Enhancing persistence while managing cytokine release syndrome to embrace next-generation CAR-T cell therapy.

Introduction

Cancer continues to pose a major global health burden, with rising incidence and mortality rates. Advances in understanding antitumor immunity have established immunotherapy as a critical treatment modality. Among these approaches, CAR-T cell therapy has emerged as a transformative strategy, like a modern Perseus equipped to defeat Medusa of hematological malignancies.

To date, several CAR-T cell products targeting antigens such as CD19 and B cell maturation antigen (BCMA) have been approved by the US Food and Drug Administration (FDA) [ 1 , 2 ]. Additionally, several CAR-T products and targets are under evaluation in clinical trials, including Obecabtagene autoleucel [ 2 – 5 ], Relmacabtagene autoleucel (targeting CD19) [ 2 , 4 – 6 ], and therapies directed against CD22 [ 2 , 4 , 5 , 7 ] and CD30 [ 2 , 4 , 5 , 8 ]. The scope of CAR-T cell therapy is progressively expanding beyond hematological cancers into clinical investigations for solid tumors [ 2 , 4 , 5 , 9 ], infectious diseases [ 2 , 4 , 5 , 10 ], autoimmune disorders [ 2 , 4 , 5 , 11 , 12 ]. Additionally, this therapy has shown exploratory promise in animal models of asthma [ 2 , 4 , 5 , 13 , 14 ].

Despite these advantages, the CAR-T therapy is constrained by the treatment-related toxicity and insufficient persistence. Toxicities such as CRS, immune effector cell-associated neurotoxicity syndrome (ICANS) and other noncanonical toxicities narrow the therapeutic window [ 15 ]. Moreover, premature exhaustion caused by tonic signaling and functional decline of CAR-T cells in vivo compromises their persistence, directly leading to disease relapse [ 16 ]. In this review, we propose that insufficient persistence and treatment-related toxicity represent two manifestations of imbalanced intracellular signaling networks. We consider that coordinated modulation of this integrated network is essential for developing the next-generation CAR-T cell therapies. We systematically dissect CAR-T cell signaling pathways, covering receptor design, signal transduction, functional output, and interaction with the tumor microenvironment (TME). We examine how the dynamic interplay of signaling pathways affects CAR-T cell function. We also discuss how tuning these signals can steer CAR-T cells toward a more desirable state in the future. The ideal state is characterized by the combination of long-term persistence and manageable toxicity. Ultimately, we hope to paint a blueprint to inform the rational design of safer and more persistent CAR-T cell therapies.

Signaling pathway initiation, transmission, and functional output

Basic CAR structure: the foundation of molecular design

The CAR is a synthetically constructed fusion protein engineered through genetic recombination techniques. Structurally, it is composed of four functional domains. These domains include an extracellular antigen-binding domain, a hinge region, a transmembrane domain, and an intracellular signaling domain. The extracellular domain is responsible for recognizing and binding tumor-associated antigens on the target cell surface. The hinge region provides necessary structural flexibility. The transmembrane domain not only serves as a structural anchor, but also plays a key role in transducing extracellular ligand-binding signals to the intracellular region.

The antigen-binding domain typically comprises a single-chain variable fragment (scFv), which is formed by linking the variable heavy (VH) and light (VL) chains of a monoclonal antibody. This scFv enables MHC-independent antigen recognition. The affinity of the scFv critically influences CAR-T cell efficacy. Low-affinity CAR-T cells display heightened sensitivity to the PD-1/PD-L1 axis. Therefore, their therapeutic activity can be substantially enhanced by checkpoint blockade, with a more favorable safety profile. Conversely, although high-affinity CAR-T cells are capable of resisting PD-L1-mediated inhibition and exhibiting potent initial antitumor responses, they are frequently associated with CAR-T cell exhaustion [ 17 ]. Moreover, CAR constructs may exhibit tonic signaling, defined as constitutive and persistent activation in the absence of ligand engagement. Excessive tonic signaling impairs CAR-T cell persistence, promotes premature exhaustion, and limits long-term survival in vivo [ 18 ]. Molecules such as CD5, Ly6C, and Nur77-GFP serve as biomarkers of tonic signaling [ 19 ].

The hinge and transmembrane domains collectively transmit extracellular signals to the intracellular domains. Hinge composition and length influence both the toxicity and persistence of CAR-T cells. For instance, the extracellular domain of CD27 is a natural substrate for matrix metalloproteinases, using it as the hinge region renders the CAR susceptible to cleavage. This cleavage generates soluble CD27, compromising CAR surface stability and function. Substituting the CD27 hinge with the CD8-derived hinge improves in vivo efficacy [ 20 ]. The transmembrane domain is typically derived from single-pass transmembrane proteins such as CD28, CD8, CD4, or B7-1. The selection of this domain affects CAR stability and function [ 21 ]. Moreover, recent evidence indicates that the transmembrane domain can also participate in trogocytosis processes, modulating the cytotoxic functions of CAR-T cells [ 22 ].

The intracellular domain primarily mediates signal transduction. Structurally, it comprises an activation module derived from the CD3ζ chain, integrated with one or more co-stimulatory domains. The CD3ζ chain has three immunoreceptor tyrosine-based activation motifs (ITAMs). When these ITAMs are phosphorylated, they recruit and activate the ζ-associated protein of 70 kDa (ZAP70), triggering downstream signaling cascades [ 23 ]. CD28 and 4-1BB are the most frequently utilized co-stimulatory domains. The intracellular region of CD28 harbors three conserved signaling motifs (YMNM, PYAP, PRRP). Upon activation with CD80/CD86, CD28 recruits signaling molecules such as LCK and PKCθ via the PYAP motif. This recruitment synergistically activates transcription factors such as NF-κB, AP-1, and NFAT. This process promotes CAR-T cell survival, proliferation, and cytokine production [ 24 – 26 ]. In CAR-T cells, 4-1BB signaling recruits TRAF1, TRAF2 and TRAF3. This recruitment can result in the activation of the NF-κB pathway and the enhancement of CAR-T cell persistence [ 27 , 28 ]. Other co-stimulatory domains, such as OX40, ICOS, and CD27, are also employed. Specifically, CD27 and OX40 engage TRAF2 and TRAF5 to activate both the canonical NF-κB and MAPK signaling pathways [ 29 – 31 ].

Lipid metabolism has been identified as a key pathway for sustaining memory CAR-T cells. Augmenting lipid metabolism improves CAR-T cell persistence and antitumor efficacy [ 32 ]. The 4-1BB co-stimulatory domain promotes lipid metabolism and fosters a central-memory phenotype, enhancing persistence. Conversely, CD28 co-stimulation drives glycolytic metabolism and fosters effector-memory differentiation at the expense of long-term survival [ 28 ].

The first-generation CAR only contains the CD3ζ-derived signaling domain. This design enables basic killing of target cells with MHC-independent T-cell activation, representing the first proof of concept for CAR-T cell technology. Compared to later generations, it lacks sufficient persistence and antitumor efficacy. The second-generation CAR has a CD3ζ activation domain and a co-stimulatory domain for transmitting auxiliary stimulatory signals. The co-stimulatory domain is usually from CD28 or 4-1BB, as represented by the approved products Kymriah (4-1BB, Tisagenlecleucel) and Yescarta (CD28, Axicabtagene ciloleucel). The second-generation CAR-T cells are hindered by limited persistence and subsequent relapse of the disease [ 2 , 4 , 5 ]. The third-generation designs add two or more co-stimulatory domains alongside CD3ζ to further refine this architecture. This design can generate synergistic signaling that confers broader functional capabilities on CAR-T cells, contributing to favorable persistence. The overactivation of multiple co-stimulatory domains leads to a higher incidence of severe side effects such as CRS and more rapid exhaustion [ 2 , 26 ]. Cytokines also play key roles in CAR-T therapy. Effector cytokines such as IL-2, IL-12, IL-15, and IFN-γ can enhance the cytotoxicity of lymphocytes. IL-15 can optimize the in vitro expansion and metabolic fitness of CAR-T cells by modulating mTORC1 activity. These insights led to the emergence of the fourth-generation CAR, which typically places cytokine genes under the control of constitutive promoters or inducible elements such as the NFAT promoter, enabling sustained or localized cytokine release, respectively. This design can remodel the TME and enhance persistence. However, the fourth-generation CAR-T cells still exhibit markedly reduced efficacy against solid tumors and carry the risk of on-target off-tumor toxicity due to transgenic cytokine release in healthy tissues [ 2 , 33 – 35 ]. The fifth-generation CAR integrates intracellular domains from distinct signaling pathways (JAK-STAT pathway) to enhance cellular performance. By incorporating the IL-2 Rβ cytoplasmic domain and an YXXQ motif, this CAR design enables JAK-STAT3/5 activation. This design utilizes triple-signaling to promote more persistent CAR-T cells. However, the fifth-generation CAR-T cells exhibit off-target and low tumour specificity [ 2 , 36 – 38 ]. Figure 1 illustrates the evolution of CAR structure across five generations. and Table 1 summarizes the structural evolution, key advantages, limitations, and representative products of different CAR generations.

CAR-T signaling pathways

The CAR coding sequence is delivered to T cells through lentiviral vectors or non-viral methods, such as lipid nanoparticles (LNPs) carrying mRNA or DNA. Numerous researches demonstrate that the choice of delivery system can influence the gene transfer efficiency of the CAR construct [ 39 , 40 ]. Following delivery, the CAR transgene is transcribed and translated by the host cell machinery. This process leads to expression of the CAR protein on the T cell surface and the generation of functional CAR-T cells. This manufacturing process is typically performed ex vivo. Furthermore, the source of the T cells used for engineering can also affect the overall efficacy.

Initiation of CAR signaling: from immune synapse formation to initial activation

CAR-T cells recognize and bind tumor cell surface antigens through their chimeric antigen receptors (CARs), leading to the formation of an immune synapse (IS). The IS serves as the structural platform for T cell cytotoxicity, which enables cytoskeletal reorganization, directional release of cytotoxic granules, and activation of downstream signaling pathways. The IS formation depends on coordinated ligand and receptor interactions, including CD58/CD2, ICAM-1/LFA-1, and CD99 [ 41 – 43 ].

For intracellular signal transduction, CARs mimic signaling mechanisms from the native T cell receptor (TCR). CAR signaling is progressively amplified by kinases such as LCK and ZAP-70. The signaling ultimately activates the key downstream pathways including NFAT and NF-κB [ 44 ]. Upon activation, CAR-T cells exert cytotoxic effects primarily through three mechanisms, which are the release of perforin and granzymes, activation of the death-receptor pathway, and secretion of cytokines such as IFN-γ [ 45 , 46 ]. Additionally, the dominant killing mechanisms differ between tumor types. In hematological malignancies, CAR-T cells predominantly utilize granzyme B (GZMB) and granzyme A (GZMA). In solid tumors, CAR-T cells exhibit a greater dependence on GZMB and the FasL-mediated death-receptor pathway [ 47 ].

CAR-T cells recognize target antigens through a CAR that integrates monoclonal antibody specificity with TCR-CD3 signaling machinery. Upon antigen binding, the Src-family kinase LCK typically phosphorylates ITAMs on the CD3ζ chain, initiating downstream signaling. However, in the absence of LCK, another Src-family kinase, FYN, can phosphorylate ITAMs and sustain signal transduction. LCK signaling typically induces potent immune activation that may lead to excessive CAR-T cell activation and exhaustion, while FYN signaling preferentially supports proliferation and survival [ 48 ].

ITAMs function as core signaling modules. Their number correlates closely with signal strength. Increasing the ITAM count lowers the antigen-recognition threshold and enhances CAR signaling, while reducing the number of ITAMs diminishes responses to tumors with low antigen density [ 49 ]. Phosphorylated ITAMs provide docking sites for ZAP-70. Activated ZAP-70 phosphorylates adapter proteins such as LAT and SLP-76, forming a signaling complex that recruits PLCγ1 to initiate downstream pathways. In CAR-T cells, LAT is not indispensable for signal transduction, as CARs can partially activate downstream signaling independently of LAT. However, its absence impairs specific functions, including IL-2 production [ 50 , 51 ].

In terms of pathway activation, CAR-T cells can engage three canonical downstream axes. They are Ca 2 + -NFAT, MAPK, and NF-κB. Current evidence indicates that ZAP-70 participates directly in NF-κB activation. However, its precise contribution to Ca 2 + signaling and MAPK pathway activation remains unclear, and the mechanisms need further elucidation [ 52 – 54 ].

Transmission of core signaling pathways

MAPK/ERK and NFAT pathways

During T cell activation, ZAP-70 recruits the GRB2–SOS1 complex to activate membrane-associated RAS, initiating the RAF-MAPK cascade and enhancing the activity of the transcription factor AP-1 [ 55 , 56 ]. Simultaneously, PLCγ1 hydrolyzes PIP₂ to generate IP₃ and DAG. IP₃ triggers intracellular Ca 2 + release and NFAT pathway activation, and DAG modulates T cell function through the PKCθ-NF-κB pathway and the RAS-ERK pathway. AP-1 and NFAT act synergistically to promote the expression of genes such as IL-2 [ 57 , 58 ]. These pathways ultimately converge in the nucleus to govern T cell activation, proliferation, and effector functions.

In CAR-T cells, enhanced MAPK signaling promotes AP-1 activity. Sustained MAPK signaling supports CAR-T cell proliferation and suppresses exhaustion [ 56 , 59 ]. CAR-T cells upregulate phosphatases such as DUSP2 and DUSP4, which negatively regulate MAPK/ERK signaling. Exploiting this regulatory mechanism, exogenous ERK2 inhibitors can further improve CAR-T cell transduction efficiency [ 60 ]. The CD28 co-stimulatory domain markedly accelerates CD3ζ phosphorylation via LCK recruitment, amplifying MAPK signaling. However, phosphorylated CD3ζ ITAMs also recruit the phosphatase SHP-1, which dephosphorylates multiple pathway components and attenuates signal propagation. Research suggests that reducing the number of ITAMs on CD3ζ may accelerate MAPK pathway activation kinetics, promoting proliferation and restraining exhaustion [ 61 ]. In the absence of AP-1, NFAT may instead drive CAR-T cell exhaustion and cooperate with repressors such as TOX and TOX2, collectively impairing function [ 62 , 63 ]. Analogous to native T cell signaling, it is inferred that in CAR-T cells, particularly those containing the 4-1BB co-stimulatory domain, canonical NF-κB activation occurs via mechanisms downstream of PKC signaling [ 44 ]. The Ca 2 + -NFAT pathway in CAR-T cells, involving PIP₂ hydrolysis to generate IP₃, elevating intracellular Ca 2 + levels and inducing NFAT dephosphorylation, is similar to that in native T cells [ 64 ].

Furthermore, from a bioengineering perspective, CAR-T cells expressing membrane-anchored SLP-76 can bypass the requirement for ZAP-70 recruitment. When these CAR-T cells are activated, they exhibit markedly enhanced PLCγ1-dependent NFAT transcriptional activity, inducing robust expression of cytokines such as IL-2. However, the precise mechanisms remain to be fully elucidated [ 65 ].

NF-κB signaling

The activation status and dynamic regulation of the NF-κB pathway critically shape CAR-T cell function and fate. The 4-1BB co-stimulatory domain efficiently activates NF-κB signaling via the recruitment of TRAF molecules, substantially enhancing CAR-T cell persistence [ 28 ]. In addition to co-stimulatory signals, upstream-activated ZAP-70 also induce NF-κB pathway, augmenting cytotoxicity [ 54 ].

The precise mechanisms of NF-κB signaling in CAR-T cells involve the activation of both the canonical and non-canonical pathways. Evidence indicates that the 4-1BB domain specifically activates the non-canonical NF-κB pathway. This process relies on stabilization of NF-κB inducing kinase (NIK), phosphorylation of p100, and its processing to p52, ultimately driving nuclear translocation of the RelB-p52 dimer. This cascade suppresses the expression of the pro-apoptotic protein Bim, enhancing cell survival and long-term persistence [ 66 ]. CAR-T cells co-expressing SMAD7 effectively suppress the canonical NF-κB pathway, as evidenced in reduced phosphorylation of TAK1, IκB, and p65. This alleviates adverse effects associated with excessive NF-κB activation and identifies SMAD7 as a negative regulator of the canonical NF-κB pathway [ 67 ]. Furthermore, Dou et al. revealed that under antigen-independent conditions, canonical NF-κB activation mediated by 4-1BB induces ICAM-1 overexpression, leading to cell aggregation. Coupled with sequestration of the negative regulator A20, this alleviates inhibition of the receptor-interacting serine/threonine kinase 1-receptor-interacting serine/threonine kinase 3-mixed lineage kinase domain like pseudokinase (RIPK1/RIPK3/MLKL) pathway, synergistically promoting apoptosis and necroptosis [ 68 ]. The canonical activation supports CAR-T cell cytotoxicity, and the non-canonical activation primarily enhances persistence [ 66 , 67 , 69 ].

PI3K-AKT-mTOR signaling

In CAR-T cells, the PI3K-AKT-mTOR axis functions as a critical regulatory pathway. The signaling is primarily activated by the CAR molecule in conjunction with the CD28 costimulatory domain and further augmented by cytokines such as IL-2. This pathway governs critical processes including T cell proliferation, survival, migration, and memory formation [ 70 ].

At the molecular level, in native T cells, activated PI3K generates PIP₃, which recruits AKT to the plasma membrane for phosphorylation by mTORC2. This process results in full AKT activation and initiation of mTORC1 signaling. This pathway ultimately drives metabolic reprogramming characterized by enhanced glycolysis, protein synthesis, and lipogenesis [ 71 , 72 ]. In CAR-T cells, the activation of AKT and mTOR is essential for robust cellular activation, although the precise downstream signaling networks have yet to be systematically elucidated [ 73 ].

