Serum and Synovial Fluid Serum Amyloid A Response in Equine Models of Synovitis and Septic Arthritis

Objective To investigate the serum and synovial fluid serum amyloid A (SAA) response in equine models of synovitis and septic arthritis and to compare handheld and validated immunoturbidometric assays for SAA quantification. Study Design Controlled, experimental study. Animals Healthy adult horses (n = 9). Methods Synovitis (n = 4) and septic arthritis (n = 5) were induced using lipopolysaccharide and Staphylococcus aureus, respectively, and serial serum and synovial fluid samples were collected. Serial synovial fluid cytology was performed for both models and synovial fluid from the septic arthritis model was submitted for bacterial culture. Serum and synovial fluid SAA were quantified by handheld test and immunoturbidometric assay. Cytologic and SAA data were compared within and between models (mixed model ANOVA) and results of SAA assays were compared using category-by-category analysis (weighted kappa coefficient). Results Synovial fluid total nucleated cell counts and total protein increased significantly following induction of both models. Serum and synovial fluid SAA remained normal in synovitis horses and increased significantly in septic arthritis horses. Serum SAA increased more rapidly than synovial fluid SAA. Agreement was 98% when SAA concentrations were low (<50 μg/mL) but the assays diverged when concentrations were greater than ∼100 μg/mL. Overall, there was good category-by-category agreement between SAA assays (weighted kappa = 0.824). Conclusion Serum and synovial fluid SAA may be useful adjuncts in diagnosing septic arthritis in horses. SAA concentrations for the assays diverged and examination using a larger sample size is needed before direct numeric comparisons between the assays can be made.