Abstract
Background: The purpose of this study was
to assess the rate of union, time to union, and complications associated with
the extended slide trochanteric osteotomy. We also evaluated how
outcomes were influenced by the preoperative cortical-bone thickness,
the preoperative cancellous-bone quality of the greater trochanter,
the number of cables used to reattach the trochanteric osteotomy
fragment, and the use of cortical strut augmentation.
Methods: We reviewed the results for forty-six
hips in forty-five patients who underwent a revision total hip arthroplasty
with an extended slide trochanteric osteotomy between December 1991
and December 1996. Twenty-three patients were men, and twenty-two
were women; the mean age at the time of the operation was 66.3 years.
Two hips had an isolated acetabular revision, fifteen had an isolated femoral
revision, and twenty-nine had acetabular and femoral revisions.
One patient (one hip) was lost to follow-up.
Results: At a mean of forty-four months after
the operation, the rate of union of the distal osteotomy site was
98 percent (forty-four of forty-five hips), with no change in the
femoral component position. The time to union was not significantly
correlated with the number of cables, the preoperative cortical-bone
thickness, or the preoperative cancellous-bone quality of the greater
trochanter. Interestingly, the time to bridging-callus union was
significantly longer in the hips with a strut allograft than in
the hips without a strut allograft (p = 0.04, t test for independent
samples). Two fractures of the osteotomy fragment occurred, but
neither necessitated another revision.
Conclusions: The extended slide trochanteric
osteotomy allows extensive acetabular and femoral exposure, facilitates
removal of distal cement or a well fixed porous-coated stem, and
allows reliable reattachment and healing of the trochanteric fragment.
The number of patients requiring revision of a failed hip
arthroplasty is increasing2. In
such revisions, removal of the existing component, reconstruction
of the bone-stock deficiency, and achievement of solid fixation
with a new component can be technically demanding. Successful revision
requires a clear preoperative plan for adequate operative exposure,
removal of the prosthesis, and reconstruction. Exposure of the hip
by means of a standard trochanteric osteotomy has been widely used
in the past. However, complications associated with reattachment
of the fragment, such as nonunion, fibrous union, migration or fragmentation
of the trochanter, and wire breakage, have been common1,3.
Indications for the extended slide trochanteric osteotomy include:
(1) osteolysis or osteopenia of the greater trochanter rendering
the bone inadequate for either wire fixation or cable fixation after a
conventional osteotomy, (2) proximal-medial bone loss requiring
distal cable fixation for trochanteric reattachment, (3) the need
to revise a well fixed cemented or extensively porous-coated cementless
component, and (4) angular deformity of the proximal part of the
femur that obstructs cement removal or component reimplantation.
Advantages of the extended slide trochanteric osteotomy include
extensive exposure, preservation of soft-tissue attachments to the
trochanteric bone fragment, and maintenance or adjustment of abductor-muscle
tension.
To our knowledge, the results of this technique have not been
widely published. The purpose of this study was to assess the rate
of union, time to union, and complications associated with the extended
slide trochanteric osteotomy. Additionally, we evaluated how the
preoperative cortical-bone thickness, the preoperative cancellous-bone
quality of the greater trochanter, the number of cables used to
reattach the trochanteric osteotomy fragment, and the use of cortical
strut augmentation influenced the outcomes.
Between December 1991 and December 1996, forty-five patients
with forty-six failed hip arthroplasties underwent an extended slide
trochanteric osteotomy during a revision operation. There were twenty-three
men and twenty-two women, and the mean age at the time of the index
operation was 66.3 years (range, thirty-six to eighty-four years). Of
the forty-six hips, two had an isolated acetabular revision, fifteen
had an isolated femoral revision, and twenty-nine had both the acetabular
and the femoral component revised. An extensively porous-coated
stem (DePuy, Warsaw, Indiana) was used for all femoral revisions.
The main indication for the revision operation was aseptic loosening
of the acetabular or femoral component in forty-three hips (93 percent).
