Kyverna closed its scleroderma CAR T trial after three patients, and left the hypothesis untested

What was tried

Kyverna Therapeutics ran NCT06400303, KYSA-5, an open label, multicentre Phase 1/2 study of KYV-101 in adults with systemic sclerosis. KYV-101, also called mivocabtagene autoleucel or miv-cel, is an autologous fully human anti-CD19 chimeric antigen receptor T cell product. The design was sequential and non-randomised, a Phase 1 dose finding stage then a Phase 2 stage at the recommended dose, with cyclophosphamide and fludarabine lymphodepletion before infusion.

Eligibility was narrow: a 2013 ACR/EULAR classification of systemic sclerosis, diffuse cutaneous disease, under six years since the first non-Raynaud sign, and active disease. Clinically significant interstitial lung disease was an exclusion, as were prior cell or gene therapy and stem cell transplant. Phase 1 primary endpoints were adverse events, laboratory abnormalities and dose limiting toxicity frequency. The Phase 2 primary endpoint was the revised Composite Response Index in Systemic Sclerosis, rCRISS 30/5, at 52 weeks. Secondary endpoints covered CAR positive T cell counts, B cell levels, serum cytokines, rCRISS at 12, 24 and 52 weeks, and immunogenicity.

The study opened on 6 August 2024. Actual primary completion and completion were both 16 June 2026, at an actual enrolment of 3. ClinicalTrials.gov posted the terminated status on 31 August 2026 with the sponsor reason "Study discontinued due to sponsor decision", and no results were posted. On the same date the registry recorded the termination of NCT06342960, KYSA-3, the sister study of KYV-101 in refractory lupus nephritis, enrolment 2, same stated reason.

The biological hypothesis

Systemic sclerosis is characterised by skin and organ fibrosis, vasculopathy, and autoantibodies such as anti-Scl-70 and anti-RNA polymerase III. B cells are implicated in that pathology (10.1093/rheumatology/keac578). The rationale for CD19 CAR T over antibody mediated depletion is reach. Rituximab clears circulating B cells but leaves tissue resident and lymphoid populations relatively intact, and DESIRES reported only modest skin score effects in systemic sclerosis (10.1016/S2665-9913(21)00107-7). A CAR T product is designed to deplete B cells in lymphoid and non-lymphoid tissue, allowing a reconstituted naive repertoire rather than a trough.

The clinical precedent was real but small. A case series reported durable remissions after CD19 CAR T across lupus, myositis and systemic sclerosis (10.1056/NEJMoa2308917), an allogeneic CD19 CAR T was reported in myositis and systemic sclerosis (10.1016/j.cell.2024.06.027), and an iPSC derived CD19/BCMA CAR NK therapy in one systemic sclerosis patient (10.1016/j.cell.2025.05.038). The field was moving on case reports, not controlled data.

Open Targets scores the CD19 and systemic sclerosis association at 0.294: literature 0.556, clinical evidence 0.455, and no genetic association or somatic mutation datatype. The target is a lineage marker rather than a disease gene, so the association carries no human genetic anchor here.

What actually happened

Three participants were enrolled over 22 months against an endpoint requiring 52 weeks of follow-up per patient. The trial closed before it could produce an rCRISS readout, and no safety, pharmacodynamic or efficacy data were posted. openFDA FAERS holds no reports naming KYV-101. The class reference is instructive: FAERS returns 1,470 cytokine release syndrome and 373 neurotoxicity syndrome reports for tisagenlecleucel.

The sponsor's disclosures explain the decision better than the registry does. At the January 2026 J.P. Morgan conference Kyverna set out 2026 priorities around stiff person syndrome and generalized myasthenia gravis, and reported a December 2025 financing of roughly $105million.TheQ22026updatedescribesarollingBLAformiv-celinstiffpersonsyndrome,aPhase3gMGtrialenrolling,anRMATdesignationinnon-activesecondaryprogressivemultiplesclerosis,and$199.4 million in cash at 30 June 2026. Neither release mentions KYSA-3 or KYSA-5. The rheumatology indications were not defended and not disparaged. They were dropped.

