Alternatives to Autogenous Bone Graft:

Efficacy and Indications

André R. Gazdag, MD, Joseph M. Lane, MD,
David Glaser, MD, and Robert A. Forster, MD

Abstract

Bone grafting is frequently used to augment bone healing with the numerous lining have osteogenic potential; and
approaches to reconstructing or replacing skeletal defects. Autologous cancellous (3) the bone graft and the adjacent
bone graft remains the most effective grafting material because it provides the clot contain a family of growth fac-
three elements required for bone regeneration: osteoconduction, osteoinduction, tors, most notably bone mor-
and osteogenic cells. Autologous cortical bone graft provides these three compo- phogenic protein (BMP) and
nents to a limited extent as well and also provides the structural integrity impor- transforming growth factor-beta,
tant in reconstruction of larger defects. However, because autogenous grafting is which have the ability to induce the
associated with several shortcomings and complications, including limited quan- regenerative process as well as to
tities of bone for harvest and donor-site morbidity, alternatives have been used in augment the process to completion.
a wide range of orthopaedic pathologic conditions. Grafting substitutes currently Autologous cortical bone provides
available include cancellous and cortical allograft bone, ceramics, demineralized these elements to a more limited
bone matrix, bone marrow, and composite grafts. No single alternative graft extent, but its structure confers
material provides all three components for bone regeneration. The clinical appli- strength when needed to fill larger
cations for each type of material are dictated by its particular structural and bio- defects. Any alternative to autoge-
chemical properties. Composite grafts consisting of several materials are often nous bone graft should be judged in
used to maximize bone healing, especially where the grafting site is compromised. terms of its ability to provide these
J Am Acad Orthop Surg 1995;3:1-8

Dr. Gazdag is a Resident in the Department of

Bone grafting is commonly used to
augment bone healing in the surgical
treatment of a broad spectrum of
musculoskeletal disorders.1 Bone
grafts have been used to reconstruct
or replace skeletal defects, to aug-
ment fracture repair, to strengthen
arthrodeses, and to fill defects after
the treatment of bone tumors. For
over 100 years, autologous cancel-
lous bone grafting has been the stan-
dard of care.
Autogenous Grafts
Autogenous grafts can be cancel-
lous, nonvascularized cortical, or
vascularized cortical; each type has
different biologic activities (Table 1).
Ideally, graft substitutes should pro-
vide four elements: an osteoconduc-
tive matrix, which is a nonviable
scaffolding conducive to bone
ingrowth; osteoinductive factors,
which are the chemical agents that
induce the various stages of bone
regeneration and repair; osteogenic
cells, which have the potential to dif-
ferentiate and facilitate the various
stages of bone regeneration; and
structural integrity.
Autogenous cancellous bone
graft contains three of these compo-
nents: (1) the hydroxyapatite and
collagen are well suited to serve as
an osteoconductive framework; (2)
numerous stromal cells within the
Orthopaedics, University of California, Los
Angeles, School of Medicine. Dr. Lane is Profes-
sor and Chairman, Department of Orthopaedics,
UCLA School of Medicine. Dr. Glaser is a Resi-
dent in the Department of Orthopaedic Surgery,
Hospital of the University of Pennsylvania,
Philadelphia. Dr. Forster is a Resident in the
Department of Emergency Medicine, Carolinas
Medical Center, Charlotte, NC.
Reprint requests: Dr. Lane, Department of
Orthopaedic Surgery, UCLA School of Medicine,
10833 LeConte Avenue, Los Angeles, CA
90024-6902.
One or more of the authors or the departments
with which they are affiliated have received some-
thing of value from a commercial or other party
related directly or indirectly to the subject of this
article.
Copyright 1995 by the American Academy of
Orthopaedic Surgeons.

Vol 3, No 1, Jan/Feb 1995 1

Alternatives to Autogenous Bone Graft
Table 1
Properties of Types of Autologous Bone Grafts
Property Cancellous
Osteoconduction ++++
Osteoinduction ++
Osteoprogenitor cells +++
Immediate strength - +++
Strength at 6 months --
Strength at 1 year ---
three components, as well as struc-
tural integrity when applicable.
Bone regeneration initiated by
autogenous cancellous bone occurs
in three major steps. First, the undif-
ferentiated osteoprogenitor cells are
recruited. Then, by osteoinduction,
these cells differentiate to give rise to
osteoblasts and chondrocytes.
Finally, a suitable scaffold on which
active osteoprogenitor cells can pro-
duce new bone is established.
Osteoinduction is mediated by
numerous growth factors provided
by the bone matrix itself. The most
notable group are the BMPs, which
are low-molecular-weight proteins
that initiate endochondral bone for-
mation, presumably by stimulating
local progenitor cells of osteoblast
lineage and by enhancing bone col-
lagen synthesis. Transforming
growth factor-beta, shown to be
closely related to BMP by sequence
homology, is synthesized in many
tissues, including bone, and appears
to stimulate bone formation simi-
larly. Others include fibroblast
growth factors, which are angio-
genic factors important in neovascu-
larization and wound healing, and
platelet-derived growth factor,
which acts as a local tissue-growth
regulator. Platelet-derived growth
factor was initially isolated in blood
platelets, underscoring one of
the important roles of the clot
(hematoma) after a fracture or graft-
ing, but recently other tissues,
2 Journal of the American Academy of Orthopaedic Surgeons
Nonvascularized Vascularized
Cortical Cortical
+ +
+/- +/-
- +
+++
++ +++
+++ ++++
including bone, have been shown to
synthesize it as well. Insulinlike
growth factors and microglobulin-
beta are other examples of bone-
matrix-synthesized growth factors.2
Cancellous bone graft starts with
no structural integrity, but this
rapidly changes due to bone aug-
mentation and union (osteointegra-
tion) with preexisting osseous
structures. The bone strength
increases as the bone mass accumu-
lates, and the construct is remodeled
along the lines of stress. The con-
verse occurs with cortical bone. The
graft initially conveys structural
strength as it undergoes osteointe-
gration at its ends. It then undergoes
a remodeling phase, in which the
nonviable bone is removed by osteo-
clast tunneling and resorption. Dur-
ing this resorptive phase, which can
last from 6 to 18 months, the bone
can lose up to one third of its
strength, as demonstrated by Ennek-
ing et al. 3 The cortical bone will
retain significant islands of nonre-
placed nonviable bone throughout
the life of the individual. The major
advantage of cortical bone over can-
cellous bone is that it offers initial
structure and can provide compres-
sive strength to the graft construct.
Free vascularized cortical grafts
most commonly involve the fibula,
although other bones, such as the
ribs and the iliac crest, have been
used. In this process, the bone does
not undergo significant cell necrosis
and remains viable through its arte-
rial and venous anastomoses, which
avoids some of the problems of non-
vascularized cortical bone. Biome-
chanical studies have demonstrated
that it is superior to nonvascularized
cortical graft for approximately 6
months, after which time no differ-
ence can be demonstrated as mea-
sured by torque, bending, and
tension studies. The disadvantages
of vascularized grafts include
donor-site morbidity, which is
minor in most cases; prolonged
operating time; and greater utiliza-
tion of resources. Vascularized
grafts are clearly superior to non-
vascularized cortical grafts when the
bridging area is more than 12 cm.
Reported stress-fracture rates for
this distance in nonvascularized cor-
tical bones approach 50%, while the
rate of fracture for vascularized
grafts is less than 25%. The vascu-
larized graft also has a greater ability
to heal stress-related fractures and to
enhance its girth.
The advantages of autogenous
cancellous/cortical bone grafts are
that they are histocompatible, do not
transfer disease, and retain viable
osteoblasts that participate in the
formation of new bone. Although
autogenous bone grafting is effec-
tive, it is associated with several
shortcomings and potential compli-
cations. Its disadvantages are that a
limited quantity of bone is available
for harvest and there is significant
donor-site morbidity (rates as high
as 25%),4 including infections and
pain, increased anesthesia time, and
significantly increased operative
blood loss.
Alternatives to autogenous bone
graft have been sought in an effort to
increase the quantity of bone
obtained and decrease the morbidity
of the grafting process. The ensuing
discussion will address currently
available grafting substitutes that
have been approved by the Food
and Drug Administration (FDA),
 André R. Gazdag, MD, et al

including allograft, ceramic, de-
mineralized bone matrix, bone mar-
row, and composite grafts (Table 2).
Allografts
Allografts are available fresh,
frozen, or freeze-dried. With fresh
allografts, no preservation is
required. However, the speed with
which the grafting transfers need to
be performed leaves little time to test
for disease or sterility. Fresh allo-
graft evokes an intense immune
response, making it clearly inferior
to autografts. It is not currently a
mainstay in grafting, and its applica-
tions are limited to joint resurfacing.
Most allografts are either frozen or
freeze-dried. Frozen allografts are
maintained at temperatures below
–60°C to diminish degradation by
enzymes, affording decreased
immunogenicity without changes in
biomechanical properties. Osteo-
chondral allografts undergo a much
more controlled slow freeze with use
of a cryopreservative (glycerol or
dimethylsulfoxide) for the cartilage.
There is controversy regarding the
viability of frozen cartilage, as stud-
ies have demonstrated viability val-
ues ranging from 20% to 70% when
these preservation techniques were
used. 5 Freeze-drying (lyophiliza-
tion) involves the removal of water
from the frozen tissue, after which
the tissues are vacuum-packed and
stored at room temperature for up to
5 years. These methods decrease
antigenicity even further, produce
almost no biochemical changes, and
do not affect the limited osteoin-
ductive properties. These grafts
undergo biomechanical alteration,
however, with loss of hoop strength
and compressive strength on rehy-
dration. In all these techniques, the
osteoprogenitor cells are destroyed,
the osteoconductive properties are
largely retained in terms of their can-
cellous and cortical structure, and
the deeply bound, limited osteoin-
ductive material present in the graft
may be only partially retained.
Allografts can be used for non-
structural purposes, such as recon-
structing defects after curettage of
benign neoplasms and periarticular
bone cysts at the time of arthro-
plasty. Morcellation of cancellous
and cortical chips can be carried out
for this purpose. Some clinicians
have recommended mixing allograft
bone with autogenous tissue to
enhance osteoinduction and/or
mixing it with autogenous bone
marrow to introduce osteoprogeni-
tor cells. Data on the efficacy of
these processes are not yet available.
The structural roles of allografts
include use as an intercalary seg-
ment to reconstruct a diaphyseal
defect of long bone and use in
arthrodeses about the ankle, hip, cer-
vical spine, and lumbar spine. Large
segments can be modeled to replace
acetabular, femoral, and tibial
defects. Osteochondral allografts
have also been used for the dual pur-
pose of replacing resected bone and
providing a biologic joint surface.
Various allograft structures are
available, including iliac bicortical
and tricortical strips, patellar tricorti-
cal strips, cancellous cortical dowels,
fibular shafts and wedges, femoral
cross sections, and ribs. Structures
limited to frozen preservation due to
size include whole or partial tibia,
humerus, femur, talus, acetabulum,
ilium, and hemipelvis. Complica-
tions of the structural use of large
allografts include nonunion (10% of
cases), fractures (5% to 15%), and
infection (10% to 15%).6 Morcellated

Table 2
Properties of Bone-Graft Alternatives

Immediate
Grafting Osteo- Osteo- Osteoprogenitor Immuno- Donor-Site Torque
Material conduction induction Cells genicity Morbidity Strength

Cancellous autologous
graft
Cortical autologous
graft
Fresh allograft
Frozen allograft
Freeze-dried allograft
Ceramics
Demineralized bone
matrix
Bone marrow
Particulate ceramic
with bone marrow
++++ ++
+ +/-
+ +/-
+ +/-
+ +/-
+ -
+ ++
- +/-
++ +/-
+++ - + -
+/- - + ++
- ++ - ++
- + - ++
- +/- - +
- - - +/-
- - - -
++ - - -
++ - - -

Vol 3, No 1, Jan/Feb 1995 3

Alternatives to Autogenous Bone Graft
allograft lacks the osteoinduction
potential and osteoprogenitor cells of
autologous bone graft and has been
used largely as a filler or extender of
graft, except in individuals who have
a very high potential for bone regen-
eration (e.g., children), where the
grafts are being used without autolo-
gous augmentation.
One of the main concerns with
use of allograft bone is transmission
of infection, most notably hepatitis
and acquired immunodeficiency
syndrome (AIDS). Since 1976, most
tissue banks in the United States
have been represented by the Amer-
ican Association of Tissue Banks
(AATB), which evaluates members
for compliance with a comprehen-
sive set of standards. On December
14, 1993, the FDA mandated that
every national tissue bank must
comply with governmental regula-
tions that essentially parallel the
AATB comprehensive screening
standards. These regulations in-
clude donor screening, repeated
infectious disease testing, labeling
requirements, long-term tracking of
the graft, and inspections of facili-
ties. Many local in-hospital bone
banks at first had difficulty in fulfill-
ing these obligations, but they are
now required by law to comply.
Strict donor-screening and tissue-
testing techniques have significantly
lowered the risk of disease transmis-
sion. The AATB records indicate
that, of the 3 million tissue trans-
plants performed since the identifica-
tion of the human immunodeficiency
virus, only two donors’ tissues have
been linked with documented trans-
mission of the AIDS virus.7 Both
cases involved transplantation of
unprocessed, fresh-frozen allografts.
Attempts at sterilization of allografts
have compromised the tissue. Ethyl-
ene oxide and radiation alter some of
the structural properties as well as
the biochemical properties of the
graft. In one well-documented
instance of a donor with AIDS, grafts
4 Journal of the American Academy of Orthopaedic Surgeons
that were lyophilized and irradiated
did not give rise to AIDS, while fresh-
frozen grafts did. Thus, there is a sug-
gestion that such processing of bone
may destroy the AIDS virus.
Ceramics
Ceramics have been utilized solely as
osteoconductive bone-graft matrices.
Most calcium phosphate ceramics
currently under investigation are syn-
thetic and are composed of hydroxy-
apatite (Ca10[PO4]6[OH]2), tricalcium
phosphate (TCP) (Ca3[PO4]2), or com-
binations of the two. These biomate-
rials are produced commercially as
porous implants, nonporous dense
implants, and granular particles
with pores. Most calcium phosphate
ceramics are created with the use of
a high-temperature process called
sintering along with high-pressure
compaction techniques.
The chemical composition of the
ceramic profoundly affects its rate of
bioresorption. Studies indicate that
TCP, which is more porous than
hydroxyapatite, undergoes biologic
degradation 10 to 20 times faster
than hydroxyapatite.8 In clinical tri-
als, TCP was totally resorbed in
some circumstances but lasted a
number of years in others.9,10 Once
in the body, TCP is partially con-
verted to hydroxyapatite, which is
degraded slowly. The resorbing cell
for hydroxyapatite is the foreign-
body giant cell (not the osteoclast),
which stops after resorbing 2 to 10
µm of hydroxyapatite. Thus, large
segments of hydroxyapatite will
remain in place in the body for peri-
ods of up to 10 years.
In clinical applications, TCP is
remodeled better than hydroxyapatite
due to its porosity, but it is mechani-
cally weaker because it is resorbed so
quickly. Therefore, it is not ideal in
compression, unlike hydroxyapatite.
The combination of the two is used
clinically to offer both advantages.
Material factors such as the surface
area affect biologic degradation; in
general, the larger the surface area, the
greater the bioresorption. Dense
ceramic blocks with small surface
areas biodegrade more slowly than
porous implants. Thus, the shape and
architecture of the ceramic have a pro-
found effect on its resorption rate.
Ceramics are brittle and have very
little tensile strength. Use of ceram-
ics in applications requiring signifi-
cant impact, torsional, bending, or
shear stress seems impractical at pres-
ent. However, the mechanical prop-
erties of porous calcium phosphate
materials are comparable to those of
cancellous bone once they have been
incorporated and remodeled.
Ceramics must be shielded from
loading forces until bone ingrowth
has occurred. Rigid stabilization of
surrounding bone and non-weight-
bearing are required during this
period because the ceramics them-
selves tolerate minimal bending and
torque loads before failing.
The optimal osteoconductive pore
size for ceramics appears to be
between 150 and 500 µm. Ceramics
appear to have no early adverse
effects such as inflammation, and for-
eign-body responses to ceramics are
practically nonexistent when they are
in a structural arrangement.9-11 How-
ever, small granules of material have
been shown to elicit a foreign
body–giant cell reaction (at least in
the rodent) and partial resorption.
When ceramics are used, radio-
graphs demonstrate continued
presence of the ceramic for a pro-
longed period of time due to the fail-
ure of complete remodeling. A
persistently dense radiographic
appearance creates difficulty in
determining the degree of bone
ingrowth and incorporation into the
implant.12 Tricalcium phosphate,
which is more biodegradable, loses
more of its radiodensity and appears
to be more incorporated into the
bone.

The replamineform ceramics are
porous hydroxyapatite materials
derived from the calcium carbonate
skeletal structure of sea coral. They
are produced from a marine coral
specimen using a hydrothermal
exchange method that replaces the
original carbonate of the coral with
calcium phosphate replicas.13 In con-
trast to the random pore structure cre-
ated in totally synthetic porous
materials, the pore structure of the
coralline calcium phosphate implants
is highly organized and is similar to
that of human cancellous bone (Fig.
1). The pore size of these materials is
determined by the genus of the coral
used. Coralline hydroxyapatite
derived from the genus Gonipora has
large pores measuring from 500 to 600
µm in diameter, with interconnec-
tions measuring 220 to 260 µm.9,12
The coral genus Porites has a
microstructure that appears similar to
that of interstitial cortical bone, with
its smaller pore diameter of 200 to 250
µm, its parallel channels intercon-
nected by 190-µm fenestrations, and
its porosity of 66%. 9,12 Coralline
hydroxyapatites are available com-
mercially as Pro Osteon Implant 500
and Pro Osteon Implant 200 (Inter-
pore Orthopaedics, Irvine, Calif),
with average pore sizes of 500 and 200
µm, respectively.
Experimental animal studies have
consistently demonstrated the supe-
A B
Fig. 1 The porous structure and composition of coralline hydroxyapatite (A) and human
cancellous bone (B) are very similar. (Courtesy of Interpore Orthopaedics, Inc, Irvine, Calif.)
Vol 3, No 1, Jan/Feb 1995
rior performance of autologous bone
grafts when compared with ceramic
implants alone.8,9,11 However, some
studies have yielded promising
results when certain specific condi-
tions were met. Coralline hydroxy-
apatite performed quite favorably as
a defect filler in proximal tibial
defects in dogs when compared with
corticocancellous autogenous bone.14
Clinically, the first successful
results were reported in dentistry
and reconstructive craniofacial
surgery. In orthopaedics, Bucholz
et al15 demonstrated a similar effi-
cacy between calcium phosphate
ceramics and autogenous grafts for
certain applications, particularly to
fill defects under tibial plateau frac-
tures where the material was under
compression. In studies comparing
the use of coralline hydroxyapatite
and cancellous bone in the tibia
(including the tibial plateau), they
reported no difference in functional
outcome, and on histologic analysis,
the hydroxyapatite implants re-
vealed bone ingrowth with both
cortical and cancellous bone in
appropriate locations. Bucholz et
al9 also studied TCP and found it
comparable to autogenous bone for
filling defects secondary to trauma,
benign tumors, and cysts. Studies
performed by other individuals,
including Altermatt et al, 10 have
indicated that granular hydroxy-
André R. Gazdag, MD, et al
apatite and TCP, particularly when
used in bone defects, can be quite
efficacious.
An advantageous property of
ceramics when used as a filler to
restore volume in cavities is that the
osteoconductive hydroxyapatite
bonds well to bone. Ceramics alone
do not have osteoinductive proper-
ties. However, there is some sug-
gestion that hydroxyapatite has
significant chemical affinity for local
growth factors that act in the regen-
eration process. Ohgushi et al 16
found that ceramics can be filled
with bone marrow prior to use, at
least in animal studies, and that bone
marrow grows well within ceramics
and results in a composite. This has
not been attempted in humans, how-
ever. It can be concluded that ceram-
ics can serve as a bone graft
expander and/or filler, particularly
in compressive applications. Be-
cause a ceramic material is brittle
and has no initial hoop strength or
shear strength, the bone must be
protected while the ceramic is incor-
porated. It does not stimulate new
bone formation and so is not as
attractive for “jump-starting” the
healing process in the treatment of
nonunions.
Demineralized Bone
Matrix
Demineralized bone matrix (DBM)
is formed by means of acid extrac-
tion of bone, which leaves noncol-
lagenous proteins, bone growth
factors, and collagen in continuity in
a composite. Demineralized bone
matrix is prepared by bone banks,
and a chemically processed form is
produced commercially under the
name Grafton Allogenic Bone
Matrix (Osteotech, Shrewsbury, NJ).
Currently, DBM is available freeze-
dried and is processed from cortical
or corticocancellous bone as a pow-
der, as crushed granules, and as
5
Alternatives to Autogenous Bone Graft

chips. Grafton is also available as a

gel; it is packaged in a syringe, from
which it can be applied directly
intraoperatively. All four forms are
easy to mold intraoperatively.
Demineralized bone matrix has
been utilized to promote bone
regeneration, mainly within well-
supported, stable skeletal defects.
The results in clinical trials have
been excellent. 17 The enhanced
osteoinductive capability of DBM is
afforded most notably by BMP,
although the amount of BMP within
demineralized grafts is far lower
than in recombinant BMP studies.
The FDA requires sterilization of the
DBM prepared by bone banks, 7
which may decrease the viability of
the available BMP. Grafton is
A B C
processed from human bone by
means of a patented technique that Fig. 2 Radiographs of a grade II comminuted left femoral fracture treated with a supra-
incorporates a permeation treat- condylar plate and a three-part composite graft consisting of demineralized allograft bone
gel (Grafton), demineralized allograft chips, and autogenous bone marrow. A, Preoperative
ment that does not expose tissue to view. B, Radiograph obtained immediately postoperatively shows fixation and three-part
ethylene oxide or gamma radiation, composite graft in place. C, Good bone formation bridging two segments at 8 months.
which may protect more of the BMP. (Courtesy of Paul G. Kleinman, MD, East Meadow, NY.)
Although DBM offers no structural
strength, it has proved useful in facil-
itating the development of bone that
is comparable in mechanical strength cient grafting bed. A number of genitor cells could be easily increased
to autograft. It has been most suc- studies by Connelly and Healey in number and concentration.
cessfully used in conjunction with have demonstrated that bone
internal fixation (Fig. 2) and as an marrow can successfully treat
adjunct to other grafting materials. Its nonunions when provided in ade- Composite Grafts
applications include augmentation of quate amounts.
autogenous and traditional allograft Bone marrow should be har- The desire to incorporate the favor-
bone grafts in repairing cysts, frac- vested in aliquots of approximately able properties of different materials
tures, nonunions, and stable fusions. 2.5 to 3 ml per site (cancellous bone into a single graft compound has led
from the proximal humerus or, to the proliferation of various com-
preferably, the ilium). The marrow posite grafts. Composite grafts can
Bone Marrow is then diluted with blood, after be defined as any combination of
which it should be used immedi- materials that includes both an
Bone marrow contains osteoprogen- ately to maintain its viability. It has osteoconductive matrix and an
itor cells on the order of 1 per 50,000 had only limited reported clinical osteogenic or osteoinductive mater-
nucleated cells, and certain tech- use; however, it does offer the ability ial. For example, composites of TCP
niques have increased that number to augment all the other synthetic and BMP are currently being used in
fivefold. Burwell18 and Salama and grafts and allografts that are cur- craniofacial reconstruction.20 The
Weissman19 have utilized bone mar- rently more widely used, as well as ceramic maintains soft-tissue posi-
row, either by itself or in combina- to reestablish a more normal fracture tion and provides an osteoconduc-
tion with an inorganic matrix for milieu after extensive irrigation. tive matrix, and the proteins
clinical application. It can grow into There is essentially no morbidity stimulate osteoinduction.
ceramics and can be used to bring from obtaining bone marrow, but it Collagraft (Zimmer, Warsaw, Ind;
osteoprogenitor cells back to a defi- would be desirable if the osteopro- and Collagen Corporation, Palo Alto,

