Synovitis was induced in rabbits sensitized to bovine serum albumin (BSA) by a modification of the method of Dumonde and Glynn. Agents were administered starting on the day of initial BSA joint challenge and every weekday (14-17 doses) for 17 to 21 day periods. Synovial tissues were then excised and evaluated either (1) histologically for inflammatory cell infiltration, or (2) both histologically and for total IgG, anti-BSA, and beta-glucuronidase levels in homogenates of portions of the same tissues. By the intraperitoneal and oral routes, methylprednisolone (1 mg/kg/day), azathioprine (25-40 mg/kg/day) and cyclophosphamide (10 mg/kg/day) produced significantly decreases of 40-100% in some or all of the parameters measured. Non-steroidal anti-inflammatory drugs, including phenylbutazone, ibuprofen and acetylsalicylic acid at oral doses of 75 or 100 mg/kg/day, were ineffective as were colchicine at 1 mg/kg/day and indomethacin at 5 mg/kg/day. Thus, as we have measured it, this model of rheumatoid arthritis is not affected by those agents considered to be of limited effectiveness in this chronic disease, but is ameliorated by corticoids and some slow acting agents.

Immune synovitis in rabbits was investigated as a potential in vivo model for evaluating new antiinflammatory agents. Antigen-induced increases in knee width as well as P-glucuronidase and acid phosphatase activities in exudates were observed. Histologically, polymorphonuclear leukocytes appeared within hours in synovial tissues and reached maximum infiltration at about 24 hours. Subsequently, mononuclear cells, including plasma cells, appeared. The 6-hour Arthus-like phase of synovitis can be depressed by some antiinflammatory agents, colchicine and steroids being particularly effective. It is suggested that this model can be utilized to define potentially more effective antiinflammatory drugs.