Patients who had been treated with a subtrochanteric femoral valgus
osteotomy by the senior one of us (R.H.G.) were identified by a review of
surgical codes. The hospital and office charts and radiographs of all of those
patients were then reviewed. The indications for the surgical procedure,
findings of preoperative and postoperative examinations, complications, and
surgical results at the time of follow-up were determined from the medical
record. A telephone survey was administered to the patients' caregivers to
further ascertain the preoperative and postoperative condition of the patient
(see Appendix). This survey included questions regarding hip pain, sitting
tolerance, perineal care, postoperative complications, and comorbidities. To
ensure that the survey results were as unbiased as possible, a medical student
rather than the treating surgeon contacted the caregivers. Institutional
review board approval was obtained for the entire study.
Symptomatically dislocated hips in patients who were unable to walk were
treated surgically. The most common indication for surgery was pain
(twenty-two of the twenty-four patients) followed by problems with perineal
care and positioning (twelve patients). As the majority of these patients were
nonverbal and had severe neuromuscular spasticity, pain was identified by
observing the patient's reaction to changes in position both while seated and
during transfers. Grimacing, posturing, and increasing agitation are examples
of behavior exhibited by these patients when in pain. The caregivers, who
interacted with the patients daily, often provided the best interpretation of
the patient's changing response and pain with prolonged sitting, transfers,
and diaper changes. Only one patient had pressure sores preoperatively. Twelve
patients had concurrent surgical procedures, including phenol injections (five
patients), adductor releases (five), removal of hardware (two), femoral head
resection (one), tendon transfers in the wrist (one), excision of heterotopic
ossification (one), and hamstring-lengthening (one). The need for concurrent
soft-tissue releases was determined intraoperatively. An inability to abduct
the distal fragment of the osteotomy because of soft-tissue tension was the
indication for an abductor release. Simultaneous bilateral procedures were
performed in seven patients. No procedure was performed on the contralateral
hip in the absence of pain or difficulty with positioning, even if the hip was
dislocated.
Surgical Procedure
A subtrochanteric femoral osteotomy was performed with modifications of the
technique first described by
Hass15
(Fig. 1). A curvilinear
incision was made from the tip of the dislocated femoral head, coursing gently
posteriorly toward the greater trochanter and extending distally along the
lateral aspect of the thigh. The incision was carried down through the
iliotibial band. A circumferential subperiosteal dissection of the proximal
part of the femur was performed at the level of the lesser trochanter. At this
stage, it is very important to release the iliopsoas tendon insertion from the
lesser trochanter. A longitudinal incision was then made in the hip capsule to
expose the hip joint and the entire proximal part of the femur to the level of
the acetabulum.
A second capsular incision was made parallel to the acetabulum in a
"T" shape to allow visualization of the femoral head. The proximal
femoral valgus osteotomy was performed by making an oblique cut from the
lateral cortex to a point on the inferior margin of the lesser trochanter. A
second cut of the proximal part of the femur created a laterally based wedge
that, when removed, allowed abduction or valgus correction. Typically, a wedge
of 45° to 60° was removed (Fig.
2).
Fixation of the osteotomy site has evolved over time. Originally, a single
3.5-mm small-fragment plate was used, but problems with hardware failure led
to a change to a contoured 4.5-mm large-fragment plate. However, the
large-fragment plate is more difficult to contour, and recently this method
was further modified. Currently, we use two 3.5-mm small-fragment plates to
secure the osteotomy site; the first is contoured and placed laterally, and a
second, straight plate is placed on the anterior aspect of the femur
(Fig. 3). The proximal fragment
(femoral head and neck) is positioned in a valgus orientation such that the
lesser trochanter articulates with the acetabulum. The positions of the
osteotomy and plates were verified intraoperatively with the image intensifier
and postoperatively with radiographs (Figs.
4-A and
4-B).
Postoperatively, patients received physical therapy for hip motion, and the
family was instructed about performing daily stretching exercises to maintain
abduction. Initially, skin traction was maintained for the first week
following the procedure, but more recently we have used only an abduction
pillow between the legs and early motion exercises. Patients were typically
discharged on the third postoperative day.
