The One-Shot Reset: How Kyverna's CAR-T Data Could Rewrite the Rules for Autoimmune Disease

There is a category of clinical result that arrives so cleanly, so comprehensively, that it forces a recalibration of what the field believes is possible. Kyverna Therapeutics delivered exactly that kind of data on April 21, 2026, when it presented the full primary analysis from KYSA-8, its registrational Phase 2 trial of miv-cel in stiff person syndrome, at the American Academy of Neurology Annual Meeting in Chicago. The numbers were not merely good. They were the kind of numbers that make physicians stop and read the table twice.

Stiff person syndrome is a rare, progressive neurologic autoimmune disease affecting an estimated 6,000 Americans. It is defined by relentless muscle stiffness and painful spasms that progressively rob patients of mobility, independence, and quality of life. Up to 80 percent of patients eventually require walking aids or wheelchair use. The disease is driven by autoantibodies, most commonly against the glutamic acid decarboxylase 65 enzyme, that disrupt the inhibitory neurotransmission responsible for normal muscle relaxation. And until now, it has had no FDA-approved treatment of any kind. Patients and their physicians have been navigating a landscape of off-label immunotherapies, symptomatic medications, and supportive care, with most patients showing minimal or no meaningful improvement over time.

What the KYSA-8 Data Actually Show

In KYSA-8, 26 patients with stiff person syndrome who had an inadequate response to prior immunomodulatory therapies received a single dose of miv-cel, Kyverna's autologous CD19-targeting CAR-T cell therapy. The primary endpoint was the change from baseline in the Timed 25-Foot Walk at Week 16. The trial met that endpoint with a statistically significant median improvement of 46 percent from baseline, with a p-value of 0.0003. Eighty-one percent of patients achieved a clinically meaningful improvement, defined as at least a 20 percent reduction from baseline. Of the 12 patients who required a walking aid at the start of the trial, 67 percent no longer needed one at Week 16.

Every secondary endpoint was also met, with statistically significant improvements across the Modified Rankin Scale, the Hauser Ambulation Index, the Distribution of Stiffness Index, and the Heightened Sensitivity Scale, all at p less than 0.0001. Exploratory endpoints reinforced the picture, including significant reductions in GAD65 autoantibody titers consistent with the drug's mechanism of action, and improvements in physical and mental functioning that approached healthy population benchmarks on the SF-36 health survey. Perhaps most strikingly, all 26 patients were able to discontinue all chronic immunotherapies by Week 16 and remained off them through their last follow-up. That outcome, freedom from lifelong immunosuppression, has not been observed before in this disease.

The Mechanism Behind the Milestone

Miv-cel works by reprogramming a patient's own T cells to recognize and destroy CD19-positive B cells, the immune cells responsible for producing the pathogenic autoantibodies that drive stiff person syndrome. The result is a deep, transient depletion of B cells that allows the immune system to reset. The concept of using CAR-T cell therapy to achieve an immune reset in autoimmune disease has been building in the scientific literature for several years, with early case reports and small studies generating considerable excitement. What KYSA-8 has done is validate that concept in a rigorous, multicenter, registrational trial, with consistent results across all measured endpoints and a follow-up period that extends to more than a year for the earliest-treated patients.

The safety profile was notably clean for a CAR-T therapy. No high-grade cytokine release syndrome or immune effector cell-associated neurotoxicity syndrome was observed. Grade 3 or 4 neutropenia, a known side effect of the lymphodepletion conditioning required before CAR-T infusion, occurred in four patients and was manageable. Serious treatment-related adverse events occurred in three patients and resolved fully without lasting consequences. For a modality that has historically been associated with significant toxicity in oncology settings, the tolerability profile in KYSA-8 is a meaningful signal that CAR-T therapy in autoimmune disease may carry a different risk calculus than in cancer.

A Historic Threshold for the Field

The significance of what Kyverna is attempting extends well beyond stiff person syndrome. CAR-T cell therapies have been approved for blood cancers since 2017, but no CAR-T therapy has ever received regulatory approval for an autoimmune disease. If Kyverna's BLA submission, planned for the first half of 2026, results in FDA approval, miv-cel would become the first. That is not a minor distinction. It would represent the opening of an entirely new therapeutic category, one that applies the immune-engineering principles developed in oncology to the far larger universe of autoimmune conditions.

The implications are substantial. Kyverna is already running trials of miv-cel in generalized myasthenia gravis, where updated Phase 2 data presented at the same AAN meeting showed a 100 percent response rate in seven patients, with deepening responses sustained for up to a year. The company is also evaluating miv-cel in multiple sclerosis, lupus nephritis, and systemic sclerosis. The CD19-targeting mechanism that drives miv-cel's effect in stiff person syndrome is relevant across all B-cell-mediated autoimmune diseases, which is a large and heterogeneous population. If the immune reset concept holds across indications, the addressable market for this approach is orders of magnitude larger than the 6,000 Americans with stiff person syndrome.

The Regulatory Question That Remains

The one genuine uncertainty hanging over miv-cel's path to approval is whether the FDA will accept results from an open-label, single-arm study as sufficient evidence for full approval. The agency has recently signaled tougher standards for rare disease therapies that rely on external control cohorts or uncontrolled trial designs, most visibly in its handling of uniQure's Huntington's disease gene therapy application. Kyverna's CEO Warner Biddle confirmed that the FDA has never asked the company to run a randomized trial for miv-cel in stiff person syndrome, and the company is pursuing full approval rather than accelerated approval. The natural history data presented alongside KYSA-8, showing that 70 percent of patients experienced less than 20 percent improvement in walking speed over a decade of off-label treatment, provides important context for interpreting the single-arm results. Whether that context is sufficient will be one of the central questions of the regulatory review.

The broader lesson from KYSA-8 may be about what becomes possible when a therapeutic modality developed for one disease is applied with genuine scientific rigor to another. CAR-T cell therapy was built for cancer. The immune engineering principles it relies on, the ability to redirect T cells against specific cellular targets, turn out to be directly applicable to autoimmune diseases where the problem is not malignant cells but misdirected immune activity. The KYSA-8 data suggest that a single infusion can achieve what years of chronic immunosuppression cannot: a durable reset of the immune system that frees patients from both their disease and their treatment burden simultaneously. That is a genuinely different kind of therapeutic promise, and the field will be watching closely to see whether the FDA agrees.