Molecular Diagnosis of : Review Series Guest Editors: J.P. Neoptolemos (Liverpool); T.M. Gress (Marburg)

Pancreatology 2009;9:34–44 Published online: December 12, 2008 DOI: 10.1159/000178873

Identification of Malignancy Factors by Analyzing Cystic Tumors of the Pancreas

a, d b a b b Andrea Bauer Jörg Kleeff Melanie Bier Martin Wirtz Hany Kayed c e d a Irene Esposito Murray Korc Mathias Hafner Jörg D. Hoheisel b Helmut Friess a b Division of Functional Genome Analysis, Deutsches Krebsforschungszentrum, Division of General Surgery, c d and Department of Pathology, University of Heidelberg, Heidelberg , Division of Biotechnology, Mannheim e University of Applied Science, Mannheim , Germany; Departments of Medicine, and Pharmacology and Toxicology, Dartmouth Hitchcock Medical Center and Dartmouth Medical School, Lebanon, N.H. ,USA

Key Words analysis of rare types of pancreatic cancer, which are less fre- Pancreatic ؒ Cystic lesions ؒ Expression profiling ؒ quent in terms of disease, variations could be identified that siRNA silencing ؒ Fas-activated serine/threonine kinase could be critical for the regulation of malignancy and thus relevant to the treatment of also the majority of pancreatic tumors. Copyright © 2008 S. Karger AG, Basel and IAP Abstract Aim: The diversity in the aggressiveness of cystic tumors of the pancreas – ranging from the usually benign serous cyst- adenoma to lesions of variable degrees of malignancy – was Introduction utilized for the identification of molecular factors that are involved in the occurrence of malignancy. Methods: We an- Pancreatic is the fifth most frequent alyzed the transcript profiles of different cystic tumor types. cause of cancer-related death in the industrialized world. The results were confirmed at the protein level by immuno- The highly malignant ductal adenocarcinoma (PDAC) histochemistry. Also, functional studies with siRNA silencing accounts for approximately 90% of pancreatic tumors. were performed. Results: Expression variations at the RNA However, other less frequent tumor entities exist, which and protein level were identified that are closely correlated have a generally much better prognosis [1] . While repre- with the degree of malignancy. Besides, all tumors could be senting with about 5% only a small portion of pancreatic classified effectively by this means. Many of the identified tumors, cystic neoplasms are of particular interest, since factors had not previously been known to be associated with they actually constitute a variety of lesions with a wide malignant cystic lesions. siRNA silencing of the gene with spectrum of aggressiveness and clinicopathological fea- the most prominent variation – the anti-apoptotic factor tures [2, 3] . Therefore, these tumors could be suitable for FASTK (Fas-activated serine/threonine kinase) – revealed a identifying factors, which contribute to disease progres- regulative effect on several genes known to be relevant to sion. Cystic neoplasms are classified into three subgroups: the development of tumors. Conclusion: By a molecular serous cystic neoplasms (SCA), mucinous cystic neo-

© 2008 S. Karger AG, Basel and IAP Andrea Bauer 1424–3903/09/0092–0034$26.00/0 Functional Genome Analysis, Deutsches Krebsforschungszentrum Fax +41 61 306 12 34 Im Neuenheimer Feld 580 E-Mail [email protected] Accessible online at: DE–69120 Heidelberg (Germany) www.karger.com www.karger.com/pan Tel. +49 6221 424 882, Fax +49 6221 424 687, E-Mail [email protected] plasms (mucinous cystadenoma – MCA, and mucinous plicate spots on each array, four hybridizations per sample inclu- cystadenocarcinoma – MCAC) and intraductal papillary sive dye-swap) were accumulated. Quantification of the signal in- tensities was done with GenePix Pro 4.1 analysis software (Axon mucinous tumors (intraductal papillary mucinous neo-