FormalPara Key Points

Patients with certain brain tumors require screening for the BRAF V600E mutation.

BRAF V600E-mutant tumors need to be considered in the context with other genetic alterations (e.g., coexisting gain-of-function mutation of EGFR or KIAA1549-BRAF fusion).

Dual inhibition of both RAF and MEK is expected to provide more potent and durable effects than anti-BRAF monotherapy.

BRAF-targeted therapy in brain tumors should be limited to the setting of a clinical trial.

1 Introduction

Primary brain tumors remain the leading cause of mortality from malignant neoplasms in children and young adults. Glioblastoma (GBM), the most common brain tumor, is characterized by a median survival of < 21 months, despite surgical resection, radiation therapy, high-dose chemotherapy, and alternative approaches such as Tumor Treating Fields (TTFields) [1,2,3,4]. Within the central nervous system (CNS), immune cells follow different principles. The blood–brain barrier (BBB) not only restricts the movement of soluble mediators and leukocytes from the periphery [5], but also prevents the brain uptake of most neurotherapeutics [6]. Finally, brain neoplasms are exceptionally heterogeneous, further hindering the development of successful treatment modalities [7].

The mitogen-activated protein kinase (MAPK) is an essential signaling pathway in a number of malignancies. Alterations in various components of the MAPK pathway, especially the BRAF gene, have been thoroughly described in melanoma, colorectal cancer, thyroid cancer, non-small-cell lung cancer (NSCLC), and hairy cell leukemia [8,9,10,11]. BRAF encodes the B-Raf kinase that activates MAPK signaling through phosphorylation of MAPK kinase (MEK) and subsequently MAPK. Activating mutations of BRAF lead to constitutive downstream activation of RAF-MEK-MAPK signaling cascade, promoting cell proliferation and survival while inhibiting apoptosis, and eventually driving tumor growth [12, 13].

Currently available combinations of RAF and MEK inhibitors approved by the FDA include vemurafenib/cobimetinib, dabrafenib/trametinib, and encorafenib/binimetinib in melanoma and dabrafenib/trametinib in NSCLC [14,15,16,17]. Given the remarkable responses seen in these patients, BRAF-targeted approaches have attracted significant attention in the field of neuro-oncology.

The possible application of BRAF-targeted therapy in CNS tumors grows continuously. While clinical trials are still ongoing, there is preliminary evidence for a range of positive responses in certain brain tumor types harboring BRAF V600E mutation. Herein, we propose a management algorithm for brain tumor patients who could benefit from BRAF-targeted therapy.

2 Relevance of V600E Among BRAF Mutations

In vivo xenograft studies confirm the previously described role of BRAF in MAPK signaling regulation within CNS tumors [18, 19]. To date, over 30 BRAF alterations have been associated with human cancers. They have been grouped into three classes according to their kinase activity, rat sarcoma protein (RAS) dependency, and dimerization status. Although they all lead to MAPK activation, only class I mutations are sensitive to currently available BRAF inhibitors.

Class I mutations are independent of both upstream RAS activation and the need for dimerization. This class is represented by four V600 subtypes (V600E, V600D, V600K, and V600R). V600E is a single nucleotide mutation at codon 600 resulting in substitution of glutamic acid (E) for valine (V). The glutamate residue interacts with glycine-rich loop that ordinarily suppresses the activity of BRAF. The loss of this inhibitory effect results in an increase in BRAF basal activity and contributes to oncogenesis [8, 20].

Compared with V600E, the remaining class I mutations occur far less frequently and, therefore, their clinical relevance is harder to assess [21]. V600E is present in a significant subset of CNS tumors (Table 1). Its highest incidence is observed in papillary craniopharyngioma and pleomorphic xanthoastrocytoma (PXA), in 95% and up to 78% of cases, respectively [26, 50]. It also frequently occurs in pilocytic astrocytoma (PA), ganglioglioma (GG), and pediatric low-grade astrocytoma. While this mutation is rare in GBM in general, it is relatively common among young adults and when diagnosed as an epithelioid type [34].

Table 1 Published reports of brain tumor patients treated with BRAF-targeted therapy
Full size table

Class II mutations include several point mutations and fusions that activate MEK through RAS-independent dimerization. The most common point mutations include K601E/N/T, L597Q/V, and G469A/V/R; however, their relative frequency in brain tumors remains unknown [21]. Among fusions, the most common is KIAA1549-BRAF, which is predominantly identified in low-grade gliomas.

