RESUMEN
PURPOSE: A pancreaticoduodenectomy is the procedure of choice in patients with periampullary cancers. Recently, marked improvements in both morbidity and mortality rates have been reported following a pancreaticoduodenectomy. However, leakage at the pancreaticoenteric anastomosis remains a common and serious complication following a pancreaticoduodenectomy. Various measures directed towards prevention of pancreatic leakage have been studied over the past few decades. The aim of this study was to identify the risk factors for and prevention of pancreatic leakage following a pancreaticoduodenectomy. METHODS: A retrospective review of 75 patients, which had undergone a pancreaticoduodenectomy by the same surgeon between September 1997 and October 2004, was conducted. RESULTS: Ten, 30, 19 and 3 patients were treated for pancreatic head, common bile duct, ampullary, duodenal cancers, with a further 13 treated for benign diseases. Standard and pylorus-preserving pancreaticoduodenectomies were performed in 11 (14.7%) and 64 patients (85.3%), respectively. Postoperative complications were observed in 46 cases (61.3%), with 13 (17.3%) of these developing postoperative pancreatic fistula. The operative mortality rate was 1.3% (1 case). The pancreatic fistula with no peritoneal signs (which were observed in 12 patients) healed under simple drainage, without the need for reoperation. The analysis of factors for pancreatic fistula indicated no significant differences in age, sex, symptoms and signs, combined diseases, laboratory values, preoperative biliary drainage, operative times, blood transfusion, types of pancreatic stent or pathologic diagnosis when comparing those with and without pancreatic leakage; however, the pancreatic fistula rate (29.6%) during the early period (1997 to 2000) was significantly higher (10.4%) than during the latter period (2001 to 2004). CONCLUSION: This study suggests that accumulated experience of pancreaticoduodenectomies may reduce the incidence of pancreatic fistula and contribute to the prevention of pancreaticojejunal leakage complications. The number of cases enrolled in this study was small, as well as the data having heterogeneity, resulted in a limited statistical power and ability to interpret negative results. Therefore, to further understand the risk factors that contribute to pancreatic fistula, the need for a randomized controlled study is suggested.