Design considerations and life prediction of metal-on-metal bearings: the effect of clearance.

BACKGROUND Clinical observations suggest that metal-on-metal arthroplasties that have been implanted for more than twenty years do fail. It is proposed that there are not two, but three distinct phases of wear life for any metal-on-metal implant system: bedding-in, steady state, and end point. In this study, we asked two questions: can we explain late failure due to wear, and will there be a late failure mechanism due to a change in the frictional torque? METHODS In order to characterize wear failure, an analysis was made of five retrieved metal-on-metal couples that were mapped with use of a roundness machine. A geometrical model was developed on the basis of these observations, and wear at the end point was calculated. The literature on first-generation metal-on-metal implants retrieved for aseptic loosening was reviewed to assess the agreement with the retrieval findings as well as the wear model. RESULTS A wear patch of an appreciable and constant size could be measured in all five retrieved couples. The end point of revision was observed to occur when the wear progression reached a contact area corresponding to approximately 75% of the projected diameter of the ball. The wear volume was calculated from the geometry. The available literature describing the wear characteristics of retrieved bearings after successful clinical use showed good agreement with the calculated wear model. CONCLUSIONS During the implant life of long-term successful metal-on-metal devices, a wear patch develops, as evident from retrieved failed devices. Failure often occurs through loosening, and the observed wear patch is similar in size for devices measured by us and for those described in the literature. We hypothesized that failure by loosening occurs through the accumulation of wear, which eventually leads to high friction within the bearing and increased torsional forces across the joint and its fixation.