Inconsistencies in the MRI Evaluation of Supraspinatus Volume After Repair

BACKGROUND: Reversibility of rotator cuff atrophy after surgical repair is controversial. Traditionally, the cross-sectional area (CSA) of the rotator cuff was measured in conventional Y-view (CYV) via magnetic resonance imaging (MRI) to evaluate reversibility. However, it has been suggested that scanning axis inconsistency in CYV was overlooked and that the CSA in CYV reflects not only atrophy but also rotator cuff retraction. HYPOTHESIS: Discrepancies between scanning axes in CYV cause significant errors when one is evaluating changes in the CSA of the supraspinatus (SS) using preoperative and postoperative MRI scans. A more medial section than the Y-view is not influenced as much by retraction recovery after repair. STUDY DESIGN: Cohort study (diagnosis); Level of evidence, 3. METHODS: The study included 36 patients with full-thickness SS tear and retraction who underwent arthroscopic complete repair with preoperative MRI and immediate postoperative MRI (within 5 days after rotator cuff repair). Angles between CYV planes in the preoperative and immediate postoperative MRI scans were measured. MRI scans were reconstructed perpendicular to the scapular axes by multiplanar reconstruction. Differences between the CSAs of the SS in preoperative and postoperative Y-view on the original and reconstructed MRI scans were compared, and changes in CSAs of the SS muscles after repair in 2 sections medial to the reconstructed Y-view (RYV) were compared. RESULTS: The mean angle between CYV planes in preoperative and postoperative MRI scans was 13.1° ± 7.1°. Mean pre- to postoperative increase in the CSA of the SS was greater in CYV than in RYV (95 ± 72 vs 75 ± 62 mm(2); P = .024). Furthermore, pre- to postoperative CSA differences in the 2 medial sections were less than in RYV. For the most medial section, crossing the omohyoid origin, the CSA differences were not significant (434 ± 98 vs 448 ± 98 mm(2); P = .061). CONCLUSION: Scanning axes inconsistencies in CYV cause unacceptable errors in CSA measurements of the SS after repair. We recommend reconstruction along a consistent axis by multiplanar reconstruction when evaluating postoperative changes in SS atrophy and the use of sections more medial than the scapular Y-view to reduce errors caused by tendon retraction.