Targeted Therapy for Brain Tumors: New Approaches to an Old Problem
The modern era of targeted therapy for cancer began in the 1990s with the development of imatinib, a selective inhibitor of the BCR/ABL fusion protein tyrosine kinase produced by tumor cells in 95 percent of patients with chronic myelogenous leukemia (CML). The path from initial reports of imatinib’s in vitro growth inhibiting activity in 1996, to FDA approval of the drug in 2001, was unusually rapid, but the reason was abundantly clear— this drug was a home run. When the results of the initial phase III CML imatinib trial were reported at the 2002 ASCO meeting, imatinib was superior to standard therapy (interferon- Ara C) on every clinical endpoint: 91 percent complete hematologic response compared to 49 percent with interferon-Ara C, complete cytogenetic response rates were increased by almost tenfold (68 percent versus 7 percent), longer progression free survival, and lower toxicity. The discussant of the trial, a specialist in CML standard therapy, declared imatinib was “clearly now the therapy of choice for newly diagnosed CML,” and many in attendance had the sense that a new era in cancer treatment had abruptly become a reality.1 A few years later, the five-year survival for CML in the United States had doubled, from 31 percent in 1993, to 63 percent in 2005-2011.2
That was 15 years ago, and despite significant progress in many other cancer types with targeted agents, the targeted therapy era has brought little actual change to brain tumor treatment. Targeted therapies now play a major role in initial treatment of many common cancers—traztuzumab (Herceptin) for HER-2 positive breast cancer, vismodegib for basal cell skin cancer, sunitinib and everolimus for renal cell cancer, dabrafenib and trametinib for BRAF V600E-positive melanoma, erlotinib for EGFR-mutant lung adenocarcinoma, and many more. In contrast, early reports of responses to EGFR kinase inhibitors in glioblastoma, particularly in patients with expression of the constitutively active EGFR mutant EGFRvIII,3 did not lead to the development of an effective drug.
The reasons for this disappointing lack of success are manifold. Poor penetration of the blood-brain barrier by many targeted agents and lack of cytotoxic effect with some forms of growth factor blockade are widely recognized. Glioblastomas, the most common adult malignant intraparenchymal tumor, are not uniform in molecular pathology, with only some patients’ tumors expressing EGFRvIII, and no single, consistent dependence known on a single growth factor or tyrosine kinase. Even within individual glioblastomas, intratumoral heterogeneity manifests at the single-cell level; adjacent cells within the same tumor can have widely differing expression of multiple tyrosine kinases.4 Thus, even if a targeted therapy were to have a significant cytotoxic effect on one subpopulation of glioblastoma cells, other subclones within the same tumor can rapidly expand, and the overall tumor growth dynamic is little affected. Although treatment resistance is less well-understood in medulloblastoma, the smoothened (SMO) antagonist vismodegib similarly evokes responses in a minority of sonic hedgehog-subgroup medulloblastoma tumors, and responses seen are often short-lived.5 More research on intratumoral heterogeneity of medulloblastoma and other brain tumors is badly needed.
These disappointments in treating the most common adult and pediatric malignant brain tumors may have obscured some significant progress in treating other, less common brain tumors with targeted agents. One successful paradigm for targeted treatment in solid tumors was used against rare, but genetically more homogenous tumors that lack the high somatic mutation rates and intratumoral heterogeneity characteristic of glioblastoma. Among the earliest successes in treating solid tumors with targeted agents was the discovery that a rare form of sarcoma, gastrointestinal stromal tumor (GIST), commonly expresses a constitutively activated KIT tyrosine kinase that is efficiently blocked by imatinib. Clinical activity of the drug with frequent relief of symptoms and imaging responses was seen in phase I testing in advanced GIST tumors,6 and the drug received FDA approval for this use in 2002. Although imatinib resistance eventually occurs in some GISTs after prolonged treatment, often through emergence of second-site mutated KIT or KIT overexpression, the drug almost immediately improved GIST survival in the United States population, similar to its success in CML.7
A parallel example in brain tumor treatment is the success of everolimus against subependymal giant cell astrocytoma (SEGA), a signature lesion of the tuberous sclerosis complex (TSC). Everolimus is an inhibitor of mammalian target of rapamycin, or mTOR, which is consistently activated in tuberous sclerosis-related tumors due to loss of function of TSC1- or TSC2- encoded proteins. An mTOR inhibitor thus would be expected to interrupt the hyperactivated pathway through its action downstream to the activating lesion. Indeed, everolimus does cause clinically significant regression in SEGAs in most patients, with responses that are both rapid and durable.8 In phase III testing, everolimus caused regression of SEGAs, TSC-related skin lesions, and renal angiomyolipomas9 with little toxicity. Subsequent testing has shown clinical benefits of everolimus in TSC, including reduction in treatment-refractory seizures,10 and perhaps in other associated neurobehavioral problems as well.
