Natural History of Intrameniscal Signal Intensity on Knee MR Images: Six Years of Data from the Osteoarthritis Initiative

Purpose To assess the natural history of intrameniscal signal intensity on magnetic resonance (MR) images of the medial compartment. Materials and Methods Both knees of 269 participants (55% women, aged 45–55 years) in the Osteoarthritis Initiative without radiographic knee osteoarthritis (OA) and without medial meniscal tear at baseline were studied. One radiologist assessed 3-T MR images from baseline and 24-, 48-, and 72-month follow-up for intrameniscal signal intensity and tears. A complementary log-log model with random effect was used to evaluate the risk of medial meniscal tear, adjusting for age, sex, body mass index, and knee side. Results At baseline, linear intrameniscal signal intensity in the medial compartment was present in 140 knees (26%). Once present, regression only in a single knee was observed. In 31 knees (19%) with linear intrameniscal signal intensity at any of the first three time points, the signal intensity progressed to a tear in the same segment, and in a single knee, the tear occurred in an adjacent segment. The corresponding number of tears without prior finding of intrameniscal signal intensities was 11 (3%). In the adjusted model, the hazard ratio for developing medial meniscal tear was 18.2 (95% confidence interval: 8.3, 39.8) if linear intrameniscal signal intensity was present, compared when there was no linear signal intensity. There was only one of 43 knees with injury reported in conjunction with the incident tear. Conclusion In middle-aged persons without OA, linear intrameniscal signal intensity on MR images is highly unlikely to resolve and should be considered a risk factor for medial degenerative meniscal tear. © RSNA, 2015 Online supplemental material is available for this article.