Drug-Resistant Epilepsy: Options When Medications Fail

Research-informed explainer — last updated April 12, 2026

About one in three people with epilepsy never achieve seizure control with medications alone. Trying two seizure drugs and failing both is the formal threshold for "drug-resistant epilepsy" — a point at which continuing to add medications has sharply diminishing returns. The good news is that surgery, implanted devices, dietary therapy, and several newer medications offer meaningful options for many people at that threshold, and the research on when to act is clearer than most patients realize.

This explainer draws on published research from four epilepsy specialists in the Convene directory. Jacqueline French, M.D., at NYU Langone, led the ILAE consensus definition of drug-resistant epilepsy and the epilepsy classification system that guides treatment decisions [1][2]. Robert Fisher, M.D., at Stanford, published the operational classification of seizure types that determines which patients are candidates for surgical and device-based treatment [3][4]. Gregory Cascino, M.D., at Mayo Clinic, published the landmark JAMA review of resective surgery outcomes demonstrating 60 to 80 percent seizure freedom in appropriately selected patients, and his MRI volumetry work helped identify which patients are surgical candidates [5][6]. Elizabeth Thiele, M.D., Ph.D., at Massachusetts General Hospital, published foundational studies on tuberous sclerosis complex and epilepsy, including genetic severity data for TSC1 and TSC2 mutations [7][8].

What drug-resistant epilepsy means

The formal ILAE definition, established in 2009, states that drug-resistant epilepsy is the failure of two adequately chosen, tolerated, and correctly dosed antiseizure medications to achieve sustained seizure freedom [1]. This threshold matters because data consistently shows that after two drug failures, the likelihood of achieving seizure freedom with a third or fourth medication drops below 10 percent. Continuing to try drug after drug while avoiding other options wastes time and allows ongoing neurological damage.

Drug resistance does not mean that medications have no role — many patients require ongoing drug treatment even after surgery or device implantation. But it does mean that medication alone is unlikely to be the path to freedom from seizures.

How your seizure type affects your options

The classification of seizure type — focal (starting in one region of the brain) versus generalized (involving both hemispheres from the onset) — directly determines which non-drug options are viable [3]. Patients with focal epilepsy whose seizures start in an identifiable brain region are the best surgical candidates. Patients with generalized or multifocal epilepsy, or those whose seizure focus cannot be clearly located, generally rely on devices or dietary therapy instead.

Several epilepsy syndromes have specific targeted treatments. Tuberous sclerosis complex (TSC), for example, is caused by mutations in TSC1 or TSC2 genes and causes brain tubers that trigger severe seizures from infancy [7]. Everolimus, an mTOR inhibitor, is FDA-approved specifically for TSC-related seizures — a case where knowing the genetic cause leads directly to a targeted therapy [8].

Surgical options

Resective surgery — removing or disconnecting the brain region responsible for seizures — is the most effective long-term treatment for drug-resistant focal epilepsy. A 2015 JAMA review found that across randomized clinical trials, epilepsy surgery reduced seizure activity significantly compared with continued medication therapy, with 60 to 80 percent of appropriately selected patients achieving seizure freedom [5].

The most common resective procedure is anterior temporal lobectomy for temporal lobe epilepsy, where detailed pre-surgical evaluation including MRI, EEG video monitoring, and neuropsychological testing determines whether surgery is safe and likely to work. For non-temporal focal epilepsy, outcomes vary more widely depending on the size and location of the seizure focus.

MRI-based hippocampal volume measurements help predict surgical outcomes — patients with mesial temporal sclerosis (scarring of the hippocampus visible on MRI) have particularly good response rates to anterior temporal lobectomy [6].

Laser interstitial thermal therapy (LITT), a minimally invasive approach using MRI-guided laser probes, now offers an option for patients who are not candidates for open resection or who prefer a less invasive procedure.

Device-based options

Vagus nerve stimulation (VNS): An implanted device that delivers electrical pulses to the left vagus nerve in the neck. VNS reduces seizure frequency by roughly 50 percent in about half of patients. It does not typically eliminate seizures completely, but many patients report shorter, less severe seizures and a useful hand-held magnet that can abort or lessen an impending seizure.

Responsive neurostimulation (RNS): An implanted device that continuously monitors brain activity and delivers electrical stimulation directly to the seizure focus when it detects abnormal activity. RNS is particularly useful for patients with seizure foci in eloquent brain regions (areas controlling language or movement) where resection would cause unacceptable deficits. Studies show a roughly 70 percent reduction in seizure frequency after several years of use.

Deep brain stimulation (DBS) of the anterior thalamus: FDA-approved for drug-resistant focal epilepsy, thalamic DBS modulates the brain's seizure-generating networks. Like VNS and RNS, it reduces rather than eliminates seizures in most patients.

Dietary therapy

The ketogenic diet — a medically supervised, high-fat, low-carbohydrate regimen — has been used for epilepsy since the 1920s and remains a valid option, particularly for children with drug-resistant epilepsy. Roughly 50 percent of patients on the ketogenic diet achieve a 50 percent reduction in seizures. Modified versions, including the modified Atkins diet and low glycemic index treatment, are better tolerated by older children and adults.

Newer medications for specific syndromes

Cannabidiol (Epidiolex), derived from cannabis, is FDA-approved for seizures in Dravet syndrome and Lennox-Gastaut syndrome — two severe drug-resistant epilepsy syndromes of childhood. Trials showed significant seizure reduction compared with placebo, giving families whose children had exhausted most options a new evidence-based choice.

Fenfluramine (Fintepla) is also approved for Dravet syndrome after Phase 3 trials showed it reduced convulsive seizures by more than 60 percent versus placebo. Sodium channel blockers like lacosamide and newer agents including cenobamate (which was FDA-approved in 2019 and showed complete seizure freedom in 21 percent of patients in its Phase 2b trial) expand the medication toolkit for focal epilepsy.

The referral gap

Despite good evidence that surgery can dramatically help the right patients, referrals to epilepsy surgery centers remain far too infrequent. One reason is that many patients and their primary neurologists assume drug resistance is permanent and continue cycling through medications. Another is fear of brain surgery. The data from randomized trials and long-term observational studies suggests that the risk of surgery for well-selected patients is substantially lower than the ongoing risk of uncontrolled seizures — including injury, sudden unexpected death in epilepsy (SUDEP), and progressive neurological decline.

Questions to ask your doctor

• Have I failed two appropriately chosen seizure medications? Does that mean I should be referred to a comprehensive epilepsy center?
• Has anyone evaluated whether my seizures come from a specific location in my brain, and whether that could be removed or treated?
• Am I a candidate for resective surgery, LITT, RNS, VNS, or DBS — and how would I find out?
• Is there a genetic cause for my epilepsy that could open the door to a targeted treatment?
• Is the ketogenic diet or a modified version a reasonable option for me?
• Are there newer medications, such as cenobamate or cannabidiol, that might work when others have not?

The bottom line

Drug-resistant epilepsy does not mean untreatable epilepsy. The formal definition — failing two medications — is the starting point for a broader evaluation, not a dead end. Surgery frees 60 to 80 percent of carefully selected patients from seizures entirely. Implanted devices substantially reduce seizures for many who are not surgical candidates. Dietary therapy and newer targeted medications expand the toolkit further. If you or someone you care for has failed two seizure drugs, the most important next step is an evaluation at a comprehensive epilepsy center with the full range of diagnostic and treatment options.