In total hip arthroplasty, aseptic loosening remains the most important cause of long-term implant failure. Chronic radiographically documented migration and progressive periprosthetic radiolucencies remain the clinical hallmarks of aseptic loosening. For the practicing surgeon, the goal of follow-up is to identify the symptomatic “moving” implant. The current literature describes several methods for studying how various implant systems perform clinically. These methods include national implant registries, prospective randomized clinical trials, retrospective case series, biomechanical bench and theoretical studies, and tell-tale case reports. Lack of generalizability is the major weakness of each of these approaches, and the most important editorial response of this journal is to limit the reach of authors’ conclusions.
This article is “classic” in that it takes a well-accepted scientific method for assessing implant loosening and carefully chronicles clinical experience with the acetabular component of a total hip replacement over a minimum of ten years. RSA, also known as roentgenographic stereophotogrammetry, has an accuracy of 300 μm or better, and this method has been standardized by an international research consortium1,2. The authors’ values for the RSA precision of this study indicate a “target grouping” of approximately 0.2 to 0.5 mm for migration and 0.7° to 0.9° for cup inclination, which means that virtually every clinical measurement falls within these error parameters. Given this precision, the biases involved in interpretation of radiographs are virtually eliminated. Large numbers of patients are no longer needed to generate statistically significant conclusions. Additionally, we can begin to understand clinical performance based solely on how much demonstrated movement is identified over time.
Prior to this study, practicing surgeons knew that failing cups migrate superiorly and commonly have an increasing cup inclination. Aspenberg et al. demonstrated that a clear dichotomy in results could be identified within two years, with one group of acetabular components showing much greater migration than the remaining components3. Similar RSA studies performed with other types of implants (such as cemented femoral components and the tibial base plate in total knee arthroplasty) have demonstrated parallel findings in which some implants were quite stable from the outset but others that were doomed to long-term clinical failure showed early increased migration4.
The results of this study demonstrated long-term aseptic loosening in approximately 25% of the conventional cemented all-polyethylene cups at a mean of seventy-six months of follow-up. Examination of the RSA data acquired at two years of follow-up revealed that the implants destined to long-term failure migrated superiorly by a mean of approximately 2 mm and tilted more vertically by approximately 4.5°. Without exception, the trend for late loosening was consistent with these early RSA findings. The obvious conclusion from this study is that implants that demonstrate substantial early micro-instability will not stabilize and will become late failures. Furthermore, for the practicing surgeon, early radiographically documented cephalad cup migration and increasing cup inclination presage late loosening.
Although this study may offer important knowledge regarding this particular cup model and technique, we may not generalize the results to all components or schemes of fixation. Another important qualification is the fact that other problems, such as polyethylene-caused osteolysis or periprosthetic fractures resulting from trauma, may not be predicted by an RSA study. Those statements notwithstanding, the results of this paper suggest that we may monitor new devices with RSA and draw conclusions regarding possible ultimate late success or failure. These conclusions may be made on the basis of the early experience with the implant, before long-term findings are available from national registries.