Meta-analysis of treatment-emergent seizures in glioma vaccine trials.

purposeEpilepsy is commonly associated with gliomas, with glioma-neuron hyperexcitable interactions promoting tumor growth. Glioma vaccines are a promising form of immunotherapy, however immune-mediated effects on hyperexcitability are poorly understood. Seizures induce a proinflammatory state, whereas gliomas manipulate the microenvironment to evade adaptive immunity. This meta-analysis aimed to evaluate the incidence of seizures after glioma vaccine therapy as a clinical marker of peritumoral hyperexcitability.

methodsStudies were identified by query of the MEDLINE, Embase, and Web of Science databases with filtering for clinical trials and review of references. Inclusion criteria were prospective trials of cancer vaccines in patients with histologically confirmed diffuse glioma. Studies were excluded if seizure incidence was not reported or if data for all adverse events (AEs) regardless of treatment attribution was unavailable. Serious AEs were defined as grade 3 or higher by the Common Terminology Criteria for Adverse Events when available. Seizure incidence was calculated by pooled proportions with random effects using a generalized linear mixed model. Binary outcome meta-analysis of controlled trials was performed using the Mantel-Haenszel method with random effects.

resultsAfter applying pre-screening exclusion criteria, there were 125 studies screened for seizure-related AEs, and 53 studies with 2002 vaccine-treated patients were included in the final analysis. Treatment-emergent seizures of any grade occurred in 19.3% of patients (95% CI: 16.7–22.2%). Serious seizure AEs occurred in 4.3% (95% CI: 2.9–6.4%). In controlled trials, seizure rates of any grade were similar between vaccine and control groups (RR 1.20, 95% CI: 0.89–1.62, I2 = 8.5%), as well as for serious seizure AEs (RR 1.06, 95% CI: 0.65–1.73, I2 = 0%). There were no differences observed in seizure risk between dendritic cell, tumor cell, and peptide vaccines (p = 0.67).

conclusionSeizure-related data were not available for most glioma vaccine trials, likely due to lack of attribution to vaccine treatment. Using AE reporting, approximately 19% of patients had seizures during vaccine therapy and trial monitoring. In controlled trials, there was moderate certainty of similar seizure risk associated with vaccine versus control treatment. Integrating standardized detailed seizure tracking into future glioma trials may improve response assessments.