Review Article

Legg-Calvé-Perthes Disease

Abstract
Harry K. W. Kim, MD Legg-Calvé-Perthes disease is an idiopathic hip disorder that
produces ischemic necrosis of the growing femoral head.
Permanent femoral head deformity is the most significant sequela.
Experimental studies indicate that the pathologic repair process,
which is marked by an imbalance of bone resorption and formation,
contributes to the pathogenesis of femoral head deformity.
Important prognostic factors include degree of deformity, age at
disease onset, extent of head involvement, head-at-risk signs, and
lateral pillar collapse. Treatment should be guided by age at
disease onset, current best evidence, and prognostic factors.
Patients aged <6 years at onset are best managed nonsurgically,
whereas older patients may benefit from surgical treatment. Good
surgical results have been reported in 40% to 60% of older
patients (>8 years), indicating the need to develop more effective
treatments based on the pathobiology of the disease.

L egg-Calvé-Perthes disease (LCPD)

From the Center for Excellence in
Hip Disorders, Texas Scottish Rite
Hospital, UT Southwestern Medical
Center, Dallas, TX.
Neither Dr. Kim nor any immediate
family member has received
anything of value from or owns
stock in a commercial company or
institution related directly or
indirectly to the subject of this
article.
J Am Acad Orthop Surg 2010;18:
676-686
Copyright 2010 by the American
Academy of Orthopaedic Surgeons.
is a childhood hip disorder of un-
known etiology that can produce per-
manent deformity of the femoral head.
This condition was first reported as a
disease entity separate from tuberculo-
sis in 1910 by three independent au-
thors: Legg,1 Calvé,2 and Perthes.3
Since then, controversies regarding
the etiology, pathogenesis, and man-
agement of LCPD have arisen, many
of which remain unresolved. Recent
genetic studies on a type II collagen
mutation as a cause of LCPD4,5 and
studies on the role of inherited
thrombophilia6-8 on LCPD represent
advancement, but the cause of ische-
mic necrosis remains unknown. Fur-
ther insight into the pathogenesis of
femoral head deformity has been
gained through experimental stud-
ies.9-11 These studies reveal a patho-
logic repair process in which an im-
balance of bone resorption and
formation contributes to the devel-
opment of the deformity. Long-term
studies suggest that although the
femoral head deformity is relatively
well-tolerated in the short and inter-
mediate term, >50% of patients de-
velop disabling arthritis in the sixth
decade of life.12 Thus, the overall
goal of treatment should be to pre-
vent or minimize femoral head defor-
mity. Two recent multicenter pro-
spective studies constitute the best
current evidence (level II) to guide
treatment of patients in different age
groups,13,14 along with recent large
retrospective studies.15-17
Etiology
Two recent genetic studies of families
with inherited bilateral osteonecrosis
of the femoral head have provided
insight into the etiology of LCPD.
Asian families in which multiple
members were affected in an auto-
somal dominant fashion were found
to have a missense mutation in the
type II collagen gene (ie, replacement
of glycine with serine at codon 1170

