We previously reported the three-to-six-year1 and seven-to-eleven-year2 results following the insertion of a porous-coated acetabular metal shell without cement (Harris-Galante I; Zimmer, Warsaw, Indiana) in a consecutive series of 138 revision total hip arthroplasties performed in 132 patients by four surgeons at our institution. The acetabular metal shell is hemispherical, made of commercially pure titanium, and fully coated with sintered titanium fiber-metal mesh. The surgical technique included reaming of the acetabulum with hemispherical reamers to the same diameter as the component to be inserted and the use of three to six bone screws for fixation. The median size of the acetabular components was 62 mm (range, 46 to 78 mm). A bulk allograft was used in two hips with segmental defects, and cancellous grafts were used in 107 hips with contained defects. In eighteen hips, only the acetabular component was revised. In the remaining hips, the femoral component was revised with a variety of components: forty-eight cementless Harris-Galante stems, forty-three cementless Gustilo-Kyle stems, twenty-one cementless BIAS stems, seven cemented Harris Precoat stems, and one allograft-prosthesis composite (all manufactured by Zimmer). The mean age of the patients in this cohort was fifty-five years old (range, twenty to eighty-two years old) at the time of the index revision surgery.
At the time of the most recent report2, no acetabular component had been revised because of aseptic loosening, there had been no operations because of wear of the polyethylene liner or for the treatment of osteolysis, and no cups were loose as seen radiographically. Nonprogressive periacetabular osteolysis was seen in four (4%) of the hips that had been followed radiographically; no osteolytic lesion was >4 mm in diameter, and all of the lesions were located at the periphery of the component. As complications such as aseptic loosening and osteolysis tend to occur more frequently with longer follow-up, the goal of this update was to present the longer-term outcomes of revision total hip arthroplasty with this implant.
The patients were evaluated prospectively, and the data were retrospectively reviewed at a minimum of fifteen years postoperatively. Institutional review board approval was obtained for this study. Clinical evaluation was performed with use of the Harris hip score3. A clinical score was considered to be excellent if it was >90 points, good if it was between 80 and 89 points, fair if it was between 70 and 79 points, and poor if it was <70 points. The techniques that were utilized for radiographic evaluation were previously described by Martell et al.4. Osteolysis was defined by the criteria described by Archibeck et al.5, and it was classified according to the location and size (with small considered to be ≤1 cm in diameter and large, >1 cm in diameter). Preoperative bone defects were classified according to the system of Paprosky et al.6; there were seven Type-1, forty-six Type-2A, forty Type-2B, eleven Type-2C, seventeen Type-3A, and two Type-3B acetabula. The original preoperative radiographs were not available for fifteen hips. Kaplan-Meier survival analyses7 were performed with use of the following end points: revision of the acetabular metal shell because of aseptic loosening or radiographic evidence of definite loosening, revision of the acetabular component for any reason, radiographic identification of periacetabular osteolysis, revision of the femoral component because of aseptic loosening, and revision of the femoral component for any reason. All radiographic and clinical data available for all 138 hips were used for the calculation of survivorship.
SPSS for Windows (version 11; SPSS, Chicago, Illinois) was used for data management and statistical analysis. Because the time from the surgery to the time of the first detection of osteolysis and the time from the surgery to the time of the first reoperation were often censored, survival analysis methods7 were used to analyze these data. For the analysis of the first detection of lysis, patients were censored at death, at the time of replacement of the acetabular component, and at the end of follow-up. For the analysis of the survival until the first reoperation, patients were censored at death and the end of follow-up. The log-rank test was done to compare independent groups with respect to the times until these end points. Bivariate Cox proportional hazards regression (with likelihood ratio p values) was used to investigate whether these times were related to gender, age at surgery, weight and body mass index (weight in kilograms divided by the square of the height in centimeters) at surgery, thickness of the polyethylene liner, type of femoral head (titanium or non-titanium), or acetabular inclination angle. Backward elimination Cox regression was done to evaluate multivariable relationships between these times and gender, age, body mass index, and liner thickness and to evaluate interactions among gender, age, body mass index, and liner thickness. Interactions were represented by cross-product terms, and categorical variables were represented by indicator variables. For these analyses and for the Kaplan-Meier curves, only one hip in each patient (the first hip that was operated on in the patients with bilateral revision) was used to ensure statistical independence and, hence, the validity of the statistical tests and confidence intervals. A 0.05 significance level was used for all statistical tests. No one-sided statistical tests were done.
