2.1 Viral infections

2.1.1 Human immunodeficiency virus

Acquired immunodeficiency syndrome (AIDS), a chronic disease resulting from infection with the human immunodeficiency virus (HIV), has emerged as a significant global public health issue.^{57, 58} Despite antiretroviral therapy (ART) reducing the viral load in the plasma of HIV-infected individuals to undetectable levels, ART alone is insufficient to completely eradicate the virus, as integration of HIV into the host cell genome leads to the formation of viral reservoirs.^{59, 60} Preclinical studies and early-phase clinical trials have provided preliminary evidence supporting CAR-T therapy in combating AIDS. Currently, there are two main strategies for designing CAR constructs targeting HIV infection: one that focuses on targeting the interaction between CD4 and gp120, and another that utilises broadly neutralising antibodies (BNAbs) to target the virus's envelope glycoproteins. The earliest CAR-T cells designed for HIV infection involved the incorporation of the CD4 extracellular domain on CD8+ T cells, enabling them to bind to and lyse cells expressing HIV-1 gp120 in vitro.⁶¹ However, despite the ability of these CAR-T cells to inhibit viral replication, clear HIV-infected cells, and persist for extended periods in vivo, they did not significantly impact the viral reservoirs.⁶² By incorporating the co-stimulatory domain CD28, CD4-CAR T cells effectively target and eliminate HIV Env-expressing cells in cellular models and, through reactivation of latent HIV, identify and kill latently infected cells.⁶³ Research based on an HIV-infected humanised mouse model has shown that replacing the CD28 co-stimulatory domain with the 4-1BB co-stimulatory domain not only greatly enhances CD4-CAR T cells activity but also improves their ability to control HIV spread following ART treatment.⁶⁴ CAR-T therapies utilising BNAbs also exhibit effective antiviral activity,⁶⁵ and ongoing improvements have enabled these cells to recognise and eliminate latently infected cells.^{66, 67} Additionally, bispecific CAR-T cells combining CD4 and BNAbs have been investigated, showing stronger effects and more specific binding to target cells.⁶⁸ Several clinical studies are assessing the safety and efficacy of CAR-T therapy for treating AIDS.^69-71 For example, Deeks et al. initiated a clinical trial (NCT04648046) aiming to recruit 18 individuals to evaluate the safety and anti-HIV effects of T cells expressing bispecific anti-gp120 CAR molecules (duoCAR; Figure 3A). In a prior preclinical study, they had already evaluated the capacity of duoCAR-T cells to identify and destroy HIV-infected cells in vitro and in vivo models, laying the groundwork for the current clinical trial.⁶⁹ In June 2023, the research team provided a brief report on the first two patients treated with duoCAR-T, with preliminary results indicating that the participants were in good condition and no related adverse events were observed.⁷⁰ Additionally, another clinical trial led by Mao et al. aimed to assess the performance and safety profile of versatile M10 CAR-T cells in treating HIV-1-infected patients.⁷¹ The multifunctional M10 CAR-T cells are engineered to perform three key biological roles through the expression of a bispecific CAR molecule, a 10E8scFv-Fc fusion antibody and the receptor CXCR5 for follicular homing. These functions include broad cytotoxicity against cells harbouring HIV, neutralisation of viruses in a cell-free state and promotion of cell migration to germinal centres to target viral reservoirs (Figure 3B). The study demonstrated that, following M10 CAR-T infusion, both plasma viral load and cell-associated HIV-1 RNA levels were significantly reduced, with no severe adverse events reported.

NK cells play a critical role in the innate immune response to viral infections.⁷² Compared to CAR-T cells, CAR-NK cells offer unique advantages, such as a lower risk of graft-versus-host disease (GVHD) and reduced long-term toxicity.⁷³ As a result, CAR-NK cells are being explored as a viable option for HIV treatment. As early as the 1990s, researchers explored the use of retroviral transduction to genetically modify human NK cells for high expression of CD4ζ.⁷⁴ This study demonstrated that these CD4ζ+ CAR-NK cells could effectively lyse gp120-expressing or HIV-infected T cells. Unfortunately, in vivo experiments in mice revealed that CD4ζ+ CAR-NK cells did not exhibit superior HIV-suppressing efficacy compared to unmodified NK cells.⁷⁵ Recently, Lim et al. designed a CAR-NK cell capable of recognising multiple epitopes of gp160 from different HIV-1 subtypes, potentially addressing the challenge of HIV diversity.⁷⁶ This universal CAR-NK cell expresses a CAR structure targeting 2,4-dinitrophenyl (DNP) and utilises DNP-conjugated anti-gp160 antibodies as bridging molecules to target various gp160 epitopes (Figure 3C).