6 Journal of the American Academy of Orthopaedic Surgeons


Calif) is a commercially prepared
composite consisting of suspended
deantigenated bovine fibrillar colla-
gen and porous calcium phosphate
ceramic, of which 65% is hydroxyap-
atite and 35% is TCP. The collagen is
purified from bovine dermis and is
95% type I and 5% type III. Calcium
phosphate consists of granules with
70% porosity and a pore diameter
ranging from 500 to 1,000 µm. The
mixture is nonosteoinductive; the
addition of autogenous bone mar-
row provides osteoprogenitor cells
and a limited amount of growth fac-
tors, such as platelet-derived growth
factor and transforming growth fac-
tor-beta within the bone marrow
clot.
In a prospective, randomized
multicenter trial, Cornell et al 21
compared a composite graft consist-
ing of Collagraft plus autogenous
marrow with a cancellous iliac bone
graft in acute long-bone fractures
and found no significant differences
in functional result or radiographic
appearance. The use of Collagraft
significantly shortened operative
time and avoided the complications
and morbidity of autograft harvest-
ing. This study, however, did not
include a control group treated with-
out grafting, and additional trials
against such controls are needed.
Collagraft is currently available
only as a paste or in soft strips and
therefore provides no structural
strength. In addition, it has a ten-
dency to flow if there is continued
bleeding at the site of the fracture.
The material must be carefully main-
tained in the location of use until the
clot has formed. Biopsy specimens
from patients in whom Collagraft was
used have demonstrated some slight
inflammation at the site of the gran-
ules, but there were no infections in
over 139 patients treated with the
material, compared with five infec-
tions in 128 patients treated with auto-
genous bone graft.21 This appears to
be a material that can be used as a
Vol 3, No 1, Jan/Feb 1995
bone-graft expander or as a graft sub-
stitute for stabilized fractures that are
protected by internal fixation but
require grafting due to extensive com-
minution or segmental bone loss.
Pharmaceuticals, such as antibiotics
and antineoplastic agents, can also be
combined with Collagraft to create a
delivery system that would treat bone
disorders locally. Its use is con-
traindicated in intra-articular frac-
tures because of potential migration
of granules into the joint.
Summary
A number of grafting materials are
available as alternatives to autoge-
nous bone graft for a wide range of
clinical applications (Table 3). Allo-
grafts can provide structure and
osteoconduction; however, they offer
limited osteoinduction and no osteo-
progenitor cells. Their indications are
similar to those of autologous bone,
including repair of nonunions, pro-
motion of arthrodesis, and segmental
replacement of long bones. However,
if the grafting bed is unfavorable (e.g.,
after infection or if there is poor soft-
tissue coverage), the allograft bone
must be augmented with either auto-
graft or another graft substitute that
provides growth factors and osteo-
progenitor cells. Allograft alone
would be contraindicated in treating a
4-cm humeral defect that developed
from an infected nonunion. Concerns
regarding allografts include fracture,
osteointegration, transmission of dis-
ease, and infection.
Ceramics, available in powders,
granules, and blocks, are excellent in
compression and confer critical struc-
tural support. However, they are brit-
tle and have little strength in bending,
shear, and tension until incorporated
into the existing adjacent bone.
Because ceramics are exclusively
osteoconductive, they are contraindi-
cated for use by themselves. They
must be combined with autograft or
André R. Gazdag, MD, et al
have access to a rich bone marrow,
but they are effective graft fillers or
expanders when patching defects
after tumor resection or in a
depressed tibial plateau fracture.
Demineralized bone matrix is a
limited source of BMP and can be
used as an adjunct in the regeneration
process. Despite its osteoconductive
Table 3
Clinical Applications of
Bioalternative Grafts*
Reconstruct diaphyseal defect
(6 cm)
1. Vascularized cortical autograft
2. Nonvascularized cortical
autograft
3. Frozen cortical allograft†
4. Freeze-dried allograft
Augmentation of autologous bone
graft
1. Bone marrow mixed with
either ceramics or allograft
2. Ceramics or morcellated
allograft
3. Demineralized bone matrix or
bone marrow
Expander to fill defects or cavities
1. Autograft or ceramics mixed
with bone marrow
2. Ceramics or morcellated
allograft alone
3. Demineralized bone matrix or
bone marrow
Reconstruct short defects (2-4 cm)
or perform arthrodesis of the
cervical spine
1. Autograft
2. Frozen cortical allograft‡
3. Freeze-dried allograft
4. Ceramics
Nonunion
1. Autograft
2. Demineralized bone matrix
3. Bone marrow
4. Morcellated allograft
* Options are listed in order of efficacy,
with the first being considered the
most efficacious.
†
Much, much preferable to the follow-
ing option.
‡
Much preferable to the following op-
tion.
7
Alternatives to Autogenous Bone Graft
potential, DBM provides no immedi-
ate torque or compressive strength;
thus, its use as the sole material
would be contraindicated when
grafting large cortical segmental
defects. Its clinical applications
include augmentation of autogenous
and allograft bone for repairing frac-
tures, packing cysts, and promoting
arthrodesis, and it can be used in both
posterolateral lumbar fusions and
hip fusions with instrumentation.
Bone marrow is best used as an
adjunct to existing allograft or
biosynthetic ceramics to provide
osteoprogenitor cells to compro-
mised grafting beds. Because it pro-
vides no structural strength and is
References
1. Friedlaender GE, Goldberg VM (eds):
Bone and Cartilage Allografts: Biology and
Clinical Applications. Park Ridge, Ill:
American Academy of Orthopaedic
Surgeons, 1991.
2. Canalis E: Regulation of bone remodel-
ing, in Favus MJ (ed): Primer on the
Metabolic Bone Diseases and Disorders of
Mineral Metabolism. Kelseyville, Calif:
American Society for Bone and Mineral
Research, 1990, pp 24-25.
3. Enneking WF, Burchardt H, Puhl JJ, et
al: Physical and biological aspects of
repair in dog cortical-bone transplants.
J Bone Joint Surg Am 1975;57:237-252.
4. Summers BN, Eisenstein SM: Donor site
pain from the ilium. J Bone Joint Surg Br
1989;71:677-680.
5. Schachar NS, Henry WB Jr, Wadsworth
P, et al: Fate of massive osteochondral
allografts in a feline model, in Fried-
laender GE, Mankin HJ, Sell KW (eds):
Osteochondral Allografts: Biology, Bank-
ing, and Clinical Applications. Boston:
Little Brown, 1981, pp 81-101.
6. Sim FH, Frassica FJ: Use of allografts
following resection of tumors of the
musculoskeletal system. Instr Course
Lect 1993;42:405-413.
8 Journal of the American Academy of Orthopaedic Surgeons
available only in small amounts, it
should never be used alone in
attempts to fill gaps or span segmen-
tal defects. It is strictly an adjunct to
other grafts and works well to jump-
start the healing of nonunions.
Composite grafts consisting of
ceramics, collagen, and bone mar-
row have been used successfully,
but since they are in a form without
structure, they must be protected
until they have been osteointe-
grated. They have a role in aug-
menting limited autogenous bone
graft.
Bone morphogenic protein is not
currently available clinically in a
highly purified or recombinant
7. AATB Information Alert, vol 3, no. 6.
McLean, Va: American Association of
Tissue Banks, Dec 15, 1993.
8. Jarcho M: Calcium phosphate ceramics
as hard tissue prosthetics. Clin Orthop
1981;157:259-278.
9. Bucholz RW, Carlton A, Holmes RE:
Hydroxyapatite and tricalcium phos-
phate bone graft substitutes. Orthop
Clin North Am 1987;18(2):323-334.
10. Altermatt S, Schwöbel M, Pochon JP:
Operative treatment of solitary bone
cysts with tricalcium phosphate
ceramic: A 1 to 7 year follow-up. Eur J
Pediatr Surg 1992;2:180-182.
11. Moore DC, Chapman MW, Manske D:
The evaluation of a biphasic calcium
phosphate ceramic for use in grafting
long-bone diaphyseal defects. J Orthop
Res 1987;5:356-365.
12. Sartoris DJ, Holmes RE, Resnick D:
Coralline hydroxyapatite bone graft
substitutes: Radiographic evaluation. J
Foot Surg 1992;31:301-313.
13. Light M, Kanat IO: The possible use of
coralline hydroxyapatite as a bone
implant. J Foot Surg 1991;30:472-476.
14. Holmes RE, Bucholz RW, Mooney V:
Porous hydroxyapatite as a bone graft
form. The closest alternative is
DBM, which is readily available
from bone banks. Recombinant
BMP is still in clinical trials, but it is
anticipated that it will be more easily
accessible to the orthopaedic sur-
geon in the near future.
When the grafting site is compro-
mised and all three components of
osteoconduction, osteoinduction, and
osteoprogenitor cells are required,
autogenous bone graft is probably
superior. However, a composite of
particulate ceramic, bone marrow,
and DBM that incorporates all three
regenerative components may be just
as effective. Clinical trials are needed
to further define relative efficacy.
substitute in diaphyseal defects: A histo-
metric study. J Orthop Res 1987;5:114-121.
15. Bucholz RW, Carlton A, Holmes R:
Interporous hydroxyapatite as a bone
graft substitute in tibial plateau frac-
tures. Clin Orthop 1989;240:53-62.
16. Ohgushi H, Goldberg VM, Caplan AI:
Heterotopic osteogenesis in porous
ceramics induced by marrow cells. J
Orthop Res 1989;7:568-578.
17. Pals SD, Wilkins RM: Giant cell tumor
treated by curettage, cementation, and
bone grafting. Orthopedics 1992;15:703-
708.
18. Burwell RG: The function of bone mar-
row in the incorporation of a bone graft.
Clin Orthop 1985;200:125-141.
19. Salama R, Weissman SL: The clinical
use of combined xenografts of bone and
autologous red marrow: A preliminary
report. J Bone Joint Surg Br 1978;60:111-
115.
20. Desilets CP, Marden LJ, Patterson AL, et
al: Development of synthetic bone-repair
materials for craniofacial reconstruction.
J Craniofac Surg 1990;1:150-153.
21. Cornell CN, Lane JM, Chapman M, et al:
Multicenter trial of Collagraft as bone graft
substitute. J Orthop Trauma 1991;5:1-8
 Isolated Medial Collateral Ligament Injuries in the Knee
Peter A. Indelicato, MD

Abstract
The management of ligament injuries in the knee has progressed significantly over oblique ligament) as the “semimem-
the past two decades as a result of both laboratory and clinical studies that better branosus corner.” According to him,
define the healing capacity of these supporting structures. The intracapsular lig- the semimembranosus muscle,
aments (the anterior and posterior cruciate ligaments) appear to have limited through its various attachment sites,
potential for spontaneous healing and frequently require surgical reconstruction. “dynamizes” this area of the knee
The extracapsular ligaments (the medial and lateral collateral ligaments), how- during active contracture and pro-
ever, appear to have a fairly robust potential for healing. As a result, the need for vides support even with the knee in
surgical intervention is limited to specific clinical situations. flexion. In addition, he states that
J Am Acad Orthop Surg 1995;3:9-14 surgical repair of damage to this area
is essential to eliminate any valgus
laxity that exists with the knee in full
extension. Finally, he points out the
The management of ligament injuries insertion. Immediately deep to the anatomic connection between the
to the knee continues to evolve. Bet- MCL is the medial capsular ligament. vastus medialis muscle and the MCL
ter understanding of the biomechan- Posterior to the MCL is a thickening and states that the MCL is also
ical function and physiologic healing of the capsular ligament specifically dynamized by contraction of the vas-
process of these important structures referred to as the posterior oblique tus medialis during active extension
affords the clinician a greater oppor- ligament (Fig. 1, A). (Fig. 1, B).
tunity to secure a successful outcome The main medial stabilizer that Grood et al,1 Warren et al,2 and
when major damage to the cruciate resists valgus loading is the MCL.1,2 Müller3 have pointed out the contri-
and collateral ligaments of the knee Grood et al1 have shown that the bution of the MCL and the posterior
is encountered. MCL contributes 78% to the restrain- oblique ligament in resisting abnor-
In this article I will review the ing force on the medial side of the mal external tibial rotation. How-
anatomy, clinical examination, and knee. Because of its parallel collagen ever, the degree of external rotation
management of isolated complete arrangement, only 5 to 8 mm of that results from sectioning these
medial collateral ligament (MCL) increased opening indicates a com- two ligaments appears to be mini-
tears in the knee. plete failure of the ligament. Even mal and should not be confused
though this difference in laxity with the more anteromedial rotatory
between the injured knee and the instability that occurs with com-
Functional and Surgical uninjured knee is small, significant bined failure of the MCL and the
Anatomy damage has occurred to the primary anterior cruciate ligament (ACL).
restraint on the medial side of the
The nomenclature describing the knee, a fact not stressed enough
medial ligamentous structures of the when teaching physical examination Dr. Indelicato is Professor of Orthopaedics and
knee has been somewhat confusing. In of the knee. The midmedial portion Huizenga Professor of Sports Medicine, Univer-
the past, the MCL has been called the of the medial capsular ligament, sity of Florida, Gainesville.
“superficial medial collateral ligament” deep to the MCL, provides a firm
or the “tibial collateral ligament.” attachment site for the medial Reprint requests: Dr. Indelicato, University of
Florida Sports Medicine Center, 200 SW 62nd
The MCL is attached proximally to meniscus, but does not provide sig- Boulevard, Suite B, Gainesville, FL 32607.
the medial femoral condyle and dis- nificant restraint against direct val-
tally to the metaphyseal area of the gus stress. Copyright 1995 by the American Academy of
tibia, 4 or 5 cm distal to the medial Müller3 refers to the posterome- Orthopaedic Surgeons.
joint line beneath the pes anserinus dial aspect of the knee (posterior

Vol 3, No 1, Jan/Feb 1995 9

Medial Collateral Ligament Injuries
Vastus
medialis
Medial
capsular
ligament
Pes anserinus
A B
Fig. 1 A, The medial capsuloligamentous complex. B, According to Müller,3 the MCL is dynamized by the connecting fibers of the vas-
tus medialis, and the posterior oblique ligament is dynamized by the semimembranosus.
It is critical to understand the sur-
gical anatomy in order to appreciate
the relationship the various impor-
tant stabilizing structures have with
the MCL. Warren and Marshall4
have defined three distinct layers,
each containing important struc-
tures and landmarks. The most
superficial layer (layer I) consists of
the deep fascia encompassing the
patellar tendon anteriorly and the
popliteal fossa posteriorly and
enveloping the medial hamstrings.
Layer II is mainly composed of the
MCL, which blends posteriorly with
layer III, the posteromedial capsule
(posterior oblique ligament). Layer
III consists of the medial capsular
ligament and blends posteriorly
with layer II to form the posterior
oblique ligament. There is a bursa
that separates layer II, containing the
MCL, from the medial capsular liga-
ment immediately beneath it.
According to Müller, this bursa
10
Posterior oblique
ligament
Semimembranosus
tendon
Medial collateral
ligament
allows the MCL to glide for the
required 1.5-cm anteroposterior
excursion during flexion/extension
of the knee.3 This bursa can be used
as a familiar landmark for proper
orientation when surgical repair is
required.
Clinical Examination
A critical part of the history is to
determine the mechanism of a
potential MCL injury in the knee.
The vast majority of significant MCL
injuries involve a valgus force
applied laterally to the lower thigh
or upper leg. Although complete
tears can occur from a rotational
force alone, such as occurs in snow
skiing, significant tears are more
likely in contact sports, such as foot-
ball and rugby. Since most forces are
coupled (i.e., valgus and external
rotation), it is not unusual to find
Journal of the American Academy of Orthopaedic Surgeons
Connecting fibers
between vastus
medialis and
medial collateral
ligament
Medial meniscus,
posterior oblique
ligament, and
semimembranosus
unit
combined damage to both the MCL
and the posterior oblique ligament.
Müller stated that a pure valgus
force to the lateral aspect of the knee
without any rotational component
will mainly damage the MCL.3 If,
however, a combination of valgus
and external rotational force is sus-
tained, tears primarily to the poste-
rior oblique ligament and possibly to
the ACL can occur before the MCL is
significantly damaged.
When performing the physical
examination, it is critical to keep in
mind some basic principles. The
patient must be relaxed and the con-
tralateral knee must be used as a con-
trol to determine the presence and
degree of asymmetrical medial joint
opening. The key element of the
physical examination is the applica-
tion of a gentle valgus force with the
patient’s knee in slight flexion (30
degrees). When examining a large
patient, one should let the thigh rest

on the table and drop the lower leg off
to the side while supporting the foot
and ankle (Fig. 2). This gives the
patient the opportunity to relax and
keep the thigh musculature flaccid, a
prerequisite for an accurate assess-
ment in the presence of abnormal lig-
amentous laxity. For a brief period of
time following the injury, a patient
may be relatively pain-free even if the
MCL is completely torn. Such a situ-
ation provides a golden opportunity
to perform a very thorough and reli-
able physical examination. The
degree of medial joint opening rela-
tive to the uninjured knee is a direct
measure of damage to the MCL. It is
worth emphasizing that a difference
of only 5 to 8 mm is indicative of sig-
nificant structural damage to the
MCL.1
The second part of the examina-
tion, which evaluates the extent of
medial soft-tissue damage, is per-
formed with the knee in full exten-
sion. The valgus force is repeated,
Fig. 2 Placing the injured leg over the side
of the table will help the patient relax while
allowing the knee to flex the necessary 30
degrees.
Vol 3, No 1, Jan/Feb 1995
and once again the degree of medial
joint opening is compared with that
in the uninjured knee. Asymmetrical
opening in full extension is indicative
of combined MCL and posterior
oblique ligament damage and should
alert the examiner to the possibility of
an associated ACL or posterior cruci-
ate ligament injury. Thus, gross
medial opening in full extension
strongly indicates damage to the cru-
ciate ligaments. If the knee is stable in
full extension, one can safely assume
that there is no significant damage to
the posterior oblique ligament.
A more subtle examiner skill is
the evaluation of the “quality of the
end point.” When a complete MCL
tear is present, no firm end point is
encountered. In isolated complete
tears of the MCL, the end point in the
valgus stressed knee in slight flexion
is the intact ACL. However, this end
point is encountered quite beyond
the normal medial opening as deter-
mined by comparison to the normal
contralateral knee.
Finally, it is important to distin-
guish between localized soft-tissue
swelling and a hemarthrosis. The
former is commonly seen in associa-
tion with MCL damage; the latter is
seen more frequently with associ-
ated ACL rupture. When combined
damage to both the MCL and the
ACL is present, the size of the
hemarthrosis may be minimal
because of the extravasation of blood
outside the knee through the tear of
the medial capsular ligament.
Although the history and physi-
cal examination are critical to deter-
mination of the type and degree of
ligamentous damage to the knee, it
may be difficult to perform the
appropriate examination due to
pain, swelling, and muscle spasm.
In this situation, it is helpful to splint
the extremity and reexamine the
patient a few days later when the
swelling and pain have diminished.
If this proves unsuccessful, magnetic
resonance (MR) imaging may be
Peter A. Indelicato, MD
considered to highlight the location
and extent of ligamentous damage
present, as well as to determine
whether there are coexisting menis-
cal abnormalities. Because of the
development of MR imaging, diag-
nostic arthroscopy in this clinical sit-
uation is now used less frequently.
Management
Grade I and II Injuries
The extent of ligamentous dam-
age determined on clinical examina-
tion dictates treatment. Bergfeld 5
has outlined the management of
incomplete tears of the MCL. Dur-
ing the first 48 hours, ice, compres-
sion, and elevation should be used as
much as possible. In general, incom-
plete tears of the MCL are treated
with temporary immobilization and
the use of crutches for pain control.
Isometric, isotonic, and eventually
isokinetic progressive resistive exer-
cises are begun within a few days of
the subsidence of pain and swelling.
Weight-bearing is encouraged, the
rate being dictated by the level of
pain. Occasionally, persistent pain
over the posterior oblique ligament
prevents full, pain-free extension; in
this setting, the recovery can be
somewhat prolonged.
Derscheid and Garrick6 treated 51
incomplete tears of the MCL in a
prospective study performed in col-
lege football players. A specific
rehabilitation protocol, similar to
that outlined above, was followed.
All players with grade I injuries
returned to full, unprotected partici-
pation an average of 10.6 days after
injury; for players with grade II
injuries, the return took 19.5 days.
Although a slight amount of resid-
ual laxity remained, it proved to be
of no functional significance.
Grade III Injuries
In 1950, O’Donoghue7 advocated
the immediate repair of all complete
11
Medial Collateral Ligament Injuries
tears of the MCL whether they
occurred as an isolated injury or in
conjunction with other major liga-
ment damage: “The knee which
demonstrates a serious or complete
rupture of the medial collateral liga-
ment, a fracture of the meniscus, or a
tear in the cruciate, or any combina-
tion of these, should have early and
careful repair not of one or two, but
of all damaged structures.” In 1983,
Hughston and Barrett8 advocated
primary repair of all torn medial
structures, including the MCL and
the posterior oblique ligament,
when complete disruption is discov-
ered on clinical examination. They
believed that anterior advancement
of the posterior oblique ligament
was the key to restoring medial sta-
bility and advocated proper tension-
ing of each suture under direct
visualization. Müller3 performed a
primary repair of the MCL using
various methods of fixation and
reported good or excellent results in
86% of cases.
Other authors have advocated
nonoperative intervention when the
MCL is completely torn. In 1974, Ell-
sasser et al 9 observed that severe
tears of the MCL in professional foot-
ball players could be treated nonop-
eratively with a high degree of
success. In 1978, Fetto and Mar-
shall10 reported equally satisfactory
results in isolated complete tears of
the MCL, irrespective of whether
they were repaired primarily or
treated conservatively. However,
they stressed the importance of iden-
tifying any associated ligamentous
damage (particularly to the ACL)
and stated that the results were uni-
versally poor when combined liga-
mentous injuries were treated
nonoperatively. More recent studies
have highlighted the importance of
identifying coexisting ACL damage
and the deleterious effect it may have
on MCL healing.11-13 For example, in
1988, Kannus14 reported poor long-
term results of nonoperative treat-
12
ment of complete MCL tears in 27
patients.
Jones et al15 reported the results
of nonoperative treatment of com-
plete isolated MCL tears in 24 high
school football players. The players
were able to return to competition
an average of 34 days after injury
even though some had some mild
residual medial laxity on physical
examination.
A recent study performed by Rei-
der et al16 supports the concept of
early motion and functional rehabil-
itation in the management of iso-
lated MCL injuries in athletes. The
data on 35 patients were reviewed
(average follow-up, over 5 years).
The overall results were very good.
Nineteen of 34 patients estimated
their time to full recovery to be
under 2 months. Thirty-three
patients were able to return to full
participation in the preinjury sport.
Sixteen of 19 football players
returned to the sport within 4 weeks
after the injury.
My preferred method for treating
isolated complete MCL tears has not
changed very much since 1983. 17
Provided structural damage to
either cruciate ligament has been
excluded by clinical examination or
MR imaging, I believe that primary
repair of the MCL is not necessary,
and a structured program of rehabil-
itation can provide predictably good
results in the vast majority of cases.18
What has changed over the past few
years is the rate of progression of the
rehabilitation program. Compared
with the original protocol published
in 1983, the program has become
much more aggressive and driven
by patient comfort and performance
rather than by any predetermined
period of time.
I also believe that arthroscopy is
no longer indicated in every case.
An accurate clinical examination
with particular emphasis on the
Lachman test is sufficient to evaluate
the extent of damage to the ACL.
Journal of the American Academy of Orthopaedic Surgeons
During examination of the knee
soon after injury, the presence of
joint-line tenderness does not neces-
sarily mean a substance tear of the
medial meniscus. Occasionally, MR
imaging is helpful in determining
whether there is any coexisting dam-
age to the meniscus, although it has
been my experience that it is unusual
to have significant damage to the
body of the meniscus when there is a
complete disruption of the MCL.
Once it has been established that
no structural damage to other liga-
ments has occurred, a structured
supervised program is begun. If the
knee is not too painful, a hinged brace
is used, and quadriceps-strengthen-
ing exercises and straight leg raises
are encouraged immediately. If the
knee is painful, it is placed in full
extension in an immobilizer without
hinges. When the initial pain and
swelling subside, the patient is
instructed to remove the immobilizer
five times daily and to perform a
range-of-motion program, the limits
of motion being dictated by comfort,
optimally for 5 minutes each set.
Usually, the patient can easily flex
the knee beyond 90 degrees within 10
to 14 days, and the immobilizer can
then be removed. When the patient
can walk without a noticeable limp,
the crutches are discontinued. No
form of brace is recommended at this
point. Once the knee can be flexed
beyond 100 degrees, the use of a sta-
tionary bicycle is started. This
encourages further motion and
builds strength.
It is important to note that it is
unusual for an effusion to develop
during the rehabilitation program.
If a recurrent effusion is noted, par-
ticularly more than once, the possi-
bility of undiagnosed associated
meniscal and/or articular cartilage
damage should be considered.
As stated previously, a small
group of patients develop postero-
medial pain that prevents comfort-
able full extension at the beginning of

their rehabilitation program. How-
ever, this is much more likely to occur
in patients with incomplete tears of
the posterior oblique ligament and
associated vastus medialis damage
than it is in patients with complete
tears of the MCL not involving the
posterior oblique ligament.
When the patient regains 60% of
his quadriceps strength, as deter-
mined on isokinetic testing, he is
allowed to start a straight-ahead jog-
ging program. Usually this takes
place within the first 3 weeks of ther-
apy. Jogging may be started sooner
on a trampoline or in a pool in chest-
deep water. Usually within a few
more weeks the quadriceps and ham-
string strength will be 80% of that on
the contralateral leg. When this
occurs, the patient is provided with a
functional orthosis, and an agility
program is begun. I prefer an off-the-
shelf model, the type usually recom-
mended following ACL surgery.
Return to contact sports is permitted
whenever the patient can perform an
agility program equivalent to that
needed to play his sport. I encourage
the use of a functional brace for the
remainder of the season because
players feel that such bracing pro-
vides additional protection. How-
ever, continued use of the brace is
discouraged when the player returns
the following season. Because pro-
phylactic knee bracing remains con-
troversial, routine use is not
encouraged.
References
1. Grood ES, Noyes FR, Butler DL, et al:
Ligamentous and capsular restraints
preventing straight medial and lateral
laxity in intact human cadaver knees.
J Bone Joint Surg Am 1981;63:1257-
1269.
2. Warren LF, Marshall JL, Girgis F: The
prime static stabilizer of the medial side
of the knee. J Bone Joint Surg Am
1974;56:665-674.
Vol 3, No 1, Jan/Feb 1995
The obvious question arises:
“When, if ever, should primary repair
of a complete MCL tear be per-
formed?” There still remains some
controversy over this issue.
Recently, Shelbourne and Porter 19
reported the results in 68 patients
who had tears of both the MCL and
the ACL. The ACL tear was treated
with a primary reconstruction using
autogenous patellar tendon, and the
MCL tear was managed nonopera-
tively. They concluded that good to
excellent results could be achieved
without primary repair of the MCL
rupture and saw no advantage to the
additional surgery.
My approach to combined ACL-
MCL damage is to decide intraoper-
atively whether the MCL rupture
requires primary repair. After the
ACL reconstruction is complete, the
medial laxity is reassessed. If the
knee continues to be unstable in full
extension or slight flexion (grade II
or III), the MCL and the posterior
oblique ligament are exposed, and a
primary repair is performed. Unlike
Hughston and Barrett,8 however, I
avoid reefing or advancement of the
posterior oblique ligament during
the repair in order to avoid the risk
of a flexion contracture. If the degree
of medial laxity has been reduced to
grade I after the ACL reconstruction,
primary repair of the medial liga-
ment damage is not necessary.
Regardless of whether a primary
medial repair has been performed,
3. Müller W: The Knee: Form, Function, and
Ligament Reconstruction. New York:
Springer-Verlag, 1983.
4. Warren LF, Marshall JL: The support-
ing structures and layers on the medial
side of the knee: An anatomical analysis.
J Bone Joint Surg Am 1979;61:56-62.
5. Bergfeld J: First-, second-, and third-
degree sprains. Am J Sports Med 1979;7:
207-209.
Peter A. Indelicato, MD
the knee is placed in full extension,
and a range-of-motion program is
started early. The rehabilitation pro-
gram is focused mainly on the ACL
rather than the MCL.
Summary
Clinical evaluation of patients with
complete tears of the MCL will dic-
tate the course of action necessary
for a successful outcome. When
associated damage to the ACL can
be ruled out, primary repair is usu-
ally not necessary. Although it is
important to document the mecha-
nism of injury, the correct diagnosis
is dependent on the results of the
physical examination. Magnetic
resonance imaging and diagnostic
arthroscopy should be reserved for
those situations in which the extent
of damage remains questionable.
There is little controversy regarding
the management of incomplete
tears of the MCL and the posterior
oblique ligament. Although in the
past some authors recommended
primary repair for all complete
tears of the MCL, the pendulum
appears to have swung toward non-
operative management of these
lesions in most cases. There still
remains some controversy regard-
ing the management of these tears
when they are discovered in con-
junction with tears of either cruciate
ligament.
6. Derscheid GL, Garrick JG: Medial collat-
eral ligament injuries in football: Nonop-
erative management of grade I and grade
II sprains. Am J Sports Med 1981;9:365-368.
7. O’Donoghue DH: Surgical treatment
of fresh injuries to the major ligaments
of the knee. J Bone Joint Surg Am
1950;32:721-738.
8. Hughston JC, Barrett GR: Acute antero-
medial rotatory instability: Long-term
13
Medial Collateral Ligament Injuries
results of surgical repair. J Bone Joint
Surg Am 1983;65:145-153.
9. Ellsasser JC, Reynolds FC, Omohundro
JR: The non-operative treatment of col-
lateral ligament injuries of the knee in
professional football players: An analy-
sis of seventy-four injuries treated non-
operatively and twenty-four injuries
treated surgically. J Bone Joint Surg Am
1974;56:1185-1190.
10. Fetto JF, Marshall JL: Medial collateral
ligament injuries of the knee: A ratio-
nale for treatment. Clin Orthop 1978;132:
206-218.
11. Anderson DR, Weiss JA, Takai S, et al:
Healing of the medial collateral ligament
following a triad injury: A biomechanical
and histological study of the knee in rab-
bits. J Orthop Res 1992;10:485-495.
12. Inoue M, McGurk-Burleson E, Hollis
JM, et al: Treatment of the medial col-
14
lateral ligament injury: I. The impor-
tance of anterior cruciate ligament on
the varus-valgus knee laxity. Am J
Sports Med 1987;15:15-21.
13. Woo SLY, Inoue M, McGurk-Burleson
E, et el: Treatment of the medial collat-
eral ligament injury: II. Structure and
function of canine knees in response to
differing treatment regimens. Am J
Sports Med 1987;15:22-29.
14. Kannus P: Long-term results of conser-
vatively treated medial collateral liga-
ment injuries of the knee joint. Clin
Orthop 1988;226:103-112.
15. Jones RE, Henley MB, Francis P: Non-
operative management of isolated
grade III collateral ligament injury in
high school football players. Clin
Orthop 1986;213:137-140.
16. Reider B, Sathy MR, Talkington J, et al:
Treatment of isolated medial collateral
Journal of the American Academy of Orthopaedic Surgeons
ligament injuries in athletes with early
functional rehabilitation: A five-year
follow-up study. Am J Sports Med
1993;22:470-477.
17. Indelicato PA: Non-operative treat-
ment of complete tears of the medial
collateral ligament of the knee. J Bone
Joint Surg Am 1983;65:323-329.
18. Indelicato PA, Hermansdorfer J, Huegel
M: Nonoperative management of com-
plete tears of the medial collateral liga-
ment of the knee in intercollegiate
football players. Clin Orthop 1990;256:
174-177.
19. Shelbourne KD, Porter DA: Anterior
cruciate ligament-medial collateral liga-
ment injury: Nonoperative manage-
ment of medial collateral ligament tears
with anterior cruciate ligament recon-
struction: A preliminary report. Am J
Sports Med 1992;20:283-286.
 Revision Total Hip Arthroplasty:
The Acetabular Component
Pasquale Petrera, MD, and Harry E. Rubash, MD

