Effect of tumor necrosis factor antibody on synovial fluid cytokine activities in equine antebrachiocarpal joints injected with endotoxin.

Six horses received intra-articular injections of a mixture of 1 micrograms of endotoxin/5 mg of equine tumor necrosis factor (eqTNF) monoclonal antibody in 1 antebrachiocarpal joint and an equal volume (2 ml) of 1 micrograms of endotoxin/5 mg of control antibody in the opposite joint. Synovial fluid sample collection (1 ml) was accomplished by use of an indwelling, intra-articular catheter at postinjection hours (PIH) 0, 1, 1.5, 2, 5, and 8, and by arthrocentesis at PIH 24. Joint fluid samples were analyzed for nucleated cell count, protein concentration, and TNF, interleukin 6 (IL-6), IL-1, and IL-1-inhibitory activities. To monitor local inflammation, each carpus was graded semiquantitatively for swelling prior to each sample collection. Tumor necrosis factor, IL-1, or IL-1-inhibitory activity was not detected in any synovial fluid sample collected before endotoxin/antibody was administered. However, low IL-6 activity (< 100 U/ml) was found in 2 of 12 preinjection samples. In joints injected with endotoxin/control antibody mixture, maximal mean +/- SEM activities for TNF (1,019 +/- 310 U/ml), IL-1 (173 +/- 102 U/ml), and IL-6 (10.8 +/- 3.1 x 10(4) U/ml) were observed at PIH 2, 5, and 8, respectively. Tumor necrosis factor and IL-1 activities returned to baseline values by PIH 8 and 24, respectively; however, IL-6 activity remained high. Interleukin 1 inhibitory activity (27.4 +/- 2.25 IU/ml) was detected in all PIH-24 samples from control joints, but was not detected at any other time in control joints (limit of detection, 20 IU/ml). Tumor necrosis factor activity was not detected in any synovial fluid sample from joints treated with endotoxin/eqTNF antibody. In contrast, endotoxin IL-1 inhibitory activity (PIH 24) was higher in eqTNF antibody-treated joints (41.0 +/- 7.7 IU/ml) than in control joints, but the difference was not significant. Mean WBC count and protein concentration in control and treated joints were maximal at PIH 8. The curves for mean values of WBC count and total protein concentration were not significantly different in treated versus control joints. Swelling in each treated joint was either less than or the same as that in the opposite control joint at even, time in the initial 8 PIH. There was significant (P = 0.043) difference between treated and control joints at PIH 5 and 8. These results describe a profile of synovial fluid TNF, IL-1, IL-6 bioactivities, and IL-1-inhibitory activity during the initial 24 hours of synovitis induced by intra-articular administration of endotoxin in horses. Our eqTNF monoclonal antibody was effective in neutralizing TNF activity in synovial fluid when administered intra-articularly with endotoxin in horses. The induction of IL-1, IL-1 inhibitory activity IL-6, WBC, and total protein concentration responses are largely independent of TNF activity in synovial fluid of horses receiving endotoxin intra-articularly.