How do MRI-detected subchondral bone marrow lesions (BMLs) on two different MRI sequences correlate with clinically important outcomes?

The aim of this study is to describe the association of bone marrow lesions (BMLs) present on two different MRI sequences with clinical outcomes, cartilage defect progression, cartilage volume loss over 2.7 years, and total knee replacement (TKR) over 13.3 years. 394 participants (50-80 years) were assessed at baseline and 2.7 years. BML presence at baseline was scored on T1-weighted fat-suppressed 3D gradient-recalled acquisition (T1) and T2-weighted fat-suppressed 2D fast spin-echo (T2) sequences. Knee pain, function, and stiffness were assessed using WOMAC. Cartilage volume and defects were assessed using validated methods. Incident TKR was determined by data linkage. BMLs were mostly present on both MRI sequences (86%). BMLs present on T2, T1, and both sequences were associated with greater knee pain and functional limitation (odds ratio = 1.49 to 1.70; all p < 0.05). Longitudinally, BMLs present on T2, T1, and both sequences were associated with worsening knee pain (β = 1.12 to 1.37, respectively; p < 0.05) and worsening stiffness (β = 0.45 to 0.52, respectively; all p < 0.05) but not worsening functional limitation or total WOMAC. BMLs present on T2, T1, and both sequences predicted site-specific cartilage defect progression (relative risk = 1.22 to 4.63; all p < 0.05) except at the medial tibial and inferior patellar sites. Lateral tibial and superior patellar BMLs present on T2, T1, and both sequences predicted site-specific cartilage volume loss (β = - 174.77 to - 140.67; p < 0.05). BMLs present on T2, T1, and both sequences were strongly associated with incident TKR. BMLs can be assessed on either T2- or T1-weighted sequences with no clinical predictive advantage of either sequence.