We reviewed the results in thirty-one hips in twenty-four patients who had
been treated with a subtrochanteric femoral valgus osteotomy at our
institution between January 1995 and February 2005 by the senior one of us
(R.H.G.). The average age at the time of the surgery was fourteen years
(range, 4.5 to thirty-one years). The average duration of follow-up following
the surgery was 44.5 months (range, eight to eighty-five months). All but two
of our patients were nonverbal. Twenty had been diagnosed with the spastic
quadriplegic form of cerebral palsy. Other conditions included, in one patient
each, spina bifida, spastic hemiplegia, leukodystrophy with spasticity, and
Sturge-Weber syndrome. The patient with spina bifida had developmental delay
and a shunted hydrocephalus. Although it is unusual for a child with spina
bifida, this child had lower-extremity spasticity. A tethered cord was
identified and released two months following the hip surgery, but the
spasticity did not completely resolve. The patient with Sturge-Weber syndrome
had muscular spasticity of the upper and lower extremities.
Fifteen (63%) of the twenty-four patients had at least one complication
(see Appendix), with the complications typically occurring within the first
few months following the surgery. Two patients continued to have pain
following surgery. The plates and screws were eventually removed and the
femoral head was excised, and this led to pain resolution in both patients.
Another patient who had excellent relief of pain in the immediate
postoperative period had severe unilateral hip pain beginning a month
following bilateral osteotomy. Six weeks later, the pain resolved
spontaneously and the patient has continued to do well. Three patients
required revision because of hardware failure. One of these patients was the
only one in the series in whom a pressure sore developed. All of the hardware
failures occurred while we were using a single 3.5-mm small-fragment plate for
fixation. In another patient, pain developed fourteen months after the
osteotomy. It was treated successfully with the removal of protruding screws.
There was one intraoperative femoral fracture. Two patients had heterotopic
bone formation, which was massive in one of them, subsequently resulting in
fusion of the hip (Fig. 5).
This patient initially had had a small amount of heterotopic ossification
which had been removed at the time of the initial surgery. In the other
patient, heterotopic ossification developed following a wound infection.
Four patients had a postoperative infection: a urinary tract infection
developed in one; pneumonia, in two; and a wound infection, in two (one of
whom also had pneumonia). Two patients had a seroma requiring surgical
drainage. One patient with pneumonia required intubation and a prolonged stay
in the intensive-care unit, before being discharged home on the seventieth
postoperative day. Another patient went home on the day of the surgery.
Excluding these outliers, the average hospital stay was 3.5 days.
Despite these complications, most patients were doing well at an average of
forty-four months postoperatively. According to the clinic notes, twenty-one
of the twenty-four patients had improved sitting. One patient had a recurrence
of adduction contractures related to his inability to receive physical
therapy, and he ultimately died from an unrelated condition. Two patients had
persistent pain with prolonged sitting, but each of them had scoliosis
exceeding 120°. Scoliosis surgery had not been performed because of
pulmonary compromise in one and the family's refusal to accept blood products
for religious reasons in the other.
During the proximal femoral surgery, the lesser trochanter is directed into
the acetabulum. This position was successfully obtained in all but three of
the thirty-one hips. There was no clear association between the final position
of the lesser trochanter and postoperative pain.
Sixteen caretakers of the patients participated in the follow-up telephone
survey. This survey was performed retrospectively at a variable length of time
after the surgical procedure. In all cases, the person who completed the
survey was the primary caregiver for the patient, both before and after the
procedure. Five caregivers could not be contacted and two patients had died.
One patient was in a nursing home, and the nurses refused to complete the
survey because of privacy concerns. The remainder of the patients continued to
live at home with their caregiver.
Prior to the surgery, twelve of the sixteen patients could not sit for more
than one hour without pain. One could not sit at all, and six could sit
comfortably for less than thirty minutes. Following the surgery, eleven of the
patients could sit for at least two hours; however, many could sit for as long
as eight hours without any pain or discomfort. The other patients could sit
for anywhere between one and two hours without showing any signs of pain.
