Evaluation of changes in prostate specific antigen in clinically localized prostate cancer managed without initial therapy

Purpose: We define changes in prostate specific antigen (PSA) measurements with time in 49 men 71.9 +/- 7.0 years old (mean plus or minus standard deviation) with clinically localized prostate cancer who remain untreated. Materials and Methods: We retrospectively analyzed PSA changes in prostate cancer patients managed by watchful waiting. In all patients a minimum of 3 PSA levels were measured at intervals of at least 6 months after malignancy was diagnosed. The rate of change in serum PSA level with time (PSA velocity) was determined using an exponential, log linear model. Results: In 49 patients treated conservatively mean initial PSA level plus or minus standard deviation was 12.3 +/- 11.1 ng./ml. and mean PSA followup during which no therapy for prostate cancer was introduced was 32.1 +/- 13.2 months. PSA levels decreased during the observation period in 11 of the 49 patients (22%) and median PSA doubling time in the remaining 38 was 55.7 months (range 15.1 to 994.5). There was no significant correlation between age at diagnosis, Gleason sum, initial PSA level or clinical stage and PSA velocity. The short-term rate of change in PSA during the first 9 months after prostate cancer was diagnosed correlated poorly with overall PSA velocity. The short-term rate of PSA change was greater than the overall rate of change in 14 of 37 patients (38%). Conclusions: There is significant variability in the rate of change of PSA with time in men with clinically localized prostate cancer who remain untreated. The usefulness of serial PSA measurements in the management of watchful waiting is unclear. Changes in PSA may not be helpful or appropriate in determining the need for therapy after a period of observation.