPH

Rheumatic Disease Unit
Massachusetts General Hospital
Boston, MA 02115, USA
E-mail address: jhstone@partners.org
Rheum Dis Clin N Am 33 (2007) 691–739

Vasculitis: A Collection of Pearls

and Myths
John H. Stone, MD, MPH
Rheumatology Unit, Massachusetts General Hospital, 55 Fruit Street, Boston,
MA 02114, USA

In 1866, Kussmaul and Maier reported the first case of systemic vasculitis
(periarteritis nodosa) in an unfortunate 27-year-old tailor’s apprentice
(Fig. 1) [1,2]. In an earlier brief communication, Kussmaul and Maier [3]
had also promulgated the first vasculitis myth: that the malady in question
was caused by a worm (trichinae) infestation, an ‘‘aneurysma verminosum
hominis’’ (human worm aneurysm).
Describing the postmortem examination, Kussmaul and Maier wrote of
encountering ‘‘.numerous objects deceivingly similar to undeveloped filar-
iae in the microscopic sections that were removed from the flesh, blood, and
intestinal mucus. We felt justified in regarding these as worms after numer-
ous experts had taken them to be such.’’ [2]. Kussmaul and Maier [1] re-
tracted this initial assertion several months later in their full report of
periarteritis nodosa.
Since the two publications of Kussmaul and Maier on this index case of
vasculitis, the field has seen many changes. What was considered for decades
to be only a single disorder is now known to consist of at least 15 to 20 dis-
tinct conditions. Important strides have been made in unraveling the path-
ophysiology of some individual forms of vasculitis, but many mysteries
remain. Over time, numerous new myths and occasional pearls have arisen
from the care of patients with these disorders.
In this collection of pearls and myths, I have gathered lessons about the
status of clinical care of vasculitis patients in the year 2008. The work is an
extension of a previous publication related to vasculitis pearls and myths
that appeared in these pages in 2001 [4]. I have tried not to duplicate the
lessons outlined there, with exceptions only for points that deserved to be
re-emphasized. This article is also not an attempt to be comprehensive,
and focuses on only a few of the major forms of vasculitis.

E-mail address: jhstone@partners.org

0889-857X/07/$ - see front matter Ó 2007 Elsevier Inc. All rights reserved.
doi:10.1016/j.rdc.2007.09.002 rheumatic.theclinics.com

Fig. 1. Title of the report of periarteritis nodosa by Kussmaul and Maier in the Deutsches
Archiv für Klinische Medizin [3], with translation. (From Kussmaul A, Maier R. Aneurysma ver-
minosum hominis: verlaufige nachricht. Deutsches Archiv fur klinische Medizin 1866;1:125.)

Vasculitis therapies
Every attempt should be made to discontinue prednisone in patients who
have systemic vasculitis.

Despite advances in the treatment of systemic vasculitis, several forms are
likely to recur after stopping conventional therapies. For example, up to
80% of patients who have Wegener granulomatosis (WG) eventually suffer
disease flares, even if their disease remissions were induced by prolonged
courses of cyclophosphamide (CYC) and glucocorticoids [5–7].
Clinicians are usually successful in suppressing the inflammatory pro-
cess underlying vasculitis, but seldom fortunate enough to abolish it en-
tirely. Differences in the flare rates in three major clinical trials involving
patients who had antineutrophil cytoplasmic antibody (ANCA)–associated
vasculitis are explained primarily by whether or not glucocorticoids were
maintained throughout the trial or discontinued entirely [8]. In most pa-
tients, attempts to discontinue prednisone after an appropriate length of
time are prudent, because a minority maintain long-term remissions off
glucocorticoids and other medications. For patients who declare their
phenotype to be one of recurrent disease flares, however, long-term,
low-dose prednisone (!5 mg/d or !10 mg/d) may be an appropriate ther-
apeutic regimen until more rational, mechanism-based therapies are

Pneumocystis carinii (jiroveci) prophylaxis can be stopped once the pa-
tient is on low doses of prednisone and off CYC.

One of the greatest errors in using CYC and prednisone is the failure to
use Pneumocystis carinii (jiroveci) pneumonia (PCP) prophylaxis (Fig. 2). In
contrast to the situation with HIV care, there are no firm guidelines for the
use of PCP prophylaxis among patients who have vasculitis. One rule of
thumb in rheumatic disease is to use PCP prophylaxis in any patient treated
with the combination of high-dose glucocorticoids and any other immuno-
suppressive agent.
Some PCP experts endorse prophylaxis in patients treated with more
than 20 mg of prednisone for more than 2 weeks, regardless of the use of
other immunosuppressive medications [9]. In practice, this suggestion is sel-
dom followed, even for patients who might be at higher risk for infection
(eg, because of older age). For example, PCP prophylaxis is not part of
the routine care of patients who have giant cell arteritis (GCA), nor should
it be, in the author’s view.
Many clinicians do not realize that the effects of CYC may persist for
many months after the medication is stopped, even if no additional agent
is used in its place [10]. Before stopping PCP prophylaxis, therefore, clini-
cians should ensure that patients are no longer profoundly lymphopenic.
CD4þ lymphocyte counts that are consistently greater than 250/mL should
signal that the discontinuation of PCP prophylaxis is safe. There are no
evidence-based guidelines on this point, however.

The dose of CYC should be adjusted in the setting of renal dysfunction.

Fig. 2. Pneumocystis carinii (jiroveci), identified in bronchoalveolar lavage fluid from a patient
who had Wegener granulomatosis.

The dose of CYC must be adjusted in renal dysfunction [11]. When oral
CYC is used, one rule of thumb is to decrease the dose by 50% for patients
who have end-stage renal disease or are elderly (eg, age O75 years). For in-
travenous CYC, generally administered according to the National Institutes
of Health lupus nephritis regimen, the recommended approach is to use 500
mg/m2 once monthly rather than 750 mg/m2 for patients who have ad-
vanced renal dysfunction.
Regardless of the precise reduction that is used, the critical issue is to fol-
low the patient’s white blood cell counts regularly (at least once every 2
weeks for patients on daily CYC) and to make dose adjustments as needed
to keep the total leukocyte count greater than 4000/mL. For intermittent
CYC, prolonged leukocyte nadirs should be avoided. Because the therapeu-
tic index for CYC is so narrow, aiming to achieve a modest degree of leuko-
penia as a treatment endpoint is not appropriate. Moreover, it is not

On days when a patient is dialyzed, a post-dialysis supplementation of
CYC should be administered.

Daily CYC is usually administered in the morning, so that the patient
can drink fluids liberally throughout the day and wash CYC metabolites
out of the bladder before going to sleep at night. CYC is partly removed
by dialysis. On days when a patient is dialyzed, a post-dialysis supple-
mentation of half the usual daily dose should be administered afterward

Previous use of CYC makes patients more susceptible to myelosuppres-
sion with other agents.

The long-held notion that CYC should be used for 1 full year after the
patient has achieved disease remission has been discarded. The standard
of care for remission induction now in diseases that require CYC is to use
this medication for 3 to 6 months, and then to replace CYC with either aza-
thioprine or methotrexate [12–14]. Despite shorter courses of CYC therapy,
the bone marrow does not forget its sight of CYC. A patient who has had
neutropenia on CYC is more likely to experience this complication with sub-
sequent immunosuppressive agents, too. The white blood cell count may
drop sharply when the patient switches from CYC to azathioprine (AZA).
Vigilance is required.

Life-threatening bone marrow toxicity associated with AZA can be pre-
dicted by either a genotyping or functional assay.

Many of the toxicities of AZA, particularly the bone marrow toxicity, are
predicted by the number of alleles for the enzyme thiopurine methyltransfer-
ase (TPMT) that one inherits [15]. Approximately 90% of individuals have
two TPMT alleles and usually tolerate AZA well from the standpoint of he-
matologic toxicity. Approximately 9% of individuals are heterozygous for
the TPMT allele and can tolerate AZA, but should be treated with lower
doses. The general approach to using AZA in patients who are TPMT
heterozygotes is to begin with 50 mg/d and increase the dose slowly, never
exceeding 150 mg/d (and usually not using more than 100 mg/d). For homo-
zygous TPMT-negative patients, AZA should not be used under any cir-
cumstances. Life-threatening bone marrow toxicity can ensue.

TPMT status does not explain all potential AZA toxicities.

Some AZA toxicities do not correlate with TPMT status. The most strik-
ing example is the AZA hypersensitivity reaction, a life-threatening compli-
cation of the medication that is not mediated by TPMT deficiency [16].
Gastrointestinal (GI) toxicity is also not predicted by TPMT status. A pa-
tient may have two TPMT alleles that function fully yet experience GI intol-
erance, usually precluding future use of the medication. Finally, the presence
of two TPMT alleles does not guarantee freedom from myelosuppression
with AZA use.

Monitor the renal function of patients on methotrexate (MTX) closely,
even if their disease does not involve the kidneys at presentation.

Any decline in renal function, be it from vasculitis or not, should trigger
a reconsideration of the MTX dose or a discontinuation of the medication
entirely. One cutoff for MTX use has been a serum creatinine of greater than
2.0 mg/dL [17]. Clinicians should remember that serum creatinine levels cor-
relate poorly with the glomerular filtration rate. In an elderly patient or one
who has small muscle mass, a serum creatinine of 1.7 mg/dL may signal
a substantial decline in renal function and preclude the use of MTX.

Plasma exchange
Plasma exchange should be instituted for patients who have systemic vas-
culitis associated with antineutrophil cytoplasmic antibodies (ANCA) who
present with pulmonary-renal syndromes.

There are two problems related to the routine use of plasma exchange in
ANCA-associated vasculitis. First, the link between ANCA titer and disease
activity is significantly weaker than in anti–glomerular basement membrane
(anti-GBM) disease, a disorder in which plasma exchange is the standard of
care (see later discussion of WG). Second, plasma exchange has potential
adverse effects, particularly the risk for infection resulting from a large-
bore, indwelling catheter.
There are three current indications for considering plasma exchange in
ANCA-associated vasculitis: (1) Patients who have anti-GBM antibodies
and ANCA, (2) alveolar hemorrhage that does not respond promptly to
the combination of CYC and high-dose glucocorticoids, and (3) end-stage
or near–end-stage renal disease at presentation.

Trimethoprim-sulfamethoxazole (TMP-SMX) is an effective therapy for

There is one clear use for TMP/SMX in WG: to prevent PCP (see earlier
discussion). Although this myth has been kicking around for 2 decades or
more, there remains no firm evidence that TMP/SMX is effective for the
treatment of active WG. There is also moderate evidence that TMP/SMX
might be effective in the prevention of upper respiratory tract disease flares
[18]. A randomized, double-blind trial from the Netherlands demonstrated
a reduction in upper respiratory tract flares compared with placebo.
Several points from that study are worth emphasizing:
 The dose of TMP/SMX used was one double-strength tablet twice daily.
This dose is significantly higher than the dose used for PCP prophylaxis.
 There was no reduction in severe disease flares.
 Even in the active prophylaxis group, 60% of patients suffered disease
flares (compared with 82% in the placebo group).
In short, double-strength TPM/SMX twice daily might be worth a try in
patients who have WG with histories of upper respiratory tract disease
flares. But excessive confidence in the efficacy of this intervention seems

unwarranted. Remember also that TMP/SMX has potentially serious side

effects: interstitial nephritis and the Stevens-Johnson syndrome, to name
two. Twenty percent of the patients in the TMP-SMX arm had to discon-
tinue the treatment because of adverse effects [18].

Polyarteritis nodosa
Classic polyarteritis nodosa (PAN) is often a curable illness.

In contrast to WG, GCA, and some other forms of vasculitis, classic
PANdthe disorder described by Kussmaul and Maier [1]dis usually a cur-
able disease. Approximately 50% of PAN cases can be cured with glucocor-
ticoid treatment alone, usually in courses lasting a total of 9 to 12 months
[19]. The remainder of patients who have PAN require CYC to induce dis-
ease remission. Once classic PAN is in remission, however, it usually stays
there without requiring long-term maintenance therapy.
The propensity of classic PAN to be cured was illustrated by a study from
the French Vasculitis Study Group that evaluated the outcomes of 351 pa-
tients who had classic PAN [20]. More than two thirds of the patients were
cured by their initial courses of therapy.

When attempting to confirm the diagnosis of vasculitis by peripheral
nerve biopsy, target an affected nerve but not a dead one.

Several forms of systemic vasculitis are associated frequently with vascu-
litic neuropathy. In addition to PAN, these are the Churg-Strauss syn-
drome, microscopic polyangiitis, WG, and cryoglobulinemic vasculitis
(Fig. 3).
Which nerve offers the highest yield for biopsy in vasculitic neuropathy?
Biopsies of peripheral nerves that are electrophysiologically normal are
rarely useful, because such nerves are histologically normal, too. Under
ideal circumstances, the surgeon biopsies a nerve that is affected but not
dead. If the sural nerves demonstrate no action potentials, the superficial pe-
roneal nerve and peroneus brevis muscle are other possible targets [21]. In
addition, upper extremity nerves (eg, the lateral antebrachial cutaneous
nerve or other nerves in the upper extremities) are also candidates for bi-
opsy, with simultaneous sampling of the underlying muscle.
Even when a given peripheral nerve is demonstrated to be abnormal by
nerve conduction studies, the likelihood of confirming the diagnosis by bi-
opsy of the nerve is on the order of 50%. Nevertheless, a peripheral nerve
biopsy usually provides useful information, as illustrated by a study of

Fig. 3. Muscle wasting caused by nerve infarction in vasculitic neuropathy. Wasting of the
interosseus muscles between the thumb and first finger in a patient who had microscopic

93 cases of vasculitic neuropathy [22]. The following observations were

 Eighty-four peripheral nerve biopsies were performed (82 sural, 2 super-
ficial radial).
 Forty-nine biopsies (58%) confirmed the diagnosis of vasculitic
 In an additional 31 cases, the histopathologic results were consistent
with but not diagnostic of vasculitic neuropathy, demonstrating asym-
metric acute axonal loss and differential fascicular involvement with
or without active Wallerian degeneration or regenerating fibers.
In summary, 95% of nerve biopsies contributed information important
to the diagnosis of vasculitic neuropathy. In the absence of another clear
site to biopsy, it is usually worth biopsying a peripheral nerve demonstrated
to be abnormal by nerve conduction velocities (NCVs) along with the adja-
cent skeletal muscle. In many cases, the diagnosis can be inferred from
a highly consistent clinical history, NCVs that reveal an asymmetric,
length-dependent process, and compatible (even if not diagnostic) histo-
pathologic findings.

Examination of the flexor hallucis brevis muscle with electromyography
(EMG) and NCVs is important in the evaluation of length-dependent

Length-dependent axonopathies are those diseases of the nerves that af-
fect first the longest nerves in the body, usually those of the feet. Vasculitic
neuropathy is typically a length-dependent axonopathy. EMG/NCVs of the

flexor hallucis brevis can establish the presence of motor involvement that is
not otherwise apparent. This finding may alter the differential diagnosis sub-
stantially. Fibrillation potentials of the flexor hallucis brevis muscle indicate
subclinical involvement of motor fibers, thereby pointing to a sensorimotor
multiple mononeuropathy if sensory nerve action potentials are also absent
in the lower extremities [23].
When evaluating a patient for vasculitic neuropathy, the electromyogra-
phy must perform EMGs of the flexor hallucis brevis.

Vasculitic neuropathy typically behaves like the rest of PAN.

Vasculitic neuropathy typically responds well to treatment and does not
return in most cases. In the series described above [22], disease relapse oc-
curred in only one of 54 patients treated with CYC. Seven of the nine deaths
in the overall series occurred in patients who were treated with CYC. This
finding underscores the need to use CYC daily.
The management of the long-term sequelae of nerve infarctions (painful
sensory neuropathies and residual muscle weakness) poses major challenges,
regardless of the form of vasculitis underlying the nerve dysfunction. Pain
control and physical and occupational therapy modalities assume greater
priority than immunosuppression once remission has been induced.

PAN is sometimes associated with ANCA.

This myth is a holdover from the early days of ANCA testing, when re-
liable enzyme immunoassays for antibodies to proteinase-3 (PR3) and mye-
loperoxidase (MPO) were not available. Classic PAN is an ANCA-negative
disease, meaning that it is not associated with either PR3- or MPO-ANCA.
Many types of disorders can be associated with ANCA as demonstrated
by immunofluorescence testing. Most of these cases are associated with an-
tibodies directed not against PR3 or MPO but rather against antigens, such
as lactoferrin, bactericidal permeability inhibitor, or others [24]. If enzyme
immunoassays for PR3- and MPO-ANCA are negative, then positive immu-
nofluorescence assays are usually false-positive results, at least with regard
to diagnoses of systemic vasculitis.

Punch biopsies of the skin are an easy way to make the diagnosis of PAN.

Classic PAN involves muscular, medium-sized arterioles and arteries.
The cutaneous location of such blood vessels is in the deep dermis or

subcutaneous fat (Fig. 4). No punch biopsy extends deep enough to reach
the culprit vessels and confirm the diagnosis of PAN [25].
In contrast, if the cutaneous lesion is one of a small-vessel vasculitis (eg,
palpable or nonpalpable purpura), a simple punch biopsy should suffice for
diagnostic purposes. Immunofluorescence studies in addition to the stan-
dard hematoxylin and eosin (H & E) stains are extremely helpful on biopsies
of skin lesions that suggest small-vessel vasculitis [25].

Among the cutaneous lesions of medium-vessel vasculitis, nodular lesions
are most likely to yield a diagnosis of vasculitis.

Medium-vessel vasculitis is associated with four primary skin lesions:
nodules, ulcers, livedo reticularis, and digital ischemia. Nodules are the le-
sion most likely to yield a diagnosis of vasculitis and should be biopsied
in their center, deep enough to capture subcutaneous fat in the sample. Un-
fortunately, nodules in cutaneous vasculitis are usually short-lived; they
quickly break down to form ulcers. The best place to biopsy an ulcer is
not in the center but at the edge. Again, a biopsy of sufficient depth must
be performed.
Compared with nodules or ulcers, livedo reticularis is associated with
a significantly lower likelihood of yielding a diagnosis. In 19 of 26 patients
who had livedo reticularis from whom two biopsy specimens were available,
no obvious pathologic change was detected in either biopsy despite careful
examination of multiple serial sections [26,27]. When biopsies of livedo re-
ticularis are attempted, the recommendation is to take more than one
deep punch biopsy (4 mm) from different areas of both the white and red
areas. The sensitivity of three biopsies taken from white areas has been es-
timated to be as high as 80% [28].

Fig. 4. Polyarteritis nodosa involving the skin, a vasculitis of medium-sized muscular arteries
located in the deep dermis and subcutaneous fat. This figure shows chronic inflammation in
an artery located at the junction of the deep dermis and subcutaneous fat.

Lesions of digital ischemia are rarely biopsied because of their low yield
and concerns about wound healing.

‘‘Nodular vasculitis’’ is a confusing term that means medium-vessel

The subject of cutaneous vasculitis is difficult enough by itself without the
use of nonspecific nomenclature. Nodular vasculitis usually refers to biopsy-
proven vasculitis of medium-sized muscular arteries and arterioles that
causes a nodular skin lesion. Many forms of systemic vasculitis (eg, PAN,
the ANCA-associated disorders, and mixed cryoglobulinemia) can be asso-
ciated with nodular cutaneous findings. Nodular vasculitis is thus a syn-
drome, not a specific diagnosis. This term should be avoided.

Classic PAN involvement of the skin can be distinguished from impor-
tant mimickers by its clinical appearance.

Diagnosing skin ulcers without an attempt by histopathology to confirm
one’s diagnostic impression is fraught with hazard. For example, cholesterol
emboli (Fig. 5A) and calciphylaxis lesions (see Fig. 5B) can mimic PAN
closely. The common objection to biopsying a skin ulcer is concern about
wound healing, but diagnostic confirmation and the ability to use the correct
therapy usually justify that risk, even if poor wound healing or extension of
the lesion through pathergy [29] are possible concerns.

The proper name for the lacy, purplish discoloration illustrated in
Fig. 6A is livedo reticularis.

Purists note correctly the consistent sloppiness in the literature regarding
the terms livedo reticularis and livedo racemosa. The literature is remark-
ably muddy here and at this point it is probably not worth trying to change.
Livedo racemosa, named by Ehrmann in 1907 [30], is a more striking cuta-
neous finding than livedo reticularis (see Fig. 6B). Ehrmann described livedo
racemosa as a ‘‘tendril-like bluish pattern reminiscent of forked lightening’’
Livedo racemosa differs from livedo reticularis in the following ways:
 Livedo racemosa is more generalized. Trunk or buttocks involvement is
present in nearly all patients. Even the face can be involved in some

Fig. 5. Mimickers of polyarteritis nodosa (PAN) that can only be differentiated from PAN by
histopathology. (A) Large ulcer over the medial malleolus caused by cholesterol embolization.
(Courtesy of A. Golding, MD, PhD, Baltimore, MD, and B. Ehst, MD.) (B) Ulcer over the up-
per lateral thigh caused by calciphylaxis.

 Livedo racemosa is nearly always associated with an underlying

 In contrast, livedo reticularis often occurs in physiologic settings (eg,
a response to cold) rather than in the setting of a disease.

Fig. 6. Livedo racemosa (more commonly referred to as livedo reticularis). (A) Lacy pattern of
discoloration over the lower extremities that does not blanch with pressure in a patient who has
polyarteritis nodosa. (B) Livedo racemosa, ‘‘a tendril-like pattern reminiscent of forked light-
ning’’ [30].

Disorders associated with livedo racemosa are not invariably vasculitic in

nature. For example, in the antiphospholipid syndrome, livedo racemosa is
frequently associated with cerebrovascular events, arterial thrombosis, and
pregnancy morbidity [31]. Other causes of livedo racemosa include essential
thrombocythemia, polycythemia vera, pernicious anemia, disseminated in-
travascular coagulation, and cryofibrinogenemia.

If leukopenia and thrombocytopenia are important components of the
clinical presentation, a primary form of vasculitis is unlikely.

Leukopenia and thrombocytopenia are not usually associated with pri-
mary vasculitic syndromes. Their presence suggests other disorders, such
as a connective tissue disorder in the spectrum of systemic lupus erythema-
tosus or a malignancy.

Giant cell arteritis

Leukocytosis does not occur in GCA before the institution of glucocor-
ticoid treatment.

Despite the often dramatic other indications of systemic inflammation,
the white blood cell count is usually normal before the patient starts pred-
nisone. Nonspecific laboratory abnormalities that can accompany GCA in-
clude a normochromic normocytic anemia, thrombocytosis, polyclonal
hypergammaglobulinemia, and hypoalbuminemia.

GCA does not occur among patients younger than 50 years of age.

Biologic and chronologic age often correlate poorly with each other, but
the statement that GCA does not occur in patients younger than 50 is re-
markably robust. Of the 1435 patients who had GCA with biopsy-proven dis-
ease included in a meta-analysis of 26 studies [32], only 2 were younger than
50. Even those 2 patients, one suspects, may not have had true GCA but
rather some other form of systemic vasculitis known to involve the temporal
arteries and mimic GCA (eg, an ANCA-associated disorder; see WG) [33].

Aspirin is an important adjunct to glucocorticoids in the therapy of

As detailed elsewhere in this issue (see Cid and colleagues, ‘‘Five Clinical
Conundrums in the Management of Giant Cell Arteritis’’), the results of
clinical trials designed to identify an effective steroid-sparing agent in
GCA have been disappointing. Two retrospective cohort studies demon-
strated that a baby aspirin along with prednisone is more effective at pre-
venting visual loss and cerebral ischemic events than is prednisone alone
[34,35]. The consistency of these two studies and the strong reduction in
the odds ratio for clinical events by aspirin makes this medication an impor-
tant adjunct to glucocorticoids. A daily baby aspirin should be part of the
standard therapy for GCA [36].

The prompt use of pulse glucocorticoids may restore vision lost in GCA.

This myth, unfortunately, is wishful thinking. Vision loss in GCA is
usually permanent. A carefully performed study that examined visual out-
comes in patients who had GCA who had lost vision concluded that essen-
tially all improvement in patients’ ability to read eye charts was related to
their accommodation of new visual difficulty rather than to qualitative
improvement in ocular function [37]. Pulse glucocorticoids may be appro-
priate in GCA-related visual loss, however, to preserve vision in the other

GCA may be associated with reversible visual hallucinations.

Anterior ischemic optic neuropathy (AION) is the most common ocular
manifestation of GCA, but far from the only one [38]. Various other ocular
lesions can occur, depending on which specific arteries are involved severely
enough to cause critical ischemia. Central retinal artery occlusion, posterior
ischemic optic neuropathy, and diplopia are other common eye complica-
tions of GCA.
The most intriguing ocular finding in GCA is the Charles Bonnet syn-
drome (pronounced B o-0 n
a). The Charles Bonnet syndrome is the occur-
rence of visual hallucinationsd‘‘release hallucinations’’din individuals
who are psychologically normal [39]. The precise mechanism of these hallu-
cinations is not defined, but the Charles Bonnet syndrome is believed to be
a consequence of ischemia to parts of the visual pathways. Reversibility of
these symptoms with glucocorticoid therapy has been reported.

GCA may present as noises in the head.

This disease feature is rarely mentioned in the literature, and its cause is
obscure. Patients have reported diverse sounds seeming to come from within
their heads, however. Examples of such sounds are turbine engines and tree

GCA does not affect the central nervous system (CNS).

If this myth were reworded to read that ‘‘GCA seldom involves the intra-
cerebral circulation yet can still affect the brain through causing extensive
extracranial disease,’’ then this statement would be correct. When transient
ischemic attacks, vertigo, hearing loss, and stroke occur in GCA, they are
usually the result of vertebral artery or internal carotid artery lesions that
are extradural.
The rarity of intracerebral disease in GCA has been a mystery for de-
cades. One proffered explanation (but almost certainly not the complete an-
swer) is that the extracranial arteries lose their internal elastic laminae,
a focal point for the inflammation of GCA, once they penetrate the dura
[40,41]. Regardless of the reason, intracerebral GCA is rare but is well-docu-
mented in a small number of patients. In a study of 463 patients from the
Mayo Clinic who had clinical diagnoses of CNS vasculitis, only 2 had con-
vincing evidence of intracranial GCA [42].

Fibrinoid necrosis does not occur in GCA.

Different vasculitides involve blood vessels of characteristic size (large,
medium, or small) and affect different categories of vessels (artery, arteriole,
capillary, venule, vein). They also often cause distinctive histopathologic
findings within the blood vessel wall in some cases. Most GCA cases (but
not all) are associated with multinucleated giant cells that cluster around
the internal elastic lamina. Intimal proliferation is prominent in advanced
stages of GCA. In contrast to PAN, ANCA-associated vasculitis, and cer-
tain other vasculitides, GCA does not cause fibrinoid necrosis (Fig. 7).
The finding of fibrinoid necrosis in a temporal artery or in a branch of
the temporal artery [43] should make the pathologist and clinician consider
other disease entities.

Ultrasonography is a useful procedure for the diagnosis of GCA in the

Fig. 7. Fibrinoid necrosis in a branch of the temporal artery. Fibrinoid necrosis does not occur
in giant cell arteritis, in contrast to polyarteritis nodosa and forms of vasculitis associated with
antineutrophil cytoplasmic antibodies (ANCA), for example. This patient had ANCA directed
against myeloperoxidase and a clinical diagnosis of microscopic polyangiitis.

Ultrasonography made a splash when the idea of using this technology to
diagnose GCA was introduced a decade ago. Ten years of the availability of
ultrasound have revealed it to be a tool that is impractical for broad use in
the community, however. Ultrasound is unlikely to have a significant impact
in the current state of its technology in the diagnosis or management of
Early studies of this imaging procedure were flawed by the failure to blind
ultrasonographers to the clinical history and physical examination data of
the patients. Subsequent studies have indicated that ultrasound of the tem-
poral arteries adds no information to the evaluation of a patient who has
possible GCA, beyond that which can be gleaned from a thorough history
and physical examination [44,45]. The technique remains highly operator-
dependent, and it seems unlikely that radiologists or other practitioners be-
yond specialty centers can be trained to use this technology in the diagnosis
and management of GCA responsibly and well.

Alternate-day glucocorticoid tapers are ineffective in GCA.

For most patients, the worst part of GCA is the prednisone [46]. Clinicians
should resist the temptation to try alternate-day glucocorticoid therapy, be-
cause it is less effective than daily prednisone in GCA [47,48] and there is little
evidence that it leads to less toxicity. In the long run, patients and clinicians
are better off controlling the disease with the lowest dose possible of predni-
sone. A daily baby aspirin is a good idea, too (see earlier discussion).

Temporal artery biopsy is an excellent way to make the diagnosis of giant
cell arteritis.

Temporal artery biopsy is the best way to make the diagnosis of GCA.
Unfortunately, it is far from perfect. Even when performed under rigorous
circumstances, the false-negative rate of temporal artery biopsies is approx-
imately 10% [49]. The false-negative rate is probably considerably higher in
many centers that do not perform as many temporal artery biopsies.
One reason is that temporal arteritis [50,51] does not always clearly in-
volve the temporal arteries. We merely use these vessels whenever possible
to make the diagnosis, because biopsy of the temporal arteries is associated
with almost no morbidity. In 1941, the systemic nature of GCA was recog-
nized and reported in the pathology literature by Gilmour under the name
‘‘giant cell chronic arteritis’’ [52]. In recent years, the frequency of large-ves-
sel involvement (aortitis, subclavian disease) has been appreciated. In one
study, temporal artery biopsies were positive in only 56% of patients who
had large-vessel disease [53].
In the proper clinical setting, MRI of the great vessels is a useful diagnos-
tic test and should be considered in patients suspected to have GCA. MRI
remains an adjunct to temporal artery biopsy, however, rather than an alter-
native. Other studies have shown that the use of MRI as a gauge of disease
activity is still not ready for prime time [54].

Unilateral temporal artery biopsies are generally sufficient for the diagno-
sis of GCA.

Unilateral biopsies in GCA are sufficient only if one is content to miss
between 20% and 40% of cases [55–57]. Although a unilateral temporal ar-
tery biopsy can establish the diagnosis of GCA in many cases, at some in-
stitutions both temporal arteries are biopsied routinely. Bilateral temporal
artery biopsies were performed on 186 patients in one study [56]. Although
only 6 patients had arteritis on only one side, this number represented 20%
of the total number of GCA cases diagnosed through biopsy. In another
study that involved 60 patients [57], 8 (13%) had temporal arteritis on
only one side. Those 8 patients composed 40% of the total number of pa-
tients diagnosed with GCA by biopsy.
Considering the potential consequences of a missed diagnosis and the low
morbidity of the procedure, the routine performance of bilateral temporal
artery biopsies is prudent, especially at institutions where frozen sections
cannot be examined intraoperatively to determine if the first biopsy is

GCA is an extremely rare disease among African Americans, particularly
African American men.

A study from Johns Hopkins evaluated 90 cases of biopsy-proven tempo-
ral arteritis and compared them to 90 controls matched for age, gender, and
year of biopsy [58]. Among the controls, 26 of the 90 patients were African
Americans. Among the cases, however, only 5 patients (5.6% percent) were
African Americans (P!.001).
The upshot of these data is that when considering the cause of an inflam-
matory illness in an African American patient, diagnoses other than GCA
should be weighted more heavily.

GCA is associated with nodular thickening of the temporal arteries.

When present, the finding of nodular, thickened temporal arteries is ex-
tremely helpful in supporting the diagnosis of GCA. Nodular thickening
of the temporal artery on physical examination is associated with an odds
ratio of 4.5 for biopsy-proven GCA [32]. Unequivocally abnormal temporal
arteries occur in a minority of GCA patients, and in approximately one
third of cases of biopsy-proven GCA the physical examination of the tem-
poral arteries is completely normal [49].

In an elderly patient who has an unexplained cough, consider the possi-
bility of GCA.

Cough is one of the many occult manifestations of GCAddisease fea-
tures that are not considered classic for the diagnosis but that, when present,
may signal the diagnosis to the astute clinician. The upper respiratory tract
contains numerous cough receptors. The cause of cough in GCA is pre-
sumed to be irritation of the cough receptors by ischemia to surrounding tis-
sues [59].

GCA is invariably associated with an elevated erythrocyte sedimentation
rate (ESR).

A population-based study of patients who had biopsy-proven GCA in
Olmsted County, Minnesota definitively debunked this myth [60]. A total

of 167 patients were diagnosed between 1950 and 1998. In that cohort, 18
patients (11%) had ESRs less than 50 mm/h, and 9 (5%) had ESRs less
than 40 mm/h. GCA can occur with a low ESR, therefore, and this event
is not rare.
Patients who have GCA who do not have an elevated ESR may be at
higher risk for visual loss. One explanation for this is that such patients’ nor-
mal ESR fails to arouse timely suspicion among clinicians. The presenting
complaint in such patients can be AION.

A normal ESR is more useful than an abnormal one when evaluating
a patient for GCA.

According to the meta-analysis described above [32], a normal ESR is
more useful in excluding GCA than a high ESR is in diagnosing it. Of the
941 biopsy-proven GCA patients who had sufficient information about
the ESR to permit analysis, only 4% had normal ESRs. A normal ESR di-
minished the probability of a positive temporal artery biopsy by a factor of
An elevated ESR by itself does not confirm a diagnosis of GCA. Simi-
larly, the presence of headache does not help distinguish patients who
have this disease from those who have another diagnosis, because headaches
are so common in the general population. An ESR greater than 100 mm/h is
a less-powerful predictor of a positive biopsy than many elements of the his-
tory and physical examination, particularly jaw claudication, diplopia, and
an abnormal (beaded or enlarged) temporal artery [32].

Acute unilateral vision loss in an elderly individual is likely to be GCA.

Only a minority of acute unilateral vision loss events in this setting are
caused by GCA. Another entity with a tendency to occur in old age, nonar-
teritic anterior ischemic optic neuropathy (NAAION), is approximately 10
times more common than GCA as a cause of unilateral vision loss.
NAAION is the most common cause of optic neuropathy (with the ex-
ception of glaucoma) [61]. As with GCA, NAAION tends to affect whites
older than 60 years of age. Unlike GCA, however, NAAION is not an in-
flammatory disorder: constitutional symptoms and elevated acute-phase re-
actants are usually not present. NAAION is associated with funduscopic
findings that help differentiate it from GCA: hyperemic disc swelling and
a morphologically small optic disc with no cup in NAAION, compared
with disc pallor, disc swelling, cotton-wool spots, and a normal optic cup
size in GCA. Table 1 compares these two entities.

Table 1
Comparison of features of arteritic and nonarteritic anterior ischemic optic neuropathy
Feature Arteritic Nonarteritic
Age (mean years) 70 60
Gender distribution Female O male Male ¼ female
Associated symptoms Headache, scalp tenderness Pain occasionally noted
Visual acuity Up to 76% ! 20/20 Up to 61% O 20/20
Disc Pale O hyperemic edema Hyperemic O pale edema
Cup size Cup normal Cup small
Mean ESR 70 mm/h 20–40 mm/h
Fluorescein angiogram Disc and choroids filling delay Disc filling delay
Natural history Improvement rare Improvement in up to 43%
Contralateral involvement Fellow eye in 95% Fellow eye in !30%
Treatment Glucocorticoids None proved

Henoch-Schönlein purpura
High-dose glucocorticoids are an effective therapy for refractory cutane-
ous Henoch-Schönlein purpura (HSP).

HSP is usually a self-limited disorder, but not always. In a small minority
of patients, palpable purpura (Fig. 8) and sometimes other major disease
manifestations (joint complaints, abdominal pain) recur regularly and re-
spond poorly to a series of conventional treatment approaches: glucocorti-
coids at doses of up to 60 mg/d, colchicine, dapsone, azathioprine, and
others. Limited anecdotal experience suggests that rituximab is also ineffec-
tive in this setting.
For reasons that remain unclear, megadoses of glucocorticoids seem to
work in a significant number of these cases. A 3-day methylprednisolone
pulse (1 g/d) can lead to disappearance of the purpura for a year or more.
In some cases, HSP returns after that time, but typically responds again
to megadoses of glucocorticoids. Because glucocorticoids have 100% oral
bioavailability, anecdotal experience has also shown that oral pulses of
prednisone can be effective: prednisone 960 mg (16 60-mg tablets) once daily
for 3 days. This treatment is appropriate, because most cases of recurrent
purpura related to HSP do not require hospitalization.
A second consideration in treatment-refractory HSP is the presence of
a monoclonal IgA gammopathy [62].

Immunofluorescence studies of skin biopsies may yield a diagnosis of vas-
culitis even when H & E stains are nondiagnostic.

Fig. 8. Chronic palpable purpura in a patient who had Henoch-Schönlein purpura. In addition
to chronic purpura, the patient had recurrent arthralgias and arthritis and abdominal pain. He
responded to megadoses of glucocorticoids (see text).

This issue is one of sampling that is obvious to pathologists but not ap-
preciated by many clinicians. The skin biopsy is divided and sent to separate
laboratories for processing. One sample may contain diagnostic information
that is absent in the other. Even if the H & E stain is nondiagnostic, the im-
munofluorescence study may demonstrate the characteristic immunoreac-
tants within blood vessel walls (ie, IgA).
The same phenomenon is true for other kinds of tissue biopsies. Kidney
biopsy samples sent for H & E and immunofluorescence studies in renal dis-
ease may lack the all-important glomeruli, yet the electron microscopy study
can reveal crescentic and necrotizing changes, findings that are critical to
rendering the diagnosis.

Wegener granulomatosis
Strawberry gums are a classic (if uncommon) sign of WG.

WG seems to be unique among the vasculitides in its ability to cause ex-
tensive gum inflammation, as that shown in Fig. 9. When evaluating a pa-
tient for possible WG, the most detailed part of the examination takes
place above the patient’s neck.

Fig. 9. Strawberry gums in Wegener granulomatosis. (Reproduced from the American College
of Rheumatology Copyright Ó 1972–2004, American College of Rheumatology Slide Collec-
tion. Used with permission.)

Prognostication about renal function early in the treatment course of
WG is fraught with hazard.

The kidneys are often the last organ to respond to therapy in WG. Even if
aggressive treatment begins when patients have relatively good renal func-
tion (eg, a serum creatinine of 1.4 mg/dL), the disease marches directly to
end-stage renal disease in a small minority of patients. The reasons for
the delay in renal response to treatment in some cases are not clear, but
are probably multifactorial and may include intercurrent processes, such
as acute tubular necrosis. A significant proportion of patients who require
dialysis in this setting are able to discontinue renal replacement therapy
within several weeks to a few months.

The peri-dialysis time period is a hazardous one for patients who have

Patients who have end-stage or near end-stage renal disease are at increased
risk for complications. In a randomized trial of plasma exchange in ANCA-
associated vasculitis in which all patients had a serum creatinine of at least
5.7 mg/dL, 19 of the 137 patients died of infection within one year, with
more than 25% of patients in the trial expiring within that time period [63].
The increased morbidity and mortality among patients who have ad-
vanced renal dysfunction has multiple causes and several implications for
therapy. The utmost precautions are required to prevent dialysis-related in-
fections. Patients’ CYC should be modified according to their degree of

renal failure (see earlier discussion of cyclophosphamide). PCP prophylaxis

should be used conscientiously.

For patients who have nasal crusting, irrigation of the nose with saline is
a cornerstone of therapy.

There is a spectrum of responses to treatment in nasal disease. Some pa-
tients’ noses clear up completely with immunosuppressive therapy. Others
settle into a pattern of chronic nasal crusting that responds poorly to immu-
nosuppressive therapy and is often difficult to parse according to active dis-
ease or damage (Fig. 10). In such settings, patients usually find that their
own daily irrigation of the nose is the most effective intervention available
for nasal crusting. Sometimes the irrigation must be performed two or
more times a day.

The eyes of a patient who has WG are much more likely to be wet than

Nasal inflammation in WG can lead to nasolacrimal duct dysfunction,
insufficient tear drainage, and watery eyes. Nasolacrimal duct obstruction
(dacryocystitis) is caused by active inflammation or damage to the nose.
A dacryocystorhinostomyda plastic tube placed surgically through the

Fig. 10. Nasal crusting in Wegener granulomatosis. The brown and bloody nasal crusts are
highly characteristic of this disease. (Courtesy of R. Lebovics, MD, New York, NY.)

tear duct of the eye and extending into the superior nosedmay ameliorate
the problem.
Nasal disease in WG has other potential implications for extranasal
symptoms. In the setting of unilateral conjunctivitis in a patient who has
WG, the underlying cause is likely to be nasolacrimal duct dysfunction. Se-
rous otitis in WG is caused by problems that prevent ventilation of the mid-
dle ear through the eustachian tube. Tympanostomy tubes may provide
prompt relief of this problem.

Splenic involvement in WG is rare.
Splenic involvement was actually recognized by Friedrich Wegener in his
first paper on this disease and illustrated nicely with a pathologic figure of
splenic infarctions [64]. Because of the decrease in the number of patients
who have WG who die from their illness and the declining frequency of
deaths studied by autopsy, splenic involvement is probably underrecognized
in WG. Cross-sectional imaging studies have reminded us that splenic in-
volvement can occur in WG, and that it usually presents with splenic infarc-
tions that are evident radiologically [65].
WG can cause painful oral lesions (Fig. 11).
The prominence of oral ulcers in WG is often underappreciated, yet oral
ulcers (along with nasal ulcers) are one of the four American College of
Rheumatology criteria for the classification of WG [66]. In the WG Etaner-
cept Trial (WGET [4]), 14% of the patients had oral ulcerations attributed
by the investigators to active disease at some point over the course of follow-
up [67].

Fig. 11. Painful tongue ulcer in Wegener granulomatosis.


WG rarely erodes the roof of the mouth.
The oral ulcers of WG, which typically involve the buccal mucosa or the
tongue, heal quickly without scarring. In contrast to upper respiratory tract
disease, oral lesions in WG are not destructive. When a patient has an oral
lesion that destroys the roof of the mouth, various other disorders must be
considered, including cocaine use (Fig. 12) and T-cell lymphomas (once
termed ‘‘lethal midline granuloma’’ [68]).
ANCA testing may be useful in differentiating cocaine-associated oral
(and nasal) lesions from true WG [69].

WG can present with headaches that mimic GCA.

The principal characteristic of a GCA headache is that it is new (ie, the
patient has not experienced such headaches before). WG can cause intense,
unrelenting headaches that are new for the patient and mimic GCA. Sources
of headache in WG include sinusitis, meningeal involvement, ocular disease
(posterior scleritis, which may be difficult to diagnose [70]), and temporal
artery inflammation [33,71].

Fig. 12. Cocaine-induced pharyngeal necrosis. Although Wegener granulomatosis can cause
destructive upper airway lesions and is often associated with oral ulcers, it rarely leads to
destructive lesions of the oral cavity.

Episcleritis (Fig. 13) is the most common eye lesion in WG.

Among the many ocular lesions of WG are scleritis, peripheral ulcerative
keratitis, orbital pseudotumor, dacryocystitis, and (more rarely) uveitis (see
article on ‘‘Scleritis and Peripheral Ulcerative Keratitis’’ by Galor and
Thorne in this issue). The most common eye problem in WG is episcleritis,
a painless ocular erythema that usually signals the presence of active disease.
Episcleritis can cause diffuse eye redness or involve the eye in a more sec-
toral fashion, as shown in Fig. 10.

Diffuse scleritis (Fig. 14) spells cyclophosphamide.

WG can be associated with any of the individual types of scleritis, which are
diffuse anterior scleritis, nodular scleritis, necrotizing anterior scleritis, poste-
rior scleritis, and scleromalacia perforans. These forms of ocular inflamma-
tion are distinct clinically, and the individual types of scleritis rarely change
categories. Whereas necrotizing anterior scleritis and posterior scleritis re-
quire intensive immunosuppression with CYC and high-dose glucocorticoids,
the other forms may respond well to milder interventions (glucocorticoids
alone, or even nonsteroidal anti-inflammatory drugs). (See article on ‘‘Scler-
itis and Peripheral Ulcerative Keratitis’’ by Galor and Thorne in this issue).

An increase in ANCA titer predicts a disease flare.

The identification of ANCA and advances in the understanding of these
antibodies are among the most important advances in the understanding of

Fig. 13. Episcleritis in a patient who had Wegener granulomatosis.


Fig. 14. Non-necrotizing scleritis in a patient who had Wegener granulomatosis.

vasculitis in the past quarter century. Assays for ANCA are tremendously
helpful as an adjunct to diagnosis. At least four prospective studies have
shown, however, that the temporal association of ANCA titer increases
with the occurrence of disease flares is poor [72–74]. Data from three of
these studies are shown in Table 2.
As a corollary to the concept that ANCA titers do not predict disease
flares in a timely manner (if at all), it is also worth quashing another
myth: that patients who have WG on the renal transplant list should be
ANCA-negative before receiving a renal allograft. There is no good evi-
dence that elevations in ANCA titers or continued ANCA positivity predict

Table 2
Prospective studies illustrating poor correlation between antineutrophil cytoplasmic antibody
titer elevations and disease flares
PR3-ANCA– collection Detection ANCA
Study N positive interval (mo) method increases Relapsesa
Boomsma 100 85 2 Direct ELISA 38 22
et al [74]
d d d Immunofluorescence 30 15
Jayne 60 39 1 Direct ELISA 27 13
et al [72]
Finkielman 156 134 3 Capture ELISA 34 9
et al [73]
For mature-PR3 d d
Capture ELISA 33 11
For mature-PR3 d d
Within the year following the increase in ANCA.

recurrence of glomerulonephritis in the allograft, and there is some evidence

to the contrary.

Dyspnea on exertion in a patient who has WG whose disease has been
quiescent suggests subglottic stenosis (Fig. 15).

Subglottic stenosis can develop even in patients who have been in appar-
ent remission. The reason for this is not entirely clear, but may relate either
to subclinical disease activity, progression to subglottic stenosis because of
scarring rather than active disease, or both. In any event, clinicians should
remember the propensity of WG to cause subglottic stenosis and evaluate
patients for this complication when they present with decreased exercise tol-
erance, dyspnea on exertion, ‘‘asthma,’’ or similar complaints.

Subglottic stenosis can be diagnosed over the telephone.

Subglottic stenosis can usually be detected by listening to patients as they
talk: because of their upper airway narrowing, they must pause slightly to
suck in air before beginning a sentence. This sound can be detected over
the telephone and in the examining room. On chest auscultation, loud inspi-
ratory sounds (sometimes audible without the stethoscope) are a finding
consistent with subglottic stenosis.

Fig. 15. Subglottic stenosis in a patient who had Wegener granulomatosis. There is a pink shelf
of fibrous tissue between the vocal cords that narrows the trachea in the subglottic region, lead-
ing to dyspnea on exertion. The patient had no evidence of disease activity in other organs at the
time this subglottic stenosis developed.

Confirming the suspicion of subglottic stenosis can be challenging. Thin-

cut computed tomography scans of the trachea can be helpful, but the diagno-
sis often requires direct bronchoscopic visualization by an otolaryngologist.

Tracheostomies are usually good solutions for patients who have WG
with difficult subglottic stenosis.

Tracheostomies should be avoided at all costs in WG. Performing a tra-
cheostomy in a patient who has WG generally makes the patient sicker than he
or she was before the procedure. Avoiding a tracheostomy often requires close
cooperation between the rheumatologist and otolaryngologist, and usually
necessitates frequent endoscopic procedures to inject glucocorticoids and mi-
tomycin C and to dilate stenotic segments of the trachea manually. Some pa-
tients undergo such procedures a half-dozen times or more before the process
is stabilized. Otolaryngologists experienced in the management of subglottic
stenosis in WG counsel strongly against use of the laser in procedures to dilate
airways, because the tissue in WG often responds by exuberant scarring.

Patients who have WG are at increased risk for venous thrombotic events

Among the 180 patients enrolled in the WGET [4], 13 (7.2%) had histories
of VTE before enrollment [75]. During the trial, an additional 16 patients de-
veloped new episodes of VTE. Approximately 1 in 6 patients in the trial thus
had a VTE either before or after trial entry. In a comparison of patients in
the WGET cohort to patients in the Johns Hopkins Lupus Cohort, the inci-
dence of VTE among patients who had WG who had not suffered a VTE be-
fore enrollment was 7.0 per 100 person-years (95% confidence interval [CI]:
0.63–77.9). This incidence rate was sevenfold higher than that of the patients
who had lupus. The median time from enrollment in WGET to VTE was 2
months, implying that VTE is a manifestation of active WG.
In contrast to some other forms of vasculitis (eg, PAN), WG is known
to involve both the venous and arterial sides of the circulation [76]. This
involvement may account for the increased risk for VTE in WG. Investiga-
tions of many conventional risk factors for thrombophilia in WG have been
unrevealing [77,78].

Churg-Strauss syndrome
The Churg-Strauss syndrome (CSS) often involves lymphadenopathy.

The CSS, often included as one of the ANCA-associated vasculitides (see
later discussion), overlaps in many ways with WG and microscopic polyan-
giitis (MPA), as recognized as early as 1954 [79]. Godman and Churg noted
that these three diseases ‘‘group themselves into a compass, [ranging from]
necrotizing and granulomatous processes with angiitis. to vasculitis with-
out granulomata’’ (with mixed forms in between) [79].
One interesting difference between the CSS and the other forms of
ANCA-associated vasculitis is the frequency with which CSS is associated
with lymphadenopathy [80], which is rarely a feature of either WG or
MPA. CSS can also cause striking salivary gland enlargement [81].

Nerve infarction in CSS signals the need to intensify treatment.

Active vasculitic neuropathy is one of the strongest indications for the use
of CYC in CSS. The development of a new nerve infarction shortly after
starting immunosuppression does not necessarily signal the need to intensify
treatment, however, particularly if other indicators of disease activity sug-
gest disease control (Fig. 16). For example, nerve infarctions may manifest
themselves even several weeks after the start of appropriate treatment (high-
dose glucocorticoids and CYC) with the abrupt occurrence of a foot or wrist
drop or other nerve lesion.
In addition, the appearance of muscle wasting secondary to nerve infarc-
tions may continue for weeks after the correct therapy is initiated. This de-
velopment is also not a sign of treatment failure and should not trigger an
intensification of therapy.

Fig. 16. Tongue wasting in a patient who had Churg-Strauss syndrome, attributable to infarc-
tion of the right XII (hypoglossal) cranial nerve. The tongue deviates to the affected side,
consistent with a cranial nerve lesion.

CSS is an ANCA-associated vasculitis.

A substantial percentage of patients who have CSSdmore than 50% in
some studiesdare ANCA-negative [82]. To some extent, the ANCA status
of patients who have CSS may depend on when testing is performed during
the course of therapy [83], but in general it seems that patients who have
CSS are significantly less likely to be ANCA-positive than are patients
who have WG or MPA. ANCA negativity should not dissuade one from
the diagnosis of CSS if other clinical findings fit.
The presence or absence of ANCA may also influence clinical phenotype
in CSS, but the literature is mixed on this point. In one study [83], CNS in-
volvement was more common among patients who were ANCA-positive
(20% versus 5%). In another [84], patients who were ANCA-positive were
significantly more likely to have glomerulonephritis and neurologic disease,
but patients who were ANCA-negative were more likely to have cardiac in-
volvement. ANCA status did not seem to affect either mortality or (as in
WG) the likelihood of relapse.

