Seon-Kwan Juhng MD, Presenter: Nothing to Disclose

Gang-Deuk Kim MD, Abstract Co-Author: Nothing to Disclose

Hye-Won Kim MD, Abstract Co-Author: Nothing to Disclose

Suhng-Ah Lee MD, Abstract Co-Author: Nothing to Disclose

Hyoung-Joon Kim MD, Abstract Co-Author: Nothing to Disclose

To determine the accuracy of CT arthrography(CTA) and MR arthrography(MRA) in the evaluation of glenoid labral and rotator cuff lesions of the shoulder.

CTA using 16 slice multidetector CT and MRA were performed in 112 patients from 2004. 5. to 2005. 12. Forty patients underwent arthroscopy. Prospective analysis of findings on CTA and MRA were done by three musculoskeletal radiologists under consensus with evaluation of the SLAP lesions, Bankart lesions(including bony Bankart lesions), Hill-Sachs defects and rotator cuff lesions. Type I SLAP lesions were regarded as negative as they most often are not clinically relevant. The results were compared with arthroscopic findings. Sensitivity, specificity and accuracy were determined for detection of each items.

At arthroscopy, there were 31 SLAP lesions(7 type I, 23 type II, 1 type III), 6 Bankart lesions, 4 Hill-Sachs lesions, 26 rotator cuff lesions(18 full thickness and 8 partial thickness tears). On CTA, sensitivity, specificity and accuracy in detection of SLAP lesions were 83%, 100% and 90%, in detection of Bankart lesions and Hill-Sachs defects were 100%, 100%, 100%, and in depicting rotator cuff tears were 92%, 100%, 98%, and in depicting full thickness supraspinatus tendon tears were 100%, 93% and 95%, and in depicting partial thickness supraspinatus tendon tears were 33%, 100% and 90%. On MRA, sensitivity, specificity and accuracy in detection of SLAP lesions were 100%, 94% and 98%, in detection of Bankart lesions were 83%, 100%, and 98%, in detection of Hill-Sachs defects were 100%, 100%, 100%, and in depicting rotator cuff tears were 96%, 100%, 99%, and in depicting full thickness supraspinatus tendon tears were 100%, 93% and 95%, and in depicting partial thickness supraspinatus tendon tears were 50%, 100% and 93%.

CTA was better than MRA in depicting bankart lesions. MRA was superior to CTA in depicting SLAP lesions and partial thickness rotator cuff tears. CTA was equally competent to MRA in demonstrating Hill-Sachs lesions and full thickness rotator cuff tears.

Sixteen slice CTA is at least equal to MRA in the evaluation of glenoid labral and rotator cuff lesions of the shoulder.

Juhng, S, Kim, G, Kim, H, Lee, S, Kim, H, Accuracy for Detection of Glenoid Labral and RC Tears: Prospective Evaluation with CT Arthrography and MR Arthrography.  Radiological Society of North America 2006 Scientific Assembly and Annual Meeting, November 26 - December 1, 2006 ,Chicago IL. http://archive.rsna.org/2006/4436757.html