Abstract
BACKGROUND: Both the material and geometry of a total knee arthroplasty (TKA) component influence the induced periprosthetic bone strain field. Strain, a measure of the local relative deformation in a structure, corresponds to the mechanical stimulus that governs bone remodeling and is therefore a useful in vitro biomechanical measure for assessing the response of bone to new implant designs and materials. A polyetheretherketone (PEEK) femoral implant has the potential to promote bone strains closer to that of natural bone as a result of its low elastic modulus compared with cobalt-chromium (CoCr). QUESTIONS/PURPOSES: In the present study, we used a Digital Image Correlation (DIC) technique to answer the following question: Does a PEEK TKA femoral component induce a more physiologically normal bone strain distribution than a CoCr component? To achieve this, a DIC test protocol was developed for periprosthetic bone strain assessment using an analog model; the protocol aimed to minimize errors in strain assessment through the selection of appropriate analysis parameters. METHODS: Three synthetic bone femurs were used in this experiment. One was implanted with a CoCr femoral component and one with a PEEK femoral component. The third (unimplanted) femur was intact and used as the physiological reference (control) model. All models were subjected to standing loads on the corresponding polyethylene (ultrahigh-molecular-weight polyethylene) tibial component, and speckle image data were acquired for surface strain analysis using DIC in six repeat tests. The strain in 16 regions of interest on the lateral surface of each of the implanted bone models was plotted for comparison with the corresponding strains in the intact case. A Wilcoxon signed-rank test was used to test for difference at the 5% significance level. RESULTS: Surface analog bone strain after CoCr implantation indicated strain shielding (R(2) = 0.6178 with slope, β = 0.4314) and was lower than the intact case (p = 0.014). The strain after implantation with the PEEK implant deviated less from the intact case (R(2) = 0.7972 with slope β = 0.939) with no difference (p = 0.231). CONCLUSIONS: The strain shielding observed with the contemporary CoCr implant, consistent with clinical bone mineral density change data reported by others, may be reduced by using a PEEK implant. CLINICAL RELEVANCE: This bone analog in vitro study suggests that a PEEK femoral component could transfer more physiologically normal bone strains with a potentially reduced stress shielding effect, which may improve long-term bone preservation. Additional studies including paired cadaver tests are necessary to test the hypothesis further.