Other indications included failed bipolar hemiarthroplasty (one
hip) and recurrent hip dislocations with malposition of the femoral
component (two hips). Sixteen hips (35 percent) had undergone a
prior conventional trochanteric osteotomy. The cancellous structure of
the greater trochanter was extensively osteolytic in twenty-nine
hips and only minimally changed in seventeen. The proximal-lateral
femoral cortical bone proximal to the planned osteotomy site was often
of poor quality. On the average, the minimum cortical thickness
of the osteotomized bone segment measured only 2.8 millimeters (range, less
than one millimeter to ten millimeters) on the anteroposterior radiograph.
The reasons for the osteotomy included: (1) exposure for a complicated acetabular
reconstruction in twenty-four hips, (2) removal of a cemented stem
with a well bonded distal cement-bone interface or cement extending down
the canal in thirty-six hips, (3) removal of a cementless stem with
extensive porous coating or osseous ingrowth distally in seven hips,
and (4) proximal femoral deformity in twelve. Most of the patients
had more than one of these indications.
Operative Technique
All hips were treated through an extended posterolateral approach,
with the patient in the lateral decubitus position. The posterior
border of the vastus lateralis muscle was identified and stripped
from the femur. Using a sagittal saw, the surgeon performed the
osteotomy along the exposed posterior aspect of the femur, just
anterolateral to the linea aspera, extending from the greater trochanter
to a level on the femoral diaphysis determined by preoperative templating.
Then, the distal transverse cut was completed down to the implant.
After the posterior and distal cuts were completed, the anterolateral
cortex was cut from distal to proximal. The distal part of the cut
was done with a sagittal saw; the cut then was completed proximally
with a one-half-inch (1.3-centimeter) osteotome. When the osteotomy
was complete, any anterior capsule attached to the anterior trochanteric
fragment was divided, allowing the fragment to be separated from
the femur, elevated, and displaced anteriorly. In this way, a sling
attachment, consisting of the gluteus medius and minimus muscles
proximally and the vastus lateralis muscle distally, was left (Fig. 1). The osteotomy
site was beveled so that the diaphyseal portion included the lateral
one-third of the periosteal surface but only the lateral one-quarter
of the endosteal surface of the femur. The mean length of the osteotomy
was 12.5 centimeters (range, seven to nineteen centimeters).
The osteotomy fragment was reattached to the femur with multiple
two-millimeter-diameter cables. The upper cable was usually placed
through a drill-hole in the lesser trochanter. If the proximal-medial
aspect of the femur was deficient, the upper cable was placed distal
to the lesser trochanter to prevent fretting. Before the cables
were secured, the range of motion of the hip was checked with the
fragment repositioned to ensure that the trochanter did not abut
the pelvis and cause instability. If anterior impingement of the trochanter
was encountered during flexion and internal rotation, a portion
of the greater trochanter was removed or the osteotomy fragment
was displaced posteriorly, overlapping the proximal part of the
posterior cortex. Finally, morseled autogenous bone graft, if available
from the revision operation, was placed at the distal osteotomy
site. If there were concerns about the adequacy of the apposition
of the osteotomy site or if the existing proximal bone stock was
poor, one or two cortical strut grafts were added across the osteotomy
site beneath the cables.
Of the forty-six hips, fifteen had one cortical strut graft and
two had two strut grafts. Two to five cables (mean, 2.8 cables)
were used for fixation. The mean length of the stem bypassing the
distal portion of the osteotomy site was 142 millimeters (range,
fifty-eight to 221 millimeters) as seen on the radiographs with
correction for magnification.
Postoperative Management
Postoperative rehabilitation depended on the stability of the
hip and on the surgeon's subjective assessment of the strength of
the reattachment of the osteotomy fragment. Passive range of motion, with
the exception of adduction, was allowed within the recorded limits
of hip stability. Hip flexion and abduction strengthening exercises were
not allowed for six weeks. All patients used a walker or two axillary
crutches with protected weight-bearing for a minimum of two months.