Class III mutations are kinase impaired and enhance MAPK signaling through RAS and subsequent CRAF activation. This class is represented by G466E, D594G, and G596D; however, their incidence in brain tumors remains unclear [21, 54].

3 Strategies for the Management of BRAF-Mutant Brain Tumors

Table 1 summarizes the outcomes gathered from experiences with BRAF-targeted therapy in primary brain tumors; a PubMed database search for valid records was conducted until April 3, 2020; keywords used were BRAF mutation, V600, V600E, CNS tumors, brain tumors, MAPK, ERK, MEK1/2, RAF, glioma, glioblastoma, ependymoma, medulloblastoma, oligodendroglioma, xanthoastrocytoma, ganglioglioma, craniopharyngioma, spindle cell oncocytoma, and astrocytoma. The search results were narrowed by selecting studies in humans published in English. Twenty-six studies mentioning BRAF-targeted therapy in primary brain tumors were included [23,24,25, 27,28,29,30, 32, 33, 35,36,37,38,39, 41,42,43,44,45,46,47,48,49, 51,52,53]. The PubMed database was chosen as it remains the most widely used resource of medical journals and indexes only peer-reviewed literature [55]. Experience with BRAF-targeting therapies in brain tumors is generally limited to case reports and series that are likely biased toward unusually positive responses. In the majority of cases, the targeted therapy was initiated after progression on standard therapy. The assessment of its efficacy is further complicated by concurrent treatments including TTF, chloroquine, or vinblastine. Also, there is mixed use of second-generation RAF inhibitors (vemurafenib, dabrafenib) and MEK inhibitors (trametinib, cobimetinib). In cerebral metastatic melanoma, the response rates are better for dabrafenib (31%) compared with vemurafenib (16%) [56,57,58,59]. This difference is probably due to smaller size and molecular structure of dabrafenib that enables better penetration of the BBB [60]. In primary brain tumors, however, there are no clinical trials that support any combination of the aforementioned agents. Authors generally agree that first-generation RAF inhibitors, such as sorafenib, have poor blood–brain penetrance and efficient response is only to be expected in tumors with significant BBB breakdown [61]. In fact, sorafenib was tested in pediatric patients with low-grade gliomas and contributed to unexpected and unprecedented acceleration of tumor growth, irrespective of the BRAF status [62].

Finally, tumor heterogeneity often affects responses in ways not yet fully understood. Whether BRAF mutations are present in all or in only a subset of tumor cells seems fundamental for the selection of appropriate treatment.

3.1 BRAF Inhibition May Be Insufficient

While MAPK signaling is essential in glioma formation, BRAF activation alone is not sufficient for the development of high-grade tumors [63]. In fact, following the positive response to V600E-targeted monotherapy, patients often succumb to disease progression [25, 29, 30, 33]. The resistance mechanisms have been described in detail in other malignancies, especially in melanoma [64]. Their common denominator is the maintenance of MAPK upregulation through alternative activators, either upstream or downstream of mutant BRAF (Fig. 1) [65,66,67].

Fig. 1
figure 1

Mechanisms of resistance to BRAF-targeted therapy. V600E and KIAA1549-BRAF fusion lead to constitutive activation of MAPK signaling in tumor cells. MAPK phosphorylates downstream nuclear effectors that ultimately enhance cell survival and proliferation. One recognized mechanism of resistance to BRAF inhibition is the upregulation of CRAF, which activates MAPK through MEK. Concurrent inhibition of MEK overcomes this resistance. Both RAF and MEK inhibition, as well as EGFR gain-of-function mutation contribute to amplification of the PI3K/AKT/mTOR pathway that represents another mechanism of resistance to BRAF-targeted therapy. EGF epidermal growth factor, EGFR epidermal growth factor receptor, MAPK mitogen-activated protein kinase, MEK MAPK kinase, mTOR mammalian target of rapamycin, PI3K phosphoinositide 3-kinase, RAS rat sarcoma protein

