HIP DISORDERS SUPPLEMENT

Recurrent Dislocations and Complete Necrosis:

The Role of Pelvic Support Osteotomy
In Ho Choi, MD, Tae-Joon Cho, MD, Won Joon Yoo, MD, and Chang Ho Shin, MD

in posture, gait, and walking tolerance to those adoles-

Abstract: The basic concepts and goals of pelvic support os- cents and young adults13 who have unstable hips showing
teotomy (PSO) are to enhance femoro-pelvic stability by prox- severe dysplastic acetabulum and partial or total absence
imal femoral valgus osteotomy and to improve hip of the femoral head and neck (Table 1).
biomechanics by displacing the center of gravity medially, which The purpose of this section of the symposium is to
results in an improvement in the mechanical efficiencies of ab- address the roles of PSO and Ilizarov hip reconstruction
ductor muscles. However, the clinical application of traditional (IHR), a modified form of PSO, which combines distal
PSO is limited due to its intrinsic shortcomings. Ilizarov de- femoral (DF) osteotomy for concomitant lengthening and
signed a modified PSO, so called “Ilizarov hip reconstruction varus angulation, for the management of recurrent dis-
(IHR),” which incorporated a second distal femoral osteotomy, locations and complete necrosis of the femoral head.14–19
to realign the knee joint and to correct limb-length discrepancy We will briefly review the histories of PSO and IHR and
and proximal femoral valgus osteotomy for pelvic support. IHR elaborate on the technical considerations and outcomes of
remains as a viable option for the salvage of severely damaged IHR, based upon an extensive review of the literature and
hips in the adolescents and young adults in whom arthrodesis or personal experience.
hip arthroplasty is not suitable. The purpose of this article is to
briefly review the evolution of PSO and to concentrate on the
technical considerations and outcomes of IHR. BRIEF HISTORY AND RATIONALE OF PSO
Key Words: pelvic support osteotomy, Ilizarov hip re- AND IHR
construction, hip instability PSO has long history in orthopaedic surgery. The
technique was developed and popularized by Lorenz,2
(J Pediatr Orthop 2013;33:S45–S55) Von Bayer,3 Schanz,4 Milch,5–8 and Henderson10 but was
rapidly replaced by total hip replacement arthroplasty
(THRA). Henry Milch5–8 expanded the concept and

T he treatment of hip instability in adolescents and

young adults is often related to severe dysplastic
acetabulum, proximal migration of the femur, and/or the
popularized the PSO in the United States during the mid-
20th century. The basic concepts and goals of PSO are to
enhance femoro-pelvic stability by PF valgus osteotomy
absence of part or all of the femoral head and neck, and and to improve hip biomechanics by displacing the center
presents a surgical challenge. Furthermore, reconstructive of gravity medially, which results in an improvement in
procedures for this difficult problem, such as, greater the mechanical efficiencies of abductor muscles.5–7,20 The
trochanteric arthroplasty, hip arthrodesis, pelvic osteot- change of joint biomechanics seems to be more important
omy, femoral osteotomy, and Girdlestone operation, are that the anatomic support to reduce painful instability
less than satisfactory.1 Different methods of proximal and to improve hip function.21 Overcorrection of PF
femoral (PF) subtrochanteric valgus angulation osteot- valgus osteotomy places the extremity in a fixed abduc-
omy, known as pelvic support osteotomy (PSO),2–11 have tion position relative to the pelvis to eliminate hip ad-
been described to salvage the damaged hips in whom duction and reduces or prevents the Trendelenburg sign
arthrodesis or hip arthroplasty are not appropriate. because the contralateral pelvis cannot drop.12,20 How-
Support of the pelvis is achieved by means of a valgus ever, the clinical application of traditional PSO is limited
osteotomy that places the superior end of the femur due to its intrinsic shortcomings. In particular, the opti-
against the lateral aspect of the pelvis.12 PSO is a useful mal extent of angulation is difficult to achieve. If the angle
surgical procedure that offers a significant improvement is too large, excessive genu valgum, fixed pelvic obliquity,
and impingement pain on adduction of the lower ex-
From the Division of Pediatric Orthopedics, Seoul National University tremity to the neutral position may ensue. Alternatively, if
Children’s Hospital, Seoul, Korea. the angle is too small, the result would be an insufficient
None of the authors received financial support for this study. improvement in hip biomechanics,6,9,20,22,23 and most
The authors declare no conflict of interest. importantly, the issue of remaining limb-length discrep-
Reprints: In Ho Choi, MD, Division of Pediatric Orthopedics, Seoul
National University Children’s Hospital, Seoul 110-744, Korea. ancy (LLD) cannot be addressed.19
E-mail: [email protected]. To overcome the shortcomings of traditional PSO,
Copyright r 2013 by Lippincott Williams & Wilkins Ilizarov designed a modified PSO, which incorporated a

J Pediatr Orthop  Volume 33, Number 1 Supplement, July/August 2013 www.pedorthopaedics.com | S45
Choi et al J Pediatr Orthop  Volume 33, Number 1 Supplement, July/August 2013

