Encouraging response rates to a dual-targeted regimen in patients with recurrent or refractory malignant brain tumors with BRAF V600E mutations were reported in 2021, representing a new potential targeted therapy strategy against glioblastoma.
Patrick Y. Wen, MD
For more than 20 years, investigators have tried numerous targeted therapy strategies against glioblastoma (GBM) that have fallen frustratingly short in establishing clinical benefit.1 That changed in November 2021 when an international team of investigators reported encouraging response rates to a dual-targeted regimen in patients with recurrent or refractory malignant brain tumors with BRAF V600E mutations.2
The combination of dabrafenib (Tafinlar), a BRAF inhibitor, and trametinib (Mekinist), a MEK inhibitor, demonstrated an objective response rate (ORR) of 33% (95% CI, 20%-49%) in patients with BRAF V600E–mutant high-grade gliomas, many of whom had GBM, according to interim data from the phase 2 Rare Oncology Agnostic Research (ROAR) trial (NCT02034110) published in Lancet Oncology.2
The findings represent the first time a targeted therapy has shown efficacy against GBM in a clinical trial, according to lead study author Patrick Y. Wen, MD, director of the Center for Neuro-Oncology at Dana-Farber Cancer Institute and a professor of neurology at Harvard Medical School, both in Boston, Massachusetts. Wen said the activity observed in the study is an encouraging sign that other targeted therapies may be effective against GBM, an aggressive malignancy that is the most prevalent primary brain cancer in adults.3-5
The prospect of a targeted therapy regimen for a subset of patients with malignant brain tumors comes at time when the molecular underpinnings of these diverse and challenging cancers are increasingly recognized. In 2021, the World Health Organization (WHO) published its sixth version of international standards for classifying tumors of the brain and spinal cord. Although histology and immunohistochemistry analyses remain central to tumor diagnosis, multiple molecular parameters have been incorporated into the WHO guidelines for more precise identification. A prominent example is the classification of adult diffuse glioma subtypes including GBM by the presence or absence of an IDH1/2 mutation.6
Molecular testing also is gaining ground in clinical decision-making for a spectrum of brain cancers. National Comprehensive Cancer Network (NCCN) guidelines recommend integrated histopathological and molecular profiling for all gliomas, with data for a menu of molecular alterations considered noteworthy in a pathology workup (FIGURE7).
Although investigators have explored many novel options, the frontline standard of care for GBM treatment remains surgical resection followed by radiotherapy and temozolomide. For recurrent disease, NCCN guidelines include protocols involving cytotoxic drugs, bevacizumab (Avastin), and regorafenib (Stivarga).7
The FDA has approved bevacizumab, a VEGF inhibitor, for recurrent glioblastoma based on progression-free survival (PFS) and response data.1,8 Regorafenib, which inhibits VEGF receptors and other kinases, demonstrated a benefit in median overall survival (OS) compared with lomustine (7.4 months vs 5.6 months; HR, 0.50; 95% CI, 0.33-0.75; P = .0009) in the phase 2 REGOMA trial (NCT02926222). However, the response rate with lomustine was lower than in other clinical trials, prompting some experts to question whether the study provides sufficient evidence of its efficacy against GBM.7,9
Overall, drug development efforts to expand FDA-approved treatment options with targeted agents have not succeeded despite a plethora of clinical trials. During a 20-year period ending in April 2020, there were 257 phase 1/2 through phase 3 studies that tested targeted regimens, according to an in-depth analysis by French investigators. Most of these studies (70%) were phase 2 trials. Only 12 of 37 studies with comparative data showed an improvement with a novel therapy, mostly in PFS.1
Cruz Da Silva et al cited 3 key factors that contribute to targeted therapy failures against GBM: (1) the impossibility of performing a full surgical resection, (2) difficulty designing drugs that cross the blood-brain barrier, and (3) intratumoral molecular heterogeneity. Additionally, GBM is a comparatively rare cancer and clinical trial enrollment remains too low.1
Nevertheless, investigators continue to pursue targeted agents against GBM and other gliomas, with multikinase inhibitors and multitargeted regimens among the key trends.1 Ongoing clinical trials include studies evaluating various pathway inhibitors as well as targeted vaccines and combinations with immunotherapy. Several research teams also are using molecularly selected populations in basket studies to test multiple targeted agents (TABLE 1).
