ESTABLISHING A NON-HUMAN PRIMATE MODEL OF OSTEOARTHRITIS USING TRANSLATIONAL ENDPOINTS

Purpose: Current animal models of osteoarthritis (OA) have somewhat limited translation value for developing disease-modifying OA drugs (DMOADs) due to species and assessment related mismatches from clinical OA. Here, we investigated the potential of a nonhuman primate (NHP) surgical model of OA as a translational model by characterizing longitudinal arthritic changes using quantitative and semi-quantitative magnetic resonance imaging (MRI) scoring, serum biomarkers, pain behavior tests and histology. Methods: Six Cynomolgous macaques underwent surgical excision of the right knee medial meniscus (MMT) and four age-/weight-matched sham animals served as controls. MRI of the right knee joint were performed longitudinally prior to surgery and at one, two, and four months after surgery on a 3T Phillips MRI scanner using high resolution T1- and T2-weighted protocols, and a dynamic contrast enhanced series. Tibial and femoral cartilage plates were manually segmented and divided into subregions for quantitative cartilage assessment. The semi-quantitative MRI Osteoarthritis Knee Scoring (MOAKS) system was used to assess the structural changes of the knee joint, and the uptake of Gd-DTPA contrast agent was quantified in the regions of synovial tissue. OA related pain behaviors including Von-frey, knee flexion, and activity monitoring were assessed using the same time schedule. Blood samples were taken every two weeks to assess serum levels of CTX-1, HA, C2C, and MMP3. Animals were sacrificed at four (3 MMT, 2 sham) and six months (the remainder) post-surgery for histological evaluation of cartilage degeneration, bone alterations, and synovial inflammation. Group differences between MMT and sham were evaluated by a linear mixed model of the change from baseline for the behavioral, cartilage morphometry, and contrast enhancement rate measurements. Wilcoxon rank sum test was used for MOAKS evaluation. Group differences in histological assessments and biochemical biomarkers were analyzed by ANOVA. Results: Quantitative cartilage analysis depicted significantly thinner cartilage in the posterior medial femur of MMT animals at all timepoints (p<0.0001) and in the central medial tibiofemoral compartment at two months post surgery (p<0.05). Based on MOAKS scores, MMT animals had more cartilage degradation than sham in the central medial tibia at all timepoints (p<0.01) and in the central medial femur at one month (p<0.05) post-surgery. Hoffa-synovitis and whole knee effusion/synovitis scores were significantly higher in the MMT group with the maximal difference occuring at three months post-surgery (p<0.01) and abating thereafter. Higher contrast uptake rate was observed in multiple subregions of synovial tissue (p<0.01) of MMT animals compared to sham, peaking at two months post-surgery. Histological analysis captured increased tibial and femoral medial cartilage degradation in the MMT group relative to sham (p<0.05) with pronounced osteophytes in all MMT animals. MMT animals demonstrated decreased mobility relative to sham post-surgery (p<0.05). No significant differences were detected in other behavioral measures or in any serum biomarkers tested. Conclusions: In this study, MRI joint assessment demonstrated cartilage loss and synovial inflammation, both hallmarks of OA progression. These findings coupled with reduced physical activity and histological confirmation of OA-specific pathology are the first reported findings in a NHP model of surgically-induced OA. Additional assessment of this model with pharmacological intervention will further elucidate the utility of this novel translational platform for the development of DMOADs.