Homologous recombination deficiency (HRD) is a common phenotypic alteration that is highly druggable with poly (ADP-ribose) polymerase inhibitors (PARPi). Although BRCA1/2 gene mutations are among the commonest genomic aberrations associated with HRD, defects in other DNA damage repair (DDR) genes also may influence clinical response to PARPi. Here, we report the case of a 51-year-old Chinese woman with extensive symptomatic brain metastases from metastatic BRCA1/2 wild-type triple-negative breast cancer (TNBC). Comprehensive genomic profiling (CGP) of resected central nervous system tumor revealed mutations in TP53 and multiple DDR pathway genes, suggesting an HRD phenotype. The patient showed a rapid and remarkable response to single-agent niraparib, and her improved condition remained stable for > 8 weeks. To the best of our knowledge, this is the first report of the use of CGP for guiding targeted therapy with PARPi in patients with TNBC, for which options have been limited. CGP may have an increasingly impactful role to predict clinical response of PARPi in patients with BRCA1/2 wild-type TNBC.
Comprehensive genomic profiling (CGP) enables molecular classification of various cancers for personalized genomic-guided therapies. Homologous recombination deficiency (HRD) is a common phenotypic alteration in cancers, and is druggable by platinum chemotherapy and poly (ADP-ribose) polymerase inhibitor (PARPi). Of various genomic aberrations that may lead to HRD, BRCA1/2 mutations are the commonest, especially in breast and ovarian cancers. The use of PARPi in patients with ovarian cancer and breast cancer with germline BRCA1/2 mutations produced remarkable improvement in progression-free survival.
Despite its commonality in breast cancers, BRCA1/2 mutations were found in only 20%-30% of patients with triple-negative breast cancers (TNBC). Besides BRCA1/2 mutations, many gene mutations (eg, RAD51, PALB2) are associated with HRD. Detection of these mutations with next-generation sequencing (NGS)–based CGP may potentially identify more patients who might benefit from PARPi therapy.
We present a case of remarkable and rapid response to niraparib, a central nervous system (CNS)-active PARPi, in a patient with recurrent, BRCA1/2 wild-type TNBC with symptomatic, high-volume brain metastases. To our knowledge, this is the first report of the use of CGP for guiding targeted therapy with PARPi in TNBC.
A 51-year-old Chinese woman with recurrent/advanced TNBC was referred to our center in August 2018 for assessment of multiple brain metastases. The patient provided written informed consent for her case to be published. She was first diagnosed with breast cancer in June 2015 (Fig. 1). In October 2017, a 2-year follow-up positron emission tomography/computed tomography (PET/CT) scan revealed multiple axillary, interpectoral, subcarinal and bilateral hilar lymph nodes and lung metastases. Recurrent TNBC was confirmed pathologically (estrogen receptor [ER]/progesterone receptor [PR]/human epidermal growth factor receptor 2 [HER-2]-negative, Ki67 60%) and treated with capecitabine from January to August 2018. In late August, the patient developed persistent dizziness. Magnetic resonance imaging (MRI) revealed multiple, high-volume brain metastases in both cerebral hemispheres, the cerebellum, and the pons (Fig. 2a). An emergency palliative excision of the left cerebellar and left basal frontal cystic metastases was performed in October 2018 in view of the high intracranial pressure.
Following craniotomy, the patient received whole brain radiation therapy in November 2018 but her general condition remained poor. Despite dexamethasone therapy, she remained largely bedbound with Karnofsky performance score (KPS) at 30, and suffered from sacral sore and recurrent sepsis. In December 2018, a repeat MRI revealed more than 10 cystic metastases, and rapid progression of the unresected right cerebellar and left frontal lobe tumors. Further surgery was deemed inadvisable because of her poor performance status. High metastatic volume and rapid progression despite whole brain radiation therapy indicated that further radiosurgery boost would be of limited value. The patient became dependent on high-dose dexamethasone due to severe cerebral edema caused by extensive brain metastases (Fig. 2b).