Although the three most common intracranial tumors—glioblastoma, meningioma, and metastasis—dominate both neurosurgical oncology practice and brain tumor research, the 2016 WHO classification of brain tumors describes 155 distinct brain tumor entities, many of which are rare and poorly characterized.11 With increasing research into these arcane tumors, an overlapping array of tumor types that harbor treatable mutations or other molecular alterations are being discovered. For example, the highly treatable BRAF V600E mutation discovered in malignant melanoma has now been found in some glioblastomas, in particular those with epithelioid or rhabdoid morphology, some pilocytic astrocytomas, dysembryoplastic neuroepithelial tumors (DNETs), angiocentric gliomas, and pleomorphic xanthoastrocytomas (PXAs), as well as many gangliogliomas, and nearly all papillary subtype craniopharyngiomas. Improved treatments for all of these tumor types are badly needed, and the potential for effective and safe therapy with targeted agents is an attractive one. Anecdotal reports of significant tumor responses to agents specific for the V600E mutated BRAF protein, as well as to inhibition of the downstream MEK protein, have been reported in many of these tumor types.
Designing and accruing to traditional prospective phase II drug trials in many of these rare entities will pose a significant logistic challenge, even in the context of NCI-supported clinical trials groups. The Pediatric Brain Tumor Consortium has recently completed a phase I study of AZD6244, an orally available MEK inhibitor, in pediatric low-grade gliomas (NCT01089101), and a phase II trial is planned. In addition, the Alliance for Clinical Trials in Oncology is developing a phase II protocol of combined oral V600E BRAF and MEK inhibitors in papillary craniopharyngioma. These, and other similar trials, will require the aggressive support of the neurosurgical oncology community if we hope to see lasting progress in the management of these uncommon, but potentially life-threatening conditions. Ultimately, building a firm foundation in molecular characterization of all brain tumor types, harnessed to the emerging panoply of novel targeted treatments developed for other tumors, will allow the repurposing of many powerful targeted treatments for tumors affecting the brain and nervous system.
1 Delmonte L. Imatinib superior to standard therapy in newly diagnosed CML. Plenary paper. Oncology Times June 2002, p. 57.
2 Leukemia—chronic myeloid—CML statistics. http:// www.cancer.net/cancer-types/leukemia-chronicmyeloid- cml/statistics. Accessed January 26, 2017.
3 Mellinghoff IK, Wang MY, Vivanco I, et al. Molecular determinants of the response of glioblastomas to EGFR kinase inhibitors. N Engl J Med. 2005 Nov 10;353(19):2012-24.
4 Snuderl M, Fazlollahi L, Le LP, Nitta M, et al. Mosaic amplification of multiple receptor tyrosine kinase genes in glioblastoma. Cancer Cell. 2011 Dec 13;20(6):810-7.
5 Robinson GW, Orr BA, Wu G, et al. Vismodegib Exerts Targeted Efficacy Against Recurrent Sonic Hedgehog-Subgroup Medulloblastoma: Results from Phase II Pediatric Brain Tumor Consortium Studies PBTC-025B and PBTC-032. J Clin Oncol. 2015 Aug 20;33(24):2646-54.
6 Van Oosterom AT, Judson I, Verweij J, Stroobants S, et al. European Organisation for Research and Treatment of Cancer Soft Tissue and Bone Sarcoma Group. Safety and efficacy of imatinib (STI571) in metastatic gastrointestinal stromal tumours: a phase I study. Lancet. 2001 Oct 27;358(9291):1421-3
7 Perez EA, Livingstone AS, Franceschi D, et al. Current incidence and outcomes of gastrointestinal mesenchymal tumors including gastrointestinal stromal tumors. J Am Coll Surg. 2006 Apr;202(4):623-9.
8 Krueger DA, Care MM, Holland K, et al. Everolimus for subependymal giant-cell astrocytomas in tuberous sclerosis. N Engl J Med. 2010 Nov 4;363(19):1801-11.
9 Franz DN, Belousova E, Sparagana S, et al. Efficacy and safety of everolimus for subependymal giant cell astrocytomas associated with tuberous sclerosis complex (EXIST-1): a multicentre, randomised, placebo-controlled phase 3 trial. Lancet. 2013 Jan 12;381(9861):125-32.
10 French JA, Lawson JA, Yapici Z, et al. Adjunctive everolimus therapy for treatment-resistant focal-onset seizures associated with tuberous sclerosis (EXIST-3): a phase 3, randomised, double-blind, placebo-controlled study. Lancet. 2016 Oct 29;388(10056):2153-2163.
11 Louis DN, Ohgaki H, Wiestler OD, et al. WHO Classification of Tumours of the Central Nervous System. Lyon: IARC, 2016