2.2 Bacterial infections

Bacterial infections have long been one of the most serious challenges in medicine, with antibiotics serving as an effective means of treatment.¹⁰⁹ However, the rising prevalence of antibiotic-resistant bacteria and the limitations of current antibiotic therapies underscore the urgent need for innovative approaches to treating bacterial infections.¹¹⁰ Currently, research on CAR-T therapy for bacterial infections is quite limited, with its most promising application being the treatment of chronic diseases such as tuberculosis. Although CAR cells have not yet been developed for tuberculosis, several potential strategies have been proposed. One approach involves identifying antibodies against bacterial antigens in latently infected tuberculosis patients, which could then be cloned to generate scFv fragments and ultimately used to create CARs for T cells and NK cells.¹¹¹ Additionally, given the critical role that NKT cells play in tuberculosis infection,^112-114 researchers have suggested their direct use in adoptive cell therapy or the incorporation of their receptors as recognition domains in CAR constructs.¹¹⁵ Another promising avenue is the development of CAR-NKT cells, which could utilise both natural receptors and CARs to recognise bacteria, thereby overcoming potential mutations in antigens recognised by the CAR.¹¹⁵ Notably, Liang et al. reported the safety and therapeutic effectiveness of Vγ9Vδ2 T cells in treating tuberculosis in a clinical study.¹¹⁶ This finding indicates that Vγ9Vδ2 T cells may be further enhanced through integration with CAR technology, offering improved efficacy in complex infectious environments. Encouragingly, while evidence supporting CAR-T therapy for bacterial infections remains limited, the efficacy of CAR-M therapy has been preliminarily validated. Given the critical role of macrophages in host defence and inflammation regulation during microbial infections,^117-119 as well as their importance in the innate immune response to Staphylococcus aureus (S. aureus),^{120, 121} researchers have attempted to engineer CAR-M cells targeting S. aureus-specific antigens to enhance their phagocytic and bactericidal capabilities against this pathogen. For instance, Li et al. developed a nanoparticle coating that enables in vivo delivery of a CAR targeting S. aureus surface protein A (SasA) and caspase-11 shRNA, ultimately leading to the localised induction of CAR-M cells in the prosthetic microenvironment¹²² (Figure 3I). These anti-SasA CAR-M cells exhibited dual functionalities: targeting and phagocytosing S. aureus through the expression of anti-SasA scFv, while suppressing caspase-11 expression, which allowed mitochondrial recruitment around phagosomes to produce reactive oxygen species for bacterial killing. Similarly, Tang et al. designed lipid nanoparticles modified with sequence CRVLRSGSC (CRV) peptides to encapsulate and deliver SasA-CAR mRNA and caspase-11 siRNA into macrophages, effectively generating CAR-M cells with potent bactericidal activity.¹²³

2.3 Fungal infections

Fungal infections annually cause around 11.5 million life-threatening cases and over 1.5 million deaths,¹²⁴ posing a significant threat to public health. At present, antifungal drugs targeting fungal pathogens, including azoles, echinocandins, polyenes, allylamines and antimetabolites, remain the first-line treatment for infected patients.¹²⁵ However, due to the continuous emergence of resistance, these drugs are becoming increasingly ineffective at treating the growing number of fungal infections, with the mortality rate for invasive fungal infections remaining high.¹²⁶ The prospects of CAR-T therapy in treating fungal infections have been explored in preliminary studies. Kumaresan et al. engineered CAR-T cells to incorporate the pattern recognition receptor Dectin-1, which can trigger T cell activation through chimeric CD28 and CD3-ζ upon engagement with carbohydrates present in the cell wall of Aspergillus fumigatus (Af) germlings (Figure 3J), ultimately inhibiting the growth of Af both in vitro and in vivo.¹²⁷ Another study designed and constructed Af-specific chimeric antigen receptor (Af-CAR) T cells targeting a conserved protein antigen in the cell wall of Af (Figure 3K), demonstrating their antifungal activity in preclinical models. These Af-CAR-expressing T cells can recognise Af strains and directly target Af hyphae with antifungal effects.¹²⁸ Additionally, CAR-T cells engineered to target glucuronoxylomannan (GXM) (Figure 3L) have shown their potential to recognise both Cryptococcus neoformans and Cryptococcus gattii.¹²⁹

