The role of center of rotation on subscapularis biomechanics in reverse shoulder arthroplasty: a simulation study

Abstract

BackgroundReverse shoulder arthroplasty (RSA) has been associated with poor internal rotation (IR) postoperatively. While studies have demonstrated lateralized glenospheres to achieve greater impingement-free range of motion (ROM), the underlying biomechanics driving functional IR loss are unknown, with conflicting results in studies that have attempted to understand the role of the subscapularis in the setting of RSA. Therefore, the purpose of this study is to use a validated computational musculoskeletal model of the shoulder to examine the influence of glenosphere lateralization and superoinferior placement on subscapularis strength, and more specifically, IR torque. MethodsSimulations were performed using an existing computational shoulder model. To simulate center of rotation (COR) placement in common RSA constructs, glenohumeral COR and humeral head position were translated medially from the native COR position −10 mm (lateralized glenoid), −20 mm, and −30 mm (medialized glenoid) along the scapular spine axis. In addition, COR was shifted 10 mm superior and inferior to native COR to assess potential impact of superoinferior placement of the glenosphere. Subscapularis IR torque, moment arm, force-generating capacity, and muscle-tendon length were computed over −40° to 40° shoulder axial rotation (IR = positive) in neutral abduction. IR torque is the product of moment arm and force. ResultsSubscapularis IR torque after simulated RSA decreased with IR, as in a native shoulder. IR torque increased with an increasingly lateralized glenosphere placement. Subscapularis moment arm was minimally affected by lateralization, as peak moment arm was only 0.7 mm larger for COR −30 mm compared to the intact shoulder. Subscapularis muscle-tendon force-generating capacity decreased markedly with more medial COR; −10 mm, −20 mm, and −30 mm COR positions exhibited a decrease in mean force of 144N, 412N, and 565N from native, respectively. Subscapularis muscle-tendon length decreased with medialized COR, most notably with increased IR. For superoinferior COR translation, subscapularis IR torque and force-generating capacity decreased with 10 mm inferior translation relative to native COR, while the same metrics increased for superior COR translation. DiscussionThese simulations suggest that medializing the glenosphere contributes to reduced subscapularis IR torque, primarily due to muscle-tendon slackening in accordance with muscle force-length behavior. These results support use of a lateralized glenosphere to improve subscapularis strength, and thus potentially increase functional IR after RSA.