Stem cell therapy

Stem cells are types of cells that have both self-renewal and multidirectional differentiation potential, and have the advantages of high proliferation, immune regulation and multidirectional differentiation. They are classified into totipotent stem cells, pluripotent stem cells and unipotent stem cells based on differentiation potential or into embryonic stem cells and adult stem cells based on developmental stage. Stem cell therapy utilises the immune regulation and multidirectional differentiation capabilities of human stem cells to transplant healthy stem cells into a patient’s body to achieve the effects of regulating immunity and repairing tissues.

Mesenchymal stem cells (MSCs)

MSCs are multipotent stem cells derived from the mesoderm with high self-renewal ability and multidirectional differentiation potential. They can not only migrate to damaged tissues, but also release growth factors, cytokines and chemokines, thereby promoting tissue regeneration (Refs Reference Fu, Liu, Halim, Ju, Luo and Song2, Reference Gnecchi, Danieli and Malpasso3). In addition, MSCs also have a wide range of immunomodulatory abilities (Refs Reference Zhu, Ma and Yang4, Reference Abumaree, Al Jumah and Kalionis5). They can inhibit the differentiation and maturation of dendritic cells and the proliferation of natural killer cells and induce the transformation of M1-like macrophages to M2-like macrophages, thereby playing anti-inflammatory and antigen presentation-reduction roles. They can promote the differentiation of Th2 cells and regulatory T cells (Treg) and the secretion of the anti-inflammatory factors IL-4, IL-10 and TGF-β, and inhibit the differentiation of Th1 cells and Th17 cells and the secretion of the pro-inflammatory factors IFN-γ and IL-17A. In addition, MSCs can inhibit B cell activation and differentiation, reduce the production of plasma cells and autoantibodies and promote Breg cell proliferation. Clinical studies on MSCs in several common AIDs have been carried out and have achieved satisfactory results.

SLE is an AID affecting multiple organs, including the kidneys, brain and lungs, and is characterised by T cell dysfunction and polyclonal B cell activation. Lupus nephritis (LN) is a major cause of death in SLE patients (Refs Reference Liang, Zhang, Hua, Wang, Lu, Shi, Hou, Zeng, Gilkeson and Sun6, Reference Koarada and Tada7). Previous animal studies showed that MSCs improved lupus symptoms and reversed multi-organ dysfunction. A pilot study on four refractory SLE patients found that MSC infusion reduced disease activity (SLEDAI), improved renal function and serological markers, with no adverse events over 18 months (Ref. Reference Sun, Akiyama, Zhang, Yamaza, Hou, Zhao, Xu, le and Shi8). These results were confirmed in 15 additional refractory SLE patients (Ref. Reference Liang, Zhang, Hua, Wang, Lu, Shi, Hou, Zeng, Gilkeson and Sun6). In 81 refractory LN patients, MSC treatment lowered SLEDAI scores, proteinuria and serum urea nitrogen and creatinine levels, with 49 achieving remission within 12 months (Ref. Reference Gu, Wang, Zhang, Feng, Gilkeson, Shi and Sun9). A 4-year study on 87 severe refractory SLE patients showed symptom improvement in most, 49% achieving complete remission, and no adverse events (Ref. Reference Wang, Zhang, Liang, Li, Feng, Wang, Hua, Liu, Lu, Gilkeson, Silver, Chen, Shi and Sun10). A subsequent multicentre trial (40 active SLE patients) reported significant reductions in SLEDAI, BILAG scores and 24-hour proteinuria, alongside improved serological markers; however, some patients relapsed after 6 months, suggesting a need for a second infusion, though all tolerated treatment well (Ref. Reference Wang, Li, Zhang, Zhang, Chen, Li, Hu, Jiang, Shi and Sun11). A recent phase I trial on six multiracial SLE patients confirmed good efficacy and safety, but larger studies are needed (Ref. Reference Kamen, Wallace, Li, Wyatt, Paulos, Wei, Wang, Wolf, Nietert and Gilkeson12). Finally, in 35 SLE patients with refractory cytopenia, MSC infusion significantly increased blood cell counts and reduced SLEDAI scores, indicating improvement of haematological abnormalities (Ref. Reference Li, Wang, Liang, Zhang and Sun13).

RA is an AID characterised by joint lesions, with synovial inflammation leading to joint destruction and deformity. Conventional medicines have limited efficacy in some patients. Wang et al. enrolled 172 refractory RA patients and found that adding MSCs to disease-modifying antirheumatic drugs (DMARDs) significantly improved outcomes after 3 months. Compared with DMARDs alone, the combination reduced HAQ scores, DAS28 scores, CRP and RF levels, increased Treg levels and improved quality of life, with no serious adverse events (Ref. Reference Wang, Wang, Cong, Liu, Zhou, Bai, Li, Bai, Li, Ji, Zhu, Wu and Liu14). This first study had only 8 months of follow-up, so the team later followed 64 patients for 3 years, confirming long-term efficacy and safety (Ref. Reference Wang, Huang, Li, Li, Shi, Bai, Wang, Zheng and Liu15). However, this was a single-centre study. Álvaro Gracia et al. conducted a multicentre, placebo-controlled phase Ib/IIa trial in 53 refractory RA patients, showing symptom improvement; transient fever was the most common adverse event, with no life-threatening events (Ref. Reference Álvaro-Gracia, Jover, García-Vicuña, Carreño, Alonso, Marsal, Blanco, Martínez-Taboada, Taylor, Martín-Martín, DelaRosa, Tagarro and Díaz-González16). The first three studies used allogeneic MSCs; the subsequent two evaluated autologous MSCs. Ghoryani et al. treated nine refractory RA patients with autologous MSCs, significantly reducing DAS28 and VAS pain scores without complications (Ref. Reference Ghoryani, Shariati-Sarabi, Tavakkol-Afshari, Ghasemi, Poursamimi and Mohammadi17). Another autologous MSC study reported sustained decreases in 66/68 joint scores for up to 52 weeks, with only four mild adverse events (Ref. Reference Vij, Stebbings, Kim, Park and Chang18). A randomised, triple-blind, placebo-controlled phase 1/2 study in 30 knee-joint RA patients found reduced pain and improved function (Ref. Reference Shadmanfar, Labibzadeh, Emadedin, Jaroughi, Azimian, Mardpour, Kakroodi, Bolurieh, Hosseini, Chehrazi, Niknejadi, Baharvand, Gharibdoost and Aghdami19). Since RA stage affects the immune microenvironment, He and Yang showed that high serum IFN-γ levels positively correlated with reduced DAS28 and may predict MSC efficacy (Ref. Reference Yang, He, Zhao, Guo, Zhu, Xing, Jiang, Liu and Xu20). Subsequently, a phase 1/2 randomised controlled study in 63 refractory RA patients found that adding IFN-γ to MSCs enhanced clinical improvement and increased the Treg/Th17 ratio, with no serious acute adverse events (Ref. Reference He, Yang, Yao, Yang, Ao, Hu, Li, Wu, Tan, Xing, Guo, Bellanti, Zheng and Xu21). Further research is needed to confirm this strategy’s safety and effectiveness.

