Christoph Schaeffeler MD, Presenter: Nothing to Disclose

Simone Waldt MD, Abstract Co-Author: Nothing to Disclose

Konstantin Holzapfel MD, Abstract Co-Author: Nothing to Disclose

Chlodwig Kirchhoff, Abstract Co-Author: Nothing to Disclose

Pia Jungmann MD, Abstract Co-Author: Nothing to Disclose

Petra Wolf, Abstract Co-Author: Nothing to Disclose

Michael Schroeder, Abstract Co-Author: Nothing to Disclose

Ernst J. Rummeny MD, Abstract Co-Author: Nothing to Disclose

Andreas Imhoff, Abstract Co-Author: Nothing to Disclose

Klaus Woertler MD, Abstract Co-Author: Nothing to Disclose

To retrospectively determine the accuracy of MR arthrography (MR-A) of the shoulder in the detection of lesions of the reflection pulley of the long head of the biceps tendon (LHBT) using arthroscopy as reference standard.

80 patients who underwent MR-A prior to arthroscopy with detailed evaluation of the pulley sling were included to the study. Arthrograms were independently evaluated by three experienced musculoskeletal radiologists, who were unfamiliar with the cases and unaware of the arthroscopic results. Diagnostic criteria were: anterior or caudal dislocation of the LHBT in relation to the subscapularis (SSC) tendon on sagittal images, medial subluxation or dislocation of the LHBT on transverse images. The visibility of the superior glenohumeral ligament (SGHL), the presence of tendinosis, and rotator cuff tears adjacent to the rotator interval were evaluated.

At arthroscopy 28 pulley lesions were found. MR-A showed an overall sensitivity for the detection of a pulley lesion of 89.3%, 82.1%, and 85.7% for reader 1, reader 2, and reader 3 with a specificity of 96.2%, 86.5%, and 98.1% and an accuracy of 93.8%, 85.0%, and 93.8%. Multirater κ revealed substantial interrater agreement (0.74). Classification-and-regression-tree (CART) analysis showed the caudal dislocation of the LHBT on sagittal images to be the strongest criterion in predicting a pulley lesion for all three readers (p<0.001, κ=0.729) with no significant improvement by other variables. The sensitivities were 85.7%, 75.0%, and 82.1% , the specificities 96.2%, 96.0%, and 98.1% and the accuracies 92.5%, 86.3%, and 92.5%. On MR-A an uninterrupted SGHL was visible in 43/52 (82.7%), 39/52 (75.0%), and 41/52 (78.9%) of arthroscopically proven intact pulley slings.

MR arthrography of the shoulder shows high sensitivity, specificity, and accuracy in the detection of pulley lesions. The caudal dislocation of the LHBT in relation to the SSC tendon is the most valuable tool in predicting a pulley lesion with substantial interrater agreement.

MR arthrography of the shoulder is efficient in the evaluation of the reflection pulley of the LHBT. The position of the LHBT on sagittal images is the key to identify even isolated pulley lesions.

Schaeffeler, C, Waldt, S, Holzapfel, K, Kirchhoff, C, Jungmann, P, Wolf, P, Schroeder, M, Rummeny, E, Imhoff, A, Woertler, K, Lesions to the Reflection Pulley of the Long Head of Biceps Tendon: Diagnostic Accuracy of MR Arthrography of the Shoulder in 80 Patients with Arthroscopic Correlation.  Radiological Society of North America 2011 Scientific Assembly and Annual Meeting, November 26 - December 2, 2011 ,Chicago IL. http://archive.rsna.org/2011/11004242.html