676 Journal of the American Academy of Orthopaedic Surgeons


Figure 1
Histopathologic changes observed in the necrotic femoral head over time. (Copyright © Texas Scottish Rite Hospital for
Children, Dallas, TX.)
of COL2A1).4,5 In children, the ra-
diographic changes typical of LCPD
were observed.4,5 In contrast to skele-
tal dysplasias, the affected persons
generally did not display skeletal ab-
normalities outside the hips. It is
speculated that the mutation may
cause weakening of the cartilage ma-
trix and compromise the blood ves-
sels within the cartilage. The muta-
tion has not yet been reported in
sporadic or nonfamilial bilateral
cases of LCPD.
Thrombophilia as a cause of LCPD
remains controversial. In a case-
control study of patients with LCPD,
Glueck et al6 reported that 75% had
a coagulation abnormality. In con-
trast, the authors of a prospective
study did not find a difference in the
prevalence of protein C, protein S, or
antithrombin III deficiencies, or in
factor V Leiden mutation between
the disease group and the estimated
population frequency.7 A more recent
prospective study found no increase
in the prevalence of protein C, pro-
tein S, or antithrombin III deficien-
cies in patients with LCPD.8 How-
ever, this study did find a higher
November 2010, Vol 18, No 11
prevalence of factor V Leiden and
anticardiolipin antibodies in the
LCPD group. Thrombotic events are
uncommon during childhood, even
in patients with inherited thrombo-
philia, and their significance in the
pathogenesis of LCPD remains un-
clear.
Pathogenesis
The etiology of LCPD remains un-
known. However, clinical and exper-
imental evidence support the notion
that disruption of blood supply to
the femoral head is a key pathogenic
event associated with the disease
process. The few histopathologic
studies available18 indicate that the
pathologic processes in LCPD affect
the articular cartilage, bony epiphy-
sis, physis, and metaphysis.
Changes in the articular cartilage
are found primarily in the middle
and deep layers. These changes in-
clude necrosis in the deep layer, ces-
sation of endochondral ossification,
separation of cartilage from the un-
derlying subchondral bone, vascular
Harry K. W. Kim, MD
invasion of the cartilage, and new ac-
cessory ossification (Figure 1). In the
bony epiphysis, necrosis of the mar-
row space and trabecular bone, com-
pression fracture of the trabeculae,
fibrovascular granulation tissue inva-
sion and osteoclastic resorption of
the necrotic bone, and thickened tra-
beculae in some areas have been re-
ported. Physeal changes are most of-
ten seen in the anterior part of the
femoral head, with focal areas of
growth cartilage columns extending
below the endochondral ossification
line. Metaphyseal changes are com-
monly seen during the early stages of
LCPD. Various tissue types have
been reported, including physeal car-
tilage columns, fibrocartilage, fat ne-
crosis, vascular proliferation, and fo-
cal fibrosis.
The lack of availability of clinical
samples for research has prompted
alternative approaches, such as the
use of animal models, to investigate
the pathogenesis of LCPD. In partic-
ular, a piglet model has allowed
more in-depth investigation of ische-
mic tissue damage and the repair
process.9 These models indicate that
677
Legg-Calvé-Perthes Disease
Figure 2
Graphic representation of the mechanical changes in the necrotic femoral
head versus a normal hip. The extent of head involvement, degree of
imbalance between bone resorption and formation, duration of healing, and
level of hip loading likely affect the deformity. The potential to remodel the
deformed head, as seen in young patients, offsets the deformity produced in
the acute phase.
the induction of ischemia produces a
decrease in the mechanical strength
of the necrotic femoral head, result-
ing in a head that is softer than nor-
mal.10 The mechanical compromise
observed in the avascular necrotic
phase may be the result of necrosis
of the deep layer of articular carti-
lage; increased mineralization of the
calcified cartilage and trabecular
bone,19 which presumably makes
them more brittle; and the possible
accumulation of microfractures in
the necrotic bone due to the absence
of viable cells to repair the micro-
fractures produced by normal wear
and tear caused by repetitive loading.
Vascular invasion and subsequent re-
sorption of necrotic bone further
compromise the mechanical proper-
ties of the infarcted head in the vas-
cular repair phase.10 It is postulated
that the weakened femoral head be-
gins to deform when its ability to re-
sist deformation falls below a critical
level surpassed by loading of the
hip joint (Figure 2). Inhibition of
bone resorption using antiresorptive
agents, such as diphosphonates, has
been shown to decrease deformity in
preclinical studies, which indicates
that the resorptive process is an im-
portant component in the pathogene-
sis of femoral head deformity follow-
678
ing ischemic necrosis.20,21 Although
clinical validation of the protective
effect of diphosphonate on LCPD is
lacking, early studies on its effects on
preserving the femoral head in adult
osteonecrosis appear promising.22,23
The hip is a major load-bearing
joint, and it is pertinent to consider
the development of femoral head de-
formity in the context of hip joint
loading. Data on the hip contact
pressures associated with various ac-
tivities of daily living in children are
not available. In adults, however, a
sophisticated femoral head prosthe-
sis equipped with a strain gauge has
been used to collect these data fol-
lowing total hip replacement.24 Sig-
nificant forces were found to act on
the femoral head with various activi-
ties. Normal walking produced a hip
contact pressure approximately 2.5
times body weight. Running on a
treadmill (8 km/h) increased the
pressure to approximately 4.5 times
body weight. In a disease in which
femoral head deformity is produced
as the result of mechanical weaken-
ing, avoidance of activities that gen-
erate a significant increase in the hip
contact pressure would seem to be
reasonable. Currently, it is unknown
what constitutes “significant” load-
ing; neither is it known what effect
Journal of the American Academy of Orthopaedic Surgeons
activity restriction has on preventing
deformity.
Clinical Features
Although LCPD can affect a wide age
range of children, it is most commonly
seen in children aged 5 to 8 years. The
male-to-female ratio is approximately
5:1, and bilateral disease occurs in
10% to 15% of patients.25 Many pa-
tients have delayed bone ages and
appear to be younger than their
chronologic age. Some patients are
clearly hyperactive. LCPD is a diag-
nosis of exclusion; thus, other causes
of osteonecrosis (eg, sickle cell dis-
ease, corticosteroid therapy) and
mimicking conditions (eg, skeletal
dysplasias) must be ruled out.
In general, patients present with mild
pain, a limp, and/or limited hip motion,
which tends to have insidious onset.
On physical examination, most pa-
tients have a mild limp. They may also
have a positive Trendelenburg sign.
The limitation of hip motion depends
on the stage of the disease. Hip motion
is generally good in the early stages, but
hip irritability may be present because
of synovitis, which can be persistent.
Abduction and internal rotation are the
earliest motions to decrease. In the
fragmentation stage, hip motion can
become severely restricted. Flexion and
adduction contractures may develop in
some patients. Motion improves dur-
ing the reossification stage, but it can
remain restricted by a severe residual
deformity. Depending on the duration
of the disease, thigh and calf muscle at-
rophy may be observed as well as limb-
length discrepancy of 1 to 2.5 cm.
Imaging Studies
Waldenström26 described four radio-
graphic stages of LCPD according to
the characteristic features of each:
initial (ie, increased radiodensity),
fragmentation, reossification, and

