Propensity score matching comparisons of postoperative complications and morbidity between digestive tract reconstruction methods after gastrectomy in gastric cancer patients with visceral obesity

BACKGROUND: Few studies have compared the prognosis of different reconstruction methods after gastrectomy for gastric cancer (GC) patients with obesity. The aim of the present study was to compare postoperative complications and overall survival (OS) between the following reconstruction methods: Billroth I (B-I), Billroth II (B-II), and Roux-en-Y (R-Y) after gastrectomy for GC patients with visceral obesity (VO). METHODS: We performed a double-institutional dataset study of 578 patients who underwent radical gastrectomy with B-I, B-II, and R-Y reconstructions between 2014 and 2016. VO was defined as a visceral fat area at the level of the umbilicus greater than 100 cm(2). Propensity score-matching analysis was performed to balance the significant variables. Postoperative complications and OS were compared between the techniques. RESULTS: VO was determined in 245 patients, of which 95, 36, and 114 underwent B-I, B-II, and R-Y reconstructions, respectively. B-II and R-Y were fused into the Non-B-I group due to the similar incidence of overall postoperative complications and OS. Therefore, 108 patients were enrolled after matching. The overall postoperative complications incidence and overall operative time in the B-I group were significantly lower than those in the non-B-I group. Further, multivariable analysis showed that B-I reconstruction was an independent protective factor for overall postoperative complications (odds ratio (OR) 0.366, P=0.017). However, no statistical difference in OS was found between the two groups (hazard ratio (HR) 0.644, P=0.216). CONCLUSIONS: B-I reconstruction was associated with decreased overall postoperative complications, rather than OS, in GC patients with VO who underwent gastrectomy.