3.1 Systemic lupus erythematosus

A hallmark of systemic lupus erythematosus (SLE) is the generation of antinuclear antibodies (ANAs), which form immune complexes, activate complement and promote the release of inflammatory cytokines, leading the immune system to mistakenly target the body's healthy tissues and cells.¹³⁵ B cells play a crucial role in the development of SLE due to their secretion of autoantibodies and pro-inflammatory cytokines, making them a key therapeutic target in the disease.¹³⁶ CD19 is a key regulator of B-cell differentiation, antibody production and memory formation, and is expressed on all mature B cells, memory B cells and plasmablasts, as well as on certain pre-B cells and plasma cells.¹³⁷ Utilising CD19 for deep B-cell depletion could reset the immune system, potentially leading to long-lasting therapeutic effects in autoimmune diseases, including SLE.¹³² Previous studies have offered compelling evidence supporting that CD19-targeted CAR-T cell therapies are both effective and safe in SLE-related animal models.^{138, 139} Even more excitingly, the first case of an SLE patient being treated with CD19 CAR-T cells was documented in 2021.¹⁴⁰ Post-treatment, the patient's autoantibodies disappeared, their condition improved, and no significant side effects were observed. Subsequently, the same research team reported results in 2022 for an additional five refractory SLE patients receiving treatment with CD19 CAR-T cells.¹⁴¹ The results showed that all patients reached remission of SLE 3 months after treatment, and during follow-up periods of up to 17 months, drug-free remission was maintained even after B cells reappeared. Most recently, this group published a follow-up case series.¹⁴² This study tracked 15 patients treated with CD19 CAR-T cells between February 2021 and May 2023, including those with severe SLE, systemic sclerosis (SSc) and idiopathic inflammatory myositis, some of whom had been previously reported. The results indicated that all patients experienced disease improvement and tolerated the treatment well, suggesting that CAR-T therapy represents a viable option for these three distinct autoimmune diseases. The significantly increased expression of B-cell maturation antigen (BCMA) across all B-cell subsets in SLE patients makes it a candidate therapeutic target.¹⁴³ In another small phase I study, the CD19/BCMA dual-target CAR-T cell therapy (Figure 4A) developed by iCell Gene Therapeutics safely induced drug-free remission in SLE patients.¹⁴⁴ Of particular note is that the first paediatric SLE patient undergoing CD19-targeted CAR-T therapy was described in 2024.¹⁴⁵ This patient, who had early-onset and rapidly progressive lupus nephritis, experienced a rapid decline in SLE disease activity, resolution of arthritis symptoms and improvement in renal function following CAR-T therapy. These findings suggest that paediatric SLE patients may also benefit from CAR-T therapy, highlighting the urgent need for further clinical trials.

The deficiency of Treg cells may play a critical role in the pathogenesis of SLE,¹⁴⁶ while antigen-specific Treg cells have shown greater efficacy and safety than polyclonal Tregs in preclinical studies.¹⁴⁷ These findings collectively provide a theoretical foundation for CAR-Treg cell therapy in SLE. Building on this foundation, Doglio et al. developed Treg cells co-expressing an anti-CD19 CAR and FoxP3 (Fox19CAR-Tregs) for the treatment of SLE.¹⁴⁸ In vitro experiments demonstrated that these Fox19CAR-Tregs effectively suppressed B-cell proliferation and activation, while in a humanised mouse model of SLE, they reduced autoantibody production and promoted the restoration of immune homeostasis.

Although CAR-T therapy has demonstrated significant efficacy in autoimmune diseases and severe toxicities are rarely reported in treated patients, the limited number of autoimmune patients treated to date necessitates cautious interpretation of the available safety data.¹⁴⁹ In contrast, CAR-NK cells exhibit clear safety advantages, with early clinical trials reporting minimal severe side effects like cytokine release syndrome (CRS) or neurotoxicity.^{31, 150} Additionally, sources such as umbilical cord blood, stem cells and cell lines can be used to generate CAR-NK cells, not only reducing manufacturing costs but also enhancing clinical feasibility.^{151, 152} Leveraging these advantages, researchers have started investigating CAR-NK therapy as a treatment option for SLE. A research team in the United States engineered NK-92 cells, a human NK cell line, to express a CAR construct containing the extracellular domain of PD-L1.¹⁵³ These modified CAR-NK cells were designed to target and eliminate PD-1-expressing follicular helper T (TFH) cells, which have been implicated in the pathogenesis of various autoimmune diseases.¹⁵⁴ In both in vitro studies and lupus-like humanised mouse models, PD-L1-CAR-NK cells selectively eliminated PD-1-high CD4+ T cells, including TFH cells, leading to a reduction in memory B-cell proliferation, differentiation and Ig antibody secretion.

3.2 Rheumatoid arthritis

Rheumatoid arthritis (RA) is a chronic, systemic autoimmune disease that mainly impacts joints and soft tissues surrounding them.¹⁵⁵ B cells are pivotal in the development of RA, and rituximab, a monoclonal antibody that targets CD20 antigen and results in the depletion of B cells, can significantly improve the symptoms and signs in RA patients. However, rituximab is linked to an elevated risk of infection and requires long-term treatment cycles.¹⁵⁶ Zhang et al. found that CAR-T cells engineered to target FITC (Figure 4B) can selectively eliminate hybridoma cells derived from antigen peptide immunisation, as well as autoreactive B cells from RA patients, through the recognition of corresponding FITC-tagged citrullinated peptide epitopes.¹⁵⁷ Recently, a clinical trial employing CD19-targeted CAR-T cells for RA treatment demonstrated that, despite early adverse events, the patient achieved drug-free remission by day 100 without any neurological sequelae.¹⁵⁸

3.3 Systemic sclerosis

SSc is an uncommon and complex autoimmune disorder marked by progressive connective tissue fibrosis, microvascular damage, innate and adaptive immune dysregulation, and multi-organ or systemic fibrosis.¹⁵⁹ At present, there is no cure for SSc, and treatment primarily aims to alleviate symptoms, reduce functional disability and improve quality of life.¹⁶⁰ The imbalance of B cells in SSc patients, along with the observed improvements in lung function and skin fibrosis with rituximab treatment, suggests that targeting CD19 with CAR-T cells might be a useful strategy.¹⁶¹ In May 2023, Georg Schett's team administered CD19 CAR-T therapy to an individual diagnosed with severe, refractory SSc involving the skin, joints, heart and lungs. Following treatment, improvements were observed in the patient's skin hardening, cardiac and joint damage and serological markers.¹⁶² Subsequently, the results of two clinical studies further supported the therapeutic potential and safety of CD19-targeted CAR-T therapy in SSc.^{163, 164} Both studies demonstrated that patients with SSc treated with CD19-targeted CAR-T cells achieved disease remission, with no severe adverse events observed.

