Comparison of 3-T MRI and Arthroscopy of Intrinsic Wrist Ligament and TFCC Tears

This study assesses the sensitivity and specificity of MRI at 3 T compared with arthroscopy for detection of intrinsic wrist ligament and triangular fibrocartilage complex (TFCC) tears. MR wrist images in 300 consecutive patients underwent consensus review by two radiologists without knowledge of arthroscopy findings. Forty-nine of 300 patients underwent arthroscopy. Thirty-five of these 49 patients also underwent MR arthrography. Scans were assessed for full-thickness scapholunate, lunatotriquetral, or TFCC tears. After consensus review of MR images, arthroscopy results were compared with MR interpretations. Of the 49 patients who underwent arthroscopy, 22 had TFCC tears, 18 had scapholunate tears, and 11 had lunatotriquetral tears (51 total tears at arthroscopy). Nineteen of 22 TFCC tears, 16 of 18 scapholunate tears, and nine of 11 lunatotriquetral tears seen at arthroscopy were seen on MRI. Four patients had both a TFCC tear and a lunatotriquetral ligament tear. One patient had a scapholunate ligament tear and a TFCC tear. There were no false-positive readings on MRI compared with arthroscopy. MR sensitivity for detection of TFCC tears was 86%, for detection of scapholunate tears was 89%, and for detection of lunatotriquetral tears was 82%. MR specificity for detection of tears was 100%. MR arthrography sensitivity for detection of ligament and TFCC tears was 100%. There were three patients in whom contrast media passed between joint spaces without an anatomic abnormality seen. These patients underwent arthroscopy with no tears seen. MRI at 3 T is sensitive and specific for detection of wrist ligament tears. MR arthrography is more sensitive for ligament evaluation but can result in false-positive findings because of microperforations.