End-to-side duct-to-mucosa pancreaticojejunostomy after pancreaticoduodenectomy A comparison trial of small versus larger jejunal incision. A single center experience.

Aim: The rates of post-operative pancreatic fistula (POPF) after pancreaticoduodenectomy (PD) are between 5% and 30%. Nowadays, pancreaticojejunostomy (PJ) represents the most common type of reconstruction after PD, but the ideal technique is still debated. Our randomized trial was conceived with the intent to evaluate if two variants of PJ could influence the post-operative outcome in term of early complications.

Material And Methods: Forty-eight consecutive patients treated with PD were randomized into 2 groups (Group 1 or Large Jejunal Incision or LJI group and Group 2 or Small Jejunal Incision or SJI group). Outcome measures were the operative time, postoperative complications, length of postoperative hospital stay, amylase content in drains.

Results: wenty-two patients were enrolled in the LJI and 26 in the SJI group. Median operative times did not differ between the 2 groups. The groups were homogeneous in respect to the median age of patients, the clinical presentation of jaundice and the presence of percutaneous biliary drainage (PBD). POPF developed in 3/22 (13.6%) and 1/26 (4%) patients among the LJI and SJI group respectively (3 grade B and 1 grade C respectively) (p=0.341). PPH occurred in 8/22 (36%) and 2/26 (8%) patients among the LJI and SJI group, respectively (p=0.018). The Amylase content in the drainage fluid measured at the 5th postoperative day showed a higher value in patients who underwent LJI anastomosis compared to those with SJI anastomosis [LJI group: 26.5 (6-254) U/l vs SJI group: 7 (0-38) U/l; p=0.051]. Delayed Gastric Emptying (DGE) was not different. The multivariate logistic regression analysis demonstrated both LJI anastomosis and DGE as independent predictors for pancreatic fistula (DGE: OR=20.04, CI 95%=1.92-208.83, P=0.012; LJI anastomosis: OR=24.58, CI 95%=1.71-354.32, P=0.019) and PPH (DGE: 30.5, CI 95%=3.02-308.16, P=0.004; LJI anastomosis: OR=12.71, CI 95%=1.23-131.55, P=0.033).

Conclusions: Based on the present results, we suggest to adopt what a "pancreas duct-oriented" approach: if pancreas duct is large a SJI-PJ is recommended; if the duct is < than 3 mm, a LJI must be preferred. Our conclusion is that the association of some surgeons to perform always the techniques with them are more confident is a concept of the past: recent data suggest that the pancreatic surgeon must have the different techniques in his "armamentarium" and varying the technique depending on local characteristic of the pancreas to allow a tailored approach to the patient.

Key Words: Pancreaticojejunostomy, Pancreatic fistula, Surgical Sutcome.