Under optimal circumstances, the patients were allowed to bear full
weight on the involved hip at three months. A brace was used if
the strength of the osteotomy repair was tenuous or if there were concerns
about hip stability or patient compliance.
The patients were followed at six weeks, three months, six months,
and annually thereafter. Standardized anteroposterior pelvic radiographs,
Judet oblique radiographs of the acetabulum, and true anteroposterior
and Lowenstein lateral radiographs of the femur were made at every
follow-up examination. The trochanteric osteotomy site was considered
healed radiographically if callus was seen bridging the site in
both the anteroposterior and the lateral plane4.
Trochanteric migration was evaluated by measuring the perpendicular
distance from the tip of the greater trochanter to the interteardrop
line. Migration up to 2.5 millimeters was within the standard error
of measurement and was considered unimportant8.
As an indicator of remodeling, the time that the osteotomy gap disappeared
was noted. Any complications were also recorded. Statistical analysis of
the results was performed with Pearson's correlation or a t test
for independent samples.
Of the forty-five patients (forty-six hips), one was lost to
follow-up and excluded from further analysis, two died less than
twenty-four months (twelve and fifteen months) after the operation,
and the remaining forty-two (forty-three hips) were followed for
a minimum of eighteen months. The forty-three hips of the surviving
patients were followed for a mean of forty-five months (range, eighteen
to eighty-eight months). Three of these surviving patients (three
hips) were followed for between eighteen and twenty months.
At a mean of forty-four months after the operation, forty-four
of the forty-five trochanteric osteotomy sites had united with bridging
callus. Only immediate postoperative and most recent follow-up radiographs
were available for one patient (one hip), and these demonstrated
that the osteotomy gap had disappeared by forty-eight months. Serial
radiographs were available to assess the time to bridging-callus
formation and disappearance of the osteotomy gap in forty-three
hips. The mean time to union of the osteotomy site was 5.0 months (range,
six weeks to fourteen months). Seventy-four percent (thirty-two)
of the forty-three sites united within three to six months. Bone-remodeling,
evidenced by disappearance of the osteotomy gap on radiographs,
occurred in 86 percent (thirty-seven) of the forty-three hips by
twelve months (Fig. 2-A and Fig. 2-B). Twelve percent (five) of the
osteotomy gaps became radiographically imperceptible between twelve
and twenty-four months. One femur had a consistently visible gap
at the time of the forty-six-month follow-up despite the fact that
bridging callus had been visible at fourteen months. All five hips
that were followed for less than twenty-four months demonstrated
bridging callus and disappearance of the osteotomy gap by the time
of their last evaluation. None of the cables broke. Although the
osteotomy fragment in one hip migrated five millimeters superiorly
in the first six weeks postoperatively, it united uneventfully within
six months without further migration.
In the seventeen patients who had cortical strut allografting,
no change in the position or integrity of the grafts was noted.
All strut grafts had united to host bone by the twelve to fifty-nine-month
follow-up visit.
The single instance of nonunion in this series occurred in a
patient who was discharged two weeks after the revision procedure
and was subsequently followed at an outside institution. During
her initial hospitalization, a low-grade fever with persistent drainage
developed. Multiple organisms, including diphtheroids and Staphylococcus
epidermidis, grew on culture of specimens from the wound. Records
indicated clinical resolution of the infection with antibiotic management
and incision and drainage. Between six and twelve months postoperatively,
the patient had recurrent hip dislocations, and radiographs demonstrated
fifteen millimeters of proximal displacement of the trochanteric
osteotomy fragment but no clear evidence of cable breakage or fracture
at the osteotomy site. There was no additional operative intervention,
and the patient did not have persistent instability or infection.
