Preoperative partial-thickness rotator cuff tears do not compromise anatomic total shoulder replacement outcomes: medium-term follow-up

Abstract

Reverse total shoulder replacement (TSR) in elderly patients with primary osteoarthritis (OA) and rotator cuff pathology is increasingly being performed. The purpose of our study was to determine the medium-term results of anatomic TSRfor OA in patients with established preoperative partial-thickness rotator cuff tears on magnetic resonance imaging (MRI) scans. We reviewed a cohort of patients who had undergone anatomicTSR for OA with a preoperative MRI diagnosis of partial-thickness rotator cuff tear. Patients were assessed with preoperative and post operative Oxford Shoulder Scores, evaluation of their range-of-movement and clinical rotator cuff assessment. Anteroposterior and axillary radiographs were used to assess for any proximal humeral migration (using the Torchia classification) and any evidence of loosening. The Lazarus score was used to grade glenoid radiolucencies. The study comprised 36 patients (14 men and 22 women) who underwent TSR and had partial-thickness rotator cuff tears on MRI; preoperatively, all showed mild to moderate fatty infiltration. The mean age of the patients was 79.2 years (range, 75-88 years); the mean follow-up period was 5.8 years (range, 5-9 years). Significant improvements in pain and range of movement were reported in all cases. At the final follow-up, the mean Oxford Shoulder Score was 42 points (range, 32-46 points), with a minimum improvement of 14 points (P = .001). External rotation (20° vs. 40°, P = .001), forward flexion (80° vs. 140°, P = .015), abduction (45° vs. 90°, P = .015), and internal rotation also improved. Lucencies were observed in 8 glenoids, with 6 showing grade 1 Lazarus changes, 2 showing grade 2, and none showing grade 3. There were no cases of implant loosening. Clinically, 4 patients had rotator cuff weakness but only 2 showed evidence of proximal migration. One patient remained satisfied, whereas the other patient, with moderate-grade proximal migration according to the Torchia classification, underwent revision for rotator cuff failure; one further patient underwent washout and DAIR (débridement, antibiotics, and implant retention) for infection. There is a paucity of literature on whether a preoperative partial-thickness rotator cuff tear has an adverse effect on the outcome of TSR. Our results show that the presence of a partial cuff tear on preoperative MRI does not significantly affect function after anatomic TSR in the medium term. With anatomic TSR having less morbidity for patients and allowing greater potential options for revision, we believe that the use of reverse shoulder arthroplasty in this cohortof patients, with partial rotator cuff tears, may not be necessary and we advocate consideration of anatomic TSR in this patient group.