Table 1
Stulberg Radiographic Classification of Legg-Calvé-Perthes Disease and Evidence of Osteoarthritis at
Follow-up31
Class Descriptive Features
I Normal hip joint
II Spherical head with enlargement,
short neck, or steep acetabulum
III Nonspherical head (ie, ovoid,
mushroom-shaped, umbrella-
shaped)
IV Flat head
V Flat head with incongruent hip joint
OA = osteoarthritis
Adapted with permission from Stulberg SD, Cooperman DR, Wallensten R: The natural history of Legg-Calvé-Perthes disease. J Bone Joint
Surg Am 1981;63:1095-1108.
healed. The fragmentation stage lasts
approximately 1 year, and the reossi-
fication stage lasts 3 to 5 years.
Older patients appear to have a lon-
ger duration of reossification than
younger patients. Femoral head de-
formity develops and progresses dur-
ing the initial and fragmentation
stages. Femoral head shape can im-
prove, worsen, or remain unchanged
during the reossification stage. In
one study, the femoral head was
more likely to undergo progressive
flattening in older patients, in those
with more severe lateral pillar in-
volvement, and in those with pro-
longed reossification.27 Although ra-
diography is useful in assessing
disease progression, it lacks the sen-
sitivity and specificity needed to
demonstrate changes in vascular re-
pair within the femoral head.
Gadolinium-enhanced MRI can de-
tect changes in bone perfusion in the
early stages when radiographic
changes are not apparent.28 The role
of MRI in the management of LCPD
is still evolving. Although enhanced
MRI techniques have been shown to
provide accurate information regard-
ing the vascular status of the femoral
head, further studies are necessary to
demonstrate their clinical and prog-
November 2010, Vol 18, No 11
Radiographic Signs of OA at
Mean 40-year Follow-up (%)
0 0
16
58
75
78
nostic relevance for patient manage-
ment. Recently, a three-dimensional
MRI assessment technique has been
described to quantify the loss of fem-
oral head sphericity in patients with
LCPD.29
Natural History and
Radiographic
Classification
Treatment of patients with LCPD re-
quires an understanding of the natu-
ral history of the disease, the prog-
nostic factors, and the effectiveness
of various treatment methods. The
few published long-term natural his-
tory studies are limited by small sam-
ple size, loss of follow-up, and the
inclusion of patients treated nonsur-
gically. Long-term studies with a
mean follow-up <40 years show that
most patients are asymptomatic and
remain active despite the presence of
femoral head deformity.30 Noticeable
deterioration in hip function has
been reported in studies with longer
follow-up. A study from the Univer-
sity of Iowa with a mean follow-up
of 47.7 years found that only 40%
of the patients maintained a good
level of function (Iowa hip rating
Harry K. W. Kim, MD
Radiographic Evidence of Joint
Space Narrowing at Mean
40-year Follow-up (%)
0
47
53
61
>80 points), 40% required arthro-
plasty, 10% had disabling pain, and
the remaining 10% had an Iowa hip
rating of <80 points.12
Stulberg et al31 reported on the re-
lationship between the shape of the
femoral head and the risk of pre-
mature osteoarthritis. Their five-
category radiographic classification,
which is based on the severity of de-
formity and on hip joint congruence
at maturity, correlated with the de-
velopment of radiographic signs of
osteoarthritis at a mean follow-up of
40 years (Table 1). These results
show a significant decline in the out-
come, from spherical heads (class I
and II [good results]) to nonspherical
heads (class III through V [fair to
poor results]). The validity of the
classification has been called into
question because of low interob-
server reliability. Interobserver reli-
ability was improved when quantita-
tive parameters were assigned to
define certain Stulberg classes.32 Stul-
berg outcome cannot be determined
until maturity, which remains a ma-
jor drawback of this system. Re-
cently, the deformity index, which is
a continuous outcomes measure, has
been shown to predict the Stulberg
outcome at 2 years into the disease.33
679
Legg-Calvé-Perthes Disease

Table 2 Figure 3
Prognostic Indicators of
Outcome in Patients With Legg-
Calvé-Perthes Disease
Extent of femoral head deformity and
loss of hip joint congruity at maturity
(Stulberg classification)
Age at onset
Extent of subchondral fracture (Salter-
Thompson classification)
Extent of head involvement at the frag-
mentation stage (Catterall classifica-
tion)
Two or more Catterall head-at-risk signs
(lateral subluxation, lateral calcifica-
tion, diffuse metaphyseal reaction,
horizontal growth plate, Gage sign)
Lateral pillar height at the fragmentation
stage (lateral pillar classification)
Premature physeal closure

