CAR T Cell Therapy for Rheumatoid Arthritis

Abstract

Chimeric Antigen Receptor (CAR) T cell therapy has revolutionised the treatment of relapsed/refractory B cell leukaemia, lymphoma and multiple myeloma through targeting of CD19 and BCMA antigens on the surface of these cells. A growing body of evidence has recently demonstrated that these cell-based therapies can also target autoimmune diseases including systemic lupus erythematosus, systemic sclerosis, neuromyelitis optica spectrum disorder, myasthenia gravis, idiopathic inflammatory myositis, multiple sclerosis and rheumatoid arthritis. To date, ten patients with rheumatoid arthritis have been treated with CAR T cells targeting CD19/CD20 or BCMA antigens on B cells. Nine patients with seropositive disease have shown remarkable responses, including depletion of circulating B cells, ablation of autoantibody levels and drug-free remission. A tenth patient with seronegative disease initially responded to CAR T cell therapy but later relapsed. This review provides in-depth analysis of these single case studies and highlights emerging in-vitro and animal model studies where T cell subsets have been engineered with CARs to fine-tune their immune responses for the treatment of rheumatoid arthritis, including targeting of autoreactive B cells, autoreactive T cells or fibroblasts. CAR T cell therapy holds enormous promise for the treatment of difficult-to-treat rheumatoid arthritis, but more research and large clinical trials are needed to confirm its efficacy and safety.

Introduction

Rheumatoid Arthritis (RA) is a chronic inflammatory disease that affects the joints of affected individuals but is increasing recognised as a systemic inflammatory disorder that impacts the wider vasculature, metabolic function and cognition [1]. Prevailing evidence strongly indicates that the pathogenesis of RA is driven by genetic predisposition that increases the autoreactivity of the host immune response for modified self-antigens that arise due to environmental influences [1]. The most important genetic risk factor for developing RA resides in the Major Histocompatability Complex class II complex comprising the highly polymorphic Human Leukocyte Antigens (HLA) DR, DP and DQ genes encoding cell surface proteins that present peptide antigens to CD4 + T cells. More specifically, disease-associated alleles within this HLA locus encode a conserved sequence of five amino acids, termed the ‘shared epitope’, that are thought to be involved in the presentation of post-translationally modified self-antigens, such as citrullinated peptides, to autoreactive CD4 + T cells via T cell receptor (TCR) signalling [2]. Autoreactive B cells may also recognise these modified antigens through their B cell receptor and present them to autoreactive T cells, driving T cell: B cell collaboration and the production of anti-citrullinated peptide antibodies (ACPAs) that perpetuate inflammatory signalling cascades via crosstalk with tissue-resident stromal cells and immune cells within the joints [1]. This form of the disease with the presence of ACPAs in serum is known as seropositive RA. The accumulation of citrullinated peptides in RA patients is strongly linked with smoking and therefore provides an intriguing insight into how host genetics and environmental factors may shape the pathogenesis of this inflammatory disease [3]. While T cells and B cells are key drivers of inflammatory cascades in RA, several other immune cells (primarily macrophages and neutrophils) and non-immune cells including fibroblast-like synoviocytes and macrophage-like synoviocytes significantly contribute to disease pathogenesis [4]. While seropositive RA is the dominant form of the disease, around 20–30% of RA patients do not have ACPAs in serum and are designated as seronegative [5]. RA is therefore a heterogenous disease with diverse and complex cellular crosstalk between immune and non-immune cells leading to tissue destruction in the joint.

As there is currently no cure for RA, therapeutic interventions are focused on decreasing inflammatory pathways though various mechanisms (Table 1). Current treatments include conventional synthetic Disease Modifying Anti-Rheumatic Drugs (csDMARDs) and glucocorticoids that have broad acting and poorly defined mechanisms of action, in addition to biological DMARDs (bDMARDs) and targeted synthetic DMARDs (tsDMARDs) that target more defined inflammatory pathways. For bDMARDs and tsDMARDs, their mechanism of action includes inhibition of TNFα signalling, inhibition of T cell co-stimulation, CD20-mediated B cell/plasma cell depletion, modulation of IL-6 signalling and inhibition of cytokine-mediated JAK-STAT signalling. Therapeutic regimens are guided by the American College of Rheumatology [6] and the European Alliance of Associations for Rheumatology [7] in the USA and Europe, respectively. Despite the plethora of available therapies, many patients do not adequately respond to first-line treatments and undergo several cycles of trial and error with next-in-line therapeutic options which is likely due to the heterogeneity of the disease. Furthermore, many patients lose response over time and/or do not achieve clinical remission with any of the currently available therapies [8]. This latter cohort are defined as ‘difficult-to-treat’. Overall, drug-free remission in RA is rare [9]. Chimeric Antigen Receptor (CAR) T cell therapy is a clinically approved treatment for different forms of B cell leukaemia, lymphoma and myeloma, and their success in targeting malignant B cells has led to their exploration as a potential therapeutic strategy for difficult-to-treat RA via depletion of B cells (key drivers of inflammation in RA) as well as other immune cell types and tissues.

Table 1.

Current treatments for RA. Therapeutic indications were sourced from the FDA and EMA drug labels (accessed February 2024), as well as the American college of rheumatology guideline for the treatment of RA [6] and the EULAR recommendations for the management of rheumatoid arthritis with synthetic and biological disease- modifying antirheumatic drugs [7]

Treatment	Mechanism of Action
Conventional DMARDs  ▪ MTX (low dose)  ▪ Sulfasalazine  ▪ Leflunomide  ▪ Hydroxychloroquine	Unknown (possibly inhibition of immune cell proliferation)
Glucocorticoids	Broad immunosuppression
Biological DMARDs  ▪ TNFα antagonists - Infliximab (+ MTX) - Etanercept (+/- MTX) - Adalimumab (+/- MTX or non-biological DMARDs) - Certolizumab pegol (+/- MTX) - Golimumab (+ MTX)  ▪ IL-1α/β antagonists - Anakinra (+/- MTX or *DMARDs other than TNFα antagonists)  ▪ IL-6 receptor antagonists - Tocilizumab (+/- MTX or *non-biological DMARDs) - Sarilumab (+/- MTX or *other conventional DMARDs)  ▪ T cell co-stimulation modulator - Abatacept (+/- MTX or DMARDs other than JAK inhibitors or biological DMARDs)  ▪ B cell depletion - Rituximab (+ MTX)	Inhibition of TNFα signalling Inhibition of IL-1 signalling Inhibition of IL-6 receptor signalling Inhibition of T cell co-stimulation Depletion of CD20 + B cells
Targeted Synthetic DMARDs  ▪ JAK inhibitors - Tofacitinib (+/- MTX) - Baricitinib (+/- MTX) - Upadacitinib (+/- MTX) - Filgotinib (+/- MTX)$	Inhibition of cytokine-mediated JAK-STAT pathway signalling

*Indicates additional information on the FDA drug label that is not apparent on the EMA drug label. ^$Indicates EMA approval only. MTX methotrexate

CAR T Cell Therapy

CAR T cells are cell-based therapies whereby the patient’s T cells are extracted from their blood and engineered to express a CAR protein on the surface of the T cells that redirects them towards target cells [10]. The engineered T cells are expanded in-vitro before being infused back into the same patient. The CAR is a non-natural synthetic receptor comprised of an extracellular single chain variable fragment (scFv) specific for antigen(s) on the target cell, linked to a transmembrane domain and intracellular signalling domains derived from the TCR (CD3 ζ chain) and the co-stimulatory receptors CD28 or 4-1BB (Fig. 1). The interaction of the CAR molecule with its antigen on the target cell activates the T cell via phosphorylation of the intracellular domains of the CD3 ζ and co-stimulatory domains by endogenous kinases, resulting in the production of perforin/granzyme and inflammatory cytokines that promote target cell death. The advantage of CARs is that they can be tailored to target essentially any antigen on the surface of the target cell and modified with different intracellular domains to fine-tune their function [10].

