CAR-T · Evidence by disease
Evidence atlas: read the denominator before the headline
A curated primary-source map, not a pooled response estimate.
In this chapter
- 2022 · Initial refractory SLE series
- 2024 · Multidisease NEJM series
- 2026 · CASTLE
- 2025 · CD19 + BCMA co-infusion
- 2026 · ICG318 extended follow-up
- 2026 · COMPARE phase 1
- 2026 · Pediatric autologous series
- 2025 · Allogeneic YTS109
- 2026 · Skin-remodeling analysis
- 2025 · AAV report
- 2024 · Sjögren’s with lymphoma
- 2026 · Descartes-08 phase 2b
- 2025–26 · In vivo reports
- 2026 · TRITON
- 2026 · A-319
- CD19-BBz CAR-NK in refractory SLE
Studies are organized by disease, design and platform. This is not a systematic review or a meta-analysis. Cohorts may overlap across clinical, mechanistic and conference publications; the rows must not be added to estimate a unique treated population.
2022 · Initial refractory SLE series
Population: 5
All five met DORIS remission at 3 months; initial follow-up was short. [1]
Interpretation: Established human proof of concept; do not count early patients again when later overlapping series are summarized.
2024 · Multidisease NEJM series
Population: 15: 8 / 4 / 3
Median follow-up 15 months, range 4–29. Disease-specific improvement and immunosuppressive withdrawal were reported. [2]
Interpretation: Endpoints differed by disease; one grade-1 ICANS and pneumonia requiring hospitalization were reported.
2026 · CASTLE
Population: 24: 10 / 9 / 5
At 24 weeks: SLE DORIS 9/10; SSc no progression 9/9; IIM moderate/major response 4/5. All 24 were off glucocorticoids and other immunosuppression during the assessment period. [3]
Interpretation: No CRS above grade 2 or ICANS in this cohort; 15 serious adverse events were recorded. “22/24 endpoints met” is not “22/24 cured.”
2025 · CD19 + BCMA co-infusion
Population: 15
Evaluated two autologous products after lymphodepletion, with early clinical and safety endpoints. [7]
Interpretation: Separate-product co-infusion; no direct comparison proving superiority over single-target CAR-T.
2026 · ICG318 extended follow-up
Population: 12
Reported stringent complete remission in 10/12. Median follow-up 30 months; maximum 67 months. [4]
Interpretation: Two patients had concomitant lymphoma; small uncontrolled cohort. Maximum follow-up must not be presented as the duration for every patient.
2026 · COMPARE phase 1
Population: 6
Three achieved DAS28-CRP remission and ACR70; follow-up 36–52 weeks. [5]
Interpretation: All had grade-1/2 CRS; one reversible grade-3 transaminase DLT. No ICANS or serious adverse events reported in these six.
2026 · Pediatric autologous series
Population: 8: 4 / 3 / 1
Published clinical experience in severe treatment-refractory childhood disease. [6]
Interpretation: Not a randomized pediatric trial or a general approval for children; do not infer efficacy for all juvenile rheumatic disorders.
2025 · Allogeneic YTS109
Population: 5
Early human feasibility of a gene-edited donor-derived CD19-targeting STAR product. [8]
Interpretation: Distinct architecture; small cohort cannot resolve rejection, durability or uncommon toxicity.
2026 · Skin-remodeling analysis
Population: Serial samples
Skin structural, fibroblast and vascular changes supported tissue-remodeling potential. [13]
Interpretation: Includes previously treated CASTLE/named-patient material; not a new independent randomized efficacy cohort.
2025 · AAV report
Population: 1
Reported anti-CD19 treatment with antibody seroconversion and marrow B-cell depletion. [14]
Interpretation: Case-level evidence; not a standard induction pathway or proof in all AAV phenotypes.
2024 · Sjögren’s with lymphoma
Population: 1
Sjögren’s improvement was described during lymphoma-directed CAR-T treatment. [15]
Interpretation: Concurrent malignancy and its treatment substantially confound disease-specific conclusions.
2026 · Descartes-08 phase 2b
Population: 36 randomized; 26 primary efficacy
The reported 3-month MGC response was 10/15 versus 3/11 in the primary efficacy population. [9]
Interpretation: Neurologic, not myositis, efficacy. The randomized and primary-efficacy denominators differ and require scrutiny.
2025–26 · In vivo reports
Population: Small / heterogeneous
Human proof of concept has been reported for RNA/LNP SLE and lentiviral neurologic approaches. [11]
Interpretation: Different delivery systems and diseases; no automatic extrapolation to routine rheumatology care.
2026 · TRITON
Population: Planned trial
A placebo-controlled Descartes-08 myositis protocol is registered. [31]
Interpretation: Study design is not an efficacy result; current enrollment must be confirmed with the site.
2026 · A-319
Population: 12 in IV arm
A CD3×CD19 T-cell engager was evaluated with 52-week follow-up. [19]
Interpretation: Comparator landscape only: this is not CAR-T and not a CAR-T head-to-head trial.
For new data, preserve screened, enrolled, apheresed, manufactured, infused and evaluable denominators whenever available. Manufacturing failure, early withdrawal, missing assessments and rescue therapy all affect interpretation. Event-free nonprogression is not synonymous with remission; a laboratory response is not automatically an organ response.
CD19-BBz CAR-NK in refractory SLE
Design: phase 1, open-label dose escalation; 5 patients.
Main observation: 4/5 met DORIS remission at final follow-up, median 12 months.
