Impact of exogenous hormones on Prostate Cancer Risk, Detection, and Outcomes in the Nationwide Veterans Affairs Health System - PROJECT ABSTRACT Prostate cancer (PC), the most common and second most lethal cancer in men, is driven by androgens (i.e. testosterone). Indeed, the primary screening for PC is by prostate specific antigen (PSA), a protein produced from the prostate in an androgen-controlled fashion. Estrogen, typically found at low levels in men, may impact PC. Indeed, we have a DOD grant to look at the translational link between estrogen and PC. Another way to study this is by examining men who take estrogen (MTE), which some men do for various reasons. Estrogen inhibits androgen production leading to castration; the impact of this on PC screening, risk, and aggressiveness is largely untested. A prior study found MTE had an ~80% lower PC risk than men with no estrogen intake (NoE). We published in JAMA in 2023 using data from the Veterans Affairs Health System (VAHS) that MTE at PC diagnosis had higher grade disease and higher PSA density than NoE, suggesting delayed PC diagnosis. Intriguingly, only 6% of MTE with PC were African American vs. 29% of NoE veterans with PC, suggesting interactions involving race and estrogen. We propose that this delayed diagnosis is in part due to the castrating effects of estrogen leading to reduced PSA levels, as identified by a pilot study we published in JAMA in 2024. As such, to avoid delayed diagnosis, the threshold to define abnormal PSA in MTE should likely be much lower than traditional values currently in use among NoE. We hypothesize: 1) MTE are screened for PC at lower rates than NoE, 2) PSA cut-offs for MTE to define abnormal should be lower than in NoE, 3) these factors coupled with the biological effect of estrogen results in lower PC risk but creates harms such as higher risk of aggressive PC, 4) There are differences in contributory factors affecting PC diagnosis among MTE based on race and neighborhood environment factors. To test this, we will use the nationwide VAHS database and identify MTE via pharmacy records and verify by chart review. First, we will determine if there are differences in PSA screening in MTE vs. NoE and establish normal PSA values for MTE. We will compare age- and race-matched PSA screening rates between MTE and NoE accounting for factors that influence PSA screening (i.e. age, race, and neighborhood environment). Among MTE who had screening PSA tests done, we will determine age-appropriate normal values. Next, we will determine the reasons for lower PC incidence among MTE by comparing PC incidence between MTE and NoE accounting for PSA screening intensity, estrogen use (as a time-dependent covariate) and other factors associated with PC risk (age and race). Ultimately, at the conclusion of our study, we will have a far better understanding of the influence of estrogen and castration on PC biology as well as have data on how best to manage MTW to reduce the risk of delayed PC diagnosis and improve outcomes.
