Evaluation of the Rotator Cuff Tearwith Magnetic Resonance Arthrography

INTRODUCTION: Rotator cuff tear is one of the common shoulder disorders and surgical repair is very effective for the improvement of shoulder functions. Current surgical technique for rotator cuff tear is an arthroscopical repair or a mini-open repair. Selection of surgical technique should be determined by the appropriate patient’s selection based on the pathology of the rotator cuff tear. Magnetic resonance imaging (MRI) has become a useful imaging modality for establishing the diagnosis, confirming or defining pathology and thereby, directing treatment of rotator cuff tears. However, studies in the literature have reported 89% of sensitivity for the evaluation of the rotator cuff tears with MRI and attention has been focused on optimizing scanning protocols to increase accuracy. Magnetic resonance arthrography (MRA), which is the intraarticular introduction of dilute gadolinium diethylenetriamine pentaacetic acid (Gd-DTPA) before MR imaging, has been used to significantly improve visualization of the anterior glenoid labrum and rotator cuff lesions. However, there is little known about the correct comparison between MRI and MRA for the evaluation of the rotator cuff lesions. Computerized image analysis system is reported to be helpful for the accurate evaluation of MR imaging. The goal of our study is to elucidate the advantage of the shoulder MRA by comparing the rotator cuff lesions among MRI and MRA and surgical findings with computerized image analysis system. MATERIALS AND METHODS: Patients: We retrospectively evaluated MRI and MRA of 14 shoulders in 14 consecutive patients with rotator cuff tears that were surgically treated. There were 8 male and 6 female who had a mean age of 61.4 years. (range, 31-73) MR Imaging: With use of a 1.5 tesla MRI (Siemens Vision 1.5T) spin-echo fatsuppressed T1-weighted images were made with or without injection of 10ml of saline solution and 10ml of radiographically contrast medium (Isovist) containing 1% Gd-DTPA into the glenohumeral joint. Computerized measurements on MR Images: Pathology and size of rotator cuff tears were analyzed with MRI and MRA images. Concerning pathological changes, we classified several types of the edge of torn rotator cuff with MRA images. Concerning rotator cuff tear size, we assessed the following two parameters with both of MRI and MRA images. (1) The maximum transverse size and (2) the maximum longitudinal size of each tear were measured by oblique sagittal images and oblique coronal images respectively with use of image analyzing soft wear (NIH image) (Fig.1). All of the measurements were performed by two of us in a blinded fashion and the average measurements were used as data. We analyzed the correlation of each computerized measurement and each actual measurement during surgical procedures. Statistical analyses were performed using paired t-test. Significance was set at the 5% level.