Covariation between rotator cuff muscle quality and shoulder morphometric bony features in B-glenoids: a statistical modeling approach.

Abstract

Rotator cuff muscle (RCM) degeneration, bone morphology, and humeral head subluxation (HHS) are known risk factors for failure of anatomic total shoulder arthroplasty in patients with B-glenoid shoulder osteoarthritis. Yet, the understanding of RCM asymmetry in these patients remains an area of active investigation, including its relation with other risk factors. We therefore aimed to characterize the variability of RCM degeneration in B-glenoids and analyze its covariation with scapular morphology and HHS. First, computed tomography images were used to quantify 3D RCM degeneration, including muscle atrophy and fatty infiltration, in sixty B-glenoids referenced against twenty-five healthy controls. Next, the 3D scapular shape of B-glenoids was quantified using a previously published statistical shape model. Thirdly, 3D HHS was quantified. Using dedicated correlation analyses covariation patterns were modeled between each of these risk factors. Results indicated that RCM degeneration in B-glenoids is primarily characterized by fatty infiltration, without any sign of asymmetric impact on the anterior versus posterior RCM. However, B-glenoids with asymmetric bone loss were found to have more RCM atrophy and fatty infiltration of the infraspinatus. We identified four significant patterns of RCM degeneration and scapular shape, explaining 90.3% of their correlation. The primary mode indicates an association between combined posterior glenoid erosion and coracoid rotation with an increased infraspinatus' fatty infiltration. Interestingly, this mode was also positively correlated with posterior HHS (r = 0.46, P < 0.01). Identification of such patterns can improve the accuracy of musculoskeletal models in predicting postoperative implant failure risks.