SjD is an AID characterised by reduced lacrimal and salivary gland secretion, potentially affecting multiple systems. Xu et al. reported that after MSC infusion,SjD patients showed significantly decreased sjögren’s syndrome disease activity index (SSDAI) and VAS dryness scores, increased salivary flow rate and no adverse events (Ref. Reference Xu, Wang, Liu, Fan, Zhang, Liu, Ding, Gao, Zhang, Ding, Bromberg, Chen, Sun and Wang22). This first demonstration confirms MSC infusion as an effective SjD therapy, supporting further exploration. Autoimmune liver disease (ALD) results from immune dysfunction (Ref. Reference Zhu, Ma and Yang4). Primary biliary cirrhosis (PBC) is a progressive ALD; some patients are refractory to existing drugs. Two clinical studies found that after MSC treatment, most PBC patients had significantly reduced symptoms, such as fatigue and itching (Refs Reference Wang, Li, Liu, Li, Fu, Sun, Xu, Lin, Wang, Lv, Chen, Zou, Li, Shi, Zhang and Wang23, Reference Wang, Han, Chen, Wang, Shan, Kong, Yang, Li, Zhang, Dong, Wang, Xu, Hu, Wang, Keating, Bi, Zhang and Zhao24), and their serum alkaline phosphatase (ALP) (Ref. Reference Wang, Li, Liu, Li, Fu, Sun, Xu, Lin, Wang, Lv, Chen, Zou, Li, Shi, Zhang and Wang23), glutamyl transferase (GGT) (Refs Reference Wang, Li, Liu, Li, Fu, Sun, Xu, Lin, Wang, Lv, Chen, Zou, Li, Shi, Zhang and Wang23, Reference Wang, Han, Chen, Wang, Shan, Kong, Yang, Li, Zhang, Dong, Wang, Xu, Hu, Wang, Keating, Bi, Zhang and Zhao24), alanine aminotransferase (ALT) (Ref. Reference Wang, Han, Chen, Wang, Shan, Kong, Yang, Li, Zhang, Dong, Wang, Xu, Hu, Wang, Keating, Bi, Zhang and Zhao24), aspartate aminotransferase (AST) (Ref. Reference Wang, Han, Chen, Wang, Shan, Kong, Yang, Li, Zhang, Dong, Wang, Xu, Hu, Wang, Keating, Bi, Zhang and Zhao24) and IgM (Ref. Reference Wang, Han, Chen, Wang, Shan, Kong, Yang, Li, Zhang, Dong, Wang, Xu, Hu, Wang, Keating, Bi, Zhang and Zhao24) levels decreased significantly. All patients tolerated the treatment well, and no obvious short-term side effects or long-term adverse events were observed during treatment and follow-up (Refs Reference Wang, Li, Liu, Li, Fu, Sun, Xu, Lin, Wang, Lv, Chen, Zou, Li, Shi, Zhang and Wang23, Reference Wang, Han, Chen, Wang, Shan, Kong, Yang, Li, Zhang, Dong, Wang, Xu, Hu, Wang, Keating, Bi, Zhang and Zhao24).

MS is an AID characterised by inflammatory demyelinating lesions in the central nervous system white matter. The first MSC treatment for an MS patient in 2009 reduced the Expanded Disability Status Scale (EDSS) score and CNS lesions (Ref. Reference Liang, Zhang, Hua, Wang, Wang, Han and Sun25). Subsequently, a phase 2a study in 10 refractory MS patients showed improved visual function and physiological/structural measures after MSC infusion (Ref. Reference Connick, Kolappan, Crawley, Webber, Patani, Michell, du, Luan, Altmann, Thompson, Compston, Scott, Miller and Chandran26). A randomised two-arm study of 23 refractory MS patients found that adding MSCs to traditional drugs significantly improved overall symptoms, EDSS scores and relapse rates compared with drugs alone, with no notable adverse reactions over 1 year (Ref. Reference Li, Zhang and Geng27). Another phase 1/2 study in 18 refractory MS patients receiving 7-day continuous MSC infusion reported improved EDSS scores, bladder/bowel/sexual dysfunction, non-dominant hand scores and walking time; 1 year later, MRI of 15 patients showed inactive lesions, and no serious adverse events occurred, indicating continuous infusion is safe and effective (Ref. Reference Riordan, Morales, Fernández, Allen, Fearnot, Leckrone, Markovich, Mansfield, Avila, Patel, Kesari and Paz Rodriguez28). Comparing infusion routes, Petrou et al. treated 48 refractory MS patients and found that both intrathecal (IT) and intravenous (IV) MSC infusions improved symptoms and quality of life, with IT showing greater efficacy (Ref. Reference Petrou, Kassis, Levin, Paul, Backner, Benoliel, Oertel, Scheel, Hallimi, Yaghmour, Hur, Ginzberg, Levy, Abramsky and Karussis29). Finally, Jamali et al. conducted a phase I/II study in 35 refractory MS patients receiving single- or double-dose combined MSC infusions. The double-dose group had more significant improvements in damage load, cortical thickness, manual dexterity and information processing speed, with good tolerance in all patients (Ref. Reference Jamali, Aldughmi, Atiani, al-Radaideh, Dahbour, Alhattab, Khwaireh, Arafat, Jaghbeer, Rahmeh, Abu Moshref, Bawaneh, Hassuneh, Hourani, Ababneh, Alghwiri and Awidi30).

T1DM is an immune-mediated disease that destroys pancreatic β cells. Hu et al. enrolled 29 newly diagnosed T1DM patients and found that MSC treatment lowered glycated haemoglobin (HbA1c) and insulin dose but raised C-peptide, with no adverse events over 2 years (Ref. Reference Hu, Yu, Wang, Wang, Wang, Gao, Chen, Zhao, Jia, Yan and Wang31). Cai et al. enrolled 42 patients with T1DM and showed that MSCs plus autologous bone marrow mononuclear cells improved insulin secretion and C-peptide measures (Ref. Reference Cai, Wu, Xu, Liao, Chen, Huang, Wu, Luo, Wu, Pugliese, Pileggi, Ricordi and Tan32). Over 8 years, the treatment group had lower rates of neuropathy, nephropathy and retinopathy, with no serious adverse events (Ref. Reference Wu, Xu, Cai, Chen, Huang, Wu, Pugliese, Li, Ricordi and Tan33).

Inflammatory bowel disease (IBD) is a chronic inflammatory gastrointestinal disorder. A phase 2 study in 15 refractory Crohn’s disease (CD) patients reported that MSC infusion led to clinical remission in eight patients and endoscopic improvement in seven patients, but lacked a control group (Ref. Reference Forbes, Sturm, Leong, Sparrow, Segarajasingam, Cummins, Phillips and Herrmann34). A subsequent randomised controlled trial in 82 refractory CD patients showed that MSC treatment significantly reduced CD activity index (CDAI), Harvey-Bradshaw index (HBI) and corticosteroid use, with four patients developing fever (Ref. Reference Zhang, Lv, Liu, Song and Shi35). A dose-escalation study found that low-dose MSC achieved a higher fistula healing rate than high-dose (Ref. Reference Molendijk, Bonsing and Roelofs36). Four-year follow-up confirmed sustained fistula healing and improved quality of life (Ref. Reference Barnhoorn, Wasser, Roelofs, Maljaars, Molendijk, Bonsing, Oosten, Dijkstra, van der Woude, Roelen, Zwaginga, Verspaget, Fibbe, Hommes, Peeters and van der Meulen-de Jong37). A phase 3 trial in 212 CD patients with perianal fistulas reported higher fistula closure with MSC; common adverse events were anal abscess and rectal pain (Ref. Reference Panés, García-Olmo and Van Assche38). A phase IB/IIA study in rectovaginal fistulas showed that 50% of patients achieved complete clinical and radiological healing after 6 months (Ref. Reference Lightner, Reese, Ream, Nachand, Dadgar, Adams, VanDenBossche, Pineiro and Hull39). A recent phase I paediatric study demonstrated decreased CDAI, Wexner and Van Assche scores, with 83% complete healing; five patients experienced perianal pain. No serious adverse reactions occurred in these studies (Ref. Reference Lightner, Otero-Pineiro, Reese, Ream, Nachand, Adams, VanDenBossche and Kurowski40).