At a minimum of fifteen years postoperatively, forty-one patients (forty-three hips) had died, seven patients (eight hips) had been lost to follow-up, and twenty patients (twenty hips) had undergone a repeat revision of the acetabular component. This left sixty-seven hips in sixty-four patients for the present study. The mean age of these sixty-four patients at the time of surgery was fifty-five years old. Fifty-four patients (fifty-seven hips) were evaluated both clinically and radiographically, and ten patients (ten hips) had a clinical evaluation only. The ten patients completed a telephone questionnaire but either could not or would not have new radiographs made. The latest clinical and radiographic evaluations were performed at a mean of 205 months (range, 180 to 231 months) postoperatively. A comparison of the results between the current and previous studies is shown in Table I.
The Harris hip score for the patients who did not have a repeat revision improved from a mean of 49 points preoperatively to a mean of 82 points (range, 38 to 100 points) at the most recent evaluation (p < 0.001). A comparison of the hip scores in the present study with those in the previous studies is shown in Table II. Of the fourteen patients with a poor score, seven had severe low-back pain, two had polyarticular arthritis, and one had a cervical spine disorder; the poor scores for the remaining four patients were directly related to a poorly functioning hip replacement.
Of the 138 hips in the entire cohort of 132 patients, twenty (14%) underwent repeat revision of the acetabular component. Seven repeat revisions (5%) were performed because of recurrent dislocation; five of these procedures followed the index revision, and two followed a later femoral revision. All of these metal shells were well fixed. Seven metal shells, six of which were well fixed, were removed at the time of femoral revision surgery, and six were revised because of infection. Thus, at the time of the repeat revision, nineteen of the twenty components were deemed to be well fixed. The one patient with a loose metal shell underwent revision at 122 months because of aseptic loosening of the femoral component. At the time of surgery, the screws were removed and the metal shell was found to be fixed by fibrous tissue only and was removed. Preoperative radiographs showed a radiolucent line of <2 mm in thickness in four of the five zones surrounding this component. In the entire cohort of sixty-seven hips that were followed for a minimum of fifteen years, thirty-eight did not require a reoperation and still had the original femoral component, acetabular component, and polyethylene liner in place at the most recent evaluation.
Two patients underwent a reoperation on the hip for a problem directly related to the acetabular component but did not require revision of the metal shell. One of these patients, who was twenty-six years old at the time of the index revision, underwent the reoperation because of excessive wear of the polyethylene liner at 142 months postoperatively; the metal shell was found to be well fixed at the reoperation and was retained. The polyethylene liner in this patient was 9.2 mm in thickness. The other patient, who was thirty-five years old at the time of the index revision, underwent a repeat revision because of wear of the polyethylene liner and osteolysis at 189 months; the metal shell was found to be well fixed at the reoperation and was retained. The polyethylene liner in this patient was 6.2 mm in thickness. A comparison of the repeat acetabular revisions reported in this study and our previous studies is shown in Table III.
In the entire cohort of 138 hips, two metal shells (1.4%) were definitely loose as seen radiographically. One patient died before revision surgery was required, and the other patient was asymptomatic and elderly at the time of writing. Of the fifty-seven metal shells that were followed radiographically for more than fifteen years, one was definitely loose, six were possibly loose, and fifty were deemed to be well fixed (Table IV). All six of the possibly loose components had previously been classified as possibly loose, and none showed evidence of progressive radiolucent lines. Four of these six components were in patients with a Harris hip score of >90 points, and the two remaining components were in the same patient, who, at the time of writing, was ninety years of age and had multiple medical comorbidities and musculoskeletal problems.
Twenty-three (40%) of the fifty-seven metal shells that survived and were followed radiographically for more than fifteen years were associated with osteolysis; nine (16%) of the osteolytic lesions were >1 cm in diameter. In the entire cohort of 138 metal shells, twenty-seven (20%) were associated with radiographic evidence of osteolysis, and eleven of the lesions were >1 cm in diameter (8% prevalence in the series). Of the twenty-nine osteolytic lesions, fifteen were small peripheral lesions, eleven were large peripheral lesions, one was a small retroacetabular lesion, and two (including the one in the patient who had a reoperation because of the osteolysis) were large retroacetabular lesions. Osteolysis was first identified radiographically at a mean of 137 months (range, fifty-nine to 223 months) after the index operation.
