CORR Insights®: RSA-tested TKA Implants on Average Have Lower Mean 10-year Revision Rates Than Non-RSA-tested Designs

Where Are We Now?

New total joint implants are introduced into the market for a variety of reasons. These may be scientific (such as new innovations, different technologies, or specific design changes), economic (such as jobs, market share, or value), or clinical (such as failure rate, increased surgical options, or perceived performance). Thus, many implants with a proven long-term success rate may be replaced by newer, clinically unproven ones [3]. Indeed, once a new prosthesis is designed, manufactured, delivered, implanted, and studied over the long-term, years may pass and eventually the product may become obsolete, replaced by a “new and improved” implant. This process may involve preclinical trials and governmental approval before the prosthesis is actually introduced into the marketplace. Joint implant registries also collect and disseminate information, both good and bad. The common goal in all of these processes is to detect any early problems with a new prosthesis, such as early loosening.

Early prosthetic migration within the first 24 months has been shown to predict eventual failure from aseptic loosening with some implant designs [3, 4]. Recall, however, that a relatively small amount of initial postoperative component migration is not unusual with some implants [4] and this should be studied sometime after the first several months postoperatively. Roentgen stereophotogrammetric analysis (RSA) remains one the most-accurate methods to measure migration of a prosthesis. It has an accuracy of 0.2 mm, which is 10- to 20-times greater than that using ordinary radiographs [2]. To diminish the number of revision operations, by identifying prostheses that demonstrate high levels of early migration and subsequent loosening, some authors have advocated implementing mandatory 2-year RSA testing before a new prosthesis is introduced into the marketplace [3]. In the current study, Hasan et al. [1] examined data from seven national registries and concluded that the RSA-tested design had a lower (less than 1%) revision rate at 5 and 10 years than did non-RSA-tested implants. This conclusion appears to support the importance of RSA testing before the introduction of a new implant. These implants might be eliminated, or the use of prostheses that may be predisposed to aseptic loosening might be diminished.

Where Do We Need to Go?

Subjective evaluation (pain scores, function scores, and patient-reported outcomes) as well as objective metrics (ROM, stability, and radiographic measurements, including RSA data) are commonly used to determine the success of a total joint arthroplasty. Ideally, all of these data should be collected before the release of an implant. One metric alone is not sufficient. RSA data must be evaluated in terms of other subjective and objective scores in order to completely understand the importance of RSA. Not all non-RSA-tested prostheses are inferior. In fact, the current study [1] actually demonstrated that some RSA-tested implants did worse that some non-RSA-tested implants in terms of revision rates. Furthermore, although the difference in 5- and 10-year revision rates between the two groups was small, there was a wide variation in revision rates. There are many indications for revision arthroplasty. Causes are influenced not only by prosthesis design factors (such as polyethylene wear, patellofemoral kinematics, or cement-prosthesis interface problems) and patient factors (such as BMI, level of activity, trauma, or diabetes), but also by surgeon factors (such as operative technique, prosthesis choice, or cement technique). Clearly, it is difficult, at best, to determine these causes by reviewing registry data alone.

With respect to the RSA-tested and non-tested prostheses, does a small difference in revision rates exist between cruciate-retaining and posterior-stabilized prostheses? Between fixed-bearing and mobile-bearing devices? Between cemented and uncemented implants? Between modular and non-modular tibial implants? Between femoral and tibial components? How do extensor mechanism complications influence the revision rates? The reader should not simply conclude that all revisions of non-RSA-tested implants are secondary to aseptic loosening alone. It is wrong to assume that if non-RSA-tested implants are, in fact, RSA-tested, then the revision rate would be diminished because the inferior designs have been made unavailable. This conclusion has yet to be supported. Yet, it does appear that RSA testing is an important tool in the evaluation of new implants.

How Do We Get There?

In the Hasan et al. study [1], the goal was to diminish the number of revision knee arthroplasties by identifying implants that are subject to premature loosening by using RSA testing. Because many patient and surgeon factors may lead to aseptic loosening, each factor must be controlled between the RSA-tested and non-tested groups. Additionally, each prosthesis must be evaluated individually. Large, multicenter trials using RSA are needed during the prerelease phase of the investigational trial for each different prosthetic design. Generally, these trials are initiated by the designing surgeons who supervise the data collection and evaluation. Many (and perhaps most) implants may perform well with or without the scrutiny of RSA testing. If RSA testing shows that migration continues beyond the early months after implantation, it is ultimately up to the surgeon, the regional standard of care, and the orthopaedic community at large to decide what is, or what is not, an acceptable aseptic loosening rate.