CSS is caused by the use of leukotriene inhibitors.

This myth, approximately as old as leukotriene inhibitors themselves, is
a classic case of confounding. Leukotrienes are highly effective asthma med-
ications and permit the taper of prednisone in glucocorticoid-dependent
asthma. Most patients who have CSS have asthma as their presenting fea-
ture and are usually misdiagnosed initially as having simple reactive airway
disease. The tapering of prednisone permitted by a leukotriene inhibitor
eventually unmasks the underlying CSS [85].
CSS-related vasculitis recurs in only about 25% of all patients. After the
vasculitic phase of CSS has been treated effectively, most patients are left
with glucocorticoid-dependent asthma and are unable to taper their predni-
sone to less than 8 mg/d [86]. Managing this refractory reactive airway dis-
ease is a major challenge in caring for patients who have CSS over the long
term. Some clinicians who are also experienced in the care of patients who
have CSS are sufficiently comfortable that leukotriene inhibitors do not
cause the disease that they actually use these agents to treat the glucocorti-
coid-dependent asthma of CSS [83].

Microscopic polyangiitis
Interstitial fibrosis is common pulmonary manifestation of MPA.

Interstitial lung disease is an important and underrecognized complica-
tion of vasculitis associated with ANCA directed against myeloperoxidase
(MPO-ANCA) (Fig. 17) [87,88]. MPO-ANCA can occur in WG and CSS
and in MPA, but interstitial lung disease usually occurs in patients whose
clinical phenotype is most consistent with MPA (Table 3). Although the
chest radiologic appearance of this disorder resembles usual interstitial
pneumonitis (UIP), there are subtle pathologic differences between the in-
terstitial lung disease of MPA and UIP [89]. Most importantly, the inter-
stitial lung disease of MPA is strikingly responsive to therapy in many
cases, a marked contrast with other forms of idiopathic interstitial

Arthritis is a common presentation of a flare in ANCA-associated

Arthritis and arthralgias are frequently overlooked signs of ANCA-asso-
ciated vasculitis, perhaps because they improve so quickly after the institu-
tion of therapy. The arthritis commonly involves large joints in a migratory,
asymmetric, oligoarticular pattern, one day involving a knee and an ankle,
the next day a shoulder. Small joints may also be involved. Arthritis and ar-
thralgias are a common tip-off to a disease flare.

MPA is less likely than WG to flare.

Fig. 17. Interstitial lung disease in microscopic polyangiitis. The patient, a 55-year-old man,
presented with dyspnea, dry rales, and segmental necrotizing (pauci-immune) glomerulonephri-
tis. He had antineutrophil cytoplasmic antibodies directed against myeloperoxidase.

Table 3
Unique disease features of the three antineutrophil cytoplasmic antibody–associated forms of
Disease Features
Wegener granulomatosis Destructive upper respiratory tract disease
Orbital pseudotumor
Strawberry gums
Hypertrophic pachymeningitis
Microscopic polyangiitis Interstitial lung disease
Churg-Strauss syndrome Asthma

In the CYCAZAREM trial [12], disease flares occurred in 17 of the 92
patients who had WG (18%), compared with 4 of 52 (8%) of those who
had MPA. These findings are consistent with other studies that involved lon-
ger follow-up. In a case series of 85 patients who had MPA with longer fol-
low-up, disease flares occurred in approximately 33% [90]. In contrast, up to
80% of patients who have WG flare over time [5,6].

Scurvy can mimic ANCA-associated vasculitis, particularly MPA.

Scurvy can cause perifollicular petechiae that mimic purpura, extensive
gum inflammation, and alveolar hemorrhage, all of which could be confused
with MPA [91]. Think of scurvy when a severely malnourished patient is sus-
pected of having an ANCA-associated vasculitis.

Co-occurrence of ANCA and anti-GBM antibodies is associated with
a worse renal prognosis.

Up to 30% of patients diagnosed with anti-GBM disease may also have
ANCA, usually directed against myeloperoxidase and associated with a peri-
nuclear (P-ANCA) pattern of immunofluorescence [92]. Conversely, up to
10% of patients who have ANCA-associated vasculitis may also have
anti-GBM antibodies. Both types of autoantibodies should be sought in pul-
monary-renal syndromes. The presence of anti-GBM antibodies is an indi-
cation for plasma exchange in addition to immunosuppressive agents.

Cryoglobulinemic vasculitis
Hepatitis C–associated cryoglobulinemia can present with a pulmonary-
renal syndrome.

A proliferative glomerulonephritis is a well-known complication of cryo-
globulinemic vasculitis. Less familiar complications of hepatitis C virus
(HCV) infections can involve the lung. Pulmonary hemorrhage and intersti-
tial fibrosis have been reported in patients who have HCV [93]. The poten-
tial association between HCV and lung dysfunction is worth keeping in
mind when the pieces of a puzzling pulmonary case do not add up.

False-negative results in testing for cryoglobulins are common.

Sensitive testing for cryoglobulins requires an experienced laboratory
that is set up to perform the collection in the proper conditions. While the
patient is fasting (lipids can interfere with the assay), at least 20 mL of blood
should be drawn into a tube that has not been treated with anticoagulant.
The specimen should be transported and centrifuged at 37 C, and then
kept for more than 72 hours at 4 C to allow for the precipitation of cryoglo-
bulins [94].
Persistence (ie, repeat testing and coordination with the laboratory) may
be required to obtain a good assay. Clues that suggest a harder look for cry-
oglobulins is in order include positive HCV serologies and a serum C4 level
that is depressed out of proportion to the decrease in C3.

Most patients who have serum cryoglobulins are at risk for an array of
inflammatory conditions that are complications of vasculitis.

Although cryoglobulinemia is present in 40% of patients who have
chronic hepatitis C, vasculitic sequelae are seen in only 2% to 3% of these
patients [95,96]. The risk factors for the development of cryoglobulinemic
vasculitis remain poorly defined, but duration of HCV infection seems to
correlate highly with the risk for vasculitis. Cryoglobulinemic vasculitis typ-
ically occurs in patients who have longstanding HCV infections. Most pa-
tients who develop cryoglobulinemia are unaware that they have HCV.

The cryocrit correlates well with disease activity in cryoglobulinemic

Cryocrits tend to be higher in type I cryoglobulinemia compared with
those found in types II and III, and can continue to be frighteningly high
even when patients’ clinical symptoms (eg, digital ischemia) have improved

in response to treatment. The cryocrit level should not be used as a gauge for

Cryoglobulinemic symptoms are exacerbated by the winter and other
conditions of cold.

The cryoglobulins associated with HCV (types II and III) are immune
complexes that precipitate in cold laboratory conditions (eg, refrigeration
for several days at 4 C). This condition does not occur under physiologic
conditions because of the maintenance of body temperature at approxi-
mately 35 to 37 C, even in the acral extremities.
In contrast, type I cryoglobulins can precipitate at temperatures achieved
in the distal extremities, nose, ears, and elsewhere. Type I cryoglobulins of-
ten present with cold-induced symptoms, and are then termed cryoprecipi-
tating monoclonal gammopathies.

Purpura in immune complex–mediated vasculitides, such as cryoglobuli-
nemia, is generally more extensive than in ANCA-associated vasculitis.

Immune complex–mediated disease is likely to be associated with exten-
sive, confluent purpura involving large areas of the lower extremities, but-
tocks, and other parts of the body, even extending in some cases onto the
trunk, arms, and face (Fig. 18A). In contrast, patients who have WG can
have overwhelming inflammation in certain organ systems (eg, the kidney)
yet have little or no purpura (Fig. 18B). An assessment of the extent of

Fig. 18. Purpura in immune complex–mediated vasculitis as opposed to pauci-immune vascu-

litis. (A) Extensive purpura in a patient who had Henoch-Schönlein purpura, even including
lesions on the face. The patient had diffuse purpuric lesions of her extremities and trunk
also. (B) Subtle purpura in a patient who had Wegener granulomatosis who had intense inflam-
mation in other organ systems (active glomerulonephritis refractory to cyclophosphamide and

cutaneous disease helps focus the differential diagnosis until more specific
information from serologic testing or biopsies is available.

Takayasu arteritis
Takayasu arteritis can be associated with extremely high platelet counts.

In a young woman who has an extremely high platelet count, a hematocrit
compatible with anemia of chronic disease, and a bone marrow examination
that shows reactive hyperplasia, check to make sure her pulses are palpable
in all extremities (particularly her arms). Platelet counts greater than 1 mil-
lion/mL in Takayasu arteritis may be observed. Platelets are acute-phase re-
actants and in Takayasu arteritis they reflect the degree of inflammation
within the walls of the great vessels and their primary branches.

Headaches in Takayasu arteritis are bad news.

Takayasu arteritis tends to cause bilateral subclavian artery stenoses and
stenosis of the infrarenal aorta or narrowing of the large vessels to both
lower extremities. Unfortunately, it can also cause renal artery stenosis
(Fig. 19), leading to renin-mediated hypertension. The combined effects of
these various stenoses over time are that blood pressure measurements in
the arms and even the legs do not provide accurate reflections of the central
aortic pressure, which may become dangerously elevated.

Fig. 19. Renal artery stenosis in Takayasu arteritis leading to renin-mediated hypertension in
a 24-year-old woman.

The patient who has Takayasu arteritis and intractable headaches re-
quires a serum renin measurement, a thorough investigation of the circula-
tion to all extremities, precise imaging of the renal arteries, and an accurate
determination of the central aortic pressure. Reliable central aortic pressure
measurements in this setting require catheterization.

Many attempts at revascularization in large-vessel vasculitis are unneces-
sary or even ill-advised.

In Takayasu arteritis, everything that narrows does not need to be fixed.
The wisdom of potential interventions should be considered carefully. Ste-
notic lesions in large blood vessels, such as the subclavian artery, take
many months or years to form. The slow development of these lesions per-
mits dramatic neovascularization to occur (Fig. 20). The spiderwebs of
blood vessels that form in large-vessel vasculitis are not associated with pal-
pable pulses at the brachial or radial sitesdhence the term ‘‘pulseless dis-
ease’’ to describe Takayasu arteritis. This collateral circulation is sufficient
to ensure the continuation of critical blood flow to extremities, however.
In the days of glucocorticoid therapy for patients who have Takayasu ar-
teritis, the loss of digits or other parts of the extremities to amputation is
a rare event. A major function of prednisone may be to check inflammation
in primary branches long enough to allow the establishment of collateral
Although critical stenoses to vital organs in which collateral circulation is
poor need to be addressed with bypass surgery or some other attempted re-
vascularization, many attempts at angioplasty or stent placement in the long
stenoses of well-collateralized lesions are either doomed to fail or superflu-
ous from the start.

Fig. 20. Extensive collateral circulation around the right subclavian artery in Takayasu

Cogan syndrome
Any kind of ocular inflammation can occur in Cogan syndrome.

Nonsyphilitic interstitial keratitis is the classic ocular manifestation of
Cogan syndrome, but just about any type of ocular inflammation can occur
in this disorder. In addition to interstitial keratitis and scleritis, other ocular
lesions in Cogan syndrome include conjunctivitis, episcleritis, uveitis, retinal
vasculitis, orbital pseudotumor, and a pan-ophthalmitis that resembles or-
bital pseudotumor [97]. Patients who do not have the classic interstitial ker-
atitis at diagnosis are sometimes referred to as having atypical Cogan
syndrome, but it is not clear that their courses differ from those of patients
who have the more textbook presentation. If an atypical eye lesion develops
before the onset of sensorineural hearing loss, the diagnosis of Cogan syn-
drome is impossible to make until both ocular and auditory findings are

If disease remission is achieved in Cogan syndrome, treatment should
never be stopped completely.

If the ability of clinicians to control most forms ocular inflammation with
immunosuppression is usually rewarding, then the challenge of preventing
deafness in Cogan syndrome is humbling. The largest available series of pa-
tients who have Cogan syndrome paint discouraging pictures of the likeli-
hood of retaining useful hearing over time. In a study from the Mayo
Clinic [98], 42 of 60 patients became deaf in at least one ear, and 31 of 60
became completely deaf over time. In a study from France [99], 30 of 32 pa-
tients developed significant auditory dysfunction and 11 became completely
These reports and anecdotal experiences inform the bias that patients who
achieve disease control should never discontinue their treatment entirely, be-
cause ground lost in hearing is usually not regained. The optimal remission
maintenance regimen is far from clear (ie, it is entirely empiric), and must
be devised individually for each patient. It may include classic immunomodu-
lating agents, such as methotrexate, azathioprine, or mycophenolate mofetil,
but may also require low-dose glucocorticoids over the long term.

Central nervous system vasculitis

Most cases of CNS vasculitis are really vasospasm.

In recent years, an important synthesis of knowledge related to CNS vas-
culitis has occurred. This synthesis involves the conclusion that most cases
of CNS vasculitis not documented by histopathology (the great majority
of all cases in this day of multiple imaging modalities) are not vasculitis
at all but rather vasospasm [100]. A host of entities known by other names
are often confused with and mistreated as CNS vasculitis. These include be-
nign angiopathy of the CNS, cases of vasculitis associated with pseudoephe-
drine use, thunderclap headache, and others.
Absent a stroke that is evident on MRI or histopathologic confirmation
of vasculitis, it is wise usually to begin with the assumption that the patient
who has CNS vasculitis does not have primary angiitis of the CNS, but
rather a vasculopathy of some other name associated primarily with vaso-
spasm rather than the inflammatory destruction of blood vessel walls.

The critical test in the evaluation of a patient who has CNS vasculitis is
a repeat angiogram in 4 to 6 weeks.

Because most cases of CNS vasculitis are really vasospasm, a repeat CNS
angiogram a few weeks after presentation may show dramatic improvement
or complete resolution of earlier changes. Such improvement argues
strongly for a vasospastic cause of the findings.

Behçet disease
In Behçet disease, uveitis does not resolve completely between clinical

The implications of this are that the posterior uveitis of Behçet disease
must be monitored carefully and therapy stopped only after great delibera-
tion and the experience of longitudinal follow-up. Although randomized
controlled trials are still lacking, tumor necrosis factor inhibition has be-
come the standard of care now for the treatment of posterior uveitis in Beh-
çet disease, replacing cyclophosphamide, which offered certain toxicity but
unclear benefit.
With careful follow-up, it is probably safe to stop immunosuppressive
therapies in at least some patients who have Behçet disease. Some patients
who have Behçet disease go into complete remission with the passage of
time [101].

Genital ulcers in male patients who have Behçet syndrome are usually
found on the scrotum.

This feature helps distinguish Behçet ulcers in males from those of Reiter
syndrome or venereal diseases, which usually occur on the penis. Genital ul-
cers in Behçet, like the oral ulcers, are usually painful. In female patients
who have Behçet disease, genital ulcers usually occur on the labiadthe em-
bryological counterpart in women of the scrotum. In women, the genital ul-
cers are not so easy to distinguish from herpetic lesions, because both are
painful. Syphilitic chancres, in contrast, are usually painless.

Behçet disease patients can get acneiform lesions that are indistinguish-
able from acne vulgaris in appearance and histology.

The distinguishing feature between these entities is that in Behçet disease,
these lesions occur on the arms and legs rather than on the face, back, and
chest. The company that the lesions keep is also critical, of course, because
oral ulcers are virtually a sine qua non of Behçet disease.

Behçet disease is a form of spondyloarthropathy.

This myth dates from the early 1970s and the discovery of HLA-B27,
when it was believed that this antigen would become the Rosetta stone
for understanding many types of rheumatic disease. The increased preva-
lence of sacroiliac joint involvement reported in Behçet disease has not
been confirmed in controlled studies [102].

Despite the frequency of venous thrombotic events in Behçet disease,
thromboembolism is rare.

This pearl may be true but is difficult to reference with any clinical data.
The suggestion is that thromboembolism occurs rarely in Behçet disease be-
cause the thrombi adhere tightly to diseased veins [103,104].
The management of patients who have Behçet disease who develop
thrombidanticoagulation versus immunosuppressiondis often debated.
The consensus approach of Behçet disease experts is not to anticoagulate
these patients in most cases, because the cause of the clots is intense vessel

wall inflammation that is treated more effectively with immunosuppression


Buerger disease
Buerger disease is a systemic vasculitis.

Although often classified as a systemic vasculitis, there are several impor-
tant distinctions between thromboangiitis obliterans and other forms of vas-
culitis [106]. Buerger disease is characterized pathologically by a highly
cellular thrombus with relative sparing of the blood vessel walls. Multinucle-
ated giant cells can even be observed within the clot in Buerger disease. The
architecture of the vascular wall is preserved, and there is no fibrinoid necro-
sis. The histopathologic findings in this disease provide significant insights
into why systemic immunosuppressive therapy is relatively useless in
Buerger disease.

Buerger disease is a disease of young men.

When Horton from the Mayo Clinic wrote about the disease in 1938
[107], less than 1% of patients affected were women. As the proportion of
smokers who are women has grown, the percentage of Buerger disease pa-
tients who are women has also grown. More recent estimates of the propor-
tion of patients who have Buerger disease who are women have been as high
as 23% [108,109]. Buerger disease has a lot more to do with whether or not
one smokes than whether or not a Y chromosome is present.

Cannabis arteritis is a different disease than Buerger disease.

It has been suggested that digital ischemia can arise from the use of mar-
ijuana, independent of tobacco [110]. Scrutiny of these papers indicates that
cannabis arteritis is similar to Buerger disease clinically, that the reported
differences in pathologic lesions are unconvincing, and that most of the pa-
tients smoked cigarettes, too [111,112]. Cannabis arteritis is probably really
just Buerger disease.

Superficial thrombophlebitis is observed in 40% to 60% of cases.

The superficial thrombophlebitis of Buerger disease is migratory, recur-
rent, and may affect the arms and legs. Migrating phlebitis (phlebitis saltans)
in young patients is therefore highly suggestive of thromboangiitis obliter-
ans. Deep venous thrombosis in Buerger disease is unusual.
Biopsy of a subcutaneous nodule can be used to confirm the presence of
superficial thrombophlebitis at an early phase of Buerger disease, before the
onset of larger vessel thrombosis [113].

Buerger disease may begin with joint manifestations.

Recurrent episodes of large-joint arthritis and transient, migratory single-
joint episodes accompanied by local signs of inflammation have been
described in Buerger disease [114]. When present, joint symptoms usually pre-
cede the onset of digital ischemia. In this way, Buerger disease may resemble
other forms of vasculitis, which are often heralded by musculoskeletal com-
plaints (see earlier discussion of microscopic polyangiitis). In Buerger disease,
the arthritis disappears definitively with the appearance of ischemic signs.

Extra-extremity disease may occur in Buerger disease.

Joints aside, the chief characteristic of the presentation of Buerger disease
is the intense ischemia of the extremities that stands in stark contrast to the
relative absence of ischemia in other vascular beds. The rare cases of
Buerger disease involving the CNS or gastrointestinal tract are generally
poorly documented [106]. If a patient presents with significant internal or-
gan dysfunction in addition to digital ischemia, the diagnosis is probably
not Buerger disease.

Buerger disease does not occur in only one limb.

Buerger disease virtually always occurs in more than one limb, often in all
four, even if there is not clinical evidence of multilimb involvement [115]. If
called to see a patient who has intense ischemia in only one limb and no an-
giographic evidence of disease in other extremities, Buerger disease is not the

Buerger disease may present with Raynaud phenomenon.

Buerger disease causes digital ischemia equaled by few other disorders,
but it is not because of Raynaud phenomenon. Raynaud phenomenon, so
typical of systemic sclerosis, systemic lupus erythematosus, and other con-
nective tissue disorders, is caused by vasospasm usually induced by cold.
Conversely, Buerger disease is caused by a highly inflammatory clot in the
medium-sized arteries at the level of the wrists and ankles. The term
Raynaud’s phenomenon should not be applied to patients who have Buerger
disease, nor to any other form of vasculitis that causes digital ischemia.

Patients who continue to smoke are the ones who lose limbs.

An analysis of 850 patients identified by a national study in Japan
showed that the risk for amputation was 2.73 times higher (95% CI:
1.86–4.01) in patients who continued to smoke [116]. Patients who have
Buerger disease have to choose between their cigarettes and their distal ex-
tremities (Fig. 21).

Buerger disease may present with subungual splinter hemorrhages.

Endocarditis is not the only disorder that causes splinter hemorrhages. In
addition to Buerger disease (Fig. 22) [117], several forms of vasculitis are
known to do this. Other examples include ANCA-associated vasculitis
and cryoglobulinemic vasculitis.

Fig. 21. Patients who have Buerger disease who continue to smoke continue to lose digits and
other portions of their extremities. This patient lost every finger on both hands and had bilateral
below-the-knee amputations.

Fig. 22. Splinter hemorrhages at presentation in a patient who had Buerger disease (the patient
in Fig. 21 before finger amputation).

Rheumatoid vasculitis
Distal, symmetric polyneuropathy occurring in a patient who has long-
standing rheumatoid arthritis (RA) is likely to be a vasculitic neuropathy.

The inclination is to lump most patients who have symmetrical sensory
symptoms into the category of distal peripheral neuropathy, along with
the 400 or so other causes of this disorder (including diabetic neuropathy
and alcoholic neuropathy). In patients who have RA who are the typical
substrate for rheumatoid vasculitis, however, the cause of such symptoms
is often vasculitic neuropathy [118]. Careful evaluations of these patients
and an increase in their RA therapy (if not treatment targeting the vasculitis
specifically) may be required.

[1] Kussmaul A, Maier R. Über eine bisher nicht beschriebene eigenthumliche arterienerkran-
kung (periarteritis nodosa), die mit morbus brightii und rapid fortschreitender allgemeiner
muskellahmung einhergeht. Dtsch Arch Klin Med 1866;1:484–518.
[2] Matteson EL. Polyarteritis nodosa: commemorative translation of the 130-year anniver-
sary of the original article by Adolf Kussmaul and Rudolf Maier. Rochester (MN): Mayo
Foundation; 1996.
[3] Kussmaul A, Maier R. Aneurysma verminosum hominis: verlaufige nachricht. Deutsches
Archiv fur klinische Medizin 1866;1:125–6.
[4] Stone JH, Calabrese LH, Hoffman GS, et al. Vasculitis: a collection of pearls and myths.
Rheum Dis Clin NA 2001;27:677–728.
[5] The WGET Research Group. Etanercept plus standard therapy in patients with Wegener’s
granulomatosis. N Engl J Med 2005;352:351–61.

[6] Reinhold-Keller E, Beuge N, Latza U, et al. An interdisciplinary approach to the care of

patients with Wegener’s granulomatosis. Long term outcome in 155 patients. Arthritis
Rheum 2000;43(5):1021–32.
[7] Hoffman G, Kerr G, Leavitt R, et al. Wegener’s granulomatosis: an analysis of 158 patients.
Ann Intern Med 1992;116:488–98.
[8] Goek ON, Stone JH. Randomized controlled trials in ANCA-associated vasculitis. Curr
Opin Rheumatol 2005;17:257–64.
[9] Thomas CF Jr, Limper AH. Pneumocystis pneumonia. N Engl J Med 2004;350:2487–98.
[10] Fauci AS, Dale DC, Wolff SM. Cyclophosphamide and lymphocyte subpopulations in We-
gener’s granulomatosis. Arthritis Rheum 1974;17(4):355–61.
[11] Haubitz M, Brunkhorst R, Bohnenstengel F, et al. Cyclophosphamide phramacokinetics
and dose requirements in patients with renal insufficiency. Kidney Int 2002;61:1495–501.
[12] Jayne D, Rasmussen N, Bacon PA, et al. Randomised trial of cyclophosphamide versus
azathioprine during remission in ANCA-associated systemic vasculitis. N Engl J Med
[13] The WGET Research Group. Design of the Wegener’s granulomatosis etanercept trial
(WGET). Control Clin Trials 2002;23:450–68.
[14] Langford C, Talar-Williams C, Barron K, et al. A staged approach to the treatment of We-
gener’s granulomatosis: induction of remission with glucocorticoids and daily cyclophos-
phamide switching to methotrexate for remission maintenance. Arthritis Rheum 1999;42:
[15] Teml A, Schaeffeler E, Herrlinger KR. Thiopurine treatment in inflammatory bowel dis-
ease: clinical pharmacology and implication of pharmacogenetically guided dosing. Clin
Pharmacokinet 2007;46:187–208.
[16] Sinico RA, Sabadini E, Borlandelli S, et al. Azathioprine hypersensitivity: report of two
cases and review of the literature. J Nephrol 2003;16:272–6.
[17] Sneller MC, Hoffmann GS, Talar-Williams C, et al. An analysis of forty-two Wegener’s
granulomatosis patients treated with methotrexate and prednisone. Arthritis Rheum
[18] Stegeman CA, Tervaert JW, deJong PE, et al. Trimethoprim-sulfamethoxazole (co-trimox-
azole) for the prevention of relapse in Wegener’s granulomatosis. N Engl J Med 1996;335:
[19] Guillevin L, Lhote F. Treatment of polyarteritis nodosa and microscopic polyangiitis-
Arthritis Rheum 1998;41:2100–5.
[20] Pagnoux C, Seror R, Mahr A, et al. Polyarteritis nodosa: clinical description and long-term
follow-up of the 351 patients from the French Vasculitis Study Group. Arthritis Rheum
[21] Collins MP, Mendell JR, Periquet MI, et al. Superficial peroneal nerve/peroneus brevis
muscle biopsy in vasculitic neuropathy. Neurology 2000;55:636–43.
[22] Mathew L, Talbot K, Love S, et al. Treatment of vasculitic peripheral neuropathy: a retro-
spective analysis of outcome. QJM 2007;100:41–57.
[23] Vallat J-M, Cros DP, Hedley-White ET. Case 9-2007: a 27 year-old woman with pain and
swelling of the legs. N Engl J Med 2007;356:1252–9.
[24] Talor MV, Stone JH, Stebbing JJ, et al. Antibodies to selected minor target antigens in pa-
tients with anti-neutrophil cytoplasmic antibodies (ANCA) by immunofluorescence anti-
bodies to minor target antigens in patients with anti-neutrophil cytoplasmic antibodies.
Clin Exp Immunol 2007;150:42–8.
[25] Stone JH, Nousari HC. ‘‘Essential’’ cutaneous vasculitis: what every rheumatologist should
know about vasculitis of the skin. Curr Opin Rheumatol 2001;13:23–34.
[26] Frances C, Niang S, Laffitte E, et al. Dermatologic manifestations of the antiphospholipid
syndrome. Arthritis Rheum 2005;52:1785–93.
[27] Uthman IW, Khamashta MA. Livedo racemosa: a striking dermatological sign for the anti-
phospholipid syndrome. J Rheumatol 2006;33:2379–82.

[28] Stockhammer GJ, Felber SR, Zelger B, et al. Sneddon’s syndrome: diagnosis by skin biopsy
and MRI in 17 patients. Stroke 1993;24:685–90.
[29] Yazici H, Chamberlain MA, Tuzun Y. A comparative study of the pathergy reaction
among Turkish and British patients with Behcet’s disease. Ann Rheum Dis 1984;43:74–5.
[30] Ehrmann S. Ein gefaessprozess bei lues. Wien Med Wochenschr 1907;57:777–80.
[31] Frances C, Papo T, Wechser B, et al. Sneddon’s syndrome with and without aPL. Medicine
[32] Smetana GW, Shmerling RH. Does this patient have temporal arteritis? JAMA 2002;287:
[33] Small P, Brisson M-L. Wegener’s granulomatosis presenting as temporal arteritis. Arthritis
Rheum 1991;34:220–3.
[34] Lee MS, Smith SD, Gabor A. Antiplatelet and anticoagulant therapy in patients with giant
cell arteritis. Arthritis Rheum 2006;54:3306–9.
[35] Nesher G, Berkun Y, Mates M. Low-dose aspirin and prevention of cranial ischemic com-
plications in giant cell arteritis. Arthritis Rheum 2004;50:1332–7.
[36] Stone JH. Antiplatelet versus anticoagulant therapy in patients with giant cell arteritis:
which is best? Nat Clin Pract Rheumatol 2007;3:136–7.
[37] Hayreh SS, Zimmerman B, Kardon RH. Visual improvement with corticosteroid therapy
in giant cell arteritis. Report of a large study and review of literature. Acta Ophthalmol
Scand 2002;80:355–67.
[38] Miller NR. Visual manifestations of temporal arteritis. Rheum Dis Clin North Am 2001;27:
[39] Razavi M, Jones RD, Manzel, et al. Steroid-responsive Charles Bonnet syndrome in tem-
poral arteritis. J Neuropsychiatry Clin Neurosci 2004;16:505–8.
[40] Wilkinson IM, Russell RW. Arteries of the head and neck in giant cell arteritis: a patholog-
ical study to show the pattern of arterial involvement. Arch Neurol 1972;27:378–91.
[41] Caselli RJ, Hunder GG, Whisnant JP. Neurologic disease in biopsy-proven giant cell (tem-
poral) arteritis. Neurology 1988;38:352–9.
[42] Salvarani C, Giannini C, Miller DV, et al. Giant cell arteritis: involvement of intracranial
arteries. Arthritis Rheum 2006;55:985–9.
[43] Esteban MJ, Font C, Hernandex-Rodriguez J, et al. Small-vessel vasculitis surrounding
a spared temporal artery: clinical and pathological findings in a series of twenty-eight pa-
tients. Arthritis Rheum 2001;44:1387–95.
[44] Karahaliou M, Vaiaopoulos G, Papaspyrous S, et al. Colour duplex sonography of tempo-
ral arteries before decision for biopsy: a prospective study in 55 patients with suspected gi-
ant cell arteritis. Arthritis Res Ther 2006;8:R116.
[45] Salvarani C, Silingardi M, Ghirarduzzi A, et al. Is duplex ultrasonography useful for the
diagnosis of giant-cell arteritis? Ann Intern Med 2002;137:232–8.
[46] Hellmann DB, Uhlfelder ML, Stone JH, et al. Domains of quality of life important to pa-
tients with giant cell arteritis. Arthritis Rheum 2003;49(6):819–25.
[47] Hunder GG, Sheps SG, Allen GL, et al. Daily and alternate-day corticosteroid regimens in
treatment of giant cell arteritis: comparison in a prospective study. Ann Intern Med 1975;
[48] Hoffman GS, Cid MC, Hellmann DB, et al. A multicenter, randomized, double-blind, pla-
cebo-controlled trial of adjuvant methotrexate treatment for giant cell arteritis. Arthritis
Rheum 2002;46:1309–18.
[49] Hall S, Lie J, Kurland L, et al. The therapeutic impact of temporal artery biopsy. Lancet
[50] Horton BT, Magath TB, Brown GE. Arteritis of temporal vessels: report of 7 cases. Proc
Staff Meet Mayo Clin 1937;12:548–53.
[51] Horton BT, Magath TB, Brown GE. Undescribed forms of arteritis of temporal vessels.
Proc Staff Meet Mayo Clin 1932;7:700–1.
[52] Gilmour JR. Giant-cell chronic arteritis. J Pathol 1941;53:263–77.

[53] Brack A, Martinez-Taboada VM, Rodriguez-Valverde V, et al. Disease pattern in cranial

and large-vessel giant cell arteritis. Arthritis Rheum 1999;42:311–7.
[54] Tso E, Flamm SD, White RD, et al. Takayasu arteritis: utility and limitations of mag-
netic resonance imaging in diagnosis and treatment. Arthritis Rheum 2002;46(6):
[55] Seo P, Stone JH. Large-vessel vasculitis. Arthritis Care Res 2002;51:128–39.
[56] Boyev LR, Miller NR, Green WR. Efficacy of unilateral versus bilateral temporal artery
biopsies for the diagnosis of giant cell arteritis. Am J Ophthalmol 1999;128:211–5.
[57] Pless M, Rizzo JF 3rd, Lamkin JC, et al. Concordance of bilateral temporal artery biopsy in
giant cell arteritis. J Neuroophthalmol 2000;20:216–8.
[58] Regan MJ, Green WR, Stone JH. Ethnic disparity in the incidence of temporal arteritis:
a 32-year experience at an urban medical center. Arthritis Rheum 2000;43:S128.
[59] Olopade CO, Sekosan M, Schraufnagel DE. Giant cell arteritis manifesting as chronic
cough and fever of unknown origin. Mayo Clin Proc 1997;72:1048–50.
[60] Salvarani C, Hunder GG. Giant cell arteritis with a low erythrocyte sedimentation rate: fre-
quency of occurence in a population-based study. Arthritis Rheum 2001;45:140–5.
[61] Lessell S. Nonarteritic anterior ischemic optic neuropathy: enigma variations. Arch Oph-
thalmol 1999;117:386–8.
[62] Birchmore D, Sweeney C, Choudhury D, et al. IgA multiple myeloma presenting as HSP/
polyarteritis nodosa overlap. Arthritis Rheum 1996;39:698–703.
[63] Jayne DR, Gaskin G, Rasmussen N, et al. Randomized trial of plasma exchange or high-
dosage methylprednisolone as adjunctive therapy for severe renal vasculitis. J Am Soc
Nephrol 2007;18:2180–8.
[64] Wegener F. Uber generalisierte, septische gefasserkrankungen. Verh Dtsch Ges Pathol
[65] Papaionnides A, Nikas SN, Fotinou M, et al. Asymptomatic splenic infarction in Wege-
ner’s granulomatosis. Ann Rheum Dis 2002;61:185–6.
[66] Leavitt RY, Fauci AS, Bloch DA, et al. The American College of Rheumatology 1990
criteria for the classification of Wegener’s granulomatosis. Arthritis Rheum 1990;33:
[67] Mahr AD, Neogi T, LaValley MP, et al. Assessment of the item selection and weighting in
the Birmingham vasculitis activity score for Wegener’s granulomatosis (BVAS/WG)
[68] Al-Hakeem DA, Fedele S, Carlos R, et al. Extranodal NK/T-cell lymphoma, nasal type.
Oral Oncol 2007;43:4–14.
[69] Wiesner O, Russell KA, Lee AS, et al. Antineutrophil cytoplasmic antibodies reacting with
human neutrophil elastase as a diagnostic marker for cocaine-induced midline destructive
lesions but not autoimmune vasculitis. Arthritis Rheum 2004;50:2954–65.
[70] Sadiq Z, Shepherd R, Vardy S. Posterior scleritis: an unusual presentation of facial pain.
Br J Oral Maxillofac Surg 2003;41:62–3.
[71] Genereau T, Lortholary O, Pottier M, et al. Temporal artery biopsy. A diagnostic tool for
systemic necrotizing vasculitis. Arthritis Rheum 1999;42:2674–81.
[72] Jayne DR, Gaskin G, Pusey CD, et al. ANCA and predicting relapse in systemic vasculitis.
QJM 1995;88:127–33.
[73] Finkielman JD, Merkel PA, Schroeder D, et al, for the WGET Research Group. Antineu-
trophil cytoplasmic antibodies against proteinase 3 do not predict disease relapses in We-
gener’s granulomatosis. Ann Intern Med 2007;147(9):611–9.
[74] Boomsma MM, Stegeman CA, van der Leij MJ, et al. Prediction of relapses in Wegener’s
granulomatosis by measurement of antineutrophil cytoplasmic antibody levels. A prospec-
tive study. Arthritis Rheum 2000;43:2025–33.
[75] Merkel PA, Lo GH, Holbrook JT, et al, for the WGET Research Group. Incidence of ve-
nous thrombotic events among patients with Wegener’s granulomatosis. Ann Intern Med

[76] Wegener F. Über eine eigenartige rhinogene granulomatose mit besonderer Beteiligung des
arteriensystems und der Nieren. Beitr Pathol 1939;102:36–40.
[77] Sebastian JK, Voetsch B, Stone JH, et al. Anti-endothelial cell antibodies in patients with
Wegener’s granulomatosis: prevalence and correlation with disease activity and manifesta-
tions. J Rheumatol 2007;34:1027–31.
[78] Sebastian JK, Voetsch B, Stone JH, et al. The frequency of anticardiolipin antibodies and
genetic mutations associated with hypercoagulability among patients with Wegener’s gran-
ulomatosis with and without a history of a thrombotic event. J Rheumatol 2007 [epub
ahead of print].
[79] Godman G, Churg J. Wegener’s granulomatosis: pathology and review of the literature.
Arch Pathol Lab Med 1954;58:533–51.
[80] Churg A. Recent advances in the diagnosis of Churg-Strauss syndrome. Mod Pathol 2001;
[81] Boin F, Sciubba JJ, Stone JH. Churg-Strauss syndrome presenting with salivary gland en-
largement and respiratory distress. Arthritis Rheum 2006;55:167–70.
[82] Guillevin L, Cohen P, Gayraud M, et al. Churg-Strauss syndrome. Clinical study and long-
term follow-up of 96 patients. Medicine (Baltimore) 1999;78:26–37.
[83] Keogh KA, Specks U. Churg-Strauss syndrome: clinical presentation, antineutrophil
cytoplasmic antibodies, and leukotriene receptor antagonists. Am J Med 2003;115:
[84] Sablé-Fourtassou R, Cohen P, Mahr A, et al. Antineutrophil cytoplasmic antibodies and
the churg-strauss syndrome. Ann Intern Med 2005;143:632–8.
[85] Wechsler ME, Finn D, Gunawardena D, et al. Churg-Strauss syndrome in patients receiv-
ing montelukast as treatment for asthma. Chest 2000;117:708–13.
[86] Pagnoux C, Guilpain P, Guillevin L. Churg-Strauss syndrome. Curr Opin Rheumatol 2007;
[87] Eschun GM, Mink SN, Sharma S. Pulmonary interstitial fibrosis as a presenting manifes-
tation in perinuclear antineutrophilic cytoplasmic antibody microscopic polyangiitis. Chest
[88] Homma S, Matsushita H, Nakata K. Pulmonary fibrosis in myeloperoxidase antineutro-
phil cytoplasmic antibody-associated vasculitides. Respirology 2004;9(2):190–6.
[89] Birnbaum J, Danoff S, Askin FB, et al. Microscopic polyangiitis presenting as a ‘pulmo-
nary-muscle’ syndrome: does alveolar hemorrhage contribute to the mechanism of pulmo-
nary fibrosis? Arthritis Rheum 2007;56(6):20–71.
[90] Guillevin L, Durand-Gasselin B, Cevallos R. Microscopic polyangiitis: clinical and labora-
tory findings in eighty-five patients. Arthritis Rheum 1999;42:421–30.
[91] Duggan CP, Westra SJ, Rosenberg AE. Case records of the Massachusetts General Hospi-
tal. Case 23–2007. A 9-year-old boy with bone pain, rash, and gingival hypertrophy. N Engl
J Med 2007;357:392–400.
[92] Bosch X, Guilabert A, Font J. Antineutrophil cytoplasmic antibodies. Lancet 2006;368:
[93] Prasad M, Buller GK, Mena CI, et al. Sum of the parts. N Engl J Med 2006;355:2468–73.
[94] Musset L, Diemert MC, Taibi F, et al. Characterization of cryoglobulins by immunoblot-
ting. Clin Chem 1992;38:798–802.
[95] Cacoub P, Poynard T, Ghillani P, et al. Extrahepatic manifestations of chronic hepatitis C.
Arthritis Rheum 1999;42:2204–12.
[96] Vassilopoulos D, Calabrese LH. Hepatitis C virus infection and vasculitis: implications of
antiviral and immunosuppressive therapies. Arthritis Rheum 2002;46:585–97.
[97] Vollertsen R, McDonald T, Younge B, et al. Cogan’s syndrome: 18 cases and a review of the
literature. Mayo Clin Proc 1986;61:344–61.
[98] Gluth MB, Baratz KH, Matteson EL, et al. Cogan syndrome: a retrospective review of 60
patients throughout a half century. Mayo Clin Proc 2006;81:483–8.

[99] Grasland A, Pouchot P, Hachulla E, et al. Typical and atypical Cogan’s syndrome: 32 cases
and review of the literature. Rheumatology 2004;43:1007–15.
[100] Calabrese LH, Dodick DW, Schwedt TJ, et al. Narrative review: reversible cerebral vaso-
constriction syndromes. Ann Intern Med 2007;146:34–44.
[101] Kural-Seyahi E, Fresko I, Seyahi N, et al. The long-term mortality and morbidity of Behcet
syndrome: a 2-decade outcome survey of 387 patients followed at a dedicated center. Med-
icine (Baltimore) 2003;82:60–76.
[102] Yazici H, Tuzlaci M, Yurdakul S, et al. A controlled survery of sacroiliitis in Behcet’s dis-
ease. Ann Rheum Dis 1981;40:558–9.
[103] Koc Y, Gullu I, Akpek G, et al. Vascular involvement in Behcet’s disease. J Rheumatol
[104] Lie JT. Vascular involvement in Behcet’s disease: arterial and venous and vessels of all sizes.
J Rheumatol 1992;19:341–3.
[105] Kahraman O. Long-term course of deep venous thrombosis in patients with Behcet’s dis-
ease. Arthritis Rheum 2003;48(Suppl):S385.
[106] Puechal X, Fiessinger JN. Thromboangiitis obliterans or Buerger’s disease: challenges for
the rheumatologist. Rheumatology 2007;46:192–9.
[107] Horton BT. The outlook in thromboangiitis obliterans. JAMA 1938;111:2184–9.
[108] Lie JT. Thromboangiitis obliterans (Buerger’s disease) in women. Medicine (Baltimore)
[109] Olin J. Thromboangiitis obliterans (Buerger’s disease). N Engl J Med 2000;343:864–70.
[110] Sterne J, Ducastaing C. Arteritis caused by cannabis indica. Arch Mal Coeur 1960;53:
[111] Combemale P, Consort T, Denis-Thelis L, et al. Cannabis arteritis. Br J Dermatol 2005;152:
[112] Disdier P, Swiader L, Jouglard J, et al. Arterite du cannabis versus maladie de leo buerger.
Discussion nosologique a propos de deux nouveaux cas. Presse Med 1999;28:71–4.
[113] Hill GL. A rational basis for management of patients with the Buerger syndrome. Br J Surg
[114] Puechal X, Fiessinger JN, Kahan A, et al. Rheumatic manifestations in patients with
thromboangiitis obliterans (Buerger’s disease). J Rheumatol 1999;26:1764–8.
[115] Hirai M, Shionoya S. Arterial obstruction of the upper limb in Buerger’s disease: its inci-
dence and primary lesion. Br J Surg 1979;66:124–8.
[116] Sasaki S, Sakuma M, Yasuda K. Current status of thromboangiitis obliterans (Buerger’s
disease) in Japan. Int J Cardiol 2000;75(Suppl 1):S175–81.
[117] Quenneville JG, Prat A, Gossard D. Subungual splinter hemorrhage: an early sigh of
thromboangiitis obliterans. Angiology 1981;32:424–32.
[118] Puechal X, Said G, Hilliquin P, et al. Peripheral neuropathy with necrotizing vasculitis in
rheumatoid arthritis. A clinicopathologic and prognostic study of thirty-two patients-
Arthritis Rheum 1995;38:1618–29.
Rheum Dis Clin N Am 33 (2007) 741–754

Targeting B Lymphocytes as Therapy

for ANCA-Associated Vasculitis
Jason M. Golbin, DO, MS, Ulrich Specks, MD*
Division of Pulmonary and Critical Care Medicine, Mayo Clinic Foundation,
200 First Street SW, Rochester, MN 55905, USA

The term anti-neutrophil cytoplasmic antibody (ANCA)–associated vascu-

litis (AAV) designates a group of three heterogeneous syndromes, Wegener’s
granulomatosis (WG), microscopic polyangiitis (MPA), and the Churg-
Strauss syndrome (CSS) [1]. Primary necrotizing small vessel vasculitis
that demonstrates a predilection for the kidneys, lungs, and peripheral ner-
vous system is the main common feature of these three disorders. WG is dis-
tinguished from MPA by the presence of necrotizing granulomatous
inflammation, which affects the upper and lower respiratory tracts and fre-
quently precedes other disease manifestations. CSS is unique in that it is
characterized by the presence of asthma and eosinophilia of tissue and pe-
ripheral blood in addition to the features of small vessel vasculitis. Because
of these commonalities, WG and MPA are frequently studied together in
clinical trials [2–4].
Most patients with these syndromes have ANCAs detectable in the serum
at the time of initial presentation [5–7]. Two types of ANCAs are of clinical
significance in patients who have vasculitis [8]. The first type of ANCA rel-
evant to systemic vasculitis is directed against proteinase 3 (PR3-ANCA),
a neutrophil granule enzyme. PR3-ANCA causes a cytoplasmic immunoflu-
orescence pattern on ethanol-fixed neutrophils. These autoantibodies occur
in more than 80% of patients with WG.
The second type of ANCA that is relevant to systemic vasculitis is di-
rected against myeloperoxidase (MPO-ANCA), another neutrophil granule
enzyme. MPO-ANCA causes a perinuclear immunofluorescence pattern on
ethanol-fixed neutrophils. MPO-ANCA is detected in less than 10% of pa-
tients with WG but occurs in the majority of patients with MPA. In addi-
tion, MPO-ANCA is more common than PR3-ANCA in patients with CSS.

* Corresponding author.
E-mail address: specks.ulrich@mayo.edu (U. Specks).

0889-857X/07/$ - see front matter Ó 2007 Elsevier Inc. All rights reserved.
doi:10.1016/j.rdc.2007.09.001 rheumatic.theclinics.com

These three syndromes respond to similar immunosuppressive treatment

regimens. The introduction of cyclophosphamide for the treatment of WG
and subsequently other types of AAV brought fundamental changes to
the prognosis of these conditions, converting them from invariably fatal dis-
orders to ones that are now usually manageable but often associated with
chronically relapsing illness [9,10]. The substantial toxicities of cyclophos-
phamide, particularly those associated with the long-term and repeated
use of the agent, have emerged as the principle challenge in the management
of these patients [11,12].
The introduction of methotrexate as the standard agent for remission in-
duction in patients with less severe disease has been the first significant result
of the search for better-tolerated alternatives to cyclophosphamide [7,13].
The second important advance has been the recognition that after 3 to 6
months of remission induction therapy, cyclophosphamide can be replaced
with methotrexate or azathioprine for long-term remission maintenance
[6,14]. To date, there is no proven alternative to cyclophosphamide for re-
mission induction in patients with severe AAV.
This review focuses on the initial results achieved with the more selective
immunosuppressive approach of B-lymphocyte depletion in patients who
fail cyclophosphamide or have contraindications for its use. This novel ap-
proach has sparked hope for patients and physicians in their search for ef-
fective, well-tolerated therapy for AAV. B-cell depletion is now undergoing
rigorous investigation in randomized clinical trials.

Effect of rituximab on B lymphocytes

Rituximab is a chimeric monoclonal antibody directed against the cell
surface protein CD20 [15]. This antibody is comprised of the variable re-
gions derived from a murine anti-CD20 antibody fused to human IgG con-
stant regions. The target antigen CD20 is a 297–amino acid phosphoprotein
with four transmembrane domains. CD20 is expressed only on cells of B-
lymphocyte lineage [16]. The molecule is thought to function as a calcium
channel subunit and to have a role in B-lymphocyte activation, prolifera-
tion, and differentiation [17].
There is no known natural ligand for CD20. Antibody binding to CD20
does not down-regulate its expression or cause shedding or internalization
of CD20 [18]. Binding of rituximab to CD20 causes death of the target
cell by a variety of mechanisms including antibody-dependent cellular cyto-
toxicity via macrophages and natural killer cells, complement-mediated
B-cell lysis, apoptosis, and sensitization to cytotoxic agents or glucocorti-
coids [19–23]. All of these mechanisms have been demonstrated in vitro,
but the exact combination of each in vivo remains unclear [24].
Circulating B lymphocytes remain undetectable in the peripheral blood
for about 6 to 12 months following treatment with rituximab [25,26];
however, it remains unclear to what degree circulating peripheral blood

B-lymphocyte numbers reflect the presence or depletion of these cells in the

bone marrow, spleen, lymph nodes, or pathologic inflammatory tissue. The
absence of CD20 on bone marrow stem cells and the earliest B-lymphocyte
precursors allows for the regeneration of naı̈ve B lymphocytes following
rituximab therapy, with the potential of profound amelioration of
Recent studies in patients with rheumatoid arthritis (RA) and systemic
lupus erythematosus (SLE) support this concept. In clinical trials, peripheral
blood memory B-lymphocyte reconstitution in patients with RA and SLE
was found to be heterogeneous among patients and substantially delayed.
The degree of delay in reconstitution seems to be related to the duration
of clinical disease response [27,28].
Although the humoral immune response is impaired after rituximab ther-
apy [29,30], serum IgG levels in general and humoral immunity to specific
antigens acquired before therapy do not seem to be affected because plasma
cells are not depleted by CD20-targeted therapy [31].

Therapeutic use of rituximab

In November 1997, rituximab was approved by the US Food and Drug
Administration (FDA) for the treatment of relapsed or refractory low-grade
or follicular B-cell non-Hodgkin’s lymphoma. This approval was based on
the results of a large-scale, multicenter trial that showed that 48% of pa-
tients with relapsed low-grade or follicular lymphoma responded to rituxi-
mab [32]. Over the course of the last decade, rituximab has been used in
more than 500,000 patients with hematologic malignancies [33].
The use of rituximab for autoimmune disease began when a patient expe-
rienced dramatic improvement of generalized severe inflammatory arthritis
after receiving rituximab therapy for lymphoma [34]. This clinical observa-
tion invigorated investigations focusing on the role of B lymphocytes in the
pathogenesis of autoimmune disease and led to a plethora of formal clinical
trials assessing the efficacy and safety of B-lymphocyte depletion as therapy
for RA, SLE, and a variety of other autoimmune diseases. In 2006, rituxi-
mab was approved by the FDA for the treatment of RA refractory to inhib-
itors of tumor necrosis factor (TNF).

