Quantitative measures of bone shape, cartilage morphometry and joint alignment are associated with disease in an ACLT and MMx rat model of osteoarthritis

Multi-scale, subject-specific quantitative methods to characterize and monitor osteoarthritis in animal models and therapeutic treatments could help reveal causal relationships in disease development and distinguish treatment strategies. In this work, we demonstrate a reproducible and sensitive quantitative image analysis to characterize bone, cartilage and joint measures describing a rat model of post-traumatic osteoarthritis. Eleven 3-month-old male Wistar rats underwent medial anterior cruciate ligament (ACL) transection and medial meniscectomy on the right knee to destabilise the right tibiofemoral joint. They were sacrificed 6 weeks post-surgery and a silicon-based micro-bead contrast agent was injected in the joint space, before scanning with micro-computed tomography (microCT). Subsequently, 3D quantitative morphometric analysis (QMA), previously developed for rabbit joints, was performed. This included cartilage, subchondral cortical and epiphyseal bone measures, as well as novel tibiofemoral joint metrics. Semi-quantitative evaluation was performed on matching two-dimensional (2D) histology and microCT images. Reproducibility of the QMA was tested on eleven age-matched additional joints. The results indicate the QMA method is accurate and reproducible and that microCT-derived cartilage measurements are valid for the analysis of rat joints. The pathologic changes caused by transection of the ACL and medial meniscectomy were reflected in measurements of bone shape, cartilage morphology, and joint alignment. Furthermore, we were able to identify model-specific predictive parameters based on morphometric parameters measured with the QMA. •Demonstration of quantitative morphometric analysis for characterizing structural change in a rat post-traumatic OA model.•Strong and significant reproducibility and accuracy against the gold-standard, histology.•Sensitivity to pathological differences, and identification of model-specific predictive parameters.