TABLE 1. Indications of Pelvic Support Osteotomy

TECHNICAL CONSIDERATIONS OF IHR
Careful preoperative surgical planning, based on
Hip instability: severe dysplastic acetabulum data obtained from clinical and radiographic assessments,
DDH: neglected, unsuccessfully treated
Traumatic hip dislocation with instability is essential to achieve a level pelvis and to restore the
Paralytic or spastic dislocation (postpoliomyelitis, cerebral palsy, mechanical axis of the lower limbs perpendicular to the
muscular dystrophy) horizontal line of the pelvis in bipedal stance.1,12,13 Im-
Partial or total absence of femoral head and neck portant technical considerations of IHR are summarized
Severe sequelae of septic arthritis (Choi type IV)
Skeletal dysplasia (SED, Morquio, etc.)
as follows.1,12–14,19
Severe AVN 1. Although some authors have recommended proximal
Post-Girdlestone resection arthroplasty osteotomy, in which the acetabulum rests on the lesser
AVN indicates avascular necrosis; SED, spondylo-epiphyseal dysplasia.
trochanter,3,11 most prefer more distal osteotomy in
the anticipation of abductor mechanical benefits due
to displacement of the center of gravity medially. To
second DF osteotomy, to realign the knee joint and to determine the optimal level of PF valgus osteotomy,
correct LLD, and PF valgus osteotomy for pelvic sup- the femoral shaft should be fully adducted against the
port. Russian literature indicates that Ilizarov and his lateral wall of the pelvis, which is usually situated
associates started to use a modified PSO technique in the somewhere between the infracotyloid recess and the
early or mid-1970s.15–17,22,23 Ilizarov emphasized the im- ischial tuberosity. This renders longer proximal seg-
portance of PF extension to correct the fixed flexion de- ment and provides better hip stabilization and an
formity of the hip and to permit locking of the hip joint optimal location for soft tissue interposition to
by stabilizing the hip in the sagittal plane during single produce a weight-bearing surface without direct abut-
stance.12,14,18,19 ment between the PF osteotomy and the pelvis (Fig. 1).
IHR is considered a breakthrough in terms of re- 2. How much valgus angulation is desirable? This issue is
solving the inherent problems of PSO, as the treatment highly controversial. In the era of conventional PSO,
goals for normal gait are to obtain stability by re- Henry Milch6,8 emphasized that the postosteotomy
constructing a stable fulcrum, to improve energy effi- angle (“b-angle”: the obtuse angle formed between a
ciency by restoring abductor mechanism, and to improve line drawn along the inner aspect of the cortex of the
cosmetic appearance by eliminating shortening/joint distal fragment and an oblique line drawn from the
contracture-related problems. upper end of the distal fragment tangent to the most
medial projection of the proximal fragment, which
may be considered as the neck angle of the osteotom-
INDICATIONS OF IHR ized femur) must not be permitted to far exceed the
IHR is most suitable for skeletally mature adoles- angle of inclination of outer pelvic wall in the level
cents or young adults that present with an unstable hip pelvis (“a–angle”: the obtuse angle formed between an
that is mobile and associated with a Trendelenburg gait oblique line connecting the 2 points of the outer edge
(T-gait) and a large LLD. IHR is highly effective at of the acetabular roof and the outer edge of the ischial
eliminating T-gait, particularly when there is good ab- tuberosity and the mechanical axis line, passing
ductor muscle function before surgery. The 2 most fre- through the outer edge of the acetabular roof,
quent indications are a neglected or an unsuccessfully perpendicular to the horizontal pelvic reference line).
treated developmental dislocation of the hip24–26 and se- He reported that the mean a-angle was 206.4 ± 4.7
vere septic hip sequelae.1,12,26–28 IHR is also indicated for degrees in females and 211.4 ± 5.8 degrees in males.
the treatment of hip instability related to paralytic sub- He insisted that postosteotomy angle should be
luxation/dislocation, posttraumatic hip subluxation/dis- between 210 and 230 degrees, because excessive valgus
location, spondylo-epiphyseal dysplasia,29 osteonecrosis at the osteotomy site leads to PF abutment against the
of the femoral head,30 and postexcision arthroplasty31 pelvis, and even pelvic tilt, when the patient tries to
(Table 2). bring the involved extremity into a neutral adduction/
It seems that IHR is not ideal for young children, abduction position.6,8,13 However, with contemporary
because in accordance with Wolff’s law, gradual IHR, irrespective of the size of the overcorrection
straightening of the PF tends to occur at the site of valgus of PF valgus angulation, much of the abduction
angulation and this results in loss of pelvic support. Al- deformity caused by PF valgus angulation can be
though IHR is not contraindicated in young children, for compensated, if the second DF osteotomy restores the
example, when hip instability is associated with marked position of knee joint inclination to parallel the
LLD due to multiple lower-limb growth disturbances horizontal line of the pelvis. The mean PF valgus
secondary to neonatal sepsis, one should expect repeat angulation reported in the literature varied widely
IHR at or near skeletal maturity. Another alternative is to between 35 and 60 degrees.12,20,26,29,32–37 Paley12,19
undergo femoral lengthening without a PSO, inserting (Fig. 1A) recommended overcorrection of 15 to 20
half pins into the pelvis to prevent proximal migration of degrees during PF valgus osteotomy to eliminate hip
the femur, in the younger age, and subsequently perform adduction in addition to pelvic drop angle (the angle
IHR when the patients are near skeletal maturity.12 between the line perpendicular to iliac crest pelvic line

S46 | www.pedorthopaedics.com r 2013 Lippincott Williams & Wilkins

J Pediatr Orthop  Volume 33, Number 1 Supplement, July/August 2013 The Role of Pelvic Support Osteotomy

