Circulating Cytokine/Inhibitor Profiles Reshape the Understanding of the SIRS/CARS Continuum in Sepsis and Predict Mortality

Mortality in sepsis remains unacceptably high and attempts to modulate the inflammatory response failed to improve survival. Previous reports postulated that the sepsis-triggered immunological cascade is multimodal: initial systemic inflammatory response syndrome (SIRS; excessive pro-, but no/low anti-inflammatory plasma mediators), intermediate homeostasis with a mixed anti-inflammatory response syndrome (MARS; both pro- and anti-inflammatory mediators) and final compensatory anti-inflammatory response syndrome (CARS; excessive anti-, but no/low proinflammatory mediators). To verify this, we examined the evolution of the inflammatory response during the early phase of murine sepsis by repetitive blood sampling of septic animals. Increased plasma concentrations of proinflammatory (IL-6, TNF, IL-1beta, KC, MIP-2, MCP-1, and eotaxin) and anti-inflammatory (TNF soluble receptors, IL-10, IL-1 receptor antagonist) cytokines were observed in early deaths (days 1-5). These elevations occurred simultaneously for both the pro- and anti-inflammatory mediators. Plasma levels of IL-6 (26 ng/ml), TNF-alpha (12 ng/ml), KC (33 ng/ml), MIP-2 (14 ng/ml), IL-1 receptor antagonist (65 ng/ml), TNF soluble receptor I (3 ng/ml), and TNF soluble receptor II (14 ng/ml) accurately predicted mortality within 24 h. In contrast, these parameters were not elevated in either the late-deaths (day 6-28) or survivors. Surprisingly, either pro- or anti-inflammatory cytokines were also reliable in predicting mortality up to 48 h before outcome. These data demonstrate that the initial inflammatory response directly correlates to early but not late sepsis mortality. This multifaceted response questions the use of a simple proinflammatory cytokine measurement for classifying the inflammatory status during sepsis.