Abstract
Background: Recent studies have indicated that the rate of
successful long-term outcome after primary total hip arthroplasty for patients
with osteonecrosis may be improved with the use of later-generation
porous-coated prostheses (biologic ingrowth fixation) and cement techniques
(cement fixation). Less is known about the long-term outcome after revision
arthroplasty in the same patient population. The purpose of this study was to
characterize the clinical and radiographic outcomes of revision total hip
arthroplasty in patients with osteonecrosis.
Methods: We evaluated thirty-four osteonecrotic hips in thirty
patients who had undergone revision of a femoral component of a prior total
hip arthroplasty. There were nineteen men (twenty-two hips) and eleven women
(twelve hips) with a mean age of forty-six years. Thirty one of thirty-four
hips were implanted without cement. The cementless prostheses were of
different stem lengths, but thirty of thirty-one were proximally porous
coated. The mean duration of follow-up was 8.2 years. Prerevision radiographs
were used to determine the degree of femoral bone loss according to the
classification system of Della Valle and Paprosky. The need for revision was
analyzed for correlation to known risk factors for osteonecrosis, age and
gender of the patient, and degree of prerevision femoral deficiency. A
clinical and radiographic evaluation of outcome was performed.
Results: This was the first revision for twenty-seven hips, the
second for five hips, and the third for two hips. Preoperatively, the defects
included four Type I, nine Type II, fifteen Type IIIA, two Type IIIB, one Type
IV, and three unknown. The femoral component was rerevised in twelve of the
thirty-four hips. One of the failures was the only fully porous-coated stem
that was implanted. One of the three cemented implants failed, as compared
with eleven of the thirty-one noncemented implants. Survival rates were 90.9%
at five years, 54.8% at ten years, 54.8% at fifteen years, and 27.4% at twenty
years. With the small sample size, no relationship could be identified with
regard to frequency of re-revision and defects, associated risk factors,
patient age, or gender.
Conclusions: There was a high failure rate of revised, uncemented,
proximally coated femoral components in patients with osteonecrosis of the
femoral head at the time of the intermediate-term follow-up. The cause of
failure could not be correlated with patient age, gender, risk factors for
osteonecrosis, or femoral bone stock.
Level of Evidence: Therapeutic Level IV. See Instructions
to Authors on
jbjs.org for a
complete description of levels of evidence.
Total hip arthroplasty has become a reliable and durable solution for
patients with arthritis of the hip. Excellent results have been reported with
both cemented and cementless prostheses. Callaghan et al. have reported 98%
survivorship at nineteen years and 90% survivorship at thirty years with the
cemented Charnley
prosthesis1, and
Grübl et al. have reported 99% survivorship at ten years with the
cementless Zweymüller
design2.
Early reports of total hip arthroplasty in patients with osteonecrosis of
the femoral head showed unfavorable results. Of the twenty-seven articles
reviewed by Mont and Hunger-ford in 1995, only two showed acceptable
survivorship compared with age-matched
controls3. More
recently, studies in which modern stem designs and techniques were used have
had more promising
results4-11.
It now appears that in this patient population, the results of primary total
hip arthroplasty with use of modern stem designs are no worse than the results
seen in patients with other diagnoses and with similar age and functional
demand10.
The results of revision total hip arthroplasty in patients with
osteonecrosis of the femoral head are featured in general reports of revision
total hip
arthroplasty12,13.
A review of the literature revealed only one study, by Wei et
al.14, that
directly compared the results obtained after femoral revision with use of a
cemented prosthesis in patients with either osteonecrosis or osteoarthritis.
The authors did not find any difference in the failure rate of femoral
revisions between the two
groups14.
The reasons for failure of revision arthroplasty include factors that are
extrinsic to the patient, such as the implant design and material or the
surgical technique, as well as factors that are intrinsic to the patient, such
as age, weight, activity level, the quality of bone stock remaining at the
time of revision, or the presence of infection. In patients with
osteonecrosis, additional risk factors for failure of revision total hip
arthroplasty, such as corticosteroid dependence, disease involving the femoral
metaphysis, or alcoholism, may contribute to revision failure.
