Pathological formation of subcoracoid bursa effusion on magnetic resonance imaging studies

The subcoracoid space includes the subcoracoid bursa (SCB) and subscapular bursa (SSB). We aimed to clarify the relationship between the presence or amount of SCB effusion and other structures around the SCB on magnetic resonance imaging (MRI) and to discuss the pathological formation of SCB effusion. Medical records and MR images of patients who had undergone MRI examinations for shoulder pain were retrospectively reviewed. The presence or absence of SCB, SSB, and subacromial-subdeltoid bursa (SASDB) effusion was evaluated. If SCB effusion was present, the largest diameter of the effusion was measured on sagittal images to represent the amount of SCB effusion. The presence or absence of communication between the SCB and SASDB or SSB effusion were also evaluated on sagittal and axial MRI. Eighty shoulders in 70 patients were analyzed. The mean age was 70.4 ± 10.4 (range, 50-87) years. Thirty-three of the 80 shoulders (41.3%) showed SCB effusion on MRI. The clinical diagnoses of these 33 shoulders were rotator cuff tear (RCT), n = 23; frozen shoulder, n = 6; subacromial impingement, n = 3; and calcific tendinopathy, n = 1. Multivariate logistic regression analysis showed that RCT (P = 0.015) and SSB effusion (P = 0.036) were significantly associated with the presence of SCB effusion, but SASDB effusion was not. In shoulders with RCT, the SCB communicated with the SASDB in 65.2%, and with the SSB in 4.3%. In other shoulders, the SCB communicated with the SASDB in 60.0%, and with the SSB in 40.0%. The rate of SCB-SSB communication was significantly higher in shoulders without RCT than in shoulders with RCT (P = 0.021). The largest diameter of SCB effusion was normally distributed in 33 shoulders (4.7–34.8 mm), and mean 19.6 ± 7.4 mm. The largest diameter of SCB effusion was 21.9 ± 6.3 mm in 23 shoulders with RCT, and 13.5 ± 6.8 mm in 10 other shoulders (P < 0.05). Multiple regression analysis showed that RCT (P = 0.002) and SSB effusion (P = 0.029) were significantly associated with the largest diameter of SCB effusion, but SASDB effusion was not. SCB effusion can be recognized and extended by inflow from SASDB effusion in RCT. Without RCT, SCB effusion may occasionally be visible due to inflow from SSB effusion.