TABLE 2. Summary of the English Literature on Ilizarov Hip Reconstruction

Type of Hip Function* Postop
Support Trendelenburg
Etiology Age at Op. (No. Sign (No. Complications and
References (No. Patient) (y) Patient) Preop. Postop. Patient) Follow-up Comments
Samchukov DDH (1) 15 Subacetab. NA NA None 5 mo NA
and Birch20
Kocaoglu DDH (11); 20 (12-33) Subacetab. 64 (42-72) 84 (68-92) None (17); P (3) 68 mo (55-81) 3 PTI 3; 1
et al32 PFFD (1); hypocorrection
MMC (1);
paralytic
dislocat. (1)
Manzotti Septic hip (15) 21.1 (14-36) Acetabular NA 1.94 (W) None (9); P (6) 108 mo (38-178) 3 knee sublux.; 2 loss
et al28 (9); of support; 2 pin
subace- substitute; 1
tab. (5); peroneal n. palsy;
pubic 1 regenerate fx.;
ramus (1) 1 foot ER
Inan et al25 DDH (17) 24.8 (17-39) Subacetab. NA NA None (12); P (5) 36.3 mo (21-65) Many PTI; 2 fx.
Rozbruch Septic hip (8) 11.2 (7.8-14.2) Subacetab. 51 (21-67) 73 (64-79) None (6); P (2) 5 y (1.9-9.8) 3 PTI; 2 knee
et al12 stiffness; 1 knee
sublux.; 1
premature
consolidation; 1
prox. migration of
femur
Inan and DDH (12); 25.3 (17-39) Subacetab. 50 (32-73) 87.6 (67-98) None (12); P (4) 52.5 mo (26-84) 2 PTI; 2 delayed
Bowen33 septic hip (3); consolidation; 1
traumatic fracture; 1 knee
lux. (1) stiffness; 1
obturator n.
entrapment
Inan et al34 DDH (11) 25.2 (13-39) Subacetab. 52 (32-73) 92 (77-98) None (6); P (5) 36 mo (23-59) NA; many PTI
El-Mowafi24 DDH (12); 22.4 (19-35) Subacetab. 55 (40-78) 81 (65-90) None (20); P (5) 4.5 y (2-7) 4 of 5 patients with
septic hip (5); persistent T-sign
AVN (5); was RP
RP (4)
Emara31 Excision 51.9 (45-61) Subacetab. 43.5 (31-50) 70.9 (65-80) None (11) Minimum 2 y 7 PTI; 2 knee
arthroplasty stiffness; 2 residual
after infected knee valgus; 3
hip delayed
arthroplasty consolidation
(11)
Shetty et al29 Spondylo- 16.4 (9-25) Subacetab. 67.9 (54-85) 79.1 (68-87) None (6 hips); P 25.9 mo 3 PTI; 3 knee
epiphyseal (10 hips) stiffness; 1 delayed
dysplasia consolidation
congenita (8
bilat.)
Gursu et al35 DDH (12 unilat., 22.6 (12-34) Subacetab. 48.2 (28-79) 80.1 (60-93) None (13 hips); P 33.5 mo (16-45) 15 PTI; 2 prox. fx.; 2
1 bilat.); septic (8 hips) delayed union; 6
hip (8) knee stiffness; 1
depression
Krieg et al36 Spina bifida (1) 14 Subacetab. NA NA Improved 12 mo None; prox.: locking
compression plate,
distal: Fitbone
Marimuthu DDH (1); 23 (13-32) Subacetab. 44.3 (14-73) 70.8 (60-86) None (9); P (3) 59.4 mo (38-86) PTI almost all
et al26 nonunion
(NOF) (2);
septic hip (6);
tuberculois (3)
Mahran DDH (9); septic 21.5 (14-30) Subacetab. NA NA None (13); P (7) 6 mo PTI almost all; 2
et al27 hip (9); RP (2) regenerate fx.; 2
premature
consolidation; 4
delayed
consolidation; 3
residual
LLD > 3 cm

r 2013 Lippincott Williams & Wilkins www.pedorthopaedics.com | S47

Choi et al J Pediatr Orthop  Volume 33, Number 1 Supplement, July/August 2013

TABLE 2. (continued)
Type of Hip Function* Postop
Support Trendelenburg
Etiology Age at Op. (No. Sign (No. Complications and
References (No. Patient) (y) Patient) Preop. Postop. Patient) Follow-up Comments
Sabharwal Osteonecrosis 13.5 Subacetab. 26 88 None 3y Knee
and after leukemia stiffnessZquadri-
Macleod 30 (1) cepsplasty
Choi et al Septic hip (11); 13.4 (6.4-16) Subacetab. 12 (W) 6 (W) None (8); P (5) 5 y (1-10.2) Repeat IHR in 3
(unpublish- DDH (2) patients; 4 PTI; 4
ed data) loss of support; 1
prox. fx.
*Harris Hip Score or a modified Harris Hip Score.12
AVN indicates avascular necrosis of the femoral head; bilat., bilateral; DDH, developmental dislocation of the hip; dislocat., dislocation; ER, external rotation; fx.,
fracture; IHR, Ilizarov hip reconstruction; LLD, limb-length discrepancy; lux., luxation; MMC, meningomyelocele; n., nerve; NA, unknown; NOF, nonunion of femur
neck fracture; op., operation; P, reduced but persistent T-sign; PFFD, proximal femoral focal deficiency; postop., postoperative; preop.; preoperative; prox., proximal; PTI,
pin-tract infection; RP, residual poliomyelitis; subacetab., subacetabular; T, Trendelenburg; unilat., unilateral; W, WOMAC score.