It remains to be seen whether this

method proves to be a reliable pre-
dictor when used by other investiga-
tors.
A long-standing dilemma for the
treating surgeon is discerning at the
time of presentation who will benefit
from surgical treatment. Various
clinical and radiographic features of
the disease have been identified as
prognosticators of outcome (Table
2). Ideally, a prognosticator should
be applicable at the time of presenta-
tion, easy to use, reliable, and repro-
ducible. The Salter-Thompson classi-
fication is a two-category system
based on the extent of subchondral
fracture in the early stage of frag-
mentation.34 Its application is re-
stricted by the absence of subchon-
dral fracture in many patients at the
time of presentation and follow-up.
The Catterall classification is based
on the extent of head involvement:
group I, involvement of the anterior
head only; group II, anterior head in-
volvement only and sequestrum with
a clear junction; group III, only a
small part of the epiphysis is not
involved; group IV, total head in-
volvement35 (Figure 3). Catterall also
The Catterall classification of Legg-Calvé-Perthes disease. In group I there is
involvement (hatched areas) of the anterior head only, no sequestrum, and
no collapse of the epiphysis. In group II, only the anterior head is involved,
and there is a sequestrum with a clear junction. In group III only a small part
of the epiphysis is not involved. In group IV there is total head involvement.
(Reproduced from Skaggs DL, Tolo VT: Legg-Calvé-Perthes disease. J Am
Acad Orthop Surg 1996;4:9-16.)
described “head at risk” signs associ- early fragmentation stage, when the
ated with a poor outcome. The ma- femoral head cannot be classified
jor criticism of this classification is correctly. Assigning a lateral pillar
its low interobserver reliability. classification based on the presenting
Lateral pillar classification was radiographs has been found to be
originally designed as a three- premature in 33% of patients, whose
category system (group A, B, and C), hips showed worsening of the lateral
but it was recently modified to in- pillar height over time.37 One ap-
clude a fourth group: B/C border.32 proach has been to wait until the pa-
The classification is based on the tient can be classified before institut-
height of the lateral 15% to 30% of ing treatment. An argument for this
the epiphysis, which is called the lat- “wait-and-classify” approach is that
eral pillar. The three-category lateral it prevents the likelihood of operat-
pillar classification has been shown ing on a patient who would not oth-
to have better interobserver reliabil- erwise have needed surgery (lateral
ity than the Catterall classification.36 pillar A) or who would not have
Both the Catterall and the lateral benefited from surgery (lateral pillar
pillar classification are applicable C). However, if the main goal of
during the fragmentation stage when treatment is to prevent deformity,
femoral head deformity occurs. This then treatment should be instituted
poses a dilemma for patients who early in the older patient (>8 years)
present at the initial stage or the rather than postponed until the head