Abstract
Intermediate and long-term results of revision total hip arthroplasty performed In 1976, DeLee and Charnley 4
with the use of a cemented acetabular component have been disappointing, with reviewed the results in 141 patients
high rates of radiographic and clinical failure. Other methods of acetabular revi- with cemented cups after an average
sion involving the use of threaded cups and bipolar implants have also met with of 10 years. These patients were
high failure rates. Although the long-term results of revision arthroplasty with derived from Charnley’s initial expe-
uncemented acetabular components, especially in terms of polyethylene wear and rience with the all-polyethylene cup
pelvic osteolysis, are not yet available, the intermediate results have been excellent. from 1962 through 1965. Over 69% of
J Am Acad Orthop Surg 1995;3:15-21 the cups showed radiolucent demar-
cations of varying thickness between
the cement and the bone of the
acetabulum, and 13% of the implants
The need for revision total hip arthro- ylene had better wear characteristics had migrated. Overall, 9.2% of the
plasty (THA) continues to grow than PTFE. Acetabular components cups in the total series had loosened
yearly. Fixation of the acetabular com- were fashioned from this material at the 10-year surveillance. DeLee
ponent in this circumstance remains a and first implanted in 1962.1 and Charnley attributed this often
challenge to the orthopaedic surgeon, The technique for implanting symptomless demarcation to techni-
and many methods have been tried polyethylene cups required ream- cal causes during initial implantation
with varying success. To understand ing of the native acetabulum, with and believed that improved surgical
fully the problems that can arise with removal of all articular cartilage and techniques could halt it.
acetabular revisions, the history and subchondral bone. Later the tech-
results of primary acetabular replace- nique evolved to include multiple 6-
ment must be partially reviewed. mm-deep anchor holes to improve Review of Recent Literature
cement fixation. Other investigators
demonstrated the importance of the In a retrieval study in 1992, Schmalz-
Historical Review subchondral acetabular bone as a ried et al5 documented a cause for
weight-bearing structure and rec- the radiolucencies around cemented
John Charnley’s initial attempt to ommended preservation of this
find a bearing surface for his low-fric- bone in acetabular preparation.2
tion arthroplasty was a failure. Poly- Within 6 months after implanting Dr. Petrera is in private practice with Peninsula
tetrafluoroethylene (PTFE), despite the polyethylene cup, Charnley noted Orthopaedic Associates, Salisbury, Md. Dr.
its characteristic slipperiness, was ill some radiographic demarcation Rubash is Associate Professor and Clinical Vice
suited to the demands of a human hip between acetabular bone and cement Chairman, Department of Orthopaedic Surgery,
and Chief, Division of Adult Reconstructive
joint. Cemented PTFE cups wore and described this as a radiolucency. Surgery, University of Pittsburgh Medical Cen-
quickly, and the wear particles were Radiolucencies were not present ter, Pittsburgh.
associated with large, amorphous around the femoral component, and
periprosthetic granulomas. The Charnley worried that they might be Reprint requests: Dr. Rubash, Kaufmann Build-
resultant osteolysis resulted in cata- indicative of early cup loosening or ing, Suite 1010, 3471 Fifth Avenue, Pittsburgh,
PA 15213.
strophic bone destruction and early infection. Despite his concerns, the
implant loosening. Discouraged, but patients with radiolucencies did well Copyright 1995 by the American Academy of
not defeated, Charnley discovered clinically, and initially the radiolucent Orthopaedic Surgeons.
that high-molecular-weight polyeth- lines did not appear to be progressive.3

Vol 3, No 1, Jan/Feb 1995 15

Revision Total Hip Arthroplasty

acetabular cups. They examined ing. This finding points out the rela-
implants from 23 revisions, most of tive “forgiveness” of the acetabulum
which were performed because of in THA and the fact that a loose cup
pain, and 11 autopsy specimens does not necessarily indicate a need
from patients who had been for revision (Table 1). However,
asymptomatic. All had some evi- longer follow-up has corresponded
dence of periprosthetic bone loss. with higher failure rates for
Loosening appeared to be biologic cemented all-polyethylene acetabu-
in nature, and the investigators lar components, as documented by
concluded that it was a result of a radiographic evidence of loosening
physiologic reaction to small par- (Fig. 1) and the need for revision
ticulate polyethylene debris. They surgery.7,8,11-13
observed a “cutting wedge” of
bone resorption, beginning at the
periphery of the cup and contigu- Acetabular Revision With
ous with the joint space. The areas Cement
that corresponded to radiographic
radiolucencies were filled with a Soon after the popularization of
fibrous membrane. Histologic THA, revision procedures became
examination of the membranes necessary. Initially, acetabular
revealed polyethylene wear debris revisions relied on recementing a
Fig. 1 Ten-year follow-up radiograph of a
within macrophages. Ultimately new polyethylene component into loosened cemented all-polyethylene acetab-
these membranes extended to the the pelvis with the use of the same ular component in a patient scheduled for
dome of the component associated technique employed for primary revision THA. Note the continuous bone-
cement radiolucency, the crack in the bone
with eventual component loosen- arthroplasty. Cemented acetabular cement, and the evidence of superior cup
ing. This study and others have revision was a less-than-optimal migration (zone III).
substantiated Charnley’s initial treatment for several reasons. The
concern about acetabular radiolu- acetabular bed was frequently scle-
cencies. rotic without normal trabecular
In 1988, Hodgkinson et al 6 architecture for cement fixation. component fixation, which can be
reported a definite correlation Osteolysis and motion of the loose difficult to achieve under optimal
between radiographic demarcation component within the acetabular conditions in a primary THA, was
and component loosening. They bed often had destroyed large bound to have more frequent fail-
found that 94% of cups with a con- amounts of the anterior and poste- ures after revision procedures.
tinuous radiolucent line were loose rior column and dome of the Later investigations have con-
at subsequent revision and con- acetabulum. Therefore, acetabular firmed this prediction (Table 2).
cluded that radiolucencies around a
cemented socket were a sign of even-
tual failure. These findings, coupled
with those of Schmalzried et al5 and Table 1
others,7-13 changed opinions about Results With Charnley Cemented All-Polyethylene Acetabular
radiolucencies and the eventual fail- Components
ure of cemented acetabular cups.
No. of Follow-up, Rate of Revision
Despite the dire predictions, most
Study Patients yr Loosening, % Rate, %
total hip prostheses with cemented
all-polyethylene cups that have not Eftekhar9 138 7-8 0 1.5
migrated have functioned well. At a Griffith et al10 547 8.3 3.3 0.91
minimum follow-up of 20 years, Stauffer8 231 10 (mean) 11.3 3.0
Schulte et al7 examined 94 cemented Hozack et al11 1,041 10 (mean) 25 1.65
Charnley hips. While 22% of the Older12 153 10-12 15 2.0
acetabula showed radiographic evi- Kavanagh et al13 166 15 (mean) 14 7.2
dence of loosening, only 10% Schulte et al7 98 20 (minimum) 22 10
required revision for aseptic loosen-

16 Journal of the American Academy of Orthopaedic Surgeons


Table 2
Results of Cemented Acetabular Revision
No. of
Study Patients
Kavanagh et al14 166
Amstutz et al15 66
Pellicci et al16 110
Callaghan et al17 139
Snorrason and
Kärrholm18 15
Engelbrecht et al19 138
Goodman and
Schatzker20 32
Marti et al21 60
*Values are means except as noted.
Kavanagh et al14 evaluated 166
cemented acetabular components
with an average follow-up of 4.5
years after revision. Fewer than 2%
required another revision, but 25%
were characterized as probably
loose at this early review. Using the
stricter criterion of Hodgkinson et
al,6 which requires only a complete
radiolucency independent of its
thickness or location, 53% of these
cups would be defined as loose.
In a review of 66 revisions for
aseptically loosened total hip pros-
theses, Amstutz et al15 noted a 3%
rate of further revision after a mean
of only 2 years. Even more alarming
was the observation that 10% of
patients had a complete radiolu-
cency around the cup immediately
after surgery. By 2 years, this rate
had increased to 71%.
Similarly, Pellicci et al16 reviewed
110 revisions at a mean of 3.4 years
and found that 1.8% required further
revision. However, radiographs
taken immediately after the revision
procedure revealed radiolucent
lines in all but nine patients.
At a mean follow-up of 3.6 years,
Callaghan et al17 found that 34.2% of
revised cemented acetabular compo-
nents had circumferential radiolu-
Vol 3, No 1, Jan/Feb 1995
Pasquale Petrera, MD, and Harry E. Rubash, MD
that 61% of the threaded cups they
had placed were loose at an average
of 3.4 years; four required rerevision.
Follow-up, Rate of Revision
Others 25,26 obtained equally poor
yr* Loosening, % Rate, % results using threaded cups for
revision.
4.5 53 2
2.1 71 3
3.4 ... 1.8
3.6 34.2 ... Bipolar Implants
2 93 0 As a possible solution, Scott27 and
7.4 54 ... others proposed the use of bipolar
implants with acetabular bone graft-
3 7 0
5-14 (range) 17 5
ing. However, 61% of acetabular
reconstructions performed with the
use of bipolar prostheses at the Hos-
pital for Special Surgery had failed
by 3 years, and the technique was
recommended for salvage proce-
cencies; 9% of these cups had dures only.28 Emerson et al24 found
migrated. Similar findings were a 49% rate of migration and a 68%
noted in the studies by Snorrason incidence of loosening in bipolar
and Kärrholm,18 Engelbrecht et al,19 components at a mean of only 28
Goodman and Schatzker, 20 and months after revision; clinical results
Marti et al. 21 The findings of were also poor. Scott et al29 noted
Kavanagh and Fitzgerald 22 in that 10 of 19 components had
patients requiring multiple revisions migrated after 2 to 4 years.
were even more alarming: 69% of Outstanding results recently
the cups they examined had radio- were reported for cemented fe-
graphic signs of loosening after a moral components affixed with the
second revision, and the loosening use of new cementing techniques
rate among patients requiring a in both primary and revision
third revision was 100%. Clearly, arthroplasty.30-32 Although these
other methods were needed for reports have increased enthusiasm
acetabular component fixation in for cemented femoral components,
revision procedures. the results obtained with newer
cement techniques and metal-
backed sockets have not shown
Threaded Cups similar improvement. 33-35
Threaded metal cups that screwed
into the acetabular bed gained popu- Acetabular Revision With
larity in the 1980s, but initial enthusi- Bone-Ingrowth Prostheses
asm was soon tempered by early
failures. Engh et al23 found that 45% of Bone-ingrowth prostheses became
such cups of various designs were popular in the United States for both
loose an average of 4.4 years after primary and revision procedures in
implantation. Engelbrecht et al 19 the early 1980s and were seen as the
noted that 83% of the titanium- solution to the problem of the mis-
threaded cups used in revision arthro- named “cement disease.” The pub-
plasty had migrated after an average lished results of revisions with
of 7.4 years. Emerson et al24 found uncemented components (Table 3)
17
Revision Total Hip Arthroplasty
Table 3
Results of Uncemented Acetabular Revision
Mean
No. of Follow-up Rate of Revision
Study Patients mo Loosening, % Rate, %
Hedley et al36 61 20.7 6.6 1.6
Emerson et al24 46 22 15.2 0 Bone grafting was necessary in 127
Engh et al23 34 52.8 2.9 0 revisions; in most of these cases,
Harris et al37 60 17 0 0
Tanzer et al38 140 41 1.4 0.7
Padgett et al39 124 44 0 0
favor the use of hemispheric compo- cup (PCA Hip; Howmedica, Ruther-
nents with some additional form of ford, NJ). Bone slurry was placed in
supplemental fixation, be it fins, the bed of each acetabulum before
spikes, or screws (Fig. 2). impaction of the component. Al-
Hedley et al 36 performed 61 most half of the procedures also
acetabular revisions for infection required structural bone grafts.
and mechanical loosening using the After 20 months, four cups (6.6%)
porous-coated anatomic ingrowth were loose, but only one (1.6%)
required further surgery. Clinical
results were excellent or good in 56
patients despite the presence of radio-
lucencies at the bone-implant inter-
face in 60.7% of cases.
Emerson et al 24 reviewed the
results of 46 acetabular revisions in
which a hemispheric porous-
coated, titanium plasma–spray cup
with four fins had been used. After
an average of 22 months, seven
cups (15.2%) had migrated mini-
mally; none of these was revised.
Four of the seven had required
structural allografting at the index
procedure.
Engh et al23 reported the results in
34 revision THAs in which a hemi-
spheric porous-coated acetabular
component with three spikes had
been used. They found that only one
acetabular component was loose
after an average of 4.4 years.
Fig. 2 Anteroposterior radiograph of a tita-
nium-mesh ingrowth-cup (HGP I) unce-
Harris et al37 identified a need for
mented acetabular component used for particulate bone grafting to the
acetabular revision. At 8-year follow-up, acetabular bed in more than 80% of
there is no evidence of migration, radiolu-
cencies, or pelvic osteolysis. Note the asym-
60 acetabular revisions in which a
metric polyethylene wear of the cup liner titanium-mesh ingrowth cup (HGP
with the 32-mm femoral head. I; Zimmer, Warsaw, Ind) had been
used. After follow-up averaging 17
18 Journal of the American Academy of Orthopaedic Surgeons
months, only one cup had a com-
plete radiolucent line.
Tanzer et al 38 reported the 41-
month results for 140 acetabular
revisions in which titanium-mesh
ingrowth prostheses with supple-
mental acetabular screws had been
used (HGP I and HGP II; Zimmer).
contained defects were filled with
particulate graft material. None of
these grafts was used for major
structural support of the implant.
Only two cups (1.4%) were loose,
and in both cases the patients had
major pelvic discontinuity at the
time of the operation. Five compo-
nents demonstrated a continuous
radiolucency, but none had mi-
grated; these were not considered
loose.
Padgett et al 39 conducted a
prospective study of 124 consecu-
tive acetabular revisions in which a
titanium-fiber metal cup with
screws (HGP I) was used. At 44
months, no revisions for loosening
had been performed, but 4% of
arthroplasties had a continuous
radiolucent line.
The use of cemented acetabula
for primary THA still has its sup-
porters. Long-term review (for
more than 10 years) of the use of
ingrowth cups for primary arthro-
plasty is awaited, but at intermedi-
ate follow-up of 5 to 7 years,
ingrowth cups are producing
results equal to or better than those
obtained with cemented cups. 40
Although the short length of fol-
low-up remains a point of con-
tention in the comparison of
cemented and ingrowth primary
acetabular reconstruction, the same
does not appear to be true for revi-
sions. Although the short-term
results with recemented acetabula
have been disappointing, they still
are used in rare circumstances (e.g.,
patients who require extensive allo-
grafts or who previously under-

went irradiation of the acetabular
bed). In comparable follow-up
periods of 3 to 5 years, ingrowth
cups are performing better in revi-
sion circumstances. Initial enthusi-
asm, however, must be tempered
by the fact that these results are
early. Early experience with
cemented total joint implants and
the nearly exponential failure of
cemented all-polyethylene cups
after 10 years indicate that low early
failure rates cannot be extrapolated
to 20 years.7,8
Modes of Failure
In addition to aseptic loosening
and mechanical failure, other
modes of failure are occurring with
uncemented implants. A recent
investigation by Maloney et al 41
documented the occurrence of
severe pelvic osteolysis related to
uncemented acetabular reconstruc-
tion performed with a variety of
implants. The largely asympto-
matic osteolysis was noted an aver-
age of 5.5 years after the initial
arthroplasty, and particulate poly-
ethylene was implicated as its
cause. Radiographic evaluation
revealed wear of the polyethylene
liner in 80% of the hips. The liner
was 8 mm thick or less in 80% of the
hips, and the femoral head size was
32 mm in 11 of 15 patients. Pelvic
osteolysis is becoming a more fre-
quent indication for revision, and
its incidence will most likely con-
tinue to increase with long-term
studies.
Modular-cup polyethylene as-
semblies with thin (measuring less
than 8 mm) press-fit liners and 32-
mm femoral heads are all potential
causes of increased polyethylene
wear. In particular, holes in cups
may provide access for particulate
polyethylene to migrate to the
pelvic area from the dome of the
acetabulum. With uncemented
cups, access to the implant-bone
interface can also occur at the
Vol 3, No 1, Jan/Feb 1995
Pasquale Petrera, MD, and Harry E. Rubash, MD
periphery of the acetabulum. This
may be one of the reasons for the
continued occurrence of peripheral
radiolucencies in the reported stud-
ies on uncemented components and
is partially responsible for the recent
popularity of underreaming and
press-fitting acetabular components
into the slightly smaller acetabular
bed.
Modular polyethylene assem-
blies are also available with aug-
mented and extended-lip liners.
These constructs have been advo-
cated to improve femoral-head cov-
erage and hip stability. In some
instances, they may be a liability
because they can decrease the effec-
tive range of motion and increase
the chance of dislocation by
impingement of the femoral neck on
the buildup.
While the long-term results with
ingrowth cups are pending, thus far
their use has proved to be a success-
ful and reproducible method of revi-
sion acetabular reconstruction.
Surgical Technique for Porous
Ingrowth Revision
Our technique for acetabular
revision using a hemispheric tita-
nium-mesh cup is as follows: A
modified lateral approach to the hip
is used. Trochanteric osteotomy is
not routine unless it is needed to
facilitate exposure and dislocation.
If the femoral component is unsta-
ble, it is removed; if it is stable, it is
displaced posteriorly to allow
access to the acetabulum. All
fibrous capsular tissue surrounding
the component is removed to pro-
vide access to the bone-cement-
implant interface. Osteotomes are
used to separate the polyethylene
cup from the underlying cement.
The contour of the cup is followed
closely to prevent unnecessary bone
loss or intrapelvic perforations.
Osteotomes, chisels, and curettes
are used to remove all cement from
the acetabular side. Cement within
anchor holes is also removed with
great care. The acetabular mem-
brane is totally removed. Every
attempt is made to preserve the
acetabular rim.
Power reamers are used to
enlarge but not deepen the acetabu-
lum. Once rim contact has been
obtained, fitting of the component is
tried with the size of the last reamer
used. Depending on the bone qual-
ity and the size of reamer, a cup that
is 2 to 4 mm larger than the reamed
acetabulum is chosen. Before final
cup placement, all acetabular
defects are filled with particulate
bone graft (autograft or allograft)
and reverse-reamed for concentric-
ity. A positioning device is used to
impact the component into the
acetabulum in 45 degrees of abduc-
tion and 15 to 20 degrees of antever-
sion. “Bottoming out” of the dome
of the component into the acetabu-
lum is recommended, along with
the rim press-fit.
Stability is then assessed by man-
ual manipulation. If the cup appears
to be unstable, two screws are placed
into the safe zones of Wasielewski et
al.42 A 28-mm-head polyethylene
liner is then placed. A smaller head
size may be needed for smaller cup
diameters. Each system is different,
and the surgeon must be familiar
with the options available.
This technique, although not
greatly different from our primary
acetabular reconstruction technique,
can have pitfalls in revisions.
Overzealous reaming can result in loss
of the anterior and/or posterior
columns. Padgett et al39 found that
retention of the load-bearing posterior
column was of the utmost importance
for successful reconstruction. Ante-
rior-column defects and contained
cavitary defects, whether preexisting
or reamer-induced, can be managed
with a particulate bone graft. Struc-
tural defects of the posterior column
or dome often require large (jumbo)
acetabular components, bulk allo-
19
Revision Total Hip Arthroplasty
graft, and ancillary internal fixation of
the acetabular column and a graft.
Pelvic discontinuity remains an
unsolved problem in revision surgery.
At present, acetabular compo-
nents for primary THA are being
produced without screw holes. The
References
1. Charnley J: Low Friction Arthroplasty of
the Hip: Theory and Practice. New York:
Springer-Verlag, 1979, pp 3-15.
2. Kobayashi S, Terayama K: Radiology
of low-friction arthroplasty of the hip:
A comparison of socket fixation tech-
niques. J Bone Joint Surg Br 1990;72:
439-443.
3. Waugh W: John Charnley: The Man and
the Hip. New York: Springer-Verlag,
1990, pp 113-138.
4. DeLee JG, Charnley J: Radiological
demarcation of cemented sockets in
total hip replacement. Clin Orthop
1976;121:20-32.
5. Schmalzried TP, Jasty M, Harris WH:
Periprosthetic bone loss in total hip
arthroplasty: Polyethylene wear debris
and the concept of the effective joint
space. J Bone Joint Surg Am 1992;74:
849-863.
6. Hodgkinson JP, Shelley P, Wroblewski
BM: The correlation between the
roentgenographic appearance and
operative findings at the bone-cement
junction of the socket in Charnley low
friction arthroplasties. Clin Orthop
1988;228:105-109.
7. Schulte KR, Callaghan JJ, Kelley SS, et
al: The outcome of Charnley total hip
arthroplasty with cement after a mini-
mum twenty-year follow-up: The
results of one surgeon. J Bone Joint Surg
Am 1993;75:961-975.
8. Stauffer RN: Ten-year follow-up
study of total hip replacement: With
particular reference to roentgeno-
g rap hic lo o se ning o f the c o m p o-
nents. J Bone Joint Surg Am 1982;64:
983-990.
9. Eftekhar N: Charnley “low friction
torque” arthroplasty: A study of long-
term results. Clin Orthop 1971;81:
93-104.
10. Griffith MJ, Seidenstein MK, Williams
D, et al: Eight year results of Charnley
arthroplasties of the hip with special
reference to the behavior of cement.
Clin Orthop 1978;137:24-36.
20
rationale is to prevent polyethylene
wear debris from gaining central
access to the bone-implant interface
through holes. Investigations are
proceeding to assess the efficacy of
these press-fit acetabula without
screws in revision procedures.
11. Hozack WJ, Rothman RH, Booth RE Jr,
et al: Survivorship analysis of 1,041
Charnley total hip arthroplasties. J
Arthroplasty 1990;5:41-47.
12. Older J: Low-friction arthroplasty of
the hip: A 10–12-year follow-up study.
Clin Orthop 1986;211:36-42.
13. Kavanagh BF, Dewitz MA, Ilstrup
DM, et al: Charnley total hip arthro-
plasty with cement: Fifteen-year
results. J Bone Joint Surg Am 1989;71:
1496-1503.
14. Kavanagh BF, Ilstrup DM, Fitzgerald
RH Jr: Revision total hip arthro-
plasty. J Bone Joint Surg Am 1985;
67:517-526.
15. Amstutz HC, Ma SM, Jinnah RH, et al:
Revision of aseptic loose total hip
arthroplasties. Clin Orthop 1982;170:
21-33.
16. Pellicci PM, Wilson PD Jr, Sledge CB, et
al: Revision total hip arthroplasty. Clin
Orthop 1982;170:34-41.
17. Callaghan JJ, Salvati EA, Pellicci PM, et
al: Results of revision for mechanical
failure after cemented total hip replace-
ment, 1979 to 1982: A two to five-year
follow-up. J Bone Joint Surg Am
1985;67:1074-1085.
18. Snorrason F, Kärrholm J: Early loos-
ening of revision hip arthroplasty:
A roentgen stereophotogrammet-
r i c analysis. J Arthroplasty 1990;
5:217-229.
19. Engelbrecht DJ, Weber FA, Sweet MBE,
et al: Long-term results of revision
total hip arthroplasty. J Bone Joint Surg
Br 1990;72:41-45.
20. Goodman SB, Schatzker J: Revision
hip surgery using the straight-stem
Muller prosthesis. J Arthroplasty
1987;2:83-88.
21. Marti RK, Schüller HM, Besselaar PP, et
al: Results of revision of hip arthro-
plasty with cement: A five to fourteen-
year follow-up study. J Bone Joint Surg
Am 1990;72:346-354.
22. Kavanagh BF, Fitzgerald RH Jr: Mul-
tiple revisions for failed total hip
Journal of the American Academy of Orthopaedic Surgeons
Conclusion
We are entering a very exciting
period for THA. The next 10 years
will bring forth answers to questions
asked by Charnley more than a gen-
eration ago.
arthroplasty not associated with infec-
tion. J Bone Joint Surg Am 1987;69:
1144-1149.
23. Engh CA, Glassman AH, Griffin WL, et
al: Results of cementless revision for
failed cemented total hip arthroplasty.
Clin Orthop 1988;235:91-110.
24. Emerson RH Jr, Head WC, Berklacich
FM, et al: Noncemented acetabular
revision arthroplasty using allograft
bone. Clin Orthop 1989;249:30-43.
25. Apel DM, Smith DG, Schwartz CM, et
al: Threaded cup acetabuloplasty:
Early clinical experience. Clin Orthop
1989;241:183-189.
26. Shaw JA, Bailey JH, Bruno A, et al:
Threaded acetabular components for
primary and revision total hip
arthroplasty. J Arthroplasty 1990;5:
201-215.
27. Scott RD: Use of a bipolar prosthesis
with bone grafting in acetabular recon-
struction. Contemp Orthop 1984;9:35-41.
28. Brien WW, Bruce WJ, Salvati EA, et al:
Acetabular reconstruction with a bipo-
lar prosthesis and morseled bone
grafts. J Bone Joint Surg Am 1990;72:
1230-1235.
29. Scott RD, Pomeroy D, Oser E, et al: The
results and technique of bipolar revi-
sion hip arthroplasty combined with
acetabular grafting. Orthop Trans
1987;11:450.
30. Mulroy RD Jr, Sedlacek RC, O’Connor
DO, et al: Technique to detect migra-
tion of femoral components of total hip
arthroplasties on conventional radi-
ographs. J Arthroplasty 1991;6(suppl):
S1-S4.
31. Rubash HE, Harris WH: Revision of
nonseptic, loose, cemented femoral
components using modern cementing
techniques. J Arthroplasty 1988;3:
241-248.
32. Oishi CS, Walker RH, Colwell CW Jr:
The femoral component in total hip
arthroplasty: Six to eight-year follow-
up of one hundred consecutive patients
after use of a third-generation cement-

ing technique. J Bone Joint Surg Am
1994;76:1130-1136.
33. Mulroy RD Jr, Harris WH: The effect of
improved cementing techniques on com-
ponent loosening in total hip replacement:
An 11-year radiographic review. J Bone
Joint Surg Br 1990;72:757-760.
34. Harris WH, Penenberg BL: Further fol-
low-up on socket fixation using a
metal-backed acetabular component
for total hip replacement: A minimum
ten-year follow-up study. J Bone Joint
Surg Am 1987;69:1140-1143.
35. Ritter MA, Keating EM, Faris PM, et al:
Metal-backed acetabular cups in total
hip arthroplasty. J Bone Joint Surg Am
1990;72:672-677.
Vol 3, No 1, Jan/Feb 1995
Pasquale Petrera, MD, and Harry E. Rubash, MD
36. Hedley AK, Gruen TA, Ruoff DP: Revi-
sion of failed total hip arthroplasties
with uncemented porous-coated
anatomic components. Clin Orthop
1988;235:75-90.
37. Harris WH, Krushell RJ, Galante JO:
Results of cementless revisions of total
hip arthroplasties using the Harris-
Galante prosthesis. Clin Orthop 1988;
235:120-126.
38. Tanzer M, Drucker D, Jasty M, et al:
Revision of the acetabular compo-
nent with an uncemented Harris-
Galante porous-coated prosthesis. J
Bone Joint Surg Am 1992;74:987-994.
39. Padgett DE, Kull L, Rosenberg A, et
al: Revision of the acetabular compo-
nent without cement after total hip
arthroplasty: Three to six-year follow-
up. J Bone Joint Surg Am 1993;75:
663-673.
40. Petrera P, Rubash HE: Fixing the cup.
Instr Course Lect 1994;3:393-407.
41. Maloney WJ, Peters P, Engh CA, et al:
Severe osteolysis of the pelvis in asso-
ciation with acetabular replacement
without cement. J Bone Joint Surg Am
1993;75:1627-1635.
42. Wasielewski RC, Cooperstein LA,
Kruger MP, et al: Acetabular
anatomy and the transacetabular fix-
ation of screws in total hip arthro-
plasty. J Bone Joint Surg Am 1990;72:
501-508.
21
 Scapular Fractures and Dislocations:
Diagnosis and Treatment
Thomas P. Goss, MD

Abstract

Traumatic injuries of the scapula and the scapulothoracic articulation have Indirect forces may also cause a
received little attention in the literature, since they are uncommon. Scapular frac- variety of avulsion fractures at mus-
tures constitute only 1% of fractures in general, and scapulothoracic dissociations culotendinous and ligamentous
and dislocations are extremely rare. The vast majority can and should be managed attachment sites, such as the superior
nonoperatively. However, recent experience has shown that injuries that involve scapular angle (insertion of the leva-
significant displacement can have long-term adverse functional consequences for tor scapulae), the superior scapular
both the shoulder complex and the upper extremity as a whole. In these situations, border (omohyoid muscle attach-
surgery should at least be considered. Various scapular fractures and dislocations ment), the tip of the coracoid process
are discussed, with particular emphasis on those requiring operative care. (attachment of the conjoined tendon),
J Am Acad Orthop Surg 1995;3:22-33 the superior border of the coracoid
process (attachment of the coraco-
clavicular ligaments), the acromial
margin (origin of the deltoid muscle),
Scapular fractures and dislocations fractures of the glenoid neck, 25%; and the inferior angle of the scapula
can result in considerable morbidity fractures of the glenoid cavity, 10%; (insertion of the serratus anterior).
and therefore deserve more respect and fractures of the acromial and Scapular fractures caused by forceful
and effort than they have been ac- coracoid processes, 7% each. Any muscle contractions associated with
corded in the past. The purpose of fracture line that runs from the pos- seizures, electroconvulsive treat-
this review is to summarize the prin- terior margin of the scapular spine ment, and electrical injuries are well
ciples of the evaluation and treat- or acromion to the undersurface of documented. In addition, fatigue and
ment of these challenging injuries. the acromion all the way to the deep- stress fractures are occasionally seen.
est point of the spinoglenoid interval
is considered an acromial fracture.
Incidence and Mechanisms Scapular fractures are usually Diagnosis
of Injury1-5 caused by high-energy trauma.
Direct forces are most common, al- The physician’s attention is initially
Scapular fractures are uncommon, though indirect mechanisms can be drawn to the scapular region by the
constituting only 1% of all fractures, responsible, such as a fall on the arm patient’s complaints of discomfort in
3% of shoulder-girdle injuries, and that causes the humeral head to the area, often accompanied by
5% of all shoulder fractures. A vari- impact the glenoid cavity. As a
ety of reasons have been offered for result, 80% to 95% of these fractures
this low frequency, of which the are associated with other injuries, Dr. Goss is Professor of Orthopedic Surgery,
most important are that (1) the which may be multiple, major, and Department of Orthopedics, University of Mas-
scapula is protected by the rib cage even life-threatening. Consequently, sachusetts Medical Center, Worcester.
and thoracic cavity anteriorly and a scapular fractures are often diag-
Reprint requests: Dr. Goss, Department of Ortho-
thick covering of soft tissues poster- nosed late, and definitive treatment is
pedics, University of Massachusetts Medical Cen-
iorly, and (2) the mobility of the often delayed. This fact, combined ter, 55 Lake Avenue North, Worcester, MA 01655.
scapula allows considerable dissipa- with the possibility of injury to adja-
tion of traumatic forces. Fractures of cent osseous and/or soft-tissue struc- Copyright 1995 by the American Academy of
the scapular body and spine make tures, may compromise the patient’s Orthopaedic Surgeons.
up approximately 50% of the total; final functional result.