and the femoral shaft in single stance or during from the midline of the body axis in bipedal stance.
maximum attempted adduction when supine). On Furthermore, if there is no compensatory DF varus
the basis of our experiences, we recommend at least angulation, unequal knee distances from midline may
25 degrees of overcorrection,1 and >30 degrees in cause secondary pelvic obliquity despite well-per-
preadolescents. Pafilas and Nayagam13 (Fig. 1B) pro- formed pelvic support.
posed overcorrection of 30 to 40 degrees of extravalgus 4. The amount of extension should be adjusted to correct
in addition to the sum of maximum adduction range hip flexion contracture and pelvic tilt, and the
plus adduction contracture plus another 9 degrees to sacrofemoral angle (the angle formed by the inter-
bring the femur parallel to the vertical axis perpendic- section of a line drawn along the top of the sacrum to
ular to the pelvic line. The remaining 21 to 31 degrees locate the pelvic inclination and a second line drawn
of abduction will take the femur away from midline. along the shaft of the femur on the standing lateral
Their 30 to 40 degrees corresponds to 21 to 31 degrees radiograph). In normal children and adolescents, the
of extravalgus in addition to the pelvic drop angle. sacrofemoral angle measures between 50 and 65
Overcorrection is entirely empirical in anticipation of degrees.38
remodeling at the valgus osteotomy and some atrophy 5. The amount of derotation should also be determined,
of the soft tissue interposed between the femur and based on foot progression angle and the amount and
lateral pelvic wall. direction of rotation during passive maximum adduc-
3. The next important issue is how to determine the level tion. The amount of varization is also controversial.
of DF osteotomy. Paley12,19 used the CORA method, Although Pafilas and Nayagam13 proposed bringing
which utilizes an imaginary proximal mechanical axis the femoral shaft parallel to the vertical midline axis,
line (Fig. 2A). He stated that proximal mechanical axis this may cause 9 degree of valgus inclination at the
line corresponds to a line perpendicular to the knee. We believe that the knee joint should be
horizontal pelvic line, passing through the point of realigned to its physiological position even after DF
1/3 to 1/2 the distance lateral to the medial edge of varus angulation.
the proximal fragment. In contrast, Kadykalo and 6. The amount of lengthening should be recalculated
Kuftyev22 presented a formula that took into consid- after IHR during lengthening using a woodblock or
eration the amount of PF valgus and DF varus scanogram to obtain a level pelvis. Overlengthening is
angulation (Fig. 2B; Table 3). The mean DF varus poorly tolerated in hips that are already in full
angulations reported in the literature varied in the adduction after IHR.
range between 10 and 22 degrees.20,29,30,32 Other 7. Resection of the femoral head and neck remnant may be
important point of consideration when determining indicated, if the hip is painful and stiff, regardless of pre-
the level of distal osteotomy concerns the equalization existing osteoarthritis. The concept of resection-angula-
of distances between the midline of the body axis and tion osteotomy was originally proposed by Milch7 in
the centers of the knee joints of affected and normal 1955 for restoring hip mobility and pain relief.
contralateral limbs, which was emphasized by Pafilas
and Nayagam13 (Fig. 2C). This suggests that the level
of DF osteotomy relies on the level of PF valgus SUMMARY OF THE ENGLISH LITERATURE
osteotomy, that is, the higher the level of PF valgus ON IHR
angulation, the more proximally located DF osteot- At the time of writing, only 12 original articles and 3
omy should be to equalize the distances of knee joints case reports have been published in the English literature

S48 | www.pedorthopaedics.com r 2013 Lippincott Williams & Wilkins

J Pediatr Orthop  Volume 33, Number 1 Supplement, July/August 2013 The Role of Pelvic Support Osteotomy

FIGURE 1. Determination of valgus angulation of the proximal femur for Ilizarov hip reconstruction. A, Paley12,19 recommended
overcorrection of 15 to 20 degrees during valgus osteotomy to eliminate hip adduction and pelvic drop angle, which is an angle
between the line perpendicular to iliac crest pelvic line and the femoral shaft in single stance or during maximum attempted
adduction in the supine position. On the basis of our experiences, we recommend at least 25 degrees of overcorrection1 and >30
degrees in preadolescents. B, Pafilas and Nayagam13 proposed overcorrection of 30 to 40 degrees of extravalgus in addition to
the sum of maximum adduction range plus adduction contracture plus another 9 degrees to bring the femur parallel to the
vertical axis perpendicular to the pelvic line. The remaining 21 to 31 degrees of abduction will take the femur away from midline.
Their 30 to 40 degrees corresponds to 21 to 31 degrees (b a) of extravalgus in addition to the pelvic drop angle. Overcorrection
is entirely empirical in anticipation of remodeling at the valgus osteotomy and some atrophy of the soft tissue interposed between
the femur and lateral pelvic wall.

on the merits of IHR (Table 2). Samchukov and Birch20 with a mean LLD of 4.4 cm. The clinical outcome was
reported the successful use of IHR to treat a fixed, irre- satisfactory: pain subsided in all patients, the Trendelen-
ducible congenital dislocation of the hip associated with burg sign became negative in all but 3 patients, no patient
other multiple lower extremity growth disturbances sec- had LLD, and lower limb alignment was reestablished.
ondary to neonatal multifocal osteomyelitis. Kocaoglu Manzotti et al28 reviewed their experience with IHR to
et al32 used IHR for the treatment of hip dislocation in 14 treat late sequelae of septic arthritis of the hip in 15 pa-
patients in adolescents and young adults (mean age, 20 y) tients (mean age, 21.1 y) with a mean LLD of 6.5 cm. At

