🔬 NEW IN SOLEIL RESEARCH™
Levetiracetam in Diffuse Midline Glioma: What the New Nature Medicine Study Found
Could a widely used antiseizure medication interfere with one of the electrical signaling pathways that helps diffuse midline glioma grow?
A new open-access study published in Nature Medicine on September 17, 2026, reports a striking translational signal. In retrospective clinical cohorts, children with diffuse midline glioma (DMG) who had received levetiracetam lived longer than those without documented levetiracetam use. In orthotopic tumor models, the drug reduced tumor growth and extended survival. Electrophysiology experiments then pointed to a possible mechanism: attenuation of GABAergic neuron-to-glioma synaptic signaling.
The convergence is scientifically compelling. But the human evidence remains observational, the exposed groups are small, and the study cannot establish that levetiracetam itself caused the survival difference.
Plain-Language Summary
Diffuse midline glioma is an aggressive tumor that develops in midline structures of the brain or spinal cord. Researchers reviewed clinical records from several institutions and found that children with DMG who had taken levetiracetam, a medication commonly used to treat seizures, tended to live longer than children who had not taken it.
The investigators also tested levetiracetam in laboratory models of DMG. In mice carrying patient-derived tumors, treatment reduced tumor burden and cell proliferation and prolonged survival. Additional experiments suggested that levetiracetam may weaken GABAergic synaptic signals sent from neurons to glioma cells—signals that can promote DMG growth.
These findings do not prove that levetiracetam is an effective anticancer treatment for patients with DMG. The clinical data were retrospective rather than randomized, and only a small number of children in the cohorts had received levetiracetam. A prospective clinical study is needed before the drug can be considered a validated DMG-directed therapy.
Why This Study Matters
Gliomas do not grow in isolation. A growing body of research shows that malignant glioma cells can integrate into neural circuits and exploit normal neuronal signaling. DMG appears to use both glutamatergic and GABAergic neuron-to-glioma communication to support proliferation.
That creates an unusual therapeutic possibility: instead of targeting only a tumor-intrinsic mutation or signaling pathway, researchers may be able to disrupt the communication between neurons and cancer cells.
Levetiracetam is especially interesting in this context because it is already widely used as an antiseizure medication in patients with brain tumors. The question raised by this study is whether it may have a second, tumor-relevant effect in a biologically specific glioma subtype.
What the Researchers Did
The study combined several layers of evidence rather than relying on a single experiment.
First, investigators analyzed retrospective records from Stanford University and the University of Michigan. The overall pediatric high-grade glioma dataset included 218 patients. Within that group, 119 had DMG: 15 had a history of levetiracetam use and 104 did not.
The researchers then examined two additional DMG validation cohorts from University Medical Center Hamburg-Eppendorf and the University of California, San Francisco. They also analyzed children with non-DMG hemispheric high-grade glioma as a biologically informative comparator.
Second, the team tested levetiracetam in patient-derived orthotopic DMG xenografts and another genetically engineered DMG model. These experiments asked whether the clinical association could be reproduced under controlled preclinical conditions.
Third, whole-cell patch-clamp electrophysiology and neuron-glioma co-culture experiments were used to investigate how levetiracetam changes synaptic signaling between GABAergic neurons and DMG cells.
What the Human Data Found
In the primary DMG cohort, children with a history of levetiracetam use had a median overall survival of 20.96 months, compared with 9.92 months among children without levetiracetam use. The reported log-rank comparison was statistically significant (P = 0.015).
The association was not seen in the hemispheric pediatric high-grade glioma comparator group: median overall survival was 24.0 months among 40 children with levetiracetam exposure and 22.2 months among 59 children without exposure (P = 0.96).
The two external DMG cohorts showed the same directional pattern. In Hamburg, six children with levetiracetam exposure had a median survival of 24.0 months versus 11.0 months among 19 without exposure. At UCSF, five exposed children had a median survival of 15.5 months versus 12.2 months among 30 unexposed children. The combined validation dataset included only 11 levetiracetam-exposed children, so the replication is encouraging but still based on very small numbers.
The investigators also used elastic net-regularized Cox regression for covariate selection in the larger retrospective dataset, with 20-fold cross-validation. That strengthens the statistical analysis, but multivariable modeling cannot remove unmeasured confounding from a nonrandomized historical cohort.
What the Preclinical Models Found
In mice bearing patient-derived orthotopic H3K27M-positive DMG xenografts, levetiracetam treatment prolonged survival, reduced tumor burden, and reduced malignant-cell proliferation. The antiproliferative effect was observed across multiple independent DMG models.
Importantly, levetiracetam did not reduce DMG cell proliferation when the tumor cells were grown alone in monoculture. That suggests the effect depends on interactions with the neural microenvironment rather than a simple direct toxic effect on glioma cells.
The same pattern of benefit was not observed in three hemispheric H3/IDH-wild-type high-grade glioma models. This tumor-subtype specificity parallels the retrospective clinical findings and strengthens the biological coherence of the study.
A Possible Mechanism: GABAergic Neuron-to-Glioma Synapses
Patch-clamp recordings showed that levetiracetam progressively reduced low-frequency GABAergic postsynaptic currents in DMG cells. Under the same experimental conditions, the drug did not produce the same effect on normal neuronal GABAergic currents.
The investigators also explored whether this effect depended on SV2A, the canonical molecular target responsible for much of levetiracetam’s antiseizure activity. Brivaracetam, a more selective SV2A ligand, did not reproduce the same effect on DMG GABAergic currents, and reducing SV2A expression in GABAergic neurons did not eliminate levetiracetam’s antiproliferative effect in neuron-glioma co-culture.