Genomic profiling via somatic NGS gene panel (ACTOnco+, ACT Genomics Laboratory, Taiwan), programmed death-ligand 1 protein (PD-L1) immunohistochemistry (22C3 pharmDx assay, Agilent Technologies, Santa Clara, CA), and germline mutation testing (ACTRISK, ACT Genomics Laboratory, Taiwan) were conducted in parallel to identify possible systemic treatment strategies, although the extracranial disease load was still relatively small. Somatic NGS profiling of the resected CNS tumor tissue revealed TP53 R213 mutation and multiple heterozygous deletions of known cell-cycle, checkpoint, and DDR genes, including BAP1, CHEK1, ERCC1, MLH1, and RAD50 (Table 1). Copy number increase or amplification in other cancer-related genes that are also associated with the HR pathway, such as MYC,SMARCA4, and BRD4, was also found. Germline mutation testing showed no BRCA1/2 mutation or other common pathogenic variants. Nonetheless, these test findings suggest a possible HRD phenotype. PD-L1 expression scores—tumor proportion score and combined positive score—were < 1%.
Based on CGP of the brain metastases, two treatment options were proposed: combination therapy with chemotherapy plus atezolizumab (a PD-L1 immune checkpoint inhibitor [ICI]), or single-agent PARPi. Because of the patient's poor general condition at that time (KPS 30), the family perceived chemotherapy would be too toxic to the patient. Of the two PARPi available in Hong Kong at this time, niraparib and olaparib, niraparib was chosen because of its higher CNS penetrance and accumulation compared with olaparib in breast and ovarian cancer xenograft models. In December 2018, oral niraparib was initiated at 100 mg daily, which was 50% of the recommended initial dose.
The patient responded remarkably to niraparib 100 mg daily within 2 weeks and was able to discontinue dexamethasone completely within 4 weeks despite high-volume intracranial metastases. Her KPS score improved to approximately 70, and dysphasia was largely resolved. In the next 2 months, intracranial lesions showed moderate decrease in size and extracranial disease remained stable; these results were confirmed by PET/CT and MRI scans in February 2019 (Fig. 2c). The patient tolerated niraparib treatment very well except for an episode of grade 3 thrombocytopenia that resolved spontaneously.
Unfortunately, the control of intracranial disease was not durable. In April 2019, approximately 4 months after initiating niraparib treatment, the patient experienced deterioration in clinical condition (KPS decreased to 40). The patient became very weak and largely chair-bound. MRI scans showed mixed response with slight increase in the size of cerebellar cystic mass despite reduction in the size of cerebral tumor (Fig. 2d).
A change of systemic treatment regimen was considered and included the option of palliative chemotherapy. However, the patient preferred to continue niraparib as palliative treatment due to its tolerable side-effect profile. In April 2019, niraparib dose was increased to 200 mg/100 mg on alternating days. The patient remained in a stable condition until late July 2019, when she experienced further decline in neurological function due to progressive brain metastases (Fig. 2e). Extracranial disease remained stable. The patient was transferred to hospice facilities in August 2019.
This case report illustrates the application of NGS-based CGP for selection of PARPi treatment of BRCA1/2 wild-type TNBC. Niraparib monotherapy has been studied in only one report (QUADRA) of patients with ovarian cancer and HRD-positive tumors with or without BRCA mutations. The present case showed that this strategy may have wider applications in breast cancers with an “HRD phenotype,” as in this patient with mutations in HR pathway oncogenes other than BRCA1/2. Intriguingly, there is evidence that brain metastases from breast cancers have greater HRD relative to the corresponding primary tumors. This supports further investigation of PARPi with appreciable CNS activity, such as niraparib, in treating brain metastases from breast cancer.