The purpose of this study was to evaluate the results of revision total hip
arthroplasty with use of cementless femoral components in patients with
osteonecrosis of the femoral head. Although a number of the patients who
required rerevision of the femoral component also required acetabular
rerevision, we have restricted our analysis to the survivorship of the femoral
implant because the osteonecrosis in these patients only affected the proximal
portion of the femur.
Between March 1984 and January 2001, thirty-four revision total hip
arthroplasty procedures were performed by one of two surgeons (D.S.H. or
M.W.H.) on thirty patients who had a diagnosis of osteonecrosis of the femoral
head. In this cohort, there were nineteen men (twenty-two hips) and eleven
women (twelve hips). The mean age at the time of revision was forty-six years
(range, twenty-eight to sixty-nine years). Risk factors for osteonecrosis
included corticosteroid use (fifteen hips), excessive alcohol intake (eight
hips), and trauma (seven hips). There were no known risk factors for four
hips. The average duration of clinical follow-up after the index revision was
8.2 years (range, 0.1 to 19.8 years). The average time from the primary total
hip arthroplasty procedure to the first revision was nine years (range, two to
nineteen years) for patients who were undergoing their first revision.
Clinical and radiographic assessment prior to surgery included a detailed
history and physical examination and the calculation of a Harris hip score.
Prior to revision, all patients had development of thigh pain and radiographic
changes consistent with loosening of the femoral component. Additionally,
eight of the patients had evidence of polyethylene wear or acetabular
loosening and underwent acetabular revision at the time of femoral revision.
In addition to femoral loosening, one patient also had development of deep
sepsis. This was the first revision for twenty-seven hips, the second for five
hips, and the third for two hips.
The preoperative radiographs were analyzed and the femoral deficiency was
classified according to the system described by Della Valle and
Paprosky15, and the
results are shown in Table I.
Four hips were classified as Paprosky type I; nine hips, as type II; fifteen
hips, as type IIIA; two hips, as type IIIB; and one hip, as type IV. The
preoperative radiographs could not be located for three patients and thus the
exact type of deficiency was unknown; however, a review of the operative notes
did not indicate massive bone loss.
All revision total hip arthroplasty procedures were performed through a
direct lateral approach. Four hips required the use of an additional
trochanteric osteotomy for exposure. The osteotomy was repaired with
Dall-Miles cables (Stryker, Mahwah, New Jersey) in these hips. The abductor
musculature was reattached using the Krackow suturing
technique16, with
Ethibond suture (Ethicon, Cincinnati, Ohio) in all cases. Morselized allograft
was used in thirty-two of thirty-four hips to fill femoral and acetabular
deficiencies. Only one patient underwent structural bone-grafting with a
femoral strut graft.
Components used in the femoral revision included the Precision cemented
revision stem (two hips) (Howmedica, Mahwah, New Jersey), a cemented Omnifit
stem (one hip) (Osteonics, Mahwah, New Jersey), the PCA long stem cementless
revision stem (eight hips) (Howmedica), the PCA midstem (twenty hips)
(Howmedica), the PCA Primary (two hips) (Howmedica), and the Sentry-S (one
hip) (Howmedica).
Postoperatively, patients were allowed only foot-flat weight-bearing with
crutches for a period of six weeks. Thereafter, they progressed to full
weight-bearing as tolerated. Patients were also required to wear an abduction
brace for the first six weeks after surgery to prevent dislocation.
Postoperative thromboprophylaxis consisted of early mobilization, compression
stockings, sequential compression devices while in the hospital, and aspirin
(325 mg per day for six weeks).
Postoperative follow-up evaluations were performed at six weeks, three
months, and yearly thereafter. At each follow-up appointment, a Harris hip
score and radiographic analysis was performed. Radiographic failure was
defined as progressive stem migration, either in subsidence or angulation, or
increasing osteolysis on successive yearly follow-up examinations. To analyze
the factors that led to the need for revision, the hips in each cohort were
stratified according to whether or not failure had occurred.
