Chemotherapy and immunotherapy of brain tumors: What the epileptologist must know


  • Lisa R. Rogers

    Corresponding author
    1. Department of Neurology, Seidman Cancer Center, University Hospitals Case Medical Center, Cleveland, Ohio, U.S.A
    • Address correspondence to Lisa R. Rogers, Department of Neurology, University Hospitals Case Medical Center, 11100 Euclid Avenue, Cleveland, OH 44106, U.S.A. E-mail:

    Search for more papers by this author


I present an overview of therapy for the most common brain tumors encountered in clinical practice if adult patients. Current therapy paradigms and evolving therapies are reviewed. The introduction of non–enzyme-inducing antiepileptic drugs (NEIADs) has simplified the approach to combined medical treatment of epilepsy and brain tumors, but the major interactions between enzyme-inducing antiepileptic drugs (EIAEDs) are included, to serve as guidance in selecting these medications if they are required.


Gliomas are the most common adult primary brain tumors. Glioblastoma (GBM) is the most common glioma and is associated with a poor prognosis, despite aggressive multimodality therapy with surgery, radiation, and chemotherapy. More effective treatment strategies will rely on targeting underlying biologic abnormalities within these tumors that are associated with tumor proliferation and invasion, and this is an active area of current research. The most common anaplastic gliomas (AGs) are anaplastic astrocytoma (AA), anaplastic oligodendroglioma (AO), and mixed anaplastic oligoastrocytoma (AOA). The prognosis in these tumors is considerably better than in GBM, especially in AO with codeletions of chromosomes 1p and 19q. Low-grade gliomas (LGGs), most commonly astrocytoma and less commonly oligodendroglioma, occur most often in young adults, who typically present with headache or seizures and have a normal neurologic examination. The long-term progression-free survival and overall survival of LGG patients is significantly better than those with GBM or AG, but malignant transformation occurs in up to 50% of such patients, and close monitoring is required.

Molecular markers

In addition to standard histomorphology features to identify different grades of gliomas, efforts are underway to predict and prognosticate outcome based on molecular abnormalities identified in tumor tissue. The first glioma molecular classification was based on the presence or absence of the chromosomal arms of 1p and 19q in AO (Cairncross et al., 1998). AO patients with codeletion of these chromosomes have a significantly better response to all therapies and survival than those without. Jenkins et al. (2006) have also shown that codeletion of 1p and 19q is associated with superior survival in low-grade oligodendrogliomas as well.

Intense research is ongoing to identify molecular markers in GBM that are predictive and prognostic of outcome, but these have not yet been identified. The Cancer Genome Atlas (TCGA) project has identified four subclasses of GBM: proneural, neural, classical, and mesenchymal. Each subclass has characteristic genomic abnormalities that may lead to a deeper understanding of which molecular “drivers” result in sustained tumor growth. It is unproven whether these classifications have predictive and prognostic significance. Isocitrate dehydrogenase (IDH)1 and IDH2 mutations in GBM are also currently being investigated as markers of outcome. There is some evidence that tissues harboring IDH1 mutations are correlated with a superior outcome as compared to wild-type IDH1 (Wang et al., 2013). The discovery of molecular “drivers” will hopefully lead to identification of targets for therapy. Currently, these types of molecular studies are performed in GBM tissue predominantly in the clinical research setting, in order to correlate molecular status with treatment outcome, and they are not used to dictate therapy in clinical practice.


Standard therapy for newly diagnosed GBM is maximal resection, and external beam radiation of 60 Gy over 6 weeks in combination with daily temozolomide followed by temozolomide for 6 months. Temozolomide is a DNA-methylating agent with high cerebrospinal fluid penetrance. In a prospective randomized trial in patients with newly diagnosed GBM, the median survival with radiation and temozolomide was 14.6 months versus 12.1 months with radiation alone. The 2-year survival rate was also superior with the combined regimen (26.5%) versus radiation alone (10.4%) (Stupp et al., 2005). Advantages of this drug, as compared with prior chemotherapies administered for GBM, include bioavailability with oral dosing, a convenient treatment schedule, acceptable patient tolerance, and a good safety profile, although a small percentage of patients experience severe and life-threatening myelosuppression. In addition, there are no significant interactions between temozolomide and EIAEDs. Current phase 0, 1, 2, and 3 therapeutic trials for patients with newly diagnosed GBM incorporate the addition of other treatments into the standard radiation and temozolomide regimen described above.