Full size image

MEK inhibitors effectively prevent reactivation of MAPK signaling in various BRAF-mutated malignancies [14,15,16,17]. Hence, dual BRAF and MEK inhibition is associated with both prolonged clinical endpoints and improved side effect profiles compared with the BRAF inhibition alone. To date, multiple combinations of BRAF and MEK inhibitors have been approved by the FDA for the treatment of patients with melanoma and NSCLC [14,15,16,17]. In primary brain tumors, adequate brain distribution of both compounds is essential. While BRAF inhibitors show some level of BBB penetrance in melanoma patients with intracranial metastasis, brain distribution of MEK inhibitors may be significantly limited. A possible reason for that is the P-glycoprotein (P-gp) that was reported to efficiently restrict brain distribution of trametinib [68]. Therefore, inhibiting the efflux transporters represents an attractive approach to further increasing the therapeutic efficacy of BRAF-targeted therapy.

In V600E-mutated glioma cell lines, epidermal growth factor receptor (EGFR) amplification causes resistance to BRAF inhibition and blocking EGFR overcomes this resistance [69]. Presumably, this resistance emanates from adaptive feedback reactivation of MAPK signaling, in which EGFR activates other kinases, such as CRAF, which are resistant to BRAF inhibitors. One case report supports this observation, where a patient with GBM and concurrent gain of function EGFR mutation experienced no clinical benefit, despite RAF and MEK inhibition [38].

Also, the KIAA1549-BRAF fusion may hinder the effect of targeted treatment [30]. Although this fusion occurs in a mutually exclusive pattern with other activating mutations in the MAPK signaling pathway [70, 71], a few cases were reported with both the fusion and V600E mutations [30, 72]. In such patients, KIAA1549-BRAF fusion is believed to function as a homodimer that not only renders cells resistant to BRAF inhibition, but also displays paradoxical activation of MAPK signaling [73]. It might be an explanation for rapid progression of low-grade astrocytoma in pediatric patients with both BRAF alterations after treatment with selumetinib [30].

Mutations in receptor tyrosine kinase (RTK) growth factor receptors resulting in constitutive downstream signaling of the canonical MAPK and PI3K/Akt pathways are found in almost all astrocytic tumors [63]. While MEK/MAPK signaling represents one arm of the tumor growth and survival network in gliomas, the MAPK and PI3K pathways regulate a common set of downstream apoptotic regulators, such as Bad, Mcl-1, Bcl-xL and Bcl2. In such cases, the combination of a BH3 mimetic with MEK/RAF and PI3K inhibitors may prove necessary.

3.2 BRAF and the Oncogene-Induced Senescence (OIS)

BRAF activation in neural stem and progenitor cells not only promotes tumor growth, but also mediates oncogene-induced senescence in low-grade brain tumors [74]. This explains the relatively high frequency of BRAF-mutant brain tumors associated with favorable prognosis [71]. Up to 78% of PXAs WHO grade II and only 12% of anaplastic astrocytoma WHO grade III harbor the V600E mutation [26, 40]. This phenomenon suggests that BRAF activation entails the oncogene-induced senescence (OIS) in PXA [74]. The major role of OIS is to abolish the increased mitogenic signaling and protect from tumor progression [74]. The potential protective role of BRAF in brain tumors might put under question its suitability as a molecular target. Nonetheless, recent results from an open-label, nonrandomized, multicohort study for BRAFV600-mutant nonmelanoma cancers supports the detrimental effect of BRAF activation in PXA. From 24 patients treated with vemurafenib, seven with grade II PXA achieved the highest response rate [25]. Another three case reports support the suitability of BRAF inhibitors against grade III PXA (Table 1) [38, 39, 41]. On the other hand, OIS may potentially promote carcinogenesis through alternative combinations of downstream effectors [75], and whether it occurs in brain tumors remains to be seen.

3.3 Current Recommendations Regarding BRAF Mutations in CNS Tumors

Only a few molecular markers to determine prognosis and guide treatment have been integrated into the clinical care of brain tumor patients. This applies especially for MGMT promoter methylation and TERT promoter mutation in IDH-wild type gliomas. To date, however, the validity of this testing is not supported by strong evidence and no large consensus exists on the optimal method of assessment of these alterations [76, 77]. National Comprehensive Cancer Network (NCCN) Guidelines for central nervous system cancers are the only ones to mention gain of function BRAF mutation in brain tumors. The guidelines recommend treatment with BRAF and MEK inhibitors in V600E-mutated PA, PXA and GG in the adjuvant setting [78].

