OBJECTIVE. The purposeof this study was to determinethe positivepredictivevalue (PPV) for diagnosis of discoid lateral meniscal tear using MR imaging and to describe various patterns of such tears in the knee.SUBJECTS AND METHODS. MR reportsof 77 patients(10â€"67 yearsold) who under went prospective MR imaging that led to a diagnosis of discoid lateral meniscal tear were cor related with arthroscopic results. MR images obtained in 71 patients confirmed to have discoid lateral meniscial tear were retrospectively reviewed for the presence, site, and pattern of discoid lateral meniscal tear, including type of displacement of the torn segment. MR ab normalities were correlated with arthroscopic findings. RESULTS. For the prospective MR interpretations, the PPV for discoidmeniscus was92%. ppv for discoid meniscal tear was 57%. PPVs for individual types ofdiscoid meniscal tears were 46% (peripheral tear, 19/41), 76% (peripheral tear with horizontal tears, 16/21), 56% (horizontal tear, 5/9), 50% (transverse tear, 1/2), 67% (horizontal tear combined with transverse tear, 2/3), and 100% (longitudinal tear, 1/1). Peripheral tear alone and peripheral tear with horizontal tear were the most common types oftears (n = 20, 28%). Multiple tears (n = 34, 48%) were common. Displacement ofthe torn segments was seen in 51 patients (72%).CONCLUSION. MR imaging has a low PPV for diagnosingdiscoid lateral meniscal tear. Peripheral tear alone and peripheral tear with horizontal tear were the most common types of tears, and displacement of the torn segment was frequent. Because of its abnormal structure, the discoid meniscus sustains increased stress, which in creases the incidence of tears and cystic degen eration [15].
MMR imaging of the knee is commonly used to diagnosediscoidmenisci;however,to our knowledge, its reliability has not been criti cally studied. Because of regional anatomic features, the MR imaging pattern of such a tear may be atypical [16].In our practice in Korea, we have encountered repeated instances in which MR findings it ported to represent discoid lateral meniscal tear have not correlated with arthroscopic abnoimali ties. To address these issues, we designed a study to calculate the positive predictive value (PPV) ofMR imagingandtodetermine thevarious pat tems of discoid lateral meniscal tear, with ar throscopy as the standard of reference.
Subjects and MethodsFrom 1991 through 1996, 1802 patients at our institution underwent MR imaging of the knee. A database search yielded reports of 156(8.1%) who had evidence of a discoid lateral meniscus with a tearon prospectiveMR imaging.Arthroscopywas subsequently performed in 11 patients (49%, 11 knees) at our institution; these patients (38 males, 39 females; 10â€"61 years old; mean age, 27 years old) form the basis of this study. The right knee was examined in 31 patients and the left knee in 40 patients.