If you or someone you love has epilepsy, you have probably heard the reassuring version: most people get their seizures under control with medicine. That part is true, about two thirds of people do, often completely. This article is about the other third. For them, the first drug did not work, the second did not either, and once that happens, doctors call it drug resistant epilepsy. That group has barely gotten smaller in decades, even though the number of approved medicines has passed 25.
About 3.4 million Americans live with active epilepsy. This article covers what treatment looks like today, what researchers are testing for the third of people who still have seizures, why progress here has been slow, and where clinical trials fit in. If you are looking on behalf of a child, we also have a guide to new epilepsy treatments for children.
Looking for treatment options for epilepsy?
Browse open epilepsy clinical trials near you! We'll handle the follow-up.
What epilepsy treatment includes today
Seizure medicines, the first thing tried. Doctors call these antiseizure medications, or ASMs. There are more than 25 of them, and they work in different ways: some calm down overactive signals in nerve cells, some boost a natural calming chemical in the brain called GABA, and some act on the machinery cells use to send signals to each other. Which one a doctor picks depends on the type of seizure, age, other health conditions, and for women, pregnancy plans. Most people start on one drug and get good control. The newest medicine for focal seizures in adults, cenobamate, was approved in 2019 for focal seizures and helped up to about a quarter of people become seizure free even after other drugs had failed them, a notably higher rate than older add on drugs typically achieve.
Surgery. When seizures start in one spot in the brain that can be safely removed, surgery can stop them completely. For temporal lobe epilepsy, the most common type treated this way, roughly two out of three people become seizure free afterward. Surgery is used less than it should be, many people who could benefit are never even sent to an epilepsy center to be evaluated for it.
Implanted devices. For people who are not good candidates for surgery, three devices are approved: vagus nerve stimulation (a pacemaker like device placed in the chest), responsive neurostimulation (an implant that detects unusual brain activity and responds to it), and deep brain stimulation. These rarely stop seizures completely, but they often cut them down a lot, and they tend to work even better over time.
Diet as treatment. The ketogenic diet, very high fat and very low carbohydrate, is over a hundred years old and still one of the most effective treatments for certain childhood epilepsies. It is hard to stick to, which is why researchers keep testing easier versions of it.
What researchers are studying now
A new kind of pill, and the biggest trial result of 2026. Azetukalner, made by Xenon Pharmaceuticals, works on a channel in brain cells called KV7 that no approved seizure drug touches yet. In a major trial presented in April 2026, 380 adults with severe focal epilepsy, who had already tried a median of five medicines, took it alongside their current drugs. Their monthly seizures dropped 53 percent on the higher dose, compared to just 10 percent on placebo, and over half cut their seizures by at least half. In a long term extension of an earlier trial, among people who stayed on the drug, about 38 percent went a full year without a seizure. The drug does not need a slow buildup in dose, and it stays in the body long enough that missing a pill matters less. The company plans to file with the FDA in the third quarter of 2026, so it is not approved yet, it is only available through trials right now.
Gene therapy: fixing the source, just once. For temporal lobe epilepsy that starts on one side of the brain, a gene therapy study called AMT-260 delivers a one time treatment straight to where the seizures start, aiming to calm those overactive cells for good. It is still early, and results are still coming in, but it is the first real attempt to fix the actual source of seizures instead of calming the whole brain.
Cell therapy: adding cells that calm the brain down. The NRTX-1001 study injects lab grown cells, whose job is to quiet nearby activity, directly into the part of the brain where seizures start. Like the gene therapy above, this is for drug resistant temporal lobe epilepsy in people who might otherwise need surgery.
Treatments that do not require surgery. A trial of low intensity focused ultrasound is testing whether sound waves aimed at the seizure source from outside the head can reduce seizures, with no implant and no cutting involved. A separate study is working out how to get better results from deep brain stimulation devices that are already implanted.
Trying to predict seizures before they happen. Not knowing when a seizure will hit is often the hardest part for people living with epilepsy. A study combining EEG readings with wearable devices is testing whether seizures can be forecast ahead of time, almost like a weather forecast, so people could plan their day and doctors could time medicine better. If this works, it could change daily life more than most new drugs.
Studying diet the right way. A trial testing the ketogenic diet in children with new onset absence epilepsy, and a study testing whether the diet can prevent spasms in infants with genetic forms of epilepsy, are both asking whether one of the oldest treatments out there can be used earlier and more precisely.
Why epilepsy research takes time
The counting problem. There is no blood test that shows whether seizures are under control. Trials rely on people keeping a seizure diary, and seizures happen unpredictably, sometimes in clusters, so every trial needs weeks of just counting seizures before treatment even starts, then months more of watching afterward. Wearable devices that detect seizures automatically may eventually fix this, which is part of why the forecasting studies matter for the whole field, not just for patients.
