The antigen recognition domain consists of an ScFv formed by linking the variable heavy chain (VH) and light chain (VL) of a monoclonal antibody through a linker. The primary function of the antigen recognition domain is the specific recognition of TAAs present on the surface of target cells, which subsequently facilitates the activation of the T cell bearing the CAR (2, 3).

Importantly, the recognition of target cells by CAR-T cell rely strongly on the high affinity of the ScFv component for the TAA (4–6). Therefore, the selection of appropriate ScFvs is crucial for antigen targeting. Initially, it is important to select an antibody with high affinity for the TAA and to determine its VH and VL sequences. Subsequently, CARs targeting the antigen can be constructed using ScFv sequences, after which in vitro cytotoxicity assays can be used to screen for CARs with cytotoxic functionality (5).

However, it is important to note that the affinity of the ScFv for the target antigen and its efficacy for CAR-T are not necessarily linearly correlated. In fact, excessive affinity can cause toxicity, as the CAR-T cells attack normal cells with low antigen density as well as the target tumor tissues, ultimately leading to damage to healthy tissue. Rather, researchers have found that the use of low-affinity ScFvs allows optimal CAR function (4–6). The ability to appropriately reduce the affinity of ScFvs to mitigate CAR-T cell therapy-induced tumor-associated toxicity has led researchers to speculate that there is an “affinity threshold” for CARs. This threshold represents the optimal balance where the ScFvs can effectively target tumor cells without attacking normal tissues with lower antigen density. Current research is actively exploring this concept by attempting to engineer ScFvs with affinities that fall within this ideal range. Scientists are using various screening and optimization techniques to fine-tune the affinity of ScFvs, ensuring they are strong enough to bind to tumor antigens while avoiding off-target effects on normal cells. This ongoing work aims to enhance the safety and efficacy of CAR-T cell therapies (5).

Almost all CAR constructs use ScFvs derived from monoclonal antibodies. However, the low folding stability of VH and VL could lead to aggregation or misfolding of ScFv, which may weaken the target effect and lead to exhaustion of CAR-T cells. To overcome these problems caused by conventional scFv, Xie et al. developed convenient and structurally stable CAR-T cell antigen-binding domains. This study designed a high-affinity binding protein (called “binder”) replacing typical ScFv to target tumor antigens. This binder CAR-T cells showed a stronger solid tumor treatment effect, providing a potential therapeutic benefit for CAR T cell therapy (7).

The hinge region connects the ScFv to the transmembrane structural domain. The hinge functions to enhance the flexibility of the ScFv, overcoming steric hindrance and facilitating the binding of the CAR to target antigens (8).

In the early, the hinge region is mainly derived from variants of immunoglobulin G (IgG), such as IgG1 and IgG4; therefore, one drawback to the use of these hinge regions is their potential interactions with the Fcγ receptor. Such interactions can lead to CAR-T cell depletion and ultimately adversely affect the durability of the therapy (4, 9). As an alternative, hinge regions from CD28 and CD8α have been adopted. In CD19 CAR-T cells, investigations have revealed that CD8α outperforms CD28 as a hinge region because CD8α stimulates a lower level of cytokine release and activation-induced cell death than CD28 (10). Moving forward, further optimization of hinge regions, including exploring other potential alternatives or modifications to CD8α, could enhance the safety and effectiveness of CAR-T therapies.

The TMD represents the intermediary section of the CAR, comprising a hydrophobic α-helix that spans the cellular membrane and connects the extracellular and intracellular domains. Its roles involve tethering the CAR to the cell membrane and transmitting activating signals to the intracellular domain (4, 11).

Historically, typical sources of TMDs have included CD3ζ, CD8α, CD28, and ICOS. However, the CD3ζ transmembrane structural domain is being abandoned because CARs containing this domain can form homodimers or heterodimers with the T-cell receptor (TCR), leading to strong T-cell activation in the absence of appropriate ligand binding. This uncontrolled activation can result in excessive immune responses, causing inflammation and potential damage to healthy tissues, making it a significant safety concern in CAR-T cell therapy. Conversely, CD8α and CD28 have become more frequently employed in clinical applications. Several investigations have revealed that the use of the CD8α TMD in CD19 CAR-T cells can attenuate the cell death caused by T-cell activation, and the CD8α TMD exhibits improved tolerability compared to that of CD28. In addition, the ICOS TMD has demonstrated superior efficacy and sustained antitumor activity compared to that of CD8 (10).

The TMD has attracted less attention in terms of structural innovations than other CAR domains. Nonetheless, researchers have shown that manipulating the length of the TMD can decrease CAR-T cell proliferation without compromising CAR-T cell antitumor capabilities. This characteristic offers more durable management of inflammatory cytokines and mitigation of the inflammation associated with the therapy (12). Furthermore, a recently introduced engineered TMD referred to as a programmable membrane protein (proMP) was reported to modulate the functionality of CAR receptors. As an entirely novel sequence, proMPs can form transmembrane homodimers, resulting in the generation of proCAR constructs. Compared to native CARs, ProCAR-T cells offer a more predictable functional range in vivo and significantly reduced release of inflammatory cytokines (13). Consequently, further research is warranted to explore the role and optimization of TMDs in enhancing this therapy.