Rituximab therapy for Wegener’s granulomatosis

and microscopic polyangiitis
The first use of rituximab in WG was reported in 2001 [35]. A 66-year-old
man had initially been diagnosed with WG in 1994. During subsequent
treatment of disease flares, he experienced cyclophosphamide toxicity pre-
cluding its repeated use. Prednisone in combination with azathioprine or
mycophenolate mofetil failed to restore remission. The decision was made
to treat the patient on a compassionate-use basis with four weekly doses

of rituximab at 375 mg/m2 accompanied by a prednisone taper. This first pa-

tient was treated under the hypothesis that B-lymphocyte depletion would
result in rapid removal of potentially pathogenic ANCA. By the time of
the fourth infusion, ANCAs were undetectable in the serum, and 2 months
later, the patient was in remission. Nine months after completion of the first
course of rituximab when the patient’s ANCA titer began to rise, he was
treated with another course of rituximab (this time without prednisone).
He remained in remission and did not experience significant adverse events.
To date, this patient has received six courses of rituximab, each triggered by
a rise in his serum ANCA titer, and remains in remission.
This successful treatment of the index patient with refractory WG led to
the compassionate use of rituximab in an additional ten patients with severe
refractory AAV [36]. All patients had either active severe disease not con-
trolled by cyclophosphamide or contraindications to its use. The patients
were treated with rituximab (four weekly courses of 375 mg/m2) in combi-
nation with prednisone. The majority of these patients also received intrave-
nous methylprednisolone at the start of their therapy, and three received
plasma exchange immediately before rituximab therapy to control life- or
organ-threatening disease manifestations.
In this compassionate-use study, clinical remission was accompanied by
a significant decrease in ANCA levels in all patients, and glucocorticoids
were successfully tapered in all patients over the course of 5 months. Two
patients experienced minor flares, one at 7 months and the other at 12
months. These flares responded to retreatment with rituximab and predni-
sone. Three patients were retreated prophylactically (with rituximab alone)
when their ANCA titers demonstrated a consistent pattern of increase. All
three of these patients remained in remission. Adverse effects were minimal
and mostly related to the first infusion.
As a result of this favorable experience, a prospective open-label pilot
trial of rituximab in conjunction with an oral glucocorticoid tapering regi-
men based on a strict protocol was conducted [26]. Ten patients with active
severe refractory AAV were treated with four weekly doses of rituximab
(375 mg/m2). As in the compassionate-use trial, all patients achieved remis-
sion by 3 months. One patient who had a relapse following the return of pe-
ripheral blood B lymphocytes and ANCA was retreated successfully with
a second course of prednisone and rituximab. Five patients were retreated
with rituximab alone in the presence of rising ANCA levels and remained
in remission. Adverse effects were minimal. Two patients experienced herpes
zoster infections, but both had received other immunosuppressive agents
before inclusion in the trial, and herpes zoster eruptions are a recognized
problem for WG patients during their treatment courses [37].
Several other case reports and small case series of AAV patients who re-
ceived treatment with rituximab have been published over the last couple of
years (Table 1). Erikkson and colleagues [38] described the use of rituximab
in nine patients with cyclophosphamide-resistant or frequently relapsing
Table 1
Rituximab use in anti-neutrophil cytoplasmic antibody-associated vasculitis: review of the published literature
Number of RIT doses (number
patients, of patients  number Concomitant Follow-up Number
Series type Reference diagnosis of doses  amount) drugs Remission (months) of relapses
Case series Keogh [36] 10, WG 11 – 4  375 mg/m2 11, prednisone 11, complete 14, median 2
1, MPA
Prospective Keogh [26] 10, WG 10 – 4  375 mg/m2 10, prednisone 10, complete 12, median 1
Prospective Erikkson [38] 7, WG 6 – 4  500 mg 2, CYC 8, complete 19, median 2
2, MPA 2 – 2  500 mg 4, MMF 1, partial


1 – 4  375 mg/m2 1, MMF þ SIR
1, AZA
1, none
Case series Omdal [40] 3, WG 2 – 4  375 mg/m2 Unknown 3, complete 12, median 3
1 – 1  375 mg/m2
Case series Aries [46] 8, WG 8 – 4  375 mg/m2 q4 wks 5, CYC 2, complete 18, median 0
2, MTX 1, partial
1, MMF 5, failed
Case series Gottenberg [39] 2, WG 1 – 3  375 mg/m2 1, MP 1, complete 16.5, median 0
1 – 15  375 mg/m2 1, MTX þ MMF 1, failed
Prospective long-term Stasi [51] 8, WG 10 – 4  375 mg/m2 Prednisone 9, complete 33.5, median 3
2, MPA 1, partial
Prospective long-term Smith [52] 5, WG 11 – 4  375 mg/m2 11, CYC (1 dose) 9, complete 23, median 5
5, MPA 1, partial
1, CSS 1, failed
Case series Brihaye [47] 8, WG 7 – 4  375 mg/m2 4, MTX 3, complete Unknown 1
121g 3, MMF 3, partial
1, AZA 2, failed
5, prednisone
Case series Henes [54] 6, WG 6 – 4  375 mg/m2 4, prednisolone 5, complete 16, mean 1
5, leflunomide 1, partial

Abbreviations: AZA, azathioprine; CYC, cyclophosphamide; MMF, mycophenolate mofetil; MP, methylprednisolone; MTX, methotrexate; RIT, ritux-
imab; SIR, sirolimus.

AAV. In that series, rituximab doses varied, with the majority of patients
receiving four weekly doses of 500 mg (substantially lower than doses of
375 mg/m2). In the interest of preventing the formation of human anti-chi-
meric antibodies to rituximab, eight of the nine patients were given concom-
itant treatment with other immunosuppressive medications. By 6 months,
eight of the patients had achieved remission, and the ninth was in partial re-
mission. Two patients sustained minor relapses after 1 year and were treated
successfully with glucocorticoids alone. No serious adverse effects of ritux-
imab therapy were noted.
A report from France on the use of rituximab in 43 patients with systemic
autoimmune diseases included two patients with WG [39]. Both patients had
failed treatment with cyclophosphamide and other immunosuppressive
agents. Only one of the WG patients was reported to have achieved remis-
sion; the nature of the persistent disease activity in the other was not stated.
Another report of successful remission induction in three patients with re-
fractory WG comes from Norway [40]. This report contains no details about
other immunosuppressive medications used concomitantly. Successive re-
lapses were treated with rituximab again, with varying doses. Rituximab ap-
peared to have a greater effect on the inflammatory aspects of WG, whereas
granulomatous lesions (particularly those in the orbits and sinuses) ap-
peared more resistant to treatment.
Several other reports of patients refractory to standard therapies have
been published since 2005, all describing successful treatment of AAV
with rituximab [41–45]. Some patients in these reports had unusual clinical
manifestations, including scleritis, cerebral vasculitis, endobronchial disease,
orbital pseudotumor, pachymeningitis, and foot drop, in addition to more
typical ear, nose and throat, and lung and kidney disease involvement.
Some reports suggest that rituximab may be less effective for certain dis-
ease manifestations of refractory AAV. Aries and colleagues published
a study of eight patients with active refractory WG despite treatment with
cyclophosphamide and TNF blockade [46]. These patients had predomi-
nantly granulomatous lesions, five of which were characterized by orbital
pseudotumors. The rituximab dosing regimen these patients received dif-
fered from all previous reports. The patients received 375 mg/m2 per infu-
sion, but instead of receiving four weekly infusions they received four
monthly infusions. Despite this substantially lower dose application, two pa-
tients achieved complete remission and one a partial remission. Orbital
pseudotumors remained unchanged in three patients and progressed in two.
In this study, which involved monthly rather than weekly intervals of rit-
uximab administration, peripheral blood B lymphocytes became undetect-
able in all patients but ANCA levels declined in only two. The lack of
decline in ANCA levels, as well as the lack of clinical response, proffers
the question as to whether the dosing regimen used in this study was suffi-
cient to allow for remission induction when compared with the standard
lymphoma dosing regimen of four weekly infusions of 375 mg/m2, or

whether, as suggested by the authors, granulomatous inflammation of WG

is more resistant to rituximab therapy when compared with more vasculitic
disease manifestations.
Brihaye and colleagues [47] reported on eight patients with refractory or
relapsing WG treated with four weekly doses of rituximab at 375 mg/m2. At
6 months, three patients were in complete remission, three were considered
to be in partial remission with persistent granulomatous lung nodules, and
two did not respond to treatment. The two patients who were non-re-
sponders had orbital pseudotumors that decreased only slightly in size; how-
ever, pain relief was obtained. In the three patients with partial remission
due to persistent lung nodules, the nodules eventually disappeared over
a prolonged period of time extending out to 24 months.
Additional data presented at the 13th International Vasculitis & ANCA
Workshop suggest that two 1000-mg doses of rituximab administered 2
weeks apart might be as effective in patients (n ¼ 4) with predominantly
granulomatous disease [48] as the standard lymphoma regimen (n ¼ 8) [49].
In addition, Sanchez-Cano and colleagues [50] described a 35-year-old
woman with WG who had granulomatous inflammation causing central dia-
betes insipidus that was resistant to conventional treatment but responsive to
subsequent rituximab use at the lymphoma dose.
At least some reports indicate that not all types of granulomatous lesions
in WG respond equally well to B-cell depletion. Interpretation of the reports
is complicated by variations in the dosing regimens used. It is conceivable
that some types of granulomatous inflammation in WG, particularly orbital
pseudotumors, are refractory to B-cell depletion. If such refractoriness ex-
ists, the lesions may respond more fully to lengthier periods of B-cell deple-
tion. Additional studies of this question are required.
In contrast to remission induction, only a few reports have focused on the
long-term results of rituximab use in AAV and on remission maintenance.
Stasi and colleagues [51] described 10 patients (8 with WG, 2 with MPA)
who received the lymphoma dosing regimen for refractory disease or for dis-
ease requiring multiple agents for control. All patients received oral predni-
sone at 2 mg/kg/d followed by a dose taper with clinical improvement,
similar to the original reports from the Mayo Clinic. The median follow-up
was 33.5 months, and all patients were in complete remission at 6 months,
except for one in partial remission. Three patients sustained clinical relapse
and were successfully retreated with rituximab following the same dosing
schedule, and all responded again. ANCA levels decreased significantly in
all patients, and all three relapses were preceded by elevations in ANCA
levels. Adverse effects were minimal and mainly consisted of first-infusion
Another report with a median follow-up of 23 months included 11 pa-
tients with active or refractory AAV (5 with WG, 5 with MPA, and 1
with CSS) and 11 patients with SLE [52]. Patients received the lymphoma
dosing regimen of rituximab, but concurrent with the first infusion 500 mg

of intravenous cyclophosphamide was administered. Nine of the 11 AAV

patients achieved complete response, 1 had a partial response, and 1 WG
patient failed treatment and experienced progressive nasal, sinus, and en-
dobronchial disease requiring treatment with another agent. Five relapses
were observed, and all were successfully retreated with rituximab. In this
series, the median time to peripheral B-lymphocyte return was 9 months.
The authors’ long-term experience over a mean of 35 months of follow-
up in 28 patients who received at least two repeated courses of rituximab for
refractory chronically relapsing WG suggests that long-term B-lymphocyte
depletion is a promising novel approach to remission maintenance associ-
ated with few adverse events [53].
Most recently, Henes and colleagues [54] reported on six patients with re-
fractory or relapsing WG who received rituximab for remission induction.
All patients received the lymphoma dosing regimen in conjunction with a ta-
pering oral glucocorticoid regimen. Five patients were in complete remission
at 6 months, and one had achieved partial remission. Five of the six patients
subsequently were given leflunomide (20 mg orally daily) to enhance remis-
sion maintenance over a mean follow-up of 16 months. In this series, leflu-
nomide was well tolerated, and only one patient had a relapse, which
responded promptly to retreatment with rituximab.

Rituximab therapy for Churg-Strauss syndrome

Four cases of CSS treated with rituximab have been reported to date
[52,55,56]. The first case was included among the 11 patients with AAV
treated by Smith and colleagues [52]. That patient achieved remission 1
month after treatment with rituximab and relapsed at 6 months following
the return of peripheral blood B lymphocytes. Remission was restored 2
months following retreatment, and another relapse occurred 7 months later,
again after the return of B cells. The patient underwent a third rituximab
course and at the time of this writing has been in remission 12 months after
the last treatment.
Kasushik and colleagues [55] described a 49-year-old man with refractory
disease who received only three (instead of four) doses of rituximab at the
lymphoma dosing regimen. Nonetheless, he responded well and achieved
clinical remission.
Koukoulaki and colleagues [56] reported two cases of successful rituxi-
mab use in CSS. The first patient was a 37-year-old woman who initially re-
sponded to intravenous cyclophosphamide but subsequently had multiple
relapses. She responded well to the lymphoma regimen and, with relapses
of eosinophilia, nasal symptoms, and asthma at 7 and 16 months, was re-
treated successfully with rituximab. The second patient was a 35-year-old
woman who was treated with two infusions of 1000 mg 2 weeks apart.
She responded well, and after B-cell recovery at 9 months, remained in
remission until the time of publication.

These encouraging initial results in patients with CSS provide the ratio-
nale for an ongoing formal prospective pilot trial designed to assess the
safety and efficacy of rituximab in CSS.

Safety of rituximab therapy for anti-neutrophil cytoplasmic

antibody–associated vasculitis
Rituximab appears to be well tolerated by most patients. Infusion-related
adverse events are typically limited to the first infusion and occur in as many
as 10% of subjects. These reactions are usually mild, resolve with interrup-
tion of the infusion or slowing of the infusion rate, and rarely preclude com-
pletion of the infusions. The concomitant use of high-dose glucocorticoids
given during remission induction therapy for AAV may help prevent
some rituximab-related reactions. Reactivation of latent viral infections
has been reported in patients treated with rituximab [57–60]; however,
even its repeated use does not seem to portend a higher risk for infections
than that associated with other more commonly used immunosuppressive
agents. Definitive conclusions about the safety of rituximab in AAV cannot
be derived from the existing literature because the number of reported cases
is too small to detect a potential increased frequency of rare adverse events.
An ongoing prospective, multicenter, double-blind, placebo-controlled
phase II/III trial of rituximab for AAV (RAVE trial, www.clinicaltrials.
gov) is being conducted in 200 patients and is designed to determine whether
rituximab can replace cyclophosphamide for remission induction in AAV.
This trial will provide more conclusive insights into the safety of this prom-
ising agent.

Other B lymphocyte-targeting approaches

B-cell activating factor inhibition
B-cell activating factor (BAFF) is a 285–amino acid member of the TNF
family. It is also called BLyS, which stands for B lymphocyte stimulator.
BAFF has a crucial role in B-cell development, survival of B cells, and im-
munoglobulin production [61]. Recently, BAFF has received attention in the
field of autoimmune diseases for a number of reasons. First, overexpression
of BAFF in mice induces an SLE-like phenotype [61]. Second, BAFF-block-
ing agents are successful treatment agents in several animal models of auto-
immunity [62]. Third, elevated serum levels have been detected in various
rheumatologic diseases including WG [63–65]. A crucial biologic property
of BAFF is its ability to inhibit B-cell apoptosis and its function as a B-lym-
phocyte survival factor [66]. In particular, peripheral autoantigen-binding B
lymphocytes seem to be dependent on the presence of BAFF for their sur-
vival, supporting the hypothesis that elevated BAFF levels promote the
development of autoimmune disease [67].

Moreover, recent clinical observations in patients treated with rituximab

are consistent with the concept of BAFF expression having a significant role
in the regulation of B-lymphocyte homeostasis. In RA, circulating BAFF
levels rise quickly following rituximab infusion and B-cell depletion, and
they remain elevated until B cells return in the peripheral blood, at which
time BAFF levels return back to baseline [68]. Similarly, in patients with
Sjögren’s syndrome, serum levels of BAFF were found to be inversely cor-
related with the duration of rituximab-induced B-cell depletion; the higher
the BAFF levels, the shorter the duration of B-cell depletion [69].
Consequently, BAFF inhibitors may be used in future clinical applica-
tions with the goal being to ‘‘starve B lymphocytes to death’’ [66] or in at-
tempts to prolong the duration of B-lymphocyte depletion induced by
agents like rituximab.

Epratuzumab is a monoclonal antibody that targets the cell surface anti-
gen CD22. This agent has been shown to be effective against non-Hodgkin’s
lymphoma as well as SLE and Sjögren’s syndrome [70–73]. Epratuzumab
appears to have a different mechanism of action than rituximab, acting
more as an immunomodulatory agent as opposed to the cytotoxic activity
of rituximab [74]. An open-label pilot trial of 14 patients with active SLE
suggested efficacy, with greater than 70% of patients experiencing more
than a 50% improvement in disease scores with minimal adverse effects
[73]. This agent may also merit further investigation in AAV.

Despite its well-recognized toxicity, cyclophosphamide remains the stan-
dard agent for remission induction in patients with severe AAV. A significant
pathogenic role of B lymphocytes beyond their responsibility for autoanti-
body production is emerging in autoimmune diseases, including AAV. Prelim-
inary data indicate that B-lymphocyte depletion with rituximab is an effective
and safe mechanism-based treatment approach for remission induction and
remission maintenance in refractory AAV. Even repeated courses of rituxi-
mab therapy seem to be associated with surprisingly few infections. Whether
B-cell depletion can replace cyclophosphamide as first-line therapy for newly
diagnosed patients with AAV remains under investigation.

[1] Jennette JC, Falk RJ, Andrassy K, et al. Nomenclature of systemic vasculitides: proposal of
an international consensus conference. Arthritis Rheum 1994;37(2):187–92.
[2] Jayne D, Rasmussen N, Andrassy K, et al. A randomized trial of maintenance therapy for
vasculitis associated with antineutrophil cytoplasmic autoantibodies. N Engl J Med 2003;

[3] De Groot K, Rasmussen N, Bacon PA, et al. Randomized trial of cyclophosphamide versus
methotrexate for induction of remission in early systemic antineutrophil cytoplasmic anti-
body-associated vasculitis. Arthritis Rheum 2005;52(8):2461–9.
[4] Jayne DR, Gaskin G, Rasmussen N, et al. Randomized trial of plasma exchange or high-dos-
age methylprednisolone as adjunctive therapy for severe renal vasculitis. J Am Soc Nephrol
[5] Guillevin L, Durand-Gasselin B, Cevallos R, et al. Microscopic polyangiitis: clinical and lab-
oratory findings in eighty-five patients. Arthritis Rheum 1999;42(3):421–30.
[6] Keogh KA, Specks U. Churg-Strauss syndrome: clinical presentation, antineutrophil cyto-
plasmic antibodies, and leukotriene receptor antagonists. Am J Med 2003;115(4):284–90.
[7] Finkielman JD, Lee AS, Hummel AM, et al. ANCA are detectable in nearly all patients with
active severe Wegener’s granulomatosis. Am J Med 2007;120(7):643.
[8] Hoffman GS, Specks U. Anti-neutrophil cytoplasmic antibodies. Arthritis Rheum 1998;41:
[9] Walton EW. Giant-cell granuloma of the respiratory tract (Wegener’s granulomatosis).
Br Med J 1958;2:265–70.
[10] Fauci AS, Haynes BF, Katz P, et al. Wegener’s granulomatosis: prospective clinical and ther-
apeutic experience with 85 patients for 21 years. Ann Intern Med 1983;98:76–85.
[11] Hoffman GS, Kerr GS, Leavitt RY, et al. Wegener granulomatosis: an analysis of 158 pa-
tients. Ann Intern Med 1992;116:488–98.
[12] Talar-Williams C, Hijazi YM, Walther MM, et al. Cyclophosphamide-induced cystitis and
bladder cancer in patients with Wegener granulomatosis. Ann Intern Med 1996;124:477–84.
[13] Sneller MC, Hoffman GW, Talar-Williams C, et al. An analysis of forty-two Wegener’s gran-
ulomatosis patients treated with methotrexate and prednisone. Arthritis Rheum 1995;38:
[14] Langford CA, Talar-Williams C, Barron KS, et al. A staged approach to the treatment of
Wegener’s granulomatosis: induction of remission with glucocorticoids and daily cyclophos-
phamide switching to methotrexate for remission maintenance. Arthritis Rheum 1999;
[15] Grillo-Lopez AJ, White CA, Dallaire BK, et al. Rituximab: the first monoclonal antibody
approved for the treatment of lymphoma. Curr Pharm Biotechnol 2000;1(1):1–9.
[16] Einfeld DA, Brown JP, Valentine MA, et al. Molecular cloning of the human B cell CD20
receptor predicts a hydrophobic protein with multiple transmembrane domains. EMBO J
[17] Tedder TF, Engel P. CD20: a regulator of cell-cycle progression of B lymphocytes. Immunol
Today 1994;15(9):450–4.
[18] Silverman GJ, Weisman S. Rituximab therapy and autoimmune disorders: prospects for
anti-B cell therapy. Arthritis Rheum 2003;48(6):1484–92.
[19] Demidem A, Lam T, Alas S, et al. Chimeric anti-CD20 (IDEC-C2B8) monoclonal antibody
sensitizes a B cell lymphoma cell line to cell killing by cytotoxic drugs. Cancer Biother Radio-
pharm 1997;12(3):177–86.
[20] Golay J, Zaffaroni L, Vaccari T, et al. Biologic response of B lymphoma cells to anti-CD20
monoclonal antibody rituximab in vitro: CD55 and CD59 regulate complement-mediated
cell lysis. Blood 2000;95(12):3900–8.
[21] Shan D, Ledbetter JA, Press OW. Signaling events involved in anti–CD20-induced apoptosis
of malignant human B cells. Cancer Immunol Immunother 2000;48(12):673–83.
[22] Johnson P, Glennie M. The mechanisms of action of rituximab in the elimination of tumor
cells. Semin Oncol 2003;30(1 Suppl 2):3–8.
[23] Uchida J, Hamaguchi Y, Oliver JA, et al. The innate mononuclear phagocyte network
depletes B lymphocytes through Fc receptor-dependent mechanisms during anti-CD20
antibody immunotherapy. J Exp Med 2004;199(12):1659–69.
[24] Eisenberg R, Looney RJ. The therapeutic potential of anti-CD20: ‘‘what do B-cells do?’’.
Clin Immunol 2005;117(3):207–13.

[25] Anolik JH, Barnard J, Cappione A, et al. Rituximab improves peripheral B cell abnormal-
ities in human systemic lupus erythematosus. Arthritis Rheum 2004;50(11):3580–90.
[26] Keogh KA, Ytterberg SR, Fervenza FC, et al. Rituximab for refractory Wegener’s granulo-
matosis: report of a prospective, open-label pilot trial. Am J Respir Crit Care Med 2006;
[27] Roll P, Palanichamy A, Kneitz C, et al. Regeneration of B cell subsets after transient B cell
depletion using anti-CD20 antibodies in rheumatoid arthritis. Arthritis Rheum 2006;54(8):
[28] Anolik JH, Barnard J, Owen T, et al. Delayed memory B cell recovery in peripheral blood
and lymphoid tissue in systemic lupus erythematosus after B cell depletion therapy. Arthritis
Rheum 2007;56(9):3044–56.
[29] Gonzalez-Stawinski GV, Yu PB, Love SD, et al. Hapten-induced primary and memory hu-
moral responses are inhibited by the infusion of anti-CD20 monoclonal antibody (IDEC-
C2B8, Rituximab). Clin Immunol 2001;98(2):175–9.
[30] van der Kolk LE, Baars JW, Prins MH, et al. Rituximab treatment results in impaired sec-
ondary humoral immune responsiveness. Blood 2002;100(6):2257–9.
[31] Browning JL. B cells move to centre stage: novel opportunities for autoimmune disease treat-
ment. Nat Rev Drug Discov 2006;5(7):564–76.
[32] McLaughlin P, Grillo-Lopez AJ, Link BK, et al. Rituximab chimeric anti-CD20 monoclonal
antibody therapy for relapsed indolent lymphoma: half of patients respond to a four-dose
treatment program. J Clin Oncol 1998;16(8):2825–33.
[33] Salama AD, Pusey CD. Drug insight: rituximab in renal disease and transplantation. Nat
Clin Pract Nephrol 2006;2(4):221–30.
[34] Protheroe A, Edwards JC, Simmons A, et al. Remission of inflammatory arthropathy in as-
sociation with anti-CD20 therapy for non-Hodgkin’s lymphoma. Rheumatology (Oxford)
[35] Specks U, Fervenza FC, McDonald TJ, et al. Response of Wegener’s granulomatosis to anti-
CD20 chimeric monoclonal antibody therapy. Arthritis Rheum 2001;44(12):2836–40.
[36] Keogh KA, Wylam ME, Stone JH, et al. Induction of remission by B lymphocyte depletion
in eleven patients with refractory antineutrophil cytoplasmic antibody-associated vasculitis.
Arthritis Rheum 2005;52(1):262–8.
[37] Wung PK, Holbrook JT, Hoffman GS, et al. Herpes zoster in immunocompromised
patients: incidence, timing, and risk factors. Am J Med 2005;118(12):1416.
[38] Erikkson P. Nine patients with anti-neutrophil cytoplasmic antibody-positive vasculitis suc-
cessfully treated with rituximab. J Intern Med 2005;257:540–8.
[39] Gottenberg JE, Guillevin L, Lambotte O, et al. Tolerance and short term efficacy of rituxi-
mab in 43 patients with systemic autoimmune diseases. Ann Rheum Dis 2005;64(6):913–20.
[40] Omdal R, Wildhagen K, Hansen T, et al. Anti-CD20 therapy of treatment-resistant Wegener’s
granulomatosis: favourable but temporary response. Scand J Rheumatol 2005;34(3):229–32.
[41] Ferraro AJ, Day CJ, Drayson MT, et al. Effective therapeutic use of rituximab in refractory
Wegener’s granulomatosis. Nephrol Dial Transplant 2005;20(3):622–65.
[42] Cheung CM, Murray PI, Savage CO. Successful treatment of Wegener’s granulomatosis as-
sociated scleritis with rituximab. Br J Ophthalmol 2005;89(11):1542.
[43] Memet B, Rudinskaya A, Krebs T, et al. Wegener granulomatosis with massive intracerebral
hemorrhage: remission of disease in response to rituximab. J Clin Rheumatol 2005;11(6):
[44] Maher LV, Wilson JG. Successful treatment of rheumatoid vasculitis–associated foot drop
with rituximab. Rheumatology (Oxford) 2006;45(11):1450–1.
[45] Tamura N, Matsudaira R, Hirashima M, et al. Two cases of refractory Wegener’s granulo-
matosis successfully treated with rituximab. Intern Med 2007;46(7):409–14.
[46] Aries PM, Hellmich B, Voswinkel J, et al. Lack of efficacy of rituximab in Wegener’s gran-
ulomatosis with refractory granulomatous manifestations. Ann Rheum Dis 2006;65(7):

[47] Brihaye B, Aouba A, Pagnoux C, et al. Adjunction of rituximab to steroids and immunosup-
pressants for refractory/relapsing Wegener’s granulomatosis: a study on 8 patients. Clin Exp
Rheumatol 2007;25(1 Suppl 44):S23–7.
[48] Flores-Suarez L. Rituximab is useful in Wegener granulomatosis patients with predominant
granulomatous disease. Clin Exp Rheumatol 2007;25(1 Suppl 44):S113.
[49] Seo P, Specks U, Keogh KA. Efficacy of rituximab in limited Wegener’s granulomatosis with
refractory granulomatous manifestations. Clin Exp Rheumatol 2007;25(1 Suppl 44):S111.
[50] Sanchez-Cano D, Callejas-Rubio JL, Rios-Fernandez R, et al. Use of rituximab in Wege-
ner’s granulomatosis: comment on the article by Wong. Nephrol Dial Transplant 2007;
22(3):958–9 [author reply: 959].
[51] Stasi R, Pagano A, Stipa E, et al. Rituximab chimeric anti-CD20 monoclonal antibody
treatment for adults with chronic idiopathic thrombocytopenic purpura. Blood 2001;
[52] Smith KG, Jones RB, Burns SM, et al. Long-term comparison of rituximab treatment for
refractory systemic lupus erythematosus and vasculitis: remission, relapse, and re-treatment.
Arthritis Rheum 2006;54(9):2970–82.
[53] Golbin JM, Keogh KA, Fervenza FC, et al. Update on rituximab use in patients with chron-
ically relapsing Wegener’s granulomatosis. Clin Exp Rheumatol 2007;25(1 Suppl 44):S111.
[54] Henes JC, Fritz J, Koch S, et al. Rituximab for treatment-resistant extensive Wegener’s gran-
ulomatosis: additive effects of a maintenance treatment with leflunomide. Clin Rheumatol
2007, in press.
[55] Kaushik VV, Reddy HV, Bucknall RC. Successful use of rituximab in a patient with recal-
citrant Churg-Strauss syndrome. Ann Rheum Dis 2006;65(8):1116–7.
[56] Koukoulaki M, Smith KG, Jayne DR. Rituximab in Churg-Strauss syndrome. Ann Rheum
Dis 2006;65(4):557–9.
[57] Dai MS, Chao TY, Kao WY, et al. Delayed hepatitis B virus reactivation after cessation of
preemptive lamivudine in lymphoma patients treated with rituximab plus CHOP. Ann Hem-
atol 2004;83(12):769–74.
[58] Tsutsumi Y, Kawamura T, Saitoh S, et al. Hepatitis B virus reactivation in a case of
non-Hodgkin’s lymphoma treated with chemotherapy and rituximab: necessity of prophy-
laxis for hepatitis B virus reactivation in rituximab therapy. Leuk Lymphoma 2004;45(3):
[59] Freim Wahl SG, Folvik MR, Torp SH. Progressive multifocal leukoencephalopathy in
a lymphoma patient with complete remission after treatment with cytostatics and rituximab:
case report and review of the literature. Clin Neuropathol 2007;26(2):68–73.
[60] Garcia-Suarez J, de Miguel D, Krsnik I, et al. Changes in the natural history of progressive
multifocal leukoencephalopathy in HIV-negative lymphoproliferative disorders: impact of
novel therapies. Am J Hematol 2005;80(4):271–81.
[61] Mackay F, Schneider P, Rennert P, et al. BAFF and APRIL: a tutorial on B cell survival.
Annu Rev Immunol 2003;21:231–64.
[62] Baker KP. BLySdan essential survival factor for B cells: basic biology, links to pathology
and therapeutic target. Autoimmun Rev 2004;3(5):368–75.
[63] Cheema GS, Roschke V, Hilbert DM, et al. Elevated serum B lymphocyte stimulator levels
in patients with systemic immune-based rheumatic diseases. Arthritis Rheum 2001;44(6):
[64] Mariette X, Roux S, Zhang J, et al. The level of BLyS (BAFF) correlates with the titre of
autoantibodies in human Sjogren’s syndrome. Ann Rheum Dis 2003;62(2):168–71.
[65] Krumbholz M, Specks U, Wick M, et al. BAFF is elevated in serum of patients with Wege-
ner’s granulomatosis. J Autoimmun 2005;25(4):298–302.
[66] Stohl W, Looney RJ. B cell depletion therapy in systemic rheumatic diseases: different
strokes for different folks? Clin Immunol 2006;121(1):1–12.
[67] Lesley R, Xu Y, Kalled SL, et al. Reduced competitiveness of autoantigen-engaged B cells
due to increased dependence on BAFF. Immunity 2004;20(4):441–53.

[68] Cambridge G, Stohl W, Leandro MJ, et al. Circulating levels of B lymphocyte stimulator in
patients with rheumatoid arthritis following rituximab treatment: relationships with B cell
depletion, circulating antibodies, and clinical relapse. Arthritis Rheum 2006;54(3):723–32.
[69] Pers JO, Devauchelle V, Daridon C, et al. BAFF-modulated repopulation of B lymphocytes
in the blood and salivary glands of rituximab-treated patients with Sjogren’s syndrome.
Arthritis Rheum 2007;56(5):1464–77.
[70] Leonard JP, Coleman M, Ketas JC, et al. Epratuzumab, a humanized anti-CD22 antibody,
in aggressive non-Hodgkin’s lymphoma: phase I/II clinical trial results. Clin Cancer Res
[71] Leonard JP, Coleman M, Ketas J, et al. Combination antibody therapy with epratuzumab
and rituximab in relapsed or refractory non-Hodgkin’s lymphoma. J Clin Oncol 2005;
[72] Steinfeld SD, Tant L, Burmester GR, et al. Epratuzumab (humanized anti-CD22 antibody)
in primary Sjogren’s syndrome: an open-label phase I/II study. Arthritis Res Ther 2006;8(4):
[73] Dorner T, Kaufmann J, Wegener WA, et al. Initial clinical trial of epratuzumab (humanized
anti-CD22 antibody) for immunotherapy of systemic lupus erythematosus. Arthritis Res
Ther 2006;8(3):R74.
[74] Carnahan J, Stein R, Qu Z, et al. Epratuzumab, a CD22-targeting recombinant humanized
antibody with a different mode of action from rituximab. Mol Immunol 2007;44(6):1331–41.
Rheum Dis Clin N Am 33 (2007) 755–775

Airway Involvement in Wegener’s

Vlassis S. Polychronopoulos, MDa,b,
Udaya B.S. Prakash, MDb, Jason M. Golbin, DO, MSb,
Eric S. Edell, MDb, Ulrich Specks, MDb,*
Third Pulmonary Department, Sismanaglion General Hospital, Athens, Greece
Division of Pulmonary and Critical Care Medicine, Mayo Clinic,
Rochester, MN 55905, USA

Wegener’s granulomatosis is characterized by necrotizing granulomatous

inflammation and necrotizing vasculitis affecting predominantly small
arteries, arterioles, capillaries, and venules [1–4]. Larger vessels are rarely
affected. The disease has a predilection for the upper respiratory tract, lungs,
and kidneys, but any organ can be involved. The spectrum of disease ranges
from limited disease affecting the nasal and paranasal tract to aggressive
multisystem inflammation leading to multiple organ damage and failure
[5,6]. Immunosuppressive therapy has improved the prognosis from being
universally fatal with a median survival of 5 months to a treatable, chroni-
cally relapsing disease with a median survival of 21.7 years (95% confidence
interval, 15.6–27.86 years) after diagnosis [6,7]. The etiology of the disease
remains unknown. Its response to immunosuppressive therapy and the pres-
ence of antineutrophil cytoplasmic antibodies (ANCA) in most patients sug-
gest autoimmune mechanisms of pathogenesis [8].
In contrast to the well-described pulmonary parenchymal involvement of
Wegener’s granulomatosis, the lower airway (tracheobronchial) disease
manifestations are less well recognized by clinicians. Consequently, mild dis-
ease of the airways is easily missed. Airway involvement usually occurs in
conjunction with disease manifestations elsewhere, but occasionally it may
be the only manifestation of the disease [9]. Early in the course of the disease
symptoms from active inflammatory lesions of the airways may be absent
or very subtle. Frequently, they are insidious in onset and sometimes

* Corresponding author.
E-mail address: specks.ulrich@mayo.edu (U. Specks).

0889-857X/07/$ - see front matter Ó 2007 Elsevier Inc. All rights reserved.
doi:10.1016/j.rdc.2007.09.004 rheumatic.theclinics.com

misdiagnosed as asthma or other common airway diseases. This may result

in significant delays in arriving at the correct diagnosis [10]. Tracheobron-
chial disease manifestations may also occur after remission has been
achieved with appropriate immunosuppressive therapy, and the airway dis-
ease may proceed to airway scarring and stenosis [11].
There is a relative paucity of published information on various tracheo-
bronchial manifestations of Wegener’s granulomatosis. Many of the reports
are based on single cases or small case series. Since the publication of the
authors’ experience with tracheobronchial manifestations of Wegener’s
granulomatosis in 51 patients [9], additional insights on this topic have
been gained at their institution and other medical centers. This article pro-
vides a comprehensive review of the diagnosis and management of the infra-
glottic tracheobronchial disease manifestations.

Clinical presentation and differential diagnosis

The ease with which the diagnosis of Wegener’s granulomatosis is made
depends on the clinical presentation. When the patient presents with typical
and well-defined features of Wegener’s granulomatosis and corroborating
blood test abnormalities including positive ANCA serology, the diagnosis
is readily apparent. The difficulty arises when clinical features are subtle
or nonspecific, and several key elements in the diagnosis are either absent
or not easily discernible.
Tracheobronchial involvement is usually associated with disease affecting
the supraglottic structures, pulmonary parenchyma, and other organ sys-
tems [9,11,12]. Airway involvement is found in 15% to 55% of patients
with Wegener’s granulomatosis [9,11,13,14]. On rare occasions, however,
airway involvement may be the only or the presenting feature of Wegener’s
granulomatosis. In such cases a positive ANCA test result can point to the
right diagnosis, but ANCA are undetectable in about 20% of patients with
Wegener’s granulomatosis limited to the respiratory tract [15].
Patients younger than 30 years of age are more prone to develop airway
involvement. In one report on 158 patients with Wegener’s granulomatosis,
23 patients were younger than 20 years of age, and 11 of them (approxi-
mately 50%) developed upper tracheal involvement [16]. Women also are
more likely to develop airway disease than men [13,17]. One report has sug-
gested that pregnancy may contribute both to the relapse of Wegener’s gran-
ulomatosis and the occurrence of tracheal stenosis [18].
The common symptoms caused by tracheobronchial Wegener’s granulo-
matosis include hoarseness, cough, hemoptysis, dyspnea, stridor, and
wheezing [9,11,19]. Hoarseness and cough are not as frequent as dyspnea
and wheezing, the former symptoms being reported in less than 10% of pa-
tients with airway stenosis caused by Wegener’s granulomatosis [9]. Hemop-
tysis is a common symptom in Wegener’s granulomatosis and has several
anatomic sources including cavitated pulmonary parenchymal lesions,

alveolar hemorrhage, bronchiectasis, and expectoration of blood aspirated

from supraglottic airways, the last two being uncommon sources of hemop-
tysis in Wegener’s granulomatosis.
Progressive respiratory difficulty with stridor can be the first manifesta-
tion of airway involvement and may be sudden or progressive. Subglottic
stenosis is the most common cause of stridor in Wegener’s granulomatosis.
Conversely, the differential diagnosis in any patient with stridor should
include Wegener’s granulomatosis of the upper airways. When stridor is
the first manifestation of the disease, differential diagnoses include foreign
body aspiration, trauma, acute paralysis of vocal cords, papillomatosis,
and malignancy.
If cough, wheezing, and dyspnea are the main symptoms considerable de-
lay in diagnosing Wegener’s granulomatosis may result from patients being
misdiagnosed as asthmatics. Involvement of nasal passages and paranasal
sinuses is common in Wegener’s granulomatosis. The resultant symptoms
may also be erroneously attributed to asthma complex. In a large cohort
(N ¼ 701) of patients with Wegener’s granulomatosis, up to 15% of patients
indicated that their first symptoms had initially been attributed to asthma
[10]. Substantial delays (sometimes years) in the diagnosis have been
reported in patients in whom airway symptoms were not accompanied by
other more severe disease manifestations, such as glomerulonephritis
[10,20]. For these reasons, it is essential for physicians to consider Wegener’s
granulomatosis in patients with airway symptoms who are initially thought
to have asthma, but are not promptly responding to optimal asthma therapy
[11,17]. Moreover, it is important to reiterate that slowly progressive airway
disease occurs in a significant number of patients with documented Wege-
ner’s granulomatosis who have been optimally treated with immunosuppres-
sive agents and are considered to be in remission [21].

Physiologic features
Pulmonary function testing is an integral part of the comprehensive
approach to the diagnosis and management of airway disease caused by
Wegener’s granulomatosis [9,11,19]. Large airway involvement (trachea
and main stem bronchi) is best assessed with inspiratory and expiratory
flow-volume tracings. They are helpful in determining both the severity of
the tracheobronchial stenosis and in assessing the therapeutic success.
Abnormal flow-volume tracings have led to the diagnosis of airway disease
caused by Wegener’s granulomatosis in patients previously treated for pre-
sumed asthma [11]. Narrowing or stenosis of intrathoracic and extrathoracic
airways shows typical flattening of the inspiratory and expiratory flow-vol-
ume tracings (Fig. 1). In patients with chronic and recurrent airway stenosis
flow-volume tracings obtained before and after interventional procedures
help better to manage the airway problem (Fig. 2).

Fig. 1. Flow-volume tracings in a patient with significant midtracheal stenosis (inset) show typ-
ical features of a fixed major airway obstruction with flattening of both the inspiratory and
expiratory flow-volume curves. Interrupted lines show predicted normal tracing.

In addition to the flow-volume tracings, spirometric analysis of flow

rates, lung volumes, and other parameters of lung function help determine
the presence or absence of obstructive airways disease. Obstructive disease
of smaller airways occurs in patients with Wegener’s granulomatosis. The
mechanisms responsible include diffuse or localized areas of bronchial
luminal compromise, bronchiectasis, peribronchial scarring, and broncho-
malacia. When smaller airways (lobar, segmental, or subsegmental) are
involved, many physiologic abnormalities typical of chronic obstructive
pulmonary disease are seen. A study of 22 patients with Wegener’s granu-
lomatosis found that forced expiratory volume in 1 second and maximal ex-
piratory flow were reduced in more than 50% of patients with airway
involvement [22]. Pulmonary function testing in another group of 41 pa-
tients observed that 24% of patients had pulmonary symptoms and the
same percent of patients exhibited reduced diffusing capacity for carbon
monoxide, whereas forced expiratory volume in 1 second was reduced in
only 10% [23]. Reductions in lung volumes and diffusing capacity for car-
bon monoxide are more likely to be caused by pulmonary parenchymal
Pulmonary function and flow-volume tracings are not uniformly abnor-
mal in all patients with airway disease caused by Wegener’s granulomatosis.
Pulmonary function test abnormalities and their severity depend on the ex-
tent of compromise of the airway lumina. Particularly early in the disease
process, when active inflammation has not yet caused significant scarring
or stenosis, even severe tracheobronchial inflammation may be reflected
only in subtle abnormalities of the flow-volume tracing. Furthermore,

Fig. 2. Selected flow-volume tracings and corresponding bronchoscopic findings obtained at

different intervals in a 55-year-old man with Wegener’s granulomatosis complicated by severe
stenosis of left main bronchus (LMB), and mild stenoses of right main bronchus and subglottic
trachea, who required 14 bronchoscopy procedures during a 4-year period for symptomatic
relief of dyspnea. Interrupted lines show predicted normal tracing. YAG, yttrium-aluminum-
garnet laser; BD, balloon dilatation.

with mild degrees of tracheal stenosis or main stem bronchial narrowing

flow-volume tracings may appear near-normal.

Imaging studies
Chest radiographs combined with CT of the chest demonstrate abnor-
malities during the course of the disease in 85% of patients with Wegener’s
granulomatosis. The abnormal findings include lung nodules (often cavi-
tated); consolidation; ground glass opacification; and less often pleural dis-
ease, mediastinal lymphadenopathy, and interstitial processes [24]. Even
though airway involvement occurs in approximately 50% of patients with
Wegener’s granulomatosis [14], these findings are less frequently appreciated
radiographically [12,24,25]. Large airway involvement by Wegener’s granu-
lomatosis may consist of focal or elongated segments of stenosis, and intra-
luminal and extraluminal soft tissue masses or thickening with or without
lobar or segmental atelectasis [24,26]. In patients with significant airway dis-
ease, airway narrowing as depicted by CT can be extensive. Other findings

include calcification and thickening of the tracheal rings [24]. In one report
on 30 patients with Wegener’s granulomatosis, large airways were found to
be abnormal in nine (30%) patients, and bronchial wall thickening in the
segmental and subsegmental bronchi was discovered in 22 (73%) patients
[27]. The airway lesions showed complete or partial improvement with treat-
ment. In this series of patients, bronchiectasis was found in four (13%)
patients [27]. Bronchial abnormalities, including bronchiectasis and peri-
bronchial thickening of the small airways, have been reported in approxi-
mately 40% of cases [28–30].
Virtual bronchoscopic imaging derived from CT technique has been used
to create three-dimensional projections of airway abnormalities in patients
with Wegener’s granulomatosis (Fig. 3). In a study of 11 patients with
Wegener’s granulomatosis, 32 (80%) of 40 stenoses were detected by virtual
bronchoscopy [31]. Even if the resolution of this novel CT technique
improves further in the future, it is unlikely to replace the visual assessment
of the mucosa by fiberoptic bronchoscopy in the diagnosis of airway lesions
associated with Wegener’s granulomatosis, and it does not allow for sam-
pling of diagnostic specimens.
Both linear tomography and MRI of the trachea have been used in the
accurate assessment of airway involvement [16,32]. MRI of upper airways
may demonstrate the diffuse thickening of tracheobronchial submucosal tis-
sues and the luminal narrowing, but such abnormalities are nonspecific and
can be seen in other disorders, such as postintubation trauma, amyloidosis,
sarcoidosis, tracheobronchopathia, relapsing polychondritis, and infectious
tracheitis [32].

Fig. 3. Virtual bronchoscopic imaging derived from CT technique in a patient with Wegener’s
granulomatosis shows significant stenosis of the right upper lobe bronchus (downward arrow)
and bronchus intermedius (upward arrow).

The imaging procedures are helpful in the overall assessment of pulmo-

nary involvement by Wegener’s granulomatosis. The clinical applicability
of imaging in the management of airway involvement is limited.

Endoscopic visualization of the airways remains the major diagnostic
procedure in the evaluation, diagnosis, and management of airway disease
in Wegener’s granulomatosis. The various airway abnormalities are listed
in Table 1, and Table 2 summarizes the most important case series reports
focusing on airway pathology in Wegener’s granulomatosis. The supraglot-
tic airways can be assessed by examining the nasal passages and obtaining
tissue biopsies for histologic documentation of the disease. Laryngoscopy
permits detailed examination of oropharynx, nasopharynx, epiglottic, glot-
tis, and larynx. Bronchoscopy is the most important diagnostic procedure in
the assessment of location and the degree of airway involvement below the
vocal cords, the nature of luminal abnormalities, and the feasibility of bron-
choscopic therapy to restore functional airway patency. Bronchoscopy is
also used for diagnostic tissue sampling, and it is valuable in the evaluation
of the efficacy of both pharmacologic and interventional therapy of the air-
way involvement in Wegener’s granulomatosis.
The symptoms related to airway disease from Wegener’s granulomatosis
may remain undiagnosed or misdiagnosed until clinical suspicion prompts
bronchoscopy. Similarly, a cursory bronchoscopy can easily overlook subtle
changes in the airways, particularly in the subglottic trachea. The bronchos-
copist should be cognizant of the possibility of Wegener’s granulomatosis
affecting the tracheobronchial tree and carefully examine the airway lumen
and describe the abnormalities. If suspicious mucosal abnormalities are
detected, adequate tissue samples should be obtained for histologic docu-
mentation of the diagnosis. A significant number of patients with known
and treated airway disease from Wegener’s granulomatosis require repeated,
periodic, bronchoscopies for the re-evaluation of airway lumen and to detect
localized disease recurrence. Ideally, the findings should be recorded in the
report and by photographic documentation.
The following paragraphs provide more detailed information regarding
the types of airway involvement, bronchoscopic findings, and management
options to treat symptomatic airway involvement.

Types of airway involvement

Airway involvement is present in 15% to 55% of patients with Wegener’s
granulomatosis [11,13,14,33–37]. Because the airway manifestations in this
disease are varied, it is difficult to classify them in any specific manner.
The following discussions encompass the common and uncommon airway

Table 1
Types of airway abnormalities described in Wegener’s granulomatosis
Supraglottic airways
Nasal passages Mucosal ulceration
Hemorrhagic mucosal lesions
Crusting of mucosa
Nose Saddle nose deformity
Sinuses Sinusitis
Ears Otitis media
Oropharynx Gingivitis
Ulceration and hemorrhage
Epiglottis Ulceration and hemorrhage
Glottis and larynx Edema
Ulceration and hemorrhage
Nodules and masses
Trachea and bronchi Mucosal edema and erythema
Ulceration and hemorrhage
Cobble-stone mucosa
Nodule and mass (pseudotumor)
Polyps and pseudopolyps
Submucosal tunnels
Luminal synechial bands
Membrane formation
Mucosal lattice
Stenotic (transluminal)
Stenosis (subglottic)a
Simple (localized)
Complex (multiple)
Stenosis (tracheobronchial)a
Simple (localized)
Complex (multiple)
Necrosis of airway
Cartilaginous deformities
Tracheoesophageal fistula
Large airway stenosis can be reversible or fixed.

abnormalities encountered in this disease (see Tables 1 and 2). The type and
severity of symptoms are dependent on a variety of factors, including the
specific nature of the lesion, the extent of airway luminal involvement and
its location, and the sequelae airway obstruction.

Mucosal abnormalities
In Wegener’s granulomatosis, mucosal abnormalities can occur any-
where in the tracheobronchial tree; they are usually patchy in distribution
and may occur as isolated lesions. The most common airway abnormality
Table 2
Selected series of patients with Wegener’s granulomatosis with bronchoscopic findings
Author No. Age Gender of Lesions:
(Ref #) Patients (y) M/F N (%) Bronchoscopic findings (N) Airway therapy (N) Outcome (N)
Daum [9] 51 59 31/20 TB: 25 (49) Ulcerating TB-itis (18) Pseudotumor removal (1) Repeat bronchoscopy
for symptoms (7)


SGS: 5 (9.8) TB stenosis (4) YAG laser (1) New infiltrates (2)
Alveolar hemorrhage (2) Dil þ Stent (2)
Dil alone (3)
Gluth [11] 27 40 11/16 SGS: 27 (100) Mature scar þ SGS (20) Tracheostomy (11) Persistent tracheostomy (3)
Friable SGS (7) LTR (3) Multiple procedures (13)
Partial tracheal resection (1)
CO2 laser (12)
CS injection (11)
Utzig [46] 15 49 7/6 SGS: 7 (47) Smooth, firm SGS (7) Stent (1) N/A
TB: 1 (7) Lower tracheal granuloma (1) Dil (3)
Dil þ injection (3)
SGS resection (1)
Hoffman [23] 21 39 5/16 SGS: 21 (100) SGS (21) ILCD (64) Mean 2.4 procedures: initial
Tx group (12)
Mean 4.1 procedures: prior
Tx group (9)
McDonald [19] 108 NA 54/54 SGS: 17 (16) Upper tracheal stricture (17) Tracheostomy (9) Decannulation (4)
Cordier [14] 77 46 39/38 TB: 41 (55) TB stenosis (13) N/A N/A
Pseudotumor (7)
Inflammatory lesion (10)
Alveolar hemorrhage (10)
Purulent secretions (1)
Abbreviations: CO2, carbon dioxide; CS, corticosteroids; Dil, dilatation; ILCD, intralesional corticosteroids and dilatation; LTR, laryngotracheal recon-
struction; N/A, information not available; SGS, subglottic stenosis; TB, tracheobronchial; Tx, treatment; YAG, yttrium-aluminum-garnet laser.


in Wegener’s granulomatosis consists of mucosal edema, erythema, thick-

ening, and granularity of mucosal surface. Shallow mucosal ulcers may
be interspersed throughout the tracheobronchial tree. Ulcerated lesions
are responsible for hemoptysis, but the extent of hemoptysis is not as
voluminous as sometimes associated with cavitated lesions and alveolar
hemorrhage. In patients with ulcerating tracheobronchial lesions, the
inflammation is often associated with the narrowing of tracheal or bron-
chial lumen [9]. Cobble-stoning of the mucosa is also the result of active
inflammation. It is very likely that prolonged mucosal inflammation fol-
lowed by fibrosis is responsible for the eventual formation of bronchial
The authors have observed unusual mucosal abnormalities, such as small
shallow diverticula in the trachea and bronchi (Fig. 4, middle panel). An-
other peculiar abnormality is the formation of submucosal tunnels in the
trachea and main stem bronchi [38]. Bronchoscopy reveals a pin hole–like
proximal entrance to the submucosal tunnel, and when a small diameter
probe is passed, the probe emerges from distal pin-hole opening of the tun-
nel (Fig. 5, left and middle panels). The authors theorize that these tunnels
result from excessive mucosal pseudomembrane formation, which becomes
incorporated into the normal mucosa by epithelialization. They have also
observed formation of thick synechial bands across the bronchial lumen
(see Fig. 5, right panel). These bands and mucosal tunnels may represent
the same process. Occasionally, mucosal plaques of yellow-green hue are
seen [39]. Purulent secretions and excessive mucous production in the tra-
cheobronchial tree is also common in active airway disease caused by Wege-
ner’s granulomatosis.
Given the similarities of the respiratory epithelium, it is not surprising
that the inflammatory processes affecting the tracheobronchial tree have
many similarities with the well-described abnormalities of the nasal mucosa

Fig. 4. Left panel: Bronchoscopic appearance of subglottic stenosis in a patient with Wegener’s
granulomatosis of 3 years duration. Circumferential stenosis is typical. Middle panel: Severe
circumferential stenosis of midtrachea with mucosal pits and shallow diverticula and mucosal
inflammation in Wegener’s granulomatosis. Right panel: Proximal tracheal involvement in
Wegener’s granulomatosis demonstrating formation of pseudomembrane. Removal of the
pseudomembrane may reveal shallow ulcerations.