Failure mechanism, best guess

The archetype is strategic_reprioritization, and the evidence for that label is unusually clean. Two trials in two autoimmune indications were terminated on the same date with identical wording, by a company simultaneously reporting a strengthened cash position and an accelerating neurology programme. This was portfolio allocation, not a stopping rule or a safety signal.

The mechanism behind it is competition for scarce manufacturing and clinical operations capacity. An autologous CAR T programme cannot run five indications at the enrolment rates seen here. Stiff person syndrome and myasthenia gravis offered RMAT designation, a defined responder population and a nearer filing. Systemic sclerosis offered a 52 week composite endpoint, a heterogeneous disease and no accepted surrogate. A sponsor with finite apheresis slots picks the first pair. KYSA-5 leaves no interpretable data behind, so the CD19 CAR T hypothesis in scleroderma is neither supported nor refuted.

How to prevent this next time

No endpoint level data were released, so no effect size, posterior or power calculation is supportable. The available levers are structural.

Design early autoimmune CAR T trials so a truncated run still yields interpretable biology. B cell depletion depth, CAR persistence and autoantibody titres are measurable within weeks and were already secondary endpoints here. Committing to publish them on termination converts an abandoned trial into a mechanistic data point.

Match endpoint horizon to portfolio horizon. A 52 week composite in a rare fibrotic disease fits badly with a single asset company facing a strategic review before the readout arrives. Either add an interim analysis or partner the indication at the outset.

Treat sponsor concentration as a scored risk. Both terminated indications depended on one company and one manufacturing chain, and investigator initiated studies carry a mechanism forward when the sponsor reallocates.

The single highest leverage change would have been a prespecified commitment to report B cell depletion, CAR persistence and autoantibody data on early termination, so three dosed patients still informed the field.

What this means for similar programs

The Claidex graph now holds three CD19 claims and the target mechanism risk score rises to 62 out of 100, orange band. Components: phase burden 14.92, archetype severity 12.16, recency 9.42, genetic deficit 10.60, saturation 14.66 from 19 distinct programmes. Read the composition carefully. Only one is an efficacy failure, LY3541860 in rheumatoid arthritis, a non-depleting antibody rather than a CAR T. The other two, KYSA-1 in refractory lupus nephritis and KYSA-5 here, are sponsor decisions that closed without readouts. The score is high because the record is crowded and shallow, not because CD19 depletion has been tested and found wanting.

For anyone planning a CD19 directed autoimmune programme, the signal is that saturation and sponsor concentration, not biology, retired two of the three programmes documented here.

Open questions

Did any of the three KYSA-5 participants achieve B cell aplasia, and for how long? Were rCRISS trajectories recorded at week 12 or 24 before closure? Will Kyverna publish or license the scleroderma data? Does the neurology pipeline preclude a partnered return to rheumatology, or merely defer it?

Sources

1. - ClinicalTrials.gov, NCT06400303 (KYSA-5): https://clinicaltrials.gov/study/NCT06400303 - ClinicalTrials.gov, NCT06342960 (KYSA-3, terminated 31 August 2026) - Open Targets Platform API: CD19 (ENSG00000177455) and MONDO_0005100 - ChEMBL API, mechanism records for CD19 (CHEMBL3390821) - openFDA FAERS drug event API: KYV-101 and tisagenlecleucel - N Engl J Med 2024, - Cell 2024, - Cell 2025, - Lancet Rheumatol 2021, - Rheumatology 2023, - Kyverna Therapeutics, J.P. Morgan corporate update, January 2026 - Kyverna Therapeutics, second quarter 2026 results - Claidex claims: kyv-101-cd19-car-t-refractory-lupus-nephritis-kysa-1-strategic-reprioritization, ly3541860-cd19-rheumatoid-arthritis-phase2a-efficacy-failure.