Non-invasive neuromodulation for drug-resistant epilepsy: a contemporary structured review of mechanisms, clinical evidence and future directions.

Drug-resistant epilepsy (DRE) affects nearly one-third of people with epilepsy and remains a major cause of disability despite continued advances in antiseizure medications and surgical therapies. For patients who are not candidates for resective surgery or invasive neuromodulation, non-invasive neuromodulation (NIN) has emerged as a promising strategy to modulate epileptic networks while avoiding the risks associated with implanted devices. This structured narrative review examines current evidence across the principal NIN modalities, including repetitive transcranial magnetic stimulation (rTMS), transcranial electrical stimulation (transcranial direct current (tDCS), alternating current, and temporal interference stimulation), transcutaneous vagus nerve stimulation, trigeminal nerve stimulation, and transcranial focused ultrasound. We integrate current knowledge on mechanisms of action, clinical efficacy, safety, and practical considerations while highlighting recent advances that may shape future clinical implementation. Available evidence indicates that low-frequency rTMS and cathodal tDCS provide the most consistent reductions in seizure frequency and interictal epileptiform discharges, particularly in focal DRE. Transcutaneous vagus nerve stimulation has shown encouraging results with an excellent safety profile, whereas focused ultrasound, trigeminal nerve stimulation, and temporal interference stimulation remain investigational and require further clinical validation. Across modalities, interpretation of current evidence is limited by heterogeneous stimulation protocols, small study populations, and variable outcome measures. Emerging approaches, including individualized targeting, connectome-guided stimulation, computational modeling, neurophysiological biomarkers, and adaptive closed-loop neuromodulation, are expected to improve treatment precision and patient selection. Although high-quality multicenter randomized trials remain necessary, current evidence supports NIN as an increasingly relevant component of personalized epilepsy care and a promising avenue for expanding therapeutic options in DRE.