Regulation of serum tumor necrosis factor in glucocorticoid-sensitive and -resistant rodent endotoxin shock models.

Bolus injection of lethal or sublethal doses of endotoxin or lipopolysaccharide (LPS) results in the rapid and transient rise in tumor necrosis factor (TNF) levels in serum in mammals. TNF levels peak between 1 and 2 h after LPS injection in mice and guinea pigs and approach basal levels by 6 h. Although the kinetics of TNF in serum appear similar between these two species, guinea pigs respond to a lethal dose of LPS of 20 mg/kg by producing approximately 10-fold more TNF than mice do. These two endotoxin shock models also differ in their sensitivity to glucocorticoids. TNF levels in serum are not reduced in the lethal endotoxin shock model in guinea pigs after treatment with dexamethasone at 25 mg/kg. In contrast, TNF levels in mouse serum are inhibited by more than 90% after treatment with dexamethasone at 3 mg/kg. Coincident with the TNF peak in serum is a leukopenia which approaches control levels by 6 h in dexamethasone-treated mice, while remaining depressed in dexamethasone-treated guinea pigs. Treatment with dexamethasone at 25 mg/kg did not save guinea pigs from endotoxin lethality, whereas long-term survival of mice under identical conditions was apparent. These results suggest that the relative glucocorticoid resistance observed in guinea pigs is also apparent in a lethal endotoxin shock model in which dexamethasone does not modulate TNF levels or result in increased survival as occurs in mice. The lack of clear efficacy for steroid therapy in human clinical septic shock trials would suggest that the guinea pig endotoxin model may be a more predictive system than the mouse model for the identification of novel agents useful in the treatment of endotoxin shock.