Abstract
BACKGROUND: Bolus antithrombin III (AT) improved sepsis/organ dysfunction and survival in lipopolysaccharide/monomicrobial infusion pretreatment animal models; however, AT failed in clinical trials. Because insults and drug administration schedules differed between preclinical and clinical settings, we re-examined AT using a clinically relevant polymicrobial insult (cecal ligation and puncture, CLP) and a new method to continuously infuse AT after animals became ill. METHODS: Mice were catheterized with saline-filled osmotic minipumps. During CLP surgery we inserted AT- or saline-containing minipumps. We created and validated a ~6 h delay between sepsis induction and treatment. We compared delayed, continuous AT infusion with a conventional bolus AT injection using survival studies and 48-h studies. RESULTS: 6-h delayed, continuous AT infusion significantly improved 7-d survival versus saline infusion (65% vs. 29%, n = 21, P = 0.018) and versus a single injection of AT (65% vs. 19%, n = 21, P = 0.003). Delayed, continuous AT attenuated liver but not kidney or lung injury. Vascular leakage and inflammatory cytokines were suppressed only in liver. The highest accumulation of bacteria and thrombin at 48 h was in liver. AT did not change organ bacterial counts. CONCLUSIONS: Delayed, continuous AT infusion improved 7-d survival after CLP compared to single bolus AT injection or continuous vehicle. Liver may be critical in abdominal sepsis because of bacterial accumulation and subsequent thrombin generation. AT may be protective due to attenuation of thrombin-induced vascular leakage, inflammation, and liver injury during CLP sepsis. Because other organs were unprotected, AT may be combined with drugs protecting different organs.