Inhibiting the PI3K-AKT-mTOR pathway exerts divergent effects in CAR-T cells. The suppression of AKT or mTOR can compromise effector CAR-T cell proliferation and antitumor function. Such inhibition has conversely been shown to enhance CAR expression, foster the acquisition of a memory phenotype, and mitigate CAR-T cell exhaustion [ 74 , 75 ]. For example, the PI3Kδ inhibitor idelalisib or the dual PI3Kδ/γ inhibitor duvelisib improves CAR-T cell differentiation status, reduces the expression of exhaustion markers, and balances the CD4/CD8 ratio without compromising viability or expansion. These effects augment CAR-T cell persistence and efficacy [ 76 , 77 ]. Similarly, rapamycin (an mTORC1 inhibitor) during in vitro expansion can counteract mTORC1-mediated downregulation of CXCR4, potentiating bone marrow homing and antitumor activity [ 78 ].

Based on the above research, precise regulation of the PI3K-AKT-mTOR pathway critically dictates CAR-T cell fate. Its activation augments effector function and antitumor potency, while moderate inhibition promotes stem properties and delays differentiation and exhaustion. These insights suggest that the temporal and context-dependent modulation of this axis may prove important for optimizing next-generation CAR-T cell therapies.

Canonical Wnt signaling pathway

In T cells, the canonical Wnt pathway governs cell fate. In the absence of Wnt signaling, β-catenin is phosphorylated by a destruction complex (comprising APC, GSK3β, and other scaffold proteins), leading to its ubiquitination and proteasomal degradation. This mechanism ensures that β-catenin remains low cytoplasmic levels. Upon activation, the binding of Wnt ligands to Frizzled and LRP receptors signals via Dishevelled to disrupt this complex. This event stabilizes β-catenin, facilitating its cytoplasmic accumulation and subsequent nuclear translocation. In the nucleus, β-catenin interacts with TCF/LEF transcription factors to induce the expression of key target genes, such as c-Myc and Cyclin D1. This mechanism promotes proliferation, differentiation, and survival [ 79 , 80 ].

Although the precise regulatory mechanisms of the Wnt pathway in CAR-T cells have not been fully elucidated, recent studies highlight that it have potential significance. Zhu et al. reported that Class I histone deacetylase (HDAC) inhibitors such as M344 and chidamide, increase histone H3K27 acetylation, epigenetically activating Wnt pathway gene expression. This promotes β-catenin dephosphorylation and enhances downstream signaling. As a result, it bolsters a memory phenotype and delays exhaustion in CAR-T cells [ 81 ]. Additionally, the VEGFR inhibitor Axitinib has been demonstrated to suppress CAR-T cell exhaustion through the activation of the canonical Wnt pathway [ 82 ]. Collectively, current evidence indicates that the Wnt pathway plays a critical role in preserving memory traits and mitigating exhaustion in CAR-T cells. Targeted modulation of this pathway may represent a promising strategy for enhancing the functional persistence of CAR-T cell therapies [ 81 ].

Functional output

Tumor cell killing

Death-receptor-mediated killing pathways

CAR-T cell-mediated tumor killing involves multiple death-receptor pathways, primarily Fas-FasL, TRAIL-TRAILR, and TNFα-TNFR1. The Fas-FasL pathway mediates cytotoxicity against not only antigen-positive (Ag + ) tumor cells but also antigen-negative (Ag − ) cells. In a model containing both Ag + and Ag − tumor cells, Upadhyay et al. demonstrated that CAR-T cells can engage Fas on target cells via FasL, activating downstream apoptotic signaling [ 83 ]. The binding of FasL to Fas recruits FADD, which activates procaspase-8 and −10 via its death effector domain. This cascade culminates in caspase-3 activation and the induction of caspase-dependent apoptosis in target cells [ 84 , 85 ]. As CAR-T cells themselves express Fas, surface FasL may bind to Fas on neighboring CAR-T cells, resulting in fratricide and impaired persistence. To address this limitation, the inhibition of Fas signaling in CAR-T cells has been shown to prevent fratricide and enhance CAR-T cell antitumor efficacy [ 86 – 88 ].

The TNF signaling pathway also plays a key role. Activated CAR-T cells express TNF family ligands that bind to TNFR1 (encoded by TNFRSF1A ) on tumor cells, initiating apoptosis. During TNFR1-mediated cytotoxicity by CAR-T cells, FADD functions as a central adapter for apoptosis signal transduction. RIPK1 acts as a key downstream kinase modulated by inhibitor of apoptosis proteins. SMAC mimetics stabilize RIPK1 by antagonizing these inhibitory proteins, potentiating the TNF-RIPK1 apoptotic pathway. The knockout of RIPK1 or TNFRSF1A in tumor cells abolishes the sensitizing effect of SMAC mimetics. This confirms the RIPK1–TNFR1 axis as a critical mechanism by which pharmacological intervention augments CAR-T cell cytotoxicity [ 84 ].

Similarly, the TRAIL-TRAIL-R pathway contributes to CAR-T cell cytotoxicity. In native T cell cytotoxicity, the binding of TRAIL to its receptor assembles the death-inducing signaling complex (DISC), directly activating the caspase cascade and inducing p53-independent apoptosis [ 89 ]. In CAR-T cells, the scFv domain of the CAR can directly activate TRAIL-R1, promoting receptor clustering, FADD recruitment, and caspase-8 activation. This cascade event triggers the caspase cascade and programmed tumor cell death [ 46 , 90 ].

Perforin/Granzyme cytotoxic pathway

CAR-T cells primarily exert cytotoxicity against target cells directly via the perforin/granzyme pathway [ 91 ]. Following CAR engagement with tumor cell surface antigens, the intracellular CD3ζ domain is activated. This initiates downstream signaling and triggers the release of cytotoxic granules containing perforin and granzymes, such as GZMB. Perforin forms pores in the target cell membrane, facilitating the entry of GZMB into the cytoplasm. In the cytoplasm, GZMB activates the caspase cascade and/or directly cleaves critical substrates, irreversibly inducing apoptosis [ 46 ]. In addition to apoptosis, CAR-T cells can induce tumor cell pyroptosis through GZMB. GZMB activates caspase-3, which cleaves the highly expressed protein gasdermin E, enhancing antitumor activity. However, pyroptosis can exacerbate the severity of CRS [ 92 ].

IFN-γ and other cytotoxic mechanisms

In addition to the aforementioned pathways, IFN-γ-mediated cytotoxicity is significant in CAR-T cells. For example, CD4 + CAR-T cell cytotoxicity relies predominantly on IFN-γ secretion rather than perforin release. The binding of IFN-γ to its receptor activates downstream signaling, upregulating pro-apoptotic genes such as STAT1 and XAF1 . However, the efficacy depends on the sensitivity of tumor cells to IFN-γ [ 93 ]. IFN-γ has been reported to restrict expansion of CAR-T cells incorporating the CD28 co-stimulatory domain [ 94 ].

CRS

According to the American Society for Transplantation and Cellular Therapy (ASTCT) consensus, CRS is classified into five grades based on the severity of hypotension and hypoxia, with fever (≥38℃) required for diagnosis though fever may be masked after intervention. Grade 1 involves fever without hypotension or hypoxia. Escalation to Grade 2 marks the onset of hypotension not requiring vasopressors and/or hypoxia requiring low-flow nasal cannula (≤6 L/min). Grade 3 is characterized by hypotension requiring one vasopressor and/or hypoxia requiring high-flow nasal cannula (>6 L/min), facemask, non-rebreather mask, or venturi mask. Advancement to Grade 4 represents a life-threatening condition, characterized by hypotension requiring multiple vasopressors (excluding vasopressin) or hypoxia requiring positive pressure ventilation. Grade 5 is defined as death directly attributable to CRS [ 95 ].

CRS is regarded as a systemic inflammatory response driven by excessive CAR-T cell activation. Following tumor antigen recognition, activated CAR-T cells undergo rapid proliferation and release cytokines such as GM-CSF, IFN-γ and TNF-α. These cytokines further activate endogenous immune cells, triggering the substantial release of additional inflammatory mediators including IL-1, IL-2, IL-6 and soluble IL-6 R (sIL-6 R). This inflammatory cascade precipitates CRS [ 23 , 96 ]. CD4 + T cells appear to contribute more to CRS induction than CD8 + T cells [ 97 ].

Myeloid cells like macrophages and monocytes serve as central amplifiers in CRS progression. They are activated through two principal routes. The first route is mediated by cytokines such as IFN-γ and GM-CSF released from CAR-T cells [ 98 ], and the second route is through the ligation of CD40 on myeloid cells by CD40L on CAR-T cells. This activation results in abundant production of inflammatory mediators including IL-6, IL-1, and iNOS [ 98 – 100 ]. Moreover, recent studies reveal a critical positive feedback loop, where activated macrophages synthesize and secrete catecholamine. These molecules bind to adrenergic receptors on the macrophages in an autocrine manner to further potentiate cytokine release. This leads to exponential amplification of inflammatory signals and ultimately precipitates severe CRS [ 101 ].

IL-6 serves as a central mediator of CRS. IL-6 is secreted primarily by macrophages and monocytes. The peak expansion of CAR-T cells in peripheral blood and serum IL-6 concentrations correlate positively with CRS severity [ 96 ]. In canonical signaling, IL-6 binds to the membrane-bound IL-6 R, inducing gp130 dimerization and activating the downstream JAK-STAT3 and SHP2-MAPK pathways [ 102 , 103 ]. In the context of CAR-T cell therapy, IL-6 signaling is particularly relevant. The host myeloid cells activated by CAR-T cells produce IL-6 and sIL-6 R. The IL-6 and sIL-6 R complex that can activate cells in the TME, including CAR-T cells, via gp130. This can markedly enhance their expansion and inflammatory functions and exacerbate CRS [ 104 ]. As cytokine levels rise and CAR-T cells expand, the systemic inflammation intensifies, causing vascular endothelial injury and increased permeability. This is clinically characterized by hypotension, hypoxia, and organ dysfunction. Following tumor clearance and antigen stimulation wanes, CAR-T cell numbers and cytokine levels gradually decline. Ultimately, CRS resolves. The patients may retain CAR-T memory cells, enabling long-term persistence [ 105 , 106 ]. However, moderate to severe CRS requires prompt intervention and cannot wait to resolve spontaneously, otherwise the risk of CRS is extremely high.

From the perspective of clinical management, CRS is always associated with CAR-T cell activity, and a certain degree of CRS is often regarded as an indicator of treatment efficacy. Hence, balancing efficacy with CRS constitutes a central challenge [ 107 ]. IL-6 has limited impact on CAR-T cell intrinsic antitumor function, which provides a rationale for using IL-6 signaling blockade such as tocilizumab to manage CRS symptoms without substantially compromising antitumor efficacy [ 64 ]. Supportive care and targeted interventions are essential to maximize therapeutic benefit while maintaining severe toxicity within manageable levels.

The persistence and toxicity of CAR-T cells are also governed by the restriction of TME. Therefore, comprehensively dissecting the interactions between TME components and CAR-T cells is paramount for enhancing the persistence and safety profile.

Impact of the immune microenvironment on CAR-T cells

TGF-β in the TME

In the TME, elevated levels of TGF-β exert a complex influence on CAR-T cell function. SMAD2 and SMAD3 function as key downstream transcription factors in the TGF-β signaling pathway. They act as central immunosuppressive mediators. These factors substantially inhibit CAR-T cell persistence and drive CAR-T cell exhaustion. The targeted degradation of SMAD2 and SMAD3 abrogates TGF-β signal transduction, counteracting the immunosuppressive influence of the TME on CAR-T cells [ 108 ]. TGF-β inhibits CAR-T cell activation and accelerates exhaustion, thereby compromising persistence. However, as part of a negative feedback mechanism, TGF-β can induce the expression of its own inhibitors, SMAD6 and SMAD7. Exploiting this pathway, the co-expression of SMAD7 in CAR-T cells has been shown to effectively suppress TGF-β signal transduction, abrogating its suppressive effects. This results in enhanced antitumor efficacy and persistence, along with mitigated toxicity [ 67 , 109 , 110 ]. SMAD4 is a central transcriptional regulator and plays a distinct dual role. It not only transduces canonical inhibitory signals but also transcriptionally upregulates LCK. This upregulation activates the downstream PI3K/AKT pathway, promoting CAR-T cell proliferation and survival [ 111 , 112 ]. Research by Tang et al. demonstrated that TGF-β in the TME markedly suppresses CAR-T cells via two distinct mechanisms. First, it upregulates the transcription factor FOXP3, inducing iTreg phenotype in CAR-T cells, impairing cytotoxicity and conferring immunosuppressive properties. Second, TGF-β drives CAR-T cell exhaustion, significantly upregulating key immune checkpoint molecules such as PD-1 and TIM-3. These immune checkpoints result in CAR-T cells functional decline and restricted proliferation [ 113 ]. These findings suggest that precise modulation of the TGF-β signaling axis will be a promising strategy for optimizing CAR-T cell therapy.

Immune checkpoints

Multiple immune checkpoints, including PD-1, CTLA-4, LAG-3, TIGIT, and TIM-3, constrain the persistence and function of CAR-T cells. Immune checkpoints also represent a major cause of poor responses or relapse [ 114 , 115 ].

The PD-1/PD-L1 pathway is among the most extensively studied inhibitory mechanisms. Binding to PD-L1 expressed on tumor cells, PD-1 on CAR-T cells delivers a strong inhibitory signal. A study found that this pathway critically impairs the persistence of CAR-T cells with 4-1BB in the TME. Mechanistically, PD-1 signaling counteracts the 4-1BB-mediated enhancement of oxidative phosphorylation capacity. This inhibitory signal hinders differentiation toward a central memory phenotype and compromises long-term immune persistence [ 116 ].

CTLA-4 inhibits T cell activation by binding B7 ligands (CD80/CD86) on antigen-presenting cells, out-competing the co-stimulatory receptor CD28. Owing to the fact that CTLA-4 has a higher affinity for B7, it restricts CD28-mediated co-stimulatory signals, suppressing T cell activation and proliferation [ 117 ]. In CAR-T cell design, genetic deletion of CTLA-4 amplifies downstream signaling mediated by the CD28 domain, equipping CAR-T cells with enhanced proliferation and antitumor efficacy [ 115 ].

LAG-3 is an inhibitory receptor highly expressed on exhausted T cells. In native T cells, it associates with the T cell receptor complex, disrupting the interaction of CD4/CD8 with the tyrosine kinase LCK. This limits phosphorylation of CD3ζ and ZAP70, impairing downstream signaling [ 118 ]. LAG-3 ligands include MHC-II, galectin-3, and fibrinogen-like protein 1 [ 119 ]. Genetic disruption of LAG-3 has been demonstrated to rejuvenate the antitumor response of CAR-T cells [ 120 ].

TIGIT is another inhibitory receptor expressed on activated T cells and NK cells. When TIGIT binds to the ligands CD155 and CD112 on tumor or antigen-presenting cells, it transmits inhibitory signals that attenuate T cell antitumor efficacy [ 121 – 123 ]. Targeting TIGIT is considered a promising strategy to potentially prevent relapse following CAR-T cell therapy [ 124 ].

TIM-3 has a dynamic regulatory mechanism. In the native T cells, its intracellular domain binds BAT3 and LCK in the absence of ligand, facilitating activation. When binding ligands such as galectin-9, TIM-3 recruits multiple TCR signaling components (FYN, LCK, and GRB2) and initiates inhibitory signaling [ 125 ]. In CAR-T cells, TIM-3 activation directly leads to functional decline and exhaustion. Conversely, inhibiting TIM-3 signaling enhances persistence and cytotoxicity [ 126 , 127 ].

Trogocytosis

Trogocytosis involves the active transfer of membrane fragments and surface proteins between cells in contact. This process serves as a significant mechanism contributing to tumor immune escape and impaired CAR-T cell function [ 128 ].

Following binding to tumor cells, CAR-T cells acquire target antigens from the tumor cell surface via trogocytosis and incorporate them on CAR-T cell membrane. This process leads to two detrimental consequences. One is that tumor cells downregulate antigen expression by having surface antigens stripped, which contributes to evasion of CAR-T cell recognition. The other is that CAR-T cells acquiring tumor antigens become susceptible to lysis by neighboring CAR-T cells, depleting the effector pool and compromising therapeutic efficacy. This phenomenon is known as fratricide between CAR-T cells [ 129 ].

Moreover, trogocytosis is a bidirectional process. Tumor cells can also acquire CAR molecules from the CAR-T cell membrane. These transferred CAR molecules displayed on the tumor cell surface can mask the original antigen, interfering with normal recognition and cytotoxicity by CAR-T cells. Concurrently, the loss of CAR molecules from CAR-T cells attenuates their activation and functional integrity, further contributing to immune escape [ 130 ].

Tumor-associated macrophages

Tumor-associated macrophages (TAMs) within the TME exert a dual regulatory influence on CAR-T cell therapy. Myeloid lineage cells, particularly TAMs, can be continuously activated by substantial quantities of tumor cell debris. This activation leads to the hypersecretion of pro-inflammatory cytokines, including IL-6, IL-1β, and TNF-α. This process contributes to CRS and establishes a positive feedback loop, where the immune activation and systemic inflammatory responses mutually reinforce each other. The severity of this response is positively correlated with the CAR-T cells infusion dosage [ 131 – 133 ].

The polarization status of macrophages influences the persistence of CAR-T cell therapy. In response to chronic inflammatory stimulation, macrophages within the TME frequently polarize from the antitumor M1 phenotype macrophages to the immunosuppressive M2 phenotype macrophages [ 134 ]. These M2 macrophages secrete immunosuppressive cytokines, such as TGF-β and IL-10, which compromise CAR-T cell function and simultaneously promote angiogenesis and tumor progression. Therefore, this immunosuppressive environment limits long-term therapeutic persistence and promotes tumor recurrence [ 135 ]. M1 macrophages possess intrinsic pro-inflammatory properties and can contribute to CRS in certain conditions. While the M1 macrophage polarization alone does not necessarily cause CRS, the key factor is the mode of induction and the microenvironmental context. In the context of excessive CAR-T activation, M1 polarization is driven by IFN-γ released from overactivated CAR-T cells. This process creates a positive inflammatory feedback loop that amplifies systemic inflammation [ 136 ]. Pharmacologically induced M1 repolarization such as via CSF-1 R inhibitors does not lead to CRS [ 137 ].

