Instability of the long head of the biceps tendon in patients with rotator cuff tear: evaluation on magnetic resonance arthrography of the shoulder with arthroscopic correlation

Objectives: To evaluate the diagnostic value of MR arthrography (MRA) in diagnosing instability of the LHBT in patients with rotator cuff tendon tear. Materials and methods: The MR arthrograms of 101 patients were retrospectively reviewed and correlated with arthroscopic findings as the standard reference. Images were evaluated for (1) the integrity of the LHBT, (2) the position of the LHBT (subluxation/dislocation on axial images, inferior displacement on oblique sagittal image) and (3) the integrity of the biceps pulley (SGHL, supraspinatus and subscapularis tendon adjacent to the rotator interval). Results: The integrity of the LHBT was correctly classified in 74.3% (75/101) and 66.3% (67/101) by readers 1 and 2, respectively. The diagnosis of LHBT instability could be made on axial images with a sensitivity of 82.6% and 73.9% and specificity of 69.9% and 87.7%, whereas the displacement sign on sagittal images had a sensitivity of 73.9% and 78.3% and a specificity of 64.4% and 61.6%, respectively. Assessing the integrity of the SGHL had a sensitivity of 60.9 and 93.3% and a specificity of 70.4 and 75.0%, respectively. By combining the different image findings, the accuracy in assessing LHBT instability was 80.9 and 90.5% with a sensitivity of 60.9 and 86.7% and specificity of 83.1 and 91.8%, respectively. Conclusion: Individual image findings may have a limited role in diagnosing LHBT instability in patients with rotator cuff tendon tear. The accuracy of MRA may be improved by assessing the integrity of the biceps pulley structures along with the position of the LHBT on both axial and sagittal images.