An In-Silico Assessment of Stemless Shoulder Arthroplasty: from CT to Predicted Bone Response

Despite the emergence of stemless humeral implants that utilize short fixation features to gain purchase solely in the metaphysis, the literature contains little information regardingthe morphology and mechanical properties of the humerus’ proximal trabecular-canal, and how stemless implants impact bone response. The present work employs in-silicotools, including CT-based and Finite Element (FE) methods, to define parameters that may influence stemless implant design. The density and morphology of the proximal humerus were assessed using CT-derived point clouds of the trabecular-canal. Bone density was found to diminish 15-20mm beneath the humeral head resection and was greater peripherally. The depth, path and bounding diameters of the proximal trabecular-canal were also quantified and established the spatial constraints in which implants should be designed. To address the lack of consensus regarding the FE modelling of humeral trabecular- stiffness, eight (8) FE models were constructed then duplicated six different trabecular- stiffness relationships. The deviation induced in FE outcomes by stiffness relationship selection was quantified. It was determined that inhomogeneous stiffness definition is important; however, the anatomic site from which the stiffness is defined induced minor deviations in the implant-bone contact area, the change in bone stresses and the potential bone response following stemless reconstruction. Finally, with humeral FE modelling parameters defined, a series of ten generic stemless implants were designed with fixation features that were primarily central, peripheral or boundary-crossing. A population of five (5) cadaveric humeral FE models were constructed for each implant. Tradeoffs were found, with central implants producing the least resorbing potential, and peripheral implants maintaining the most implant-bone contact. Regardless of fixation feature design, predicted bone changes were most prominent within the lateral quadrant of the humerus, directly beneath the humeral head resection. The present work advances the understanding of stemless humeral arthroplasty. The morphological parameters defined provide a spatial definition of the region in which stemless implants function. Through the development of humeral FE models, general trends in bone response following stemless reconstruction were noted; along with tradeoffs regarding the placement of stemless fixation features. These methods can be applied in the design of future stemless implants.