Relationship among initial serum prostate specific antigen, prostate specific antigen progression and prostate cancer detection at repeat screening visits

PURPOSE: We evaluated the probability of positive serum PSA (3 ng/ml or greater) and CaP detection at annual followup visits in men with negative initial PSA (less than 3 ng/ml) to optimize the re-screening schedule. MATERIALS AND METHODS: Data on 5,387 men 45 to 80 years old with negative PSA and no CaP diagnosis at the first screening visit were obtained from the Laval University Prostate Cancer Screening Program database. Accelerated failure time regressions were fitted to time from baseline to positive PSA and to time from positive PSA to CaP detection. The models were combined to estimate the cumulative probability of positive PSA followed by CaP detection at re-screening. RESULTS: The 5-year cumulative probability of detecting CaP at annual visits in men with baseline PSA up to 1.5 ng/ml remained below 0.8%, while it was 1.3%, 4.8% and 8.3% in men with PSA 1.5 to less than 2, 2 to less than 2.5 and 2.5 to less than 3 ng/ml, respectively. Time to positive PSA significantly decreased with increasing baseline PSA and age, while the time between positive PSA and CaP detection depended only on age. Men with PSA below 1.0 ng/ml could wait for 4 to 5 years before being re-tested, while men with PSA between 1.0 and 1.5 ng/ml should be screened every second year and men with PSA 1.5 ng/ml or greater should be screened every year. CONCLUSIONS: The proposed retesting schedule using current PSA and age decreases the number of visits by 38.1%, while delaying the detection of only 2.4% of CaPs that would have been detected using annual PSA testing.