Statistical analysis was performed with use of a chi-square test with Yates
correction to compare frequencies and the Student t test (parametric data)
with use of computer programs for epidemiologic analysis (PEPI, version 2;
USD, Stone Mountain, Georgia). Statistical significance was accepted at the p
< 0.05 level. A Kaplan-Meier survivorship curve was generated (PEPI,
version 2; USD) with failure defined as the need for revision.
The average length of follow-up was 8.2 years (range, 0.1 to twenty years).
At the time of final follow-up, twelve (35%) of the thirty-four hips had been
revised. Reasons for revision were loosening (ten hips), massive osteolysis
(one hip), and deep infection (one hip). The average time to rerevision in the
aseptic hips was 101 months (range, one to 237 months). A Kaplan-Meier
survivorship analysis, with failure defined as the need for revision, is given
in Figure 1. The highest rate
of failure was seen in the period between fifty and 120 months
postoperatively. Survival rates were 90.9% at five years, 54.8% at ten years,
54.8% at fifteen years, and 27.4% at twenty years.
Of the twenty-two hips still functioning at the time of the final
follow-up, one patient had a Harris hip score of 49 points. This low score was
due to systemic lupus erythematosus and chronic debilitation and recurrent
acetabular failure. One patient had a score of 78 points due to pain from
arthritis of the knee and ankle on the contralateral side. The remaining
patients had Harris hip scores of 85 points or higher.
In comparing the survivors to the failed group, given the small sample
size, there was no significant difference with regard to age (forty-four
versus forty-eight years) or gender. One of three cemented stems failed, and
eleven of thirty-three cementless stems failed. Only one fully porous-coated
stem was used in this series, and it failed.
When comparing the grade of preoperative femoral deficiency, no significant
difference could be detected between the surviving and the failed hips
(Table II). In the group of
twelve hips that failed, two hips had a grade-I deficiency prior to the index
procedure, one hip had grade II, eight hips had grade IIIA, and one hip had
grade IIIB deficiency. Given the small sample size, the frequency of
high-grade deficiency (Grade IIIA or IIIB) was not statistically different
between the hips that failed and those that did not (p = 0.207). Similarly,
the associated risk factors for osteonecrosis of the femoral head did not
appear to affect outcome. Whether the patient was undergoing the first,
second, or third revision was of no consequence with regard to
survivorship.
Radiographic analysis of the twenty-two surviving hips revealed that two
patients had nonprogressive subsidence (5 and 8 mm), two patients had a
medullary pedestal, and one patient had nonprogressive bead-shedding. One
patient had calcar osteolysis, but the hip was otherwise stable and the lysis
was nonprogressive. The one hip with a cemented stem that failed had a
circumferential radiolucency, but that hip was also stable. Eight patients had
no significant radiographic abnormalities at the time of final follow-up. One
patient had a circumferential radiolucency, progressive subsidence (7 mm), and
migration of the stem. This hip was considered a radiographic failure,
although the patient continued to function well clinically.
Many reasons for aseptic loosening of revised cementless prostheses have
been cited in the literature. These reasons include stem
design17, quality
of the host bone, bone
loss18, the age and
functional demands of the patient, and comorbid conditions, such as renal
dialysis19 and
corticosteroid
use20.
While more recent reports of primary total hip arthroplasty in patients
with osteonecrosis of the femoral head have yielded results comparable with
those in patients with osteoarthritis, little has been written about revision
total hip replacement in this population. Wei et al. published their results
of cemented revisions for a group of patients with
osteonecrosis14.
They compared thirty-four revisions in patients with osteonecrosis to
thirty-five revisions in patients with osteoarthritis. The failure rate in the
osteonecrosis group was 18% compared with 16% in the osteoarthritis group at
an average duration of follow-up of seven years. Statistical analysis did not
reveal any differences between the groups. The authors concluded that patients
with osteonecrosis could expect revision success rates that were comparable
with those of patients with osteoarthritis.
While the failure rate in our series was relatively high at 35%, the
prosthesis was retained in >50% of the cohort for at least ten years. The
reasons for failure did not correlate with any factors known to affect the
failure rate of primary total hip arthroplasty, including age, gender, type of
fixation, or the degree of preoperative bone loss. However, the numbers used
for comparison for some variables were quite small; thus, a larger series
might reveal an effect not recognized in our analysis.
