Modified binding pancreaticogastrostomy modified Blumgart pancreaticojejunostomy after laparoscopic pancreaticoduodenectomy for pancreatic or periampullary tumors.

Background: Laparoscopic pancreaticoenteric anastomosis is one of the technically challenging steps of minimally invasive pancreaticoduodenectomy (PD), especially during the learning curve. Despite multiple randomized controlled trials and meta-analyses, the type of pancreatico-enteric anastomosis as a risk factor for post-pancreatectomy complications is debatable. Also, the ideal technique of pancreatic reconstruction during the learning curve of laparoscopic PD has not been well studied.

Aim: To compare the short-term outcomes of modified binding pancreaticogastrostomy (PG) and Blumgart pancreaticojejunostomy (PJ) during learning curve of laparoscopic PD.

Methods: The first 25 patients with resectable pancreatic or periampullary tumors who underwent laparoscopic PD with modified binding PG or modified Blumgart PJ between January 2015 and May 2020 were retrospectively analyzed to compare perioperative outcomes during the same learning curve. A single layer of the full-thickness purse-string suture was placed around the posterior gastrotomy in the modified binding PG. In the modified Blumgart technique, only a single transpancreatic horizontal mattress suture was placed on either side of the pancreatic duct (total two sutures) to secure the pancreatic parenchyma to the jejunum. Also, on the ventral surface, the knot is tied on the jejunal wall without going through the pancreatic parenchyma. Post pancreatectomy complications are graded as per the International Study Group for Pancreatic Surgery criteria.

Results: During the study period, modified binding PG was performed in 27 patients and modified Blumgart PJ in 29 patients. The demographic and clinical parameters of the first 25 patients included in both groups were comparable. Lower end cholangiocarcinoma and ampullary adenocarcinoma were the primary indications for laparoscopic PD in both groups (32/50, 64%). The median operative time for pancreatic reconstruction was significantly lower in the binding PG group (42 58 min, = 0.01). The clinically relevant (Grade B/C) postoperative pancreatic fistula (POPF) was significantly more in the modified PJ group (28% 4%, = 0.04). In contrast, intraluminal postpancreatectomy hemorrhage (PPH) was more in the binding PG group (32% 4%, = 0.02). There was no significant difference in the incidence of delayed gastric emptying between the two groups.

Conclusion: During the learning curve of laparoscopic PD, modified binding PG reduces POPF but is associated with increased intraluminal PPH compared to PJ using the modified Blumgart technique.