NAUTILUS
Responsive stimulation of the thalamus for idiopathic generalized epilepsy: Results of the randomized controlled NAUTILUS trial through 18 months
Clinical Question
Does responsive closed-loop stimulation of the centromedian thalamus safely and effectively reduce generalized tonic-clonic seizures in patients with drug-resistant idiopathic generalized epilepsy?
Bottom Line
Responsive thalamic (centromedian nucleus) stimulation with the RNS System missed its prespecified primary effectiveness endpoint but demonstrated an acceptable safety profile (SADE 6.9%) and clinically meaningful, durable seizure reductions at 18 months (76.8% median GTCS reduction, 62.5% responder rate, 40% GTCS-free), offering a much-needed neuromodulation option for drug-resistant idiopathic generalized epilepsy.
Major Points
- First randomized controlled neuromodulation trial for drug-resistant IGE
- Primary safety endpoint met: SADE rate 6.9% at 84 days, well below 25% performance goal (p<.0001)
- Prespecified primary effectiveness endpoint (time-to-second-GTCS during EEP) was NOT met
- Post hoc mixed-effects model with monthly seizure counts showed 61% GTCS reduction in Active vs 49% in Sham (p=.030)
- 18-month outcomes: 76.8% median GTCS reduction, 62.5% responder rate, 40% GTCS-free
- 77.8% median reduction in days with any generalized seizure
- >90% of patients and 86% of physicians reported improvement on Global Impression of Change scales
- No adverse effects on cognition, mood, or sleep
- One definite SUDEP event; combined RNS clinical trial dataset (N=693) showed SUDEP rate of 2.9/1000 patient-years, lower than expected rates in surgical candidates (9.3/1000) and ASM placebo arms (6.1/1000)
Design
Study Type: Prospective, multicenter, single-blind, randomized, sham-controlled pivotal trial
Randomization: 1
Blinding: Single-blind (patients blinded to randomization assignment until final 24-month study visit); Sham patients underwent simulated stimulation programming to maintain blinding
Allocation: Randomized to Active (stimulation enabled) or Sham (stimulation disabled) one month postimplant, balanced for study site, baseline seizure frequency, and prior VNS use
Follow-up Duration: 18 months (with 24-month total study duration)
Centers: 23
Countries: United States
Sample Size: 87
Analyzed: 87
Analysis: Intent-to-treat (ITT, N=87) for primary safety and effectiveness; per-protocol (PP, N=82) for additional analyses
Registration: NCT05147571
Inclusion Criteria
- Age ≥12 years
- Confirmed IGE diagnosis by history and EEG
- Drug-resistant epilepsy defined as failure of ≥2 appropriate antiseizure medications (ASMs)
- GTCSs with or without absence or myoclonic seizures
- ≥2 GTCSs during the 3-month baseline period
Exclusion Criteria
- Focal epilepsy or EEG findings suggestive of focal onset
- Nonepileptic seizures that could not be distinguished from epileptic seizures
- Active psychosis, major depression, or suicidal ideation
- Concurrent therapy with another active stimulation device for epilepsy
Arms
| Field | Active | Control |
|---|---|---|
| N | 44 | 43 |
| Intervention | Responsive closed-loop stimulation with RNS System; bilateral depth leads targeted to the centromedian nucleus of the thalamus; stimulation enabled with 2-month optimization period for detection and stimulation parameters | Bilateral depth leads implanted with stimulation disabled; detection parameters adjusted as needed; simulated stimulation programming to maintain blinding; transitioned to open-label active stimulation after second GTCS in EEP or other transition event |
| Duration | Through 18-month follow-up (24-month total study) | Sham until transition event (second GTCS in EEP, 1 year without second GTCS, or 60th patient with second GTCS), then open-label active |
Outcomes
| Outcome | Type | Control | Intervention | HR / OR / RR | P-value |
|---|---|---|---|---|---|