r 2013 Lippincott Williams & Wilkins www.pedorthopaedics.com | S49

Choi et al J Pediatr Orthop  Volume 33, Number 1 Supplement, July/August 2013

FIGURE 2. Determination of varus angulation of the distal femur for Ilizarov hip reconstruction. A, The CORA method recom-
mended by Paley,12,19 which utilizes an imaginary proximal mechanical axis line. The proximal mechanical axis line corresponds
to a line perpendicular to the horizontal pelvic line, passing through the point of 1/3 to 1/2 the distance lateral to the medial edge
of the proximal fragment. B, Preoperative simulation of the Ilizarov hip reconstruction, which utilizes a trigonometric equation,
recommended by Kadykalo and Kuftyev,22 taking into consideration of proximal femoral (PF) valgus angulation and distal femoral
varus angulation and the biomechanical limb axis after reconstruction (Table 3). C, When determining the level of distal
osteotomy, the equalization of distances between the midline of the body axis and the centers of the knee joints of affected and
normal contralateral limbs should be considered, which was emphasized by Pafilas and Nayagam.13 The level of distal femoral
osteotomy relies on the level of PF valgus osteotomy, that is, the higher the level of PF valgus angulation, the more proximally
located distal femoral osteotomy should be to equalize the distances of knee joints from the midline of the body axis in bipedal
stance.

an average follow-up of 108 months after removal of the

TABLE 3. K-Value in the Design of the Biomechanical Axis22 fixator, 10 patients had an excellent or good result,
(Refer to Fig. 2B, a = K L) whereas 5 patients had a fair (LLD > 2.5 cm, de-
a (Deg.) formity > 10 degrees, decreased hip and/or knee range
of motion between 10 and 20 degrees, a positive
b (Deg.) 5 10 15 20 25 30 35 40
Trendelenburg sign, or mild pain) or poor(LLD > 5 cm,
5 0.50 0.67 0.76 0.81 0.85 0.87 0.89 0.91 deformity > 10 degrees, loss of hip and/or knee range
10 0.34 0.51 0.61 0.68 0.74 0.78 0.81 0.84
15 0.26 0.41 0.52 0.60 0.66 0.71 0.75 0.78
of motion >20 degrees, or continuous pain) result.
20 0.21 0.35 0.45 0.53 0.60 0.65 0.71 0.74 Rozbruch et al12 similarly reported the results of IHR for
25 0.17 0.30 0.40 0.48 0.55 0.61 0.66 0.71 the treatment of the late sequelae of infantile hip infection
30 0.15 0.27 0.37 0.45 0.51 0.58 0.63 0.68 in 8 patients (mean age, 11.2 y) with improvement in pain,
35 0.14 0.25 0.34 0.42 0.49 0.55 0.61 0.67 gait, LLD, and lower-limb alignment. At a mean follow-
40 0.12 0.23 0.31 0.40 0.47 0.53 0.59 0.65
up of 5 years (range, 1.9 to 9.8 y), LLD improved from a

S50 | www.pedorthopaedics.com r 2013 Lippincott Williams & Wilkins

J Pediatr Orthop  Volume 33, Number 1 Supplement, July/August 2013 The Role of Pelvic Support Osteotomy

FIGURE 3. A case of Ilizarov hip reconstruction complicated by refracture at the level of valgus angulation. A, Preoperative
radiograph of the pelvis taken in the standing position in a 12-year and 5-month-old boy with complete absence of the femoral
head and neck after infantile septic arthritis of the hip. B, A radiograph showing Ilizarov hip reconstruction and 6 cm length gain
with use of circular fixator. C, A radiograph showing refracture at the level of valgus angulation at age 15 years. D, A radiograph
showing repeat Ilizarov hip reconstruction using a monolateral fixator. Three centimeter length gain was obtained. E, A tele-
radiogram of the lower extremities taken at age 21 years showing a well-performed “pelvic support,” in which the acetabulum is
rested on the lesser trochanter.

mean of 4.6 (range, 0.6 to 6.4 cm) to 0.8 cm (range, 0 to LLD. Four of 5 patients with persistent Trendelenburg
1.2 cm). Modified hip score improved from a mean of 51 sign had residual poliomyelitis. Marimuthu et al26 and
(range, 21 to 67) to 73 points (range, 64 to 79). Gait Mahran et al27 also reported overall satisfactory results of
analysis data, which was performed in 5 patients, revealed IHR for chronically dislocated hips or destructed hips due
that the mean stance-time asymmetry improved from to various causes. Gursu et al35 studied the influence of
16% to 5.4%, and the mean ground-reaction force (sec- etiology of hip instability on the results of IHR. Preceding
ond peak) improved from 102% of body weight to 122% pathology included neglected congenital dislocation of
of body weight. El-Mowafi et al24 reported on the use of the hip in 13 hips (12 patients) and septic hip sequelae in 8
IHR in a study of 25 patients (mean, 22.4 y) with unstable hips. At a mean follow-up of 33.5 months, there was no
hips. At a mean follow-up of 4.5 months, all patients were significant difference between the final Harris Hip Scores
reported to be pan free with improvement in gait and of the 2 groups. They observed that complications were

r 2013 Lippincott Williams & Wilkins www.pedorthopaedics.com | S51

Choi et al J Pediatr Orthop  Volume 33, Number 1 Supplement, July/August 2013

FIGURE 4. The newly reconstructed weight-bearing area is not absolute, nor a true joint or false articulation. A, A radiograph of
Ilizarov hip reconstruction in a 15-year and 1-month-old boy. B, When determining the center of rotation during adduction/
abduction motion, the main fulcrum is seemingly located around the lesser trochanter and not around the apex of the valgus
angulation adjacent to the ischial tuberosity during passive abduction and adduction. This suggests that the center of rotation
seems to vary with the position/direction of motion of the lower limbs and depends on the soft-tissue interpositional weight-
bearing surface between the PF osteotomy and the pelvis.

