Anastomotic Leak in Colorectal Cancer Surgery: From Diagnosis to Management or Failure - A Retrospective Cohort Study.

Abstract Background Anastomotic leakage (AL) after colorectal resections is a common surgical experience and the most frequent major adverse outcome. Early recognition of AL is critical to reduce mortality. We aim to evaluate the incidence, diagnostic criteria, morbidity, and mortality related with AL. Methods This is a cohort, descriptive retrospective, single-centred study of consecutive patients who underwent surgery with a colorectal anastomosis due to colorectal cancer, over a 4-year period (2013–2016). Results From 2013 to 2016, a total of 480 patients were included. A total of 37 (7.7%) had an anastomotic leakage. AL was diagnosed after 6.8 days in average (range 2–17), but most frequently on day 5. 25 out of the 37 patients were diagnosed based on clinical criteria, and 12 had a CT scan of the abdomen; 3 (25%) did not show unequivocal signs of AL. From all AL patients, 6 were managed non-operatively. 24 out of 31 patients (64.8%) were submitted to anastomotic takedown and Hartmann-type of procedure. The rate of Clavien-Dindo grade III and IV complications was significantly higher in the AL-patient group (70.2 vs. 7.7%, p < 0.0005). Mortality was higher in the leakage group (21.6% vs. 4.7%, p < 0.0005). Conclusions In this study, most patients were diagnosed earlier based on clinical criteria and the remaining patients had an abdomen-pelvic CT scan, with 25% of false negatives and a significant delay in diagnosis. The leakage group had higher morbidity and mortality, longer hospital stays and rate of reoperations. Both systematic use of scores in AL diagnosis and early reoperation, may have a positive impact in FTR rate reduction, and for this, additional prospective studies are needed.