Keith D. Lillemoe, M. P. Kim, John L. Cameron, Kurtis A. Campbell, Patricia K. Sauter, J. A. Coleman, Charles J. Yeo, Indiana University School of Medicine, Department of Surgery, Indianapolis, IN, Johns Hopkins University, Baltimore, MD

Purpose: Despite significant improvements in perioperative mortality, various complications such as the development of pancreatic fistula remain a common occurrence following pancreaticoduodenectomy. A number of uncontrolled series have reported that the use of fibrin glue sealant may decrease the rate of pancreatic anastomotic leak. It was the objective of this study to evaluate the role of fibrin glue sealant to decrease the rate of pancreatic fistula after pancreaticoduodenectomy. Methods: Between August 2000 and May 2002, 124 patients were entered into this study. Criteria for entry included patients undergoing pancreaticodudenal resection with pancreaticojejunal anastomosis which was at a high risk for development of a pancreatic anastomotic leak (soft, normal texture gland, nondilated pancreatic duct). After completion of the pancreaticojejunal anastomosis, the patients were randomized to topical application of fibrin glue or no such application. The remainder of the operative and perioperative management were as directed by the attending surgeon. The primary endpoints in this study were pancreatic fistula, total complications, death, and length of hospital stay. Results: A total of 58 patients were randomized to the fibrin glue arm, whereas 66 patients were randomized to the control arm and did not receive fibrin glue application. There were no significant differences noted between the two groups with respect to patient demographics, pathologic diagnosis, and operation performed. There was one death in the series due to a postoperative myocardial infarction in a patient in the control arm. The pancreatic fistula rate in the fibrin glue arm of the study was 26% vs. 30% in the control group (p = non-significant). The overall length of stay for all patients randomized was similar (fibrin glue = 12.2 days, control = 13.6 days) and the length of stay in patients that developed pancreatic fistula was also not different (fibrin glue = 18.9 days, control = 21.7 days). There were no differences with respect to total complications or specific complications such as postoperative bleeding, infection, or delayed gastric emptying. Conclusion: These data demonstrate that the topical application of fibrin glue sealant to the surface of the pancreatic anastomosis in this patient population undergoing high risk pancreatico- jejunal anastomosis did not reduce the incidence of pancreatic fistula or total complications after pancreaticoduodenectomy.