680 Journal of the American Academy of Orthopaedic Surgeons


Table 3
Stulberg Outcome of Five Treatments Reported on by the Perthes Study
Group13
Stulberg Radiographic Outcome (%)
Age at Onset (yr)
6 to 8
No treatment
Range of motion
Brace
Innominate osteotomy
Femoral osteotomy
>8
No treatment
Range of motion
Brace
Innominate osteotomy
Femoral osteotomy
Adapted with permission from Herring JA, Kim HT, Browne R: Legg-Calvé-Perthes disease:
Part II. Prospective multicenter study of the effect of treatment on outcome. J Bone Joint
Surg Am 2004;86(10):2121-2134.
deforms and is classified, because
these patients have less remodeling
potential than do younger patients.
Such arguments underscore the limi-
tations of these classifications, which
are not sufficiently prospective to de-
termine the prognosis before the de-
velopment of deformity. These classi-
fications may not be applicable for
patients aged >12 years, in whom
femoral head collapse and the lack of
remodeling is more like that of adult
patients with osteonecrosis.38
Management
Two recent multicenter prospective
cohort studies, one from the Perthes
Study Group (PSG)13 and the other
from Norway,14 provide the highest
level of evidence (level II) to date on
the treatment of LCPD. Because
these and other studies show a differ-
ence in outcome depending on pa-
tient age at disease onset, for the pur-
poses of discussion we will group
patients based on age at onset: <6
years, 6 to 8 years, and >8 years. The
November 2010, Vol 18, No 11
I or II III, IV, or V
27 73
48 52
62 38
69 31
68 32
25 75
30 70
36 64
41 59
62 38
following discussion pertains to pa-
tients at the stage of increased ra-
diodensity or fragmentation. The pa-
tient in the stage of reossification, or
healed stage, does not generally re-
quire active treatment unless he or
she is symptomatic, has hinge abduc-
tion, or develops late sequelae, such
as central head osteochondritis disse-
cans or anterior femoroacetabular
impingement.39
Age at Onset <6 Years
Most patients in whom disease onset
occurs earlier than age 6 years
achieve Stulberg class I/II hips at ma-
turity. In a large retrospective study,
80% of hips were found to have
good results with symptomatic or
nonsurgical management only.15 An-
other recent retrospective study com-
paring the results of Salter innomi-
nate osteotomy with nonsurgical
management of Catterall group
III/IV hips found no significant dif-
ference between treatments.16
The study by Wiig et al14 currently
offers the best level of evidence for
Harry K. W. Kim, MD
nonsurgical treatment of patients
aged <6 years. Patients with >50%
head involvement (Catterall III/IV)
who were treated with physiother-
apy, Scottish Rite orthosis (SRO), or
femoral varus osteotomy produced
53%, 46%, and 52% Stulberg I/II
hips, respectively, at 5-year follow-
up. No significant difference was
found between the treatments.
Not all patients in this age group
have a good radiographic outcome.
According to the studies noted
above, one to two patients in five
developed Stulberg III or worse
hips.14,15 These results raise the ques-
tion of how to identify persons with
a poor prognosis and treat them
more effectively. These studies show
no added benefits in outcome with
surgical management; thus, it ap-
pears that currently, patients in this
age group are best treated nonsurgi-
cally.
Age at Onset 6 to 8 Years
The treatment results for children
aged 6 to 8 years are less clear, with
two prospective studies indicating
different findings. The PSG study
showed no statistically significant
difference between hips in the non-
surgical and surgical group.13 How-
ever, the rate of good outcome varied
noticeably between the treatment
groups (Table 3). The lower success
rate in the “no treatment” group
(27%) compared with a much higher
success rate in the bracing (62%)
and surgical groups (68% and 69%)
suggests that the study may have
been underpowered.
In the study by Wiig et al,14 pa-
tients treated with femoral varus os-
teotomy had significantly better ra-
diographic results (ie, Stulberg I/II
hips) than patients treated with ei-
ther SRO (43% versus 20%, respec-
tively; P = 0.001) or physiotherapy
(43% versus 33%, respectively; P =
0.001). A few obvious differences be-
681
Legg-Calvé-Perthes Disease
tween this study and the PSG study
are noteworthy. This study did not
stratify patients into groups by age
(6 to 8 years, >8 years), which makes
it difficult to compare the results.
Another difference is that in this
study the final follow-up was 5
years, when the healed stage was
reached,14 whereas in the study by
PSG the final follow-up was at skele-
tal maturity.13 Finally, Wiig et al14 ini-
tiated treatment at the stage of frag-
mentation, whereas the PSG group
did so at the stage of increased ra-
diodensity or early fragmentation in
>95% of their patients.13 Although
this may not explain why Wiig et al14
found a significant difference be-
tween surgical and nonsurgical man-
agement while the PSG study did
not, it may explain the lower per-
centage with a good result in patients
treated with femoral varus osteot-
omy in the former versus the latter
(43% versus 68%, respectively). A
retrospective study of 640 patients
suggests that timing of femoral varus
osteotomy is important and that re-
sults are better with early surgery.17
The evidence from one prospective
study favors femoral varus osteot-
omy over both physiotherapy and
SRO,14 whereas the other study
found no difference between nonsur-
gical and surgical treatments in pa-
tients aged 6 to 8 years at disease on-
set.13 Femoral varus osteotomy
performed at the stage of fragmenta-
tion had a modest effect on achieving
Stulberg I/II hips at 5-year follow-up
(43%).14
Age at Onset >8 Years
According to the PSG study, Stulberg
I/II hips were reported in 25% of pa-
tients with no treatment, 30%
treated with range of motion, 36%
treated with SRO, 41% treated with
innominate osteotomy, and 62%
treated with femoral varus osteot-
omy13 (Table 3). Although these re-
682
sults suggest superiority of surgical
treatments, especially femoral varus
osteotomy, the difference was not
found to be statistically significant.
Insufficient power of the study due
to the relatively small sample sizes
cannot be ruled out. An analysis of
the results based on the lateral pillar
classification did show a beneficial
effect with surgery compared with
nonsurgical management for the lat-
eral pillar B and B/C border groups
but not for persons classified as
group C. Clinical applicability of the
treatment recommendations based
on the lateral pillar classification,
however, is controversial because the
surgical treatments in the study were
rendered at early stages when the
classification could not be applied.
The possibility that different treat-
ments affect the lateral pillar height
differently cannot be ruled out.40
These observations raise the question
whether patients aged >8 years
should be treated with early surgery
or whether surgery should be post-
poned until the lateral pillar classifi-
cation can be determined. The effec-
tiveness of surgery on achieving
Stulberg I/II hips in this age group is
also modest: 41% with Salter innom-
inate osteotomy and 62% with fem-
oral varus osteotomy.13
The results of the surgical treat-
ments raise the question of why good
results were not obtained in all pa-
tients. Both osteotomies are based on
the concept of containment, which
proposes that to prevent deformities
of the affected femoral head, the
head must be contained within the
depth of the acetabulum, thereby
equalizing the pressure on the head
and subjecting it to the molding ac-
tion of the acetabulum. The mechan-
ical concept does not directly address
the pathologic repair process in the
head, such as the predominance of
bone resorption and delayed bone
formation. Although the mechanical
effect of the surgeries and other un-
Journal of the American Academy of Orthopaedic Surgeons
identified factors associated with sur-
gery may be adequate for some pa-
tients, it may be insufficient for those
with slower, impaired healing, who
may have limited potential for re-
modeling of the deformed head. Al-
though some have advocated more
extensive containment procedures,
such as combined femoral varus and
Salter innominate osteotomies or a
triple pelvic osteotomy, there is cur-
rently no evidence demonstrating
that an aggressive mechanical ap-
proach produces better results. Alter-
natively, others have advocated a
longer period of protected weight
bearing postoperatively and make
the decision to allow return to nor-
mal weight bearing based on healing
of the femoral head.
Other Treatment Methods
Petrie casting is one method of non-
surgical containment. The effective-
ness of this prolonged treatment has
been reported only in retrospective
studies.41 These studies show results
comparable to those with other
treatments, including pelvic osteot-
omy and femoral varus osteotomy.
Currently, no standardized treatment
protocol exists, but we have found
that shortening the duration of cast-
ing to 6 weeks can lead to a recur-
rence of stiffness. The authors’ pre-
ferred approach incorporates a wide
abduction brace called an “A-
frame,” which serves as an alterna-
tive method to maintain femoral
head containment following 6 weeks
of Petrie casting (Figure 4). The
A-frame brace is used initially for 12
hours at night for 3 months, after
which bracing is tapered to 8 hours
at night to maintain hip abduction.
The brace is left off during the day to
allow ambulation using a walker or
crutches and to allow leg motion.
Non–weight-bearing status of the af-
fected leg and nighttime bracing are
 Harry K. W. Kim, MD

