Prostate cancer is the most common malignancy among males in the United States. Very few molecular prognostic markers are available to assist in clinical management and classify patients for clinical trials. Whereas for other cancers large-scale screening of tissues for oncogene rearrangement or abnormal expression have identified genes that may play a pathogenic role in those cancers, no oncogenes have been found to be associated with for prostate cancer. Our experiments are aimed at identifying molecular markers and activated oncogenes that indicate poor prognosis in prostate cancer. We have identified several growth factor genes that are expressed in hormone-independent, but not hormone-dependent cell lines. Cell lines manifest phenotypic differences which characterize malignant progression in clinical cancer. Molecular markers that correlate with malignant progression in vitro will be tested for expression in tumor specimens. We will perform PCR analysis of prostate tumor tissue cDNA to screen for the expression of these factors in fresh tissues. We will perform immunohistochemistry on paraffin-- embedded tumor tissue to look for erbB-2 and vimentin expression in prostate cancer cells. We will analyze prostate tumor DNA to detect mutations or rearrangements of three oncogenes, p53, lyt- 10, and hox- 11. Marker expression and immunohistochemical analysis will be correlated with pathologic findings and histologic grade of the tumor specimens. Although we will readily test material supplied to us through the cooperative network, we actively accrue freshly frozen surgical material to a prostate tissue bank which is cross-indexed to demographic information and pathology records and blocks. The tumor bank is comprised of specimens from both black (35%) and white (65%) men. Identification of specimens by racial origin will allow us to explore any racial differences in the relevance of molecular markers or oncogene activation.
