5-Alpha Reductase Inhibitor Use and Prostate Cancer Prevention: A Victim of the Times?

The story of 5-alpha reductase inhibitors (5-ARI) and their potential role in prostate cancer prevention has journeyed a winding road. From splash excitement, through realization of potential harms, to reasoned scientific rebuttal, to health authority black box, then gradual demonstration of long-term safety, to now the potential irrelevance because of the drastically changed way we currently diagnose and treat prostate cancers. Sadly, after decades of study, deciding whether to start a 5-ARI solely to prevent prostate cancer remains a challenge.

In this issue of Cancer Epidemiology, Biomarkers & Prevention, Vaselkiv and colleagues present, in my view, the strongest evidence to date of long-term safety of 5-ARI use (1). Capitalizing on the robust prospective cohort that is the Health Professionals Follow-up Study, they demonstrated that self-reported 5-ARI use was not associated with an increased risk of developing an advanced prostate cancer, nor dying of prostate cancer. Moreover, they showed what we have come to expect of 5-ARI use: a reduced risk of developing low-grade prostate cancer.

There are several strengths of this study worth mentioning. Because of the serial participant questionnaires, the authors were able to perform analyses based on duration of use and confirmed that even among longer-term users (≥4 years), no increased risk of advanced disease or prostate cancer death was seen. They had detailed data on PSA screening history, PSA testing intensity, prior prostate exams, and biopsies. This enabled multivariate adjustment and a separate sensitivity analysis to address the potential issue that 5-ARI users may have increased contact with urologists and the health care system in general that could lead to biased results. With detailed family history data, they performed an analysis restricted to this higher-risk group (i.e., family history positive) and found similar results to the overall analysis. Finally, by observing no preventive association with alpha-blocker use, another class of medications used to treat benign prostate symptoms, they strengthened their observed 5-ARI benefits as being less likely due to confounding by indication.

Despite representing the most compelling safety data to date, the study is not without limitations. Even among this group of health professionals, arguably a more homogenous group, there are clear differences between 5-ARI users and nonusers. Measured differences include substantially higher PSA screening (72% vs. 57%), prior prostate exam (78% vs. 69%) and prior prostate biopsy (19% vs. 9%). While the ability to adjust for these measured differences is a strength, the risk of residual confounding is real. This is particularly true when looking at distal outcomes like mortality where years of more active health seeking could theoretically overcome a biological disadvantage induced by 5-ARI use.

That said, there exists now enough published data to state that 5-ARI use: (i) decreases risk of low-grade prostate cancer; (ii) has no effect on intermediate-grade disease; and (iii) has either no effect or slight increase in risk of high-grade disease, but insufficient population attributable risk to negatively influence mortality.

Why then are 5-ARIs not used more for cancer prevention. Are 5-ARIs victims of the times? In 2003, when the Prostate Cancer Prevention Trial (PCPT) investigators first announced a 25% reduction in prostate cancer incidence in the finasteride arm (1), the “culture” of prostate cancer diagnosis was moving toward lower and lower PSA thresholds triggering biopsies with the perceived benefits of finding all cancers as early as possible. Thus, there was understandable excitement.

However, this was quickly doused by the concerns 5-ARIs may induce high-grade tumors, as reported in the PCPT and the companion Reduction by Dutasteride of PCa Events (REDUCE) trial (2, 3).

While tempted to prescribe 5-ARIs, clinicians had concerns about potentially inducing high-grade disease in otherwise healthy men (4). Over time, the high-grade concerns were diminished—explained away by two plausible theories: (i) 5-ARIs shrink prostate volume, making it more likely to find high-grade disease when it is present (5); and (ii) 5-ARIs, by reducing confounding from benign prostatic hyperplasia (BPH), heighten the sensitivity of PSA and DRE in detection of high-grade disease (6, 7). However, just as clinicians were perhaps feeling more comfortable with the concept, the FDA Oncology Drugs Advisory Committee reviewed the PCPT and REDUCE results and concluded the risks of high-grade cancer were likely real and could not be explained away by the proposed theories (8). This led to altering the labels of finasteride and dutasteride to reflect the risk of high-grade disease.

With the passage of time, longer term follow-up on larger cohorts of 5-ARI users has shown no adverse safety signal. Data from Sweden (9–11), the United Kingdom (12), and the United States (13, 14) endorse no increased grade and no worse survival. While one large U.S. study did show 5-ARIs were associated with higher-grade disease and worse prostate cancer–specific mortality, this study has been devalued as it appeared the 5-ARI group had significant delays in diagnoses likely because clinicians in the VA system did not account for the halving of PSA that occurs with 5-ARIs (15).

So, with the addition of the current study by Vaselkiv and colleagues, will now the world feel more comfortable prescribing 5-ARIs for cancer prevention? Would health authorities reconsider their black box warnings? Should clinicians hesitant to prescribe 5-ARIs for BPH or male pattern hair loss relax their concerns?

For the latter, with the building evidence in favor of safety, clinicians should feel more at ease prescribing 5-ARIs for the approved indications of BPH and hair loss without concerns of inducing high-grade prostate cancer.

For cancer prevention though, perhaps the point is moot: the bigger question is whether preventing low-grade disease matters anymore. The answer maybe “no,” for two reasons. First, increasingly, the world is moving toward surveillance as the preferred management for low-grade disease. A first glance at U.S. data may underwhelm this fact, with utilization of active surveillance for low-risk prostate cancer varying from 40% to 60% depending on geography (16, 17). However, utilization in the U.S. VA system is much higher at 70% to 80% (18), and mirrors practice in other countries such as Sweden (74%) and Canada (70%) where active surveillance is the predominant management choice (19, 20).

Second, with increased incorporation of MRI and blood and urine-based biomarkers into prebiopsy risk stratification (21), the focus of urology practice has shifted from trying to find all cancers to finding only clinically significant cancers and avoiding biopsy of clinically insignificant tumors (i.e., those that would be prevented by 5-ARIs).

As it appears we have finally mounted enough evidence to support the utilization of 5-ARIs for prostate cancer prevention, the field may have moved on. Exposing healthy men to a medication, for years, only to prevent something we increasingly care less to find, seems inappropriate. Perhaps 5-ARIs are a victim of the times.

No disclosures were reported.