Receptor discordance in breast cancer brain metastases: when knowledge is power

See article by Hulsbergen et al. in this issue, pp. 1173–1181.

In the article by Hulsbergen et al, “Subtype Switching in Breast Cancer Brain Metastases: A Multicenter Analysis,” we gain new insights into receptor discordance between intracranial metastases, extracranial metastases, and primary breast tumors.¹ Collective findings from this important analysis provide clinically actionable information based on intracranial tumor receptor status, which offer new treatment directions for individual patients. While prior studies have illustrated subtype discordance between breast cancer brain metastases and their matched primary tumors, these studies were largely limited by smaller sample size and lacked information on matched extracranial metastases.^2–4 Moreover, clinical data accompanying the histopathologic data in the current analysis illustrate that knowledge of discordance, largely gain of targeted receptor expression in intracranial tumors, corresponds with improved patient survival and, in some cases, can be predicted based on baseline clinicopathologic characteristics.

In the current analysis, Hulsbergen et al conducted a multicenter study, to our knowledge the largest of its kind, to compare estrogen receptor (ER), progesterone receptor (PR), and human epidermal growth factor receptor 2 (HER2) status from brain and nonbrain metastases, and primary breast tumors from 219 patients, of which 193 had complete data on all 3 receptors, across 4 academic institutions. Nearly half of patients had extracranial metastases at the time of brain metastases diagnosis (58.9%). Key results of this analysis showed (i) 36.3% of breast cancer and brain metastases cases were discordant for ER, PR, or HER2, (ii) 22.8% of cases changed subtype as a result of receptor alteration, (iii) gain in HER2 expression in the brain metastases was more common than loss (8.0% vs 2.5%, respectively), and (iv) loss of ER or PR was more common in the brain metastases than gain (14.8% vs 1.9% and 22.4% vs 2.9%, respectively). When comparing receptor status between brain versus nonbrain metastases, the discordance rate was 63.6%; in contrast, receptor concordance between nonbrain metastases and primary breast tumor was 59.1%. Finally, loss of ER in the brain metastases predicted worse survival, while gain of HER2 showed a trend toward improved survival, and receipt of subtype-targeted therapy, particularly HER2-directed therapy, translated to improved patient outcome.

While knowledge of receptor status in brain metastases tissue is desirable, the current standard of care for patients with breast cancer brain metastases involves neurosurgical resection only in a minority of patients.⁵ As with any invasive surgery, the risks of neurosurgical resection include bleeding, infection, and risks associated with anesthesia, and are not to be taken lightly. Historically, neurosurgical resection has typically been reserved for patients with a good performance status who present with solitary, large metastases (≥3 cm), resulting in significant symptom burden. Another scenario supporting upfront neurosurgical resection is in patients who present with intracranial metastases as the only site of disease, thus are in need of a histopathologic diagnosis. More commonly, stereotactic radiosurgery remains the preferred local therapy treatment modality in the majority of cases, reserving whole brain radiotherapy for multiple, diffuse intracranial metastases.⁵

Data provided by Hulsbergen et al, however, challenge the current paradigm of “operating in the dark” with regard to receptor status in breast cancer brain metastases. As brain metastases subtype differs from extracranial sites of disease in 1 and 5 cases, with accompanying changes in targeted systemic therapy, less invasive ways to obtain this knowledge are urgently needed. Options might include analysis of circulating tumor DNA from cerebral spinal fluid (CSF) obtained via lumbar puncture, which has been shown to reflect intracranial tumor biology more accurately than that of primary brain tumors.^6,7 Another consideration would be a less invasive brain biopsy that has been increasingly employed as part of laser interstitial thermal therapy for treatment of radiation therapy necrosis.⁸ Continued research to identify procedures less invasive than traditional craniotomy will be critical to obtaining precise molecular information from intracranial tumors in a manner that is safest and most accurate for our patients.

With brain metastases receptor status in hand, how would this change our therapeutic recommendations in the modern era? As recently as several years ago, receptor and/or molecular information obtained from brain metastases tissue biopsies may not have been “actionable.” We are now in an era, however, where receptor status and molecular information can inform treatment decisions in tangible ways. One such approach is the Alliance for Clinical Trials in Oncology “Genomically-Guided Treatment Trial in Brain Metastases” (NCT03994796), which assigns patients to one of several brain-permeable, targeted agents: an inhibitor of cyclin-dependent kinase 4/6 (abemaciclib), a dual inhibitor of mammalian target of rapamycin/phosphatidylinositol 3-kinase (GDC-0084), or an inhibitor of neurotrophic tyrosine receptor kinase (entrectinib; lung only). Brastianos et al have shown that knowledge of brain metastases molecular characteristics provides targetable information, unique to the intracranial tumor, in over half of cases.⁹ In this protocol, treatment assignment is based on the molecular characteristics of the brain metastasis itself, thus is limited to patients who have undergone a resection or biopsy of intracranial tumor tissue as per standard of care. A second example is in advanced, HER2-positive breast cancer and brain metastases as per the HER2Climb study.¹⁰ Combination therapy of tucatinib, a brain-permeable small-molecule inhibitor of HER2, with capecitabine and trastuzumab was recently FDA approved for patients with HER2-positive brain metastases following one prior HER2-directed therapy. Receptor discordance in brain metastases tissue involves gain of HER2 in brain metastases in 1 of 10 patients. In the absence of this knowledge, eligible patients would be missing an opportunity for an effective, brain-permeable, targeted therapy with associated survival advantages.

In summary, the data provided by Hulsbergen et al continue to build upon our growing knowledge of breast cancer brain metastases biology and the changes that are occurring in the central nervous system space. While information from brain metastases tissue is valuable and can lead to novel treatment opportunities, this does not mean all patients with brain metastases arising from breast cancer should undergo an invasive procedure given the risks associated with intracranial biopsy and/or resection. This does point toward the need for validated, non-invasive methods (eg, through circulating CSF or blood, novel imaging techniques) in the short term to help us understand the molecular characteristics of brain metastases to provide our patients with optimal care while minimizing risks.

The text is the sole product of the author(s) and no third party had input or gave support to its writing.

Conflict of interest statement. CKA: Research funding: PUMA, Lilly, Merck, Seattle Genetics, Nektar, Tesaro, G1-Therapeutics; compensated consultant role: Genentech, Eisai, IPSEN, Seattle Genetics; AstraZeneca; Royalties: UpToDate, Jones and Bartlett.

SS: Research funding: Astra Zeneca, Eli Lilly; compensated consultant role: Novartis, Daiichi Sankyo, Foundation Medicine, Sermonix.