MSCs offer a promising regenerative approach for AIDs due to unique immunoregulatory mechanisms. Key advantages include: 1) dynamic inflammation-sensitive regulation – MSCs suppress overactive T cells, B cells and NK cells while expanding regulatory T cells, converting a pro-inflammatory Th1/Th17 phenotype to an anti-inflammatory state without indiscriminate immunosuppression; 2) low immunogenicity – low expression of MHC class II and co-stimulatory molecules allows allogeneic use with generally mild adverse reactions (e.g., transient fever); 3) tissue repair – MSCs home to damaged sites and secrete neurotrophic and vascular endothelial growth factors, inhibiting fibrosis and apoptosis.

MSCs source, donor age, culture methods, passage number and cryopreservation affect function. Lack of standardised preparation and potency testing leads to variable efficacy. Some studies report cancer risk, e.g., MSCs promoting myeloma growth (Ref. Reference Xu, Menu and De Becker41). Not all patients respond; microenvironment differences (e.g., cytokine storms, immune memory) may influence outcomes. In one trial, 9 of 12 MSC-treated patients improved, but 5 of 6 placebo patients also improved, and 1 MSC recipient died from severe pneumonia (Ref. Reference Deng, Zhang, Guo and Lim42). High titres of autoantibodies or complement activation in some patients with AIDs may rapidly clear the infused MSCs via antibody-dependent cell-mediated cytotoxicity or complement-dependent cytotoxicity, leading to treatment failure (Ref. Reference Gilkeson43). Additionally, high costs and variable regulations limit accessibility. Future directions should focus on genetically modified ‘third-generation’ MSCs, optimised regimens (local injection, combination with immunosuppressants) and functional marker-based quality control.

MSC-derived extracellular vesicles (MSC-EVs)

MSC-EVs are double-layered membrane structures secreted by MSCs, containing various proteins, lipids and nucleic acids. They can be divided into three categories based on size and biological components: exosomes, apoptotic bodies and microvesicles (Refs Reference Seo, Kim and Hong44, Reference Abels and Breakefield45). MSC-EVs have similar biological functions to MSCs, but they are more stable and can largely avoid adverse events, such as immune response and capillary embolism (Refs Reference Seo, Kim and Hong44, Reference Withrow, Murphy, Liu, Hunter, Fulzele and Hamrick46). In addition, some studies have shown that the effects of MSCs are mainly attributed to the effects of MSC-EVs (Ref. Reference Harrell, Jovicic, Djonov, Arsenijevic and Volarevic47). At present, the research on MSC-EVs in AIDs is mainly preclinical.

Studies show that MSC-EVs exert therapeutic effects in multiple AID models. In SLE, MSC-EVs inhibit CD4⁺T cell proliferation, promote TGF-β1 production, regulate macrophage phenotype and reduce alveolar haemorrhage (Refs Reference Xie, Li, She, Wu, Mao, Hun, Luo, Wan, Tian and Wen48, Reference Chen, Wei, Sun, Zhang, Yang, Tao and Nong49). In RA, MSC-EVs reduce CD4⁺/CD8⁺T cells (Refs Reference Cosenza, Toupet, Maumus, Luz-Crawford, Blanc-Brude, Jorgensen and Noël50, Reference Ma, Xu, Zhang, Liu, Gao, Tian, Wei, Li and Zhang51), downregulate CXCL9 to suppress fibroblast-like synoviocyte (FLS) activation (Ref. Reference Meng and Qiu52) and deliver miR-34a to inhibit cyclinI/ATM/ATR/p53 signalling, reducing proliferation and promoting apoptosis (Ref. Reference Wu, Zhou, Wang, Cheng, Hu, Wang, Luo, Huang and Gu53). In SjD, MSC-EVs inhibit Th17 cells and pro-inflammatory factors, promote Treg cells and anti-inflammatory factors, restore salivary gland function (Refs Reference Hai, Shigemoto-Kuroda, Zhao, Lee and Liu54, Reference Li, Xing, Gan, He and Hua55), and in a dry eye rabbit model, induce M2 macrophage polarisation and Treg differentiation via miR-100-5p (Ref. Reference Li, Gao, Zhao, du, Ma, Nian and Wei56). MSC-EVs also decreased the expression of IL-17 (Refs Reference Ahmadvand Koohsari, Absalan and Azadi57–Reference Heidari, Abbasi-Kenarsari, Namaki, Baghaei, Zali, Ghaffari Khaligh and Hashemi59), IL-12 (Ref. Reference Heidari, Abbasi-Kenarsari, Namaki, Baghaei, Zali, Ghaffari Khaligh and Hashemi59), IFN-γ (Refs Reference Ahmadvand Koohsari, Absalan and Azadi57–Reference Heidari, Abbasi-Kenarsari, Namaki, Baghaei, Zali, Ghaffari Khaligh and Hashemi59) and TNF-α (Refs Reference Ahmadvand Koohsari, Absalan and Azadi57, Reference Heidari, Abbasi-Kenarsari, Namaki, Baghaei, Zali, Ghaffari Khaligh and Hashemi59) as well as proliferation of Th1 (Refs Reference Hosseini Shamili, Alibolandi, Rafatpanah, Abnous, Mahmoudi, Kalantari, Taghdisi and Ramezani60–Reference Tian, Zhu, Zhang, Bian, Hong, Shen, Xu, Rui, Yin and Wang62) and Th17 cells (Refs Reference Shigemoto-Kuroda, Oh, Kim, Jeong, Park, Lee, Park, Kim, An, Prockop and Lee61, Reference Tian, Zhu, Zhang, Bian, Hong, Shen, Xu, Rui, Yin and Wang62) in MS, T1DM and IBD. Moreover, they increased the expression of IL-4 (Refs Reference Ahmadvand Koohsari, Absalan and Azadi57–Reference Heidari, Abbasi-Kenarsari, Namaki, Baghaei, Zali, Ghaffari Khaligh and Hashemi59), IL-10 (Refs Reference Ahmadvand Koohsari, Absalan and Azadi57–Reference Heidari, Abbasi-Kenarsari, Namaki, Baghaei, Zali, Ghaffari Khaligh and Hashemi59) and TGF-β (Refs Reference Nojehdehi, Soudi, Hesampour, Rasouli, Soleimani and Hashemi58, Reference Heidari, Abbasi-Kenarsari, Namaki, Baghaei, Zali, Ghaffari Khaligh and Hashemi59) as well as the proliferation of Treg cells (Refs Reference Ahmadvand Koohsari, Absalan and Azadi57, Reference Nojehdehi, Soudi, Hesampour, Rasouli, Soleimani and Hashemi58, Reference Hosseini Shamili, Alibolandi, Rafatpanah, Abnous, Mahmoudi, Kalantari, Taghdisi and Ramezani60, Reference Tolomeo, Castagliuolo, Piccoli, Grassi, Magarotto, de Lazzari, Malvicini, Caicci, Franzin, Scarpa, Macchi, de Caro, Angriman, Viola, Porzionato, Pozzobon and Muraca63). In addition, MSC-EVs reduced neuroinflammation and promoted myelin regeneration in the nervous system in an MS model mice (Refs Reference Zhang, Buller, Zhang, Zhang, Lu, Rosene, Medalla, Moore and Chopp64–Reference Clark, Zhang, Barthe, Kumar, Pivetti, Kreutzberg, Reed, Wang, Paxton, Farmer, Guo and Wang66) in addition to significantly up-regulating plasma insulin concentration in T1DM (Ref. Reference Shigemoto-Kuroda, Oh, Kim, Jeong, Park, Lee, Park, Kim, An, Prockop and Lee61). Further, in an IBD model, MSC-EVs regulated the polarisation of macrophages from the M1 to the M2 phenotype (Refs Reference Tolomeo, Castagliuolo, Piccoli, Grassi, Magarotto, de Lazzari, Malvicini, Caicci, Franzin, Scarpa, Macchi, de Caro, Angriman, Viola, Porzionato, Pozzobon and Muraca63, Reference Cao, Xu, Wang, Liu, Tian and Yuan67, Reference Liu, Liang, Wang, Zhou, Zheng, Hu, He, Wu and Lan68), promoted intestinal function recovery and maintained intestinal barrier integrity (Ref. Reference Yu, Yang, Xiao, Xu, Yang, Xue, Li, Wang and Zhao69).