Fig. 1.

Current clinical approach for manufacturing CAR T cell therapies. T cells are isolated from a cancer patient and engineered with viral vectors encoding the CD19 or BCMA-specific CAR. The CAR T cells are expanded and then infused back into the same patient. scFV = single-chain variable fragment; TM = transmembrane domain. Figure 1 was created in Biorender

To date, there are seven CAR T cell therapies approved by the Food and Drug Administration (FDA) and European Medicines Agency (EMA) and they all target CD19 or BCMA surface antigens on B cells or plasma cells for the treatment of leukaemia, lymphoma or multiple myeloma (Table 2). These are highly personalised, cell-derived living drugs that are made for individual patients and have shown unprecedented clinical results, particularly in patients who have failed conventional forms of treatment including different regimens of chemotherapy [11]. All current FDA/EMA-approved therapies use second generation CAR constructs comprising the intracellular domain of CD3 ζ and the intracellular domain of CD28 or 4-1BB co-stimulatory receptors (Table 2). First-generation CAR constructs lacked the co-stimulatory domains necessary for full T cell activation and were ineffective in clinical trials [10], whereas third and fourth generation CAR constructs, comprising two intracellular co-stimulatory domains within a single construct (i.e. CD28 and 4-1BB domains) or incorporating co-expression of inflammatory cytokines/antibodies, respectively (Fig. 2), are currently under evaluation in pre-clinical and clinical studies [12]. Clinically approved CAR T cell production utilises retroviral or lentiviral viral vectors to genetically engineer the cells, as these vectors integrate the CAR gene into the T cell genome to mitigate against gene loss following T cell expansion prior to infusion back into the patient (Table 2). The primary mechanism of action of these cellular therapies is depletion of CD19 + or BCMA + B cells or plasma cells, resulting in B cell aplasia and elimination of malignant clones. Although these products have demonstrated impressive clinical responses across various B cell malignancies, they have significant and life-threatening side effects, most notably Cytokine Release Syndrome (CRS) and neurological toxicity. Interestingly, the anti-rheumatic bDMARD tocilizumab, an IL-6 receptor antagonist, is approved for the treatment of CRS for six of the seven CAR T cell products and must be available for use prior to reinfusion of the engineered cells in the clinical treatment centre to rapidly manage this side effect, if required. CAR T cell therapies also carry a black box warning of possible secondary T cell malignancies which may be due to the integrating nature of the viral vectors that disrupts the genome of the genetically engineered T cells [13]. Although CAR T cell therapy has demonstrated impressive results for B cell leukaemias, lymphomas and multiple myeloma, they have had limited success in targeting solid tumours [11]. 

Table 2.

Clinically approved CAR T cell therapies for cancer. The clinical indications were obtained from the FDA drug labels (accessed February 2025). 

CAR T cell product	Target antigen	Viral vector	Co-stimulatory receptor	Indication
Tisagenlecleucel (Kymriah)	CD19	Lentiviral	4-1BB	• Patients up to 25 years of age with B-cell precursor ALL that is refractory or in second or later relapse. • Adult patients with relapsed or refractory large B-cell lymphoma after two or more lines of systemic therapy, including diffuse large B-cell lymphoma (DLBCL) not otherwise specified, high grade B-cell lymphoma and DLBCL arising from follicular lymphoma. • Adult patients with relapsed or refractory follicular lymphoma after two or more lines of systemic therapy*.
Axicabtagene Ciloleucel (Yescarta)	CD19	Retroviral	CD28	• Adult patients with large B-cell lymphoma refractory to first-line chemoimmunotherapy or that relapses within 12 months of first-line chemoimmunotherapy. • Adult patients with relapsed or refractory large B-cell lymphoma after two or more lines of systemic therapy, including diffuse large B-cell lymphoma (DLBCL) not otherwise specified, primary mediastinal large B-cell lymphoma, high grade B-cell lymphoma, and DLBCL arising from follicular lymphoma. • Adult patients with relapsed or refractory follicular lymphoma (FL) after two or more lines of systemic therapy*.
Brexucabtagene autoleucel (Tecartus)	CD19	Retroviral	CD28	• Adult patients with relapsed or refractory mantle cell lymphoma (MCL)*. • Adult patients with relapsed or refractory B-cell precursor ALL.
Lisocabtagene Maraleucel (Breyanzi)	CD19	Lentiviral	4-1BB	• Adult patients with large B-cell lymphoma (LBCL), including diffuse large B-cell lymphoma (DLBCL) not otherwise specified (including DLBCL arising from indolent lymphoma), high-grade B cell lymphoma, primary mediastinal large B-cell lymphoma, and follicular lymphoma grade 3B, who have:  - Refractory disease to first-line chemoimmunotherapy or relapse within 12 months of first-line chemoimmunotherapy; or  - Refractory disease to first-line chemoimmunotherapy or relapse after first-line chemoimmunotherapy and are not eligible for hematopoietic stem cell transplantation (HSCT) due to comorbidities or age; or  - Relapsed or refractory disease after 2 or more lines of systemic therapy. • Adult patients with relapsed or refractory chronic lymphocytic leukemia (CLL) or small lymphocytic lymphoma (SLL) who have received at least 2 prior lines of therapy, including a Bruton tyrosine kinase (BTK) inhibitor and a B-cell lymphoma 2 (BCL-2) inhibitor*. • Adult patients with relapsed or refractory follicular lymphoma (FL) who have received 2 or more prior lines of systemic therapy*. • Adult patients with relapsed or refractory mantle cell lymphoma (MCL) who have received at least 2 prior lines of systemic therapy, including a Bruton tyrosine kinase (BTK) inhibitor.
Idecabtagene Vicleucel (Abecma)	BCMA	Lentiviral	4-1BB	• Treatment of adult patients with relapsed or refractory multiple myeloma after two or more prior lines of therapy including an immunomodulatory agent, a proteasome inhibitor, and an anti-CD38 monoclonal antibody.
Ciltacabtagene autoleucel (Carvykti)	BCMA	Lentiviral	4-1BB	• Treatment of adult patients with relapsed or refractory multiple myeloma who have received at least one prior line of therapy, including a proteasome inhibitor and an immunomodulatory agent, and are refractory to lenalidomide.
Obecabtagene autoleucel (Aucatzyl)	CD19	Lentiviral	4-1BB	• Treatment of adults with relapsed or refractory B-cell ALL.

*Indicates accelerated approval at the current time with continued approval contingent upon verification and description of clinical benefit in confirmatory trial(s). The trade name of each CAR T cell product is in brackets in the first column. ALL acute lymphoblastic leukaemia

Fig. 2.