The costimulatory domain, situated within the T-cell activation region, is pivotal for T-cell activation because it facilitates dual activation through costimulatory molecules and intracellular signaling. This dual activation is essential as it involves two key pathways: the first pathway provides the primary signal through the TCR for antigen recognition, while the second pathway, mediated by costimulatory molecules, enhances the T-cell’s response and promotes survival, proliferation, and cytokine production. Both pathways are necessary to ensure a robust and sustained immune response, preventing premature T-cell exhaustion and maximizing the effectiveness of CAR-T cell therapy (14, 15).

The most common costimulatory structural domains are CD28 and 4-1BB. CARs incorporating the CD28 costimulatory structural domain stimulate the differentiation of T cells into effector memory T cells and foster aerobic glycolysis. On the other hand, CARs containing the 4-1BB costimulatory structural domain drive the differentiation of T cells into central memory T cells, facilitating mitochondrial production, enhancing respiration, and promoting fatty acid oxidation (16, 17). Notably, CD28 exhibits faster and more robust signaling activity, while 4-1BB has a slower and more gradual profile that extends the lifespan of T cells and sustains their anticancer effects (16, 17). Furthermore, the CD28 family comprises additional costimulatory structural domains, such as ICOS, while the tumor necrosis factor receptor family includes OX40 and CD27. These additional costimulatory domains have been explored in various preclinical and clinical studies. For instance, ICOS has been tested and shown to enhance T-cell function and survival, suggesting its potential to improve CAR-T cell efficacy. Similarly, OX40 and CD27 have been investigated for their roles in enhancing T-cell proliferation and longevity. Incorporating these domains into CAR designs could further optimize the therapeutic outcomes of CAR-T cell therapy, though ongoing research is needed to fully understand their benefits and any associated risks (18).

Moreover, in recent years, the development of the pioneering in vivo gene delivery system approach, in which CAR genes are directly delivered and expressed within the patient’s body, has offered significant time and cost advantages over traditional manufacturing procedures for CAR-T cell production. Unlike conventional ex vivo methods that require T cells to be extracted, modified in the lab, and then reinfused into the patient, this in vivo system streamlines the process by modifying T cells directly within the body, reducing both production time and associated cost (19). For example, CD3-targeted lentiviral vectors (CD3-LVs) can achieve selective gene transfer into T cells specifically, thereby enabling in vivo gene delivery and T-cell expansion and allowing direct generation of CAR-T cells within the patient (20, 21). As a result, this approach has the potential to greatly simplify and accelerate the CAR-T therapy process, making it more accessible and scalable, and opening new possibilities for treating a broader range of diseases.

CARs of this generation consist of an ScFv, a hinge region, a transmembrane domain, and an intracellular signaling domain, such as CD3-ζ or FcγRI, that act as major transducers of TCR signaling and induce signaling cascades. The immunoreceptor tyrosine-based activation motif (ITAM) region of CD3ζ has been the most popular intracellular signaling domain due to the presence of three ITAMs in CD3ζ rather than one ITAM in FcϵRI (23, 24), although CAR-T cells containing the 4-1BB signaling domain demonstrate significantly better antitumor activity and persistence than those with CD3ζ (25). Leveraging the CD3ζ signaling domain, CAR-T cells of this generation trigger T-cell activation, resulting in cytotoxicity and cytokine secretion (e.g., IL-2) independent of human leukocyte antigen (HLA). Consequently, tumor cells are effectively eliminated by these CAR-T cells (26).

Nevertheless, CAR-T cells of this generation have several limitations. First, they do not produce sufficient IL-2 to support their long-term proliferation and sustain their cytotoxic activity against cancer cells, so exogenous IL-2 must be used to achieve effective tumor cell eradication (9). Second, the intracellular domains of CAR-T cells are deficient in costimulatory signals that are necessary for robust T-cell activation, limiting the efficacy of T-cell activation. Third, although these CAR-T cells exhibit conventional T-cell cytotoxicity, their proliferation is transient, and they secrete limited amounts of cytokines, ultimately leading to T-cell apoptosis and an inability to proliferate and survive long-term in vivo, greatly limiting their antitumor efficacy (27, 28).

Due to these limitations, CAR-T cells of this generation failed to achieve the expected efficacy in clinical applications. This failure propelled the development of second-generation CAR-T cell therapies aiming to enhance overall treatment efficacy by extending the in vivo persistence of CAR-T cells and augmenting their cytotoxic capabilities.

The major advance of second-generation CAR-T cells over the first-generation therapies is the incorporation of a costimulatory domain, predominantly CD28 or 4-1BB, linked with CD3-ζ (14, 29). This modification allows T cells to receive both antibody stimulation signals and costimulatory signals (14), promoting IL-2 synthesis and facilitating T-cell activation. Second-generation CAR-T cells have been further optimized to address issues concerning T-cell proliferation, survival, cytotoxicity, and persistence (29–32).