In ROAR, US-based investigators joined with researchers from community and academic centers in 12 countries to study the combination of dabrafenib (150 mg twice daily) plus trametinib (2 mg daily) in 9 cohorts of patients with rare advanced BRAF V600E–mutated cancers.
There were 2 cohorts for patients with gliomas (low-grade and high-grade tumors) as well as arms for anaplastic thyroid cancer, biliary tract cancer, gastrointestinal stromal tumor, hairy cell leukemia, multiple myeloma, and adenocarcinoma of the small intestine. Investigators also planned a cohort for patients with germ cell tumors but no one enrolled. In all, 206 patients 18 years and older enrolled from April 2014 through July 2018.2
The rationale for testing the targeted therapy combination against BRAF V600E–mutated gliomas stemmed from prior clinical data showing antitumor activity with single-agent BRAF inhibition in pediatric and adult patients with that tumor type. Investigators noted that dual blockade of MAPK pathway signaling with BRAF and MEK inhibitors has become a standard of care in BRAF V600-mutant melanoma, non–small cell lung cancer, and anaplastic thyroid cancer (TABLE 210).2
Understanding of the role of BRAF alterations also is growing across several brain cancer subtypes. BRAF V600 mutations that activate the MAPK pathway have been detected in 5% to 15% of low-grade gliomas but are less common in highgrade gliomas, including GBM, where they have been identified in approximately 3% of tumors.2
In the ROAR trial, investigators recruited 45 patients with recurrent high-grade glioma classified as WHO grades 3 or 4 by 2007 criteria and 13 participants with low-grade glioma of WHO grade 1 or 2. The primary end point was investigator-assessed ORR using Response Assessment in Neuro-Oncology (RANO) criteria. BRAF V600E mutation status was confirmed on tumor tissue, with a range of testing methods permitted.
In the high-grade cohort, the average age of participants was 42 years (range, 18-72). Most patients had undergone resection or craniotomy (93%) and were previously treated with chemotherapy (93%) and/or radiation therapy (RT; 98%) regimens. The most prevalent histologies were GBM in 31 patients (69%), followed by anaplastic pleomorphic xanthoastrocytoma and anaplastic astrocytoma, with 5 patients each (11% each).2
In the low-grade cohort, the average age was 33 years (range, 18-58). All patients had surgery and many had prior RT (62%) and/or chemotherapy (38%). The most common histologies were ganglioglioma in 4 patients (31%), followed by diffuse astrocytoma and pleomorphic xanthoastrocytoma, with 2 patients each (15% each).
After a median follow-up of 12.7 months, 15 patients in the high-grade cohort exhibited a response, including 3 with complete responses (CRs) and 12 with partial responses (PRs), for an ORR of 33% (95% CI, 20%-49%). Among the 31 patients with GBM, 10 patients responded, including 2 with CRs and 8 with PRs, for an ORR of 32% (95% CI, 17%-51%).2
Among patients with low-grade gliomas, there were 9 responses, including 1 CR, 6 PRs, and 2 minor responses, for an ORR of 69% (95% CI, 39%-91%). The median follow-up was 32.2 months.
The combination also demonstrated clinical benefit in the secondary end points of PFS and OS. In the high-grade glioma cohort, median PFS was 3.8 months (95% CI, 1.8-9.2) and median OS was 17.6 months (95% CI, 9.5-45.2). In the GBM subset, median PFS was 2.8 months (95% CI, 28.213.7) and median OS was 13.7 months (95% CI, 8.4-25.6). In the low-grade group, the median was not reached (NR) for PFS (95% CI, 7.4-NR) and OS (95% CI, 11.6-NR).2
Moreover, the safety profile for the combination was consistent with clinical experience in other indications, investigators noted. Adverse events of grade 3 or worse severity were observed in 53% of patients, with fatigue (9%), decreased neutrophil count (9%), and headache and neutropenia (5% each) as the most frequently reported.
