Abstracts

Rationale: Recent evidence indicates a bidirectional relationship between gliomagenesis and epileptogenesis involving the upregulation of peritumoral glutamatergic activity, associated with neuronal-glial synaptogenesis, excitotoxicity, and tumor growth. The role of this mechanism in the pathogenesis of human high-grade glioma and tumor-associated seizures remains poorly understood. The development of biomarkers for glutamatergic hyperexcitability is necessary to translate metabolic alterations into targeted treatments. We implemented a novel method of mass spectrometry imaging (MSI) of small molecules (m/z 50-1500) and drugs combined with analysis of high frequency oscillations (HFOs) to spatially correlate molecular data with pathologic neuronal activity, and to evaluate the effect of targeted therapy with perampanel, a noncompetitive AMPA glutamate receptor antagonist.

Methods: Participants with radiologic evidence of newly diagnosed high-grade glioma were enrolled into the ongoing NCT04497142 trial, consisting of a perampanel treatment cohort (6mg pre-operative loading dose followed by 4mg daily maintenance) and a standard of care seizure treatment cohort (peri-operative levetiracetam as clinically indicated). All patients underwent intraoperative electrocorticography (ECoG), sampled at 1000 Hz/channel, with a 4x5 contact subdural grid overlying the maximal exposed tumor margin. Electrode coordinates and tissue biopsy sites were coregistered to patient space with the neuronavigation system. The rate of HFOs (80-500 Hz) were quantified using an automated Hilbert envelope detection method and visually validated. The spatial distribution of metabolites and perampanel/levetiracetam in tissue was analyzed using previously validated matrix-assisted laser desorption ionization (MALDI) MSI protocols.

Results: Analysis of ECoG and molecular imaging data was performed in four cases (three WHO Grade 4 IDH-wildtype glioblastoma, one WHO Grade 3, IDH-mutated, 1p/19q co-deleted anaplastic oligodendroglioma), revealing focal anatomic clustering of HFOs in each case. HFO rate was higher peritumorally < 1cm from the contrast-enhancing margin (mean 0.58 per minute [range 0-2.8]) and in cortex >1cm away (0.52 [0-5.5]), compared to contacts directly overlying enhancing tumor (0.11 [0-1.3], margin/cortex vs tumor p=0.007). Within-subject MSI analysis confirmed the feasibility of correlating glutamate signal intensity with HFO rate at adjacent ECoG contacts. MALDI MSI demonstrated detection of perampanel in tissue specimens at drug concentrations below 1 μM, with qualitatively homogenous distribution across biopsy sites (range 0.24-0.30 μM) in integrated analysis of one participant.

Conclusions: The methodology described here demonstrates the ability to correlate local tissue metabolic data with electrophysiologic biomarkers of peritumoral neuronal hyperexcitability in patients with high-grade glioma. This platform has the potential to identify biomarkers for the targeted treatment and prevention of glioma-related epilepsy.

Funding: Please list any funding that was received in support of this abstract.: NCI P50CA165962
Eisai, Inc FYC-IIS-M001-1100.