3.4 Type 1 diabetes

Type 1 diabetes (T1D) is a chronic autoimmune disease where pancreatic β-cells are destroyed, causing a shortage of insulin.¹⁶⁵ Initial studies have suggested the potential practicality of CAR-based therapies for treating T1D. In previous research, Zhang et al. produced a monoclonal antibody named mAb287, which selectively recognises complexes formed by the major autoantigen insulin peptide (B:9-23) and major histocompatibility complex (MHC) class II molecules (I-A^g7) in register 3 (R3) in non-obese diabetic (NOD) mice.¹⁶⁶ Through this mechanism, mAb287 blocks the binding of I-A^g7-B:9-23(R3) complexes to T cells, thereby inhibiting the activation of insulin-specific CD4 T cells and significantly delaying the onset of diabetes, without affecting the presentation of other antigens. Subsequently, the team further introduced this antibody into a CAR structure, creating cytotoxic 287-CAR-T cells (Figure 4C). These cells can selectively bind to specific I-A^g7-B:9-23(R3) complexes and eliminate the corresponding antigen-presenting cells, thereby delaying the development of T1D.¹⁶⁷ Additionally, Kobayashi et al. described a first-generation biomimetic five-module chimeric antigen receptor (^5MCAR) aimed at targeting autoimmune CD4+ T cells in a diabetic mouse model. This biomimetic ^5MCAR comprises a chimeric receptor module, three cluster of differentiation (CD) signalling modules and a surrogate co-receptor module (Figure 4D). The chimeric receptor module is constructed by fusing the antigen targeted by pathogenic T cells with T-cell receptor (TCR)-associated elements, while the surrogate co-receptor module is formed by the fusion of CD80 and Lck to enhance signalling transduction. The results showed that ^5MCAR-T cells were capable of both preventing and alleviating T1D in NOD mice and persisting in vivo for an extended period.¹⁶⁸

The functional impairment of Tregs in patients with T1D suggests their potential involvement in the pathogenesis of the disease.^{169, 170} Adoptive transfer of Tregs is considered a potential therapeutic strategy^171-173; however, studies have shown that islet antigen-specific Tregs are significantly more effective than polyclonal Tregs in preventing or delaying disease progression.^174-177 Nevertheless, due to the limited availability of antigen-specific Tregs under autoimmune conditions, their isolation and expansion pose significant challenges. CAR engineering offers an ideal solution to overcome this limitation. Tenspolde et al. proposed the first CAR-Treg designed to improve and treat T1D by transducing a second-generation CAR and the FOXP3 gene into CD4+ T cells, resulting in insulin-specific CAR-Treg cells¹⁷⁸ (Figure 4E). These insulin-specific CAR-Treg cells exhibited a phenotype similar to that of natural Treg cells and effectively suppressed antigen-specific T cells in vitro. Unfortunately, their infusion into NOD mice did not prevent the onset of diabetes. This lack of efficacy may be related to the secretion form and metabolic rate of insulin in vivo, suggesting that selecting more appropriate targets may be necessary. In another study, Spanier et al. developed CAR-Treg cells targeting the insulin B chain 10–23 peptide presented by the I-Ag7 MHC II allele and demonstrated that these cells effectively prevented disease onset in a mouse model of diabetes.¹⁷⁹

3.5 Immune thrombocytopenia

Immune thrombocytopenia (ITP) is an autoimmune bleeding disorder characterised by a reduced platelet count in peripheral blood, mainly caused by enhanced platelet clearance and reduced platelet production.¹⁸⁰ Among the underlying mechanisms, B-cell-derived autoantibody-mediated platelet destruction is considered a key driver of ITP pathogenesis.¹⁸¹ In July 2024, the first case of SLE-associated ITP managed using CD19-targeted CAR-T cell therapy was reported.¹⁸² The patient, diagnosed with SLE, presented with persistent thrombocytopenia. After receiving CAR-T cell therapy, the patient showed a substantial rise in platelet count alongside a notable decrease in ANA titres. This case suggests that CD19-targeted CAR-T therapy may be effective in treating SLE-associated ITP and further implies its potential application in primary ITP. This hypothesis has recently received preliminary validation. A German research team reported a case of primary ITP treated with CD19-targeted CAR-T cells.¹⁸³ After treatment, the patient's platelet count increased significantly, and no recurrence of ITP was observed for approximately 3 months after discontinuation of all therapies, with platelet counts remaining within the normal range.