First, second, third and fourth generation CAR T cells

The effectiveness of CAR T cell therapy for the treatment of B cell cancers has led to their exploration for the treatment of inflammatory and autoimmune diseases including systemic lupus erythematosus (SLE), systemic sclerosis (SSc), neuromyelitis optica spectrum disorder (NMOSD), myasthenia gravis (MG), multiple sclerosis (MS) and RA where similar mechanisms of action are desired i.e. depletion of autoreactive B cells [14, 15]. Beyond these conventional autoimmune diseases, CAR T cell therapy has shown promising results in rare autoimmune disorders such as antiphospholipid syndrome [16], severe idiopathic Lambert-Eaton myasthenic syndrome [17] and dermatomyositis [18] Hence it is anticipated that the strategy of depleting CD19 + or BCMA + B cells with CAR T cell therapy in cancer patients will result in elimination of autoreactive B cells in RA patients. In many ways, their introduction into the clinic mimics rituximab which is clinically approved for the treatment of RA and haematological cancers, with a common mechanism of action of depletion of B cells via targeting of antigens on the surface of the cell [14]. Promising case studies in RA patients have recently shown, however, that CAR T cells may provide superior therapeutic advantages over rituximab as a single infusion of these cells can induce sustained drug-free remission in RA (see case studies below). Furthermore, the T cells used for CAR T cell manufacture can be enriched for specific T cell subsets to tailor immune responses (e.g. CD4 + helper T cells, CD8 + cytotoxic T cells, naïve, effector, memory cells central, effector, stem cell-like subtypes or regulatory T cells [19]), while the CAR itself can be engineered to target defined subsets of pathogenic cells including antigen-specific B cells or T cells and thus avoiding the destruction of bystander B cells. These studies and the exciting potential of CAR T cell therapy for the treatment of RA is discussed in the next section and summarised in Table 3.

Table 3.

Summary of patient characteristics, treatment approaches with CAR T cell therapies and clinical results

Age/Sex	Clinical History	Prior Treatments	CAR T cell product	B cell status & disease activity scores pre/post CAR treatment	Side Effects (CRS/ICANS)	Ref
CD19/CD20-directed CAR T cell therapy
73/M	RA (2011) DLBCL (2023)	GC, MTX, HC, Toc (for RA) R-CNOP (for DLBCL)	Zamtocabtagene autoleucel (CD20-CD19 tandem CAR with 4-1BB co-stimulatory domain)	CD19 + B cell aplasia until day 180, followed by recovery DAS-28: 6.43*/0 ACPA (U/ml): 31*/low RF (U/ml): 1200/13	Grade 1 CRS (treated with antipyretics)	[20]
62/F	SS (2000) RA (2009) DLBCL (2020)	GC, MTX, anti-TNFα (x2), Toc, Abeta, Tofa (for SS and RA) EPOCH-R (for DLBCL)	Yescarta (CD19 CAR with CD28 co-stimulatory domain)	No detectable CD19 + B cells in peripheral blood before or > 2 years post therapy CDAI: 9*/0 ESR: 28*/14 CRP: >10*/2.4 Anti-CCP (RU/ml): 149*/128 RF: 128*/<10	None reported	[22]
37/F	MG (2013) RA (2020)	Thy, ACI, AZ, Ig, R, EC (for MG)	KYV-101 (CD19 CAR with CD28 co-stimulatory domain)	No detectable CD19 + B cells in peripheral blood from days 4–150, with slow reconstitution thereafter CDAI: 24/0 DAS-28-ESR: 6/1.9 DAS-28-CRP: 4/1 ACPA (U/ml): >35/0	Grade 1 CRS	[23]
32/F	SSc (> 6 years) RA (> 6 years)	GC, MTX, AZ, HC, Toc, Nin, Sil, Nif, Ilo.	KYV-101 (CD19 CAR with CD28 co-stimulatory domain)	CD19 + B cell aplasia until day 56, followed by recovery ACPA (U/ml): 60.2/11.9	Grade 2 CRS (treated with Toc and steroids)	[24]
32/F	MS (2012) RA (2017)	IFN β−1a, Fingo, Alemtuz (for MS) GC, MTX, LEF, R, Abeta, Ana, Toc, anti-TNFα (x3), Bar, Upa, Tofa, Fil, Sec (for RA)	MB-CART19.1 (CD19 CAR with 4-1BB co-stimulatory domain)	CD19 + B cells very low at baseline and increased to normal levels at day 120 post CAR T cell therapy DAS-28-CRP: 6/<3 CRP (mg/dL): 4.8–12/low	Grade 2 CRS (treated with Toc)	[25]
39/F	RA (> 20 years)	MTX, CYC, SLZ, LEF, PLQ, anti-TNFα (x5), Toc/Sar, Ana, Abeta, R (x2), Bar, Upa, Tofa	FMC63-28-CD3ζ CAR (CD19 CAR with CD28 co-stimulatory domain)	CD19 + B cells not reported DAS-28-CRP: 7.46/2.5 ACPA (U/ml): 489/low RF (U/ml): 339/low	Grade 3 CRS Grade 4 ICANS (treated with Toc, Ana and steroids)	[27]
49/F	RA (> 20 years)	MTX, LEF, HC, GC, anti-TNFα (x2), Tofa	Fourth generation CAR (CD19 CAR with 4-1BB co-stimulatory domain with co-expression of scFvs targeting IL-6 and TNFα)	CD19 + B cell aplasia from days 3–60, followed by recovery CDAI: 23/1.5 DAS-28-CRP: 4.67/2.26 ESR (mm/h): 120/78 CRP (mg/L): 63.7/low Anti-CCP (RU/ml): 399/1 RF (U/ml): 290/low	None reported	[28]
52/F	RA (> 7 years)	MTX, Igur, GC, anti-TNFα, Bar	Fourth generation CAR (CD19 CAR with 4-1BB co-stimulatory domain with co-expression of scFvs targeting IL-6 and TNFα)	CD19 + B cell aplasia from days 7–120, followed by recovery CDAI: 18/2 DAS-28-CRP: 4.04/2.67 ESR (mm/h): 120/34 CRP (mg/L): 20.9/low Anti-CCP (RU/ml): 865/1 RF (U/ml): 1110/low	None reported	[28]
56/F	RA (> 3 years)	MTX, LEF, HC, Igur, GC, Bar, Abeta	Fourth generation CAR (CD19 CAR with 4-1BB co-stimulatory domain with co-expression of scFvs targeting IL-6 and TNFα)	CD19 + B cell aplasia from days 3–120, followed by recovery CDAI: 28/2 DAS-28-CRP: 4.61/1.56 ESR (mm/h): 60/16 CRP (mg/L): 3.2/low Anti-CCP (RU/ml): 1604/45 RF (U/ml): 263/low	None reported	[28]
BCMA-directed CAR T cell therapy
67/F	NMOSD	MTX, AZ, FK506 (for NOSD)	CT103A Second generation CAR (BCMA CAR with 4-1BB co-stimulatory domain)	BCMA + B cells not reported DAS-28-ESR: 4.698/1.987	Grade 2 CRS	[29]

