FLUID RESUSCITATION IMPROVES SURVIVAL OF ENDOTOXEMIC OR SEPTICEMIC RATS - POSSIBLE CONTRIBUTION OF TUMOR-NECROSIS-FACTOR

The present study was designed to investigate the effects of fluid adm inistration on survival in endotoxemic or septicemic male Sprague-Dawl ey rats. Endotoxemia was induced by intravenous injection of Escherich ia coli lipopolysaccharide (LPS), and septicemia produced by cecal lig ation and puncture (CLP). In endotoxemic animals deprived of fluid res uscitation, 7-day survival following injection of LPS at doses of 1, 3 , or 10 mg/kg LPS were 70 % (n = 10), 30 % (n = 10), and 0 % (n = 10), respectively. In rats resuscitated with 3.3 ml/kg/h of 0.9% NaCl, the dose-response curve for survival was shifted 5-fold rightward in a pa rallel manner (p < 0.001, between the fluid-resuscitated and nonfluid resuscitated LPS groups), indicating a reduced sensitivity to the effe cts of LPS following fluid resuscitation. LPS increased serum tumor ne crosis factor (TNFalpha) concentrations in fluid-resuscitated endotoxe mic animals from a baseline value of 20 U/ml to 2,350 U/ml at 1 h, whi ch returned to 200 U/ml at 2 h. In endotoxemic animals not receiving f luid resuscitation, serum TNFalpha levels at 1 and 2 h were 5-fold and 27-fold higher, respectively, than in fluid-resuscitated animals. The re were no differences in arterial blood pressure or heart rate betwee n the two groups of endotoxemic animals; total peripheral resistance w as significantly lower at 1 h, and cardiac index was significantly gre ater at 3 h in the fluid-resuscitated LPS group; otherwise there were no further differences in hemodynamic parameters between the two group s. The survival rate at 4 days following CLP without fluid resuscitati on was 14 %, whereas CLP with fluid resuscitation improved survival to 74 % (p < 0.01). TNFalpha was undetectable (i.e., < 20 U/ml) in the s erum of animals subjected to CLP. The improvement in survival with flu id infusion in the LPS and CLP models cannot be attributed to catheter implantation, or to improved hemodynamic parameters in the LPS model. The improvement in survival in the LPS model with fluid infusion was associated with attenuated increases in TNFalpha levels. Furthermore. these studies illustrate that fluid-resuscitated and nonfluid-resuscit ated experimental animal models should not be considered equivalent.