The relation between the biochemical composition of knee articular cartilage and quantitative MRI: a systematic review and meta-analysis.

Summary Objective Early and non-invasive detection of osteoarthritis (OA) is required to enable early treatment and monitoring of interventions. Some of the earliest signs of OA are the change in proteoglycan and collagen composition. The aim of this study is to establish the relations between quantitative magnetic resonance imaging (MRI) and biochemical concentration and organization in knee articular cartilage. Methods A preregistered systematic literature review was performed using the databases PubMed and Embase. Papers were included if quantitative MRI and a biochemical assay or polarized light microscopy (PLM) was performed on knee articular cartilage, and a quantified correlation was described. The extracted correlations were pooled using a random effects model. Results 21 papers were identified. The strongest pooled correlation was found for delayed gadolinium-enhanced MRI of cartilage (dGEMRIC) vs. proteoglycan concentration (r=0.59). T1ρ relaxation times are inversely correlated to proteoglycan concentration (r=-0.54). A weak correlation between T2 relaxation times and proteoglycans was found (r=-0.38). No correlation between T2 relaxation time and collagen concentration was found (r=-0.02). A heterogeneous set of correlations between T2 relaxation times and PLM were identified, including strong correlations to anisotropy. Conclusion DGEMRIC measures are significantly correlated to proteoglycan concentration. The needed contrast agent is however a disadvantage; the T1ρ sequence was found as a non-invasive alternative. Remarkably, no correlation was found between T2 relaxation times and collagen concentration. T2 relaxation times is related to organization, rather than concentration of collagen fibers. Prospero ID CRD42020168337.