Impact of anastomotic leak on long-term survival in patients undergoing gastrectomy for gastric cancer

BACKGROUND: The impact of anastomotic leak (AL) on long-term outcomes after gastrectomy for gastric adenocarcinoma is poorly understood. This study determined whether AL contributes to poor overall survival. METHODS: Consecutive patients undergoing gastrectomy in a single high-volume unit between 1997 and 2016 were evaluated. Clinicopathological characteristics, oncological and postoperative outcomes were stratified according to whether patients had no AL, non-severe AL or severe AL. Severe AL was defined as anastomotic leakage associated with Clavien-Dindo Grade III-IV complications. RESULTS: The study included 969 patients, of whom 58 (6·0 per cent) developed AL; 15 of the 58 patients developed severe leakage. Severe AL was associated with prolonged hospital stay (median 50, 30 and 13 days for patients with severe AL, non-severe AL and no AL respectively; P < 0·001) and critical care stay (median 11, 0 and 0 days; P < 0·001). There were no significant differences between groups in number of lymph nodes harvested (median 29, 30 and 28; P = 0·528) and R1 resection rates (7, 5 and 6·5 per cent; P = 0·891). Cox multivariable regression analysis showed that severe AL was independently associated with overall survival (hazard ratio 3·96, 95 per cent c.i. 2·11 to 7·44; P < 0·001) but not recurrence-free survival. In sensitivity analysis, the results for patients who had neoadjuvant therapy then gastrectomy were similar to those for the entire cohort. CONCLUSION: AL prolongs hospital stay and is associated with compromised long-term overall survival.