22 Journal of the American Academy of Orthopaedic Surgeons


abnormal physical findings such as
swelling and ecchymosis, or by
abnormalities noted on a chest radio-
graph. If a scapular fracture is noted
or suspected, true scapular antero-
posterior (AP) and lateral views as
well as a true glenohumeral axillary
projection make up the routine
“trauma series.” The scapular body
and spine, the three processes (the
acromial, coracoid, and glenoid
processes), and the three articula-
tions (the scapulothoracic, gleno-
humeral, and acromioclavicular
[AC] articulations) must be evalu-
ated. Oblique views may be helpful
in certain situations, and a stress AP
projection with weights should be
obtained if an injury to the AC artic-
ulation is suspected. Because of the
complex osseous anatomy in the
area, computed tomographic (CT)
scanning with reconstructions is
often necessary to accurately detect
and define the extent of injury.
Nonoperative Treatment
The vast majority (more than 90%) of
scapular fractures are only mini-
mally or acceptably displaced, pri-
marily because of the thick, strong
support provided by the surround-
ing soft tissues. Treatment is symp-
tomatic. Short-term immobilization
in a sling and swathe bandage is pro-
vided for comfort. Early progressive
range-of-motion exercises and use of
the shoulder out of the sling within
clearly defined limits are begun as
pain subsides. In some cases, close
radiographic follow-up is necessary
to ensure that unacceptable dis-
placement does not occur. Most
scapular fractures heal completely
by 6 weeks, and all external support
is discontinued at this time. Pro-
gressive use of the upper extremity
is encouraged. Range-of-motion
exercises continue until full shoul-
der mobility is recovered. As range
of motion improves, progressive
Vol 3, No 1, Jan/Feb 1995
strengthening exercises are added.
It can be anticipated that full func-
tional recovery will take several
months. Ultimately, the prognosis
for these fractures is excellent.
Approximately 50% of scapular
fractures involve the scapular body
and spine. Both avulsion fractures
caused by indirect forces and injuries
caused by direct trauma have been
described. The latter may be severely
comminuted and displaced. Despite
sporadic reports describing operative
management, there seems to be little
enthusiasm for surgical treatment.
There are two reasons for this reluc-
tance: (1) there is little substantial
bone stock for internal fixation aside
from the scapular spine and lateral
scapular border, and (2) these frac-
tures seem to heal reliably with a
good functional result without surgi-
cal treatment. If painful scapulotho-
racic impingement occurs at a later
date, bone prominences over the ven-
tral scapular surface can be removed
surgically.
Operative Indications
While the vast majority of scapular
fractures are managed quite success-
fully without surgery, most agree
that surgical management should be
considered for severely displaced
injuries. Since these fractures are so
rare, large personal series and
strictly controlled studies compar-
ing nonoperatively and operatively
managed injuries are unavailable.
However, the literature does contain
relevant case reports and the per-
sonal experience of a number of
investigators. This review will draw
on this information as well as the
experience of the author to provide
the orthopaedist with operative
guidelines (indications and technical
principles) for the management of
these challenging fractures.
The following injuries occur with
enough frequency to merit discus-
Thomas P. Goss, MD
sion: (1) significantly displaced frac-
tures of the glenoid cavity (glenoid
rim and glenoid fossa), (2) signifi-
cantly displaced fractures of the
glenoid neck, and (3) double disrup-
tions of the superior shoulder sus-
pensory complex (SSSC) in which
one or more elements of the scapula
are significantly displaced.
The following injuries are quite rare
and will not be discussed: fracture of
the scapular body with a lateral spike
protruding into the glenohumeral
joint; significantly displaced, func-
tionally important avulsion fracture
of the scapula; displaced coracoid
fracture associated with neurovascu-
lar compression; coracoid fracture in
the area of the suprascapular notch
with suprascapular nerve paralysis;
significantly displaced fracture of the
distal coracoid process in which the
coracoclavicular ligaments are
attached to the distal fragment; com-
bined rotator cuff tear and acromial
fracture caused by traumatic superior
displacement of the humeral head;
and isolated but significantly dis-
placed fracture of the acromial
process.
Significantly Displaced
Fractures of the Glenoid
Cavity (Rim and Fossa) 6,7
Fractures of the glenoid cavity make
up 10% of scapular fractures, no more
than 10% of which are significantly
displaced. Figure 1 offers a classifica-
tion scheme that outlines the various
mechanisms of injury and fracture
patterns that can occur. For the pur-
pose of this discussion, one need con-
sider only whether the glenoid rim or
the glenoid fossa is fractured.
Fractures of the glenoid rim occur
when a laterally applied high-
energy force drives the humeral
head against the glenoid margin.
Surgical management is indicated if
the fracture results in persistent sub-
luxation of the humeral head,
23
Scapular Fractures and Dislocations

depending on the vector of the trau-

matic force. The degree of resultant
incongruity of the articular surface is
of prime concern. Hardegger et al9
reported that “if there is significant
displacement, conservative treatment
alone cannot restore congruence” and
that “stiffness and pain may result . . .
for this reason, open reduction and
stabilization is indicated.”
Kavanagh et al11 reported on their
experience at the Mayo Clinic with
ten displaced intra-articular frac-
tures of the glenoid fossa treated
with open reduction and internal
fixation (ORIF). They found ORIF to
be “a useful and safe technique for
the treatment of selected displaced
fractures of the glenoid fossa,”
which can “restore excellent func-
tion of the shoulder.” In their series,
the range of displacement of the
major intra-articular fracture frag-
ments was 4 to 8 mm. They empha-
sized that it remained uncertain
how much incongruity of the gle-
noid articular surface could be
accepted without long-term sequelae
of pain, stiffness, and/or traumatic
osteoarthritis.
Soslowsky et al12 found the max-
imal depth of displacement of the
glenoid articular cartilage to be 5
Fig. 1 Scheme for classification of fractures of the glenoid cavity: type Ia, anterior rim frac- mm. Consequently, if displace-
ture; type Ib, posterior rim fracture; type II, fracture line through the glenoid fossa exiting at ment at the fracture site is 5 mm or
the lateral border of the scapula; type III, fracture line through the glenoid fossa exiting at the
superior border of the scapula; type IV, fracture line through the glenoid fossa exiting at the more, subchondral bone is ex-
medial border of the scapula; type Va, combination of types II and IV; type Vb, combination posed, making posttraumatic
of types III and IV; type Vc, combination of types II, III, and IV; type VI, comminuted fracture. osteoarthritis a possibility. Case
reports by Aulicino et al 13 and
Aston and Gregory14 lend support
to the role of surgery in the man-
defined as failure of the humeral fixation of the fragment is indicated agement of significantly displaced
head to lie concentrically within the to prevent recurrent or permanent glenoid fossa fractures. Rüedi
glenoid fossa, or if the reduction is dislocation of the shoulder.” Gut- and Chapman 2 have stated that
unstable. DePalma8 has stated that tentag and Rechtine10 and Butters5 “grossly displaced intra-articular
instability can be expected if the agreed with these recommendations. fractures of the glenoid that render
fracture is displaced 10 mm or more Fractures of the glenoid fossa occur the joint incongruent and unstable
and if at least one fourth of the ante- when a laterally applied high-energy profit from operative reconstruc-
rior aspect of the glenoid cavity or force drives the humeral head directly tion and internal fixation as incon-
one third of the posterior aspect of into the glenoid cavity. The fracture gruities will result in osteoarthritic
the glenoid cavity is involved. generally begins as a transverse dis- changes.” Rowe15 has also advo-
Hardegger et al 9 concurred and ruption, which then propagates in cated surgical management of
stated that “operative reduction and one of several possible directions severely displaced injuries.

24 Journal of the American Academy of Orthopaedic Surgeons

 Thomas P. Goss, MD

On the basis of these reports, it

seems reasonable to conclude that a
fracture of the glenoid fossa with an
articular step-off of 5 mm or more
must be considered for surgical
intervention to restore articular con-
gruity and that displacement of 10
mm or more is a definite indication.
Other indications for surgical man-
agement include (1) glenoid fossa
fractures that result in significant
displacement of the humeral head
such that it fails to lie in the center of
the glenoid cavity, thereby resulting
in glenohumeral instability (Fig. 2);
A B
and (2) fractures of the glenoid fossa
with such severe separation of the Fig. 3 Radiographs of a patient who sustained a type IV fracture of the glenoid cavity. A,
fracture fragments that a nonunion Preoperative AP radiograph shows severe separation of the superior and inferior segments
is likely to occur (Fig. 3). of the glenoid fossa and scapular body. B, Postoperative AP radiograph shows anatomic
reduction and stabilization of the superior and inferior segments of the glenoid fossa and
To detect and define these frac- scapular body with restoration of articular congruity.
tures, a true AP view of the gleno-
humeral joint should be obtained.
This will allow the best visualization
of disruptions to the glenoid fossa osseous anatomy in the region. lateral scapular border, and the cora-
and associated articular incongruity Axial CT images demonstrate frac- coid process. Fixation can be
and/or separation. A true axillary tures of the glenoid rim precisely, achieved with a variety of devices.
radiograph of the glenohumeral while reconstructions in the coronal However, the most useful are appro-
joint will define fractures of the gle- plane are necessary for assessment priately contoured 3.5-mm recon-
noid rim and will indicate whether of glenoid fossa fractures. struction plates and 3.5-mm
the humeral head is subluxated and If ORIF is necessary, 16,17 four interfragmentary compression
whether the reduction is stable. regions of substantial bone stock are screws. I have found cannulated
However, CT scanning is usually available for internal fixation: the screws extremely useful. The choice
necessary because of the complex glenoid neck, the scapular spine, the of implants depends on the surgeon’s
experience and preference and the
available bone stock. Rigid internal
fixation is the desired result, but
inability to achieve rigid fixation does
not necessarily preclude an excellent
anatomic and functional result.

Treatment of Fractures of the

Glenoid Rim
Surgery is indicated if a glenoid
rim fracture results in persistent sub-
luxation of the humeral head or if the
reduction is unstable. As previously
noted, instability is anticipated if the
A B fracture is displaced by 10 mm or
more and if at least one fourth of the
Fig. 2 Radiographs of a patient who sustained a type II fracture of the glenoid cavity. A,
Preoperative AP radiograph shows significant displacement of the inferior glenoid fragment anterior aspect of the cavity or one
and a severe articular step-off. B, Postoperative AP radiograph shows anatomic reduction third of the posterior aspect of the
and stabilization of the inferior glenoid fragment with restoration of articular congruity. cavity is involved. The goal of oper-
ative intervention is to reestablish

Vol 3, No 1, Jan/Feb 1995 25

Scapular Fractures and Dislocations
osseous stability, thereby preventing
chronic glenohumeral instability.
Fractures of the anterior rim are
approached anteriorly, and fractures
of the posterior rim are approached
posteriorly. The fracture fragment is
reduced anatomically and fixed to
the glenoid process with an interfrag-
mentary compression screw. If the
fragment is severely comminuted, it
is excised, and an appropriately
shaped tricortical graft, harvested
from the iliac crest, is rigidly fixed
into the osseous defect. Alterna-
tively, the periarticular soft tissues
can be sutured to the glenoid process,
thereby obliterating the osseous
defect.
Treatment of Fractures of the
Glenoid Fossa
Surgical management should be
considered if there is (1) an articular
step-off of 5 mm or more (this value
represents a relative indication; a
step-off of 10 mm or more is a defi-
nite indication); (2) enough separa-
tion between the glenoid fragments
to make a nonunion likely; and (3)
significant displacement of the gle-
noid fragment such that the humeral
head follows and fails to lie in the
center of the glenoid cavity. The
goals of operative intervention are
to prevent posttraumatic gleno-
humeral osteoarthritis, to avoid
chronic glenohumeral instability,
and to prevent a nonunion at the
fracture site.
All fractures of the glenoid fossa
are approached posteriorly. If access
to the inferior portion of the glenoid
process or the lateral scapular bor-
der is necessary, the interval
between the infraspinatus and teres
minor muscles is developed. If a
superior glenoid fragment is present
and is significantly displaced, a
superior approach to the glenoid
process is added. The major fracture
fragment (or fragments) is reduced
as anatomically as possible and is
internally fixed as securely as possi-
26
ble with use of either an interfrag-
mentary compression screw or a
contoured reconstruction plate.
Associated SSSC disruptions (e.g.,
clavicular or acromial fractures) are
surgically addressed if unacceptable
displacement remains.
Treatment of Type VI Fractures
These injuries include all commi-
nuted fractures of the glenoid cavity.
Nonoperative care is usually indi-
cated because attempts at ORIF can
disrupt what little soft-tissue sup-
port remains. The shoulder is placed
in a position that maximizes articu-
lar congruity. The choices are sling-
and-swathe immobilization, an
abduction brace, and overhead olec-
ranon-pin traction. Early range-of-
motion exercises are begun in an
effort to mold the articular frag-
ments into as normal a relationship
to each other as possible. At 2 weeks
sling-and-swathe immobilization is
used in all cases. By 6 weeks osseous
union is complete. Physical therapy
is continued until range of motion
and strength have been maximized. Fig. 4 Classification of fractures of the gle-
Type VI fractures pose the greatest noid neck. Type I includes all minimally
risk of late symptomatic degenera- displaced fractures. Type II includes all sig-
nificantly displaced fractures (either trans-
tive joint disease and glenohumeral lational or angulatory displacement).
instability.
Significantly Displaced arm, or a fall on the superior aspect of
Fractures of the Glenoid the shoulder. Displacement may
Neck3,18 occur if the fracture is complete, with
the fracture line exiting through both
Fractures of the glenoid neck make up the lateral and superior scapular
25% of scapular fractures; of that margins. If the superior support
number, 10% or fewer (2.5% of the structures (the clavicle–AC joint–
total) are significantly displaced. Fig- acromion strut or the coracoid
ure 4 offers a classification scheme process–coracoclavicular ligaments
that is based on whether these injuries linkage) are disrupted, displacement
are minimally or significantly dis- is especially likely.
placed. If significant displacement DePalma8 has stated that severe
exists, it may be in either the transla- angulation of the articular surface of
tional or the rotatory plane. the glenoid fossa must be corrected
Fractures of the glenoid neck may because it interferes with normal
be caused by a direct blow over the glenohumeral motion and may pre-
anterior or posterior aspect of the dispose to subluxation or dislocation
shoulder, a fall on an outstretched of the joint. He reported that, in gen-
Journal of the American Academy of Orthopaedic Surgeons

eral, marked displacement should be
treated more aggressively. Bateman19
has asserted that significant displace-
ment can interfere with abduction
and that significant angulation can
lead to instability. Nordqvist and
Petersson20 evaluated 37 glenoid neck
fractures treated nonoperatively and
found the functional results at 10- to
20-year follow-up to be fair or poor in
32% of cases. They believed that in
some cases early ORIF might have
improved the result.
Ada and Miller21 retrospectively
reviewed 16 displaced glenoid neck
fractures characterized by transla-
tional displacement greater than or
equal to 1 cm or angulatory defor-
mity greater than or equal to 40
degrees in either the transverse or
the coronal plane. The average
postinjury follow-up period was 36
months. They found that 20% of
patients had decreased range of
motion, 50% had pain (of whom 75%
had night pain), 40% had weakness
with exertion, and 25% noted pop-
ping. In particular, they found that
these individuals frequently had
shoulder abductor weakness and
subacromial pain due at least in part
to rotator cuff dysfunction. They
recommended ORIF for glenoid
neck fractures with this degree of
displacement.
Hardegger et al9 noted that dis-
placed glenoid neck fractures result
in functional imbalance because the
relationship of the glenohumeral
joint with the acromion and nearby
muscle origins is altered. They con-
cluded that in terms of restoration of
normal function, operative treat-
ment is preferable to conservative
management.
There is, therefore, reasonable
support in the literature to suggest
that surgery is indicated or should at
least be considered for significantly
displaced fractures of the glenoid
neck (translational displacement
greater than or equal to 1 cm and/or
angulatory displacement greater
Vol 3, No 1, Jan/Feb 1995
Thomas P. Goss, MD
than or equal to 40 degrees in either tures of the scapular body, and they
the transverse or the coronal plane). should be treated as such).
The basic radiographs necessary to If surgical management is indi-
detect and define these fractures cated (type II fractures),17,18 the gle-
include true AP and lateral views of noid neck is approached posteriorly,
the scapula and a true axillary view of and the interval between the infra-
the glenohumeral joint. However, spinatus and teres minor muscles is
CT scanning is usually necessary to developed to gain access to the infe-
determine whether a complete frac- rior glenoid process and the lateral
ture of the glenoid neck is present, to scapular border. A superior exten-
define the degree of translational or sion can be added to gain access to the
angulatory displacement, and to superior aspect of the glenoid process.
reveal injury to adjacent osseous After reduction of the fracture, fixa-
structures. Three distinct patterns tion is generally achieved with a 3.5-
may be seen: (1) fractures of the mm contoured reconstruction plate
anatomic neck (exiting through the applied along the posterior aspect of
lateral scapular border and the supe- the glenoid fragment and the lateral
rior scapular border lateral to the border of the scapula (Fig. 5). Tempo-
coracoid process); (2) fractures of the rary and supplemental fixation can be
surgical neck (exiting through the lat- provided by Kirschner wires or inter-
eral scapular border and the superior fragmentary screws passed between
scapular border medial to the cora- the glenoid fragment and the adjacent
coid process); and (3) fractures osseous structures.
through the inferior glenoid neck that Occasionally, comminution of the
then run along or through the inferior scapular body or spine can be so
border of the scapular spine before severe or the size of the glenoid frag-
finally exiting out the medial or supe- ment so small as to preclude plate fix-
rior border of the scapula (these frac- ation. In these cases, Kirschner-wire
tures frequently look like displaced or interfragmentary-screw fixation
fractures of the glenoid neck on plain can be used to secure the reduced gle-
radiographs; however, CT scanning noid fragment to adjacent osseous
shows that these are primarily frac- structures, including the acromial
A B
Fig. 5 Radiographs of a patient who sustained a type II fracture of the glenoid neck. A,
Preoperative AP radiograph shows significant angulatory displacement of the glenoid frag-
ment. B, Postoperative AP radiograph shows reduction and stabilization of the glenoid
fragment.
27
Scapular Fractures and Dislocations
process and the distal clavicle. In
those rare instances in which the
scapular body and spine, the acromial
process, and the distal clavicle are all
severely comminuted, overhead olec-
ranon-pin traction must be consid-
ered, or displacement of the glenoid
neck fracture must be accepted.
If a disruption of the clavicle–AC
joint–acromion strut is also present
(most commonly a fracture of the
clavicle), fixation of that injury may
indirectly reduce and stabilize the
glenoid neck fracture in a satisfac-
tory position. If significant displace-
ment persists, the glenoid neck
fracture must be addressed. Con-
versely, ORIF of the glenoid neck
fracture may satisfactorily reduce
and stabilize the clavicle–AC joint–
acromion strut. If not, the associated
disruption must be addressed. Dis-
ruptions of the coracoid pro-
cess–coracoclavicular ligaments
linkage are indirectly managed by
reducing and stabilizing the glenoid
neck fracture and restoring the
integrity of the clavicle–AC joint–
acromion strut.
A B
Fig. 6 Superior shoulder suspensory complex A, AP view of the bone–soft-tissue ring and superior and inferior bone struts. B, Lateral
view of the bone–soft-tissue ring.
28
Double Disruptions of the
SSSC With Significant
Displacement of One or
More Scapular Elements22
The SSSC is a bone–soft-tissue ring
at the end of a superior and an infe-
rior bone strut (Fig. 6). The ring is
composed of the glenoid process, the
coracoid process, the coracoclavicu-
lar ligaments, the distal clavicle, the
AC joint, and the acromial process.
The superior strut is the middle third
of the clavicle. The inferior strut is
the lateral scapular body and spine.
Each individual structure has its
own particular functions. The com-
plex as a whole maintains a normal
stable relationship between the
scapula and upper extremity and the
axial skeleton, allows limited motion
to occur through the AC joint and
the coracoclavicular ligaments, and
provides a firm point of attachment
for several soft-tissue structures.
Traumatic disruptions of one of the
components of the SSSC (Fig. 7) are
common. They tend to be minor
injuries, however, since such single
Journal of the American Academy of Orthopaedic Surgeons
disruptions usually do not signifi-
cantly compromise the overall
integrity of the complex. If the trau-
matic force is sufficiently severe or
adversely directed, the ring may fail
in two or more places (termed a
“double disruption”), a situation in
which significant displacement at
both the individual sites and of the
SSSC as a whole frequently occurs.
Similarly, a disruption of one portion
of the ring combined with a fracture
of one of the struts or fractures of
both struts also creates a potentially
unstable anatomic situation. This, in
turn, often leads to adverse long-
term functional consequences,
including delayed union, nonunion,
and malunion; subacromial impinge-
ment; decreased strength and mus-
cle-fatigue discomfort due to altered
shoulder mechanics; neurovascular
compromise due to a drooping
shoulder; and glenohumeral degen-
erative joint disease. Consequently,
injuries to the SSSC need to be care-
fully evaluated for the presence of a
double disruption. Computed
tomography with reconstructions is

A B
C D
F G
often necessary to make a definitive
diagnosis. If unacceptable displace-
ment is present, surgical reduction
and stabilization of one or more of
the injury sites is necessary. Fre-
quently, operative management of
one of the injury sites will satisfacto-
rily reduce and stabilize the second
disruption indirectly.
Fractures of the glenoid, coracoid,
and acromial processes may each be
part of a double disruption and require
surgical management. All of the vari-
ous combinations cannot be detailed,
and some are extremely rare. How-
ever, some of the more commonly seen
disruptions will be described to illus-
trate the double-disruption principle
as it applies to scapular fractures.
Fractures of the Glenoid Process
Fractures of the Glenoid Neck With
Another Disruption of the SSSC18
Each of these disruptions in isola-
tion is usually minimally displaced
and is therefore treated nonopera-
tively. However, when a fracture of
the glenoid neck is combined with
another SSSC disruption (e.g., an
associated fracture of the middle
third of the clavicle), together they
constitute an anatomically unstable
Vol 3, No 1, Jan/Feb 1995
Fig. 7 Types of traumatic
ring/strut disruptions. Sin-
gle disruptions of the bone–
soft-tissue ring may be a
break (A) or a ligament dis-
ruption (B). Double disrup-
tions of the bone–soft-tissue
ring may be a double-liga-
ment disruption (C), a
double break (D), or a com-
bination of a bone break and
a ligament disruption (E).
E
Other double disruptions
may be a break of both struts
(F) or a break of one strut
and a ring disruption (G).
situation that some have called a
“floating shoulder.” The glenoid
neck fracture allows significant dis-
placement to occur at the clavicular
fracture site, and vice versa.
If displacement at the clavicular
fracture site is unacceptable, surgical
reduction and stabilization is indi-
cated, most commonly with the use
of plate fixation. This may indirectly
reduce and stabilize the glenoid
neck fracture satisfactorily. If not,
the glenoid neck fracture may also
need to be addressed surgically.
Leung and Lam 23 reported on 15
patients with surgically treated frac-
tures (average follow-up period, 25
months). In 14 of the 15 patients the
fractures healed with a good or
excellent functional result.
Herscovici et al 24 reported the
results in nine patients with ipsilateral
clavicular and glenoid neck fractures
(average follow-up period, 48.5
months). Seven patients were surgi-
cally treated with plate fixation of the
clavicular fracture and achieved excel-
lent results. Two patients were treated
without surgery and were found to
have decreased range of motion as
well as “drooping” of the involved
shoulder. The authors strongly rec-
Thomas P. Goss, MD
ommended ORIF of the clavicle to pre-
vent glenoid neck malunion.
Fractures of the Glenoid Cavity With
Another Disruption of the SSSC6
A type I fracture of the distal third
of the clavicle in isolation is usually
minimally displaced and treated
nonoperatively, as is a fracture of the
glenoid cavity in which the superior
aspect of the glenoid process and the
coracoid process are a separate frag-
ment. In combination, however,
each disruption may lead to unac-
ceptable displacement at the other
fracture site. The glenoid fracture
may allow the clavicular fracture to
displace widely, while the clavicular
fracture may allow the superior gle-
noid fragment to displace laterally,
creating an articular step-off that can
result in late traumatic degenerative
joint disease.
If displacement at the clavicular
fracture site is unacceptable, surgical
reduction and stabilization is indi-
cated, usually with a Kirschner
wire–tension band fixation construct.
Since the proximal clavicular seg-
ment is attached to the superior gle-
noid–coracoid process fragment by
means of the coracoclavicular liga-
ments, this may indirectly reduce and
stabilize the glenoid cavity fracture
satisfactorily. If not, the glenoid frac-
ture may also need to be addressed
surgically, using the surgical tech-
niques previously described.
Fractures of the Acromial or
Coracoid Process With Another
Disruption of the SSSC
Isolated fractures of the acromial
and coracoid processes are almost
always minimally displaced and are
therefore managed nonoperatively.
If they are combined with another
disruption of the SSSC (e.g., a frac-
ture of both processes), an unstable
anatomic situation is created.
If displacement at either or both
sites is unacceptable, surgical man-
agement is indicated. Generally,
29
Scapular Fractures and Dislocations