MSC-EVs mediate the paracrine effects of MSCs and represent a promising ‘cell-free’ therapeutic strategy (Ref. Reference Martinez-Arroyo, Ortega, Forner and Cortes70). Compared with MSCs, MSC-EVs offer several advantages: 1) Safety – they lack self-replicating ability, avoiding tumorigenic risk; low expression of MHC-II reduces immunogenicity and rejection. 2) Scalable production – EVs can be mass-produced from expanded MSCs and purified by ultracentrifugation or tangential flow filtration, enabling standardisation. 3) Potent immunoregulation – EVs are rich in miRNAs and proteins, acting on multiple immune targets (Refs Reference Cosenza, Ruiz, Maumus, Jorgensen and Noël71, Reference Liu, Li, Liu, Yang, Yin and Xie72). 4) Superior biodistribution – their small size allows crossing of biological barriers (e.g., blood–brain barrier) to precisely reach damaged tissues, avoiding pulmonary entrapment seen with MSCs (Ref. Reference Liu, Li, Liu, Yang, Yin and Xie72). 5) Easier logistics – EVs circumvent cold-chain requirements, simplifying storage and transport.

First, MSCs source, culture conditions (2D/3D, hypoxia) and isolation methods cause batch-to-batch variability; no internationally accepted production or potency standards exist. Second, homing mechanisms, pharmacokinetics and in vivo fate are unclear, hampering rational dosing. Third, EVs may be rapidly cleared by the mononuclear phagocyte system, requiring strategies to improve circulatory stability. Fourth, mass production still requires large numbers of MSCs; bioreactor-based 3D culture with microcarriers could improve efficiency but remains challenging. Finally, no unified efficacy evaluation standard exists; MSC-EVs cannot be defined solely by physicochemical properties and require complex bioactivity assays. Regulatory pathways need further refinement. Future directions should integrate gene editing and materials science to develop engineered, targeted, stable, ‘off-the-shelf’ MSC-EVs products.

Engineered MSCs (E-MSCs)

Although MSCs have the ability to migrate to damaged tissues, they are easily trapped by pulmonary capillaries, causing ectopic osteogenesis, tumour formation, immune rejection and reduced efficacy caused by local microenvironment factors. In order to improve the targeted homing and therapeutic efficiency of MSCs, MSCs are modified through various technical means, such as gene modification to induce the expression of different proteins and soluble factors and improve their inherent properties; surface modification to increase the number of MSCs transported to the target through covalent or non-covalent interactions and binding to overexpressed proteins in diseased or damaged tissues and to enhance the therapeutic effect; and tissue engineering using MSCs as seed cells and combining them with biological scaffolds to promote tissue construction. At present, the research on engineered MSCs in AIDs is mainly preclinical.

In SLE mice, MSCs overexpressing IL-37 had a strong immunosuppressive effect (Ref. Reference Xu, Chen, Li, Lian, Huang, Lai, Li and Huang73), and MSCs encapsulated in porous adhesive hydrogel microparticles can prolong survival time (Ref. Reference Nie, Chen, Zhao, Gan, Alip, Zhao and Sun74). They can reduce the expression of proinflammatory factors as well as total antibodies and autoantibodies in serum and urine more than MSCs (Ref. Reference Xu, Chen, Li, Lian, Huang, Lai, Li and Huang73). They can also promote the transformation of inflammatory macrophages to anti-inflammatory phenotypes and promote the expression of anti-inflammatory factors (Ref. Reference Nie, Chen, Zhao, Gan, Alip, Zhao and Sun74), thereby improving the disease progression of SLE mice. In RA mice, infusion of transgenic MSCs expressing human soluble TNFR2 can alleviate arthritis symptoms and reduce circulating TNF-α levels more effectively than the TNF-α inhibitor etanercept (Ref. Reference Liu, Wang, Hendricks, Lee, Bohnlein, Junker and Mosca75). In addition, after nano-modification of MSCs, their migration, anti-inflammatory, antioxidant and chondrogenic abilities are enhanced, promoting the polarisation of macrophages from the M1 type to the M2 type, thereby effectively inhibiting synovial inflammation in RA rats and reducing cartilage erosion (Ref. Reference Lu, Li, Li, Chen, Xu, Lin, Zhang, Zhu, Ding and Mao76). Infusion of ICAM-1-overexpressing MSCs (Ref. Reference Li, Wang, Ding, Wang, Zhao, Mao, Wu, Wang, Zhu and Ning77) and IL-25-induced MSCs (Ref. Reference Cheng, Su and Hu78) into IBD mice inhibited the proliferation of Th1 (Ref. Reference Li, Wang, Ding, Wang, Zhao, Mao, Wu, Wang, Zhu and Ning77) and Th17 cells (Refs Reference Li, Wang, Ding, Wang, Zhao, Mao, Wu, Wang, Zhu and Ning77, Reference Cheng, Su and Hu78); promoted the production of Treg cells (Ref. Reference Li, Wang, Ding, Wang, Zhao, Mao, Wu, Wang, Zhu and Ning77); reduced local inflammatory responses (Ref. Reference Li, Wang, Ding, Wang, Zhao, Mao, Wu, Wang, Zhu and Ning77) and promoted the proliferation and migration of intestinal epithelial cells, with better effects than before modification (Ref. Reference Su, Xie and Fan79).