From the research presented, macrophage polarization toward the M1 macrophages is associated with tumor regression, and the M2 macrophages fosters tumor progression. Reprogramming macrophage plasticity constitutes a critical strategy to disrupt the interplay between local immunosuppression and systemic inflammation, augmenting the long-term persistence of CAR-T cells [ 138 , 139 ].

Overcoming insufficient persistence of CAR-T cells

The persistence of CAR-T cell therapy is governed by both intracellular transcriptional networks and the external TME, illustrating that it takes two to tango. Current strategies to improve persistence focus on three primary domains. They are optimizing intracellular signaling, overcoming TME immunosuppression, and improving CAR molecule transfection.

Optimizing intracellular signaling pathways

Structural engineering

CAR-T cells can exhibit spontaneous (tonic) activation without antigen. Moderate basal signaling supports CAR-T cell survival and persistence. However, excessive signaling induces exhaustion and insufficient signaling compromises efficacy. In the absence of antigen stimulation, CAR-T cells may exhibit tonic signaling. The tonic signaling is due to the self-aggregation mediated by positively charged patches on the surface of the scFv domain. Rational modulation of this region can optimize tonic signaling without impairing antigen recognition [ 140 ]. In addition to refining tonic signaling, novel strategies for CAR structural engineering have been developed to enhance persistence. For example, in the scFv domain, the bispecific Tan-CAR targeting both CD19 and CD22, can eliminate tumor cells and reduce antigen-negative relapse. However, clinical trials continue to report relapses associated with insufficient persistence [ 141 , 142 ]. E3K CAR-T cells targeting endosialin (CD248) can indirectly suppress tumor growth and metastasis by disrupting perivascular stroma [ 143 ]. In the intracellular domain, Zhang et al. anchored IL-7 to the CAR-T cell membrane via a transmembrane domain. Their study found that the CD8 transmembrane domain provided an optimal balance of high membrane expression and low shedding of IL-7, boosting cell function and persistence [ 21 ]. Furthermore, the incorporation of the functional domain of CD99 is a key membrane protein regulating actin-microtubule interactions at the IS. Incorporating CD99 can improve CAR-T cell antitumor efficacy [ 42 ]. In preclinical models, CAR-T cells co-expressing IL-15 showed superior persistence and function [ 143 , 144 ]. However, in clinical trials, CAR-T cell therapy co-expressing IL-15 elevated the risk of CRS [ 143 , 145 ]. MC-CAR-T cells, which incorporate MyD88 and CD40 as co-stimulatory domains, exhibited superior persistence in animal and non-human primate models [ 143 , 146 , 147 ]. Reengineering the primary signaling architecture offers a strategy to improve specificity. For instance, traditional CD3ζ signaling domains cannot reliably distinguish tumor cells from normal tissue. CAR designs using ZAP-70 or PLCγ1 as signaling domains can activate downstream signals independently of CD3ζ, enhancing antigen-dependent signaling and persistence [ 50 ].

Signal pathway modulation

Pharmacologically, dasatinib inhibits LCK to induce a rapid and reversible functional pause in CAR-T cells, offering a tunable mechanism to modulate their activity. When the drug is withdrawn, the LCK blockade is promptly relieved, allowing for the full restoration of antitumor function. This strategy maintains CAR-T cell antitumor efficacy while simultaneously controlling CRS [ 143 , 148 ]. However, mutating the PYAP motif in CD28 to abrogate LCK recruitment reduces IL-2 production and enhances resistance to Treg suppression. Nonetheless, this mutation may impair overall CAR-T cell efficacy [ 27 ]. These findings indicate that excessive LCK signaling drives terminal differentiation and exhaustion, while insufficient signaling compromises effector function. Novel signaling strategies such as chimeric adaptor proteins (CAPs) and CAR tyrosine kinases (CAR-TKs) demonstrate superior signaling persistence and antitumor function. CAR-TK activation is LCK-independent and is designed to bypass upstream inhibition and directly activate downstream pathways, prolonging signal duration. Moreover, CAR-TK containing only the intracellular ZAP70 domain exhibits low basal signaling and high antigen-dependent transduction, enhancing CAR-T cell persistence and antitumor efficacy [ 143 , 149 ].

Under low-antigen conditions, attenuated signaling through the T cell activation linker protein (LAT) leads to insufficient activation of the downstream MAPK/ERK pathway, impairing CAR-T cell persistence. To mitigate this, the ALA-CAR-T cell (Adjunctive LAT-activating CAR T cell) bicistronic CAR platform was developed. The CAR-T cells combine a second-generation BBzCAR with a LAT-CAR that enhances LAT phosphorylation. By upregulating MAPK signaling and potentiating NFAT/AP-1 synergy, the LAT-CAR augments CAR-T cell activation and long-term persistence [ 56 , 143 ].

Numerous studies have investigated the modulation of CAR-T cell activation kinases and signaling pathways. Specifically, MEK inhibitors (trametinib, cobimetinib, binimetinib) have been shown to attenuate antigen-induced exhaustion and differentiation of CAR-T cells in preclinical models. This is achieved by downregulating the effector molecules c-Fos and JunB, suppressing genes linked to exhaustion, differentiation, glycolysis, and apoptosis [ 143 , 150 ]. In a clinical trial for GD2-positive melanoma, the combination of CAR-T cell therapy and BRAF/MEK inhibitors resulted in a partial response. While discerning the specific contributions of CAR-T cells versus targeted agents proved challenging [ 151 ]. The engineering of transcription factors offers an intrinsic method to sustain effector function. Basic leucine zipper ATF-like transcription factor forms heterodimers with Jun family transcription factors and cooperates with interferon regulatory factor 4 (IRF4) to bind AP1-IRF sites. This process can antagonize exhaustion by NFAT in the absence of AP-1, and enhance persistence [ 63 , 152 ].

Metabolic reprogramming

The metabolic state of T cells is closely associated with their differentiation fate, including CAR-T cells. Naive and memory T cells primarily rely on mitochondrial oxidative phosphorylation (OXPHOS) and fatty acid oxidation (FAO), exhibiting low glycolytic activity, high mitochondrial spare respiratory capacity (SRC), and long lifespan. Effector T cells primarily rely on high glycolysis to produce ATP rapidly to support proliferation and cytotoxicity. This metabolic program results in a short lifespan and a propensity toward exhaustion [ 153 , 154 ].

The PI3K-AKT-mTOR axis is impaired by hypoxia and glucose deprivation. To address this challenge, overexpressing glucose transporters 1 (GLUT1) and GLUT3 in CAR-T cells enhances glucose uptake and correlates with upregulated PI3K signaling, improving survival and function under metabolic stress. Furthermore, overexpression of GLUT1 and GLUT3 favors differentiation toward a Th17 phenotype and mitigates exhaustion. Ultimately, this method can enhance persistence in vivo [ 155 – 158 ]. Particularly, CAR-T cells with GLUT1 overexpression secrete more proinflammatory cytokines, enhancing potency, but the experiment did not assess CRS risk [ 155 – 157 ]. Metabolically engineering CAR-T cells to express the fructose transporter GLUT5 enables them to utilize fructose, augmenting antitumor efficacy under limited glucose conditions [ 159 ]. Additionally, mitophagy activators such as rapamycin, UrA, and metformin further enhance metabolic fitness and long-term persistence by promoting mitochondrial turnover and reactive oxygen species clearance [ 160 ]. CD38 expression is positively correlated with exhaustion. Inhibiting CD38 downregulates glycolysis and NFAT signaling, improves mitochondrial function, and promotes a memory phenotype [ 161 ].

FOXO1 is a key transcription factor downstream of PI3K. It localizes to the nucleus when unphosphorylated and upregulates memory-associated genes. When the PI3K pathway is activated, FOXO1 is phosphorylated and translocates to the cytoplasm, resulting in the loss of transcriptional activity [ 162 ]. FOXO1 deficiency leads to loss of the memory phenotype and accelerated exhaustion, and its overexpression enhances persistence, antitumor efficacy [ 163 , 164 ].

Activation of the Wnt pathway also attenuates exhaustion. Both class-I HDAC inhibitors (M344, chidamide) and the VEGFR inhibitor axitinib enhance memory formation and resistance to exhaustion through this pathway [ 81 , 82 ]. Urolithin A (UrA) induces mitophagy to activate Wnt signaling, promoting stem-like T cells formation [ 165 ]. The GLP-1 R agonist semaglutide and UrA inhibit mTOR and activate mitophagy, inducing metabolic reprogramming that enhances persistence [ 166 ]. Metformin modulates the ERK/CREB/PGC-1α signaling and balances STAT3 and STAT5 signaling. This can promote oxidative phosphorylation and fatty-acid oxidation to maintain a memory phenotype and augment in vivo persistence [ 167 ].

β-hydroxybutyrate enters the tricarboxylic acid cycle preferentially, driving OXPHOS and ATP production. This metabolic reprogramming ultimately improves CAR-T cells in vivo persistence and antitumor efficacy [ 168 ]. Overexpression of PGC-1α boosts OXPHOS activity, proliferation, and memory formation. IDH2 knockout or inhibition indirectly increases fatty acid oxidation and promote the expression of memory genes. Collectively, these strategies enhance mitochondrial biogenesis and OXPHOS, effectively promoting memory phenotype generation and antitumor function in CAR-T cells [ 154 , 169 , 170 ]. Single-gene editing can only partially restore CAR-T cell function and fails to counteract the synergistic effects of multiple metabolic barriers in the TME. The metabolic network engineering, which involves simultaneous editing of multiple genes within the same pathway or across distinct pathways, achieves a 1 + 1 ≥ 2 synergistic effect in restoring antitumor function [ 171 ]. Exemplifying this strategy, the combination of semaglutide and UrA enhances autophagy and mitophagy in CAR-T cells by inhibiting mTOR and activating Atg4b respectively, significantly improving CAR-T cell persistence and antitumor efficacy [ 166 ].

The gut microbiota and its metabolites also participate in the metabolic reprogramming of CAR-T cells. The gut microbiota produces metabolites, particularly succinate. These metabolites enhance the spare respiratory capacity of CAR-T cells and promote central memory T cell differentiation, enhancing CAR-T cell persistence and antitumor efficacy in vivo [ 172 ]. Intestinal Akkermansia spp. metabolize tryptophan into indole derivatives, which engage the aryl hydrocarbon receptor on CAR-T cells, promoting CAR-T cell persistence in the bone marrow and enhancing antitumor efficacy [ 173 ].

Overcoming tumor microenvironment suppression

Overcoming signaling inhibition

TGF-β is a potent immunosuppressive cytokine in the TME. CRISPR/Cas9 knockout of TGFBR2 abrogates TGF-β signaling, enhancing CAR-T cell tumor infiltration and persistence [ 113 ]. Dual-targeting CAR-T cells targeting IL-13 Rα2 and TGF-β can convert the inhibitory TGF-β signal into an immunostimulatory signal in glioblastoma models, remodeling the TME [ 174 ]. In prostate cancer and glioblastoma, co-expression of a dominant-negative TGF-β receptor II competitively binds TGF-β, inhibits SMAD signaling, and preserves CAR-T cell function in TGF-β environments. This approach exhibits a preclinical safety profile. The dominant-negative TGF-β receptor is a disabled receptor that blocks the TGF-β signaling pathway. As a result, this receptor can prevent TGF-β immunosuppressive effects [ 175 , 176 ]. Anti-TGF-β antibodies can neutralize all TGF-β isoforms (TGF-β1, TGF-β2, TGF-β3) in the TME, preventing downstream signaling and markedly enhancing CAR-T cell infiltration and persistence [ 177 ]. Lane et al. engineered a bifunctional fusion protein designated SySR, which recruits E3 ligases to target SMAD proteins, facilitating the conditional degradation of SMAD2 and SMAD3. CAR-T cells expressing this degrader exhibited resistance to TGF-β suppression and enhanced proliferation and antitumor efficacy both in vitro and in vivo [ 108 ].

Trogocytosis can induce CAR-T cell dysfunction and antigen loss. Combining CD19 CAR-T cells with anti-CD19 antibodies reduces antigen uptake, preserving CAR surface levels and efficacy [ 178 ]. The ATF3–CH25H-25HC axis plays a key role in regulating trogocytosis. CAR-T cells co-expressing CH25H inhibit this process via the production of 25-HC, enhancing persistence and cytotoxicity [ 179 ]. TNF-mediated fratricide also limits CAR-T cell efficacy. In certain acute myeloid leukemia and solid tumor models, the N-terminal fragment of TNF acts as a CAR target. Activated CAR-T cells endogenously express TNF, resulting in fratricide. Genetic ablation of the TNF gene abrogates this phenomenon. This approach synergistically enhances cell survival and function [ 180 ]. The FasL/Fas pathway acts as another negative regulator of persistence. Engineering CAR-T cells to express a dominant-negative Fas receptor can inhibit CAR-T cell apoptosis induced by Fas-L. These engineered CAR-T cells enhance survival and antitumor efficacy [ 86 ]. In T cell acute lymphoblastic leukemia, co-expression of an anti-CD7 CAR with a protein expression blocker effectively mitigates fratricide and augments persistence, with mild toxicities [ 181 ].

Enhancing the persistence of CAR-T cells can be also achieved by reprogramming macrophage polarization in the TME. This strategy involves polarization into the antitumor M1 phenotype and simultaneously inhibiting the differentiation into the immunosuppressive M2 phenotype. A study demonstrates that the co-administration of a folate-targeted TLR7 agonist and CAR-T cells effectively reprograms the TME by driving the polarization of TAMs into an M1 phenotype. This approach enhances the antitumor persistence and efficacy of CAR-T cells without exacerbating toxicity [ 182 ]. To further explore this paradigm, a study demonstrates that CAR-T cells targeting folate receptor β-positive TAMs can induce persistent remodeling of the TME. Mechanistically, this process enables the sustained elimination of M2 macrophages and the promotion of M1 polarization, augmenting the recruitment of endogenous T cells and pro-inflammatory myeloid cells [ 183 ].

Combination immunotherapy

The modulation of immune checkpoints is a critical strategy for optimizing CAR-T cell efficacy. ShRNA-mediated knockdown of PD-1 in BCMA CAR-T cells reduced CAR-T cell exhaustion, enhanced memory T cell formation, and improved antitumor activity both in vitro and in vivo, with a preliminary clinical safety profile and efficacy observed in relapsed or refractory multiple myeloma patients [ 29 ]. In patients with B-cell lymphoma, combining CAR-T cell therapy with the anti-PD-1 antibody pembrolizumab attenuated exhaustion and improved persistence [ 184 ]. A phase I study demonstrated that the efficacy of combining anti-PD-L1 antibody durvalumab with CD19 CAR-T cell therapy is critically dependent on the timing of administration. Infusion prior to CAR-T cells suppresses early CAR-T cell function, while continuous durvalumab after CAR-T cell infusion can enhance CAR-T cell expansion. Importantly, the study found this strategy did not increase the risk of CRS [ 185 ]. Additionally, the deletion of CTLA-4 enhances CD28 signaling and mitigates exhaustion [ 115 ]. Silencing LAG-3 gene augments cytotoxicity and persistence [ 186 ]. Targeting the TIM-3, via CAR-T cells secreting TIM-3 decoy receptors or the small-molecule inhibitor ML-T7 reverses exhaustion and enhances antitumor efficacy and persistence [ 127 , 187 ]. Evidence suggests that TIGIT plays a critical role in CAR-T cell exhaustion in multiple myeloma. However, a study combining antibody blockade, secretion of a blocking scFv, and CRISPR/Cas9-mediated knockout demonstrated that single-target TIGIT inhibition failed to enhance the antitumor efficacy of BCMA CAR-T cells in preclinical models. Thus, further investigation is needed [ 188 ]. Figure 2 illustrates the key immune checkpoints that regulate CAR-T cell function.

The combination of oncolytic viruses (OVs) with CAR-T therapy has emerged as a novel therapeutic strategy. OVs are natural or genetically modified viruses that selectively infect and replicate in tumor cells, lysing them without damaging normal cells. They can also trigger an adaptive immune response targeting the tumor [ 189 ]. OVs can enhance the persistence of CAR-T cells through multiple mechanisms, including direct oncolysis, remodeling the TME, recruiting CAR-T cells via chemokine like CXCL11 or RANTES, upregulating tumor antigens, downregulating immune checkpoints and expressing cytokines like IL-7 or IL-15 [ 190 – 193 ]. A novel oncolytic virus, through encapsulation in tumor cell membrane-derived nanovesicles and bioorthogonal chemical modification, significantly enhances the persistence and antitumor function of CAR-T cells [ 194 ].

Cytokine engineering represents another critical strategy. The CAR-enhancer (CAR-E) platform fuses a target antigen (BCMA) with a low-affinity immunomodulator (mutant IL-2), synergistically enhancing persistence via dual signaling without additional genetic modification [ 195 ]. In glioblastoma and prostate cancer models, combining intratumoral IL-12 delivery with CAR-T cells improved efficacy and persistence [ 196 , 197 ].

CAR-T cells can be combined with radiotherapy or chemotherapy. The former enhances CAR-T cell infiltration, activation and persistence by inducing chemokine gradients and upregulating tumor-associated antigen expression. The latter promotes CAR-T cell recruitment, infiltration, and accumulation by remodeling the tumor stroma, reducing immunosuppressive cells, and creating a proinflammatory environment [ 198 ].

Improving CAR transfection

Current viral-vector transfection is hindered by safety profile and cost constraints, and conventional electroporation is associated with cytotoxicity and functional impairment of CAR-T cells [ 199 ]. Novel strategies are emerging to improve CAR transfection. For example, the VivoVec platform enables the generation of functional CAR-T cells via in vivo T cell transfection, obviating the need for ex vivo manufacturing. This platform utilizes a rapamycin-activated cytokine receptor system as an alternative to lymphodepleting chemotherapy, and the preclinical study has demonstrated a favorable safety profile and antitumor activity [ 200 ].

