One-Sided Brain Stimulation Calms Drug-Resistant Epilepsy in Landmark Case

For roughly one in three people with epilepsy, medications simply do not stop the seizures. When drugs fail, neurosurgeons can turn to deep brain stimulation, implanting electrodes that deliver rhythmic electrical pulses to quiet overactive brain circuits. The most common target is the anterior nucleus of the thalamus, a small relay station deep in the brain that helps orchestrate seizure traffic between the cortex and limbic system. Since the landmark SANTÉ trial, this approach has been performed almost universally on both sides of the brain, and long-term data show seizure reductions of around 75 percent after a decade of stimulation. Now, a surgical team in Japan has reported something unusual: a patient whose disabling seizures were substantially controlled with stimulation of the thalamic nucleus on one side only, a result that challenges the assumption that bilateral electrodes are always necessary.

The case, published in Frontiers in Neuroscience by neurosurgeons at Nara Medical University and collaborators at the Montreal Neurological Institute, describes a 29-year-old woman whose epilepsy had resisted every conventional therapy. Her medical history was extraordinarily complex. At age seven she was diagnosed with a suprasellar germ cell tumor, a rare growth near the pituitary region, and underwent craniotomy for resection, high-dose chemotherapy, whole-brain radiation totaling 50 Gy with a local boost, and stereotactic radiation to the anterior spinal cord at 24 Gy. The tumor went into remission, but the aggressive treatment left her with panhypopituitarism and severe intellectual disability, with a full-scale intelligence quotient of 42 measured at age 16. A ventriculoperitoneal shunt was later inserted to manage secondary hydrocephalus, a buildup of cerebrospinal fluid, but the device caused intracranial hemorrhage from overshunting and had to be ligated.

Seizures emerged at age 22 after an episode of nonconvulsive status epilepticus triggered by severe hyponatremia, a sodium level of 123 mmol/L against a reference range of 135 to 145. Magnetic resonance imaging revealed abnormal lesions in both temporal lobes, and recurrent seizures followed. Over the next seven years she tried levetiracetam, carbamazepine, perampanel, lacosamide, valproic acid, topiramate, and brivaracetam, with either no benefit or intolerable side effects. Vagus nerve stimulation, implanted at age 23, reduced dramatic falls from several per week to one or two, but daily seizures persisted at a rate of roughly ten per day. By the time she reached the Nara epilepsy clinic at 29, her team faced a patient whose brain had been reshaped by tumor, surgery, radiation, and shunt hardware, and whose seizures continued unabated.

The diagnostic workup was exhaustive. Neurological examination showed right-sided hemiparesis with manual muscle strength of grade 3, no meaningful speech, and an inability to follow verbal commands, which ruled out formal neuropsychological testing. Her seizures followed a distinctive pattern: asymmetric tonic posturing lasting 10 to 30 seconds, with eyes deviating to the left, the left arm extending, and the right arm flexing, followed by a flaccid atonic phase in which all four limbs went limp and the trunk tilted leftward. Computed tomography revealed brain atrophy, hydrocephalus, calcifications in both basal ganglia, and post-craniotomy changes in the right frontal region. Magnetic resonance imaging showed marked atrophy of the left hemisphere, so severe that the left anterior thalamic nucleus could not be identified at all.

Video-electroencephalography over three days, using scalp electrodes placed according to the International 10-20 system, painted a detailed electrical portrait. Between seizures, the awake recording showed a disorganized posterior dominant rhythm of 6 to 7 hertz, frequent slow and sharp waves over the F4 and C4 electrode positions on the right, and spikes with high-amplitude fast activity at sleep onset. Sleep architecture was disrupted, without obvious slow-wave sleep, and intermixed abnormal theta and delta activity was maximal over the left frontal region, with independent interictal epileptiform discharges also appearing in the right frontotemporal region. Crucially, the team captured around ten habitual seizures, and ictal recordings showed high-amplitude bursts of spiky fast waves originating at F4 and spreading across the right hemisphere. Frequent subclinical electrographic seizures occurred during sleep. The diagnosis: focal epilepsy arising from a widespread epileptogenic network centered on the right frontal and temporal lobes.

That network localization shaped the surgical decision. Because the epileptogenic zone was distributed rather than confined to a resectable focus, focal removal was not indicated; resective surgery can render roughly two-thirds of medial temporal lobe epilepsy patients seizure-free and about half of those with neocortical epilepsy, but only when the abnormal tissue can be cleanly delineated. Deep brain stimulation of the anterior thalamic nucleus was chosen instead. The original plan was conventional bilateral implantation, but anatomy intervened. The left nucleus could not be visualized on imaging, even when compared against ultrahigh-resolution 200-micrometer sections from the BigBrain human atlas. Severe left hemispheric atrophy had widened the subdural space excessively, and a shunt valve sat at the anticipated burr-hole site beneath thinned scalp. With the epileptiform abnormalities predominantly right-sided, and with prior animal studies and human responsive neurostimulation cases suggesting unilateral thalamic stimulation could work, the team implanted a single electrode in the right anterior nucleus alone.

