Association between extrinsic and intrinsic carpal ligament injuries at MR arthrography and carpal instability at...

To retrospectively compare the presence or absence of carpal instability on radiographs with the findings of magnetic resonance (MR) arthrographic evaluation of intrinsic and extrinsic ligament tears in patients with chronic wrist pain.

The institutional review board approved this study and did not require informed consent. Signs of carpal instability were assessed on static and dynamic radiographs of the wrist obtained in 72 patients (24 female, 48 male; mean age, 36 years; age range, 14-59 years) with posttraumatic wrist pain. MR arthrography was subsequently performed. Two musculoskeletal radiologists independently analyzed the radiographs and MR images. Each intrinsic and extrinsic ligament was individually evaluated for the presence of a ligament tear. The extent of the tear also was recorded. Interobserver agreement regarding MR arthrographic findings was tested by calculating kappa statistics. Statistical comparison between radiography and MR arthrography was performed by using the Fisher exact test.

Twenty-five triangular fibrocartilage complex, 18 (five partial, 13 complete) scapholunate ligament, and 25 (10 partial, 15 complete) lunotriquetral ligament tears were visualized. Twenty-two (all complete) extrinsic ligament tears were detected: two radial collateral ligament, 10 radioscaphocapitate ligament, and 10 radiolunotriquetral ligament tears. Interobserver agreement regarding intrinsic and extrinsic ligament tear detection at MR arthrography was excellent (kappa = 0.80). Nineteen patients had evidence of carpal instability on radiographs. Fourteen (52%) of 27 patients with at least one complete intrinsic lesion had no sign of carpal instability. On the other hand, the association of scapholunate ligament and/or lunotriquetral ligament and extrinsic ligament tears was significantly correlated (P < .001) with carpal instability at radiography.

The presence or absence of carpal instability on radiographs depends on the association between intrinsic and extrinsic ligament tears-even partial ones-rather than on the presence of intrinsic ligament tears alone, even when the tears are complete.