Abstract
Purpose: We aimed to measure the early migration pattern of a titanium alloy, tapered, plasma and hydroxyapatite coated femoral stem and any factors associated with subsidence.
Method: Between January 2005–June 2007, 387 Accolade cementless femoral stems (Stryker, Allendale NJ) were implanted at our institution. Seventy-seven had a minimum of two years post operative follow up and a complete set of pre and postoperative radiographs for analysis. Our group inlcuded 45 females with a mean age of 71.4 years, and 32 males with a mean age of 68.5 years. The primary diagnosis was degenerative osteoarthritis in 71 patients, avascular necrosis in two, and post fracture in four patients. The average BMI was 27.1. We measured the canal index to assess bone quality and the canal calcar index to assess the proximal femoral morphology. Immediate postoperative radiographs were assessed for canal fill of the prosthesis and implantation varus/valgus angles. The EBRA-FCA software was used to obtain migration curves for each stem. Best fit curve of subsidence over time was calculated and the data was analysed using a Kaplan Meier survivorship with 1.5 mm of subsidence as an endpoint. We then performed a multivariate and univariate regression analysis for predictors of subsidence.
Results: The mean follow up was 29.3months (24–48). The mean canal index was 0.55 (0.36–0.68) with a mean canal calcar index of 0.54 (0.39–0.79). The average canal fill index at the midpoint of the stem was 80 in 40 stems. A total of 414 radiographs were analysed for the EBRA measurements. Of these 21 (5%) were discarded by the software as they did not meet the criteria for comparability. This led to the exclusion of seven patients, leaving the final study group of 77. All remaining patients had a minimum of four radiographs with an average of 4.6 for analysis. The average subsidence at 24 months was 2 mm and this had risen to 2.4 mm by 36 months postoperatively. When analysed using a Kaplan Meier curve using 1.5 mm as an end point we found a survivorship of 63.4% (52.3–74.5) at 24 months and this had worsened to 41.6% (26.6–56.5) by 36 months. Multivariate and univariate regression analysis of measured variables did not reveal any significant hazard for any factor other than the larger stem sizes doing worse.
Conclusion: Although several cementless tapered stem designs have had an excellent track record, our migration analysis of the Accolade stem is somewhat concerning. Thirty three percent of stems had reached the 1.5 mm subsidence point by two years. This is of concern as work has previously shown this to predict failure of stems with aseptic loosening at ten years with an accuracy of 79%. If these stems go on to fail at the predicted rate this would represent an unacceptably high level of failure. Our data raises serious concerns about the overall clinical performance of this stem design due to poor initial stability and integration.
Correspondence should be addressed to: COA, 4150 Ste. Catherine St. West Suite 360, Westmount, QC H3Z 2Y5, Canada. Email: meetings@canorth.org
