Mucosal damage in the esophageal remnant after esophagectomy and gastric transposition.

Objective: To assess development of mucosal damage in the esophageal remnant in regard to the level of the esophagogastrostomy reconstruction either in a right chest or in a left neck position.

Summary Background Data: Esophagectomy with gastric interposition creates an in vivo human model of pathologic esophageal reflux with the potential for long-term reflux disease complications.

Methods: Eighty-four esophagectomy patients were assessed over time by symptoms, endoscopy and biopsies of their esophageal remnant after the operation. The anastomosis was in the right upper chest (n = 36) or in a left cervical position (n = 48). Visual quantification of damage, details of histopathology, and time period since surgery were recorded.

Results: Twenty-nine patients (81%) with a right chest reconstruction had reflux symptoms when compared with 25 patients (53%) with a neck reconstruction (P = 0.007). Visualized reflux esophagitis was observed in 31 patients (81%) with chest anastomoses and in 22 patients (46%) with cervical anastomoses (P = 0.006). Documented mucosal damage and columnar lined metaplasia were significantly more frequent in the chest anastomosis group than the cervical group. The median of all mucosal damage and columnar lined metaplastic-free evolution were 13 +/- 3 and 20.5 +/- 6 months for the intrathoracic anastomosis, and 22 +/- 6 months and 40 +/- 8 months for the cervical anastomosis (P = 0.087). Two factors affecting the development of metaplasia were included in the multivariate analysis: an intrathoracic anastomosis (P = 0.018) and the presence of a previous Barrett esophagus (P = 0.064).

Conclusions: When a gastric transplant is used after esophagectomy, a high prevalence of mucosal damage is observed in the esophageal remnant independently of the level of reconstruction. A left cervical anastomosis favors less reflux symptoms, less visualized damage, and delays the development of mucosal damage over time.