| Time to second GTCS during the 9-month effectiveness evaluation period (EEP) — prespecified analysis (Cox proportional hazards stratified by baseline GTCS frequency and prior VNS); primary safety endpoint was rate of serious adverse device-related events (SADEs) at 84 days postimplant | Primary | Sham group (n=43) | Active group (n=44) | Prespecified primary effectiveness endpoint NOT significant. Primary safety endpoint: SADE rate 6.9% (6/87), 95% CI 2.9%-14.5%, well below 25% performance goal | Safety: p<.0001; Effectiveness primary endpoint: not significant (NS) |
| Post hoc mixed-effects model (negative binomial) with monthly seizure counts from randomization through 12 months, difference-of-differences approach | Secondary | .030 | |||
| Median percent reduction in GTCSs | Secondary | Time Point: 18 months · 76.8% | |||
| 50% responder rate (GTCS) | Secondary | Time Point: 18 months · 62.5% | |||
| GTCS-free rate | Secondary | Time Point: 18 months · 40% | |||
| Median reduction in days with any generalized seizure | Secondary | Time Point: 18 months · 77.8% | |||
| Patient Global Impression of Change — reported improvement | Secondary | Time Point: 18 months · >90% of patients reported improvement | |||
| Physician Global Impression of Change — reported improvement | Secondary | Time Point: 18 months · 86% of physicians reported improvement | |||
| Primary safety endpoint: SADE rate at 84 days postimplant | Safety | 6.9% (6/87 patients), 95% CI 2.9%-14.5%, p<.0001 vs 25% performance goal | |||
| Total SADEs through 18 months | Safety | 9 SADEs across 7 patients | |||
| Neurocognitive, mood, and sleep outcomes (NIH Toolbox Cognition Battery, BDI-II, Pittsburgh Sleep Quality Index) | Safety | No adverse effects on cognition, mood, or sleep observed through 18 months | |||
| SUDEP rate (combined across all prospective RNS System clinical trials) | Safety | 1 definite SUDEP in NAUTILUS; combined N=693 patients, 3502.0 implant years → 2.9/1000 patient-years (95% CI 1.5-5.3), lower than surgical candidates (9.3/1000) and ASM placebo arms (6.1/1000) | |||
| Deaths through 18 months | Safety | 2 deaths — 1 definite SUDEP (stimulation enabled) and 1 from a plane crash | |||
| SADEs (18 months) | Adverse | 9 events in 7 patients; specific events noted include implant site erosion and headache | |||
| SADE rate at 84 days | Adverse | 6.9% (6/87) | |||
| Cognition | Adverse | No adverse effects | |||
| Mood | Adverse | No adverse effects | |||
| Sleep | Adverse | No adverse effects | |||
Subgroup Analysis
Randomization was balanced for study site, baseline GTCS frequency (≤2 vs >2 per month), and prior VNS use; primary effectiveness Cox model was stratified by these factors. Exploratory analyses included ASM changes (Kruskal-Wallis), rescue benzodiazepine use (GEE binomial models), seizure duration and postictal recovery (cumulative link mixed model on 1679 seizures from 80 subjects), and injury event rates (Poisson regression with within-subject pairwise design).
Criticisms
- Prespecified primary effectiveness endpoint (time-to-second-GTCS) was not met; positive effectiveness conclusions rely on a post hoc mixed-effects model
- Single-blind design (patients blinded but not investigators)
- Sham patients transitioned to open-label active stimulation after a second GTCS, limiting long-term comparison between groups
- Single-country (US-only) trial with predominantly White (80.5%) and non-Hispanic (85.1%) population may limit generalizability
- Reliance on patient-reported electronic seizure diaries for outcome ascertainment
- Industry-sponsored trial (NeuroPace) with multiple authors affiliated with the sponsor
Funding
NeuroPace, Inc.
Based on: NAUTILUS (Epilepsia, 2026)
Authors: Uysal U, Landazuri P, Burdette DE, ..., Morrell MJ; Nautilus Study Group
Citation: Uysal U, et al. Responsive stimulation of the thalamus for idiopathic generalized epilepsy: Results of the randomized controlled NAUTILUS trial through 18 months. Epilepsia. 2026;00:1-14. doi:10.1002/epi.70321
Content summarized and formatted by NeuroTrials.ai.