Can MRI visualise mechanical knee load in osteoarthritic knees? A systematic review with meta-analysis

Background: Bone marrow lesions (BML) are considered as a predictor of pain, disability and structural progression of knee osteoarthritis. A correlation between BML and pain exists: larger BMLs were found in OA patients with pain compared to OA patients without pain. Moreover, resolution of BMLs was related to a decrease in pain and a reduced loss of cartilage volume. This suggests that BMLs might be a modifiable feature for knee pain and knee OA. Therefore, BMLs are of particular interest as a target for preventive and therapeutic interventions to counter knee OA. They have been suggested as a biomarker for increased knee load but no comprehensive review on the subject exists.The relationship between knee loading and BML is not yet completely understood.Purpose: To summarize the available evidence regarding the relationship between joint load and the prevalence and progression of BMLs in the tibiofemoral joint.Methods: We systematically screened 3 databases (PubMed, Web of Science, The Cochrane Library) for studies encompassing BML and changes in knee load. Methodological quality assessment was done and a meta-analysis computing overall odd’s ratios was performed where possible.Results: Twenty-nine studies, involving 7641 participants, were included. Mechanical loading was categorized as: body weight & composition, compartmental load, structural lesion and physical activity. High compartmental load and structural lesion increased the risk for BML (overall ORs ranging from 1.56, 95% C.I. (1.13,2.15), p=.006 to 8.2, 95% C.I. (4.4,15.1)). Body weight increased the risk for BML to a lesser extent (overall OR=1.03, 95%C.I. (1.01, 1.05), p=.007). Contradictory results for the effect of physical activity on BML were found. Conclusion: Augmented compartmental loads and structural lesions increased the risk of presence or progression of BML. Body weight increased the risk for BML to a lesser extent. Contradictory results for the effect of physical activity on BML may be explained by a dose-response relationship, knee alignment and structural lesions.Implication: It has been shown that unloading the knee temporarily, may induce beneficial effects on osteoarthritis related structural changes. Therefore, an early recognition of BML in the aging athlete’s knee may provide information to counter the onset and aggravation of symptomatic knee osteoarthritis by reducing the knee load. What is known about the subject: A BML is generally defined as an area of either hyper- or hypo-intense signal in the trabecular subchondral bone on respectively T2-weighted fat-suppressed or T1- MR images. The exact pathophysiology of BML is still under debate. What this study adds to existing knowledge: This study is the first comprehensive review including a meta-analysis that investigates the effect of knee load on BML and that takes a broad spectrum of types of loading into account.