Factors Affecting Late-Period Retear That Occurs More than 2 Years After Rotator Cuff Repair

Background: While the majority of retears of the rotator cuff occur in the early postoperative period, some tears occur several years after repair. The risk factors associated with late retears have not been studied. Purposes: To evaluate the frequency of retears in the late postoperative period after primary rotator cuff repair (RCR) compared with that in the early postoperative period and analyze various demographic and radiologic factors associated with late-period retears. Study Design: Case-control study; Level of evidence, 3. Methods: Among 2463 patients who underwent RCR from 2011 to 2017, 485 patients were diagnosed with rotator cuff retears using magnetic resonance imaging at a single institution, and 345 patients sustained a retear at ≤6 months (group E) or >2 years after repair (group L). To reduce the selection bias, propensity score matching (PSM) was performed using age, sex, and initial mediolateral (ML) tear size. The association between demographic factors, including smoking (non/current/ex), diabetes mellitus (DM, noncontrolled/uncontrolled), and radiologic factors, including initial tear size and late-period retear, were analyzed. Results: Group E had 314 (91%) and group L had 31 (9%) patients. Multiple regression analysis using the Firth method after PSM identified that controlled DM (odds ratio [OR], 4.42; 95% CI, 1.06-18.39; P = .04) and uncontrolled DM (OR, 8.81; 95% CI, 1.82-42.55; P = .007) were associated with late-period retear. Using the receiver operating characteristic curve, the group was divided into groups A (≤ cutoff) and B (> cutoff) based on the cutoff value (2.75 cm) of the ML size. Multiple regression analysis was performed in each group, revealing that current smoking (OR, 8.68; 95% CI, 1.35-55.67; P = .02) and uncontrolled DM (OR, 13.60; 95% CI, 1.88-98.46; P = .01) were significantly associated with late-period retear in group A. Conclusion: Although the incidence was lower than that in the early period after RCR, retears also occurred >2 years after RCR. Compared with early-period retears, DM was significantly associated with late-period retears. In addition, the risk of late-period retear was high among patients who were current smokers and those with uncontrolled DM in the presence of a small-sized ML tear.