ORIGINAL ARTICLE

Advanced Containment Methods for Legg-Calve-Perthes

Disease: Results of Triple Pelvic Osteotomy
Dennis R. Wenger, MD,*w Maya E. Pring, MD,*w Harish S. Hosalkar, MD,*
Christine B. Caltoum, MD,z Francois D. Lalonde, MD,y and Tracey P. Bastrom, MA*

Background: Although surgical containment has become a

mainstay for the treatment of Legg-Calve-Perthes (LCP)
disease; traditional operations (varus osteotomy of femur, Salter
osteotomy) have certain limitations, sometimes resulting in a
prolonged limp or inadequate containment. This paper presents
the surgical method and results of triple pelvic osteotomy for
containment treatment of LCP disease.
Methods: This retrospective review of 39 children (40 hips; age 5
to 13 y) with LCP disease treated with triple pelvic osteotomy
(1995 to 2005) included preoperative lateral pillar assessment
and other measurements. Final follow-up lms (minimum
3 y, range 3 to 9 y) were assessed using the modied Stulberg
classication. Clinical follow-up evaluation assessed limp, limblength inequality, range of motion, and activity level.
Results: Twenty-one (53%) hips were graded as lateral pillar B
and 19 (48%) were lateral pillar C. Four patients required
further treatment before the nal follow-up. At nal follow-up,
42% had a good outcome (Stulberg I/II), 47% had a fair
outcome (Stulberg III), and 11% had a poor outcome. Thus,
89% of patients had satisfactory (good or fair) results. There
was a signicant dierence in outcome based on the preoperative lateral pillar, with B hips more likely to have a good
outcome (65%) compared with lateral pillar C hips (12.5%)
(P = 0.002). There were no lateral pillar B patients with a poor
outcome. Seventeen percent of the lateral pillar C patients more
than or equal to age 8 had a poor outcome compared with 50%
being more than age 8 with a poor outcome. Four patients (3
lateral pillar C, 1 lateral pillar B) required further surgery.
Conclusions: Triple pelvic osteotomy resulted in maintenance of
head shape in lateral pillar B patients of all ages and in younger
lateral pillar C patients. Lateral pillar C patients over age 8 were
more dicult to treat, however, we still advise containment for
these cases because methods are now available to deal with
containment failure. Triple pelvic osteotomy is an eective

From the *Department of Orthopedics and Scoliosis, Rady Childrens

Hospital-San Diego; wDepartment of Orthopedic Surgery, University
of California-San Diego; yDepartment of Orthopedics, Childrens
Hospital of Orange County, Orange, CA; and zDepartment of
Orthopedics, Riley Childrens Hospital, Indianapolis, IN.
This project was supported in part by the Childrens Specialists
Orthopedic Education and Research Fund. No other funding was
received for this study.
Study conducted at Rady Childrens Hospital, San Diego, CA.
Reprints: Dennis R. Wenger, MD, 3030 Childrens Way, Ste 410,
San Diego, CA 92123. E-mail: orthoedu@rchsd.org.
Copyright r 2010 by Lippincott Williams & Wilkins

J Pediatr Orthop

Volume 30, Number 8, December 2010

treatment method for LCP patients with lateral pillar B disease

and younger patients with lateral pillar C disease. This method
provides eective containment, which allows prolonged remodeling while avoiding the limitations of femoral varus osteotomy
(limp, short limb) and Salter osteotomy (incomplete containment).
Level of Evidence: Level IV.
Key Words: Legg-Calve-Perthes, surgical containment, triple
pelvic osteotomy
(J Pediatr Orthop 2010;30:749757)

egg-Calve-Perthes (LCP) disease is a childhood hip

disease in which the femoral head unexpectedly loses
a portion of its blood supply with resulting avascular
necrosis. Subsequent revascularization leads to early
softening of the femoral head (biologic plasticity)1
followed by femoral head shape change, attening and
subluxation (in severe cases). These changes can lead to
premature hip arthritis.26
Risk factors for femoral head deformity in LCP
disease include age at onset and the extent of proximal
femoral epiphyseal involvement.2,4 The greater the
femoral head deformity and joint incongruity at skeletal
maturity, the worse the prognosis for long term hip
function.2 Accordingly, the treatment is designed to
minimize femoral head deformity with a goal of delaying
the onset of premature degenerative joint disease.
Containment treatment is designed to center the
femoral head within the acetabulum during the fragmentation and reossication phase. This allows the acetabulum to
serve as a mold during the healing (revascularization) phase
when the biologically plastic femoral head4 is at risk for
subluxation, hinge abduction, and permanent femoral head
deformation.
Containment can be achieved nonoperatively or
operatively. Nonoperative containment measures, using
either abduction casts or bracing have been ineective
owing to impracticality (prolonged casting), lack of
cooperation by the patient (child takes brace o), or
poor brace design.7 Accordingly, surgical containment
has become widely accepted as the best method for
sustained femoral head containment.8,9 The most common methods for surgical containment have been either
proximal femoral varus osteotomy5 or Salter innominate
www.pedorthopaedics.com |

