We previously showed that a tyrosine kinase inhibitor, tyrphostin AG556 reduced serum tumor necrosis factor levels and improved survival in a canine model of intraperitoneal E. coli sepsis. Prior to the introduction of this drug clinically however, we performed additional studies, first in a rat model and then again in a canine model to determine the effects of this agent with S. aureus. In these studies, tyrphostin AG556 did not have consistent beneficial effects. Plots of control mortality and the effects of AG556 suggested however that severity of infection explained much of the variablity we had observed in the effects of AG556. With very lethal infection AG556 improved survival but with mild infection it was harmful. To test this possibility further, we did studies in rats assessing the effects of AG556 with doses of either E. coli, S. aureus, or endotoxin produc-ing low, medium, or high control mortalities. Analysis of this data showed that the effects of AG556 for each challenge were closely associated with the severity of infection as we had seen earlier. These studies suggest that tyrphostin AG556 may only be beneficial in patients with a high likelihood of dying related to their underlying infection. It may be harmful in patients with a low likelihood of dying. Additional studies are underway to determine the basis for the variable effects of AG556 in these studies.