Stimulation began seven days after surgery at an intensity of 2 milliamperes, and the results were striking. Seizure frequency plummeted from roughly ten per day to about one, an early drop the authors attribute largely to a microlesioning effect, the temporary suppression that often follows electrode implantation itself rather than the stimulation. Once that effect settled, sustained stimulation produced a stable seizure reduction of approximately 60 percent, maintained throughout nine months of follow-up. Antiseizure medications were left unchanged, isolating the stimulation as the variable. Seizure duration also shortened by about half, reducing interruptions to daily life, and the patient showed improved daytime alertness, possibly reflecting better nighttime sleep quality, which in turn eased the burden on her caregivers. Final stimulation parameters were set at 2 mA, 145 pulses per second, 90-microsecond pulse width, and a duty cycle of one minute on and five minutes off, while vagus nerve stimulation was reduced to 1 mA. Routine postoperative electroencephalography showed a measurable reduction in interictal epileptiform discharges, and no surgery- or stimulation-related adverse events occurred.

The result fits a broader pattern in neuromodulation. In movement disorders such as Parkinson’s disease, unilateral deep brain stimulation of the subthalamic nucleus or internal globus pallidus improves contralateral symptoms by up to 75 percent but ipsilateral symptoms by only about 28 percent, and a recent systematic review found unilateral stimulation improved Unified Parkinson’s Disease Rating Scale Part III scores by 37 percent compared with up to 66 percent for bilateral approaches. The principle is that one electrode primarily treats the hemisphere in which it sits. Epilepsy complicates this logic, because seizures can propagate through the thalamus to widespread networks on both sides, which is precisely why bilateral anterior thalamic stimulation became standard. Yet animal work supports the one-sided approach in the right circumstances: in rats with unilateral amygdala kindling, high-frequency stimulation of the ipsilateral anterior nucleus significantly reduced seizure frequency and duration, while contralateral stimulation did not, although in a pilocarpine model of secondarily generalized seizures only bilateral stimulation or lesioning suppressed seizures.

Human case reports of responsive neurostimulation, a closed-loop device that fires only when it detects seizure activity, point the same way. One patient with refractory genetic generalized epilepsy achieved a 90 to 95 percent seizure reduction over two years with stimulation of the right anterior nucleus and prefrontal cortex, and three patients with multifocal epilepsy treated with unilateral thalamic responsive neurostimulation improved by 50 to 56 percent over 33 to 35 months. Although closed-loop and open-loop stimulation differ mechanistically, both experiences underscore the importance of targeting the hub of the epileptogenic network. The Nara case strengthens that hypothesis because open-loop stimulation alone, delivered continuously between seizures, achieved meaningful control, and its effects may continue to evolve over longer periods. The authors also note that the anterior nucleus and medial temporal structures are both components of the limbic system, and that stronger functional connectivity between the thalamus and the seizure onset zone has been linked to better stimulation outcomes, which may explain why unilateral stimulation worked when the network was predominantly ipsilateral.

The team is careful about the limits of a single case. The follow-up was only nine months, so durability remains unknown, and the pre-existing vagus nerve stimulator complicates attribution, since a synergistic effect between the two devices cannot be excluded. Still, the clinical message is concrete: when the predominant epileptogenic network is confined to one hemisphere, and especially when anatomical constraints such as atrophy, shunt hardware, or distorted anatomy prevent placing electrodes on both sides, unilateral anterior thalamic stimulation is a reasonable and potentially worthwhile option. The authors caution that bilateral stimulation remains necessary when seizures arise from or spread through both hemispheres, as in post-encephalitic epilepsy, and that larger cohorts with longer follow-up are needed to define the indications, efficacy, and durability of the unilateral approach. For now, the case offers a carefully documented proof of principle that in epilepsy surgery, as in so much of neuroscience, matching the intervention to the individual patient’s network may matter more than following the standard template.

Subject of Research: Unilateral deep brain stimulation of the anterior thalamic nucleus for drug-resistant focal epilepsy

Article Title: Case Report: Successful treatment of drug-resistant epilepsy with unilateral deep brain stimulation of the anterior nucleus of the thalamus

Article References: Sasaki, R., Yamada, K., Tamura, K., Kinoshita, M., Park, Y.-S., Sadikot, A. F., & Nakagawa, I. (2026). Case Report: Successful treatment of drug-resistant epilepsy with unilateral deep brain stimulation of the anterior nucleus of the thalamus. Frontiers in Neuroscience, 20, Article 1978183. https://doi.org/10.3389/fnins.2026.1978183

Image Credits: AI Generated

DOI: 10.3389/fnins.2026.1978183

Keywords: deep brain stimulation, drug-resistant epilepsy, anterior nucleus of the thalamus, neuromodulation, focal epilepsy, case report, electroencephalography, vagus nerve stimulation, thalamus, epilepsy surgery, responsive neurostimulation, neuroscience