Boundary: CAR-NK, not CAR-T; no controlled superiority comparison or inferred drug-free remission rate. [40]
November 2026 · COMPARE week-36 follow-up
Population: 6 patients already included in the published COMPARE phase 1 cohort; not an independent new cohort.
Four patients reached DAS28-CRP remission and ACR70 by week 36; this differs from remission at latest follow-up and should not be substituted for the published endpoint. ACR abstract.
Interpretation: Abstract-level longitudinal update, uncontrolled cohort; the timepoint, treatment exposure and flare history determine the interpretation.
November 2026 · Rese-cel organ-specific signals
Lupus nephritis: five of five participants who elected repeat biopsy had histologic response or remission and DORIS remission off immunomodulators. Systemic sclerosis: five of seven evaluated at one year met rCRISS-50. Cabaletta company report.
Interpretation: Different disease-specific endpoints and selected denominators, not a comparative treatment effect or evidence of universal organ reversal.
November 2026 · Long-term relapses after CD19 CAR-T
Population: 50 recipients, including 28 SLE, 14 SSc and 8 inflammatory myopathy patients. Seven relapses were reported (1, 3 and 3 respectively), at a median of 18 months after infusion. ACR 2026 abstract.
Interpretation: Selected, heterogeneous single-center experience with at least six months of follow-up; not a general population estimate and not proof of a standard rescue treatment sequence.
Sources · 16
Source titles are shown in their original language.
[1] Primary research · 2022-09-15
Mackensen et al. Anti-CD19 CAR T cell therapy for refractory systemic lupus erythematosus. Nature Medicine.
Open original source ↗[2] Primary research · 2024-02-21
Müller et al. CD19 CAR T-Cell Therapy in Autoimmune Disease — A Case Series with Follow-up. NEJM.
Open original source ↗[3] Primary research · 2026-01-07
CD19 CAR-T cells for treatment-refractory autoimmune diseases: the phase 1/2 CASTLE basket trial. Nature Medicine.
Open original source ↗[4] Primary research · 2026-01-29
Hong et al. BCMA-CD19 armored compound CAR T cells in systemic lupus erythematosus: extended follow-up of a phase 1 clinical trial. J Hematol Oncol.
Open original source ↗[5] Primary research · 2026-08-27
Albach et al. CD19 CAR T cell therapy for treatment-refractory seropositive rheumatoid arthritis: a phase 1 trial. Nature Medicine.
Publisher abstract and publication metadata reviewed; no claim of full supplementary-data review.
Open original source ↗[6] Primary research · 2026-02-05
Becilli et al. Anti-CD19 CAR T cells for pediatric patients with treatment-refractory autoimmune diseases. Nature Medicine.
Publisher/PubMed abstract and metadata; detailed numerical efficacy extraction deliberately limited.
Open original source ↗[7] Primary research · 2025-09-24
Feng et al. Co-infusion of CD19-targeting and BCMA-targeting CAR-T cells for treatment-refractory SLE: a phase 1 trial. Nature Medicine.
Primary abstract reviewed; two separately manufactured products, not a single dual-target construct.
Open original source ↗[8] Primary research · 2025-08-27
Wang et al. Allogeneic CD19-targeting T cells for treatment-refractory SLE: a phase 1 trial. Nature Medicine.
YTS109 uses a synthetic TCR and antigen receptor (STAR); do not describe all engineered T-cell products as identical conventional CARs.
Open original source ↗[9] Primary research · 2026-01-09
Vu et al. BCMA-directed mRNA CAR T cell therapy for myasthenia gravis: randomized, double-blind, placebo-controlled phase 2b trial. Nature Medicine.
Open original source ↗[11] Primary research · 2025-09-17
Wang et al. In Vivo CD19 CAR T-Cell Therapy for Refractory Systemic Lupus Erythematosus. NEJM.
Primary correspondence/abstract; early human proof of concept, not confirmatory trial evidence.
Open original source ↗[13] Primary research · 2026-05-23
Rius Rigau et al. Deep phenotyping of skin tissue remodeling in patients with systemic sclerosis treated with CD19-CAR T cells. Nature Communications.
Mechanistic tissue study; includes CASTLE/named-patient material, not wholly independent treatment cohort.
Open original source ↗[14] Primary research · 2025-01-31
Minopoulou et al. Anti-CD19 CAR T cell therapy induces antibody seroconversion and complete B cell depletion in bone marrow in therapy-refractory ANCA-associated vasculitis. Ann Rheum Dis.
Human case report; not a controlled vasculitis trial.
Open original source ↗[15] Primary research · 2024
CAR T-cell therapy in a patient with Sjögren’s syndrome and diffuse large B-cell lymphoma. Primary report, PubMed 38173731.
Concurrent lymphoma materially limits inference about Sjögren’s-specific efficacy.
Open original source ↗[19] Primary research · 2026-09-10
A bispecific CD3×CD19 antibody for systemic lupus erythematosus: a phase 1 trial. Nature Medicine.
A-319 is a bispecific antibody/T-cell engager, not CAR T-cell therapy.
Open original source ↗[31] Trial-center registry record · 2026
UC Irvine clinical trial record. Descartes-08 in dermatomyositis and antisynthetase syndrome (TRITON), NCT07391605.
Protocol identity/design verified; live enrollment availability not independently confirmed with the site.
Open original source ↗[40] Primary research · 2026-07-22
Zeng et al. Safety and efficacy of low-dose, cord blood-derived CD19 CAR-NK cells with 4-1BB co-stimulation in refractory systemic lupus erythematosus. Med.
Primary abstract reviewed; August 14, 2026 issue. Five-patient phase 1 cohort; CAR-NK, not CAR-T.
Open original source ↗