Study Design and Patient Cohorts
Between November 1, 2000 and October 31, 2003, forty-three metal-on-metal
total resurfacing procedures were performed in thirty-seven patients with
osteonecrosis of the femoral head at our institution. One patient (one hip)
who had been doing well clinically and had a Harris hip score of 86 points was
lost to follow-up at six months after surgery. All patients were followed both
clinically and radiographically for a minimum of two years, with a mean
duration of follow-up of forty-one months (range, twenty-four months to
sixty-one months). There were twenty-five men and eleven women who had a mean
age of forty years (range, eighteen to sixty-four years). The mean body mass
index was 27.6 (range, 18.8 to 36.4). The hospital records for all patients
were reviewed, including data from preoperative studies, operative reports,
and postoperative office visits. In October 2000, institutional review board
(IRB) consent was obtained to prospectively follow all total hip resurfacings
as part of the United States Food and Drug Administration multicenter
investigational device exemption study of the Conserve Plus prosthesis (Wright
Medical Technology, Arlington, Tennessee). The IRB approval for this study was
then renewed annually.
The most common associated factors for osteonecrosis of the femoral head in
our study population were trauma (eleven hips) and corticosteroid use (eleven
hips), which together affected 52%. Other conditions associated with
osteonecrosis were neoplasm (four hips, 10%), systemic lupus erythematosus
(three hips, 7%), and Legg-Calvé-Perthes disease (three hips, 7%). The
cause of osteonecrosis was believed to be idiopathic in three hips, associated
with a history of chronic obstructive pulmonary disease in two hips and a
history of chronic renal failure in two other hips, and associated with sickle
cell disease in one hip, multiple sclerosis in one hip, and liver disease in
one hip. Twelve of the patients (29% of the hips) had had a prior hip
operation, which included four core decompressions, one pinning, two free
vascularized fibular grafts, two fractures, an osteotomy, an arthroscopic hip
exploration, and removal of an osteoid osteoma.
The forty-two hips with osteonecrosis were matched with the same number of
hips in a cohort of patients who underwent total hip resurfacing for
osteoarthritis during the same time frame. Matching was done for surgeon,
surgical approach, prosthesis, age within five years, and gender. There were
twenty-eight men and thirteen women who had a mean age of forty-four years
(range, twenty-two to sixty-four years). The mean body mass index was 29.2
(range, 19.1 to 42.9).
Surgical Technique and Postoperative Management
All surgeries were performed by one of us (M.A.M.) through an anterolateral
approach10.
Standard instruments for the Conserve Plus hybrid metal-on-metal surface
replacement prosthesis were used for all patients in this study. The
acetabular component is nearly hemispherical with a sintered porous coating
for cementless press-fit fixation. The femoral component has a short
metaphyseal stem to facilitate accurate component alignment.
All procedures were performed through an anterolateral
approach11. After
placing the patient in the lateral decubitus position, an incision was carried
through Scarpa's fascia to the fascia lata, which then was divided to expose
the body of the gluteus medius muscle. The anterior one-third of the gluteus
medius was divided and reflected from the greater trochanter along with the
underlying gluteus minimus. The capsule was fully excised and the hip was
dislocated. The acetabulum was reamed to accommodate the acetabular component,
and the cup was implanted in a press-fit manner. The pin-shaft angle was
measured with a goniometer to achieve a target of 140° (±5°).
All femoral components were cemented.
Thromboembolic prophylaxis with subcutaneous heparin or enoxaparin was
started on the day of the operation and was continued for fourteen consecutive
days. On the first postoperative day, all patients were allowed to stand and
take a few steps, and physical therapy, including strengthening exercises, was
initiated. Patients were restricted to 20% weight-bearing (with the aid of two
crutches or a walker) until the sixth postoperative week, at which time they
were allowed to advance to 50% weight-bearing (cane or crutch in opposite
hand). At twelve weeks postoperatively, all patients were advanced to full
weight-bearing as tolerated. Patients were encouraged to continue
hip-strengthening exercises three times per week.