ORIF of the acromial fracture is all
that is required because this will
indirectly reduce and stabilize the
coracoid fracture satisfactorily and
is less difficult than ORIF of the cora-
coid fracture (Fig. 8). Fractures of
the distal acromion are generally sta-
bilized using the dorsal tension-
band technique. Disruptions of the
proximal acromion are more
amenable to plate fixation. Fractures
of the coracoid process are stabilized
with interfragmentary-screw fixa-
tion if the distal fragment is suffi-
ciently large and noncomminuted.
Otherwise, the fragment and con-
joined tendon are reattached with
use of a heavy nonabsorbable suture
placed in a Bunnell fashion through
the tendon and then through a drill
hole in the coracoid process proxi-
mal to the fracture site.
tinued, and progressive functional use
of the articulation is encouraged. If
transfixing Kirschner wires have been
placed, they are removed at this time.
Physical therapy is continued until
range of motion and strength have
been maximized. The initial emphasis
is on regaining range of motion. As
range of motion improves, progres-
sive strengthening exercises are
added.
The patient must be encouraged to
continue to work diligently on his
rehabilitation program, since range
of motion and strength can still
improve, and often the end result is
not achieved for approximately 6
months to 1 year after injury. The
functional outcome after operative
treatment of significantly displaced
scapular fractures is dependent on
the specifics of the injury, the ade-
quacy of the reduction, the quality of
the fixation, and the rigor of the post-
operative rehabilitation program.
Scapulothoracic
Dissociation (Lateral
Dislocation of the
Scapula)25
Scapulothoracic dissociation (Fig. 9)
is a rare traumatic disruption of the
scapulothoracic articulation caused
by a severe direct force over the

Postoperative Management
and Rehabilitation

The postoperative care of surgically

treated scapular fractures depends
on the degree of stability achieved.
Rigidly fixed fractures are protected
in a sling and a swathe bandage.
Early progressive range-of-motion A B
exercises are begun, and functional
use of the shoulder out of the sling is
permitted as symptoms allow. If sta-
bilization is less than rigid, full-time
postoperative immobilization in a
sling and a swathe bandage, an
abduction splint, or even overhead
olecranon-pin traction for 7 to 14
days may be necessary before initiat-
ing the rehabilitation program.
By 2 weeks, most fractures need
only a sling and a swathe bandage for
protection. Progressive range-of-
motion exercises are begun at this C D
point, and gradually increasing func- Fig. 8 Images of a patient who sustained fractures of both the coracoid and the acromial
tional use of the arm is permitted processes. A, Preoperative AP radiograph of the involved area. B, Axial CT image of the
within clearly defined limits. The acromion. Note the wide separation between the acromial fragments due to the associated
coracoid process fracture. C, Axial CT image shows the fractured coracoid process. D,
patient is carefully monitored. At 6 AP radiograph of the shoulder after ORIF of the acromial fracture using the dorsal tension-
weeks, bone union is usually com- band technique.
plete, all external protection is discon-

30 Journal of the American Academy of Orthopaedic Surgeons

 Thomas P. Goss, MD

A B C

Fig. 9 Images of a patient who sustained a left scapulothoracic dissociation. A, Preoperative AP radiograph shows significant lateral dis-
placement of the scapula and a significantly displaced fracture of the distal clavicle. B (top), CT image shows significantly increased dis-
tance between the left scapula and the rib cage as compared with the opposite (uninjured) side. B (bottom), Arteriogram shows disruption
of the subclavian artery. C, Postoperative AP radiograph shows reduction and stabilization of the distal clavicle (and secondarily the scapu-
lothoracic articulation) obtained with use of a tension-band technique.

shoulder accompanied by traction
applied to the upper extremity.
Although the skin remains intact,
the scapula is torn away from the
posterior chest wall, prompting
some to call this injury a “closed
traumatic forequarter amputation.”
Due to the violent forces involved,
any of the three bones in the shoul-
der complex (the clavicle, the
scapula, and the proximal humerus)
may be fractured, and any of the
remaining three articulations (the
glenohumeral, AC, and sternoclav-
icular joints) may be disrupted.
Neurovascular injury is common.
Disruption of the subclavian or axil-
lary artery (most frequently the
former) and complete or partial dis-
ruption of the brachial plexus also
occur.
In addition, there may be severe
damage to the soft-tissue support-
ing structures, especially those that
run from the chest wall to the
scapula or from the chest wall to
the humerus. Complete and par-
tial tears of the trapezius, levator
scapulae, rhomboids, pectoralis
minor, and latissimus dorsi have
all been described.
A presumptive diagnosis is
based on a history of violent
trauma and the presence of mas-
sive soft-tissue swelling over the
shoulder girdle. A pulseless upper
extremity, indicating a complete
vascular disruption, and a com-
plete or partial neurologic deficit,
indicating injury to the brachial
plexus, are quite suggestive. Sig-
nificant lateral displacement of the
scapula seen on a nonrotated chest
radiograph confirms the diagnosis.
As with all rare injuries, awareness
of the clinical entity is critical to
making the correct diagnosis.
Treatment recommendations
have focused on care of the accom-
panying neurovascular injury. If
the vascular integrity of the
extremity is in question, an emer-
gency arteriogram is performed,
followed by surgical repair if nec-
essary. The brachial plexus is
explored at the same time. If a neu-
rologic deficit is present, elec-
tromyographic testing is performed
3 weeks after injury to determine
the extent of injury and to assess
the degree of recovery, if any. Cer-
vical myelography can be per-
formed at 6 weeks.
If nerve avulsions or a complete
neurologic deficit is present, the
prognosis for a functional recovery
is poor. However, partial plexus
injuries have a good prognosis,
and most patients achieve com-
plete recovery or regain functional
use of the extremity. If some por-
tions of the plexus are intact and
others are disrupted, neurologic
repair is a possibility. Late recon-
structive efforts are guided by the
degree of neurologic return, and
musculotendinous transfers are
performed as needed.
Care of the surrounding soft-tis-
sue supportive structures (muscu-
lotendinous and ligamentous) has
been nonoperative, consisting of
immobilization of the shoulder
complex for 6 weeks to allow heal-
ing, followed by a closely moni-
tored physical therapy program
designed to restore range of

Vol 3, No 1, Jan/Feb 1995 31

Scapular Fractures and Dislocations
motion initially, followed by
strength. Magnetic resonance imag-
ing of the involved area now offers
the ability to visualize important
disruptions that may be amenable
to surgical repair.
Injury to the sternal-clavicular-
acromial linkage (a disruption of
the sternoclavicular or AC joint or
a fracture of the clavicle) is fre-
quently, if not invariably, present
in these injuries, permitting pos-
terolateral displacement of the
scapula. This component of scapu-
lothoracic dissociation has been
largely ignored in terms of both
diagnosis and treatment.
Of the three possible disrup-
tions, fracture of the clavicle seems
to be the most common. This con-
stitutes a very unstable anatomic
situation, with the clavicular frac-
ture allowing maximal displace-
ment of the scapula. The unstable
scapulothoracic articulation also
causes significant displacement at
the clavicular fracture site. Conse-
References
1. Norris TR: Fractures and dislocations of
the glenohumeral complex, in Chap-
man MW, Madison M (eds): Operative
Orthopaedics. Philadelphia: JB Lippin-
cott, 1988, vol 1, pp 203-220.
2. Rüedi T, Chapman MW: Fractures of
the scapula and clavicle, in Chapman
MW, Madison M (eds): Operative
Orthopaedics. Philadelphia: JB Lippin-
cott, 1988, vol 1, pp 197-202.
3. Miller ME, Ada JR: Fractures of the
scapula, clavicle, and glenoid, in
Browner BD, Jupiter JB, Levine AM, et
al (eds): Skeletal Trauma: Fractures, Dis-
locations, Ligamentous Injuries. Philadel-
phia: WB Saunders, 1992, vol 2, pp
1291-1310.
4. Neer CS II, Rockwood CA Jr: Fractures
and dislocations of the shoulder, in
Rockwood CA Jr, Green DP (eds): Frac-
tures in Adults, 2nd ed. Philadelphia: JB
Lippincott, 1984, vol 1, pp 713-721.
5. Butters KP: The scapula, in Rockwood
CA Jr, Matsen FA II (eds): The Shoulder.
32
quently, surgical ORIF of the clav-
icular fracture site should be con-
sidered (1) to avoid a delayed
union or nonunion at the clavicular
fracture site; (2) to restore as much
stability as possible to the shoulder
complex, in order to avoid adverse
functional consequences; and (3) to
protect the brachial plexus and the
subclavian and axillary vessels
from further injury caused by ten-
sile forces.
Similar therapeutic reasoning
would apply to scapulothoracic
dissociations accompanied by dis-
ruption of the AC joint. Open
reduction and internal fixation of a
sternoclavicular disruption has
less appeal because of the technical
difficulty involved.
Summary
Scapular fractures and disruptions
are decidedly uncommon. Because
of the complex osseous and articular
Philadelphia: WB Saunders, 1990, vol 1,
pp 335-366.
6. Goss TP: Fractures of the glenoid cav-
ity. J Bone Joint Surg Am 1992;74:299-
305.
7. Ideberg R: Fractures of the scapula
involving the glenoid fossa, in Bateman
JE, Welsh RP (eds): Surgery of the Shoul-
der. Philadelphia: BC Decker, 1984, pp
63-66.
8. DePalma AF: Surgery of the Shoulder, 3rd
ed. Philadelphia: JB Lippincott, 1983.
9. Hardegger FH, Simpson LA, Weber BG:
The operative treatment of scapular
fractures. J Bone Joint Surg Br 1984;
66:725-731.
10. Guttentag IJ, Rechtine GR: Fractures of
the scapula: A review of the literature.
Orthop Rev 1988;17:147-158.
11. Kavanagh BF, Bradway JK, Cofield RH:
Open reduction and internal fixation of
displaced intra-articular fractures of the
glenoid fossa. J Bone Joint Surg Am
1993;75:479-484.
Journal of the American Academy of Orthopaedic Surgeons
anatomy in the area, CT scanning
with or without reconstructions is
usually needed to detect and accu-
rately define these injuries.
Most scapular fractures are not
significantly displaced, and nonop-
erative treatment will reliably yield
a good to excellent functional result.
Fractures that are significantly dis-
placed can result in adverse healing
and long-term functional conse-
quences and should therefore be
considered for ORIF.
Scapulothoracic dissociations
associated with a vascular disrup-
tion must be diagnosed promptly, so
that the appropriate vascular repair
can be performed. Reestablishing
the integrity of the sternal-clavicu-
lar-acromial linkage is advisable if
possible.
In all scapular fractures and disloca-
tions, an optimal end result is depen-
dent on the severity of the injury, the
adequacy of the reduction, the quality
of the fixation, and the rigor of the
postoperative rehabilitation program.
12. Soslowsky LJ, Flatow EL, Bigliani LU,
et al: Articular geometry of the gleno-
humeral joint. Clin Orthop 1992;285:
181-190.
13. Aulicino PL, Reinert C, Kornberg M, et
al: Displaced intra-articular glenoid
fractures treated by open reduction and
internal fixation. J Trauma 1986;26:
1137-1141.
14. Aston JW Jr, Gregory CF: Dislocation of
the shoulder with significant fracture of
the glenoid. J Bone Joint Surg Am
1973;55:1531-1533.
15. Rowe CR (ed): The Shoulder. New York:
Churchill Livingstone, 1988.
16. Goss TP: Fractures of the glenoid cav-
ity: Operative principles and tech-
niques. Techniques Orthop 1994;8:
199-204.
17. Goss TP: Fractures of the glenoid
cavity [videotape]. Rosemont, Ill:
American Academy of Orthopaedic
Surgeons Physician Videotape Library,
1994.

18. Goss TP: Fractures of the glenoid neck.
J Shoulder Elbow Surg 1994;3:42-52.
19. Bateman JE: The Shoulder and Neck,
2nd ed. Philadelphia: WB Saunders,
1978.
20. Nordqvist A, Petersson C: Fracture of
the body, neck, or spine of the scapula:
A long-term follow-up study. Clin
Orthop 1992;283:139-144.
Vol 3, No 1, Jan/Feb 1995
21. Ada JR, Miller ME: Scapular fractures:
Analysis of 113 cases. Clin Orthop
1991;269:174-180.
22. Goss TP: Double disruptions of the
superior shoulder suspensory complex.
J Orthop Trauma 1993;7:99-106.
23. Leung KS, Lam TP: Open reduction
and internal fixation of ipsilateral frac-
tures of the scapular neck and clavicle.
Thomas P. Goss, MD
J Bone Joint Surg Am 1993;75:1015-
1018.
24. Herscovici D Jr, Fiennes AGTW, Ruedi
TP: The floating shoulder: Ipsilateral
clavicle and scapular neck fractures. J
Orthop Trauma 1992;6:499.
25. Ebraheim NA, An HS, Jackson WT, et al:
Scapulothoracic dissociation. J Bone
Joint Surg Am 1988;70:428-432.
33
 Disorders of the First Metatarsophalangeal Joint
Roger A. Mann, MD

Abstract

The two most common disorders of the first metatarsophalangeal (MTP) joint are metatarsalgia and deformity of the
hallux valgus and hallux rigidus. The hallux valgus deformity has been the sub- great toe due to lack of stability. Fur-
ject of numerous clinical studies in the past decade. This information has enabled thermore, many physically active
the creation of an algorithm to assist the clinician in evaluating the patient with patients had poor function.
hallux valgus and selecting the appropriate surgical procedure. The technical To alleviate the chief complaint of
aspects of various operative procedures and the most common complications are pain over the medial eminence, Sil-
reviewed. The other major disorder of the first MTP joint is arthrosis, which ver2 advised excision of the medial
results in hallux rigidus. As the arthrosis progresses, there is often proliferation eminence and plication of the medial
of bone on the dorsal aspect of the metatarsal head, which results in impingement joint capsule. This relatively simple
of the proximal phalanx during dorsiflexion. The impingement causes jamming, procedure resulted in alleviation of
instead of gliding, of the proximal phalanx on the metatarsal head, which results the painful bunion, but the correction
in pain. The treatment for this condition consists of debridement of the MTP joint of the first MTP joint abnormality was
to relieve the dorsal impingement and, in most cases, the pain. If the arthrosis is not achieved in most cases, except in
advanced in an active individual, arthrodesis is indicated. those with a minimal deformity.
J Am Acad Orthop Surg 1995;3:34-43 It became recognized that there
were some patients in whom the
problem was more than just a defor-
mity at the MTP joint, and that the
Hallux valgus and hallux rigidus are the basis for rational decisions regard- intermetatarsal (IM) angle must also
the most common disorders of the ing treatment of the deformity. There be corrected. The McBride proce-
first metatarsophalangeal (MTP) is no single operative procedure to cor- dure3 consists of excision of the fibu-
joint. Selection of the optimal treat- rect all types of hallux valgus deformi- lar sesamoid, removal of the medial
ment for either of these painful con- ties, since no two bunion deformities eminence, plication of the medial joint
ditions must be based on careful are exactly alike. The surgeon must be capsule, and implantation of the
consideration not only of objective very specific in the selection of the adductor tendon into the first
clinical and radiographic findings but operative procedure, which must then metatarsal in order to reduce the IM
also of subjective factors, such as the be carried out in a technically correct angle. This procedure provided bet-
patient’s lifestyle and expectations. manner and carefully followed post- ter long-term results than just excising
operatively to obtain the maximum the medial eminence and plicating the
correction and to minimize com-
Hallux Valgus plications.

Dr. Mann is Associate Clinical Professor of

The hallux valgus deformity can pre-
sent in many forms. The main problem
may be a prominent medial eminence
(bunion) or may be due to the lateral
deviation of the proximal phalanx on
the first metatarsal, which results in a
second-toe problem. The precise type
of the deformity, the patient’s clinical
complaints, and the physical examina-
tion and radiographic findings form
Historical Overview of Surgical
Treatment
Keller1 approached hallux valgus
by decompression of the MTP joint
through resection of the base of the
proximal phalanx and removal of the
medial eminence. Unfortunately,
this approach destabilizes the first
MTP joint due to loss of the windlass
mechanism and results in transfer
Orthopaedic Surgery, University of California
School of Medicine, San Francisco, and is Direc-
tor of the Foot Fellowship Program in his private
practice in Oakland, Calif.
Reprint requests: Dr. Mann, 3300 Webster
Street, No. 1200, Oakland, CA 94609.
Copyright 1995 by the American Academy of
Orthopaedic Surgeons.

34 Journal of the American Academy of Orthopaedic Surgeons


capsule, since it did address, to a cer-
tain extent, the increased IM angle.
Overall results were satisfactory, but
overcorrection of the MTP joint
occurred, resulting in hallux varus,
probably mainly due to the imbalance
brought about by excision of the fibu-
lar sesamoid and plication of the
medial joint-capsule structures.
To diminish the incidence of
varus, DuVries4 proposed that the
adductor tendon no longer be placed
into the metatarsal, but rather
sutured adjacent to it, which
resulted in satisfactory alignment
but inconsistent results. Osteotomy
then began to be used increasingly.
Osteotomies
The recognition that the abnormal
IM angle must also be addressed
resulted in metatarsal osteotomies of
many shapes and forms designed to
address this component of the defor-
mity. Ideally, the most useful osteot-
omies correct the IM angle with
minimal shortening and provide ade-
quate stability of the osteotomy site.
Most metatarsal osteotomies are car-
ried out either distally or proximally,
although the degree of correction that
can be obtained with a distal osteotomy
is not as great as with a proximal
osteotomy. As a rule, the distal
osteotomy is used for the mild to low-
moderate deformity, and the proximal
osteotomy is used for the more severe
deformity. Regardless of the site, 2 to 3
mm of shortening is inevitable.
Further, it has been recognized that
metatarsal osteotomy alone is insuffi-
cient to correct the entire deformity.
A distal soft-tissue procedure must
be added, including release of the lat-
eral soft-tissue contracture. This
release involves the adductor hallucis
from the proximal phalanx and fibu-
lar sesamoid, the lateral joint capsule,
and the transverse metatarsal liga-
ment. The medial eminence is
exposed and excised in line with the
medial aspect of the metatarsal shaft.
The medial joint capsule is plicated.
Vol 3, No 1, Jan/Feb 1995
Occasionally, some dorsiflexion
results after osteotomy. Precisely how
much shortening and dorsiflexion can
be tolerated without causing a clinical
symptom remains uncertain. Dorsi-
flexion of the first MTP joint does
bring about plantar flexion of the first
metatarsal, which, in turn, will allow
for a certain degree of shortening
and/or elevation of the metatarsal. In
a recent study of our patients with a
proximal osteotomy,5 about 30% were
found to have some element of dorsi-
flexion of the metatarsal, but 43 of 48
patients with a preexisting lesion
beneath the second metatarsal had
relief of symptoms, and no transfer
lesion developed in any case.
Fusions
Instability of the metatarso-
cuneiform joint is a factor in 2% to 3%
of patients with a hallux valgus defor-
mity.6 If there is marked instability, a
metatarsocuneiform arthrodesis is
necessary. Without arthrodesis, an
early recurrence of the deformity
results. This procedure has limited
application and should probably not
be used in the patient with high ath-
letic ambitions.
Arthrodesis of the MTP joint will
produce excellent correction of a
severe deformity or can be used to sal-
vage a failed operative procedure.
The result will not deteriorate with
time, and patients can resume most
activities. This procedure is not rec-
ommended for persons engaged in
competitive athletics. It is not neces-
sary to correct even a significant
deformity of the IM angle when an
arthrodesis is carried out; the metatar-
socuneiform joint will self-correct
after arthrodesis due to the motion
that normally occurs at the joint.7
Clinical Evaluation
In addition to the medical history,
it must be determined whether the
patient’s main concern is cosmesis,
transfer metatarsalgia, second-toe
deformity, problems with shoe fit, or
Roger A. Mann, MD
pain. The patient’s occupational and
recreational requirements are also
important; professional dancers and
high-performance athletes need
very special consideration before
any type of foot surgery.
A critical factor is patient expecta-
tions. It is crucial that the patient
understand precisely what can be
achieved surgically and what cannot.
Unfortunately, patients may have been
led to believe there is a “quick fix” for
many foot problems. In a study of our
patients, we found that prior to surgery
for hallux valgus only one third of the
patients could wear the type of shoe
they desired. After surgery, two thirds
achieved their goal, which unfortu-
nately left one third unsatisfied.
Physical Examination
Physical examination begins with
the patient standing in order to
observe the alignment of the lower
extremities, the longitudinal arch,
and the great toe and the relationship
of the lesser toes to it. With the
patient in a sitting position, it usually
is impossible to fully appreciate the
dynamic instability in the foot. The
range of motion of the ankle and the
subtalar and transverse tarsal joints
of the forefoot is assessed. Overall
tissue elasticity is evaluated, and the
presence of a tight Achilles tendon is
sought, especially in the juvenile.
The range of motion of the first
MTP joint is carefully observed by
dorsiflexing and plantar flexing the
joint while attempting to realign the
great toe. The degree of restriction of
dorsiflexion indicates how much cor-
rection can be obtained surgically. If
the great toe cannot be brought into
fairly good alignment without severe
restriction of dorsiflexion, the sur-
geon and the patient should under-
stand that full motion will not be
achievable. The degree of pronation
of the great toe should also be noted.
As a general rule, the more severe the
deformity, the greater the degree of
pronation.
35
Disorders of the First Metatarsophalangeal Joint
Other physical findings of note
include synovial thickening, dorsal
osteophytes, sesamoid pain, and crepi-
tation. Significant callosities under the
first metatarsal head are secondary to
prominence of the tibial sesamoid;
under the second metatarsal, they are
due to the increased weight-bearing
brought about by instability of the hal-
lux. The second MTP joint must be
evaluated for instability, medial devia-
tion, and the presence of a hammer toe.
Not infrequently, the second toe is
more symptomatic, in terms of pain
and deformity, than the hallux even
though the great toe has initiated the
problem.
The stability of the first metatarso-
cuneiform joint is evaluated by hold-
ing the second metatarsal head in
one hand and the first metatarsal in
the other. The first metatarsal head
is deviated dorsomedially and then
plantarward and laterally. Unfortu-
nately, because there are no precise
guidelines for instability, the exami-
nation cannot be quantified. Signif-
icant instability is observed in about
2% to 3% of patients with hallux val-
gus, often associated with a moder-
ate to severe flatfoot deformity.
Finally, the neurovascular status
of the foot should be evaluated. If
there is any doubt about the
integrity of the circulation, further
studies should be obtained before
any surgery is contemplated.
Radiographic Evaluation
The foot should always be evalu-
ated with weight-bearing radio-
graphs, because non-weight-bearing
radiographs often fail to indicate the
severity of a deformity. The follow-
ing factors should be assessed:
Hallux Valgus Angle
The extent of hallux valgus defor-
mity is determined by measuring
the angle between the long axes of
the proximal phalanx and the first
metatarsal (Fig. 1). The normal
value is less than 15 degrees.
36
imal phalanx and the long axis of the
proximal phalanx and the relationship
between the long axis of the proximal
phalanx and that of the distal phalanx
(Fig. 3). A deformity may exist at one
or more levels. The normal value is less
than 10 degrees of lateral deviation.
Obliquity of the Metatarsocuneiform
Joint
The angle of the metatarsocuneiform
joint is quite variable radiographically.
If there appears to be more than 15
degrees of medial deviation, one
should reevaluate the patient for possi-
ble instability of this joint (Fig. 4).
Fig. 1 The hallux valgus (HV) angle is
formed by the intersection of lines bisecting
the proximal phalanx and the first
metatarsal. Normal is less than 15 degrees.
In this case, it measures 40 degrees. The IM
angle is formed by the intersection of the
lines that bisect the first and second
metatarsals. Normal is less than 9 degrees.
In this case, it measures 16 degrees.
Intermetatarsal Angle
The IM angle is the angle between
the first and second metatarsals (Fig.
1). The normal value is less than 9
degrees.
Distal Metatarsal Articular (DMA)
Angle
The DMA angle describes the rela-
tionship between the distal articular
surface and the long axis of the first
metatarsal (Fig. 2). 8 The normal
value is less than 10 degrees of lat-
eral deviation.
Hallux Valgus Interphalangeus
Hallux valgus interphalangeus is a
deformity involving only the pha-
langes of the hallux. The presence of Fig. 2 Top, The normal DMA angle is less
this abnormality is identified on the than 10 degrees of lateral deviation. In this
case, it is 0 degrees. Bottom, An abnormal
basis of the relationship between the DMA angle of 27 degrees.
articular surface of the base of the prox-
Journal of the American Academy of Orthopaedic Surgeons

Fig. 3 Hallux valgus interphalangeus is
identified on the basis of the angle between
the lines bisecting the proximal and distal
phalanges of the metatarsal. Normal is less
than 10 degrees. In this case, it is 30 degrees.
Sesamoid Position
The position of the sesamoids in
relation to the metatarsal head pro-
vides information regarding the
Fig. 4 The metatarsocuneiform joint
demonstrates marked obliquity, which may
indicate instability. A radiographic appear-
ance such as this indicates that the patient
should be reevaluated clinically.
Vol 3, No 1, Jan/Feb 1995
severity of the hallux valgus defor-
mity, the degree of pronation of the
hallux, and possible pathologic
changes in the sesamoids.
Arthrosis of the MTP Joint
Arthrosis is not common, but may
occur in hallux valgus and can influ-
ence the treatment outcome.
Determination of the Congruence of the
MTP Joint
A congruent MTP joint has no lat-
eral subluxation of the proximal
phalanx on the metatarsal head.9 An
incongruent or subluxated joint
demonstrates lateral deviation of
the proximal phalanx from the
metatarsal head (Fig. 5).
Decision Making
Essentially, all hallux valgus
deformities can be treated conserva-
tively with use of a wide, soft shoe
that provides an adequate toe box
and sufficient insole padding to
make the patient comfortable. A
surgical option is considered if the
patient is not satisfied with that con-
servative approach.
The most important factor in
determining the surgical approach
is the congruence of the MTP joint.
A congruent joint is one in which
the articular surfaces of the proxi-
mal phalanx and metatarsal head
are parallel. The operative proce-
dure should protect the integrity of
an anatomically aligned joint. If
there is incongruence, or lateral
subluxation, the surgical correction
should attempt to bring the proxi-
mal phalanx back onto the
metatarsal head, thereby recreating
a congruent joint.
The other major determining fac-
tor is the presence of arthrosis of the
joint. Realignment of the joint will
result in satisfactory correction of
the deformity; unfortunately, how-
ever, significant joint stiffness and
usually pain will result. In the 2% to
Roger A. Mann, MD
3% of patients who have hypermo-
bility of the metatarsocuneiform
joint, stabilization of that joint must
be achieved to obtain a satisfactory
long-term result.6
The algorithm presented in Figure
6 presents an approach the surgeon
may find useful in the selection of an
operative procedure from among
the more than 130 options. The
choices presented in the algorithm
have evolved over a period of time
and have undergone extensive clini-
cal evaluation. The selection of a
procedure is guided by the severity
of the deformity. The measurements
given in the algorithm to assess the
deformity are guidelines, and there
certainly is a moderate degree of lee-
way in the selection process. The
physical and radiographic findings
must always be correlated when
making a decision.
For the patient with significant
arthrosis of the MTP joint, an
arthrodesis is recommended, al-
though a carefully performed
Keller procedure can give a satis-
factory result in the patient with
limited ambulatory capacity. Cur-
rently, the use of a prosthesis in
primary bunion surgery is not re-
commended because of the less-
than-optimal long-term results and
the silicone-related problems that
often occur, such as significant syno-
vitis, osteolysis, and migration of
silicone particles to the regional
lymph nodes.10,11
For the patient with a congruent
joint, the chevron procedure, the dis-
tal soft-tissue procedure, and the
Akin procedure with excision of the
medial eminence are alternatives.
The chevron procedure is probably
the most reliable, particularly in the
patient with a large medial emi-
nence.
Incongruent deformities are classi-
fied as mild (hallux valgus angle less
than 30 degrees and IM angle less
than 13 degrees), moderate (hallux
valgus angle less than 40 degrees and
37
Disorders of the First Metatarsophalangeal Joint

an excellent alternative, but it is tech-

nically somewhat more difficult to
perform, and the osteotomy is not as
stable as that created by the chevron
or more proximal metatarsal
osteotomies.
For advanced moderate and
severe deformities, a distal soft-tis-
sue procedure with a proximal
osteotomy will give a reproducible
satisfactory result in most cases,
although it is technically demand-
ing. The MTP arthrodesis is an
excellent procedure for treating a
severe hallux valgus deformity, par-
ticularly in older patients and those
with rheumatoid arthritis, spasticity,
A B or arthrosis. Occasionally, a Keller
Fig. 5 A, In a congruent joint, the joint surfaces are parallel, and there is no lateral subluxa-
procedure, particularly in the less
tion of the proximal phalanx on the metatarsal head. B, In an incongruent or subluxated joint, active individual, or a double-
the joint surfaces are no longer parallel, and there is lateral subluxation of the proximal pha- stemmed hinged silicone prosthesis
lanx on the metatarsal head.
can be considered.
When there is hypermobility of
the metatarsocuneiform joint,
metatarsocuneiform arthrodesis is
IM angle greater than 13 degrees), deformity, the chevron procedure used along with the complete distal
and severe (hallux valgus angle usually will yield a satisfactory result, soft-tissue procedure. Complica-
greater than 40 degrees and IM angle as will the distal soft-tissue proce- tions of arthrodesis are stiffness of
greater than 20 degrees). As a general dure. However, if the indications are the foot and nonunion.
rule, most patients have a mild to stretched, an incomplete correction The various operative procedures
moderate deformity. For the mild can result. The Mitchell procedure is presented in the algorithm will now

Hallux valgus

Congruent joint Incongruent joint Degenerative joint disease

Akin procedure and Fusion, or

exostectomy, or Keller procedure, or
chevron, or prosthesis
IM angle <13˚ IM angle >13˚ IM angle >20˚ Hypermobile
DSTP
HV angle <30˚ HV angle <40˚ HV angle >40˚ first MC joint

Chevron (age <50 yr), or DSTP with proximal DSTP with proximal Fusion of first MC
DSTP with or without crescentic osteotomy, or crescentic osteotomy, or joint and DSTP
proximal crescentic Mitchell procedure MTP joint fusion
osteotomy, or
Mitchell procedure

Fig. 6 Algorithm for selecting the operative approach to treatment of hallux valgus (HV). DSTP = distal soft-tissue procedure; MC =
metatarsocuneiform. (Adapted with permission from Mann RA: Decision-making in bunion surgery. Instr Course Lect 1990;39:3-13.)

