Residency · Residency · Neurology
Neuromodulation for Drug-Resistant Epilepsy
Overview
Neuromodulation therapies offer palliative seizure reduction for patients with drug-resistant epilepsy who are not candidates for resective surgery. Three FDA-approved devices are currently available: vagus nerve stimulation (VNS), responsive neurostimulation (RNS), and deep brain stimulation (DBS) of the anterior nucleus of the thalamus. These devices reduce seizure frequency by 40-70% over time, though seizure freedom rates remain modest at 5-15%. Patient selection depends on the type of epilepsy, whether seizures can be localized, and individual treatment goals.
Vagus Nerve Stimulation (VNS)
Mechanism
VNS delivers intermittent electrical stimulation to the left vagus nerve in the neck. Afferent vagal fibers project to the nucleus tractus solitarius, then to the locus coeruleus, and from there to widespread cortical projections. This modulates noradrenergic and serotonergic systems and alters thalamocortical synchrony. The exact antiseizure mechanism remains incompletely understood.
Device and Procedure
A pulse generator is implanted in the left chest wall, similar to a cardiac pacemaker. A lead is wrapped around the left vagus nerve in the neck (the left side is chosen to avoid cardiac effects). The procedure is outpatient and takes approximately one hour. Programmable parameters include output current (0.25-3.5 mA), frequency (20-30 Hz), pulse width, and the on-time/off-time duty cycle. Newer models incorporate automatic stimulation triggered by heart rate changes through ictal tachycardia detection. A patient-activated magnet allows the patient or caregiver to trigger additional stimulation at seizure onset.
Evidence
The pivotal E03 and E05 trials demonstrated a 25-30% responder rate (defined as 50% or greater seizure reduction) at 3 months. Importantly, efficacy improves over time, reaching approximately 50% responder rate at 2-3 years, suggesting neuroplastic mechanisms contribute to the benefit. Seizure freedom is rare, occurring in only about 5-8% of patients. VNS also has emerging evidence for mood improvement and holds separate FDA approval for treatment-resistant depression.
Side Effects
Hoarseness and voice changes during stimulation are the most common side effects and are usually tolerable. Cough, throat discomfort, and dyspnea during stimulation also occur. VNS is generally well-tolerated and notably has no cognitive side effects, which represents a major advantage. Device infection or lead malfunction is uncommon. Obstructive sleep apnea may worsen and should be monitored.
Patient Selection
VNS has a broad indication for drug-resistant focal or generalized epilepsy in patients aged 4 years and older. No seizure localization is required, which is a significant advantage over RNS and resective surgery. VNS is particularly useful when surgery is not feasible due to multifocal, generalized, or non-localizable epilepsy. It can be used alongside ASMs without requiring invasive monitoring.
Responsive Neurostimulation (RNS System)
Mechanism
RNS is a closed-loop (responsive) system that continuously records electrocorticographic (ECoG) activity from the seizure focus. When abnormal patterns such as seizure onset or interictal epileptiform discharges are detected, the device delivers brief electrical stimulation to the focus, designed to abort seizures at their onset rather than prevent them. A unique advantage is that RNS provides chronic ambulatory ECoG data for ongoing seizure monitoring and management optimization.
Device and Procedure
The neurostimulator is implanted in a craniectomy flush with the skull. One or two depth electrodes and/or cortical strip electrodes are placed at the identified seizure onset zone(s), allowing stimulation of up to two independent foci. Presurgical evaluation to identify the seizure onset zone (from Phase I or Phase II data) is required. Detection algorithms are programmed and refined by the epileptologist using cloud-based data review.
Evidence
The pivotal RCT by Morrell and colleagues (2011) demonstrated significant seizure reduction versus sham stimulation at 3 months. Long-term data show a median seizure reduction of 50-60% at 2 years and approximately 70% at 9 years. Seizure freedom is achieved in approximately 15% at long-term follow-up. Like VNS, efficacy improves progressively over years, suggesting neuroplastic remodeling. The device is effective for both mesial temporal and neocortical onset seizures.
Advantages
As a responsive closed-loop system, RNS stimulates only when needed. The chronic ECoG recording provides valuable long-term data on seizure patterns, medication effects, and circadian trends. It can target up to two independent seizure foci. Because stimulation is focal and brief, there are no systemic side effects.
Limitations
RNS requires precise localization of the seizure onset zone, making it unsuitable for non-localizable or diffuse epilepsy. It is an intracranial procedure with higher surgical risk than VNS. Battery replacement is needed every 3-4 years, though a rechargeable generator is available. Cost and access remain barriers.
Deep Brain Stimulation (DBS) of the Anterior Nucleus of the Thalamus
Mechanism
Bilateral electrodes are placed in the anterior nucleus of the thalamus (ANT), which serves as a relay in the circuit of Papez (mammillary bodies to ANT to cingulate gyrus to hippocampus). Stimulation modulates thalamocortical networks involved in seizure propagation. The system uses open-loop (scheduled) stimulation, though newer paradigms include responsive modes.
SANTE Trial (2010)
The multicenter randomized controlled SANTE trial evaluated bilateral ANT DBS for drug-resistant focal epilepsy. During the 3-month blinded phase, active stimulation produced a 40% seizure reduction compared to 14.5% with sham, a statistically significant difference. Long-term open-label follow-up at 7 years demonstrated a median 75% seizure reduction, with 16% of patients achieving seizure freedom. Efficacy improved progressively over years. The device was most effective for seizures of temporal lobe origin.
Side Effects
Depression and memory complaints were reported in 5-15% of patients, related to the proximity of the ANT to limbic circuits. Implant site infection or hemorrhage occurred in approximately 5%. Paresthesias during stimulation are usually transient. Formal cognitive testing in most studies did not show significant decline.
