Complex and unstable proximal humerus fractures (PHF) in the elderly are challenging to treat with considerable rates of failures even with state-of-the-art locking plates. A large part of these complications is related to mechanical factors, with loss of reduction via screw perforation and cut-out being the predominant failure modes. Alleviating the risk of mechanical failures would contribute to improved outcomes of PHF. Computer simulations such as finite element (FE) analyses could help understand the reasons behind the multi-factorial problem of mechanical failures and aid in the improved use of currently existing fixation technologies or develop novel advanced designs ensuring higher stability. Towards this aim, we have developed an FE simulation framework for PHF that allowed incorporation of anatomical variance with nearly 50 virtual subjects, seven different fracture patterns, freely configurable locking plate fixations and various loading schemes. We validated the underlying FE analysis technology to predict cyclic screw cut-out failure experimentally measured in instrumented cadaveric human PHF specimens (R 2 =90). A series of in silico studies was then performed with the computational framework to systematically address clinically relevant questions related to the type and position of the plate, as well as the length, configuration and cement reinforcement of the locking screws, with each of these involving 24–42 virtual specimens and requiring 504–4608 simulations. The analyses provided clinically relevant findings suggesting maximizing screw length and spread, prioritizing calcar screws for augmentation and aiming towards a proximal plate position. Beyond the use of existing implants, we utilized the in silico tool to optimize screw trajectories in a study involving 19 digital subjects and more than 5000 analyses. The resulting design was confirmed in a subsequent biomechanical study to provide significantly improved stability compared to the standard implant. A follow-up in silico study has shown only moderate but significant benefits for patient-specific implants versus standardized optimized ones. These studies demonstrated the potential of validated in silico tools for improved care of PHFs, although the findings require clinical corroboration.
Варга et al. (Mon,) studied this question.