Background: The pancreatic anastomosis is still the Achilles heel in partial pancreatoduodenectomy (PPD).
Methods: This study describes retrospectively a series of 441 patients who underwent standard or extended PPD and reconstruction by either pancreatogastrostomy or pancreatojejunostomy over a period of 13 years (1988-2000).
Results: Reconstruction of the pancreatic remnant was achieved by pancreatogastrostomy in 250 patients (56.7 per cent) and by pancreatojejunostomy in 191 patients (43.3 per cent). The leakage rate of the pancreatic anastomosis was 2.8 per cent after pancreatogastrostomy versus 12.6 per cent after pancreatojejunostomy (P < 0.001), whereas other surgical complications (bile leakage, haemorrhage, pancreatitis) were identical in the two groups. The leakage rate after standard PPD with or without vascular reconstruction was 2.0 per cent (four of 205 patients) after pancreatogastrostomy and 11.5 per cent (18 of 156) after pancreatojejunostomy (P < 0.001); following extended PPD it was 6.7 per cent (three of 45) after pancreatogastrostomy and 17.1 per cent (six of 35) after pancreatojejunostomy. The mortality rate due to leakage was 1.6 per cent (four of 250 patients) after pancreatogastrostomy versus 5.2 per cent (ten of 191) after pancreatojejunostomy (P = 0.037).
Conclusion: Pancreatogastrostomy is a safe and reliable method of reconstruction after PPD that may be associated with a lower leakage and mortality rate than pancreatojejunostomy.