OBJECTIVES: Missense mutations in androgen receptor ligand binding domain (AR-LBD-mut) and AR copy number amplifications (AR-CN-amp) are potential mechanisms of resistance to second-generation antiandrogen (ARSi-II) therapies among mCRPC patients. There is limited real world evidence (RWE) on these double-mutant mCRPC patients. Thus, we analyzed the genomic profile and clinical outcomes of these patients as compared to patients with no detectable AR mutations, in first line (1L) RWE setting.

METHODS: The GuardantINFORM database was used to identify adult mCRPC patients who received cell-free circulating tumor testing done via Guardant360 between June 1, 2014, and June 30, 2023. The AR-aberration–positive cohort (case) consists of patients with both AR-LBD-mut and AR-CN-amp. The absence of any aberration in the AR region was defined as AR–negative (control). Patients with index date (first Guardant360 test post mCRPC ascertainment) within 90 days of 1L treatment start were included. We compared the mutation profile and 1L RW time to treatment discontinuation (rwTTD) among the propensity score weighted cohorts (variables: age, race, Elixhauser Comorbidity Index, ctDNA level, index year and prior ARSi-II use).

RESULTS: 62 cases and 655 controls were included in the analysis. The patients in both groups had mean age of 70 years with similar comorbidities and prior ARSi-II use; however, differ significantly in their race and ctDNA levels (p<.0001). Matched cases showed shorter median rwTTD compared to the control cohort [4.4 months (95% CI 3.1-5.7) vs 5.6 months (95% CI 4.6-7.4), p= 0.017]. The oncoprint of the matched cases are enriched with FGFR1, CCNE1 and CDK6 as co-occurring mutations.

CONCLUSIONS: In this RW analysis, these double mutant mCRPC patients exhibit worse clinical outcomes and different genomic profile relative to patients without these mutations. Further investigation is needed to elucidate the significance of these AR-aberrations and improve prognosis for these patients.