Rotator cuff fatty infiltration and muscle atrophy: relation to glenoid deformity in primary glenohumeral osteoarthritis

Muscle atrophy (MA) and fatty infiltration (FI) are degenerative processes of the rotator cuff musculature that have incompletely understood relationships with the development of eccentric glenoid wear in the setting of primary glenohumeral osteoarthritis (GHOA). All patients with GHOA and an intact rotator cuff who underwent both magnetic resonance imaging and computed tomography scans of the affected shoulder prior to total shoulder arthroplasty between 2015 and 2020 were identified from a prospectively maintained registry. Rotator cuff MA was measured quantitatively on sequential sagittal magnetic resonance images, whereas FI was assessed on sagittal magnetic resonance imaging slices using the Goutallier classification. Preoperative computed tomography scans were reconstructed using automated 3-dimensional software to determine glenoid retroversion, glenoid inclination, and humeral head subluxation. Glenoid deformity was classified according to the Walch classification. Univariate and multivariable regression analyses were performed to characterize associations between age, sex, muscle area, FI, and glenoid morphology. Among the 127 included patients, significant associations were found between male sex and larger overall rotator cuff musculature (P < .01), increased ratio of the posterior rotator cuff (PRC) to the subscapularis area (P = .01), and glenoid retroversion (19° vs. 14°, P < .01). Larger supraspinatus and PRC muscle size was correlated with increased retroversion (r = 0.23 [P = .006] for supraspinatus and r = 0.25 [P = .004] for PRC) and humeral head subluxation (r = 0.25 [P = .004] for supraspinatus and r = 0.28 [P = .001] for PRC). The ratio of PRC muscle size to anterior rotator cuff muscle size was not associated with evidence of eccentric glenoid wear (P > .05). After we controlled for confounding factors, increasing glenoid retroversion was associated with high-grade infraspinatus FI (β, 6.8; 95% confidence interval, 2.9-10.7; P < .01) whereas larger PRC musculature was predictive of a Walch type B (vs. type A) glenoid (odds ratio, 1.3; 95% confidence interval, 1.0-1.5; P = .04). Patients with eccentric glenoid wear in the setting of primary GHOA and an intact rotator cuff appear to have both larger PRC musculature and higher rates of infraspinatus FI. Although the temporal and causal relationships of these associations remain ambiguous, MA and FI should be considered 2 discrete processes in the natural history of GHOA.