Fig. 5. Submucosal tunnel in the trachea of a patient with Wegener’s granulomatosis. The left
panel reveals a pin-hole like proximal entrance to the submucosal tunnel, and the middle panel
shows a probe passed through the aperture emerging from the distal opening of the tunnel.
Right panel: Thick synechial band across the bronchial lumen in a patient with Wegener’s

in Wegener’s granulomatosis. Furthermore, many of the mucosal abnormal-

ities described previously are not specific for Wegener’s granulomatosis.
Similar findings can be encountered in airway involvement in acute infec-
tious or noninfectious tracheobronchitis, submucosal spread of malignancy,
endobronchial sarcoidosis, tracheobronchial amyloidosis, and fungal

Masses and polyps

Wegener’s granulomatosis can cause tracheal or bronchial mass lesions,
which have also been referred to as ‘‘inflammatory pseudotumor’’ [9].
Such lesions mimic malignancy and call for optimal biopsy specimens to es-
tablish Wegener’s granulomatosis as the cause [9]. Case reports of tracheal
mass protruding into the airway lumen have been described [40]. Polypoid
lesions and mucosal projections into the airway lumen also occur and mimic
malignancy (Fig. 6). These lesions may ulcerate and cause hemoptysis. The

Fig. 6. Left panel: A polypoid lesion in the distal main bronchus in a patient with Wegener’s
granulomatosis. These lesions are friable and usually reveal granulation tissue on histologic
analysis. Middle panel: Distal tracheal mass-like lesion (inflammatory pseudotumor) caused
by Wegener’s granulomatosis. Right panel: The airway obstruction required rigid broncho-
scopic insertion of a silicone stent to provide symptomatic relief.

tracheobronchial luminal masses can be present at the time of presentation

of the disease or at the time of relapse [41]. They are generally a reflection of
active disease. On rare occasions, rapidly proliferative tissue growth may
lead to acute laryngotracheal airway obstruction and death [42]. Biopsy of
these lesions may show characteristic histopathologic features of Wegener’s
granulomatosis, various degrees of fibrosis, or granulation tissue [36,42].

Subglottic stenosis
Subglottic stenosis is the most frequent airway manifestation in Wege-
ner’s granulomatosis (see Fig. 4, left panel). Depending on the acuity of on-
set and the degree of airway compromise, it can be life threatening.
Subglottic stenosis may occur at initial presentation or later in the course
of the disease. When the airway obstruction is caused by fibrotic scarring
(damage occurring in the process of healing inflammation) rather than by
active inflammation, it develops independently of other features of Wege-
ner’s granulomatosis and is unresponsive to systemic immunosuppressive
therapy [43].
The reported frequency of subglottic stenosis has ranged from 8.5% to
50% of patients [44,45]. The numbers derived from reports focusing on air-
way manifestations of Wegener’s granulomatosis are difficult to interpret
because of highly variable selection bias. For instance, in the Mayo Clinic
series of 51 patients with Wegener’s granulomatosis, subglottic stenosis
was found in 10%, but only patients who had undergone bronchoscopy
were included, thus not including those subglottic stenosis detected on lar-
yngoscopy [9]. In two other studies subglottic stenosis, tracheobronchitis,
and tracheal or bronchial narrowing were identified at bronchoscopy in
55% and 59% of patients, respectively [14,25]. In a group of 108 patients
with Wegener’s granulomatosis, direct laryngoscopy in 17 patients showed
subglottic stenosis in all [19]. In another group of 15 patients with Wegener’s
granulomatosis, eight had tracheal involvement and all but one showed in-
volvement of the subglottic area [46].
Information derived from large cohorts of patients with Wegener’s gran-
ulomatosis does not provide more clarity. Among the 158 patients with We-
gener’s granulomatosis seen at the National Institute of Health, 25 (16%)
had subglottic stenosis, but the degree of inflammatory activity is not spec-
ified [16]. Among the 180 patients enrolled with newly diagnosed or relaps-
ing disease into the Wegener’s granulomatosis etanercept trial, 8.3% had
documented active subglottic inflammation at the time of enrollment [47].
A subsequent analysis of damage in this cohort found that 17.8% had sub-
glottic stenosis scored as a damage item at completion of the trial, but it is
not stated how many of the patients enrolled with a relapse already had sub-
glottic stenosis as a damage item scored at enrollment [48].
Based on individual personal observations and the interpretation of the
literature the following concept emerges. Subglottic (tracheobronchial)

inflammation is a common feature at the time of active disease. At this stage

erythematous, edematous, and friable mucosa and ulcerations can be de-
tected, and the lesions are typically circumferential. In most cases symptoms
are mild or absent because the vocal cords are usually not affected, and the
airway is not compromised. Consequently, the early inflammatory lesions
may remain undetected, unless specifically looked for. If the inflammation
is not treated and if fibrotic scarring occurs during the resolution of in-
flammation, significant compromise of the airway lumen develops, causing
stridor and dyspnea (see Fig. 4, left panel). In the absence of persistent
active inflammation, the development of subglottic stenosis should not be
misinterpreted to indicate failure of immunosuppressive therapy. Serial pul-
monary function tests and inspiratory and expiratory flow-volume tracings
may aid in follow-up of patients with subglottic inflammation and stenosis
(see Fig. 2).
Subglottic involvement can occasionally be an isolated finding [49]. In
such patients the attribution of this lesion to underlying Wegener’s granulo-
matosis may only be possible with a diagnostic biopsy specimen or a positive
ANCA test result. In the absence of evidence for Wegener’s granulomatosis
in other organs, other causes of isolated subglottic stenosis should be ex-
cluded. These include postintubation stenosis, idiopathic subglottic stenosis,
postinfectious stenosis, hypocomplementemic urticarial vasculitis, and
extrinsic compression [50,51].

Tracheobronchial stenosis
Similar inflammatory processes described for the subglottic region can
occur at any level of the tracheobronchial tree and lead to stenoses. These
can be simple (a localized tracheal segmental stenosis) or complex (multiple
tracheal segmental stenoses with luminal distortion). Most of the tracheal
and bronchial stenoses are circumferential, and the mucosa in the stenotic
segment and just above and below reveals mucosal thickening (see Fig. 4,
middle panel). In acute and active disease, the stenotic area may exhibit
erythematous and edematous mucosa, with formation of pseudomembrane
and ulceration (see Fig. 4, right panel). Tracheal stenosis produces symp-
toms similar to those in patients with subglottic stenosis. If stenosis is
accompanied by significant mucosal inflammation and ulceration, hemopty-
sis occurs. Excessive instrumentation and biopsy may lead to hemorrhage,
but this is usually easily controlled.
In isolated cases of tracheal stenosis caused by Wegener’s granulomato-
sis, the following causes of focal stenosis should be excluded: postintubation
stenosis; postinfectious stenosis; posttransplantation stenosis; and various
systemic diseases that may also involve the airways and lead to focal stenosis
including Crohn’s disease, sarcoidosis, and Behçet’s syndrome. In the
rare event of diffuse tracheal involvement, other diseases more likely to
cause diffuse tracheal disease, such as relapsing polychondritis,

tracheobronchopathia osteochondroplastica, tracheobronchial amyloidosis,

and papillomatosis, need to be excluded [39].
Bronchial inflammation and stenoses are less frequent than subglottic
disease, and almost always associated with disease activity elsewhere [52].
Isolated bronchial stenosis, however, has been described [53]. Stenosis in-
volving the main stem bronchi can be mild to severe [54]. The stenoses
can be simple or complex, and typically vary in their severity from one
bronchus to another in the same lobe (Fig. 7, upper panels). Rarely, in-
flammation and necrosis can be so severe that pneumonectomy is required,
such as in a case in which extensive necrotizing and granulomatous bron-
chial inflammation with vasculitis of the bronchial arteries and the pulmo-
nary vein caused complete loss of unilateral lung perfusion and ventilation

Fig. 7. Upper panels: Wegener’s granulomatosis leading to localized stenosis of the bronchus to
the right middle lobe. The left upper panel shows the stenotic bronchus at the top and normal
bronchial lumen to lower lobe below. The right upper panel is a close of the stenotic bronchus.
Left lower panel: Wegener’s granulomatosis complicated by a pin-hole stenosis of the bronchus
to the superior segment of the right lower lobe (right-center). Mucosal erythema is evident.
Right lower panel: Irreversibly stenotic and occluded segmental bronchi in a patient with
chronic Wegener’s granulomatosis.

Symptoms of patients with bronchial inflammation or stenosis may in-

clude cough; wheezing (often localized or asymmetric); and dyspnea (if
the mainstem bronchi or multiple segments are affected). As in patients
with tracheal stenosis, hemoptysis is more likely in acute inflammatory dis-
ease with significant mucosal involvement, ulceration, and nodular or polyp-
oid changes.
Lesions occurring in the lobar or segmental bronchi lead to concentric
narrowing of the lumen, which becomes progressively more severe (see
Fig. 7). In patients with chronic and active Wegener’s granulomatosis
involving the bronchial tree, bronchoscopy may demonstrate early stages
of the stenotic process in the form of thin semitranslucent membranes
covering the bronchial lumen. Most of the severely narrowed bronchi
become irreversibly occluded from fibrotic scarring (see Fig. 7, right lower

Other airway abnormalities

Wegener’s granulomatosis affecting the airways occasionally leads to tra-
cheomalacia or bronchomalacia [9]. Even though CT imaging has shown
bronchiectasis in patients with Wegener’s granulomatosis, significant clini-
cal features of bronchiectasis are described only occasionally [56,57]. An
unusual case of Wegener’s granulomatosis presenting as paratracheal medi-
astinal lesion with tracheal wall invasion, which responded dramatically to
corticosteroid treatment, has been described [58]. Cases of right middle lobe
syndrome and acquired tracheoesophageal fistula caused by Wegener’s
granulomatosis have also been described [59,60].

In a report on pathologic features in 87 open lung biopsies from 67
patients with Wegener’s granulomatosis, the following were detected in
the airways: bronchial or bronchiolar lesions including acute and chronic
bronchiolitis (51% and 64%); follicular bronchiolitis (28%); and bronchio-
litis obliterans (31%). Bronchial stenosis was found in three specimens [36].
In contrast to open-lung biopsy, bronchoscopic tissue sampling of tracheo-
bronchial luminal abnormalities reveals typical histologic features of Wege-
ner’s granulomatosis only in a minority of cases. In previously reported
experience, biopsies of tracheobronchial mucosa in 17 patients with Wege-
ner’s granulomatosis showed changes diagnostic of the disease in only three
patients with an additional seven patients showing features that were inter-
preted as supportive of the diagnosis in the given clinical context [9]. Simi-
larly, biopsy of mass-like lesions and polypoid lesions in the airways of
patients with active disease may reveal typical features of Wegener’s granu-
lomatosis. In many cases, however, the tissue analysis reveals granulation
tissue or nonspecific inflammation.

Treatment of airway disease in Wegener’s granulomatosis can be broadly
classified into medical and nonmedical types.
Glucocorticoids (systemic, topical, intralesional)
Cytotoxic agents
Airway debridement (laser, electrocautery, and so forth)
Airway dilatation
Stent insertion
Resection or reanastomosis of stenotic segments
Many patients respond well to appropriate medical therapy, with resolu-
tion of the airway disease. In a significant number of patients, however,
tracheobronchial stenosis occurs, persists, or recurs, despite prolonged im-
munosuppressive therapy. This group of patients requires nonmedical (in-
terventional) approaches for the relief of their symptoms.

The general principles of medical therapy for Wegener’s granulomatosis
apply also to patients who have tracheobronchial involvement. The severity
of disease governs the choice of the initial remission induction regimen. If the
disease severity is judged to be life threatening or putting an affected organ at
risk for irreversible damage, glucocorticoids in combination with cyclophos-
phamide remains the treatment of choice. In less severe cases, methotrexate
is the preferred alternative to cyclophosphamide. Successful remission in-
duction is followed by less aggressive remission maintenance therapy with
either methotrexate or azathioprine. Regardless of the severity of other
organ manifestations, severe tracheobronchial disease should initially be
treated with the combination of oral glucocorticoids and cyclophosphamide.
Methotrexate should only be considered as remission induction agent if there
is no compromise of airway lumen, and the bronchoscopically apparent
‘‘inflammatory burden’’ is mild or affected isolated segmental bronchi.
In addition, in the authors’ practice patients with documented tracheo-
bronchial disease are treated with high-dose inhaled glucocorticoids, such
as fluticasone, 440 to 880 mg twice daily. Inhaled glucocorticoid therapy is
usually initiated when the oral glucocorticoid dose has been tapered to daily
doses less than 30 mg. The role of intralesional injection of long-acting
glucocorticoids is described later.

The management of subglottic disease in Wegener’s granulomatosis
remains a therapeutic dilemma. This complication often occurs indepen-
dently of other features of active Wegener’s granulomatosis and is fre-
quently unresponsive to systemic immunosuppressive therapy. Almost all
nonmedical therapies consist of invasive procedures aimed at mechanically
relieving the airway obstruction. The types of treatments have included in-
tralesional injection of corticosteroids; balloon dilatation; mechanical sub-
glottic dilation (bouginage); laser ablation; intraluminal stent placement;
tracheostomy; and surgical resection and reanastomosis [12,19,43,44,
61–64]. Some of these procedures must be repeated to provide symptomatic
relief. In patients with critical airway stenosis that is unresponsive to medi-
cal and dilatational therapies, tracheostomy provides long-term relief. In
one series of 27 patients with Wegener’s granulomatosis and subglottic ste-
nosis, 11 (41%) underwent tracheotomy and 13 (48%) required multiple
surgical procedures [11].
Tracheostomy has been recommended as the first-line surgical treatment
in the presence of airway obstruction [17]. This recommendation should be
tempered with assessment of the severity of respiratory distress. It has been
suggested that airway manipulation be minimal or avoided during disease
activity because this may aggravate upper airway obstruction [11]. If clinical
evaluation does not indicate the need for urgent tracheostomy, appropriate
immunosuppressive therapy first should be administered. Induction of re-
mission with glucocorticoids and cytotoxic agents, however, does not always
prevent the progression of subglottic stenosis. Patients with known active
subglottic inflammation need to be observed and followed carefully.
Intralesional long-acting corticosteroid injection combined with luminal
dilatation has been used to treat subglottic stenosis caused by Wegener’s
granulomatosis [13,44,46,65]. Among 21 patients who underwent 64 proce-
dures, methylprednisolone acetate was injected directly into the stenotic
segment, followed by microsurgical lysis of the stenotic ring and serial
dilatation with Maloney bougies or Fogarty catheter balloon [13]. The
procedure was repeated at a later date if restenosis occurred. The mean
follow-up was 40.6 months. In patients without luminal scarring from pre-
vious procedures, a mean of 2.4 procedures were required at mean intervals
of 11.6 months to maintain subglottic patency; complications included two
cases of pneumothorax [13]. More recently, some laryngologists have also
locally injected mitomycin C because of its antifibrotic properties, but the
outcomes of this approach have not been formally assessed.
Bronchoscopic intervention to treat airway involvement in Wegener’s
granulomatosis is an important therapeutic option for patients with signif-
icant or progressive airway symptoms. Bronchoscopic applications include
mechanical debulking of endoluminal masses and tissue, rigid broncho-
scopic dilatation of major airway stenosis, flexible bronchoscopic balloon

dilatation laser ablation, and endoluminal stent insertion (see Figs. 2 and 6,
right panel) [9]. Successful bronchoscopic dilatation of airway stenosis
caused by Wegener’s granulomatosis has been reported in a small number
of patients [66]. Many patients require repeated therapeutic interventions
because of recurrent airway stenosis. Of all the bronchoscopic interventions,
silicone stents seem to provide the most long-lasting symptomatic relief [9].
Insertion of these stents requires expertise in rigid bronchoscopy. Because
there is no consensus regarding the optimal type of bronchoscopic therapy
for the airway disease caused by Wegener’s granulomatosis, each case
should be evaluated on an individual basis and suitable therapy selected.
Ideally, manipulations of the airways in patients with Wegener’s granuloma-
tosis should only be performed at specialized centers with extensive experi-
ence with both the disease and interventional bronchoscopy.
Extensive surgical options are not recommended in patients with active
disease. In appropriate cases, resection and anastomosis have been success-
fully performed [46,62]. Among 158 patients with Wegener’s granulomatosis
seen at the National Institutes of Health, 25 (16%) had subglottic stenosis
and their symptoms varied from minimal to life-threatening. Sixteen
(80%) of 20 patients with fixed subglottic stenosis required surgical inter-
vention, including manual dilations, carbon-dioxide laser resections, and
laryngotracheoplasty with and without microvascular reconstruction [16].
Following resection of the stenotic lesion, disease reactivation can occur
in the remaining subglottis and may lead to dehiscence of anastomotic sites
and recurrent stenosis [19]. In unusual cases of severe airway involvement,
such as bronchial necrosis, treatment has required pneumonectomy [55,67].
Anecdotal patients with Wegener’s granulomatosis have been treated
with external beam radiation therapy for severe airway inflammation
[34,68]. This form of treatment is not routinely used, however, and cannot
be recommended.

Wegener’s granulomatosis affects the tracheobronchial tree in more than
half of the patients with the disease. The airway involvement can be an iso-
lated presentation, part of the fully developed multisystem disease, or
a chronic complication of progressive disease or disease in remission. Symp-
toms are dependent on the degree of airway narrowing and may include
wheezing, dyspnea, cough, and hemoptysis. The severity can vary from
mild wheezing and dyspnea to life-threatening stridor. Clinical features, pul-
monary function test, imaging studies, and bronchoscopy are helpful in the
diagnosis and management of airway complications in patients with Wege-
ner’s granulomatosis. Bronchoscopic findings include mucosal inflamma-
tion, ulceration and hemorrhage, formation of pseudomembrane, mucosal
tunneling, luminal masses, simple or complex airway stenosis, and tracheo-
bronchomalacia. Therapeutic options for the treatment of airway

involvement caused by Wegener’s granulomatosis include pharmacologic

agents, bronchoscopic interventions, and surgical techniques.

[1] Wegener F. On generalised septic vessel diseases. Thorax 1987;42:918–9.
[2] Godman GC, Churg J. Wegener’s granulomatosis: pathology and review of the literature.
AMA Arch Path 1954;58:533–53.
[3] Leavitt RY, Fauci AS, Bloch DA, et al. The American College of Rheumatology 1990
criteria for the classification of Wegener’s granulomatosis. Arthritis Rheum 1990;33:1101–7.
[4] Jennette JC, Falk RJ, Andrassy K, et al. Nomenclature of systemic vasculitides: the proposal
of an international consensus conference. Arthritis Rheum 1994;37:187–92.
[5] Hoffman GS, Kerr GS, Leavitt RY, et al. Wegener granulomatosis: an analysis of 158
patients. Ann Intern Med 1992;116:488–98.
[6] Reinhold-Keller E, Beuge N, Latza U, et al. An interdisciplinary approach to the care of
patients with Wegener’s granulomatosis: long-term outcome in 155 patients. Arthritis
Rheum 2000;43(5):1021–32.
[7] Walton EW. Giant-cell granuloma of the respiratory tract (Wegener’s granulomatosis).
Br Med J 1958;2:265–70.
[8] Hewins P, Tervaert JW, Savage CO, et al. Is Wegener’s granulomatosis an autoimmune dis-
ease? Curr Opin Rheumatol 2000;12(1):3–10.
[9] Daum TE, Specks U, Colby TV, et al. Tracheobronchial involvement in Wegener’s granu-
lomatosis. Am J Respir Crit Care Med 1995;151(2 Pt 1):522–6.
[10] Abdou NI, Kullman GJ, Hoffman GS, et al. Wegener’s granulomatosis: survey of 701
patients in North America. Changes in outcome in the 1990s. J Rheumatol 2002;29(2):309–16.
[11] Gluth MB, Shinners PA, Kasperbauer JL. Subglottic stenosis associated with Wegener’s
granulomatosis. Laryngoscope 2003;113(8):1304–7.
[12] Shapshay SM, Valdez TA. Bronchoscopic management of benign stenosis. Chest Surg Clin
N Am 2001;11(4):749–68.
[13] Hoffman GS, Thomas-Golbanov CK, Chan J, et al. Treatment of subglottic stenosis, due to
Wegener’s granulomatosis, with intralesional corticosteroids and dilation. J Rheumatol
[14] Cordier JF, Valeyre D, Guillevin L, et al. Pulmonary Wegener’s granulomatosis: a clinical
and imaging study of 77 cases. Chest 1990;97(4):906–12.
[15] Lee AS, Finkielman JD, Peikert T, et al. Agreement of anti-neutrophil cytoplasmic antibody
measurements obtained from serum and plasma. Clin Exp Immunol 2006;146(1):15–20.
[16] Lebovics RS, Hoffman GS, Leavitt RY, et al. The management of subglottic stenosis in
patients with Wegener’s granulomatosis. Laryngoscope 1992;102(12 Pt 1):1341–5.
[17] Alaani A, Hogg RP, Drake Lee AB. Wegener’s granulomatosis and subglottic stenosis: man-
agement of the airway. Journal of Laryngology & Otology 2004;118(10):786–90.
[18] Pauzner R, Mayan H, Hershko E, et al. Exacerbation of Wegener’s granulomatosis during
pregnancy: report of a case with tracheal stenosis and literature review. J Rheumatol 1994;
[19] McDonald TJ, Neel HB III, DeRemee RA. Wegener’s granulomatosis of the subglottis and
the upper portion of the trachea. Ann Otol Rhinol Laryngol 1982;91(6 Pt 1):588–92.
[20] D’Cruz DP, Baguley E, Asherson RA, et al. Ear, nose, and throat symptoms in subacute
Wegener’s granulomatosis. BMJ 1989;299(6696):419–22.
[21] Boubenider SA, Akhtar M, Alfurayh O, et al. Late recurrence of Wegener’s granulomatosis
presenting as tracheal stenosis in a renal transplant patient. Clin Transplant 1994;8(1):5–9.
[22] Rosenberg DM, Weinberger SE, Fulmer JD, et al. Functional correlates of lung involvement
in Wegener’s granulomatosis: use of pulmonary function tests in staging and follow-up. Am
J Med 1980;69(3):387–94.

[23] Koldingsnes W, Jacobsen EA, Sildnes T, et al. Pulmonary function and high-resolution CT
findings five years after disease onset in patients with Wegener’s granulomatosis. Scand
J Rheumatol 2005;34(3):220–8.
[24] Sheehan RE, Flint JD, Muller NL. Computed tomography features of the thoracic manifes-
tations of Wegener granulomatosis. J Thorac Imaging 2003;18(1):34–41.
[25] Kuhlman JE, Hruban RH, Fishman EK. Wegener granulomatosis: CT features of parenchy-
mal lung disease. J Comput Assist Tomogr 1991;15(6):948–52.
[26] Aberle DR, Gamsu G, Lynch D. Thoracic manifestations of Wegener granulomatosis: diag-
nosis and course. Radiology 1990;174(3 Pt 1):703–9.
[27] Lee KS, Kim TS, Fujimoto K, et al. Thoracic manifestation of Wegener’s granulomatosis:
CT findings in 30 patients. Eur Radiol 2003;13(1):43–51.
[28] Foo SS, Weisbrod GL, Herman SJ, et al. Wegener granulomatosis presenting on CT with
atypical bronchovasocentric distribution. J Comput Assist Tomogr 1990;14(6):1004–6.
[29] Papiris SA, Manoussakis MN, Drosos AA, et al. Imaging of thoracic Wegener’s granuloma-
tosis: the computed tomographic appearance. Am J Med 1992;93(5):529–36.
[30] Maskell GF, Lockwood CM, Flower CD. Computed tomography of the lung in Wegener’s
granulomatosis. Clin Radiol 1993;48(6):377–80.
[31] Summers RM, Aggarwal NR, Sneller MC, et al. CT virtual bronchoscopy of the central air-
ways in patients with Wegener’s granulomatosis. Chest 2002;121(1):242–50.
[32] Park KJ, Bergin CJ, Harrell J. MR findings of tracheal involvement in Wegener’s granulo-
matosis. AJR Am J Roentgenol 1998;171(2):524–5.
[33] Maguire R, Fauci AS, Doppman JL, et al. Unusual radiographic features of Wegener’s gran-
ulomatosis. AJR Am J Roentgenol 1978;130(2):233–8.
[34] Eagleton LE, Rosher RB, Hawe A, et al. Radiation therapy and mechanical dilation of en-
dobronchial obstruction secondary to Wegener’s granulomatosis. Chest 1979;76(5):609–10.
[35] Hellmann D, Laing T, Petri M, et al. Wegener’s granulomatosis: isolated involvement of the
trachea and larynx. Ann Rheum Dis 1987;46(8):628–31.
[36] Travis WD, Hoffman GS, Leavitt RY, et al. Surgical pathology of the lung in Wegener’s
granulomatosis: review of 87 open lung biopsies from 67 patients. Am J Surg Pathol 1991;
[37] Amin R. Endobronchial involvement in Wegener’s granulomatosis. Postgrad Med J 1983;
[38] Prakash UBS. Bronchoscopy. In: Gold WM, Murray JF, Nadel JA, editors. Atlas of proce-
dures in respiratory medicine. Philadelphia: WB Saunders; 2002. p. 214–65.
[39] Prince JS, Duhamel DR, Levin DL, et al. Nonneoplastic lesions of the tracheobronchial wall:
radiologic findings with bronchoscopic correlation. Radiographics 2002;22:S215–30.
[40] Bohlman ME, Ensor RE, Goldman SM. Primary Wegener’s granulomatosis of the trachea:
radiologic manifestations. South Med J 1984;77(10):1318–9.
[41] Yilmaz A, Damadoglu E, Aksoy F, et al. A relapsing case of Wegener’s granulomatosis pre-
senting as an endobronchial mass. Tuberk Toraks 2006;54(1):56–60.
[42] Matt BH. Wegener’s granulomatosis, acute laryngotracheal airway obstruction and death in
a 17-year-old female: case report and review of the literature. Int J Pediatr Otorhinolaryngol
[43] Strange C, Halstead L, Baumann M, et al. Subglottic stenosis in Wegener’s granulomatosis:
development during cyclophosphamide treatment with response to carbon dioxide laser
therapy. Thorax 1990;45(4):300–1.
[44] Langford CA, Sneller MC, Hallahan CW, et al. Clinical features and therapeutic manage-
ment of subglottic stenosis in patients with Wegener’s granulomatosis. Arthritis Rheum
[45] Screaton NJ, Sivasothy P, Flower CD, et al. Tracheal involvement in Wegener’s granuloma-
tosis: evaluation using spiral CT. Clin Radiol 1998;53(11):809–15.
[46] Utzig MJ, Warzelhan J, Wertzel H, et al. Role of thoracic surgery and interventional bron-
choscopy in Wegener’s granulomatosis. Ann Thorac Surg 2002;74(6):1948–52.

[47] Stone JH. Limited versus severe Wegener’s granulomatosis: baseline data on patients in the
Wegener’s granulomatosis etanercept trial. Arthritis Rheum 2003;48(8):2299–309.
[48] Seo P, Min YI, Holbrook JT, et al. Damage caused by Wegener’s granulomatosis and its
treatment: prospective data from the Wegener’s Granulomatosis Etanercept Trial
(WGET). Arthritis Rheum 2005;52(7):2168–78.
[49] Guerrero M, Gall E, Tilahun E, et al. Wegener’s granulomatosis presenting as subglottic ste-
nosis. J Clin Rheumatol 2001;7(2):91–6.
[50] Valdez TA, Shapshay SM. Idiopathic subglottic stenosis revisited. Ann Otol Rhinol Lar-
yngol 2002;111(8):690–5.
[51] Chen HJ, Bloch KJ. Hypocomplementemic urticarial vasculitis, Jaccoud’s arthropathy, val-
vular heart disease, and reversible tracheal stenosis: a surfeit of syndromes. J Rheumatol
[52] Hervier B, Pagnoux C, Renaudin K, et al. Stenoses endobronchiques au cours de la granu-
lomatose de Wegener. Revue de Medecine Interne 2006;27(6):453–7.
[53] Hirsch MM, Houssiau FA, Collard P, et al. A rare case of bronchial stenosis in Wegener’s
granulomatosis: dramatic response to intravenous cyclophosphamide and oral cotrimoxa-
zole. J Rheumatol 1992;19(5):821–4.
[54] Stavngaard T, Mortensen J. Wegener’s granulomatosis causing main stem bronchial stenosis
and large perfusion-ventilation defects. Clin Nucl Med 2002;27(6):440–1.
[55] Schneider P, Grone J, Braun J, et al. Wegener’s granulomatosis presenting as necrosis of the
left mainstem bronchus. Asian Cardiovasc Thorac Ann 2002;10(3):277–9.
[56] Woywodt A, Goebel U. Bronchial stenosis and extensive bronchiectasis due to Wegener’s
granulomatosis. Nephron 2000;85(4):366–7.
[57] Sitara D, Hoffbrand BI. Chronic bronchial suppuration and antineutrophil cytoplasmic an-
tibody (ANCA) positive systemic vasculitis. Postgrad Med J 1990;66(778):669–71.
[58] Song YH, Kim TH, Lee IH, et al. Wegener’s granulomatosis presenting as mediastinal soft
tissue mass invading the tracheal wall. Clin Rheumatol 2000;19(6):495–8.
[59] Lee SJ, Berry GJ, Husari AW. Wegener’s granulomatosis presenting as right middle lobe ob-
struction. Chest 1993;103(5):1623–4.
[60] Conces DJ Jr, Kesler KA, Datzman M, et al. Tracheoesophageal fistula due to Wegener
granulomatosis. J Thorac Imaging 1995;10(2):126–8.
[61] Watters K, Russell J. Subglottic stenosis in Wegener’s granulomatosis and the nitinol stent
[see comment]. Laryngoscope 2003;113(12):2222–4.
[62] Herridge MS, Pearson FG, Downey GP. Subglottic stenosis complicating Wegener’s gran-
ulomatosis: surgical repair as a viable treatment option. J Thorac Cardiovasc Surg 1996;
[63] Eliachar I, Chan J, Akst L. New approaches to the management of subglottic stenosis in
Wegener’s granulomatosis. Cleve Clin J Med 2002;69:SII149–51.
[64] Sheski FD, Mathur PN. Long-term results of fiberoptic bronchoscopic balloon dilation in
the management of benign tracheobronchial stenosis. Chest 1998;114(3):796–800.
[65] Stappaerts I, Van Laer C, Deschepper K, et al. Endoscopic management of severe subglottic
stenosis in Wegener’s granulomatosis. Clin Rheumatol 2000;19(4):315–7.
[66] Mayse ML, Greenheck J, Friedman M, et al. Successful bronchoscopic balloon dilation of
nonmalignant tracheobronchial obstruction without fluoroscopy. Chest 2004;126(2):634–7.
[67] Chandran PK, First MR, Weiss MA, et al. Wegener’s granulomatosis: prolonged pa-
tient survival after pneumonectomy and renal transplantation. Am J Nephrol 1982;
[68] Neviani CB, Carvalho Hde A, Hossamu C, et al. Radiation therapy as an option for upper
airway obstruction due to Wegener’s granulomatosis. Otolaryngol Head Neck Surg 2002;
Rheum Dis Clin N Am 33 (2007) 777–786

Takayasu Arteritis: What is the

Long-Term Prognosis?
Kathleen Maksimowicz-McKinnon, DOa,*,
Gary S. Hoffman, MDb
Division of Rheumatology and Clinical Immunology, University of Pittsburgh,
3500 Terrace Street, BST S718, Pittsburgh, PA 15261, USA
Rheumatic and Immunologic Diseases, Center for Vasculitis Care and Research,
Lerner College of Medicine, Cleveland Clinic, 9500 Euclid Avenue,
Cleveland, OH 44195, USA

Takayasu arteritis (TA) is a chronic, inflammatory vascular disease that

primarily affects the aorta and its primary branches. The initial description
of TA is usually credited to Mikito Takayasu [1], a Japanese ophthalmolo-
gist, who reported a 21-year-old woman who had sudden visual loss and an
anastomotic wreath of vessels around the optic disc. In fact, an even earlier
report of disease that was consistent with TA came from Morgagni [2] in
1761, who noted large-vessel aneurysms and stenoses at the postmortem ex-
amination of a 40-year-old woman. Doctors had noted the absence of radial
pulses in Morgagni’s patient years before her demise. In addition, Savory
reported a 22-year-old English woman who had large-vessel bruits, absent
radial and carotid pulses, and blindness in one eye [3]. At postmortem, all
of the major upper extremity arteries had been reduced to solid cords. These
references emphasize facts that have become increasingly apparent in recent
years: TA is not just a disease of young Asian women. Although women are
affected disproportionately by this disease, TA afflicts individuals in a wide
variety of ethnic and racial populations worldwide.
Before the availability of noninvasive vascular imaging, little was known
about the extent and progression of vascular involvement in TA. Long de-
lays in diagnosis (months to years) were the rule, and the extent and progres-
sion of vascular lesions were typically underappreciated [4]. For patient
cohorts in which sequential vascular imaging studies were not performed,
TA was perceived as relatively benign disease that eventually became

* Corresponding author.
E-mail address: mckinnonk@dom.pitt.edu (K. Maksimowicz-McKinnon).

0889-857X/07/$ - see front matter Ó 2007 Elsevier Inc. All rights reserved.
doi:10.1016/j.rdc.2007.07.014 rheumatic.theclinics.com

‘‘burned out’’ in most patients [5,6]. Longitudinal imaging data from the
National Institutes of Health (NIH) and the Cleveland Clinic (CC) have
confirmed that most patients who have TA experience new lesions over
time, often in the absence of symptoms attributable clearly to TA [4,7].

Disease chronicity
Reports from India and Mexico in the era just preceding the availability
of noninvasive vascular imaging indicated a high prevalence of patients in
a chronic phase, among whom only a minority required ongoing immuno-
suppressive therapy [5,6]. This finding led to the perception of TA as
a chronic but self-limited disease in most patients.
Earlier studies from Asia, however, had reported morbidity stemming
from the progression of vascular lesions and the potential for complications
of TA to cause death [8,9]. In those earlier studies, a subset of patients who
had TA was recognized to have persistent disease associated with progres-
sive vascular dysfunction and ongoing systemic inflammation, including
arthralgias, fever, and fatigue.
One common finding in TA is the persistence of vascular inflammation in
patients who appear to be clinically quiescent. Surgical specimens from pa-
tients believed to be in clinical remission have revealed histologic evidence
of vasculitis in more than 40% of cases [4,10]. In a postmortem series reported
by Sharma and colleagues [11], active inflammatory lesions were demon-
strated even in patients believed to be in the chronic, fibrotic phase of disease.
In the NIH and CC cohorts, monophasic disease patterns were reported in
only 20% and 7% of all patients, respectively [4,7]. It is now clear that
most patients who have TA have disease that is chronic or relapsing in nature.

Diagnostic guidelines
The age of the population at risk for TA is broader than generally appre-
ciated. The American College of Rheumatology (ACR) classification criteria
for TA and Ishikawa’s [12,13] criteria reflect historical perceptions of TA as
a disease that predominantly affects young women. In a cohort of 104 pa-
tients from Italy who had TA, however, 17% had disease onset after age
40 [14]. Similar findings were observed in two United States cohorts [4,7],
and Asian investigators have reported the diagnosis of TA in patients in
their sixth and seventh decades of life [15,16]. Although some diagnoses in
patients older than 40 years of age probably reflect failures to diagnose sub-
clinical disease before that age, it is also now clear that at least in a signifi-
cant minority of patients the onset of clinically evident disease occurs after
the age of 40.
Fig. 1 demonstrates the frequency of vascular involvement during the
course of TA in American, Italian, Japanese, Indian, and Mexican cohorts.
Although it is clear that upper extremity arterial involvement is the most


% Affected





Left iliac
Left renal
Aortic arch


Right iliac
Left carotid

Right renal
Right carotid
Thoracic aorta

Left subclavian
Abdominal aorta

Right subclavian

CCF NIH Italy Japan India Mexico

Fig. 1. Frequency of vascular involvement. Instances in which data were not provided for
a cohort are indicated by absence of a comparison bar. (From Maksimowicz-McKinnon K,
Clark TM, Hoffman GS. Limitations of therapy and a guarded prognosis in an American
cohort of Takayasu arteritis patients. Arthritis Rheum 2007;56:1000–9; with permission. Copy-
right Ó 2007 American College of Rheumatology. Reprinted with permission of Wiley-Liss,
Inc., a subsidiary of John Wiley & Sons, Inc.)

common disease pattern, visceral or lower extremity arterial involvement

can be found in up to half of patients who have TA.

Disease-associated morbidity
The types of morbidity associated with TA are summarized in Fig. 2. We
now review the potential for TA to cause morbidity in a variety of organ

Hypertension (HTN) is a common complication of TA that often results
from stenotic disease of the renal arteries or aorta. In cohorts of patients
from India and Mexico, HTN was reported as a presenting feature in ap-
proximately one half and three quarters of all patients, respectively [5,6].
Aortic valve insufficiency, generally caused by dilatation of the aortic
root, has been reported in up to one quarter of patients. Valve ring expan-
sion may necessitate root reconstruction and valvuloplasty. Congestive
heart failure (CHF), present in up to one fourth of patients who have
TA, usually occurs as a consequence of uncontrolled HTN or aortic regur-
gitation. Clinically evident ischemic heart disease occurs in up to 10% of


% Affected











l sy


CCF NIH Italy India Mexico

Fig. 2. Prevalence of associated morbidities in Takayasu’s arteritis. Instances in which data

were not provided for an individual cohort are indicated by absence of a comparison bar.
(Adapted from Maksimowicz-McKinnon K, Clark TM, Hoffman GS. Limitations of therapy
and a guarded prognosis in an American cohort of Takayasu arteritis patients. Arthritis Rheum
2007;56:1000–9; with permission. Copyright Ó 2007 American College of Rheumatology. Re-
printed with permission of Wiley-Liss, Inc., a subsidiary of John Wiley & Sons, Inc.)

cases [4–7]. Vascular claudication has been reported in more than one third
of patients and was of sufficient severity to limit routine daily activities in
60% of patients in the CC cohort [4,6,7].

Visual disturbances, such as blurring or amaurosis fugax, occur in 8% to
13% of patients who have TA. Permanent loss of vision is unusual in this
disease, however [4,5,7,14]. Stroke and transient ischemic attacks may result
from progressive arterial stenosis or from uncontrolled HTN, and occur in
5% and 20% of cases, respectively [4–9,14–17]. In the NIH cohort, two of
the five patients who experienced strokes had residual hemiparesis and
one developed unilateral blindness [4].

Data on the outcomes of pregnancy among patients who have TA are
limited but do not necessarily reflect either a tendency for worse pregnancy
outcomes compared with patients who do not have TA or the occurrence of
disease exacerbations during pregnancy among patients who have TA. In
a study from Italy [14], the pregnancy outcomes of 18 women (24 pregnan-
cies) who became pregnant after the diagnosis of TA were compared with

those of 39 women (94 pregnancies) whose pregnancies occurred before

symptoms of TA became evident. The number of spontaneous abortions in
the two groups, 21% and 11%, respectively, were not statistically different.
In a cohort of 88 Indian patients [18], 7 women had 11 pregnancies, 8 of
which led to live births and 3 of which were associated with stillbirths.
Among patients in the combined NIH and CC cohorts, 9 pregnancies led
to 8 live births and one spontaneous pregnancy loss at 25 weeks of gestation
[4,7]. Worsening of TA during pregnancy was reported in 1 patient [4]. One
patient in this cohort had worsening symptoms of disease during pregnancy.

Disease-associated morbidities frequently impair patients’ abilities to per-
form routine daily activities and to maintain employment. In the NIH co-
hort, only 26% reported not being functionally affected by their disease
[4]. Forty-seven percent of those patients were fully disabled. Findings
from the CC cohort were similar; 60% stated that vascular claudication im-
paired their ability to carry out routine daily activities and 23% were fully
disabled. The likelihood of disability correlated with the number of docu-
mented disease relapses [7].

Lupi-Herrera and colleagues reported a 15% mortality rate among pa-
tients who had TA in whom long-term follow-up data were available [17].
Approximately half of the deaths in the cohort resulted from CHF. Among
a cohort of 88 patients from India who had TA, there was a 5-year survival
rate of 91% and 10-year survival rate of 84%. Forty percent of deaths in the
Indian cohort were the result of CHF [18].
CHF was also the most common reported cause of death in a second se-
ries from India that included 69 patients [5]. Mortality was reported as 17%
in this group, with about half of the patients dying of CHF and one fourth
of renal failure. Among 120 Japanese patients who had TA followed for
a median of 13 years, 13% mortality was reported [16]. The median age
at death was 48 years. In that series, the cause of death was identified in
14 patients: CHF (5 patients), stroke (4 patients), postoperative complica-
tions (3 patients), and myocardial infarction (1 patient).
In considering morbidity and mortality in TA, one needs to be cognizant
of two points. First, the mean age at diagnosis for patients is only about 26
years. Disease- and treatment-related morbidities may thus take their toll on
patients over decades, and mortality means a tremendous loss of potentially
productive years of life. Second, in most TA patient cohorts, reported fol-
low-up is usually relatively brief compared with the patients’ potential life-
span. As a consequence, the full impact of the disease and its therapy may
not be recorded entirely during the period covered by a published report.

Response to medical therapy

In most cohorts, most patients who have active disease attain disease con-
trol with glucocorticoid (GC) therapy. Even among patients who achieve
complete remissions on GC therapy, however, relapse is common on GC ta-
per or withdrawal. This phenomenon is reflected in the prevalence of use of
other immunosuppressive agents in the Italian (54%), Korean (40%), and
American (NIH 42%, CC 73%) cohorts [4,7,14,15]. In the CC longitudinal
cohort, 96% of patients who achieved disease remissions experienced at
least one relapse, with a mean of 2.83 relapses per patient over a median pe-
riod of 3 years [7]. Of the 54 relapses documented, 63% occurred while pa-
tients were being treated with immunosuppressive agents other than GC.
Half of the disease relapses occurred while patients were either taking meth-
otrexate (MTX) alone (12%) or MTX with GC (38%). The median doses of
MTX and GC at relapse were 17.5 mg/wk and 10 mg/d, respectively.
The limitations of current therapies are also reflected in the duration of
remission. The NIH reported that only 63% of 382 patient-years of fol-
low-up were spent in remission [4]. In the CC cohort, only 28% of patients
achieved sustained remission, defined as: (1) the absence of clinical or labo-
ratory markers of active disease and absence of radiographic evidence of
new vascular lesions for more than 6 months, and (2) treatment with less
than 10 mg/d of prednisone [7].

Interventional therapy outcomes

Angioplasty/vascular stents
Initial studies that examined the outcomes of percutaneous transluminal
angioplasty (PCTA) in TA seemed promising. In 1992, Sharma and col-
leagues reported an initial success rate of 82% for renal artery PTCA in
the treatment of 33 renal artery lesions in 20 patients who had TA [19].
The patency rate at a mean follow-up of 8 months was 79% [19]. Similar
findings were reported by Tyagi and colleagues [20] in a study of 45 patients
who had TA, with an initial success rate of 89% and a patency rate of 79%
at a mean follow-up of 43 months. More recent data from United States co-
horts in which serial catheter-directed or MR-based vascular imaging stud-
ies were performed as part of longitudinal disease surveillance revealed a less
optimistic view of PCTA in TA, however. In the NIH cohort, 11 patients
underwent 20 PCTA procedures, most frequently of the subclavian and re-
nal arteries. The initial success rate was only 56%, and complications (reste-
nosis, thrombosis, hemorrhage, infection, or aneurysm formation) were
reported in 45% of the procedures [4]. In the CC cohort, 20 PTCA proce-
dures were performed, 10 of which utilized stents. Restenosis occurred in
78% (14 out of 18) of the procedures that were initially successful; 13 of
14 patients required additional interventions [7].

In summary, although PTCA can achieve high rates of initial success in

TA, the longer-term view of these procedures in this disease is less optimis-
tic. Because of the propensity of patients who have TA to develop exuberant
collateral circulation that often bypasses (to a significant degree) the lesions
in larger vessels, the usefulness of performing PTCA should be considered
carefully for most patients who have this disease.
Outcomes of arterial bypass and reconstruction procedures in TA have
demonstrated better sustained patency than PTCA procedures. Long-term
results reveal the limitations of these procedures also. Lupi-Herrera and col-
leagues [17] reported a patency rate of only 33% at 2 years in 9 patients who
had TA who underwent arterial bypass. The NIH cohort described outcomes
in 23 patients who underwent a total of 50 bypass procedures. In this group,
24% had restenosis and 4% suffered bypass graft thrombosis [4]. Complica-
tions (restenosis, thrombosis, hemorrhage, or infection) occurred in 36% of
synthetic grafts and 9% of autologous grafts. At the CC, 44 bypass procedures
were performed in 17 patients [7]. Of note, 21 (46%) of these procedures were
performed following the failure of an earlier interventional procedure. Thirty-
six percent of patients undergoing bypass procedures had vascular restenosis
or occlusion, generally requiring an additional procedural intervention.
Two of 60 patients in the NIH cohort and 5 of 30 patients in the CC co-
hort required aortic valve replacement; 1 patient developed prosthetic valve
endocarditis (NIH) and 1 required revision after 10 years [4,7].

Factors affecting the care of patients who have Takayasu arteritis:

what can be done to improve outcomes?
Challenges in disease activity assessment
As in many other disease states, unremitting disease activity can lead to
poor outcomes (arterial stenosis and end-organ damage) in TA. A major
limitation in the early detection of disease activity of TA is the paucity of
reliable clinical symptoms and signs of disease activity at presentation and
disease relapse.
Correlation between the levels of acute phase reactants (ie, the erythro-
cyte sedimentation rate and C-reactive protein) and radiologic findings is of-
ten poor in TA. For example, in one study of patients at routine visits, 46%
were noted to have either elevated acute-phase reactants but no changes on
MR studies or normal acute-phase reactants but the detection of new vessel
wall enhancement or edema by MR imaging [7].
Pharmacologic pitfalls
Longitudinal cohort studies demonstrate the limitations of current ther-
apies in TA. Although most patients are able to attain disease control with

high-dose GC therapy, most patients suffer relapses with dose reductions. In

addition, data from the CC cohort has demonstrated a high frequency of
disease relapse even among patients receiving multiple immunosuppressive
Anti-TNF agents may be useful. Hoffman and colleagues reported out-
comes in 15 patients who had refractory, relapsing TA following treatment
with infliximab (8 patients) and etanercept (7 patients, 3 of whom later were
changed to infliximab) [21]. Before the institution of anti-TNF therapy,
patients required a median daily dose of 20 mg of prednisone to maintain
disease remission. Ten patients achieved complete remission and were able
to discontinue GC therapy, whereas 4 patients achieved partial remission
(at least 50% reduction in daily GC dose). Similarly encouraging findings
were noted in a follow up study, in which 15 of 25 patients who had refrac-
tory TA treated with anti-TNF therapy were able to discontinue or taper
prednisone to less than 10 mg/day, and 9 of 18 patients were able to discon-
tinue or taper the dose of additional immunosuppressive agents [22].
Despite the encouraging data from these case series, rigorous assessment
of targeted TNF inhibition in a randomized clinical trial is required before
the full clinical usefulness of this therapeutic approach can be understood.

Interventional pitfalls
Patients who have TA, when compared with patients undergoing inter-
ventions for atherosclerosis or other vascular diseases, have decreased rates
of sustained vessel patency for PTCA and arterial bypass procedures
[20,23,24]. One factor that has been significantly associated with the proce-
dural failure is the presence of active disease at the time of surgery. Fields
and colleagues [25] followed 42 patients who had TA following arterial
bypass procedures and found that none of the patients believed to have
quiescent disease at the time of the procedure (ie, no maintenance GC
requirement, and absence of signs and symptoms of active disease) required
surgical revisions at 5 and 10 years. In contrast, among the patients on GC
therapy for the treatment of active disease at the time of surgery, 43%
required surgical revisions at 5 years.

Longitudinal studies that have included sequential vascular imaging have
revealed that in most patients TA is a chronic, relapsing disease that is
associated with significant morbidity and disability. The scope of vascular
involvement is broader than appreciated two decades ago, before the avail-
ability of noninvasive large-vessel imaging studies. High rates of restenosis
following vascular interventional procedures, especially during times of
overtly active disease, reinforce recommendations to try to postpone surger-
ies until TA is in remission. For bypass, autologous vascular grafts are

superior to synthetic materials. Disease activity may be subclinical and not

recognized by laboratory surrogate markers in up to half of all patients. The
frequency of disease relapse and disability in TA, despite treatment with or
more immunosuppressive agents, demands our attention and mandates
efforts toward identifying better therapies.

