Peritoneal cytokines predict mortality after surgical treatment of secondary peritonitis in the rat

The study aimed to analyze if peritoneal cytokine levels can predict survival in an experimental model for peritonitis. Early identification of patients most at risk for adverse outcomes would facilitate the decision for aggressive therapy in order to maximally exploit their chance for survival.

Peritonitis was induced by intraperitoneal injection of a feces/bacteria mixture in 175 rats. Surgical debridement was performed after 1 hour. Abdominal fluid samples were taken after 24 and 72 hours for the measurement of interleukin (IL)-6, IL-10, and tumor necrosis factor (TNF)-alpha. Surviving animals were sacrificed after 5 days and correlations between cytokine levels and survival were analyzed.

Altogether, 60 animals died prematurely, 12 before the first sampling of cytokines. So, 48 nonsurvivors and 115 survivors were analyzed. Peritoneal cytokine levels were much higher (p < 0.0001) in nonsurvivors than in survivors. At 24 hours there were strong correlations between cytokine levels, especially between IL-6 and IL-10 (r = 0.93). Peritoneal cytokines at 24 hours also discriminated between animals dying within the next 24 hours and those dying later. A strongly (p < 0.0001) increased mortality was observed if IL-6, IL-10, or TNF-alpha levels exceeded 2, 1, or 0.2 ng/mL, respectively. Receiver operating characteristic curves were promising for all 3, but IL-10 showed the best characteristics, with an area under the curve of 0.94 and 67% sensitivity at 95% specificity, obtained at a cut-off value of 1.26 ng/mL.

These data should generate renewed interest to examine the peritoneal cytokines as early markers for adverse outcomes in patients with secondary peritonitis. Possibly, combinations of peritoneal cytokines with other markers can lead to much needed, reliable early prediction of disease severity.