Abstract 6627: Clonal architecture analysis of response to PARP inhibition in prostate cancer

Abstract Background: Prior studies have charted the clonal and subclonal evolutionary dynamics (“clonal architecture”) of prostate cancer (PC), although the relationship between PC clonal architecture and therapeutic response remains unclear. PARP inhibition (PARPi) is an approved genomic strategy for treating PCs deficient in homologous recombination (HRD) as determined by specific gene alterations. As clinical responses to PARPi are heterogeneous and not exclusively associated with specific gene mutations, and HRD can exhibit discrete mutational and copy number signatures, we hypothesized that deriving overall tumor and clone-specific mutational signatures in prostate tumors may reveal novel associations between clonal architecture and clinical response to PARPi. Methods: We analyzed two independent cohorts of PC whole exome sequencing data that passed quality control metrics (Cohort 1: N=845 patients and 1 sample per patient, Cohort 2: N= 185 patients and 21 patients having multiple samples, enriched for patients with known deleterious BRCA2 alterations). For each tumor/normal paired sample, mutations and copy-number alterations were identified and used as input for phylogenetic reconstruction as well as single-nucleotide and copy-number mutational signatures. We examined the relationship between single gene germline or somatic alterations, tumor and clone-specific mutational signatures of HRD such as COSMIC mutational signature 3 (Sig3), and clinical response to PARP inhibition or immune checkpoint blockade. Here we report features of tumor and clone-specific Sig3 positivity (Sig3+) and association with clinical response in these two cohorts. Results: In Cohort 1, evidence of HRD at the tumor level by Sig3+ was detected in 2% (11/560) of primary tumors and 23.6% (67/285) of metastatic tumors. 31 Cohort 1 patients had germline or somatic alterations in BRCA2. 51.2% (40/78) of Sig3+ tumors had detectable Sig3+ at the individual and clone-specific level, including 12 tumors with only clone-specific Sig3+. Of the 37 patients in cohort 1 treated with PARPi and Sig3+ information, patients with detectable Sig3+ had a lower hazard for progression with treatment at the tumor level (HR = 0.60, p = 0.31), individual clone level (HR = 0.45, p = 0.095), or either (HR = 0.59, p = 0.20) compared to patient tumors with no detectable Sig3+. In Cohort 2, 68.4% (63/92) of primary and 76.7% (89/116) of metastatic tumors were Sig3+. 30 primary and 59 metastatic samples in Cohort 2, the majority of patients in each category, harbored deleterious BRCA2 alterations. 90.1% (137/152) of Sig3+ tumors had detectable Sig3+ at the individual clone level, including 18 with only clone-specific Sig3+. Analysis of clinical outcomes in Cohort 2 is ongoing. Conclusions: Clone-specific evidence of HRD using clonal architecture and mutational signature analysis is detectable and may identify additional PC patients who could benefit from PARPi. Citation Format: Matias Vergara, Alok K. Tewari, Jake R. Conway, Sabrina Y. Camp, Jihye Park, Brendan Reardon, Wassim Abida, David B. Solit, Nikolaus Schultz, Keyan Salari, Eliezer M. Van Allen. Clonal architecture analysis of response to PARP inhibition in prostate cancer [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2025; Part 1 (Regular Abstracts); 2025 Apr 25-30; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2025;85(8_Suppl_1):Abstract nr 6627.