MRI-detected subchondral bone marrow lesions: does using two mri sequences improve detection and clinical relevance?

Abstract

Purpose: Fluid-sensitive MRIs are preferred in scoring BMLs compared to gradient recalled echo (GRE)-type MRI sequences. However, whether one sequence correlates better with clinical osteoarthritis (OA) outcomes than the other is unknown. We aim to describe the association of bone marrow lesions (BMLs) present on two different MRI sequences with pain, physical function limitation, stiffness, cartilage defect progression, cartilage volume loss over 2.7 years, and total knee replacement (TKR) over 13.3 years in older adults. Methods: 394 community-dwelling adults aged 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-w GRE) and T2-weighted fat-suppressed 2D fast spin-echo (T2-w FSE) MRI sequences, at the tibial, femoral, and patellar sites. Knee pain, physical function limitation, and stiffness were assessed using Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC) scales. Cartilage volume and defect scores were assessed using validated methods. The incidence of TKR was determined by data linkage to the Australian Orthopaedic Association National Joint Replacement Registry. Ordinal logistic, linear, log binomial, and mixed-effects regression were used for analysis. Results: BMLs were mostly present on both MRI sequences (86%) and appeared slightly larger on T2-w FSE MRI. BMLs present on T2-w FSE, T1-w GRE, and both MRI sequences were associated with increased odds of moving to a higher category of knee pain and physical function limitation (odds ratio (OR) = 1.49 to 1.70; all P < 0.05) but not stiffness. Longitudinally, BMLs present on T2-w FSE, T1-w GRE, and both MRI 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 physical function limitation or total WOMAC. BMLs present on T2-w FSE, T1-w GRE, and both MRI sequences predicted site-specific cartilage defect progression (relative risk (RR) = 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-w FSE, T1-w GRE, and both MRI sequences predicted site-specific cartilage volume loss (β = −174.77 to −140.67; P < 0.05). BMLs present on T2-w FSE, T1-w GRE, and both MRI sequences were strongly associated with incident TKR. Conclusions: BMLs can be assessed on either T2-w FSE or T1-w GRE MRI sequence alone with no predictive advantage of performing both sequences, even though BMLs appear smaller on T1-w GRE MRI.