3.4 Screening for V600E in CNS Tumors

Anti-BRAF V600E (clone VE1), a mouse monoclonal antibody, is used in the immunohistochemical identification of mutant BRAF V600E protein and recently has been successfully validated in CNS tumors [79,80,81]. Because of its unsuitability for detecting V600E in particular neoplasms, such as pituitary adenomas [82], it is generally recommended to perform sequencing validation in positive cases prior to application for targeted treatment [34].

Behling et al. suggest routine screening for V600E for glioblastomas WHO grade IV below the age of 30 years, especially those with epithelioid features, PXAs, and GGs [34]. We propose additional testing in papillary craniopharyngiomas, pediatric low-grade astrocytomas, and PAs (Fig. 2) due to significant subsets of V600E positive cases within these groups (Table 1).

Fig. 2
figure 2

Management algorithm for brain tumor patients—the candidates for BRAF-targeted therapy. The initial step in primary brain tumor treatment is for the neurosurgeon to remove as much of the tumor as possible while minimally disturbing the surrounding brain tissue. Patients with GBM WHO grade IV below the age of 30 years (especially those with epithelioid features), PLGA, PA, PXA, GG, and PCP should be screened for BRAF V600E mutation using the anti-BRAF V600E. The positive IHC result needs to be confirmed by sequencing. BRAF-targeted therapy should be limited to the setting of a clinical trial. If the patient harboring a V600E mutation does not qualify for a trial, multidirectional treatment is recommended. Dual inhibition of both RAF and MEK is expected to provide more potent and durable effect than anti-BRAF monotherapy. First-generation RAF inhibitors, such as sorafenib, are not recommended. The oncologist that is directly involved in treatment of the patient should decide whether and when to test for gain-of-function mutation of EGFR and KIAA fusion, since both of these alterations may compromise BRAF-targeted therapy. BBB blood–brain barrier, eGBM epithelioid glioblastoma, GBM glioblastoma, GG ganglioglioma, IHC immunohistochemical, PA pilocytic astrocytoma, PCP papillary craniopharyngioma, PLGA pediatric low grade astrocytoma, PXA pleomorphic xanthoastrocytoma

Full size image

3.5 Selection of Patients with V600E for Targeted Therapy

There is no strong evidence of benefit from BRAF-targeted therapy as a first-line treatment in CNS tumors. Regardless of tumor histology, maximal safe resection is recommended whenever possible. BRAF-targeted therapy may improve survival in some patients, however its initiation requires a multidisciplinary consultation.

Schreck et al. suggest limiting the BRAF-targeted therapy to the setting of a clinical trial and considering combination therapy only in patients who are not eligible for a trial [21]. The NCCN Guidelines recommend single-modality therapies only for patients with poor performance status (Karnofsky Performance Score [KPS] < 60), since they frequently do not tolerate the toxicity associated with combination regimens [78].

Since the therapy must be tailored to the specific mutational context and distinct mechanisms of action of the mutant kinase, we recommend considering additional testing for the gain of function EGFR mutation and KIAA1549-BRAF fusion beforehand (Fig. 2). KIAA1549-BRAF fusion can be detected with RNA-Seq. FISH is a less reliable method of identifying BRAF fusions because the KIAA1549 and BRAF genes lie in close proximity and it is difficult to distinguish fusion signal from a normal signal [83]. If, however, the RNA extraction was unsuccessful, FISH may represent a useful adjunct.

Ongoing trials will further clarify targeted treatment options in the presence of BRAF-mutant brain tumors. Three phase II clinical trials are evaluating the biological activity of trametinib (NCT03363217), MEK 162 (NCT02285439), and a combination of dabrafenib with trametinib (NCT02684058) in pediatric gliomas. In adults, two phase II clinical trials using binimetinib with encorafenib in HGG (NCT03973918) and vemurafenib with cobimetinib in craniopharyngioma (NCT03224767) are currently underway.

4 Conclusions

BRAF mutations that result in MAPK pathway activation are common events in several types of brain tumors. The efficacy of targeted therapy for BRAF-mutant brain tumors is currently being investigated. Given the encouraging early results, in the not-too-distant future, such approaches may shift the treatment paradigms and increase survival of patients with CNS tumors. The proposed algorithm was designed to help oncologists to provide the best therapeutic options for brain tumor patients.