Abstract
BACKGROUND: Agent Orange exposure (AOE) is considered a presumptive cause of prostate cancer (PCa) in the Veterans Affairs (VA) population; however, cohort studies reported inconsistent associations of AOE and PCa incidence and outcomes. In this nationwide cohort study, Vietnam veterans who received VA care were evaluated for associations of AOE and PCa incidence and adverse outcomes. METHODS: Vietnam veterans 17-25 years old during military service (1962-1971) who received VA health care between 2005 and 2020 were stratified by presumptive AOE assigned by the VA and followed until death from any cause or censoring on September 31, 2023. Multivariable Cox models permitted estimation of adjusted hazard ratios (aHRs) of AOE with PCa incidence, de novo metastasis (DNM), any metastasis, metastatic castration-resistant PCa (mCRPC), all-cause mortality (ACM), or PCa-specific mortality (PCSM). RESULTS: Among 2.6 million Vietnam veterans at risk for PCa, 779,472 (30%) had AOE. Compared to unexposed veterans, AOE veterans had higher PCa risk (aHR, 1.15; 95% confidence interval [CI], 1.15-1.16), higher DNM (aHR, 1.17; 95% CI, 1.16-1.17), any metastasis (aHR, 1.17; 95% CI, 1.16-1.17), mCRPC (aHR, 1.17; 95% CI, 1.16-1.17), ACM (aHR, 1.41; 95% CI, 1.41-1.42), and PCSM (aHR, 1.17; 95% CI, 1.16-1.17) in adjusted models. Sensitivity analyses suggested robustness of association between AOE and all-cause mortality, but selection bias could explain associations with PCa outcomes. CONCLUSIONS: Presumptive AOE was associated with higher PCa incidence, mortality, and adverse outcomes. Although associations may not be causal, AOE may predict worse PCa outcomes in the Veterans Affairs.