Second-generation CAR-T cells featuring CD3ζ and costimulatory structural domains have been extensively utilized in clinical practice, with CAR-T cell products containing 4-1BB as a costimulatory molecule gradually becoming mainstream. Currently, all approved CAR-T cell therapies on the market are second-generation CAR-T products targeting hematologic malignancies. The Food and Drug Administration (FDA) has approved six CAR-T cell products, with two targeting BCMA (B-cell Maturation Antigen) (see Glossary) and four targeting CD19 ( Table 1 ). For instance, the FDA first approved Tisagenlecleucel (Kymriah) in 2017, a CD19-targeting CAR-T cell therapy widely used for treating B-cell acute lymphoblastic leukemia and B-cell lymphoma, particularly in patients under 25 years old. Kymriah is currently approved for three indications and remains the only CAR-T cell therapy approved for both adult and pediatric populations (33). Similarly, Axicabtagene ciloleucel (Yescarta) received its first FDA approval in 2017 for the treatment of relapsed or refractory diffuse large B-cell lymphoma (DLBCL) in adult patients and is now approved for three indications (34). In March 2021, Abecma became the first FDA-approved BCMA-targeting CAR-T cell therapy for adults with relapsed or refractory multiple myeloma after at least four prior lines of therapy, marking a significant advancement in multiple myeloma treatment (34).

Several studies have demonstrated that the costimulatory molecule CD28 imparts potent activation signals, allowing T cells to rapidly attain high levels of cytotoxic activity while also contributing significantly to the generation of memory and effector cells, albeit with a shorter duration. On the other hand, the activation signals mediated by 4-1BB are more persistent, playing a vital role in regulating lymphocyte proliferation and survival and enhancing IL-2 production but with limited cytotoxic capabilities (41–43).

Unfortunately, studies have shown that a single costimulatory domain fails to resolve the issues of CAR-T cell therapy persistence and relapse. The use of retroviral vectors for gene transfer into T cells, a prevailing method for generating second-generation CAR-T cells, constrains the size of the gene fragments they can accommodate, and consequently, it is not feasible to transfect the ITAM regions of both CD28 and CD137 costimulatory molecules into T lymphocytes simultaneously. As a result, researchers have been forced to choose between achieving persistence and preventing relapse during therapy with the use of second-generation CAR-T cells.

The development of third-generation CAR-T cells has focused on using specific transduction techniques to endow them with two or more costimulatory domains linked with CD3ζ. These approaches typically involve advanced gene delivery methods such as lentiviral vectors or transposon systems, which allow for the precise insertion of multiple costimulatory signals into the CAR-T cells. These techniques enable the enhancement of T-cell activation and persistence by integrating additional signaling domains (10, 44). CAR-T cells of the third generation are embodied by the CAR-T cell construct, which arises from the fusion of CD3ζ and multiple costimulatory domains. The viral vectors used for third-generation CAR-T cells are mostly lentiviral, which allows larger gene fragments to be carried into T lymphocytes, thus allowing CAR-T cells of this generation to contain more costimulatory structural domains, such as CD28, 4-1BB, and CD134 (OX40). At present, the most widely utilized third-generation CAR-T cell construct is the CD3ζ-CD28-4-1BB construct (44–49). The costimulatory molecules in these CAR-T cells can activate signaling pathways within T cells, including the JNK, ERK, and NF-κB pathways, thereby enhancing T-cell antitumor activity, proliferation capacity, and survival time, as well as the secretion of cytokines such as IL-2, TNF-α, and IFN-γ (27).

Although clinical data indicate that these CAR-T cells show superior anticancer potency relative to that of second-generation CAR-T cells, as demonstrated by their prolonged duration and increased expansion in B-cell non-Hodgkin lymphoma patients (50, 51), the safety profile of these CAR-T cells is statistically equivalent to or even worse than that of second-generation CAR-T cells. Hence, the simple addition of costimulatory molecules within the ITAM domain does not necessarily guarantee enhanced therapeutic efficacy (52, 53).

Known as precision CAR-T cell therapy, fourth-generation CAR-T cell therapy is an advanced approach that offers improved targeting and a reduced risk of tumor recurrence. Building upon the foundation of second-generation CAR-T cells, precision CAR-T cells include regulated suicide genes within CAR-T cells to control their longevity within the organism. This innovative modification aims to circumvent the adverse effects associated with cytokine storms while still precisely targeting tumor cells within the host system. These CAR-T cells are known as TRUCK or armored CAR-T cells (27, 54–56).

These CAR-T cells also harbor an element that activates the transcription of the nuclear factor of activated T cells (NFAT) gene, enabling the secretion of specific cytokines (primarily IL-12) within the TME to recruit and activate other immune cells for an immune response (57). Additionally, the secretion of these cytokines promotes T lymphocyte infiltration within tumor tissues, thereby offering therapeutic advantages for solid tumors (as opposed to hematologic malignancies) (58).

Earlier generations of CAR-T cell therapies, while effective, face challenges such as immune rejection by the host and the risk of graft-versus-host disease (GVHD) when infused into patients. To address these issues, a promising strategy has emerged that involves precise genetic modifications of HLA and TCR genes in T cells obtained from healthy donors. This fifth-generation approach aims to enhance CAR-T cells by reducing the likelihood of immune rejection and preventing GVHD, making the therapy safer and more effective for a broader range of patients. Notably, the inherent advantage of this approach is that it eliminates the need for patient-specific alterations, thereby offering a promising avenue for the treatment of diverse patients with the same therapy. This innovation can benefit a broader range of patients, including those with different types of cancers or diseases and individuals who were previously excluded from CAR-T cell therapy due to compatibility issues or concerns about GVHD (59). Universal chimeric antigen receptor T cells (UCAR-T cells) are prototypical examples of fifth-generation CAR-T cells. Unlike conventional CAR-T cell therapy, which involves extracting, genetically modifying, amplifying, and transferring T cells from patients (45, 60–62), UCAR-T cell therapy relies on T cells acquired from healthy allogeneic donors. Subsequently, endogenous TCRs and HLA class I molecules are knocked out of these cells through gene editing techniques, such as ZFN, TALEN, and CRISPR-Cas9, enabling large-scale production and therapy without the constraints of individual compatibility (63).

