Fatigue failure of plated osteoporotic proximal humerus fractures is predicted by the strain around the proximal screws.

The high rate of required reoperation indicates that treatment of fragility fractures at the proximal humerus still remains a major challenge in trauma surgery. Improved fixation approaches are needed. Several limitations of the conventional implant development process involving experimental testing could be overcome by using computer models that would allow systematic and efficient analyses. However, such models require experimental validation. This study investigated if linear elastic continuum finite element (FE) models can predict experimental fatigue failure in locking plate fixation of osteoporotic proximal humerus fractures. Three-part fractures were created in twenty fresh-frozen proximal humeri of elderly donors, stabilized with angular stable plate osteosythesis and tested to failure in a previously developed experimental setup using a cyclic loading protocol with increasing peak load. Case-specific, linear elastic FE models of the instrumented samples were created from CT images and loaded virtually by mimicking the experimental conditions. Average principal strains were evaluated in cylindrical regions around the proximal screws. Parametric sensitivity analysis was performed to investigate the effects of specific model parameters on the results. The number of cycles to failure was 10500 ± 3300 (mean ± SD, range: 3100 - 16400) and showed a strong logarithmic correlation with the average compressive principal strain around the screws (R2 = 0.90). These results suggest that the latter parameter may be used as a surrogate estimate for construct stability under cyclic loading. The computationally cheap linear elastic continuum FE analysis could be used as an efficient screening tool for optimization and development of implants. Further work is required to investigate if the findings of this study apply to other loading modes and bone-implant constructs.