RA Rheumatoid Arthritis, DLBCL Diffuse large B cell lymphoma, MG Myasthenia gravis, SS Sjogren’s syndrome, SSc Systemic sclerosis, NMOSD Neuromyelitis Optica Spectrum Disorder, GC Glucocorticoid, MTX Methotrexate, HC Hydroxychloroquine, R Rituximab, Toc Tocilizumab, Nin Nintedanib, Sil Sildenafil, Nif Nifedipine, Ilo Iloprost, Ana Anakinra, Sar Sarilumab, ACI Acetylcholinesterase inhibitors, AZ Azathioprine, Ig Immunoglobulins, Bar Baricitinib, Upa Upadacitinib, Tof Tofacitinib, Fil Filgotinib, Cyc Cyclosporin, LEF Leflunomide, IFN β-1a Interferon β−1a, Fingo Fingolimod, Alemtuz Alemtuzumab, PLQ Plaquenil, SLZ Sulfasalazine, Igur Iguratimod, EC Eculizumab, Thy thymectomy, Dex Dexamethasone, FK506 Tacrolimus, R-CNOP rituximab, cyclophosphamide, mitoxantrone, vincristine and prednisolone, EPOCH-R etoposide, prednisone, vincristine, cyclophosphamide, doxorubicin, and rituximab. CRS Cytokine Release Syndrome, ICANS Immune Effector Cell-Associated Neurotoxicity Syndrome, CDAI Clinical disease activity index, DAS28 Disease Activity Score in 28 joints, ESR Erythrocyte Sedimentation Rate, CRP C-reactive protein, ACPA Anti-citrullinated peptide antibodies, Anti-CCP anti-Cyclic Citrullinated Peptides, RF Rheumatoid Factor

*Indicates historical disease activity markers prior to CAR T cell infusion (most recent is indicated here)

‘Low’ refers to baseline/undetectable values from graphs (actual values not reported)

CAR T Cell Therapy for the Treatment of RA

Clinical Studies: CD19 and Tandem CD20-CD19-directed CAR T Cell Therapy

Several case studies have reported remarkable success with CAR T cells targeting CD19 + and/or CD20 + B cells in patients with RA. A recent case study [20] reported a 73-year-old male patient diagnosed with highly active seropositive RA in 2011. The patient was treated with low-dose glucocorticoid, MTX, hydroxychloroquine and tocilizumab (anti-IL6 receptor antagonist) and remained on treatment until 2018 with low/moderate disease activity. The patient was subsequently diagnosed with Germinal Centre B cell-like-DLBCL in 2023 and was treated with rituximab, cyclophosphamide, mitoxantrone, vincristine and prednisolone, but was deemed refractory to treatment. On the basis of the patient’s refractory DLBCL, they were enrolled in a phase II randomised, multicentre, open-label clinical trial to study the effect of a novel bispecific CD20-CD19 directed CAR T cell therapy. Since loss of CD19 surface antigen is common in B cell malignancies and appears to be a mechanism for immune evasion with CD19-specific CAR T cell therapy [21] a dual approach of targeting both CD19 and CD20 in tandem aimed to mitigate against a reduction in CAR T cell efficacy should loss of CD19 occur on the target cells. The patient’s T cells were engineered with zamtocabtagene autoleucel (zamto-cel) which is an investigational lentiviral vector encoding a CAR construct specific for both CD20 and CD19 in tandem, linked to the intracellular domains of CD3 ζ and 4-1BB (Fig. 3). The CAR T cell product was well tolerated with mild CRS (grade 1) that was subsequently treated with anti-fever medication. Deep CD19 + B cell aplasia was noted until 180 days post CAR T cell infusion and was concomitant with a reduction in disease activity scores and autoantibody levels (Table 3). As of October 2024, the patient remains in partial response/stable disease for DLBCL one year post-therapy and, intriguingly, is in complete drug-free clinical remission for RA. Notably, the re-emergence of B cells after 180 days post-therapy has not led to the reappearance of RA, indicating that an ‘immune reset’ may have taken place with elimination of the autoreactive B cell clones [20]. A very similar case study highlighted a 62-year-old woman diagnosed with Sjogren’s syndrome in 2000 and seropositive RA later in 2009, the latter which was difficult-to-treat [22]. The patient’s treatment for RA consisted of long-term glucocorticoids, MTX, two anti-TNFα therapies, anti-IL-6 receptor therapy (toculizumab), a co-stimulation modulator (abatacept) and a JAK inhibitor (tofacitinib). However, both the patient’s Sjogren’s syndrome and RA remained active and unfortunately DLBCL was also diagnosed in 2020. The patient was treated with the clinically approved CD19-targeting CAR T cell therapy axicabtagene ciloleucel (Yescarta) in April 2022 after relapse of DLBCL following standard-of-care rituximab in combination with chemotherapy. Remission of the patient’s lymphoma and RA was achieved with CAR T cell therapy in October 2022 and December 2022, respectively. Curiously, CD19 + B cells were undetectable both before CAR T cell therapy and up to 2 years after infusion. Disease activity scores and autoantibody levels for Rheumatoid Factor were reduced after CAR T cell infusion when compared to historical values (Table 3). As of December 2024 (i.e. 2 years), the patient remains drug free from RA-directed therapies including corticosteroids, csDMARDs, bDMARDs and tsDMARDs [22]. The patient’s SSA and SSB antibodies remained elevated following CAR T cell therapy, indicating persistent Sjogren’s disease. In addition, this patient’s long-standing Sjogren’s had resulted in chronic severe sicca complex with minimal tear/saliva production and parotid atrophy and these manifestations remained largely unchanged following CAR T cell therapy [22].

Fig. 3.

Tandem CD20-CD19-directed CAR T cell therapy. Zamtocabtagene autoleucel (zamto-cel) is an investigational CAR T cell therapy targeting both CD20 and CD19 on the same target B cell [20]. Figure 2 was created in Biorender

Three case studies have shown excellent results for the treatment of RA co-existing with other inflammatory diseases. In the first of these studies [23], a 37-year-old woman first diagnosed in 2013 with MG (a B cell driven autoimmune disease) subsequently developed ACPA-positive RA in 2020. As treatment was focused on MG and involved the use of glucocorticoids and rituximab (treatments that can also be used for RA), no additional therapies were considered for RA. The patient did not respond to these conventional therapies, as well as other therapeutic regimens (Table 3), and had high levels of disease activity for both MG and RA concomitant with recurrent infections and a poor quality of life. The patient’s T cells were isolated and engineered with a lentiviral vector encoding a second generation CD19-specific CAR T construct comprising a fully human CD19 binding domain with CD3 ζ and CD28 costimulatory domains (KYV-101, Kyverna Therapeutics). The CAR T cell product was mainly comprised of CD4 + helper T cells and was well tolerated with grade I CRS following infusion. Disease activity scores for both MG and RA indicated disease remission (Table 3), with undetectable ACPA levels reported for RA. Circulating B cells were undetectable at day 4 post CAR T cell infusion and slowly returned by day 150. The patient was free of joint pain, was symptom free and was subsequently able to exercise for an hour. Clinical and immunological data was reported up to 150 days post treatment and therefore it be will interesting whether long-term drug free remission is possible. In the second case study [24], a 32-year-old female patient with a six-year history of SSc and ACPA-positive RA was treated with autologous T cells engineered with KYV-101. The patient developed grade II CRS that was treated with tocilizumab and corticosteroids. Circulating B cell counts were undetectable for 56 days and reappeared in the blood and bone marrow as naïve/transitional B cell subsets by week 16. The level of ACPAs declined below the cut-off values within 21 days (Table 3) and have remained below baseline for 6 months. As of February 2025, the patient remains in remission for RA [24]. This patient’s SSc-associated complications, namely vasculopathy and fibrosis, improved despite discontinuation of therapy. Lung function also improved slightly at six months post CAR T cell therapy. However, the patient’s anti-Scl-70 antibodies have remained stable during this period. The third case study [25] reported a 32 year old female patient diagnosed with relapsing-remitting MS at age 20 who received sequential treatments of interferon β1a, fingolimod and the CD52-targeting antibody alemtuzumab, with the latter used to deplete circulating T cells and B cells. Remission of the patient’s MS was achieved with alemtuzumab in 2016 but the development of symmetric, erosive polyarthritis in 2017 resulted in a diagnosis of seronegative RA. Over the next 6 years, the patient’s treatment was focused towards their RA via glucocorticoids and csDMARDS (MTX and leflunomide), bDMARDS (rituximab, a co-stimulation modulator, anti-IL-1, an anti-IL-6 receptor and 3 x anti-TNFα), tsDMARDS (4 JAK inhibitors) and an anti-IL-17 A antibody (approved for psoriatic arthritis, non-radiographic axial spondyloarthritis and enthesitis-related arthritis). As these treatments were ineffective, CD19-directed CAR therapy was initiated in May 2023 where the patient’s T cells were engineered with MB-CART19.1, an investigational lentiviral viral vector encoding a CD3 ζ and 4-1BB costimulatory domain that has been used successfully to treat patients with refractory SLE [26]. The CAR T cell product was comprised of CD4 + helper T cells and CD8 + cytotoxic T cells. A grade II CRS was reported shortly after infusion of the cells and was treated with tocilizumab. Disease activity scores improved (Table 3), and remission was achieved on day 28 after administration of the CAR T cells. However, there was a severe re-occurrence of synovitis at 8 months post CAR T cell therapy concurrent with an increase in RA disease activity scores, indicating disease relapse. Interestingly, the infusion of an autologous CD34 + haematopoietic stem cell transplant in April 2024 resulted in a decline of the RA disease activity scores and remission [25]. It is important to highlight that the decision to use CD19-directed CAR T cell therapy was probably unusual in this case study, given that the patient’s refractory RA was seronegative and there were very low levels of circulating B cells. Hence the decision to target CD19 + B cells may have been based on very limited remaining treatment options for this patient. Indeed, the decline of the T cell population, and not the B cell population, after haematopoietic stem cell transplantation correlated with remission and suggests that this patient’s relapsed RA may have been driven by T cells rather than B cells.

