Clinical risk factors for anastomotic leakage after laparoscopic anterior resection for rectal cancer: a systematic review and meta-analysis.

Abstract

Anastomotic leakage (AL) is a serious complication in laparoscopic rectal cancer surgery, and risk factors for AL are not well defined. Herein, we conducted a systematic review to quantify the clinicopathologic factors predictive for AL in patients who underwent laparoscopic anterior resection (LAR) for rectal cancer. A systematic search of electronic databases (PubMed, Embase, Cochrane CENTRAL, Scopus Database, and Wanfang Database) for studies published until August 2014 was performed. Cohort, case-control studies, and randomized controlled trials that examined clinical risk factors for AL were included. Fourteen studies (seven prospective and seven retrospective studies) involving 4580 patients met final inclusion criteria. From the pooled analyses, five demographic factors were found to be significantly associated with the development of AL, including male gender (OR 2.04, 95% CI 1.50-2.77), BMI ≥25 kg/m(2) (OR 1.46, 95% CI 1.00-2.14), ASA score >2 (OR 1.74, 95% CI 1.04-2.93, P = 0.04), tumor size >5 cm (OR 1.63, 95% CI 1.01-2.64, P = 0.05), and preoperative chemotherapy (OR 1.67, 95% CI 1.10-2.55, P = 0.02). Four operative factors were significantly associated with increased risk of AL, including longer operative time (95% CI 1.71-5.77, P = 0.0002), number of stapler firings ≥3 (OR 0.17, 95% CI 0.07-0.41, P < 0.001), intra-operative transfusions/blood loss >100 mL (OR 3.79, 95% CI 2.48-5.49, P < 0.001), and anastomosis level within 5 cm from the anal verge (OR 9.63, 95% CI 3.05-30.43, P = 0.0001), while pelvic drain (OR 0.43, 95% CI 0.19-0.94, P = 0.04) was significantly associated with a lower AL rate. Our analysis identified several clinicopathologic factors associated with AL in patients who underwent LAR. The knowledge of these risk factors may influence treatment- and procedure-related decisions and possibly reduce the leakage rate.