Year-long walking gains in stiff person syndrome, not a CAR-T rewrite

In Short. Stiff person syndrome is a rare autoimmune illness that turns muscles rigid and can steal walking. A one-time treatment that engineers immune cells to wipe out B cells just showed year-long walking gains in a 26-patient study, with no high-grade inflammatory nerve or cytokine crises of the kinds that recently paused some Big Pharma peers. Promising for that rare disease package, not a rewritten rule for every autoimmune CAR-T.

---

Stiff person syndrome is a rare autoimmune disease. The immune system attacks molecules that help nerves quiet muscle tone. Muscles lock into painful rigidity and spasms. Walking fails. Many patients eventually need a cane, walker, or wheelchair. There is no approved disease-modifying drug. Off-label immunotherapies are the usual stopgap, and they often fail to halt the slide.

The treatment idea under test is a one-time cell therapy. Doctors take a patient’s own immune cells, engineer them to hunt B cells that carry the surface marker CD19 (the cells that help drive the autoimmunity), and return that single infusion. The goal is deep B-cell wipeout and a lasting immune reset, not lifelong infusions.

Kyverna Therapeutics reported one-year topline results from KYSA-8, its registrational single-arm trial of that approach, branded miv-cel (mivocabtagene autoleucel), in 26 adults with stiff person syndrome. On the timed 25-foot walk (T25FW), a clocked walk used to track mobility, median improvement from baseline was 46% at week 16 and 49% at month 12 (p<0.0001 for the sustained month-12 result). Of patients who had already hit a clinically meaningful gain of at least 20% at the primary analysis, 95% kept that level at one year. Among the 12 who needed a walking aid before treatment, 67% still did not need one. Ninety-two percent remained off chronic stiff person immunotherapies. The company says it will fold the one-year package into its rolling biologics license application, with completion aimed at the fourth quarter of 2026. (Kyverna)

The April primary cut, filed with the SEC and presented at the American Academy of Neurology meeting, had already shown the week-16 walk win (then p=0.0003) and no high-grade cytokine release syndrome (CRS, a systemic inflammatory reaction to engineered T cells) or immune effector cell-associated neurotoxicity syndrome (ICANS, brain and nerve toxicity tied to the same therapies). The one-year update still reported no high-grade CRS or ICANS, and zero cases of immune effector cell-associated hemophagocytic lymphohistiocytosis-like syndrome (IEC-HS, a severe inflammatory bone-marrow and organ crisis). (SEC EX-99.1)

That safety contrast matters because larger autoimmune CAR-T programs have stumbled. Novartis held rapcabtagene autoleucel after three IEC-HS deaths. Bristol Myers Squibb paused zolacabtagene autoleucel after inflammatory events. Kyverna’s chief executive argues that miv-cel’s fully human CD19 binder, CD28 co-stimulatory design, and conventional manufacturing differ from those peers. That is a company claim, not independent mechanistic proof that construct alone explains why those high-grade crises did not appear here. (Fierce Biotech)

A smaller secondary update sits beside the stiff person package. In the Phase 2 portion of KYSA-6 in generalized myasthenia gravis, all seven patients had meaningful gains on daily-living and strength scores at 24 weeks. Five with at least a year of follow-up kept clinically meaningful improvements. Minimal symptom expression (a daily-living score of 0 or 1) held in 57% at last follow-up. Phase 3 enrollment is targeted for mid-2027. Useful context, not a second lead.

Miv-cel is not approved. KYSA-8 is a 26-patient single-arm package in one rare neurologic disease, now durable to a year and heading into a rolling license file. It is promising for that package. It is not a rewritten safety or efficacy rule for every autoimmune CAR-T. ■