Brain Stimulation Therapies Offer New Hope for Drug-Resistant Epilepsy
A comprehensive review of VNS, deep brain stimulation, and responsive neurostimulation reveals improving seizure control over time for refractory epilepsy.
Summary
Roughly 20-30% of epilepsy patients fail drug therapy and need alternative treatments. This review from Sapporo Medical University systematically covers three neuromodulation approaches: vagus nerve stimulation (VNS), deep brain stimulation (DBS) targeting multiple brain regions, and responsive neurostimulation (RNS). VNS offers broad applicability without craniotomy but modest efficacy (~50% responder rate). Anterior nucleus of thalamus DBS, now covered by Japanese insurance since 2023, achieves a 75% median seizure reduction at 7 years. RNS delivers on-demand stimulation triggered by detected seizure activity, reaching 75% median seizure reduction at 9 years. Each modality has distinct indications, benefits, and side-effect profiles, offering neurologists and neurosurgeons a growing toolkit for individualized treatment.
Detailed Summary
Epilepsy affects 4-8 per 1,000 people globally, and approximately 20-30% of patients develop drug-resistant disease. For these individuals, surgical resection is only possible when the epileptogenic focus is clearly localized and removable without neurological harm. Neuromodulation therapies—VNS, DBS, and RNS—have emerged as palliative alternatives, and this 2025 review from Sapporo Medical University synthesizes the current evidence for each.
VNS, the longest-established modality (FDA-approved 1997, Japan-approved 2010), delivers cyclic or demand-triggered electrical pulses to the left vagus nerve without requiring craniotomy. Pivotal trials (E03, E05) showed high-stimulation groups achieved 24-28% seizure reduction versus 6-15% for low stimulation. Meta-analyses of over 5,000 patients show responder rates (~50%) that improve progressively over 5 years. Newer AutoStim models detect heart-rate surges as seizure proxies, adding further benefit. Side effects are generally mild (hoarseness, throat pain), making VNS widely applicable regardless of seizure type.
Among DBS targets, the anterior nucleus of the thalamus (ANT) has the strongest evidence base. The landmark SANTE trial (110 patients) showed 40.4% median seizure reduction at 3 months versus 14.5% for controls, improving to 56% at 2 years, 69% at 5 years, and 75% at 7 years. ANT-DBS was approved in the US in 2018 and Japan in 2023. Psychiatric side effects (depression ~15%, memory impairment ~13%) are a concern given ANT's role in the Papez circuit. The centromedian nucleus (CM) is particularly effective for generalized epilepsies including Lennox-Gastaut syndrome, with one prospective study showing ≥50% seizure reduction in 90% of patients. The ESTEL trial demonstrated significant EEG seizure reduction (89% responder rate) with CM-DBS. Other investigated targets—hippocampus (temporal lobe epilepsy), subthalamic nucleus (motor seizures), pulvinar nucleus (occipital/temporal epilepsy), posteromedial hypothalamus, nucleus accumbens, and cerebellum—show variable or preliminary efficacy and require larger controlled studies.
RNS represents a closed-loop system: intracranially implanted electrodes continuously record EEG and automatically deliver brief stimulation bursts upon seizure detection. FDA-approved in 2013, it is especially suited for patients with eloquent cortex foci or bilateral temporal lobe epilepsy. The pivotal RCT (191 patients) demonstrated 37.9% seizure reduction versus 17.3% for sham. Long-term outcomes are compelling—75% median seizure reduction at 9 years, with 73% achieving ≥50% reduction and 18.4% becoming seizure-free for ≥1 year. A unique advantage is continuous long-term EEG recording, enabling detection of seizure cycles and informing future surgical planning. Emerging data also support RNS targeting deep structures like ANT and CM, with one cohort showing 82.6% mean seizure reduction at 1 year.
The review highlights that Japan lags behind Western nations in device availability—RNS is not yet approved in Japan, and CM-DBS lacks insurance coverage in both Japan and the US. Choosing among modalities requires individualized assessment of seizure type, focus localizability, patient age, and risk tolerance. Efficacy generally improves with longer stimulation duration across all three modalities, underscoring the importance of sustained treatment commitment.
Key Findings
- ANT-DBS achieves 75% median seizure reduction at 7 years, now covered by Japanese national insurance since 2023.
- VNS yields ~50% responder rates over 5 years and requires no craniotomy, making it broadly applicable.
- CM-DBS shows 90% of patients with ≥50% seizure reduction in generalized epilepsies including Lennox-Gastaut syndrome.
- RNS achieves 75% median seizure reduction at 9 years with 18.4% becoming seizure-free for ≥1 year.
- Hippocampal DBS demonstrates up to 95% seizure reduction in temporal lobe epilepsy patients with normal MRI.
Methodology
This is a narrative review article from a single academic neurosurgery department (Sapporo Medical University) synthesizing evidence from multiple RCTs, prospective studies, retrospective cohorts, and meta-analyses on VNS, DBS, and RNS. The review spans foundational pivotal trials to long-term follow-up registries and post-marketing data. It does not perform a formal systematic review or meta-analysis.
Study Limitations
As a narrative review without systematic search methodology or meta-analytic pooling, selection bias in included studies is possible. Many DBS targets (hippocampus, STN, pulvinar, cerebellum) are supported only by small, heterogeneous studies lacking sham-controlled designs. RNS and several DBS targets remain unapproved in Japan, limiting generalizability of the review's clinical recommendations to the Japanese context.
Enjoyed this summary?
Get the latest longevity research delivered to your inbox every week.
Enter your email to subscribe: