Demographics
After obtaining institutional review board approval, we identified all
patients undergoing total hip arthroplasty for osteonecrosis of the femoral
head between January 2002 through January 2004. During the years of this
study, 412 total hip arthroplasties in 325 patients were performed by the
senior author (M.A.M.), of which fifty-two arthroplasties in forty-one
patients were performed for an underlying diagnosis of osteonecrosis of the
femoral head. There were thirty men and eleven women with a mean age of
thirty-eight years (range, fifteen to fifty-three years) in the osteonecrosis
group. These fifty-two osteonecrotic hips were matched to hips from a cohort
of osteoarthritic hips that received a total hip arthroplasty from the same
surgeon, with the same prosthesis, during the same time-period. The matching
was done for closest age at surgery, gender, surgeon, arthroplasty components
used, method of fixation, surgical approach, and the method of anesthesia.
There were thirty men and ten women with a mean age of forty-two years (range,
eighteen to sixty-one years) in the osteoarthritis group. A summary of the
demographic data for both groups can be found in
Table I.
For the patients with osteonecrosis of the femoral head, the associated
conditions and risk factors were corticosteroid use in nine (17%), human
immunodeficiency virus infection in eight (15%), alcohol abuse in eight (15%),
systemic lupus erythematosus in seven (13%), none (idiopathic osteonecrosis of
the femoral head) in seven (13%), tobacco abuse in six (12%), sickle cell
disease in four (8%), and hepatitis C in three hips (6%). Of the remaining
hips, one each had an underlying diagnosis of diabetes, hyperlipidemia,
leukemia, inflammatory bowel disease, or chronic obstructive pulmonary
disease. Some patients may have had more than one associated risk factor.
Surgical Data
All surgeries were performed by the senior author (M.A.M.) with use of an
anterolateral
approach19. The
components utilized in all patients were a collarless, tapered, proximally
hydroxyapatite-coated femoral stem (Accolade, Stryker Orthopaedics, Mahwah,
New Jersey) and an uncemented hemispherical acetabular component (Stryker
Orthopaedics). The type of bearing surface utilized was metal-on-highly
cross-linked polyethylene (Crossfire; Stryker Orthopaedics) in eighty-one hips
and ceramic-on-ceramic (Trident, Stryker Orthopaedics) in twenty-three
hips.
Postoperative Protocol
Postoperative care was standard and was the same for patients in both
groups. A second-generation antibiotic (Ancef [cefazolin], GlaxoSmithKline,
Research Triangle Park, North Carolina) was administered for twenty-four
hours. Thromboembolic prophylaxis included intraoperative heparin
administration, postoperative aspirin, and compression stockings. Patients
were encouraged and assisted to commence walking on the first postoperative
day. Patients were limited to 50% weight-bearing for the first six weeks and
then were permitted to advance to full weight-bearing as tolerated.
Hip-strengthening exercises were begun six weeks after surgery.
Clinical Evaluation
Clinical outcome was measured with use of the Harris hip
score20. The
outcome was categorized as excellent or good (hip score =80 points, no use
of a walking aid, and a nonpainful hip), fair (hip score of 70 to 79 points,
occasional use of a walking aid and/or mild hip pain), or poor (latter
category not satisfied).
Radiographic Analysis
Serial anteroposterior and lateral radiographs made prior to and after the
hip arthroplasty were evaluated in detail to assess osseointegration,
component positioning, gross wear (eccentric positioning of the femoral head),
and loosening. The femur was divided into seven
zones21 and the
acetabulum into three
zones22 to evaluate
the location of lucent lines. Definitely loose components were defined as
those that demonstrated a complete lucent line on any radiograph, femoral
subsidence of 2 mm or more, or acetabular component
migration23.
Possibly loose components were defined as those with a >50% but <100%
lucent line on any radiograph or those with a progressive radiolucent line.
Radiographs were also evaluated for the presence of heterotopic ossification,
which was classified according to the system of Brooker et
al.24. In addition,
the femoral offset, femoral height, cup offset, and cup inclination angle were
recorded during each radiographic evaluation.
Follow-up
Clinical and radiographic data on all patients were collected
prospectively. Patients were contacted on a regular basis, which in most cases
included examination at three months, one year, two years, and every year
thereafter. All patients were followed for a minimum of two years, until
failure of the pros-thesis or until death. Follow-up averaged 37 months
(range, 24 to 50 months) both clinically and radiographically. No patients
were lost to follow-up. There were no deaths during the interval under
study.
Statistical 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). Student t tests and the chi-square test were utilized to compare
variables between the two study groups. Kaplan-Meier survivorship analysis was
used to evaluate time from surgery to revision, and the log-rank test was used
to compare survival
probabilities25.
Survival rates were then calculated, with revision of the femoral stem for any
cause and revision for aseptic loosening as the end points. All p values
<0.05 were considered significant.
Clinical Outcome
There was a significant improvement in function as measured with the Harris
hip score. The mean preoperative Harris hip score of 30 points (range, 5 to 57
points) for osteonecrosis and 35 points (range, 4 to 71 points) for
osteoarthritis improved to a mean of 92 points (range, 69 to 100 points) and
93 points (range, 66 to 100 points) for the two groups, respectively (p <
0.178 and p = 0.269, respectively). With the numbers studied, there was no
difference in the improvement between the two groups (p = 0.308). At the time
of the latest follow-up, the outcome was categorized as good to excellent in
94% (forty-nine of fifty-two) of the osteonecrotic hips and 96% (fifty of
fifty-two) of the osteoarthritic hips. The outcome was fair in two and poor in
one of the osteonecrotic hips and fair in one and poor in another
osteoarthritic hip. Of the five hips in the fair or poor category, three
patients (three hips) had adverse effects on hip scores because of symptomatic
degenerative joint disease involving other joints, and two patients (two hips)
had systemic illness that impacted both walking ability and functional
scores.
Implant Survival
With revision for any reason as the end point, the Kaplan-Meier survival
probability for the two groups was similar (p = 0.575)
(Fig. 1). At fifty months
postoperatively, the survival rate was 96.1% (95% confidence interval, 93.4%
to 98.8%) for the osteonecrotic hips and 98% (95% confidence interval, 96.1%
to 99.9%) for the osteoarthritic hips.
Radiographic Findings
Evaluation of radiographs at the time of the latest follow-up revealed
nonprogressive radiolucent lines about the femur in one hip in the group with
osteonecrosis and no radiolucent lines in the group with osteoarthritis. In no
hip was the femoral component outside the optimal position of 2° of varus
or valgus. The position of the acetabular component was optimal (abduction
angle between 35° and 50°) in forty-five hips (87%) in the group with
osteonecrosis and in forty-three hips (83%) in the group with osteoarthritis.
The cups in the remaining six osteonecrotic hips and in eight of the
osteoarthritic hips were <10° from the conventional optimal position.
There was no evidence of gross wear of the acetabular liner in any of the hips
at the time of the latest follow-up.
Heterotopic ossification was noted in five of the osteonecrotic hips and
was categorized as Brooker class I in four hips and class III in one hip.
Heterotopic ossification also occurred in five of the osteoarthritic hips and
was classified as Brooker class I in two hips, class-II in two hips, and
class-III in one hip. As of the time of writing, none of the patients had
undergone surgical excision of heterotopic bone.
Complications
Three complications, consisting of dislocation, deep infection, and aseptic
loosening of the femoral stem, occurred in the group with osteonecrosis. In
contrast, there was only one complication, consisting of a dislocation, in the
group with osteoarthritis.
In the group with osteonecrosis, a forty-six-year-old man with diabetes
mellitus was noted to have subsidence of the femoral component and underwent
revision surgery with conversion to a bigger femoral stem four months after
the index procedure. This patient had no associated risk factors, such as
corticosteroid usage, and underwent an uncomplicated femoral component
exchange. He was doing well at the time of the final follow-up at twenty
months and had a Harris hip score of 90 points. A thirty-one-year-old man
experienced septic loosening of the femoral component at three months
postoperatively. Following the index surgery and prior to the development of
the infection, the patient had undergone a dental procedure and a colonoscopy
with polypectomy. Subsequent to those procedures, a low-grade fever developed
and a purulent mass formed on the anterior aspect of the thigh. The patient
underwent surgical irrigation and débridement of the mass and the hip
joint, removal of the prosthesis, and placement of an antibiotic cement spacer
as the first of a two-stage revision procedure. Twelve weeks after the first
stage, the antibiotic cement spacer was removed and a second-stage revision
with reimplantation of a total hip pros-thesis was performed. At the time of
follow-up at twenty-two months, this patient had a Harris hip score of 80
points. A forty-nine-year-old man had a hip dislocation that was managed
successfully with closed reduction and physical therapy. There had been no
recurrence of the dislocation by thirty-one months, and the patient had a
Harris hip score of 98 points.
In the group with osteoarthritis, a thirty-seven-year-old man presented to
the outpatient clinic with radiographic evidence of dislocation of the right
hip. He underwent an unsuccessful attempt at a closed reduction. The patient
then underwent open reduction of the dislocated hip, excision of heterotopic
ossification, and placement of a locking polyethylene liner. At the time of
follow-up at twenty-six months, the patient had excellent hip stability and a
Harris hip score of 86 points.
There are conflicting reports regarding the outcome of total hip
arthroplasty for patients with osteonecrosis of the femoral head (see
Appendix). Standard total hip arthroplasty historically has been shown to have
less than optimal results in patients with osteonecrosis as compared with
patients with
osteoarthritis3-8.
Other studies have refuted these findings and have reported that patients with
osteonecrosis and patients with osteoarthritis had an equally successful
outcome after total hip
arthroplasty14,17,26-28.
Many reasons may explain the observed discrepancy in the literature. Most
studies have not made an effort to separate patients on the basis of
associated risk factors. It is possible that a total hip prosthesis that has
been placed as treatment for osteonecrosis of the femoral head resulting from
immunosuppressive therapy for transplantation or systemic lupus erythematosus
may behave differently than a total hip prosthesis that has been placed as
treatment for osteonecrosis of the femoral head that is associated with drugs,
trauma, or
alcohol29. Thus,
combining the results of various patients who have osteonecrosis of the
femoral head may introduce a bias. A second important factor is that many
previous studies reported on the outcome of total hip arthroplasty in which
first-generation cementing techniques or first-generation uncemented
prosthetic designs were
used3,5,18,30,
making it difficult to compare those studies with studies in which
contemporary prosthetic designs and modern cementing techniques were used.
Patient age may be an important factor. Patients with osteonecrosis of the
femoral head tend to be very young, which means that the outcome of a cemented
or hybrid total hip arthroplasty can be expected to be
suboptimal16,18,31.
The inferior results for patients with osteonecrosis of the femoral head
compared with patients with osteoarthritis may in some instances relate to the
younger age of the patients affected by osteonecrosis of the femoral head
17. Finally, many
of the studies that reported inferior outcomes are from an era when medical
treatment, anesthetic management, and surgical care were not optimal. The last
decade has witnessed an immense improvement in prosthetic designs, bearing
surfaces, and general delivery of surgical care that is likely to lead to
better outcomes for all patients, including younger patients with
osteonecrosis of the femoral head.
The results of the present study, although at short term, are encouraging.
We have demonstrated that the results of total hip arthroplasty with a
hydroxyapatite-coated prosthesis in young patients with osteonecrosis of the
femoral head were similar to the results seen in a matched group of patients
with osteoarthritis. However, we did observe a slightly higher rate of
complications and reoperations in the group with osteonecrosis, which may be
an incidental observation. Longer-term results are needed to assess the
longevity of total hip arthroplasty. It is important to point out, however,
that the survivorship of 96% for the hips in our extremely young patients is
better than has been reported previously. For example, Saito and
coworkers3 analyzed
the outcome of twenty-nine osteonecrotic hips that had been treated with a
cemented total hip prosthesis. At a mean follow-up of eighty-four months, the
clinical success rate was only 52%. Cornell and
coworkers8 found
similar poor results in a study of twenty-four patients (twenty-eight hips) at
a mean duration of follow-up of ninety-one months (range, sixty to 120
months). At the time of final follow-up, eleven hips required revision total
hip arthroplasty, indicating a 39% failure rate.
Some recent studies have been associated with better results. In a
prospective study of clinical and radiographic outcome after total hip
arthroplasty in 200 hips with osteonecrosis of the femoral head, Kim and
colleagues15
compared the use of so-called "third-generation" cementing
techniques (n = 100) with the use of second-generation cementless techniques
(n = 100). At the time of the final follow-up at 122 months, the survival rate
was 98% for both groups. Although the outcome associated with both the
cementless and the cemented technique was excellent, it must be understood
that this study contained highly selected patient cohorts with no obvious
adverse associated risk factors. In addition, the mean age of the patients was
forty-seven years (range, twenty-six to fifty-five years), which is high for a
patient population with osteonecrosis. When evaluating studies that report the
outcome of total hip arthroplasty in patients with osteonecrosis, it is
important to consider any variability between groups with regard to prosthetic
design, fixation technique, associated risk factors, patient age, and patient
activity level. An overview of results of total hip arthroplasty in patients
with osteonecrosis who have an increased risk of implant failure is shown in a
table in the Appendix.
In our current study, we sought to minimize the influence of confounding
variables by selecting a homogeneous cohort of patients who received the same
type of uncemented total hip prosthesis and who were operated on by a single
surgeon. This study was also limited to young adults (mean age, thirty-eight
years). A matched group of patients with osteoarthritis was selected for
comparison. The study has some limitations. First, the follow-up in the cohort
is relatively short and longer-term results are awaited before firm
conclusions can be made. Second, the study contains a relatively small number
of patients and, therefore, the higher complication rate in the group with
osteonecrosis may not be related to the underlying disease. Because of the
relatively small number of patients, a difference of one patient in both
groups would have had a significant impact on these results. Despite the
aforementioned limitations, the findings of this study are important because
they suggest that there can be an excellent early survivorship for
contemporary uncemented total hip prostheses for young adults with
osteonecrosis of the femoral head.
Tables showing previously published articles on this subject 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 CDROM (call our subscription department, at 781-449-9780, to
order the CD-ROM). ?