3.6 Inflammatory bowel disease

Inflammatory bowel disease (IBD) represents a category of progressive, chronic inflammatory conditions affecting the gastrointestinal tract, classified as autoimmune disorders, with the main types being Crohn's disease and ulcerative colitis.^{184, 185} Over the past decades, the incidence of IBD has risen sharply, with a trend towards an earlier age of onset.¹⁸⁶ Unlike most autoimmune diseases, the therapeutic potential of CD19-targeted CAR-T therapy in IBD remains unclear, which may be related to the fact that the positive role of CD20 antibodies in IBD has not yet been proven.¹⁸⁷ However, given that Treg cell dysfunction is considered a key mechanism in IBD pathogenesis,^{188, 189} researchers have actively explored the potential of CAR-Tregs in IBD and have made preliminary progress. As early as 2008, Elinav et al. demonstrated that CAR-Tregs specifically targeting 2,4,6-trinitrophenol (TNP) could be activated by exogenous TNP in a 2,4,6-trinitrobenzenesulfonic acid (TNBS)-induced colitis mouse model, accumulate at colitis lesion sites and ameliorate intestinal inflammation by suppressing effector T cells.¹⁹⁰ Given the elevated expression of carcinoembryonic antigen (CEA) in human colitis and colorectal cancer, Blat et al. designed CAR-Tregs targeting CEA and validated their ability to reduce colitis severity and inhibit colorectal cancer development in mouse models.¹⁹¹ Furthermore, Cui et al. developed CAR-Tregs targeting the interleukin-23 receptor (IL23R).¹⁹² Their findings showed that these IL23R-CAR-Tregs could migrate to IL23R-expressing tissues in humanised mice and be activated in vitro by cells derived from intestinal biopsies of Crohn's disease patients.

3.7 Neuromyelitis optica spectrum disorder

Neuromyelitis optica spectrum disorder (NMOSD) is an uncommon autoimmune disease of the central nervous system that often results in severe sequelae or even death.¹⁹³ Its characteristic features include recurrent acute episodes and the detection of antibodies specific to the AQP4 protein, namely, AQP4-IgG.¹⁹³ B cells contribute to the pathogenesis of NMOSD through mechanisms such as maturing into plasmablasts and plasma cells that produce pathogenic AQP4-IgG, secreting pro-inflammatory cytokines like IL-6 that enhance inflammatory immune responses, and functioning as antigen-presenting cells to activate autoimmune T cells.¹⁹⁴ Given the excellent performance of CAR-T therapy in eliminating B cells, it is being considered for the treatment of NMOSD. Qin et al. reported a preliminary clinical study using CAR-T cells targeting BCMA in a cohort of 12 patients with recurrent or refractory AQP4-IgG seropositive NMOSD. The results indicated that CAR-T cell therapy was well tolerated and exhibited a favourable safety profile, with 11 of 12 patients remaining relapse-free over a median follow-up period of 5.5 months. Considerable improvements in functional scores and quality of life were also observed.¹⁹⁵

3.8 Myasthenia gravis

Myasthenia gravis (MG) is an autoimmune disease that impairs neuromuscular junction (NMJ) transmission, primarily driven by antibodies targeting acetylcholine receptors (AChR), muscle-specific kinase (MuSK) or other AChR-associated proteins expressed on the postsynaptic muscle membrane.¹⁹⁶ Recently, Granit's research team utilised RNA transfection to develop CAR proteins targeting BCMA, which can be continuously expressed over the course of a week for the treatment of MG, aiming to reduce the inherent risks of conventional CAR-T cell therapy. The study concluded that this therapy significantly improved the disease condition in MG patients without causing serious adverse reactions.¹⁹⁷ The 12-month follow-up results of patients receiving this treatment were further reported in a preprint.¹⁹⁸ The data showed that five out of seven participants maintained clinical improvement at 12 months.

3.9 Pemphigus vulgaris

Pemphigus is a chronic autoimmune blistering disease that poses a life-threatening risk, primarily due to autoantibodies targeting desmoglein (DSG) 1 and DSG3.¹⁹⁹ Pemphigus vulgaris (PV) is the predominant type of pemphigus, representing more than 80% of cases.²⁰⁰ A previous study demonstrated that DSG3 chimeric autoantibody receptor T (DSG3-CAART) cells (Figure 4F), which are engineered T cells expressing a chimeric autoantibody receptor composed of the PV autoantigen DSG3 and CD137-CD3ζ signalling domains, specifically targeted B lymphocytes expressing DSG3 B-cell receptors (BCRs) in vitro and showed expansion and persistence in vivo.²⁰¹ Further research by Lee et al. found that DSG3-CAART cells could specifically lyse anti-DSG3 human B cells from PV patients. In an active immune model of PV with physiological levels of anti-DSG3 IgG, these DSG3-CAART cells effectively suppressed antibody responses targeting pathogenic DSG3 epitopes and blocked the binding of autoantibodies to epithelial tissues, ultimately achieving clinical and histological resolution of blisters.²⁰²