The five case studies highlighted above showcase the potential for CAR T cell therapies to treat RA. These treatments, however, were complicated by co-existing B cell malignancies or autoimmunity in all five patients. Two additional case studies have been recently reported in difficult-to-treat RA patients without additional malignancies or inflammatory disease [27, 28]. The first patient was a 39-year-old woman with erosive seropositive RA who had failed a variety of csDMARDs, bDMARDs (5 x anti-TNFα, 2 x rituximab, 2 x anti-IL-6 receptor, anti-IL-1 and a co-stimulation modulator) and tsDMARDS (3 JAK inhibitors) over a 20-year disease period [27]. These therapies resulted in depletion of B cells in peripheral blood yet > 70% CD19 + B cells remained in synovial biopsies. The patient’s T cells were extracted and engineered with a second-generation retroviral vector encoding CD19-specific CAR linked to the intracellular domains from CD3 ζ and CD28. The patient developed grade III CRS and grade IV neurotoxicity one day after infusion of the CAR T cells which required treatment with tocilizumab, anakinra and high dose corticosteroids. After 100 days post-treatment, inflammatory markers declined, and autoantibody levels were reduced by 80% (Table 3). The patient is currently in drug-free remission [27]. The second study reported three seropositive RA patients that were refractory to a variety of csDMARDs and bDMARDs (notably these patients had not been treated with rituximab) [28]. The patients’ T cells were engineered with a fourth-generation CAR comprising a CD19-targeting scFv, intracellular domains from CD3 ζ and 4-1BB and scFv sequences derived from neutralising anti-IL-6 (sirukumab) and anti-TNFα (adalimumab) to direct secretion of soluble cytokine targeting antibodies. The rationale for investigating this advanced CAR construct was that targeting both B cells and RA-associated inflammatory cytokines may provide additional therapeutic benefit in difficult-to-treat patients. The therapy was tolerated well in all three patients and no serious adverse events including CRS or neurotoxicity were recorded. CD19 + B cells were depleted in all three patients and this correlated with proliferation of the CAR + T cells in vivo. Clinical responses were impressive, with a reduction in disease activity markers and ablation of autoantibody levels correlating with an improvement of synovitis and reduction in joint swelling (Table 3). Similar to studies already highlighted above [20, 23, 24], circulating B cells re-emerged at day 60–120 post CAR T cell infusion and no relapse in RA was observed at 9 months.

Clinical Studies: BCMA-directed CAR T Cell Therapy

12 patients with NMOSD - an autoimmune, inflammatory disorder of the central nervous system - were treated with CAR T cells targeting the BCMA antigen expressed on normal B cells and plasma cells [29]. One of these patients was a 67 year old female with NMOSD co-existing with RA. The patient’s T cells were transduced with a lentiviral vector expressing a second generation BCMA-specific CAR linked to the intracellular domains from CD3 ζ and 4-1BB co-stimulatory domain (CT103A). Although the intention-to treat-was focused on NMOSD, the patient showed clinical improvement in their disease activity scores for RA at week 12, indicating remission (Table 3), in addition to improvement in NMOSD. It was not disclosed whether this patient had treatment refractory RA and whether the disease was seropositive or seronegative. Grade 2 CRS was observed following infusion of the CAR T cells [29].

CAR T Cells Targeting Other Antigens for the Treatment of RA

CAR T cells targeting CD19/CD20 on malignant or autoreactive B cells also target healthy B cells, resulting in B cell aplasia and hypogammaglobulinemia. The targeting of all circulating CD19 + B cells by CAR T cell therapy has led to the exploration for novel antigens that are selectively expressed on malignant or autoreactive cells relative to healthy cells. Zhang and colleagues elegantly demonstrated that human CAR T cells could be engineered to recognise autoreactive B cells from RA patients in-vitro [30]. Primary human T cells were transduced with a lentiviral vector encoding a CAR specific for the fluorescent reporter molecule FITC (Fig. 4a). These engineered T cells were subsequently co-cultured in-vitro with B cells from RA patients in the presence of FITC-labelled citrullinated peptides to direct the CAR T cells to the autoreactive B cells. The CAR T cells specifically lysed peptide-specific autoreactive B cells in the presence of FITC-labelled peptide but had minimal effects on B cells incubated with an irrelevant FITC-labelled peptide [30]. In a separate study, the Rosloniec group [31] took a different approach and engineered a novel CAR construct with HLA DR1 (a common antigen-presentation molecule expressed in RA patients) on the extracellular domain that was covalently linked to an autoantigenic peptide from type II collagen (Fig. 4b). Hence, the HLA DR1/collagen peptide replaced the scFv that is typically found on most CAR molecules. As the HLA DR1 molecule is comprised of two peptide chains (DR1A and DR1B), each chain was linked with the intracellular signalling domain of CD3 ζ and CD28. The rationale of this study was to demonstrate that expression of this CAR construct in murine CD8 + cytotoxic T cells via retroviral expression could directly target and eliminate autoreactive CD4 + T cells with TCRs specific for this HLA/peptide complex. In vitro assays demonstrated that the CAR T cells efficiently recognised and killed CD4 + T cells specific for type II collagen, whereas CD4 + T cells that recognise other antigens were not targeted. Furthermore, the infusion of these CD8 + CAR T cells into a humanised mouse model of autoimmune arthritis reduced disease severity and/or delayed onset, as well as reducing antigen-specific T cell proliferation and autoantibody levels. Thus, these CAR T cells were clearly capable of targeting autoreactive T cells and B cells. The CAR T cell population did not proliferate in vivo, however, and the numbers of engineered cells declined after 10 days post administration which may be due to the low number of target cells in-vivo [31].

