Editors' Note: Effectiveness of Antiseizure Medication Duotherapies in Patients With Glioma: A Multicenter Observational Cohort Study.

Abstract

In a multicenter retrospective cohort study of 1,435 patients with glioma followed up to 36 months, Dr. van der Meer et al. examined whether levetiracetam combined with valproic acid (LEV + VPA), a commonly prescribed duotherapy for uncontrolled seizures, is more effective than other duotherapy combinations that included either LEV or VPA. Of 355 patients receiving duotherapy, 66% received LEV + VPA, and patients receiving other duotherapy had a higher risk of treatment failure due to uncontrolled seizures but not due to adverse effects. The authors concluded that LEV + VPA had better efficacy than other antiseizure medication combinations in their patient population. In response, Dr. Zhao et al. suggest reclassifying the patients' glioma grades according to the updated 2021 World Health Organization guidelines instead of the 2016 guidelines that they had used, given the potentially different implications for treatment and epilepsy risk. They also argue that differences in the use of corticosteroids and surgical resection between the LEV + VPA group and the comparison group may have influenced the severity of epilepsy. Responding to these comments, the authors note that their patients were treated between 2004 and 2018 when the 2016 WHO criteria were in use, and they contend that reclassifying the patients would not change their results, pointing to the absence of evidence that antiseizure medications would have different efficacies in low-grade vs high-grade glioma. They acknowledge the lack of data on corticosteroid use for their cohort, but point to complexities in incorporating such data into analyses even if they were available. They note that they did adjust for surgical resection as a potential confounder in their Cox proportional hazards model but acknowledge the need for validation of their findings in a randomized controlled trial. This exchange demonstrates important confounders that can arise in observational studies of epilepsy treatment in patients with brain tumors such as gliomas. In a multicenter retrospective cohort study of 1,435 patients with glioma followed up to 36 months, Dr. van der Meer et al. examined whether levetiracetam combined with valproic acid (LEV + VPA), a commonly prescribed duotherapy for uncontrolled seizures, is more effective than other duotherapy combinations that included either LEV or VPA. Of 355 patients receiving duotherapy, 66% received LEV + VPA, and patients receiving other duotherapy had a higher risk of treatment failure due to uncontrolled seizures but not due to adverse effects. The authors concluded that LEV + VPA had better efficacy than other antiseizure medication combinations in their patient population. In response, Dr. Zhao et al. suggest reclassifying the patients' glioma grades according to the updated 2021 World Health Organization guidelines instead of the 2016 guidelines that they had used, given the potentially different implications for treatment and epilepsy risk. They also argue that differences in the use of corticosteroids and surgical resection between the LEV + VPA group and the comparison group may have influenced the severity of epilepsy. Responding to these comments, the authors note that their patients were treated between 2004 and 2018 when the 2016 WHO criteria were in use, and they contend that reclassifying the patients would not change their results, pointing to the absence of evidence that antiseizure medications would have different efficacies in low-grade vs high-grade glioma. They acknowledge the lack of data on corticosteroid use for their cohort, but point to complexities in incorporating such data into analyses even if they were available. They note that they did adjust for surgical resection as a potential confounder in their Cox proportional hazards model but acknowledge the need for validation of their findings in a randomized controlled trial. This exchange demonstrates important confounders that can arise in observational studies of epilepsy treatment in patients with brain tumors such as gliomas.