3.10 Other autoimmune diseases

In antisynthetase syndrome (ASS), Georg Schett's team reported on a patient who underwent CD19 CAR-T therapy after failing to respond to multiple immunosuppressants and biologics. After treatment, the patient's condition significantly improved, with nearly complete resolution of pulmonary, joint and muscle inflammation, and gradual recovery of muscle strength and endurance.²⁰³ Subsequently, Pecher et al. reported another successful case of ASS managed with CD19 CAR-T therapy.²⁰⁴ In autoimmune encephalitis, Reincke et al. developed N-methyl-D-aspartate receptor (NMDAR)-CAART cells (Figure 4G) to selectively eliminate anti-NMDAR B cells and pathogenic autoantibodies.²⁰⁵ In juvenile dermatomyositis, a case report published by Nicolai et al. demonstrated that the patient achieved significant clinical remission starting at week 4 after receiving a single dose of CD19-targeted CAR-T cells, and this remission persisted even after B-cell recovery.²⁰⁶ In addition, recent advancements have been made in applying CAR-T therapy for multiple sclerosis (MS) treatment. Kyverna Therapeutics first reported the use of KYV-101, a fully human CD19 CAR-T therapy, in two patients with progressive MS.²⁰⁷ Following treatment, one patient showed a significant reduction in intrathecal antibody production in the cerebrospinal fluid, and neither patient exhibited severe adverse reactions. To further collect data, Kyverna Therapeutics is conducting clinical trials independently or in collaboration (NCT06138132, NCT06384976). CAR-Tregs are now being investigated for their potential in treating autoimmune skin disorders, with vitiligo being a notable example. Mukhatayev et al. explored the use of GD3-specific Treg cells (Figure 4H) for the treatment of vitiligo,²⁰⁸ based on previous studies demonstrating that ganglioside D3 (GD3) is upregulated in perilesional epidermal cells, including melanocytes.²⁰⁹ They found that GD3-specific Treg cells exhibited enhanced IL-10 secretion upon antigen stimulation, effectively controlled cytotoxic responses against melanocytes and significantly delayed depigmentation in a spontaneous vitiligo mouse model.

10.1 Efficacy and safety

Similar to its application in cancer treatment, the durability and effectiveness of CAR-T treatment for non-oncological diseases remain significant challenges. Data show that 40%–60% of patients with haematological malignancies experience disease relapse after CAR-T therapy,²⁷⁷ potentially due to the gradual exhaustion of these engineered cells in vivo, which weakens their antitumour activity.²⁷⁸ Likewise, such exhaustion may compromise the effectiveness of CAR-T treatment for non-oncological diseases, raising the risk of disease relapse or progression. Furthermore, since CAR-T therapy typically targets specific antigens, it may face challenges in addressing disease heterogeneity, potentially leading to incomplete clearance of pathogenic cell populations and making disease remission difficult or prone to relapse.^{278, 279} To overcome these challenges, researchers have proposed various innovative strategies to improve the persistence and effectiveness of CAR-T cells.^280-282 However, the clinical feasibility and long-term safety of these approaches require further in-depth investigation and validation.

One major concern regarding CAR-T therapy is its safety. Patients undergoing CAR-T therapy still face some risks, such as CRS,²⁸³ neurotoxicity,²⁸⁴ and increased susceptibility to infections.²⁸⁵ Existing data suggest that severe toxicities are rare in autoimmune disease patients treated with CAR-T cells, possibly due to their lower B-cell burden compared to those with B-cell malignancies.¹⁴⁹ However, the limited data are insufficient to fully confirm the safety of CAR-T therapy, making the improvement of its safety a primary focus. Controlling toxicity by enabling transient CAR expression through mRNA gene transfer appears to be an effective strategy.¹⁹⁷ Another promising approach is the localised delivery of CAR-T cells to concentrate them at the target site and limit their cytotoxic range, thus enhancing therapeutic outcomes while minimising potential toxicity.²⁸⁶ Furthermore, the therapeutic index of CAR-T cells continues to be an important research priority, as precise dose control is vital to addressing safety concerns.²⁸⁷

10.2 Target selection

Target selection is a key factor influencing both the effectiveness and safety of CAR-T therapy. Currently, CD19 is predominantly used as a single target in CAR-T therapy for most autoimmune disease patients. CD19 is highly specific to the B-cell lineage and widely present throughout various stages of B-cell differentiation.¹³² In contrast, although CD20 has shown potential in immunotherapy for autoimmune diseases (e.g., rituximab), it is nearly absent in plasmablasts and plasma cells.²⁸⁸ BCMA and CD38 are two additional potential targets for autoimmune diseases. BCMA is mainly found on plasma cells and a subset of memory B cells, positioning it as a candidate target for antibody-driven autoimmune diseases.^{195, 197, 289} CD38, on the other hand, is expressed on multiple immune cell populations, such as T cells, NK cells, plasma cells and early B cells.²⁹⁰ Although monoclonal antibodies targeting CD38 have shown benefits in SLE, its safety as a CAR-T target requires further evaluation.²⁹¹ In addition, multi-target strategies are being actively explored. For example, dual-target CAR-T cells directed against both CD19 and BCMA have been developed and preliminarily validated for efficacy in SLE patients, though further research is required to evaluate their long-term efficacy and safety.¹⁴⁴ In non-oncological diseases other than autoimmune diseases, current research primarily focuses on targeting single disease-related antigens. The single-target strategy offers advantages such as relatively simple design and mature technology, but it carries the risk of antigen escape. In contrast, multi-target CAR-T therapy may have potential advantages in enhancing efficacy and reducing the risk of relapse; however, its manufacturing process is more complex, and its long-term safety and efficacy still require further validation.

