Focused Ultrasound Lesions Achieved in Three of Four Epilepsy Patients

Definite lesion formation was confirmed in three of four drug-resistant epilepsy patients treated with magnetic resonance-guided focused ultrasound (MRgFUS) in a Taiwan FDA-regulated phase 1 trial, with no serious permanent neurological adverse events during follow-up. The interim report, published in Epilepsy Research and registered at ClinicalTrials.gov as NCT06292494, includes the first reported applications of MRgFUS to the posterior corpus callosum (CC) and the centromedian nucleus (CM) for epilepsy.

What did the trial involve?

The prospective, open-label interventional study was conducted at Taipei Veterans General Hospital under institutional review board approval (No. 2022-08-012B). Fourteen patients underwent screening; 10 (71.4%) were excluded because their skull density ratio (SDR, a measure of how efficiently ultrasound passes through the skull) fell below 0.30. The four treated patients had a mean age of 34.7 years and met all inclusion criteria.

Treatment targets differed across patients: the anterior nucleus of the thalamus (ANT), a hypothalamic hamartoma (HH, a benign congenital lesion in the hypothalamus), the centromedian nucleus, and the splenium of the corpus callosum. The authors report clinical outcomes at 3 to 12 months after treatment. Each patient reached the decision to proceed only after a thorough discussion of alternatives, including completion of corpus callosotomy or deep brain stimulation.

In the ANT case, the procedure was discontinued before therapeutic temperatures were reached because the patient could not tolerate treatment under local anesthesia, partly attributed to baseline factors described by the investigators. Treatment feasibility depended strongly on patient selection, and skull characteristics were a determining factor, the authors concluded.

Why an incisionless option matters

Epilepsy affects nearly 1% of the world's population, and roughly 13–37% of patients develop drug-resistant epilepsy (DRE), defined as failure to achieve sustained seizure freedom after adequate trials of two tolerated, appropriately chosen antiseizure medications. Patients with DRE are candidates for surgical or neuromodulatory treatment, yet referral is often delayed for years after epilepsy onset because of concerns about craniotomy, infection, and device implantation.

The evidence base the authors cite frames the bar MRgFUS must clear:

• Resective surgery: seizure freedom rates of approximately 70% for focal cortical dysplasia and 86.5% for cavernous malformations.
• SEEG-guided radiofrequency ablation and MR-guided laser interstitial thermal therapy (MRgLITT): positive outcomes in roughly 60% of cases.
• Deep brain stimulation (DBS): a mean 64.8% seizure-frequency reduction in meta-analysis.
• Vagus nerve stimulation (VNS): a mean 48.3% reduction in the same meta-analysis.
• Corpus callosotomy: significant reduction of severe seizures such as drop attacks, though total seizure freedom is less common.

MRgFUS creates intracranial lesions with considerable precision in centrally located deep structures relevant to epilepsy, including the thalamus, hypothalamic hamartomas, and hippocampus, without incision. Its safety and efficacy are well established for movement disorders. Prior studies have reported seizure freedom after ablation of the ANT and hypothalamic hamartoma, and seizure reduction in mesial temporal lobe epilepsy despite incomplete hippocampal lesioning. The Taiwan trial extends this evaluation to a broader range of targets and, notably, to an Asian cohort.

What are the limitations?

The authors are explicit about the constraints of the interim data. The study included only four patients, each with a different treatment target, and no sham control. Epilepsy surgery typically requires at least two years of follow-up to assess efficacy; this series offers 3 to 15 months. The heterogeneous sample and short follow-up mean the observed changes in seizure frequency should be read as preliminary clinical observations rather than evidence of sustained efficacy. Postoperative cognitive outcomes also remain to be characterized.

What does this change for the reading room and epilepsy monitoring unit? For now, nothing in routine practice — MRgFUS for epilepsy remains investigational work, not an FDA-cleared indication in the sense that movement-disorder applications are established. But electroencephalographers and surgical committees should note the modality's trajectory: the SDR threshold of 0.30 excluded seven of ten screened candidates, meaning skull imaging will gate eligibility at any center adopting the technique, much as it already does for essential tremor programs.

The study was funded by Taipei Veterans General Hospital (grants V113C-080, V114C-031, V115C-108), which supported personnel, clinical database management, and focused ultrasound supplies. The authors declare no competing interests. Longer follow-up of this cohort, and enrollment beyond the first four patients, will determine whether incisionless thalamic and callosal lesioning can hold a durable place alongside resection, MRgLITT, and neuromodulation in the drug-resistant epilepsy pathway.