RESUMEN
Objective To analyze MRI features of real tears and pseudotears of posterior horn of the lateral meniscus (PHLM) at the insertion site of meniscofemoral ligament(MFL), and to discuss main points of differential diagnosis. Methods MR images of 32 patients with PHLM tears and 30 patients with anterior cruciate ligament tears but without PHLM tears confirmed by arthroscopy were analyzed retrospectively. Another 20 asymptomatic volunteers as controls underwent MR examination and analyzed. The number of consecutive slices displaying longitudinal increased signal in sagittal images and the length in axial images were evaluated. The one?way analysis of variance, χ2 test and ROC curve were used to analyze diagnostic value of different MRI findings. Results Longitudinal line with abnormal increased signal (pseudotear) was found in 82.0% (41/50) normal insertion site of MFLs. The typical MRI finding of real tears was peripheral longitudinal linear high signal in PHLM which reached the margin of articular surface. In sagittal images, longitudinal linear high signal was shown in (5.8 ± 1.2) slices in knees of real tears, which was more than (2.6±1.1) slices and (2.7±1.4) slices in pseudo?tear groups (F=60.9, P<0.01). The area under ROC curve was 0.96 for differentiating real tear from pseudo-tear using the number of consecutive slices displaying longitudinal increased signal in sagittal images. With a threshold of five or more consecutive images with abnormal longitudinal increased signal as the positive standard of continuous?line sign, the overall sensitivity, specificity and accuracy for diagnosing real tear were 90.6%(29/32), 90.2%(37/41) and 90.4%(66/73), respectively. The axial images showed that the length of increased signal line in the outer of PHLM was (16.4±4.9) mm in patients with real tears, which was longer than pseudo?tear groups with length of (8.1 ± 3.2) mm and (6.0 ± 3.1) mm (F=17.0, P<0.01). The area under ROC curve was 0.92 for differentiating real tear from pseudo?tear using the length in axial images. The zip sign was defined when its length was not less than 10 mm. The sensitivity, specificity and accuracy of the zip sign was 84.4%(27/32), 90.2%(37/41) and 87.7%(64/73) respectively. In coronal images, high signal of MFL attachment insertion was shown in 71.9%(23/32), 60.0%(15/25) and 10/16 cases, there was no significant difference (χ2=0.98, P=0.61). Conclusion The continuous?line sign and zip sign are characteristic findings of PHLM tears at the insertion site of MFL attachment, which are valuable for differential diagnosis with pseudotears at the insertion site of MFL.