Abstract 1260: Nonclinical evaluation of rucaparib in tumors with mutations in non-BRCA1/2 homologous recombination repair (HRR) genes

Abstract

Abstract Background: Rucaparib is a PARP inhibitor approved to treat ovarian cancer (OC) and metastatic castration-resistant prostate cancer (mCRPC). Deficiencies in HRR resulting from genetic or epigenetic alterations in BRCA1/2 sensitize tumors to rucaparib through synthetic lethality. Inactivation of other HRR genes may also confer sensitivity to PARP inhibition. Here, we evaluated rucaparib efficacy in preclinical models with genomic or epigenetic alterations in an HRR gene panel. Methods: BRCA1/2 and 14 additional genes were selected for their role in DNA repair and mutation frequency in OC and mCRPC. Small interfering RNA (siRNA) knockdowns were performed in OC (OVCAR-3, OAW28, SK-OV-3) and mCRPC (DU145, PC-3) cell lines to model their inactivation and assess impact on rucaparib sensitivity (IC50) in cell viability assays. Patient-derived xenografts (PDX) harboring deleterious alterations in BRCA1/2, BARD1, BRIP1, NBN, PALB2, RAD51B, RAD51C, or RAD51D, or BRCA1/RAD51C promoter hypermethylation were treated with up to 150 mg/kg rucaparib BID for 25-42 days and tumor growth was monitored. Mutational status was confirmed by next-generation sequencing. Results: In addition to BRCA1/2 knockdown, depletion of BARD1, PALB2, or RAD51 resulted in potent rucaparib activity in OC and mCRPC cells, where the rucaparib IC50 was reduced by >70% vs control in at least 4 of the 5 cell lines tested. Rucaparib also displayed greater potency in cells with siRNA-mediated decreased expression of FANCA, NBN, RAD51C, or RAD54L, where at least 1 cell line had >50% decrease in IC50. Rucaparib treatment in 19/33 HRR-deficient PDX models resulted in significant tumor growth inhibition (TGI). These included BRCA1/2, PALB2, NBN, RAD51B, RAD51C, or RAD51D-mutated and BRCA1/RAD51C-hypermethylated models. Efficacy was strongly associated with biallelic inactivation of BRCA1/2, PALB2, RAD51C, or RAD51D with 87% vs 29% mean TGI for monoallelic BRCA1/2 alterations (P<0.0001). In addition, a single PDX model each for the exploratory genes NBN and RAD51B had 101% and 80% TGI, respectively. No significant difference was observed between PDX models with BRCA1/2 vs non-BRCA1/2 (PALB2, NBN, RAD51B, RAD51C, or RAD51D) biallelic alterations (98% vs 80% mean TGI, respectively; P=0.21). In support of these nonclinical findings, we highlight a patient with RAD51B-mutant mCRPC treated with rucaparib in the TRITON2 trial (NCT02952534) who had confirmed radiographic and PSA responses. Conclusions: Cell lines and PDX models with biallelic alterations in HRR genes other than BRCA1/2, including RAD51B, RAD51C, RAD51D, PALB2, and NBN, are comparably sensitive to rucaparib. Rucaparib is being evaluated in the LODESTAR trial (NCT04171700) in patients with tumors associated with deleterious alterations in BRCA1/2, PALB2, RAD51C, and RAD51D (cohort A) and BARD1, BRIP1, FANCA, NBN, RAD51, and RAD51B (exploratory cohort B). Citation Format: Liliane Robillard, Kevin K. Lin, Andrea Loehr, Tanya Kwan, Rachel Dusek, Andrew D. Simmons, Thomas C. Harding, Brieuc Sautois, Minh Nguyen. Nonclinical evaluation of rucaparib in tumors with mutations in non-BRCA1/2 homologous recombination repair (HRR) genes [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2022; 2022 Apr 8-13. Philadelphia (PA): AACR; Cancer Res 2022;82(12_Suppl):Abstract nr 1260.