Figure 4

A, AP (top) and lateral (bottom) radiographs of a boy

aged 8 years 6 months who presented with hip pain,
stiffness, and adduction and flexion contractures. He
had been treated previously with shelf acetabuloplasty.
The patient was treated with a few days of bed rest
followed by adductor tenotomy and Petrie casting (B,
top) for 6 weeks. Subsequently, a night-time A-frame
abduction brace was used for 1 year (B, bottom). C, AP
(top) and lateral (bottom) radiographs obtained at 4-year
follow-up demonstrating the affected femoral head in the
healed stage with restoration of the Shenton line and an
improvement in femoral head shape.

maintained for approximately 1 year. problems, such as limb shortening, from overcoverage. Major disadvan-
The proposed advantages include abductor muscle weakening, and tages are prolonged duration of
avoidance of surgery and related femoral head impingement resulting treatment, the cumbersome nature of

November 2010, Vol 18, No 11 683

Legg-Calvé-Perthes Disease

Figure 5

An 11-year-old girl with hip stiffness, no passive hip abduction, lateral subluxation, and severe flattening of the femoral
head was treated with adductor tenotomy and Petrie casting for 6 weeks, followed by proximal femoral varus
osteotomy and protected weight-bearing postoperatively for an extended period. AP radiographs at the time of
presentation (A) and after proximal femoral varus osteotomy (B). AP (C) and lateral (D) follow-up radiographs at age
19 years, demonstrating improvement in the position of the femoral head and femoral head flattening compared with
the pretreatment state.

the A-frame, and loss of compliance
over time.
Shelf acetabuloplasty is used by
some surgeons to improve acetabular
coverage of the femoral head. The
procedure may also stimulate an in-
crease in acetabular depth.42 Prospec-
tive studies examining the results of
this procedure are lacking. In a study
reviewing the results of patients aged
between 8 and 13 years at onset
treated with shelf acetabuloplasty, 14
of 27 hips were reported to be Stul-
berg I/II at maturity.43
In preliminary studies, hip distrac-
tion using an external fixator for 4
to 5 months has been shown to have
either a protective effect on the fem-
oral head when applied in the early
stages or a restorative effect on the
femoral head height when applied
at the fragmentation stage.44 A
follow-up study showed that the
femoral head height gained with dis-
traction was subsequently lost after
the removal of the distractor, with 7
of 10 hips having Stulberg IV hips.45
No one best treatment has been
identified for patients who present
with a deformed head and hinge ab-
duction. In the active phase, Petrie
casting followed by either an
A-frame brace or surgical contain-
ment remains an option (Figure 5). It
is unknown whether one treatment is
superior to another. For patients in
the reossification or healed stage,
valgus femoral osteotomy has been
shown to improve function and hip
scores at a mean follow-up of 5 to 7
years.46,47 In a 10-year follow-up
study consisting of 48 patients, 4 re-

