Purpose: Studies have shown that finasteride decreases prostate specific antigen (PSA) by approximately 50% during the first 12 months of use. We estimated the long-term effects of finasteride on PSA in men with and without a prostate cancer diagnosis at the end of the study.
Materials and methods: We analyzed serial PSA in participants in the Prostate Cancer Prevention Trial who had an end of study biopsy (928 with cancer and 8,620 with negative biopsy) or an interim diagnosis of prostate cancer (671). Linear mixed effects regression models were fit to longitudinal PSA values beginning 1 year after randomization.
Results: In subjects with no cancer in the end of study biopsy PSA in the finasteride arm showed a median annual decrease of 2% [corrected] after year 1, while PSA in the control arm showed an annual increase of 3% (p <0.001). In end of study cases PSA increased annually by 6% (placebo) and 7% (finasteride). In those with interim diagnoses PSA increased by 11% (placebo) and 15% (finasteride) each year prior to diagnosis. Cases with high grade disease (Gleason 7 and above) had greater PSA increases than cases with low grade disease (p <0.001).
Conclusions: In men who have been receiving finasteride for more than 1 year time varying adjustment factors may be needed to determine whether PSA is in the normal range. In the Prostate Cancer Prevention Trial cohort the adjustment factor required to preserve median PSA increased from 2 at 24 months to 2.5 at 7 years after the initiation of finasteride.