Transfers can be difficult for these patients. Prior to surgery, all but
three of our patients had difficulty with transfers. Following surgery, twelve
patients displayed no signs of discomfort during transfers, whereas the
remaining four grimaced while being moved in and out of a chair. Only one
caregiver reported pressure sores prior to the surgery, and this was unchanged
after the surgery.
Prior to the surgery, two patients had recurrent urinary tract infections
and five had pneumonia; one patient had pneumonia nine times. After the hip
surgery (and the immediate postoperative period), there were no reported
urinary tract infections and only one case of pneumonia.
Despite the complications reported here, fourteen caregivers stated that
hip motion had been improved and pain had been decreased after the surgery. Of
fifteen caregivers (one caregiver did not understand the question and did not
answer), fourteen stated that they would have the surgery done again if they
had the decision to make. They were satisfied with the results of the
procedure. The one patient whose caregivers stated that they would not have
the procedure done again had the development of severe scoliosis, which was
not amenable to surgery because of the parents' refusal to allow the child to
receive blood products. The child had pain with sitting and positioning.
Subtrochanteric femoral osteotomy to treat congenital hip dislocation was
first described in 1838 by Bouvier in Paris, but it did not become popular
until the early
1900s16. Use of a
valgus osteotomy for chronically dislocated hips was described by Haas in 1924
and was described in the English-language literature in
194315. As
originally described, the procedure was designed to restore hip stability,
relieve pain, and maintain mobility. It was based on the case of a
forty-two-year-old woman with unrecognized bilateral traumatic hip
dislocation. With time, the proximal parts of the femora had migrated
superiorly and the lesser trochanters were articulating with the acetabula.
The patient had no hip pain. The original operative technique, as described by
Hass, was a locking triangular osteotomy positioned so that the lesser
trochanter was directed into the acetabulum. No internal fixation was
utilized, but a spica cast was worn for three months. Hass reported his
follow-up results in eighteen patients (twenty-two hips) with congenital hip
dislocation; he did not mention neuromuscular
disease15. Hass
also did not report the amounts of postoperative hip motion, but he did
comment that "all patients had adequate motion with the exception of one
who had had limited motion before operation." This osteotomy was
considered a modification of an oblique bifurcation osteotomy, described by
Lorenz in 1917, which placed the distal spike of bone into the
acetabulum16,17.
A similar procedure had also been described by Schanz in the German literature
in 1922, but the position of the described osteotomy was slightly more distal,
at the level of the ischial
tuberosity16,18.
Schanz also placed long screws (Schanz pins) into the femur proximal and
distal to the osteotomy site. The pins were used to measure angulation and to
control osteotomy alignment. These pins protruded from the skin and were
incorporated into the spica cast to maintain the position of the osteotomy.
They were removed four weeks later through a window in the cast; the spica
cast remained in place for an additional two
months19. Blount
and others later described the use of internal fixation for these
osteotomies16,20.
In America, Milch described the biomechanics of these pelvic support
osteotomies and proposed additional
modifications17,21,22.
On the basis of the original descriptions of these osteotomies, the procedure
described in the present paper is most similar to that described by Hass in
194315, although it
was modified by the addition of internal fixation and elimination of the
postoperative spica cast.
Most reports on proximal femoral valgus osteotomy indicated that it was a
useful surgical option for a missed congenital hip dislocation and, prior to
the advent of the modern total hip replacement, for osteoarthritis in an
ambulatory young adult. In 1983, in Germany, Kuesswetter et al. reviewed their
eleven-year follow-up results of a subtrochanteric angulation procedure with
resection of the femoral
head23. Most
patients did surprisingly well, with the majority able to go for daily walks.
Poor results were attributed to the development of secondary arthritic changes
between the acetabulum and the lesser trochanter. In a study of twenty-eight
subtrochanteric femoral valgus osteotomies without femoral head resection,
Schiltenwolf et al. reported that, at an average of seventeen years, most
patients had maintained abduction and improvements in
gait24.
It is not clear when the subtrochanteric valgus femoral osteotomy was first
used for children with neuromuscular hip dislocations, but it is included in
the review, by Samilson et al., of treatments used from 1960 to 1970 for hip
dislocation associated with cerebral
palsy25. More
recently, McHale et al. described a valgus femoral osteotomy combined with a
femoral head resection in a series of six hips in children with cerebral
palsy12. They
reported excellent pain relief and few complications.
