Wrist ligament tears: Evaluation of MRI and combined MDCT and MR arthrography

OBJECTIVE. The purpose of this study was to evaluate the performance of MRI and a combination of MDCT arthrography and MR arthrography in the diagnosis of tears and cartilage abnormalities of the wrist ligaments. SUBJECTS AND METHODS. The feasibility of combining MDCT arthrography and MR arthrography and performing them with an optimized contrast solution was evaluated in vitro and in vivo. Forty-five consecutively enrolled subjects with suspected wrist ligament tears underwent MRI and a combined MDCT and MR arthrographic procedure. Two observers reviewed the images for evidence of tears and cartilage abnormalities of the scapholunate and lunotriquetral ligaments and triangular fibrocartilaginous complex. Interobserver agreement was determined with kappa statistics, and the diagnostic accuracy of each technique was calculated. RESULTS. A 1: 1 solution of 2.5 mmol/ L tetraazacyclododecanetetraacetic acid ( DOTA)- gadolinium and 300 mg I/mL iopamidol provided adequate contrast enhancement for both in vitro and in vivo MDCT arthrographic and MR arthrographic images. Interobserver agreement was substantial for MRI (kappa = 0.61) and MR arthrography (kappa = 0.71) and almost perfect for MDCT arthrography (kappa = 0.93). The sensitivity and specificity of MRI, MDCT arthrography, and MR arthrography for tears of the scapholunate ligament were 59% and 70%, 95% and 96%, and 68% and 87% for the first observer and 77% and 83%, 95% and 100%, and 77% and 87% for the second observer. For tears of the lunotriquetral ligament, these values were 30% and 94%, 100% and 94%, and 60% and 97% for the first observer and 50% and 97%, 90% and 100%, and 50% and 94% for the second observer. The three techniques appeared equivalent for complete tears of the scapholunate and lunotriquetral ligaments, but partial tears were significantly better visualized with MDCT arthrography. The sensitivity and specificity of MRI, MDCT arthrography, and MR arthrography for triangular fibrocartilaginous complex tears were 27% and 100%, 100% and 100%, and 82% and 100% for the first observer and 45% and 100%, 100% and 100%, and 82% and 100% for the second observer. For cartilage abnormalities, these values were 30% and 100%, 100% and 100%, and 30% and 100% for the first observer and 10% and 100%, 100% and 100%, and 40% and 100% for the second observer. CONCLUSION. MDCT arthrography appears more accurate than MRI and MR arthrography, particularly for discerning partial tears of the scapholunate and lunotriquetral ligaments that do not necessitate surgical therapy.