Among non-viral vectors, LNPs show promise for mRNA delivery in the CAR-T cell therapy by improving transfection efficiency, though transient expression limits long-term persistence [ 201 ]. Bimbo et al. developed a novel DNA-based targeted LNP system containing minicircle DNA and SB100x transposase mRNA, enabling direct in vivo generation of CAR-T cells and bypassing ex vivo processing [ 202 ]. Physical transfection methods also show promise. Photothermal nanomaterial-mediated photoporation enables the efficient delivery of mRNA, siRNA, and proteins [ 203 ]. Electroactive nanoinjection achieves higher CAR expression with lower metabolic impact compared to traditional electroporation [ 204 ].

In the context of cell sorting and processing, microfluidic chips facilitate the efficient enrichment of high-purity CD8 + T cells, improving CAR-T cell manufacturing efficiency and antitumor capability [ 205 ]. Multidimensional double spiral inertial microfluidics effectively enriches activated T cells, increasing transduction efficiency [ 206 ]. Prospectively, sonofluidics may enable efficient, controllable delivery of functional nanomaterials [ 207 ]. Non-viral vector delivery platforms, including polymer-based and lipid-based nanosystems, also offer diverse routes to efficient CAR transfection [ 199 ]. Fig. 3 illustrates these advanced engineering and transfection platforms.

Strategies to mitigating CRS in CAR-T cell therapy

CRS is a major clinical toxicity in CAR-T cell therapy, representing its Achilles’ heel in wider application. It is primarily triggered by the excessive CAR-T cells and their interaction with the host immune system. It is a case where too much of a good thing turns therapeutic potency into a self-destructive storm. Current approaches to reduce the incidence and severity of CRS focus on three main strategies. They are optimizing CAR molecular design, blocking key inflammatory pathways and using novel technology.

Optimization of CAR molecular structure

When designing a CAR, a rational design to modulate the intensity of CAR-T cell activation can mitigate CRS and enhance persistence. For instance, in the extracellular domain, low-affinity antigen-binding domains enable CAR-T cells to maintain antitumor activity with lower toxicity compared to high-affinity antigen-binding domains [ 208 , 209 ]. A novel study demonstrates that non-immune cells (mesenchymal stromal cells, MSCs) engineered to express a deficient signaling CAR containing only the scFv (lacking signaling domains) induce internalization and lysosomal degradation of target-cell surface antigens (CD19). This leads to the depletion of essential signaling molecules for survival and triggers apoptosis. To a degree, this process operates independently of T cell activation and may prevent CRS at its source [ 210 ].

In the area of hinge and transmembrane domain engineering, preclinical deletion of two glycines in the CD8 hinge region reduces severe CRS incidence without compromising antitumor efficacy. This effect is potentially attributed to reduced hinge flexibility [ 211 ].

Regarding intracellular signaling domain optimization, Wang et al. systematically compared ITAMs derived from different CD3 subunits and found that CAR-T cells incorporating CD3ε-derived ITAMs released the lowest level of cytokines. The CD3ε-derived ITAMs can also delay CAR-T cell exhaustion and promote memory phenotype differentiation [ 212 ]. In clinical trials, CARs incorporating the 4-1BB exhibited superior safety profiles compared to those with CD28 [ 213 – 215 ]. The adoptive transfer of a 1:1 mix of CD4 + T cells (4-1BB) and CD8 + T cells (CD28) reduces CRS severity and maintains therapeutic efficacy [ 216 ]. Adjusting the ratio of CD4 + /CD8 + to 1:1 is also regarded as an effective strategy for modulating cytokines and mitigating CRS [ 97 , 217 , 218 ].

For systematic molecular engineering, Ying et al. developed a novel CD19-BBz(86 aa) CAR based on the prototype CD19-BBz (71 aa), featuring a longer extracellular domain (76 aa) and intracellular sequence (7 aa). In a phase I trial, none of the 25 patients experienced CRS exceeding grade 1. In the high-dose cohort (2–4 × 10 8 cells), 6 patients achieved complete response with sustained remission, suggesting this design combines low cytokine release with long-term persistence [ 219 ]. The CD19 CAR-T cell product AUTO1 exhibits reduced scFv affinity for CD19 due to a fast off rate. The AUTO1 attenuates CAR-T cell activation and shows no CRS exceeding grade 3 in clinical use [ 142 , 220 ]. Figure 4 summarizes the evolution of CAR structural optimization, from enhancing persistence to mitigating CRS, and ultimately achieving both simultaneously through various innovative designs.

Targeted intervention strategies for inflammatory pathways

Regarding the management of cytokine release, the IL-6 receptor blocker tocilizumab is a first-line agent for CRS exceeding grade 2 [ 221 ]. Clinical data show that in patients with B-cell acute lymphoblastic leukemia and multiple myeloma, tocilizumab shortens CRS duration without compromising CAR-T cell efficacy [ 222 , 223 ]. Although tocilizumab will elevate the peripheral blood IL-6 level and potentially increase neurotoxicity risk [ 224 ], a report indicates it does not raise the incidence of ICANS [ 225 ]. The antibody received FDA approval for severe CRS [ 226 ]. To reduce the severe CRS incidence, Lin et al. engineered CAR-T cells to express a tocilizumab-derived scFv, enabling autocrine secretion of a protein called Toci. This autocrine delivery suppressed CRS more effectively than exogenous administration [ 227 ]. For CRS cases resistant to tocilizumab, the direct IL-6-neutralizing antibody siltuximab shows therapeutic potential. However, further evidence is needed to evaluate it in pediatric patients [ 228 – 230 ]. Given the central role of IL-6 in CRS, CAR-T cells with IL-6 knockdown by shRNA exhibit reduced autocrine cytokine secretion and lower severe CRS toxicity in the preclinical model [ 231 , 232 ]. Furthermore, IL-6 knockout CAR-T cells have demonstrated safety in clinical trials [ 233 ]. Interestingly, pharmacological modulation of IL-6 synthesis provides another method for intervention. The β-blocker metoprolol suppresses IL-6 translation by regulating eEF2 phosphorylation. It reduces IL-6 levels without compromising antitumor efficacy [ 234 ].

GM-CSF is a critical mediator of CRS, driving inflammation and exacerbating inflammatory damage [ 235 ]. Neutralizing GM-CSF has emerged as a therapeutic strategy [ 236 ]. Lenzilumab, a monoclonal antibody neutralizing human GM-CSF, can attenuate CRS and enhance antitumor efficacy. Similarly, GM-CSF -deficient CAR-T cells engineered via gene knockout exhibit enhanced antitumor efficacy [ 237 ].

IFN-γ, another core cytokine of CRS, can be neutralized by the monoclonal antibody emapalumab. This antibody mitigates CRS in preclinical models and clinical case reports without compromising therapeutic outcomes [ 238 – 241 ]. Compared with tocilizumab, IFN-γ blockade represents a more upstream intervention and is particularly suitable for patients refractory to tocilizumab or with concurrent hemophagocytic lymphohistiocytosis/macrophage-activation syndrome [ 242 ].

TNF-α also contributes to CRS. The anti-TNF-α monoclonal antibodies golimumab and adalimumab mitigate toxicity by clearing CAR-T cells expressing membrane-bound TNF-α. This intervention does not compromise antitumor efficacy [ 209 , 243 ]. Anti-TNF-α antibody therapy represents a promising exogenous strategy to control CRS without compromising antitumor function.

The IL-1 receptor antagonist anakinra serves as a third-line agent and is often given for patients with CRS refractory to high-dose corticosteroids (intravenous dexamethasone 10 mg) and grade 4 CRS [ 221 , 244 ]. The IL-1 signaling stimulates IL-6 production, and anakinra suppresses CRS progression by blocking the IL-1 receptor to reduce IL-6 levels [ 98 ]. However, to prophylactically prevent CRS with anakinra, frequent dosing is required to achieve significant efficacy [ 245 ]. Theodoros et al. engineered CAR-T cells that are capable of continuously secreting an IL-1 receptor antagonist, and can effectively prevent severe CRS toxicity [ 98 ].

Compared with antibody monotherapy, simultaneous neutralization of multiple key inflammatory mediators (IFN-γ, TNF-α, IL-2 and IL-6) has shown superior efficacy against CRS [ 246 ]. Fig. 5 outlines the key inflammatory pathways of CRS and therapeutic interventions.

Targeting key signaling pathways offers a promising strategy to mitigate CRS in CAR-T cell therapy. For instance, kinase inhibitors represent a novel strategy to mitigate CRS toxicity [ 247 ]. JAK inhibitors (itacitinib and ruxolitinib) alleviate CRS by blocking downstream cytokine signaling without impairing CAR-T cell function [ 248 , 249 ]. mTOR inhibitors reduce cytokine release without compromising cytotoxicity. Interestingly, the traditional Chinese medicine herbal remedy tetrandrine regulates secretion of multiple cytokines by inhibiting the AKT and NF-κB pathways [ 247 , 250 ].

Upregulation of CD40L on CAR-T cells stimulates IL-6 release from monocytes, exacerbating CRS [ 98 ]. Ala et al. employed CRISPR-Cas9 to knock out the CD40L and/or CSF2 in CAR-T cells, which reduced monocyte-derived IL-6 release and mitigated CRS [ 251 ].

Atrial natriuretic peptide and α-methyltyrosine can inhibit catecholamine synthesis. This pharmacological inhibition can alleviate CRS in animal models without compromising CAR-T cell efficacy [ 101 ]. Fig. 6 distinguishes the signaling pathways involved in CAR-T cell persistence and CRS.

Novel technological strategies

Inhibiting over-activated CAR-T cells is a critical approach to manage therapeutic toxicity, and novel technologies are being developed for this purpose. Molecular switches and synthetic biology play important roles in novel technologies aimed at reducing the risk of CRS. The molecular switches can regulate CAR-T cells at multiple molecular levels, including DNA transcription, translation, and protein level.

From the DNA transcription level, Miller et al. introduced a photothermal-inducible gene switch into CAR-T cells, enabling these cells to express immunostimulatory molecules only upon localized heating at the tumor site. This design allows expression to be turned on and off on demand and repeatedly activated through photothermal stimulation. The photothermal control achieves both spatial and temporal control, avoiding systemic immune activation and reducing CRS risk [ 252 ]. From the translational level, Liu et al. developed a BOCK-inducible genetic code expansion system. In the absence of the unnatural amino acid N-ε-((tert-butoxy) carbonyl)-l-lysine (BOCK), the CAR protein translation is terminated prematurely, preventing complete translation of the CAR protein and maintaining the system in an OFF state. In the presence of BOCK, BOCK binds to the orthogonal tRNA/aminoacyl-tRNA synthetase pair, enabling full translation of the CAR protein and switching the system to an ON state. This strategy achieves reversible and rapid regulation of CAR-T cell activity, providing an effective approach for mitigating CRS [ 253 ]. Molecular drugs can regulate CAR-T cells at the protein level. The inducible caspase-9 (iCasp9) is a clinically validated suicide gene system that has shown particular promise. The iCasp9 safety switch can effectively mitigate CRS in the preclinical model. iCasp9 allows the rapid elimination of CAR-T cells following rimiducid administration in cases of severe toxicity. Clinical studies have confirmed their safety profile and efficacy [ 254 , 255 ]. Utilizing the FDA-approved antiviral drug grazoprevir (GZV) reversibly controls CAR-T cell activity. In the ON VIPER (Versatile ProteasE Regulatable CARs) CAR, GZV inhibits the self-cleavage activity of the Non-Structural Protein 3 (NS3) protease, preserving CAR integrity and enabling CAR-T cell activation. In the OFF VIPER CAR, GZV competitively binds to the NS3 protease, disrupting the interaction between CAR components and rendering the CAR inactive. This design offers the potential to rapidly reverse CRS and mitigate toxicity [ 256 , 257 ].

Synthetic Notch (synNotch) receptors function as programmable logic gates within CAR-T cells. The system employs an “AND gate” mechanism that necessitates the simultaneous recognition of antigen A and antigen B to achieve highly specific dual-antigen targeting. Alternatively, the system can be designed as an “OR gate” where the detection of either antigen is sufficient to induce CAR expression. This design can effectively address tumor antigen heterogeneity. “IF-THEN gate” is designed such that initial activation by antigen A triggers a secondary verification step through antigen B prior to the initiation of cytotoxic activity. “NOT gate” represents a novel immunotherapeutic strategy that merges high targeting precision with flexible regulatory capabilities [ 258 , 259 ].

Clinical management strategies for CRS

To mitigate the severe toxicity of CRS, mechanistic insights from preclinical CAR-T studies have been successfully translated into clinical practice. This bench-to-bedside approach has established tocilizumab as the cornerstone of CRS management.

For patients with grade 1 CRS, supportive care with symptomatic treatment is the mainstay, including close monitoring of vital signs, antipyretic therapy, and fluid and electrolyte management. Low-risk patients may be followed in an outpatient setting, whereas hospitalization is advised for high-risk patients.

For patients with grade 2 CRS, supportive care remains the primary approach, including antipyretic therapy, intravenous fluid administration for hypotension, and low-flow nasal cannula oxygen for hypoxia. Close monitoring of vital signs and clinical status is essential. Tocilizumab is recommended and administered intravenously at a dose of 8 mg/kg (maximum 800 mg). If no clinical improvement is observed after 8–12 hours, a repeat dose may be given, with a maximum of three total doses. Corticosteroids are generally reserved for patients who progress to higher grades or do not respond adequately to initial management [ 260 , 261 ].

For grade 3 CRS, early administration of tocilizumab and prompt use of corticosteroids are crucial. Tocilizumab is administered intravenously at 8 mg/kg (maximum 800 mg), with the same dosing and repeat schedule as grade 2 CRS. Corticosteroid therapy should be initiated promptly. Dexamethasone 10 mg intravenously every 6 hours is recommended, with taper upon clinical improvement. Patients typically require ICU admission for vasopressor support and high-flow oxygen or facemask oxygen therapy [ 260 , 261 ].

Grade 4 CRS represents a life-threatening condition requiring immediate intensive care unit admission with multiple vasopressor support and/or positive pressure ventilation. First-line therapy consists of tocilizumab (same dosing as grade 3) combined with high-dose corticosteroids (500 mg every 12 hours) [ 262 ]. It is worth noting that corticosteroids inhibit T cell function. They suppress untransduced T cells more strongly than CAR-T cells, and have a greater inhibitory effect on CD4 + T cells than on CD8 + T cells. Therefore, repeated high-dose use should be avoided in clinical practice [ 263 , 264 ].

For corticosteroid-refractory severe CRS, methylprednisolone may be increased to 1,000 mg twice daily. Alternative therapies may be considered, including anakinra, siltuximab, ruxolitinib, emapalumab, dasatinib, blood filtration, plasma exchange, mechanical ventilation, and surgical intervention [ 229 , 260 , 265 – 267 ]. Figure 7 illustrates the stepwise treatment strategy for CRS based on ASTCT grading.

Grade 1: Supportive care (monitoring, antipyretics, fluids and electrolytes). Outpatient for low-risk, hospitalization for high-risk.

Grade 2: Supportive care+early tocilizumab. Corticosteroids reserved for progression.

Grade 3: Tocilizumab+dexamethasone. ICU for vasopressor and oxygen support.

Grade 4: Immediate ICU. Tocilizumab+high-dose methylprednisolone.

Refractory: High-dose corticosteroid. Anakinra, siltuximab, ruxolitinib, emapalumab, dasatinib, blood filtration, plasma exchange, ventilation, surgery.

Achieving enhanced persistence alongside reduced CRS

In CAR-T cell therapy, efficacy and toxicity are often considered a trade-off, creating a major clinical challenge in striking a balance between the two factors [ 107 ]. While the emerging evidence suggests that long-term persistence and CRS are not governed by an absolute trade-off, emerging technological innovations make it possible to maintain CAR-T cell persistence while mitigating CRS severity. This dynamic mirrors the Daoist philosophy of “Yin and Yang”, where seemingly opposing forces are interdependent and regulated to achieve a state of equilibrium.

Strategies that simultaneously address persistence and toxicity provide critical insights for designing safer and more effective next-generation CAR-T cell therapies, and realizing this outcome presents a formidable challenge. The primary challenge in enhancing CAR-T cell persistence while mitigating severe CRS arises from several inherent difficulties. First of all, enhanced cytotoxicity and persistence are intrinsically linked to heightened inflammatory cytokine production through shared intracellular signaling pathways. Moreover, sustained antigen exposure is necessary for long-term persistence, and excessive stimulation leads to CAR-T cell exhaustion and recurrent cytokine storms [ 268 ]. Lastly, in clinical management, interventions for CRS (tocilizumab) can inadvertently suppress CAR-T cell function, creating a direct trade-off between toxicity control and persistence. Despite these hurdles, recent research breakthroughs have identified specific strategies that show promise in enhancing persistence while reducing CRS severity.

In the context of CAR molecule engineering, CAR-T cells co-expressing SMAD7 concurrently inhibit both the TGF-β and NF-κB signaling pathways, effectively counteracting TME-induced exhaustion and reducing the risk of CRS. The CAR-T cells can synergistically improve both efficacy and safety profile in solid tumors [ 67 ]. Multifunctional CAR is one of the synthetic CAR categories, integrating multiple functional modules. A dual-functional chimeric receptor is designed as a CAR consisting of GP130, IL-6 R, and constitutively active IL-7 R with the M452L mutation. This receptor fuses two functional modules, which are the extracellular domain of the IL-6 receptor and the intracellular domain of a constitutively active IL-7 receptor. The former captures and reduces IL-6 to alleviate CRS, while the latter continuously activates JAK-STAT signaling to enhance CAR-T cells proliferation and persistence [ 269 ]. Multifunctional CAR-T cells co-expressing the NKG2D-based multi-targeting CAR along with IL-12 and IFN-α2 can achieve safe and durable antitumor immunity [ 270 ]. Adaptor CAR is a novel CAR design, bridging CAR-T cells and tumor cells via exogenous soluble adaptor molecules. This modular platform enables reversible and precise spatiotemporal control of CAR-T cell activity via adaptor dose adjustment to mitigate CRS. Sustained and sufficient adaptor binding to tumor cells is essential for CAR-T persistence [ 271 , 272 ].

