Stomach cancer: a prospective study of anastomotic failure following total gastrectomy.

The incidence and cause of esophagojejunal anastomotic leakage were prospectively studied in 350 patients following total gastrectomy for stomach cancer. Anastomotic leak was demonstrated in 30 patients (8.6%), ten of whom underwent reoperation. Nine of the 30 patients died. Statistical analysis revealed that the outcome was better for patients with stapled esophagojejunostomy than when the anastomosis was hand-sutured. The odds for leakage were 2.37 times higher in patients with hand-sutured, than in those with stapled anastomosis. Antibiotic prophylaxis may be significant in preventing leakage. No association was found between anastomotic leakage and type of hospital, patient age or sex, preoperative weight loss, concomitant splenectomy or residual tumor in the resection margin.