Management of esophagogastric anastomotic leaks☆

Abstract

We read with interest Junemann-Ramirez and colleagues’ article [1] on esophagogastric anastomotic leaks and their analysis of predictive factors, management and survival. We agree with the authors that esophagogastric anastomotic leaks are the most devastating of complications after esophageal resection and that they carry high morbidity and mortality rates. The choice of surgical procedure has a lot to do with the ultimate outcome of the patient. We feel that the consequences of an anastomotic leak are far worse with an intrathoracic anastomosis compared with a neck anastomosis. The Ivor Lewis esophagectomy has neither the advantages of a transhiatal esophagectomy (in avoiding a thoracotomy) nor does it have the advantages of a transthoracic (three-hole) total esophagectomy (where a neck anastomosis avoids the morbidity of mediastinitis in case of a leak). A three-hole esophagectomy also enables radical supracarinal mediastinal lymphadenectomy. The authors’ conclusions that patients treated conservatively fared better than those who underwent surgical intervention could clearly be explained by a selection bias as it is likely that clinically stable patients would have been treated conservatively and sicker patients with mediastinitis would have been reoperated. We also fail to understand how a pyloric drainage procedure could reduce anastomotic leaks—decompression of the gastric tube could easily and effectively be achieved by nasogastric tube decompression. In our unit (we perform an average of 160 esophagectomies a year), we perform a three-hole total esophagectomy in all patients with a neck anastomosis, but do not perform a formal pyloric drainage procedure. None of the leaks we encountered could be explained by non-performance of a gastric drainage procedure. Most true anastomotic leaks in our unit are managed conservatively as the leak gets localized in the neck with no signs of mediastinitis. Gastric tube necrosis or ischemia, on the other hand, are managed aggressively by immediate surgical reoperation with disconnection of the anastomosis with reconstruction by a coloplasty as an interval surgery [2]. Needless to state, the latter group fare worse than the former.