Influencing Factors of Symptomatic Anastomotic Leakage after Anterior Resection of the Rectum for Cancer.

Zhi-jie Cong,Chuan-gang Fu,Han-tao Wang,Lian-jie Liu,Wei Zhang,Hao Wang

DOI: https://doi.org/10.1007/s00268-009-0008-4

IF: 3.282

2009-01-01

World Journal of Surgery

Abstract:BACKGROUND:The aim of the present study was to analyze the factors associated with anastomotic leakage after anterior resection for rectal cancer.METHODS:Retrospectively collected consecutive data of 738 rectal cancer patients who underwent anterior resection in our hospital between 2005 and 2008 were reviewed. The associations between 15 patient-related and surgery-related variables and anastomotic leakage were studied with both the univariate chi-square test and multivariate logistic regression analysis.RESULTS:Univariate analysis showed that risk factors associated with anastomotic leakage were low rectal cancer (located 5 cm or less above the dentate line) (5.9% vs. 0.9%; P = 0.003), non-specialized surgeon (3.9% vs. 11.3%; P = 0.031), and defunctioning transanal catheter placement (14.5% vs. 3.6%; P < 0.001). It should be noted that the mean surgeon case volumes of anterior resection of colorectal surgeons and non-specialized general surgeons were 43 per year and 2 per year, respectively (P < 0.001). In addition, there was a tendency for colorectal surgeons to operate on a greater proportion of low rectal cancers (72.1% vs. 52.8%; P = 0.003). In the multivariate analysis, besides low rectal cancer, non-specialized surgeon, and transanal catheter placement, three other factors were associated with anastomotic leakage: diabetes mellitus (P = 0.027), free distal margins less than 1 cm (P = 0.009), and a defunctioning stoma (P = 0.031). In a further analysis of 522 patients with low rectal cancer, the leakage rate in patients with a defunctioning stoma was significantly lower (2.9% vs. 8.5%; P = 0.007). By contrast, the leakage rate in the transanal catheter placement group was higher (15.1% vs. 4.9%; P = 0.008), because of its poor protective effect as well as the selection bias.CONCLUSIONS:From the findings of this study, we believe that low rectal cancer, non-specialized surgeons, and diabetes mellitus are risk factors for anastomotic leakage after rectal surgery, and that a defunctioning stoma could significantly reduce the incidence of leakage in low rectal cancer patients.