O6-methylguanine-DNA methyltransferase (MGMT) is a DNA repair gene that reduces the efficacy of temozolomide and other DNA-damaging treatments for cancer. Methylation of the MGMT promoter in tumor tissue silences this gene and results in improved survival in GBM (Hegi et al., 2005). Determination of the promoter status of MGMT is often obtained as a part of clinical research trials, but it is rarely used in clinical practice.

Anaplastic gliomas are treated by maximal surgical resection, followed by external-beam radiation. Despite the overall better prognosis for AGs, the role of chemotherapy has not been as well established as in GBM, but it is frequently recommended as adjuvant therapy. One AG that is exquisitely sensitive to chemotherapy is the AO with codeletions of 1p and 19q. Small clinical trials have evaluated the benefit of chemotherapy alone in these patients, and deferring radiation until the time of tumor progression, in an attempt to reduce the potential central nervous system (CNS) toxicity associated with brain radiation. The benefit of chemotherapy when added to radiation in newly diagnosed AGs, with or without codeletion of 1p and 19q, is currently under investigation via a cooperative group trial.

Prior to the use of temozolomide, the standard chemotherapy regimen for malignant gliomas, including GBM, was the “PCV” regimen consisting of procarbazine, an oral alkylating agent, CCNU (lomustine), an oral alkylating agent, and vincristine, a microtubule disrupter administered intravenously. This combination is associated with a moderate risk of myelosuppression and peripheral neuropathy. There is no significant effect of enzyme-inducing antiepileptic drugs (EIAEDs) on the pharmacokinetics of procarbazine (Grossman et al., 2006); however, there is an increased risk of procarbazine hypersensitivity reactions with this combination (Lehmann et al., 1997). There is a suggestion that administration of EIAEDs may decrease the efficacy of lomustine (Oberndorfer et al., 2005). EIAEDs can increase the clearance of vincristine, thereby decreasing the efficacy of this chemotherapy (Villikka et al., 1999).

The standard use of the PCV regimen declined significantly in recent years following the introduction of temozolomide. However, the late results of a cooperative group study in which the efficacy of radiation and PCV versus radiation alone to treat newly diagnosed AGs has led to a reconsideration that PCV may be a very effective regimen in the subset of AO patients with codeletions of 1p and 19q (van den Bent et al., 2013). A significantly improved median survival of such patients treated with radiation therapy and PCV (14.7 years) versus radiation therapy alone (7.3 years) has led to a redesigned study of an ongoing multisite international trial of radiation therapy with temozolomide versus radiation therapy alone to include treatment arms including PCV, in order to compare the efficacy of temozolomide in this setting.

Patients with LGGs should be treated initially with surgical resection. The appropriate aggressiveness of additional therapy was questioned for many years in LGG patients until a trial conducted by the European Organization for Research and Treatment of Cancer (EORTC) demonstrated that patients can be subdivided into good and bad prognosis, based on several risk factors (Karim et al., 2002). Patients <40 years of age who have undergone a complete tumor resection have the best prognosis, and radiation can be withheld in this subgroup. Radiation therapy does not affect overall survival, but delays the time to tumor progression in patients with poor prognostic factors. The role of chemotherapy in addition to radiation has not been established and is currently under investigation in a cooperative group trial.

Irinotecan is a camptothecin derivative that functions as a topoisomerase I inhibitor. It is used occasionally in combination with the anti-angiogenic agent bevacizumab (see below) and only rarely as a single agent. There is a significant decrease in the plasma levels of irinotecan and its metabolites when EIAEDs are administered, requiring an increase in irinotecan dose if these medications are used in combination.