[1] Takayasu M. A case with peculiar changes of the retinal central vessels. Acta Societatis
Opthalmological Japonicae 1908;12:554–5.
[2] Morgagni GB. De sedibus et causis morborum per anatomen indagatis [Letter 30]. 1761.
Article 12.
[3] Savoy WE. Case of a young woman in whom the main arteries of both upper extremities and
the left side of the neck were throughout completely obliterated. Med Chir Thrans London
[4] Kerr GS, Hallahan CW, Giordano J, et al. Takayasu arteritis. Ann Int Med 1994;120:
[5] Jain S, Kumari S, Ganguly NK, et al. Current status of Takayasu arteritis in India. Int J Car-
diol 1996;54(Suppl):S111–6.
[6] Dabague J, Reyes P. Takayasu arteritis in Mexico: a 38-year clinical perspective through lit-
erature review. Int J Cardiol 1996;54:S87–93.
[7] Maksimowicz-McKinnon K, Clark TM, Hoffman GS. Limitations of therapy and a guarded
prognosis in an American cohort of Takayasu arteritis patients. Arthritis Rheum 2007;56:
[8] Ishikawa K. Natural history and classification of occlusive thromboaortopathy (Takayasu’s
disease). Circulation 1978;57:27–35.
[9] Ueda H, Morooka S, Ito I, et al. Clinical observation of 52 cases of aortitis syndrome. Jpn
Heart J 1969;10:277–88.
[10] Lagneau P, Michel JB, Vuong PN. Surgical treatment of Takayasu’s disease. Ann Surg 1987;
[11] Sharma BK, Jain S, Radotra BD. An autopsy study of Takayasu’s arteritis in India. Int
J Cardiol 1998;66(Suppl):S85–90.
[12] Arend WP, Michel BA, Bloch DA, et al. The American College of Rheumatology 1990 cri-
teria for the classification of Takayasu’s arteritis. Arthritis Rheum 1990;33:1129–34.
[13] Ishikawa K. Diagnostic approach and proposed criteria for the clinical diagnosis of Takayasu’s
arteriopathy. J Am Coll Cardiol 1988;12:964–72.
[14] Vanoli M, Daina E, Salvarani C, et al. Takayasu’s arteritis: a study of 104 Italian patients.
Arthritis Rheum (Arthritis Care and Research) 2005;53:100–7.
[15] Park MC, Lee SW, Chung NS, et al. Clinical characteristics and outcomes of Takayasu’s ar-
teritis: analysis of 108 patients using standardized criteria for diagnosis, activity assessment,
and angiographic classification. Scand J Rheumatol 2005;34:284–92.
[16] Ishikawa K, Maetani S. Long-term outcome for 120 Japanese patients with Takayasu’s dis-
ease. Clinical and statistical analysis of related prognostic factors. Circulation 1994;90:
[17] Lupi-Herrera E, Sanchez-Torres G, Marcushamer J, et al. Takayasu’s arteritis: clinical study
of 107 cases. Am Heart J 1977;93:94–103.
[18] Subramanyan R, Joy J, Balakrishnan K. Natural history of aortoarteritis (Takayasu’s dis-
ease). Circulation 1989;80:429–37.
[19] Sharma S, Saxena A, Talwar KK, et al. Renal artery stenosis caused by nonspecific arteritis
(Takayasu disease): results of treatment with percutaneous transluminal angioplasty. Am
J Roentgenol 1992;158:417–22.

[20] Tyagi S, Gambhir DS, Kaul UA, et al. A decade of subclavian angioplasty: aortoarteritis
versus atherosclerosis. Indian Heart J 1996;48:667–71.
[21] Hoffman GS, Merkel PA, Brasington RD, et al. Anti-tumor necrosis factor therapy in
patients with difficult to treat Takayasu arteritis. Arthritis Rheum 2004;50:2296–304.
[22] Molloy ES, Langford CA, Clark TM, et al. Durable remission in patients with refractory
Takayasu’s arteritis treated with infliximab and etanercept. Arthritis Rheum 2006;54:4093.
[23] Fitzgibbon GM, Kafka HP, Leach AJ, et al. Coronary bypass graft fate and patient
outcome: angiographic follow up of 5065 grafts related to survival and reoperation in
1388 patients during 25 years. J Am Coll Cardiol 1996;28:616–26.
[24] Novick AC, Ziegelbaum M, Bidt DG, et al. Trends in surgical revascularization for renal
artery disease. Ten years’ experience. JAMA 1987;257:498–501.
[25] Fields C, Bower T, Cooper L, et al. Takayasu’s arteritis: operative results and influence of
disease activity. J Vasc Surg 2006;43:64–71.
Rheum Dis Clin N Am 33 (2007) 787–802

Pyoderma Gangrenosum: An Update

Jeffrey P. Callen, MD*, J. Mark Jackson, MD
Division of Dermatology, University of Louisville, 310 East Broadway,
Louisville, KY 40202, USA

Pyoderma gangrenosum (PG) is an inflammatory condition of the skin

first described by Brunsting and colleagues in 1930 [1]. The cause of PG is
unknown, but some studies have suggested that abnormal neutrophil
chemotaxis is the primary process [2]. PG is more frequent in adults, but
children may be affected on rare occasions [3,4].
PG begins as a pustule or vesiculopustule that progresses to an ulcer or
deep erosion with violaceous overhanging or undermined borders [5–7].
There are three major clinically distinct variations of PG [8]:
Classic PG, characterized by ulcers usually located on the legs
Atypical PG, characterized by more superficial ulcers or deep erosions
with a blue-gray, bullous border, usually found on the hands, arms,
or face
Peristomal PG, typified by the occurrence of lesions around a stoma (eg, one
formed following surgery for inflammatory bowel disease [IBD]).
The course of PG can be acute (uniphasic), relapsing, or chronic. Relaps-
ing or chronic courses are more likely to be associated with an underlying

Classic PG (Fig. 1) usually occurs on the lower extremities, but may occur
anywhere, including the abdomen, genitalia, trunk, head, and neck [9]. The

Dr. Jackson has received funding for research, honoraria, consulting, and/or other sup-
port from the following companies: 3M, Abbott, Amgen, Biogen, Centocor, CollaGenex,
Connetics, Ferndale, Galderma, Genentech, Novartis, and Roche. Dr. Callen has received
honoraria, consulted for, or served as a safety monitor for Amgen, Abbott, Centocor,
Genmab, Electrical Optical Sciences, and Stiefel.
* Corresponding author.
E-mail address: jefca@aol.com (J.P. Callen).

0889-857X/07/$ - see front matter Ó 2007 Elsevier Inc. All rights reserved.
doi:10.1016/j.rdc.2007.07.016 rheumatic.theclinics.com

Fig. 1. Classic pyoderma gangrenosum manifest by a large, tender ulceration on the leg. This
patient did not have an associated disease and healed fully with mycophenolate mofetil therapy
over a period of 8 months. She has been followed for more than 5 years and has not developed
further disease.

disorders most commonly associated with classic PG are IBD and rheumatoid
arthritis [10]. It has been suggested that the ulcerations in patients who have
classic PG associated with arthritis are more refractory to treatment than in
patients who have PG without associated arthritis [11]. The cutaneous mani-
festations may precede, follow, or occur simultaneously with the associated
condition. In some patients who have IBD, control of the bowel inflammation
results in control of the PG, but this is not true in all patients.
Atypical PG (Fig. 2) manifests itself as a more superficial ulceration, most
frequently on the upper extremities, head, or neck. The lesions begin as pus-
tules that spread into plaques that may be studded with pustules, particu-
larly on the dorsum of the hands [12,13]. This variant resembles Sweet

Fig. 2. Atypical pyoderma gangrenosum. This patient has superficial ulcers on the dorsal hands
with a bullous blue-gray margin. She was otherwise healthy.

syndrome. Other terms that have been used for this variant include ‘‘pustu-
lar vasculitis’’ and ‘‘neutrophilic dermatoses of the dorsal hands.’’ Similar to
the classic form of PG, patients who have atypical PG are commonly mis-
diagnosed as having an infection or cellulitis [14,15]. The most common dis-
ease associations among patients who have atypical PG are myelogenous
leukemia, myelodysplastic disorders, refractory anemias, and IgA parapro-
teinemia [10,16,17]. In addition, some patients who have blind loops created
by ulcer surgery or bowel bypass may manifest either the atypical variant of
PG or pustular lesions alone.
A third type of PG occurs when ulcerations occur in a peristomal
location (Fig. 3) [18]. The lesion resembles those seen in patients who
have classic PG. Peristomal PG occurs most often in patients who have
IBD, particularly Crohn disease. Patients who have a history of ulcerative
colitis or stomas created for other reasons, including malignancies of the
gastrointestinal or urinary tract, should have an evaluation for the possibil-
ities of either active Crohn disease or malignancy recurrence.
The ulcerative lesions of PG occur occasionally on the oropharynx (pyos-
tomatitis vegetans), vulva, or penis [19–22]. Such lesions have not been
classified as separate and distinct subsets of PG, however. Lesions occurring
on mucosal sites may be difficult to differentiate from Behçet disease.

Clinical features
The pathergy phenomenon
Pathergy is defined as the development of skin lesions at sites of injury
(Figs. 4 and 5). This development may occur with minor injuries, including
cuts, or may follow needle sticks or surgery [23]. Patients who have PG are
often initially misdiagnosed and treated erroneously for an infection, the

Fig. 3. Peristomal pyoderma gangrenosum. This young woman developed these ulcerations fol-
lowing removal of her diseased colon (Crohn disease). In addition to the peristomal disease
there is an ulcer in the midline incision.

Fig. 4. This patient developed pyoderma gangrenosum following surgery to release her fascia
because of a compartment syndrome. This case demonstrates pathergy.

therapy of which might include débridement of the area. PG lesions often

worsen after débridement, but do not always do so. In addition, some
patients believe that they have been bitten by a brown recluse spider at
the onset of the disorder without ever noting the spider or an actual bite.
In areas of the United States endemic for brown recluse spiders, some
authorities believe that some PG lesions are a result of such bites.

Extracutaneous disease
Extracutaneous infiltration by neutrophils may complicate the course of
PG. The extracutaneous lesions do not contain pathogenic microorganisms
and have thus been termed ‘‘sterile’’ neutrophilic abscesses. As is true with

Fig. 5. This patient’s lesions continued to worsen with multiple débridements, eventually result-
ing in a loss of her entire right breast. The initial lesion was believed to be an infection.
Although she was asymptomatic, colonoscopy revealed a small area of inflamed bowel that
on biopsy was compatible with ulcerative colitis.

the cutaneous lesions of PG, these infiltrates must be distinguished from in-
fections before initiating glucocorticoid or immunosuppressive therapies.
The most common extracutaneous involvement is in the lungs [24]. Bone
lesions have been noted and in one patient treated by us, the cutaneous dis-
ease developed only after a bone biopsy had been performed. Bony lesions
are often termed ‘‘multifocal sterile recurrent osteomyelitis’’ [25,26]. Other
sites of involvement include the cornea [27], liver and spleen [28], heart
[29], skeletal muscles [30], and the central nervous system [31]. All of these
manifestations, including pulmonary involvement, are rare.

Differential diagnosis
The diagnosis of PG involves the exclusion of other diseases that cause ero-
sive or ulcerative skin lesions. The differential diagnosis of PG thus includes
vasculitis, thrombophilic conditions, malignancies, other inflammatory dis-
eases, and infections [32]. The forms of systemic vasculitis most likely to cause
skin ulcers are polyarteritis nodosa, the types of vasculitis associated with
antineutrophil cytoplasmic antibodies (ANCA), and mixed cryoglobulinemia
[33] (usually associated with hepatitis C). Thrombophilic states that can be
confused with PG include livedoid vasculitis (atrophie blanche) and the anti-
phospholipid syndrome [34], factor V Leiden mutation (Fig. 6), and methy-
lene tetrahydrofolate reductase polymorphisms (Fig. 7) [35].
Malignancies that can mimic PG include squamous cell carcinoma, cutane-
ous lymphoma, leukemia cutis, and metastatic carcinoma. Crohn disease with
cutaneous involvement (metastatic Crohn disease, a condition distinct from
PG) and localized immunobullous disorders must be excluded. Infections
that can mimic PG include cellulitis, herpetic ulcers, Buruli ulcer, atypical

Fig. 6. This patient has a factor V Leiden mutation. Careful examination of his biopsy revealed
thromboses in small vessels.

Fig. 7. This young man has mutations of his methylene tetrahydrofolate reductase genes. He
had been misdiagnosed as having pyoderma gangrenosum. Therapy with folic acid led to
complete healing of these lesions. Unfortunately, when he stopped his medication he relapsed,
but again remitted when the folic acid was restarted.

mycobacterial infections, cutaneous tuberculosis, leishmaniasis, sporotricho-

sis and other deep fungal infections [36,37], and ecthyma gangrenosum [38].
Factitious ulcerations are another consideration in some cases [39]. The
difficulty with a diagnosis of factitial ulceration is that it is also a diagnosis
of exclusion. Unless there is a clear indication that the patient is inducing
the lesions, the treating physician may not be able to make this diagnosis
firmly. In addition, because patients who have true PG often have pa-
thergy, the patient who has factitial tendencies can sometimes create new
lesions easily. Findings such as perfectly straight lines or sharply angular
lesions should alert the clinician to the possibility that the lesions are

Evaluation of possible pyoderma gangrenosum

Diagnostic criteria have been proposed for PG, but none have been
validated (Box 1) [40,41]. We recommend that patients who have PG have
a complete history and physical examination, with particular emphasis on
symptoms or signs of arthritis, gastrointestinal disease, and malignancy.
In addition, a history of thrombophilic processes, including superficial or
deep venous thrombophlebitis, pulmonary emboli, transient ischemic
attacks, or recurrent fetal loss, should be included.
A biopsy of the affected area for routine processing and special stains for
microorganisms should be performed in almost all patients. The histopath-
ologic findings in PG are not specific and the biopsy cannot be diagnostic
(in and of itself) of PG. The principal importance of biopsydwhich should
not be deferred because of concerns of potentially extending lesions through

Box 1. Criteria for the diagnosis of pyoderma gangrenosum

Major criteria
1. Rapida progression of a painfulb necrolytic cutaneous ulcerc
with an irregular, violaceous, and undermined border
2. Exclusion of other causes of cutaneous ulceration
Minor criteria
1. History suggestive of pathergyd or clinical finding of cribriform
2. Systemic diseases associated with PGe
3. Histopathologic findings (sterile dermal neutrophilia ± mixed
inflammation ± lymphocytic vasculitis)
4. Treatment response (rapid response to systemic
glucocorticoid treatment)f
Characteristic margin expansion of 1 to 2 cm/d, or a 50% increase in ulcer
size within 1 month.
Pain is usually out of proportion to the size of the ulceration.
Typically preceded by a papule, pustule, or bulla.
Ulcer development at sites of minor cutaneous injury.
Inflammatory bowel disease, polyarthritis, myelocytic leukemia, or
Generally responds to a dosage of 1 to 2 mg/kg/d, with a 50% decrease in size
within 1 month.
Adapted from Su WPD, Davis MD, Weenig RH, et al. Pyoderma gangrenosum:
clinicopathologic correlation and proposed diagnostic criteria. Int J Dermatol

pathergydis to exclude other diseases. Although vascular inflammation in

lesions of PG is not uncommon, the histology is not that of a true vasculitis
[42]. In some cases, biopsies of PG lesions from patients who have myelog-
enous leukemia or preleukemic states contain atypical cells in the infiltrate,
thereby unmasking a malignant or premalignant lesion [43]. Tissue culture
for bacteria, mycobacteria, and fungi should be performed, and viral
cultures should be performed in selected individuals [44].
Laboratory evaluation should include a complete blood count, chemistry
profile, hepatitis panel, serum and urine protein electrophoresis, and
urinalysis. Additional tests should include antinuclear antibodies, ANCA,
rheumatoid factor, and antiphospholipid antibody assays.
Gastrointestinal evaluations should be performed in all patients who
have PG. Patients who have peristomal PG should be evaluated carefully
for active bowel disease or malignancy if there is a previous history of gas-
trointestinal malignancy. Chest radiography is important to exclude infec-
tions and forms of systemic vasculitis associated with pulmonary

Management of PG is divided into local wound care, topical therapy, and
systemic therapy [45,46]. Local care of ulcerations in PG includes moist
dressings (such as Vaseline gauze) to prevent trauma to the underlying tissue
on removal. Aggressive debridement or grafting should be avoided. Bioen-
gineered skin dressings may be beneficial in covering ulcers and preventing
the need for skin grafts [47]. When extensive necrosis of the skin is present or
vital tissues, such as tendons and ligaments, are exposed at the ulcer bed,
debridement and skin grafting may be necessary. In such cases, concomitant
systemic therapy is required to halt the inflammatory process.
When superficial bacterial cultures reveal a mixture of Gram-negative
and Gram-positive bacteria, the clinician may be misled into believing
that the ulcer is secondary to an infection. In these situations, the bacteria
usually represent colonizers and not pathogens. Treatment of the underlying
inflammatory process with immunosuppressive medications is usually suffi-
cient to heal the ulcer without antibiotic therapy.
Topical therapy is helpful in some cases and successful use of tacrolimus
[20,48,49], pimecrolimus, potent topical glucocorticoids [50], nicotine [51],
or cromolyn sodium [52] has been reported in individual patients. Another
local therapy is the injection of glucocorticoids into the lesions of PG.
Despite several individual reports of success, we do not endorse this
approach as first-line therapy because of the risk for exacerbating the lesions
through pathergy.
Successful systemic therapies for PG have included glucocorticoids, aza-
thioprine, cyclosporine [53–55], tacrolimus, mycophenolate mofetil [56–58],
methotrexate [59], chlorambucil [60], thalidomide [61], colchicine [62] cyclo-
phosphamide [63,64], and dapsone, minocycline, or sulfapyridine [65].
Tumor necrosis factor-a (TNF-a) inhibitors (etanercept, infliximab, and
adalimumab) seem effective in PG (Fig. 8) [66–75]. Infliximab was studied in

Fig. 8. (A and B) This patient developed multiple lesions following colectomy for Crohn dis-
ease. The pictures represent complete healing at 12 months following three infusions of inflix-
imab (weeks 0, 2, and 6).

a double-blind, placebo-controlled trial and was demonstrated to be signif-

icantly more effective than placebo in patients who had idiopathic PG
and PG associated with Crohn disease [71]. This was a small clinical trial
of 30 patients treated for a relatively short time (three infusions over
a 6-week period). Serious adverse effects occurred in 2 patients, 1 of which
was fatal (staphylococcal septicemia). Because etanercept and adalimumab
are administered by subcutaneous injection and therefore pose a risk for
pathergy, infliximab is the drug of choice when targeted TNF inhibition is
Many patients respond initially to the systemic glucocorticoids (eg, pred-
nisone 0.5 to 1.0 mg/kg/d) but the long-term use of glucocorticoids is asso-
ciated with important toxicity, including osteoporosis, bone fracture,
diabetes mellitus, cataracts, and an increased risk for infections. Steroid-
sparing agents should be considered early in the course of PG [45]. Pulse
methylprednisolone, pulse cyclophosphamide, and intravenous immuno-
globulin have also shown benefit in some patients [76–78]. Dosing recom-
mendations for the above agents are included in Table 1.
Surgical treatment of patients who have PG should be reserved for
patients in extreme situations and, if possible, for those on sufficient therapy
to prevent the development of PG at donor sites. Patients who have peristo-
mal PG have developed recurrent PG at the site of the new stoma placement
[19]. There are times when debridement of superficial necrotic tissue is
necessary for healing. In addition, when tendons, ligaments, or bones are ex-
posed, surgical intervention is unavoidable (Fig. 9).
In patients who have underlying disease and no contraindications to
high-dose glucocorticoids, short courses of oral prednisone (40–60 mg/d,
tapered off over a 6-week period) are sometimes sufficient for an acute
episode. This treatment leads to improvement in pain and inflammation
in most patients. In those who have a more chronic course, it is important
to add an immunosuppressive agent, such as cyclosporine, mycophenolate
mofetil, azathioprine, or infliximab.

Many patients who have PG develop a single episode that resolves with
a short course of therapy. Of these patients, some never develop the process
again, and others remain recurrence-free for years between episodes. Other
patients have a chronic or relapsing course requiring long-term therapy.
Treatment of these patients is more complex, because they may require
combination therapy and close monitoring for toxicity from therapy.
Some patients lose efficacy on one form of therapy and require new medica-
tions added or used in replacement.
The long-term prognosis for most patients is good, especially in those
who have no other underlying associated disease. Concomitant disorders
Table 1

Systemic therapeutic options for pyoderma gangrenosum
Agent Dose range Benefits Risks Helpful hints and pearls
Glucocorticoids 0.5–1 mg/kg/d Quick acting on pain Long-term risk for Check PPD before starting,
(prednisone or and ulcer resolution infections, insulin Use for 2–4 weeks
equivalent) resistance, HTN, maximum if possible
glaucoma, and adrenal
Cyclosporine 4–5 mg/kg/d Fairly quick acting HTN, renal toxicity, Start at 4–5 mg/kg initially
hypertriglyceridemia, then taper (do not start
hypertrichosis too low)
Azathioprine 100–300 mg/d Ease of long-term Slow onset of action, Bone Check thiopurine
administration in most marrow suppression, methyltransferase levels


patients GI intolerance, to prevent toxicity in
patients lacking the
enzyme or with low levels
Dapsone 50–200 mg/d Better long-term safety Slow onset of action, does not Watch for drop in WBC
profile work in more severe cases, count and hemoglobin
methemoglobinemia, (normal drop is about
anemia, neuropathy 10%)
Methotrexate 10–30 mg/wk Familiar long-term safety Hepatoxicity, and bone Check hepatitis C levels
issues, good tolerance marrow suppression before therapy even in
in most patients who have normal
LFTs. Folate may reduce
hepatoxicity and GI
Mycophenolate mofetil 2–3 g/d Good long-term safety Slow onset of action, Best if initiated or added to
profile neutropenia glucocorticoids because
of slow onset of effect
Sulfapyridine 1 g bid Good long-term safety
profile, some potential
effect in arthritis patients
Thalidomide 50–200 mg/d A potential steroid-sparing Birth defects, somnolence, Dose at bedtime because
agent neuropathy, of somnolence, consider
coagulopathy nerve conduction studies
Colchicine 0.6 mg bid to tid Safe if tolerated Diarrhea Start low initially and
titrate up because of
potential for GI
intolerance initially
Cyclophosphamide 1.5–3.0 mg/kg/d Hemorrhagic cystitis, Good hydration before
azoospermia and during therapy
Etanercept 50 mg SQ q wk Well tolerated, no drug May not be potent enough Works best if starting


to biweekly interactions, good for inhibitor of TNF-a at 50 mg twice weekly
patients who have RA, in severe cases, Watch closely for
Crohn disease, or colitis contraindicated in pathergy at injection
a patients who have sites
CHF, TB, or multiple
Infliximab 5 mg/kg/wk weeks Fast onset of action, may Contraindicated in patients Pretherapy evaluation to
0, 2, 6, then use as monotherapy, well who have CHF, TB, or exclude tuberculosis
q 6–8 wk tolerated, no drug multiple sclerosis Don’t infuse too
interactions, good for quickly to prevent
patients who have RA, infusion reactions
Crohn disease, or colitis
Adalimumab 40 mg SQ No drug interactions, good Contraindicated in patients Watch closely for
qo wk to q wk for patients who have RA, who have CHF, TB, or pathergy at
Crohn disease, or colitis multiple sclerosis injection sites
Tacrolimus 0.1–0.15 mg/kg Rapid onset of action Hypertension and renal
Antibiotics (minocycline/ 100 mg bid to tid Well tolerated, good if Slow onset of action, does Best if added to another
doxycycline) secondary staphylococcal not work often therapeutic option
Abbreviations: CHF, congestive heart failure; GI, gastrointestinal; HTN, hypertension; LFT, liver function test; PPD, purified protein derivative;

RA, rheumatoid arthritis; TB, tuberculosis; WBC, white blood cell.

Fig. 9. This patient failed to respond to glucocorticoids and azathioprine. Within 3 months of
his initial presentation his lesion progressed and exposed tendons were visible in the base of the
ulcer. Successful therapy included gentle cleansing and oral chlorambucil. His disease went into
remission within 8 months of initiation.

can alter the long-term outcome significantly, however, because of PG that

is more difficult to treat and because of the occurrence of comorbidities. Pa-
tients who present with atypical PG need regular follow-up to evaluate for
the development of myeloproliferative disorders, even though most patients
who develop a myeloproliferative disorder associated with PG present be-
fore or simultaneously with the diagnosis of PG.

PG can be differentiated into classic and atypical forms. The classic form
is characterized by ulcers and the atypical form by deep erosions with bul-
lous blue-gray margins. PG lesions usually begin as pustules that progress
rapidly to either an ulceration with an overhanging violaceous border or
a deep erosion with a bullous blue-gray margin. Pathergy, the development
of cutaneous lesions at sites of trauma, is a common feature of both forms of
PG. Approximately 50% of patients who have PG have underlying systemic
diseases, most commonly IBD, myeloproliferative disorders, and various
forms of inflammatory arthritis. The diagnosis of PG is one of exclusion.
The management of this disorder begins with treatment of any underlying
disease and local or systemic glucocorticoids or immunomodulating

[1] Brunsting LA, Goeckerman WH, O’Leary PA. Pyoderma gangrenosum: clinical and exper-
imental observations in five cases occurring in adults. Arch Derm Syphilol 1930;22:655–80.
[2] Adachi Y, Kindzelskii AL, Cookingham G, et al. Aberrant neutrophil trafficking and met-
abolic oscillations in severe pyoderma gangrenosum. J Invest Dermatol 1998;111:259–68.

[3] Takeuchi K, Kyoko H, Hachiya M, et al. Pyoderma gangrenosum of the skin and respiratory
tract in a 5 year-old girl. Eur J Pediatr 2003;162:344–5.
[4] Torrelo A, Colmenero I, Serrano C, et al. Pyoderma gangrenosum in an infant. Pediatr Der-
matol 2006;23:338–41.
[5] Callen JP. Pyoderma gangrenosum. Lancet 1998;351:581–5.
[6] Jackson JM, Callen JP. Pyoderma gangrenosum: an expert commentary. Expert Rev Der-
matol 2006;1:391–400.
[7] Brooklyn T, Dunnill G, Robert C. Diagnosis and treatment of pyoderma gangrenosum.
BMJ 2006;333:181–4.
[8] Powell FC, Collins S. Pyoderma gangrenosum. Clin Dermatol 2000;18:283–93.
[9] Vidal D, Puig L, Gilaberte M, et al. Review of 26 cases of classical pyoderma gangrenosum:
clinical and therapeutic features. J Dermatolog Treat 2004;15:146–52.
[10] Bennett ML, Jackson JM, Jorizzo JL, et al. Pyoderma gangrenosum. A comparison of
typical and atypical forms with an emphasis on time to remission. Case review of 86 patients
from 2 institutions. Medicine 2000;79:37–46.
[11] Charles CA, Bialy TL, Falabella AF, et al. Poor prognosis of arthritis-associated pyoderma
gangrenosum. Arch Dermatol 2004;140:861–4.
[12] Shankar S, Sterling JC, Rytina E. Pustular pyoderma gangrenosum. Clin Exp Dermatol
[13] Walling HW, Snipes CJ, Gerami P, et al. The relationship between neutrophilic dermatosis
of the dorsal hands and Sweet syndrome: report of 9 cases and comparison to atypical
pyoderma gangrenosum. Arch Dermatol 2006;142:57–63.
[14] Louis DS, Jebson PJ. Mimickers of hand infections. Hand Clin 1998;4:519–29.
[15] Huish SB, de La Paz EM, Ellis PR, et al. Pyoderma gangrenosum of the hand: a case series
and review of the literature. J Hand Surg Am 2001;26:679–85.
[16] Isomura I, Miyawaki S, Morita A. Pyoderma gangrenosum associated with nasal
septal perforation, oropharyngeal ulcers and IgA paraproteinemia. J Dermatol 2005;
[17] Jackson JM, Callen JP. Pyoderma gangrenosum and IgA myeloma. J Cutan Med Surg 1997;
[18] Hughes AP, Jackson JM, Callen JP. Clinical features and treatment of peristomal pyoderma
gangrenosum. JAMA 2000;284:1546–8.
[19] Lally A, Hollowood K, Bunker CB, et al. Penile pyoderma gangrenosum treated with topical
tacrolimus. Arch Dermatol 2005;141:1175–6.
[20] Hernandez-Martin A, Arias-Paloma D, Hermida G, et al. Oral pyoderma gangrenosum. Br J
Dermatol 2003;149:663–4.
[21] Setterfield JF, Shirlaw PJ, Challacombe SJ, et al. Pyoderma gangrenosum associated with
severe oropharyngeal involvement and IgA paraproteinemia. Br J Dermatol 2001;144:
[22] Lachapelle JM, Marot L, Jablonska S. Superficial granulomatous pyoderma gangrenosum
of the face successfully treated by cyclosporin: a long-term follow-up. Dermatology 2001;
[23] Banga F, Schuitemaker N, Meijer P. Pyoderma gangrenosum after caesarean section: a case
report. Reprod Health 2006;3:1–5.
[24] Brown TS, Marshall GS, Callen JP. Cavitating pulmonary infiltrate in an adolescent with
pyoderma gangrenosum: a rarely recognized extracutaneous manifestation of a neutrophilic
dermatosis. J Am Acad Dermatol 2000;43:108–12.
[25] Yoshida C, Kojima H, Shigaki T, et al. Association of pyoderma gangrenosum and sterile
osteomyelitis in a patient having myelodysplastic syndrome. Eur J Heamatol 2004;72:
[26] Nurre LD, Rabalais GP, Callen JP. Neutrophilic dermatosis-associated sterile chronic
multifocal osteomyelitis in pediatric patients: case report and review. Pediatr Dermatol

[27] Wilson DM, John GR, Callen JP. Peripheral ulcerative keratitis- an extracutaneous neutro-
philic disorder: report of a patient with rheumatoid arthritis, pustular vasculitis, pyoderma
gangrenosum, and Sweet’s syndrome with an excellent response to cyclosporine therapy.
J Am Acad Dermatol 1999;40:331–4.
[28] Vadillo M, Jucgla A, Podzamczer D, et al. Pyoderma gangrenosum with liver, spleen, and
bone involvement in a patient with chronic myelomonocytic leukemia. Br J Dermatol
[29] Muster AJ, Bhaarati S, Herman JJ, et al. Fatal cardiovascular disease and cutis laxa follow-
ing acute febrile neutrophilic dermatosis. J Pediatr 1983;102:243–8.
[30] Marie I, Levesque H, Joly P, et al. Neutrophilic myositis as an extracutaneous manifestation
of neutrophilic dermatosis. J Am Acad Dermatol 2001;44(1):137–9, 2001.
[31] Vignon-Pennamen MD. The extracutaneous involvement in the neutrophilic dermatoses.
Clin Dermatol 2000;18(3):339–47.
[32] Trent JT, Kirsner RS. Diagnosing pyoderma gangrenosum. Adv Skin Wound Care 2001;14:
[33] Schmid MH, Hary C, Marstaller B, et al. Pyoderma gangrenosum associated with secondary
antiphospholipid syndrome. Eur J Dermatol 1998;1:45–7.
[34] Smith JB, Shenefelt PD, Soto O, et al. Pyoderma gangrenosum in a patient with cryoglobu-
linemia and hepatitis C successfully treated with interferon alfa. J Am Acad Dermatol 1996;
[35] Barrio VR, Sanfilippo AM, Malone JC, et al. Nonhealing ulcer secondary to factor V Leiden
mutation and cryofibrinogenemia. J Am Acad Dermatol 2004;51(Suppl 5):S194–6.
[36] Byrd DR, El-Azhary RA, Gibson LE, et al. Sporotrichosis masquerading as pyoderma
gangrenosum: case report and review of 19 cases of sporotrichosis. J Eur Acad Dermatol
Venereol 2001;15:581–4.
[37] Laochumroonvorapong P, DiCostanzo DR, Wu H, et al. Disseminated histoplasmosis
presenting as pyoderma gangrenosum-like lesions in a patient with acquired immunodefi-
ciency syndrome. Int J Dermatol 2001;40:518–21.
[38] Yang CC, Hsieh FS, Lee JY. Pyoderma gangrenosum complicated by ecthyma gangreno-
sum. Br J Dermatol 2004;150:1025–6.
[39] Oh CC, McKenna DB, McLaren KM, et al. Factitious panniculitis masquerading as
pyoderma gangrenosum. Clin and Exp Dermatol 2005;30:253–5.
[40] Weenig RH, Davis MD, Dahl PR, et al. Skin ulcers misdiagnosed as pyoderma gangreno-
sum. N Engl J Med 2002;347(18):1412–8.
[41] Su WPD, Davis MD, Weenig RH, et al. Pyoderma gangrenosum: clinicopathologic correla-
tion and proposed diagnostic criteria. Int J Dermatol 2004;43:790–800.
[42] Malone JC, Slone SP, Wills-Frank LA, et al. Vascular inflammation (vasculitis) in Sweet
syndrome: a clinicopathologic study of 28 biopsy specimens from 21 patients. Arch Derma-
tol 2002;138(3):345–9.
[43] Farina MC, Requena L, Domine M, et al. Histopathology of cutaneous reaction to granu-
locyte colony-stimulating factor: another pseudomalignancy. J Cutan Pathol 1998;25(10):
[44] Brown TS, Callen JP. Atypical presentation of herpes simplex virus in a patient with chronic
lymphocytic leukemia. Cutis 1999;64(2):123–5.
[45] Gettler S, Rothe M, Grin C, et al. Optimal treatment of pyoderma gangrenosum. Am J Clin
Dermatol 2003;4:597–608.
[46] Powell FC, O’Kane M. Management of pyoderma gangrenosum. Dermatol Clin 2002;20:
[47] de Imus G, Golomb C, Wilkel C, et al. Accelerated healing of pyoderma gangrenosum
treated with bioengineered skin and concomitant immunosuppression. J Am Acad Dermatol
[48] Chiba T, Isomura I, Suzuki A, et al. Topical tacrolimus therapy for pyoderma gangrenosum.
J Dermatol 2005;32:199–203.

[49] Vidal D, Alomar A. Successful treatment of peristomal pyoderma gangrenosum using

topical tacrolimus. Br J Dermatol 2004;150:387–8.
[50] Nybaek H, Olsen AG, Karlsmark T, et al. Topical therapy for peristomal pyoderma
gangrenosum. J Cutan Med Surg 2004;8:220–3.
[51] Patel GK, Thodes JR, Evans B, et al. Successful treatment of pyoderma gangrenosum with
topical 0.5% nicotine cream. J Dermatolog Treat 2004;15:122–5.
[52] Langenbach N, Goetz A, Hohenleutner U, et al. Effectiveness of 4% disodium cromoglycate
in the treatment of disseminated pyoderma gangrenosum. Acta Derm Venereol 1996;76:
[53] Zakhireh M, Rockwell WB, Fryer RH. Stabilization of pyoderma gangrenosum ulcer with
oral cyclosporine prior to skin grafting. Plast Reconstr Surg 2004;113:1417–20.
[54] Patrone P, Bragadin G, De Francesco V, et al. Pyoderma gangrenosum of the scalp treated
with cyclopsorine A. Int J Dermatol 2002;41:916–8.
[55] Schofer H, Baur S. Successful treatment of postoperative pyoderma gangrenosum with
cyclosporin. J Eur Acad Dermatol Venereol 2002;16:148–51.
[56] Daniels NH, Callen JP. Mycophenolate mofetil is an effective treatment for peristomal
pyoderma gangrenosum. Arch Dermatol 2004;140:1427–9.
[57] Lee MR, Cooper AJ. Mycophenolate mofetil in pyoderma gangrenosum. J Dermatolog
Treat 2004;15:303–7.
[58] Nousari HC, Lynch W, Anhalt GJ, et al. The effectiveness of mycophenolate mofetil in
refractory pyoderma gangrenosum. Arch Dermatol 1998;134:1509–11.
[59] Teitel AD. Treatment of pyoderma gangrenosum with methotrexate. Cutis 1996;57:326–8.
[60] Burruss JB, Farmer ER, Callen JP. Chlorambucil is an effective corticosteroid-sparing agent
for recalcitrant pyoderma gangrenosum. J Am Acad Dermatol 1996;5:720–4.
[61] Federman GL, Federman DG. Recalcitrant pyoderma gangrenosum treated with thalido-
mide. Mayo Clin Proc 2000;75:842–4.
[62] Kontochristopoulos GJ, Stavropoulos PG, Gregoriou S, et al. Treatment of pyoderma
gangrenosum with low-dose colchicine. Dermatology 2004;209:233–6.
[63] Reynoso-von Drateln C, Perla-Navarro AV, Gamez-Nava JI, et al. Intravenous cyclophos-
phamide pulses in pyoderma gangrenosum: an open trial. J Rheumatol 1997;24:689–93.
[64] Fox LP, Pandya AG. Pulse intravenous cyclophosphamide therapy for dermatologic
disorders. Dermatol Clin 2000;18(3):459–73.
[65] Shenefelt PD. Pyoderma gangrenosum associated with cystic acne and hidradenitis suppu-
rativa controlled by adding minocycline and sulfasalazine to the treatment regimen. Cutis
[66] Mcgowan JW, Johnson CA, Lynn A. Treatment of pyoderma gangrenosum with etanercept.
J Drugs Dermatol 2004;3:441–4.
[67] Jenne L, Sauter B, Thurmann P, et al. Successful treatment of therapy-resistant chronic veg-
etating pyoderma gangrenosum with infliximab. Br J Dermatol 2004;150:380–2.
[68] Mimouni D, Anhalt GJ, Kouba DJ, et al. Infliximab for peristomal pyoderma gangrenosum.
Br J Dermatol 2003;148:813–6.
[69] Stichweh DS, Punaro M, Pascual V. Dramatic improvement of pyoderma gangrenosum with
infliximab in a patient with PAPA syndrome. Pediatr Dermatol 2005;22:262–5.
[70] Bens G, Laharie D, Beylot-Barry M, et al. Successful treatment with infliximab and metho-
trexate of pyostomatitis vegetans associated with Crohn’s disease. Br J Dermatol 2003;
[71] Brooklyn TN, Dunnill GS, Shetty A, et al. Infliximab for the treatment of pyoderma gangre-
nosum: a randomised, double-blind placebo-controlled trial. Gut 2006;55:505–9.
[72] Hewitt D, Tan C. Use of infliximab in pyoderma gangrenosum. Australas J Dermatol 2007;
[73] Kouklakis G, Moschos J, Leontiadis GI, et al. Infliximab for treatment of pyoderma gangre-
nosum associated with clinically inactive Crohn’s disease: a case report. Rom J Gastroen-
terol 2005;14:401–3.

[74] Rispo A, Scarpa R, Di Girolamo E, et al. Infliximab in the treatment of extra-intestinal

manifestations of Crohn’s disease. Scand J Rheumatol 2005;34:387–91.
[75] Hubbard VG, Friedmann AC, Goldsmith P. Systemic pyoderma gangrenosum responding
to infliximab and adalimumab. Br J Dermatol 2005;152:1059–61.
[76] Hagman JH, Carrozzo AM, Campione E, et al. The use of high-dose immunoglobulin in the
treatment of pyoderma gangrenosum. J Dermatolog Treat 2001;12:19–22.
[77] Dirschka T, Kastner U, Behrens S, et al. Successful treatment of pyoderma gangrenosum
with intravenous human immunoglobulin. J Am Acad Dermatol 1998;39:789–90.
[78] Zhang X, He Y, Zhou H, et al. A case of pyoderma gangrenosum responding to high-dose
intravenous immunoglobulin therapy. Chin Med J 2006;119:1230–2.
Rheum Dis Clin N Am 33 (2007) 803–817

Retroperitoneal Fibrosis: Evolving

Augusto Vaglio, MDa,*, Alessandra Palmisano, MDa,
Domenico Corradi, MDb, Carlo Salvarani, MDc,
Carlo Buzio, MDa
Dipartimento di Clinica Medica, Nefrologia e Scienze della Prevenzione,
Università degli Studi di Parma, Via Gramsci 14, 43100 Parma, Italy
Dipartimento di Patologia e Medicina di Laboratorio, Sezione di Anatomia Patologica,
Università degli Studi di Parma, Via Gramsci 14, 43100 Parma, Italy
Servizio di Reumatologia, Arcispedale Santa Maria Nuova,
Viale Risorgimento, 80, 42100, Reggio Emilia, Italy

The term ‘‘retroperitoneal fibrosis’’ (RPF) commonly refers to a clinico-

pathologic entity characterized by a mainly sclerotic tissue that develops in
the periaortic and peri-iliac retroperitoneum and often encases structures,
such as the ureters and the inferior vena cava [1,2]. RPF does not identify
a single disease entity but rather includes a wide spectrum of diseases that
may or may not have known causes: RPF is idiopathic in most cases but
can also be secondary to certain medications, malignancies, infections,
and other triggering factors [3]. In addition, idiopathic RPF may be isolated,
or develop in the setting of a systemic disorder characterized by fibrosis in
other organs.
RPF is associated with an inflammatory infiltrate with peculiar histo-
pathologic aspects, rich in lymphocytes, plasma cells, and macrophages
[4], along with other abnormalities, such as the high frequency of circulating
autoantibodies [5]. As described below, idiopathic RPF also has an HLA
association (HLA-DRB1*03) [6]. These features lend support to the idea
that RPF involves a multifaceted systemic disease whose pathogenesis is
mediated by immunopathologic mechanisms [7], rather than merely a scle-
rotic reaction to some still-undefined insult that happens to be localized
to the retroperitoneum.

Drs. Salvarani and Buzio share senior authorship.

* Corresponding author.
E-mail address: augusto.vaglio@virgilio.it (A. Vaglio).

0889-857X/07/$ - see front matter Ó 2007 Elsevier Inc. All rights reserved.
doi:10.1016/j.rdc.2007.07.013 rheumatic.theclinics.com
804 VAGLIO et al

We review the clinical and pathogenetic aspects of RPF and discuss the
most relevant diagnostic and therapeutic strategies for this unusual

Nosologic considerations
Clear definitions for the different forms included in the spectrum of RPF
have not been developed. This lack of clarity probably reflects the rarity of
this condition and the consequent lack of classification or diagnostic crite-
ria. When approaching a patient suspected of having RPF, however, it is
mandatory to differentiate the idiopathic from the secondary form of the
disease, because the treatment implications differ substantially. Several dif-
ferent conditions are potential causes of secondary RPF: the use of certain
medications (eg, methysergide, ergot alkaloids, dopamine agonists), primary
retroperitoneal cancer (eg, lymphoma, sarcoma), retroperitoneal metastatic
disease (eg, carcinoid, various carcinomas), trauma, radiotherapy, major
abdominal surgery, and infections [1,3,8].
Idiopathic RPF does not have any identifiable causes. It usually presents
as a retroperitoneal mass surrounding an undilated abdominal aorta or
encasing adjacent structures. Idiopathic RPF is included in the spectrum
of chronic periaortitis along with certain other conditions of unknown
cause, such as inflammatory abdominal aortic aneurysms and perianeurys-
mal retroperitoneal fibrosis [9]. All of these clinical entities not only demon-
strate substantial overlap in their symptoms, signs, and laboratory features,
such as abdominal or lumbar pain and high inflammatory markers, but also
share a common histopathologic appearance.
Among the several forms of chronic periaortitis distinguished from each
other, one feature separates idiopathic RPF from the other different entities
nosologically: in idiopathic RPF, the abdominal aorta is not dilated. In con-
trast, the retroperitoneal fibrotic reactions associated with inflammatory
aneurysms and perianeurysmal fibrosis are, by definition, situated around
an aneurysmal aorta [9]. The pathophysiologic implications of the presence
of an aneurysm are considered in later discussion.
Some authors propose to distinguish perianeurysmal fibrosis from
inflammatory aneurysms on the basis of the presence of obstructive compli-
cations in adjacent structures in perianeurysmal fibrosis, such as the ureters
[10]. The authors believe that these conditions may simply be referred to as
aneurysmal forms of chronic periaortitis, however, as opposed to the non-
aneurysmal counterpart (ie, idiopathic RPF) [1,11].
Idiopathic RPF has been reported in several patients suffering from
autoimmune or inflammatory diseases involving other organs or structures,
such as autoimmune thyroiditis, autoimmune pancreatitis, and systemic
lupus erythematosus [5,12–14]. It has been suggested that in these cases
the overlapping diseases have common pathogenetic aspects, but a clear
causal relationship has not been demonstrated. The authors thus prefer to

define these forms of idiopathic RPF as associated with (rather than second-
ary to) other autoimmune diseases. Finally, in a subset of patients,
idiopathic RPF arises in the context of a systemic sclerosing disorder that
can also feature sclerosing (autoimmune) pancreatitis, mediastinal fibrosis,
pseudotumor of the orbit, Riedel’s thyroiditis, chronic sialoadenitis, or scle-
rosing cholangitis. This rare condition has been termed multifocal fibroscle-
rosis [12,15,16]. The recent findings of high serum levels of IgG4 and
abundant IgG4-bearing plasma cells in the inflammatory infiltrates of the
diseased sites led some investigators to coin the term ‘‘hyper-IgG4
syndrome’’ [17].

Epidemiology, genetic factors, and environmental factors

The main epidemiologic characteristics of idiopathic RPF were investi-
gated in a 2004 study conducted in Finland, which reported an annual inci-
dence of 0.1 per 100,000 people and a prevalence of 1.4 per 100,000
inhabitants [18]. The incidence peaked in patients 40 to 60 years of age.
There seems to be a male predominance in this disease, with a male/female
ratio of between 2:1 and 3:1 [18,19]. No data are available concerning the
prevalence and incidence of secondary RPF. Secondary cases are believed
to represent less than one third of all RPF cases [3].
Immunogenetic factors may play a role in RPF. In a recent case-control
study, the HLA-DRB1*03 and (to a lesser extent) HLA-B*08 alleles were
associated with chronic periaortitis [6]. In addition, a study that focused
on the perianeurysmal forms of the disease (ie, inflammatory abdominal
aortic aneurysms) found a genetic risk determinant at the HLA-DRB1 locus
and identified HLA-DRB1*15 and B1*0404 as predisposing alleles [20].
Taken together, these findings provide evidence for a role of the HLA
system in conferring susceptibility to RPF.
In addition to genetic factors, environmental or exogenous agents may
contribute to the development of RPF. A recent case-control study demon-
strated that smoking and occupational exposure to asbestos significantly
increase the risk for developing RPF. Moreover, the relative risk for RPF
increased according to cumulative asbestos exposure [18].

Clinical manifestations and laboratory findings

Idiopathic RPF patients usually present with abdominal, lumbar, or
flank pain, which can be persistent, insidious, and dull, but sometimes (par-
ticularly in patients who have severe ureteral involvement) acute and
colicky. The pain usually responds to nonsteroidal anti-inflammatory drugs
(NSAIDs), but the efficacy of these medications in intervention is typically
short-lived [1,19]. In addition to pain, constitutional symptoms, such as
low-grade fever, weight loss, anorexia, and fatigue, often herald the onset
of idiopathic RPF and are believed to reflect the systemic inflammatory
806 VAGLIO et al

nature of the disease [7]. Less common manifestations include constipation,

testicular pain, varicocele, hydrocele, claudication, edema, and deep vein
thrombosis of the lower limbs. All of these manifestations are usually attrib-
uted to the compressive effects of the retroperitoneal mass on arterial,
venous, and lymphatic retroperitoneal structures. The main clinical manifes-
tations of idiopathic RPF are summarized in Box 1.
In cases with advanced bilateral obstructive uropathy, manifestations
related to oliguria and uremia may predominate: hypertension, fluid and
electrolyte disturbances, anemia, nausea, and vomiting [21]. Patients may
also complain of urinary frequency and dysuria, even in the absence of
urinary tract infections [22]. Obstruction of both ureters occurs in a high
percentage of casesd50% or greater in two separate studies [22,23]. The ob-
structions may occur in either a simultaneous or metachronous manner.
Some patients who have idiopathic RPF have atrophic kidneys at the
time of diagnosis, probably as a result of asymptomatic but longstanding
hydronephrosis. The retroperitoneal tissue in this disease often involves
the origin of the renal arteries. Renal artery stenosis must also be excluded,
particularly if the patient is hypertensive.
Laboratory findings are often nonspecific in RPF. The erythrocyte sedi-
mentation rates (ESR) and C-reactive protein (CRP) levels are usually
elevateddsometimes strikingly sodand other acute-phase reactant levels,
such as ferritin, may also be high [1,19,22]. Polyclonal hypergammaglobuli-
nemia is observed in some patients [19]. The degree of renal failure generally
depends on the extent of ureteral involvement, but anecdotal cases of renal

Box 1. Major clinical manifestations of idiopathic retroperitoneal

Pain (back, flank, abdominal)
Constitutional symptoms (fatigue, weight loss, anorexia,
low-grade fever)
Moderately frequent
Testicular pain, varicocele, hydrocele
Nausea, vomiting
Leg swelling, deep vein thrombosis of the lower limbs
Urinary frequency, polyuria
Erectile dysfunction

parenchymal disease (eg, rapidly progressive glomerulonephritis) causing

functional renal impairment have also been reported in association with
idiopathic RPF [24]. Renal dysfunction and the systemic inflammatory
response contribute to the development of a normochromic, normocytic
Autoantibodies are often detected in patients who have idiopathic RPF.
Antinuclear antibodies (ANA) are found in up to 60% of the cases [5], but
anti–smooth muscle antibodies and rheumatoid factor (RF) are also not
uncommon. In most cases the autoantibodies are not organ specific. Their
titers are generally low, and their presence contributes little to narrowing
the differential diagnosis. Such autoantibodies probably reflect global de-
rangement in self-tolerance [1,5,21]. The absence of certain autoantibodies
is useful in excluding certain systemic conditions with which RPF can be
associated. (For example, the absence of ANA essentially excludes systemic
lupus erythematosus).
Several case reports have described the occurrence of antineutrophil
cytoplasmic antibodies (ANCA) in RPF. Only a few of these patients
have had clinical features recognizable as Wegener granulomatosis or micro-
scopic polyangiitis [24–26]. Anti-thyroid microsomal and anti-thyroglobulin
antibodies are positive in a subset of patients who have idiopathic RPF,
sometimes indicating the presence of autoimmune thyroiditis [27]. Finally,
other autoantibodies, such as those directed against dsDNA and the
Ro(SSA)/La(SSB) antigens, have been reported in patients who have
idiopathic RPF associated with systemic lupus erythematosus [13] and
Sjögren’s syndrome [28], respectively.
Unfortunately, no laboratory findings have proved to be helpful in differ-
entiating idiopathic from secondary RPF. As discussed later, imaging stud-
ies and histopathology are ultimately required to define the subtype of RPF
more precisely.

