Effects of intra-articularly administered endotoxin on clinical signs of disease and synovial fluid tumor necrosis factor, interleukin 6, and prostaglanadin E2 values in horses

In each of 4 horses, sterile synovitis was induced by intra-articular injection of 3 micrograms of Escherichia coli endotoxin (lipopolysaccharide, LPS) into one antebrachiocarpal joint; an equal volume (2 ml) of phosphate-buffered saline solution (PBSS) was injected into the opposite, control carpus. Blood and 1.5 ml of synovial fluid were obtained at postinjection hours (PIH) 0, 2, 4, 8, 12, 18, 42, 66, and 144. Synovial fluid sample collection was accomplished by use of an indwelling, intra-articular catheter through PIH 12, and by arthrocentesis subsequently. Joint fluid samples were analyzed for cell counts, protein concentration, cytologic variables, and tumor necrosis factor (TNF), interleukin 6 (IL-6), and prostaglandin E2 (PGE2) values. Tumor necrosis factor and IL-6 activities and WBC count were also measured in blood. To monitor local inflammation, skin temperature of each carpus was imaged, using a thermographic scanner prior to each sample collection time. Horses had minimal systemic effects. Mean (SEM) rectal temperature increased significantly to 39.02 +/- 0.15 C only at PIH 18 after intra-articular injection of LPS. One horse had signs of mild depression from PIH 7 to 18, but its vital signs did not change appreciably. Each horse had mild signs of discomfort in the LPS-injected limb from PIH 1 to 3 until PIH 8 to 10. Mean peak surface temperature of the LPS-injected carpi was significantly higher than that of control carpi from PIH 8 to 144 (P < 0.05). Mean synovial fluid WBC count in the LPS-injected and control carpi increased after injection and peaked by PIH 8 (193,125 +/- 8,528 cells/microliter, LPS-treated; 16,425 +/- 8,336 cells/microliter, controls). Mean values for LPS-treated (principal) joints were significantly greater than values for control joints from PIH 2 until after PIH 66 (P < 0.05). Mean synovial fluid protein concentration increased in the principal and control joints, with values for the principal joints remaining significantly greater than values for control joints from PIH to 144 (P < 0.05). Mean TNF activity in synovial fluid was maximal at PIH 2 (10,573 +/- 5,924 U/ml). Interleukin-6 activity increased more gradually and peaked at PIH 8 (1.78 +/- 0.71 X 10(6) U/ml). Tumor necrosis factor activity did not increase above the minimal detectable value of 6 U/ml in the control joints. Mean PGE2 concentration in the principal joints peaked by PIH 2 (3.6 +/- 0.37 ng/ml) and remained significantly (P < 0.05) greater than the value for control joints from PIH 2 through 66. These results indicate that a humane model of synovitis was created by intra-articular injection of LPS and can be used to establish the basic responses of synovial TNF, IL-6, and PGE, in horses with early inflammatory joint disease.