Objectives: In watchful waiting patients with localized prostate cancer, the prostate-specific antigen doubling time (PSADT) has been suggested to correlate with disease progression and is often used to determine the need for therapy. In these studies, all available prostate-specific antigen data throughout the observation period were used to determine the PSADT. However, the correlation between the short-term and long-term assessments of PSADT in this population is unknown. Our aim was to determine the correlation between short-term PSADT measurements and long-term PSADT among men with prostate cancer on observation.
Methods: In a prospective cohort of 108 watchful waiting patients, the correlation of the overall PSADT with the PSADT measured during short-term intervals was analyzed. The level of agreement in classifying patients as having a rapid PSADT (24 months or less) was also evaluated.
Results: The median interval of observation was 36 months. According to the various methods used, the median PSADT was 82 months using all prostate-specific antigen values, 128 months using the first three values (early), 78 months using the last three values (late), and 144 months using the first and last values only (2-point). The PSADT measured over short-term intervals demonstrated a weak correlation with the overall PSADT (Spearman's rho 0.53 for early, 0.56 for late, and 0.50 for 2-point PSADT). A statistically significant agreement was noted in the classification of rapid PSADT between the overall PSADT measurement and early PSADT measurement (kappa 0.67). Less agreement was found in the classification, however, with the late and 2-point PSADT and the overall PSADT.
Conclusions: In untreated patients with localized prostate cancer, the PSADT varies considerably depending on the interval evaluated.