Anastomotic leaks after bariatric surgery: it is the host response that matters

Abstract

Anastomotic leaks after bariatric surgery can lead to severe complications and adverse outcomes. We tested the hypothesis that not all patients with an anastomotic leak after bariatric surgery present with clinical symptoms and that their outcome is dependent on the aggressiveness of the host inflammatory response. This was a retrospective analysis of prospectively collected clinical data from 2384 bariatric surgeries from 1983 to 2006. All anastomotic leaks were identified from the database, and the vital signs, hematologic and biochemical data, mode of diagnosis, treatment, and outcome were recorded and analyzed. We identified 55 anastomotic leaks (2.3%) at a median of 4 days (range 1-26) after surgery. In 37 patients (67.3%), the leaks were identified at a median of 5 days (range 1-26) postoperatively because of clinical signs and symptoms of a systemic inflammatory response (SIRS leaks). In contrast, in 18 patients (32.7%), the leaks were identified at a median of 1.5 days (range 1-16) postoperatively only after routine contrast studies (non-SIRS leaks). Treatment included antibiotics and open drainage in 41.8%, laparoscopic drainage in 21.8%, computed tomography-guided drainage in 12.7%, conservative treatment in 14.5%, and other in 9.2%. All 6 deaths (4 men and 2 women, 10.9%) occurred in the SIRS group. Using logistic regression analysis, temperature (inflammatory response) and body mass index were independent predictors of mortality. The results of our study have shown that one third of patients with anastomotic leaks after bariatric surgery present with minimal clinical symptoms (non-SIRS) and are only detected if contrast studies are performed. Such leaks are unlikely to lead to death. Two thirds of patients with anastomotic leaks present with a systemic inflammatory response to the leak. Such leaks require urgent treatment that might not always prevent death.