Current musculoskeletal inverse dynamics shoulder models have two limitations to use in the context of nonconforming total shoulder arthroplasty (NC-TSA). First, the ball and socket glenohumeral (GH) joint simplification avoids any humeral head translations. Second, there is no contact at the GH joint to compute the contact area and the center of pressure (COP) between the two components of NC-TSA. In this paper, we adapted the AnyBody™ shoulder model by introducing humeral head translations and contact between the two components of an NC-TSA. Abduction in the scapular plane was considered. The main objective of this study was to adapt the AnyBody™ shoulder model to a NC-TSA context and to compare the results of our model (translations, COP, contact area, GH joint reaction forces (GH-JRFs), and muscular forces) with previous numerical, experimental, and clinical studies. Humeral head translations and contact were successfully introduced in our adapted shoulder model with strong support for our findings by previous studies.

Prosthetic components' mismatch and subscapularis (SC) tear are determining factors for glenoid failure complication in nonconforming total shoulder arthroplasty (NC-TSA). Risk factors are linked to glenoid prosthetic loading. However, the mechanisms underlying the clinical observations remain unclear. This study assessed the combined impact of mismatch and subscapularis tear on glenoid loading. It was assumed that adequate glenoid loading was associated with minimal, but non-null, humeral head translations and contact pressure, as well as with maximal glenoid contact area, and that the center of pressure (COP) on the glenoid would have a centered displacement pattern. A numerical model was used to achieve two objectives. The first was to verify whether an optimum mismatch existed, for which failure risk would be minimal. The second was to explore the effect of subscapularis tear on the position of applied forces on the glenoid. A shoulder AnyBody musculoskeletal model was adapted to the arthroplasty context by introducing humeral head translations and contact between implants. Ten simulations were computed to compare combinations of varying mismatches (1.4 mm, 3.4 mm, 6.4 mm, 8.6 mm, and 9 mm) with two shoulder conditions (intact-muscle or subscapularis tear). Humeral head translations, center-of-pressure, contact area, contact pressure, and glenohumeral joint contact forces were numerically estimated. Mismatches between 3.4 mm and 6.4 mm were associated with the most minimal humeral translations and contact pressure, as well as with maximal contact area. Center of pressure displacement pattern differed according to shoulder condition, with an outward anterior tendency in presence of tear.