Clinical Evaluation
Patients returned for postoperative evaluation at six weeks, six months,
one year, and annually thereafter. Pain, function, and deformity were
evaluated with use of a Harris hip-rating
system12 and both
the mental component summary and the physical summary of the Short Form-12
(SF-12)
questionnaire13
prior to surgery and at each postoperative clinical visit. A Harris hip score
of 90 points or more was defined as an excellent outcome; 80 to 89 points, a
good outcome; 70 to 79 points, a fair outcome; and fewer than 70 points, a
poor outcome. Excellent and good results were classified as successful
outcomes, whereas fair or poor results were classified as unsuccessful
outcomes.
Radiographic Analysis
Standard anteroposterior and cross-table lateral radiographs were obtained
at each clinical visit for all patients. Acetabular radiolucencies were
evaluated and classified according to the zones of DeLee and
Charnley14
(Fig. 1). To assess
radiolucencies at the bone interface of the femoral component, six zones
around the short stem were delineated (Fig.
1). All radiolucent lines of >1 mm were recorded. Heterotopic
ossification was graded in accordance with the standard Brooker
classification15.
In addition, the stem-shaft angle, the cup-inclination angle, and the femoral
neck-shaft angle were recorded during each radiographic evaluation. Because of
the possible introduction of error in the radiographic measurements that were
made by two of the authors (G.A.M. and T.M.S.), an assessment of interobserver
and intraobserver error in radiographic evaluation was made before initiating
the study. The intraobserver reliability of the two physicians was excellent,
with agreement in all of the cases. The inter-observer agreement was an exact
match in 90% of the cases. To avoid the problem of intraobserver and
interobserver variability in assessing the various radiographic parameters,
all radiographs were evaluated independently by two of the authors who were
not involved with the surgery (G.A.M. and T.M.S.). If there was a
disagreement, a third author interpreted the radiographs until a unanimous
decision could be made.
Data Analysis
The data were subjected to averaging and analysis with use of Statistical
Package for the Social Sciences (SPSS) software (version 13.0; SPSS, Chicago,
Illinois). Paired Student t tests were utilized to compare variables between
the two study groups. Kaplan-Meier survivorship curves were used to evaluate
time from surgery to revision, and the log-rank test was used to compare
survival
probabilities16.
The end point of survival was defined as either revision (removal or exchange
of one or more components) for any reason, or as revision for aseptic
loosening. All p values <0.05 were considered significant. In addition, we
correlated various demographic factors, such as body mass index, age, gender,
and previous surgery, with the functional outcome of metal-on-metal total hip
resurfacing.
Clinical Results
At the time of the final follow-up evaluation, both patient populations
exhibited similar clinical success rates with thirty-nine of forty-two (93%)
osteonecrotic hips and forty of forty-two (98%) osteoarthritic hips having a
Harris hip score of =80 points. The mean preoperative Harris hip score of
52 points (range, 42 to 60 points) for osteonecrosis was less than the mean
57-point score (range, 31 to 73 points) for osteoarthritis (p = 0.008). At the
time of the final postoperative follow-up, the scores were similar (p = 0.941)
with a mean of 91 points (range, 57 to 100 points) for the osteonecrosis group
and 91 points (range, 49 to 96 points) for the osteoarthritis group. On the
basis of the Harris hip scores, the outcomes in the osteonecrosis group were
considered excellent in twenty-eight, good in eleven, fair in one, and poor in
two hips; in the osteoarthritis group, the outcomes were considered excellent
in twenty-nine, good in twelve, and poor in one hip.
The mean SF-12 mental-component summary scores were also similar for the
two groups at the time of final follow-up, with both groups reporting a mean
score of 56 points (p = 0.774). However, the mean SF-12 physical component
summary score of the osteoarthritis group was superior (p = 0.008) with a mean
of 53 points (range, 33 to 62 points) compared with the osteonecrosis mean
score of 49 points (range, 31 to 60 points). The clinical results are
summarized in Table I.
Kaplan-Meier survivorship analysis indicated that the overall final
survival rate was similar (p = 0.977) for both groups
(Fig. 2). At the time of the
final follow-up, the survival rate was 94.5% (95% confidence interval, 90.6%
to 98.3%) for the patients with osteonecrosis and 95.2% (95% confidence
interval, 92.0% to 98.5%) for the patients with osteoarthritis.
