Objective: To examine the association between anastomotic leak and oncologic outcome after anterior resection, stratifying for defunctioning stoma.
Background: It has been hypothesized that anastomotic leak predisposes rectal cancer patients to local recurrence. Many have a defunctioning stoma to reduce risk of clinically significant leakage.
Methods: The records of patients undergoing low anterior resection (1991-2010) for rectal adenocarcinoma (≤15 cm from anal verge) were retrospectively analyzed using a prospectively collected colorectal database. Data (age, gender, stage, defunctioning stoma, neoadjuvant treatment, distance from anal verge, anastomotic leak) were collected. Clinical leakage was defined as anastomotic complication requiring intervention or interventional radiology within 60 days of surgery. Estimated local recurrence, overall survival, and disease-specific survival were compared using log-rank method and Cox regression analysis.
Results: 1127 patients were included, with 5.6-year median follow-up. The incidence of clinical anastomotic leak was 3.5%. Sixteen of 677 with defunctioning stoma (2.2%) developed clinical leak; 24 of 450 without stoma (6.3%) developed leak (P = 0.005). There were no perioperative deaths among patients with clinical leakage. When stratified for defunctioning stoma, there was no association between clinical leak and local recurrence, disease-free survival, or overall survival. On multivariable analysis, when controlling for neoadjuvant therapy, distance of tumor from anal verge, defunctioning stoma, and pathologic stage, clinical leak was not associated with time to local recurrence, disease-free survival, or overall survival.
Conclusions: In this cohort, anastomotic leakage was not associated with risk of local recurrence. Defunctioning stoma was associated with lower incidence of clinical leakage but not with difference in oncologic outcome. Careful patient selection for defunctioning stoma helps reduce risk of clinically significant anastomotic leak.