In comparison to the report by Wei et
al.14, in which the
hips were revised with use of cemented stems, most of the hips in the present
study were revised with use of cementless stems. Wei et al. did not
characterize the degree of proximal femoral bone loss at the time of revision.
Also, no survival analysis or radiographic analysis of impending failures was
presented, which made a direct comparison difficult.
Most of the revisions in our series were performed with use of proximally
coated, chromium cobalt stems (porous-coated anatomic [PCA] stem, PCA midstem,
and PCA revision long stem; Howmedica, Rutherford, New Jersey). These
proximally coated, press-fit stems require a good mechanical fit in the
metaphysis for short-term fixation and bone ingrowth or ongrowth in the
metaphyseal region for long-term fixation. In the revised femur, the
metaphysis is often sclerotic and avascular. Other reports of revisions using
this fixation concept have also yielded unsatisfactory
results21,22.
To our knowledge, no reports of results of revision with PCA long stems exist
in the peer-reviewed literature, but we have published preliminary results of
this stem on two occasions. In 1988, at an average duration of follow-up of
two years, we reported two failures due to aseptic loosening of thirty-two
revised hips23.
Subsequently, we reported two failures in a group of fifty-four cementless
revisions that were performed between 1983 and 1987 and were followed for an
average of two
years24. While the
revision rate is lower, the length of follow-up is also much shorter than that
reported in the current series.
Other factors that were not studied in this group included weight,
functional demand, or other disease-related factors, such as continued
corticosteroid use or necrosis of the femoral metaphysis, which may contribute
to early failure.
Since the metaphysis frequently has limited ingrowth potential in these
patients, we believe that alternative types of fixation, such as diaphyseal
press-fit or cemented fixation, should be used routinely in revision of the
femoral stem in patients with osteonecrosis of the femoral head.
In conclusion, patients with osteonecrosis of the fe-moral head who
underwent revision total hip arthroplasty experienced a high failure rate in
this series. No relationship between outcome and age, gender, associated risk
factors, cementless or cemented prosthesis, or the extent of the femoral
deficiency was found in this study. We believe that cementless, proximally
coated, chromium cobalt stems should not be used for revision total hip
replacement in patients with osteonecrosis of the femoral head. ?
Callaghan JJ, Templeton JE, Liu SS,
Pedersen DR, Goetz DD, Sullivan PM, Johnston RC. Results of Charnley total hip
arthroplasty at a minimum of thirty years. A concise follow-up of a previous
report. J Bone Joint Surg Am.
2004;86:
690-5.86690
2004
[PubMed]
Grûbl A, Chiari C, Gruber M,
Kaider A, Gottsauner-Wolf F. Cementless total hip arthroplasty with a tapered,
rectangular titanium stem and a threaded cup: a minimum ten-year follow-up.
J Bone Joint Surg Am.
2002;84:
425-31.84425
2002
[PubMed]
Mont MA, Hungerford DS. Non-traumatic
avascular necrosis of the femoral head. J Bone Joint Surg Am.
1995;77:
459-74.77459
1995
[PubMed]
Garino JP, Steinberg ME. Total hip
arthroplasty in patients with avascular necrosis of the femoral head: a 2- to
10-year follow-up. Clin Orthop Relat Res.
1997;334:
108-15.334108
1997
[PubMed]
Fye MA, Huo MH, Zatorski LE, Keggi KJ.
Total hip arthroplasty performed without cement in patients with femoral head
osteonecrosis who are less than 50 years old. J Arthroplasty.
1998;13:
876-81.13876
1998
[PubMed][CrossRef]
Babis GC, Soucacos PN. Effectiveness of
total hip arthroplasty in the management of hip osteonecrosis. Orthop
Clin North Am. 2004;35:
359-64, x.35359
2004
[CrossRef]
Kim YH, Oh SH, Kim JS, Koo KH.
Contemporary total hip arthroplasty with and without cement in patients with
osteonecrosis of the femoral head. J Bone Joint Surg Am.