more frequent after congenital dislocation of the hip, and IHR in 16 patients (mean age, 25.3 y). At a mean follow-
patient satisfaction was higher in cases with sequelae of up of 52.5 months, the mean Harris Hip Score increased
sepsis. from 50 points (range, 32 to 73 points) preoperatively to
Emara31 reported the results of IHR in 11 patients 87.6 points (range, 67 to 98 points). Four patients retained
who had undergone excision hip arthroplasty to treat a positive Trendelenburg sign. We have also experienced
resistant hip infection. Harris Hip Scores improved in all that monolateral fixator that enables multiplanar angular
patients: the average score preoperatively was 43.5 (range, correction and translation (Dyna-ATC; BK Meditech,
31 to 50), whereas at final follow-up, the average score Seoul, Korea) were suitable for IHR. Recently, Krieg
was 70.9 (range, 65 to 80). Shetty et al29 reported the et al36 proposed a new technique of IHR with internal
efficacy of a modified PSO using hybrid external fixator in systems exclusively. A locking compression plate is ap-
8 patients (mean age, 16.4 y) with severe bilateral hip in- plied to fix the PF valgus-extension osteotomy and a
volvement in spondylo-epiphyseal dysplasia congenita. motorized retrograde intramedullary lengthening device
After a mean follow-up of 25.9 months, the mean modi- (Fitbone; Wittenstein, Igersheim, Germany) for the distal,
fied Harris Hip Score had improved from 67.9 to 79.1 lengthening-varization osteotomy. Although the use of
points. Waddling gait was absent in 3 patients, reduced in internal implants for IHR is a feasible and patient-friendly
4 patients, and was the same in 1 at the last follow-up. alternative to traditional methods, their use, however, may
Mean LLD was >0.5 cm, and the mechanical axis was be restricted by geometric preconditions.
realigned in all. Sabharwal and Macleod30 reported a case Reported complications of IHR in the literature
of successful IHR for the management of advanced os- (Table 2) included knee stiffness, pin-tract infection, de-
teonecrosis of the proximal femur after chemotherapy for layed consolidation, refracture (Fig. 3), obturator nerve
acute lymphoblastic leukemia in an adolescent. They entrapment, straightening of proximal valgus angulation,
performed the 2-level femoral osteotomy with acute val- and persistent T-gait. We think that there is possibility of
gus-extension angulation at the PF osteotomy site and the ischiofemoral impingement if the PF valgus angulation
DF osteotomy for gradual lengthening and varus angu- site directly abuts the ischium.
lation using Taylor Spatial Frame (Smith and Nephew,
Memphis, TN). DISCUSSION
Monolateral external fixation is usually more com-
fortable than circular frames for the patient with de- How to Avoid or Reduce Remodeling of
formity around the hip and proximal femur, Inan and Proximal Valgus Angulation?
Bowen33 reported on the use of monolateral external IHR is safe to perform at an older age, preferably
fixator (Orthofix S.R.L.; Bussolengo, Verona, Italy) for after peak growth spurt. Rozbruch et al12 observed that

S52 | www.pedorthopaedics.com r 2013 Lippincott Williams & Wilkins

J Pediatr Orthop  Volume 33, Number 1 Supplement, July/August 2013 The Role of Pelvic Support Osteotomy

FIGURE 5. Total hip replacement arthroplasty (THRA) performed in a Mongolian 34-year-old lady who underwent Ilizarov hip
reconstruction for treatment of neglected developmental dislocation of the hip at 24 years of age in Russian Federation. She
underwent THRA because of diffuse pain around the buttock and hip that had lasted after Ilizarov hip reconstruction. A and B,
Preoperative radiographs of the hip. The deformed femoral head is still in close contact with the secondary socket (false ace-
tabulum) of the dislocated hip, which may explain the source of pain. The pelvic support osteotomy without resection of the
femoral head might have been contraindicated in this regard. C, Three-dimensional computed tomography of the hip. D and E,
Postoperative radiograph of THRA with a straight stem inserted after 4.5 cm of segmental shortening at the subtrochanteric level.

when IHR was performed at a younger age before ado- may also be beneficial. Most importantly, postoperative
lescence, the PF valgus osteotomy site completely re- strenuous abductor muscle exercise is essential.
modeled, demonstrating no evidence of the pelvic support
within 1 or 2 years after the operation. We also experi-
enced the same phenomenon of remodeling (straighten-
ing) of the proximal femur when IHR was performed in Where Is the Weight-bearing Fulcrum?
the preadolescent age. One should consider adding ex- We agree with others12,13 that the weight-bearing
travalgus angulation at the PF osteotomy site, when area is not absolute, nor a true joint or false articulation.
performing IHR to address marked LLD in a younger The center of rotation seems to vary with the position/
age. As mentioned previously, one should expect repeat direction of motion of the lower limbs and depends on the
IHR at or near skeletal maturity to obtain a level pelvis soft-tissue interpositional weight-bearing surface between
and to eliminate residual LLD. Another alternative is to the PF osteotomy and the pelvis. To identify the center of
perform simple femoral lengthening with extension of the rotation during adduction/abduction motion, we used
external fixation to the pelvis at a younger age and to cineradiography in a patient who underwent Shanz-type
reserve PSO for the second lengthening when the patients PSO and managed to figure out that the center of rotation
are near skeletal maturity.12 was located around the lesser trochanter and not around
In our experience, translation of the proximal the apex of the valgus angulation adjacent to the ischial
fragment medially relative to the distal fragment helps tuberosity during passive abduction and adduction
facilitate and maintain valgization. Prebent plate fixation (Fig. 4).

