1812P Exploration of Germline (g) Vs Somatic (s) Origin of Homologous Recombination Repair (HRR) Gene Alterations and Potential Associations with Antitumor Activity in the HRR-deficient Population from TALAPRO-2

TALAPRO-2 evaluated the PARP inhibitor talazoparib (TALA) combined with the androgen receptor pathway inhibitor enzalutamide (ENZA) as 1st-line treatment in patients (pts) with metastatic castration-resistant prostate cancer (mCRPC). The primary endpoint was radiographic PFS (rPFS) assessed centrally. Exploratory biomarker analyses assessed g vs s origin of HRR gene alterations and potential associations with efficacy in the HRR-deficient population from TALAPRO-2. Alterations were defined as known/likely pathogenic variants. The frequency of gHRR alterations (gHRRm) was assessed in pts evaluable for saliva sequencing using the Ambry CustomNext-Cancer panel (sequenced 9/12 genes from the HRR12 tumor panel). Association of g vs s tumor HRRm and efficacy was assessed in the ITT population using saliva data (or solid tumor s-g-zygosity prediction when adequate saliva results unavailable) to annotate tumor alteration origin (limited to short variants). Associations with rPFS were evaluated using a stratified log-rank test. Data cutoff: October 3, 2022. 91/302 pts (30.1%) evaluable for gHRR exhibited ≥1 HRR alteration in their saliva: BRCA2 (14.9%); CHEK2 (7.3%); ATM (4.0%); PALB2 (2.3%); BRCA1 (1.0%); NBN (1.0%); MRE11A (0.7%); MLH1 (0.3%). Where evaluable, solid tumor and prospectively tested baseline ctDNA HRR alterations were typically of s origin (eg BRCA2 [44 g vs 50 s], CDK12 [1 g vs 27 s], ATM [14 g vs 49 s]) with some exceptions (eg BRCA1 [7 g vs 4 s], CHEK2 [24 g vs 2 s], NBN [5 g vs 2 s]). rPFS was comparable for pts with g and s alterations in the TALA + ENZA arm (HR 0.74, 95% CI, 0.35–1.58; 2-sided P value = 0.44), but relatively shorter for gHRRm pts in the placebo (PBO) + ENZA arm (HR 1.87, 95% CI, 1.06–3.30; P = 0.03). GHRRm incidence in mCRPC pts with tumor HRRm was 30.1%. No association with differential antitumor activity for TALA + ENZA was observed in tumors bearing HRRm of g vs s origin. Though the analyses were exploratory, a shorter rPFS in PBO + ENZA-treated gHRRm pts may reflect an intrinsically poor prognosis subgroup that could be prioritized for TALA + ENZA treatment. Further study is warranted.