Characteristics of clinical sepsis reflected in a reliable and reproducible rodent sepsis model.

ABSTRACT

BACKGROUND: Sepsis models are frequently based on induction of peritonitis, with cecal ligation and puncture reflecting the prototypical model. However, there is an ongoing discussion about the limitations of these models due to their variability in progression and outcome. Since standardization is a cornerstone of experimental models, we aimed to develop a reliable and reproducible procedure for induction of peritonitis. MATERIALS AND METHODS: A human stool batch was processed for -80° storage. For induction of peritonitis in fluid-resuscitated rats, a defined volume of stool suspension from this batch was injected intraperitoneally. For characterization of the model, physiologic and inflammatory changes were evaluated after sepsis induction. Survival analyses with the same batch were repeated in four independent experiments over a time period of 16 mo. RESULTS: The polymicrobial infection resulted in severe peritoneal inflammation with a systemic increase in cytokines. The mortality rate at 15 h was 29% and this was reproducible over a 16 mo time period. If antibiotic treatment was applied, a 50% survival was achieved. Laboratory markers indicated a progressive multi-organ dysfunction, while blood gas analysis showed respiratory compensation of a metabolic acidosis, and maintenance of PaO(2). Intravital microscopy of the liver revealed an impaired microcirculation. A decreased hemostatic potential was demonstrated by rotational thromboelastometry. Despite clinical recovery within 3 d, surviving animals showed laboratory and histologic signs of persisting inflammation even after 2 wk. CONCLUSIONS: This model reflects many features of human sepsis. Application of an infectious focus that is both quantitatively and qualitatively defined assures high reproducibility. Moreover, the procedure is simple and can be easily standardized.