J. M. Doniec

Background and Study Aims: The mortality rate for surgical revision of gastroesophageal anastomotic leakage after resection for cancer approximates 60 %. The efficacy of endoscopically placed covered metallic stents for treatment of gastroesophageal leakage was evaluated. Patients and Methods: Between June 1996 and June 2002 we treated 21 patients with proven gastroesophageal leakage; 18 had anastomotic leakage and three patients had perforation for different reasons. The extent of the leaks ranged from one-quarter of the intestinal circumference to its complete dehiscence. The average time from surgery to detection of leakage was 6.1 days (range 3 - 15 days). Mortality, healing rate, length of hospital stay, and complications were assessed. Results: The insertion of stents was performed endoscopically under radiological guidance without any complication in all patients. In 9.5 % (2/21) of patients complete sealing of the leak was not achieved. The mortality associated with anastomotic leakage was 23.8 % (5/21). In 80.1 % (17/21) patients complete healing of the leakage was achieved. The average hospital stay was 67 days (range 14 - 158 days). Of 23 stents, 13 (56.5 %) were removed, and three patients developed stenosis after removal. Conclusion: The treatment of gastroesophageal leakage with covered stents appears to reduce mortality and the complication rate associated with major leakage. Therefore this technique seems to be a reasonable alternative in the treatment of clinically relevant anastomotic leakage.