By Padmaja Kandula, MD

Assistant Professor of Neurology and Neuroscience, Comprehensive Epilepsy Center, Weill Cornell Medical College

Dr. Kandula reports no financial relationships relevant to this field of study.

SYNOPSIS: In this multicenter, observational study, the authors assessed the prevalence of neuropsychiatric side effects from medications in subjects with tumor-related epilepsy. Levetiracetam was found to have the highest prevalence of such side effects.

SOURCE: Bedetti C, et al. Neuropsychiatric adverse events of antiepileptic drugs in brain tumour-related epilepsy: An Italian multicentre prospective observational study. Eur J Neurol 2017;24:1283-1289.

The treatment of brain tumor-related epilepsy is a challenging paradigm, often requiring a fine balancing act between efficacy and side effect profile. Up to 50% of brain tumor patients may experience a seizure in the course of their disease. Hence, the variables associated with optimal efficacy and minimal adverse effects should dictate early appropriate anticonvulsant drug selection in these patients. This multicenter, prospective, observational study evaluated the tolerability of anticonvulsant use based on tumor localization, tumor subtype, and oncologic treatment.

Researchers recruited 259 patients with brain tumor-related epilepsy from regional health centers across Italy from 2010 to 2014. All patients were older than 18 years of age, had documented epilepsy secondary to either primary or metastatic brain tumor, were on anticonvulsant monotherapy, and had no previous exposure to prior anticonvulsants. Patients on anticonvulsant monotherapy, prior personal or family psychiatric history, inoperable brain tumor, end-stage renal disease, or pregnancy were excluded. All brain tumors were classified according to the 2007 WHO Classification of Tumors. Seizures were classified according to the International League Against Epilepsy Guidelines. Seizure frequency at baseline and at five-month follow-up were recorded. Anticonvulsant efficacy was defined as 50% reduction in seizure counts from baseline or seizure freedom. Practitioners were allowed to titrate anticonvulsants according to standard clinical practice, with titration periods not exceeding 14 days. The Standardized 12-item Neuropsychiatric Inventory Test was administered to all patients or caregivers at the time of inclusion and at the five-month mark.

Approximately 55% of enrolled patients were male. Median age was 52 years. Glioblastoma was the most common tumor pathology (43%), followed by astrocytoma (17%), oligodendroglioma (15%), meningioma (9.7%), metastasis (7%), oligoastrocytoma (7%), and the remaining patients comprising hemangiopericytoma, craniopharyngioma, and central neurocytoma.

Nearly 50% of tumors were in the frontal lobe. The temporal lobe was the second most common region for tumor localization (21%). The majority of the patients experienced focal seizures with or without impaired awareness (60%). Nearly 58% of patients were prescribed levetiracetam monotherapy, followed by phenobarbital (14.8%), oxcarbazepine (8.5%), and valproate (7%).

All patients had > 50% reduction in seizures from baseline. In addition, nearly 95% of patients had 75% seizure reduction rate. All patients underwent surgical resection, with 42% undergoing dexamethasone treatment, 48% chemotherapy with temozolomide, 46% radiotherapy, and 33% with combination dexamethasone, temozolomide, and radiation.

A multivariate analysis showed no statistically significant difference in neuropsychiatric prevalence and severity with regard to brain tumor side, age, gender, steroid dose, or chemotherapy or radiation alone or in combination. However, patients with frontal lobe tumors had a higher prevalence and severity of neuropsychiatric effects (odds ratio [OR], 7.73; P < 0.001) vs. non-frontal tumors. Among patients treated with levetiracetam, those with frontal lobe tumors again had a higher prevalence (OR, 5.00) and magnitude (P < 0.01) of neuropsychiatric side effects. There was no dose correlation between levetiracetam and side effects. Multivariate analysis specifically targeting anticonvulsant use revealed that patients receiving levetiracetam had higher prevalence and magnitude of neuropsychiatric effects, regardless of dose, compared to other anticonvulsants (OR, 7.94; P < 0.01). Non-neurobehavioral side effects were more common in first-generation anticonvulsants such as phenytoin, phenobarbital, and valproic acid.


Because of ease of administration, broad-spectrum coverage, and relative lack of drug interactions, levetiracetam increasingly has been used in the treatment of epilepsy and, specifically, in those with brain tumors. However, despite increasing use there has not been a large-scale effort to study the brain tumor population who may be susceptible to anticonvulsant side effects. These study results are not surprising, since levetiracetam has been implicated in a higher incidence of neuropsychiatric side effects. However, it is intriguing that in patients with brain tumors, specifically frontal lobe brain tumors, this effect was not dose dependent. Bedetti et al also raised the possibility of a synergistic effect of frontal lobe tumor localization and levetiracetam treatment. Patients were excluded based on familial or personal psychiatric history. However, those with pre-existing comorbid psychiatric diagnosis or family history may be even more susceptible to levetiracetam effects. Lastly, the authors hypothesized that SV2A gene expression in tumor tissue may be predictive of levetiracetam treatment response and risk of neuropsychiatric side effects. The authors’ proposed future histologic analysis of brain tumor genetic markers may be an important factor for treatment success and serve as a goal in this age of precision medicine.