A novel magnetic resonance imaging-based classification of subscapularis muscle atrophy: reproducibility and clinical relevance.

Background: Muscle atrophy and fatty degeneration of rotator cuff are critical factors influencing surgical indications for rotator-cuff tears. However, no widely accepted method exists for the evaluation of subscapularis (SSC) atrophy. Based on our previous clinical experience, we hypothesized that SSC atrophy extends from the cranial side and proposed a novel magnetic resonance imaging-based classification of SSC atrophy. This study aimed to investigate the validity of the proposed classification and its relationship with clinical parameters. Methods: In total, 70 patients (72 shoulders; mean age, 62.3 ± 11.5 years) who underwent arthroscopic rotator-cuff repair or reverse shoulder arthroplasty were included. SSC atrophy was evaluated preoperatively on T1-weighted oblique sagittal images in the Y-shaped view, in which the lateral scapular spine and coracoid base were visualized. Atrophy was classified as follows: none (superior margin of the SSC extends beyond the anteroinferior edge of the coracoid base), mild (superior margin tapers below this edge), moderate (anteroposterior width is reduced at the scapular body-spine intersection), or severe (anteroposterior width is reduced at the inferior tip of the scapular body). Reproducibility was assessed by 2 shoulder surgeons using intraclass correlation coefficients (ICCs). To evaluate clinical relevance, correlations between this classification and the SSC-tear size, fatty degeneration, and internal rotation strength were examined. In addition, sensitivity and specificity to detect the presence of a comma sign (retracted SSC tendon stump) were calculated using this classification. Results: The classification was applicable to all cases. Intrarater and inter-rater reliability was excellent (ICC [1, 2] = 0.89, ICC [2, 1] = 0.85, respectively). Higher atrophy grades of SSC was significantly correlated with greater tear size, higher fatty degeneration grade, and lower internal rotation strength (Spearman ρ = 0.65, 0.64 and -0.42, respectively, Discussion: This classification, which was based on the craniocaudal extension of SSC atrophy, utilizes clear bony landmarks that enable a high reproducible and simple evaluation. Its significant associations with tear size, fatty degeneration, and internal rotation strength support its value as a clinically relevant indicator. Moreover, its ability to predict the presence of the comma sign suggests that the classification reflects both muscle atrophy and tendon retraction. Conclusion: The proposed magnetic resonance imaging-based classification of SSC atrophy demonstrated high reproducibility and validity, effectively reflecting tear size, tendon retraction, and internal rotation strength and thereby establishing its clinical relevance.