Idiopathic RPF grossly appears as a firm grayish mass surrounding the
abdominal aorta and the iliac arteries and usually extends from the origin
of the renal arteries to the caudal portion of the common iliac vessels.
The soft tissues surrounding the thoracic aorta may also be involved by
a process termed mediastinal fibrosis. In addition, in a minority of cases
localized fibrosis occurs in atypical areas, such as the peri-iliac, peri-splenic,
and peri-pancreatic regions [2,4,29].
Microscopic examination of retroperitoneal specimens in idiopathic RPF
shows the coexistence of fibrous and inflammatory components (Fig. 1A).
The fibrous part consists of an abundant, sclerotic matrix composed of
type I collagen and a population of spindle-shaped cells whose immunohis-
tochemical characteristics are those of fibroblasts and myofibroblasts. The
spindle-shaped cells stain positively for vimentin and, in the more cellular
808 VAGLIO et al

Fig. 1. (A) Light microscopic examination of a retroperitoneal specimen taken from a patient
who has idiopathic retroperitoneal fibrosis shows an admixture of sclerotic tissue and mononu-
clear cell inflammatory infiltrate; the inflammatory cells are diffuse and arranged in perivascular
aggregates (asterisk). Hematoxylin and eosin. (B) Higher magnification of the area marked with
the asterisk in panel A: an aggregate of small lymphocytes, plasma cells and scattered eosino-
phils is organized around a retroperitoneal vessel (arrow). Hematoxylin and eosin. (C) Mono-
nuclear inflammatory cells are diffusely interspersed within thick collagen bundles. Hematoxylin
and eosin.

areas, smooth-muscle actin [29]. The fibroblast population does not show
any signs of active proliferation. The Ki-67 proliferation marker is usually
negligible. In addition, these cells do not express ALK-1, b-catenin, or des-
min, the types of markers characteristic of other fibrous tissue-producing
disorders. The fibrous tissue seems to be distributed randomly, but also en-
circles small retroperitoneal vessels and nerves.
The inflammatory component consists of lymphocytes, plasma cells,
macrophages, and scattered eosinophils [2,9,29]. Granulocytes are rare.
Two main inflammatory patterns, namely perivascular and diffuse, usually
coexist. In the former, inflammatory cells are aggregated around small
retroperitoneal vessels (see Fig. 1B). Immunohistochemical studies show
a target-like appearance of these aggregates, with a central core of B cells
and a periphery of CD4þ and CD8þ T cells. In some cases, the inflamma-
tory aggregates show germinal center reactions. In contrast, the diffuse

pattern is characterized by inflammatory infiltration of the narrow spaces

within the collagen bundles (see Fig. 1C), with T cells outnumbering B cells
A recent histopathologic study on 24 idiopathic RPF patients showed
vasculitis of small retroperitoneal vessels in about half of the cases. In
addition, 90% of the CD138þ plasma cells identified also stained for
IgG4, which normally represents the rarest IgG subclass [29]. The main his-
topathologic and immunohistochemical features of idiopathic RPF are
summarized in Box 2.
The aortic wall also shows pathologic changes in idiopathic RPF and in
the other forms included in the spectrum of chronic periaortitis. These
changes include prominent adventitial inflammation (with an inflammatory
infiltrate that is strikingly similar to the retroperitoneal findings), medial
thinning, and, in most cases, advanced medial and intimal atherosclerosis
In the secondary forms of RPF, a careful search for neoplastic cells may
steer the diagnosis toward a primary neoplastic process. Clonality of the
inflammatory infiltrate or the presence of Reed-Sternberg cells suggests
the presence of an underlying lymphoma [29]. In addition, finding even
a few lipoblasts should heighten suspicion of an underlying liposarcoma.
The presence of granulomas, which is exceedingly rare in idiopathic RPF,
usually indicates a primary process (eg, Wegener granulomatosis or tubercu-
losis) [25,31].

The pathogenesis of idiopathic RPF is discussed separately from that of
secondary disease.

Box 2. Main histopathologic and immunohistochemical features

of idiopathic retroperitoneal fibrosis
Abundant fibrous background of type 1 collagen, with fibroblasts/
myofibroblasts without atypia or signs of active proliferation
Chronic inflammatory infiltrate, coexistence of diffuse and
nodular (perivascular) inflammatory patterns
Target-like appearance of lymphoid aggregates (core of B cells,
periphery of T cells)
IgG4-bearing plasma cells
Vasculitis of small retroperitoneal vessels
Perivascular and perineural fibrosis
Negative staining for ALK-1, b-catenin, desmin
810 VAGLIO et al

Idiopathic retroperitoneal fibrosis

The pathogenesis of idiopathic RPF is largely unknown. A series of
studies led by Parums and Mitchinson [9,30,32] found a correlation between
advanced atherosclerosis of the abdominal aorta, idiopathic RPF, and other
forms of chronic periaortitis. These investigators hypothesized that the
pathogenesis of several subtypes of RPF may relate in part to an inflamma-
tory reaction against antigens within atherosclerotic plaques. The proposed
antigens include ceroid and oxidized low density lipoproteins (LDL)
[9,30,32]. Assuming the thinning or breach of the medial aortic layer, these
antigens could be processed by adventitial macrophages and then presented
to lymphocytes, leading to lymphocyte activation and immunoglobulin
production. This same research group demonstrated the deposition of IgG
in proximity to extracellular ceroid, and reported serum levels of antibodies
to oxidized-LDL and ceroid that were higher in patients who had chronic
periaortitis compared with controls [32,33].
Other factors certainly contribute to the pathogenesis of idiopathic RPF.
Indeed, the concept of a local inflammatory response to atherosclerosis
cannot explain why a large proportion of patients present with features of
a systemic disease (eg, constitutional symptoms such as fatigue, weight
loss, fever, and high acute-phase reactants) and autoimmunity involving
other organs (eg, the thyroid) [5]. In addition, the disease also develops in
patients who do not have evidence of atherosclerosis, and pediatric cases
have been reported also [34]. Idiopathic RPF probably results from an inter-
action between genetic and environmental determinants.
The finding of a strong association with HLA-DRB1*03 suggests that the
HLA repertoire is restricted and thus that only selected antigens trigger the
disease [6]. HLA-DRB1*03 is also a risk factor for several autoimmune con-
ditions: autoimmune thyroiditis, type 1 diabetes mellitus, and systemic lupus
erythematosus. Its association with idiopathic RPF thus lends support to
the hypothesis of an autoimmune origin of the disease. Among microbial
agents, cytomegalovirus was described as one of the potential etiologic
agents in inflammatory abdominal aortic aneurysms [35], but additional
work is warranted to address this issue.
Patients who have idiopathic RPF often have autoantibodies in their
serum, and most autoimmune inflammatory disorders that overlap with
RPF also demonstrate autoantibody production. Examples include autoim-
mune thyroid disease and ANCA-associated vasculitis [24,27]. In addition,
B cells are abundant in retroperitoneal lesions [29]. Together with the
evidence of an association with HLA-DRB1*03, these findings suggest
that B cells are likely to play a central role in the pathogenesis of the disease.
In a recent study, we have found that autoantibodies reacting against fibro-
blasts, which have already been described in systemic sclerosis [36], are
detectable in the sera of about 30% of patients who have idiopathic RPF.
These antibodies are internalized in vitro by retroperitoneal fibroblasts

and enhance the production of interleukin-6 in a dose-dependent manner

(N. Ronda and colleagues, unpublished data, 2007).
Idiopathic RPF and the other chronic periaortitis forms are not necessar-
ily limited to the abdominal aorta and iliac arteries but can also involve the
thoracic aorta and the origin of its major branches [37]. This vascular
pattern is similar to that of large-vessel vasculitic diseases, such as Takaya-
su’s arteritis and giant cell arteritis [38]. Furthermore, pronounced adventi-
tial inflammation and perivascular infiltrates surrounding aortic vasa
vasorum have been observed in chronic periaortitis and in vascular biopsies
from patients who have Takayasu’s arteritis and giant cell arteritis [39].
These findings raise the question of whether chronic periaortitis originates
as a primary aortitis and subsequently induces a fibroinflammatory reaction
in the surrounding soft tissues [1]. A recent study of endothelial damage in
chronic periaortitis demonstrated that the number of circulating endothelial
cells was higher in chronic periaortitis than in healthy individuals and
patients who had atherosclerosis, and that the number correlated with
disease activity [40].
Finally, a potential pathogenetic role for IgG4-bearing plasma cells in
idiopathic RPF has been suggested. High serum IgG4 levels were detected
initially in patients who had isolated autoimmune (sclerosing) pancreatitis
[41] and in cases of that disorder associated with RPF [16]. Subsequent stud-
ies have demonstrated numerous infiltrating IgG4-positive plasma cells at
disease sites of patients who had multifocal fibrosclerosis of pancreatic,
bile, and salivary ducts [12]. We have also reported recently that such
plasma cells are abundant in the retroperitoneal biopsies of patients who
have idiopathic RPF patients [29]. No data are available concerning the
serum IgG4 levels of patients who have idiopathic RPF but do not have
other fibrotic lesions. The overall interpretation of these findings points to
a pathogenetic link between IgG4-positive plasma cells and the development
of sclerotic lesions. These findings require confirmation in other studies.

Secondary retroperitoneal fibrosis

The pathogenesis of the secondary forms of RPF largely depends on the
underlying cause. Primary retroperitoneal neoplasms (eg, lymphoma,
sarcoma) and metastatic retroperitoneal tumors (eg, breast and colon carci-
nomas) may show a pronounced desmoplastic reaction and thus mimic
RPF. In some cases, the neoplastic disorder does not metastasize to the
retroperitoneum but rather (as in the case of carcinoid) causes RPF by
releasing fibrogenic compounds, such as serotonin and growth factors
[1,8]. The use of certain drugs, such as methysergide and other ergot deriv-
atives, has traditionally been associated with the development of secondary
RPF; the pathophysiologic mechanism is uncertain but is likely to involve
a serotonin-mediated pathway [1]. Finally, RPF may also arise as a sclerotic
response to radiotherapy, trauma, or surgical injury. Less common causes
812 VAGLIO et al

include rare infiltrative diseases, such as the non-Langerhans histiocytosis

form named Erdheim-Chester disease [42].

Imaging findings and differential diagnosis

Computed tomography (CT) and magnetic resonance imaging (MRI) are
considered the modalities of choice for the diagnosis of idiopathic RPF. Ret-
roperitoneal biopsy is usually performed only in cases with atypical
localizations or when clinical or laboratory findings suggest an underlying in-
fection or malignancy. Biopsy is also performed for diagnostic confirmation
whenever patients undergo surgery for ureterolysis or aneurysmectomy [1].
CT shows the presence of a periaortic mass, which usually develops
around the anterior and lateral sides of the abdominal aorta and often
causes medial deviation and extrinsic compression of the ureters. The
craniocaudal extension of the mass commonly ranges from the origin of
the renal arteries to the common iliac vessels, but in some cases may also
involve the superior mesenteric artery or the internal and external iliac ves-
sels. The inferior vena cava is often encased and compressed, and collateral
venous circles are not uncommon [43,44]. On CT, idiopathic RPF appears
isodense to muscle and shows pronounced enhancement with contrast
administration, particularly in early disease stages when the inflammatory
component is robust and the mass more cellular and vascular (Fig. 2).
MRI in patients who have idiopathic RPF demonstrates hypointensity
on T1-weighted images but high intensity on T2-weighted images in the
early, active stages [43]. The degree to which either contrast enhancement
on CT or signal intensity on MRI T2-weighted images correlates with
disease activity remains unclear, however.
Although not useful for the diagnosis, sonography is a noninvasive
modality that is used routinely to monitor the extent of

Fig. 2. (A) Computed tomography of the abdomen in a patient who has idiopathic retroperi-
toneal fibrosis shows a thick and homogeneous periaortic tissue (arrow), isodense to muscle,
which develops around the anterior and lateral sides of the abdominal aorta. (B) The mass
shown in panel A extends caudally to encircle both common iliac arteries (arrow).

ureterohydronephrosis and is therefore frequently performed to indirectly

assess response to treatment. In addition, sonography may be useful in peri-
aneurysmal forms of chronic periaortitis as a means of tracking the size of
aortic aneurysms.
In recent years, 18F-fluorodeoxyglucose (FDG) positron emission tomog-
raphy (PET) has emerged as a potentially useful imaging tool in several
inflammatory disorders, including large vessel vasculitides and RPF
[45,46]. Although prospective studies are lacking, 18FDG-PET is noninva-
sive and safe (albeit expensive). In addition, because it allows whole-body
imaging, it may reveal other diseased sites, such as in cases of thoracic aorta
involvement [37] or when other autoimmune-inflammatory disorders (eg,
thyroiditis) are associated with RPF [47]. More importantly, whole-body
F-FDG-PET may reveal neoplastic or infectious processes to which
RPF may be secondary. The proper interpretation of increased 18FDG
uptake in most clinical situations associated with RPF is far from clear,
however, and further studies are required.
CT and MRI also yield insights that may help differentiate idiopathic
from secondary RPF. It has been suggested that an inhomogeneous signal
on T2-weighted images implies malignant RPF [48]. In addition, anterior
displacement of the aorta and lateral displacement of the ureters by the
retroperitoneal mass more often occur in malignant than in idiopathic cases
[43]. Findings such as infiltration of adjacent bone or muscle structures
should also heighten suspicion of a secondary form.
A wide spectrum of fibroinflammatory disorders can mimic idiopathic
RPF on imaging studies. Inflammatory myofibroblastic tumors (which
occur more frequently in children than in adults) usually appear as an
infiltrative soft-tissue mass with variable location and heterogeneous atten-
uation on CT studies. Histology and immunohistochemistry are required to
confirm the diagnosis. Immunohistochemistry studies in the setting of
myofibroblastic tumors reveal positive staining for ALK-1 and desmin
[29]; such studies are consistently negative in idiopathic RPF.
Desmoid-type fibromatosis, frequently associated with Gardner syn-
drome (a variant of familial adenomatous polyposis), commonly presents
as an intra-abdominal soft tissue mass with mass effect that displaces adja-
cent structures. Histology and immunohistochemistry are also required in
these cases. Distinctive findings include the relative absence of inflammation
and positive staining for b-catenin [29,49]. Sclerosing mesenteritis, often
considered a variant of idiopathic RPF, typically involves the small bowel
mesentery and lacks the periaortic disposition of idiopathic RPF; moreover,
it causes small bowel fixation and bowel obstruction but usually spares the
ureters [49]. Finally, Erdheim-Chester disease, a rare form of non–Langer-
hans cell histiocytosis, involves not only the periaortic soft tissues but also
the perirenal space. In such cases, bone scintigraphy may reveal long bone
involvement and biopsy of the affected tissue shows foamy histiocytes that
stain positive for CD68 but negative for CD1a and S-100 [42].
814 VAGLIO et al

Treatment and prognosis

First-line therapies
The first goal of treatment in idiopathic RPF is the relief of ureteral
obstruction, if significant impairment in the function of one or both kidneys
has been documented. This relief can be achieved by way of an open surgical
approach or by conservative techniques, such as percutaneous nephrostomy
tubes or double-J ureteral stents placed at cystoscopy [1]. The conservative
approach followed by administration of medical therapy (eg, glucocorti-
coids) has recently become popular [22]. Surgery, which usually consists
of ureterolysis followed by omental wrapping of the ureters, has been advo-
cated by others because it offers the opportunity to perform multiple
biopsies of the mass and probably reduces the risk for recurrent ureteral
obstruction [50].
Once ureteral obstruction has been relieved, medical treatment is usually
given. The aims of medical therapy are to switch off the acute-phase
systemic inflammatory response, reduce retroperitoneal inflammation, and
inhibit the progression of fibrosis [1]. Although no prospective randomized
trials have been performed so far, glucocorticoids are the mainstay of ther-
apy for idiopathic RPF. The optimal dose and duration of glucocorticoid
therapy still need to be established. Different regimens have been proposed:
in a case series involving 11 patients, glucocorticoids were administered at
an initial dose of 60 mg every other day for 2 months, and tapered over 5
months to 5 mg daily. In that study, 9 patients reported remission of symp-
toms and regression of the retroperitoneal mass [23]. In a recent study on
a larger cohort of patients, glucocorticoids were given at an initial dose of
60 mg daily for 6 weeks and tapered over the following 2 to 3 months to
10 mg daily. The overall duration of treatment was 1 year. Twenty-two of
the 24 enrolled patients reported significant reduction of symptoms and
19 experienced partial or complete regression of the mass. Disease relapses
occurred in 13 patients following the withdrawal of glucocorticoids [22].

Glucocorticoid-sparing agents
Some authors have also proposed treatment regimens using glucocorti-
coids in combination with immunosuppressive agents, such as mycopheno-
late mofetil, cyclophosphamide, and azathioprine. Other experts reserve
these additional medications for patients who have glucocorticoid-resistant
disease, whose disease relapses on tapering, or who have contraindications
to high-dose glucocorticoid treatment.
In a study of 26 patients treated with prednisone in combination with
either cyclophosphamide or azathioprine for 6 to 12 months, all except
one patient (who died of pneumonia) had stable resolution of symptoms
and urinary tract obstruction. Several patients suffered severe, treatment-
related side effects, including leukopenia and sepsis [51]. Recent papers on

single cases or small case series have reported the efficacy and safety of
mycophenolate mofetil in combination with glucocorticoids [52,53]. Finally,
several publications have reported the efficacy of and tolerability of tamox-
ifen, including a recent description of 19 patients, 15 of whom seemed to
benefit from this medication [54].

Monitoring treatment response

Response to treatment can be monitored by the patient’s symptoms, the
levels of acute-phase reactants, such as ESR and CRP, the degree of hydro-
nephrosis on ultrasound, and the serum creatinine levels. A CT/MRI study
is usually performed 2 to 4 months after the beginning of treatment. Failure
to detect radiologic improvement at this time should prompt re-evaluation
of the diagnosis. In some cases, however, clinical improvement may occur
despite the absence of any regression of the mass on cross-sectional imaging
studies [22]. Given its potentially relapsing/remitting course, RPF needs to
be monitored even after treatment withdrawal; relapses may occur up to
several years after treatment has been stopped [1].

We are indebted to Dr. Pietro Schianchi for his help in the preparation of
the figures and for his excellent administrative assistance. We thank all
of our colleagues who participate in clinical research and management of
retroperitoneal fibrosis patients.

[1] Vaglio A, Salvarani C, Buzio C. Retroperitoneal fibrosis. Lancet 2006;367(9506):241–51.
[2] Mitchinson MJ. The pathology of idiopathic retroperitoneal fibrosis. J Clin Pathol 1970;
[3] Koep L, Zuidema GD. The clinical significance of retroperitoneal fibrosis. Surgery 1977;
[4] Hughes D, Buckley PJ. Idiopathic retroperitoneal fibrosis is a macrophage-rich process.
Implications for its pathogenesis and treatment. Am J Surg Pathol 1993;17(5):482–90.
[5] Vaglio A, Corradi D, Manenti L, et al. Evidence of autoimmunity in chronic periaortitis:
a prospective study. Am J Med 2003;114(6):454–62.
[6] Martorana D, Vaglio A, Greco P, et al. Chronic periaortitis and HLA-DRB1*03: another
clue to an autoimmune origin. Arthritis Rheum 2006;55(1):126–30.
[7] Vaglio A, Greco P, Corradi D, et al. Autoimmune aspects of chronic periaortitis. Autoim-
mun Rev 2006;5(7):458–64.
[8] Wu J, Catalano E, Coppola D. Retroperitoneal fibrosis (Ormond’s disease): clinical patho-
logic study of eight cases. Cancer Control 2002;9(5):432–7.
[9] Parums DV. The spectrum of chronic periaortitis. Histopathology 1990;16(5):423–31.
[10] Serra RM, Engle JE, Jones RE, et al. Perianeurysmal retroperitoneal fibrosis. An unusual
cause of renal failure. Am J Med 1980;68(1):149–53.
[11] Vaglio A, Buzio C. Chronic periaortitis: a spectrum of diseases. Curr Opin Rheumatol 2005;
816 VAGLIO et al

[12] Ohara H, Nakazawa T, Ando T, et al. Systemic extrapancreatic lesions associated with
autoimmune pancreatitis. J Gastroenterol 2007;42(Suppl 18):15–21.
[13] Okada H, Takahira S, Sugahara S, et al. Retroperitoneal fibrosis and systemic lupus eryth-
ematosus. Nephrol Dial Transplant 1999;14(5):1300–2.
[14] Demko TM, Diamond JR, Groff J. Obstructive nephropathy as a result of retroperitoneal
fibrosis: a review of its pathogenesis and associations. J Am Soc Nephrol 1997;8(4):684–8.
[15] Kamisawa T, Funata N, Hayashi Y, et al. Close relationship between autoimmune pancre-
atitis and multifocal fibrosclerosis. Gut 2003;52(5):683–7.
[16] Hamano H, Kawa S, Ochi Y, et al. Hydronephrosis associated with retroperitoneal fibrosis
and sclerosing pancreatitis. Lancet 2002;359(9315):1403–4.
[17] Neild GH, Rodriguez-Justo M, Wall C, et al. Hyper-IgG4 disease: report and characterisa-
tion of a new disease. BMC Med 2006;4:23.
[18] Uibu T, Oksa P, Auvinen A, et al. Asbestos exposure as a risk factor for retroperitoneal
fibrosis. Lancet 2004;363(9419):1422–6.
[19] Gilkeson GS, Allen NB. Retroperitoneal fibrosis. A true connective tissue disease. Rheum
Dis Clin North Am 1996;22(1):23–38.
[20] Rasmussen TE, Hallett JW Jr, Metzger RL, et al. Genetic risk factors in inflammatory ab-
dominal aortic aneurysms: polymorphic residue 70 in the HLA-DR B1 gene as a key genetic
element. J Vasc Surg 1997;25(2):356–64.
[21] Jois RN, Gaffney K, Marshall T, et al. Chronic periaortitis. Rheumatology (Oxford) 2004;
[22] van Bommel EF, Siemes C, Hak LE, et al. Long-term renal and patient outcome in idio-
pathic retroperitoneal fibrosis treated with prednisone. Am J Kidney Dis 2007;49(5):615–25.
[23] Kardar AH, Kattan S, Lindstedt E, et al. Steroid therapy for idiopathic retroperitoneal
fibrosis: dose and duration. J Urol 2002;168(2):550–5.
[24] Vaglio A, Manenti L, Allegri L, et al. ANCA-positive periaortic vasculitis: does it fall within
the spectrum of vasculitis? J Intern Med 2002;251(3):268–71.
[25] Izzedine H, Servais A, Launay-Vacher V, et al. Retroperitoneal fibrosis due to Wegener’s
granulomatosis: a misdiagnosis as tuberculosis. Am J Med 2002;113(2):164–6.
[26] Mavragani CP, Voulgarelis M. Retroperitoneal fibrosis and c-ANCA positivity. Clin Rheu-
matol 2007;26(1):115–6.
[27] Armigliato M, Paolini R, Bianchini E, et al. Hashimoto’s thyroiditis and Graves’ disease
associated with retroperitoneal fibrosis. Thyroid 2002;12(9):829–31.
[28] Ghinoi A, Pipitone N, Boiardi L, et al. Primary Sjogren’s syndrome associated with chronic
periaortitis. Rheumatology (Oxford) 2007;46(4):719–20.
[29] Corradi D, Maestri R, Palmisano A, et al. Idiopathic retroperitoneal fibrosis: clinicopatho-
logic features and differential diagnosis. Kidney Int 2007; [epub ahead of print].
[30] Mitchinson MJ. Chronic periaortitis and periarteritis. Histopathology 1984;8(4):589–600.
[31] Greco P, Vaglio A, Corradi D, et al. Tuberculosis as a trigger of retroperitoneal fibrosis. Clin
Infect Dis 2005;41(7):e72–5.
[32] Parums DV, Brown DL, Mitchinson MJ. Serum antibodies to oxidized low-density lipopro-
tein and ceroid in chronic periaortitis. Arch Pathol Lab Med 1990;114(4):383–7.
[33] Parums DV, Chadwick DR, Mitchinson MJ. The localisation of immunoglobulin in chronic
periaortitis. Atherosclerosis 1986;61(2):117–23.
[34] Miller OF, Smith LJ, Ferrara EX, et al. Presentation of idiopathic retroperitoneal fibrosis in
the pediatric population. J Pediatr Surg 2003;38(11):1685–8.
[35] Yonemitsu Y, Nakagawa K, Tanaka S, et al. In situ detection of frequent and active
infection of human cytomegalovirus in inflammatory abdominal aortic aneurysms: pos-
sible pathogenic role in sustained chronic inflammatory reaction. Lab Invest 1996;74(4):
[36] Chizzolini C, Raschi E, Rezzonico R, et al. Autoantibodies to fibroblasts induce a proadhe-
sive and proinflammatory fibroblast phenotype in patients with systemic sclerosis. Arthritis
Rheum 2002;46(6):1602–13.

[37] Salvarani C, Pipitone N, Versari A, et al. Positron emission tomography (PET): evaluation
of chronic periaortitis. Arthritis Rheum 2005;53(2):298–303.
[38] Vanoli M, Daina E, Salvarani C, et al. Takayasu’s arteritis: a study of 104 Italian patients.
Arthritis Rheum 2005;53(1):100–7.
[39] Ramshaw AL, Roskell DE, Parums DV. Cytokine gene expression in aortic adventitial in-
flammation associated with advanced atherosclerosis (chronic periaortitis). J Clin Pathol
[40] Moroni G, Del Papa N, Moronetti LM, et al. Increased levels of circulating endothelial cells
in chronic periaortitis as a marker of active disease. Kidney Int 2005;68(2):562–8.
[41] Hamano H, Kawa S, Horiuchi A, et al. High serum IgG4 concentrations in patients with
sclerosing pancreatitis. N Engl J Med 2001;344(10):732–8.
[42] Veyssier-Belot C, Cacoub P, Caparros-Lefebvre D, et al. Erdheim-Chester disease. Clinical
and radiologic characteristics of 59 cases. Medicine (Baltimore) 1996;75(3):157–69.
[43] Kottra JJ, Dunnick NR. Retroperitoneal fibrosis. Radiol Clin North Am 1996;34(6):
[44] Vivas I, Nicolas AI, Velazquez P, et al. Retroperitoneal fibrosis: typical and atypical mani-
festations. Br J Radiol 2000;73(866):214–22.
[45] Blockmans D. The use of (18F)fluoro-deoxyglucose positron emission tomography in the
assessment of large vessel vasculitis. Clin Exp Rheumatol 2003;21(Suppl 32):S15–22.
[46] Vaglio A, Versari A, Fraternali A, et al. (18)F-fluorodeoxyglucose positron emission tomog-
raphy in the diagnosis and followup of idiopathic retroperitoneal fibrosis. Arthritis Rheum
[47] Drieskens O, Blockmans D, Van den Bruel A, et al. Riedel’s thyroiditis and retroperitoneal
fibrosis in multifocal fibrosclerosis: positron emission tomographic findings. Clin Nucl Med
[48] Arrive L, Hricak H, Tavares NJ, et al. Malignant versus nonmalignant retroperitoneal fibro-
sis: differentiation with MR imaging. Radiology 1989;172(1):139–43.
[49] Pickhardt PJ, Bhalla S. Unusual nonneoplastic peritoneal and subperitoneal conditions: CT
findings. Radiographics 2005;25(3):719–30.
[50] Baker LR, Mallinson WJ, Gregory MC, et al. Idiopathic retroperitoneal fibrosis. A retro-
spective analysis of 60 cases. Br J Urol 1987;60(6):497–503.
[51] Marcolongo R, Tavolini IM, Laveder F, et al. Immunosuppressive therapy for idiopathic
retroperitoneal fibrosis: a retrospective analysis of 26 cases. Am J Med 2004;116(3):194–7.
[52] Jois RN, Kerrigan N, Scott DG. Mycophenolate mofetil for maintenance of remission in
idiopathic retroperitoneal fibrosis. Rheumatology (Oxford) 2007;46(4):717–8.
[53] Scheel PJ Jr, Piccini J, Rahman MH, et al. Combined prednisone and mycophenolate mofetil
treatment for retroperitoneal fibrosis. J Urol 2007;178(1):140–3 [discussion: 143–4].
[54] van Bommel EF, Hendriksz TR, Huiskes AW, et al. Brief communication: tamoxifen
therapy for nonmalignant retroperitoneal fibrosis. Ann Intern Med 2006;144(2):101–6.
Rheum Dis Clin N Am 33 (2007) 819–834

Five Clinical Conundrums

in the Management of Giant
Cell Arteritis
Maria C. Cid, MDa,*, Ana Garcı́a-Martı́nez, MDb,
Ester Lozano, PhDa, Georgina Espı́gol-Frigolé, MDa,
José Hernández-Rodrı́guez, MDa
Vasculitis Research Unit, Department of Internal Medicine, Hospital Clı´nic,
University of Barcelona, Institut d’Investigacions Biome`diques August Pi i Sunyer
(IDIBAPS), Villarroel 170, 08036 Barcelona, Spain
Department of Emergency Medicine, Hospital Clı´nic, Villarroel 170,
08036 Barcelona, Spain

Patients with giant cell arteritis (GCA) usually experience a dramatic im-
provement with glucocorticoid treatment. The symptomatic relief that most
patients experience is so remarkable that rapid response to treatment is one
of the features required to diagnose GCA in patients for whom histopatho-
logic confirmation is lacking [1–3]. Moreover, although the complications of
GCA are occasionally fatal [4–7], most studies indicate that the disease does
not diminish the life spans of most patients [8,9]. The treatment of GCA is
usually viewed as extremely gratifying.
GCA may deeply impair the patient’s quality of life [10], and the usually
dramatic initial recovery is often followed by substantial difficulties that
arise during follow-up as a result of both the disease and its therapy. These
difficulties challenge the concept that GCA is an easy disease and that re-
sponse to glucocorticoid treatment is great. Visual impairment still occurs
in 15% to 20% of patients, and this frequency has remained stable for
the last decades [11–14]. Reversibility or significant improvement with glu-
cocorticoids, including intravenous megadoses, remains anecdotal [15–19].

Work for this article was supported by the Ministerio de Educación y Ciencia and Fondo
Europeo de Desarrollo Regional (FEDER) (SAF 05/06250) and Marató TV3 (06/0710).
Dr. Cid has received consulting/lecturing fees from Centocor.
* Corresponding author.
E-mail address: mccid@clinic.ub.es (M.C. Cid).

0889-857X/07/$ - see front matter Ó 2007 Elsevier Inc. All rights reserved.
doi:10.1016/j.rdc.2007.08.001 rheumatic.theclinics.com
820 CID et al

Approximately 10% to 15% of these patients may experience further visual

deterioration during the first 1 to 2 weeks after beginning glucocorticoid
therapy [16–20]. In addition, despite glucocorticoid treatment, significant
aortic structural damage leading to aneurysm, dissection, or aortic valve in-
sufficiency develops in approximately 18% to 22% of patients with GCA
[21,22], and about 5% to 15% of patients have symptomatic stenoses of
large vessels [22,23].
In 10% to 15% of patients, prednisone or an equivalent cannot be re-
duced below 10 to 15 mg/d without relapse or smoldering activity, and
the addition of immunosuppressive drugs need to be considered [24].
Approximately 40% to 50% of patients who can achieve a tapering of pred-
nisone to physiologic levels cannot tolerate complete withdrawal after the
initially anticipated treatment period of 2 to 3 years [24,25]. Although glu-
cocorticoids induce an important symptomatic improvement, they appear
to be insufficient to completely suppress essential pathways involved in dis-
ease persistence. Consequently, glucocorticoid-related side effects are fre-
quent during the follow-up of GCA patients.
This review addresses unresolved issues in the management of GCA.
These issues include the inability to reverse and prevent disease-derived
damage (ischemic events, aortic structural damage, and large vessel steno-
ses) and the failure to go beyond symptomatic relief and efficiently abrogate
disease activity.

Conundrum 1: visual loss

Visual loss is the most frequent ischemic event in the active phase of
GCA. Anterior ischemic optic neuropathy accounts for visual loss in 80%
to 90% of cases, and occlusion of central retinal artery branches or ciliore-
tinal arteries is found in about 10% [15,17]. Corticosteroid treatment is
highly effective in preventing visual loss. Although there are no clinical trials
supporting this point, visual impairment infrequently occurs after glucocor-
ticoid treatment has been initiated, and the frequency of GCA-related ische-
mic events has decreased from 60% in the pre-steroid era [26,27] to 15% to
20% in recent series [11–14]. Moreover, glucocorticoids preclude the pro-
gression and irreversibility of visual loss in patients presenting with partial
defects or transient events, both well-known risk factors for irreversible
complications [15,17,18]. Increased physician awareness and better and eas-
ier access of patients to medical care have probably contributed to early di-
agnosis and treatment and, consequently, to a decrease in the frequency of
ischemic complications.
Early diagnosis and treatment are the only effective ways to decrease ir-
reversible vascular occlusive events in patients with GCA; however, ischemic
complications cannot always be attributed to a delayed diagnosis. In 26% of
patients, visual loss occurs shortly after the onset of symptoms, and in a sig-
nificant subset of patients, this complication is the main or even the only

cranial manifestation of GCA [28,29]. Ischemic events tend to cluster in cer-

tain individuals, suggesting that some GCA patients are particularly predis-
posed to such complications [11]. In addition, several studies indicate that
visual loss is more frequent among patients who do not have a striking sys-
temic inflammatory response, making early diagnosis more difficult [11–14].

Pulse glucocorticoids
Although the level of evidence is low, pulse intravenous methylpredniso-
lone is reasonable when visual loss occurs; however, the main purpose of
a glucocorticoid pulse in this setting is to prevent further impairment in
patients with partial defects rather than to improve existing deficits. Most
improvement in visual acuity following pulse glucocorticoids results from
the patient’s learned ability to use their remaining vision more efficiently.
When measured accurately, improvement of visual field defects occurs in
only about 4% of GCA patients [15–17].
Once glucocorticoid treatment is established, the risk of further sight de-
terioration dramatically decreases. Additional visual loss during the first 1
to 2 weeks after the beginning of treatment has been reported in 10% to
15% of patients who presented with visual symptoms but is unlikely to
occur among patients who did not have visual manifestations at presenta-
tion and is very infrequent in any patient beyond the first 2 weeks of treat-
ment [15–19]. Under the current standard-of-care therapy, only 1% of
patients with GCA experienced visual loss during a follow-up time of 5 years

Two recent retrospective studies have shown that previous treatment with
antiplatelet therapy (primarily aspirin) is associated with a significantly
lower frequency of ischemic events in patients in whom GCA develops while
on antiplatelet treatment [30,31]. In addition, the risk of further visual dete-
rioration after the start of glucocorticoid treatment was reduced among pa-
tients on aspirin, although the total number of patients with visual loss after
glucocorticoid treatment was too small to draw definitive conclusions [30].
The extrapolation of these results, which are strictly applicable to only
those patients on aspirin at diagnosis, to all patients with GCA may translate
into a reduction of the number of patients who lose vision after the beginning
of glucocorticoid treatment. It seems prudent to recommend low-dose aspirin
therapy in patients without contraindications to this intervention.

Conundrum 2: delayed large vessel complications

In 1995, Evans and colleagues [32] reported an increased frequency of
aortic aneurysm, particularly in the thoracic aorta, among patients with
GCA. Thoracic and abdominal aneurysms were found to be 17.3 and 2.4
822 CID et al

times more frequent among patients with GCA than among the general pop-
ulation. A delayed development of aortic aneurysms has been confirmed in
additional retrospective studies by other investigators. Reported frequencies
range from 9.5% during a median follow-up period of 38.5 months to 18%
during a median-follow-up of 7.6 years [21,32,33].
With the increase in life expectancy in developed countries, extended fol-
low-up of these patients, and the more widespread use of cross-sectional im-
aging techniques such as MRI with angiography, it is likely that the
incidence of this complication will grow. In a recent study that employed
a systematic simple screening protocol (abdominal ultrasonography and
a chest radiograph followed by CT when appropriate), the authors detected
either an aortic aneurysm or significant aortic dilatation in 12 of 54 patients
(22.2%) whose median follow-up was only 5.4 years [34]. As observed in
other series, abnormalities were much more frequent in the thoracic than
in the abdominal aorta (11 patients versus 1 patient, respectively). Five
patients in whom the abnormalities were discovered were candidates for sur-
gical repair because of the aneurysm’s size or ensuing aortic valve
Based on the findings of aortic inflammatory infiltrates in surgical or nec-
ropsy specimens of GCA patients, it has been postulated that aneurysms de-
velop as a consequence of persisting subclinical activity [21,22,32]; however,
based on necropsy and imaging studies, although the thoracic and the ab-
dominal portions of the aorta are equally affected during active disease,
the development of significant dilatation is much more frequent in the tho-
racic aorta (Fig. 1) [35,36]. This observation suggests that, in addition to

Fig. 1. Spiral CT scan of the chest in a patient with GCA, revealing dilatation of the ascending
aorta at the time of diagnosis.

persistent inflammatory activity, factors related to the nature of the target

tissue, perhaps structural differences or the high pressure acting on the tho-
racic aorta, may have a role in the dilatation of a weakened wall [37]. In-
deed, Lie [38] observed that intense granulomatous lesions were found
mainly in surgical or necropsy specimens obtained early in the course of
the disease, whereas the intensity of inflammatory infiltrates was compara-
tively less among patients who had been treated with glucocorticoids
when aortic complications occurred (Fig. 2).
In the authors’ study, aneurysms were not more frequent among patients
with either a relapsing course or subclinical activity as suggested by persistent
moderate elevation of acute phase proteins [34]. Although these results need to
be interpreted with caution and confirmed in larger series, they suggest that
aneurysm formation is more likely to be a delayed consequence of the initial
structural damage than a result of persistent subclinical inflammatory activ-
ity. By the time treatment begins, the damage may already have been done
(see Fig. 1).
Glucocorticoid treatment is guided by clinical and laboratory response. It
is not clear whether the discovery of aortic aneurysms in patients apparently
in remission should imply any change in their therapeutic management.
Whether the development of aortic structural abnormalities can be reduced
or prevented by changes in the current treatment policies in terms of the
dose and duration of glucocorticoid therapy is presently unknown.
According to necropsy studies, inflammatory involvement of supra-aortic
branches is very common in GCA [35]. More recently, positron emission to-
mography imaging has demonstrated increased 18F- fluorodeoxyglucose
(FDG) uptake by supra-aortic branches in 74% of patients at diagnosis
[36]. Subclavian, axillary, brachial, and cervical arteries are the ones most
commonly involved (Fig. 3). FDG uptake in femoral arteries is somewhat

Fig. 2. Scattered residual inflammatory infiltrates in the medial layer of the thoracic aorta in
a surgically removed aneurysm from a patient with GCA, 6.5 years after diagnosis (hematox-
ylin-eosin staining).
824 CID et al

Fig. 3. CT angiogram showing occlusion of the left subclavian artery (arrow) in a patient with

less common but still occurs in approximately 37% of patients with active
Involvement of aortic branches remains asymptomatic unless inflamma-
tion-induced intimal hyperplasia leads to lumen occlusion. Symptoms in-
clude stroke, arm claudication, distal pallor on exercise, and blood
pressure measurements in the extremities that do not reflect true central aor-
tic pressure. Severe ischemic complications are infrequent in limbs because
compensatory collateral flow usually develops. Symptomatic lower limb
involvement is more difficult to interpret, particularly among smokers, be-
cause atherosclerotic arteriopathy is so common in advanced age. Retro-
spective studies have shown that symptomatic stenosis occurs in about
5% to 15% of patients [21–23]; however, the true prevalence of large vessel
stenosis is probably higher, because patients have not been screened rou-
tinely for minor features such as arterial bruits or the absence or asymmetry
of pulses.
Glucocorticoid treatment usually improves symptoms, but normal blood
flow through tightly stenotic vessels is generally not restored. As a result,
some patients are referred for angioplasty or bypass surgery. The reported
experience with treating large vessel stenoses in GCA by angioplasty is lim-
ited. A recent study showed that balloon angioplasty of stenotic lesions of
the upper extremities in 10 GCA patients was successful in 65% of the le-
sions. Restenosis was frequent, but repetition up to three times resulted in
an overall successful patency rate of almost 90% [39].

Intimal hyperplasia leading to vessel occlusion develops as a consequence

of myointimal cell proliferation, migration of cells to the lumen, and exces-
sive matrix production. Among several growth factors known to be pro-
duced in GCA, platelet-derived growth factor (PDGF) is the most potent
inducer of myointimal cell proliferation and migration [40]. Imatinib mesy-
late (Gleevec), a potent inhibitor of the tyrosine kinase activity of the PDGF
receptor, has been shown to inhibit myointimal cell outgrowth from ex-
planted temporal arteries. Whether imatinib would be able to prevent or re-
verse vascular occlusion in GCA is an interesting possibility, albeit, untested

Conundrum 3: do all patients with giant cell arteritis require

the same treatment approach?
Patients with GCA are usually given 40 to 60 mg of prednisone or an
equivalent a day during the first weeks after diagnosis [41–44]. There ap-
pears to be no difference between oral or intravenous glucocorticoids in
terms of efficacy or the likelihood of preventing subsequent visual loss
[18]. Initial doses higher and lower than 40 to 60 mg/d are also employed
to induce remission in some patients [42–44]. Tapering schedules applied af-
ter the first months are heterogeneous but usually aimed to achieve a main-
tenance dose of 5 to 7.5 mg of prednisone per day 8 to 12 months after
diagnosis [42,43]. Several studies have shown that alternate day administra-
tion of glucocorticoids is less efficient in maintaining remission in GCA and
may be harmful because it fails to control the disease as effectively as daily
treatment [45,46]. An unusually high proportion of patients suffering new-
onset visual impairment during follow-up (13.8%) was observed in a clinical
trial in which an alternate day schedule was used [2]. Delayed visual loss oc-
curs in about 1% of patients treated with a regimen of daily glucocorticoids
[19]. In the trial that featured alternate day prednisone [2], the percentage of
patients who relapsed (83%) was also higher than in other trials (64%–
75.5%) [3,47].
During the maintenance phase of therapy, a spectrum of patient re-
sponses exists. Some patients need higher maintenance doses than 5.0 to
7.5 mg of prednisone per day. This observation has motivated the search
for glucocorticoid-sparing agents [24]. On the other hand, some patients en-
ter prolonged remissions after only short glucocorticoid courses. In fact, be-
fore the introduction of glucocorticoids in the 1950s, GCA was considered
to be a self-limited disease even though as many as 60% of patients lost vi-
sion without therapy [27,48,49]. The validity of GCA as a self-limited dis-
ease in at least a subset of patients is impossible to test today because of
the major potential consequences of untreated disease in the short term
and the availability of effective therapy.
Some contemporary evidence suggests that cases of self-limited (or at
least subclinical) GCA exist. In a survey of 100 unselected necropsies, Lie
826 CID et al

and colleagues [49] found unequivocal evidence of obsolescent inflammatory

lesions in the temporal arteries of two patients who were asymptomatic dur-
ing life. The authors have recently reported two patients with spontaneous
disease remission, closely followed for 4 and 10 years, respectively, who
never received glucocorticoid treatment and have not developed any compli-
cation attributable to GCA to date [50].
Recent clinical trials testing potential glucocorticoid-sparing agents have
shown that approximately 16% to 25% of patients in the placebo arms do
not relapse after glucocorticoid-tapering schedules that are significantly
shorter than the standard of care [3,47]; therefore, it seems that some pa-
tients enter sustained GCA remissions with only a few months of treatment
or even no treatment at all. Whether these patients will eventually sustain
delayed complications such as aortic aneurysm or dilatation or large vessel
stenosis is uncertain. These data suggest that the risks and benefits of long-
term daily glucocorticoid therapy must be weighed carefully for each pa-
tient, and that clinicians should question heavy-handed treatment regimens.

Conundrum 4: limiting glucocorticoid exposure

Although the initial response to prednisone is highly rewarding, a sub-
stantial number of patients have side effects during follow-up. Nesher and
colleagues [51] found that 58% of a series of 43 patients experienced major
glucocorticoid-related complications (infection or fracture) during a median
of 3 years. A more recent study by Proven and colleagues [43] performed on
120 patients disclosed that 86% of patients sustained side effects, with 58%
experiencing more than one adverse effect during a median follow-up of 10
years [43].
In the study by Proven and colleagues [43], the glucocorticoid doses were
lower than those employed by other investigators, particularly ophthalmol-
ogy groups [15–18]. Glucocorticoids were reduced to a physiologic dose at
a median of 6.5 months and completely discontinued at a median of less
than 2 years [43]. Nevertheless, the frequency of adverse events was substan-
tial. The time to first event was relatively long, indicating that the number of
side effects may increase over time and needs to be assessed on a long-term
basis [43]. Glucocorticoid-related adverse effects are more frequent in patients
with higher cumulative doses in most, but not all, studies, indicating that in-
dividual susceptibility may have a role (Fig. 4) [42–44]. Patients with low bone
density at diagnosis or patients with underlying metabolic syndrome are prob-
ably at greater risk for an exacerbation of these problems on glucocorticoid
treatment [52,53].
The high incidence of glucocorticoid complications along with the fact that
40% to 50% of patients are unable to taper prednisone completely after 2 to 3
years has motivated an active search for glucocorticoid-sparing agents. Three
options have been investigated in randomized clinical trials: methotrexate,

Fig. 4. MRI study disclosing multiple vertebral fractures in a glucocorticoid-resistant patient

with GCA.

pulse intravenous methylprednisolone at the start of therapy, and targeted tu-

mor necrosis factor (TNF) alpha inhibition with infliximab [2,3,47,54–56]. All
of these treatments have been tried in newly diagnosed patients. No published
trials have been conducted on patients with GCA that is refractory or relaps-
ing-remitting, a large subset of patients that comprises an important unmet

The efficacy of methotrexate (7.5 to 20 mg weekly) to maintain glucocor-
ticoid-induced remission and to spare glucocorticoids has been tested in
three randomized, double-blind, placebo-controlled trials [2,47,54]. The re-
sults of these trials were conflicting. Nevertheless, a recent meta-analysis
of individual patient data from the three trials indicated that methotrexate
treatment reduced the likelihood of disease relapse when compared with pla-
cebo, but the magnitude of the effect was modest [55]. The addition of meth-
otrexate was estimated to save each patient treated a total of approximately
800 mg of prednisone; however, no significant differences were observed in
glucocorticoid side effects, raising doubts about the clinical relevance of the
amount of prednisone spared. As mentioned previously, it may not be rea-
sonable to expect a reduction in glucocorticoid-related adverse events within
the relatively short time frame of most clinical trials. The upshot of these
three randomized trials is that methotrexate may be of some help in
828 CID et al

relapsing patients or in the management of patients with severe glucocorti-

coid-related side effects.

Pulse methylprednisolone
A recent randomized double-blind, placebo-controlled trial demonstrated
that treatment with intravenous methylprednisolone (1 g/d for 3 consecutive
days) followed by prednisone resulted in a significantly higher proportion of
patients who achieved a daily dose of prednisone less than or equal to 5 mg/d
by week 36, a significantly lower median daily prednisone dose at week 78, and
a higher percentage of patients in sustained remission after the discon-
tinuation of prednisone [56]. Nevertheless, the size of this cohort was small
(27 patients), and the definition of flare was loose, including isolated
increases in C-reactive protein or erythrocyte sedimentation rate [56]. More-
over, the cumulative dose of prednisone was lower in the methylpredniso-
lone arm only if the initial 3-g ‘‘loading dose’’ was excluded. The ultimate
clinical utility of using methylprednisolone pulses at the outset of therapy
in GCA requires further study.

A recent international multicenter randomized, double-blind, placebo-
controlled trial including 44 patients assessed the efficacy of blocking TNF
with infliximab (5 mg/kg) [3]. Unfortunately, this trial failed to demonstrate
any benefit of infliximab over placebo in the maintenance of disease remission,
and the trial was terminated following an interim analysis at week 22. The
results of this trial were disappointing given the demonstrated efficacy of
infliximab in several chronic inflammatory and granulomatous disorders.
Similar findings were achieved by a similarly designed trial testing lower
doses (3 mg/kg) in patients with new-onset polymyalgia rheumatica [57].

Azathioprine (100–200 mg/d) was tested in a randomized, placebo-con-
trolled clinical trial in 31 patients with glucocorticoid complications [58]. Al-
though this was, indeed, a pioneering study, it included a mixed population
of GCA and polymyalgia rheumatica patients, making reliable conclusions
difficult. The validity of the modest prednisolone-sparing effect demon-
strated in this trial (a mean of 4.2 mg/d in the azathioprine arm versus
1.9 mg/d in the placebo arm at week 52) requires confirmation.

Based on the anti-inflammatory effects of statins and their benefit on pri-
mary and secondary prevention of vascular occlusive events in atherosclerosis,

two retrospective studies have assessed the potential glucocorticoid-sparing

effect of these medications [59,60]. Both studies employed statin doses within
the low-moderate range, with negative results. Besides the retrospective char-
acter of these studies, additional limitations include the concomitant use of
glucocorticoids according to the standard of care, which may have masked
a slight effect. Whether statins at higher doses with a more aggressive glucocor-
ticoid tapering schedule would allow an earlier reduction or discontinuation of
glucocorticoids is not known.

Conundrum 5: what do glucocorticoids not effect (what therapeutic

targets are left?)
Glucocorticoids remain the cornerstone of GCA treatment. The quick
and exquisite remission that high glucocorticoid doses exert on GCA symp-
toms suggests that this intervention efficiently suppresses the immuno-
regulatory pathways that orchestrate this disease. As noted, short
glucocorticoid courses and low doses are not sufficient to suppress essential
pathways, which are eventually able to re-activate the inflammatory process
fully in the majority of cases. In addition, increased serum concentrations of
inflammatory markers such as TNF, interleukin (IL)-6, C-reactive protein,
and the von Willebrand’s factor antigen can be detected even in patients
in apparent clinical remission, suggesting subclinical inflammatory activity
Which pathways are crucial to maintain disease activity? Does apparent
clinical benefit always imply disease suppression? Which pathways do gluco-
corticoids suppress and which ones remain active? These important ques-
tions still await satisfactory answers.
Current understanding of the immunopathogenic mechanisms involved
in the generation of the inflammatory lesions in GCA is based on low levels
of experimental evidence. We assume that certain cytokines, receptors, pro-
teases, or growth factors are important in triggering or perpetuating inflam-
matory cascades or in promoting vessel injury or remodeling in GCA.
Nevertheless, our convictions about these mediators’ roles are based primar-
ily on comparisons between inflamed and normal arteries [65–68], on extrap-
olation of their known biologic activities [65–69], and on their association
with certain phenotypes or outcomes [68–71].
Cross-sectional comparisons of several inflammatory molecules in treat-
ment-naı̈ve patients versus those who receive glucocorticoid treatment at
the time of temporal artery biopsy suggest down-regulation of several prod-
ucts by glucocorticoid treatment, including the IL-2 receptor [72], endothe-
lial adhesion molecules [67], and gelatinases [66]. Analysis of temporal artery
biopsies engrafted onto SCID mice reveals down-regulation of IL-6 and
IL-1 mRNAs after 3 weeks of treatment with high-dose dexamethasone,
whereas the reduction in other transcripts is less remarkable (ie, interferon
gamma) or nonexistent (ie, transforming growth factor beta) [73,74]. The
830 CID et al

effects of treatment at the protein level have not been investigated in this
model. Unfortunately, none of these interesting observations provide defin-
itive evidence that any specific mediator is the ‘‘main player’’ in the patho-
genesis of the disease.
Although disappointing from the clinical point of view, the failure of the
recent multicenter trial assessing the efficacy of TNF blockade with inflixi-
mab was a useful learning experience [3]. TNF is highly expressed in GCA
(Fig. 5) [70] and has potent pro-inflammatory activities that are crucial in
various animal models of human inflammatory disease [75]. Its expression
correlates with the intensity of the systemic inflammatory response and glu-
cocorticoid requirements in patients with GCA [64,70]. Moreover, blocking
TNF induces remission in several chronic inflammatory and granulomatous
diseases in humans [75]. Despite of all these pieces of evidence, blocking
TNF was insufficient to suppress disease activity in GCA [3].
Targeting activated dendritic cells with the monoclonal antibody anti-
CD86 is the only experimental intervention that has been shown to reduce
T-cell infiltrates in temporal arteries from patients with GCA engrafted
into NOD-SCID mice [76]. This finding suggests that interfering with early
events in antigen presentation/T-cell activation might abrogate vascular in-
flammatory lesions in GCA. Although intriguing, these data need to be in-
terpreted with caution given the small number of samples analyzed and the
segmental nature of inflammatory lesions in GCA.
Currently, no therapeutic intervention has been found to achieve the clin-
ical benefit elicited by glucocorticoids in GCA [77,78]. Optimization of glu-
cocorticoid therapy might also be a useful option. New glucocorticoid
therapies with improved risk-benefit ratios are in the horizon. These thera-
pies include nitrosoglucocorticoids, selective glucocorticoid receptor ago-
nists, and long circulating liposomal glucocorticoids [79,80].

