Old Arthritis Drug Could Fix Cancer Immunotherapy’s Scariest Side Effect

CAR T-cell therapy has given new hope to patients whose blood cancers stopped responding to everything else, yet it carries a serious, sometimes deadly side effect: a brain condition that can cause confusion, seizures, and death. Researchers at the University of California, Irvine now argue that an old arthritis drug deserves serious investigation as a potential solution, and their reasoning is hard to dismiss.

Their perspective paper, published in the journal Frontiers in Pharmacology, proposes that leflunomide, a drug widely used to treat rheumatoid arthritis, could be repurposed to prevent and treat a neurological complication called immune effector cell-associated neurotoxicity syndrome, or ICANS. Severe forms of the condition appear in up to 10% of CAR T-cell therapy patients and remain one of the most serious barriers to wider use of an otherwise life-changing cancer treatment. Doctors currently manage it with high-dose steroids, drugs that carry their own severe risks and fail a meaningful portion of patients.

What Is ICANS and Why Is It So Hard to Treat?

CAR T-cell therapy works by genetically re-engineering a patient’s own immune cells to find and destroy cancer. It has produced remarkable results for certain blood cancers that no longer respond to standard treatments. But once infused, these engineered immune cells can trigger a massive inflammatory response throughout the body, and sometimes that inflammation reaches the brain.

ICANS exists on a spectrum. On the mild end, patients experience confusion and difficulty finding words. On the severe end, it can escalate to seizures, dangerous brain swelling, and death. Standard response is large doses of corticosteroids, drugs that broadly suppress the immune system. Those steroids bring serious problems of their own: dangerous blood sugar spikes, sudden muscle weakness, and severe psychiatric episodes including psychosis. A subset of patients develop ICANS that does not respond to steroids at all, and for those patients, outcomes are poor.

Several other approaches are under investigation, including targeted drugs that block specific inflammatory proteins and modified versions of CAR T-cell technology designed to be less toxic. But none have become a universal standard of care.

The Mitochondria Connection in ICANS

This new paper focuses on a mechanism most researchers have not yet prioritized: the role of mitochondria, the tiny structures inside cells that generate energy. Mitochondria also regulate inflammation, and according to the paper’s authors, this dual role is central to why ICANS is so difficult to stop.

When inflammatory proteins triggered by CAR T-cell therapy begin damaging the brain, they also impair mitochondrial function. The authors propose that damaged mitochondria then release molecular alarm signals into the surrounding tissue, triggering more inflammation and, in turn, more mitochondrial damage. If that cycle takes hold during ICANS, a self-sustaining feedback loop could keep driving the condition even after doctors have gotten the bloodstream inflammation under control.

That distinction matters clinically. It may explain why some patients keep deteriorating despite treatment: the blood inflammation is managed, but inside the brain, a separate inflammatory cycle could keep running on its own.

Researchers describe this as “a feedforward amplifier of neuroinflammation” and argue it represents “a mechanistically distinct and therapeutically targetable pathway.” Brain cells are especially vulnerable because they depend almost entirely on their mitochondria for energy and cannot easily compensate when those structures fail.

Why Leflunomide Could Prevent Brain Toxicity From CAR T-Cell Therapy

Rather than simply naming a candidate drug, the UC Irvine team built a structured, four-part checklist for evaluating any drug being considered for this purpose. A candidate must show relevance to the specific cellular processes involved in ICANS, a proven ability to cross from the bloodstream into the brain, a safety profile that would not undermine the CAR T-cells doing the cancer-fighting work, and existing evidence from other inflammatory brain conditions suggesting it might work.

Applying this framework pointed toward several drug classes: those that block a mitochondrial enzyme involved in immune cell replication, antioxidants that target mitochondria directly, drugs that inhibit a protein complex that triggers inflammation, and drugs that improve how mitochondria manage energy output. To the authors’ knowledge, none are currently being formally tested for ICANS.

Leflunomide satisfied all four criteria. Its active form in the body blocks a mitochondrial enzyme that rapidly dividing immune cells depend on to multiply. By cutting off that supply chain, it selectively slows the overactive immune cells driving inflammation, while sparing cells that can get what they need through other pathways.

Critically, leflunomide’s active metabolite, teriflunomide, is already FDA-approved to treat relapsing multiple sclerosis, and studies in those patients have measured meaningful concentrations of the drug in the fluid surrounding the brain and spinal cord. That is an advantage many of the larger biological drugs currently being studied for ICANS cannot claim. Doctors also have years of experience managing leflunomide’s known side effects, primarily liver enzyme elevations requiring monitoring and, in some patients, risks requiring contraceptive counseling. As a bonus, the drug has demonstrated antiviral activity against cytomegalovirus, a meaningful benefit for patients who are already immunocompromised after CAR T-cell therapy.

One concern with any immune-suppressing drug used alongside CAR T-cell therapy is that it might blunt the very cells fighting the cancer. The authors address this directly, noting that leflunomide’s mechanism is reversible and can be pharmacologically counteracted if needed, and that careful timing of administration could preserve the therapy’s effectiveness while still managing brain toxicity.

Beyond leflunomide itself, the paper proposes a staged evaluation pathway, starting with laboratory validation and working toward a randomized clinical trial comparing leflunomide to current standard-of-care approaches. The authors also call for a broader push to systematically evaluate the entire class of mitochondria-targeting drugs for ICANS, and suggest that artificial intelligence-driven drug discovery could help identify additional candidates faster. They envision a future where blood-based protein measurements could identify patients at highest risk before brain symptoms begin, allowing preventive treatment to start earlier.

Steroid toxicity is not a minor inconvenience. For patients already battling relapsed or treatment-resistant cancer, developing steroid-induced psychosis, acute muscle weakness, or uncontrolled blood sugar on top of everything else is a genuine crisis. If an approved, well-characterized drug like leflunomide can prevent ICANS without those consequences and without undermining the cancer therapy itself, the case for testing it formally is hard to argue against. The authors argue that the shortcomings of steroids make the search for steroid-sparing alternatives increasingly urgent.

Disclaimer: This article describes a perspective paper, a type of publication in which researchers propose and argue for a hypothesis rather than report results from a clinical trial or laboratory experiment. Leflunomide has not been tested in patients for ICANS, and the mitochondrial mechanism described here is the authors’ proposed explanation, drawn from evidence in related conditions rather than demonstrated in ICANS patients. Nothing in this article is medical advice. Patients and caregivers with questions about CAR T-cell therapy or its side effects should consult their oncology care team.

---