The design and pre-clinical evaluation of knee replacements for osteoarthritis

Abstract

One of the concepts that Rik Huiskes promoted was that implants such as knee and hip replacements could be analyzed and optimized using numerical models such as finite element analysis, or by experimental testing, an area he called pre-clinical testing. The design itself could be formulated or improved by defining a specific goal or asking a key question. These propositions are examined in the light of almost five decades of experience with knee implants. Achieving the required laxity and stability, was achieved by attempting to reproduce anatomical values by suitable radii of curvature and selective ligament retention. Obtaining durable fixation was based on testing many configurations to obtain the most uniform stress distribution at the implant-bone interface. Achieving the best overall kinematics has yet to be fully solved due to the variations in activities and patients. These and many other factors have usually been addressed individually rather than as a composite, although as time has gone on, successful features have gradually been assimilated into most designs. But even a systematic approach has been flawed because some unrecognized response was not accounted for in the pre-clinical model, a limitation of models in general. In terms of the design process, so far no method has emerged for systematically reaching an optimal solution from all aspects, although this is possible in principle. Overall however, predictive numerical or physical models should be an essential element in the design of new or improved knee replacements, a part of the design process itself.