Patient Selection
DBS of the ANT is FDA-approved for drug-resistant focal epilepsy in patients aged 18 and older with partial-onset seizures who have failed 3 or more ASMs. It does not require precise localization of the seizure onset zone because it targets a network node rather than the focus itself. This makes it useful for multifocal or bilateral temporal epilepsy where resection is not feasible. Research is also exploring stimulation of the centromedian nucleus for generalized epilepsy in conditions like LGS.
Comparing Neuromodulation Devices
| Feature | VNS | RNS | DBS (ANT) |
|---|---|---|---|
| Stimulation type | Open-loop (scheduled) | Closed-loop (responsive) | Open-loop (scheduled) |
| Target | Left vagus nerve (peripheral) | Seizure focus (1-2 sites, intracranial) | Anterior nucleus of thalamus (bilateral) |
| Seizure localization required | No | Yes (precise) | No |
| Long-term responder rate (≥50% reduction) | ~50% | ~60-70% | ~70-75% |
| Seizure freedom rate | 5-8% | ~15% | ~16% |
| Cognitive side effects | None | None | Possible mood/memory effects (5-15%) |
| Surgical risk | Low (extracranial) | Moderate (intracranial) | Moderate (intracranial) |
| Unique advantage | Simplicity; broad eligibility | Chronic ECoG data; responsive stimulation | Network modulation without focus localization |
| FDA-approved age | ≥4 years | ≥18 years | ≥18 years |
VNS is an open-loop peripheral nerve stimulator targeting the left vagus nerve. It requires no seizure localization, achieves about a 50% long-term responder rate with 5-8% seizure freedom, has no cognitive side effects, and carries low surgical risk as an extracranial procedure. Its unique advantage is simplicity and broad eligibility.
RNS is a closed-loop intracranial system targeting the seizure focus at one or two sites. It requires precise localization, achieves a 60-70% long-term responder rate with approximately 15% seizure freedom, has no cognitive side effects, and carries moderate surgical risk. Its unique advantages are chronic ECoG data and responsive stimulation.
DBS of the ANT is an open-loop intracranial system targeting the anterior nucleus of the thalamus bilaterally. It does not require precise focus localization, achieves a 70-75% long-term responder rate with approximately 16% seizure freedom, may cause mood or memory effects, and carries moderate surgical risk. Its unique advantage is network modulation without requiring focus localization.
Emerging Neuromodulation Approaches
Chronic subthreshold cortical stimulation delivers continuous low-level stimulation below the seizure threshold. Transcranial magnetic stimulation (TMS) is non-invasive but has limited evidence for epilepsy. Transcranial direct current stimulation (tDCS) remains investigational. Focused ultrasound offers non-invasive neuromodulation or ablation and is in early-phase trials. Centromedian nucleus DBS for generalized epilepsy, particularly Lennox-Gastaut syndrome, shows promising open-label data.
<image>A side-by-side comparison illustration of the three FDA-approved neuromodulation devices for epilepsy. Panel 1 (VNS): shows the left cervical vagus nerve with a stimulating lead connected to a pulse generator in the chest wall, with a pathway diagram showing vagal afferents to NTS, locus coeruleus, and cortical projections. Panel 2 (RNS): shows a skull-mounted neurostimulator connected to depth and strip electrodes at the seizure focus, with a schematic of the closed-loop detection and stimulation cycle. Panel 3 (DBS): shows bilateral depth electrodes in the anterior nucleus of the thalamus connected to an abdominal/chest pulse generator, with the circuit of Papez highlighted (ANT -> cingulate -> hippocampus -> mammillary bodies). A comparison table summarizing key differences is shown below.</image>
<image>A graph showing the long-term efficacy trajectories of VNS, RNS, and DBS for drug-resistant epilepsy. The x-axis shows time from implantation (0 to 9 years). The y-axis shows median seizure frequency reduction (%). All three devices show a progressive improvement curve, with initial modest reductions at 3-6 months that increase to 50-75% reduction by 5-9 years. Seizure freedom percentages at long-term follow-up are annotated for each device. A text box explains the concept of progressive efficacy and hypothesized neuroplastic mechanisms.</image>
Clinical Pearls
All three neuromodulation devices demonstrate progressive improvement over years, so patients should be counseled that the full benefit may not become apparent for 2-3 years. VNS is the simplest and most broadly applicable device, requiring no seizure localization, and should be considered as first-line neuromodulation. RNS is the only device that provides chronic ambulatory ECoG data, which can inform ongoing management decisions; however, it requires precise seizure localization. DBS of the ANT is most effective for temporal lobe onset seizures but does not require precise focus localization. Neuromodulation is palliative, not curative, and realistic expectations must be set: seizure freedom is achieved in only 5-16% of patients. VNS has no cognitive side effects, making it particularly attractive in patients with cognitive concerns. Neuromodulation should be considered when resective surgery is not feasible, when the patient declines surgery, or as an adjunct to partial resection.
References
- Morris GL, et al. Evidence-based guideline update: vagus nerve stimulation for the treatment of epilepsy. Neurology. 2013;81(16):1453-1459.
- Morrell MJ, et al. Responsive cortical stimulation for the treatment of medically intractable partial epilepsy. Neurology. 2011;77(13):1295-1304.
- Fisher R, et al. Electrical stimulation of the anterior nucleus of thalamus for treatment of refractory epilepsy (SANTE). Epilepsia. 2010;51(5):899-908.
- Salanova V, et al. Long-term efficacy and safety of thalamic stimulation for drug-resistant partial epilepsy. Neurology. 2015;84(10):1017-1025.
- Nair DR, et al. Nine-year prospective efficacy and safety of brain-responsive neurostimulation for focal epilepsy. Neurology. 2020;95(9):e1244-e1256.
