Arthritis Drug Tofacitinib Stops Seizures in Mice for Months After Treatment Ends

A 10-day course of tofacitinib, an FDA-approved arthritis drug, kept mice with chronic epilepsy seizure-free for at least two months after treatment ended, according to a University of Wisconsin-Madison study published March 25, 2025 in Science Translational Medicine. If the result translates to human patients, the drug would become the first to provide lasting seizure relief that persists after patients stop taking it.

The study, led by postdoctoral researcher Olivia Hoffman and senior author Avtar Roopra, a neuroscience professor in the UW-Madison School of Medicine and Public Health, also found that tofacitinib restored short-term and working memory in the mice and reduced brain inflammation driven by the disease.

"It ticks all the boxes of everything we've been looking for," Roopra said.

What did the researchers actually do?

The team first used recent data science methods to compare gene expression across millions of cells in the brains of mice with and without epilepsy. That analysis pointed to a protein called STAT3, a key node in a cell signaling pathway known as JAK, at the center of activity in seizure-affected mouse brains. When the researchers ran the same analysis on brain tissue removed from human epilepsy patients, they found the same STAT3-driven signature.

A separate clue came from Taiwan: a study of tens of thousands of arthritis patients found epilepsy more common among them than in people without arthritis — but less common than expected among those who had taken anti-inflammatory drugs for more than five and a half years. Long-term rheumatoid arthritis patients are typically prescribed JAK-inhibitors, drugs targeting precisely the pathway the Wisconsin lab suspected.

The researchers then dosed mice with tofacitinib after administering a brain-damaging drug that triggers repeated seizures. The first attempt failed. The mice still developed epilepsy, mirroring the human condition, in which seizures often emerge only months or years after a brain injury.

The key came from timing. In the mouse model, weeks of relatively normal behavior separate the initial injury from what the researchers call "reignition" of seizures. The team designed a 10-day tofacitinib course starting at that reignition point.

"Honestly, I didn't think it was going to work," Hoffman said. "But we believe that initial event sort of primes this pathway in the brain for trouble. And when we stepped in at that reignition point, the animals responded."

Collaborators at Tufts University and Emory University replicated the seizure-free results in mouse models of slightly different epilepsy variants. Roopra's lab has since followed mice that remained seizure-free for four and five months.

What does this change for clinical practice?

For now, nothing directly: these are animal data, not human trial results. But the findings identify the JAK-STAT3 pathway as a disease-modifying target rather than a merely symptomatic one, a distinction from current anti-seizure drugs, which address symptoms without halting the underlying process. Existing drugs also leave roughly one-third of epilepsy patients without an effective medication, according to Hoffman; for the most damaging seizures, the only current option is surgical removal of the brain region where the disruptive activity originates.

The memory findings carry particular weight. "These animals are having many seizures a day. They cannot navigate mazes. Behaviorally, they are bereft," Roopra said. "We gave them that drug, and the seizures disappear. But their cognition also comes back online, which is astounding."

Roopra said the drug appears to act on multiple brain systems simultaneously, unlike existing drugs that force a single component back under control. Epilepsy affects more than 50 million people worldwide and often follows a brain injury such as physical trauma or stroke.

How close is a human trial?

Tofacitinib already holds FDA approval as safe for human use in arthritis, which could shorten the path from animal studies to clinical trials compared with a novel compound. That path is currently blocked: next steps toward human patients largely await NIH review of new studies, paused indefinitely amid changes at the agency.

The research drew funding from the National Institutes of Health (grants R01NS108756, R21NS093364, R01NS112308, NS112350, R01NS105628, R01NS102937 and R21NS120868), with early support from Madison-based Lily's Fund for Epilepsy Research and CURE Epilepsy. Hoffman and Roopra have filed a patent on the use of tofacitinib in epilepsy.

The researchers now aim to identify which brain cell types tofacitinib shifts back to healthy behavior and to test the drug in animal models of additional epilepsy subtypes, work that will set the stage for the human trials the field is waiting to see.