Granulocyte colony-stimulating factor but not peritoneal lavage increases survival rate after experimental abdominal contamination and infection

Abstract

Abstract Background The value of peritoneal lavage for intra-abdominal contamination and infection has never been proven scientifically. In contrast, the stimulation of host defence mechanisms with cytokines such as granulocyte colony-stimulating factor (G-CSF) has appeared promising in recent clinical trials. Methods Clinic modelling randomized trials (CMRTs), which model the complexity of the clinical reality, were used in rats in which peritoneal contamination and infection (PCI) was produced with human stool bacteria. The following groups were compared: trial 1, intraoperative peritoneal lavage with saline versus taurolin (18 rats per group); trial 2, no lavage versus saline lavage versus saline lavage plus subcutaneous administration of G-CSF (18 rats per group); trial 3, lavage with saline versus no lavage (30 rats per group). The primary endpoint was mortality at 120 h. Secondary endpoints were the phagocytic activity of granulocytes, and systemic and peritoneal cytokine levels. Results In trial 1 lavage with taurolin was not superior to that with saline (five of 18 versus eight of 18 animals survived; P = 0·32). In trial 2, six of 18 animals having no lavage and three of 18 receiving saline lavage survived. The combination of lavage and G-CSF increased the number of animals surviving to 11 of 18 (P < 0·05). Lavage combined with G-CSF stimulated granulocyte phagocytic activity (P < 0·01) and reduced the levels of interleukin (IL) 6 (P < 0·01) and tumour necrosis factor α (P < 0·05) in peritoneal fluid, as well as plasma levels of IL-6 (P < 0·05) and IL-10 (P < 0·01). In trial 3, survival was not significantly different in animals having lavage (14 of 30) and no lavage (19 of 30) (P = 0·14). Conclusion In these CMRTs of intra-abdominal contamination and infection, peritoneal lavage was not beneficial, but when lavage was combined with subcutaneous administration of G-CSF mortality was reduced and the local and systemic cytokine response was downgraded. Results from these CMRTs were used directly to define the trial conditions of a randomized clinical trial with G-CSF. Peritoneal lavage is not recommended.