Background: In recent years, pancreaticoduodenectomy has been used more than eyer to resect a variety of benign and malignant diseases in the periampullary region.
Methods: The authors reviewed retrospectively metrical records of 169 patients who underwent pancraticoduodenectomy in the 1990s at the Korea University Medical Center, Department of Surgery. Statistical analyses of data were peformed using a statistical software package, Statistics ver 5.1.
Results: The 169 patients had a mean age of 57±12.1, with a male to female ratio of 1.46:1. Final diagnoses by pathologic examination showed malignancy in 88.2%(n=149) and benign diseases 11.8%(n=20). The. surgical procedures involved pyloric preservation in 63.4%, total pancreatectomy 6.5%, combined vascular resection(portal vein, superior mesenteric artery and vein) 2.4%, and adjacenct organ combined resection(colon, spleen, liver and etc.) 8.3%. Pancreaticoenteric reconstruction was done by pancreaticojejunostomy in 91.7% and by pancreaticogastrostomy in 1.8%. In both procedures, anastomosis was performed by mucosa to mucosa methods in 66.3%, and Dunking methods in 24.3%. The amount of median blood loss in operation was 800㎖, and that of mean transfusion was 2units. The median operative time was 410 minutes. Operative mortality was 2.4%(n=4). The operative complication rate was 46.7%, with wound infection the most common complication in 23%(n=39), pancreatic fistula in 21.9%(n=37), early delayed gastric emptying in 2.4%(n=4) which were pyloric-preserving cases. Five patients(3.5%) required reoperation in the immediate postoperative period for delayed bleeding, intarabdominal abscess and wound dehiscence. Mean hospitalization period was 28.1 days. A multivarate analysis indicated that the most powerful independent predictors of poor survival were malignant pathology, stage Ⅲ of disease Bloom-Richardson’s grade 3 of cellular differentiation and complications such as bleeding, acute renal failure and multiple organ dysfunction. Complicating pancreatic fistula prolonged hospital stay and increased morbidity, but did not influence overall survival. The incidence of pancreatic fistula was increased in the situation of preoperative biliary non-decompression, severe leukocytosis, low alkaline phosphatase level and smaller pancreatic ducts. It was not influenced by operative time, intraoperative blood transfusion and amount of blood loss.
Conclusion: The pancreaticdoudenectomy can be safely performed with reasonable morbidity and mortality in arious conditions of the periampullary region. Overall survival was determined largely by the various clinicopathologic characteristics, not by pancreatic fistula which had been suggested as the cause of poor prognosis and mortality. The pancreatic fistula is not a determinating factor for long-term as well as short-term survival as previously thought.
|