Real-world prevalence of homologous recombination repair alterations (HRRa) and poly (ADP-ribose) polymerase inhibitor (PARPi) use/outcomes in patients (pts) with metastatic prostate cancer (mPC) by race and ethnicity.

Abstract

5077 Background: Racial/ethnic (R/E) disparities in prostate cancer incidence and outcomes have long been noted. Increasing evidence points to underlying genetics/biology as potential drivers, which could influence response to targeted tx such as PARPi. Here, we used real-world evidence (RWE) to assess R/E variations in the prevalence of HRRa and outcomes of PARPi therapies in pts with mPC. Methods: This retrospective, RWE study used GuardantINFORM, a deidentified clinical genomic database combining claims and genomic data reported from the liquid biopsies, Guardant360 (G360) CDx (HRRa: BRCA1/2, ATM, CDK12, MLH1 ) and LDT (HRRa: BRCA1/2, ATM, CDK12, CHECK2, FANCA, MLH1, PALB2, RAD51D ). Pts with mPC, age >18, and tested between July 2014 - Sep 2024 were included and grouped into non-Hispanic Black (NHB), non-Hispanic White (NHW), Asian/Other, and Hispanic (any race) cohorts. HRRa and frequency of PARPi treatment (tx) were assessed with two-sided Fisher’s exact tests. Outcomes were assessed for PARPi tx overall, as monotherapy, or in combination with androgen receptor pathway inhibitor (ARPI) using pairwise comparisons and log-rank tests. Results: 30,913 pts with G360 liquid testing were identified. The rates of HRRa on the LDT panel were significantly higher for NHW (22.8%) when compared to NHB (18.1%, p<0.0001) and Hispanic (19.6%, p=0.017) cohorts. This was similarly seen on the CDx panel for NHW (16.7%) vs. NHB cohorts (15.0%, p=0.018). Similar BRCA1/2 detection rates (range 5.2-5.8%) and frequencies of PARPi use were observed among R/E groups. No significant differences were observed in real world overall survival (rwOS), time to tx discontinuation, or time to next tx among R/E groups. While not significant, there was a numerical rwOS improvement for NHB when compared to NHW pts treated with PARPi/ARPI combinations (CDx panel, median NR vs. 20.5m, p=0.09), with no NHB pts known to be deceased at the end of follow up (median 10.5 months). Conclusions: Using a RWE dataset, we demonstrate distinct rates of HRRa between R/E groups identified using ctDNA analysis. Similar PARPi use among R/E groups suggest tx equity for pts receiving genomic sequencing. The numerically improved rwOS for NHB pts on combination PARPi/ARPI tx is hypothesis generating. This may support other studies demonstrating distinct outcomes in Black vs. White pts treated with ARPI, with potential added benefit in a combination paradigm. Cohort G360, n LTD:HRRa, % LTD: PARPi use in HRRa pts, % CDx: HRRa, % CDx: PARPi use in HRRa pts, % LTD+CDx: BRCA1/2+, % LTD+CDx: PARPi use in BRCA 1/2+ pts, % All mPC 30,913 21.5 28.0 16.2 29.1 5.6 37.5 NHB 3,216 18.1 30.5 15.0 33.1 5.8 41.9 NHW 19,482 22.8 28.1 16.7 28.8 5.6 38.6 Asian/Other 1,150 20.3 28.9 15.7 28.0 5.7 36.1 Hispanic (any race) 2,152 19.6 30.0 15.3 32.1 5.2 37.5