Compared with autologous CAR-T cells, allogeneic CAR-T cells possess distinct advantages, including broad individual compatibility and potentially scalable manufacturing and treatment. However, despite their promise, current clinical trials of UCAR-T cells face certain challenges that may limit their clinical application. For instance, their CAR recognition capacity is limited, with affinity restricted to only one or two targets. Nonetheless, allogeneic CAR-T cell therapy has great potential for addressing issues related to the lengthy production cycle and high costs associated with CAR-T cell therapy, underscoring the need for extensive exploration of this approach (64–67).

The design, preparation, and application of CAR-T cells exhibit considerable heterogeneity, as their effectiveness varies depending on the tumor type. Notably, first-generation CAR-T cells have exhibited substantial therapeutic efficacy in specific hematologic malignancies but have shown limited effectiveness in targeting solid tumors. Second- and third-generation CAR-T cells have significantly enhanced the viability, effectiveness, and safety profile of CAR-T cells, thereby improving therapeutic outcomes in clinical trials. Although third-, fourth-, and fifth-generation CAR-T cell technologies are still in the developmental stage and the cost of first- and second-generation CAR-T cell treatments remains high, it is reasonably presumed that these technologies will progressively enhance the stability and effectiveness of CAR-T cells, reduce their cost, and mitigate their adverse effects (68).

Antigen escape describes the situation in which the expression of target antigens by malignant tumor cells in the body is partially or completely inhibited. This event plays a pivotal role in the development of acquired resistance against CAR-T cells in solid tumors (79, 80).

To overcome this limitation and minimize damage to healthy tissues, researchers have explored the utilization of dual-CAR or tandem-CAR systems that target multiple antigens simultaneously (4). Several studies have reported encouraging results from the application of dual CAR-T cells that target both CD19 and CD22 to treat patients with acute lymphoblastic leukemia (ALL), reporting that this therapy is more effective than CD19 CAR-T cell therapy alone (81). Additionally, in this situation, CAR-T cells are activated upon recognizing multiple antigens, thereby reducing the likelihood of accidental activation in healthy tissues and mitigating adverse effects (82).

Neoplasm-targeting allele-sensing CAR (NASCAR), is designed to target a mutation commonly found in the cancer transformation process called clonal loss of heterozygosity (LOH). LOH is observed in 90% of human cancers and often involves tumor suppressor genes, providing a clear distinction between normal and cancer cells. Previous therapeutic approaches targeting this distinction, such as antisense oligonucleotides and siRNA, have generally been ineffective. However, programming CAR-T cells to recognize LOH can significantly enhance their precision. This is achieved by introducing both CAR and an inhibitory CAR (iCAR) into CAR-T cells, with iCAR carrying intracellular domains like PD-1 or CTLA-4. In cancer cell lines engineered with LOH using CRISPR technology, NASCAR-T cells have demonstrated strong antitumor efficacy, highlighting their potential for precise and effective tumor targeting (83).

Above, we introduced several representative strategies focusing primarily on enhancing the targeting capabilities of CAR-T cells. However, researchers are also making significant efforts to improve other aspects of CAR-T cell functionality, such as the controllability of their activity and cytotoxicity. These enhancements include the addition of drug-responsive interfaces to T cells, allowing for the precise modulation of CAR-T cell activity through external drug administration. Looking ahead, more advanced versions of CAR-T cells will likely emerge, taking into greater account the specific needs and conditions of patients (78).

Prolonged proliferation of CAR-T cells is a key requirement for achieving an effective antitumor response within patients (84, 85), and T-cell exhaustion therefore poses a significant hurdle in CAR-T cell therapy (86, 87). Continuous stimulation of T cells by antigens within the tumor environment renders them susceptible to exhaustion, leading to inadequate production of IL-2 (88, 89).

Recent studies have provided compelling evidence indicating that CAR-T cell expansion and persistence can be augmented through the judicious selection of costimulatory signaling molecules, such as CD28, ICOS, OX40, and 4-1BB (29, 90–92). CAR-T cells incorporating CD28 displayed greater persistence than those containing 4-1BB, thus mitigating T-cell exhaustion and accelerating the expression of effector molecules after antigenic stimulation (37, 93).

Additionally, genetic engineering methods have emerged as promising strategies to address T-cell exhaustion. For example, enhancing the expression of the cytokine IL-7 or IL-15 has been shown to promote T-cell proliferation and, when combined with the downregulation of PD-1 expression, prevent CAR-T cell exhaustion (94, 95). Furthermore, gene manipulation to induce the overexpression of mutated Fas variants within T cells has been used to effectively suppress FasL-mediated cellular apoptosis (96). Modulating the expression of inhibitory receptors, such as PD-1 and TIM-3, in CAR-T cells is another approach that has been successful in counteracting CAR-T cell exhaustion (97, 98).

