Clinical impact of the new "twin U-stitch method" of pancreaticogastrostomy in pancreaticoduodenectomy.

PURPOSE The optimal pancreaticogastrostomy (PG) method for reducing pancreatic fistula (PF) incidence remains unclear. This retrospective review aimed to evaluate the clinical impact of the "twin U-stitch method" and compared it with the conventional invagination method. METHODS Data of 183 consecutive patients who underwent PG after pancreaticoduodenectomy (PD) between January 2015 and November 2020 were evaluated. PF incidence was compared between patients who experienced twin U-stitch PG (twin U-stitch group) and those who experienced conventional invagination PG (conventional PG group). RESULTS The twin U-stitch and conventional PG methods were performed in 97 and 86 patients, respectively. The time required for twin U-stitch PG was shorter than conventional PG (9.3 min vs 20.0 min, P < 0.001). The twin U-stitch group showed a lower incidence of PF than the conventional PG group (8% vs. 19%, P = 0.038). Multivariate analysis confirmed that twin U-stitch PG was significantly correlated with a decreased risk of PF (odds ratio, 0.23; P = 0.006), independent of the texture of the pancreas. Subgroup analysis of patients with soft-textured pancreas showed that the median drain amylase levels in the twin U-stitch group on postoperative days (POD) 1 and 3 were significantly lower than those in the conventional PG group (POD 1: 1,335 vs. 5,991 U/L, P < 0.001; POD 3: 212 vs. 518, P = 0.001). CONCLUSION The twin U-stitch method was simple and preferable to the conventional method for preventing PF in patients with PD.