38 Journal of the American Academy of Orthopaedic Surgeons


be discussed in terms of their indica-
tions, contraindications, main compli-
cations, and special technical aspects.
Proximal Phalangeal Osteotomy
(Akin Procedure)12,13
Indications
The indications for this proce-
dure are hallux valgus interpha-
langeus, a congruent joint with a
large medial eminence with a DMA
angle of less than 10 degrees, and
use as a secondary procedure if a
primary procedure (e.g., chevron
or distal soft-tissue procedure) did
not provide sufficient correction
due to a large DMA angle or hallux
valgus interphalangeus.
Contraindications
The contraindications are an
incongruent joint and hallux valgus
with a DMA angle greater than 15
degrees.
Complications
Hallux valgus deformity will fre-
quently increase if the procedure is
attempted on an incongruent joint.13
Technical Aspects
The osteotomy must be distal to
the concavity of the proximal pha-
lanx to prevent violation of the joint.
Comment
The Akin procedure occupies a
small niche in foot surgery. Its
main indication is to treat hallux
valgus interphalangeus or to aug-
ment a chevron or distal soft-tissue
procedure in order to gain complete
correction. It is not recommended
as a primary bunion procedure for
an incongruent joint since the cor-
rection will deteriorate with time.
Chevron Procedure 14
Indications
This procedure is used to treat a
mild deformity (hallux valgus angle
Vol 3, No 1, Jan/Feb 1995
of less than 30 degrees and IM angle
of less than 13 degrees).
Contraindications
The sole contraindication is a
deformity with a hallux valgus
angle greater than 35 degrees and an
IM angle greater than 15 degrees.15
Complications
Possible complications include
incomplete correction if the hallux
valgus deformity is too advanced,
loss of position of the capital frag-
ment secondary to lack of internal
fixation, and avascular necrosis (in
1% to 2% of cases).15-17
Technical Aspects
A medial approach is used to
avoid the dorsal medial cutaneous
nerve (Fig. 7). Soft-tissue stripping
is limited. Internal fixation is pre-
ferred. If the DMA angle is greater
than 10 to 15 degrees, a medial clos-
ing-wedge osteotomy is added to
rotate the articular surface from
excessive lateral deviation into the
normal range.
Comment
The chevron osteotomy is an
excellent procedure for the mild to
moderate hallux valgus deformity.
Complications are most likely to
occur when the indications are
exceeded. Incomplete correction
will occur if the DMA angle is
greater than 10 to 15 degrees and
more bone needs to be removed
from the medial aspect of the prox-
imal metatarsal, thereby creating a
medial closing-wedge effect to cor-
rect the articular surface. This
eliminates the need for the addi-
tion of the Akin procedure to gain
full correction. The medial closing
wedge is achieved by removing an
additional 1 to 3 mm of bone from
the proximal portion of the
osteotomy so that the articular sur-
face can be rotated medially. 18 The
prevalence of avascular necrosis is
Roger A. Mann, MD
probably 1% to 2%, but can be min-
imized by limiting soft-tissue
stripping.
Distal Soft-Tissue Procedure18
Indications
This procedure is recommended
in the treatment of mild hallux val-
gus deformities (hallux valgus angle
less than 30 degrees and IM angle
less than 13 degrees).
Contraindications
The contraindications include (1)
a deformity with a hallux valgus
angle greater than 35 degrees and an
IM angle greater than 15 degrees, (2)
a deformity with a DMA angle
greater than 15 degrees (a chevron
procedure with a medial closing-
wedge osteotomy would be prefer-
able), and (3) arthrosis of the MTP
joint.
Complications
Recurrence of the deformity is
due to failure to include a metatarsal
osteotomy, inadequate lateral
release, or the poor quality of the
medial capsular tissue. Hallux
varus may occur, especially if too
much of the medial eminence is
excised or if the fibular sesamoid is
removed.
Technical Aspects
Two incisions should be used:
one in the first web space and the
other on the medial side of the joint.
The medial eminence should be
excised in line with the medial
aspect of the metatarsal. A
metatarsal osteotomy should be
added if the first and second
metatarsals do not close down eas-
ily at the time of surgery. The post-
operative dressings, which consist
of 2-inch conforming gauze (Kling,
Johnson & Johnson Medical Prod-
ucts, Arlington, Texas) and 0.5-inch
adhesive tape, are used for 8 weeks.
The dressing binds the metatarsal
39
Disorders of the First Metatarsophalangeal Joint
A B
heads together and maintains the
hallux in correct alignment.
Comment
This procedure will give satis-
factory results when used within
its limitations. Failure to add a
metatarsal osteotomy, however,
often results in recurrence. The
soft tissues on the lateral aspect
must be completely released for
the medial plication to be effective.
Excision of too much of the
metatarsal head can result in a hal-
lux varus deformity.
Distal Soft-Tissue Procedure Plus
Proximal Metatarsal Osteotomy5
Indications
The indication for this procedure
is a deformity with a hallux valgus
angle greater than 30 degrees
and an IM angle greater than 13
degrees.
Contraindications
The contraindications are a con-
gruent joint and the presence of
arthrosis.
40
Complications
Excessive lateral deviation of the
metatarsal head will result in hallux
varus. Excessive dorsiflexion of the
osteotomy and excessive shorten-
ing may also occur.
Technical Aspects
I prefer a crescentic osteotomy
carried out through a dorsal inci-
sion, with the concavity directed
proximally toward the heel. Internal
fixation of the osteotomy site is rec-
ommended. The postoperative
dressing requires careful manage-
ment, as described for the distal soft-
tissue procedure.
Comment
The distal soft-tissue procedure
with a proximal osteotomy can be
used for most hallux valgus defor-
mities. A variety of osteotomies
have been advocated, ranging
from a proximal crescentic os-
teotomy to a horizontal chevron
or an opening or closing wedge.
The basic principles of the
osteotomy are to correct the IM
angle with minimal shortening and
Journal of the American Academy of Orthopaedic Surgeons
Fig. 7 Biplanar chevron
osteotomy to correct a
laterally sloping distal
metatarsal articular sur-
face. A, Placement of the
normal chevron osteo-
tomy. B, Removal of
medial bone wedge mea-
suring 1 to 3 mm. This
permits medial devia-
tion of the articular sur-
face of the metatarsal
head, thereby correcting
the DMA angle. C, Dis-
tal articular surface after
correction. The osteo-
tomy site should be fixed
with a pin.
C
to avoid overdisplacement of the
metatarsal head laterally, which
would result in an incongruent
joint and possibly hallux varus
deformity.
Mitchell Procedure19,20
Indications
This procedure can be useful in
treating a deformity with a hallux
valgus angle greater than 30 degrees
and an IM angle greater than 13
degrees.
Contraindications
This procedure is not appropri-
ate in the presence of a congruent
joint or a deformity with a hallux
valgus angle greater than 40
degrees and an IM angle greater
than 20 degrees.
Complications
Loss of position of the capital
fragment can occur due to dorsal
migration, and transfer meta -
tarsalgia can occur secondary to
dorsiflexion of the metatarsal
head. Other possible complica-

tions are avascular necrosis and
nonunion.
Technical Aspects
Soft-tissue stripping should be
minimized. Adequate fixation of the
osteotomy site is necessary.
Comment
The Mitchell procedure will
result in satisfactory correction
when used within its limitations.
However, it is technically more
demanding than the chevron proce-
dure and the distal soft-tissue pro-
cedure with osteotomy, and the
osteotomy site is somewhat less sta-
ble. For these reasons, the proce-
dure has not been used as much in
recent years.
Distal Soft-Tissue Procedure
With Metatarsocuneiform
Fusion6,21
Indications
This procedure is appropriate in
the presence of an unstable metatar-
socuneiform joint or a deformity
with a hallux valgus angle greater
than 35 degrees and an IM angle
greater than 15 degrees.
Contraindications
This procedure is contraindicated
in the presence of a congruent joint.
It is also not appropriate for an ath-
letic younger patient or for a patient
with normal metatarsocuneiform
joint stability.
Complications
Possible complications include
increased stiffness of the foot sec-
ondary to the arthrodesis, nonunion
of the arthrodesis site, and overcor-
rection of the metatarsal head later-
ally, resulting in hallux varus.
Technical Aspects
The metatarsocuneiform joint
should be redirected plantarward
Vol 3, No 1, Jan/Feb 1995
and laterally along the lines of the
normal movement of the articular
surfaces. The arthrodesis is carried
out with the joint in this position,
and the joint is fixed with a screw or
plate. Local bone graft may be used.
The distal soft-tissue procedure is
also carried out to correct the hallux
valgus deformity.
Comment
A metatarsocuneiform joint
arthrodesis is indicated in the 2% to
3% of patients with hallux valgus
who have an unstable metatarso-
cuneiform joint. The procedure is
technically difficult. It also adds to
the stiffness of the foot and should
not, therefore, be overutilized, par-
ticularly in the athletic patient.
Metatarsophalangeal
Arthrodesis22
Indications
This procedure can be useful in
the presence of arthrosis of the MTP
joint or a severe hallux valgus defor-
mity with dislocation of the MTP
joint. It may also be useful in the
patient with spasticity or as a sal-
vage procedure.
Contraindications
Use of this procedure is not appro-
priate for less severe deformities that
can be managed with other proce-
dures.
Complications
Possible complications include
nonunion, malalignment, and de-
generative changes in the interpha-
langeal joint.23
Technical Aspects
Alignment is critical; the joint
should be placed into approximately
15 degrees of dorsiflexion in relation
to the floor, 15 degrees of valgus,
and neutral rotation. Rigid internal
fixation should be used. I prefer the
Roger A. Mann, MD
use of an interfragmentary screw
and a dorsal plate.
Comment
The arthrodesis is an excellent
choice in patients with advanced
arthrosis or rheumatoid arthritis.24 It
is also useful as a salvage procedure.
The long-term results are excellent,
although changes may occur at the
interphalangeal joint, especially if
there is malalignment. The arthrode-
sis is an integral part of the recon-
struction of the rheumatoid forefoot.
Keller Procedure1
Indications
This procedure is most useful in
less active patients with advanced
hallux valgus deformity or with
arthrosis of the MTP joint (as an
alternative to fusion).
Contraindications
This procedure is not appropriate
for active individuals.
Complications
Possible complications include
instability of the medial aspect of
the foot, due to loss of the windlass
mechanism; drifting of the hallux
into varus or valgus rotation or dor-
siflexion; transfer metatarsalgia;
and significant shortening of the
hallux.
Technical Aspects
An attempt should be made to
insert the intrinsic muscles back into
the proximal phalanx to prevent a
cock-up deformity. The joint should
be stabilized with a pin for 3 to 4
weeks to permit scarring around the
MTP joint to occur.
Comment
Because this procedure destabi-
lizes the first MTP joint, it should
not be used for patients who are
very active. There are other proce-
41
Disorders of the First Metatarsophalangeal Joint
dures that will give a more stable
foot with fewer complications. It is
an excellent procedure for the
housebound ambulator or for the
patient who places fewer demands
on the foot.
Juvenile Hallux Valgus
Symptomatic hallux valgus
deformity is uncommon in chil-
dren, but it does occur. Unfortu-
nately, surgical correction of the
juvenile form is associated with
a significant rate of recurrence and
variable clinical outcomes com-
pared with the adult form. 19,25
Most surgeons advocate delay-
ing surgery until skeletal maturity
has been achieved unless an
unusual degree of pain and defor-
mity significantly interferes with
activities.
The evaluation of the patient
with juvenile hallux valgus is
extremely critical. There is a high
prevalence of pes planus and liga-
mentous laxity. There also appears
to be an increased incidence of lat-
eral deviation of the distal articular
surface of the first MTP joint,
which may account for the high
incidence of failure in the juvenile
patient. When there is an unrecog-
nized lateral slope of the distal
articular surface, either the defor-
mity recurs or the joint becomes
stiff despite maintenance of cor-
rection.
When considering treatment for
the patient with juvenile hallux val-
gus, one can follow the same deci-
sion-making precepts based on the
severity of the deformity already
outlined in the algorithm in Figure
6. However, if there is an open
metatarsal epiphysis at the time of
surgery, it should be avoided to
prevent possible growth distur-
bance. One should even be cau-
tious in placing pins across the
epiphysis, since this can theoreti-
cally result in closure or alteration
of its growth.
42
Hallux Rigidus24,26,27
Hallux rigidus is a painful affliction
of the first MTP joint secondary to
arthrosis and is associated with
restriction of dorsiflexion. The con-
dition can occur in adolescence,
although it is uncommon; in those
instances, it is usually associated
with an osteochondritic lesion.28
Clinical Evaluation
The condition usually occurs
insidiously without a history of
trauma. The main complaints are
pain, loss of dorsiflexion, and
increased bulk of the joint, which
makes shoe wearing difficult. The
patient’s level of activity must be
carefully evaluated, particularly the
limitations that have been necessi-
tated by the condition.
Physical examination reveals
increased bulk of the joint and loss
of dorsiflexion, which should be
quantified. Marginal osteophytes
are typically present dorsally and
laterally. Forced dorsiflexion will
usually reproduce the patient’s pain,
as will lateral deviation if a lateral
osteophyte is present. Often, the
dorsal medial cutaneous nerve is
sensitive.
The radiographic evaluation
includes weight-bearing anteropos-
terior, lateral, and oblique views.
Bone proliferation along the lateral
aspect of the joint is evaluated on the
anteroposterior radiograph; that
along the dorsal aspect, on the lateral
radiograph. The extent of joint nar-
rowing is determined from the
oblique radiograph.
Conservative management con-
sists of use of a shoe with adequate
width and depth to accommodate
the increased bulk of the joint and
with a rigid rocker sole to diminish
joint motion. If there is significant
bone proliferation or pain with dor-
siflexion, a cheilectomy or debride-
ment of the MTP joint should be
considered. It is important to ex-
Journal of the American Academy of Orthopaedic Surgeons
plain the expected outcome before
surgery, since an arthritic joint will
still be present, which may be symp-
tomatic when stressed.
I have observed that patients with a
relatively mild to moderate degree of
restriction of dorsiflexion tend to do
very well with a cheilectomy, which I
believe is due to reestablishing some
of the normal gliding that occurs at
the MTP joint. More severe and
advanced degenerative changes, par-
ticularly marked osteophytes, are
associated with less certain outcomes,
and a certain degree of residual dis-
comfort can be expected. The alterna-
tive is to carry out an arthrodesis of
the joint, which will eliminate the pain
and permit a return to most activities
at a nonprofessional level.
The Keller procedure or an
implant arthroplasty can be consid-
ered in less active individuals with
significant arthrosis, although the
long-term results of these procedures
do not compare favorably with those
of cheilectomy or fusion.11
Operative Procedures
Cheilectomy is carried out through
a dorsal approach, with the incision
being carried down through the
extensor hood on either side of the
MTP joint. The capsule is opened,
and a complete synovectomy is car-
ried out. Proliferative bone is
removed along the lateral side of the
metatarsal head, in line with the long
axis of the metatarsal, and over the
dorsal aspect by removing 20% to
30% of the metatarsal head. Dorsal
bone must be removed until approx-
imately 60 to 70 degrees of dorsiflex-
ion at the MTP joint has been
achieved. If less bone is removed,
dorsal impingement will persist, and
the patient will usually not be satis-
fied with the result. The average
increase in dorsiflexion is about 25
degrees, but it varies considerably.
The main benefit is relief of the dorsal
impingement, which is the main
source of pain.

The patient wears a postoperative
shoe until the wound is healed (at
about 10 days), after which active
range of motion is encouraged. As a
general rule, maximum improve-
ment will occur by approximately 3
to 4 months. The advantage of
cheilectomy is that if the procedure
fails, one can still carry out a fusion,
possibly a Keller procedure or
implant arthroplasty.
Occasionally, a patient with hal-
lux rigidus still does not have ade-
quate dorsiflexion after surgery,
which may be a source of pain.
Under these circumstances, one
References
1. Keller WL: The surgical treatment of
bunions and hallux valgus. N Y Med J
1904;80:741-742.
2. Silver D: The operative treatment of hal-
lux valgus. J Bone Joint Surg 1923;5:
225-232.
3. McBride ED: The conservative opera-
tion for “bunions”: End results and
refinements of technic. JAMA 1935;105:
1164-1168.
4. DuVries HL: Surgery of the Foot. St
Louis: CV Mosby, 1959, p 381.
5. Mann RA, Rudicel S, Graves SC: Repair
of hallux valgus with a distal soft-tissue
procedure and proximal metatarsal
osteotomy: A long-term followup. J
Bone Joint Surg Am 1992;74:124-129.
6. Sangeorzan BJ, Hansen ST Jr: Modified
Lapidus procedure for hallux valgus.
Foot Ankle 1989;9:262-266.
7. Mann RA, Katcherian DA: Relationship
of metatarsophalangeal joint fusion on
the intermetatarsal angle. Foot Ankle
1989;10:8-11.
8. Richardson EG, Graves SC, McClure JT,
et al: First metatarsal head-shaft angle:
A method of determination. Foot Ankle
1993;14:181-185.
9. Piggott H: The natural history of hal-
lux valgus in adolescence and early
adult life. J Bone Joint Surg Br 1960;42:
749-760.
10. Granberry WM, Noble PC, Bishop JO, et
al: Use of a hinged silicone prosthesis
for replacement arthroplasty of the first
metatarsophalangeal joint. J Bone Joint
Surg Am 1991;73:1453-1459.
Vol 3, No 1, Jan/Feb 1995
might consider doing a dorsal clos-
ing-wedge osteotomy of the proxi-
mal phalanx (Moberg procedure),29
which makes use of the fact that
plantar flexion can still occur at the
MTP joint. With this procedure, one
is able to gain approximately 25
degrees of dorsiflexion, which
decreases the stress on the MTP
joint.
As a general rule, proximal pha-
langeal osteotomy should not be car-
ried out at the same time as
cheilectomy, since the cheilectomy
requires early motion and the proxi-
mal phalangeal osteotomy requires
11. Shereff MJ, Jahss MH: Complications of
Silastic implant arthroplasty in the hal-
lux. Foot Ankle 1980;1:95-101.
12. Akin OF: The treatment of hallux val-
gus: A new operative procedure and its
results. Med Sentinel 1925;33:678-679.
13. Plattner PF, Van Manen JW: Results of
Akin type proximal phalangeal osteotomy
for correction of hallux valgus deformity.
Orthopedics 1990;13:989-996.
14. Austin DW, Leventen EO: A new
osteotomy for hallux valgus: A horizon-
tally directed “V” displacement osteotomy
of the metatarsal head for hallux valgus
and primus varus. Clin Orthop 1981;
157:25-30.
15. Hattrup SJ, Johnson KA: Chevron
osteotomy: Analysis of factors in
patients’ dissatisfaction. Foot Ankle
1985;5:327-332.
16. Meier PJ, Kenzora JE: The risks and ben-
efits of distal first metatarsal osteotomies.
Foot Ankle 1985;6:7-17.
17. Johnson JE, Clanton TO, Baxter DE, et al:
Comparison of chevron osteotomy and
modified McBride bunionectomy for cor-
rection of mild to moderate hallux valgus
deformity. Foot Ankle 1991;12:61-68.
18. Mann RA, Coughlin MJ: Adult hallux
valgus, in Mann RA, Coughlin MJ (eds):
Surgery of the Foot and Ankle, 6th ed. St
Louis: CV Mosby, 1993, vol 1, pp 204-216.
19. Canale PB, Aronsson DD, Lamont RL, et
al: The Mitchell procedure for the treat-
ment of adolescent hallux valgus: A
long-term study. J Bone Joint Surg Am
1993;75:1610-1618.
Roger A. Mann, MD
a period of immobilization for ade-
quate healing.
Summary
Satisfactory bunion surgery is
predicated on the integration of the
patient’s chief complaints, the find-
ings from a thorough physical and
radiographic evaluation, and selec-
tion of the appropriate surgical
procedure. The postoperative
management must be meticulously
carried out to ensure optimal align-
ment of the hallux.
20. Hawkins FB, Mitchell CL, Hedrick DW:
Correction of hallux valgus by metatarsal
osteotomy. J Bone Joint Surg Am 1945;7:
387-394.
21. Mauldin DM, Sanders M, Whitmer
WW: Correction of hallux valgus with
metatarsocuneiform stabilization. Foot
Ankle 1990;11:59-66.
22. Coughlin MJ: Arthrodesis of the first
metatarsophalangeal joint with mini-
fragment plate fixation. Orthopedics
1990;13:1037-1044.
23. Fitzgerald JAW: A review of long term
results of arthrodesis of the first
metatarso-phalangeal joint. J Bone Joint
Surg Br 1969;51:488-493.
24. Thompson FM, Mann RA: Arthritides,
in Mann RA, Coughlin MJ (eds):
Surgery of the Foot and Ankle, 6th ed. St
Louis: CV Mosby, 1993, vol 1, pp 622-
634.
25. Coughlin MJ: Juvenile bunions, in
Mann RA, Coughlin MJ (eds): Surgery of
the Foot and Ankle, 6th ed. St Louis: CV
Mosby, 1993, vol 1, pp 297-339.
26. Mann RA, Clanton TO: Hallux rigidus:
Treatment by cheilectomy. J Bone Joint
Surg Am 1988;70:400-406.
27. Mann RA, Coughlin MJ, DuVries HL:
Hallux rigidus: A review of the litera-
ture and a method of treatment. Clin
Orthop 1979;142:57-63.
28. Kessel L, Bonney G: Hallux rigidus in
the adolescent. J Bone Joint Surg Br
1958;40:668-673.
29. Moberg E: A simple operation for hal-
lux rigidus. Clin Orthop 1979;142:55-56.
43
 Flexor Tendon Injuries: I. Foundations of Treatment

James W. Strickland, MD

Abstract

During the past 20 years, the difficult process of reestablishing satisfactory function face usually has a layer of uniform col-
after primary repair of flexor tendons has evolved from scientifically unsupported lagen fibers, and elastin and individ-
trial-and-error efforts to protocols based on sound laboratory and clinical investiga- ual fascicles are capable of sliding past
tions. Enhanced appreciation of tendon structure, nutrition, and biomechanical each other with no apparent direct
properties and investigation of factors involved in tendon healing and adhesion for- attachments or cellular communica-
mation have had significant clinical applications. In particular, it has been found that tions. The surface of the individual
repaired tendons subjected to early motion stress will increase in strength more bundles of collagen is covered by the
rapidly and develop fewer adhesions than immobilized repairs. As a result, new and endotenon and a fine fibrous and cel-
stronger tendon repair techniques have evolved, permitting the application of early lular outer layer, the epitenon, which
passive and even light active forces. The author reviews the most recent and clini- is continuous with the endotenon. In
cally pertinent research in flexor tendon surgery and discusses repair techniques and the hand, flexor tendon fascicles are
rehabilitation protocols based on the information provided by these studies. covered by a thin visceral and parietal
J Am Acad Orthop Surg 1995;3:44-54 adventitia called the paratenon, which
contains fluid similar to synovial fluid.
In the digits, the flexor tendons are
enclosed in sheaths lined by visceral
The return of satisfactory digital dons, their function, the biomechanics and parietal synovial layers.1 The A2
motion after severance of the flexor of their action at the joints they move, and A4 annular pulleys arise from the
tendons has long challenged sur- and their biologic response to injury periosteum of the proximal half of the
geons dealing with upper extremity and repair. These investigative efforts proximal phalanx and the midportion
trauma. Flexor lacerations in the fin- have given rise to improved methods of the middle phalanx, respectively
ger were once found to perform so of tendon repair, a greater emphasis on (Fig. 1). The A1, A3, and A5 pulleys are
poorly after primary repair that the flexor sheath preservation and restora- joint pulleys that arise successively
digital sheath was referred to as a sur- tion, and an emphasis on the early from the palmar plates of the metacar-
gical no-man’s-land. Surgeons were application of stress as a means of pophalangeal, proximal interpha-
advised not to repair tendon injuries more rapidly increasing the strength langeal, and distal interphalangeal
in this zone and to resort to free-ten- and gliding capacity of repaired ten- joints. The palmar aponeurosis pulley
don grafting at a later date to achieve dons. This article will review some of is composed of the transverse and ver-
the best results. While that advice has the most recent and relevant research
now been discredited and primary and discuss the current clinical
flexor tendon repair has been the approaches to flexor tendon repair that Dr. Strickland is Clinical Professor of
accepted procedure for over two have resulted from these efforts. Orthopaedic Surgery, Indiana University School
of Medicine, Indianapolis; Chairman, Depart-
decades, the techniques for repair
ment of Hand Surgery, St. Vincent Hospital,
and the postoperative management Indianapolis; and Senior Staff Surgeon, Indiana
programs have varied greatly from Basic Science of Flexor Hand Center, Indianapolis.
surgeon to surgeon. Most clinical Tendons
approaches have been based to a Reprint requests: Dr. Strickland, PO Box 80434,
8501 Harcourt Road, Indianapolis, IN 46280-
large extent on individual experience Structure and Anatomy
0434.
with little or no scientific support. Tendons are composed of fascicles
Considerable research has been con- of long, narrow, spiraling bundles of Copyright 1995 by the American Academy of
ducted in recent years in an effort to mature fibroblasts (tenocytes) and Orthopaedic Surgeons.
better understand the structure of ten- type I collagen fibers. The fascicle sur-

