Shagran Mohammed Binkhamis MBBS, FRCPC, Presenter: Nothing to Disclose

Mark E. Schweitzer MD, Abstract Co-Author: Consultant, Paradigm Spine LLC Consultant, Pfizer Inc

Paul E. Beaule MD, Abstract Co-Author: Consultant, Wright Medical Group, Inc, Memphis, TN Consultant, GETINGE AB, Ontario, Canada

Robert James Nairn MBBS, Abstract Co-Author: Nothing to Disclose

Kawan S. Rakhra MD, Abstract Co-Author: Nothing to Disclose

Many authors believe that direct MR arthrography is required to evaluate for internal derangements of the hip. However, since indirect MR arthrography (I-MRA) is more convenient and less invasive than direct arthrography, we sought to determine the accuracy of hip I-MRA in detecting acetabular labral tears.  

Eighteen 1.5T hip I-MRAs which had surgical correlation were retrospectively evaluated. Intravenous gadolinium (0.1 mmol/kg) was administered followed by 20 minutes of gentle exercise prior to imaging. The imaging protocol included oblique sagittal, oblique coronal and oblique axial fat-saturated T1 weighted sequences, and oblique axial fat-saturated proton density weighted images. The I-MRAs were evaluated by 3 independent, blinded musculoskeletal radiologists. The labrum was divided into three zones, for a total of fifty-four labral zones [Anterior (2-5 o'clock), Superior (10-2 o'clock) and Posterior (7-10 o'clock)]. Labral abnormalities were characterized as intrasubstance, substance tears or detachments. The radiologists’ interpretations were correlated with the intraoperative findings.

A total of 18 labral tears were diagnosed at surgery. The tears were distributed with 9 in the anterior zone, 8 in the superior zone and 1 in the posterior zone. For all three zones combined, the sensitivity, specificity, PPV, NPV and accuracy of indirect MRA for detecting labral tears for the three readers ranged from 50.0-55.6%, 55.6-61.1%, 36.0-39.1%, 29.0-31.0% and 53.7-57.4% respectively. The kappa correlation values between the readers’ interpretations and the surgical findings were 0.10, 0.05, and 0.10 (poor agreement). The interclass correlation value for all three readers was 0.28 (fair agreement).  

Indirect MR arthrography at 1.5T has moderate accuracy and fair interobserver agreement, but poor correlation with surgical assessment in diagnosis of hip labral tears.  

We do not feel at this time that indirect MR arthrography of the hip at 1.5T can be used routinely clinically due to the variability in interpretations between radiologists and surgeons.

Binkhamis, S, Schweitzer, M, Beaule, P, Nairn, R, Rakhra, K, The Accuracy of Indirect MR Arthrography of the Hip in Detecting Acetabular Labral Tears.  Radiological Society of North America 2011 Scientific Assembly and Annual Meeting, November 26 - December 2, 2011 ,Chicago IL. http://archive.rsna.org/2011/11034223.html