RESUMO
The aim of this study was to investigate the performance of a cementless osseointegrated tibial tray (P.F.C. ® Sigma®, Depuy® Inc, USA) in a general population using finite element (FE) analysis. Computational testing of total knee replacements (TKRs) typically only use a model of a single patient and assume the results can be extrapolated to the general population. In this study, two statistical models (SMs) were used; one of the shape and elastic modulus of the tibia, and one of the tibiofemoral joint loads over a gait cycle, to generate a population of FE models. A method was developed to automatically size, position and implant the tibial tray in each tibia, and 328 models were successfully implanted and analysed. The peak strain in the bone of the resected surface was examined and the percentage surface area of bone above yield strain (PSAY) was used to determine the risk of failure of a model. Using an arbitrary threshold of 10% PSAY, the models were divided into two groups ('higher risk' and 'lower risk') in order to explore factors that may influence potential failure. In this study, 17% of models were in the 'higher risk' group and it was found that these models had a lower elastic modulus (mean 275.7MPa), a higher weight (mean 85.3kg), and larger peak loads, of which the axial force was the most significant. This study showed the mean peak strain of the resected surface and PSAY were not significantly different between implant sizes.