The correlation of various demographic factors with the outcome of total
hip resurfacing revealed a significant difference for body mass index and
postoperative Harris hip scores in the osteoarthritis group. Obese patients
(body mass index =30) had a lower Harris hip score compared with nonobese
patients (87 points vs 93 points; p = 0.005). Obesity appeared to have no
significant effect on the outcomes in the osteonecrosis group. Of note, two of
the patients who required conversion to a standard total hip arthroplasty had
a body mass index of >30. With the numbers studied, the comparison with
regard to age, gender, and previous surgery showed no significant difference
between groups.
For the osteonecrosis group, the range of motion improved from an average
of 88° (range, 45° to 100°) of flexion, 39° (range, 5° to
75°) of abduction-adduction measured in extension, and 36° (range,
0° to 95°) in rotation arc measured in extension to an average of
125° (range, 50° to 135°), 57° (range, 30° to 95°),
and 49° (range, 15° to 80°), respectively. The osteoarthritis
group showed similar results, improving from an average of 84° (range,
0° to 120°) of flexion, 31° (range, 0° to 60°) of
abduction-adduction measured in extension, and 33° (range, 0° to
75°) in rotation arc measured in extension to an average of 122°
(range, 75° to 130°), 61° (range, 25° to 90°), and 51°
(range, 20° to 105°), respectively. A comparison of the measurements
in the two groups did not reach significance.
Two patients in each group had complications that resulted in the need for
revision to standard total hip arthroplasty. In the osteonecrosis group, one
patient fractured the femoral neck at three months after surgery while
externally rotating the leg and putting increased weight on the operated hip
while getting out of a car. This patient had a body mass index of 32. The
patient underwent an uncomplicated standard total hip arthroplasty and was
doing well at twenty-four months with a Harris hip score of 94 points. Another
patient, who had a core decompression of the femoral head prior to
metal-on-metal resurfacing, experienced aseptic loosening of the femoral
component at thirty months postoperatively and underwent a successful revision
to a standard prosthetic stem with a large modular head. That patient had a
Harris hip score of 95 points at the time of the six-month follow-up.
In the osteoarthritis group, both complications were fractures of the
femoral neck. One patient experienced a femoral neck fracture while attempting
to lift and move a vehicle eighteen months after surgery. The patient had a
body mass index of 32.7. To avoid a revision of the well-fixed, large
acetabular cup, the resurfacing arthroplasty was converted to a standard,
proximally coated femoral stem with a large modular head. At the time of the
final follow-up (thirty-six months after the revision arthroplasty), the
Harris hip score was 90 points. The second patient who required revision to a
total hip arthroplasty fractured the femoral neck while playing soccer twenty
months after a metal-on-metal resurfacing procedure. The patient also
under-went conversion to a total hip arthroplasty with use of a large femoral
head, thus permitting the acetabular cup to be left in place. The Harris hip
score was 96 points at thirty-nine months after the revision.
Radiographic Results
One patient in each group had Brooker-class-IV heterotopic ossification.
The patient in the osteonecrosis group had bilateral heterotopic ossification
and was also diagnosed with multifocal osteonecrosis and severe systemic lupus
erythematosus postoperatively. Within six months after surgery, the patient
showed a decreased arc of flexion of 45°, which was less than the average
of 106°, but continued to have functional arcs of external and internal
rotation and abduction-adduction. After undergoing fluoroscopic examination of
the hips, the patient underwent surgical excision of heterotopic bone
formation on the left hip at fifteen months after surgery. At the time of the
most recent follow-up, the patient had left groin pain with prolonged standing
and the Harris hip score continued to be low (63 points). The patient with
osteoarthritis also underwent surgical removal of a heterotopic ossification.
There was a presumed infection in the right hip; cultures demonstrated growth
of a coagulase-positive Staphylococcus species. Following débridement
of the right hip joint and a regimen of intravenous antibiotics, the patient
demonstrated improved flexion (from 75° to 106°), abduction (from
10° to 45°), and adduction (from 5° to 15°). Five other
patients with osteonecrosis had heterotopic bone formation (one
Brooker-class-III, one Brooker-class-II, and three Brooker-class-I). In the
osteoarthritis group, two patients had Brooker-class-I and one patient had
Brooker-class-III heterotopic ossification.
