Background: MRI-detected bone marrow lesions (BMLs) are associated with symptom severity and structural degeneration in knee osteoarthritis (OA). What BMLs represent at the tissue level is poorly described. The study aim was to characterise the cartilage-subchondral bone features corresponding to BMLs detected using two different MRI sequences for a knee OA cohort.
Methods: Whole tibial plateaus were retrieved from 54 patients (27-female, 27-male), aged 5186 years, undergoing total knee replacement surgery for OA. To identify BMLs ex-vivo, 3T-MRI scans were performed using T1 and PDFS-weighted sequences. MRI images were used for cartilage volume measurement. Micro-CT was used to assess microstructure of subchondral bone plate (SBP) and trabecular bone (STB). Cartilage and subchondral bone were assessed by OARSI and histopathology. Bone turnover indices were quantitated.
Results: BMLs were detected in 78% of patients (remainder formed No-BML group). Of all BMLs, BML-1 group (BML detected by PDFS only) represented 62%; BML-2 group (BML detected by PDFS and T1) represented 38%. BML-2 had reduced cartilage volume (P=0.007) with increased OARSI degenerative changes (P=0.009) compared to No-BML. BML-2 SBP was thicker and had lower porosity compared with No-BML (P<0.0001). BML-2 STB had higher bone volume (P=0.003), thicker (P=0.002) and more plate-like trabeculae (P=0.0004). SBP and STB osteoid volume and thickness were increased for BML-2 compared to No-BML (P<0.0001). More bone marrow oedema, necrosis and fibrosis was present in BML-2 compared to BML-1 and No-BML (P=0.03).
Conclusion: Knee OA BMLs are associated with loss and degeneration of the overlying cartilage, together with more sclerotic bone morphology. These relationships are more significant for BML-2, suggesting that BML-2 type lesions represent a later stage of OA disease. BMLs detected with specific MRI sequences may act as potential MRI biomarkers for identification of individuals at high risk of progressive OA and inform development and monitoring of new therapies.
Disclosure: The authors declared no competing interests. This work was supported by the National Health and Medical Research Council of Australia (APP1042482).