E-MSCs offer several advantages. First, engineering enhances targeting efficiency – e.g., coupling chemokine receptors or peptides improves homing to inflamed tissues (Refs Reference Liu, Wang, Hendricks, Lee, Bohnlein, Junker and Mosca75, Reference Lu, Li, Li, Chen, Xu, Lin, Zhang, Zhu, Ding and Mao76). Second, E-MSCs reduce MHC-I expression or overexpress anti-apoptotic genes, delaying immune clearance and prolonging half-life (Ref. Reference Nie, Chen, Zhao, Gan, Alip, Zhao and Sun74). Third, by knocking out pro-inflammatory pathways or overexpressing immunoregulatory molecules, E-MSCs achieve more stable and potent therapeutic effects (Ref. Reference Su, Xie and Fan79). Fourth, introducing ‘standardised’ functional modules masks inherent MSC variability, making the final product more controllable and regulatory-compliant.

Gene editing raises safety concerns; viral vectors may cause oncogene activation, and long-term safety needs verification. Over-modification may compromise MSCs’ proliferation and stemness. Chemically coupled or non-integrating modifications may degrade rapidly in vivo. E-MSCs also increase process complexity and quality control difficulty. Their in vivo behaviour can be more unpredictable, and they face dual regulatory requirements as both cell and gene therapy products. Future directions should focus on safer editing tools (non-viral vectors, mRNA electroporation), controllable ‘safety switches’, standardised production, universal off-the-shelf products and detailed pharmacokinetic/pharmacodynamic studies using precise preclinical models. Balancing safety, efficacy, controllability and accessibility is essential for translating E-MSCs into clinical AID treatment.

Haematopoietic stem cells (HSCs)

HSCs are the source of all blood and immune cells and have the potential for self-renewal, multidirectional differentiation and homing (i.e., directional migration to haematopoietic tissues and organs). They can differentiate into red blood cells, white blood cells and platelets and are multipotent (Refs Reference Eaves80, Reference Bujko, Kucia, Ratajczak and Ratajczak81). HSC transplantation (HSCT) is a procedure that uses high-dose radiotherapy and chemotherapy to pre-treat the recipient to remove tumours or abnormal cells from the recipient’s body, and then transplants autologous or allogeneic HSCs to the recipient, allowing the recipient to reconstruct a normal haematopoietic and immune system (Ref. Reference Alexander, Greco and Snowden82).

Since 1997, HSCT has been used to treat AIDs (Ref. Reference Dao, Vu, Nguyen, Hoang and Nguyen83). In SLE patients, HSCT reduced SLEDAI, proteinuria and autoantibodies, with 5-year progression-free survival of 67.9–91%, 10-year survival of 86% and overall survival of 86–95.2% (Refs Reference Leng, Jiang, Zhou, Tian, Li, Wang, Zhao, Shen, Zeng, Zhang, Tang, Dong and Zhao84–Reference Huang, Chen, Ren, Zhao, Guo, Gong, Zeng, Hu and Liu86). In RA patients, disease activity significantly decreased; some achieved ACR50-70, with effects lasting 2–5 years (Refs Reference Moghaddam, Mousavi and Ghotloo87–Reference Teng, Verburg, Sont, van den Hout, Breedveld and van Laar91). SjDpatients experienced relief of fatigue, dry mouth and joint pain, with normalised lymphocyte subsets and platelets (Refs Reference Yang and Xu92, Reference Zhiyi, Rongbin and Ying93). In MS patients, HSCT improved EDSS scores, quality of life and neurological function; 4-year relapse-free survival was 80%, and 5-year disability progression rate was 61.7% (Refs Reference Burt, Loh, Cohen, Stefosky, Balabanov, Katsamakis, Oyama, Russell, Stern, Muraro, Rose, Testori, Bucha, Jovanovic, Milanetti, Storek, Voltarelli and Burns94–Reference Boffa, Signori, Massacesi, Mariottini, Sbragia, Cottone, Amato, Gasperini, Moiola, Meletti, Repice, Brescia Morra, Salemi, Patti, Filippi, de Luca, Lus, Zaffaroni, Sola, Conte, Nistri, Aguglia, Granella, Galgani, Caniatti, Lugaresi, Romano, Iaffaldano, Cocco, Saccardi, Angelucci, Trojano, Mancardi, Sormani and Inglese96). T1DM patients showed increased C-peptide, reduced HbA1c and decreased insulin dose (Refs Reference Pastore, Assi, Ben Nasr, Bolla, Maestroni, Usuelli, Loretelli, Seelam, Abdelsalam, Zuccotti, D’Addio and Fiorina97–Reference Li, Shen, Ouyang, Hu, Hu, Cui, Zhang, Zhuge, Chen, Xu and Zhu100). CD patients achieved clinical and endoscopic remission within 6 months, with drug-free remission >60%, CDAI remission 70–80% and steroid-free survival 80% (Refs Reference Hasselblatt, Drognitz, Potthoff, Bertz, Kruis, Schmidt, Stallmach, Schmitt-Graeff, Finke and Kreisel101, Reference Burt, Craig, Milanetti, Quigley, Gozdziak, Bucha, Testori, Halverson, Verda, de Villiers, Jovanovic and Oyama102).

HSCT offers several advantages for severe, refractory AIDs. First, it provides long-term drug-free remission lasting 5–10 years or more, unattainable with conventional therapies. Second, HSCT has mature clinical protocols, complication management and long-term follow-up systems. Third, as a single intensive treatment, it can discontinue chronic medications and improve quality of life, offering long-term cost-effectiveness. Fourth, HSCT is now incorporated into multiple domestic and international AID treatment guidelines.

However, preconditioning requires high-dose chemotherapy or irradiation, causing infection risk or irreversible organ damage. Not all patients achieve long-term remission; 5-year progression-free survival varies, with relapse rates of 30–50% due to incomplete immune clearance or disease re-triggering. Reinfused grafts may still contain autoreactive cells. Allogeneic grafts risk graft-versus-host disease and higher mortality (Ref. Reference Zhang, Yuan, Meng, Hou, Li, Xue, Li, Wang, Nan, Jin and Zhang103). HSCT also requires large multidisciplinary centres, is expensive and has poor accessibility. Future directions include optimising preconditioning regimens to reduce toxicity, combining novel immunomodulators to prevent relapse, exploring allogeneic transplantation in high-risk populations and establishing better patient screening and long-term follow-up registries for precise risk-stratified management.

Immune cell therapy

Immune cell therapy refers to the collection of the body’s own immune cells (such as peripheral blood mononuclear cells), culturing them in vitro to increase their number and then returning them to the human body to induce immune tolerance and regulate the body’s immune ability.

Tregs

Tregs are a subset of T cells that control the body’s autoimmune reactivity. They are produced at the thymus and exported to the periphery. They inhibit the activation and proliferation of potential autoreactive T cells in the body through active regulation, thereby regulating the body’s immunity (Ref. Reference Sakaguchi, Mikami, Wing, Tanaka, Ichiyama and Ohkura104). Based on their surface markers, produced cytokines and mechanism of action, Treg cells can be divided into two categories: naturally occurring Treg cells (nTreg) and induced adaptive Treg cells (iTreg).