10.3 Lymphodepletion

Before using CAR-T therapy for cancer, lymphodepletion is generally considered necessary to enhance CAR-T cell efficacy^292-294 through mechanisms such as the removal of Treg cells, increased availability of serum cytokines and elimination of bone marrow-derived suppressor cells.^295-297 Pre-treatment lymphodepletion may provide temporary benefits in CAR-T therapy for autoimmune conditions by reducing immune cell counts, including B cells.^{132, 141} Currently, the most widely used lymphodepletion regimen in B-cell malignancies is a combination of cyclophosphamide (.75–1.5 g/m²) and fludarabine (75–90 mg/m²).²⁹⁸ To date, all patients with autoimmune diseases who have undergone CAR-T therapy have received lymphodepletion based on fludarabine and/or cyclophosphamide.¹⁴⁹ However, the optimal lymphodepletion strategy for autoimmune diseases remains to be defined. A previous study suggested that lower doses of cyclophosphamide and fludarabine did not significantly impair CAR-T cell efficacy in SLE,²⁹⁹ indicating the potential for dose reduction. This could help mitigate treatment-related toxicity while maintaining therapeutic efficacy. Furthermore, optimising lymphodepletion regimens involves not only dose adjustments but also the selection of alternative agents. For example, bendamustine has been proposed for lymphodepletion prior to CAR-T therapy to reduce inflammatory responses and lower toxicity risks.³⁰⁰ Future studies should further investigate how different lymphodepletion regimens influence both the initial and prolonged effectiveness of CAR-T therapy to determine the optimal strategy.

10.4 Costs

A significant challenge associated with CAR-T therapy is its high cost. These costs primarily stem from the preparation of CAR-T cells, outpatient or inpatient care and potential treatments for adverse events.³⁰¹ Clearly, the high cost of CAR-T therapy limits its accessibility in low-income regions. Moreover, many low-income and underdeveloped areas face shortages of medical resources, including trained personnel and infrastructure, further constraining the application of CAR-T therapy in these regions. Automating and streamlining the autologous CAR-T cell production process, as well as reducing the time required for manufacturing, is one approach to lowering its costs. Another potential avenue is the in vivo targeting and reprogramming of T cells to express CARs, eliminating the need for ex vivo manufacturing. The in vivo editing approach not only removes the reliance on specialised facilities and personnel required for ex vivo manufacturing but also reduces the demand for cold-chain systems and resources during transportation and storage. Beyond these two strategies, universal CAR-T therapy seems to be another promising solution for reducing costs and improving accessibility in low-income regions. Universal CAR-T therapy involves the batch production of T cells derived from healthy donors, enabling long-term storage and rapid distribution. This approach also minimises the dependence on specialised facilities and skilled personnel, making it particularly suitable for resource-limited settings. Notably, CAR-T therapy, if capable of providing long-term or permanent remission, may be cost-effective for chronic disease management compared to the cumulative expenses of long-term drug therapy.

10.5 Patient selection and treatment criteria

The patient criteria, treatment prerequisites and timing of CAR-T therapy may vary across different disease conditions. For instance, in some viral infections, such as HBV infection in the liver, the widespread distribution of the virus in specific organs may lead to large-scale killing of infected cells by CAR-T cells, potentially causing acute or irreversible organ damage.³⁰² In the treatment of certain diseases, such as cardiac diseases, a comprehensive assessment may be required before administering CAR-T therapy to determine whether the disease has reached an irreversible stage; otherwise, the treatment may no longer be beneficial and could even become counterproductive.²⁸⁷ Moreover, it is crucial to assess how much benefit patients can derive from CAR-T cell therapy compared to existing treatment options. The conditions, timing and position of CAR-T cell therapy in managing different diseases still require further clarification through experimental and clinical research.

10.6 Ethics considerations

CAR-T therapy presents a range of ethical challenges, including treatment fairness, patient informed consent, ethical concerns regarding donor cells and challenges in clinical trials.³⁰³ Due to the high dependency on specialised equipment and personnel, the production and transportation of CAR-T products remain limited, meaning that patients often have to wait a long time or may even miss the optimal window for treatment.^{304, 305} Although there has been some discussion on prioritising patients,^304-306 determining how to more reasonably ensure fairness in treatment remains a key ethical challenge in CAR-T therapy. Additionally, the potential adverse effects and uncertainty of treatment outcomes also pose significant risks for patients. One important ethical challenge is whether clinical researchers and physicians adequately inform research participants and patients about these risks and ensure that they make decisions based on a full understanding. Universal CAR-T therapy, which utilises T cells from healthy donors, raises further ethical questions regarding donor consent and awareness. It is essential to ensure that donors are fully informed about the intended use of their cells and the possibility of their involvement in commercial applications. Furthermore, the risk of genetic information from donor cells being leaked during research or commercial use introduces significant privacy and data security concerns. To address these ethical challenges, a multifaceted approach is required, including robust policy interventions, stricter regulation of the informed consent process and the establishment of comprehensive mechanisms to safeguard donor privacy and manage data security effectively.