749

Wenger et al

osteotomy1,1015 with little dierence found in the radiographic or functional outcomes after treatment with
either of these 2 methods.
Experience has suggested certain limitations to
the application of either femoral osteotomy or Salter
osteotomy for the treatment of LCP disease. The degree
of femoral varus required to contain the femoral head
further shortens the limb and may lead to a permanent
limp, particularly in older children.
Sponseller et al14 found that the center edge angle of
Wiberg, limb lengths, and range of motion were closer to
normal after innominate osteotomy, however Rab found
that the Salter osteotomy does not provide adequate
acetabular rotation to contain the femoral head in more
severe cases.16 This can lead to iatrogenic hinge abduction
(if the femoral head is not fully contained) in severe
cases.17
As of the above limitations, advanced containment
methods have been developed for more severe cases
including: Salter innominate osteotomy combined with a
proximal femoral varus osteotomy,18,19 triple pelvic
osteotomy,20 and shelf acetabuloplasty.21 As of the
problems we had experienced associated with Salter
osteotomy and femoral varus osteotomy, we began using
triple pelvic osteotomy as a means of surgical containment for LCP patients requiring prolonged containment
treatment (Fig. 1). We theorized that we could achieve
more complete femoral head containment than could be
achieved with Salter osteotomy alone and decrease the
risk for limp and limb leg discrepancy associated with
femoral varus osteotomy.

J Pediatr Orthop

Volume 30, Number 8, December 2010

In this paper, we present the clinical and radiographic results of triple pelvic osteotomy in more severe
forms of LCP disease (lateral pillar grade B, B/C border,
or C). We compare our results with published results of
the natural history of the disease and other reported
results for surgical containment.

METHODS
This study provides a retrospective review of the
patients with LCP disease treated with triple pelvic
osteotomy at our institution between 1995 and 2005.
The study was approved by the Institutional Review
Board. Of the 63 patients treated with this method, 40
hips (39 patients) had a minimum of 3 year follow-up
with a complete set of radiographs. A signicant number
of patients had moved from the region leaving 39 patients
(40 hips) with a full data set.
The study group included 36 male patients and 3
female patients with a mean age at diagnosis of 7.5 years
(range 5 to 13) and a mean age at surgery of 8.3 years
(range 5 to 13). One patient had bilateral involvement.
Mean follow-up was 64 months (range 32 to 113 mo). All
hips underwent a diagnostic arthrogram, adductor
tenotomy, and Petrie casting (4 weeks) before performance of the osteotomy.
All radiographs were evaluated with key measurements made on the immediate preoperative and the
nal follow-up lms. Waldenstrom stage22 at the time
of surgery was graded in each patient. Preoperative extent of disease was graded according to the Catterall

FIGURE 1. A, AP and frog view of the pelvis in a 7-year-old male patient with left Perthes diseaseFHerring C classification.
B, Arthrogram carried out with the hip in abduction and internal rotation shows marked medial dye pooling with the lateral
portion of the femoral head pushing the labrum upward. The child was placed in Petrie casts for 4 weeks to begin head reshaping
before triple pelvic osteotomy. C, AP view of the pelvis (in hip spica) immediately after the triple pelvic osteotomy. D, AP and frog
view of the pelvis carried out 7 years after triple pelvic osteotomy. Femoral head sphericity has been maintained.

750 | www.pedorthopaedics.com

2010 Lippincott Williams & Wilkins

J Pediatr Orthop

Volume 30, Number 8, December 2010

classication2 and the lateral pillar classication of

Herring et al.23
These standard measurements were obtained preoperatively and at latest follow-up: center-edge angle of
Wiberg,24 extrusion index,25 lateral acetabular shape,
articular-trochanteric distance ratio, and neck-shaft angle.
The 3-fold classication system of Tonnis was used for the
lateral acetabular shape: group 1 = normal concave
acetabular roof; group 2 = a horizontal at roof; and
group 3 = a rounded and steep roof.
At the latest follow-up, radiographic outcome was
assessed using the modied Stulberg classication proposed by Herring et al.23 According to this modied
Stulberg classication, an outcome of grade I or II is
considered good, an outcome of grade III is fair, and
grade IV/V as poor. Clinically, the patients were
evaluated for limp, limb-length inequality, range of
motion, and restriction of activities.

Results of Triple Pelvic Osteotomy for Perthes

Surgical Method
Our surgical technique is somewhat similar to the
method described by Carlioz,26 which was described in
the French literature in 1982. The supraacetabular cut is
identical to that of Salter. Our superior pubic ramus cut is
made through the medial groin incision because we found
that this approach is less risky as there is little or no
retraction of the femoral neurovascular bundle. The groin
incision is made in a transverse fashion distal to the groin
crease, centering over the adductor longus tendon. The
ischial cut is made just below the acetabulum. Our
approach allows all cuts to be made relatively near the
acetabulum which allows free acetabular rotation, as was
noted by Tonnis. In contrast to the Tonnis procedure,
we avoid a separate postero-lateral incision with our
approach.