Further advancements in understanding and manipulating these and other exhaustion pathways could lead to more robust and enduring CAR-T cell therapies, ultimately improving patient outcomes and expanding the applicability of this treatment across various cancers.

A significant challenge confronting solid tumor CAR-T cell therapy is the widespread presence of TAA in normal tissues, which makes it more difficult to target the desired effects specifically to tumor cells. For instance, trastuzumab-based (anti-HER2) CAR-T cell therapy exhibited fatal cardiopulmonary toxicity in a patient diagnosed with metastatic colon cancer (47). Current approaches to address this issue include targeting posttranslational modifications that are specific to tumor cells (99, 100). Examples of these modifications include phosphorylation patterns unique to oncogenic signaling pathways, glycosylation changes that alter cell surface antigens, and acetylation modifications affecting tumor suppressor proteins. By focusing on these tumor-specific modifications, therapies can more precisely target cancer cells while sparing normal tissues. For instance, certain truncated O-glycans, such as Tn (GalNAcα1-O-Ser/Thr) and sialyl-Tn (STn) (NeuAcα2-6-GalNAcα1-O-Ser/Thr), are overexpressed in solid tumors, providing opportunities for targeted therapy. Therapies targeting these specific glycan modifications are currently in various stages of development. Some have progressed to early clinical trials, where they have demonstrated initial promise in selectively targeting cancer cells while minimizing damage to normal tissues. However, many of these approaches are still in the preclinical or experimental phase, where they are being explored for their potential to enhance specificity and efficacy in cancer treatment (99).

Another approach taken by some researchers has been to endow T cells with two engineered receptors: a conventional CAR and a chimeric costimulatory receptor (CCR). This method is related to the evolution seen in the different generations of CAR-T cell therapy. Specifically, while second and third-generation CAR-T cells incorporate single and multiple costimulatory domains directly into the CAR structure to enhance T-cell activation and persistence, the strategy of using separate CCRs represents a novel enhancement beyond these generations. This dual-receptor strategy differs from the dual activation mentioned earlier in that the CCR is designed to provide an additional, distinct costimulatory signal only upon encountering a specific antigen in the tumor microenvironment. This ensures that T cells are fully activated only when they bind both the tumor-associated antigen via the CAR and the second antigen via the CCR, thereby increasing precision and reducing the likelihood of off-target effects (101). For example, these dual-receptor T cells continuously expand and persist for longer durations in multiple myeloma and low-antigen leukemia. This phenomenon concurrently prolongs patient survival and delays cancer progression (102).

Masked CAR-T cell (mCAR-T) therapy is another strategy to reduce the off-tumor effects. It has a CAR receptor with an additional masking peptide and a protease-sensitive linker. The masking peptide covers the antigen recognition domain and is removed through protease cleavage, which is triggered by the high protease levels typically found in the tumor microenvironment. This approach enhances targeting precision, and the antitumor efficacy of mCAR-T cells is comparable to that of conventional CAR-T cells (103).

Despite the advancements in CAR-T cell therapies, targeting solid tumors remains particularly challenging. CAR-T cell delivery to solid tumors encounters significant challenges related to the tumor microenvironment (TME), tumor stroma, and disparities between the chemokine receptors expressed by CAR-T cells and the chemokines released by tumor tissue. These factors collectively impede the effective infiltration and activity of CAR-T cells within solid tumors (104).

Therefore, specific strategies are needed to enhance the delivery of CAR-T cells to solid tumors. Some approaches that have been used are localized administration to facilitate CAR-T localization at the tumor site and engineering CAR-T cells to express chemokine receptors that are compatible with tumor-derived chemokine factors (105). For example, a recent study found that compared to intravenous infusion, directly injecting CAR-T cells into the pleural cavity significantly improves therapeutic outcomes and substantially reduces the required T cell dosage. Based on these promising results, the researchers have initiated a clinical trial to evaluate the safety of locally administered mesothelin-targeted CAR-T cell therapy in patients with malignant pleural tumors (106). Moreover, studies focusing on chemokine factors have revealed that CAR-T cells targeting the tumor-associated αvβ6 integrin and expressing CXCR1/2 are effectively enriched at the tumor site (107–109). Additionally, for neuroblastoma, which exhibits high CCL2 expression, the use of CAR-T cells targeting GD2 and CCR2b has yielded promising outcomes in terms of transportation (110).

Another obstacle to effective CAR-T cell transportation stems from the vasculature in the TME (111). Tumors often exploit abnormal blood vessel formation to acquire nutrients and promote infiltration, and these abnormalities can hinder therapeutic efficacy. Consequently, specifically targeting abnormal vessels represent a promising approach for enhancing CAR-T cell penetration (112, 113). Considering the pivotal role of vascular endothelial growth factors and their receptors in angiogenic signaling pathways, studies have suggested that inhibiting these factors and receptors could be employed as an antiangiogenic therapy for treating such aberrant vessels (113). Furthermore, considering that heparan sulfate proteoglycans are major constituents of the tumor stroma, the expression of heparinase, an enzyme capable of degrading heparan sulfate proteoglycans, in CAR-T cells has been suggested as a potential strategy (114).

In summary, overcoming the challenges of CAR-T cell trafficking to solid tumors is crucial for enhancing their therapeutic effectiveness. Strategies such as modifying chemokine receptors, degrading tumor stroma components, and employing innovative delivery methods are key to improving CAR-T cell infiltration and persistence within the tumor microenvironment.

