Is botulinum toxin injection of the pylorus during Ivor-Lewis esophagogastrectomy the optimal drainage strategy?

Background: The optimal management of the pylorus during esophagogastrectomy is unknown. Pyloromyotomy and pyloroplasty cause early edema and risk long-term bile reflux; however, the lack of pyloric drainage might risk early aspiration. Methods: We performed a retrospective study with a prospective database on patients with esophageal cancer or high-grade dysplasia who underwent Ivor-Lewis esophagogastrectomy. All had one surgeon and similar stomach tubularization, hand-sewn anastomoses, nasogastric tube duration, and postoperative prokinetic agents. Outcomes of postoperative gastric emptying, aspiration, and swallowing symptoms were compared. Results: Between January 1997 and June 2008, there were 221 patients. Seventy-one patients had a pyloromyotomy, and gastric emptying judged on postoperative day 4 was delayed in 93%(52% had any morbidity and 14% had respiratory morbidity). Fifty-four patients had no drainage procedure, and gastric emptying was delayed in 96% (59% had any morbidity and 22% had respiratory morbidity). Twenty-eight patients underwent pyloroplasty, and 96% had delayed gastric emptying (50% had any morbidity and 32% had respiratory morbidity). Sixty-eight patients had botulinum toxin injection into the pylorus. Gastric emptying was delayed in only 59% (P = .002, 44% had any morbidity and 13% had respiratory morbidity). Hospital length of stay (P = .015) and operative times (P = .037) were shorter in the botulinum toxin group. Follow-up (mean, 40 months) showed symptoms of biliary reflux to be lowest in the botulinum toxin group (P = .024). Conclusion: Injection of the pylorus with botulinum toxin at the time of esophagogastrectomy is safe and decreases operative time when compared with pyloroplasty or pyloromyotomy. In addition, it can improve early gastric emptying, decrease respiratory complications, shorten hospital stay, and reduce late bile reflux. A prospective multi-institutional randomized trial is needed.