In Vitro Characterization and Clinical Correlation of BRCAness as a Personalized Biomarker for Radiosensitization With Homologous Recombination Directed Therapies

Abstract

BRCA1/2-mutant tumors demonstrate sensitivity to cisplatin, a cross-linking agent, but only mild sensitivity to ionizing radiation (IR). Cisplatin with concurrent IR is one of the most common methods to improve outcomes with radiotherapy. However the mechanism of synergy and the population of patients who benefit is poorly understood. Recently, BRCAness (homologous recombination deficiency; HRD) has been described in the absence of germline BRCA1/2 mutations. We hypothesize that BRCAness tumors exhibit sensitivity to cisplatin + IR (c-IR), and that HRD is a significant driver of the observed clinical benefit of combination therapy. Sensitivity to c-IR was examined in vitro in isogenic cell line pairs and sporadic breast cancer cell lines. HR function was assessed in H1299 cells with BRCA1 inactivation and DLD1 cells with BRCA2 knockout using 3 assays: radiation-induced Rad51 foci, a flow-based DR-GFP DNA repair reporter assay and a clonogenic survival (CSA) with cisplatin +/- IR. Interaction ratios (INTR) were calculated to identify super-additivity as follows: (cisplatin survival x IR survival)/c-IR survival. Four sporadic, non-BRCA1/2 mutated breast cancer cell lines were identified using the Cancer Cell Line Encyclopedia (CCLE) and evaluated for HRD with these 3 assays. Since higher rates of HRD are observed in triple-negative breast cancer (TNBC), we also examined 10 TNBC patients treated with neoadjuvant c-IR on a clinical trial and assessed HRD by ex-vivo radiation-induced formation of Rad51 foci. The BRCA1/2 isogenic pairs demonstrated functional HRD based on the Rad51 and DRGFP assays. CSA results demonstrate no super-additivity of c-IR in HR proficient lines (INTR 0.5), but super-additivity in HRD cell lines (INTR 17.3). A genomic scar score, LST (large scale transition), was generated for all breast cancer cell lines based on CCLE data, with higher LST scores suggesting HRD. Of the 4 selected cell lines, the 2 with low LST demonstrated functional HR and additive c-IR sensitivity (INTR 1.4, 1.7), while the 2 with high LST demonstrated HRD and super-additive sensitivity to c-IR (INTR 3.4, 4.9). Of 6 evaluable TNBC patients treated with neoadjuvant c-IR, three each demonstrated complete pathological response and partial response. Multiple in vitro assays suggest that BRCA1/2-mutated tumors are much more sensitive to the widely used combination of c-IR as compared with HR proficient tumors. BRCAness cell lines demonstrate a similar but milder sensitivity to c-IR. Therefore, reliable methods for the determination of “BRCAness” should identify a subpopulation of non-germline BRCA1/2-mutated breast cancer patients who may benefit from therapy with c-IR. The value of BRCAness testing is likely to extend beyond breast and ovarian cancer as additional cancers with HRD are identified.