Bone stress relaxation in press-fit femoral knee implant fixation: A combined experimental and computational analysis.

Abstract

Aseptic loosening is the leading cause of revision in cementless total knee arthroplasty (TKA), emphasizing the need for strong initial stability. Pre-clinical evaluations are essential for understanding implant fixation mechanics. Finite element (FE) models often rely on linear elastic bone material models, which inadequately represent bone behavior. Incorporating bone plasticity enhances predictions of primary stability, but may overestimate fixation by neglecting bone viscoelasticity and abrasion. Stress relaxation reduces press-fit forces, while bone abrasion weakens fixation strength. This study incorporated bone relaxation and abrasion into FE simulations of femoral TKA fixation and compared results to experimental measurements. Cadaveric femurs were implanted with femoral components, and strain gauges recorded outer surface strain over 24 h. A pull-off test evaluated fixation strength. FE models, using a plastic-viscoelastic bone material model, simulated these phases with varying levels of bone abrasion. The computational model without abrasion predicted implant strain 350 % higher than experimental measurements. 0.75 mm of virtual abrasion reduced overestimation to 150 %. Experimental strain reductions due to bone relaxation varied per reconstruction from 17 % to 37 %, while computational predictions were on average 41 % lower. Substantial variability was observed in both the experimental and simulated pull-off forces. Simulations without abrasion showed 180 %-360 % higher pull-off forces than the experiments. Including 0.5 mm of virtual abrasion reduced pull-off forces, aligning them more closely with experimental values, though 0.75 mm led to underestimation in some cases. These results highlight the need to include bone relaxation and abrasion in simulations to optimize implant designs and surgical outcomes.