Ered at the ampulla of Vater. Though classified by the World Health Organization as cancers of the extrahepatic bile duct, inhibitor ampullary adenocarcinomas have better prognosis when compared 10781694 to similarly staged pancreatic or biliary adenocarcinomas. [1?] Three distinct epithelial linings (duodenal, biliary, and pancreatic) converge at the ampulla of Vater, with pancreatic andbiliary epithelium merging within the ampulla of Vater to form a true ampullary epithelium. Thus, it is uncertain whether adenocarcinomas originating at the ampulla of Vater represent a homogenous carcinoma group reflective of a true ampullary epithelium or a heterogeneous group reflective of these various epithelial origins. Given the uncertain epithelial origin of ampullary adenocarcinomas, a number of studies have attempted to identify prognostically differing subtypes. The first approach to subtypeGene Profiling of Periampullary Carcinomasampullary adenocarcinomas was based upon segregating cases by histology as either pancreaticobiliary type or intestinal type. [4] Though a number of studies have found this approach to have statistically significant prognostic impact [5?], other studies have not [9?1]. More recently studies have investigated additional markers such cytokeratin expression, mucin expression, microsatellite instability, and intestinal-specific markers to identify prognostically distinct subgroups of ampullary adenocarcinomas. [5,7,10?6] For example, expression of the intestinal markers, CDX-2 and CDX-1, were recently shown to correlate with improved OS in a cohort of 53 patients [13], but this finding was not validated in subsequent studies [5,12]. Though these studies taken together have been suggestive of heterogeneity within ampullary adenocarcinomas, interpretation of these results has been limited by small sample size and variability in classification methodology. Thus, at present, no single method has consistently identified prognostically relevant subgroups of ampullary adenocarcinomas. In order to improve the understanding of the heterogeneity within ampullary adenocarcinomas, we sought to classify ampullary adenocarcinomas at a molecular level by comparing the mRNA gene expression from clinically-annotated specimens of ampullary adenocarcinomas to the expression patterns of pancreatic, duodenal, and biliary adenocarcinomas. In addition transcriptional profiles were compared to patient characteristics and clinical outcomes. The patterns of the expression and activation of proteins in signaling networks were also assessed using reverse phase protein arrays (RPPA). This study shows a molecular distinction between ampullary and pancreatic adenocarcinomas, identifies robust prognostic subgroups of ampullary adenocarcinomas, and implicates a number of targetable signaling pathways in the pathogenesis of these tumors.labeled, and 10mg of cRNA was hybridized to the HG-U133 Plus 2.0 Affymetrix GeneChip array according the manufacturer’s protocol (Affymetrix, Santa Clara, CA). RMA (Robust Multichip Average) expression values were calculated from the microarray data and the hierarchical clustering was performed using ward linkage and Pearson correlation distance with probsets that were called present on at least 3 arrays. [17] One-way ANOVA was used to identify genes that are differentially expressed in at least one tissue type. The p-values from one-way ANOVA were modeled using a beta-uniform inhibitor mixture (BUM) model. With the use of a false discovery rate (F.Ered at the ampulla of Vater. Though classified by the World Health Organization as cancers of the extrahepatic bile duct, ampullary adenocarcinomas have better prognosis when compared 10781694 to similarly staged pancreatic or biliary adenocarcinomas. [1?] Three distinct epithelial linings (duodenal, biliary, and pancreatic) converge at the ampulla of Vater, with pancreatic andbiliary epithelium merging within the ampulla of Vater to form a true ampullary epithelium. Thus, it is uncertain whether adenocarcinomas originating at the ampulla of Vater represent a homogenous carcinoma group reflective of a true ampullary epithelium or a heterogeneous group reflective of these various epithelial origins. Given the uncertain epithelial origin of ampullary adenocarcinomas, a number of studies have attempted to identify prognostically differing subtypes. The first approach to subtypeGene Profiling of Periampullary Carcinomasampullary adenocarcinomas was based upon segregating cases by histology as either pancreaticobiliary type or intestinal type. [4] Though a number of studies have found this approach to have statistically significant prognostic impact [5?], other studies have not [9?1]. More recently studies have investigated additional markers such cytokeratin expression, mucin expression, microsatellite instability, and intestinal-specific markers to identify prognostically distinct subgroups of ampullary adenocarcinomas. [5,7,10?6] For example, expression of the intestinal markers, CDX-2 and CDX-1, were recently shown to correlate with improved OS in a cohort of 53 patients [13], but this finding was not validated in subsequent studies [5,12]. Though these studies taken together have been suggestive of heterogeneity within ampullary adenocarcinomas, interpretation of these results has been limited by small sample size and variability in classification methodology. Thus, at present, no single method has consistently identified prognostically relevant subgroups of ampullary adenocarcinomas. In order to improve the understanding of the heterogeneity within ampullary adenocarcinomas, we sought to classify ampullary adenocarcinomas at a molecular level by comparing the mRNA gene expression from clinically-annotated specimens of ampullary adenocarcinomas to the expression patterns of pancreatic, duodenal, and biliary adenocarcinomas. In addition transcriptional profiles were compared to patient characteristics and clinical outcomes. The patterns of the expression and activation of proteins in signaling networks were also assessed using reverse phase protein arrays (RPPA). This study shows a molecular distinction between ampullary and pancreatic adenocarcinomas, identifies robust prognostic subgroups of ampullary adenocarcinomas, and implicates a number of targetable signaling pathways in the pathogenesis of these tumors.labeled, and 10mg of cRNA was hybridized to the HG-U133 Plus 2.0 Affymetrix GeneChip array according the manufacturer’s protocol (Affymetrix, Santa Clara, CA). RMA (Robust Multichip Average) expression values were calculated from the microarray data and the hierarchical clustering was performed using ward linkage and Pearson correlation distance with probsets that were called present on at least 3 arrays. [17] One-way ANOVA was used to identify genes that are differentially expressed in at least one tissue type. The p-values from one-way ANOVA were modeled using a beta-uniform mixture (BUM) model. With the use of a false discovery rate (F.