BRCA functional domains associated with PARP-inhibitors in patients with prostate cancer: The PROGRESS study.

10543 Background: Homologous recombination repair (HRR) defects are driver mutational imprints and actionable biomarkers in prostate cancer (PrC) patients. However, recent research provides evidence of PrC onset also in the context of mismatch repair deficiency (MMRd). In the current report, we investigated whether the position of pathogenic variants (PVs) in the functional domains (FDs) of HRR genes could be associated with reduced or increased sensitivity to PARP-inhibitors (PARPis). We also questioned whether the HRR FDs had a prognostic significance and affected the development of second primary tumor (SPT) in males with a genetic predisposition. Methods: This was a real-world, observational, study including PrC patients undergoing germline, somatic, and liquid biopsy testing between January 2020 and December 2023. Prognostic factors, SPT history, and PARPi benefit were evaluated according to mutated genes ( BRCA1, BRCA2,No- BRCAHRR), PV location within the FDs, and germline (g) or somatic (s) PV. The BRCA1FDs were RING, DNA-binding domain (DBD), BRCA1 C terminus (BRCT), or other location, and the BRCA2 FDs were RAD51-BD, DBD, or other location. Results: A total of 833 metastatic PrC patients, aging 40 to 91, were included in this study; 169 (20.3%) were carriers of germline (n.48, 5.8%) or somatic (n.121, 14.5%) PVs in HRR genes: 17 in BRCA1 (g BRCA1: n.4, 2.4%; s BRCA1: n.13, 7.7% ), 102 in BRCA2 (g BRCA2: n.33, 19.5%; s BRCA2: n.69, 40.8% ), and 53 in no- BRCA HRR genes (gHRR: n.11, 6.5%; sHRR: n.42, 24.8%). CHEK2was the HRR gene most frequently affected by germline PVs (27.3%), while the most frequent somatic PVs were in the ATM (40.5%). Liquid biopsy identified PVs in 33 patients out of 185 tested, including n.10 (5.4%) not detected through somatic and germline testing, considerably expanding the therapeutic window for the use of PARPi. Notably, differences in PFS rates at 48 months were observed when comparing sub-groups according to PV location in the FDs of BRCA2.Surprisingly, when we investigated the SPT in the cohorts of patients with and without germline PVs, the difference was not significant (14.5% vs 12.4%, respectively) and, in the subgroup without germline PVs, the sites of SPT were predominantly colorectal, gastric, bladder or other tumors classically dominated by MMRd and recognized as Lynch Syndrome-spectrum tumors. This observation may result in unexplored forms of hypermutable tumor phenotypes, potentially impacting cancer risk management and increasing therapeutic opportunities. Conclusions: Our results suggest preliminary evidence that not all BRCA2-mutated PrC are equally sensitive to PARPis, and the response could depend on the PV location within FDs. The study of SPT in patients with no-HRR-associated PrC highlights the need to expand the genes in multigene panel testing, including MMR along with HRR genes, to capture unrecognized constitutional predisposition.