MR diagnosis of meniscal tears of the knee: importance of high signal in the meniscus that extends to the surface.

OBJECTIVE

The presence of intrameniscal signal in contact with the surface is a commonly used criterion for the diagnosis of meniscal tear. This signal presumably represents the actual tear in the meniscus. However, some menisci with signal that contacts the surface are noted to be intact at arthroscopy whereas other menisci that have no signal in contact with the surface are found to be torn. We investigated the relationship between the presence of a meniscal tear at arthroscopy and the location within the menisci of signal that contacted the surface. We hypothesized that patterns were present that would improve the accuracy of MR diagnosis of a meniscal tear.

MATERIALS AND METHODS

We reviewed the MR and arthroscopic findings from 200 consecutive patients who had both knee MR examinations and knee arthroscopy. There were 108 medial and 58 lateral meniscal tears on arthroscopy. Each MR examination was reviewed for three aspects of intrameniscal signal: the number of images showing signal possibly or definitely contacting the surface, the specific surfaces involved, and the signal location. The coronal and sagittal images were evaluated separately. We correlated each of these features with the presence of a meniscal tear at arthroscopy.

RESULTS

Menisci with signal possibly contacting the surface had the same frequency of tears (three tears in 33 menisci) as menisci without signal contacting the surface (15 tears in 194 menisci). More than 90% of menisci with signal contacting the surface on more than one image were torn, but only 55% of medial and 30% of lateral menisci with such signal on only one image were torn. In the torn menisci with signal contacting the surface, such signal was seen only on sagittal MR images in 31% of the medial menisci and 45% of the lateral menisci. Sixteen percent of the torn lateral menisci had signal contacting the meniscal surface in only the anterior two thirds of the meniscus, whereas this was true in only 2% of the torn medial menisci. Distinct patterns were not seen in the association between tears and signal contacting either the superior or the inferior surface.

CONCLUSION

We found definite patterns in the location of intrameniscal signal that comes in contact with the meniscal surface. These patterns vary in the frequency of associated meniscal tears. Although menisci with internal signal in contact with the surface are usually torn, a tear is less likely if such signal is present on only one image. Tears may be identifiable on only one image plane. Tears in the anterior horn of the lateral meniscus are not uncommon. Knowledge of these patterns should help in the MR diagnosis of meniscal tears.