Evaluation of rotator cuff tears with magnetic resonance arthrography

The size and morphologic features of rotator cuff tears may influence treatment selection and affect final outcomes. Magnetic resonance arthrography allows observation of these features and other intraarticular structures. To assess the utility of magnetic resonance imaging in assessing size and morphologic features, we retrospectively reviewed observations on 41 shoulders in 37 consecutive surgically treated patients (mean age, 63.2 years) who had magnetic resonance imaging followed by magnetic resonance arthrography. The maximum rotator cuff defect size in the anteroposterior direction defined transverse size, and the maximum rotator cuff defect size in the mediolateral direction defined longitudinal size. Sensitivities for detecting full-thickness rotator cuff tears by magnetic resonance imaging and magnetic resonance arthrography were 90.2% and 100%, respectively. Maximum longitudinal and transverse dimensions of the tear as shown by magnetic resonance arthrography correlated better with intraoperative measurements (r(2) = 0.85 transversely, 0.92 longitudinally) than magnetic resonance imaging measurements (r(2) = 0.47 transversely, 0.26 longitudinally). The reproducibility of the two methods is similar. Magnetic resonance arthrography also allowed morphologic classification of the torn tendon as blunt end, tapering end, indistinct end, horizontal tear, and global tear. There was good agreement in classifying torn edges; the imaging findings agreed with findings at surgery. Magnetic resonance arthrography was more accurate in evaluating rotator cuff tear size and morphologic features than conventional magnetic resonance imaging. Level of Evidence: Diagnostic study, Level II (development of diagnostic criteria on basis of consecutive patients-with universally applied reference "gold" standard). See the Guidelines for Authors for a complete description of levels of evidence.