Fig. 5. Intense TNF-a expression in active GCA lesions. Unfortunately, blocking TNF is not
sufficient to abrogate disease activity in newly diagnosed patients. Immunostaining was per-
formed with a polyclonal rabbit anti-human TNF (R & D Systems, Minneapolis, Minnesota).

Despite inducing a dramatic symptomatic relief, glucocorticoids are un-
able to completely suppress disease activity in GCA. The essential pathways
involved in maintaining disease activity have not been identified precisely.
We do not know which pathways are efficiently suppressed by glucocorti-
coids and which ones remain active with the potential for restoring the in-
flammatory process.
With currently available models, one can only assess the effects of inter-
vention on particular mediators or endpoints that are thought to be impor-
tant. The lack of true animal models hampers our ability to determine the
effects of manipulating specific pathways on the outcome of the disease. Co-
ordinated international efforts are needed to efficiently address the current
challenges in the management of GCA by formulating the appropriate ques-
tions and by testing them in properly designed and conducted clinical trials.

[1] Ellis ME, Ralston S. The ESR in the diagnosis and management of the polymyalgia rheuma-
tica/giant cell arteritis syndrome. Ann Rheum Dis 1983;42:168–70.
[2] Hoffman GS, Cid MC, Hellmann DB, et al. A multicenter, randomized, double-blind, pla-
cebo-controlled trial of adjuvant methotrexate treatment for giant cell arteritis. Arthritis
Rheum 2002;46:1309–18.
[3] Hoffman GS, Cid MC, Rendt-Zagar KE, et al. Infliximab for maintenance of glucocorticos-
teroid-induced remission of giant cell arteritis. Ann Intern Med 2007;146:621–30.
[4] Save-Soderbergh J, Malmvall BE, Andersson R, et al. Giant-cell arteritis as a cause of death.
JAMA 1986;255:493–6.
[5] Nuenninghoff DM, Hunder GG, Christianson TJH, et al. Mortality of large-artery compli-
cation (aortic aneurysm, aortic dissection, and/or large-artery stenosis) in patients with giant
cell arteritis: a population-based study over 50 years. Arthritis Rheum 2003;48:3532–7.
[6] Uddhamar D, Eriksson AL, Nystrom L, et al. Increased mortality due to cardiovascular dis-
ease in patients with giant cell arteritis in Northern Sweden. J Rheumatol 2002;29:737–42.
[7] Nordborg E, Bengtsson BA. Death rates and causes of death in 284 consecutive patients with
giant cell arteritis confirmed by biopsy. Br Med J 1989;299:549–50.
[8] Matteson EL, Gold KN, Bloch DA, et al. Long-term survival of patients with giant cell
arteritis in the American College of Rheumatology giant cell arteritis classification criteria
cohort. Am J Med 1996;100:193–6.
[9] Salvarani C, Crowson CS, O’Fallon WM, et al. Reappraisal of the epidemiology of giant cell
arteritis in Olmsted County, Minnesota over a fifty year period. Arthritis Rheum 2004;51:
[10] Hellmann D, Uhlfelder M, Stone JH, et al. Domains of quality of life important to patients
with giant cell arteritis. Arthritis Rheum 2003;49:819–25.
[11] Cid MC, Font C, Oristrell J, et al. Association between strong inflammatory response and
low risk of developing visual loss and other cranial ischemic complications in giant cell (tem-
poral) arteritis. Arthritis Rheum 1998;41:26–32.
[12] González-Gay MA, Garcı́a-Porrua C, Llorca J, et al. Visual manifestations of giant cell ar-
teritis: trends and clinical spectrum in 161 patients. Medicine (Baltimore) 2000;79:283–92.
[13] Nesher G, Berkun Y, Mates M, et al. Risk factors for cranial ischemic complications in giant-
cell arteritis. Medicine (Baltimore) 2004;83:114–22.
[14] Salvarani C, Cimino L, Machioni P, et al. Risk factors for visual loss in an Italian popula-
tion-based cohort of patients with GCA. Arthritis Rheum 2005;53:293–7.
832 CID et al

[15] Foroozan R, Deramo VA, Buono LM, et al. Recovery of visual function in patients with
biopsy-proven giant cell arteritis. Ophthalmology 2003;110:539–42.
[16] Danesh-Meyer H, Savino PJ, Gamble GC. Poor visual outcome after visual loss from giant
cell arteritis. Ophthalmology 2005;112:1098–103.
[17] Hayreh SS, Zimmerman B, Kardon RH, et al. Visual improvement with glucocorticoid ther-
apy in giant-cell arteritis: report of a large study and review of literature. Acta Ophthalmol
Scand 2002;80:353–67.
[18] Hayreh SS, Zimmerman B. Visual deterioration in giant cell arteritis patients while on high
doses of glucocorticoid therapy. Ophthalmology 2003;110:1204–15.
[19] Aiello PD, Trautmann JC, McPhee TJ, et al. Visual prognosis in giant cell arteritis. Ophthal-
mology 1993;100:550–5.
[20] Liu GT, Glaser JS, Schatz NJ, et al. Visual morbidity in giant cell arteritis: clinical charac-
teristics and prognosis for vision. Ophthalmology 1994;101:1779–85.
[21] Nuenninghoff DM, Hunder GG, Christianson TJH, et al. Incidence and predictors of large-
artery complication (aortic aneurysm, aortic dissection, and/or large-artery stenosis) in
patients with giant cell arteritis: a population-based study over 50 years. Arthritis Rheum
[22] Bongartz T, Matteson EL. Large-vessel involvement in giant cell arteritis. Curr Opin Rheu-
matol 2006;18:10–7.
[23] Klein RG, Hunder GG, Stanson AW, et al. Large artery involvement in giant-cell (temporal)
arteritis. Ann Intern Med 1975;83:806–12.
[24] Wilke WS, Hoffman GS. Treatment of glucocorticoid-resistant giant cell arteritis. Rheum
Clin Dis North Am 1995;21:59–71.
[25] Andersson R, Malmvall BE, Bengtsson BA. Long-term glucocorticoid treatment in giant-
cell arteritis. Acta Med Scand 1986;220:465–9.
[26] Hollenhorst RW, Brown JR, Wagener HP, et al. Neurologic aspects of temporal arteritis.
Neurology 1960;10:490–8.
[27] Hunder GG. The early history of giant cell arteritis and polymyalgia rheuamatica: first de-
scriptions to 1970. Mayo Clin Proc 2006;81(8):1071–83.
[28] Font C, Cid MC, Coll-Vinent B, et al. Clinical features in patients with permanent visual loss
due to biopsy-proven giant cell arteritis. Br J Rheumatol 1997;36:251–4.
[29] Hernández-Rodrı́guez J, Font C, Garcı́a-Martı́nez A, et al. Development of ischemic
complications in patients with giant cell arteritis presenting with apparently isolated
polymyalgia rheumatica: study of a series of 100 patients. Medicine (Baltimore) 2007;
[30] Nesher G, Berkun Y, Mates M, et al. Low-dose aspirin and prevention of cranial ischemic
complications in giant cell arteritis. Arthritis Rheum 2004;50:1332–7.
[31] Lee MS, Smith SD, Galor A, et al. Antiplatelet and anticoagulant therapy in patients with
giant cell arteritis. Arthritis Rheum 2006;54:3306–9.
[32] Evans JM, O’Fallon WM, Hunder GG. Increased incidence of aortic aneurysm and dissec-
tion in giant cell (temporal) arteritis. Ann Intern Med 1995;122:502–7.
[33] González-Gay MA, Garcı́a-Porrua C, Piñeiro A, et al. Aortic aneurysm and dissection in pa-
tients with biopsy-proven giant cell arteritis from northwestern Spain. Medicine (Baltimore)
[34] Garcı́a-Martı́nez A, Hernández-Rodrı́guez J, Arguis P. Development of aortic aneurysm/
dilatation during the follow-up of patients with giant cell arteritis: a cross-sectional screen-
ing of 54 prospectively followed patients. Arthritis Rheum 2008, in press.
[35] Östberg G. Morphological changes in the large arteries in polymyalgia arteritica. Acta Med
Scand 1972;533(Suppl):133–59.
[36] Blockmans D, de Ceuninck L, Vanderschueren S, et al. Repetitive 18F-fluorodeoxyglucose
positron emission tomography in giant cell arteritis: a prospective study of 35 patients.
Arthritis Rheum 2006;55:131–7.
[37] Hoffman GS. Large-vessel vasculitis: unresolved issues. Arthritis Rheum 2003;48:2406–14.

[38] Lie JT. Aortic and extracranial large vessel giant cell arteritis: a review of 72 cases with his-
topathologic documentation. Semin Arthritis Rheum 1995;24:422–31.
[39] Both M, Aries PM, Muller-Hulsbeck S, et al. Balloon angioplasty of arteries of the upper
extremities in patients with extracranial giant cell arteritis. Ann Rheum Dis 2006;65:
[40] Lozano E, Segarra M, Garcı́a-Martı́nez A, et al. Imatinib mesylate inhibits in vitro and ex
vivo biologic responses related to vascular occlusion in giant cell arteritis. Ann Rheum
Dis 2007 [Epub ahead of print].
[41] Salvarani C, Cantini F, Boiardi L, et al. Polymyalgia rheumatica and giant cell arteritis.
N Engl J Med 2002;347:261–71.
[42] Nordborg E, Nordborg C. Giant-cell arteritis: epidemiological clues to its pathogenesis and
update on its treatment. Rheumatology 2003;42:413–21.
[43] Proven A, Gabriel S, Orces C, et al. Glucocorticoid therapy in giant cell arteritis: duration
and adverse outcomes. Arthritis Rheum 2003;49:703–8.
[44] Delecoeuillerie G, Joly P, Cohen de Lara A, et al. Polymyalgia rheumatica and temporal ar-
teritis: a retrospective analysis of prognostic features and different glucocorticoid regimens
(11 year survey of 210 patients). Ann Rheum Dis 1988;47:733–9.
[45] Hunder GG, Sheps SG, Allen GL, et al. Daily and alternate-day glucocorticoid regimen in
maintenance therapy of giant cell arteritis: comparison in a prospective study. Ann Intern
Med 1975;82:613–8.
[46] Bengtsson BA, Malmvall BE. An alternate-day glucocorticoid regimen in maintenance ther-
apy of giant cell arteritis. Acta Med Scand 1981;209:347–50.
[47] Jover JA, Hernández-Garcı́a C, Morado IC, et al. Combined treatment of giant cell arteritis
with methotrexate and prednisone. Ann Intern Med 2001;134:106–14.
[48] Horton BT, Magath TB, Brown GE. An undescribed form of arteritis of the temporal ves-
sels. Mayo Clin Proc 1932;7:700–1.
[49] Lie JT, Brown AL Jr, Carter ET. Spectrum of aging changes in temporal arteritis: its signif-
icance in interpretation of biopsy of temporal artery. Arch Pathol 1970;90:278–85.
[50] Hernández-Rodrı́guez J, Garcı́a-Martı́nez A, Espı́gol-Frigolé G, et al. Sustained spontane-
ous clinical remission in giant cell arteritis: report of two cases with long-term follow-up.
Arthritis Rheum 2006;55:160–2.
[51] Nesher G, Sonnenblick M, Friendlander Y. Analysis of steroid-related complications
and mortality in temporal arteritis: a 15-year survey of 43 patients. J Rheumatol 1994;
[52] Queralt C, Hernández-Rodrı́guez J, Garcı́a-Martı́nez A, et al. Baseline osteopenia is associ-
ated with subsequent bone loss and fracture in glucocorticoid-treated patients with giant cell
arteritis receiving calcium and vitamin D supplements. Arthritis Rheum 2000;93(Suppl):
[53] Uzu T, Harada T, Sakaguchi M, et al. Glucocorticoid-induced diabetes mellitus: prevalence
and risk factors in primary renal diseases. Nephron Clin Pract 2007;105:C54–7.
[54] Spiera RF, Mitnick HJ, Kupersmith M, et al. A prospective, double-blind, randomized, pla-
cebo controlled trial of methotrexate in the treatment of giant cell arteritis (GCA). Clin Exp
Rheumatol 2001;19:495–501.
[55] Mahr A, Jover JA, Spiera RF, et al. Adjunctive methotrexate for treatment of giant cell ar-
teritis: an individual patient data meta-analysis. Arthritis Rheum 2007;56:2789–97.
[56] Mazlumzadeh M, Hunder GG, Easley KA, et al. Treatment of giant cell arteritis using
induction therapy with high-dose glucocorticoids: a double-blind, placebo-controlled,
randomized, prospective clinical trial. Arthritis Rheum 2006;54:3310–8.
[57] Salvarani C, Macchioni P, Manzini C, et al. Infliximab plus prednisone or placebo plus pred-
nisone in the initial treatment of polymyalgia rheumatica: a randomized trial. Ann Intern
Med 2007;146:631–9.
[58] De Silva M, Hazleman BL. Azathioprine in giant cell arteritis/polymyalgia rheumatica:
a double-blind study. Ann Rheum Dis 1986;45:136–8.
834 CID et al

[59] Garcı́a-Martı́nez A, Hernández-Rodrı́guez J, Grau JM, et al. Treatment with statins does
not exhibit a clinically relevant glucocorticoid-sparing effect in patients with giant cell arter-
itis. Arthritis Rheum 2004;51:674–8.
[60] Narváez J, Bernad B, Nolla JM, et al. Statin therapy does not seem to benefit giant cell
arteritis. Semin Arthritis Rheum 2007;36:322–7.
[61] Cid MC, Monteagudo J, Oristrell J, et al. Von Willebrand factor in the outcome of temporal
arteritis. Ann Rheum Dis 1996;55:927–30.
[62] Weyand CM, Fulbright JW, Hunder GG, et al. Treatment of giant cell arteritis: interleukin-6
as a biologic marker of disease activity. Arthritis Rheum 2000;43:1041–8.
[63] Weyand CM, Hicok KC, Hunder GG, et al. Tissue cytokine pattern in patients with poly-
myalgia rheumatica and giant cell arteritis. Ann Intern Med 1994;121:484–91.
[64] Garcı́a-Martı́nez A, Hernández-Rodrı́guez J, Segarra M, et al. Clinical relevance of persis-
tently elevated circulating cytokines (TNFa and IL-6) in the long-term follow-up of patients
with giant cell arteritis (GCA). Arthritis Rheum 2006;54(Suppl):S763.
[65] Rodrı́guez-Pla A, Bosch-Gil JA, Rosselló-Urgell J, et al. Metalloproteinase-2 and -9 in giant
cell arteritis: involvement in vascular remodeling. Circulation 2005;112:264–9.
[66] Segarra M, Garcı́a-Martı́nez A, Sánchez M, et al. Gelatinase expression and proteolytic ac-
tivity in giant cell arteritis. Ann Rheum Dis 2007;66:1429–35.
[67] Cid MC, Cebrián M, Font C, et al. Cell adhesion molecules in the development of inflamma-
tory infiltrates in giant cell arteritis: inflammation-induced angiogenesis as the preferential
site of leukocyte-endothelial cell interactions. Arthritis Rheum 2000;43:184–94.
[68] Weyand CM, Tetzlaff N, Björnsson J, et al. Disease patterns and tissue cytokine profiles in
giant cell arteritis. Arthritis Rheum 1997;40:19–26.
[69] Hernández-Rodrı́guez J, Segarra M, Vilardell C, et al. Elevated production of IL-6 is asso-
ciated with a lower incidence of disease-related ischemic events in patients with giant cell ar-
teritis: angiogenic activity of IL-6 as a potential protective mechanism. Circulation 2003;107:
[70] Hernández-Rodrı́guez J, Segarra M, Vilardell C, et al. Tissue production of proinflamma-
tory cytokines (IL-1b, TNFa, and IL-6) correlates with the intensity of the systemic inflam-
matory response and with glucocorticoid requirements in giant cell arteritis. Rheumatology
[71] Cid MC, Hoffman MP, Hernández-Rodrı́guez J, et al. Genes related to the maintenance
of inflammatory lesions in giant cell arteritis: association between increased CCL2 (MCP-1)
expression and persistence of disease activity. Rheumatology 2006;45:1356–63.
[72] Cid MC, Campo E, Ercilla MG, et al. Immunohistochemical analysis of lymphoid and mac-
rophage cell subsets and their immunological activation markers in temporal arteritis: influ-
ence of glucocorticoid treatment. Arthritis Rheum 1989;32:884–93.
[73] Brack A, Rittner HL, Younge BR, et al. Glucocorticoid-mediated repression of cytokine
gene transcription in human arteritis-SCID chimeras. J Clin Invest 1997;99:2842–50.
[74] Weyand CM, Kaiser M, Yang H, Younge B, et al. Therapeutic effects of acetylsalicylic acid
in giant cell arteritis. Arthritis Rheum 2002;46:457–66.
[75] Feldman M, Maini RN. Anti-TNF alpha therapy of rheumatoid arthritis: have we learned?
Ann Rev Immunol 2001;19:163–96.
[76] Ma-Krupa W, Jeon MS, Spoerl S, et al. Activation of arterial wall dendritic cells and break-
down of self-tolerance in giant cell arteritis. J Exp Med 2004;199:173–83.
[77] Spiera R, Spiera H. Therapy for giant cell arteritis: can we do better? Arthritis Rheum 2006;
[78] Luqmani R. Treatment of polymyalgia rheumatica and giant cell arteritis: are we any further
forward? Ann Intern Med 2007;146:674–6.
[79] Stahn C, Löwenberg H, Hommes DW, et al. Molecular mechanisms of glucocorticoid action
and selective glucocorticoid receptor agonists. Mol Cell Endocrinol 2007 [Epub ahead of print].
[80] Buttgereit F, Burmester GR, Lipworth BJ. Optimised glucocorticoid therapy: the sharpen-
ing of an old spear. Lancet 2005;365:801–3.
Rheum Dis Clin N Am 33 (2007) 835–854

Scleritis and Peripheral Ulcerative

Anat Galor, MDa, Jennifer E. Thorne, MD, PhDa,b,*
Department of Ophthalmology, Wilmer Eye Institute, Johns Hopkins University,
School of Medicine, 550 North Broadway, Suite 700, Baltimore, MD 21205, USA
Department of Epidemiology, Center for Clinical Trials, Johns Hopkins University
Bloomberg School of Public Health, Baltimore, MD, USA

Scleritis and peripheral ulcerative keratitis (PUK) are two ocular disor-
ders that require urgent attention for the purpose of diagnosis, treatment,
and detection of underlying systemic inflammatory diseases.

Anatomy of the sclera
The sclera, which serves as the protective outer layer of the eye (Fig. 1), is
composed of connective tissuedcollagen, elastin, and proteoglycans. The
sclera starts at the limbus, where it is continuous with the cornea, and
ends at the optic canal, where it is continuous with the dura. The six extra-
ocular muscles insert into the sclera. Scleral tissue receives its sensory inner-
vation from the short posterior and long ciliary nerves [1]. The sclera, an
avascular structure, receives its nourishment from adjacent vascularized tis-
suesdthe choroidal plexus beneath and the episcleral plexus above. The
episclera has two vascular plexi, one superficial whose vessels are arranged
radially and another deep whose vessels adhere to the sclera.

Epidemiology of scleritis
Scleritis can occur in any age group but usually presents between the ages
of 30 and 50 years [2,3]. Women are affected approximately twice as often as

Work for this article was supported by grant EY-13707 (Dr. Thorne) from the National
Eye Institute, Bethesda, Maryland, and by unrestricted funds from Research to Prevent
Blindness (Dr. Galor).
* Corresponding author. Wilmer Eye Institute, 550 North Broadway, Suite 700, Baltimore,
MD 21205.
E-mail address: jthorne@jhmi.edu (J.E. Thorne).

0889-857X/07/$ - see front matter Ó 2007 Elsevier Inc. All rights reserved.
doi:10.1016/j.rdc.2007.08.002 rheumatic.theclinics.com

Episclera, sclera



Optic Lens

Dura Limbus


Choroid Sclera

Fig. 1. Anatomy of the eye.

men, and there is no racial or geographic predilection [2,3]. There is no

known HLA association [4]. The prevalence of scleritis in the general pop-
ulation is estimated to be 6 cases per 100,000 persons, but it has been de-
scribed in 0.2% to 6.3% of patients with rheumatoid arthritis and up to
7% of those with Wegener’s granulomatosis [5–7].

Although studies on the pathogenesis of scleritis are limited, the data
available support an important if not predominant role for T cells in the in-
flammatory process [8]. Inflammatory cells, mostly T cells and macro-
phages, are found in biopsy specimens [8,9]. Evidence of vasculitis with
neutrophil invasion and necrosis of the vessel wall was observed in one his-
topathologic study but not in a second one [8,10]. Antibody deposition has
also been described [8]. One group that divided scleritis into categories by
morphology found that zonal necrotizing granulomas were more common
in patients with associated systemic autoimmune conditions and that

non-zonal diffuse scleral inflammation was more common in patients with-

out an associated systemic condition [11].

Clinical history
The chief complaint of patients with scleritis is pain that responds poorly
to analgesics. The pain is described as dull, aching, or boring and may be
severe and constant. Because the extraocular muscles insert into the sclera,
the pain typically worsens with eye movement. The pain often awakens pa-
tients from sleep and is typically worse in the early morning. The pain can
spread to the periorbital region, brow, forehead, temple, ear, or jaw, and
may be disproportionate to the clinical findings.
As many as 20% of patients with scleritis have little or no pain [12]. The
absence of pain is observed in three settings: (1) milder disease (eg, diffuse
anterior or nodular scleritis as opposed to the necrotizing type); (2) patients
already on immunosuppressive medications at the time their symptoms be-
gin; and (3) scleromalacia perforans, a rare complication of advanced rheu-
matoid arthritis. Other complaints may include tearing, photophobia, and
decreased vision. Patients with anterior scleritis usually notice eye redness.
Patients with posterior scleritis may complain of ptosis and swelling of
the periorbital tissue.

Clinical findings
Patients with anterior scleritis present with a red eye. In some cases, the
redness is located underneath the lid and may be easily overlooked without
a focused eye examination. The redness has a violaceous hue that is best
seen in natural sunlight. Slit lamp examination reveals scleral edema and di-
latation of the superficial and deep episcleral vessels. The eye in patients
with scleritis is tender to palpation. Ocular erythema persists after the appli-
cation of topical phenylephrine. Approximately 25% of patients have bilat-
eral disease at presentation, but scleritis ultimately becomes bilateral in
about 50% of patients [3,12,13].

Classification of scleritis
The Watson and Hayreh classification of scleritis [12] divides the disorder
into anterior and posterior types based on the anatomic distribution of dis-
ease. Anterior scleritis is further subdivided into diffuse, nodular, necrotizing
with inflammation, and necrotizing without inflammation (scleromalacia per-
forans). Although these forms of scleritis correspond roughly to different de-
grees of severity, it is unusual for a case of scleritis to evolve from one type to
another, such as from diffuse anterior scleritis to necrotizing scleritis.

Diffuse anterior scleritis

Diffuse anterior disease, the most common clinical presentation of scler-
itis, occurs in approximately 60% of patients [3]. Although the onset of

disease is insidious, ocular inflammation may involve a substantial propor-

tion or all of the anterior sclera at presentation (Fig. 2). The normal radial
pattern of the episcleral vessels is lost, and one may see beading and tortu-
osity of the vessels. Signs of corneal inflammation such as peripheral corneal
infiltrates or mild corneal thinning may be present [13]. Frank corneal ulcer-
ation is not typical but is more common in necrotizing scleral disease. Pro-
gression to nodular or necrotizing scleritis has been observed but is
uncommon [3,14]. Resolution of the ocular inflammation in diffuse anterior
scleritis may leave a bluish gray hue caused by rearrangement of the scleral
collagen fibers (Fig. 3). This change does not represent scleral thinning,
which is an uncommon sequela of anterior scleritis. Systemic disorders
have been described in as many as 45% of patients with diffuse anterior
scleritis, most commonly rheumatoid arthritis [13,15].

Nodular anterior scleritis

Nodular scleritis is the second most common clinical presentation of an-
terior scleritis (Fig. 4), accounting for approximately 20% of cases [3]. Pa-
tients with nodular scleritis present with a firm, immobile, and tender
nodule that typically is found close to the limbus. Similar to diffuse anterior
scleritis, progression to other forms of scleritis is uncommon [3,14]. A sys-
temic disease is diagnosed in 40% to 50% of patients [13].

Necrotizing anterior scleritis with inflammation

Necrotizing scleritis is the most serious clinical presentation of anterior
scleritis. This condition has an older age of onset when compared with
the other types of scleritis, and a higher proportion of patients (50% to
80%) have an underlying systemic disease [14]. The two diseases most often
associated with necrotizing anterior scleritis are Wegener’s granulomatosis
and rheumatoid arthritis [13,16]. Necrotizing scleritis typically requires ther-
apy with glucocorticoids or immunosuppressive drugs to control the disease

Fig. 2. Diffuse anterior scleritis. In addition to the bright red episcleral vessels, there is a deep
violaceous hue to the sclera that indicates scleral inflammation.

Fig. 3. Active (left) and resolved (right) diffuse anterior scleritis in a 50-year-old man with We-
gener’s granulomatosis after treatment with prednisone and cyclophosphamide. Resolution of
the scleral inflammation has left a bluish gray hue that represents rearrangement of the collagen
fibers in the sclera.

[3]. Patients present with typical signs of anterior scleritis combined with
areas of white sclera surrounded by edema and congestion (Fig. 5). The
areas of white sclera represent capillary closure of the episcleral vasculature,
with infarction and necrosis of the underlying sclera. Involvement of adja-
cent ocular structures including the cornea, ciliary body, and trabecular
meshwork is observed with secondary corneal ulceration, uveitis, or in-
creased intraocular pressure [3]. After resolution of scleritis with appropriate
treatment, there is thinning of the sclera with translucency, and the choroid
often is seen through the sclera (Fig. 6). Despite such thinning, rupture of
the globe is rare.

Scleromalacia perforans (necrotizing anterior scleritis

without inflammation)
Scleromalacia perforans, also known as necrotizing anterior scleritis
without inflammation, is a rare but severe form of scleritis that tends to

Fig. 4. Nodular anterior scleritis in a 44-year-old man with hepatitis C. Along with areas
of diffuse scleral inflammation, a discrete, raised, scleral nodule is seen near the limbus at the
5 o’clock position.

Fig. 5. Necrotizing anterior scleritis in a 49-year-old woman with hepatitis C–associated cryo-
globulinemia. Along with nasal episcleral and scleral inflammation, there is an area of scleral
whitening that represents an area of necrosis.

involve both eyes and presents without redness, pain, or edema. Often, there
is thinning and atrophy of the episclera and loss of the episcleral vasculature
(Fig. 7). Localized areas of yellow-white infarcted tissue are seen. Such areas
are demarcated clearly from surrounding tissue. Thinning may become so
pronounced that the choroid is covered by conjunctivae alone. Given the
lack of typical symptoms, patients often present with discolored sclera or
blurred vision secondary to astigmatism due to scleral thinning. The classic
patient with scleromalacia perforans is an elderly woman with long-standing
rheumatoid arthritis [13].

Posterior scleritis
Posterior scleritis is defined as involvement of the sclera by inflammation
posterior to the insertion of the medial and lateral rectus muscles. This form

Fig. 6. Resolution of idiopathic necrotizing anterior scleritis in a 36-year-old woman after treat-
ment with prednisone and mycophenolate mofetil. The sclera is thin and translucent in the area
of previous necrotizing inflammation. The choroidal hue can be appreciated through this area.

Fig. 7. Necrotizing anterior scleritis without inflammation (scleromalacia perforans). The pa-
tient presented with scleral necrosis and minimal pain. The bluish scleral hue is created by
the transmission of choroidal pigment through the thin sclera.

of scleritis is difficult to diagnose because of the secluded nature of its ocular

inflammation and its nonspecific clinical features [13]. Patients may present
with deep-seated headaches, pain on movements of the eye, proptosis, pto-
sis, double vision, and reduced visual acuity. Ocular examination may be
normal or may disclose localized or generalized exudative retinal detach-
ment (21% in one study of 99 patients with posterior scleritis), optic nerve
edema (18%), subretinal mass lesion (13%), choroidal effusion (4%), uveitis
(2%), and vasculitis (2%) [2]. Posterior scleritis may also be accompanied by
macular edema, retinal striae, and choroidal folds.
Patients with posterior scleritis are less likely to have an underlying sys-
temic disease when compared with those with anterior scleritis [3,12]. Diag-
nosis can be made based on the typical B-scan ultrasonography appearance
of thickened sclera with fluid in Tenon’s space (Fig. 8). An ultrasonographic
finding referred to as the T sign is created by hypoechogenic fluid that forms
an interface between the optic nerve and the sclera.

Differential diagnosis of scleritis

A variety of other etiologies of red eyes must be considered in the differ-
ential diagnosis of scleritis (Box 1). Episcleritis refers to inflammation of the
superficial vessels within the episclera. Ocular erythema in episcleritis is lim-
ited usually to only a sector of the eye and is not associated with pain, vision
changes, or discharge. Examination reveals episcleral injection, which is
more likely to be bright red than deeply violaceous in color.
The vessels involved in episcleritis are more superficial than those in scler-
itis, have a radial pattern, and can be moved with a cotton tip applicator.
Furthermore, palpation of the area does not result in pain. Phenylephrine
blanches the episcleral vessels in episcleritis but will not do so in scleritis.
The condition tends to be self-limited but intermittently recurrent.

Fig. 8. B-scan ultrasonography of idiopathic posterior scleritis in a 51-year-old woman. The fig-
ure demonstrates thickened sclera with hypoechogenic fluid in the space behind the sclera.

Episcleritis is less likely to be associated with a systemic autoimmune condi-

tion than is scleritis [17]. Ocular complications are also less likely in episcler-
itis, with vision loss occurring in none of the 37 patients in one study and in
2 of 94 (not related directly to episcleritis) in another [3,17]. Treatment is not
necessary unless the redness is bothersome to the patient. Therapy may in-
clude topical glucocorticoids and nonsteroidal anti-inflammatory drugs
Conjunctivitis can be caused by a variety of processes including bacterial
or viral infections, allergies, and prolonged contact lens use. Symptoms in-
clude tearing, discharge, and a foreign body sensation. Examination reveals
a follicular or papillary reaction of the palpebral conjunctivae. Treatment of
bacterial conjunctivitis consists of antibiotic eyedrops. Treatment of viral
conjunctivitis is supportive, with cool compresses and artificial tears. Aller-
gic conjunctivitis can also be treated with cool compresses as well as artifi-
cial tears and a topical antihistamine or mast cell inhibitor eyedrops.

Box 1. Differential diagnosis of scleritis

Keratoconjunctivitis sicca
Anterior uveitis

Blepharitis refers to inflammation of the eyelid margins. Symptoms of

blepharitis include burning, tearing, and crusting of the eyelashes. Symp-
toms are typically worse in the morning. Ocular examination reveals crust-
ing around the eyelashes, oil inspissation, and telangiectasias of the eyelid
margin. Treatment for blepharitis involves warm compresses followed by
gentle massage to clean the eyelashes and oil gland openings. Oral antibi-
otics from the tetracycline class are used in more severe cases.
Keratoconjunctivitis sicca can also present with red eyes. Symptoms in-
clude foreign body sensation and burning, typically worse at the end of
the day. Although it is paradoxical, the eyes may also tear. Examination re-
veals a decreased tear lake and roughness of the corneal surface. Treatment
includes tear replacement therapy, plugs in the cannuliculi, and topical
Anterior uveitis refers to inflammation of the iris and ciliary body. Symp-
toms include pain, sensitivity to light, and blurred vision. The redness is
more pronounced over the perilimbal (circumcorneal) region, which overlies
the inflamed ciliary body. The pupil can be miotic and poorly reactive to
light. Examination reveals cells and flare in the anterior chamber. The
first-line therapy for anterior uveitis consists of topical glucocorticoid eye-
drops and a mydriatic agent. Dilating the pupils is essential for the preven-
tion of synechiae between the iris and lens. More severe cases are treated
with oral glucocorticoids or immunosuppressive drug therapy.

Systemic conditions associated with scleritis

Fifty percent of patients with scleritis are diagnosed with an associated
systemic condition [3,18]. Disorders associated with scleritis can be divided
three categories: inflammatory, infectious, and other (Table 1). Autoim-
mune conditions are found in approximately 40% of patients and infections
in approximately 7% [3,18]. The most commonly associated rheumatic dis-
eases are rheumatoid arthritis (18%–33%); systemic vasculitis (7%–19%),
of which Wegener’s granulomatosis is the most common; systemic lupus er-
ythematosus (4%–7%); inflammatory bowel disease (4%–7%); and relaps-
ing polychondritis (3%) [3,17]. Less commonly associated conditions
include sarcoidosis [19,20], cryoglobulinemia [21], and hypocomplemente-
mic urticarial vasculitis [22]. The most commonly associated infection is
herpes zoster [18].
In patients with scleritis and an associated systemic condition, the under-
lying disease is already known in about 80% of patients at the time scleritis
presents. In one study, approximately 15% of patients were diagnosed with
a new systemic condition as a result of the initial evaluation, and in an ad-
ditional 8% a systemic disease developed during follow-up [18]. A systemic
vasculitis was more likely to be discovered by the initial diagnostic evalua-
tion than were other rheumatic diseases [18]. Among patients with no sys-
temic diagnosis after initial evaluation, a new rheumatic disease was

Table 1
Systemic diseases associated with scleritis
Autoimmune diseases Infections Other
Rheumatoid arthritisa Viruses (varicella zoster virus, Surgically induced
herpes simplex virus, hepatitis C)
Inflammatory bowel disease Bacteria Drug induced
Relapsing polychondritis Mycobacteria Pamidronate
Systemic lupus Spirochetes (Treponema pallidum, Fluvirin
erythematosus Borrelia burgdorferi)
Spondyloarthropathies Fungi Trauma
Sarcoidosis Amoeba
Vasculitides Parasite
Wegener’s granulomatosisa
Polyarteritis nodosa
Cogan’s syndrome
urticarial vasculitis
Temporal arteritis
Takayasu arteritis
Most commonly associated conditions.
Data from Okhravi N, Odufuwa B, McCluskey P, et al. Scleritis. Surv Ophthalmol 2005;

diagnosed at a rate of 4% per person-year [18]. This incidence appears to be

higher than the incidence of developing a new rheumatic disease in the
general population [23]. Among patients with apparent scleritis who have
a known underlying inflammatory condition, an important consideration
is the exclusion of an infectious complication of immunosuppression.

Clinical, radiologic, and laboratory evaluation

Evaluation should begin with a careful history, review of systems, and
physical examination. Specific questions should be directed toward symp-
toms of systemic conditions commonly associated with scleritis including
rheumatoid arthritis, Wegener’s granulomatosis, and inflammatory bowel
disease. Infectious etiologies also should be considered, and a history of
trauma or surgical insult should be sought.
Routine testing typically includes a complete blood count and metabolic
panel and a urinalysis with microscopy (Table 2). In addition, serologic test-
ing for antinuclear antibodies, antineutrophil cytoplasmic antibodies
(ANCA), rheumatoid factor, and antibodies to cyclic citrullinated peptides
should also be performed. Additional serologic assays designed to exclude
inflammatory conditions (eg, for complement levels or other types of auto-
antibodies) may be useful if dictated by the presence of specific clinical fea-
tures or by the results of initial serologic testing.

Table 2
Clinical, radiologic, and laboratory evaluation for patients with scleritis
Directed based on history
Standard and physical examination
Complete blood count Tuberculin skin test
Complete metabolic panel Sacroiliac joint radiographs
Urinalysis with microscopic analysis Sinus imaging
Antineutrophil cytoplasmic antibody assay Viral hepatitis panel
Antinuclear antibody assay
Rheumatoid factor
Anticyclic citrullinated peptide antibodies
Chest radiograph IgE levels
Rapid plasma regain Gastrointestinal evaluation
Fluorescent treponemal antibody absorption Cultures for bacteria, virus, fungi
(FTA/ABS) assay
Lyme antibody Scleral biopsy

Spirochetal infections should be excluded with a rapid plasma reagin test,

a fluorescent treponemal antibody (FTA-ABS) assay, and serological testing
for Lyme disease. Chest imaging should be performed with a radiograph or
CT scan. The latter is preferred if the suspicion for Wegener’s granulomato-
sis is strong. Other potential directed tests may include a tuberculin skin test,
sacroiliac joint radiographs (for spondyloarthropathy), a CT scan of the si-
nuses, and serologies for a viral hepatitis panel (hepatitis B and C). When an
infectious form of scleritis is suspected, cultures or a scleral biopsy or both
may be needed to secure the diagnosis.
If posterior scleritis is suspected, ultrasonography is useful in evaluating
the posterior globe for scleral thickening and fluid in Tenon’s capsule
(Fig. 8). Fluid in Tenon’s capsule seen in the plane of the optic nerve is
called the T sign. A CT scan of the orbits also can be of use in posterior
scleritis, showing thickening of the sclera or orbital inflammation.

Complications of scleritis
Scleritis can be associated with significant eye morbidity [13]. Visual loss
can result from a number of complications including uveitis, cataracts, cor-
neal melts, glaucoma, and posterior segment disease. Specific complications
of posterior scleritis can include optic disk edema, cystoid macular edema
(swelling of the macula), exudative retinal detachment, and choroidal folds
(wrinkles in the choroid, the layer of the eye interposed between the retina
and sclera). Other less common complications include scleral thinning and
globe rupture with minor trauma.
In a retrospective study of 172 patients, anterior uveitis was found in
42% of patients with scleritis, cataracts were detected in 17%, PUK in
14%, glaucoma in 13%, and posterior segment disease in 6% [17]. In that
same study, 37% of patients experienced a decrease in vision due to scleritis,
defined as a loss of two or more lines of vision during follow-up or a visual

acuity of 20/80 or worse at presentation. Vision loss was more commonly

seen in patient with necrotizing scleritis (82%) [17]. Another retrospective
study of 290 patients found that vision loss (defined as a permanent drop
in Snellen acuity of two or more lines) occurred in 9% of patients with dif-
fuse anterior scleritis, 26% of patients with nodular scleritis, 74% of those
with necrotizing disease, and 84% of those with posterior scleritis [14]. Vi-
sion loss (loss of 2 or more Snellen lines from initial best-corrected acuity)
is much more likely in patients with posterior scleritis (31% [2]) or an under-
lying systemic disorder (27% [3]).

In a retrospective study of 97 patients with scleritis, 30% required
NSAIDs alone, 31% required oral glucocorticoids, and 26% required im-
munosuppressive drugs in addition to glucocorticoids to control their dis-
ease [3]. Treatments varied with the specific type of scleritis. Nodular
anterior scleritis responded more often to NSAIDs (57%), whereas necrotiz-
ing scleritis often required systemic immunosuppressive drugs (70%). Poste-
rior scleritis was more often treated with oral glucocorticoids (83%) and
occasionally immunosuppressive drugs (17%) [3].
A step-ladder approach typically is used in the treatment of scleritis. The
first-line therapy in treating non-necrotizing scleritis that is not caused by an
infection is an NSAID. Two types of NSAIDs have been shown to be effec-
tive: flurbiprofen, 100 mg three times daily, and indomethacin, 25 to 50 mg
three times daily [3,13]. If one NSAID does not relieve the pain and inflam-
mation, another may be tried.
Systemic glucocorticoids are used in three general settings: (1) when
NSAIDs prove ineffective, (2) in cases of necrotizing anterior scleritis, and
(3) in cases of posterior scleritis. The usual starting dose is 1 mg/kg/d (max-
imum of 60 mg/daily) followed by a tapering schedule based on clinical re-
sponse. Table 3 describes the typical tapering schedule used in the authors’
clinic. In patients whose disease appears particularly aggressive, pulse meth-
ylprednisolone can be administered intravenously at a dose of 1 g/d for 3
days, followed by the initiation of prednisone, 60 mg/d. The role of depot
glucocorticoid treatment is controversial given the potential risk for scleral
perforation with this therapy [24–26].

Table 3
Typical prednisone taper
Dose Duration
1 mg/kg (maximum 60 mg/d) 1 mo
40–60 mg/d Taper by 10 mg/wk
20–40 mg/d Taper by 5 mg/wk
10–20 mg/d Taper by 2.5 mg/wk
0–10 mg/d Taper by 1–2.5 mg/wk

Immunosuppressive drug therapy is instituted under several clinical cir-

cumstances: for necrotizing scleritis, which usually requires cyclophospha-
mide plus glucocorticoids from the outset of treatment; when other types
of scleritis are not controlled by 1 month of high-dose glucocorticoids;
when more than 10 mg/d of prednisone is required to maintain disease con-
trol; and when glucocorticoid-related adverse effects become an issue [3,27].
A variety of glucocorticoid-sparing agents have been used in the treat-
ment of scleritis (Table 4). Cyclophosphamide (up to 250 mg/kg/d) is the
drug of choice for patients with necrotizing scleritis and in patients with
scleritis associated with a systemic vasculitis such as Wegener’s granuloma-
tosis [3]. The justification for the use of an alkylating agent in such cases is
the high risk for progressive ocular damage, extraocular vasculitic lesions,
and death [28]. For patients with non-necrotizing scleritis who require a
glucocorticoid-sparing agent, first-line treatment typically consists of
methotrexate (up to 25 mg/wk), azathioprine (up to 200 mg/d), or mycophe-
nolate mofetil (1 g twice daily). In a retrospective study that examined the
clinical outcomes of 50 patients treated with these agents, 46% achieved dis-
ease quiescence and were able to taper their prednisone to 10 mg/d or less
(Galor and colleagues, unpublished data, 2007). Depending on the severity
of disease, treatment typically is continued for 1 to 2 years after control of
Possible second-line agents for scleritis include calcineurin inhibitors (cy-
closporine or tacrolimus), infliximab, or rituximab [29–33]; however, none
of these agents has been studied rigorously to date. In some cases of

Table 4
Treatments for non-infectious scleritis
Antimetabolites T-cell inhibitors Alkylating agents Biologics
Methotrexatea, start Cyclosporinec, start Cyclophosphamideb, Infliximabc, start
15 mg weekly; 2 mg/kg twice start up to 5 mg/kg every
maximum, 25 mg daily; maximum, 250 mg/kg daily; 6 wks; maximum,
weekly 2.5 mg/kg twice maximum, 200 10 mg/kg every
daily mg/daily 4 wks
Azathioprine, start Tacrolimus, start Chlorambucil, start Rituximab, optimal
2 mg/kg daily; 1 mg twice daily; 0.1 mg/kg daily; dose uncertain
maximum, 200 mg increase dose until maximum,
daily therapeutic 12 mg/daily
mofetila, start 1 g
twice daily;
maximum, 1.5 g
twice daily
First-line treatment in cases in which inability to control disease on desired prednisone
dose. Generally, start for active disease on R10 mg prednisone/d.
First-line agent in cases of systemic vasculitis.
Second-line agents added if maximal tolerated dose of antimetabolite alone does not allow
for control of disease on desired prednisone dose.

necrotizing anterior scleritis or scleromalacia perforans, surgical therapy is

required to address extensive scleral thinning and avoid globe rupture.
Scleral grafting surgery may be performed with donor sclera, periosteum,
or fascia lata. Simultaneous efforts to control the underlying inflammation
with medical therapy are essential when surgery is required.
The treatment of scleritis with immunosuppressive medications is
fraught with the potential for treatment-related morbidity and mortality.
Patients on high-dose glucocorticoids and other immunosuppressive
agents require careful monitoring with frequent clinic visits and blood
work. Patients on cyclophosphamide should have a complete blood count
checked not less often than every 2 weeks. Prophylaxis against Pneumo-
cystis carinii (jiroveci) pneumonia (PCP) is also critical, particularly
when the combination of glucocorticoids and another immunosuppressive
agent is used. One common approach to PCP prophylaxis is to use tri-
methoprim-sulfamethoxazole (one single-strength tablet daily). Evaluation
for and prophylaxis against glucocorticoid-induced bone loss must also be

Peripheral ulcerative keratitis

PUK is a form of ocular inflammation that involves the outer portions of
the cornea and may be associated with systemic conditions such as rheuma-
toid arthritis, Wegener’s granulomatosis, and other systemic conditions. The
most severe complication of PUK, corneal melt syndrome, is the progres-
sion of marginal corneal thinning to perforation of the cornea. Corneal
melt syndromes can lead to the abrupt and permanent loss of vision in
the involved eye.

Anatomy of the cornea

The cornea is a transparent avascular structure that allows light into the
eye and serves as the eye’s main focusing instrument (Fig. 8). The peripheral
cornea, the site of disease in PUK, is distinct from the central cornea in its
anatomic and physiologic characteristics [34]. The peripheral cornea is near
the capillary bed and partly derives its nutrients from the capillary arcades
[35]. Proximity to this vascular and lymphatic arcade may explain the pe-
ripheral location of PUK, because inflammatory cells and mediators can
gain access to this part of the cornea more readily.

Epidemiology of peripheral ulcerative keratitis

PUK is less common that scleritis, with an incidence of 3 cases per 1 mil-
lion persons per year in one study from England [36]. Women and men ap-
pear to be affected equally [37,38].

Pathogenesis of peripheral ulcerative keratitis

The pathogenesis of PUK has not been elucidated completely, but both
T-cell and antibody-mediated pathways have been implicated in the disease
process [39]. Abnormal T-cell responses have been found in several studies
on PUK [39–41]. It is hypothesized that T cells lead to antibody production
and the formation of immune complexes that deposit in the peripheral cor-
nea [42]. The complement pathway is activated with recruitment of inflam-
matory cells to the cornea. Collagenases and other proteases are secreted by
neutrophils and macrophages, which leads to destruction of the peripheral
corneal stroma [39]. Histopathologic examinations of cornea and conjunc-
tiva from patients with PUK reveal a multitude of inflammatory cells in-
cluding plasma cells, neutrophils, mast cells, and eosinophils [39].

Clinical history
Patients with PUK present with ocular pain and redness. Other symp-
toms may include tearing, photophobia, and decreased vision due to astig-
matism or corneal opacity.

Clinical findings
PUK typically presents as a crescent-shaped corneal ulcer found within
2 mm of the limbus (Fig. 9). The epithelium is absent overlying the ulcer,
and the underlying stroma is thinned. There is a variable amount of cellular
infiltrate within the stroma and inflammation involving the contiguous con-
junctiva and sclera. In a review of 47 PUK patients, scleritis was present si-
multaneously in 36% [37]. Conversely, in a retrospective study of patients
with scleritis, 14% of patients also had PUK. PUK is observed most often
in patients with necrotizing scleritis and is bilateral in as many as 40% of
cases [17,37,38].

Fig. 9. Idiopathic PUK in a 43-year-old man. A crescent-shaped corneal ulcer with minimal
discharge is seen at the nasal limbus.

A Mooren’s ulcer is a subtype of PUK typified by extreme eye pain, no

associated scleral involvement or systemic findings, and an overhanging cor-
neal edge.

Differential diagnosis
The differential diagnosis of PUK includes other conditions that affect
the peripheral cornea and lead to peripheral corneal thinning or scarring.
These conditions include non-inflammatory conditions (eg, Terrien’s mar-
ginal degeneration, pellucid marginal degeneration, and senile furrows) as
well as others associated with inflammation or infection (eg, staphylococcal
marginal keratitis, phlyctenulosis, vernal keratoconjunctivitis). Other local
insults may also lead to peripheral corneal pathology, including poorly fit-
ting contact lenses, corneal exposure, misdirected lashes (trichiasis), and
complications of ocular surgical procedures. The presence of keratoconjunc-
tivitis sicca and meibomian gland dysfunction must be evaluated in every
patient with corneal thinning, because these factors may cause or contribute
to peripheral corneal abnormalities.

Systemic conditions associated with peripheral

ulcerative keratitis
As is true in scleritis, PUK may be associated with systemic conditions
and may be an early manifestation of an underlying vasculitis [36,43]. One
retrospective study from a tertiary care referral center reported that an un-
derlying systemic disease was found in 25 of 47 patients [37]. Although most
systemic conditions were known before the time of diagnosis, approximately
one quarter of the predisposing conditions were undiagnosed, reinforcing
the importance of a careful medical history, comprehensive review of sys-
tems, and appropriate laboratory testing [37]. Rheumatoid arthritis, found
in 34% of patients in one study and 42% in another, was the most common
associated disease [37,38]. Other systemic conditions associated with PUK
include the ANCA-associated vasculitides (Wegener’s granulomatosis,
Churg-Strauss syndrome, and microscopic polyangiitis), polyarteritis no-
dosa, relapsing polychondritis, and systemic lupus erythematosus [37,38].
In rheumatoid arthritis, corneal melting usually occurs in a patient with
long-standing disease, that is, the same type of patient who is prone to the
development of rheumatoid vasculitisdnodular, erosive, and rheumatoid
factor positive. In one study, there was a mean of 19.6 years between the di-
agnosis of rheumatoid arthritis and PUK [36]. In Wegener’s granulomatosis
and other forms of systemic vasculitis, corneal melting can occur earlier in
the disease course. Regardless of the underlying condition with which it is
associated, corneal melting can occur swiftly once inflammation begins
within the cornea. Visual loss can ensue within days.
Ocular and systemic infections may also cause or be associated with PUK.
Microbial pathogens implicated in the etiology of PUK include bacteria

(Staphylococcus and Streptococcus sp), spirochetes (Treponema pallidum), my-

cobacteria (tuberculosis), viruses (hepatitis C, herpes simplex virus, varicella
zoster virus), Acanthamoeba, and fungi [39].

Laboratory evaluation
The clinical, radiologic, and laboratory evaluation in PUK is identical to
that of scleritis (see Table 2).

