? Our hypothesis is that the patterns of low molecular weight proteins and peptide fragments in serum are altered by the presence of prostate cancer even when the cancer is still contained within the prostate. We propose to use an artificial intelligence-based pattern recognition algorithm to identify expression patterns of serum protein and peptide fragments that discriminate men with prostate cancer from those men with benign prostates. The goal of this study is to develop a serum proteomic based detection method that will help determine the need for prostate biopsy for men with serum PSA levels between 2.5 and 10.0 ng/ml. ? ? Aim 1: Determine the protein chip surface and mass spectrometer instrument that generates the serum proteomic profile that best discriminates men with prostate cancer from those with benign prostates. ? ? Aim 2: Determine the sensitivity, specificity and positive predictive value for biopsy-detected prostate cancer of the optimized serum proteomic algorithm among men with serum total PSA levels of 2.5 to 10.0 ng/ml. ? ? Aim 3: Compare the performance of serum proteomic profiling to PSA density and percent free PSA measurements for men with serum total PSA levels of 2.5 to 10.0 ng/ml. ? ? Aim 4: Determine the biological variability of serum proteomic patterns by analyzing serum samples collected from the same patient on 3 consecutive days. ? ?