Fig. 4.

Pre-clinical CAR T cell studies for the treatment of RA via targeting autoreactive B cells and T cells. (a) CAR T cells recognising FITC-labelled citrullinated peptides [30] (b) CAR T cells expressing a HLA DR1/collagen peptide complex to target autoreactive CD4 + T cells [31]. Figure 3 was created in Biorender

As already mentioned, specific T cell subsets can be enriched for CAR T cell manufacture to tailor immune responses [19]. One such example is the use of regulatory T cells (Tregs) for CAR T cell therapy as these cells negate inflammatory responses via (a) inhibitory cytokine signalling (IL-10, TGFβ and IL-35 pathways) (b) direct cell: cell contact that suppresses proinflammatory T cells (c) production of metabolites that dampen immunity or (d) sequestration of co-stimulatory signals on antigen presenting cells [32]. An unpublished study reported engineering of human Tregs with a CAR containing an extracellular binding domain specific for citrullinated vimentin which is a protein found in the synovium of RA patients (Fig. 5). The CAR expressing Treg cells proliferated in response to citrullinated vimentin (but not an unmodified protein), produced IL-10 and suppressed the proliferation of CD4 + and CD8 + T cells. Furthermore, the CAR Tregs were activated when incubated with synovial fluid from RA patients.

Fig. 5.

Pre-clinical CAR Treg cells and FAP-targeting CAR T cells for the treatment of RA

Another interesting target that could potentially be targeted with CAR T cells is fibroblast-associated protein (FAP). FAP is a type II transmembrane serine protease that cleaves peptide bonds between proline and other amino acids (reviewed in [33]). FAP expression is normally restricted to foetal tissue and is not expressed in healthy adult tissue other than bone marrow derived mesenchymal stem cells. Interestingly, FAP expression is upregulated in stromal fibroblasts of more than 90% of epithelial cancers, as well as wound healing and fibrotic diseases [33]. The restricted nature of FAP expression makes it a promising target for CAR T cell therapy because it may allow for selective targeting of cancer cells or fibrotic (scar) tissue, with minimal damage to healthy tissues (Fig. 4b). FAP-targeting CAR T cells have shown promising results in vitro for the treatment of solid cancers [34, 35], and infusion of FAP-targeting CAR T cells into mice with hypertensive cardiac injury (and concomitant expression of FAP) eliminated activated fibroblasts, significantly reduced cardiac fibrosis and restored cardiac function [36, 37]. FAP is also highly expressed in the synovium of RA patients, making it an intriguing antigen to target with CAR T cells [33] (Fig. 5). FAP is not expressed in the synovium of healthy mice or in models of osteoarthritis, suggesting that the targeting this protein should have minimal effects on non-activated fibroblasts and healthy tissues. FAP-targeting CAR T cells have not been evaluated in pre-clinical models of RA to date, but if successful, they could provide the first RA therapy that does not directly target immune cells and inflammation. In summary, these studies demonstrate the potential for selectively targeting autoreactive B cells or T cells, as well as enriching for T cell subsets to tailor immune responses for the treatment of RA.

Discussion and Future Studies

CAR T cell therapy has shown unprecedented results for the treatment of seropositive RA in difficult-to-treat patients via depletion of CD19+/CD20 + and BCMA + B cells. The clinical data, while preliminary, suggests that a single infusion of these cell-based therapies could mediate long term suppression of inflammatory disease, an ‘immune reset’ and drug-free remission [20, 22–24, 27, 28]. These encouraging results were reported in patients with difficult-to-treat seropositive RA alone, as well as seropositive RA patients with co-existing inflammatory diseases or lymphoma. One patient with seronegative RA is the only reported case with this form of the disease who was treated with CAR T cell therapy [25]. Although this patient responded well to CAR T cell therapy initially, their disease relapsed. Seronegative RA is a distinct form of the disease with a different pathogenesis and therefore more experience is needed in these patients. In total, ten patients have been treated with CAR T cells, with nine of the patients receiving CD19/CD20-directed therapy and one patient receiving BCMA-directed therapy. Clinical efficacy was observed with CD19-targeting CAR T cells expressing FDA/EMA approved constructs and manufacturing protocols (Yescarta), as well as investigational CAR T cells expressing either second generation constructs (with different co-stimulatory domains) or fourth generation constructs. Depletion of CD19 + B cells in peripheral blood was reported for most of these case studies [20, 22–24, 28] alongside reduction of disease activity markers, correlating with reduction in symptoms and drug-free remission. Intriguingly, the CD19 + peripheral B cell subsets recovered over time, and the patients did not experience classical RA symptoms suggesting elimination of autoimmune B cell subsets.

Despite these promising results, the data is preliminary as only ten RA patients have been treated with CD19/CD20 or BCMA-directed CAR T cell therapy to date, and these were mostly individual case studies where follow-up was over a relatively short period. Drawing meaningful conclusions on clinical efficacy from these case studies is therefore limited. Firstly, two of the patients that were treated with CD19-directed CAR T cell therapy were patients with difficult-to-treat seropositive RA who subsequently developed lymphoma [20, 22]. While the intention of their CAR T cell therapy was to treat their lymphoma, the resolution of their RA strongly suggests that this cell-based treatment was the sole factor behind this clinical outcome. However, treatment of these patients with rituximab-based chemotherapy regimens (R-CNOP and EPOCH-R) and steroids for their cancer prior to CAR T cell therapy could have impacted the disease course of RA. In addition, all ten RA patients (including the patient with seronegative RA) who were treated with CAR T cells received the standard pre-conditioning regimen of fludarabine and cyclophosphamide as a lymphodepletion strategy to facilitate engraftment of the infused cells. While it could be argued that such pre-conditioning may also be a contributing factor to the success of CAR T cell therapy for B cell cancers, pre-conditioning with fludarabine and cyclophosphamide cannot be ruled out as contributing factors in the depletion of autoimmune B cells. Secondly, these case studies lacked control patient groups treated with standard of care or placebo to confirm efficacy of CAR T cell therapy for RA. Hence these case studies are biased as only positive outcomes have been reported to date. Thirdly, the long-term outcome of these patients is currently unknown as they have only been followed for relatively short periods of time, ranging from 3 to 12 months [20, 23, 24, 27, 28] to less than 2 years [22]. Long-term outcomes for these patients in terms of clinical efficacy, safety (due to profound B cell depletion and infection risk), drug-free remission and the risk of relapse is unknown. Future RA studies will therefore need to evaluate larger cohorts of patients in randomised clinical trials where CAR T cell therapy is compared to standard-of-care treatments, with long-term monitoring to assess efficacy and drug-free remission, safety and potential for disease relapse. Indeed, several clinical trials are currently underway or will shortly recruit RA patients to evaluate CAR T cell therapy (Table 4). One of these trials (NCT06475495) will engineer T cells with the investigational viral vector KYV-101 encoding the CD19-specific CAR and compare their efficacy to standard-of-care rituximab treatment. It will be interesting to determine whether infusion of CAR T cells provides long-term remission of RA in a direct head-to-head comparison with rituximab. Rituximab has been reported as only partially depleting synovial B cells in RA patients [38, 39] and therefore it will be useful to determine if the impressive results with CAR T cell therapy correlates with trafficking into the joints and depletion of synovial B cells. A recent study demonstrated that CAR T cell therapy induced deep tissue depletion of CD19 + B cells in both the blood and lymph nodes of patients with SSc and SLE, whereas rituximab only depleted CD19 + B cells in the blood [40]. One possible explanation for the efficacy of CD19-targeting CAR T cells for RA is that the CD19 antigen is expressed on early B-cells, antibody-producing plasmablasts and plasma cells, whereas CD20 (the target of rituximab) is more highly expressed on mature B cells. CAR T cells may therefore target a wider B cell population compared to rituximab. Alternatively, the effector mechanisms of these personalised living drugs (cytotoxic activity and inflammatory cytokines) may be superior to antibody-based therapies.

