MRI severity of common extensor origin damage associated with prior glucocorticoid injection in chronic lateral epicondylitis: a retrospective multicenter analysis

Abstract Background Lateral epicondylitis (LE) is a degenerative tendinopathy of the common extensor origin (CEO). Although local glucocorticoid injections are widely used for symptom relief, concerns exist regarding their potential adverse effects on tendon structure. This study aimed to investigate the association between prior glucocorticoid injection and magnetic resonance imaging (MRI)-based structural CEO damage in patients with chronic LE. Methods In this retrospective, multicenter cross-sectional study, 147 patients with clinically diagnosed LE (symptom duration ≥ 6 months) who underwent elbow MRI between 2015 and 2023 were included. The primary exposure was prior local glucocorticoid injection at the CEO (yes/no and number of injections). Structural tendon damage was assessed using the Walz classification and quantitative lesion size diameters measured in coronal, axial, and sagittal MRI planes. Associations were analyzed using proportional odds ordinal logistic regression and generalized linear models, adjusted for age, sex, occupational workload, study center, and symptom duration. Results Sixty-three patients (42.9%) had received at least one prior glucocorticoid injection. Prior injection was independently associated with more severe MRI-based tendon damage. Patients with prior injection had higher odds of being classified in a higher Walz grade compared with non-injected patients (adjusted odds ratio (OR) 2.78, 95% confidence interval (CI) 1.30–5.97; p = .008). Longer symptom duration was also associated with higher Walz grades (adjusted OR 1.02 per month, 95% CI 1.00–1.04; p = .031). Among injected patients, the number of injections was not significantly associated with Walz grade. Prior injection was further associated with larger lesion size diameters: +2.07 mm in the coronal plane (95% CI 0.80–3.34; p = .001), + 1.49 mm in the axial plane (95% CI 0.21 − 2.76; p = .022), and a 94% larger sagittal diameter (ratio 1.94, 95% CI 1.15–3.29; p = .013). Conclusions Prior local glucocorticoid injection at the CEO was associated with more severe MRI-detected structural tendon abnormalities in patients with chronic LE. However, the retrospective cross-sectional design and potential for confounding by indication preclude causal attribution of these structural differences to glucocorticoid exposure. These findings provide imaging-based evidence supporting a cautious and selective approach to glucocorticoid injections.