Structural severity outweighs laboratory biomarkers in predicting short-term failure under nonoperative management for partial-thickness supraspinatus tears: a prospective cohort study.

BACKGROUND: Nonoperative treatment, including platelet-rich plasma injection is increasingly used for partial-thickness rotator cuff tears, yet predictors of treatment failure and the incremental prognostic value of laboratory biomarkers remain unclear. We aimed to identify predictors of short-term treatment failure under nonoperative management and to evaluate whether inflammatory and metabolic biomarkers improve risk stratification beyond clinical and imaging variables. METHODS: We conducted a prospective cohort study at Hai Phong International Hospital from 2020 to 2025 including patients with magnetic resonance imaging–confirmed partial-thickness supraspinatus tears. Patients received either platelet-rich plasma injection or conservative treatment and were followed for 6 months. Clinical failure was defined as less than 30% improvement in pain (Visual Analog Scale) or function (Quick Disabilities of the Arm, Shoulder and Hand score). Multivariable logistic regression models were developed in the overall nonoperative cohort to identify independent predictors of failure. Model discrimination was assessed using the area under the receiver operating characteristic curve, and calibration was evaluated using calibration slope and intercept. RESULTS: Of 2,502 screened patients, 939 (284 platelet-rich plasma; 655 conservative treatment) completed follow-up and were included in analyses. In the analytic cohort, failure occurred in 33.8% of platelet-rich plasma–treated patients and 64.9% of conservatively managed patients, although continuous improvements were broadly comparable between groups.In adjusted analyses, baseline tear size was the strongest and most consistent predictor of clinical failure, with approximately threefold higher odds per 1-SD increase. Inflammatory and metabolic biomarkers did not show independent associations after multivariable adjustment. The clinical–imaging model demonstrated acceptable discrimination (area under the curve 0.721, 95% confidence interval 0.687–0.755) and good calibration (slope 0.993; intercept 0.001). Addition of inflammatory and metabolic biomarkers did not meaningfully improve discrimination or calibration. CONCLUSIONS: Structural severity, particularly tear size, was the strongest predictor of short-term clinical failure in partial-thickness supraspinatus tears managed nonoperatively. Routine inflammatory and metabolic biomarkers provided limited incremental prognostic value beyond clinical and imaging variables. These findings are based on internal validation and require external validation before broader clinical application.