Comparing the efficacy of abiraterone and enzalutamide in United States veterans with metastatic castration-sensitive prostate cancer (mCSPC) by estimating rates of tumor growth (g-rate)

5076 Background: The choice of 1 st line novel hormonal therapy in mCSPC is often based on comorbidities/physician preference due to the lack of data comparing abiraterone (Abi) and enzalutamide (Enza) in clinical trials and in real-world settings; especially in Black American (BA) patients. We have developed a novel method to estimate rates of tumor growth ( g-rate) and have shown that the g-rate is a robust biomarker of drug efficacy and overall survival in prostate cancer (Wilkerson, 2017; Leuva, 2019). We performed a 1:1 matched analysis to compare Abi and Enza outcomes in 1 st line mCSPC, using g-rate and median overall survival (mOS) estimates. Methods: All patients with mCSPC diagnosis in national VA Corporate Data Warehouse were eligible for comparison. The cohort date range was from 7/2017 to 6/2022. Drug efficacy was established by estimating g-rates using PSA values while on therapy via the TUMGr package for R. For the matched analyses, we created 1:1 comparison cohorts for Abi and Enza, matched on age (±10 years), race, total number of therapies received (1/≥2), drug start year (<2020/≥2020), PSA at diagnosis (<20/≥20), Gleason score (GS) (<8/≥8), Charlson comorbidity index excluding cancer diagnosis (<5/≥5) and therapy setting (urban/rural). Results: We identified a total of 1712 and 464 Veterans with mCSPC who received Abi and Enza as 1 st line, with median follow-ups of 19 and 12 mo, respectively. Fewer Veterans and shorter duration of follow up with Enza than Abi reflect Abi’s earlier approval in mCSPC. The Abi and Enza cohorts had similar median g-rates 0.000114/d and 0.000110/d, respectively (p=0.76). The mOS was 43 mo for Abi versus not reached (NR) for Enza (p=0.33). Our cohort had 425(25%) and 119(26%) BA veterans who received Abi and Enza, respectively. Compared to Caucasians, mOS was similar with Abi (42.7 vs 41.9 mo, HR-0.96, p=0.6) and Enza (NR, HR-0.85, p=0.4) for BA. In 1:1 matched analysis, the Abi and Enza arms had 391 patients and both had median follow-ups of 13 mo. We did not observe a statistically significant difference in median g-rate and mOS for either the whole matched cohort or BA subgroup when comparing Abi to Enza. The matched cohort had 203 (52%) with GS ≥8 and 228 (58%) Veterans with PSA ≥20 at diagnosis; there was no significant mOS difference in these subgroups. Conclusions: Real-world data from 2176 Veterans show comparable efficacy for Abi and Enza as 1 st line for mCSPC, with similar g-rates and mOS, even when matched on various factors, including comorbidities, and in aggressive/high volume matched cohorts. BA patients also had similar g-rates and mOS. [Table: see text]