Accuracy of magnetic resonance imaging in detecting biceps pathology in patients with rotator cuff disorders: comparison with arthroscopy

Abstract

There is limited information on the validity of magnetic resonance imaging (MRI) in detection of biceps disease. The purpose of this study was to examine the measurement properties of noncontrasted MRI in diagnosis of biceps disease using arthroscopic surgery as the "gold standard." Prospectively collected surgical data of patients with impingement syndrome or rotator cuff tear, with biceps disease (study group) or without biceps disease (control group), were reviewed. MRI reports of radiologists with fellowship training in musculoskeletal imaging were retrospectively reviewed and compared with surgical findings. Data of 183 (130 study and 53 control) patients (73 women [40%], 110 men [60%]; mean age, 62 years [standard deviation, 9]) who had undergone arthroscopic rotator cuff-related surgery during a period of 11 years were used for analysis. Sensitivity and specificity of MRI for detection of full tears of the biceps tendon were 0.54 and 0.98, respectively. Sensitivity and specificity were 0.27 and 0.86 for partial tears of the biceps tendon, respectively. For biceps subluxation or dislocation, sensitivity was 1.00 and specificity was 0.83. The areas under the receiver operating characteristic curves, which quantify the overall accuracy of the tests, were 0.57, 0.75, and 0.92 for partial tear, full tear, and instability of the biceps tendon, respectively. Noncontrasted MRI has a low sensitivity and high specificity for detection of full-thickness tears of the biceps tendon. It is highly sensitive for diagnosis of instability of the long head of the biceps. However, its usefulness for diagnosis of partial tears of the biceps tendon remains limited.