Rachel Detreille, Abstract Co-Author: Nothing to Disclose

Marc Soudant, Abstract Co-Author: Nothing to Disclose

Toufik Batch MD, Abstract Co-Author: Nothing to Disclose

Marie-Rita Monzel, Abstract Co-Author: Nothing to Disclose

Francis Guillemin MD, Abstract Co-Author: Nothing to Disclose

Alain Gilbert Blum MD, Presenter: Nothing to Disclose

to determine the agreement of MR-arthrography (MR-A) and of CT-arthrography (CT-A) in the diagnosis of shoulder disorders in young adults compared to the combined reading of both techniques.

This study was performed on 108 consecutive patients aged from 18 to 40 y.o. referred for MR-A and CT-A of the shoulder with a single joint puncture and an injection of Artirem®  and Iodixanol 320 mg I/ml. Our institutional review board approved this study and informed consents were obtained. Two independent senior radiologists, one specialized in MSK (reader 1), the other with general practice (reader 2) read images independently. Their findings were compared to the combined reading of MR-A and CT-A by an 18 years experienced MSK radiologist (reader 3). Each individual lesion was quoted. The final diagnosis was classified in 6 categories: rotator cuff  lesions (RC), instability (I), SLAP, postero-superior impingement (PSI), neuropathy, other relevant anomalies, normal examination. PABAK were calculated between readers. The confidence in diagnosis of each reader was also evaluated for each patient and each technique using a VAS. An evaluation of procedure tolerance was obtained from each patient and for each technique.

107 patients were recruited (one exclusion). CT-A was better tolerated (mean VAS score, 0.4 +/-0.7) than the joint puncture (mean VAS score, 1.4 +/-1.5) and MR-A (mean VAS score, 2.1 +/-2.1). A combined reading highlighted more lesions than CT-A or MR-A alone. PABAK between readers 1 and 3 were respectively 0.85, 0.85, 0.89, 0.91, 0.42 and 0.63 for CT-A and 0.74, 0.81, 0.91, 0.91, 0.51, 0.70 for MR-A for RC, I, PSI, SLAP, other, normal. PABAK between readers 2 and 3 were respectively 0.64, 0.63, 0.85, 0.72, 0.23 and 0.59 for CT-A and 0.57, 0.76, 0.89, 0.78, 0.25, 0.79 for MR-A for RC, I, PSI, SLAP, other, normal. Confidence in diagnosis was higher with CT-A than with MR-A for reader 1 (p= 0.043) while equal for reader 2.  

In young adults, a combined reading highlights more lesions than CT-A or MR-A alone. CT-A is better tolerated that MR-A. For both techniques, the agreement is higher for the experienced MSK reader. For the diagnosis of the most frequent lesions, the agreement for the experienced MSK reader is good or excellent for both techniques

Experience improves the results for both CT-A and MR-A.

Detreille, R, Soudant, M, Batch, T, Monzel, M, Guillemin, F, Blum, A, Agreement between MR-Arthrography and CT-Arthrography Alone with Combined Techniques in the Diagnosis of Shoulder Disorders in Young Adults.  Radiological Society of North America 2009 Scientific Assembly and Annual Meeting, November 29 - December 4, 2009 ,Chicago IL. http://archive.rsna.org/2009/8015236.html