Analysis of Matched Primary and Recurrent BRCA1/2 Mutation-Associated Tumors Identifies Recurrence-Specific Drivers

Patients with inherited germline mutations in BRCA1/2 carry a drastically increased risk of early-onset breast and ovarian cancers. BRCA1/2 mutation-associated tumors respond to therapies that exploit their inherent homologous recombination deficiency, including platinum-based chemotherapy and poly(ADP-ribose) polymerase inhibitors (PARPi). However, these tumors frequently return as lethal, therapy-resistant recurrences. Outside of somatic BRCA1/2 reversions, the mechanisms underlying acquired therapeutic resistance and recurrence remain unknown. To address this gap in knowledge, we performed whole exome, targeted, and RNA sequencing on paired primary and recurrent breast and ovarian tumors from 27 BRCA1/2 mutation carriers. The main outcomes of the study were somatic variants, copy number variation, BRCA1/2 loss of heterozygosity, differential gene expression, and differential transcript usage. One key finding was a high prevalence of copy number gains and amplifications in PARP1. We detected PARP1 gains across primary and recurrent BRCA1/2 mutation-associated breast and ovarian tumors, with increased expression observed at the mRNA and protein levels. Our results suggest that PARP1 gains could be an under-appreciated mechanism of endogenous PARPi resistance in BRCA1/2 mutation-associated and sporadic breast and ovarian tumors. We also assessed allele-specific BRCA1/2 loss of heterozygosity (LOH) across the cohort. In general, LOH status was concordant between paired primary and recurrent tumors. However, seven tumors underwent LOH transitions over the course of recurrence, suggesting that selective pressure drove most cancers towards biallelic BRCA1/2 loss but some towards BRCA1/2 proficiency. Lastly, we found that recurrent tumors express a shorter BRCA2 transcript. This non-canonical isoform is protein-coding and differs only in the 3’ UTR. Expression of the alternative BRCA2 transcript was significantly associated with reduced overall survival in BRCA1/2 mutation carriers with breast cancer (median 87 vs. 121 months). Our results indicate that the shorter BRCA2 isoform may represent a novel driver of recurrence in BRCA1/2 mutation-associated breast tumors. Ultimately, these findings improve our understanding of tumor evolution in BRCA1/2 mutation-associated cancers, including conserved drivers and other features that may contribute to therapeutic resistance. Ultimately, this work will improve our understanding of late-stage disease in BRCA1/2 mutation carriers, as well as inform new treatment options for these patients