Sequence and Technique

We carry out the triple pelvic osteotomy in this
sequence; (1) iliac osteotomy, (2) pubic osteotomy, and
(3) ischial osteotomy (Fig. 2).
I. The Salter-type iliac cut is made through a typical
antero-lateral incision. Intramuscular lengthening of
the psoas is carried out at the pelvic brim to decrease
joint forces over the hip joint. Specially designed Rang
retractors placed in the sciatic notch make passing the
Gigli exible wire saw easier.
II. The pubic osteotomy is carried out through a transverse
medial groin incision (that can also be used for the
ischial osteotomy) in most patients. A Foley urinary
catheter can be placed (to decrease the risk for bladder
injury during the pubic osteotomy) but is not routinely
used in our center. A separate 2 to 3-cm transverse
incision (parallel to the inguinal ligament) can be used
for the pubic osteotomy in very large patients with a
subsequent separate ischial incision.
III. The ischial osteotomy is usually carried out through
the single medial incision which is extended posteriorly
toward the ischial tuberosity. An assistant exes the
r

2010 Lippincott Williams & Wilkins

FIGURE 2. A, Drawing illustrating our sequence for carrying

out triple pelvic osteotomy. The first cut is a typical Salter
innominate osteotomy, the second is made in the superior
pubic ramus, and the third in the ischium, just below the
acetabulum. The ischial cut is the most difficult to carry out
and therefore, is done last, allowing the surgeon to sense
freedom of the acetabular segment once the posterior ischial
cortex has been cut. B, Radiograph showing the principles
of acetabular redirection with triple pelvic osteotomy. The
acetabulum is rotated in the coronal plane, with the pubis
moved upward and inward and the ischium moved medially.
The effect is to improve anterior and lateral coverage of the
femoral head without lateralizing the hip joint. Care is taken to
avoid creating acetabular retroversion.

hip 90 degrees to allow visualization and the adductor

magnus origin is detached using an electro-cautery.
A Cobb elevator is used to follow the ischium up to
its base just below the acetabulum with 3 Hohman
www.pedorthopaedics.com |

751

J Pediatr Orthop

Wenger et al

retractors providing retraction. The ischial cut is

made 1 to 2 cm below the acetabulum (conrmed by
image intensier) using a long osteotome (required
because of depth).
Correct acetabular positioning is critical to the
success of the operation (Fig. 2B). To guide the
acetabular segment, a Schanz screw is placed in the ilium
above the joint line with a ball-point pusher placed in the
pubic ramus just lateral to the pubic osteotomy. The
pubis is pushed upward and medially with the ball point
pusher whereas the Schanz screw is used to lever and
rotate the acetabulum forward and laterally in the
coronal plane. Care is given to avoid iatrogenic acetabular retroversion which can create pathologic acetabular mechanics. Image intensier views are carefully
assessed to avoid creating retroversion, to provide an
approximately 20 degrees angle of the teardrop gure,
and to create a horizontal sourcil (Fig. 2B).
A triangular wedge of bone is removed from the
iliac crest and fashioned to t tightly in the gap of the iliac
osteotomy. The graft is smaller than that used for a Salter
osteotomy because with a triple osteotomy, rotation also
occurs at the pubic and ischial cuts. The osteotomy is rst
xed with temporary k-wires, with acetabular position
checked by uoroscopy to conrm containment and
proper acetabular position. Two or three 4.5 mm fully
threaded cortical screws are then placed to x the iliac
osteotomy. In younger children (or any patient with a
small bone structure), threaded K-wires can be used as an
alternative to screw xation. In older children (or any
very large patient), an oblique screw or a small
reconstruction plate can be placed across the superior
pubic osteotomy for more secure xation.
The wounds are closed in layers over a hemovac
drain and a single leg hip spica applied (in most cases for
6 weeks). In older children with adequate xation,
including xation of the pubic cut, the spica cast can be
avoided.

RESULTS
Forty hips in 39 patients met the inclusion criteria.
Preoperatively, all the patients reported pain and all but 1
presented with gait abnormalities. At the time of surgery,
12/40 (30%) of hips were Waldenstrom stage I, 22 (55%)
were Waldenstrom stage II, and 6 (15%) were stage III.
The preoperative Catterall classication included 9 (22.5%)
group 2 hips, 20 (50%) group 3 hips, and 11 (27.5%) group
4 hips. The lateral pillar classication included: 21 hips
(53%) graded as lateral pillar B and19 hips (48%) graded
as lateral pillar C. There were no B/C border patients.
Assessment of lateral acetabular shape (Tonnis) showed
that a majority of the hips were graded as 1 (37%) or 2
(60%), with only 1 patient (3%) with a grade of 3 (Table
1Ffull list of patients).
Details on age at diagnosis, sex, lateral pillar grade,
and outcome are seen in Table 1. Of the 40 hips, 4
required further surgery and will be discussed separately

