PubMed2000
AIM: To assess the accuracy of the absent bow tie sign in diagnosing bucket handle meniscal tears (BHT) of the knee menisci. MATERIALS AND METHODS: During a 3-year period, we correlated the MRI and arthroscopic findings and the presence of the various signs. One hundred and seven knees were reviewed: 74 where either MRI or arthroscopy had identified a BHT and 33 which were either normal (31), or a simple tear was identified (2). All cases were reviewed by a single radiologist with a musculoskeletal interest blinded to the original results. Each was assessed for the presence of (1) a central meniscal fragment, (2) the double posterior cruciate ligament (PCL) sign, (3) the bow tie sign and (4) the contribution of a 3D-volume sequence. RESULTS: Optimal results were obtained using standard sequences and a 3D-volume sequence, giving a sensitivity of 74% and positive predictive value of 89%. The bow tie sign gave a sensitivity of 71% and positive predictive value of 76%, significantly less than previous reports. The 18 BHTs diagnosed by arthroscopy but missed by MRI showed other abnormal findings at MRI and were not reported as normal. CONCLUSION: We were not able to reproduce the previously reported high sensitivity and specificity of the absent bow tie sign. Despite optimization of all factors, the accurate diagnosis of a bucket handle tear remains difficult, and is most reliably made by identifying a central meniscal fragment, rather than relying on secondary signs such as the absent bow tie sign.