In terms of signaling pathway modulation, dasatinib inhibits Src and LCK phosphorylation. It can block CAR-T cell activation signals, reverse CAR-T cell differentiation and exhaustion, exhibiting promise in combination regimens [ 273 , 274 ]. This tyrosine kinase inhibitor dasatinib suppresses CAR-T cell activation and reduces CRS in mouse models by inhibiting ZAP70 phosphorylation and blocking CAR signaling. Furthermore, its pharmacological switch property can also delay exhaustion [ 148 ]. Clinical case reports demonstrate that dasatinib can reverse grade 3 CRS [ 275 ]. Thus, dasatinib holds potential for simultaneously enhancing CAR‑T persistence and mitigating CRS risk.

Metabolic reprogramming primarily enhances CAR-T cell persistence, and also contributes to safety profile. From a metabolic perspective, Akhtar et al. revealed that repeated tumor stimulation impairs autophagy in CAR-T cells. Combined administration of a GLP-1 receptor agonist and UrA synergistically activates autophagy and mitophagy pathways, effectively clearing damaged mitochondria and promoting memory T cell formation. CAR-T cells subjected to this metabolic reprogramming exhibited persistence and reduced severe CRS risk in preclinical models. This offers a crucial strategy for designing next-generation CAR-T cell therapies that combine high persistence with low toxicity [ 166 ].

From the perspective of synthetic biology, the “IF-THEN” logic-gated design represents a promising strategy to simultaneously enhance the persistence and safety of CAR-T cells [ 276 – 278 ]. Jambon et al. developed an “IF-THEN” logic-gated SynNotch CAR‑T strategy. In this design, full antitumor activation is dependent on the simultaneous recognition of both CD33 and CD123 antigens. This design effectively eliminated AML tumor cells while sparing normal hematopoietic stem cells. Furthermore, the approach attenuated CAR-T cell exhaustion, promoted the memory phenotype, and prevented severe CRS [ 277 ]. Zhu et al. engineered a CD19-SynNotch-gated PDbody-CAR T cell system. Employing “IF-THEN” logic, the expression of a PD-L1-targeting PDbody-CAR is induced only on the recognition of CD19 on tumor cells by SynNotch, achieving precise cytotoxicity. Animal models demonstrated that this strategy effectively enhanced both T cell persistence and mitigates off-target toxicity, including CRS [ 278 ]. Beyond “IF-THEN” logic-gated strategies, CAR-T cells that specifically secrete extracellular matrix (ECM)-degrading enzymes locally within the tumor can simultaneously enhance both persistence and safety [ 279 ]. The engineering of cytokine expression under the control of a synNotch receptor enables tumor localized, autocrine secretion. This secretion is independent of CAR or TCR signaling, establishing another synthetic design with enhanced persistence and safety [ 280 ].

Artificial intelligence (AI) is systematically driving innovation in CAR-T cell therapy across two key dimensions: (1) CAR intelligent manufacturing, (2) leveraging big data for predictive and discovery tasks. One key dimension is CAR intelligent manufacturing and delivery, where AI enhances both the persistence and safety profile of CAR-T cells by enabling the analysis of complex, high-dimensional datasets [ 281 ]. Within this dimension, three advances stand out. First, AI can employ deep learning to predict tumor neoantigens and construct high-affinity scFv through protein design tools. This accelerates and optimizes the entire process from target screening to CAR design, holding the potential to overcome the bottlenecks of solid tumor therapy and drive treatment toward a personalized, persistent, and safe direction [ 282 , 283 ]. AlphaFold2 combined with other computational methods, enables the accurate prediction and screening of high-affinity scFv [ 284 ]. Second, AI can help mitigate two major bottlenecks in CAR construction, namely tonic signaling imbalance and cell exhaustion. The AI-driven tool CAR-Toner finely predicts and modulates tonic signaling intensity, addressing poor persistence from insufficient tonic signaling and exhaustion from excessive tonic signaling [ 285 ]. The predictive model CARMSeD identifies CAR constructs with low risk of spontaneous activation, mitigating toxicity from aberrant signaling. When combined with a PROTAC-based module targeting AKT3 for degradation, this strategy significantly improves CAR-T cell persistence in vivo [ 286 ]. Third, the emerging framework of smart CAR-T cell nano-symbionts leverages AI to optimize nanocarrier design, dynamically modulate CAR-T cell function, and monitor potential toxicity in real time [ 287 ]. Another key dimension is leveraging big data for predictive and discovery tasks, where AI has demonstrated unique strengths. For example, AI can predict early relapse risk within six months based solely on patient age and six routine laboratory indicators, facilitating the identification of high-risk patients and the planning of tailored interventions [ 288 ]. It can also integrate multi-omics data with large language models to forecast therapeutic responses and adverse events [ 289 ]. Furthermore, through deep learning-powered virtual screening, AI can efficiently identify optimal hit compounds. For example, BC1618 was screened from 45 candidate molecules and identified as a potent mitophagy agonist [ 290 ]. We believe that these synergistic AI-integrated approaches will become new methods for optimizing the persistence and safety profile of CAR-T cell therapy in the near future.

We posit that the enhancement of CAR-T cell persistence need not come at the cost of exacerbating CRS. Drawing inspiration from the Daoist philosophical framework of “Yin and Yang”, Yang symbolizes enhancing persistence and Yin represents mitigating CRS. We propose that these factors are not inherently antagonistic. Instead, they are interdependent components of a unified system. By employing integrated strategies, including CAR structure optimization, intracellular signaling pathways modulation, and the emerging technologies application, it is possible to engineer CAR-T cells that attain a therapeutic “Yin-Yang harmony” state. This approach offers a promising pathway to advance CAR-T cell therapy toward sustained persistence and superior safety profile.

Conclusion

This review highlights that insufficient persistence and CRS are two major clinical limitations. The cellular signaling networks are highly interconnected. As a Chinese proverb goes, “Know yourself and know your enemy, and you will win a hundred battles.” The signal mechanism in CAR-T cells is the foundation of CAR-T cell design. And our goal is to make the next-generation CAR-T cell persistent and safe.

Future efforts to improve persistence should aim to engineer a kind of intermediate CAR-T cells between memory and terminal effector cells that combines robust proliferative potential, sustained cytotoxicity, and metabolic adaptability. This phenotype holds the potential to exhibit the in vivo longevity of CAR-T cells while mitigating the risk of severe CRS. Furthermore, CAR-T cell therapy is also likely to evolve toward personalized design, tailoring the selection of co-stimulatory domains, signaling components, and ex vivo culture protocols to individual patients based on disease type and the T cell status in vivo.

Future strategies for toxicity management should prioritize reversibility and prevention. The next-generation safe systems could enable temporary pharmacological pausing of CAR-T cells during toxic episodes. This pausing would allow function to be restored upon drug withdrawal, offering reversible control. Regarding prevention, while tocilizumab reduces CRS severity, its cost and side effects remain significant concerns. Prognostic scoring systems such as EASIX, s-EASIX and modified EASIX, have shown promise in predicting the severity of CRS and ICANS. Thus prognostic scoring systems facilitate early intervention and personalized treatment planning in adult, adolescent and paediatric populations. However, validation in larger cohorts is required to confirm the systems’ accuracy [ 291 , 292 ]. Furthermore, the absence of real-time cytokines monitoring compels clinicians to rely on indirect markers (ferritin, C-reactive protein and lymphocyte counts) to assess CAR-T cell activity. This monitoring gap hinders precise regulation and can lead to inaccuracies in toxicity reporting. The development of advanced CAR-T cell-tracking technologies is urgently needed to enhance CRS control [ 15 ].

Despite CAR-T cell therapy is limited by two challenges, CAR-T cell therapy still remains a transformative strategy for treating refractory hematological malignancies and promising frontier for tumors. Continued research and precise modulation of signaling networks are essential to ensure this therapy broader application. Ultimately, solidifying the role of CAR-T cell therapy in the future of oncology depends on engineering cells with long-term persistence and safety profile, a modern quest not like threading the needle between Scylla and Charybdis.

For the limitations of the study, this review proposes a framework of “Yin and Yang” that categorizes existing strategies into those that enhance CAR-T persistence (Yang) and those that mitigate CRS (Yin). However, it should be noted that our analysis is primarily based on currently known, clinically approved, or late-stage preclinical experiments. Therefore, this framework largely reflects the Yin-Yang harmony state achieved through existing technologies. Within this scope, three important areas are beyond this review. Several emerging physical delivery technologies, such as Viscoelastic Mechanoporation, have been reported for use in native T cells but have not been applied to the CAR-T cells [ 293 ]. No evidence demonstrates whether or how these technologies could achieve a Yin-Yang harmony state. Additionally, CAR-NK cell therapy, as a potential alternative to CAR-T cell therapy, has an advantage of not inducing CRS, but its major challenge is limited persistence. This combination of low toxicity and limited persistence presents an alternative approach to achieving a Yin-Yang harmony state [ 294 ]. However, CAR-NK cells are different from CAR-T cells in terms of signaling and cell type, and this review focuses on CAR-T cells. Therefore, CAR-NK cell therapy is not discussed. Lastly, our review is constrained by a mixing of animal models and preclinical experiments, rather than focusing solely on clinical trials. This mixed composition may result in some conclusion bias. These three areas will be important directions for achieving a Yin-Yang harmony state in the future.

Method

In this study, an initial search was conducted using the core terms “CAR-T cell”, “pathway”, “persistence”, and “Cytokine Release Syndrome” across PubMed database, targeting the abstract, title, and keyword fields. The search period was restricted to January 2020 through March 2026, and the majority of retrieved literature was published from 2023 onward. If necessary, we may also cite a small number of high-quality publications prior to 2020 that provide essential background or foundational insights. To build a comprehensive understanding of signaling pathways, foundational literature on CAR-T cell signal transduction was initially reviewed. This material serves strictly to introduce CAR-T cell signaling. Subsequently, a focused search regarding the “CAR-T cell signaling pathway” was performed. Our research team established the clear screening criteria. We need the original research explicitly elucidating the unique signaling mechanisms of CAR-T cells and the relationship between signaling mechanisms and persistence or CRS. While CAR signaling shares mechanistic features with native T cell receptor signaling, the core aim of this review is to systematically dissect the specific signaling network shaped by engineered CAR structures. Thus, studies addressing only native T cell signaling without a connection to CAR design or function were excluded. The screening also excluded non-English publications and reports with incomplete data, such as the studies focusing solely on TCR-T cell therapy without CAR-T cell relevance.

The presentation of signaling pathways is structured to progress from antigen recognition to functional cellular output. First of all, we detail the roles of different CAR domains, followed by the downstream core activated pathways. By focusing on “CAR-T cell”, “persistence”, and “cytokine release syndrome”, we summarize how interventions targeting these pathways differentially modulate cellular persistence and treatment toxicity. To visually represent the intricate signaling networks and optimization strategies outlined, schematic diagrams were generated. These figures elucidate the regulatory mechanisms spanning from CAR structure to downstream functional output. The figures in this article were created using the Generic Diagramming Platform for Biomedical Graphics (BioGDP) [ 295 ].

Moreover, our study proposes an integrative perspective by framing insufficient persistence and CRS as dual clinical manifestations of imbalanced intracellular signaling networks. In this context, we emphasize multi-target strategies that synergistically enhance persistence while mitigating toxicity, painting a blueprint for the rational design of next-generation CAR-T cell therapies.

Acknowledgements

Not applicable.

Abbreviation

25-Hydroxycholesterol

Antigen Negative

Antigen Positive

Artificial Intelligence

Protein Kinase B

Adjunctive LAT-Activating CAR

Acute Lymphoblastic Leukemia

Acute Myeloid Leukemia

Activator Protein 1

American Society for Transplantation and Cellular Therapy

Activating Transcription Factor 3

CD80

CD86

B-cell Acute Lymphoblastic Leukemia

B-Cell Maturation Antigen

Bcl-2-Interacting Mediator of cell death

N-ε-((tert-butoxy)carbonyl)-L-lysine

Chimeric Adaptor Protein

CAR-Enhancer

CAR Tyrosine Kinase

CD40 Ligand

Cholesterol 25-Hydroxylase

Cytokine Release Syndrome

Colony Stimulating Factor 2

Cytotoxic T-Lymphocyte Antigen 4

C-X-C Chemokine Receptor Type 4

Diacylglycerol

Dominant-Negative TGF-β Receptor II

Dual Specificity Phosphatase 2

Dual Specificity Phosphatase 4

Endothelial Activation and Stress Index

Extracellular Matrix

Extracellular Signal-Regulated Kinase

Fas-Associated Protein with Death Domain

Fatty Acid Oxidation

Fas Ligand

Food and Drug Administration

Forkhead Box O1

Forkhead Box P3

Galectin-9

Disialoganglioside 2

Glucagon-Like Peptide-1

Glucose Transporter

Granulocyte-Macrophage Colony-Stimulating Factor

Glycoprotein 130

Glypican 3

Granzyme A

Grazoprevir

Histone Deacetylase

Herpes Simplex Virus Thymidine Kinase

Intercellular Adhesion Molecule 1

Immune Effector Cell-Associated Neurotoxicity Syndrome

Inducible Caspase 9

Interferon Gamma

Interleukin-13 Receptor Alpha 2

Inositol Triphosphate

Immunoreceptor Tyrosine-Based Activation Motif

Janus Kinase

Lymphocyte Activation Gene 3

Linker for Activation of T cells

Lymphocyte Function-Associated Antigen 1

Lipid Nanoparticle

Low-Density Lipoprotein Receptor-Related Protein

Macrophage phenotype classification

Mitogen-Activated Protein Kinase

Modified EASIX

Major Histocompatibility Complex

Mixed Lineage Kinase Domain-Like Pseudokinase

Mammalian Target of Rapamycin

Nuclear Factor of Activated T cells

Nuclear Factor Kappa B

NF-κB-Inducing Kinase

Natural Killer

Non-Structural Protein 3

Oxidative Phosphorylation

Programmed Death-1

Programmed Death-Ligand 1

Phosphatidylinositol 3-Kinase

Phosphatidylinositol 4,5-Bisphosphate

Phosphatidylinositol 3,4,5-Trisphosphate

Protein Kinase C Theta

Phospholipase C Gamma 1

Proteolysis Targeting Chimera

Receptor-Interacting Serine/Threonine Kinase

Receptor-Interacting Serine/Threonine Kinase 3

Single-Chain Variable Fragment

Simplified EASIX

Short Hairpin RNA

Soluble Interleukin-6 Receptor

SH2 Domain-Containing Leukocyte-Specific Phosphoprotein of 76 kDa

Suppressor of Mother against Decapentaplegic

Signal Transducer and Activator of Transcription

Synthetic Notch

T cell Factor

Lymphoid Enhancer Factor

T Cell Receptor

Transforming Growth Factor β

T cell Immunoreceptor with Ig and ITIM domains

T cell Immunoglobulin and Mucin-domain containing-3

Toll-Like Receptor 7

Tumor Microenvironment

TNF Receptor

Tumor Necrosis Factor Alpha

Thymocyte Selection-Associated High Mobility Group Box Protein

TNF Receptor-Associated Factor

TNF-Related Apoptosis-Inducing Ligand

TRAIL Receptor

Urolithin A

Vascular Endothelial Growth Factor Receptor

Variable Heavy

Variable Light

Zeta-Chain Associated Protein Kinase of 70 kDa

Author contributions

Yunlin Yang: Conceptualization, Methodology, Writing-original draft, Visualization, Writing-Reviewing and Editing. Yuanjie Sun: Supervision, Validation and Writing-original draft. Haobo Kang: Methodology and Writing-original draft. Enqi Guan: Writing-original draft, Visualization, Writing-Reviewing and Editing. Rui Liu: Investigation, Funding acquisition and Project administration. Kun Yang: Project administration and Supervision. Dongbo Jiang: Conceptualization, Project administration, Funding acquisition, Supervision, Writing-Reviewing and Editing.

Funding

The present study was supported by the National Natural Science Foundation of China (No.82203510 to DJ), the Xi’an Medical Key Project (No.24YXYJ0005 to DJ), and the Youth Promotion Project of Xi’an (No.959202313100 to DJ).

Data availability

Not applicable.

Declarations

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Competing interests

The authors have declared no conflict of interest.

Footnotes

Publisher’s Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Yulin Yang and Yuanjie Sun contributed equally to this work.

Contributor Information

Rui Liu, Email: tdliurui@fmmu.edu.cn, Email: 289741154@qq.com, Email: speciallr@163.com.

Kun Yang, Email: yangkunkun@fmmu.edu.cn.

Dongbo Jiang, Email: superjames1991@foxmail.com, Email: superjames1991@fmmu.edu.cn.

References

1. Brudno JN, Kochenderfer JN. Current understanding and management of CAR T cell-associated toxicities. Nat Rev Clin Oncol. 2024;21(7):501–21. 10.1038/s41571-024-00903-0.

2. Asmamaw Dejenie T, Tiruneh G/Medhin M, Dessie Terefe G, Tadele Admasu F, Wale Tesega WChekol Abebe E. Chekol Abebe E: current updates on generations, approvals, and clinical trials of CAR T-cell therapy. Hum Vaccines Immunotherapeutics. 2022;18(6):2114254. 10.1080/21645515.2022.2114254.

3. Roddie C, Sandhu KS, Tholouli E, Logan AC, Shaughnessy P, Barba P, et al. Obecabtagene autoleucel in adults with B-Cell acute lymphoblastic leukemia. N Engl J Med. 2024;391(23):2219–30. 10.1056/NEJMoa2406526.

4. Abreu TR, Fonseca NA, Gon N, Moreira JON. Current challenges and emerging opportunities of CAR-T cell therapies. J Control Release. 2020;319:246–61. 10.1016/j.jconrel.2019.12.047.