Statistical Findings
The number of cables used for fixation was not significantly
correlated with the time to union of the osteotomy site (r = 0.24
and p = 0.12, Pearson's correlation). Preoperative bone quality,
measured as the minimum cortical thickness, and cancellous bone
quality also did not influence the time to union (r = -0.21 and
p = 0.17, Pearson's correlation, and p = 0.76, t test for independent samples).
The mean time to bridging-callus union was 4.3 months (range, six
weeks to nine months) in the hips without a strut allograft and
6.1 months (range, three to fourteen months) in the hips with a strut
allograft. The time to bridging-callus union was significantly longer
in the hips with a strut allograft than in those without a strut
allograft (p = 0.04, t test for independent samples).
Complications
A total of eleven complications (24 percent) occurred in the
forty-five hips. In addition to the one hip that had the nonunion
described earlier, four hips had dislocation; two, fracture of the
osteotomy fragment; two, trochanteric bursitis; one, sciatic nerve
injury; and one, deep infection. Three hips (7 percent) had a reoperation.
Of the five patients (11 percent) with hip dislocation, two had
no additional dislocations after the initial closed reduction and
three had recurrent dislocations. The extended osteotomy was considered
to be a contributing factor only in the hip that had the nonunion
described previously.
Of the two fractures of the osteotomy fragment, one occurred
through the superior edge of the cortical strut allograft and the
other occurred at the level of the first cable. The fractured trochanteric fragments
migrated five and nine millimeters proximally. Because the migration
was not progressive, operative intervention was not required.
In complex revision total hip arthroplasties, a trochanteric
osteotomy may be necessary to achieve adequate exposure5,10. However, a conventional trochanteric
osteotomy does not have the advantage of facilitating the removal
of distal cement or a well fixed porous-coated component. Also,
the conventional method does not minimize the risks of perforations
and fractures in a patient with a femoral deformity. To our knowledge,
Younger et al.12 were the first
to describe the operative technique for the extended slide trochanteric
osteotomy with preservation of the gluteus medius and minimus and
vastus lateralis attachments on the osteotomized bone. They reported
excellent removal of the cement and components, optimal implantation of
the revision component, and reliable healing in twenty patients.
One weakness of the current study is that five hips in five patients
(two deceased and three surviving) were not followed for a minimum
of two years. Since the final outcome of the osteotomy (union) was
clear in all five hips (which were followed for twelve to twenty
months), they were included in the analysis. We recognize that such
short-term data can provide valuable information regarding the outcome
of the osteotomy but not of the revision procedure itself.
Healing after the revision hip arthroplasties with an extended
slide trochanteric osteotomy in our series was excellent. The rate
of union of the distal osteotomy site was 98 percent. Most (86 percent) of
the osteotomy sites united within six months, and these results
compare favorably with the union rates reported after conventional
trochanteric osteotomies1,9,11.
In contrast to a conventional trochanteric osteotomy6,7, the extended slide trochanteric
osteotomy preserves the blood supply from the anterior soft tissue
and the vastus lateralis muscle while providing a large surface
area for reattachment. The effect of these factors may contribute
to the high union rate associated with the extended slide trochanteric
osteotomy.
Infection may have contributed to the single instance of nonunion
in this series. The displacement of the trochanteric fragment and
subsequent hip instability demonstrate the morbidity that this complication
can produce.
Of the remaining forty-four hips, two had a fracture of the osteotomy
fragment and a third, which did not fracture, had five millimeters
of early proximal migration of the osteotomy fragment with subsequent
union. One fracture occurred through the superior edge of the strut
allograft, and the other occurred through the first cable. Both
were located distal to the vastus ridge, and both displaced less
than one centimeter. Maintaining the attachment of the vastus lateralis
to the vastus ridge may provide some protection against the proximal
pull of the abductors. This attachment could explain why there was
only minimal proximal migration of the two fractured trochanteric
fragments.
All stems bypassed the most distal portion of the osteotomy site
by more than five centimeters (range, 5.8 to 22.1 centimeters).
