HRD signature and HRD genomic landscape of tumors from 896 patients with early-stage breast cancer (BC).

Abstract

539 Background: Pathologic mutations in BRCA1, BRCA2 and PALB2 lead to impaired homologous recombination repair (HRR). Cells with defects in HRR are dependent on non-homologous DNA end joining for DNA repair through the poly(ADP-ribose) polymerase (PARP) catalytic activity. PARP inhibitors (olaparib, talazoparib) are FDA-approved therapies that are toxic to tumors with HRR deficiency (HRD) and are recommended as adjuvant therapy in high-risk germline BRCA (g BRCA) positive BC. Despite the clinical implications of HRD, the prevalence of HRR alterations and HRD signature in early-stage primary BC and its association with hormone receptor (HR) status has not been well characterized. Methods: This study used the nationwide (US-based, ~280 US cancer clinics) de-identified Flatiron Health-Foundation Medicine BC clinico-genomic database and included patients (pts) who underwent tissue comprehensive genomic profiling (FoundationOne/FoundationOneCDx) between 2014 and 2022. The analysis included all pts with BC who presented with “early” stage I-III disease), had a primary tissue specimen collected within 3 months of diagnosis, and had a reportable novel scar-based HRD signature (HRDsig). Results: A total of 896 primary BC pts with stage I-III met our inclusion criteria. Of these early BC pts 21% (188/896) were HRDsig+ (stage I: 17% (32/188); stage II: 22% (84/376); and stage III: 22% (72/332)). For early BC pts, high rates of HRDsig+ were seen in g BRCA (87% [34/39]), somatic BRCA (s BRCA, 78% [14/18]), and g PALB2 (71% [5/7]) mutated tumors, with a lower rate seen with other HRR genes [16% (10/63); ATM, BARD1, BRIP1, CDK12, CHEK1, CHEK2, FANCL, sPALB2, RAD51B, RAD51C, RAD51D and RAD54L]. HRDsig+ was also detected in 16% (135/832) of BC without g/s BRCA or g PALB2 mutations. BC pts who presented with stage I-III disease has similar prevalence of HRDsig+ when compared to those with stage IV disease (21% [188/896] vs 23% [177/767], p = 0.31), with similar trends in HR+ disease (HR+/HER2+: 7% [4/57] vs 9% [4/45], p = 0.73; and HR+/HER2-: 17% [78/466] vs 18% [82/463], p = 0.73). Conversely, in HR- cohorts, decreased prevalence of HRDsig+ was observed in the stage I-III vs stage IV cohort (HR-/HER2+: 3% [1/38] vs 18% [6/33], p = 0.04; HR-/HER2-: 31% [99/321] vs 38% [83/216], p = 0.08). Conclusions: In this study, HR+/HER2- group was the most common early BC subtype, followed by HR-/HER2-, HR+/HER2+, and HR-/HER2+. This distribution is similar to the SEER study, and closely reflects that of the general population. We observed the expected high rates of HRDsig+ in early-stage BC pts with g/s BRCA or g PALB2 mutations where clinical trial data already support the use of PARP inhibitors for early and advanced disease. Importantly, HRDsig positivity was also seen in 16% of early BC without g/s BRCA or g PALB2, and clinical trials will be needed to determine if these pts could also benefit from HRD directed therapies.