r 2013 Lippincott Williams & Wilkins www.pedorthopaedics.com | S53

Choi et al J Pediatr Orthop  Volume 33, Number 1 Supplement, July/August 2013

Why Does T-Gait Persist After PSO in Some of the external fixation to the pelvis at a younger age, and
Patients? to reserve PSO for the second lengthening. Future studies
The literature suggests that an average of 30.3% should be directed toward determining the precise loca-
(range, 0% to 62.5%) of patients have a persistent pos- tion of the newly reconstructed weight-bearing fulcrum
itive Trendelenburg sign after IHR, although it is reduced and to develop an effective means of normalizing ab-
in severity in most patients (Table 2). This persistence ductor muscle function.
may be due to abductor insufficiency related to atrophied
abductor muscles before surgery or loss of fulcrum during
ACKNOWLEDGMENTS
follow-up due to remodeling of PF valgus angulation.
Age at the time of PSO may also be an important factor The authors thank Yun Su Park, MD, Professor of
for the retention of hip function.7 Inan et al34 used the Department of Orthopaedic Surgery, Samsung Medical
magnetic resonance imaging to measure alterations in the Center, Sungkyunkwan University School of Medicine, for
length and volume of the gluteus medius muscle after providing us a case of THRA performed in a patient with
IHR in 11 patients with a history of congenital dis- previous history of IHR.
location of the hip. They reported that although the IHR
achieved a functional and painless hip in all patients, 5 REFERENCES
patients had a persistently positive Trendelenburg sign at 1. Choi IH, Shin YW, Chung CY, et al. Surgical treatment of the
a mean follow-up of 3 years after IHR. The muscle vol- severe sequelae of infantile septic arthritis of the hip. Clin Orthop
umes were restored to 43% to 89% of the muscle volumes Relat Res. 2005;434:102–109.
2. Lorenz A. On the treatment of irreducible congenital hip dislocation
on the normal contralateral side, and the postoperative and the femoral neck pseudarthrosis by means of bifurcation
muscle volume correlated significantly with the result of osteotomy (Bifurcation of the proximal femur). Wien Klin
the Trendelenburg test. T-gait disappeared with the re- Wochenschr. 1919;32:997–999.
storation of gluteus muscle volume. However, there was 3. Von Baeyer H. Surgical treatment of irreducible congenital hip
dislocation. Munster Med Wochenschrift. 1918;165:1216–1217.
no correlation between the Trendelenburg test and the 4. Schanz A. On the treatment of old congenital hip dislocation.
change in the length of the gluteus medius muscle. In their Munich Med Wochenschr. 1922;69:930–931.
series, 4 of 5 patients with a persistently positive T-gait 5. Milch H. The pelvic support osteotomy. J Bone Joint Surg.
were at least 31 years of age at the time of operation. On 1941;23:581–595.
the basis of these results, they speculated that atrophied 6. Milch H. The post-osteotomy angle. J Bone Joint Surg. 1943;25:
394–400.
muscle might not be restored by PSO in older patients. 7. Milch H. The resection-angulation operation for hip-joint disabil-
ities. J Bone Joint Surg Am. 1955;37-A:699–717.
Is It Possible to Convert to THRA Later on? 8. Milch H. Osteotomy at the Upper End of the Femur. Baltimore:
Lippincott Williams & Wilkins; 1965.
IHR does not burn the bridge in terms of preser- 9. Peltier LF. The classic. The “pelvic support” osteotomy. Clin Orthop
vation of bone stock around the PF, and thus, THRA Relat Res. 1989;249:4–11.
after PSO is still possible, although it is a technically 10. Henderson RS. Osteotomy for unreduced congenital dislocation of
demanding procedure due to the PF deformity (Fig. 5). the hip in adults. J Bone Joint Surg Br. 1970;52:468–473.
Careful attention to surgical detail is essential for suc- 11. Hass J. A subtrochanteric osteotomy for pelvic support. J Bone
Joint Surg. 1943;25-A:281–291.
cessful THRA, otherwise, the distal end of femoral stem 12. Rozbruch SR, Paley D, Bhave A, et al. Ilizarov hip reconstruction
can penetrate the cortex.37 Shiltenwolf et al21 found that for the late sequelae of infantile hip infection. J Bone Joint Surg Am.
THRA could be performed without difficulty by 2005;87-A:1007–1018.
straightening the proximal femur by osteotomy and using 13. Pafilas D, Nayagam S. The pelvic support osteotomy: indications
and preoperative planning. Strategies Trauma Limb Reconstr.
a long-stem prosthesis. 2008;3:83–92.
14. Catagni MA, Malzev V, Kirienko A. Treatment of disorders of
hip joint. In: Maiocchi AB, ed. Advances in Ilizarov Apparatus
CONCLUSIONS Assembly. Milan: Il Quadratino; 1994:119–134.
IHR effectively improves hip abductor bio- 15. Ilizarov GA, Kaplunov AG, Tereschenko VA. Modified pelvic
mechanics by PF valgus osteotomy for pelvic support, support osteotomy for restoration of the pelvic support and
and thus eliminates or reduces T-gait and, at the same simultaneous femoral lengthening. Published in Russian; 1978.
16. Ilizarov GA, Samchukov ML, Kurtov VM. Ilizarov restorative
time, corrects LLD and knee joint malalignment, if any, reconstruction surgery in the treatment of hip osteoarthritis,
by DF osteotomy. In this regard, IHR is an excellent congenital and pathological hip dislocation in children and adults.
option for the treatment of unstable, irreducible hips as- Materials of Second International Symposium on Experimental,
sociated with either severe dysplastic acetabulum or par- Theoretical, and Clinical Aspects of Transosseous Osteosynthesis.
Kurgan, USSR: Kurgan Scientific Research Institute; 1986:100–102.
tial or total absence of the femoral head and neck in 17. Ilizarov GA, Samchukov ML. Ilizarov reconstructive surgery in the
adolescents and young adults. If IHR is performed at a treatment of hip osteoarthritis. Ortop Travmatol Protez. 1988:10–13.
younger age to correct marked LLD, remodeling of PF 18. Ilizarov GA. Treatment of disorders of the hip. In: Green SA, ed.
valgus angulation is inevitable with resultant loss of pelvic Transosseous Osteosynthesis. Berlin: Springer Verlag; 1992:668–696.
support. In this situation, second IHR should be con- 19. Paley D. Hip joint considerations. Principles of Deformity Correc-
tion. Heidelberg: Springer-Verlag; 2002:647–694.
templated near or at skeletal maturity to obtain a level 20. Samchukov ML, Birch JG. Pelvic support femoral reconstruction
pelvis and to eliminate residual LLD. Another alternative using the method of Ilizarov: a case report. Bull Hosp Jt Dis.
is to perform simple femoral lengthening with extension 1992;52:7–11.

