First CAR-T Therapy for an Autoimmune Disease Edges Toward FDA Approval: What Kyverna's Stiff Person Syndrome Data Show

What This Article Covers

Every CAR-T therapy approved by the FDA to date treats blood cancer. That may be about to change. Kyverna Therapeutics is now submitting a rolling Biologics License Application for mivocabtagene autoleucel (miv-cel, also known as KYV-101) in stiff person syndrome (SPS), a rare and disabling autoimmune neurological disease — based on Phase 2 registrational data the company calls "truly remarkable." This article covers what the trial actually found, where the therapy stands in the FDA review process, and what real uncertainties remain before it could reach patients.

What Stiff Person Syndrome Is, and Why It's Hard to Treat

Stiff person syndrome is a rare autoimmune disorder, estimated to affect roughly one in a million people, in which the immune system produces autoantibodies — most often against an enzyme called GAD65 — that disrupt nerve signaling and cause progressive muscle rigidity, painful spasms, and falls severe enough to fracture bones. Because B cells are believed to drive the autoantibody production behind the disease, current treatment leans on off-label options such as high-dose intravenous immunoglobulin, benzodiazepines, and rituximab, none of which is approved specifically for SPS and many patients respond only partially or not at all.

CD19 CAR-T therapy takes a more targeted approach: it engineers a patient's own T cells to hunt down and destroy CD19-positive B cells, the source of the pathogenic autoantibodies, in a single infusion rather than ongoing infusions or immune suppression.

What the KYSA-8 Trial Found

The data come from KYSA-8, a Phase 2 registrational trial of 26 adults with SPS who had inadequate responses to available off-label therapies, each treated with a single dose of miv-cel. The results, first reported as topline data in December 2025 and later presented in more detail at the American Academy of Neurology's annual meeting, are genuinely striking for a disease this difficult to treat. On the trial's primary endpoint — time to complete a 25-foot walk test at week 16 — patients improved by an average of 46% from baseline, a highly statistically significant result (p=0.0002). Eighty-one percent of patients reached a clinically meaningful improvement of at least 20%, and two-thirds of patients who needed a walking aid at the start of the trial no longer needed one by week 16. Secondary measures of disability, muscle stiffness, and spasm frequency reportedly improved as well. On safety, the company has reported no high-grade cytokine release syndrome or neurotoxicity (ICANS) among the 26 treated patients — a meaningfully cleaner safety signal than CAR-T typically produces in cancer patients, though CRS and neurotoxicity remain recognized class risks of CD19 CAR-T therapy more broadly, alongside the risk of prolonged B-cell depletion and infection.

It's worth being precise about what this trial can and can't tell us. It was a single-arm study with no placebo or control group, so the size of the improvement is being judged against patients' own pretreatment baseline and the known natural history of SPS, not a randomized comparison. The sample is small — 26 patients — and as of the most detailed public reporting, only 16 of them had reached 24 or more weeks of follow-up, so durability beyond four months is still being established. Kyverna itself has said the ongoing follow-up will determine "how the patients do in both short-term and long-term follow-up," a fair acknowledgment that the story isn't finished.

Where the FDA Review Stands

Miv-cel already holds Regenerative Medicine Advanced Therapy (RMAT) and Orphan Drug designations from the FDA for SPS, both of which are intended to speed development and regulatory engagement for products addressing serious, unmet medical needs — they are not themselves approvals. According to Kyverna's August 2026 pipeline update, the company has begun a rolling BLA submission, with the chemistry, manufacturing, and controls module already filed, and expects to complete the full submission by the fourth quarter of 2026 while seeking priority review. If that timeline holds and the FDA agrees to review under priority status, Kyverna is positioning for a possible commercial launch in 2027 — though that depends on an FDA decision that hasn't happened yet, and the agency could ask for more data before ruling. Separately, the company reports 12-month topline data from KYSA-8 was expected in the third quarter of 2026, which should offer a clearer picture of how durable these early gains are.

The stakes extend beyond this one disease. If miv-cel is approved for SPS, it would be the first CAR-T therapy ever approved for an autoimmune condition rather than cancer — a milestone the field has been watching closely as early academic studies of CAR-T in lupus and other autoimmune diseases have generated excitement over the past few years. Regulators have signaled they intend to move carefully here: the FDA has indicated it wants to take a deliberately "tailored approach" to shepherding CAR-T into autoimmune disease, reflecting that the risk tolerance appropriate for cancer patients facing few other options doesn't automatically transfer to patients with a serious but non-fatal autoimmune disease. Kyverna is also pursuing miv-cel in a second, earlier-stage program: the FDA granted RMAT designation for non-active secondary progressive multiple sclerosis based on investigator-initiated data from Stanford, though that program remains well behind SPS and a development strategy update isn't expected until early 2027.

Bottom Line

The KYSA-8 data are a genuinely encouraging result for a disease with few good treatment options, and the regulatory path — RMAT and orphan designations, a rolling BLA submission already underway, priority review being sought — reflects real momentum toward what could be a historic first approval. But "moving toward approval" is not the same as "approved": the evidence rests on 26 patients in an open-label trial with limited long-term follow-up, and the FDA has not yet ruled. Patients and families should treat this as a promising therapy in late-stage development, not an available treatment, and watch for the completed BLA submission and FDA decision — expected at the earliest in 2027 — before drawing firmer conclusions.