TNBC is associated with less favorable prognosis than other breast cancer subtypes. Approved treatment options for TNBC are relatively limited compared with ER-positive and HER2-positive breast cancers, in which specific therapeutic targets are available. In patients without germline BRCA1/2 mutations, options are even more limited. Moreover, for the 10% to 24% of patients with breast cancer who develop brain metastases, disease prognosis will be less optimistic. The FUTURE study characterized molecular subtypes in refractory metastatic TNBC and examined the potential benefit of different precision treatments for these subtypes. Examples included PARPi-based treatment for the “basal-like immune-suppressed” subtype with germline-mutated BRCA1/2, and an ICI plus nab-paclitaxel combination for the “immunomodulatory” subtype. Interestingly, although no germline or somatic BRCA1/2 mutations were detected in our patient, her genomic alterations suggested an HRD phenotype and she derived considerable benefit from PARPi treatment for some months. It is possible that the alternative chemotherapy/ICI-based approach also could have benefited the patient if her condition had been suitable. Taken together, these observations support genomic profiling for guiding therapy selection in TNBC, but indicate that considerable work is still needed to understand the interplay between different types of genomic alterations in determining therapy response and outcomes in the metastatic setting.
Based on targetable genomic alterations identified by NGS-based CGP, several pathway-specific therapies were considered, including mammalian target of rapamycin (mTOR) inhibitors, CDK4/6 inhibitors, and PARPi. Niraparib was selected over olaparib because of its good CNS penetrance and drug accumulation in the brain, favorable tolerability, and high potency against HRD tumors. To date, three trials (NOVA; QUADRA; PRIMA) have published data on the use of niraparib in patients with HRD tumors. Importantly, niraparib showed antitumor activity regardless of BRCA status, in both mutant and wild-type BRCA tumors. In a phase 3 trial, niraparib improved median progression-free survival (12.9 months) in patients with wild-type BRCA and HRD recurrent ovarian cancer. In 2017, niraparib was approved by the US Food and Drug Administration (FDA) for maintenance therapy of patients with recurrent ovarian cancer who responded completely or partially to platinum treatment, regardless of BRCA or HRD status. PARPi, such as talazoparib and olaparib, have been FDA-approved for HER2-negative metastatic breast cancer and germline BRCA mutation, but no PARPi have been approved for TNBC. Despite the extensive high-volume brain metastases and extremely poor prognosis, the patient had a remarkable initial clinical response, accompanied by some lesion shrinkage over 2 to 3 months. Consistent with published preclinical[14–16] and clinical studies, it is likely that mutations in DDR genes and HR pathway genes were linked to the patient's response to niraparib treatment. Although the response was not durable, the patient's quality of life was maintained with niraparib as palliative therapy, even in advanced disease stages.
Thus far, no studies have reported the use of CGP-informed targeted therapies with targeted agents including mTOR inhibitor, CDK4/6 inhibitor, or PARPi in patients with TNBC. A recent study concluded that NGS-based CGP-guided therapy with PARPi, mTOR inhibitors, anti-programmed death-1 protein (anti-PD-1), and anti-HER2 in prostate cancer was a sensible approach, especially in advanced stages of the condition. Therefore, it may be worth exploring the feasibility of NGS-based CGP-guided therapy in patients with TNBC. Wider clinical applications of PARPi, such as possible combinational use with chemotherapy, radiotherapy, and immunotherapies also should be explored. Because reverse mutations of BRCA have been associated with acquired resistance to PARPi, and could explain the subsequent decline in intracranial disease control in our patient, future research should also delve deeper into the mechanism of resistance and plausible approaches for overcoming resistance.
In conclusion, we demonstrated the clinical utility of NGS-based CGP to guide treatment selection among several candidate pathway-specific therapies. Based on patient characteristics and clinical considerations, niraparib monotherapy was selected over other approaches. Niraparib monotherapy produced a rapid and remarkable response in our patient with brain metastases from BRCA1/2 wild-type TNBC. Given that brain metastasis from TNBC is associated with poor prognosis and limited effective treatment options, this case has highlighted the value of using CGP to guide the selection of pharmacotherapy in cancers with identifiable genomic alterations. NGS-based CGP has the potential to improve clinical management of TNBC and patient outcomes.
The author acknowledges The University of Hong Kong–Hong Kong Sanatorium and Hospital Molecular Tumor Board for the clinical decision support, and the scientific team of the ACT Genomics Laboratory for their assistance with genomic profiling and bioinformatics advice.
Manuscript drafting and editorial assistance was provided by Tech Observer Asia Pacific.
Source of Support: Manuscript development was supported by an educational grant from Zai Lab Hong Kong Limited to the author's institution. Conflict of Interest: None.