752 | www.pedorthopaedics.com

Volume 30, Number 8, December 2010

TABLE 1. Patients With Perthes Disease Treated With Triple

Pelvic Osteotomy
Case

Age at
Diagnosis

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21R
21L
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38

6.5
6
6
11
10.1
10.1
6
10
5.8
8
5.8
7.7
5.8
6.6
9.25
9
6.6
4
10
10
6
6
3
11
9
8
7
7
7
6
8
8
11
7
4
6
7
7
5

Sex
Male
Male
Male
Male
Male
Male
Male
Male
Male
Female
Male
Male
Male
Male
Female
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Female
Male
Male
Male

Lateral
Pillar

Further
Surgery

Stulberg
Class

B
B
C
B
B
C
C
B
B
B
B
B
B
C
B
C
C
C
C
B
B
C
B
C
C
B
C
B
B
C
C
C
C
B
C
C
B
C
B

N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
Y
N
N
N
Y
Y
N
N
N
N
N
Y
N
N
N
N
N
N
N
N

III
II
IV
II
III
IV
III
II
II
III
II
II
III
II
II
III
V
III
III
III
II
III
II
IV
IV
II
III
II
II
III
III
II
IV
III
III
III
III
III
II

as their Stulberg result reects the eects of both triple

osteotomy and a second procedure.

Clinical Outcome
Clinically, 35/40 hips were pain-free at nal followup. Five patients had a minimal limp and there were
no clinically signicant leg length discrepancies noted.
Four minor complications occurred: 1 supercial wound
infection that resolved with oral antibiotics; a readmission
for low-grade fever in which no infection was identied
and 1 patient who underwent reapplication of Petrie casts
secondary to a recurrent adduction contracture; and
temporary peroneal division neuropraxia (which resolved
spontaneously in 6 weeks).

Outcome-Stulberg Classification
In the 36 hips that did not require further corrective
surgery, 20 were graded as lateral pillar B and 16 were
lateral pillar C at the time of surgery. Stulberg classication
was assessed at the nal follow-up (average 64 23 mo).
r

2010 Lippincott Williams & Wilkins

J Pediatr Orthop

Volume 30, Number 8, December 2010

Results of Triple Pelvic Osteotomy for Perthes

TABLE 3. Stulberg Classification at Most Recent Follow-up

Based on Preoperative Lateral Pillar Grade
Stulberg Classication
Lateral pillar B
Lateral pillar C

I/II

III

IV/V

65%
12.5%

35%
62.5%

0%
25%

Change in CE Angle and Extrusion Index

FIGURE 3. Bar graph illustrating Stulberg outcome according

to age and lateral pillar involvement. All lateral pillar B patients
had a good or fair outcome, regardless of age. Lateral pillar C
results had fewer good and fair results, especially in patients
over age 8 years.

Of these 36 hips, 15 (42%) had a good outcome (Stulberg

I/II), 17 (47%) had a fair outcome (Stulberg III), and 4
(11%) had a poor outcome (Fig. 3).
There was no signicant dierence in the distribution of Stulberg outcome grades based on Catterall
groupings (P = 0.41). These data are represented in
Table 2.
There was a signicant dierence in the distribution
of Stulberg grades between lateral pillar B and C hips
(P = 0.002, Table 3). Hips that presented as lateral pillar
B were signicantly more likely to have a good outcome
(65%) compared with those graded as lateral pillar C
(12.5%). There were 4 poor outcomes observed in these
36 hips. All 4 of these poor outcomes were in lateral pillar
C (25%) hips versus 0% in the lateral pillar B hips.
We further analyzed the results with regards to
age and lateral pillar classication. The patients were
subdivided into lateral pillar B or C based on age at
diagnosis (less than 8 years or older than 8 years-Fig. 3).
All lateral pillar B patients had a good or fair result,
which seems to be independent of age. However, with the
lateral pillar C patients, a poor outcome occurred in only
17% of patients under the age of 8 versus 50% of patients
with lateral pillar C disease over the age of 8.
The distribution of good, fair, and poor outcome
was similar for patients that presented with either Tonnis
grade 1 or 2 for lateral acetabular shape (P = 0.56).

The central edge (CE) angle increased signicantly

from the preoperative (24.6 8.2) to nal postoperative
lm (41.8 9.2) for the entire cohort (Pr0.001) (Fig. 4).
Several of the abnormally high CE angle values were
concentrated early in the study before we became aware
of the risk for creating iatrogenic impingement. No
signicant interaction between change in CE angle and
lateral pillar class was observed (P = 0.89) (Fig. 4).
Extrusion index decreased signicantly from the preoperative (0.21 0.14) to the nal postoperative lm
(0.08 0.13) for the entire cohort (Pr0.001). No
signicant interaction between extrusion index and lateral
pillar class was observed (P = 0.20) (Fig. 5). We are
aware that in terms of adult CE angles, these may be
dened as being hyper-contained with the potential for
pincer type impingement. We believe that with growth
and remodeling, the CE angle returns to normal in most
cases (see discussion).