Four patients with osteonecrosis and two with osteoarthritis demonstrated
nonprogressive radiolucencies. Of the patients with osteonecrosis, one had
radiolucencies around the stem in two different zones and three patients had
radiolucencies around the acetabulum in one or more zones. Two patients with
osteoarthritis had acetabular radiolucencies in one or more zones, and one
also had cup migration in each of the acetabular zones. In all cases of
radiolucencies, the lines were small (3 mm or less) and considered to be
nonprogressive. These data are found in
Table II.
The treatment of osteonecrosis of the femoral head is challenging and
controversial, and emerging new technologies such as modern metal-on-metal
bearings are presently proposed for total hip arthroplasty. Total hip
resurfacing has generally been advocated for younger, more active patients who
are more likely to outlive the life span of a standard total hip arthroplasty
and who will likely require additional surgical treatment. The results of this
study demonstrate that, at short-term follow-up (at a mean of forty-one
months), the use of metal-on-metal total hip resurfacing achieves similar
results in patients with osteonecrosis of the femoral head as in a matched
group of patients with osteoarthritis.
Historically, standard total hip arthroplasty has been associated with less
than optimal results in patients with osteonecrosis as compared with young
patients who were being managed for other disorders. Saito et
al.17 analyzed
twenty-nine osteonecrotic hips that had been treated with a cemented total hip
prosthesis. At a mean follow-up of eight-four months, the clinical success
rate was only 52%. Similar results of cementless standard total hip
arthroplasty have been reported in studies by Dorr et
al.7 (45% failure
rate), Cornell et
al.18 (39% failure
rate), and Chandler et
al.19 (57% failure
rate). Kantor et
al.20 reported on
the use of second-generation cementing techniques on twenty-eight total hip
replacements. At a mean follow-up of ninety-two months, the survival rate in
their patient cohort was 86%. Recently, with modern designs and surgical
techniques, the results appear to be better than previously reported. Garino
and Steinberg21
reviewed the results of 123 cemented and hybrid total hip replacements in
patients with osteonecrosis of the femoral head. They reported a 96% survival
rate with a mean duration of follow-up of fifty-four months (range,
twenty-four to 120 months). Recently, Kim and
colleagues22
prospectively studied the clinical and radiographic outcome of hips after
total hip arthroplasty with so-called third-generation cementing and
second-generation cementless techniques in 100 hips with osteonecrosis of the
femoral head. At the time of the final follow-up of 122 months, there was a
survival rate of 98% for both groups. Nevertheless, patients in certain
subgroups, such as osteonecrosis secondary to systemic lupus erythematosus,
sickle cell disease, or renal transplantation, are still at increased risk of
implant failure. Acurio and
Friedman23 reported
on twenty-five total hip arthroplasties in twenty-five patients with sickle
cell disease and osteonecrosis. At a mean follow-up of 103 months (range,
twenty-four to 216 months), 40% of the arthroplasties had been revised and
nine other hips (36%) were loose according to the radiographic criteria.
Brinker and
colleagues24
reviewed patients who were younger than thirty-five years of age at the time
of the total hip arthroplasty and found a high failure rate. They found that
patients with systemic lupus erythematosus or an organ transplant had worse
results than did those with idiopathic osteonecrosis of the femoral head.
The previous high failure rate of standard total hip arthroplasty in
patients with osteonecrosis of the femoral head makes the search for other
treatment alternatives worthwhile. Limited femoral resurfacing has been
proposed as a treatment alternative. The potential advantages of limited
femoral resurfacing over total hip arthroplasty are removal of only the
damaged cartilage, preservation of bone stock, lower dislocation rates, and
easy conversion to hip arthroplasty if necessary. Mont et
al.25 compared the
outcome of hemiresurfacing to that of conventional total hip replacement for
hips with late-stage disease. At a mean follow-up time of seven years for the
hemiresurfacing group and eight years for the total hip arthroplasty group,
they found that a higher percentage of patients who had hemiresurfacing were
participating in sports (60% compared with 27%). However, more of the patients
who had resurfacing had groin pain (20% compared with 6%). The overall
survivorship was similar: 90% for the resurfacing group compared with 93% for
the total hip arthroplasty group. Beaulé and associates reported on a
series of thirty-seven hips followed for a mean of 6.5 years (range, two to
eighteen years) with conversion to total hip arthroplasty as the end point.
