Objective: To evaluate protein C and other factors associated with the septic response as predictors of mortality in a clinically relevant animal model of sepsis.
Design: Laboratory investigation.
Setting: Eli Lilly and Company discovery research laboratory.
Subjects: Forty female Sprague Dawley rats weighing 245-265 g.
Interventions: Polyethylene catheters were surgically implanted into the femoral vein and sepsis was induced by cecal ligation and puncture (CLP). A solution of 5% dextrose in 0.9 % saline was continuously infused via femoral catheters immediately following surgery. Blood sampling was done before surgery and at 6 and 20 hrs after surgery. Rats were then monitored for survival out to 4 days.
Measurements and main results: Blood collections were used to measure blood glucose, bacteremia, plasma protein C, D-dimer, hormones, chemokines, cytokines, and myoglobin (as a marker of organ damage). Mortality was categorized into three groups: early death (before 30 hrs post-CLP), late death (after 30 hrs post-CLP), and survivors (96 hrs post-CLP). Compared with survivors, early death rats had statistically significant differences in 30 variables indicative of severe inflammation, coagulopathy, and muscle damage including less bacterial clearance, hypoglycemia, lower plasma protein C, higher plasma D dimer, higher plasma cytokine/ chemokines, and higher plasma myoglobin concentrations. Twenty variables had a moderate to strong correlation with time of death. Receiver operator characteristic curves generated from a simple logistic regression model indicated that KC and macrophage inflammatory protein-2, rodent homologues of the human growth related oncogene CXC chemokine family, and protein C were the best predictors of mortality in this model.
Conclusions: The data from this study indicate that an early decrease in protein C concentration predicts poor outcome in a rat sepsis model. The data further indicate that increases in the CXC chemokines macrophage inflammatory protein-2 and KC precede poor outcome.