Back to Program | 2010 Program and Abstracts Overview | 2010 Posters
Prediction of Prognosis By Expression Profiling Analysis of Pancreatic Ductal Adenocancinoma Using Dna-Arrays
Robert GrüTzmann*1, Marco Niedergethmann2, Moritz Wente3, Helmut Friess4, Glen Kristiansen6, Christof Winter7, Marcus Bahra5, Petra Ruemmele8, Hans D. Saeger1, Christian Pilarsky1
1Surgery, University Hospital Dresden, Dresden, Germany; 2Surgery, University Hospital Mannheim, Mannheim, Germany; 3Surgery, University Hospital Heidelberg, Heidelberg, Germany; 4Surgery, University Hospital Munich, Munich, Germany; 5Surgery, University Hospital Berlin - Charité, Berlin, Germany; 6Pathology, University Hospital Zurich, Zurich, Switzerland; 7Bioinformatics, Technical University Dresden, Dresden, Germany; 8Pathology, University Hospital Regensburg, Regensburg, Germany
Prognosis for patients with pancreatic ductal adenocarcinoma (PDAC) remains poor. Despite of increasing knowledge about the molecular basis of PDAC no specific marker for early diagnosis nor a target protein for a new therapeutic approach have been identified so far. Surgery is the only potentially curative option. But even after R0-resection of PDAC, the prognosis is bad and most patients suffer from recurrences and metastases. Moreover, the individual prognosis is not known. This might be of nterestfor indication of adjuvant therapy. We were therefore interested in the analysis of differential gene expression between PDAC with relatively good and worse prognosis.We used fresh frozen tissue from 29 patients with pancreatic ductal adenocarcinoma. From every single patient, the clinical characteristics, pathological data as well as follow up has been collected prospectively. The tissues were obtained during surgery and freshly frozen. The type of each frozen tissue sample was re evaluated pathologically. The RNA was extracted using the RNeasy Mini Kit. The quality of the RNA was assessed using the Agilent Lab on a Chip System and only samples displaying a RIN > 4 were used. For hybridization we used 100 ng of total RNA, and samples were prepared according the Affymetrix two cycle amplification labelling protocol. Samples were hybridised to Affymetrix U133 2.0 plus GeneChips. The obtained data from the microarray were analysed using dCHIP.Median survival of the patients was 13 (2-53) months. Using this time point we classified the samples into two groups. Differentially gene expression analysis (FC > 2, difference of means > 100; p value < 0.05) revealed 21 probe sets. Hierarchical clustering using these 21 probe sets displayed two clusters of samples. One cluster contained only samples from patients with a survival time < 13 months. Sensitivity and specificity calculations based on the cluster data resulted in 100 % sensitivityand 73 % specificity for the detection of patient samples with a survival > 13 month. Several genes have been valedated using tissue microarrays.In conclusion, gene expression analysis of the tumor tissue of PDAC enables prediction of prognosis with high specificity and good sensitivity. The role of adjuvant treatment has to be elucidated. Moreover, using the set of differentially expressed genes we might identify new markers and therapeutic targets for pancreatic cancer.
Back to Program | 2010 Program and Abstracts Overview | 2010 Posters