684 Journal of the American Academy of Orthopaedic Surgeons


quired total hip replacement, 1 un-
derwent arthrodesis, and 6 required
repeat valgus osteotomy for recur-
rence or fixed adduction.48
Summary
LCPD was recognized as a separate
disease entity 100 years ago, and
knowledge of the disease has grown
considerably since then. Well-
recognized prognostic factors include
degree of deformity, age at onset, ex-
tent of head involvement, lateral pil-
lar collapse, and head-at-risk signs.
Although the Catterall and the lat-
eral pillar classifications are useful
guides in managing younger patients
with good remodeling potential,
their role in managing older patients
with poor remodeling potential is
controversial. The controversy un-
derscores the need to develop an
early prognosticator, such as MRI,
which can be obtained and applied
before the development of deformity.
In general, treatment should be
guided by age at disease onset, cur-
rent best evidence, and prognostic
factors. Patients aged <6 years at dis-
ease onset appear best treated non-
surgically, whereas surgical treat-
ment may benefit older patients. The
efficacy of surgical treatment in
achieving a normal hip at maturity is
modest, however, emphasizing the
need to develop new treatments that
specifically address the biologic and
mechanical aspects of the disease.
References
Evidence-based Medicine: Levels of
evidence are described in the table of
contents. In this article, reference 23 is
a level I study. References 7, 8, 13,
14, 44, and 46 are level II studies.
References 6, 15, 16, 40, and 42 are
level III studies. References 12, 17,
22, 27, 28, 30, 31, 35, 37, 38, 41,
November 2010, Vol 18, No 11
43, 45, 47, and 48 are level IV stud-
ies. In the remaining references, the
level of evidence assignment was not
applicable.
Citation numbers printed in bold
type indicate references published
within the past 5 years.
1. Legg AT: An obscure affection of the
hip-joint. Boston Med and Surg J 1910;
162:202-204.
2. Calvé J: Sur une forme particulière de
pseudo-coxalgie greffée sur des
déformations caractéristiques de
l’extrémité supérieure du fémur. Rev
Chir 1910;42:54-84.
3. Perthes G: Über arthritis deformans
juvenilis. Deutsche Zeitschr Chir 1910;
107:111-159.
4. Miyamoto Y, Matsuda T, Kitoh H, et al:
A recurrent mutation in type II collagen
gene causes Legg-Calvé-Perthes disease
in a Japanese family. Hum Genet 2007;
121(5):625-629.
5. Su P, Li R, Liu S, et al: Age at onset-
dependent presentations of premature
hip osteoarthritis, avascular necrosis of
the femoral head, or Legg-Calvé-Perthes
disease in a single family, consequent
upon a p.Gly1170Ser mutation of
COL2A1. Arthritis Rheum 2008;58(6):
1701-1706.
6. Glueck CJ, Glueck HI, Greenfield D,
et al: Protein C and S deficiency,
thrombophilia, and hypofibrinolysis:
Pathophysiologic causes of Legg-Perthes
disease. Pediatr Res 1994;35(4 pt 1):
383-388.
7. Hresko MT, McDougall PA, Gorlin JB,
Vamvakas EC, Kasser JR, Neufeld EJ:
Prospective reevaluation of the
association between thrombotic diathesis
and legg-perthes disease. J Bone Joint
Surg Am 2002;84(9):1613-1618.
8. Balasa VV, Gruppo RA, Glueck CJ, et al:
Legg-Calve-Perthes disease and
thrombophilia. J Bone Joint Surg Am
2004;86(12):2642-2647.
9. Kim HK, Su PH: Development of
flattening and apparent fragmentation
following ischemic necrosis of the capital
femoral epiphysis in a piglet model.
J Bone Joint Surg Am 2002;84(8):1329-
1334.
10. Pringle D, Koob TJ, Kim HK:
Indentation properties of growing
femoral head following ischemic
necrosis. J Orthop Res 2004;22(1):122-
130.
11. Koob TJ, Pringle D, Gedbaw E,
Meredith J, Berrios R, Kim HK:
Biomechanical properties of bone and
cartilage in growing femoral head
Harry K. W. Kim, MD
following ischemic osteonecrosis.
J Orthop Res 2007;25(6):750-757.
12. McAndrew MP, Weinstein SL: A long-
term follow-up of Legg-Calvé-Perthes
disease. J Bone Joint Surg Am 1984;
66(6):860-869.
13. Herring JA, Kim HT, Browne R: Legg-
Calvé-Perthes disease: Part II.
Prospective multicenter study of the
effect of treatment on outcome. J Bone
Joint Surg Am 2004;86(10):2121-2134.
14. Wiig O, Terjesen T, Svenningsen S:
Prognostic factors and outcome of
treatment in Perthes’ disease: A
prospective study of 368 patients with
five-year follow-up. J Bone Joint Surg Br
2008;90(10):1364-1371.
15. Rosenfeld SB, Herring JA, Chao JC:
Legg-calve-perthes disease: A review of
cases with onset before six years of age.
J Bone Joint Surg Am 2007;89(12):2712-
2722.
16. Canavese F, Dimeglio A: Perthes’
disease: Prognosis in children under six
years of age. J Bone Joint Surg Br 2008;
90(7):940-945.
17. Joseph B, Rao N, Mulpuri K, Varghese
G, Nair S: How does a femoral varus
osteotomy alter the natural evolution of
Perthes’ disease? J Pediatr Orthop B
2005;14(1):10-15.
18. Catterall A, Pringle J, Byers PD, et al: A
review of the morphology of Perthes’
disease. J Bone Joint Surg Br 1982;64(3):
269-275.
19. Hofstaetter JG, Roschger P, Klaushofer
K, Kim HK: Increased matrix
mineralization in the immature femoral
head following ischemic osteonecrosis.
Bone 2010;46(2):379-385.
20. Aya-ay J, Athavale S, Morgan-Bagley S,
