Influence of neoadjuvant radiotherapy on anastomotic leak after restorative resection for rectal cancer.

PURPOSE: This study was undertaken to determine whether neoadjuvant radiotherapy is associated with an increased risk of anastomotic leak for rectal cancer patients undergoing restorative resection.

METHODS: From 1980 to 2010, patients who underwent restorative resection for rectal cancer (tumors within 15 cm of anal verge) were identified from a prospective institutional database and grouped based on whether they received neoadjuvant radiotherapy (+RT) or not (-RT). The main outcome was anastomotic leak documented by imaging (contrast leak), intra-operative or clinical (signs of peritonitis) findings and confirmed by staff surgeon assessment. Using multivariate (MV) analysis risk factors for leak were identified, presented as OR (95 % CI).

RESULTS: One thousand eight hundred sixty-two patients were included in the analysis, 28 % in the +RT group. Eighty-six percent of +RT patients received neoadjuvant chemoradiotherapy. The overall leak rate was 6.3 %, with no significant difference in +RT and -RT groups (8 % vs 5.7 %, p = 0.06). The +RT group had a lower mean age at surgery (58 vs 63 year, p < 0.001), more male (75 % vs 62 %, p < 0.001) and more ASA 3/4 (44 % vs 35 %, p < 0.001) patients, greater use of defunctioning ostomy (87 % vs 44 %, p < 0.001) and colo-anal anastomosis (77 % vs 34 %, p < 0.001). Mean tumor distance from the anal verge was lower in +RT group (6.6 vs 9.7 cm, p < 0.001). On MV analysis, male sex (OR 1.64 (1.03-2.62), p = 0.038), ASA 4 (OR 4.70 (2.07-10.7), p < 0.001), tumor distance from anal verge ≤ 5 cm (OR 2.49 (1.37-4.52), p = 0.003), and tumor size at surgery ≥ 4 cm (OR 1.75 (1.15-2.65), p = 0.009) were independently associated with leak. +RT was not independently associated with leak (OR 1.44 (0.85-2.46), p = 0.18), while defunctioning ostomy did not reduce leak occurrence (OR 0.75 (0.44-1.28), p = 0.29).

CONCLUSIONS: The findings suggest that neoadjuvant radiotherapy is not independently associated with an anastomotic leak for rectal cancer patients undergoing restorative resection and support a selective policy towards the use of a defunctioning ostomy on a case by case basis based on intra-operative judgment and consideration of tumor location, size, and patient characteristics.