Clinical application of liquid biopsies to detect somatic BRCA1/2 mutations and guide potential therapeutic intervention for patients with metastatic breast cancer

Plasma based genotyping via cell-free DNA may identify actionable mutations for potential therapeutic intervention in patients with advanced malignancies including breast cancer. In this article, we discuss recent studies using cell-free DNA testing to identify and classify somatic BRCA1/2 mutations in metastatic breast cancer, and potential future applications for the treatment of metastatic breast cancer.


INTRODUCTION
In recent years, plasma based genotyping via cellfree DNA (cfDNA) or "liquid biopsy" has emerged as a robust means to detect actionable mutations and guide genotype-directed therapies for patients with advanced malignancies [1]. For metastatic breast cancer, the utility of cfDNA in identifying actionable mutations has been demonstrated [2], and recent studies have validated the feasibility of tumor genotyping for targeted treatment matching [3]. Notably, a PI3K inhibitor has now been approved for PI3KCA mutant hormone receptor positive (HR+)/HER2-metastatic breast cancer using diagnostic cfDNA testing to identify PI3KCA mutations [4]. CfDNA offers the advantage of being less invasive and possibly more sensitive than tumor tissue genotyping assays [5,6].
PARP inhibitors have recently been approved as a targeted therapy for the treatment of germline BRCA1/2 mutant metastatic breast cancer based on results from two pivotal phase III clinical trials. The phase III OlympiAD [7] and EMBRACA [8] studies demonstrated significant improvement in progression-free survival with olaparib and talazoparib, respectively, compared to chemotherapy, for metastatic breast cancer. PARP inhibitors may also improve patient reported outcomes and quality of life. However, germline BRCA1/2 mutations only account for about 5-10% of breast cancer [9], limiting their broad applicability.
A question that arises is whether PARP inhibitors may also be beneficial in metastatic breast cancer with somatic BRCA1/2 mutations. In a recent study led by our team [10], we demonstrated the emergence of somatic BRCA1/2 mutations detectable by cfDNA, largely in the absence of germline BRCA1/2 mutations. In our analysis of 215 patients undergoing cfDNA testing as part of care for metastatic breast cancer, we observed that 29 (13.5%) had somatic BRCA1/2 mutations detectable in cfDNA, which were seen in various subtypes, and often clonal in nature. Four percent had somatic BRCA1/2 mutations that were known germline-pathogenic, and the rest were novel variants, based on classification using extrapolation from reputable genomic databases. In addition, we demonstrated increased sensitivity to a PARP inhibitor in a CTC culture derived from a patient with a pathogenic somatic BRCA1 mutation, highlighting the unique role of PARP inhibition for patients with somatic BRCA1/2 mutant breast cancer. However, not all somatic BRCA1/2 mutations are functionally significant, i.e., pathogenic. For example, in a CTC culture derived from a patient with a novel variant BRCA2 mutation, we did not observe any impact of a PARP inhibitor. Interestingly, this patient also had widespread expression of the APOBEC mutation signature that encompassed the BRCA2 mutation itself, suggesting the novel BRCA2 mutation was likely a passenger mutation rather than a driver mutation. Based on these findings, we developed an algorithm to guide clinical assessment of somatic BRCA1/2 mutations, as well as designed a genotype-directed clinical trial for patients with metastatic breast cancer [11].
A genotype-directed clinical trial is currently ongoing at our institution and other academic centers to determine the efficacy of PARP inhibition in somatic

Research Perspective
Oncotarget 64 www.oncotarget.com cfDNA BRCA1/2-mutant metastatic breast cancer [11], and the results of this trial may help expand the population of patients who benefit from PARP inhibitors, similar to what has been observed in ovarian cancer. A combined analysis of two studies evaluating a PARP inhibitor in ovarian cancer demonstrated similar efficacy in germline BRCA1/2 mutant patients and somatic BRCA1/2 mutant patients [12].
Besides detection of germline and somatic BRCA1/2 mutations, cfDNA analysis also allows for detection of reversion BRCA1/2 mutations [13]. The acquisition of BRCA1/2 reversion mutations is a well described phenomenon [14], which restores the open reading frame (and function) of the BRCA1/2 gene, thus rendering a PARP inhibitor ineffective. In a second multicenter analysis [13], we demonstrated that routine plasma-based genotyping can be utilized to classify BRCA1/2 cfDNA mutations as germline, somatic or reversion mutations, based on specific loci and the mutant allele fraction of the BRCA1/2 mutation. Of 828 patients with advanced malignancies including breast cancer undergoing testing with a 73 gene cfDNA assay, one or more pathogenic BRCA1/2 mutation was identified in 7.2% of patients, and both somatic and germline variants were detected. Polyclonal reversion mutations were found in 21.4% of patients with germline BRCA1/2 mutations, most often in association with receipt of a prior PARP inhibitor.
Another study found that the absence of pre-existing cfDNA BRCA1/2 reversion mutations in patients with ovarian cancer who had somatic or germline BRCA1/2 mutations and were treated with rucaparib was associated with improved progression-free survival [15]. Therefore, the identification of cfDNA BRCA1/2 reversion mutations may have important implications for therapeutic response to PARP inhibitors, and will be studied in our ongoing clinical trial of a PARP inhibitor for somatic BRCA1/2 mutant metastatic breast cancer [11]. A complementary ongoing trial is also evaluating the efficacy of the PARP inhibitor, olaparib, in somatic BRCA1/2 mutant metastatic breast cancer, with initial results suggesting potential efficacy [16].
In summary, plasma-based genotyping is a promising strategy for the detection of BRCA1/2 mutations and could potentially guide triage to genotype-directed matched therapy with a PARP inhibitor. Ongoing studies will help determine the therapeutic utility of this approach and impact on long term clinical outcomes for patients with metastatic breast cancer.