Bian H, Bauss F, Kim HK: Retention,
distribution, and effects of intraosseously
administered ibandronate in the
infarcted femoral head. J Bone Miner
Res 2007;22(1):93-100.
21. Kim HK, Randall TS, Bian H, Jenkins J,
Garces A, Bauss F: Ibandronate for
prevention of femoral head deformity
after ischemic necrosis of the capital
femoral epiphysis in immature pigs.
J Bone Joint Surg Am 2005;87(3):550-
557.
22. Agarwala S, Shah S, Joshi VR: The use
of alendronate in the treatment of
avascular necrosis of the femoral head:
Follow-up to eight years. J Bone Joint
Surg Br 2009;91(8):1013-1018.
23. Lai KA, Shen WJ, Yang CY, Shao CJ,
Hsu JT, Lin RM: The use of alendronate
to prevent early collapse of the femoral
head in patients with nontraumatic
osteonecrosis: A randomized clinical
study. J Bone Joint Surg Am 2005;
87(10):2155-2159.
24. Bergmann G, Deuretzbacher G, Heller
685
Legg-Calvé-Perthes Disease
M, et al: Hip contact forces and gait
patterns from routine activities.
J Biomech 2001;34(7):859-871.
25. Molloy M, MacMahon B: Incidence of
Legg-Perthes disease (osteochondritis
deformans). N Engl J Med 1966;
275(18):988-990.
26. Waldenström H: The definitive forms of
coxa plana. Acta Radiol 1922;1:384.
27. Herring JA, Williams JJ, Neustadt JN,
Early JS: Evolution of femoral head
deformity during the healing phase of
Legg-Calvé-Perthes disease. J Pediatr
Orthop 1993;13(1):41-45.
28. Lamer S, Dorgeret S, Khairouni A, et al:
Femoral head vascularisation in Legg-
Calvé-Perthes disease: Comparison of
dynamic gadolinium-enhanced
subtraction MRI with bone scintigraphy.
Pediatr Radiol 2002;32(8):580-585.
29. Pienkowski D, Resig J, Talwalkar V,
Tylkowski C: Novel three-dimensional
MRI technique for study of cartilaginous
hip surfaces in Legg-Calvé-Perthes
disease. J Orthop Res 2009;27(8):981-
988.
30. Gower WE, Johnston RC: Legg-Perthes
disease: Long-term follow-up of thirty-
six patients. J Bone Joint Surg Am 1971;
53(4):759-768.
31. Stulberg SD, Cooperman DR, Wallensten
R: The natural history of Legg-Calvé-
Perthes disease. J Bone Joint Surg Am
1981;63(7):1095-1108.
32. Herring JA, Kim HT, Browne R: Legg-
Calve-Perthes disease: Part I.
Classification of radiographs with use of
the modified lateral pillar and Stulberg
classifications. J Bone Joint Surg Am
2004;86(10):2103-2120.
686
33. Nelson D, Zenios M, Ward K,
Ramachandran M, Little DG: The
deformity index as a predictor of final
radiological outcome in Perthes’ disease.
J Bone Joint Surg Br 2007;89(10):1369-
1374.
34. Salter RB, Thompson GH: Legg-Calvé-
Perthes disease: The prognostic
significance of the subchondral fracture
and a two-group classification of the
femoral head involvement. J Bone Joint
Surg Am 1984;66(4):479-489.
35. Catterall A: Legg-Calvé-Perthes
syndrome. Clin Orthop Relat Res 1981;
158:41-52.
36. Ritterbusch JF, Shantharam SS, Gelinas
C: Comparison of lateral pillar
classification and Catterall classification
of Legg-Calvé-Perthes’ disease. J Pediatr
Orthop 1993;13(2):200-202.
37. Lappin K, Kealey D, Cosgrove A:
Herring classification: How useful is the
initial radiograph? J Pediatr Orthop
2002;22(4):479-482.
38. Joseph B, Mulpuri K, Varghese G:
Perthes’ disease in the adolescent. J Bone
Joint Surg Br 2001;83(5):715-720.
39. Eijer H, Podeszwa DA, Ganz R, Leunig
M: Evaluation and treatment of young
adults with femoro-acetabular
impingement secondary to Perthes’
disease. Hip Int 2006;16(4):273-280.
40. Kuroda T, Mitani S, Sugimoto Y, et al:
Changes in the lateral pillar classification
in Perthes’ disease. J Pediatr Orthop B
2009;18(3):116-119.
41. Grzegorzewski A, Bowen JR, Guille JT,
Glutting J: Treatment of the collapsed
femoral head by containment in Legg-
Calve-Perthes disease. J Pediatr Orthop
2003;23(1):15-19.
Journal of the American Academy of Orthopaedic Surgeons
42. Domzalski ME, Glutting J, Bowen JR,
Littleton AG: Lateral acetabular growth
stimulation following a labral support
procedure in Legg-Calve-Perthes disease.
J Bone Joint Surg Am 2006;88(7):1458-
1466.
43. Daly K, Bruce C, Catterall A: Lateral
shelf acetabuloplasty in Perthes’ disease:
A review of the end of growth. J Bone
Joint Surg Br 1999;81(3):380-384.
44. Maxwell SL, Lappin KJ, Kealey WD,
McDowell BC, Cosgrove AP:
Arthrodiastasis in Perthes’ disease:
Preliminary results. J Bone Joint Surg Br
2004;86(2):244-250.
45. Segev E, Ezra E, Wientroub S, Yaniv M,
Hayek S, Hemo Y: Treatment of severe
late-onset Perthes’ disease with soft
tissue release and articulated hip
distraction: Revisited at skeletal
maturity. J Child Orthop 2007;1(4):229-
235.
46. Myers GJ, Mathur K, O’Hara J: Valgus
osteotomy: A solution for late
presentation of hinge abduction in Legg-
Calvé-Perthes disease. J Pediatr Orthop
2008;28(2):169-172.
47. Yoo WJ, Choi IH, Chung CY, Cho TJ,
Kim HY: Valgus femoral osteotomy for
hinge abduction in Perthes’ disease:
Decision-making and outcomes. J Bone
Joint Surg Br 2004;86(5):726-730.
48. Bankes MJ, Catterall A, Hashemi-Nejad
A: Valgus extension osteotomy for ‘hinge
abduction’ in Perthes’ disease: Results at
maturity and factors influencing the
radiological outcome. J Bone Joint Surg
Br 2000;82(4):548-554.