Preclinically, Treg cell transplantation reduces joint inflammation by migrating to the synovium, lowering anti-collagen IgG and Th17 cells, and in some contexts increasing CD4⁺IFNγ⁺Th1 cells (Refs Reference Morgan, Flierman, van Duivenvoorde, Witteveen, van Ewijk, van Laar, de vries and Toes105, Reference Sun, Yang, Huo, Zhu, Li, Jiang, Xie, Gao, Sun, Fan, Zhu and Yang106). Tregs also lower ALT/AST, restore Treg/Th1 balance and promote liver recovery in ALD mice (Ref. Reference Huang and Deng107); reduce neuroinflammation and enhance myelination in MS mice (Refs Reference Duffy, Keating, Perera, Lees, Tonkin, Makker, Carrive, Butovsky and Moalem-Taylor108, Reference Yang, Liu, Sun, Wang, Yuan, Fan, Ma, Lu, Yuan, Zou, Zhao, Li and Liu109); suppress pancreatic ICAM-1, pathogenic CD8⁺T cell migration and IFN-γ production in T1D mice (Ref. Reference Haque, Lei, Xiong, das, Ren, Fang, Salek-Ardakani, Yang and Song110) and decrease intestinal goblet cells, Th17 cells, IL-1β, TNF-α, NO and PGE2 in IBD mice (Ref. Reference Wang, Zhu and Wang111).

Clinically, Treg infusion improves cutaneous lupus activity (Ref. Reference Dall’Era, Pauli, Remedios, Taravati, Sandova, Putnam, Lares, Haemel, Tang, Hellerstein, Fitch, McNamara, Welch, Bluestone, Wofsy and Rosenblum112). In ALD patients, 22–44% of Tregs remain in the liver for 72 hours without serious adverse events (Ref. Reference Oo, Ackrill, Cole, Jenkins, Anderson, Jeffery, Jones, Jeffery, Lutz, Wawman, Athwal, Thompson, Gray, Guo, Barton, Hirschfield, Wong, Guest and Adams113). In T1DM, Treg therapy is feasible and safe, with stable persistence at 1 year (Refs Reference Bluestone, Buckner, Fitch, Gitelman, Gupta, Hellerstein, Herold, Lares, Lee, Li, Liu, Long, Masiello, Nguyen, Putnam, Rieck, Sayre and Tang114, Reference Marek-Trzonkowska, Myśliwiec, Dobyszuk, Grabowska, Derkowska, Juścińska, Owczuk, Szadkowska, Witkowski, Młynarski, Jarosz-Chobot, Bossowski, Siebert and Trzonkowski115); concurrent Treg and islet transplantation safely reduces immunosuppressive needs (Ref. Reference Bergström, Yao, Müller, Korsgren, Zur-Mühlen and Lundgren116). Most CD patients tolerate Treg infusion, and 40% show reduced CDAI (Ref. Reference Desreumaux, Foussat and Allez117). Primary sclerosing cholangitis (PSC) is an idiopathic, heterogeneous, cholestatic liver disease characterised by persistent, progressive biliary inflammation and fibrosis, often associated with IBD (Ref. Reference Lazaridis and LaRusso118). In patients with IBD and PSC, Treg infusion decreases bowel frequency, rectal bleeding, liver enzymes, DAI and Mayo scores (Ref. Reference Voskens, Stoica, Rosenberg, Vitali, Zundler, Ganslmayer, Knott, Wiesinger, Wunder, Kummer, Siegmund, Schnoy, Rath, Hartmann, Hackstein, Schuler-Thurner, Berking, Schuler, Atreya and Neurath119).

Treg therapy offers several advantages. First, Tregs are natural immune suppressors with precise, diverse mechanisms, suppressing effector T cells and modulating antigen presentation. Second, Treg therapy directly corrects the Treg/Teff imbalance central to AIDs, potentially inducing long-term tolerance and disease modification rather than merely controlling symptoms. Third, Tregs preferentially home to lymph nodes and inflamed tissues, concentrating immunosuppression at lesion sites and theoretically avoiding systemic infection or malignancy risk.

Treg stability is a core issue – inflammatory cytokines like IL-6 and TNF-α can cause Foxp3 loss and conversion into pro-inflammatory cells. Heterogeneity across Treg sources and differentiation stages complicates selection of an optimal subtype. Patients with AIDs often have reduced Treg counts or intrinsic functional defects, limiting ex vivo expansion. Clinical-grade expansion under GMP is extremely costly. Adoptively transferred Tregs typically survive only weeks to months without IL-2 support, yet IL-2 also expands effector T cells, creating a narrow therapeutic window. Future directions should focus on engineering strategies (safety switches, chemokine receptor co-expression) to improve targeting, safety and persistence; combining gene editing for large-scale production; optimising synergistic use of low-dose IL-2; and establishing patient screening systems to identify those with clear Treg defects, thereby enhancing response rates.

Chimeric antigen receptor T cells (CAR-T)

CAR-T therapy refers to the use of gene modification technology to transfer specific antigen recognition domains and T cell activation signals into T cells, so that T cells can directly bind to specific antigens on the surface of target cells and be activated. They can directly kill target cells by releasing perforin and granzyme B; at the same time, they can recruit endogenous immune cells in the human body to kill target cells by releasing cytokines, thereby helping to treat diseases (Ref. Reference Siyuan, Yan and Lu120).

CD19 is an important antigen on the surface of B cells. It undergoes a differentiation process from pre-B cells to mature B cells and can be used as a key target for CAR-T therapy. Mougiakakos et al. first treated a refractory SLE patient with active LN using autologous CD19 CAR-T cells, observing improved clinical symptoms and serology, reduced SLEDAI and no adverse events (Ref. Reference Mougiakakos, Krönke, Völkl, Kretschmann, Aigner, Kharboutli, Böltz, Manger, Mackensen and Schett121). Five additional refractory SLE patients showed consistent results (Ref. Reference Mackensen, Müller, Mougiakakos, Böltz, Wilhelm, Aigner, Völkl, Simon, Kleyer, Munoz, Kretschmann, Kharboutli, Gary, Reimann, Rösler, Uderhardt, Bang, Herrmann, Ekici, Buettner, Habenicht, Winkler, Krönke and Schett122). A 29-month study on eight refractory SLE patients reported sustained drug-free remission, with only mild upper respiratory infections and no major adverse events (Ref. Reference Müller, Taubmann, Bucci, Wilhelm, Bergmann, Völkl, Aigner, Rothe, Minopoulou, Tur, Knitza, Kharboutli, Kretschmann, Vasova, Spoerl, Reimann, Munoz, Gerlach, Schäfer, Grieshaber-Bouyer, Korganow, Farge-Bancel, Mougiakakos, Bozec, Winkler, Krönke, Mackensen and Schett123). Wang et al. treated five refractory SLE patients with CD19 CAR mRNA delivered via CD8⁺T-cell-targeted lipid nanoparticles, achieving in vivo generation of CAR-T cells for the first time. All five had decreased SLEDAI scores; two showed significant reductions in anti-nucleosome and anti-dsDNA antibodies. No major adverse events occurred (Ref. Reference Wang, Xiao, Zheng, Wang, Yang, Shi, Xiang, Wang, Zha, Schett and Chen124). Juvenile SLE is often more severe, frequently involving the kidneys, blood and nervous system (Ref. Reference Ambrose, Morgan, Galloway, Ionnoau, Beresford and Isenberg125). In two adolescent SLE patients, CD19 CAR-T infusion reduced SLEDAI and autoantibodies, improved renal function, normalised complement within 6 weeks and eliminated glomerular immune deposits (Refs Reference Krickau, Naumann-Bartsch, Aigner, Kharboutli, Kretschmann, Spoerl, Vasova, Völkl, Woelfle, Mackensen, Schett, Metzler and Müller126, Reference De Benedetti, Diomedi Camassei and Locatelli127).