Critics of the valgus osteotomy believe that it is suboptimal because of
persistent pain and postoperative problems with pressure ulcerations from the
retained femoral
head12,14.
However, these complications were not frequently reported in the original
literature on the
procedure15-19,21,26.
Recently, Aksoy and
Musdal27 reported
the results at eight to twenty-five years following fifty subtrochanteric
valgus osteotomies with retention of the femoral head performed for missed
congenital hip dislocations in young adults. All patients had improvements in
Harris hip scores and functional capacity. Four of thirty-five patients were
dissatisfied, however, because of inadequate pain relief.
Although the early complication rate was high in our series, persistent hip
pain was not a problem and only one patient had a decubitus ulcer, which had
existed prior to the surgery and did not heal postoperatively. The results of
our survey, although nonvalidated and performed retrospectively, clearly
illustrated that the majority of caregivers believed that the surgery had
increased the length of time that the patient could sit comfortably and made
daily care (transfers and diaper changes) less uncomfortable. Concurrent
femoral head resection was done in one patient, and it was performed as a
second procedure in two patients who had pain that was possibly associated
with fixation hardware or impingement of the femoral head against the pelvis.
The pain resolved following the femoral head resection and removal of the
hardware in both of those patients. However, as the majority of the patients
did not have complications related to retention of the femoral head, we do not
routinely resect it during this procedure.
Although a 63% complication rate is high, the types of patients treated in
this series can be expected to have a high complication rate regardless of the
procedure performed. For example, in a retrospective analysis of complications
of osteotomies about the hip in seventy-five patients with cerebral palsy,
Stasikelis et al. reported a 25% overall complication
rate28. Higher
rates of complications were found in patients who could not walk (29%) and
children with gastrostomies or tracheotomies (68%).
Other authors have reported even higher complication rates. Widmann et al.
reported that, of thirteen patients treated with resection arthroplasty of the
hip, eighteen had complications, including pneumonia, heterotopic
ossification, and decubitus
ulcers29. In 2005,
Leet et al. compared patients who had undergone a McHale procedure with those
treated with a femoral head resection and
traction11.
Although both groups had decreases in pain and improvements in sitting
tolerance, the patients treated with the McHale procedure had shorter hospital
stays and fewer complications (a rate of 27% compared with 55% overall). The
complication rate following hip arthrodesis was 42% in a series of fourteen
patients with cerebral palsy reported on by Fucs et
al.30. A
complication rate of 36% (four of eleven patients) was reported following
prosthetic interposition
arthroplasty31.
Total hip replacement in patients with cerebral palsy has also had comparable
rates of postoperative complications, including dislocation and component
loosening in addition to urinary tract infections, heterotopic ossification,
and decubitus
ulcers13.
In conclusion, the valgus subtrochanteric femoral osteotomy has been
criticized for providing inadequate pain relief and predisposing to pressure
sores from the retained femoral head in patients with hip dislocation
secondary to spasticity. In our patients with painful spastic hip dislocation,
this procedure reliably alleviated pain, increased sitting time, made
transfers easier, and improved perineal hygiene. Simultaneous resection of the
femoral head, as described by McHale et
al.12, was not
necessary to achieve these outcomes. Preoperatively, the caregiver should be
told that complications are common. However, these primarily early
complications do not seem to negatively affect the long-term outcome of this
procedure. Children with recently occurring spastic hip dislocations who may
have the potential to walk should be treated with tendon releases and pelvic
and femoral osteotomies with the goal of relocating the femoral head in the
acetabulum32,33.
However, the modified Hass osteotomy as described here provided excellent
midterm pain relief in the treatment of painful chronically dislocated spastic
hips in children who could not walk.
The telephone survey and a table showing clinical information on all
patients are available with the electronic versions of this article, on our
web site at
(go to
the article citation and click on "Supplementary Material") and on
our quarterly CD-ROM (call our subscription department, at 781-449-9780, to
order the CD-ROM). ?