The introduction of “targeted agents,” which are designed to deactivate oncogenic pathways, is one of the most significant advances in cancer treatment, including GBM, in recent years. Bevacizumab, a monoclonal antibody against vascular endothelial growth factor, is associated with a high response rate (6-month progression-free survival of 46%) in recurrent GBM, although the effect on overall survival is not significant (Vredenburgh et al., 2007). Bevacizumab is approved by the U.S. Food and Drug Administration for recurrent GBM and is often used as first-line treatment in this setting. This agent is also used to treat brain radiation necrosis. There is no significant interaction between bevacizumab and EIAEDs. In addition to angiogenesis, a variety of other pathways for tumor proliferation and invasion have been identified in GBM and may serve as therapeutic targets. To date, no other targeted agent has been identified to be of clinical significance, but research with such agents continues.

It has long been recognized that GBM patients have impaired immunity. This has led to a variety of treatment strategies designed to enhance the host immune response to GBM. Currently, several immunotherapies are in study for the treatment of newly diagnosed and recurrent GBM. These include peptide vaccines, including one directed toward epidermal growth factor (EGFR) variant III antigen, dendritic cell vaccines in which the host tumor is mixed and cultured with dendritic cells, and heat shock proteins derived from the tumor. Adoptive immunotherapy involves vaccination with lymphocytes from the host or the tumor and grown in the presence of immune stimulation. Several of these approaches show very promising results in clinical trials (Bregy et al., 2013). The results of phase III randomized controlled clinical trials are eagerly awaited.

Brain Metastasis

Standard therapy for patients with a single brain metastasis is surgical resection if the tumor is in a noneloquent part of the brain and if the extent of systemic disease dictates a survival of at least 4–6 months. In addition, some patients with limited systemic cancer show a survival benefit with resection of up to three metastatic tumors. Resection is typically followed by whole brain radiation or stereotactic surgery to the tumor margin. If the patient does not receive whole brain radiation, close observation with periodic magnetic resonance imaging (MRI) scans is indicated in order to assess for recurrence at the original site or at other sites in the brain. For patients presenting with multiple metastases (more than three lesions), whole brain radiation is indicated.

A variety of systemic chemotherapies show therapeutic efficacy in newly diagnosed and recurrent brain metastasis, depending primarily upon the sensitivity of the brain metastasis to the agent, rather than to delivery of the drug to the lesion. Because the blood–brain barrier is disrupted at the site of brain metastasis, the majority of the tumor is exposed to agents administered systemically.

Similarly, agents that are used for targeting growth signaling pathways in systemic cancers can be used to treat cancer that has spread to the brain. These include erlotinib, a tyrosine kinase inhibitor of EGFR used to treat some forms of lung cancer; and lapatinib, a dual tyrosine kinase inhibitor, used alone or with cytotoxic chemotherapy, to treat patients with her-2/neu negative breast cancer brain metastasis. Phenytoin reduces the blood levels and therapeutic effects of erlotinib. One of the most promising agents for treating brain metastasis in malignant melanoma is ipilimumab, an anti–cytotoxic T lympohcyte antigen-4 antibody.

Primary CNS Lymphoma

Primary CNS lymphoma (PCNSL) is a rare form of non-Hodgkin lymphoma. Chemotherapy, particularly high-dose intravenous methotrexate, given alone or in combination with other chemotherapy, is currently the mainstay of therapy based on prospective clinical trials (Batchelor et al., 2003). Combining high-dose methotrexate with standard dose brain radiation carries a significant risk of neurotoxicity, especially in elderly patients (DeAngelis et al., 2002). The use of reduced dose brain radiation following methotrexate chemotherapy is currently in clinical trial. If there is leptomeningeal lymphomatosis, intrathecal methotrexate or cytarabine may also be administered. Phenobarbital and carbamazepine reduce the cerebrospinal fluid (CSF) clearance of methotrexate and may thus result in enhanced methotrexate toxicity (Halwachs et al., 2011). Rituximab, a monoclonal antibody directed to the CD20 surface antigen, has low CSF penetrance but has demonstrated activity against newly diagnosed and recurrent PCNSL and is often used in these settings.


The author does not have any conflicts of interest to disclose. The author confirms that she has read the Journal's position on issues involved in ethical publication and affirms that this report is consistent with those guidelines.