Correlations of magnetic resonance imaging findings with clinical symptom severity and prognosis of frozen shoulder

Purpose: To evaluate the correlation between indirect magnetic resonance (MR) arthrographic imaging findings and the clinical symptoms and prognosis of patients with frozen shoulder. Methods: Indirect MR arthrography was performed for 52 patients with primary frozen shoulder (mean age 55.1 ± 9.0 years) and 52 individuals without frozen shoulder (mean age 53.1 ± 10.7 years); capsular thickening and enhancement of the axillary recess as well as soft tissue thickening of the rotator interval were evaluated. Clinical symptom severity was assessed using the Visual Analogue Scale for Pain (VAS Pain), simple shoulder test (SST), Constant score, American Shoulder and Elbow Surgeons (ASES) score, and range of motion (ROM). At 6-month follow-up, we evaluated whether MR arthrography findings correlated with the clinical symptoms and prognosis. Results: Capsular thickening and enhancement of the axillary recess as well as soft tissue thickening of the rotator interval were significantly greater in the patient group than in the controls (p < 0.001). Capsular thickening of the axillary recess did not correlate with clinical symptoms or ROM (n.s.); however, capsular enhancement correlated with clinical symptom severity according to VAS Pain (p = 0.005), SST (p = 0.046), and ASES scores (p = 0.009). Soft tissue thickening of the rotator interval did not correlate with clinical symptom severity, but was associated with external rotation limitation (p = 0.002). However, none of the parameters correlated with clinical symptoms at 6-month follow-up. Conclusions: Indirect MR arthrography provided ancillary findings, especially with capsular enhancement, for evaluating clinical symptom severity of frozen shoulder, but did not reflect the prognosis. MR findings in frozen shoulder should not replace clinical judgments regarding further prognosis and treatment decisions. Level of evidence: IV.