44 Journal of the American Academy of Orthopaedic Surgeons


tical fibers of the palmar fascia and is
clinically important when other prox-
imal components of the sheath have
been lost. The thin, condensable cru-
ciate sections of the sheath—C 1
(between A2 and A3 ), C2 (between A3
and A 4), and C 3 (between A 4 and
A5)—collapse to permit the annular
pulleys to approximate each other
during digital flexion. The flexor ten-
dons are weakly attached to the sheath
by filmy mesenteries composed of
vincula.
Flexor tendons are oval. The
flexor digitorum superficialis (FDS)
tendons usually arise from single
muscle bundles and act indepen-
dently. There is often a common
muscle origin for several flexor digi-
torum profundus (FDP) tendons,
with the result that there is simulta-
neous flexion of multiple digits. The
FDS tendons lie on the palmar side
of the FDP tendons until they enter
the A1 entrance of the digital sheath.
Within the proximal sheath, the FDS
tendon divides into two slips that
wrap around the FDP tendons;
rejoin dorsally by means of fibers
referred to as the chiasma tendinum,
or Camper’s chiasma; and terminate
as they insert along the proximal half
of the middle phalanx (Fig. 2). The
FDP tendons pass through the FDS
bifurcation to insert into the proxi-
mal aspect of the distal phalanges.
Vol 3, No 1, Jan/Feb 1995
Fig. 1 Lateral (top) and
dorsal (bottom) views of a
finger depict the compo-
nents of the digital flexor
sheath. The sturdy annular
pulleys (A1, A2, A3, A4, and
A 5 ) are important biome-
chanically in keeping the
tendons closely applied to
the phalanges. The thin, pli-
able cruciate pulleys (C1, C2,
and C3) collapse to allow full
digital flexion. A recent
addition is the palmar
aponeurosis pulley (PA),
which adds to the biome-
chanical efficiency of the
sheath system.
The FDP muscle acts as the pri-
mary digital flexor, while the FDS
and intrinsic muscles combine for
forceful flexion. Digital balance and
equilibrium during flexion and
extension require a complex integra-
tion of extrinsic and intrinsic activ-
ity. Forces of 200 N can be achieved
during power grip.
Nutrition
The vascular perfusion of the
flexor tendons includes longitudinal
vessels, which enter in the palm and
extend down intratendinous chan-
nels; vessels that enter at the level of
the proximal synovial fold in the
palm; segmental branches from the
paired digital arteries, which enter
in the tendon sheaths by means of
the long and short vincula; and ves-
sels that enter the FDS and FDP ten-
dons at their osseous insertions (Fig.
3). Both tendons have relatively
avascular segments over the proxi-
mal phalanx. The FDP tendon has
an additional short avascular zone
over the middle phalanx. Fortu-
nately, synovial fluid diffusion pro-
vides an effective alternative
nutritional and lubricating pathway
for flexor tendons. The rapid deliv-
ery of nutrients is apparently accom-
plished by a pumping mechanism
known as imbibition, in which fluid
is forced into the interstices of the
James W. Strickland, MD
tendon through small ridges, or con-
duits, in the tendon surface as the
digit is flexed and extended.
Biomechanical Properties
As much as 9 cm of flexor tendon
excursion may be required to pro-
duce composite wrist and digital
flexion, while only about 2.5 cm of
excursion is required for full digital
flexion with the wrist stabilized in a
neutral position (Fig. 4). The greater
the distance a tendon is from the
axis of joint rotation, the greater the
moment arm and the less motion
that a given muscle contraction will
generate at that joint. Conversely, a
shorter moment arm will result in
more joint rotation from the same
tendon excursion. The moment
arm, excursion, and joint rotation
produced by the flexor tendons are
governed by the constraint of the
pulley system. Loss of portions of
the digital pulleys may significantly
alter the normal integrated balance
between the flexor, intrinsic, and
extensor tendons. The A2 and A4
pulleys are the most important to
these mechanical functions; the loss
of a substantial portion of either
may diminish digital motion and
power or lead to flexion contrac-
tures of the interphalangeal joints
(Fig. 5).2,3
Several studies have been carried
out to determine the tendon forces
generated by various active and pas-
sive functions. Despite variation in
the methodology used in these inves-
Fig. 2 Early in the flexor sheath, the FDS
tendon divides and passes around the FDP
tendon. The two portions of the FDS tendon
reunite at Camper’s chiasma.
45
Flexor Tendon Injuries: Treatment Principles

on factors that relate to the injury and

the surgical repair. In the clinical set-
ting, it is impossible to isolate the two
types of healing, and the cellular
events can be viewed as similar for all
flexor tendons. After repair, tendon
healing involves an inflammatory
phase that lasts 48 to 72 hours, a
fibroblast- or collagen-producing
phase that lasts from 5 days to 4
weeks, and a remodeling phase that
continues for about 112 days. During
the inflammatory phase of tendon
healing, the strength of the repair is
almost entirely reliant on the strength
of the suture itself, with a modest con-
Fig. 3 Blood supply to the flexor tendons within the digital sheath. The segmental vascu- tribution from the fibrin and the clot
lar supply to the flexor tendons is by means of long and short vincular connections. The vin- between the tendon ends. Strength
culum brevis superficialis (VBS) and the vinculum brevis profundus (VBP) consist of small
triangular mesenteries near the insertion of the FDS and FDP tendons. The vinculum increases rapidly during the fibro-
longum to the superficialis tendon (VLS) arises from the floor of the digital sheath and the blast- and collagen-producing phase,
proximal phalanx. The vinculum longum to the profundus tendon (VLP) arises from the when granulation tissue is forming in
superficialis at the level of the proximal interphalangeal joint. The cutaway view depicts the
relative avascularity of the palmar side of the flexor tendons in zones I and II compared with the defect. When extrinsic healing
the richer blood supply on the dorsal side, which connects with the vincula. predominates, adhesions between the
tendon and its surrounding tissues
are inevitable. Healing that is largely
based on intrinsic cellular activity will
tigations, it appears that during Tendon Healing result in fewer, less dense adhesions.
unresisted passive flexion, flexor ten- After much debate, almost all In a series of definitive laboratory
dons are subjected to 2 to 4 N of force. investigators now believe that ten- experiments, Gelberman et al 6
Active flexion with mild resistance dons have both an intrinsic and an demonstrated that, compared with
may result in up to 10 N of force; extrinsic capability to heal. The rela- immobilization, the application of
moderate resisted flexion, up to 17 N; tive contribution of each will depend early passive motion stress to
and strong composite grasp, up to 70
N. Firm tip pinch can apparently
generate as much as 120 N of tensile
load on the index FDP tendon. 4,5
Forces produced by the FDS tendon
have been shown to be considerably
less than those produced by the FDP
tendon during grasp and pinch.
From a clinical perspective, it should
be remembered that these loads are
substantially increased by the resis-
tance created by stiffness and swelling
of the finger and by the increased drag
that a healing tendon may experience
within its sheath. Pressure between
the pulleys and the flexor tendons may
reach as high as 700 mg Hg during
active flexion, which perhaps explains
Fig. 4 Approximate flexor tendon excursion (measured in millimeters) necessary to pro-
the histologic alterations to the fibro- duce full flexion of digital joints at the forearm, wrist, hand, and digital levels. P = profun-
cartilagelike tissue in tendons beneath dus; S = superficialis.
annular digital pulleys.

46 Journal of the American Academy of Orthopaedic Surgeons


repaired canine tendons led to a
more rapid recovery of tensile
strength, fewer adhesions, improved
excursion, better nutrition, and min-
imal repair-site deformation. They
concluded that passive mobilization
enhances healing by simultaneously
stimulating maturation of the tendon
wound and remodeling of the ten-
don scar. From these studies and the
work of many others, it appears that
the most effective method of restor-
ing strength and excursion to
repaired tendons involves the use of
a strong, gap-resistant suture tech-
nique followed by the application of
controlled-motion stress.
Considerable research is being
conducted in an effort to understand
Vol 3, No 1, Jan/Feb 1995
Fig. 5 Function of the fin-
ger flexor tendon pulley sys-
tem. Top, The arrangement
of the annular (A1, A2, A3,
A4, and A5) and cruciate (C1,
C2, and C3) synovial pulleys
of the finger flexor tendon
sheath within the intact
fibro-osseous canal and the
normal moment arm (MA)
and FDP tendon excursion
as the proximal interpha-
langeal joint is flexed to 90
degrees. Bottom, The bio-
mechanical alteration that
results from excision of the
distal half of the A2 pulley;
the C1, A3, and C2 pulleys;
and the proximal portion of
the A4 pulley. The distance
between the distal edge of
the A2 pulley and the proxi-
mal edge of the A4 pulley is
the intra-annular pulley dis-
tance (IAPD). The moment
arm is increased and a
greater FDP tendon excur-
sion is required to produce
90 degrees of flexion
because of the bowstringing
that results from the loss of
pulley support.
the influence that soluble polypep-
tides, including mitogens (growth
factors and hormones) and chemo-
tactic and differentiating factors
(e.g., fibronectin), exert on the cellu-
lar sequence of tendon repair. These
factors have been shown to play a
role in both normal and pathologic
processes. Continuing investiga-
tions may, in time, modify current
repair methods and postrepair
motion protocols for severed flexor
tendons.6,7
Adhesion Formation and
Control
Factors that influence the for-
mation of excursion-restricting
adhesions around repaired flexor
James W. Strickland, MD
tendons include trauma to the ten-
don and its sheath from the initial
injury and the reparative surgery,
tendon ischemia, tendon immobi-
lization, gapping at the repair site,
and excision of components of the
tendon sheath. Quantitative mea-
sures have shown that adhesions
form in proportion to the amount of
tissue crushing and the number of
surface injuries to the tendon. Dis-
ruption of the vincula also has been
associated with a decrease in the
recovery of tendon excursion. There
continues to be considerable debate
about whether primary repair of the
digital sheath is favorable for adhe-
sion reduction.8,9
Various biochemical agents have
been studied in an attempt to modify
adhesion formation around tendon
repairs. Anabolic steroids, antihista-
mines, and nonsteroidal anti-inflam-
matory drugs have been subjected to
recent laboratory investigations, and
there has been some evidence that
ibuprofen 10 and indomethacin 11
may improve tendon excursion by
blocking prostaglandin synthesis
through the inhibition of the enzyme
cyclo-oxygenase at the cellular level.
While hyaluronate appears to
reduce adhesions around healing
tendons, it was found to have no sta-
tistically significant effect on digital
motion in one double-blind study.
Flexor Tendon Repair
Indications and
Contraindications
Primary repair of flexor tendons
severed in the digital sheath has now
universally replaced the “no-man’s-
land” concept, which favored sec-
ondary grafting. The concept that
flexor tendon repair should be con-
sidered a surgical emergency has
also been effectively discredited by
several studies that demonstrate
that equal or better results can usu-
ally be achieved by delayed primary
47
Flexor Tendon Injuries: Treatment Principles
flexor tendon suture. It has also
been shown that it is better to repair
both the FDP and the FDS tendons
rather than the FDP tendon alone,
which was once the preferred
option.
There are a number of important
contraindications to primary suture
of severed flexor tendons, including
severe multiple tissue injuries to the
fingers or palm, wound contamina-
tion by potentially infecting materi-
als, and significant skin loss over the
flexor system. Concomitant frac-
tures and neurovascular injuries are
not necessarily contraindications to
primary or delayed primary suture.
If the fracture can be anatomically
reduced and adequately stabilized,
it is almost always better to proceed
with flexor tendon repair and
microscopic nerve and vessel
suture, recognizing that the ultimate
results after combined tissue
injuries are not as good as those fol-
lowing tendon severance with no
associated injuries. Rejoining the
tendon at its normal length acutely
or subacutely is usually preferable
to delaying the repair for several
weeks, because of the inevitable
deterioration of the tendon ends and
shortening of the extrinsic muscle-
tendon system.
Examination and Preparation
The surgeon must carefully exam-
ine the patient’s hand to determine
the total extent of the injury. Alter-
ations in the normal resting posture
of the digits will help identify the loss
of continuity of one or both flexor
tendons, and well-known functional
tests will confirm the loss of FDP
and/or FDS action. Lacerations on
the palmar aspect of the fingers will
almost always injure the FDP tendon
before severing the FDS tendon, but
the absence of FDP function alone
does not rule out the possibility of a
near-complete FDS division.
A careful sensory evaluation of
the palmar aspect of the distal pha-
48
lanx will allow the identification of
injuries to the digital nerves and will
provide important information for
planning incisions for their exposure
and repair. A deep wound with lac-
erations of both digital nerves
almost surely indicates division of
the digital arteries as well. While the
digit will probably survive the loss
of both vessels, the viability of the
skin flaps used for exposure may be
in jeopardy. In addition, digital
ischemia may impair tendon and
nerve healing and result in severe
cold intolerance. It is important that
one or both digital arteries be
repaired in these complex injuries.
Surgical Considerations
The techniques of flexor tendon
repair and the protocols for postoper-
ative mobilization of the repaired ten-
don were, for many years, based on
anecdotes and hearsay. In retrospect,
publications on this topic often had
deficiencies in scientific methodol-
ogy. In recent years, various hand,
orthopaedic, and plastic surgery jour-
nals have published a plethora of lab-
oratory and clinical information
about methods of flexor tendon repair
and postrepair motion protocols.
These investigations stem from the
consensus that the greater the incre-
ments of repair-site stress and tendon
excursion, the faster the tendon will
achieve normal tensile strength with
fewer motion-restricting adhesions.
Trying to interpret these reports and
compare them with those from other
studies is almost impossible given the
different laboratory models used
(e.g., in vivo versus in vitro), different
testing methods, and diverse defini-
tions of failure. A thorough review of
this information does, however, per-
mit a few reasonably supportable
conclusions on which to base a clinical
protocol designed to attain the best
possible digital performance after
flexor tendon division.
While most in vivo studies of lab-
oratory animal tendon repairs sug-
Journal of the American Academy of Orthopaedic Surgeons
gest that there is a 10% to 50% loss of
the initial repair strength during the
first 5 to 21 days following injury, it
should be recognized that these
studies were carried out on immobi-
lized tendons. Some recent investi-
gations of tendon repairs in which
controlled passive motion was used
indicate that this drop may be sub-
stantially lessened by early stress
application. By converting newtons
to grams, it is possible to establish
some working numbers that allow
the determination of the strength of
various tendon repairs at the time of
surgery and throughout the healing
period. These data can then be
matched with the stress forces of
postrepair motion protocols to
determine the relative risk of tendon
rupture with each. Conservative
working numbers for tensile
demands on a normal tendon in an
unswollen finger can be estimated
as follows: passive motion, 500 g;
light grip, 1,500 g; strong grip, 5,000
g; and index finger tip pinch (FDP
tendon), 9,000 g. Corresponding
values for a finger that has under-
gone an FDS repair can be calcu-
lated as 15% to 30% of these normal
values.
Suture Techniques
Core Sutures
Numerous methods of tendon
suture (Fig. 6) have been advocated
in an effort to satisfy the six charac-
teristics of an ideal repair: (1) easy
placement of sutures in the tendon,
(2) secure suture knots, (3) smooth
juncture of tendon ends, (4) minimal
gapping at the repair site, (5) mini-
mal interference with tendon vascu-
larity, and (6) sufficient strength
throughout healing to permit the
application of early motion stress to
the tendon.
A comparison of a number of pub-
lished investigations of the charac-
teristics and performance of various
 James W. Strickland, MD

flexor tendon repairs leads to the

following general conclusions: (1)
The strength of a flexor tendon
repair is roughly proportional to the
number of suture strands that cross
the repair site. (2) Locking loops
contribute little strength to the
repair and may actually collapse
A
and lead to gapping at moderate
loads. (3) Repairs usually rupture at
the suture knots. (4) Synthetic 3-0 or
4-0 braided sutures are probably the
best for flexor tendon repair.
The observation that the number
E of suture strands crossing the
repair will determine the strength
B of the repair has been best demon-
strated by Savage, 12 who found
that a complex six-strand repair
was three times stronger than a
two-strand repair, and by three
recently published reports,13-15 in
which four-strand repairs (Fig. 7)
were found to have approximately
twice the strength of two-strand
repairs in vitro. From many repair
studies, it is possible to conserva-
tively list the initial strength of
two-, four-, and six-strand flexor
F

Fig. 6 Commonly used techniques for end-to-end tendon suture. A, Conventional Bunnell Fig. 7 Types of four-strand flexor tendon
stitch. B, Crisscross stitch. C, Mason-Allen (Chicago) stitch. D, Becker bevel (overlapped) repairs. Top, Robertson and Al-Qattan
repair. E, Kessler grasping stitch. F, Modified Kessler stitch with single knot at the repair interlock stitch. Bottom, Lee double-loop
site. G, Tajima modification of the Kessler stitch with double knots at the repair site. locking suture.

Vol 3, No 1, Jan/Feb 1995 49

Flexor Tendon Injuries: Treatment Principles
tendon repairs and predict their
strength at 1 week (–50%), 3 weeks
(–331⁄3%) and 6 weeks (+20%). This
information, based on results in
unstressed in vivo studies, can then
be plotted against the stress forces
of passive-light and moderate-
strong active motion to determine
the relative safety of each method
(Fig. 8, Table 1). From this assess-
ment, it appears that only the six-
strand repair can be considered safe
from rupture throughout the entire
period of unstressed healing.
Unfortunately, six-strand repairs,
such as those described by Sav-
age,12 are technically difficult and
may damage the tendon excessively
or compromise its nutrition or abil-
ity to heal.
Fig. 8 Estimated strength to failure (measured in grams) for two-, four-, and six-strand
flexor tendon repairs performed without the use of epitendinous sutures.
Table 1
Estimated Repair Strength Without Epitendinous Suture
Type of 0 1
Repair Weeks
Two-strand 1,800 g
Four-strand 3,600 g
Six-strand 5,400 g
50
Peripheral Epitendinous Sutures
Several studies have indicated
that gapping at the repair site
becomes the weakest part of the
tendon, unfavorably alters tendon
mechanics, and may attract adhe-
sions, resulting in decreased ten-
don excursion. The importance of
the use of a peripheral circumfer-
ential epitendinous suture at the
completion of a tendon repair has
been demonstrated by the obser-
vation that such sutures may pro-
vide a 10% to 50% increase in
flexor tendon repair strength
accompanied by a significant
reduction in gapping between the
tendon ends. These benefits have
been further confirmed by experi-
ments that apply cyclic loads to the
3 6
Week Weeks Weeks
900 g 1,200 g 2,200 g
1,800 g 2,400 g 4,200 g
2,700 g 3,600 g 6,500 g
Journal of the American Academy of Orthopaedic Surgeons
tendon repair. The running-lock
stitch and horizontal-mattress
epitenon/intrafiber methods have
been shown to be the strongest of
the peripheral suture techniques
(Fig. 9). 16-18 The gap-retarding
quality of these peripheral epi-
tendinous sutures is particularly
important in light of the finding
that gapping of tagged flexor ten-
don repairs is associated with
poorer clinical results.
On the basis of data from pub-
lished reports,16,17 it appears to be a
safe assumption that a horizontal-
mattress or running-lock peripheral
circumferential epitendinous
suture will increase the strength of
two-strand repairs by at least 40%,
or about 700 g of repair strength to
each core suture repair, and that the
improved strength will be main-
tained throughout the healing
period. The addition of this 700 g of
repair strength to the values
already presented for two-, four-,
and six-strand repairs performed
without epitendinous sutures
demonstrates the increased safety
that can be obtained (Fig. 10, Table
2). The safety of a four-strand core-
stitch repair combined with a run-
ning-lock or horizontal-mattress
circumferential epitendinous stitch
should permit even light composite
grip during the entire healing
period.
Sheath Repair
In recent years, many surgeons
have advocated repair of the flexor
tendon sheath after tendon suture.
The stated advantages of sheath
repair are that it would serve as a bar-
rier to the formation of extrinsic
adhesions, should provide a quicker
return of synovial nutrition, would
act as a mold for the remodeling ten-
don, and should result in better ten-
don-sheath biomechanics. Two
disadvantages are that sheath repair
is often technically difficult and that
 James W. Strickland, MD

the repaired sheath may narrow, Rehabilitation further loading results in tendon
restricting tendon gliding. There breakage. Because the strain resulting
have been a number of conflicting Theory of Early Postrepair from the application of a small force to
laboratory and clinical studies Motion Stress a tendon is probably the result of
regarding the biologic and biome- Splints and exercise programs are changes in the restricting scar, it
chanical benefits of sheath repair, and now routinely used early in the appears to be biologically effective to
no clear-cut advantage has yet been postrepair period in an effort to assist impart small but frequent forces in
established. A number of autoge- the functional recovery by influenc- opposite directions in an effort to
nous and synthetic materials have ing the biologic process of collagen modify and elongate restrictive ten-
been employed to restore sheath con- synthesis and degradation.21 Favor- don adhesions.6,22,23 Although some
tinuity, including tendon, fascia, able remodeling of the scar around a excellent research is being carried out,
extensor retinaculum, peritenon, healing tendon is best accomplished there is still inadequate documenta-
veins, silicone sheeting, and polyte- by applying stress to the tendon, tion of how much stress is appropri-
trafluoroethylene, but these methods which in turn transmits stress to the ate, the optimum duration and
are rarely required in the acute set- adjacent scar. Small loads result in frequency of stress application, and
ting. significant elongation of tendons. As the most advantageous methods for
the load increases, the percentage of the delivery of that stress to a finger
Suture Materials elongation rapidly decreases until after a tendon injury.
Efforts have been made to deter-
mine which are the best tendon suture
materials. A polyfilament ensheathed
by caprolactam was found to be the
strongest by one investigator. 19
Absorbable sutures developed for ten-
don repair seem advantageous
because of low long-term foreign-
body tissue reaction and reduction of
the stress-shielding effects of the host
tissue. Unfortunately, the optimal
rates of material absorption and
strength reduction have yet to be
determined. In actual practice, 3-0 and A C
4-0 braided polyester sutures are the
most commonly employed because of
their ease of placement, adequate
strength, and minimal elasticity.

Partial Tendon Lacerations

There has been debate regarding
the appropriate management of par- D
tial tendon lacerations. Initial inves-
tigations created considerable
controversy because they recom-
mended that partial flexor tendon
lacerations should not be repaired.
Recent studies have demonstrated
that partial lacerations of 60% or less
B E
need not be sutured, but that those
greater than 60% should be Fig. 9 Peripheral epitendinous suture techniques. A, Simple running stitch. B, Running-
repaired.20 The possibility of entrap- lock loop (Lin et al). C, Halsted continuous horizontal-mattress suture (Wade). D, Horizon-
tal-mattress intrafiber suture (Mashadi and Amis). E, Running-lock suture (Indiana). The
ment, rupture, and triggering of running-lock loop suture, the Halsted continuous horizontal-mattress suture, and the hori-
unrepaired partial tendon lacera- zontal-mattress intrafiber suture have been shown to be the strongest.
tions has also been reported.

Vol 3, No 1, Jan/Feb 1995 51

Flexor Tendon Injuries: Treatment Principles

Numerous techniques and modifi- imal interphalangeal joint motion added (Fig. 11, Table 3). Differential
cations of techniques have been retained about 1.3 mm (90%) of FDS excursion between the two digital
advanced in an effort to mechanically and FDP excursion per 10 degrees of flexors was also increased dramati-
alter the normal biologic sequence of flexion.24 cally by use of the synergistic splint.
tissue healing and to modify the for- The amount of tendon excursion It has been demonstrated that if an
mation of adhesions around a tendon that should occur for uninjured ten- active motion protocol is selected, the
repair. Applying early postrepair dons in the original Kleinert splint, wrist should be at 45 degrees of
motion stress to flexor tendon repairs modifications of the Kleinert splint extension with the metacarpopha-
has been shown to be beneficial for with a palmar-bar pulley (Brooke langeal joints flexed to 90 degrees in
more rapid recovery of tensile Army Splint), and an experimental order to minimize the force required
strength, fewer adhesions, improved “synergistic” dynamic tenodesis to achieve or hold full active compos-
tendon excursion, and minimal splint that permits wrist extension ite digital flexion.26
repair-site deformation in a canine (Mayo Clinic splint) has also been
model.6 The load at failure of imme- studied.25 The results demonstrate Practical Early Postrepair
diately mobilized tendons tested at 3 that improved excursion can be Motion Stress Protocols
weeks was twice that of immobilized expected from the use of a palmar bar On the basis of this information, the
tendons, while the linear slope was and that even greater excursion can best postoperative flexor tendon repair
almost three times greater and the dif- be expected if wrist extension is protocol probably (1) compensates for
ferences continued at each interval
through 12 weeks. It seems that
greater magnitudes, frequencies, and
durations of motion stress may have
an accelerating effect on tendon heal-
ing and that almost all splinting and
passive/active protocols now permit
greater interdigital motion at more
frequent intervals than was previ-
ously recommended.
Studies also have been conducted
in an effort to determine the normal
amount of flexor tendon excursion
resulting from increments of digital
joint motion and the amount of excur-
sion that may occur with the various
splints that are commonly employed
after tendon repair. It has been
observed that passive metacarpopha-
langeal joint movement produces no
relative motion of the flexor tendons.
Fig. 10 Estimated strength (measured in grams) for two-, four-, and six-strand flexor ten-
Distal interphalangeal joint motion don repairs performed with the use of epitendinous sutures.
produces excursion of the FDP tendon
of 1 to 2 mm per 10 degrees of joint
flexion, while each 10 degrees of prox-
Table 2
imal interphalangeal joint flexion
Estimated Repair Strength With Epitendinous Suture*
results in excursion of both the FDP
and the FDS tendons of about 1.5 mm. Type of 0 1 3 6
Studies measuring the excursion of Repair Weeks Week Weeks Weeks
tagged flexor tendon repairs have
demonstrated that there is a substan- Two-strand 2,500 g 1,200 g 1,700 g 2,700 g
tial decrease in the normal movement Four-strand 4,300 g 2,150 g 2,800 g 5,200 g
of the FDP tendon to an average of 0.3 Six-strand 6,000 g 3,000 g 4,000 g 7,200 g
mm per 10 degrees of distal interpha-
langeal joint flexion (36%), while prox- *Horizontal-mattress or running-lock suture.

