Chimeric antigen receptor (CAR) T cell therapy, initially developed for hematologic malignancies, has recently been applied to refractory autoimmune diseases (AID) 1. Early experience in systemic lupus erythematosus (SLE) demonstrated durable clinical remission following a single B cell–directed infusion in heavily pretreated patients 2. These findings have prompted reconsideration of the risk–benefit balance, mechanistic underpinnings, and trial design strategies specific to autoimmunity. Reports of transient CAR T expansion coinciding with sustained drug-free remission introduced the hypothesis of functional immune recalibration rather than simple B cell depletion 3. Activity observed in antisynthetase syndrome-associated myositis and interstitial lung disease supports cross-disease applicability of B lineage targeting 4. Feasibility was further demonstrated in severe systemic sclerosis; CAR T therapies showed their effectiveness in severe patients, such as candidates for lung transplantation 5. An RNA-engineered, transiently expressed CAR strategy showed promising activity in myasthenia gravis, highlighting alternative persistence-tuning approaches 6. Early anti-BCMA experience in neuromyelitis optica spectrum disorder broadened the therapeutic target space beyond CD19 7. Collectively, these uncontrolled but diverse reports support the possibility of a generalizable immune “reset.” Toxicity thresholds acceptable in oncology, where short-term mortality was high, were less tolerable in autoimmune conditions characterized by prolonged morbidity and near-normal life expectancy. For patients with AID, lower ceilings for cytokine release syndrome (CRS) and neurotoxicity were warranted, as outlined by consensus grading frameworks 8. Concerns regarding long-term sequelae, for example secondary malignancies, persistent hypogammaglobulinemia, or chronic cytopenias are further amplified by younger age at treatment. These considerations justified a stratified approach, initially prioritizing organ-threatening refractory phenotypes. Several mechanistic issues remain unresolved and must be embedded into clinical trial protocols. These include the extent and durability of B cell repertoire remodeling relative to clinical remission; minimal expansion thresholds required to destabilize autoreactive networks; the relative contributions of T follicular helper modulation versus direct B cell contraction; the role of tissue-resident immune niches in systemic recalibration; and predictive biomarkers such as plasmablast proportion, interferon signatures, and BAFF levels 9. In addition, insights from senescence-directed cellular interventions suggest that selective clearance of pathogenic stromal or immune subsets may synergize with antigen-specific tolerization strategies 10. Optimization of CAR T therapy for AID requires a departure from oncology paradigms. Priorities include the integration of modular safety switches, attenuation or redesign of lymphodepletion regimens, and reduction of immunogenicity to enable retreatment 11. Development of allogeneic, genome-edited products seeks to reduce vein-to-vein time while mitigating host-versus-graft responses and innate rejection through multiplex editing and HLA modulation 12. Emerging constructs such as chimeric autoantibody receptor (CAAR) T cells and regulatory T cell–based approaches may allow immunomodulation to be confined to culprit clones, reducing broad hypogammaglobulinemia risk. Transient expression systems or in vivo gene delivery may further refine persistence. Future trials must adopt adaptive, biomarker-rich designs. Key features should include early (≤ 90 days) safety and pharmacokinetic endpoints, intermediate (3–12 months) organ-specific indices, and long-term (≥ 24 months) durability and infection outcomes. Integrated biospecimen schedules—including single-cell profiling, BCR/TCR sequencing, and soluble mediator panels—are critical. Standardized infection prophylaxis and immunoglobulin replacement protocols should be mandated, with toxicity reporting harmonized using consensus frameworks. Bayesian or response-adaptive designs can minimize unnecessary exposure in lower mortality settings while efficiently validating efficacy signals. High manufacturing cost and limited treatment center capacity raise concerns regarding equitable access. The absence of validated predictive biomarkers increases the likelihood of overtreatment in patients who might otherwise respond to emerging biologics or small molecules. Prospective long-term registries, beginning at phase II, are essential for monitoring late cytopenias, oncogenic transformation, and persistent immune dysregulation. Transparent patient counseling must distinguish CAR T therapy as an effectively irreversible cellular intervention, in contrast to cyclical and reversible immunomodulators. Transition from experimental therapy to standard practice requires reproducible multicenter evidence of durable drug-free remission, validated predictive biomarker panels, and clearly defined long-term safety parameters. Manufacturing innovations that reduce cost and time without compromising quality will further influence feasibility. Until these conditions are met, CAR T therapy in autoimmunity should remain restricted to rigorously designed clinical trials with embedded translational analyses. Early experiences with CD19- and BCMA-directed CAR T therapy in autoimmunity suggest the possibility of single-intervention immune recalibration. Nevertheless, the biological, ethical, and economic contexts of autoimmunity differ fundamentally from oncology, requiring bespoke developmental strategies. Mechanistic clarity, safety-focused engineering, and stratified trial design will determine whether CAR T becomes a transformative therapeutic modality or remains limited to exceptional salvage cases. Conceptual framing (all authors); draft preparation (C.Y.W., H.Y.C.); critical revision (S.B.Y., C.Y.W.); final approval (all authors). The authors have nothing to report. The authors have nothing to report. The authors have nothing to report. The authors declare no conflicts of interest. Data sharing not applicable to this article as no datasets were generated or analyzed during the current study.
Wang et al. (Wed,) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: