Partial tears of the anterior cruciate ligament: diagnostic performance of isotropic three-dimensional fast spin echo (3D-FSE-Cube) MRI

Purpose To compare the performance of 3D-FSE-Cube MRI to arthroscopy, the reference test for the diagnosis of partial anterior cruciate ligament (ACL) tears. Methods A retrospective study was performed including all patients who underwent surgery for an ACL tear in our Sports Surgery Unit from January 2008 to December 2009. All patients underwent a preoperative MRI, conventional 2D or 3D-Cube. The diagnosis of a partial tear was based on the appearance of the ligament bundles and signal quality on MRI, and on the continuity of the fibers on arthroscopy and the quality of the remaining ligament. Sixty-four of the 312 included patients underwent MRI 3D-Cube and 248 conventional 2D-MRI. The series included 82 women and 223 men, mean age 33.3 ± 19.6 years. Arthroscopy did not reveal any normal ACL, 247/312 (79.2 %) complete tears, and 65/312 (20.8 %) partial tears, with 50/65 (76.9 %) involving the anteromedial bundle and 15/65 (23.1 %) the posterolateral. Results The results of MRI 3D-Cube were as follows: sensitivity 95 % CI = 62.5 ± 23.7 %, specificity 95 % CI = 93.7 ± 6.9 %, likelihood ratio LR(+) = 9.9, LR(−) = 0.4 and accuracy 85.9 %. Results of conventional 2D-MRI were as follows: sensitivity 95 % CI = 10.2 ± 8.5 %, specificity 95 % CI = 96.5 ± 2.5 %, LR(+) = 2.9, LR(−) = 0.9 and accuracy 79.4 %. The diagnostic performance of MRI 3D-Cube was better than conventional 2D-MRI. Conclusion The diagnostic performance of MRI 3D-Cube in partial ACL tears was good and significantly better than conventional 2D-MRI. The likelihood of having a positive test was 9.9 times higher in a patient with a partial tear. A negative result did not exclude this diagnosis. Level of evidence Retrospective case series, Level IV.