Postoperative risk of IDH-mutant glioma–associated seizures and their potential management with IDH-mutant inhibitors
Michael R. Drumm, … , Geoffrey T. Swanson, Craig Horbinski
Michael R. Drumm, … , Geoffrey T. Swanson, Craig Horbinski
Published April 27, 2023
Citation Information: J Clin Invest. 2023;133(12):e168035. https://doi.org/10.1172/JCI168035.
View: Text | PDF
Research Article Neuroscience Oncology

Postoperative risk of IDH-mutant glioma–associated seizures and their potential management with IDH-mutant inhibitors

Abstract

Seizures are a frequent complication of adult-type diffuse gliomas, and are often difficult to control with medications. Gliomas with mutations in isocitrate dehydrogenase 1 or 2 (IDHmut) are more likely than IDH–wild type (IDHwt) gliomas to cause seizures as part of their initial clinical presentation. However, whether IDHmut is also associated with seizures during the remaining disease course, and whether IDHmut inhibitors can reduce seizure risk, are unclear. Clinical multivariable analyses showed that preoperative seizures, glioma location, extent of resection, and glioma molecular subtype (including IDHmut status) all contributed to postoperative seizure risk in adult-type diffuse glioma patients, and that postoperative seizures were often associated with tumor recurrence. Experimentally, the metabolic product of IDHmut, d-2-hydroxyglutarate, rapidly synchronized neuronal spike firing in a seizure-like manner, but only when non-neoplastic glial cells were present. In vitro and in vivo models recapitulated IDHmut glioma–associated seizures, and IDHmut inhibitors currently being evaluated in glioma clinical trials inhibited seizures in those models, independent of their effects on glioma growth. These data show that postoperative seizure risk in adult-type diffuse gliomas varies in large part by molecular subtype, and that IDHmut inhibitors could play a key role in mitigating such risk in IDHmut glioma patients.

Authors

Michael R. Drumm, Wenxia Wang, Thomas K. Sears, Kirsten Bell-Burdett, Rodrigo Javier, Kristen Y. Cotton, Brynna Webb, Kayla Byrne, Dusten Unruh, Vineeth Thirunavu, Jordain Walshon, Alicia Steffens, Kathleen McCortney, Rimas V. Lukas, Joanna J. Phillips, Esraa Mohamed, John D. Finan, Lucas Santana-Santos, Amy B. Heimberger, Colin K. Franz, Jonathan Kurz, Jessica W. Templer, Geoffrey T. Swanson, Craig Horbinski

×

Figure 4

Modeling and treating IDHmut glioma–associated seizure-like activity with cerebral spheroids.

Options: View larger image (or click on image) Download as PowerPoint
Modeling and treating IDHmut glioma–associated seizure-like activity wit...
(A) Serial imaging demonstrates infiltration of GFP-labeled patient-derived TB09 into a spheroid over 1–48 hours. Scale bar: 1 mm. (B) Spheroids were cocultured with patient-derived glioma lines (GBM6, n = 6; GBM12, n = 32; GBM43, n = 23; TB09, n = 4; B23, n = 5), and fold change in weighted mean firing rate was calculated over time. *P < 0.05 vs. IDHwt cells via 2-tailed t test. Experiment was repeated twice for a total of 3 times. (C) Spheroids were treated with either vehicle (n = 16) or 0.1 mM D2HG (n = 4), 0.5 mM D2HG (n = 10), 1 mM D2HG (n = 10), 3 mM D2HG (n = 10), 5 mM D2HG (n = 4), or 10 mM D2HG (n = 12), and fold change of weighted mean firing rate for 10 minutes after treatment was calculated for each well. Bars represent mean ± SEM. *P < 0.05, **P < 0.01 by unpaired, 2-tailed t test. Experiment was done in triplicate. (D) Spheroids were cultured with patient-derived IDH1mut glioma TB09 and treated with either vehicle (n = 6) or 1.5 μM AGI5198 (n = 7), and fold change in weighted mean firing rate was calculated over time. Data points represent mean ± SEM. Data were analyzed with 2-way ANOVA. Experiment was done once. (E) Spheroids were either cultured alone and treated with vehicle (n = 12), or cultured with patient-derived IDH1mut glioma 905 and treated with either vehicle (n = 14) or 30 nM AG881 (n = 13). Data points represent mean ± SEM. Data were analyzed with 2-way ANOVA. Experiment was done twice.