Subchondral bone changes in three different canine models of osteoarthritis.

Abstract

OBJECTIVE: To test the hypothesis that changes in subchondral bone are significantly different among three canine models of osteoarthritis (OA). DESIGN: In 21 purpose-bred mongrel dogs, OA was induced in one knee joint via either anterior cruciate ligament transection (ACLt; n&#xa0;=&#xa0;5), medial femoral condylar groove creation (GR; n&#xa0;=&#xa0;6), or medial meniscal release (MR; n&#xa0;=&#xa0;5). Five dogs that had sham surgery (SH; n&#xa0;=&#xa0;5) in one knee joint served as controls. Lameness scoring was performed every 4 weeks. Twelve weeks after surgery, the knee joints were examined by histology and histomorphometry. RESULTS: Articular cartilage pathology as determined by Mankin scores was significantly severe in all three OA models compared to SH controls in the medial tibia (P&#xa0;<&#xa0;0.001 to P&#xa0;=&#xa0;0.026). ACLt had significantly thinner subchondral plate thickness (Sp.Th) in both the medial and lateral tibias while MR had significantly thicker Sp.Th in the medial tibia compared to SH controls (P&#xa0;<&#xa0;0.001 to P&#xa0;=&#xa0;0.011). Trabecular bone volume (BV/TV) and trabecular bone thickness (Tb.Th) for ACLt were significantly less than SH controls in the tibias (P&#xa0;<&#xa0;0.001 to P&#xa0;=&#xa0;0.011). Tibial Sp.Th, BV/TV, and Tb.Th were all moderately to strongly correlated with lameness scores obtained throughout the study period (r&#xa0;=&#xa0;-0.436 to r&#xa0;=&#xa0;-0.738, P&#xa0;<&#xa0;0.001 to P&#xa0;=&#xa0;0.047) while Mankin scores showed moderate to strong correlations with Sp.Th in each OA model (r&#xa0;=&#xa0;0.465 to r&#xa0;=&#xa0;0.816, P&#xa0;<&#xa0;0.001 to P&#xa0;=&#xa0;0.033). CONCLUSIONS: Changes in Sp.Th are associated with articular cartilage damage while tibial Sp.Th and BV/TV and Tb.Th appear to be all influenced by joint loading alterations.