S54 | www.pedorthopaedics.com r 2013 Lippincott Williams & Wilkins

J Pediatr Orthop  Volume 33, Number 1 Supplement, July/August 2013 The Role of Pelvic Support Osteotomy

21. Schiltenwolf M, Carstens C, Bernd L, et al. Late results after 29. Shetty GM, Song HR, Lee SH, et al. Bilateral valgus-extension
subtrochanteric angulation osteotomy in young patients. J Pediatr osteotomy of hip using hybrid external fixator in spondyloepiphy-
Orthop B. 1996;5:259–267. seal dysplasia: early results of a salvage procedure. J Pediatr Orthop
22. Kadykalo OA, Kuftyev LM. Some biomechanical principles of the hip B. 2008;17:21–25.
reconstruction with defect on head and neck of the femur by Ilizarov 30. Sabharwal S, Macleod R. Ilizarov hip reconstruction for the
method. The Value of General Biological Patterns in Regeneration management of advanced osteonecrosis in an adolescent with
Tissue Opened by G.A. Ilizarov. Scientific Works Collection. 13th ed. leukemia. J Pediatr Orthop B. 2012;21:252–259.
The Ministry of Health, The RSFSR (The Russian Soviet Federative 31. Emara KM. Pelvic support osteotomy in the treatment of patients
Socialist Republic) Kurgan All-Union Scientific Center: Rehabilita- with excision arthroplasty. Clin Orthop Relat Res. 2008;466:
tion Traumatology and Orthopaedics; 1988:124–129. 708–713.
23. Samchukov ML, Ivanova IV, Dolganov DV. Analysis of Pelvo- 32. Kocaoglu M, Eralp L, Sen C, et al. The Ilizarov hip reconstruction
Trochanteric Muscles and Weight-Bearing After Ilizarov Femoral osteotomy for hip dislocation: outcome after 4-7 years in 14 young
Reconstruction in Patients With Hip Osteoarthritis. USSR: The patients. Acta Orthop Scand. 2002;73:432–438.
Congress of Scientific and Technical Advances of the Young 33. Inan M, Bowen RJ. A pelvic support osteotomy and femoral
Scientists of the Urals; 1986:86–87. lengthening with monolateral fixator. Clin Orthop Relat Res.
24. El-Mowafi H. Outcome of pelvic support osteotomy with the 2005;440:192–198.
Ilizarov method in the treatment of the unstable hip joint. Acta 34. Inan M, Alkan A, Harma A. Gluteal muscle after pelvic osteotomy
Orthop Belg.. 2005;71:686–691. to treat congenital dislocation of the hip. J Bone Joint Surg. 2005;
25. Inan M, Bomar JD, Kucukkaya M, et al. A comparison between the use 87-A:2246–2252.
of a monolateral external fixator and the Ilizarov technique for pelvic 35. Gursu S, Demir B, Yildirim T, et al. The influence of aetiology of
support osteotomies. Acta Orthop Traumatol Turc. 2004;38:252–260. hip instability on the results of pelvic support osteotomy. Hip Int.
26. Marimuthu K, Joshi N, Sharma CS, et al. Ilizarov hip reconstruc- 2010;20:518–523.
tion in skeletally mature young patients with chronic unstable hip 36. Krieg AH, Lenze U, Hasler CC. Ilizarov hip reconstruction
joints. Arch Orthop Trauma Surg. 2011;131:1631–1637. without external fixation: a new technique. J Child Orthop. 2010;4:
27. Mahran MA, Elgebeily MA, Ghaly NA, et al. Pelvic support 259–266.
osteotomy by Ilizarov’s concept: is it a valuable option in managing 37. Thabet AM, Catagni MA, Guerreschi F. Total hip replace-
neglected hip problems in adolescents and young adults? Strategies ment fifteen years after pelvic support osteotomy (PSO): a case
Trauma Limb Reconstr. 2011;6:13–20. report and review of the literature. Musculoskelet Surg. 2012;96:
28. Manzotti A, Rovetta L, Pullen C, et al. Treatment of the late 141–147.
sequelae of septic arthritis of the hip. Clin Orthop Relat Res. 38. Bleck EE. Orthopaedic Management in Cerebral Palsy. London:
2003;410:203–212. Mac Keith Press; 1987.

r 2013 Lippincott Williams & Wilkins www.pedorthopaedics.com | S55