Complications of peripheral ulcerative keratitis

PUK may be associated with significant eye and systemic morbidity. In
one retrospective review of 24 patients with scleritis-associated PUK, all
24 patients had impending corneal perforation, 16 (67%) had an associated
anterior uveitis, and 20 (83%) had decreased vision defined as a decrease in
visual acuity of 2 or more Snellen lines at the end of follow-up or visual acu-
ity of 20/80 or worse at presentation [38]. In another retrospective review of
47 patients with PUK, 34% had impending or frank corneal perforations
(defined as peripheral corneal thinning of 75%–100%), 47% required a tec-
tonic graft procedure, 9% had an associated anterior uveitis, and 43% had
visual acuity of 20/400 or worse at presentation [37]. In addition to the high
risk of ocular damage from PUK, eye inflammation of this nature is a har-
binger of active inflammatory disease in other organ systems with major po-
tential for morbidity and mortality [28].

The treatment of PUK is determined by the severity of findings within the
cornea and the extent of extraocular disease. Systemic glucocorticoids are the
cornerstone of therapy for non-infectious PUK. Glucocorticoids are started at
a dose of 1 mg/kg/d (maximum of 60 mg daily), followed by a tapering sched-
ule based on clinical response. As is true in scleritis, patients in whom loss of
vision is imminent should be treated with pulse methylprednisolone, 1 g/d for
3 days. Topical glucocorticoids are not an appropriate therapy for PUK.
Alkylating agents may be used in conjunction with glucocorticoids in cases
of PUK with imminent danger of corneal perforation and in cases associated
with a systemic vasculitis. Because of the high likelihood of ocular morbidity
resulting from ineffective treatment of PUK, the usual approach is to employ
both cyclophosphamide (up to 2 mg/kg/d) in addition to high-dose glucocor-
ticoids from the outset of therapy. As with scleritis, steroid-sparing agents are
also added when an initial course of high-dose glucocorticoids does not con-
trol the disease, when disease control cannot be maintained with less than
10 mg/d of prednisone, and when significant concern related to glucocorti-
coid-related adverse effects exists [27]. Methotrexate, azathioprine, mycophe-
nolate mofetil, and the biologic agents are all employed in a fashion similar to
their use in scleritis (see Table 4).

The treatment of PUK with immunosuppressive medications is fraught with

the potential for treatment-related morbidity and mortality. Patients on high-
dose glucocorticoids and other immunosuppressive agents require careful
monitoring with frequent clinic visits and blood work. Patients on cyclophos-
phamide should have a complete blood count checked not less often than every
2 weeks. Prophylaxis against PCP pneumonia is also critical, particularly when
the combination of glucocorticoids and another immunosuppressive agent is
used. One common approach to PCP prophylaxis is to use trimethoprim-sul-
famethoxazole (one single-strength tablet daily). Evaluation for and prophy-
laxis against glucocorticoid-induced bone loss must also be considered.
Surgical resection of conjunctival tissue adjacent to PUK has been pro-
moted as a means of decreasing access of inflammatory cells and factors
to the peripheral cornea [37,44]. This treatment is controversial because it
is not uniformly agreed that resection alters the course of disease. Surgical
management of PUK is used in cases of impending perforation to preserve
globe integrity. Several surgical options exist depending on the size of per-
foration and include the use of a tissue adhesive bandage contact lens, a la-
mellar graft, and tectonic corneal grafting.
A study of 34 patients with rheumatoid arthritis–associated PUK with or
without necrotizing scleritis found decreased mortality and ocular morbidity
in patients treated with immunosuppressive medication [28]. A lower mor-
tality was seen (6% versus 53%) in patients managed with cyclophospha-
mide (12 patients) or antimetabolite agents (5 patients) when compared
with those managed with oral glucocorticoids and NSAIDS alone. Progres-
sion of ocular disease occurred less often (0% versus 76%) in the former
group as well. Furthermore, no patient treated with immunosuppressive
agents developed extraocular vasculitis while on medication. Another retro-
spective review of 12 patients with rheumatoid arthritis–associated PUK
found that 68% of patients had stable or improved visual acuities after sys-
temic treatment with cyclophosphamide or methotrexate and surgical ther-
apy (required in 63% of eyes) [45].
A retrospective study of 38 patients with PUK who underwent adjacent
conjunctival biopsy found vasculitis in 20 specimens. Eighteen of the 20 pa-
tients were treated with chemotherapy; 17 had no progression of ulceration.
The two patients not treated with chemotherapy stabilized with surgical
management. Of 18 patients without vasculitis on biopsy, 11 were treated
with chemotherapy. All remained stable or improved with combination sur-
gery and chemotherapy. Of the remaining seven patients, six remained sta-
ble after conjunctival resection alone [37].

[1] Okhravi N, Odufuwa B, McCluskey P, et al. Scleritis. Surv Ophthalmol 2005;50(4):351–63.
[2] McCluskey PJ, Watson PG, Lightman S, et al. Posterior scleritis: clinical features, systemic
associations, and outcome in a large series of patients. Ophthalmology 1999;106(12):2380–6.

[3] Jabs DA, Mudun A, Dunn JP, et al. Episcleritis and scleritis: clinical features and treatment
results. Am J Ophthalmol 2000;130(4):469–76.
[4] Joysey VC, Roger JH, Ashworth F, et al. Parallel studies of HLA antigens in patients with
rheumatic heart disease and scleritis: comparisons with three control populations. J Rheu-
matol Suppl 1977;3:84–8.
[5] Smith JR, Mackensen F, Rosenbaum JT. Therapy insight: scleritis and its relationship to sys-
temic autoimmune disease. Nat Clin Pract Rheumatol 2007;3(4):219–26.
[6] Haynes BF, Fishman ML, Fauci AS, et al. The ocular manifestations of Wegener’s
granulomatosis: fifteen years experience and review of the literature. Am J Med 1977;
[7] Bullen CL, Liesegang TJ, McDonald TJ, et al. Ocular complications of Wegener’s granulo-
matosis. Ophthalmology 1983;90(3):279–90.
[8] Fong LP, Sainz de la Maza M, Rice BA, et al. Immunopathology of scleritis. Ophthalmology
[9] Bernauer W, Buchi ER, Daicker B. Immunopathological findings in posterior scleritis. Int
Ophthalmol 1994;18(4):229–31.
[10] Bernauer W, Watson PG, Daicker B, et al. Cells perpetuating the inflammatory response in
scleritis. Br J Ophthalmol 1994;78(5):381–5.
[11] Riono WP, Hidayat AA, Rao NA. Scleritis: a clinicopathologic study of 55 cases. Ophthal-
mology 1999;106(7):1328–33.
[12] Watson PG, Hayreh SS. Scleritis and episcleritis. Br J Ophthalmol 1976;60(3):163–91.
[13] Schwam B. Scleritis. In: Krachmer J, Mannis M, Holland E, editors. Cornea and external
disease: clinical diagnosis and management. 1st edition. vol. II. St. Louis (MO): Mosby;
1997. p. 1479–91.
[14] Tuft SJ, Watson PG. Progression of scleral disease. Ophthalmology 1991;98(4):467–71.
[15] Hakin KN, Watson PG. Systemic associations of scleritis. Int Ophthalmol Clin 1991;31(3):
[16] Foster C. The sclera. New York: Springer-Verlag; 1994.
[17] Sainz de la Maza M, Jabbur NS, Foster CS. Severity of scleritis and episcleritis. Ophthalmol-
ogy 1994;101(2):389–96.
[18] Akpek EK, Thorne JE, Qazi FA, et al. Evaluation of patients with scleritis for systemic dis-
ease. Ophthalmology 2004;111(3):501–6.
[19] Dursun D, Akova YA, Bilezikci B. Scleritis associated with sarcoidosis. Ocul Immunol
Inflamm 2004;12(2):143–8.
[20] Qazi FA, Thorne JE, Jabs DA. Scleral nodule associated with sarcoidosis. Am J Ophthalmol
[21] Kedhar SR, Belair ML, Jun AS, et al. Scleritis and peripheral ulcerative keratitis with hep-
atitis C virus-related cryoglobulinemia. Arch Ophthalmol 2007;125(6):852–3.
[22] Thorne JE, Hernandez MI, Rencic A, et al. Severe scleritis and urticarial lesions. Am J Oph-
thalmol 2002;134(6):932–4.
[23] Galor A, Thorne JE, Jabs DA. Rheumatic disease and scleritis. Ophthalmology 2007;114(6):
[24] Albini TA, Zamir E, Read RW, et al. Evaluation of subconjunctival triamcinolone for non-
necrotizing anterior scleritis. Ophthalmology 2005;112(10):1814–20.
[25] Zamir E, Read RW, Smith RE, et al. A prospective evaluation of subconjunctival injection of
triamcinolone acetonide for resistant anterior scleritis. Ophthalmology 2002;109(4):798–805
[discussion: 805–7].
[26] Tu EY, Culbertson WW, Pflugfelder SC, et al. Therapy of nonnecrotizing anterior
scleritis with subconjunctival corticosteroid injection. Ophthalmology 1995;102(5):
[27] Jabs DA, Nussenblatt RB, Rosenbaum JT. Standardization of uveitis nomenclature for re-
porting clinical data: results of the First International Workshop. Am J Ophthalmol 2005;

[28] Foster CS, Forstot SL, Wilson LA. Mortality rate in rheumatoid arthritis patients develop-
ing necrotizing scleritis or peripheral ulcerative keratitis: effects of systemic immunosuppres-
sion. Ophthalmology 1984;91(10):1253–63.
[29] Murphy CC, Ayliffe WH, Booth A, et al. Tumor necrosis factor alpha blockade with inflix-
imab for refractory uveitis and scleritis. Ophthalmology 2004;111(2):352–6.
[30] Galor A, Perez VL, Hammel JP, et al. Differential effectiveness of etanercept and infliximab
in the treatment of ocular inflammation. Ophthalmology 2006;113(12):2317–23.
[31] Ahmadi-Simab K, Lamprecht P, Nolle B, et al. Successful treatment of refractory anterior
scleritis in primary Sjogren’s syndrome with rituximab. Ann Rheum Dis 2005;64(7):1087–8.
[32] Minami R, Miyamura T, Watanabe H, et al. Successful treatment of a patient with refractory
Wegener’s granulomatosis by rituximab. Nihon Rinsho Meneki Gakkai Kaishi 2007;30(2):
[33] Cheung CM, Murray PI, Savage CO. Successful treatment of Wegener’s granulomatosis
associated scleritis with rituximab. Br J Ophthalmol 2005;89(11):1542.
[34] Abelson M, Ousler G, Plummer A. An in-depth look at ‘our two corneas’. Review of Oph-
thalmology 2005;12. Available at: http://www.revophth.com/index.asp?page¼1_782.htm.
Accessed October 15, 2007.
[35] Freeman RD. Oxygen consumption by the component layers of the cornea. J Physiol 1972;
[36] McKibbin M, Isaacs JD, Morrell AJ. Incidence of corneal melting in association with sys-
temic disease in the Yorkshire Region, 1995–7. Br J Ophthalmol 1999;83(8):941–3.
[37] Tauber J, Sainz de la Maza M, Hoang-Xuan T, et al. An analysis of therapeutic decision
making regarding immunosuppressive chemotherapy for peripheral ulcerative keratitis.
Cornea 1990;9(1):66–73.
[38] Sainz de la Maza M, Foster CS, Jabbur NS, et al. Ocular characteristics and disease associ-
ations in scleritis-associated peripheral keratopathy. Arch Ophthalmol 2002;120(1):15–9.
[39] Dana M, Qian Y, Hamrah P. Twenty-five-year panorama of corneal immunology: emerging
concepts in the immunopathogenesis of microbial keratitis, peripheral ulcerative keratitis,
and corneal transplant rejection. Cornea 2000;19(5):625–43.
[40] Foster C, Kenyon K, Greiner J. The immunopathology of Mooren’s ulcer. Am J Ophthal-
mol 1979;88:149–59.
[41] Mondino B, Brown S, Rabin B. Cellular immunity in Mooren’s ulcer. Am J Ophthalmol
[42] Mondino B. Inflammatory diseases of the peripheral cornea. Ophthalmology 1988;95:
[43] Ladas JG, Mondino BJ. Systemic disorders associated with peripheral corneal ulceration.
Curr Opin Ophthalmol 2000;11(6):468–71.
[44] Messmer EM, Foster CS. Vasculitic peripheral ulcerative keratitis. Surv Ophthalmol 1999;
[45] Messmer EM, Foster CS. Destructive corneal and scleral disease associated with rheumatoid
arthritis: medical and surgical management. Cornea 1995;14(4):408–17.
Rheum Dis Clin N Am 33 (2007) 855–874

Cogan’s Syndrome: An Audiovestibular,

Ocular, and Systemic
Autoimmune Disease
Mehrdad Mazlumzadeh, MDa,b,*,
Eric L. Matteson, MD, MPHa,c
Mayo Clinic College of Medicine, Rochester, MN, USA
Division of Rheumatology, Mayo Clinic, 13400 East Shea Boulevard,
Scottsdale, AZ 85259, USA
Division of Rheumatology, Mayo Clinic, 200 First Street, SW,
Rochester, MN 55901, USA

Cogan’s syndrome (CS) is a rare systemic disease of unknown cause that

is associated with characteristic patterns of inflammation in the eyes and in-
ner ear. CS is often classified among the vasculitides. The hallmarks of the
disease are progressive inflammatory involvement of ocular and audioves-
tibular organs, but many patients also develop systemic inflammation in
blood vessels of various sizes, usually large arteries.
Mogan and Baumgartner initially reported the association of nonsyphi-
litic interstitial keratitis and audiovestibular involvement in a patient in
1934 [1]. The first description of the syndrome by an ophthalmologist
occurred in 1945, when David G. Cogan of the Massachusetts Eye and
Ear Infirmary reported a ‘‘syndrome of nonsyphilitic interstitial keratitis
and audiovestibular symptoms’’ [2] in four patients. All of the patients pro-
gressed to deafness and one became blind also. Despite the clinical
resemblance to luetic disease, none of these patients had syphilis. Cogan

Mehrdad Mazlumzadeh has received grants from the Mayo Foundation. Eric L. Matteson
is an Investigator for Amgen, Biogen-IDEC, Centocor, Genentech, Hoffmann-LaRoche,
Human Genome Sciences, and Wyeth. Dr. Matteson has received grant support from Amgen,
Aventis, Bristol-Myers-Squibb, Centocor/Johnson & Johnson, Genentech, Mayo Foundation,
Novartis, and the National Institutes of Health. Dr. Matteson has served as consultant to
and/or on the scientific advisory board of Amgen, Biogen-IDEC, Burroughs-Wellcome,
Centocor, Novartis, Regeneron, Takeda, and the Vasculitis Foundation.
* Corresponding author. Division of Rheumatology, Mayo Clinic, 13400 East Shea
Boulevard, Scottsdale, AZ 85259.
E-mail address: mazlumzadeh.mehrdad@mayo.edu (M. Mazlumzadeh).

0889-857X/07/$ - see front matter Ó 2007 Elsevier Inc. All rights reserved.
doi:10.1016/j.rdc.2007.07.015 rheumatic.theclinics.com

reported an additional four cases in 1949 [3]. Despite more than 60 years of
recognition by the medical community, CS remains a challenging disease for
the clinicians that leads to severe hearing loss and deafness in at least 50% of
patients and may also result in other important comorbidities. In many
cases, poor outcomes may be related to delay in diagnosis and failure to in-
stitute appropriate therapy in a prompt manner.

Insights into the epidemiology of CS can be gleaned from consideration
of the five major case series published in the literature (Table 1) [4–9]. These
five case series include:
 The National Institute of Health (NIH) series, reported by Haynes and
colleagues [4], which described 13 patients and reviewed an additional
111 cases reported in the medical literature.
 A Mayo Clinic series of 18 patients reported by Vollertsen and col-
leagues [5,6], which included a literature review of 60 additional patients.
 A Duke University series of patients (published in abstract form) that
included the original 13 patients from the NIH series plus an additional
34 patients [7].
 A series from the French Study Group for Cogan’s Syndrome of 32
patients and 222 additional cases reported in the literature [8].
 A series of 60 patients evaluated at the Mayo Clinic between 1940 and
2002 [9]. These patients included the ones reported earlier by Vollertsen
and colleagues [5,6].
Review of these reports indicated that there is considerable overlap in the
patients who have CS described in the medical literature. It is estimated that
250 patients with CS have been reported, but even this number probably in-
cludes some cases that were counted more than once in different
publications. On the other hand, the actual incidence and prevalence of
CS is probably higher than suggested by the reported cases in the literature.
It is likely that many cases remain unrecognized or are labeled inaccurately
as other diseases, such as idiopathic autoimmune inner ear disease,
idiopathic keratitis, viral syndromes, or idiopathic hearing loss [10].

Table 1
Epidemiologic summary of five major case reports of Cogan’s syndrome
Case series Published Total cases Mean age of presentation (range) Male/female
NIH 1980 13 22 (13–31) 1:12
Mayo Clinic 1986 18 32 (12–54) 5:13
Duke University 1992 47 28.6 (3–50) 16:31
French Group 2004 32 31.9 (5–65) 19:13
Mayo Clinic 2006 60 38 (9–70) 27:33
Data compiled from Refs. [4,6–9].

CS usually occurs in early adulthood with an average onset of about

30 years of age. CS can also develop in children and has also been described
in adults after the age of 60 [7,9,11]. The disease has no striking gender
predilection, even though the number of females reported slightly exceeds
the number of males [4–9]. There have been no reports of ethnic or racial
differences, but most of the case series did not clarify the ethnic or racial
background of the patients. In the study of 32 patients investigated by the
French Study Group for CS, there were 29 white patients and 3 blacks [7].

As is true of most autoimmune diseases, the precise cause of CS remains
elusive. Because of similarities of the ocular disease in CS to certain infectious
disorders of the eye, viral and bacterial causes have been suggested [4,6,12]. In
addition, a prodrome of upper respiratory syndrome with rhinitis, pharyngi-
tis, otitis, and sinusitis has been observed in many patients, including the re-
ported major case series [4–6,8,9]. Several laboratory investigations directed
at elucidating an infectious cause have been pursued by several authors, in-
cluding cultures or serologies for Chlamydia species, Borrelia burgdorferi,
Treponema pallidum, cytomegalovirus, herpes virus, hepatitis viruses, and
parvovirus B19. To date, however, there has been no convincing evidence
of a microorganism that contributes to the etiopathogenesis of CS [4–9].

CS is regarded as a systemic inflammatory process that is not limited to
audiovisual involvement. The presence of constitutional symptoms, the typ-
ical elevations of the erythrocyte sedimentation rate (ESR), C-reactive pro-
tein (CRP), and platelet count, along with the responsiveness of these
symptoms and laboratory signs to glucocorticoids, all underscore the
inflammatory nature of this illness [4–9,13]. Although the importance of
inflammation is clear and the organ-specific disease manifestations are man-
ifest, to date there are few data implicating strongly a role for autoimmunity
in CS (see later discussion). Among potential environmental factors, smok-
ing, a known risk factor for diseases such as rheumatoid arthritis, has been
identified as a possible contributor to CS also [9].

CS is not associated with autoantibodies or other specific immune abnor-
malities that are unique to that disorder. Antinuclear antibodies (ANAs)
were not present among patients studied by Haynes and colleagues [4],
and serum complement levels were normal in most patients described in

that series. Anemia, leukocytosis, and elevations of the ESR were typical of
CS during periods of active disease, and 3 patients (23%) had cryoglobulins
that were not characterized in detail. Among patients reported by the
French Study Group [8], ANAs were absent, only 1 patient tested positive
for rheumatoid factor (RF), 1 had anticardiolipin antibodies, and 2 had
cryoglobulins. Although 2 of the 32 patients were reported to be anti-
neutrophil cytoplasmic antibody (ANCA) positive, the antigen to which
the ANCAs were directed (ie, proteinase-3 or myeloperoxidase) was not
described. In the series of 60 patients from Mayo Clinic, only 1 of the 10
patients tested had antibodies to heat shock protein-70 (HSP-70) [9].
Although at least two case reports exist of myeloperoxidase-specific
ANCAs occurring in the setting of CS [14,15], ANCAs are found in a small
percentage of patients reported who have CS and it is possible that such
cases merely represent a true ANCA-associated vasculitis, such as Wegener
granulomatosis or microscopic polyangiitis, given the overlap in clinical
findings between ANCA-associated vasculitis and CS.
In one study, the IgM and IgG antibodies against the labyrinth and cor-
nea were detected in the sera of five patients who had CS [16]. Assays for
antibodies to inner ear antigens, available from a small number of laborato-
ries for approximately two decades, have suffered significantly from a lack
of standardization and reproducibility. Although the potential role of
such antibodies is appealing to consider, these results need to be confirmed
in other studies and larger numbers of well-characterized patients.

Clues from histopathology?

The histopathology of CS is similarly unclear. There are only a few
reports relating to the pathologic changes of the inner ear, mainly from
postmortem temporal bone examination. Most of these specimens are
from patients who had longstanding disease and died after months or years
of illness, usually treated with significant doses of immunosuppressive
medications. Attempting to decipher clues to the nature of the acute disease
features from such specimens is difficult. Pathologic changes in CS are sim-
ilar to those described in other autoimmune inner ear diseases and include
acute labyrinthitis leading to atrophy of inner ear tissues, endolymphatic
hydrops, focal and diffuse proliferation of fibrous tissue and bone, and
retrograde neuronal degradation. These pathologic features are suggestive
of an inflammatory or ischemic insult to the membranous labyrinth, but
provide no help beyond that [17].

Animal models
Animal studies suggest that immune activation, specifically of T-cells,
may play a central role in the pathogenesis of audiovisual manifestations
of CS. An experimental autoimmune keratitis has been induced in a rat

model by anti-corneal T-cell lines into the irradiated rats [18]. Similarly, lab-
yrinthitis has been produced using the T-cell lines to inner ears of another
rat model [19]. These findings support aberrancy in the immune response
as central to the pathophysiology of CS.

Clinical features
CS typically presents with ocular symptoms of interstitial keratitis that
develop suddenly and resolve gradually (Fig. 1). It is associated with red
and painful eye, photophobia, and lacrimation that can be unilateral or bi-
lateral. Ocular symptoms can develop before, simultaneously with, or after
the onset of audiovestibular symptoms. Manifestations of ocular and co-
chlear involvement usually begin within 1 to 6 months of each other. In
the presence of only eye disease without evidence of cochlear dysfunction
or vice versa, the diagnosis can be extremely challenging until clarified by
time and additional organ involvement.
Audiovestibular involvement usually behaves in a manner reminiscent of
Ménière disease, with episodes of tinnitus, vertigo, nausea, emesis, and fluc-
tuating but progressive hearing loss. Patients may present with variable sys-
temic symptoms, including constitutional symptoms, such as fever, myalgia,
fatigue, and weight loss [4–9].

Ocular manifestations
Box 1 lists the reported ocular signs and symptoms in CS. The most com-
mon symptoms of keratitis at the onset of the disease include photophobia,
pain, and blurred vision. Other associated symptoms include diplopia,

Fig. 1. 40-year-old man who has Cogan’s syndrome. The picture demonstrates interstitial
keratitis, evident as an area of oval-shaped haziness in the upper outer quadrant of the cornea.
The lesion responded to glucocorticoid therapy, but the patient suffered hearing loss and
eventually required cochlear implantation. (Courtesy of J. A. Garrity, MD, Rochester, MN.)

Box 1. Major ocular signs and symptoms in Cogan’s syndrome

Ocular pain or irritation
Blurred vision
Interstitial keratitis
Inflamed optic papilla/papilledema
Posterior uveitis
Retinal vasculitis
Excessive lacrimation
Corneal ulcerations/peripheral ulcerative keratitis

Data compiled from Refs. [4,6–9,20,21].

lacrimation, and ocular irritation with foreign body sensation. Although

the classic ocular manifestation of CS is interstitial keratitis, it is essential
to note that a wide range of ocular inflammatory lesions can occur in this
disease. In addition to interstitial keratitis, uveitis, subconjunctival or
conjunctival hemorrhage, scleritis, episcleritis, choroiditis, tenonitis, retinal
artery occlusion, retinal hemorrhage, papilledema, and exophthalmos have
all been reported in patients who have CS, with or without interstitial

Audiovestibular manifestations
Immune-mediated inner ear disease (IMIED), which can be associated
with auditory and vestibular dysfunction in CS, is identified by several dif-
ferent terms in the medical literature. For example, IMIED is also termed
‘‘autoimmune inner ear disease’’ (AIED) or ‘‘immune-mediated cochleoves-
tibular disease’’ (IMCVD) [20,22–24]. Although some experts consider
IMIED to be a complication of an autoimmune disease process, proof of

Box 2. Typical audiovestibular manifestations of Cogan’s

Hearing loss
Recurrent Ménière-like attacks
Nausea and vomiting
Abnormal corneal reflex
Bilateral deafness

Data compiled from Refs. [4,6–9,20,21].

autoimmunity as a cause of this disease manifestation is lacking. Box 2

summarizes the typical audiovestibular manifestations of CS.
Audiovestibular dysfunction in CS typically presents as sudden unilateral
involvement, but gradually progressive, bilateral sensorineural hearing loss
accompanied by vestibular disease can also occur [20,22–24]. Hearing loss
can be profound and, as discussed later, eventually leads to deafness in
50% of patients or more. In addition to the auditory disease manifestations,
vestibular components of the disease can also be enormously disabling.
These include ataxia, tinnitus, nausea, vomiting, and vertigo [4–9]. The
onset of disease may be heralded by a gradual rather than a sudden hearing
loss in a minority of patients, whereas an even smaller number of patients
may develop recurrent Ménière-like symptoms without significant hearing
loss [4].

Vascular manifestations
In approximately 15% to 20% of patients who have CS, there are overt
manifestations of systemic vasculitis. The vasculitis mainly involves the large
and medium-sized vessels, but findings of small-vessel vasculitis have also
been reported [4–10]. Occurrence of aortitis has been reported in all of the
major case series, and CS has also been reported to occur as an overlap con-
dition associated with other forms of systemic vasculitis, such as Wegener
granulomatous and polyarteritis nodosa [25,26].

Box 3. Systemic manifestations of Cogan’s syndrome

Weight loss
Head and neck
Periorbital pain
Posterior auricular pain
Systolic murmur
Aortic insufficiency
Left ventricular hypertrophy
Transient radiographic abnormalities
Abdominal pain
Gastrointestinal bleeding
Low back pain
Peripheral neuropathy
Cranial neuropathy
Mononeuritis multiplex
Central nervous system
Cerebral spinal fluid pleocytosis
Cerebral infarction
Cutaneous nodules
Nonspecific rash
Lymphatic system

Abnormal urinary sediment
Testicular pain
Renal insufficiency

Data compiled from Refs. [4,6,8,9,20,21].

Patients who have large-vessel disease present with symptoms and signs
typical of Takayasu arteritis: involvement of the aortic arch and stenosis
of its primary branches, and abdominal aortitis, aortic aneurysm, and renal
artery stenosis [4,7–10,27–30]. Clinical features of medium- or small-vessel
vasculitis can include mesenteric arteritis, gastrointestinal hemorrhage,
cutaneous leukocytoclastic vasculitis (from purpura to skin ulcers), and
glomerulonephritis [6–9,29,31].
Cardiac involvement with coronary ostial stenosis and coronary arteritis
may also occur [8,32,33]. Premature coronary artery disease should thus be
considered in patients who have CS who have signs or symptoms of cardio-
vascular disease. A case report described a patient who had CS who suffered
cerebral infarctions and underwent an angiogram revealing multiple steno-
ses of the vertebrobasilar, middle cerebral, and internal carotid arteries [34].

Other systemic manifestations

Some of the systemic manifestations of CS are a consequence of inflam-
mation. Box 3 provides a list of systemic manifestations of CS. Fever,
fatigue, and weight loss were noted in half the patients in the series reported
by Vollersten and colleagues [6]. Up to one third of patients who have CS
present with musculoskeletal complaints, such as myalgias, arthralgias,
and even inflammatory arthritis [4,6,8,9].
Gastrointestinal symptoms or signs develop in between 10% and 30% of
patients [6,8,9]. Renal and genitourinary involvement range from mild ab-
normalities on urinalysis to testicular pain, through glomerulonephritis in
rare patients and renal insufficiency related to renal artery stenosis [6,9].
Aortic sclerosis and insufficiency, left ventricular hypertrophy, and pericar-
ditis have been reported as cardiac manifestations of CS [4,6,8,9,35]. Pleur-
itis and transient abnormalities of chest radiographs have been reported in
a minority of patients, but are of uncertain cause [6,9]. Lymphadenopathy,

splenomegaly, and hepatomegaly are of unknown significance in CS, but

have been reported [4,6,8,9].
Neurologic manifestations of CS are similarly diverse and often difficult
to ascribe with certainty to the underlying disease. Nonspecific headaches,
meningeal inflammation with cerebrospinal fluid pleocytosis, encephalopa-
thy, hallucinations, and cranial or peripheral neuropathy (mononeuritis
multiplex) have been reported in a significant minority [4,6,8,9].

The diagnosis of CS relies on the presence of idiopathic interstitial kera-
titis or other ocular manifestations along with audiovestibular involvement
in the absence of other underlying disorders to explain the patients’ findings.
Below we outline a pragmatic approach to the evaluation of patients who
have suspected CS (Box 4).

Ophthalmologic and audiovestibular tests

Ophthalmologic examination, audiometric testing, and vestibular studies
are the cornerstones of diagnosis in CS. Slit-lamp examinations of the eyes
should be performed when the disease is suspected to diagnose the ocular
disease. Serial ocular and audiovestibular studies, which help to assess dis-
ease progression or response to therapy, should be performed approxi-
mately every 6 months.
Fig. 2 depicts audiograms of two representative patients who had CS be-
fore and after therapy. Most patients who have CS have abnormal audio-
grams at baseline, with hearing loss of high and low frequencies that
sometimes spares the middle frequencies [4–6,8,9,36]. Brainstem auditory
evoked potential abnormalities may signify cochlear disease. Abnormal
caloric responses implicate vestibular involvement [4,6,9,36].

Laboratory studies
Box 5 summarizes the reported laboratory abnormalities in CS. Potential
laboratory findings include anemia, leukocytosis, and elevated markers of
inflammation, including the ESR and CRP. Haynes and colleagues [4] re-
ported elevations in the ESR in all patients during the acute phase of their
disease, with declines following treatment. The search for other useful bio-
markers or characteristic laboratory features has not been fruitful, however.
Assays for ANA, RF, ANCA, and antiphospholipid antibodies are negative
in most patients, and serum complement levels are usually normal. Tests for

Box 4. The diagnostic approach in the evaluation of patients

who have possible Cogan’s syndrome
Ophthalmologic evaluation
Slit lamp
Audiovestibular evaluations
Vestibular studies
Imaging studies
Chest radiography
MRI (or CT) to rule out acoustic neuroma or other pathology
Laboratory tests
Basic labs
CBC, ESR, CRP, creatinine, urinalysis
Autoimmune serologies
ANA, ENA, RF, complements, ANCA
Tests to exclude infection
Bacterial infections
Syphilis (VDRL and FTA-Abs)
Lyme serology
Tuberculosis (PPD)
Viral infections
Epstein-Barr virus
Herpes simplex virus

Abbreviations: ANA, antinuclear antibody; ANCA, anti-neutrophil cytoplasmic

antibody; CBC, complete blood count; CRP, C-reactive protein; ENA, extractable
nuclear antibody; ESR, erythrocyte sedimentation rate; FTA-Abs, fluorescent
treponemal antibody absorption test; PPD, purified protein derivative; RF,
rheumatoid factor; VDRL, Venereal Disease Research Laboratory.
Data compiled from Refs. [6,8,9,20,21].

microbial pathogens that may mimic CS could be performed for the purpose
of exclusion.

Imaging studies
MRI studies of the brain using T1- and T2-weighted sequences and other
specific techniques have been studied in a limited number of patients who
have CS or idiopathic forms of IMIED. These studies have shown narrow-
ing or calcification and soft tissue obliteration of the vestibular labyrinth
space. MRI is effective in visualizing soft tissue intralabyrinthine disease,

Fig. 2. Audiograms of two representative patients who had Cogan’s syndrome before and after
therapy. Audiograms of a patient who had Cogan’s syndrome demonstrating typical bilateral
hearing loss (A), and some improvement on glucocorticoid therapy (B). Audiogram of a patient
who had Cogan’s syndrome (C) who responded suboptimally to glucocorticoid therapy (D).
The glucocorticoid was discontinued after 9 months of treatment. HL, hearing level (this is
the dB level that corresponds to the numbers on the ordinate of the audiogram); masked
AC: masked air conduction threshold; masked BC: masked bone conduction threshold; SL, sen-
sation level (this is the dB level above the SRT, usually 25 or 40 dB SL); SRT, speech reception
threshold (this is the lowest dB level at which patient accurately repeats common two-syllable
words, such as ‘‘airplane’’ or ‘‘ice cream’’ 50% of the time).

Fig. 2 (continued)

but the clinical applications of this capability are not clear. A major use of
MRI in the evaluation of patients who have symptoms of inner ear disease is
the exclusion of tumors of the cerebellar-pontine angle. CT scanning can de-
tect calcified obliterations of the inner ear but, as with some MRI findings,
the clinical usefulness of this is not clear [37,38].
In a pilot study of 10 patients who had IMIED that included 1 patient
who had CS, positron emission tomography (PET) demonstrated potential
for contributing to the diagnosis [39]. Such a conclusion was not verified in
a substudy conducted as part of a multicenter clinical trial in IMIED, how-
ever [40]. The role of PET scans, if any, remains to be determined.

Box 5. Potential laboratory abnormalities in Cogan’s syndrome

Relative lymphopenia
Eosinophilia (mild)
Elevated ESR >20 mm/h
Elevated CRP
Cerebrospinal fluid pleocytosis
Decreased C3
Decreased C4
Decreased total hemolytic complement
ANA (low titer)
RF (low titer)
p-ANCA immunofluorescence
c-ANCA immunofluorescence

Abbreviations: ANA, antinuclear antibody; ANCA, anti-neutrophil cytoplasmic

antibody; c-ANCA, cytoplasmic ANCA; CRP, C-reactive protein; ESR, erythrocyte
sedimentation rate; p-ANCA, perinuclear ANCA; RF, rheumatoid factor.
Data compiled from Refs. [4,6,8,9,20,21].

Differential diagnosis
The differential diagnosis of CS is varied, because inflammatory eye dis-
orders and the clinical symptoms of IMIED may be caused by a lengthy list
of conditions. Making the diagnosis of CS is particularly challenging when
only the ocular or only the audiovestibular manifestations are present by
themselves. Box 6 lists differential diagnostic considerations in evaluation
of patients who have CS.

Medical treatment
In general, the eye disease associated with CS is easier to treat and often
more responsive to therapy than is the audiovestibular component. Most
patients who have CS who develop interstitial keratitis respond well to top-
ical glucocorticoid eyedrops. Atropine eye drops are also administered for
their mydriatic effects. A minority of patients may require other topical

Box 6. The differential diagnosis of Cogan’s syndrome

Chlamydial infection
Lyme disease
Syphilis (congenital or acquired)
Whipple disease
Viral infections
Rheumatologic syndromes
Rheumatoid arthritis
Sjögren syndrome
Systemic lupus erythematosus
Relapsing polychondritis
Wegener’s granulomatosis
Polyarteritis nodosa
Takayasu arteritis
Behçet’s disease
Nonrheumatologic syndromes
Ménière’s disease
Ulcerative colitis
Crohn’s disease
Vogt-Koyanagi-Harada disease

Data compiled from Refs. [4,6,20,21].

eye drops, such as cyclosporine, to achieve control of the eye disease [41].
Interstitial keratitis rarely requires systemic glucocorticoids for disease
control, but other ocular manifestations of CS (eg, scleritis or peripheral
ulcerative keratitis) may mandate intensive immunosuppression with a cyto-
toxic agent in addition to glucocorticoids [4,6]. The therapy of scleritis and
peripheral ulcerative keratitis is discussed elsewhere in this issue of Rheu-
matic Disease Clinics of North America.
Hearing loss in CS requires systemic glucocorticoids. High-dose predni-
sone for the first 2 weeks followed by the initiation of a taper is crucial
for the prevention of further hearing loss from that particular disease flare
[4,6–9,42]. Failure to treat immune-mediated hearing loss aggressively
may lead to profound, permanent hearing loss. Despite control of hearing
damage initially, however, repeated disease flares over time gradually lead
to the loss of hearing in many patients.
The initial management of the auditory disease is with prednisone, 1 mg/kg
per day, up to 60 mg. After 2 weeks at this dose, tapering of the

medication begins in decrements of 5 to 10 mg every 1 to 2 weeks, depend-

ing on the patient’s response. The length of the initial treatment course
with prednisone is on the order of 4 to 6 months. For patients who dem-
onstrate a propensity to flare, consideration should be given to a mainte-
nance dose of prednisone (5 to 10 mg/d) and a steroid-sparing agent (see
later discussion).
Audiometric evaluation should be obtained after 2 to 4 weeks of therapy
to determine whether there has been an adequate response to therapy. If
there is both objective and subjective improvement, then the prednisone
taper should continue as outlined previously. If no audiovestibular improve-
ment is observed within the first 2 to 4 weeks of glucocorticoid therapy, then
a more rapid prednisone taper may be appropriate (unless control of the eye
disease or systemic CS manifestations dictate otherwise). A patient whose
audiovestibular disease has not responded to glucocorticoid therapy is un-
likely to respond to additional immunosuppressive treatment (eg, a cytotoxic
Methotrexate (up to 25 mg/week) is considered by many to be a reason-
able choice for a steroid-sparing agent in CS and other disorders associated
with IMIED [43–45]. Methotrexate failed to demonstrate efficacy in man-
agement of IMIED, however, in a recent double-blinded, randomized
clinical trial [46]. In the experience of the authors, the steroid-sparing benefit
of methotrexate for patients who have CS is variable; a substantial number
of patients do not respond satisfactorily.
Although cyclophosphamide is effective for the treatment of severe ocular
inflammatory manifestations of CS, the experience with this agent in the
treatment of the audiovestibular features in this disease and other forms
of IMIED is mixed [13,47]. Other immunosuppressive therapies used as
glucocorticoid-sparing agents in the management of audiovestibular
manifestations of CS include azathioprine, mycophenolate mofetil, and
cyclosporine [6,8,9,27]. The true efficacy of these agents is not known.
The effect of a tumor necrosis factor-a inhibitor, etanercept, was investi-
gated in an open-label multicenter study of 23 patients who had various
forms of IMIED. Three patients who had CS were included in this trial.
Although etanercept was tolerated well, the treatment was not effective in
preserving or improving hearing loss in the overall trial [40]. Word identifi-
cation and recognition improved in 2 of 3 patients who had CS, however.

Surgical intervention
Cochlear implantation is the most successful surgical approach to man-
agement of end-stage hearing loss in patients who have CS who do not re-
spond to medical therapy or conventional hearing augmentation. In general,
patients who have IMIED are considered good candidates for cochlear im-
plantation, because the deafness occurs in these patients following many
years of hearing experience [48]. Cochlear implants are effective in patients

who have CS, with good to excellent speech recognition without major
complications [48,49].
The overall approach to the medical and surgical management of CS is
outlined in Table 2.

The prognosis for maintenance of vision and resolution of the inflamma-
tory eye lesion is generally good in patients who have CS. Unfortunately,
a substantial number of patients develop profound hearing loss. Vollersten
and colleagues from the Mayo Clinic reported that of 18 patients, 13 expe-
rienced bilateral sensorineural deafness and 2 developed severe unilateral
deafness [7].
In the retrospective review of 60 patients who had CS at Mayo Clinic,
most patients (62%) suffered multiple relapses. Only 6 patients (10%) had
died at the time of last follow-up, however, and the mean survival in the co-
hort was 21 years. Deaths in 4 of the patients were attributable either to CS
or to effects of its treatment [9].
Experience in the French cohort with regard to hearing loss was similar.
Over a mean follow-up period of 7.3 years, 11 of 32 patients became deaf
and 19 of the remaining patients developed severe auditory dysfunction.
Of the 3 patients who died during follow-up, one death was caused by
a ruptured aortic aneurysm secondary to aortitis. Bilateral deafness was
noted in approximately 50% of the 111 patients reviewed in the same
article [8].
In the absence of other systemic disease manifestations, patients who
have CS may have a good prognosis with regard to life expectancy. Adverse
effects from medications such as glucocorticoids, the loss of productivity at
work, the need for retraining for other occupations because of hearing
impairment, and other ways in which the disease impacts patients must be
considered when assessing the disease toll.
CS is a chronic disease that may require treatment for a prolonged pe-
riod of time because patients may develop recurrent flares of the major
disease manifestations with significant risk for hearing loss and other
morbidities rather than mortality. It seems that patients who are affected

Table 2
Approach to medical and surgical management of Cogan’s syndrome
Involved sites Treatment
Eye Topical glucocorticoids and atropine
Audiovestibular Oral glucocorticoids and if responsive then methotrexate.
Cochlear implants for severe hearing loss unresponsive to medical
Systemic vasculitis Glucocorticoids  steroid-sparing agents

by systemic manifestations, especially vascular diseases including aortitis,

are particularly at increased risk for death from complications of the
In the authors’ view, if the eye disease is controlled effectively but the
hearing loss requires protracted, high doses of glucocorticoid treatment,
steroid-sparing therapies, including the use of cytotoxic agents, can be con-
sidered. Although the preservation of native hearing is desirable, the risks
and benefits of this approach must be weighed carefully, particularly in
view of the effectiveness of cochlear implantation.

CS is an immune-mediated systemic disorder characterized by ocular and
audiovestibular inflammation. Although interstitial keratitis in the absence
of an infectious cause is the classic form of eye involvement, other nonspecific
inflammatory eye diseases can also occur, often making prompt diagnosis of
the underlying disease challenging. The audiovestibular disease is difficult to
treat and can lead to profound hearing loss. A poor outcome, especially com-
plete hearing loss, can sometimes be prevented through a timely recognition
and initiation of glucocorticoid therapy at the onset of the disease.

[1] Morgan RF, Baumgartner CF. Ménière’s disease complicated by recurrent interstitial
keratitis: excellent results following cervical ganglionectomy. Western Journal of Surgery
Obstetrics and Gynecology 1934;42:628–31.
[2] Cogan DG. Syndrome of nonsyphilitic interstitial keratitis and vestibuloauditory symptoms.
Arch Ophthalmol 1945;33:144–9.
[3] Cogan DG. Nonsyphilitic interstitial keratitis with vestibuloauditory symptoms. Report of
four additional cases. Arch Ophthalmol 1949;42:42–9.
[4] Haynes BF, Kaiser-Kupfer MI, Mason P, et al. Cogan syndrome: studies in thirteen patients,
long-term follow-up, and a review of the literature. Medicine 1980;59:426–41.
[5] McDonald TJ, Vollertsen RS, Younge BR. Cogan’s syndrome: audiovestibular involvement
and prognosis in 18 patients. Laryngoscope 1985;95:650–4.
[6] Vollertsen RS, McDonald TJ, Younge BR, et al. Cogan’s syndrome: 18 cases and a review of
the literature. Mayo Clin Proc 1986;61:344–61.
[7] McCallum RM, Allen NB, Cobo LM, et al. Cogan’s syndrome: clinical features and out-
comes [abstract]. Arthritis Rheum 1992;35(Suppl):S51.
[8] Grasland A, Pouchot J, Hachulla E, et al. Typical and atypical Cogan’s syndrome: 32 cases
and review of the literature. Rheumatology 2004;43:1007–15.
[9] Gluth MB, Baratz KH, Matteson EL, et al. Cogan syndrome: a retrospective review of 60
patients throughout a half century. Mayo Clin Proc 2006;81(4):483–8.
[10] Gaubitz M, Lübben B, Seidel M, et al. Cogan’s syndrome: organ-specific autoimmune dis-
ease or systemic vasculitis? A report of two cases and review of the literature. Clin Exp Rheu-
matol 2001;19(4):463–9.
[11] Orsoni JG, Zavota L, Vincenti V, et al. Cogan syndrome in children: early diagnosis and
treatment is critical to prognosis. Am J Ophthalmol 2004;137:757–8.

[12] Darougar S, John AC, Viswalingam M, et al. Isolation of Chlamydia psittaci from a patient
with interstitial keratitis and uveitis associated with otological and cardiovascular lesions.
Br J Ophthalmol 1978;62:709–14.
[13] Watanabe K, Nishimaki T, Yoshida M, et al. Atypical Cogan’s syndrome successfully
treated with glucocorticoids and pulse cyclophosphamide therapy. Fukushima J Med Sci
[14] Ikeda M, Okazaki H, Minota S. Cogan’s syndrome with antineutrophil cytoplasmic autoan-
tibody. Ann Rheum Dis 2002;61:761–2.
[15] Suzuki M, Arimura Y, Minoshima S, et al. A case of myeloperoxidase-specific antineutro-
phil cytoplasmic antibody (MPO-ANCA)-related glomerulonephritis associated with
Cogan’s syndrome. Nippon Jinzo Gakkai Shi 1996;38:423–7.
[16] Helmchen C, Arbusow V, Jäger L, et al. Cogan’s syndrome: clinical significance of anti-
bodies against the inner ear and cornea. Acta Otolaryngol 1999;119:528–36.
[17] Schuknecht HF, Nadol JB. Temporal bone pathology in a case of Cogan’s syndrome. Laryn-
goscope 1994;104:1135–42.
[18] Verhagen C, Mor F, Kipp JB, et al. Experimental autoimmune keratitis induced in rats by
anti-cornea T-cell lines. Invest Ophthalmol Vis Sci 1999;40:2191–8.
[19] Gloddek B, Gloddek J, Arnold W. A rat T-cell line that mediates autoimmune disease of the
inner ear in the Lewis rat. ORL J Otorhinolaryngol Relat Spec 1999;61:181–7.
[20] Stone JH, Francis HW. Immune-mediated inner ear disease. Curr Opin Rheumatol 2000;12:
[21] Garcia Berrocal JR, Vargas JA, Vaquero M, et al. Cogan’s syndrome: an oculo-audiovestib-
ular disease. Postgrad Med J 1999;75:262–4.
[22] McCabe BF. Autoimmune sensorineural hearing loss. Ann Otol Rhinol Laryngol 1979;88:
[23] Hughes GB, Kinney SE, Barna BP, et al. Autoimmune reactivity in Cogan’s syndrome: a pre-
liminary report. Otolaryngol Head Neck Surg 1983;91:24–32.
[24] Hughes GB, Barna BP, Kinney SE, et al. Clinical diagnosis of immune inner-ear disease-
Laryngoscope 1988;98:251–3.
[25] Gran JT, Nordvåg BY, Storesund B. An overlap syndrome with features of atypical Cogan
syndrome and Wegener’s granulomatosis. Scand J Rheumatol 1999;28(1):62–4.
[26] Bomholt A, Knudsen JB, Permin H, et al. Profound sensorineural hearing loss in polyarter-
itis nodosa. An atypical case of Cogan’s syndrome. Arch Otorhinolaryngol 1982;236(1):
[27] Raza K, Karokis D, Kitas GD. Cogan’s syndrome with Takayasu’s arteritis. Br J Rheumatol
[28] Tseng JF, Cambria RP, Aretz HT, et al. Thoracoabdominal aortic aneurysm in Cogan’s
syndrome. J Vasc Surg 1999;30:565–8.
[29] Ho AC, Roat MI, Venbrux A, et al. Cogan’s syndrome with refractory abdominal aortitis
and mesenteric vasculitis. J Rheumatol 1999;26:1404–7.
[30] Cochrane AD, Tatoulis J. Cogan’s syndrome with aortitis, aortic regurgitation, and aortic
arch vessel stenoses. Ann Thorac Surg 1991;52:1166–7.
[31] Ochonisky S, Chosidow O, Kuentz M, et al. Cogan’s syndrome. An unusual etiology of ur-
ticarial vasculitis. Dermatologica 1991;183(3):218–20.
[32] Eisenstein B, Taubenhaus M. Nonsyphilitic interstitial keratitis and bilateral deafness
(Cogan’s syndrome) associated with cardiovascular disease. N Engl J Med 1958;258:1074–9.
[33] Fisher ER, Hellstrom HR. Cogan’s syndrome and systemic vascular disease. Analysis of
pathologic features with reference to its relationship to thromboangiitis obliterans (Buerger).
Arch Pathol 1961;72:572–92.
[34] Albayram MS, Wityk R, Yousem DM, et al. The cerebral angiographic findings in Cogan
syndrome. AJNR Am J Neuroradiol 2001;22:751–4.
[35] Pinals RS. Cogan’s syndrome with arthritis and aortic insufficiency. J Rheumatol 1978;5(3):

[36] Benitez JT, Arsenault MD, Licht JM, et al. Evidence of central vestibulo-auditory dysfunc-
tion in atypical Cogan’s syndrome: a case report. Am J Otol 1990;11:131–4.
[37] Majoor MH, Albers FW, Casselman JW. Clinical relevance of magnetic resonance imaging
and computed tomography in Cogan’s syndrome. Acta Otolaryngol 1993;113:625–31.
[38] Casselman JW, Majoor MH, Albers FW. MR of the inner ear in patients with Cogan
syndrome. AJNR Am J Neuroradiol 1994;15:131–8.
[39] Mazlumzadeh M, Lowe VJ, Mullan BP, et al. The utility of positron emission tomography in
the evaluation of autoimmune hearing loss. Otol Neurotol 2003;24:201–4.
[40] Matteson EL, Choi HK, Poe DS, et al. Etanercept therapy for immune-mediated cochleoves-
tibular disorders: a multi-center, open-label, pilot study. Arthritis Rheum 2005;53(3):
[41] Shimura M, Yasuda K, Fuse N, et al. Effective treatment with topical cyclosporine A of a pa-
tient with Cogan syndrome. Ophthalmologica 2000;214(6):429–32.
[42] Haynes BF, Pikus A, Kaiser-Kupfer M, et al. Successful treatment of sudden hearing loss in
Cogan’s syndrome with glucocorticoids. Arthritis Rheum 1981;24:501–3.
[43] Richardson B. Methotrexate therapy for hearing loss in Cogan’s syndrome. Arthritis Rheum
[44] Riente L, Taglione E, Berrettini S. Efficacy of methotrexate in Cogan’s syndrome. J Rheu-
matol 1996;23:1830–1.
[45] Matteson EL, Fabry DA, Facer GW, et al. Open trial of methotrexate as treatment for
autoimmune hearing loss. Arthritis Rheum 2001;45(2):146–50.
[46] Harris JP, Weisman MH, Derebery JM, et al. Treatment of glucocorticoid-responsive
autoimmune inner ear disease with methotrexate: a randomized controlled trial. JAMA
[47] Vaiopoulos G, Sfikakis PP, Skoumas B, et al. Lack of response to glucocorticoids and pulse
cyclophosphamide therapy in Cogan’s syndrome. Clin Rheumatol 1994;13:110–2.
[48] Quaranta N, Bartoli R, Giagnotti F, et al. Cochlear implants in systemic autoimmune
vasculitis syndromes. Acta Otolaryngol Suppl 2002;548:44–8.
[49] Pasanisi E, Vincenti V, Bacciu A, et al. Cochlear implantation and Cogan syndrome. Otol
Neurotol 2003;24:601–4.