OBJECTIVETo compare a side-to-side esophagogastric anastomosis with conventional hand-sewn or stapled esophagogastrostomy for prevention of anastomotic stricture by randomized clinical trial.

METHODSBetween November 2007 and September 2008, 160 patients with esophageal carcinoma or gastric cardia cancer were consecutively admitted and underwent surgical treatment. After excluding 5 patients (2 refused to participate in and 3 did not meet inclusion criteria), the remaining 155 patients were completely randomized to receive either a side-to-side esophagogastric anastomosis (SS group), or the conventional hand-sewn (HS group), or a circular stapled (CS group) anastomosis, after the removal of esophageal tumor. The primary outcome measured the incidence of anastomotic stricture at 3 months after the operation (defined as the diameter of the anastomotic orifice

RESULTSThere was 1 operative death (in HS group) and 1 simple exploration (in SS group). The anastomotic leakage was observed in 4 patients (2 cervical and 1 intrathoracic leaks in HS group, and 1 intrathoracic leak in CS group). The follow-up rate was 96.1% (1 patient in SS group, 3 in HS group, and 2 in CS group were lost). Finally 45 patients in SS group, 52 in HS group, and 47 in CS group were included in the analysis. The 3 groups were preoperative similar. The anastomotic stricture rate was 0% (0/45) in SS group, 9.6% (5/52) in HS group, and 19.1% (9/47) in CS group, respectively (Fisher exact probability test, P = 0.005). The reflux/regurgitation score among 3 groups was similar (chi(2) = 1.681, P = 0.432).

CONCLUSIONThe side-to-side esophagogastric anastomosis could prevent stricture formation, without increasing gastroesophageal reflux.