Evaluation of postoperative menisci with MR arthrography and routine conventional MRI

Abstract

The diagnosis of a recurrent tear can be more difficult in the postoperative meniscus. The purpose of our study was to determine the accuracy of conventional magnetic resonance imaging (MRI) and MRI with intraarticular contrast material [magnetic resonance (MR) arthrography] for detecting recurrent meniscal tears after surgery. Seventy-two patients who had arthroscopic surgery for meniscal tear and still got complaints were selected prospectively for MR arthrography and conventional MRI. Routine knee protocols with appropriate surface coil were used on 0.5 T MRI system. A 1:100 gadolinium-saline solution 30-40 ml was injected intraarticulary for MR arthrography. Of 72 patients, 45 had repeated arthroscopy for comparison of the results. The morphology of the meniscus as seen on MR images was characterized according to meniscal repair an the degree of meniscal resection into three groups. (1) less than 25% of the meniscal resection, (2) more than 25% meniscal resection, (3) meniscal repair. With routine conventional MR images, we grouped patients according to the signal intensity of menisci as Grade 1, Grade 2, Grade 3, and tear. The prevalence of recurrent meniscal tears in our study group was [68% (37/72 tears)]. In the diagnosis of recurrent meniscal tear, conventional MRI and MR arthrography had sensitivities of 54.0% and 94.5%, respectively; specificities of 75.0% and 87.5%, respectively; and accuracy of 57.7% and 93.4%, respectively. The accuracy of the conventional MRI studies was significantly less than that of the MR arthrography group (P<.05). Our study results indicate high overall accuracy for MR arthrography in the diagnosis of recurrent or residual meniscal tears. When meniscal resection is minimal, the right diagnosis could be obtained with conventional MRI. MR arthrography is necessary for patients with meniscal resection of more than 25% and who do not have sufficient joint effusion.