Pretreatment with tumor necrosis factor-α attenuates arterial hypotension and mortality induced by endotoxin in pigs

Objective: Tumor necrosis factor (TNF)-alpha administration in large amounts can induce a state of shock similar to that seen during severe sepsis. The objective of this study was to determine whether small doses of TNF-alpha might decrease the disposition for the development of shock induced by a subsequent infusion of endotoxin and to determine whether the mechanism of this protective effect of TNF-alpha pretreatment could be associated with up-regulation of the anti-inflammatory cytokine interleukin (IL)-10. Design: Prospective, randomized, experimental study. Setting: Investigative intensive care unit at an academic medical center. Subjects: A total of 14 female Yorkshire pigs, weighing 20-25 kg. Interventions: We studied two groups of animals-pigs treated with 500 ng/kg recombinant porcine TNF-alpha (n = 7) and pigs given diluent alone (n = 7). At 24 hrs after treatment, both groups of pigs were subjected to a 24-hr continuous infusion of lipopolysaccharide (LPS) at a rate of 80 ng/kg/min. Measurements and Main Results: The mortality rate was determined in both groups. Hemodynamic indices, oxygen transport variables, total and differential white cell counts, and serum concentrations of TNF and IL-10 were determined at frequent intervals before and after TNF-alpha administration and during the LPS infusion. Additionally, peripheral blood mononuclear cells were collected for determination of messenger ribonucleic acid expression of IL-10 by reverse transcription-polymerase chain reaction. The administration of TNF-alpha at the dose used in this study did not have any profound effect. No pig treated with TNF-alpha died in response to the LPS infusion. In contrast, three of seven control pigs died during the LPS infusion. Lipopolysaccharide-induced arterial hypotension and arterial hypoxemia were attenuated in the TNF-alpha-treated group. Both groups had significant increases in serum concentrations of TNF-alpha in response to LPS, with no significant difference in peak serum TNF-a between groups. Neither serum concentrations of IL-10 nor expression of IL-10 messenger ribonucleic acid in circulating mononuclear cells differed between groups. Conclusions: The administration of TNF-cu. attenuated the severity of hyperdynamic shock induced by a subsequent infusion of endotoxin. This effect could not be associated with increased expression or elaboration of the anti-inflammatory cytokine IL-10.