Summary
A Nature Medicine study links levetiracetam use with longer survival in children with diffuse midline glioma and reports reduced tumour growth in mouse models. The findings point to a tumour-selective effect on GABAergic neuron-to-glioma synapses and require prospective clinical testing.
A study in Nature Medicine has linked use of the antiseizure medicine levetiracetam with longer overall survival in children with diffuse midline glioma (DMG), a high-grade brain tumour that commonly arises in the brainstem, thalamus or spinal cord. The clinical finding came from retrospective patient records; experiments in mice and cell cultures provided evidence for a possible biological mechanism.
The researchers found that levetiracetam reduced tumour growth in several DMG models by weakening GABAergic synaptic signalling between neurons and glioma cells. The effect was seen in DMG models but not in the hemispheric high-grade glioma models examined in the study.
A survival association concentrated in DMG
The primary clinical analysis combined paediatric high-grade glioma records from Stanford University and the University of Michigan. Among 218 patients, those who had received levetiracetam had a median overall survival of 21.6 months, compared with 11.5 months among those without a recorded history of levetiracetam use.
The difference was concentrated in the DMG subgroup. Among 119 children with DMG, the 15 patients with a history of levetiracetam use had a median overall survival of 20.96 months, compared with 9.92 months for the 104 patients without such a history. The difference had a reported P value of 0.015.
A separate analysis of paediatric hemispheric high-grade glioma found similar survival with and without levetiracetam: 24.0 months among 40 treated patients and 22.2 months among 59 untreated patients, with P = 0.96.
Two additional retrospective datasets showed the same direction of association in DMG. In records from Hamburg, Germany, median survival was 24.0 months among six children with levetiracetam use and 11.0 months among 19 without it. In a University of California, San Francisco cohort, the corresponding figures were 15.5 months for five treated children and 12.2 months for 30 children without levetiracetam use.
These clinical records describe an association rather than a prospective treatment comparison. The levetiracetam-treated groups were small, and treatment was not assigned by the study investigators. The authors therefore propose prospective studies that stratify patients by tumour location and molecular subtype.
A tumour-selective synaptic mechanism
DMG cells can receive direct synaptic input from neurons. In these tumours, GABAergic signalling can be depolarising rather than inhibitory because malignant cells maintain a high intracellular chloride concentration. The resulting electrical activity promotes glioma-cell proliferation. This differs from the usual role of GABA in mature neurons, where it generally suppresses electrical activity.
In mouse experiments, the researchers implanted patient-derived H3K27M-positive DMG cells into the brain and administered levetiracetam at 20 milligrams per kilogram, five days a week for four weeks. Treatment extended survival, reduced tumour burden and lowered the proportion of proliferating malignant cells across multiple DMG models, including patient-derived xenografts and a genetically engineered mouse model.
The same treatment did not significantly reduce proliferation in three patient-derived hemispheric high-grade glioma models. Cell cultures grown without neurons also showed no direct antiproliferative response to levetiracetam, indicating that the drug’s effect depended on interactions within the brain environment rather than simply poisoning the tumour cells.
Electrophysiological recordings supplied a more specific explanation. After 30 minutes of exposure to 100 micromolar levetiracetam, GABAergic postsynaptic currents in DMG cells fell substantially. Under the same low-frequency stimulation conditions, levetiracetam did not reduce GABAergic currents in healthy neurons or glutamatergic currents between neurons and glioma cells.
Levetiracetam is commonly understood to act through SV2A, a synaptic-vesicle protein involved in neurotransmitter release. However, the study’s experiments indicated that the DMG effect was independent of SV2A. Brivaracetam, another antiseizure drug that specifically targets SV2A, did not reduce the recorded DMG GABAergic currents. Levetiracetam also continued to suppress glioma proliferation when SV2A expression in GABAergic neurons had been reduced by about 55 percent.
Why prospective testing is needed
The combined patient and laboratory results support the idea that levetiracetam may be repurposed to disrupt a growth-promoting neural circuit in DMG. They also suggest that antiseizure medicines may affect gliomas differently depending on the tumour’s neurophysiology: ethosuximide and phenytoin did not alter DMG proliferation in the tested mouse models, whereas perampanel and levetiracetam, which affect synaptic transmission, reduced growth.
The clinical evidence remains retrospective, with only 15 levetiracetam-treated patients in the main DMG cohort and 11 across the two validation cohorts. Differences in tumour location, seizure history, treatment decisions or other clinical factors could influence the survival association. Larger prospective studies, with molecular and anatomical stratification, are needed to determine whether levetiracetam itself improves outcomes.
The molecular target responsible for the tumour-selective, SV2A-independent reduction in GABAergic signalling has not yet been identified. That mechanism, and whether it can be translated into a reliable treatment benefit for children with DMG, are the study’s main directions for further research.