MR imaging of rotator cuff tears: is there a diagnostic benefit of shoulder exercise prior to imaging

The purpose of this study was to determine if shoulder exercise prior to MR imaging accentuates findings related to rotator cuff tears. In 32 consecutive patients undergoing MR imaging to evaluate the rotator cuff, after routine MR examination, the joint was moved by active or passive exercise (circumduction, including abduction, if possible) in an attempt to redistribute any joint fluid. The exercise was performed according to pain tolerance and for no longer than 4 minutes. The coronal oblique fast spin-echo T2-weighted images of these patients performed before exercise were reviewed by consensus agreement of two musculoskeletal radiologists who were blinded to clinical information. The appearance of the rotator cuff tendons and the distribution of fluid in the glenohumeral joint were determined. The identical postexercise MR images then were placed alongside the corresponding preexercise MR images, and a direct comparison of findings was made with regard to any change in the appearance of the rotator cuff or joint fluid by consensus opinion of the same two radiologists. Five patients (five shoulders) could not perform exercise because of pain. In the remaining 27 patients (27 shoulders), changes in the location of joint fluid were seen when the preexercise and postexercise images were reviewed together, the diagnosis of partial rotator cuff tear (n = 8) was changed to normal in two cases, and the diagnosis of partial tear was made with more confidence in one case. The diagnoses of normal rotator cuff (n = 16) and complete rotator cuff tear (n = 3) were unchanged. Eight patients had arthroscopy; in each of these, the preexercise and postexercise images showed similar results, and proved to be correct surgically (six normal, one partial rotator cuff tear, and one complete rotator cuff tear). Although postexercise MR images show changes in the distribution of joint fluid when compared to preexercise images, the diagnostic benefits of the postexercise images in the analysis of the rotator cuff appear to be limited.