Correlation of defect size with fatty infiltration on MRIs of rotator cuff tendinopathy.

Introduction: Sizable rotator cuff defects with limited muscle atrophy and fat replacement may represent acute traumatic ruptures that are less likely to recur after surgery to close the defect, while closure of defects with poor quality muscle are associated with defect recurrence. These distinct lesions are both referred to as tears. We analyzed the relationship between rotator cuff defect size and muscle quality to determine the relative proportion of sizable defects associated with good muscle and factors associated with muscle deterioration. Material and methods: A cohort of 230 consecutive shoulder MRIs in patients with full-thickness rotator cuff tendon defects, limited acromioclavicular arthrosis (to avoid hindrance of defect measurement), and a duration of symptoms in the radiology report from a large urban center in the United States was evaluated for the measured distance between the supraspinatus tendon edge and the greater tuberosity medial to lateral (coronal plane defect size), anterior to posterior (sagittal plane defect width), and fatty infiltration (Goutallier classification), and atrophy (Warner classification) of the supraspinatus. We sought factors independently associated with fatty infiltration and muscle atrophy in multivariable logistic regression analyses. Results: Forty-nine of 109 shoulders (45%) with a coronal plane defect >20 mm had reasonable muscle quality as defined by Goutallier grade less than 2 and Warner grade less than 2. Both greater fatty infiltration of the supraspinatus muscle and greater supraspinatus muscle atrophy were associated with older age and greater coronal plane defect size. Conclusion: The observation that supraspinatus muscle health deteriorates with age and defect size, but nearly half of the largest defects had good muscle, suggests an important distinction between relatively recent traumatic ruptures and old untreated rupture or gradual attrition that may be obfuscated by referring to all lesions as tears. Level of evidence: Level III; Retrospective diagnostic cohort.