5. Jayaraman J, Mellody MP, Hou AJ, Desai RP, Fung AW, Pham AHT, et al. CAR-T design: elements and their synergistic function. EBioMedicine. 2020;58:102931. 10.1016/j.ebiom.2020.102931.

6. Xie Y, Zhou K, Li L, Liang A, Mei H, Zou D, et al. Phase 2 study of Relmacabtagene Autoleucel (CD19 CAR-T) for Relapsed/Refractory mantle cell lymphoma in Chinese adults. Blood Adv. 2026;10(4):1134–44. 10.1182/bloodadvances.2024015763.

7. Schultz LM, Jeyakumar N, Kramer AM, Sahaf B, Srinagesh H, Shiraz P, et al. CD22 CAR T cells demonstrate high response rates and safety in pediatric and adult B-ALL: phase 1b results. Leukemia. 2024;38(5):963–68. 10.1038/s41375-024-02220-y.

8. Brudno JN, Natrakul DA, Karrs J, Patel N, Maass-Moreno R, Ahlman MA, et al. Transient responses and significant toxicities of anti-CD30 CAR T cells for CD30+ lymphomas: results of a phase 1 trial. Blood Adv. 2024;8(3):802–14. 10.1182/bloodadvances.2023011470.

9. Qi C, Liu C, Gong J, Liu D, Wang X, Zhang P, et al. Claudin18.2-specific CAR T cells in gastrointestinal cancers: phase 1 trial final results. Nat Med. 2024;30(8):2224–34. 10.1038/s41591-024-03037-z.

10. Mao Y, Liao Q, Zhu Y, Bi M, Zou J, Zheng N, et al. Efficacy and safety of novel multifunctional M10 CAR-T cells in HIV-1-infected patients: a phase I, multicenter, single-arm, open-label study. Cell Discov. 2024;10(1):49. 10.1038/s41421-024-00658-z.

11. Wang X, Wu X, Tan B, Zhu L, Zhang Y, Lin L, et al. Allogeneic CD19-targeted CAR-T therapy in patients with severe myositis and systemic sclerosis. Cell. 2024;187(18):4890–904.e9. 10.1016/j.cell.2024.06.027.

12. Pecher A, Hensen L, Klein R, Schairer R, Lutz K, Atar D, et al. CD19-targeting CAR T cells for myositis and interstitial lung disease associated with antisynthetase syndrome. JAMA. 2023;329(24):2154. 10.1001/jama.2023.8753.

13. Jin G, Liu Y, Wang L, He Z, Zhao X, Ma Y, et al. A single infusion of engineered long-lived and multifunctional T cells confers durable remission of asthma in mice. Nat Immunol. 2024;25(6):1059–72. 10.1038/s41590-024-01834-9.

14. Lambrecht BN, Hammad H. CAR T cells put the brakes on asthma. Nat Immunol. 2024;25(6):935–37. 10.1038/s41590-024-01851-8.

15. Rejeski K, Hill JA, Dahiya S, Jain MD. Noncanonical and mortality-defining toxicities of CAR T cell therapy. Nat Med. 2025;31(7):2132–46. 10.1038/s41591-025-03813-5.

16. Huang Y, Wang H. Tonic signaling in CAR-T therapy: the lever long enough to move the planet. Front Med. 2025;19(3):391–408. 10.1007/s11684-025-1130-x.

17. Andreu-Saumell I, Rodriguez-Garcia A, Mühlgrabner V, Gimenez-Alejandre M, Marzal B, Castellsagué J, et al. CAR affinity modulates the sensitivity of CAR-T cells to PD-1/PD-L1-mediated inhibition. Nat Commun. 2024;15(1):3552. 10.1038/s41467-024-47799-z.

18. Kouro T, Himuro H, Sasada T. Exhaustion of CAR T cells: potential causes and solutions. J Transl Med. 2022;20(1):239. 10.1186/s12967-022-03442-3.

19. Eggert J, Au-Yeung BB. Functional heterogeneity and adaptation of naive T cells in response to tonic TCR signals. Curr Opin Immunol. 2021;73:43–49. 10.1016/j.coi.2021.09.007.

20. Leick MB, Silva H, Scarfò I, Larson R, Choi BD, Bouffard AA, et al. Non-cleavable hinge enhances avidity and expansion of CAR-T cells for acute myeloid leukemia. Cancer Cell. 2022;40(5):494–508.e5. 10.1016/j.ccell.2022.04.001.

21. Zhang C, Liu T, Li S, Teng X, Zhu Y, Zhang G, et al. Membrane-bound IL-7 immobilized by the CD8 transmembrane region improves efficacy of CD19 CAR-T cell therapy. Mol Cancer. 2024;23(1):236. 10.1186/s12943-024-02154-0.

22. Barbera S, Schuiling MJA, Sanjaya NA, Pietilä I, Sarén T, Essand M, et al. Trogocytosis of chimeric antigen receptors between T cells is regulated by their transmembrane domains. Sci Immunol. 2025;10: eado 2054.103). 10.1126/sciimmunol.ado2054.

23. Larson RC, Maus MV. Recent advances and discoveries in the mechanisms and functions of CAR T cells. Nat Rev Cancer. 2021;21(3):145–61. 10.1038/s41568-020-00323-z.

24. Olejniczak SH, Lotze MT, Skokos D. Second signals for cancer immunotherapy. J Immunother Cancer. 2025;13(9):13. 10.1136/jitc-2024-010530.

25. Lotze MT, Olejniczak SH, Skokos D. CD28 co-stimulation: novel insights and applications in cancer immunotherapy. Nat Rev Immunol. 2024;24(12):878–95. 10.1038/s41577-024-01061-1.

26. Roselli E, Boucher JC, Li G, Kotani H, Spitler K, Reid K, et al. 4-1BB and optimized CD28 co-stimulation enhances function of human mono-specific and bi-specific third-generation CAR T cells. J Immunother Cancer. 2021;9(10):9. 10.1136/jitc-2021-003354.

27. Suryadevara CM, Desai R, Farber SH, Choi BD, Swartz AM, Shen SH, et al. Preventing lck activation in CAR T cells confers Treg resistance but requires 4-1BB signaling for them to persist and treat solid tumors in nonlymphodepleted hosts. Clin Cancer Res. 2019;25(1):358–68. 10.1158/1078-0432.CCR-18-1211.

28. Krawczyk M, Drużyńska M, Bednarska E, Winiarska M. The race between 4-1BB- and CD28-based CD19 CAR-T products in the therapy of B-cell malignancies. Biochim et Biophys Acta (BBA) - Rev Cancer. 2026;1881(1):189519. 10.1016/j.bbcan.2025.189519.

29. Ouyang W, Jin S, Xu N, Liu W, Zhao H, Zhang L, et al. PD-1 downregulation enhances CAR-T cell antitumor efficiency by preserving a cell memory phenotype and reducing exhaustion. J Immunother Cancer. 2024;12(4):12. 10.1136/jitc-2023-008429.

30. Honikel MM, Olejniczak SH. Co-stimulatory receptor signaling in CAR-T cells. Biomolecules. 2022;12(9):12. 10.3390/biom12091303.

31. Watts TH, Yeung KKM, Yu T, Lee S, Eshraghisamani R. TNF/TNFR superfamily members in costimulation of T cell responses—revisited. Annu Rev Immunol. 2025;43(1):113–42. 10.1146/annurev-immunol-082423-040557.

32. Guo M, Zhang Y, Yuan X, Wang X. Innovative approaches to CAR-T cell therapy: the role of lipid metabolism pathways. J Transl Med. 2025;23(1):670. 10.1186/s12967-025-06718-6.

33. Alizadeh D, Wong RA, Yang X, Wang D, Pecoraro JR, Kuo C, et al. IL15 enhances CAR-T cell antitumor activity by reducing mTORC1 activity and preserving their stem cell memory phenotype. Cancer Immunol Res. 2019;7(5):759–72. 10.1158/2326-6066.CIR-18-0466.

34. Tang L, Pan S, Wei X, Xu X, Wei Q. Arming CAR-T cells with cytokines and more: innovations in the fourth-generation CAR-T development. Mol Ther. 2023;31(11):3146–62. 10.1016/j.ymthe.2023.09.021.

35. Smole A, Benton A, Poussin MA, Eiva MA, Mezzanotte C, Camisa B, et al. Expression of inducible factors reprograms CAR-T cells for enhanced function and safety. Cancer Cell. 2022;40(12):1470–87.e7. 10.1016/j.ccell.2022.11.006.

36. Rabbani SA, El-Tanani M, El-Tanani Y, Kumar R, Sharma S, Khan MA, et al. Advances in adoptive cell therapies in Cancer: from mechanistic breakthroughs to clinical frontiers and Overcoming barriers. Med Sci. 2025;13(3):13. 10.3390/medsci13030190.

37. Kagoya Y, Tanaka S, Guo T, Anczurowski M, Wang C, Saso K, et al. A novel chimeric antigen receptor containing a JAK–STAT signaling domain mediates superior antitumor effects. Nat Med. 2018;24(3):352–59. 10.1038/nm.4478.

38. Bui TA, Mei H, Sang R, Ortega DG, Deng W. Advancements and challenges in developing in vivo CAR T cell therapies for cancer treatment. EBioMedicine. 2024;106:105266. 10.1016/j.ebiom.2024.105266.

39. Xiao Y, Wu B, Zhou Q, Wu F, Li N, Xu C, et al. Technologies for chimeric antigen receptor transgene delivery. Trends Mol Med. 2025;31(11):1047–64. 10.1016/j.molmed.2025.05.011.

40. Short L, Holt RA, Cullis PR, Evgin L. Direct in vivo CAR T cell engineering. Trends Pharmacological Sci. 2024;45(5):406–18. 10.1016/j.tips.2024.03.004.

41. Zhu Q, Li J, Liu N, Han L, Wu Z, Wang Y, et al. CD2 augmentation enhances CAR-T-cell efficacy via immunological synapse remodeling and T-cell exhaustion mitigation. Cell Mol Immunol. 2025;22(8):935–48. 10.1038/s41423-025-01314-6.

42. Nam G, Yeon HR, Park HB, Chang H, Kim JH, Cho B, et al. CD99-mediated immunological synapse formation potentiates CAR-T cell function. Nat Commun. 2025;16(1):7987. 10.1038/s41467-025-63184-w.

43. Chao Z, Mei Q, Yang C, Luo J, Liu P, Peng H, et al. Immunological synapse: structures, molecular mechanisms and therapeutic implications in disease. Sig Transduct Target Ther. 2025;10(1):254. 10.1038/s41392-025-02332-6.

44. Huang W, Lin W, Chen B, Zhang J, Gao P, Fan Y, et al. NFAT and NF-κB dynamically co-regulate TCR and CAR signaling responses in human T cells. Cell Rep. 2023;42(7):112663. 10.1016/j.celrep.2023.112663.

45. Ai K, Liu B, Chen X, Huang C, Yang L, Zhang W, et al. Optimizing CAR-T cell therapy for solid tumors: current challenges and potential strategies. J Hematol Oncol. 2024;17(1):105. 10.1186/s13045-024-01625-7.

46. Nai Y, Du L, Shen M, Li T, Huang J, Han X, et al. TRAIL-R1-Targeted CAR-T cells exhibit dual antitumor efficacy. Front Mol Biosci. 2021;8:756599. 10.3389/fmolb.2021.756599.

47. Montalvo MJ, Bandey IN, Rezvan A, Wu K, Saeedi A, Kulkarni R, et al. Decoding the mechanisms of chimeric antigen receptor (CAR) T cell-mediated killing of tumors: insights from granzyme and Fas inhibition. Cell Death Dis. 2024;15(2):109. 10.1038/s41419-024-06461-8.

48. Wu L, Brzostek J, Sakthi Vale PD, Wei Q, Koh CKT, Ong JXH, et al. CD28-CAR-T cell activation through FYN kinase signaling rather than LCK enhances therapeutic performance. Cell Rep. Med. 2023;4(2):100917. 10.1016/j.xcrm.2023.100917.

49. Majzner RG, Rietberg SP, Sotillo E, Dong R, Vachharajani VT, Labanieh L, et al. Tuning the antigen density requirement for CAR T-cell activity. Cancer Discov. 2020;10(5):702–23. 10.1158/2159-8290.CD-19-0945.

50. Tousley AM, Rotiroti MC, Labanieh L, Rysavy LW, Kim W, Lareau C, et al. Co-opting signalling molecules enables logic-gated control of CAR T cells. Nature. 2023;615(7952):507–16. 10.1038/s41586-023-05778-2.

51. Dong R, Libby KA, Blaeschke F, Fuchs W, Marson A, Vale RD, et al. Rewired signaling network in T cells expressing the chimeric antigen receptor (CAR). Embo J. 2020;39(16):e104730. 10.15252/embj.2020104730.

52. Meng X, Wu X, Zheng Y, Shang K, Jing R, Jiao P, et al. Exploiting Ca2+ signaling in T cells to advance cancer immunotherapy. Semin Immunol. 2020;49:101434. 10.1016/j.smim.2020.101434.

53. Salter AI, Ivey RG, Kennedy JJ, Voillet V, Rajan A, Alderman EJ, et al. Phosphoproteomic analysis of chimeric antigen receptor signaling reveals kinetic and quantitative differences that affect cell function. Sci Signal. 2018;11(544):11. 10.1126/scisignal.aat6753.

54. Ajmal I, Farooq MA, Duan Y, Yao J, Gao Y, Hui X, et al. Intrinsic ADRB2 inhibition improves CAR-T cell therapy efficacy against prostate cancer. Mol Ther. 2024;32(10):3539–57. 10.1016/j.ymthe.2024.08.028.

55. Harwood SJ, Smith CR, Lawson JD, Ketcham JM. Selected approaches to disrupting protein–protein interactions within the MAPK/RAS pathway. IJMS. 2023;24(8):24. 10.3390/ijms24087373.

56. Pham-Danis C, Novak AJ, Danis E, McClellan SM, Leach L, Yarnell MC, et al. Restoration of LAT activity improves CAR T cell sensitivity and persistence in response to antigen-low acute lymphoblastic leukemia. Cancer Cell. 2025;43(3):482–502.e9. 10.1016/j.ccell.2025.02.008.

57. Gaud G, Lesourne R, Love PE. Regulatory mechanisms in T cell receptor signalling. Nat Rev Immunol. 2018;18(8):485–97. 10.1038/s41577-018-0020-8.

58. Shah K, Al-Haidari A, Sun J, Kazi JU. T cell receptor (TCR) signaling in health and disease. Sig Transduct Target Ther. 2021;6(1):412. 10.1038/s41392-021-00823-w.

59. Gao Y, Lin H, Guo D, Cheng S, Zhou Y, Zhang L, et al. Suppression of 4.1R enhances the potency of NKG2D-CAR T cells against pancreatic carcinoma via activating ERK signaling pathway. Oncogenesis. 2021;10(9):62. 10.1038/s41389-021-00353-8.

60. Adabi E, Charitidis FT, Thalheimer FB, Guaza-Lasheras M, Clarke C, Buchholz CJ. Enhanced conversion of T cells into CAR T cells by modulation of the MAPK/ERK pathway. Cell Rep. Med. 2025;6(2):101970. 10.1016/j.xcrm.2025.101970.

61. Rohrs JA, Siegler EL, Wang P, Finley SD. ERK activation in CAR T cells is amplified by CD28-mediated increase in CD3ζ phosphorylation. iScience. 2020;23(4):101023. 10.1016/j.isci.2020.101023.

62. Martinez GJ, Pereira RM, Äijö T, Kim EY, Marangoni F, Pipkin ME, et al. The transcription factor NFAT promotes exhaustion of activated CD8+ T cells. Immunity. 2015;42(2):265–78. 10.1016/j.immuni.2015.01.006.

63. Seo H, Chen J, González-Avalos E, Samaniego-Castruita D, Das A, Wang YH, et al. TOX and TOX2 transcription factors cooperate with NR4A transcription factors to impose CD8 + T cell exhaustion. Proc Natl Acad Sci USA. 2019;116(25):12410–15. 10.1073/pnas.1905675116.

64. Mulvey A, Trueb L, Coukos G, Arber C. Novel strategies to manage CAR-T cell toxicity. Nat Rev Drug Discov. 2025;24(5):379–97. 10.1038/s41573-024-01100-5.

65. Rotiroti MC, Tousley AM, Chu H, Herrera-Barrera M, Manousopoulou A, Kim W, et al. Engineering T cells with a membrane-tethered version of SLP-76 overcomes antigen-low resistance to CAR T cell therapy. Nat Cancer. 2025;6(12):1940–54. 10.1038/s43018-025-01056-4.

66. Philipson BI, O’Connor RS, May MJ, June CH, Albelda SM, Milone MC. 4-1BB costimulation promotes CAR T cell survival through noncanonical NF-κB signaling. Sci Signal. 2020;13(625):13. 10.1126/scisignal.aay8248.

67. Liang S, Zheng R, Zuo B, Li J, Wang Y, Han Y, et al. SMAD7 expression in CAR-T cells improves persistence and safety for solid tumors. Cell Mol Immunol. 2024;21(3):213–26. 10.1038/s41423-023-01120-y.

68. Dou Z, Bonacci TR, Shou P, Landoni E, Woodcock MG, Sun C, et al. 4-1BB-encoding CAR causes cell death via sequestration of the ubiquitin-modifying enzyme A20. Cell Mol Immunol. 2024;21(8):905–17. 10.1038/s41423-024-01198-y.

69. Lim M, Jung S, Koh C, Jeong H, Moon Y, Seo H. The LXRβ/NF-κB axis reprograms CAR-T cells to resist exhaustion in the tumor microenvironment. Oncoimmunology. 2026;15(1):2611615. 10.1080/2162402X.2025.2611615.

70. Liu L, Cheng Y, Zhang F, Chen J, Tian P, Shi W, et al. IL-2/GM-CSF enhances CXCR3 expression in CAR-T cells via the PI3K/AKT and ERK1/2 pathways. J Cancer ResearchClin Oncol. 2023;149(9):5547–57. 10.1007/s00432-022-04509-w.