Outcome and Waldenstrom Stage

No correlation was found between Waldenstrom
stage at the time of surgery and the nal Stulberg
classication.

Need for Further Surgery

Four hips required further corrective surgery at an
average of 18 12 months after the triple pelvic
osteotomy. Three were lateral pillar C and 1 was lateral
pillar B at diagnosis. Two of these patients had a fair
result (Stulberg III) and 2 had a poor result (Stulberg
IV/V). The average age at diagnosis of the 4 patients
requiring further surgery was 9.5 1 years compared

TABLE 2. Stulberg Classification at Most Recent Follow-up

Based on Preoperative Catterall Grouping
Stulberg Classication
Group 2
Group 3
Group 4

I/II

III

IV/V

50%
53%
18%

37.5%
41%
64%

12.5%
6%
18%

2010 Lippincott Williams & Wilkins

FIGURE 4. Bar graph illustrating the change in CE angle

before surgery and at final follow-up.
www.pedorthopaedics.com |

753

J Pediatr Orthop

Wenger et al

Volume 30, Number 8, December 2010

FIGURE 5. Bar graph illustrating the change in femoral head

extrusion index before surgery and at final follow-up.

with 7.2 2 years in the 31 hips that did not require

further corrective surgery (P = 0.02). Procedures and
surgical indications are listed in Table 4.

DISCUSSION
After the initial description of this disease in 1910 by
Legg, Calve, and Perthes, orthopedic treatment included
prolonged nonweight bearing and then abduction casting
methods. As the abduction casting methods seemed to
improve results, extensive eorts to provide abduction by
a variety of removable abduction braces were made over
the next 50 years. Several studies have claried that brace
treatment is not very eective, either owing to brace
design or poor patient compliance.7,9 Lack of compliance
with brace treatment has become a progressively larger
problem in contemporary western culture in which issues
such as both parents working and childhood self esteem
(re: appearance in a brace) have made orthotic treatment
dicult.
Surgical containment using proximal femoral varus
osteotomy was rst introduced in 1952 by Soeure27
followed by Axer28 and Lloyd-Roberts11 and more
recently conrmed as eective by Wiig et al.9 Salter
innominate osteotomy was introduced as a method for
surgical containment of LCP disease in 1962.29 Either
of these methods may provide adequate containment in
mild to moderate LCP disease but can be problematic in
more severe cases. The frequent disadvantages of these
standard techniques led us to seek other containment
options.17

FIGURE 6. A, AP pelvis and frog view of the right hip in an

11+4 year old boy with right Perthes disease. The femoral
head shows Gage sign (erosion of the femoral head at the
margin of the femoral head and physis). Subsequent films
confirmed Herring C classification. B, AP, and frog view have
taken 1 year after surgical containment with triple pelvic
osteotomy. Note the marked involvement of the lateral pillar
(femoral head changes which often seem late in older children
with Perthes disease). C, AP and frog view of pelvis carried out
4 years after right hip pelvic osteotomy. Femoral head
sphericity has been maintained.

Proximal femoral osteotomy has the disadvantage

of shortening an already short limb as well as producing
a limp whereas Salter innominate osteotomy may not
achieve adequate acetabular rotation to cover the femoral
head in patients with extensive necrosis.16 This can lead to
iatrogenic hinge abduction and a bid head shape.

TABLE 4. Patients (4) Requiring Further Surgery After Triple Pelvic Osteotomy
1
2
3
4

Surgical Indication

Surgical Procedure

Late impingement
Hinged abduction with painful limp
Hinged abduction
Inadequate coverage

Valgus femoral osteotomy, cheilectomy, shelf procedure

Cheilectomy, bone grafting to femoral head, distal femoral stapling
Flexion/valgus proximal femoral osteotomy, medial distal femoral stapling
Varus derotational osteotomy, shelf procedure

754 | www.pedorthopaedics.com

2010 Lippincott Williams & Wilkins

J Pediatr Orthop

Volume 30, Number 8, December 2010

Results of Triple Pelvic Osteotomy for Perthes

FIGURE 7. A, AP pelvis and frog view of the right hip in a 10-year-old male patient with severe Herring C type Perthes disease of
the right hip. The lateral taper of the femoral head and extensive lucency below the physis are poor prognostic signs. B, AP pelvis
and frog view of the right hip taken 9 months after triple pelvic osteotomy. Despite the attempt to achieve containment, the
femoral head has collapsed and extruded with the lateral impingement. The child has markedly restricted hip abduction. C, AP
view of the right hip after valgus osteotomy performed to correct severe lateral impingementFcarried out at age 11.3 years.
D, The patient continued to have impingement symptoms and was treated with femoral head-neck osteoplasty and greater
trochanter advancement through a transtrochanteric hip dislocation approach at age 17 years (AP and frog view of hip)Fcarried
out 1 year after this revision surgery. He is now asymptomatic, but his hip motion is primarily flexion and extension and the
prognosis is guarded.

