LAUSR

Background The appropriate surgical procedure for patients with upper third early gastric cancer is controversial. We compared total gastrectomy (TG) with proximal gastrectomy (PG) in this patient population. Methods A multicenter, non-randomized trial was conducted, with patients treated with PG or TG. We compared short- and long-term outcomes between these procedures. Results Between 2009 and 2014, we enrolled 254 patients from 22 institutions; data from 252 were included in the analysis. These 252 patients were assigned to either the PG ( n  = 159) or TG ( n  = 93) group. Percentage of body weight loss (%BWL) at 1 year after surgery, i.e., the primary endpoint, in the PG group was significantly less than that of the TG group (− 12.8% versus − 16.9%; p  = 0.0001). For short-term outcomes, operation time was significantly shorter for PG than TG (252 min versus 303 min; p  < 0.0001), but there were no group-dependent differences in blood loss and postoperative complications. For long-term outcomes, incidence of reflux esophagitis in the PG group was significantly higher than that of the TG group (14.5% versus 5.4%; p  = 0.02), while there were no differences in the incidence of anastomotic stenosis between the two (5.7% versus 5.4%; p  = 0.92). Overall patient survival rates were similar between the two groups (3-year survival rates: 96% versus 92% in the PG and TG groups, respectively; p  = 0.49). Conclusions Patients who underwent PG were better able to control weight loss without worsening the prognosis, relative to those in the TG group. Optimization of a reconstruction method to reduce reflux in PG patients will be important.