Computational optimisation of screw orientations for improved locking plate fixation of proximal humerus fractures

Abstract

Abstract Background The implant-related failure risks of proximal humerus plates remain high, and therefore, improved solutions are needed. Systematic and efficient computational analyses can assist design optimisation of implant systems and may help reduce complication rates. Methods A previously developed validated computer simulation framework was used to optimise the screw orientations of a standard locking plate. Twenty low-density proximal humerus models with simulated unstable 3-part fractures were fixed using the Proximal Humeral Internal Locking System (PHILOS, DePuy Synthes) with six proximal screws. In a parametric analysis, the screw orientations were varied by relocating their tips within the humeral head and optimised based on two different metrics. In a first approach, average bone mineral density (BMD) around the screw tips was maximised. In a second approach, the average bone strain around the screw tips was evaluated using finite element (FE) analyses in three physiological loading situations and minimized to maximise the predicted fixation stability. Results Optimisation based on BMD did not deliver any improvement. The final FE-based optimised configuration involved the adjustment of each screw and exhibited significantly smaller peri-implant bone strain (−18.49% ​± ​9.56%, p ​ Conclusion Our findings suggest that the design of the investigated proximal humerus plate could be improved by changing the screw orientations. This study demonstrates the potential of FE analyses for implant optimisation. However, the results do not allow direct translation into a novel plate design, mainly owing to the use of a restricted set of bone samples, a single simplified fracture model, and lack of direct biomechanical confirmation of the computational findings. The translational potential of this article A cohort-specific optimised proximal humerus locking plate design was evaluated using a previously validated computer simulation framework. This study showed that screw angle optimisation solely based on BMD is not sufficient and that incorporating the mechanical aspects is required. This approach could be used in the future to inform about the best screw trajectories for variable-angle locking plates or to develop novel implants for improved treatment of proximal humerus fractures. It is therefore expected to help reduce the complication rates of the implant-based treatment of these challenging injuries.