These experiments support the hypothesis that levetiracetam may attenuate tumor-promoting GABAergic neuron-to-glioma signaling through a mechanism that is at least partly distinct from its standard antiseizure action. The exact tumor-selective molecular mechanism remains unresolved.
Methodological Quality
Overall assessment: strong translational coherence, but low-to-moderate causal certainty in humans.
The paper is unusually persuasive as a mechanistic translational study because several independent evidence streams point in the same direction: a tumor-subtype-specific clinical association, external retrospective validation cohorts, orthotopic patient-derived models, and direct electrophysiological evidence of a plausible synaptic mechanism.
That convergence substantially raises the scientific priority of prospective testing. It does not, however, convert an observational survival association into proof of treatment efficacy.
Strengths
- Multi-institution clinical evidence: the retrospective signal was examined across four institutions rather than a single database.
- Biological comparator: the absence of a similar association in hemispheric high-grade glioma provides an informative negative comparator consistent with the proposed DMG-specific mechanism.
- External validation: Hamburg and UCSF cohorts showed the same directional survival pattern.
- Translational triangulation: human data, orthotopic tumor models, co-culture systems, and electrophysiology converge on one hypothesis.
- Multiple DMG models: the antiproliferative effect was reproduced across several patient-derived and engineered models.
- Mechanistic specificity: the study tested GABAergic currents, glutamatergic currents, neuronal currents, SV2A dependence, and alternative antiseizure drugs rather than attributing the effect to a generic reduction in neuronal activity.
- Transparent reporting: source data and peer-review reports are available with the article.
Limitations and Cautions
- Retrospective clinical design: levetiracetam exposure was not randomized. Confounding by indication, tumor location, seizure biology, treatment selection, and other measured or unmeasured factors may explain part of the survival difference.
- Very small exposed groups: only 15 children with DMG received levetiracetam in the primary cohort, with 11 additional exposed children across the two validation cohorts.
- Historical heterogeneity: the primary databases span 1990–2020 at Stanford and 2012–2021 at Michigan, during which molecular classification, imaging, radiotherapy, supportive care, and systemic treatment evolved substantially.
- Incomplete molecular characterization: older cases included radiographically diagnosed DMG from periods before routine molecular testing.
- Potential time-related bias: retrospective medication exposure can be influenced by when treatment began relative to diagnosis and survival time. The report does not provide the causal protection against time-varying exposure bias that a randomized prospective design would offer.
- Small preclinical groups: several animal and electrophysiology experiments used modest sample sizes. Reproducibility across models is helpful, but translation to human treatment remains uncertain.
- No prospective therapeutic trial: the study does not establish an anticancer dose, optimal treatment timing, duration, benefit-risk profile, or whether levetiracetam should be prescribed to seizure-naive patients for tumor control.
Clinical Interpretation
The survival difference is large enough to warrant serious scientific attention, but it should be described as an association, not a treatment effect.
Levetiracetam already has an established role as an antiseizure medication, but this publication is not a randomized trial of levetiracetam as DMG therapy. It therefore should not be used by itself to justify prophylactic or off-label anticancer prescribing. The appropriate next step is prospective evaluation with clearly defined exposure, molecular and anatomic stratification, standardized concomitant treatment, and prespecified survival and safety endpoints.
Funding and Conflicts of Interest
The work was supported by multiple public, academic, philanthropic, and disease-focused research organizations, including the National Institute of Neurological Disorders and Stroke, the National Cancer Institute, the NIH Director’s Pioneer Award, ChadTough Defeat DIPG, Cancer Research UK, Cancer Grand Challenges, and several foundations and family research funds. The publication does not identify a levetiracetam manufacturer as a study funder.
Michelle Monje disclosed equity in MapLight Therapeutics and Stellaromics and service on the scientific advisory board of S1 Oncology. The other authors declared no competing interests. These relationships do not invalidate the findings, but they are relevant context for transparent evidence appraisal.
🔎 Soleil Insight
The strength of this paper is not a single survival curve. It is the convergence of a tumor-subtype-specific clinical association, orthotopic model response, and a biologically plausible synaptic mechanism.
That convergence raises the priority of a prospective clinical trial—but it does not replace one. If the survival association is reproducible under controlled prospective conditions, the study could become an important example of drug repurposing guided by cancer neuroscience. Until then, the evidence should be viewed as hypothesis-strengthening rather than practice-changing.
Bottom Line
This Nature Medicine study provides a compelling translational rationale for testing levetiracetam in diffuse midline glioma. Children with documented levetiracetam exposure had longer survival in retrospective cohorts, DMG models responded to the drug in vivo, and electrophysiology identified attenuation of GABAergic neuron-to-glioma signaling as a plausible mechanism.
The result is promising, not definitive. Prospective trials are necessary to determine whether levetiracetam truly improves patient outcomes, which patients might benefit, when treatment should begin, and whether the effect is clinically meaningful beyond seizure management.
Original Publication
Barron T, Drexler R, Mochizuki A, et al. Levetiracetam therapeutically targets GABAergic synapses in diffuse midline glioma. Nature Medicine. Published September 17, 2026. DOI: 10.1038/s41591-026-04646-6.
Featured image: Figure 5 from Barron et al., Nature Medicine (2026), showing experimental evidence that levetiracetam reduces GABAergic neuron-to-glioma synaptic currents. Reproduced from the open-access article under the Creative Commons Attribution 4.0 International License. No changes were made to the figure.
This article summarizes peer-reviewed scientific evidence for educational purposes and does not constitute individual medical advice.
Advancing Science. Empowering Humanity.