Background: Cohort 2 of the TALAPRO-2 (NCT03395197) study demonstrated benefit in radiographic progression-free survival (rPFS) with talazoparib (TALA) plus enzalutamide (ENZA) (n=200) vs placebo (PBO) + ENZA (n=199) across gene subgroups in men with HRRm receiving first-line treatment for mCRPC. Post-hoc analyses aimed to understand PROs by HRR gene clusters. Methods: PROs were assessed at day 1 and scheduled visits (every 4 weeks until week 53, then every 8 weeks) until radiographic progression using the EORTC QLQ-C30 and PR25 and BPI-SF. Prespecified PRO endpoints included overall mean change from baseline (per longitudinal repeated measures mixed-effects model) and time to definitive deterioration (TTDD) with a clinically meaningful change of ≥10-points for the EORTC QLQ-C30. Stratified log-rank test and Cox proportional hazards model were used to make TTDD between-arm comparisons. Mutually exclusive gene clustering alteration dominance hierarchy was applied in the following order: any BRCA1/2 alteration (BRCAm cluster), PALB2 (PALB2 cluster), CDK12 (CDK12 cluster), ATM (ATM cluster), then any of all other HRR7genes (MLH1, CHEK2, NBN, FANCA, ATR, RAD51C, MRE11A). Results: A significantly longer TTDD in GHS/QoL was observed for TALA + ENZA vs PBO + ENZA in the BRCAm (HR=0.54, 95% CI (0.29, 0.99); p=0.043; median NE vs 19.0) and CDK12 clusters (HR=0.43 (0.19, 0.98); p=0.019, median 30.7 vs 16.6). Overall changes from baseline in GHS/QoL and worst pain are reported in the Table. Conclusions: PRO findings by HRR gene clusters are consistent with rPFS benefit analyses by gene clusters. This was an exploratory post-hoc analysis limited by missing PRO assessments and sample sizes especially in the PALB2 cluster. Clinical trial information: NCT03395197.

Positive scores indicate better GHS/QoL; negative scores indicate better pain scores. Means are estimated least-square (LS) means from mixed effect models that adjust for baseline score.

*Sample size = (n1 = TALA + ENZA), (n2 = PBO + ENZA).