Background: With multiple phase 3 trials showing improved overall survival (OS) with doublet therapy (androgen-deprivation therapy [ADT]+ARPI), it has become the standard of care for pts with mCSPC. However, data on RW effectiveness of ARPI doublets are limited. This study assessed OS, time to progression to metastatic castration-resistant prostate cancer (mCRPC), and PSA decline to <0.2 ng/mL in pts with de novo mCSPC on ADT alone vs ADT+ARPI doublets or ADT+first-generation nonsteroidal antiandrogens (NSAAs). Methods: Men in the Veteran’s Health Administration with ≥1 ICD code for PC and de novo mCSPC diagnosis confirmed by clinical chart review were identified and categorized into first-line (1L) ADT alone, ADT+NSAA, or ADT+ARPI cohorts (study period: Feb 2018−Mar 2023). Target recruitment quotas ensured representation of each treatment (Tx) regimen. Index date was the ADT initiation date. OS and time to mCRPC were calculated using inverse probability of Tx weighting (IPTW)-adjusted Cox regression. IPTW-adjusted incidence rate ratios (IRR) of the proportion of pts with PSA decline to <0.2 ng/mL at any time during 1L Tx were estimated using Poisson regression. Results: 384 men with de novo mCSPC were identified (Table). Median follow-up range was 34.8–38.1 mo. 1L Tx duration for ADT+ARPI was over twice as long as ADT alone (Table). Importantly, pts on ADT+ARPI had a 39% lower risk of death vs ADT alone (HR: 0.61, 95% CI: 0.43–0.87); while OS was similar between ADT+NSAA vs. ADT alone (HR: 1.09, 95% CI: 0.79–1.49). Consistent with the improved OS, ADT+ARPI showed a 54% lower risk of progression to mCRPC (HR: 0.46, 95% CI: 0.33–0.66) and significant superiority in achieving PSA <0.2 ng/mL (57% vs 17%; Table) than ADT alone. Conclusions: This is one of the first RW studies to validate clinical trial findings of the effectiveness of ARPI-doublets for de novo mCSPC as evidenced by significantly improved OS, time to mCRPC, and PSA response. ADT+NSAA may have limited value vs ADT alone, and ARPI-doublet is an effective Tx for de novo mCSPC.

^aProportionality assumptions violated for comparative analysis of 3 groups; ^bADT alone, n=161; ADT+NSAA, n=100. NE, not estimable; Ref, reference.