Other factors limiting CAR-T cell efficacy include the presence of an immunosuppressive TME with a cytokine profile (such as IL-10, TGF-β, VEGF and IL-4) that preferentially recruits immunosuppressive Treg cells, myeloid-derived suppressor cells (MDSCs) and tumor-associated macrophages(TAMs) ( Figure 4 ) (115). Treg cells suppress cytotoxic T cell activity through various mechanisms, including the secretion of immunosuppressive cytokines, competitive consumption of IL-2, CTLA4-mediated inhibition of antigen-presenting cells (APCs), and the prevention of T cell activation (116). Similarly, MDSCs exert a strong immunosuppressive effect on effector T cells, significantly impairing CAR-T cell functionality. Clinical observations have linked low MDSC levels in patients receiving CD19 CAR-T therapy with favorable responses in lymphoma and leukemia treatment. TAMs, as the predominant immune-infiltrating cells in the TME, suppress T cell-mediated anti-tumor responses by secreting immunosuppressive cytokines, producing amino acid-depleting enzymes such as arginase 1 and indoleamine 2,3-dioxygenase (IDO), and promoting the recruitment of Treg cells. Moreover, immune checkpoints, particularly PD-1, contribute to the reduced efficacy of immunotherapy (117). Other elements of the TME, including physical barriers, hypoxia, and immunosuppressive factors, exert similar effects (118, 119). Targeting these immunosuppressive cells or molecular signals has the potential to enhance the therapeutic efficacy of CAR-T cells (78).

Numerous strategies have been proposed to address these challenges. First, engineered synthetic receptors have been developed to counteract suppressive cytokines, with notable targets including TGF-β and IL-4. One such innovation is the synthetic TGF-β dominant-negative receptor (DNR), a modified version of TGF-β receptor II (TGF-βRII) that lacks the cytoplasmic signaling domain. These TGF-β DNRs inhibit native TGF-β signaling by forming non-functional ligand-receptor complexes, effectively serving as a molecular sink to sequester TGF-β (120). The incorporation of immune checkpoint blockade agents into CAR-T cells also promotes their long-term persistence, leading to promising outcomes (121). Furthermore, gene-editing technologies such as CRISPR/Cas9 can be utilized to disrupt PD-1 expression, further enhancing CAR-T cell persistence (98). Designing CAR-T cells to resist immune inhibitory factors can also result in prolonged persistence, contributing to their sustained presence and efficacy in the TME (122).

A notable obstacle to the widespread adoption of CAR-T cell therapy is the toxicity associated with these approaches including cytokine release syndrome (CRS) and immune effector cell-associated neurotoxicity syndrome (ICANS) (36). CAR-T cell activation and cytokine secretion depend on the binding of the antigen-binding domain to its corresponding target antigen and reaching a specific activation threshold. This threshold refers to the amount of antigen engagement needed to sufficiently activate the CAR-T cells. Typically, it involves a certain number of antigen receptors being bound on the surface of CAR-T cells. When this threshold is met, it triggers a robust immune response. However, if the number of bound receptors or the intensity of activation signals exceeds this threshold, it can lead to excessive cytokine release and potentially cause toxicity, such as CRS or off-target effects on healthy tissues. Consequently, there has been a comprehensive exploration of modifying the CAR construct as a potential strategy to mitigate such toxicity (123).

One effective approach involves reducing the affinity of the antigen recognition domain, which consequently requires higher levels of target antigens on tumor cells to trigger T-cell activation. This alteration allows for the relative sparing of healthy cells that express lower levels of the targeted antigens (124). Another strategy to mitigate toxicity involves modifying the hinge and transmembrane regions of CAR constructs. These modifications are designed to optimize the flexibility and stability of the CAR structure, which can influence the strength and duration of the activation signal. By fine-tuning these regions, the CAR-T cells can achieve a more controlled and balanced activation response. This helps in regulating the amount of cytokine secretion, preventing excessive immune responses and reducing the risk of toxicity. Notably, modifying the CD8-α hinge and transmembrane region in anti-CD19 CAR-T cell therapy significantly reduced cytokine release (12).

Therefore, effectively managing CAR-T cell toxicity is essential to ensure the safety and success of these therapies. Current strategies focus on refining CAR designs to better regulate activation and cytokine release, thereby reducing adverse effects.

Several efforts to address the limitations arising from the adverse effects of CAR-T cell therapy have led to the development of CAR-T cells with specialized functionalities (125). For example, “self-driving” CARs are characterized by the expression of chemokine receptors that can detect and respond to the production of chemokines by tumor cells targeted for destruction. This mechanism enables the efficient trafficking of CAR-T cells to tumor sites, thereby augmenting their antitumor efficacy (45). For example, CAR-T cells carrying the tumor antigen GD2 enhance the migration of CAR-T cells to CCL2 by expressing the chemokine receptor CCR2b (126).

In the dual-CAR strategy, described above, genetic engineering techniques enable T cells to express two distinct CARs. This allows the targeting of multiple antigens simultaneously, potentially reducing patient resistance to CAR-T cell therapy and enhancing its overall effectiveness (4). Recent experimental evidence suggests that patients with ALL treated with CAR-T cells specifically targeting CD19 may experience loss or downregulation of the CD19 antigen (127, 128). However, the utilization of dual CAR-T cells that target both CD19 and CD22 has proven effective in addressing this challenge (81).