For refractory RA, CD19 CAR-T alleviated arthritis and pain, lowering anti-CCP and RF levels (Refs Reference Haghikia, Hegelmaier, Wolleschak, Böttcher, Pappa, Motte, Borie, Gold, Feist, Schett and Mougiakakos128–Reference Li, Li, Zhao, Sheng, Xue, Schett, Shi, Hu, Wang and Chen131). However, one patient developed grade 3 cytokine release syndrome and grade 4 immune effector cell-associated neurotoxicity syndrome (Ref. Reference Lidar, Rimar, David, Jacoby, Shapira-Frommer, Itzhaki, Slobodin, Stern, Eshed, Sanalla, Marcus, Shimoni, Yerushalmi, Shem Tov, Danylesko and Avigdor130), suggesting that RA patients may face higher CAR-T-related toxicity due to high tissue B cell burden and severe inflammation. In refractory SjDpatients with diffuse large B-cell lymphoma, CD19 CAR-T normalised ANA, anti-Ro-52 and cytokines, improved dry mouth, reduced ESSDAI and achieved lymphoma remission by week 4 (Ref. Reference Sheng, Zhang, Song, Jiang, Cao, Li, Yi, Weng, Chen, Wang, Wu, Wang, Zhao and Yan132). In refractory MS, CAR-T improved clinical symptoms, reduced serum IgG and decreased cerebrospinal fluid oligoclonal bands (Ref. Reference Fischbach, Richter and Pfeffer133). In antisynthetase syndrome (ASS), CAR-T normalised serum myosin, CD8⁺T subsets and inflammatory cytokines, lowered anti-Jo-1, relieved myositis and interstitial lung lesions, though transient myalgia and CK elevation occurred (Refs Reference Pecher, Hensen, Klein, Schairer, Lutz, Atar, Seitz, Stanger, Schneider, Braun, Schmidt, Horger, Bornemann, Faul, Bethge, Henes and Lengerke134–Reference Müller, Boeltz, Knitza, Aigner, Völkl, Kharboutli, Reimann, Taubmann, Kretschmann, Rösler, Manger, Wacker, Mougiakakos, Jabari, Schröder, Uder, Roemer, Krönke, Mackensen and Schett136). In systemic sclerosis (SSc), CAR-T significantly improved heart, lung, skin and joint symptoms and reduced autoantibodies (Refs Reference Bergmann, Müller, Distler, Györfi, Völkl, Aigner, Kretschmann, Reimann, Harrer, Bayerl, Boeltz, Wirsching, Taubmann, Rösler, Spriewald, Wacker, Atzinger, Uder, Kuwert, Mackensen and Schett137–Reference Müller, Hagen, Wirsching, Kharboutli, Aigner, Völkl, Kretschmann, Tascilar, Taubmann, Bucci, Raimondo, Bergmann, Rothe, Corte, Tur, Muñoz, Böltz, Schuster, Hartmann, Garantziotis, Spörl, Vasova, Gerbitz, Spriewald, Kiener, Giannarelli, Locatelli, D’Agostino, Hanssens, Miltenyi, Bozec, Grieshaber-Bouyer, Mackensen and Schett140). In myasthenia gravis (MG), CAR-T improved muscle strength, reduced fatigue and lowered anti-AchR levels (Ref. Reference Haghikia, Hegelmaier, Wolleschak, Böttcher, Desel, Borie, Motte, Schett, Schroers, Gold and Mougiakakos141).

B cell maturation antigen (BCMA) is expressed in plasmablasts and plasma cells, and the severity of the disease may be related to the increase of BCMA (Ref. Reference Siyuan, Yan and Lu120). Therefore, BCMA can be used as another therapeutic target for CAR-T therapy. In seven refractory LN patients, autologous BCMA CAR-T reduced SLEDAI and autoantibodies, improved renal function and complement, decreased immune deposits on biopsy, and five achieved complete remission (Ref. Reference Hu, Cai, Yu, Chen, Wang, Wang, Zeng, He, Shen, Yu, Zeng, Chen, Luo, Zhang, Zhang, Xiong, Bai, Ye, Lin, Zhang, Ye, Li and Dong142). In MG, BCMA CAR-T improved muscle strength and lowered anti-AchR and anti-MuSK antibodies (Refs Reference Granit, Benatar, Kurtoglu, Miljković, Chahin, Sahagian, Feinberg, Slansky, Vu, Jewell, Singer, Kalayoglu, Howard, Mozaffar, Granit, Benatar, Mozaffar, Chahin, Howard, Slansky, Feinberg, Sahagian, Vu, Pereira, Steele, Paredes, Benjamin, Komanduri, Habib, Fong, de la Cruz, Dimitrova, Chopra, Holley, DeMaria, Tenorio, Requena, Brooks, Suresh, Farias, Miljković, Kurtoglu, Ngo Casi, Chowdhury, Kamboh, Stewart, Tosun, Shan, Daniel, Duvernay, Kireeva, English, Jewell, Singer and Kalayoglu143, Reference Tian, Qin, Dong, Heming, Zhou, Wang, Cai, You, Shang, Xiao, Wang, Li, Zhang, Bu, Meyer Zu Hörste and Wang144). SLE and LN are driven by pathogenic autoantibodies produced by CD19⁺B cells and BCMA⁺ plasma cells (Refs Reference Morand, Fernandez-Ruiz, Blazer and Niewold145, Reference Kiriakidou and Ching146). Therefore, researchers have proposed a dual-target (CD19 + BCMA) combination clearance strategy to achieve a more thorough therapeutic effect. Feng et al. treated 15 refractory SLE patients with anti-CD19 and anti-BCMA CAR-T cells; all showed clinical and serological improvement, reduced SLEDAI, and 12 achieved complete remission (Ref. Reference Feng, Huo, Hong, Jin, Cao, Shao, Wen, Zhang, Zhang, Fu, Wang, Xu, Wei, Cui, Huang, Cui, Chang, Liu, Lu, Lin, Hu and Huang147). To evaluate the long-term efficacy and safety of dual-target CAR-T cell therapy, Hong et al. conducted follow-up assessments of 12 patients with refractory SLE over a period of 19–67 months. The results demonstrated sustained improvement in clinical symptoms and serological features in all patients. One patient successfully achieved a natural pregnancy and delivered a healthy baby at full term, 15 and 36 months after treatment, respectively. This is the first report of a successful live birth following dual-target CAR-T therapy. During the treatment and follow-up period, no major adverse events occurred (Ref. Reference Hong, Wang, Zeng, Ding, He, Lan, DeStefano, Wada, Pinz, Chow, Choi, Hagag, Wang, Ma, Luo, Liang, Ding, Lu, Zhang, Ma and Wang148).

