Molecular marker for the prognosis of pancreatic carcinoma
C. Pilarsky1, R. Grützmann1, D. Aust2, C. Winter3, T. Knösel4, P. Rümmele5, H D Saeger1, G. Kristiansen6
1Klinik für Chirurgie, Universitätsklinikum Dresden, Dresden
2Institut für Pathologie, Universitätsklinikum Dresden
3Institut für Bioinformatik, TU Dresden
4Institut für Pathologie, Universitätsklinikum Jena
5Institut für Pathologie, Universitätsklinikum Regensburg
6Institut für klinische Pathologie, Universitätsspital Zürich
Aims: Prognosis for patients with pancreatic carcinoma (PDAC) remains poor. Despite of increasing knowledge about the molecular basis of PDAC neither a specific marker for early diagnosis nor a target protein for a new therapeutic approach have been identified so far. However, reliable prediction of survival might help to improve the therapeutic outcome for patients with PDAC.
Methods: We analysed the gene expression profile of 29 PDAC using the Affymetrix GeneChip U133 2.0 array. We correlated gene expression values with postoperative survival time using a novel computational approach which incorporates background knowledge. Candidate genes were further validated using RT-PCR and immunohistochemistry of tissue microarrays containing cores from 458 patients with PDAC.
Results: We identified a signature of 8 genes, which were able to predict the survival of the patients with 70% accuracy. RT PCR and immunohistochemistry confirmed the findings of our gene expression experiments for most of the genes.
Conclusions: The combination of gene expression analysis, novel bioinformatic approaches and large-scale immunohistochemistry with TMAs is suited to identify novel marker genes for prognosis of PDAC patients.