When comparing femoral and innominate osteotomies, Sponseller et al14 did not nd any functional
dierences in the outcome. However, they concluded that
innominate osteotomy was a better choice when likely
closure of the growth plate (owing to severe disease) with
subsequent leg length discrepancy was anticipated. They
also found that patients over the age of 10 years at onset
of disease had a poor outcome regardless of the surgical
procedure.
Herring et al8 reported results using the more
standard osteotomies for containment in LCP disease
focused on the use of either Salter innominate osteotomy
or proximal femoral varus osteotomy. In the Texas
Scottish Rite Hospital study, 68 patients were treated
with Salter innominate osteotomy with these results:
Stulberg I or II (57%), Stulberg III (32%), and Stulberg
IV/V (10%). The femoral osteotomy treatment group
included 52 patients with these results: Stulberg I or II
(65%), Stulberg III (25%), and Stulberg IV/V (10%).
They found no dierence between the patients treated
with proximal femoral varus osteotomy or innominate
osteotomy. They concluded that patients with Lateral
pillar B or B/C disease over the age of 8 at the time of
onset have a better outcome with surgical containment.
They also noted that patients with lateral pillar C disease
at any age had poor outcomes, even when treated by
femoral or Salter osteotomy8 and suggested that treatment could not help these patients.
The Dallas study (Herring et al8) did not include
advanced containment methods. We do not withhold
treatment in these older children because we have found
that advanced containment methods (triple, triple plus
r

2010 Lippincott Williams & Wilkins

varus femoral osteotomy) can lead to satisfactory results

in some lateral pillar C patients. We also have treatment
methods to deal with type C patients who have subsequent head collapse and extrusion despite advanced
containment methods (see below).
Other authors have also described advanced containment options for more severe cases. Crutcher and
Staheli combined Salter and varus femoral osteotomies in
14 patients with severe disease. They reported 2 good
results and 12 fair results with 3 patients requiring further
operative intervention and concluded that the procedure
was safe and eective for severe cases.18 Chang et al30
recently reported the outcomes for treating advanced
LCP disease with the Staheli shelf acetabuloplasty. They
noted a marked improvement of abduction motion at
nal follow-up and found that 33% of patients achieve
spherical congruence of the hip with an ovoid femoral
head and a further 38% achieved congruence with a
mushroom-shaped femoral head. They concluded that
shelf acetabuloplasty restricts lateral displacement or
subluxation in LCP disease, resulting in good containment and acetabular coverage.
Huang and Huang compared triple pelvic osteotomy (14 pts) with shelf augmentation (14 pts) for severe
disease. Although they noted an increased number of
patients with spherical incongruity in the shelf group of
patients, they concluded that the shelf augmentation
procedure was a simpler procedure with better coverage
of the femoral head.31
Poul and Vejrostova32 evaluated the use of the Steel
triple osteotomy in hip dysplasia and LCP disease. In 12
patients with LCP disease ages 9 to 12 years treated with
www.pedorthopaedics.com |

755

Wenger et al

triple pelvic innominate osteotomy, they noted successful

containment with subsequent remodeling, even in deformed femoral heads, in 11 of the 12 cases.
Gotia et al33 compared 18 femoral osteotomies, 10
Chiari pelvic osteotomies, and 8 triple pelvic osteotomies and recommended that the triple pelvic osteotomy
be carried out more frequently in children over age 7
years with lateral extrusion. Kumar et al20 reported
on interlocking triple pelvic osteotomy in severe disease
in 21 patients and noted an average gain in acetabular
head index of 18% and gain in center edge angle of
22%. Vukasinovic et al34 reported results using triple
innominate osteotomy in 30 patients with marked
improvement in the CE angle and containment at
follow-up.
After extensive experience and frequent frustration
using traditional surgical containment methods for older
children, we became interested in a technique that
avoided femoral osteotomy, yet provided full containment even in severe cases. Having used triple pelvic
osteotomy to treat hip dysplasia, we elected to adapt the
method for containment in LCP disease. Our surgical
variation of the technique focuses on making the ischial
cut close to the acetabulum to allow more free acetabular
segment rotation. As in any advanced orthopedic operation, proper training and experience are required to
master the technique of triple pelvic osteotomy. With
appropriate patient selection and operative technique, we
have found the procedure to be an excellent treatment
method.
As triple osteotomy makes the acetabular segment
much more mobile (as compared with the Salter
procedure) care must be taken to avoid external rotation
of the acetabular fragment, thus creating acetabular
retroversion that can result in both a gait abnormality
(foot turns outward) and undesirable hip mechanics
(increased risk for arthritis). In addition, over-coverage
must be avoided to decrease the risk for late iatrogenic
impingement.
Our results show that the triple pelvic osteotomy
provides excellent containment with center edge angle
increasing by an average of 17.2 degrees (P<0.001) and
the extrusion index decreasing from 21% to 8%
(P<0.001). These are early results in some patients
(minimum 3 year follow-up) and long term follow-up will
be required to fully understand nal femoral headacetabular remodeling. We achieved predictable maintenance of head shape in Herring B patients of all ages
and Herring C patients with onset less than 8. There were
no poor results in the Herring B classication group with
80% of patients over the age of 8 having a good outcome.
In patients less than 8 with Herring C hips, 83% achieved
either a good or fair result. Patients in the above categories have remarkably good results with round femoral
heads, normal limb lengths, and rapid early return of hip
motion and function.
Older children with Herring C involvement remain
problematic (especially if the child is older than age 10
years at diagnosis). Triple pelvic osteotomy can be

