RESUMO
BACKGROUND: Accumulating evidence suggests that peritoneal cytokine concentrations may predict anastomotic leak after colorectal surgery, but previous studies have been underpowered. OBJECTIVE: We aimed to test this hypothesis by using a larger prospectively collected data set. DESIGN: This study is an analysis of prospectively collected data. SETTINGS: This study was conducted at 3 public hospitals in Auckland, New Zealand. PATIENTS: Patients undergoing colorectal surgery recruited as part of 3 previous randomized controlled trials were included. MAIN OUTCOME MEASURES: Data on peritoneal and plasma levels of interleukin-6, interleukin-8, interleukin-10, and tumor necrosis factor-α on day 1 after colorectal surgery were reanalyzed to evaluate their predictive value for clinically important anastomotic leak. Area under receiver operating characteristic curve analysis was performed. RESULTS: A total of 206 patients with complete cytokine data were included. The overall anastomotic leak rate was 8.3%. Concentration levels of peritoneal interleukin-6 and interleukin-10 on day 1 after colorectal surgery were predictive of anastomotic leak (area under receiver operating characteristic curve, 0.72 and 0.74; p = 0.006 and 0.004). Plasma cytokine levels of interleukin-6 were higher on day 1 after colorectal surgery in patients who had an anastomotic leak, but this was a poor predictor of anastomotic leak. Levels of other peritoneal and plasma cytokines were not predictive. LIMITATIONS: The study was not powered a priori for anastomotic leak prediction. Although the current data do suggest that peritoneal levels of interleukin-6 and interleukin-10 are predictive of leak, the discriminative value in clinical practice remains unclear. CONCLUSIONS: Peritoneal levels of interleukin-6 and interleukin-10 on day 1 after colorectal surgery can predict clinically important anastomotic leak.