MR Evaluation of Factors Predicting the Development of Rotator Cuff Tears Jean-Pierre Tasu, Anne Miquel, Laurence Rocher, Vale ´rie Molina, Olivier Gagey, and Michel Bléry Purpose: The purpose of this work was to assess the incidence of various factors predicting impingement in the shoulder. Method: MR examination was used to assess various anatomic parameters in 46 patients with clinical abnormalities of the shoulder and 40 asymptomatic volunteers. All patients had undergone surgery, showing no rotator cuff tear. Results: A few of the parameters evaluated were significantly different in the two groups: The acromion was more frequently curved or hook-shaped in patients than in volunteers. The acromiohumeral distances differed (4.87 mm in patients and 6.05 mm in volunteers); so did the coracohumeral distances (7.9 mm in patients and 8.9 mm in volunteers). The shape of the acromioclavicular joint and the anterior covering of the humerus, defined on the coronal view, also differed (1.07 mm in patients and 1.49 mm in volunteers). Only the shapes of the acromioclavicular joint and of the acromion were linked with age. Conclusion: All these factors reflected a decrease in the acromiohumeral space, except for the anterior covering of the acromial arch, which could be due to anterior instability. Index Terms: Shoulder—Magnetic resonance imaging—Rotator cuff tear. The etiology and relative importance of a number of factors in the pathogenesis of rotator cuff tears are still debated. However, it is commonly accepted that entrap- ment of the supraspinatus tendon between the humeral head and the coracoacromial arch (the anterior portion of the acromion, the coracoacromial ligament, and the ac- romioclavicular joint) can narrow the subacromial space, leading to damage to the rotator cuff (1,2). Predisposing factors described include a subacromial spur, hypertro- phy of the acromioclavicular joint, thickening of the gle- nohumeral joint, anatomic abnormalities of the acromi- on, os acromiale, and biceps tenosynovitis (2). Accord- ingly, a subcoracoid impingement has been defined as the result of a decrease in the volume of the coracohu- meral compartment due to reduced space or occasionally an increase in its contents (3,4). However, these anatomic variations alone do not seem fully to explain all impingement syndromes. A recent study (5) reported that the incidence of hooked acromi- ons in a symptomatic group of patients was lower than in an asymptomatic group. In a young population, impinge- ment could be due to instability or overuse, but this un- likely to be so in an older population. We have therefore postulated that there are anatomic factors in addition to those already described that may lead to symptomatic impingement. The aim of this study was to identify them. We also compared the incidence of various MR ana- tomic features of the shoulder in patients showing clini- cal signs of mechanical impingement with their inci- dence in volunteers showing no sign of rotator cuff dis- ease. All the computed or measured indexes had been correlated with age. METHODS Subjects A total of 356 MR examinations of the shoulder were performed at our center between 1996 and 1998. We excluded 15 patients, the quality of whose images was inappropriate for this study (4%). We identified a group of 122 patients who had had surgery for chronic shoulder impingement syndrome. From this group of patients, we excluded 76 with full rotator cuff tear to avoid confusion with acute traumatic shoulder lesions and because tears of the rotator cuff have been previously shown to alter the anatomy of other structures around the shoulder (6). From the Departments of Radiology (J. P. Tasu, A. Miquel, L. Rocher, and M. Bléry) and Orthopedic Surgery (V. Molina and O. Gagey), Bicêtre Hospital, Le Kremlin-Bicêtre, France. Address corre- spondence and reprint requests to Dr. J. P. Tasu at Service de Radi- ologie, Hôpital de Bicêtre, 78 rue du Général Leclerc, 94275 Le Krem- lin-Bicêtre Cedex, France. E-mail: jean-pierre.tasu@bct.ap-hop-paris.fr Journal of Computer Assisted Tomography 25(2):159–163 © 2001 Lippincott Williams & Wilkins, Inc., Philadelphia 159