756 | www.pedorthopaedics.com

J Pediatr Orthop

Volume 30, Number 8, December 2010

eective in some of these children (Fig. 6) but the result

is less predictable.
The 4 patients in our study that required additional
surgery (usually valgus osteotomy) had a mean age of 9.5
years of age at onset of symptoms. Two ended with a fair
result and 2 with a poor result owing to continued head
attening, extrusion, and hinge abduction. Containment
methods can still be used for older, type C involvement
cases, however head collapse and extrusion may still
occur, requiring further surgery (Fig. 7). Our traditional
approach to such patients has been a valgus femoral
osteotomy (sometimes combined with shelf acetabuloplasty) to correct hinge abduction. Recently we have
begun to use femoral head/neck recontouring to address
the problem.
Some surgeons might choose to avoid initial
containment in type C hips in older children and instead
select noncontainment methods. These include shelf
acetabuloplasty,21 primary valgus osteotomy,35 and hip
distraction.36
As noted in the results section, triple osteotomy can
lead to hypercontainment in some cases with a risk for
subsequent pincer impingement. Thus, advanced containment has a risk-benet ratio. In our overall current series
(now 63 patients), 2 patients have required later surgical
treatment to correct protrusio (pincer impingement).
Despite this risk, we still advise advanced containment
methods because they contribute greatly to the maintenance of a spherical femoral head which Stulberg has
shown to be so important for outcome. The few patients
who are over-contained and do not remodel can be dealt
with using newer surgical methods (acetabular rim
trim) to correct impingement.

Summary
Surgical containment by triple pelvic osteotomy
provides an eective method for the treatment of LCP
disease in patients that require prolonged containment.
Despite the more guarded prognosis in older children
(over age 8 y with Herring C involvement), we still advise
this method in most cases in early Waldenstrom stages.
We use this approach because it can produce a surprisingly spherical head, even in older children (Fig. 6). Such
cases conrm the value of femoral head containment
during the revascularization phase. In the cases in which
containment fails, one can consider valgus osteotomy,
joint distraction, or head-neck recontouring.
REFERENCES
1. Salter RB, Thompson GH. Legg-Calve-Perthes disease. The
prognostic signicance of the subchondral fracture and a two-group
classication of the femoral head involvement. J Bone Joint Surg
Am. 1984;66:479489.
2. Catterall A. The natural history of Perthes disease. J Bone Joint
Surg Br. 1971;53:3753.
3. Catterall A. Legg-Calve-Perthes syndrome. Clin Orthop Relat Res.
1981;158:4152.
4. Salter RB. The present status of surgical treatment for Legg-Perthes
disease. J Bone Joint Surg Am. 1984;66:961966.
r