The many kinds of epilepsy problem. Epilepsy is not one condition, it is really dozens of them with different causes, everything from a scar left by a head injury to a single gene change in a baby. A drug that helps focal seizures might do nothing at all for absence seizures. Each type needs its own trials, and the rarer genetic types have very few patients around to even enroll.
The add on problem. For safety, new drugs are almost always tested alongside a person's existing medicines, in people who have already tried and failed several other drugs. That is the hardest group to help, so new drugs can look weaker in these trials than they might actually be for people earlier in treatment. It also makes trials slower to fill, since participants have to keep everything else about their treatment steady.
The representation problem. Epilepsy is more common, and more often uncontrolled, in low income communities, and Black Americans have higher rates of epilepsy related death, yet trials tend to happen at well resourced academic centers that do not reach those communities well. Several current studies now name better access and diverse enrollment as goals. Here's why diversity in clinical trials matters so much.
Common myths about epilepsy
"You should put something in the mouth of someone having a seizure."
Never do this. It can break teeth or block the airway. What you should actually do: cushion their head, turn them on their side once the shaking stops, time how long it lasts, and call for help if it goes on for more than five minutes.
"Epilepsy is a mental illness."
It is not, it is a neurological condition, in the same category as migraine or stroke. It does raise the risk of depression and anxiety, which is why good epilepsy care pays attention to mood too, not just seizures.
"All seizures look like convulsions."
Many do not. A seizure can also be a blank stare that lasts a few seconds, a strange smell or feeling, or a brief moment of losing awareness. These quieter types often go unnoticed for years, which delays getting diagnosed and treated.
"If medicine has not worked, nothing will."
This is outdated. Surgery makes many people seizure free, devices can cut seizures down a lot, and a new type of medicine showed strong results in 2026 in exactly the people who had already failed the most other drugs.
Types of epilepsy, and what care usually looks like
Doctors sort seizures into two types: focal, meaning they start in one area of the brain, or generalized, meaning both sides are involved from the start. Epilepsy itself gets sorted too, into drug responsive or drug resistant, the second label only applies after two medicines have failed. Typical care looks like this: someone newly diagnosed usually starts on one medicine picked for their seizure type. If two medicines do not work, guidelines say the next step is a referral to an epilepsy center, where surgery, devices, diet, and clinical trials are all options worth discussing. Children with specific syndromes follow their own separate paths. This describes usual practice, not a recommendation, your plan belongs to you and your neurologist.
For trials, "focal onset seizures" and "drug resistant" are the words you will see in most study titles, and gene, cell, and device studies often add "temporal lobe" or "unilateral" since they target one specific spot in the brain. Your neurologist's classification and your average monthly seizure count let you filter studies in minutes.
How to find an epilepsy clinical trial
AllClinicalTrials.com lists epilepsy studies recruiting across the US, everything from new drug trials to gene and cell therapy, devices, and diet studies. Two examples recruiting right now: the AMT-260 gene therapy study for adults with drug resistant temporal lobe epilepsy on one side of the brain, and a study forecasting seizures using EEG and wearable sensors for people with drug resistant focal or generalized epilepsy.
The application takes about 5 minutes: you answer questions about your seizure type, how often they happen, and which medicines you have tried, and if a study near you looks like a match, the research team reaches out to you. Nothing is decided until you have gone through informed consent, and taking part is voluntary the whole way through. Three things to have ready: your seizure type as your neurologist has described it, your average number of seizures per month, and the list of medicines you have already tried.
Common questions
What is the most effective treatment for epilepsy? For most people, it is finding the right single seizure medicine. For drug resistant focal epilepsy that starts in a spot doctors can safely remove, surgery gives the best shot at becoming completely seizure free, around two out of three people for temporal lobe cases. The best option really depends on the type of seizures and where they start in the brain.
What is the newest medication for epilepsy? Cenobamate, approved in 2019, is the most recent medicine approved for focal seizures in adults. Azetukalner, a pill that works in a completely new way, cut seizures by 53 percent in a major 2026 trial, and its maker planned to apply for FDA approval, but it is not approved yet, it is only available through clinical trials right now.
Can epilepsy go away on its own? Some childhood epilepsies do go away as a child gets older, and some adults stay seizure free for years and can slowly come off medicine with their doctor's guidance. For many people, epilepsy is lifelong but well controlled. Stopping medicine is always a decision to make together with your neurologist, never on your own.
Do epilepsy trials use placebo? In drug trials, often yes, but the placebo or the study drug is added on top of your current medicines, which you keep taking the whole time. No one is taken off their treatment. Studies on devices, surgery, or just observing patients usually do not use a placebo at all.
See epilepsy clinical trials recruiting now:
Browse open studies, filter by location, and apply in about 5 minutes: Epilepsy Clinical Trials