10.7 Combination strategies

Significant advancements in cancer treatment have been achieved by integrating CAR-T cell therapy with other therapeutic strategies. These combination approaches aim to boost the therapeutic performance of CAR-T cells while mitigating associated toxicity, thus broadening the therapeutic scope and improving safety. For example, combining CAR-T therapy with immune checkpoint inhibitors has been found to improve the effectiveness of CAR-T treatments to some degree while maintaining manageable toxicity levels.^{307, 308} Furthermore, the combination of CAR-T therapy with monoclonal antibodies targeting cytokines involved in CRS or ICANS has also shown potential in mitigating CAR-T cell toxicity.³⁰⁹ Although combination strategies involving CAR-T therapy and other treatments for non-oncological diseases are still in their infancy, some studies provide valuable insights that encourage further exploration of CAR-T-based combination therapies. One such study indicated that combining CAR-T cells with Myrcludex B can achieve long-term control of HBV infection in both cell and animal models.³¹⁰ Another study has shown that combining CD19-targeted CAR-T therapy with Nintedanib and Mycophenolate can lead to sustained improvements in progressive pulmonary fibrosis in patients with SSc.³¹¹ The potential benefits of combining CAR-T therapy with other drugs (such as immunomodulators, anti-fibrotic agents and anti-inflammatory drugs) to enhance treatment outcomes in various disease conditions still require further investigation. Notably, to improve the safety of CAR-T treatment in non-cancer diseases, lessons from cancer-related studies can be applied, such as combining CAR-T therapy with monoclonal antibodies or small molecules targeting cytokines associated with CRS or ICANS to manage related toxicity.

10.8 The potential of different CAR therapies

For non-oncological diseases, different types of CAR-engineered cells each have their own advantages and limitations. CAR-T cells have demonstrated potent antigen-specific clearance capabilities in various non-oncological diseases, but their application is constrained by high production costs,^{301, 312} long manufacturing times,³¹³ and severe adverse reactions.³¹⁴ Although the development of universal CAR-T cells presents a possible way to overcome certain challenges, their widespread use still requires addressing issues related to allogeneic cell transplantation, such as GVHD and host immune rejection.^{315, 316} In contrast, CAR-NK cells, as cytotoxic lymphocytes, exhibit unique advantages in infectious diseases and autoimmune diseases due to their innate antiviral activity and ability to eliminate target cells. Additionally, CAR-NK cells are associated with lower toxicity and possess “off-the-shelf” potential, as they can be derived from umbilical cord blood or cell lines.^{31, 150-152} However, the relatively short lifespan of NK cells, while enhancing the safety of CAR-NK therapy, may necessitate repeated administrations to achieve long-lasting responses.^{317, 318} While the role of macrophages in host defence and their exceptional tissue infiltration capabilities^{52, 118} make CAR-M cells a promising candidate for infectious and cardiac diseases, their persistence and stability require further investigation. CAR-Treg cells, as an emerging strategy, can suppress overactivated immune responses by enhancing the function of antigen-specific Tregs,⁴⁰ offering a novel approach to inducing immune tolerance, particularly in organ transplant rejection and autoimmune diseases. Nevertheless, it remains uncertain whether CAR-Treg cells might induce adverse reactions similar to those of CAR-T cells, and further evaluation is needed to determine whether the inclusion of the CD28 co-stimulatory domain could lead to CAR-Treg exhaustion.³¹³ In brief, these CAR therapies, derived from different cell types, exhibit diverse applications due to their distinct immunological properties, potentially offering more precise and effective treatment strategies for patients.

10.9 Future research directions

To fully harness the potential of CAR therapies across various disease contexts, future research should focus on addressing current limitations and exploring innovative strategies for improvement. One key area is enhancing the safety of CAR therapies in both temporal and spatial dimensions through advanced CAR designs, such as various molecular switches, logic gates and transient CAR expression systems.^{319, 320} These approaches could minimise long-term risks like off-target effects and CRS while maintaining therapeutic efficacy. Another critical field requiring further exploration is the development of universal CAR-engineered therapies. By employing gene-editing technologies to modify one or more genes in allogeneic T cells, it is possible to effectively eliminate GVHD and immunogenicity, enabling these universal CAR-T cells to serve as “off-the-shelf” products that can be broadly applied to diverse patients without the need for time-consuming, individualised manufacturing processes.³²¹ Furthermore, combining CAR therapies with other treatment modalities in non-cancer diseases remains a promising yet underexplored area. Such combination approaches have the potential to overcome the limitations and resistance associated with monotherapies but also introduce additional complexities. Future research should focus on uncovering the synergistic mechanisms between CAR-engineered therapies and other treatments, as well as identifying optimal combination strategies to enable their broader application in non-cancer contexts.

To summarise, CAR-engineered cell therapy has not only achieved significant breakthroughs in cancer treatment but is also gradually emerging as a novel approach for addressing complex non-tumour diseases (Figure 6). With ongoing research, we can anticipate that CAR-engineered cell therapy will continue to expand its applications, offering more options for the clinical treatment of various diseases. Looking ahead, through continuous optimisation and innovation, CAR-engineered cell therapy is expected to overcome current technical and clinical challenges, achieve higher efficacy and safety, and ultimately benefit a broader range of patients.