) The objectives of this pilot study are to characterize in a statistically precise manner the behavior of long term longitudinal prostate specific antigen (PSA) levels in subjects with carcinoma of the prostate (CP), in subjects with benign prostatic hyperplasia (BPH), and in subjects without either diagnosis. Two time frames over which the statistical modeling is planned are prior to diagnosis, and prior to treatment. The subjects are chosen from the VA Normative Aging Study (VANAS), a long term Boston area cohort study of 2,200 male veterans with triannual medical exams and frozen serum samples stored for future analyses. Over 80 participants have been diagnosed with CP. A similar number of participants have been randomly selected from among subjects diagnosed with BPH, and the same number randomly selected from subjects without either diagnosis. These three groups form the study population for this proposal. Three components are required for this project: (i) completing measurement of longitudinal PSA levels in all participants in the study group, (ii) obtaining clinically relevant information via chart reviews on all study participants, and (iii) a careful analysis based on hierarchical longitudinal statistical models for PSA levels over time for each group and for each time frame. Important covariates such as age will be incorporated in all such models. The long term objectives of this research, which are outside the scope of this project, are multiple. These include determining an optimized early detection strategy for CP based on longitudinal PSA levels which utilizes the baseline PSA level jointly with the rate of increase in PSA measurements, inclusion of longitudinal """"""""free PSA"""""""" levels in the strategy, development of software which facilitates clinical implementation of such strategies, measurement and analysis of longitudinal PSA levels in the entire VANAS cohort for refined statistical characterization, and measurement of disease specific quality of life indices for prostate disease and subsequent correlation with longitudinal PSA behavior.
