Differences in structural and pain phenotypes between monoiodoacetae and meniscal transection models of osteoarthritis

Abstract

Purpose: To compare pathology in two animal models of osteoarthritis at several time points after the induction of arthritis and to investigate pain behaviour over the course of the experiment. Methods:Meniscal transection (MNX) arthritis was induced by transection of the meniscus in 180g male Sprague Dawley rats. Controls were animals subjected to sham surgery in which the collateral ligament was removed but the meniscus was left intact. Monosodium iodoacetate(MIA) arthritis was induced by the intra-articular injection of 1mg of sodium monoiodacetate dissolved in 50μl sterile saline. Separate groups of animals injected with saline were controls. Groups of animals (n=8) were sacrificed at 14, 35 and 45 days. Pain behaviour was monitored by differential weight bearing (g) and paw withdrawal thresholds (g). Pathology was scored for synovial inflammation (0-3), chondropathy (0-15), osteophytes (0-3) and the number of osteochondral channels /mm length of cartilage were counted. Data are means (95% CI). Results: Inflammation in the synovium remained constant over time. At day 49, MNX- treated knees had greater inflammation 1.7 (1.2-2.1) than Sham controls 0.2 (0.02-0.4) p<0.01, and MIA- treated knees were more inflamed 0.6 (0.2-0.8) than saline-injected controls 0.0 (0-0). MNX- treated knees were more inflamed than MIA -treated knees, p> 0.01. Osteophyte scores were greater at all time points in MNX- treated knees when compared to MIA -treated knees. At day 49 osteophytes scores were greater in MNX- treated knees 2.8 (2.6-3.0) than sham 0.23 (0.02-0.4), p<0.01, and greater in MIA-treated knees 1.0 (0.5-1.4) than in saline-injected controls 0 (0-0), p<0.01. Osteophyte scores were greater in MNX-treated knees than in MIA-treated knees (p<0.01). Chondropathy scores increased in both treatment groups with time and did not differ between MNX-treated knees and MIA-treated knees. At day 49 MNX-treated knees had higher osteophyte scores 12.7 (11.1-14.3) than sham controls 2.8 (0.8-4.8), p<0.01. Osteophyte scores in MIA-treated knees 9.7 (7.5-11.8) were higher than in saline controls 0 (0-0), p<0.01. At day 49 MNX-treated animals displayed greater numbers of channels crossing the osteochondral junction 0.25 (0.1-0.4) channels/mm compared to sham controls 0.01 (0.03-0.1), MIA-treated knees 0.1 (0.09-0.1) and saline controls 0.04 (0.01-0.2). MNX- treated animals displayed greater weight bearing asymmetry than did MIA-treated animals, whereas MIA-treated animals displayed greater hindpaw allodynia than did MNX-treated animals. Conclusions:We have compared two commonly induced models of OA in rats, within a single experiment using identical assessment tools. Although both models display features of OA, including chondropathy, osteopytosis, synovitis and pain behaviour, differences between the models may indicate that they each uniquely reflect different aspects of human OA. In particular, the MNX model is characterised by greater synovitis and osteophyte formation, and persistence of osteochondral channels, whereas the MIA model is associated with more pronounced distal allodynia. Appropriate selection of OA models depends on relevance to the research hypothesis, and to OA phenotype which it is intended to model.