2003;85:
675-81.85675
2003
[PubMed]
Lee SB, Sugano N, Nakata K, Matsui M,
Ohzono K. Comparison between bipolar hemiarthroplasty and THA for
osteonecrosis of the femoral head. Clin Orthop Relat Res.
2004;424:
161-5.424161
2004
[PubMed][CrossRef]
Mont MA, Rajadhyaksha AD, Hungerford DS.
Outcomes of limited femoral resurfacing arthroplasty compared with total hip
arthroplasty for osteonecrosis of the femoral head. J
Arthroplasty. 2001;16(8 Suppl 1):
134-9.16134
2001
[CrossRef]
Schneider W, Knahr K. Total hip
replacement in younger patients: survival rate after avascular necrosis of the
femoral head. Acta Orthop Scand.
2004;75:
142-6.75142
2004
[PubMed][CrossRef]
Xenakis TA, Gelalis J, Koukoubis TA,
Zaharis KC, Soucacos PN. Cementless hip arthroplasty in the treatment of
patients with femoral head necrosis. Clin Orthop Relat Res.
2001;386:
93-9.38693
2001
[PubMed][CrossRef]
Callaghan JJ, Salvati EA, Pellicci PM,
Wilson PD Jr, Ranawat CS. Results of revision for mechanical failure after
cemented total hip replacement, 1979 to 1982. A two to five-year follow-up.
J Bone Joint Surg Am.
1985;67:
1074-85.671074
1985
[PubMed]
Mulroy WF, Harris WH. Revision total hip
arthroplasty with use of so-called second-generation cementing techniques for
aseptic loosening of the femoral component. A fifteen-year-average follow-up
study. J Bone Joint Surg Am.
1996;78:
325-30.78325
1996
[PubMed]
Wei SY, Klimkiewicz JJ, Lai M, Garino
JP, Steinberg ME. Revision total hip arthroplasty in patients with avascular
necrosis. Orthopedics.
1999;22:
747-57.22747
1999
[PubMed]
Della Valle CJ, Paprosky WG. The femur
in revision total hip arthroplasty evaluation and classification. Clin
Orthop Relat Res. 2004;420:
55-62.42055
2004
[CrossRef]
Krackow KA, Thomas SC, Jones LC. A new
stitch for ligament-tendon fixation. Brief note. J Bone Joint Surg
Am. 1986;68:
764-6.68764
1986
Jones RE. Modular revision stems in
total hip arthroplasty. Clin Orthop Relat Res.
2004;420:
142-7.420142
2004
[PubMed][CrossRef]
Murphy SB, Rodriguez J. Revision total
hip arthroplasty with proximal bone loss. J Arthroplasty.
2004;19(4 Suppl 1):
115-9.19115
2004
[PubMed][CrossRef]
Sakalkale DP, Hozack WJ, Rothman RH.
Total hip arthroplasty in patients on long-term renal dialysis. J
Arthroplasty. 1999;14:
571-5.14571
1999
[CrossRef]
Dudkiewicz I, Covo A, Salai M, Israeli
A, Amit Y, Chechik A. Total hip arthroplasty after avascular necrosis of the
femoral head: does etiology affect the results? Arch Orthop Trauma
Surg. 2004;124:
82-5.12482
2004
[CrossRef]
Berry DJ, Harmsen WS, Ilstrup D,
Lewallen DG, Cabanela ME. Survivorship of uncemented proximally porous-coated
femoral components. Clin Orthop Relat Res.
1995;319:
168-77.319168
1995
[PubMed]
Mulliken BD, Rorabeck CH, Bourne RB.
Uncemented revision total hip arthroplasty: a 4-to-6-year review. Clin
Orthop Relat Res. 1996;325:
156-62.325156
1996
[CrossRef]
Hungerford DS, Krackow KA, Lennox DW.
The PCA primary and revision hip systems. In: Fitzgerald RH Jr, editor.
Non-cemented total hip arthroplasty. New York: Raven Press;
1988. p 433-50.433
1988
Hungerford DS, Jones LC. The rationale
of cementless revision of cemented arthroplasty failures. Clin Orthop
Relat Res. 1988;235:
12-24.23512
1988