The overall survival of hemiresurfacing in this study was 79% at five years,
59% at ten years, and 45% at fifteen
years26.
Recently, there have been a few reports emphasizing the unpredictable
results obtained with these devices. Adili and
Trousdale27
reviewed the clinical and radiographic results of twenty-nine consecutive
femoral head resurfacing procedures in twenty-eight patients. The overall
survivorship was 75.9% at three years. At the time of final follow-up, eight
hips (27.6%) had undergone conversion to a total hip arthroplasty. Cuckler et
al.28 studied
fifty-nine patients that had undergone follow-up for a mean of 4.5 years.
Femoral head resurfacing procedures in sixteen patients (27%) were considered
to have failed because of conversion to total hip arthroplasty or because of
the presence of considerable groin pain that required medication. In summary,
limited femoral resurfacing arthroplasty may be a viable option for certain
patients with late-stage osteonecrosis and minimal acetabular involvement. It
is not a permanent solution, however, because failure eventually occurs as a
result of acetabular cartilage erosion. Resurfacing of both sides of the
joint, as in metal-on-metal resurfacing, may offer a more permanent solution
than is offered by a limited resurfacing of only the femoral side.
The metal-on-metal design has made great strides forward since its
introduction in the 1960s. Grecula et
al.29 compared the
outcome of patients who were younger than fifty years of age and had
osteonecrosis of the femoral head that had been treated with either
hemiresurfacing, cementless total hip resurfacing, or total hip articular
replacement by internal eccentric shells (THARIES) with the outcome of
patients who received a standard cemented total hip prosthesis. Initially,
there were similar degrees of clinical improvement reported for the four
treatment modalities. However, the survivorship rates at ninety-six months
were 70% for the hemiresurfacing group, 15% for the cementless total hip
resurfacing group, 53% for the THARIES group, and 80% for the standard
cemented total hip replacement group. More recently, there have been several
studies showing promising short-term results with advances in metal-on-metal
bearings. Beaulé et
al.30 compared the
outcomes of fifty-six hips treated with a metal-on-metal resurfacing
arthroplasty with the outcomes of twenty-eight hips treated with a
hemiresurfacing arthroplasty. At the time of follow-up (fifty-five months for
the metal-on-metal surface arthroplasty group and fifty-three months for the
hemiresurfacing group), the University of California Los Angeles hip scores
and the Short Form-12 scores for the physical component were significantly
better in the metal-on-metal surface arthroplasty group than in the
hemiresurfacing group (p < 0.05). Amstutz et
al.31 analyzed
their experience of 400 metal-on-metal total hip arthroplasties done in 355
patients. This large series included thirty-six hips with osteonecrosis of the
femoral head. At a mean follow-up of forty-two months, the overall
survivorship was 94%. Mohamad et
al.32 reported on
their early experience with metal-on-metal total hip arthroplasty in twenty
hips. In their study, osteonecrosis of the femoral head was the underlying
diagnosis in 63% of the patients who underwent this procedure. At a mean
follow-up of only eighteen months, sixteen of nineteen patients (84%) had good
or excellent hip scores.
There have also been concerns that femoral head resurfacing can predispose
to osteonecrosis of the remaining femoral head and neck. However, there are
three reasons why we do not believe that this is a problem. Campbell et al.
reported on isolated hemiresurfacings of the femoral
head33. They
analyzed twenty-five resurfaced femoral heads histologically up to twelve
years postoperatively and found that osteonecrosis was not induced by the
procedure. In addition, after a twenty-year experience with limited
resurfacing of only the femoral head for osteonecrosis, femoral failures did
not occur. Failures only occurred as a result of cartilage erosions of the
acetabulum. Additionally, we have utilized an anterolateral approach, which
may reduce damage to the extraosseous vessels to avoid further ischemic events
to the resurfaced femoral head and neck, which could lead to further
osteonecrosis and component failure.
We are pleased with the results of the current study and the short-term
success of 93% excellent and good results in this challenging young patient
population. The results were limited by the relatively small patient
population and the limited duration of follow-up. While we await long-term
results to see if these early excellent results are maintained, on the basis
of these initial findings, we recommend metal-on-metal resurfacing as a viable
treatment option for patients with advanced stages of osteonecrosis.
?