OBJECTIVE: To determine if a viable cadaveric pancreas might be used to study viral transfection efficacy in a manner precisely mimicking in vivo human studies. DESIGN: Ex vivo gene transfer to an intact human pancreatic duct. SETTING: Molecular biology laboratory and organ procurement center. INTERVENTION: The recombinant adenoviral vector that contains the Escherichia coli beta-galactosidase (LacZ) gene driven by the human cytomegalovirus promoter, ie, AdCMVLacZ, was used to transfect the epithelial cells of the pancreatic ductal system. A human pancreas (150 g wt/wt) procured for transplantation, but subsequently found unsuitable, was used for the study. The splenic, superior mesenteric arteries and portal vein were cannulated and perfused in a heat-controlled organ procurement perfusion system. A segment of vascularized, perfused distal pancreatic duct was isolated with a balloon occlusion catheter. The recombinant adenoviral vector AdCMVLacZ was introduced into the lumen of the distal segment of the pancreatic duct and incubated for 6 hours at 25 degrees C. The proximal segment of the pancreatic duct was not exposed to the vector and served as control tissue. Tissue was harvested and processed for evaluation of beta-galactosidase activity. RESULTS: Adenoviral vector-infected pancreatic ducts exhibited intense blue staining, indicative of reporter gene expression in the epithelial cells of the pancreatic duct. The phenotype of these cells was confirmed by immunohistochemical studies using anti-annexin III polyclonal antibody. Control tissue not exposed to the adenoviral vector was subjected to an identical analysis and did not reveal evidence of expression of the reporter gene. CONCLUSIONS: This study demonstrates the first successful transfection of epithelial cells of the pancreatic duct from normal human pancreas with a recombinant adenovirus. This system will provide not only information on the efficacy of transfection but also a novel gene therapeutic approach to target pancreatic ductal adenocarcinoma.