Inhibitory CARs (I-CARs) are CARs that can suppress T-cell responses. These CAR-T cells express two different CARs, one conventional and the other carrying an I-CAR. The I-CAR acts as a self-regulating switch that inhibits T-cell activation when it specifically recognizes and binds to antigens that are present exclusively or predominantly in healthy tissues. This design ensures that CAR-T cells function only in the absence of I-CAR stimulation, preventing damage to healthy tissues. Examples include I-CARs based on CTLA-4 or PD-1, which selectively limit the cytotoxicity induced by endogenous T-cell receptors or the activation of chimeric receptors (129).

Tan-CARs are characterized by the presence of two ScFv structural domains, each capable of recognizing a different target. This configuration increases the likelihood of CAR-T cell activation. However, it also carries the risk of causing damage to healthy tissues (125).

The concept of “off-switch” CARs emerged in response to the severe adverse effects associated with CAR-T cells, such as CRS. These CAR-T cells are designed such that specific inducers can be used to halt their excessive toxicity and induce depletion, thereby reducing or preventing adverse effects (130). For instance, CD20 CAR-T cells can be depleted through the administration of rituximab (131). However, one limitation of this antibody-mediated “off-switch” approach is that it acts on a relatively long timescale; that is, the rate of depletion of the undesired cells is slow (132). Moreover, alternative methods have been developed, including the use of dasatinib, inducible Cas9, and protease-based small molecule-assisted shutdown CARs (SMASh-CARs) (133, 134). In addition to “off-switch CARs”, there are also “on-switch CARs” that remain inactive under normal circumstances and are activated only upon stimulation by exogenous molecules, thus enabling the regulation of the timing and location of CAR-T cell activity (135).

In conclusion, continued research and optimization of these specialized CAR-T cells will be crucial to maximize their therapeutic potential and extend their application to a broader range of cancers and patient populations.

While CAR-T cells have demonstrated promising efficacy, achieving complete eradication of solid tumors using CAR-T cell therapy alone is difficult (136). However, some researchers have found that CAR-T cell therapy can significantly enhance the antitumor effects of other treatment modalities. Combination therapies that have been tested in conjunction with CAR-T cells to date include synergistic approaches that activate the patient’s endogenous T cells (137), as well as chemotherapy, radiation therapy, nanoparticles (NPs), oncolytic viruses (OVs), immune checkpoint blockade (ICB) and mRNA vaccine techniques (138).

The PD-L1/PD-1 signaling pathway is a tumor hallmark, with the TME inducing the expression of PD-L1 on tumor cells, and blocking the PD-1 receptor can deactivate CAR-T cells. The binding of PD-L1 on tumor cells to PD-1 on CAR-T cells inhibits the antitumor effects of CAR-T cells (139, 140). Hence, integrating ICB into CAR-T cells by blocking the PD-1/PD-L1 signaling pathway, such as by combining CAR-T cell therapy with anti-PD-1/PD-L1 agents (121, 141, 142), has shown potential in prolonging the persistence of CAR-T cell therapy and increasing the ratio of CD8⁺/CD4⁺ T cells within the TME (121, 143–145). Clinically, combination therapy targeting both PD-1 and CD19 has yielded favorable treatment outcomes and persistence in patients with hematologic malignancies (146).

Moreover, blocking the PD-1 signaling pathway in combination with CAR-T cell therapy has significantly improved the efficacy of CAR-T treatment of solid tumors (78). Recent studies have shown that combining CAR-T cell therapy with pembrolizumab (a PD-1 inhibitor) substantially increases the longevity of chimeric antigen receptor (CAR) cells in a murine model of metastatic melanoma (147, 148). Additionally, another study revealed that CAR-T cells secreting an anti-PD-1 scFv significantly enhanced CAR-T cell persistence and prevented the toxicity associated with systemic checkpoint inhibition (149, 150). Furthermore, employing CRISPR-Cas9 to knock out the PD-1 gene in CAR-T cells enhances CAR-T cell function, prolongs persistence, and reduces cytotoxicity, although it falls short in inducing T-cell infiltration (36, 151, 152).

Other gene editing techniques can also be applied to the modification and manufacturing of CAR-T cells. Studies have shown that CRISPR/Cas9 knockout of the TCR in tumor cells enhances T-cell potency (153) and reduces the risk of GVHD. Additionally, CRISPR/Cas9 can be used in conjunction with viral vectors to achieve targeted gene integration at specific sites (154). This precise integration can enhance the efficiency and safety of CAR-T cell therapies by ensuring that the CAR gene is inserted into a known and controlled location within the genome. However, some studies have indicated that this method may inadvertently shorten the lifespan of the CAR-T cell response. This unexpected outcome could be related to the disruption of genomic stability or the altered expression of nearby genes caused by the integration process, rather than the specific gene that was knocked in. As a result, while CRISPR/Cas9 offers a powerful tool for precise gene editing, it is essential to further investigate and mitigate these potential impacts to fully harness its benefits in CAR-T cell therapy (155). In terms of manufacturing, CRISPR/Cas9-based methods are characterized by low toxicity and high efficiency.

