Prognostic impact of anastomotic leakage after elective colon resection for cancer - A propensity score matched analysis of 628 patients.

BACKGROUND: There is limited information regarding the impact of anastomotic leakage on oncologic outcome in exclusively colon cancer patients.

METHODS: The colorectal database of the Department of Surgery of the University of Heidelberg was used to assess the impact of anastomotic leakage on oncologic outcome in patients undergoing curative resection for Stage I-III colon cancer. Risk-adjusted Cox regression analysis and propensity score methods were used to assess overall, disease-free, and relative survival.

RESULTS: 628 patients of which 26 (4.1%) experienced anastomotic leakage were analysed. Anastomotic leakage was associated with significantly worse overall, disease-free and relative survival in univariate and multivariate analysis. The analysis after exact propensity score matching confirmed the negative impact of anastomotic leakage on overall (HR 2.62, 95% CI 1.33-5.18, p = .011), disease-free (HR 2.28, 95% CI 1.16-4.47, p = .027) and relative survival (HR 3.70, 95% CI 1.82-7.52, p < .001). 5-year overall survival was 51.6% (95% CI 34.5-77.2%) for patients with anastomotic leakage compared to 77.7% (95% CI 73.0-82.8%) for patients without anastomotic leakage.

CONCLUSIONS: All conceivable efforts should be made to avoid anastomotic leakage after colon resection for cancer not only to evade short-term consequences but also to allow for adequate long-term outcome.