52 Journal of the American Academy of Orthopaedic Surgeons


Fig. 11 Tendon excursion (measured in millimeters) with three types of mobilization
splints: the Kleinert splint (with no palmar bar), the Brooke Army splint (a modification of
the Kleinert splint with a palmar-bar pulley), and the Mayo Clinic “synergistic” dynamic ten-
odesis splint (which permits wrist extension).25
Table 3
Tendon Excursion (Zone II) With Three Types of Mobilization Splint
Type of Splint FDS
Kleinert (no palmar bar) 8 10
Brooke (palmar bar) 13
Mayo Synergistic 15
swelling of the finger, (2) keeps the
wrist and metacarpophalangeal joints
flexed at rest, (3) keeps proximal and
References
1. Doyle JR: Anatomy of the finger flexor
tendon sheath and pulley system. J
Hand Surg [Am] 1988;13:473-484.
2. Brand PW, Hollister A: Overview of
mechanics of the hand, in Brand PW,
Hollister A (eds): Clinical Mechanics of the
Hand, 2nd ed. St Louis: Mosby Year
Book, 1993, pp 10-12.
Vol 3, No 1, Jan/Feb 1995
Tendon Excursion, mm
FDP Differential
2.3
15 2.0
20 4.6
distal interphalangeal joints extended
at rest, (4) passively flexes all digital
joints before wrist extension, (5) per-
3. Idler RS: Anatomy and biomechanics of the
digital flexor tendons. Hand Clin 1985;1:3-11.
4. Schuind F, Garcia-Elias M, Cooney WP
III, et al: Flexor tendon forces: In vivo
measurements. J Hand Surg [Am]
1992;17:291-298.
5. Urbaniak JR, Cahill JD Jr, Mortenson
RA: Tendon suturing methods:
James W. Strickland, MD
mits active maintenance of passively
achieved digital flexion with the wrist
extended, and (6) utilizes frequent
application of motion stress.
Several programs combining a
strong four-strand repair or its
equivalent with a running-lock
loop, horizontal-mattress, or
intrafiber circumferential epi -
tendinous repair and employing
early protected passive and active
motion have now been developed,
and the results are clearly better
than with previous, more conserv-
ative techniques. 27,28
Summary
Current information supports
the use of a four-strand core
stitch or its equivalent for flexor
tendon repair combined with a
strong continuous peripheral
epitendinous suture. This suture
method should impart sufficient
strength to the repair to permit a
vigorous postrepair motion pro-
tocol, which appears to maxi -
mize the excursion of the
repaired tendon while minimiz-
ing the possibility of rupture.
Although the results of these
techniques are encouraging,
rapid advances continue to occur
in many areas of flexor tendon
surgery, and even better tech-
niques will lead to improved
results in the future.
Analysis of tensile strengths, in Amer-
ican Academy of Orthopaedic Surgeons
Symposium on Tendon Surgery in the
Hand. St Louis: CV Mosby, 1975, pp
70-80.
6. Gelberman R, Goldberg V, An KN, et al:
Tendon, in Woo SLY, Buckwalter JA
(eds): Injury and Repair of the Muscu-
53
Flexor Tendon Injuries: Treatment Principles
loskeletal Soft Tissues. Park Ridge, Ill:
American Academy of Orthopaedic
Surgeons, 1988, pp 5-40.
7. Gelberman RH, Steinberg D, Amiel D, et
al: Fibroblast chemotaxis after tendon
repair. J Hand Surg [Am] 1991;16:
686-693.
8. Gelberman RH, Woo SLY, Amiel D, et
al: Influences of flexor sheath continu-
ity and early motion on tendon healing
in dogs. J Hand Surg [Am] 1990;15:
69-77.
9. Peterson WW, Manske PR, Dunlap J, et
al: Effect of various methods of restor-
ing flexor sheath integrity on the forma-
tion of adhesions after tendon injury. J
Hand Surg [Am] 1990;15:48-56.
10. Kulick MI, Smith S, Hadler K: Oral
ibuprofen: Evaluation of its effect on
peritendinous adhesions and the break-
ing strength of a tenorrhaphy. J Hand
Surg [Am] 1986;11:110-120.
11. Carlstedt CA, Madsen K, Wredmark T:
The influence of indomethacin on bio-
mechanical and biochemical properties
of the plantaris longus tendon in the
rabbit. Arch Orthop Trauma Surg
1987;106:157-160.
12. Savage R: In vitro studies of a new
method of flexor tendon repair. J Hand
Surg [Br] 1985;10:135-141.
13. Robertson GA, Al-Qattan MM: A bio-
mechanical analysis of a new interlock
suture technique for flexor tendon
repair. J Hand Surg [Br] 1992;17:92-93.
54
14. Lee H: Double loop locking suture: A
technique of tendon repair for early
active mobilization: Part I. Evolution of
technique and experimental study. J
Hand Surg [Am] 1990;15:945-952.
15. Lee H: Double loop locking suture: A tech-
nique of tendon repair for early active
mobilization: Part II. Clinical experience. J
Hand Surg [Am] 1990;15:953-958.
16. Lin GT, An KN, Amadio PC, et al: Bio-
mechanical studies of running suture
for flexor tendon repair in dogs. J Hand
Surg [Am] 1988;13:553-558.
17. Wade PJF, Wetherell RG, Amis AA:
Flexor tendon repair: Significant gain in
strength from the Halsted peripheral
suture technique. J Hand Surg [Br]
1989;14:232-235.
18. Mashadi ZB, Amis AA: Strength of the
suture in the epitenon and within the
tendon fibres: Development of stronger
peripheral suture technique. J Hand
Surg [Br] 1992;17:172-175.
19. Ketchum LD: Suture materials and
suture techniques used in tendon repair.
Hand Clin 1985;1:43-53.
20. Bishop AT, Cooney WP III, Wood MB:
Treatment of partial flexor tendon lacer-
ations: The effect of tenorrhaphy and
early protected mobilization. J Trauma
1986;26:301-312.
21. Strickland JW: Biologic rationale, clini-
cal application, and results of early
motion following flexor tendon repair. J
Hand Ther 1989;2:71-83.
Journal of the American Academy of Orthopaedic Surgeons
22. Gelberman RH, Botte MJ, Spiegelman JJ,
et al: The excursion and deformation of
repaired flexor tendons treated with
protected early motion. J Hand Surg
[Am] 1986;11:106-110.
23. Gelberman RH, Woo SLY, Lothringer K,
et al: Effects of early intermittent pas-
sive mobilization on healing canine
flexor tendons. J Hand Surg [Am] 1982;
7:170-175.
24. Silfverskiöld KL, May EJ, Törnvall
AH: Flexor digitorum profundus
tendon excursions during con-
trolled motion after flexor tendon
repair in zone II: A prospective clin-
ical study. J Hand Surg [Am] 1992;17:
122-131.
25. Cooney WP, Lin GT, An KN:
Improved tendon excursion following
flexor tendon repair. J Hand Ther
1989;2:102-106.
26. Savage R: The influence of wrist posi-
tion on the minimum force required for
active movement of the interpha-
langeal joints. J Hand Surg [Br] 1988;13:
262-268.
27. Strickland JW: Flexor tendon repair:
Indiana method. Indiana Hand Center
Newsletter 1993;1:1-12.
28. Silfverskiöld KL, May EJ: Flexor tendon
repair in zone II with a new suture tech-
nique and an early mobilization pro-
gram combining passive and active
flexion. J Hand Surg [Am] 1994;19:
53-60.
 Flexor Tendon Injuries: II. Operative Technique

James W. Strickland, MD

Abstract

The repair of flexor tendons (zones I and II) is a technique-intensive surgical When planning the surgical expo-
undertaking. It requires a strong understanding of the anatomy of the tendon sure, the surgeon must appreciate the
sheath and the normal relationship between the pulleys and the flexor digitorum fact that the severed flexor tendon
superficialis and flexor digitorum profundus tendons in the digit. Meticulous ends will retract well away from the
exposure, careful tendon retrieval, and atraumatic repair are extremely important, laceration site. When the digit is in
and the repair should be of sufficient strength to resist gapping and permit the flexion at the time of injury, the distal
early postrepair application of motion forces. Whenever possible, the tendon stumps of the severed tendons may
sheath should be preserved or repaired, and a smooth gliding surface should be come to rest a centimeter or more dis-
reestablished. The author describes an effective method of tendon retrieval and a tal to the level of sheath disruption.
simplified technique for a four-strand tendon repair with a supplementary periph- While there is no fixed rule for the
eral running-lock suture. The repair is considered to maintain sufficient strength incisions that should be used to
throughout healing to allow a postrepair rehabilitation protocol that will impart expose the flexor tendons, there is no
passive and modest active stress forces to the repaired tendons. Complications advantage in trying to carry out these
include tendon rupture, digital joint flexion contractures, and adhesions that complicated repairs through existing
restrict tendon gliding and ultimately necessitate tenolysis. unextended wounds or through small
J Am Acad Orthop Surg 1995;3:55-62 incisions along the length of the digit.
The surgeon should select inci-
sions that will not compromise the
viability of skin flaps and that, when
The restoration of function to a digit the repair from rupture throughout healed, will not create contracting or
following flexor tendon interrup- the entire healing process. cosmetically unsightly scars. Zigzag
tion and repair may be a long and or midaxial incisions are often uti-
tedious undertaking. Strong rap- lized. The selection should be based
port must be developed between the General Considerations on the position, length, and direction
surgeon, the patient, and the thera- of the original laceration; the need to
pist. When initiating the care of a All surgeons embarking on flexor gain access to other injured struc-
patient with such an injury, the sur- tendon repairs should adhere to the tures; and the personal experience
geon must spend considerable time time-honored advice of Sterling Bun- and preference of the surgeon.
explaining the inherent problems, nell, who stated that tendon repair
the likelihood of achieving success, must be carried out using meticu-
and the fact that several procedures lous, atraumatic technique in an Dr. Strickland is Clinical Professor of
may be necessary to maximize the effort to lessen adhesion formation.1 Orthopaedic Surgery, Indiana University School
recovery of digital function. The use of ×2 to ×4 loupe magnifica- of Medicine, Indianapolis; Chairman, Depart-
ment of Hand Surgery, St. Vincent Hospital,
The exposure of the tendon tion is an important adjunct in per-
Indianapolis; and Senior Staff Surgeon, Indiana
stumps, the preservation or repair forming flexor tendon and sheath Hand Center, Indianapolis.
of the sheath, and the suturing of the repair, and small delicate instrumen-
tendon require considerable experi- tation is a prerequisite for this type of Reprint requests: Dr. Strickland, PO Box 80434,
ence and skill. Early postrepair surgery. Pinching, crushing, and 8501 Harcourt Road, Indianapolis, IN 46280-
0434.
therapy must be carried out with the excessive touching of the tendon
use of protocols that apply sufficient sheath or flexor tendons must be Copyright 1995 by the American Academy of
stress to ensure rapid healing and avoided. It usually is not necessary Orthopaedic Surgeons.
satisfactory gliding while protecting to debride or shorten tendon ends.

Vol 3, No 1, Jan/Feb 1995 55

Flexor Tendon Injuries: Operative Technique

Zone I
In zone I (distal to the insertion of the
flexor digitorum superficialis [FDS]
tendon over the middle phalanx),
when only the flexor digitorum pro-
fundus (FDP) tendon has been sev-
ered, there is usually little difficulty
in finding the proximal tendon end,
which is at least temporarily retained
in the finger by its vinculum and can A B
usually be located in the distal por-
tion of the proximal phalanx or at the
level of the proximal interphalangeal
joint. Careful dissection will expose
the distal half of the flexor tendon
sheath. The entire A4 annular pulley
should be preserved.
If the distal interphalangeal joint
was flexed at the time of injury, the
tendon will probably have a short
C D
distal stump over the base of the dis-
tal phalanx, which can be exposed
by opening the C3-A5 pulley com-
plex. It will also be necessary to
open the C2 or C1 cruciate-synovial
sheath segments proximal to the A4
pulley in order to retrieve the proxi-
mal stump of the divided FDP ten-
don. After the proximal FDP has
been delivered into the appropriate
cruciate-synovial sheath interval, a E F
core suture is placed in the tendon,
Fig. 1 Flexor digitorum repair in zone I. The patient had suffered a laceration at the level
allowing it to be passed distally of the distal interphalangeal joint with severance of the FDP. A, Appearance of the short dis-
under the A4 annular pulley without tal tendon stump. B, Core suture has been placed in the proximal FDP stump. C, Proximal
the need for further instrumentation tendon stump has been passed beneath the A4 pulley. D, Completed repair. E, Repair of the
C2 window with use of a fine monofilament suture. F, Appearance of the completed tendon
of the tendon. The proximal tendon and sheath repair. The patient had excellent recovery of digital flexion at 4 months.
is usually maintained in position by
the passage of a transversely ori-
ented 25-gauge hypodermic needle.
The repair is completed distal to the armed (straight needles) 3-0 suture is to use one of the several commer-
A4 pulley with an end-to-end tendon placed in the proximal tendon stump cially available bone-anchor sutures
suture if enough distal tendon and passed through the bone hole. In to secure the proximal stump to the
remains to accept a suture (Fig. 1). this instance, it is best to use a syn- distal phalanx. When possible, the
If the distal stump is insufficient to thetic monofilament suture placed in tendon attachment should be sup-
hold a suture, the proximal FDP a crisscross (unlocked) fashion, so plemented by sutures through the
stump may be reattached by first ele- that the suture can be pulled out after adjacent sheath or periosteum.
vating an osteoperiosteal flap from bone-tendon healing has occurred.
the base of the distal phalanx and The sutures are then used to pull
then drilling an oblique hole beneath the tendon beneath the periosteal Zone II
the flap, directed so as to penetrate flap and are tied over a cotton pad–
the dorsal cortex just beneath the button combination over the nail. In zone II (from the origin of the
proximal fingernail. A double- On some occasions, it may be helpful flexor tendon sheath to the insertion

56 Journal of the American Academy of Orthopaedic Surgeons


of the FDS tendon over the midpor-
tion of the middle phalanx) flexor
tendon repairs, it is always necessary
to make proximal and distal extend-
ing incisions that provide satisfac-
tory exposure of the repair site.
Dissection proceeds with identifica-
tion and protection of the digital
nerves and arteries. If these struc-
tures have been severed, their ends
are mobilized and brought into prox-
imity for subsequent suturing. An
assessment of the level and extent of
sheath injury should be made, as
well as assessment of the probable
position of the tendon ends.
It will then be necessary to open
either the C1 (between A2 and A3) or
C2 (between A3 and A4) cruciate-syn-
ovial sheath windows by means of
connecting incisions along one end
and one side. When opening the
intact components of the sheath,
every attempt must be taken to pre-
serve the annular components (A1,
A2, A3, and A4), which are often dif-
ficult to repair. Tendon suture
should be performed in the cruciate-
synovial sheath windows, which can
usually be restored following ten-
don suture.
By acutely flexing the distal inter-
phalangeal joint and, to some extent,
the proximal interphalangeal joint, it
is usually possible to deliver the dis-
tal FDP and FDS stumps into a cru-
ciate window. If 1 cm or more of the
tendons can be exposed, core
sutures can be placed in the FDP ten-
don and two superficial slips with-
out great difficulty. If a lesser length
of tendon end is present in the win-
dow, the next most distal cruciate-
synovial sheath interval must be
opened for core suture placement.
The actual joining of the tendon ends
can then be carried out in either win-
dow, depending on the most distal
point that can be achieved for the
proximal tendon stumps and the
length of the distal tendons and their
most proximal position during distal
interphalangeal joint flexion.
Vol 3, No 1, Jan/Feb 1995
James W. Strickland, MD
Retrieval of the proximal tendon
ends may be difficult. Repeated
blind grasps down the sheath with
an instrument will often fail to
retrieve the proximal stumps and
may, in fact, damage the synovial
lining of the pulleys and theoreti-
cally can provoke adhesions. Such
efforts are appropriate only if the
tendon or tendons can be visualized
in the sheath and are sufficiently A
close to the cruciate-synovial sheath
window to ensure that an end can be
atraumatically pinched with forceps
and delivered distally.
Many clever tactics have been
suggested to facilitate tendon cap-
ture and repositioning. These meth-
ods include proximal-to-distal
“milking” of the tendons toward the
repair site and the use of various
B
types of catheters and silicone rods,
which are sutured to the ends of the
tendon stumps in the palm and
passed through the sheath in an
effort to pull the tendons back into
their distal position. The two meth-
ods that are perhaps the most reli-
able techniques for proximal tendon
retrieval are the following:
In the first method, used if the ten-
don end is visible in the sheath, a skin C
hook is employed as described by
Morris and Martin.2 The hooked end
is slid along the surface of the sheath
until it is past the tendons. The hook
is then turned toward the tendons
and pressed into the most superficial
one. When the hook engages a ten-
don, the instrument is pulled distally.
Both tendons will usually follow.
They can then be held in position by D
a 25-gauge hypodermic needle.
The second method, described by Fig. 2 Sourmelis and McGrouther’s method
of retrieving flexor tendons. A, Polyethylene
Sourmelis and McGrouther,3 is excel- catheter is passed distal to proximal along the
lent for proximal tendon stumps that flexor tendons, which are left in situ.
cannot be visualized and are still Catheter is sutured in both tendons 2 cm
proximal to the A1 pulley through a palmar
located in the proximal sheath (Fig. 2). incision. B, Catheter is advanced distally to
A small catheter is passed from the deliver tendon ends into repair site. C, A 22-
distal wound into the palm (or vice gauge hypodermic needle is passed trans-
versely through the annular sheath to
versa) beneath the annular pulleys. maintain tendon position. D, Catheter-ten-
An important feature of this method don suture is cut in palm and withdrawn.
is that the flexor tendons are left in
57
Flexor Tendon Injuries: Operative Technique

situ in the sheath. Through a mid-

palm incision, the catheter is sutured
to both tendons several centimeters
proximal to the A 1 pulley. The
catheter is then pulled distally, easily
delivering the tendon stumps into the
distal repair site. A transversely ori-
ented needle will secure the tendons
for repair. The connecting suture can
then be severed in the palm, and the
catheter can be withdrawn. Core
sutures can then be placed in the
proximal FDP stump and the FDS
slips, and the tendons can usually be
brought into juxtaposition with the
distal stumps for repair.
When the proximal tendon ends
have retracted into the palm, it is
extremely important to reestablish
the proper anatomic relationship of
the FDP and FDS tendons. To
accomplish this goal, the FDP has to A B
be passed back through the hiatus
created by the FDS slips so that it will
lie palmar to Camper’s chiasma and
will recreate the relative positioning
that was present at the level of ten-
don laceration. Failure to restore the
correct relationship will create an
impediment to unrestricted tendon
gliding after repair. Once the proper
tendon anatomy has been reestab-
lished, a catheter passed retrograde
from the cruciate window is attached
to the tendons, and (usually with
some difficulty) the tendons are
entered into the sheath and delivered
distally, where they can be main-
tained with a transversely oriented
hypodermic needle (Fig. 3).
The technique used for the core
sutures is based on the preference of
the surgeon. Variations of the Kessler
grasping technique4,5 enjoy the most C D
popularity. If one prefers not to
employ the more complex four-strand Fig. 3 A, Altered relationship of the two proximal tendon ends in the distal palm follow-
repair methods, the simple addition of ing proximal retraction. It is important to reestablish the correct anatomic position of the
FDP and FDS stumps before passing them back into the digit. B, The proximal FDP and FDS
a horizontal-mattress suture across stumps are retrieved proximal to the A1 pulley and, following the placement of tendon
the tendon ends will complete a four- sutures, are delivered distally through the intact portion of the flexor sheath with use of a
strand repair (Fig. 4). A running-lock small catheter or an infant gastrostomy feeding tube. C, The proximal tendon stumps are
maintained at the repair site by means of a transversely placed small-gauge hypodermic nee-
stitch may then be used as a peripheral dle. Repair of the FDS slips is then carried out. D, After completed repair of both flexor ten-
epitendinous stitch (Fig. 5). This is dons, sheath can be closed.
usually easier to place than some of

58 Journal of the American Academy of Orthopaedic Surgeons


B A
E E
Fig. 4 Simplified four-strand repair in
which a basic two-strand core suture is
supplemented by a horizontal-mattress
suture and a running-lock stitch. A, Tajima
core sutures in place. Back-wall (dorsal)
running-lock peripheral epitendinous
stitch in progress. B, Back-wall suture
completed. C, Mattress core suture added
in palmar tendon gap. D, All core sutures
tied. E, Completion of running-lock
peripheral epitendinous suture. F, Repair
complete.
Vol 3, No 1, Jan/Feb 1995
James W. Strickland, MD
the more complicated methods of repairing the flexor sheath with
described in recent articles. the use of 6-0 nylon suture on a
At the conclusion of the tendon small needle. At this point, the dis-
repair, the surgeon has the option tal interphalangeal joint can be
B
C D
F
H
G
Fig. 5 Tajima core suture plus horizontal-mattress four-strand repair with running-lock
peripheral epitendinous suture was used in this patient with a laceration at the proximal
interphalangeal joint that severed both flexor tendons. A, Tendon retrieval using Sourmelis
and McGrouther’s method. B, Tajima core suture in proximal stump. C, Tajima core suture
in distal stump. D, Back-wall repair with running-lock stitch. E, Additional horizontal-mat-
tress core suture added. F, Tying of Tajima and mattress sutures results in a four-strand
repair. G, Front-wall repair with running-lock stitch. H, Sheath repaired.
59
Flexor Tendon Injuries: Operative Technique

extended to deliver the repair dis-
tally. The repaired sheath usually
serves as a smooth conduit for the
repair as it moves under an annular
pulley. If it is elected to leave the
flexor tendon sheath open, up to 1.5
cm of sheath may be excised to per-
mit unrestricted gliding of the
repaired tendon.
On occasion, the process of flexor
tendon suture is complicated by the
need to use two adjacent windows
(usually C1 and C2) for core-suture
placement. The final repair is car-
ried out in the most appropriate
window after the proximal or distal
stumps have been delivered into
position by passing the ends of their
core sutures under the interven-
ing annular pulley (usually A3) and
gently pulling the tendon ends into
position. Flexor sheath repair is
facilitated if the bulk of the tendon
juncture can be minimized and care-
ful sheath incisions have been uti-
lized. To ensure the passage of the
repaired tendon beneath the annular
pulleys, it is helpful to close the over-
lying cruciate-synovial sheath
before extending the interpha-
langeal joints (Fig. 6).
At the conclusion of the repair,
digital nerves and, if necessary, dig-
ital arteries are repaired. The skin is
then closed using fine nonab-
sorbable sutures. A large, bulky,
compressive dressing immobilizing
all the digits and the thumb is used
postoperatively with the wrist in
midflexion, the metacarpopha-
langeal joints in full flexion, and the
proximal interphalangeal and distal
interphalangeal joints in extension.
The use of antibiotics depends on the
personal philosophy and discretion
of the surgeon.
Postoperative Management
As a result of the publication of
numerous articles demonstrating
the superior results that followed
flexor tendon repairs managed by
one of several mobilization pro-
grams, almost all hand surgeons
currently employ some type of
early motion protocol. As the
development of these methods has
evolved, emphasis has been placed
on techniques that employ not only
a greater amount of composite dig-
ital motion but also the use of “syn-

A B C

Fig. 6 Situation in which more distal severance of the FDP and FDS tendons has resulted in the inability to deliver 1 cm of the distal stumps
into the C1 window. Incisions in both the C1 and C2 intervals will be necessary to provide adequate exposure for suture placement and repair.
A, Appearance of the flexor sheath after flexor tendon interruption at a level just proximal to the proximal interphalangeal joint and the A3
pulley. Passive flexion of the distal interphalangeal joint will not provide adequate length of the distal stump in the C1 window for place-
ment of the core sutures and repair. Dotted lines indicate the appropriate incisions in both the C1 and C2 windows. B, After placement of
sutures in the distal tendon stumps in the C2 window and in the proximal stumps in the C1 opening, proximal tendon stumps can be deliv-
ered distally beneath the A3 pulley. (The FDS tendon stumps have been repaired.) C, Appearance of the digit after repair of both the FDP
and FDS tendons and the C1 cruciate-synovial sheath. Repair of the C2 sheath should now be possible.

60 Journal of the American Academy of Orthopaedic Surgeons

 James W. Strickland, MD

ergistic” wrist extension in an effort
to gain the greatest amount of
excursion of the repaired tendon.
As has been discussed in the
accompanying article, it now appears
that light active digital flexion carried
out with the wrist in extension should
be relatively safe for flexor tendons
repaired with a four-strand core-
suture technique augmented by a
strong peripheral epitendinous suture
(Fig. 7). Most of these postoperative
programs permit the active mainte-
nance of passively achieved compos-
ite digital flexion once the wrist is
brought from flexion to extension.6
Complications
Rupture of one or both flexor tendon
repairs is a significant complication.
It occurs most frequently when
patients abandon the prescribed
rehabilitation protocol and attempt
to make a strong fist or lift a heavy
object. The preferred treatment for
tendon-repair rupture is prompt
reexploration and repair. The most
frequent late complication following
early postoperative mobilization
programs is the development of flex-
ion contractures at the proximal
and/or distal interphalangeal joints.
Prompt recognition of the develop-
ment of contractures, modification
of the motion program to permit
greater extension, and judicious use
of static and dynamic splints can

A B C

D E F

G H I

Fig. 7 A, Left hand of a 29-year-old man 24 hours after severance of FDS and FDP tendons in the middle and ring fingers. B, Outline of
midaxial excisions on the ulnar side of both digits. C, Sourmelis and McGrouther’s method of retrieving proximal tendons in the middle fin-
ger. D, Completed four-strand core suture plus continuous horizontal-mattress epitendinous stitch. E, Position of wrist and digits in post-
operative splint 3 days after repair, at the time of the initiation of therapy. F, Passive flexion of digits with wrist in flexion. G, Active
maintenance of composite digital flexion with wrist extension. H, Nearly full digital extension at 4 months. I, Excellent composite flexion of
both digits at 4 months.

Vol 3, No 1, Jan/Feb 1995 61

Flexor Tendon Injuries: Operative Technique
help prevent or overcome these
deformities before they progress.
In some cases, despite the best pos-
sible repair and strong cooperation on
the part of the patient, the tendons
may become adherent and fail to glide
sufficiently to reestablish adequate
digital function. A secondary tenoly-
sis procedure may be necessary if
repeated joint measurements indicate
that there has been no appreciable
improvement for several months. The
procedure should not be considered
until all wounds have reached “equi-
References
1. Bunnell S: Repair of tendons in the fin-
gers and description of two new instru-
ments. Surg Gynecol Obstet 1922;35:88-97.
2. Morris RJ, Martin DL: The use of skin
hooks and hypodermic needles in tendon
surgery. J Hand Surg [Br] 1993;18:33-34.
62
librium,” with soft, pliable skin and
subcutaneous tissues and minimal
reaction around the scars. Joint con-
tractures must have been mobilized,
and a normal or near-normal passive
range of digital motion is desirable.
Summary
Success is ultimately dependent on
the participation of a knowledgeable
surgeon, who has spent time in the
cadaver laboratory to develop the
3. Sourmelis SG, McGrouther DA:
Retrieval of the retracted flexor ten-
don. J Hand Surg [Br] 1987;12:109-
111.
4. Kessler I: The “grasping” technique for
tendon repair. Hand 1973;5:253-255.
Journal of the American Academy of Orthopaedic Surgeons
technical skills for the four-strand
repair and the running-lock periph-
eral epitendinous stitch, and a well-
trained therapist who thoroughly
understands the rationale for the
rehabilitation protocol and its pit-
falls and is prepared to spend a great
deal of time with the patient explain-
ing the details of the rigorous ther-
apy program. These methods of
primary repair cannot succeed with-
out the cooperation of an intelligent,
motivated patient who adheres
rigidly to the protocol.
5. Kirchmayr L: Zur Technik der Sehn-
ennaht. Zentralbl Chir 1917;44:
906-907.
6. Cannon NM: Post flexor tendon repair
motion protocol. Indiana Hand Center
Newsletter 1993;1:13-18.