OVs are particularly interesting as a combination therapy because themselves exhibit antitumor activity (156, 157), and moreover, combining CAR-T cell therapy with OVs provides a more effective targeting of solid tumors, leading to enhanced antitumor effects (158). The OV genome can also be modified to express tumor necrosis factor-alpha or IL-2, which has been shown to increase the quantity of CAR-T cells in the TME and thereby enhance antitumor efficacy (159, 160).

Various nanotechnologies promise to enhance cancer immunotherapy by facilitating the delivery of immune-modulating drugs. In particular, NPs can serve as carriers for CAR-T cells into the TME, preventing their inhibition by TMICs and other immunosuppressive factors (115, 161–163).

The remarkable success of the SARS-CoV-2 mRNA vaccine has spurred extensive efforts to broaden the mRNA vaccine platform for the treatment of other conditions. Recent research has demonstrated that modified mRNAs encoding a CAR designed against fibroblast activation protein (FAP) (a marker of activated fibroblasts) encapsulated in targeted LNPs can be intravenously injected to induce the generation of functionally engineered T cells in vivo, offering potential for treating cardiac injury (164). These mRNAs typically encode for CARs or TCRs, which can reprogram T cells to target specific antigens associated with disease, thereby enhancing their therapeutic potential (165–167). Furthermore, numerous experiments have shown the potential of this method for treating solid tumors. Recent findings from a pioneering phase 1/2 basket trial investigated a claudin-6 (CLDN6)-directed CAR-T cell therapy for patients with CLDN6 tumors with the use of CARVac, an innovative antigen delivery mechanism that leverages an mRNA vaccine to bolster CAR-T cell amplification and persistence (168). This strategy is attractive because it enables CAR-T cell therapy to be performed directly in vivo, that is, without removing T cells from the body for editing, which saves considerable time and money. Additionally, the ability to promptly and precisely adjust the mRNA sequence and dosage to suit various tumor types and variants offers clear practical advantages. For example, it avoids prolonged CAR-T cell activity and the resulting amplification effects, thus reducing the risk of severe side effects such as CRS and neurotoxicity. Maintaining CAR-T cell efficacy and durability through multiple sequential injections addresses challenges such as immunosuppression and drug resistance within the TME. Moreover, synergistic enhancement of CAR-T cell antitumor efficacy can be achieved by combining or coadministering various mRNA types or drugs. Overall, the integration of mRNA technology with CAR-T cell approaches has potential for ‘off-the-shelf’ universal therapies, offering a scalable and affordable solution in oncology and for other diseases, such as autoimmune disorders, cardiovascular diseases, infectious diseases, and certain genetic conditions

Recent advances have expanded the scope of immunotherapy beyond CAR-T cells alone. Innovations in cellular immunotherapy have propelled the expansion of the core CAR strategy to different cell types, such as Tregs, NK cells, and macrophages (M cells), to enhance their efficacy and broaden the therapeutic landscape. The resulting therapies hold great potential for treating cancer and immune-related disorders. These innovative approaches are described in detail below.

Following the success of CAR-T cell therapy for cancer, researchers have explored its potential in treating autoimmune diseases. Excessive B-cell activation leads to autoantibody production, driving autoimmune disorders. CAR-T cell targeting of autoreactive B cells aims to reduce autoantibodies and suppress autoimmunity. This innovative approach offers treatment promise across autoimmune disorders, enhancing patient outcomes and quality of life.

Originally utilized for GVHD treatment (246), Treg-based therapy aims to reinstate immune tolerance in the context of autoimmune diseases (188). Since its original application, the amalgamation of Treg cells with CAR technology has been considered to have considerable potential in addressing diverse autoimmune conditions, facilitating the extensive utilization of CAR-Tregs (247, 248). Clinical trials commonly employ manipulation techniques involving CD4⁺ FOXP3⁺ Treg cells (249).

Another recent area of exploration within the field of CAR-T cell and CAR-Treg therapy for autoimmune disease treatment involves modifying CAR-T cells to proficiently eliminate self-reactive B lymphocytes. Of particular interest is a novel engineered T-cell construct, termed the chimeric autoantibody receptor (CAAR). In contrast to CAR-T cells, CAAR-T cells possess an extracellular domain derived from autoantigen. This unique feature enables CAAR-T cells to identify and eradicate self-reactive B-cell receptor (BCR)-expressing B lymphocytes, granting them the ability to specifically target this population (229). Although the investigation of CAAR-T cells is still in its initial phases, it represents an enticing strategy for treating autoimmune diseases.

Pemphigus vulgaris (PV) is a severe blistering disease affecting the skin and mucous membranes caused by autoantibodies targeting desmoglein proteins (277). Therefore, the presence of anti-desmoglein BCR on memory B cells contributes to PV pathogenesis (278, 279). Recent studies have demonstrated that cell-based therapies targeting anti-desmoglein 3 (Dsg3) antibodies can be employed for treating PV. Compared with CD19 CAR-T cells, chimeric CAAR-T cells with Dsg3 extracellular domains have shown superior and more durable therapeutic effects. However, they are still associated with several adverse effects, such as CRS (280). Despite being in its nascent phase, CAAR-T-cell technology has exhibited promise in preclinical investigations and initial clinical trials. Further research endeavors and clinical studies will facilitate a comprehensive assessment and refinement of the therapeutic efficacy, safety, and long-term sustainability of this technology.