Compared with autologous CAR-T cell therapy, universal CAR-T cell therapy has the following advantages: 1) a wide source that can meet the needs of a large number of patients, 2) faster as it eliminates the delays in the manufacturing process of autologous CAR-T therapy and 3) easier to repeat the treatment and change the antigens targeted to overcome resistance to the treatment. A recent clinical study reported for the first time the use of universal CD19 CAR-T cells to treat four patients with refractory SLE. The results showed that the patients’ clinical symptoms and serological features were alleviated, SLEDAI scores decreased and no serious adverse events occurred (Ref. Reference Yang, Sun, Tan, Hu, Wan, Wang, Shi, Qin, Zhang, Zheng, Liu, Lin, du and Tong149). For other connective tissue diseases, such as patients with refractory SSc, after treatment with universal CD19 CAR-T cells, the patients’ skin hair follicles and sweat glands regenerated, and imaging examinations showed that fibrosis in the heart and lungs reversed, and they were in a state of sustained remission during the 6-month follow-up. No serious adverse events were observed during treatment and follow-up (Ref. Reference Wang, Wu and Tan150).

CAR-T therapy offers several advantages. First, anti-CD19 CAR-T cells eliminate a broader range of B cell lineages than rituximab, including pre-B, transitional, memory B cells and some long-lived plasma cells, deeply reducing autoantibody sources and enabling years of drug-free remission. Second, CAR-T cells persist as memory cells after a single infusion, providing sustained effects without chronic medication, improving compliance and quality of life while avoiding cumulative toxicity and financial burden. Third, CAR-T cells cross the blood–brain barrier and precisely target sites like joint synovium. Fourth, for refractory AIDs resistant to multiple drugs, CAR-T can induce deep remission even after rituximab failure by clearing CD19⁺CD20⁻B cells.

CAR-T can cause cytokine release syndrome (CRS) and immune effector cell-associated neurotoxicity syndrome (ICANS), requiring intensive care. B cell-targeting leads to prolonged B cell aplasia, hypogammaglobulinemia requiring IVIg, and increased infection risk (encapsulated bacteria, viral reactivation). CAR-T is expensive, requires GMP-grade vectors and specialised centres with multidisciplinary teams. Most studies are small case series; large RCTs are lacking. Optimal targets and need for lymphodepleting chemotherapy remain uncertain. Future directions include transient CAR expression via mRNA electroporation to avoid prolonged aplasia, CAR-Treg cells for precise suppression, and multicentre registries to clarify risk–benefit.

Prospects and challenges

In the field of AIDs treatment, cell therapy is gradually transitioning from traditional broad-spectrum immunosuppression towards precise immune reshaping. Currently, multiple cell therapy strategies coexist, each based on distinct mechanisms of action and at different stages of clinical translation. The following are the advantages and challenges of various cell therapies (Table 2). As an ‘off-the-shelf’ cell product, MSCs have accumulated extensive clinical research data, and their safety has been widely recognised. Their tissue repair capacity provides unique advantages in diseases accompanied by organ fibrosis or vascular lesions. Future directions include establishing standardised production systems, precise stratification of indications and optimisation of combination drug regimens. As a ‘cell-free’ alternative to MSCs, MSC-EVs avoid the tumorigenic risks associated with living cells, cold-chain transportation issues and immune rejection risks. Through engineering modifications (e.g., drug loading, targeting ligands), they can achieve a ‘smart’ upgrade as nanomedicines, holding broad prospects in areas such as acute inflammation, immune diseases and skin and soft tissue damage. Future efforts must address three bottlenecks: large-scale GMP-level production, improvement of in vivo stability and clarification of regulatory pathways. Through genetic modification, preconditioning or material integration, E-MSCs can significantly enhance homing, survival and immunoregulatory capacity. Particularly in scenarios requiring precise delivery of therapeutic factors (e.g., anti-inflammatory antibodies, enzymes), E-MSCs can serve as ‘living cell factories’. Future directions include non-viral vector development, safety switch design and large-scale production of universal products. HSCT is currently the only mature technology with immune-resetting function capable of achieving long-term drug-free remission. It has been recommended by domestic and international guidelines for refractory AIDs. Future directions include optimising preconditioning regimens to reduce toxicity, exploring non-myeloablative preconditioning to expand the eligible patient population, establishing patient risk stratification systems and evaluating the risk–benefit ratio of allogeneic transplantation in specific populations. Treg therapy most directly embodies the concept of restoring immune tolerance. It is highly physiological and targeted, and has entered early-stage clinical exploration for various diseases. Future efforts must address core issues such as efficient expansion and functional maintenance ex vivo, in vivo persistence and the development of universal allogeneic products. The application of CAR-T cell therapy in AIDs has been one of the most striking breakthroughs in recent years. Future directions include target optimisation, standardisation of toxicity management strategies, development of universal CAR-T products and clinical translation of CAR-Treg cells. If cost and safety concerns are alleviated, CAR-T therapy is expected to become a curative option for severe AIDs. First, all cell therapies face standardisation challenges regarding cell source, manufacturing processes and efficacy evaluation, which represent common bottlenecks for regulatory approval and industrial scaling. Second, cell homing, survival, functional maintenance and long-term safety (tumorigenicity, infection, autoimmune recurrence) all require support from long-term follow-up data. Third, except for MSCs, all cell therapies are high-cost treatments with insufficient insurance coverage and highly concentrated medical resources. Finally, identifying patient subgroups most suitable for a particular cell therapy is key to achieving precision medicine and improving the risk–benefit ratio. Overall, cell therapy is reshaping the treatment landscape for AIDs. In the future, with the establishment of standardised systems, maturation of engineering technologies, clarification of regulatory pathways and gradual cost reduction, the multiple cell therapies described above will form a complementary and coexisting treatment spectrum, providing ‘precise, accessible, and safe’ personalised treatment options for patients with different disease stages and pathological subtypes.

Table 2.

Advantages and challenges of various cell therapies for AIDs

Table 2. Long description

The table consists of four columns and six rows of data.

* Row 1: M S Cs. Core mechanism is Multitarget immunomodulation and tissue repair. Advantages include high safety, low immunogenicity, and strong tissue repair capacity. Challenges include high heterogeneity and unstable efficacy.

* Row 2: M S C E V s. Core mechanism is cell-free paracrine effects. Advantages include no tumorigenic risk, engineerable, and high penetrability. Challenges include difficult standardisation, poor in vivo stability, and an unclear regulatory path.

* Row 3: E M S Cs. Core mechanism is enhanced homing, survival, and function. Advantages include enhanced function, high controllability, and capability of carrying therapeutic factors. Challenges include safety concerns of gene editing, complex manufacturing, and strict ethical and regulatory oversight.

* Row 4: H S C. Core mechanism is immune resetting through myeloablation and reconstitution. Advantages include long-term drug-free remission, mature technology, and strong evidence of efficacy. Challenges include high treatment-related mortality, narrow indications, and high cost.

* Row 5: Treg. Core mechanism is restoration of immune tolerance and precise suppression. Advantages include physiological suppression, good safety profile, and high potential for engineering. Challenges include difficult ex vivo expansion, poor in vivo stability, and allograft rejection.

* Row 6: CAR T. Core mechanism is deep depletion of pathogenic immune cells. Advantages include single treatment, durable remission, and high targeting precision. Challenges include C R S / I C A N S toxicity, B cell aplasia, and extremely high cost.

Note: CAR: chimeric antigen receptor; E-MSCs: engineered mesenchymal stem cells; EVs: extracellular vesicles; HSC: haematopoietic stem cell; Treg: regulatory T cells.