In the entire cohort of 132 patients (138 hips), the fifteen-year rate of survival of the acetabular component, with failure defined as revision of the acetabular metal shell because of aseptic loosening or radiographic evidence of definite loosening, was 97%. With failure defined as a revision of the acetabular component for any reason, the fifteen-year survival rate was 81%. At fifteen years, 70% of the shells were not associated with radiographic evidence of osteolysis. Kaplan-Meier curves with 95% confidence intervals for these outcomes are shown in Figures 1, 2, and 3.
With the numbers available, log-rank tests and bivariate Cox regression did not identify a significant relationship between the time to radiographic identification of osteolysis and gender, weight or body mass index, liner thickness, femoral head type, or acetabular inclination angle. However, age was found to be significant (p = 0.012), with the risk of osteolysis decreasing as age increased. Age was the only variable that was included in the final model produced by backward-elimination Cox regression. Similar results were obtained for the time to the first reoperation: age (p = 0.001) was the only significant variable (the risk of a reoperation decreased as age increased) and the only variable kept in the final Cox regression model.
Although not the focus of this report, the outcomes with regard to the femoral components were also studied. Fifty-two (38%) of the 138 femoral components were rerevised at a mean of ninety-eight months (range, one to 203 months). Forty-two were re-revised because of loosening; five, because of infection; three, because of distal osteolysis; and two, because of periprosthetic fracture. The modular polyethylene acetabular liner was changed at the time of all fifty-two femoral revisions. Eighteen (38%) of the forty-eight cementless Harris-Galante stems, sixteen (37%) of the forty-three cementless Gustilo-Kyle stems, eight (38%) of the twenty-one cementless BIAS stems, one (14%) of the seven cemented Harris Precoat stems, and the one allograft-prosthesis composite were revised. Eight (44%) of the eighteen cemented femoral components that had been retained at the time of the index acetabular revision were revised, at a mean of 105 months (range, forty-five to 167 months) postoperatively. The rate of survival of the femoral components at fifteen years was 58% with revision because of aseptic loosening as the end point, and it was 52% with revision of the femoral component for any reason as the end point. Kaplan-Meier curves with 95% confidence intervals for these outcomes are shown in the Appendix.
Harris-Galante cementless metal shells used for acetabular revision arthroplasty performed well, with a low rate of aseptic loosening at a minimum of fifteen years postoperatively. However, since our previous reports, we found an increase in the prevalence of periacetabular osteolysis. The most common reasons for repeat revision of the metal shell were recurrent dislocation and infection. On the basis of this finding, we now routinely use larger-diameter (36 and 40-mm) femoral heads that articulate with a highly cross-linked polyethylene liner to try to decrease the rate of dislocation. Although six well-fixed shells were removed at the time of femoral revision surgery, we currently recommend retaining well-fixed and well-positioned acetabular components at the time of femoral revision surgery8.
On the basis of the results of this and our previous reports, we continue to use a hemispherical, titanium metal shell with multiple screws for fixation in the vast majority of our acetabular revisions. Although the component described in this report is no longer available, the component that we presently use (Trilogy; Zimmer) is hemispherical and has the same ingrowth surface as its predecessor. However, larger-diameter screws are used, and the locking mechanism for the modular polyethylene liner has been improved.
Kaplan-Meier survivorship curves for the femoral components are available with the electronic versions of this article, on our web site at jbjs.org (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). ▪
NOTE: The authors thank Mitchell B. Sheinkop, MD, for the contribution of cases for this series.
In support of their research or preparation of this manuscript, one or more of the authors received grants or outside funding from Zimmer. In addition, one or more of the authors received payments or other benefits or a commitment or agreement to provide such benefits from a commercial entity (Zimmer). Also, a commercial entity (Zimmer) paid or directed, or agreed to pay or direct, benefits to a research fund, foundation, educational institution, or other charitable or nonprofit organization with which the authors are affiliated or associated.
* Original Publications Padgett DE, Kull L, Rosenberg A, Sumner DR, Galante JO. Revision of the acetabular component without cement after total hip arthroplasty. Three to six-year follow-up. J Bone Joint Surg Am. 1993;75:663-73.
* Silverton CD, Rosenberg AG, Sheinkop MB, Kull LR, Galante JO. Revision of the acetabular component without cement after total hip arthroplasty. A follow-up note regarding results at seven to eleven years. J Bone Joint Surg Am. 1996;78:1366-70. Cited Here...
Investigation performed at the Department of Orthopaedic Surgery, Rush University Medical Center, Chicago, Illinois