A SYSTEMATIC APPROACH TO EARLY DIAGNOSIS OF CANCER

Prostate cancer is the second leading cause of cancer death in men, killing approximately 32,000 men annually. An urgent need exists for improved diagnostic and prognostic tools. PSA tests are widely used diagnostic tools, but they yield an unacceptable degree of false positives and false negatives. Due to a lack of prognostic tools, many men undergo unnecessary, radical treatments. We describe herein a systematic program for the discovery of secreted prostate tumor biomarkers. In Phase I, we developed a ?secretion trap?, a protocol for rapid isolation of a library of secreted sequence tags from any source. Since a ?trapped? library identifies accessible proteins, e.g. cell surface and extracellular proteins, it is a powerful means for discovering biomarkers with diagnostic and prognostic potential. We will generate a library of secreted sequence tags from prostate tumors and we will choose 100 that fulfill basic criteria required of useful biomarkers. We will determine the expression profiles of the candidate markers by automated, high-throughput in situ hybridization and immunohistochemistry. Finally, we will determine whether any of these biomarkers are present in the serum of patients with prostate cancer. These experiments will identify biomarkers that may be commercialized as diagnostic or prognostic indicators.