Outcomes in ROAR compare favorably with those for patients with recurrent high-grade gliomas in molecularly unselected populations treated with chemotherapy, investigators noted. “Response rates with chemotherapy have rarely exceeded 5%, with overall survival of 5 to 9 months and 3 to 4 months for glioblastoma after failing temozolomide and bevacizumab treatment,” they wrote. “In general, the progression-free survival for various salvage therapies in high-grade glioma was less than 3 months.”
Patients with low-grade gliomas who progress after chemotherapy also need new therapeutic options, investigators said, noting the added risk of transformation into a high-grade disease over time.
Moving forward, investigators suggested that the combination regimen be tested in a randomized clinical trial in treatment-naïve and previously treated patients with BRAF V600E–mutated high-grade or low-grade gliomas. Additionally, they recommended that BRAF V600E mutation testing be incorporated into molecular screening strategies.2
Although mutations in several genes have been implicated in primary central nervous system tumors, mutations in IDH1/2 genes, which cause epigenetic modifications, have emerged as an important marker for adult-type diffuse gliomas.6,7 In the WHO standards, the presence of an IDH mutation frequently distinguishes lowergrade gliomas (WHO grades 2 and 3) whereas the absence of the aberration helps identify GBM.6
Somatic mutations in IDH1, most frequently at codon R132 (IDH1-R132H), have been detected in more than 70% of astrocytomas, oligodendrogliomas, and GBM that developed from the lower-grade lesions. IDH2 mutations were detected in 3.5% of these tumors. In adult primary GBM samples, 6% had IDH1 mutations and none had IDH2 mutations.11
Of note, the presence of IDH1/2 mutations has been strongly associated with more favorable outcomes, including OS, in grade 2 and grade 3 gliomas.7
Despite the significance of IDH1/2 mutations, however, there are no FDA-approved IDH inhibitors for brain cancer. Ivosidenib (Tibsovo), an IDH1 inhibitor, is indicated for IDH1-mutant acute myeloid leukemia (AML) in newly diagnosed patients who are 75 years or older or have comorbidities that preclude intensive chemotherapy and for patients with relapsed AML, as well as for patients with previously treated IDH1-mutant locally advanced or metastatic cholangiocarcinoma.12 Enasidenib (Idhifa), an IDH2 inhibitor, is approved for treating adults with IDH2-mutant relapsed or refractory AML.13
Both drugs have shown low brain penetration in preclinical studies, and investigators have turned their attention to novel agents for direct inhibition of IDH in gliomas.14 Among the most advanced novel agents in clinical development is vorasidenib (AG-881), a first-in-class dual inhibitor of IDH1 and IDH2 that was designed for improved penetration across the blood-brain barrier.15
In a first-in-human study (NCT02481154), vorasidenib monotherapy elicited responses in patients with IDH1/2–mutant lower-grade gliomas (WHO grade 2 or 3). The cohort included 22 patients with nonenhancing glioma, defined as the absence of enhancement on MRI, and 30 participants with enhancing glioma. Forty-one patients with a variety of other solid tumors also were enrolled but recruitment was halted to focus development on glioma.15
Among patients with nonenhancing gliomas, there was 1 PR and 3 minor responses by RANO criteria for an ORR of 18.2% (95% CI, 5.2-40.3). An additional 16 patients (72.7%) had stable disease. None of the participants with enhancing glioma showed a response, although 17 patients (56.7%) had stable disease. With 59% of events reported, the median PFS was 36.8 months (95% CI, 11.240.8) for patients with nonenhancing glioma and 3.6 months (95% CI, 1.8-6.5) for participants with enhancing glioma.15
Vorasidenib is being compared with placebo in the randomized phase 3 INDIGO trial (NCT04164901) in patients with WHO grade 2 oligodendroglioma or astrocytoma with an IDH1 or IDH2 mutation that has recurred after surgery. The study, which is open to patients 12 years and older, has an estimated primary completion date of October 2024.