Genome-wide cDNA microarray analysis of gene expression profiles in pancreatic cancers using populations of tumor cells and normal ductal epithelial cells selected for purity by laser microdissection

To characterize molecular mechanism involved in pancreatic carcinogenesis, we analysed gene-expression profiles of 18 pancreatic tumors using a cDNA microarray representing 23,040 genes. As pancreatic ductal adenocarcinomas usually contain a low proportion of cancer cells in the tumor mass, we prepared 95% pure populations of pancreatic cancer cells by means of laser microbeam microdissection, and compared their expression profiles to those of similarly purified, normal pancreatic ductal cells. We identified 260 genes that were commonly upregulated and 346 genes that were downregulated in pancreatic cancer cells. Because of the high degree of purity in the cell populations, a large proportion of genes that we detected as upregulated or downregulated in pancreatic cancers were different from those reported in previous studies. Comparison of clinicopathological parameters with the expression profiles indicated that altered expression of 76 genes was associated with lymph-node metastasis and that of 168 genes with liver metastasis. In addition, expression levels of 30 genes were related to the recurrence of disease. These genome-wide expression profiles should provide useful information for finding candidate genes whose products might serve as specific tumor markers and/or as molecular targets for treatment of patients with pancreatic cancer.