Introduction:

The excess risk of high-grade prostate cancer in the Prostate Cancer Prevention Trial (PCPT) among men randomized to finasteride raised concerns that a chemoprevention benefit would be offset by increased mortality. Two papers published in 2019 addressed whether 5-alpha reductase inhibitors (5-ARIs) were associated with an increased risk of prostate cancer mortality with contrasting results. In the follow-up paper for PCPT, there was no excess prostate cancer mortality among men who were cancer-free at the beginning of the study period. In a study of 80,875 men with prostate cancer conducted through the Veterans Affairs (VA) database, 5-ARI use was associated with later stage at presentation and a 39% higher risk of prostate cancer death. We studied men with regular access to health care and screening to address this conflicting evidence.

Methods:

We performed two parallel analyses in the Health Professionals Follow-up Study, a cohort of male health professionals recruited in 1986. 5-ARI use was reported on biennial questionnaires starting in 1996. In line with PCPT, we studied 38,046 cancer-free men for cancer incidence through 2016 and mortality through 2018. In line with the VA, we undertook a case-only study of 4,232 men with localized/locally advanced cancers followed over a similar period. Using stratified proportional hazards models, we estimated hazard ratios (HR) and 95% confidence intervals (95% CI) accounting for screening patterns, lifestyle factors, and concomitant medications for both the overall cohort as well as the case-only study.

Results:

4,480 (11.8%) men ever used 5-ARIs. These men were slightly older than nonusers (64.0 vs. 62.8 years), and more likely to have had recent PSA screening (73% vs. 57% nonusers) and a prior prostate biopsy (20% vs. 9% nonusers). In multivariable models, we found no association between lethal prostate cancer and ever use of 5-ARIs (HR 1.04, 0.71-1.50, 482 events), nor any association for longer-term 5-ARI use (≥4 years, HR 1.03, 0.56-1.90). 5-ARI use was not associated with additional risk of advanced-stage (HR 1.03, 0.73-1.45) or high-grade prostate cancer (HR 1.02, 0.79-1.31), but men who used 5-ARIs had a 31% reduced risk of localized disease (HR 0.69, 0.58-0.81). Among men with non-metastatic tumors at diagnosis, 302 had lethal prostate cancer; there was no association between 5-ARI use and prostate cancer survival (HR 0.76, 0.46-1.26) or overall survival (HR 0.84, 0.68-1.04), even after stratifying by duration of use. 

Conclusion:

Our results align with PCPT and show no excess risk of lethal or advanced-stage prostate cancer and no excess mortality after diagnosis associated with 5-ARI use, regardless of duration of use. In fact, the men using 5-ARIs had a reduced risk of localized or low-grade disease which might not be clinically significant. This study provides evidence that these medications could be a viable option for chemoprevention as they reduce the burden of unnecessary healthcare interventions and do not lead to excess mortality. Men on 5-ARIs had more interactions with health care providers, which may have influenced their clinical outcomes. This contrasts with the results from the VA where men on 5-ARIs were diagnosed later, potentially because of suppressed PSA and, therefore, had worse survival. Taken together, the results from all three studies emphasize the importance of real-world data in framing the benefits and risks associated with 5-ARIs.

Funding: Cohort is supported by U01 167552