ABSTRACT
Objective The risk of current pancreaticojejunostomy is carefully considered from the perspective of the morphology of remnant pancreas,and we aimed to discuss the clinical outcomes of selecting different pancreaticojejunostomy techniques based on pancreatic morphology.Methods This was a prospective cohort study.The histopathology of remnant pancreatic tissues was categorized into four types based on preoperative radiological images and intraoperative palpation:Type Ⅰ:pancreas with hard texture in palpation,pancreatic atrophy,dilated pancreatic duct larger than 5 mm and remnant pancreatic surface ＜3 cm;Type Ⅱ:pancreas with hard texture in palpation,pancreatic atrophy and mild dilatation of pancreatic duct with the diameter of 3-5 mm and remnant pancreatic surface ＜3 cm;Type Ⅲ:pancreas with slightly hard texture,no atrophy,and normal or slightly dilated pancreatic duct with the diameter of 3-5 mm and remnant pancreatic surface ≥3 cm;Type Ⅳ:pancreas with soft texture,normal morphology and pancreatic duct.Results From January 2008 to August 2017,116 consecutive patients underwent pancreaticoduodenectomy in our center.Among them,10 patients with type Ⅰ underwent classic pancreatic ductal mucosa to mucosa anastomosis.19 patients with type Ⅱ underwent classic end to end invaginated pancreaticojejunostomy.45 patients with type Ⅲ underwent classic end to end invaginated pancreaticojejunostomy with overlapping U sutures;42 patients with type Ⅵ underwent total invaginated pancreaticojejunostomy.The post-operative pancreatic fistula occurred in 6 patients (5.2％) with one patient died.Postoperative bleeding occurred in 10 patients (8.6％),and gastroparesis occurred in 22 patients (19.0％).Overall complication rate was 33.6％.Conclusions Classification of pancreatic morphology based on preoperative radiological images and intraoperative palpation and the selection of corresponding pancreaticojejunostomy technique is theoretically rational and has the advantage of potentially reducing the risk of remnant pancreatic tissue.