2010 Lippincott Williams & Wilkins

J Pediatr Orthop

Volume 30, Number 8, December 2010

5. Stulberg SD, Cooperman DR, Wallensten R. The natural history

of Legg-Calve-Perthes disease. J Bone Joint Surg Am. 1981;63:
10951108.
6. Weinstein SL. Legg-Calve-Perthes disease: results of long-term
follow-up. In The Hip: Proceedings of the Thirteenth Open Scientic
Meeting of the Hip Society. St. Louis: C.V. Mosby; 1985:2837.
7. Martinez AG, Weinstein SL, Dietz FR. The weight-bearing
abduction brace for the treatment of Legg-Perthes disease. J Bone
Joint Surg Am. 1992;74:1221.
8. Herring JA, Kim HT, Browne R. Legg-Calve-Perthes disease. Part
II: prospective multicenter study of the eect of treatment on
outcome. J Bone Joint Surg Am. 2004;86-A:21212134.
9. Wiig O, Terjesen T, Svenningsen S. Prognostic factors and outcome
of treatment in Perthes disease: a prospective study of 368 patients
with ve-year follow-up. J Bone Joint Surg Br. 2008;90:13641371.
10. Hoikka V, Poussa M, Yrjonen T, et al. Intertrochanteric varus
osteotomy for Perthes disease. Radiographic changes after 2-16year follow-up of 126 hips. Acta Orthop Scand. 1991;62:549553.
11. Lloyd-Roberts GC, Catterall A, Salamon PB. A controlled study of
the indications for and the results of femoral osteotomy in Perthes
disease. J Bone Joint Surg Br. 1976;58:3136.
12. Noonan KJ, Price CT, Kupiszewski SJ, et al. Results of femoral
varus osteotomy in children older than 9 years of age with Perthes
disease. J Pediatr Orthop. 2001;21:198204.
13. Paterson DC, Leitch JM, Foster BK. Results of innominate
osteotomy in the treatment of Legg-Calve-Perthes disease. Clin
Orthop Relat Res. 1991;266:96103.
14. Sponseller PD, Desai SS, Millis MB. Comparison of femoral and
innominate osteotomies for the treatment of Legg-Calve-Perthes
disease. J Bone Joint Surg Am. 1988;70:11311139.
15. Thompson GH, Salter RB. Legg-Calve-Perthes disease. Current
concepts and controversies. Orthop Clin North Am. 1987;18:617635.
16. Rab GT. Theoretical study of subluxation in early Legg-CalvePerthes disease. J Pediatr Orthop. 2005;25:728733.
17. Wenger DR. Selective surgical containment for Legg-Perthes
disease: recognition and management of complications. J Pediatr
Orthop. 1981;1:153160.
18. Crutcher JP, Staheli LT. Combined osteotomy as a salvage
procedure for severe Legg-Calve-Perthes disease. J Pediatr Orthop.
1992;12:151156.
19. Vukasinovic Z, Slavkovic S, Milickovic S, et al. Combined salter
innominate osteotomy with femoral shortening versus other
methods of treatment for Legg-Calve-Perthes disease. J Pediatr
Orthop B. 2000;9:2833.
20. Kumar D, Bache CE, OHara JN. Interlocking triple pelvic
osteotomy in severe Legg-Calve-Perthes disease. J Pediatr Orthop.
2002;22:464470.

2010 Lippincott Williams & Wilkins

Results of Triple Pelvic Osteotomy for Perthes

21. Kruse RW, Guille JT, Bowen JR. Shelf arthroplasty in patients who
have Legg-Calve-Perthes disease. A study of long-term results.
J Bone Joint Surg Am. 1991;73:13381347.
22. Waldenstrom H. The denitive forms of coxa plana. Acta Radiol.
1922;1:384.
23. Herring JA, Kim HT, Browne R. Legg-Calve-Perthes disease. Part I:
classication of radiographs with use of the modied lateral pillar
and Stulberg classications. J Bone Joint Surg Am. 2004;86-A:
21032120.
24. Wiberg G. Studies on dysplastic acetabula and congenital subluxation of the hip joint. With special reference to the complication of Osteo-Arthritis. Acta Chir Scand. 1939;83(suppl 58):
1134.
25. Green NE, Beauchamp RD, Grin PP. Epiphyseal extrusion as a
prognostic index in Legg-Calve-Perthes disease. J Bone Joint Surg
Am. 1981;63:900905.
26. Carlioz H, Khouri N, Hulin P. Osteotomie triple juxta-cotyloidienne. Rev Chir Orthop. 1982;68:497501.
27. Soeur R, De Racker C. The anatomopathologic aspect of osteochondritis and the pathogenic theories which are relevant. I. Acta Orthop
Belg. 1952;18:57102.
28. Axer A. Subtrochanteric osteotomy in the treatment of Perthes
disease: a preliminary report. J Bone Joint Surg Br. 1965;47:
489499.
29. Salter R. Legg-Perthes disease. Treatment by innominate osteotomy. AAOS Instruc Course Lect. 1973;22:309316.
30. Chang J, Kuo K, Huang S. Outcomes in advanced LeggCalve-Perthes disease treated with the Staheli procedure. J Surg
Res. 2009. doi:10.1016/j.jss.2009.09.056.
31. Huang MJ, Huang SC. Surgical treatment of severe perthes disease:
comparison of triple osteotomy and shelf augmentation. J Formos
Med Assoc. 1999;98:183189.
32. Poul J, Vejrostova M. Triple osteotomy of the pelvis in children and
adolescents. Acta Chir Orthop Traumatol Cech. 2001;68:9398.
33. Gotia DG, Ardelean M, Galan L, et al. Surgical treatment in
osteochondritis of the femur head in children. Rev Med Chir Soc
Med Nat Iasi. 1993;97:415418.
34. Vukasinovic Z, Spasovski D, Vucetic C, et al. Triple pelvic
osteotomy in the treatment of Legg-Calve-Perthes disease. Int
Orthop. 2009;33:13771383.
35. Bankes MJ, Catterall A, Hashemi-Nejad A. Valgus extension
osteotomy for hinge abduction in Perthes disease. Results at
maturity and factors inuencing the radiological outcome. J Bone
Joint Surg Br. 2000;82:548554.
36. Segev E, Ezra E, Wientroub S, et al. Treatment of severe late onset
Perthes disease with soft tissue release and articulated hip
distraction: early results. J Pediatr Orthop B. 2004;13:158165.

www.pedorthopaedics.com |

757