Cytokine removal on extracorporeal life support for treatment of acute endotoxemia: A randomized controlled animal study

Background: We prospectively evaluated the effectiveness of resin adsorption incorporated in an extracorporeal life support (ELS) circuit in an animal model of sepsis for removal of cytokines and prevention of hemodynamic deterioration during the treatment of septic shock. Methods: Twelve female landrace pigs were randomly assigned to two groups, a study group(n = 6), treated with high-flow resin adsorption (300 mL/min) and ELS, and a control group (n = 6), treated only with ELS. Septic shock was induced by intravenous 0.02 mu g/kg/min infusion was of Escherichia coli lipopolysaccharide (LPS). Measurements were carried out in the study group at baseline, at the end of LPS injection(t(0)) at 30(t(1)), 60(t(2)), 90(t(3)) and 120 min (t(4)) and 60 min after stopping resin-adsorption (t(5)). In the control group measurements were performed at baseline (t(0)), t(1) and only t(2), as no control animal survived beyond this latter experimental timepoint. Results: The final population consisted of 9 animals, five in the study group and 4 in the control group. Plasma values of both tumor necrosis factor a (TNF-alpha) and interleukin-6 (IL-6) were reduced during resin-adsorption (t(1)-t(4)) while these mediators increased in controls undergoing ELS only. With a clearance of TNF-alpha of 15,233 pg/min and IL-6 of 10,233 mu g/min, 79.2% of TNF-alpha and 95.3% of IL-6 produced were adsorbed. Systemic vascular resistance decreased significantly in both groups at t(0). While it further was reduced during the control experiments at t(1) and t(2), it returned to normal in the study animals. Cardiac output increased at t(0), t(1) and t(2) in the control experiments. In contrast, in study animals after a peak at t(0), it returned to the baseline value and did not vary thereafter. Conclusions: Combined resin-adsorption and ELS improved hemodynamics resulting from effective removal of inflammatory mediators in a pig model of septic shock