71. Glaviano A, Foo ASC, Lam HY, Yap KCH, Jacot W, Jones RH, et al. PI3K/AKT/mTOR signaling transduction pathway and targeted therapies in cancer. Mol Cancer. 2023;22(1):138. 10.1186/s12943-023-01827-6.

72. Ramapriyan R, Vykunta VS, Vandecandelaere G, Richardson LGK, Sun J, Curry WT, et al. Altered cancer metabolism and implications for next-generation CAR T-cell therapies. Pharmacol Ther. 2024;259:108667. 10.1016/j.pharmthera.2024.108667.

73. Urak R, Walter M, Lim L, Wong CW, Budde LE, Thomas S, et al. Ex vivo Akt inhibition promotes the generation of potent CD19CAR T cells for adoptive immunotherapy. J Immunother Cancer. 2017;5(1):26. 10.1186/s40425-017-0227-4.

74. Zheng W, Jones LL, Geiger TL. Modulation of PI3K signaling to improve CAR T cell function. Oncotarget. 2018;9(88):35807–08. 10.18632/oncotarget.26334.

75. Zhang Q, Ding J, Sun S, Liu H, Lu M, Wei X, et al. Akt inhibition at the initial stage of CAR-T preparation enhances the CAR-positive expression rate, memory phenotype and in vivo efficacy. Am J Cancer Res. 2019;9(11):2379–96.

76. Stock S, Übelhart R, Schubert M, Fan F, He B, Hoffmann J, et al. Idelalisib for optimized CD19-specific chimeric antigen receptor T cells in chronic lymphocytic leukemia patients. Intl J Cancer. 2019;145(5):1312–24. 10.1002/ijc.32201.

77. Funk CR, Wang S, Chen KZ, Waller A, Sharma A, Edgar CL, et al. PI3Kδ/γ inhibition promotes human CART cell epigenetic and metabolic reprogramming to enhance antitumor cytotoxicity. Blood. 2022;139(4):523–37. 10.1182/blood.2021011597.

78. Nian Z, Zheng X, Dou Y, Du X, Zhou L, Fu B, et al. Rapamycin pretreatment rescues the bone Marrow AML cell elimination capacity of CAR-T cells. Clin Cancer Res. 2021;27(21):6026–38. 10.1158/1078-0432.CCR-21-0452.

79. Liu J, Xiao Q, Xiao J, Niu C, Li Y, Zhang X, et al. Wnt/β-catenin signalling: function, biological mechanisms, and therapeutic opportunities. Sig Transduct Target Ther. 2022;7(1):3. 10.1038/s41392-021-00762-6.

80. Song P, Gao Z, Bao Y, Chen L, Huang Y, Liu Y, et al. Wnt/β-catenin signaling pathway in carcinogenesis and cancer therapy. J Hematol Oncol. 2024;17(1):46. 10.1186/s13045-024-01563-4.

81. Zhu M, Han Y, Gu T, Wang R, Si X, Kong D, et al. Class I HDAC inhibitors enhance antitumor efficacy and persistence of CAR-T cells by activation of the Wnt pathway. Cell Rep. 2024;43(4):114065. 10.1016/j.celrep.2024.114065.

82. Wu S, Wei Y, Qiu Y, Ai K, Chen M, Wang H, et al. Inhibition of VEGF signaling prevents exhaustion and enhances anti-leukemia efficacy of CAR-T cells via Wnt/β-catenin pathway. J Transl Med. 2025;23(1):494. 10.1186/s12967-024-05907-z.

83. Upadhyay R, Boiarsky JA, Pantsulaia G, Svensson-Arvelund J, Lin MJ, Wroblewska A, et al. A critical role for Fas-mediated off-target tumor killing in T-cell immunotherapy. Cancer Discov. 2021;11(3):599–613. 10.1158/2159-8290.CD-20-0756.

84. Dufva O, Koski J, Maliniemi P, Ianevski A, Klievink J, Leitner J, et al. Integrated drug profiling and CRISPR screening identify essential pathways for CAR T-cell cytotoxicity. Blood. 2020;135(9):597–609. 10.1182/blood.2019002121.

85. Xia L, Zheng Z, Liu J, Chen Y, Ding J, Xia N, et al. EGFR-targeted CAR-T cells are potent and specific in suppressing triple-negative breast cancer both in vitro and in vivo. Clin Trans Imm. 2020;9(5):e 1135. 10.1002/cti2.1135.

86. Yi F, Cohen T, Zimmerman N, Dündar F, Zumbo P, Eltilib R, et al. CAR-engineered lymphocyte persistence is governed by a FAS ligand–FAS autoregulatory circuit. Nat Cancer. 2025;6(10):1638–55. 10.1038/s43018-025-01009-x.

87. Bajgain P, Torres Chavez AG, Balasubramanian K, Fleckenstein L, Lulla P, Heslop HE, et al. Secreted Fas decoys enhance the antitumor activity of engineered and bystander T cells in Fas Ligand–expressing solid tumors. Cancer Immunol Res. 2022;10(11):1370–85. 10.1158/2326-6066.CIR-22-0115.

88. Datlinger P, Pankevich EV, Arnold CD, Pranckevicius N, Lin J, Romanovskaia D, et al. Systematic discovery of CRISPR-boosted CAR T cell immunotherapies. Nature. 2025;646(8086):963–72. 10.1038/s41586-025-09507-9.

89. Thapa B, Kc R, Uludağ H. TRAIL therapy and prospective developments for cancer treatment. J Control Release. 2020;326:335–49. 10.1016/j.jconrel.2020.07.013.

90. Kobayashi E, Kishi H, Ozawa T, Hamana H, Nakagawa H, Jin A, et al. A chimeric antigen receptor for TRAIL-receptor 1 induces apoptosis in various types of tumor cells. Biochem Bioph Res Co. 2014;453(4):798–803. 10.1016/j.bbrc.2014.10.024.

91. Czaplicka A, Lachota M, Pączek L, Zagożdżon R, Kaleta B. Chimeric antigen receptor T cell therapy for pancreatic Cancer: a review of Current evidence. Cells. 2024;13(1):13. 10.3390/cells13010101.

92. Liu Y, Fang Y, Chen X, Wang Z, Liang X, Zhang T, et al. Gasdermin E–mediated target cell pyroptosis by CAR T cells triggers cytokine release syndrome. Sci Immunol. 2020;5(43):5. 10.1126/sciimmunol.aax7969.

93. Boulch M, Cazaux M, Cuffel A, Guerin MV, Garcia Z, Alonso R, et al. Tumor-intrinsic sensitivity to the pro-apoptotic effects of IFN-γ is a major determinant of CD4+ CAR T-cell antitumor activity. Nat Cancer. 2023;4(7):968–83. 10.1038/s43018-023-00570-7.

94. Bailey SR, Takei HN, Escobar G, Kann MC, Bouffard AA, Kienka T, et al. IFN-γ–resistant CD28 CAR T cells demonstrate increased survival, efficacy, and durability in multiple murine tumor models. Sci Transl Med. 2025;17(801):eadp 8166. 10.1126/scitranslmed.adp8166.

95. Lee DW, Santomasso BD, Locke FL, Ghobadi A, Turtle CJ, Brudno JN, et al. ASTCT consensus grading for cytokine release syndrome and neurologic toxicity associated with immune effector cells. Biol Blood Marrow Transplant. 2019;25(4):625–38. 10.1016/j.bbmt.2018.12.758.

96. Wang M, Chen S, He X, Yuan Y, Wei X. Targeting inflammation as cancer therapy. J Hematol Oncol. 2024;17(1):13. 10.1186/s13045-024-01528-7.

97. Boulch M, Cazaux M, Cuffel A, Ruggiu M, Allain V, Corre B, et al. A major role for CD4+ T cells in driving cytokine release syndrome during CAR T cell therapy. Cell Rep. Med. 2023;4(9):101161. 10.1016/j.xcrm.2023.101161.

98. Giavridis T, van der Stegen SJC, Eyquem J, Hamieh M, Piersigilli A, Sadelain M. CAR T cell–induced cytokine release syndrome is mediated by macrophages and abated by IL-1 blockade. Nat Med. 2018;24(6):731–38. 10.1038/s41591-018-0041-7.

99. Norelli M, Camisa B, Barbiera G, Falcone L, Purevdorj A, Genua M, et al. Monocyte-derived IL-1 and IL-6 are differentially required for cytokine-release syndrome and neurotoxicity due to CAR T cells. Nat Med. 2018;24(6):739–48. 10.1038/s41591-018-0036-4.

100. Dibas A, Rhiel M, Patel VB, Andrieux G, Boerries M, Cornu TI, et al. Cell-based models of ‘Cytokine release syndrome’ endorse CD40L and Granulocyte–macrophage Colony-Stimulating factor knockout in chimeric antigen receptor T cells as mitigation strategy. Cells. 2023;12(21):12. 10.3390/cells12212581.

101. Staedtke V, Bai R, Kim K, Darvas M, Davila ML, Riggins GJ, et al. Disruption of a self-amplifying catecholamine loop reduces cytokine release syndrome. Nature. 2018;564(7735):273–77. 10.1038/s41586-018-0774-y.

102. Guo H, Li R, Wang M, Hou Y, Liu S, Peng T, et al. Multiomics analysis identifies SOCS1 as restraining T cell activation and Preventing graft-versus-host disease. Adv Sci. 2022;9(21):e2200978. 10.1002/advs.202200978.

103. Aliyu M, Zohora FT, Anka AU, Ali K, Maleknia S, Saffarioun M, et al. Interleukin-6 cytokine: an overview of the immune regulation, immune dysregulation, and therapeutic approach. Int Immunopharmacol. 2022;111:109130. 10.1016/j.intimp.2022.109130.

104. Jiang Z, Liao R, Lv J, Li S, Zheng D, Qin L, et al. IL-6 trans-signaling promotes the expansion and anti-tumor activity of CAR T cells. Leukemia. 2021;35(5):1380–91. 10.1038/s41375-020-01085-1.

105. Morris EC, Neelapu SS, Giavridis T, Sadelain M. Cytokine release syndrome and associated neurotoxicity in cancer immunotherapy. Nat Rev Immunol. 2022;22(2):85–96. 10.1038/s41577-021-00547-6.

106. Wang X, He X, Zhang T, Liu J, Zhao M. Latest updates on pathogenesis mechanisms and management strategies for cytokine release syndrome, neurotoxicity, and hemophagocytic lymphohistiocytosis related to CAR-T cell therapies. Ann Hematol. 2025;104(6):3129–51. 10.1007/s00277-025-06467-y.

107. Chohan KL, Siegler EL, Kenderian SS. CAR-T cell therapy: the efficacy and toxicity balance. Curr Hematol Malig Rep. 2023;18(2):9–18. 10.1007/s11899-023-00687-7.

108. Lane IC, Kembuan G, Carreiro J, Kann MC, Lin W, Bouffard AA, et al. Genetic retargeting of E3 ligases to enhance CAR T cell therapy. Cell Chem Biol. 2024;31(2):338–48.e5. 10.1016/j.chembiol.2023.10.024.

109. Tian H, Liu C, Yu J, Han J, Du J, Liang S, et al. PHF14 enhances DNA methylation of SMAD7 gene to promote TGF-β-driven lung adenocarcinoma metastasis. Cell Discov. 2023;9(1):41. 10.1038/s41421-023-00528-0.

110. Deng Z, Fan T, Xiao C, Tian H, Zheng Y, Li C, et al. TGF-β signaling in health, disease and therapeutics. Sig Transduct Target Ther. 2024;9(1):61. 10.1038/s41392-024-01764-w.

111. Zhao H, Yang F, Yang J, Yang S. SMAD signaling in cancer: integrative roles in tumor progression, immune evasion, and therapeutic resistance. Cytokine. 2026;198:157090. 10.1016/j.cyto.2025.157090.

112. Wan R, Fu B, Fu X, Liu Z, Simayi N, Fu Y, et al. SMAD4 regulates the expression of LCK affecting chimeric antigen receptor-T cells proliferation through PI3K/Akt signaling pathway. J Cellular Physiol. 2025;240(1):e31520. 10.1002/jcp.31520.

113. Tang N, Cheng C, Zhang X, Qiao M, Li N, Mu W, et al. TGF-β inhibition via CRISPR promotes the long-term efficacy of CAR T cells against solid tumors. JCI Insight. 2020;5(4):e133977. 10.1172/jci.insight.133977.

114. Rafiq S, Yeku OO, Jackson HJ, Purdon TJ, van Leeuwen DG, Drakes DJ, et al. Targeted delivery of a PD-1-blocking scFv by CAR-T cells enhances anti-tumor efficacy in vivo. Nat Biotechnol. 2018;36(9):847–56. 10.1038/nbt.4195.

115. Agarwal S, Aznar MA, Rech AJ, Good CR, Kuramitsu S, Da T, et al. Deletion of the inhibitory co-receptor CTLA-4 enhances and invigorates chimeric antigen receptor T cells. Immunity. 2023;56(10):2388–407.e9. 10.1016/j.immuni.2023.09.001.

116. Li F, Zhang Z, Xuan Y, Zhang D, Liu J, Li A, et al. PD-1 abrogates the prolonged persistence of CD8+ CAR-T cells with 4-1BB co-stimulation. Sig Transduct Target Ther. 2020;5(1):164. 10.1038/s41392-020-00277-6.

117. Sharma P, Goswami S, Raychaudhuri D, Siddiqui BA, Singh P, Nagarajan A, et al. Immune checkpoint therapy—current perspectives and future directions. Cell. 2023;186(8):1652–69. 10.1016/j.cell.2023.03.006.

118. Aggarwal V, Workman CJ, Vignali DAA. LAG-3 as the third checkpoint inhibitor. Nat Immunol. 2023;24(9):1415–22. 10.1038/s41590-023-01569-z.

119. Leitner J, Aigner-Radakovics K, Steinberger P. LAG -3—an incompletely understood target in cancer therapy. FASEB J. 2024;38(22):e70190. 10.1096/fj.202401639R.

120. Nguyen TT, Ho P, Staudt S, Gregoire C, Ziegler-Martin K, Jassin M, et al. Fine tuning towards the next generation of engineered T cells. Nat Biomed Eng. 2025;9(10):1610–31. 10.1038/s41551-025-01492-8.

121. Chauvin J, Zarour HM. TIGIT in cancer immunotherapy. J Immunother Cancer. 2020;8(2):8. 10.1136/jitc-2020-000957.

122. Jiang VC, Hao D, Jain P, Li Y, Cai Q, Yao Y, et al. TIGIT is the central player in T-cell suppression associated with CAR T-cell relapse in mantle cell lymphoma. Mol Cancer. 2022;21(1):185. 10.1186/s12943-022-01655-0.

123. Qiao Y, Chen J, Wang X, Yan S, Tan J, Xia B, et al. Enhancement of CAR-T cell activity against cholangiocarcinoma by simultaneous knockdown of six inhibitory membrane proteins. Cancer Commun (Lond). 2023;43(7):788–807. 10.1002/cac2.12452.

124. Cui H, Hamad M, Elkord E. TIGIT in cancer: from mechanism of action to promising immunotherapeutic strategies. Cell Death Dis. 2025;16(1):664. 10.1038/s41419-025-07984-4.

125. Joller N, Anderson AC, Kuchroo VK. LAG-3, TIM-3, and TIGIT: distinct functions in immune regulation. Immunity. 2024;57(2):206–22. 10.1016/j.immuni.2024.01.010.

126. Blaeschke F, Ortner E, Stenger D, Mahdawi J, Apfelbeck A, Habjan N, et al. Design and evaluation of TIM-3-CD28 checkpoint fusion proteins to improve anti-CD19 CAR T-Cell function. Front Immunol. 2022;13:845499. 10.3389/fimmu.2022.845499.

127. Ma S, Tian Y, Peng J, Chen C, Peng X, Zhao F, et al. Identification of a small-molecule tim-3 inhibitor to potentiate T cell–mediated antitumor immunotherapy in preclinical mouse models. Sci Transl Med. 2023;15(722):eadg 6752. 10.1126/scitranslmed.adg6752.

128. Ramezani F, Panahi Meymandi AR, Akbari B, Tamtaji OR, Mirzaei H, Brown CE, et al. Outsmarting trogocytosis to boost CAR NK/T cell therapy. Mol Cancer. 2023;22(1):183. 10.1186/s12943-023-01894-9.

129. Hamieh M, Dobrin A, Cabriolu A, van der Stegen SJC, Giavridis T, Mansilla-Soto J, et al. CAR T cell trogocytosis and cooperative killing regulate tumour antigen escape. Nature. 2019;568(7750):112–16. 10.1038/s41586-019-1054-1.

130. Zhai Y, Du Y, Li G, Yu M, Hu H, Pan C, et al. Trogocytosis of CAR molecule regulates CAR-T cell dysfunction and tumor antigen escape. Sig Transduct Target Ther. 2023;8(1):457. 10.1038/s41392-023-01708-w.

131. Escobar G, Berger TR, Maus MV. CAR-T cells in solid tumors: challenges and breakthroughs. Cell Rep. Med. 2025;6(11):102353. 10.1016/j.xcrm.2025.102353.

132. Monje M, Mahdi J, Majzner R, Yeom KW, Schultz LM, Richards RM, et al. Intravenous and intracranial GD2-CAR T cells for H3K27M+ diffuse midline gliomas. Nature. 2025;637(8046):708–15. 10.1038/s41586-024-08171-9.

133. Driscoll DF, Bistrian BR. Cytokine release syndrome and CAR T cell therapy: modulating the intensity of the inflammatory response and resolution within the tumor microenvironment. Front Pharmacol. 2025;16:1615526. 10.3389/fphar.2025.1615526.

134. Mantovani A, Marchesi F, Malesci A, Laghi L, Allavena P. Tumour-associated macrophages as treatment targets in oncology. Nat Rev Clin Oncol. 2017;14(7):399–416. 10.1038/nrclinonc.2016.217.