Table 4.

Current CAR T cell clinical trials for RA. Information was obtained from the US clinical trials website Clinicaltrials.gov (accessed September 2025)

Clinical trial number	Phase	Summary	CAR T cell therapy	Trial status
NCT06475495 (Berlin, Germany)	I/II	Comparison of B-cell depletion by CD19 targeting CAR T cells or rituximab in treatment refractory RA (COMPARE)	KYV101 (CD19-targeting CAR T cells)	Not yet recruiting
NCT06428188 (Beijing, China)	I/II	Clinical study to evaluate the efficacy and safety of CAR T cells targeting BCMA or CD19, or both sequentially, in the treatment of relapsed/refractory autoimmune disease such as Sjogren’s Syndrome, SLE and others (including RA).	Unknown	Recruiting
NCT07100873 (Shanghai, China)	I	Anti-CD20 CAR-engineered Allogeneic γδ T Cells in Adults With Treatment-refractory RA	ADI-001 (Allogeneic CD20-targeting γδ CAR T cells)	Not yet recruiting
NCT07048197 (France, Germany, Singapore, Spain)	I/II	A Study to Assess Safety, Cellular Kinetics and Exploratory Efficacy of Rapcabtagene Autoleucel in RA and Sjogren’s Disease	YTB323 (CD19-targeting CAR T cells with 2 day manufacturing times)	Recruiting
NCT07085676 (Israel)	I	Study of HBI0101 CAR-T in Refractory B-Cell Autoimmune Diseases (including RA)	HBI0101 (BCMA-targeting CAR T cells)	Recruiting
NCT06417398 (Unknown)	Preliminary study	Preliminary Clinical Study of UTAA09 Injection in the Treatment of Relapsed/​Refractory Autoimmune Diseases (including RA)	UTAA09 (Allogeneic CD19-targeting γδ CAR T cells)	Not yet recruiting
NCT07115745 (US, Australia, Brazil, Czechia, France, Germany, Israel, Poland, Romania, Spain)	I	A Study of Healthy Donor CD19-targeted Allogeneic CAR T Cells in Participants With Severe, Refractory Autoimmune Diseases (including RA)	BMS-986,515 (CD19-targeting CAR T cells)	Not yet recruiting
NCT05869955 (US, Belgium, France, Germany, Italy, Spain)	I	A Study of CC-97,540, CD19-Targeted Nex-T CAR T Cells, in Participants With Severe, Refractory Autoimmune Diseases (Breakfree-1) (including RA)	CC-97,540/BMS-986,353 (CD19-Targeted Nex-T CAR T Cells with increased manufacturing times)	Recruiting
NCT06503224 (China)	I	An Exploratory Clinical Study of SCAR02 Targeting BCMA and CD19 for the Treatment of Refractory Autoimmune Diseases (SCAR02) (including RA)	SCAR02 (Fourth generation CAR T cells targeting both CD19 and BCMA)	Recruiting

Although CAR T cells have shown impressive results in B cell cancers, their serious and life-threatening side effects requires availability of the bDMARD tocilizumab in the infusion centre to rapidly manage CRS. The inclusion of black box warnings on each FDA drug label to “Do not administer [CAR T cell therapy trade name] to patients with active infection or inflammatory disorders.”, is obvious given that CAR T cell-mediated CRS would exacerbate inflammatory disease. While these warnings are applied in the context of cancer patients being considered for CAR T cell therapy, it is interesting that these therapies have now been infused in cancer patients with co-existing RA as well as RA patients without cancer. It has been reported that CD19-directed CAR T cell therapy for autoimmune disease induces low/manageable levels of CRS and neurotoxicity in patients and appears to be milder compared to cancer patients [14, 15]. However, as highlighted in the clinical case studies summarised in this review, grade II/III CRS and grade IV neurotoxicity were reported in some RA patients who received CAR T cell therapy, necessitating the use of tocilizumab to manage CRS [24, 27]. An inflammatory reaction that is distinct from CRS was described recently by the Schett group in Germany in autoimmune patients treated with second generation CD19-targeting CAR T cells [41]. This inflammatory side effect was termed Local Immune effector Cell-Associated Toxicity Syndrome (LICATS) and was observed in 30/39 patients with SLE, idiopathic inflammatory myopathy or systemic sclerosis. LICATS was distinguishable from CRS as it was localised to the organ where the autoimmunity was originally present (resulting in what could be misinterpreted as a disease relapse), whereas CRS is a systemic inflammatory reaction. LICATS coincided with CAR T cell expansion/B cell aplasia and occurred within a median timeframe of 10 days post CAR T cell infusion. This is slower than CRS which is typically observed within days after cell infusion. The median duration of LICATS was 11 days. LICATS resolved spontaneously or with short-term glucocorticoid treatment. Elsewhere, RA-like flares and symptoms were recently reported in 6 cancer patients treated with CD19-targeting CAR T cells [42, 43]. This highlights that inflammatory side effects of CAR T cell therapy, particularly CRS/ICANS, LICATS and RA-like flares will need to be monitored closely in future trials for RA. Interestingly, the three RA patients that were treated with fourth-generation CAR T cells co-expressing neutralising anti-IL-6 and anti-TNFα antibodies showed excellent responses and did not experience any side effects [28]. The use of tocilizumab to treat CRS in RA patients has added an interesting question as to whether blocking the IL-6 pathway has a dual role in controlling CRS in addition to a therapeutic benefit for the treatment of RA when used with CAR T cell therapy. Randomisation of RA patients for treatment with second-generation CAR T cells versus fourth generation CAR T cells co-expressing neutralising IL-6 antibodies may be a novel way of addressing this important clinical question in the future.

Another concern with CAR T cell therapy for treating autoimmune diseases is the prolonged depletion of autoreactive B cells as well as healthy B cells, potentially resulting in humoral immunodeficiency and the need for immunoglobulin replacement. While deep CD19 + B cell aplasia was reported across most of the case studies for RA patients treated with CAR T cells, this cell population re-emerged at different times (between days 56–180 across all patients; Table 3). Some of these case reports also documented serum immunoglobulin levels pre and post CAR T cell therapy where approximately 50% reductions in serum immunoglobulin G (IgG) were noted after CAR T therapy (reduced to ~ 5 g/L) [20, 23]. One of these studies additionally reported IgG responses to standard vaccinations as being only slightly reduced for measles, mumps, rubella, and varicella-zoster virus, whereas serum IgG responses to tetanus were reduced. Together, these studies suggest retained humoral immune memory and plasma cell functions. However, in the three RA patients treated with fourth generation CAR T cells [28], although IgG levels dropped to 6–8 g/L and recovered slightly at day 180, IgM and IgA serum antibody levels were reduced to baseline in all three patients and did not recover over this period. Overall, these limited results with serum immunoglobulins highlight that the risks of humoral immunodeficiency will need to be monitored closely over time to ensure long-term patient safety.

