Anastomotic leakage following routine mesorectal excision for rectal cancer in a national cohort of patients

Objective Mesorectal excision is successfully implemented as the standard surgical technique for rectal cancer resections in Norway. This technique has been associated with higher rates of anastomotic leakage (AL) and the purpose of this study was to examine AL in a large national cohort of patients. Methods This was a prospective national cohort study of 1958 patients undergoing rectal cancer surgery with anterior resection in Norway from November 1993 to December 1999. Results The overall rate of AL was 11.6% (228 of 1958 patients). In a multivariate analysis, the risk of AL was significantly higher in males (odds ratio (OR) 1.6, 95% confidence interval (CI) 1.1-2.2), in patients receiving pre-operative radiotherapy (OR 2.2, CI 1.0-4.7) and in low level (4-6 cm) (OR 3.5, CI 1.6-7.7) and ultra-low level (<= 3 cm) anastomoses (OR 5.4, CI 2.3-12.9). The presence of a diverting stoma was associated with a 60% reduction in the risk of AL (OR 0.4, CI 0.3-0.7) for anastomoses 6 cm and below. 30-day mortality was significantly higher for the patients with AL (7.0%, CI 3.7-10.3) compared with no AL (2.4%, CI 1.7-3.2) AL had no significant effect on local recurrence rate (log rank P = 0.608). Conclusion Low anastomoses should be defunctioned to avoid AL and the associated high perioperative mortality. No effect of AL on local recurrence was found in this large cohort.