Blumgart Anastomosis Versus Invaginating Pancreato-gastrostomy For Reconstruction After Pancreatoduodenectomy: A Randomized Controlled Trial.

OBJECTIVE To compare postoperative pancreatic fistula (POPF) rates between Blumgart anastomosis (BA) and invaginating pancreato-gastrostomy (PG) after pancreatoduodenectomy (PD). SUMMARY BACKGROUND DATA POPF rates after PD are still high. The only modifiable factor available to improve POPF is the anastomotic technique. BA and PG anastomoses have been previously shown to be feasible and safe, but they have never been compared in a randomized trial. METHODS A multicenter, randomized, controlled trial was conducted in 13 University Hospitals. Eligible patients were those presenting a pancreatic or periampullary neoplasm undergoing PD. Assignment to each group (BA or PG) was randomized by blocks and stratified by centers. The primary endpoint was the rate of POPF with special assessment of clinically relevant (B-C) POPF; secondary endpoints were postoperative complications, factors related to POPF and quality of life (QoL). RESULTS Two hundred and sixteen patients were randomized. POPF and B-C POPF were 44%-28% in BA group and 34%- 23% in PG group (P=0.39 and P=0.74, respectively). Overall complications, severe complications and mortality rates were 76%, 24%, 3.7% respectively, in the BA group, and 73%, 32%, 5.9%, in the PG group, with no significant differences. Soft pancreatic consistency, small preoperative CT Wirsung diameter, patient age, and 1st postoperative drain amylase concentration were independently associated with B-C POPF. QoL functional scales favoured the PG anastomosis at 9-months. CONCLUSIONS BA and PG showed no differences in POPF and postoperative outcomes; B-C POPF can be predicted based on several pre-intra and early postoperative parameters; QoL favoured PG anastomosis at 9 months.