The impressive results with CAR T cell therapy for refractory seropositive RA suggests that resolution of disease and drug-free remission may be possible over the long term. However, the potential for disease relapse will be a concern, with a relapse reported in a seronegative RA patient who received CD19-directed CAR T cell therapy and which was resolved following haematopoietic stem cell transplantation [25]. Emerging reports have also highlighted disease relapse following CD19-targeted CAR T cells in patients with other autoimmune diseases. In these reports, individual patients with either idiopathic inflammatory myositis [44], MG [45] or Autoimmune Hemolytic Anemia [46] experienced disease release after CAR T cell therapy. Interestingly, remission was regained with infusion of a second dose of CD19-targeting cells [45], CAR T cells targeting BCMA [44] or following treatment with a different therapeutic modality (a bispecific T cell engager) [46]. It is currently unknown whether disease relapse in autoimmune patients is similar to cancer patients whereby CD19 or BCMA negative B cells re-emerge following therapy as a mechanism of immune evasion of the malignant B cell population.

CAR T cell therapy carries a black box warning of secondary T cell malignancies in patients who received these therapies. Although it has been speculated that this may be caused by disruption of T cell tumour suppressor genes because of the integrating nature of the retroviral/lentiviral vectors used for CAR gene delivery, an alternative hypothesis is that malignant T cell clones may pre-exist prior to CAR integration and are selectively expanded during therapy [47]. If secondary T cell malignancies are caused by integrating viral vectors, transient expression of CARs in T cells may be a novel way of avoiding this, as the construct will be lost following replication of the T cells. Rheumatologists may prefer a transient, one-shot approach for the treatment of RA whereas haematologists generally aim for persistent CAR T cells to eliminate malignant B cells in cancer patients. Whether transient CAR expression is equivalent to persistent CAR T cells in terms of clinical efficacy, depletion of autoreactive B cells and long-term remission in RA will need to be tested in both pre-clinical models and human patients. Delivering CAR genes or CAR mRNA into T cells via non-viral methods including electroporation, lipid nanoparticles, polymers or peptides will not result in integration within the host T cell genome, resulting in transient expression and thus a potentially safer product. As an example, the Epstein group used lipid nanoparticles (LNP) to deliver mRNA encoding a FAP-targeting CAR into murine T cells in vivo for the treatment of cardiac fibrosis [37]. FAP was chosen as the target of the CAR T cells as it is upregulated in fibrotic/scar tissue. The LNPs encapsulated the FAP-targeting CAR mRNA and were additionally coated with antibodies to CD5 to enable targeting of T cells in-vivo, as CD5 is highly expressed on T cells and on some subsets of B cells. Up to 25% of circulating T cells in the mice were engineered with the FAP-specific CAR which correlated with regression of cardiac injury. CAR expression was transient with the loss of this receptor on T cells one-week post-infusion [37]. In a new study, CD8-targeted LNPs were developed for selective delivery of CAR mRNA into CD8 + T cells for targeting of B cell cancers or autoimmunity [48]. These LNPs delivered CD19-specific CAR mRNA into CD8 + T cells of healthy PBMC donors and autoimmune patients in-vitro and reprogrammed the cells to recognise and eliminate autologous B cells. In addition, administration of these targeted LNPs directly to humanised mice and cynomolgus monkeys in-vivo led to complete elimination of B cells within hours. In theory, FAP or CD19-targeting CAR T cells could be generated in vivo for the treatment of RA patients using LNPs where they selectively deliver the CARs into T cells to redirect the cells towards FAP (expressed on fibroblasts within the joints of RA patients) or B cells of these patients. This method of generating CAR T cells in vivo through novel delivery approaches offers a more simplistic manufacturing procedure (T cell engineering in-vivo instead of ex-vivo) using ‘Off-the-shelf’ LNPs that can be administered to all patients resulting in a potentially safer product. The disadvantages of in-vivo approaches include not being able to characterise the ‘product’ prior to infusion (since only the LNP with the encapsulated CAR construct is administered to the patient) and the likelihood for repeated dosing to achieve clinical efficacy [48]. Anti-drug antibodies have been reported in autoimmune patients treated with CAR T cell therapy [29], which suggests that repeated dosing may result in anti-drug antibodies.

Given the complex manufacturing procedures, costs and inherent safety risks associated with current CAR T cell therapies, it is likely that their exploration as advanced therapeutics for RA will continue to focus on difficult-to-treat/refractory patients with seropositive disease. RA is generally not life-threatening and there are a variety of therapeutic options that adequately control inflammation of the joints and symptoms for a lot of patients. Nonetheless, the prospect of a ‘cure’ in difficult-to-treat/refractory patient populations could result in other RA cohorts being trialled with these therapies in the future. One cohort that could benefit from CAR T cell therapy could be seropositive RA patients who have extra-articular features including interstitial lung disease and/or vasculitis which can be life threatening and refractory to conventional therapies. For RA patients who may be considered for CAR T cell therapy, it is likely that a pan B cell depletion approach (via targeting of CD19 + B cells or BCMA + B cells) will continue. This is due to the enormous body of clinical and regulatory data that exists on CAR T cell manufacture, efficacy and side-effects with these cells, much of which can be leveraged for the treatment of RA and other autoimmune diseases. Other approaches may include allogenic ‘Off-the-shelf’ CAR T cells that are being actively studied for cancer [49], ‘in-vivo’ CAR T cells (described above) or dual engineered CAR T cells that express anti-inflammatory cytokines such as IL-10 together with cytokine receptors as ‘sinks’ to trap pro-inflammatory signals [50]. This latter synthetic biology approach is functionally similar to CAR Tregs described earlier as they suppress active inflammation. Elsewhere, several clinical trials are currently underway or are planned to treat relapsed/refractory RA patients with autologous CAR T cells targeting either CD19, BCMA or both (either sequentially or together), allogenic ‘Off-the-shelf’ γδ CAR T cells (a subset of T cells with lower levels of alloreactivity against host tissues) and CAR T cells with decreased manufacturing times (YBT323 and CC-97540/BMS986353). Other strategies that could be explored in the future is the selective depletion of autoreactive B or T cells in RA patients which offers a more refined approach compared to pan B cell depletion. The pre-clinical studies highlighted in this review have demonstrated that depletion of autoreactive B/T cells, as well as FAP-expressing fibroblasts, is possible in-vitro, but translating this into the clinic will be challenging. Interestingly, a clinical study in 2023 reported a patient with Ankylosing spondylitis who was treated with an antibody targeting autoreactive T cells via recognition of a T cell receptor β chain variant that is associated with autoimmune disease [51]. The therapy was successful, did not cause systemic immunosuppression and resulted in drug-free remission (> 4 years at latest update). As this is an antibody-based approach, three doses are required per year. Given the potential of CAR T cells for inducing deep tissue depletion of CD19 + B cells with a single infusion of cells, cell-based approaches using CARs expressing conventional antibody fragments (scFvs) instead of antibodies may provide novel ways for treating RA effectively while avoiding immunosuppression.

Authors’ Contributions

Michael Freeley conceived the idea for this review and write the whole paper. The manuscript was written without the use of Generative Artificial Intelligence.

Data Availability

No datasets were generated or analysed during the current study.

Declarations

Competing interests

The authors declare no competing interests.