A new study by a brain tumor research team at CHOP supports the development of targeted treatment paradigms for BRAF-altered pediatric astrocytomas. The findings, published in April in the Proceedings of the National Academy of Sciences, also demonstrate that therapies must be tailored to the mutational context and distinct mechanisms of action of the mutant BRAF kinase.
“One of the biggest questions in science right now is how we will use the flood of gene-sequencing data to customize treatments,” says first author Angela J. Sievert, MD, MPH, a pediatric neuro-oncologist in the Cancer Center at CHOP.
“This study provides insight into the value of extremely precise characterizations of mutations in a child’s tumor in the design of new treatments.”
Along with co-first author Shih- Shan Lang, MD, Sievert works in the translational laboratory of neurosurgeon Phillip Storm, MD, and Adam Resnick, PhD. Sievert and colleagues were among several groups that first identified that mutations in the BRAF gene were highly prevalent in astrocytomas, the most common type of brain tumor in children.
Many astrocytomas are too widespread or in too delicate a site to be safely removed. A drug that can kill the tumor with low toxicity to healthy tissue is the ideal treatment. Because recently developed BRAF-targeted therapies, such as vemurafenib, have shown promise in treating melanomas in adults, there was hope that they would be effective against pediatric astrocytomas. However, follow-up studies using vemurafenib by the CHOP team were disappointing: BRAF inhibitors that were effective in adult melanomas paradoxically made brain tumors worse via drug-induced activation.
Further investigation by the team revealed how tumor behavior depended on which type of BRAF mutation was involved. First-generation anti-BRAF drugs that were effective in melanoma target V600E point mutations in BRAF. However, in most astrocytomas the mutation in the BRAF gene was different; it produced a fusion gene, KIAA1549-BRAF. When used against the fusion gene, first-generation drugs activated the MAPK signaling cascade and accelerated tumor growth.
The CHOP team characterized the distinct mechanisms of action of KIAA1549-BRAF and its differential responsiveness to first-generation BRAF inhibitors. They found that in cells expressing KIAA1549-BRAF, the fusion kinase aberrantly functions as a homodimer that is resistant to these inhibitors and is associated with a novel CRAF-independent paradoxical activation of MAPK signaling.
Mutagenesis studies demonstrated that BRAF fusion-mediated signaling is dependent on constitutive dimerization and was diminished with disruption of the BRAF kinase dimer interface. In addition, the KIAA1549-BRAF fusion displays distinct additional protein interactions that likely facilitate MEK phosphorylation by BRAF.
Based on these studies and working with the pharmaceutical industry, the investigators identified a newly developed second-generation selective BRAF inhibitor that, unlike vemurafenib, did not induce activation of BRAF in cell cultures and animals. In contrast to vemurafenib, these second generation inhibitors successfully target BRAF-fusions. Recent efforts are now characterizing MEK inhibition and combinatorial drug targeting as additional approaches to target gliomas. This preclinical work has laid a foundation for multicenter clinical trials to test the mutation-specific targeting of tumors in children with astrocytomas, Sievert says. As this effort progresses, it will benefit from CHOP’s commitment to resources and collaborations that support data-intense research efforts.
The direction of brain tumor research over the past several years reflects some of those data-driven advances, says Resnick, the senior author of the paper and principal investigator of the astrocytoma research team in the Division of Neurosurgery at CHOP. Harnessing the power of open access models for data and tissue sharing such as the Childhood Brain Tumor Tissue Consortium has allowed for novel genetic characterization of tumors. “For years, astrocytomas have been lumped together based on similar appearance to pathologists studying their structure, cell shape and other factors,” he says. “Our current discoveries show that the genetic and molecular structure of tumors provides more specific information in guiding oncologists toward customized treatments.”