The osteotomy did not predispose patients to intraoperative or late postoperative
fracture. There was no change in the position of any of the femoral
components after a mean duration of follow-up of forty-four months, but
we recognize that the follow-up of these hips is not adequate for
us to comment on the long-term stability of the components.
Although the most common reason for the osteotomy was femoral
revision, we used this technique in two hips that needed an isolated
acetabular revision because we were trying to avoid complications
associated with nonunion and migration. Despite concern that an
extended osteotomy might be detrimental to an existing stable femoral
component, the femoral components (a cemented long stem and a proximally
porous-coated stem with good osseous ingrowth) in these two hips
remained stable at thirty-six and forty-two months.
In conclusion, the extended slide trochanteric osteotomy allows
adequate exposure for complex revision, facilitates the removal
of distal cement or a well fixed porous-coated stem, corrects proximal deformities
of the femur, and makes solid reimplantation possible. Furthermore,
it provides a large surface area for reattachment, preserves blood
supply, and has a low nonunion rate. Additionally, compared with
standard osteotomy technique, it affords greater potential for adjustment
of abductor-muscle tension by proximal or distal displacement without
sacrificing stability of the reattachment. We recommend that this
technique be routinely considered in complicated revision total hip
arthroplasties.
Amstutz, H. C., and Maki, S.: Complications of trochanteric osteotomy in total hip replacement. J. Bone and Joint Surg.,60-A: 214-216, March 1978.60-A214
1978
Cabanela, M. E.: Revision hip arthroplasty:
surgical approaches. In Hip Surgery. Materials and Developments,
pp. 173-175. Edited by L. Sedel and M. E. Cabanela. London, Martin
Dunitz, 1998.
Frankel, A.; Booth, R. E., Jr.; Balderston, R. A.; Cohn, J.; and Rothman, R. H.: Complications of trochanteric osteotomy. Long-term implications. Clin. Orthop.,288: 209-213, 1993.288209
1993
[PubMed]
Glassman, A. H.; Engh, C. A.; and Bobyn, J. D.: Proximal femoral osteotomy as an adjunct in cementless
revision total hip arthroplasty. J. Arthroplasty,2: 47-63, 1987.247
1987
[PubMed]
Jensen, N. F., and Harris, W. H.: A system for trochanteric osteotomy and reattachment for
total hip arthroplasty with a ninety-nine percent union rate. Clin. Orthop.,208: 174-181, 1986.208174
1986
[PubMed]
Naito, M.; Ogata, K.; and Emoto, G.: The blood supply to the greater trochanter. Clin. Orthop.,323: 294-297, 1996.323294
1996
[PubMed]
Najima, H.; Gagey, O.; Cottias, P.; and Huten, D.: Blood supply of the greater trochanter after trochanterotomy. Clin. Orthop.,349: 235-241, 1998.349235
1998
[PubMed]
Nercessian, O. A.; Newton, P. M.; Joshi, R. P.; Sheikh, B.; and Eftekhar, N. S.: Trochanteric osteotomy and wire fixation: a comparison
of 2 techniques. Clin. Orthop.,333: 208-216, 1996.333208
1996
[PubMed]
Ritter, M. A.; Gioe, T. J.; and Stringer, E. A.: Functional significance of nonunion of the greater trochanter. Clin. Orthop.,159: 177-182, 1981.159177
1981
[PubMed]
Schutzer, S. F., and Harris, W. H.: Trochanteric osteotomy for revision total hip arthroplasty.
97% union rate using a comprehensive approach. 1988.
1988
Thompson, R. C., Jr., and Culver, J. E.: The role of trochanteric osteotomy in total hip replacement. Clin. Orthop.,106: 102-106, 1975.106102
1975
[PubMed]
Younger, T. I.; Bradford, M. S.; Magnus, R. E.; and Paprosky, W. G.: Extended proximal femoral osteotomy. A new technique for
femoral revision arthroplasty. J. Arthroplasty,10: 329-338, 1995.10329
1995
[PubMed]