135. Gao J, Liang Y, Wang L. Shaping polarization of tumor-associated macrophages in Cancer immunotherapy. Front Immunol. 2022;13:888713. 10.3389/fimmu.2022.888713.

136. Goala P, Zhang Y, Beatty N, Pavy A, McSain S, Sailer C, et al. IFN-γ–driven skewing towards Th1 over Th17 differentiation underlies CRS and neutropenia in CAR-T therapy. J Clin Investigation. 2026;136(1):136. 10.1172/JCI194631.

137. Zhu N, Chen S, Jin Y, Wang M, Fang L, Xue L, et al. Enhancing glioblastoma immunotherapy with integrated chimeric antigen receptor T cells through the Re-education of tumor-associated microglia and macrophages. ACS Nano. 2024;18(17):11165–82. 10.1021/acsnano.4c00050.

138. Zhao L, Yan F, Tang D, Li W, An N, Ren C, et al. The transition between M1 and M2 macrophage phenotypes is associated with the disease status following CD19 CAR-T therapy for B cell lymphoma/leukemia. Cell Death Dis. 2025;16(1):275. 10.1038/s41419-025-07610-3.

139. Hadjigeorgiou AG, Munn LL, Stylianopoulos T, Jain RK. Physiologically based pharmacokinetic model for CAR-T cell delivery and efficacy in solid tumors. Proc Natl Acad Sci USA. 2025;122(46):e1772333174. 10.1073/pnas.2522634122.

140. Chen J, Qiu S, Li W, Wang K, Zhang Y, Yang H, et al. Tuning charge density of chimeric antigen receptor optimizes tonic signaling and CAR-T cell fitness. Cell Res. 2023;33(5):341–54. 10.1038/s41422-023-00789-0.

141. Zanetti SR, Velasco-Hernandez T, Gutierrez-Agüera F, Díaz VM, Romecín PA, Roca-Ho H, et al. A novel and efficient tandem CD19- and CD22-directed CAR for B cell ALL. Mol Ther. 2022;30(2):550–63. 10.1016/j.ymthe.2021.08.033.

142. Ghorashian S, Lucchini G, Richardson R, Nguyen K, Terris C, Guvenel A, et al. CD19/CD22 targeting with cotransduced CAR T cells to prevent antigen-negative relapse after CAR T-cell therapy for B-cell ALL. Blood. 2024;143(2):118–23. 10.1182/blood.2023020621.

143. Ash SL, Orha R, Mole H, Dinesh-Kumar M, Lee SP, Turrell FK, et al. Targeting the activated microenvironment with endosialin (CD248)-directed CAR-T cells ablates perivascular cells to impair tumor growth and metastasis. J Immunother Cancer. 2024;12(2):e 8608. 10.1136/jitc-2023-008608.

144. Lanitis E, Rota G, Kosti P, Ronet C, Spill A, Seijo B, et al. Optimized gene engineering of murine CAR-T cells reveals the beneficial effects of IL-15 coexpression. J Exp Med. 2021;218(2):218. 10.1084/jem.20192203.

145. Steffin D, Ghatwai N, Montalbano A, Rathi P, Courtney AN, Arnett AB, et al. Interleukin-15-armoured GPC3 CAR T cells for patients with solid cancers. Nature. 2025;637(8047):940–46. 10.1038/s41586-024-08261-8.

146. Prinzing B, Schreiner P, Bell M, Fan Y, Krenciute G, Gottschalk S. MyD88/CD40 signaling retains CAR T cells in a less differentiated state. JCI Insight. 2020;5(21):e136093. 10.1172/jci.insight.136093.

147. Maynard LH, Cavanaugh EJ, Zhu H, Starke CE, Doherty SM, Einhaus TK, et al. Pooled CAR-T screening in nonhuman primates identifies designs with enhanced proliferation, trafficking, and persistence. Blood J. 2025;2025028683.

148. Mestermann K, Giavridis T, Weber J, Rydzek J, Frenz S, Nerreter T, et al. The tyrosine kinase inhibitor dasatinib acts as a pharmacologic on/off switch for CAR T cells. Sci Transl Med. 2019;11(499):eaau 5907. 10.1126/scitranslmed.aau5907.

149. Balagopalan L, Moreno T, Qin H, Angeles BC, Kondo T, Yi J, et al. Generation of antitumor chimeric antigen receptors incorporating T cell signaling motifs. Sci Signal. 2024;17(846):eadp 8569. 10.1126/scisignal.adp8569.

150. Wang X, Tao X, Chen P, Jiang P, Li W, Chang H, et al. MEK inhibition prevents CAR-T cell exhaustion and differentiation via downregulation of c-Fos and JunB. Sig Transduct Target Ther. 2024;9(1):293. 10.1038/s41392-024-01986-y.

151. Gargett T, Truong NTH, Gardam B, Yu W, Ebert LM, Johnson A, et al. Safety and biological outcomes following a phase 1 trial of GD2-specific CAR-T cells in patients with GD2-positive metastatic melanoma and other solid cancers. J Immunother Cancer. 2024;12(5):12. 10.1136/jitc-2023-008659.

152. Seo H, González-Avalos E, Zhang W, Ramchandani P, Yang C, Lio C-WJ, et al. BATF and IRF4 cooperate to counter exhaustion in tumor-infiltrating CAR T cells. Nat Immunol. 2021;22(8):983–95. 10.1038/s41590-021-00964-8.

153. Van der Vreken A, Vanderkerken K, De Bruyne E, De Veirman K, Breckpot K, Menu E. Fueling CARs: metabolic strategies to enhance CAR T-cell therapy. Exp Hematol Oncol. 2024;13(1):66. 10.1186/s40164-024-00535-1.

154. Wang MR, Mu W, Zhen A, Kitchen SG. CRISPR/Cas strategies to enhance CAR T-cell function and persistence via metabolic reprogramming. Trends Biotechnol. 2026. 10.1016/j.tibtech.2025.12.001.

155. Hu W, Li F, Liang Y, Liu S, Wang S, Shen C, et al. Glut3 overexpression improves environmental glucose uptake and antitumor efficacy of CAR-T cells in solid tumors. J Immunother Cancer. 2025;13(1):13. 10.1136/jitc-2024-010540.

156. Guerrero JA, Klysz DD, Chen Y, Malipatlolla M, Lone J, Fowler C, et al. GLUT1 overexpression in CAR-T cells induces metabolic reprogramming and enhances potency. Nat Commun. 2024;15(1):8658. 10.1038/s41467-024-52666-y.

157. Shi Y, Kotchetkov IS, Dobrin A, Hanina SA, Rajasekhar VK, Healey JH, et al. GLUT1 overexpression enhances CAR T cell metabolic fitness and anti-tumor efficacy. Mol Ther. 2024;32(7):2393–405. 10.1016/j.ymthe.2024.05.006.

158. Sun R, Liu Y, Sun Y, Zhou M, Wang Y, Shi B, et al. GPC3-targeted CAR-T cells expressing GLUT1 or AGK exhibit enhanced antitumor activity against hepatocellular carcinoma. Acta Pharmacol Sin. 2024;45(9):1937–50. 10.1038/s41401-024-01287-8.

159. Schild T, Wallisch P, Zhao Y, Wang Y, Haughton L, Chirayil R, et al. Metabolic engineering to facilitate anti-tumor immunity. Cancer Cell. 2025;43(3):552–62.e9. 10.1016/j.ccell.2025.02.004.

160. Xia J, Jin J, Dai S, Fan H, Chen K, Li J, et al. Mitophagy: a key regulator of radiotherapy resistance in the tumor immune microenvironment. Mol Aspects Med. 2025;105:101385. 10.1016/j.mam.2025.101385.

161. Huang Y, Shao M, Teng X, Si X, Wu L, Jiang P, et al. Inhibition of CD38 enzymatic activity enhances CAR-T cell immune-therapeutic efficacy by repressing glycolytic metabolism. Cell Rep. Med. 2024;5(2):101400. 10.1016/j.xcrm.2024.101400.

162. Xu K, Yin N, Peng M, Stamatiades EG, Shyu A, Li P, et al. Glycolysis fuels phosphoinositide 3-kinase signaling to bolster T cell immunity. Science. 2021;371(6527):405–10. 10.1126/science.abb2683.

163. Chan JD, Scheffler CM, Munoz I, Sek K, Lee JN, Huang Y, et al. FOXO1 enhances CAR T cell stemness, metabolic fitness and efficacy. Nature. 2024;629(8010):201–10. 10.1038/s41586-024-07242-1.

164. Doan AE, Mueller KP, Chen AY, Rouin GT, Chen Y, Daniel B, et al. FOXO1 is a master regulator of memory programming in CAR T cells. Nature. 2024;629(8010):211–18. 10.1038/s41586-024-07300-8.

165. Denk D, Petrocelli V, Conche C, Drachsler M, Ziegler PK, Braun A, et al. Expansion of T memory stem cells with superior anti-tumor immunity by Urolithin A-induced mitophagy. Immunity. 2022;55(11):2059–73.e8. 10.1016/j.immuni.2022.09.014.

166. Akhtar A, Shakir M, Ansari MS, Divya D, Faizan MI, Chauhan V, et al. Bioengineering the metabolic network of CAR T cells with GLP-1 and Urolithin a increases persistence and long-term anti-tumor activity. Cell Rep. Med. 2025;6(3):102021. 10.1016/j.xcrm.2025.102021.

167. You J, Yang X, Zhao J, Chen H, Tang Y, Ouyang D, et al. Enhancing CAR-T cell metabolic fitness and memory phenotype for improved efficacy against hepatocellular carcinoma. Int J Biol Sci. 2025;21(9):4231–51. 10.7150/ijbs.110406.

168. Liu S, Guruprasad P, Ramasubramanian R, Madhu B, Paruzzo L, Han K, et al. β-hydroxybutyrate enhances the metabolic fitness of CAR T cells in cancer. Cell. 2026;189(6):1701–16.e24. 10.1016/j.cell.2026.02.004.

169. Si X, Shao M, Teng X, Huang Y, Meng Y, Wu L, et al. Mitochondrial isocitrate dehydrogenase impedes CAR T cell function by restraining antioxidant metabolism and histone acetylation. Cell Metab. 2024;36(1):176–92.e10. 10.1016/j.cmet.2023.12.010.

170. Lontos K, Wang Y, Joshi SK, Frisch AT, Watson MJ, Kumar A, et al. Metabolic reprogramming via an engineered PGC-1α improves human chimeric antigen receptor T-cell therapy against solid tumors. J Immunother Cancer. 2023;11(3):11. 10.1136/jitc-2022-006522.

171. McPhedran SJ, Carleton GA, Lum JJ. Metabolic engineering for optimized CAR-T cell therapy. Nat Metab. 2024;6(3):396–408. 10.1038/s42255-024-00976-2.

172. Uribe-Herranz M, Oliver-Caldés A, Martínez-Micaelo N, Español-Rego M, Val-Casals M, Martínez-Soler R, et al. Microbiota shape metabolic and immune determinants of CAR-T therapy and correlate with outcomes in myeloma. Blood Cancer Discov. 2025;6(5):484–504. 10.1158/2643-3230.BCD-24-0203.

173. Marcos-Kovandzic L, Avagliano M, Ben Khelil M, Srikanthan J, Abdallah R, Petrocelli V, et al. Gut Microbiota modulation through Akkermansia spp. Supplementation increases CAR T-cell potency. Cancer Discov. 2025;15(9):1905–26. 10.1158/2159-8290.CD-24-1230.

174. Hou AJ, Shih RM, Uy BR, Shafer A, Chang ZL, Comin-Anduix B, et al. IL-13Rα2/TGF-β bispecific CAR-T cells counter TGF-β -mediated immune suppression and potentiate anti-tumor responses in glioblastoma. Neuro-Oncology. 2024;26(10):1850–66. 10.1093/neuonc/noae126.

175. Zanvit P, van Dyk D, Fazenbaker C, McGlinchey K, Luo W, Pezold JM, et al. Antitumor activity of AZD0754, a dnTgfβrii-armored, STEAP2-targeted CAR-T cell therapy, in prostate cancer. J Clin Investigation. 2023;133(22):133. 10.1172/JCI169655.

176. Li N, Rodriguez JL, Yin Y, Logun MT, Zhang L, Yu S, et al. Armored bicistronic CAR T cells with dominant-negative TGF-β receptor II to overcome resistance in glioblastoma. Mol Ther. 2024;32(10):3522–38. 10.1016/j.ymthe.2024.07.020.

177. Zhao Z, Cutmore LC, Baleeiro RB, Hartlebury JJ, Brown N, Chard-Dunmall L, et al. The combination of oncolytic virus and antibody blockade of TGF-β enhances the efficacy of αvβ6-targeting CAR T cells against pancreatic Cancer in an immunocompetent model. Cancers (Basel). 2025;17(9):17. 10.3390/cancers17091534.

178. Koh SK, Kim H, Han B, Jo H, Doh J, Park J, et al. Anti-CD19 antibody cotreatment enhances serial killing activity of anti-CD19 CAR-T/-NK cells and reduces trogocytosis. Blood. 2025;145(9):956–69. 10.1182/blood.2024025673.

179. Lu Z, McBrearty N, Chen J, Tomar VS, Zhang H, De Rosa G, et al. ATF3 and CH25H regulate effector trogocytosis and anti-tumor activities of endogenous and immunotherapeutic cytotoxic T lymphocytes. Cell Metab. 2022;34(9):1342–58.e7. 10.1016/j.cmet.2022.08.007.

180. Nakashima T, Ouchida T, Ishikawa Y, Ito Y, Hayakawa T, Yoshikawa T, et al. JAK-STAT-activated, fratricide-resistant CAR-T cells targeting membrane-bound TNF effectively treat AML and solid tumors. J Immunother Cancer. 2025;13(7):13. 10.1136/jitc-2024-011067.

181. Oh BLZ, Shimasaki N, Coustan-Smith E, Chan E, Poon L, Lee SHR, et al. Fratricide-resistant CD7-CAR T cells in T-ALL. Nat Med. 2024;30(12):3687–96. 10.1038/s41591-024-03228-8.

182. Luo W, Napoleon JV, Zhang F, Lee YG, Wang B, Putt KS, et al. Repolarization of tumor-infiltrating myeloid cells for augmentation of CAR T cell therapies. Front Immunol. 2022;13:816761. 10.3389/fimmu.2022.816761.

183. Rodriguez-Garcia A, Lynn RC, Poussin M, Eiva MA, Shaw LC, O’Connor RS, et al. CAR-T cell-mediated depletion of immunosuppressive tumor-associated macrophages promotes endogenous antitumor immunity and augments adoptive immunotherapy. Nat Commun. 2021;12(1):877. 10.1038/s41467-021-20893-2.

184. Chong EA, Alanio C, Svoboda J, Nasta SD, Landsburg DJ, Lacey SF, et al. Pembrolizumab for B-cell lymphomas relapsing after or refractory to CD19-directed CAR T-cell therapy. Blood. 2022;139(7):1026–38. 10.1182/blood.2021012634.

185. Hirayama AV, Kimble EL, Wright JH, Fiorenza S, Gauthier J, Voutsinas JM, et al. Timing of anti–PD-L1 antibody initiation affects efficacy/toxicity of CD19 CAR T-cell therapy for large B-cell lymphoma. Blood Adv. 2024;8(2):453–67. 10.1182/bloodadvances.2023011287.

186. Roman Azcona MS, Monaco G, Whitehead M, Kaufmann MM, Alzubi J, Cathomen T, et al. Sustained and specific multiplexed immune checkpoint modulation in CAR T cells induced by targeted epigenome editing. Mol Ther Nucleic Acids. 2025;36(3):102618. 10.1016/j.omtn.2025.102618.

187. Falgàs A, Lázaro-Gorines R, Zanetti SR, Rubio-Pérez L, Martínez-Moreno A, Vinyoles M, et al. A TIM-3–fc decoy secreted by engineered T cells improves CD19 CAR T-cell therapy in B-cell acute lymphoblastic leukemia. Blood. 2025;145(22):2599–613. 10.1182/blood.2024025440.

188. Oliver-Caldes A, Mañe Pujol J, Battram AM, Perez-Amill L, Bachiller M, Calderon H, et al. TIGIT blockade in the context of BCMA-CART cell therapy does not augment efficacy in a multiple myeloma mouse model. Oncoimmunology. 2025;14(1):2529632. 10.1080/2162402X.2025.2529632.

189. Ma R, Li Z, Chiocca EA, Caligiuri MA, Yu J. The emerging field of oncolytic virus-based cancer immunotherapy. Trends Cancer. 2023;9(2):122–39. 10.1016/j.trecan.2022.10.003.

190. Ponterio E, Haas TL, De Maria R. Oncolytic virus and CAR-T cell therapy in solid tumors. Front Immunol. 2024;15:1455163. 10.3389/fimmu.2024.1455163.

191. Tang Y, E Q, Ma L, Li X, Zhang Y, Wang W, et al. Oncolytic virus OVV-03 enhances CAR-T cell therapy against glioblastoma via immune modulation and specific HER2 upregulation. Oncoimmunology. 2026;15(1):2612377. 10.1080/2162402X.2025.2612377.

192. Wang G, Zhang Z, Zhong K, Wang Z, Yang N, Tang X, et al. CXCL11-armed oncolytic adenoviruses enhance CAR-T cell therapeutic efficacy and reprogram tumor microenvironment in glioblastoma. Mol Ther. 2023;31(1):134–53. 10.1016/j.ymthe.2022.08.021.

193. Xiao D, Zhang H, Liu Y, Li Y, Li G, Ning Y. Oncolytic viruses: advanced strategies in cancer therapy. Sig Transduct Target Ther. 2026;11(1):45. 10.1038/s41392-025-02343-3.

194. Huang G, He Y, Chen X, Yin T, Ma A, Zhu L, et al. Bioorthogonal oncolytic-virus nanovesicles combined bio-immunotherapy with CAR-T cells for solid tumors. Biomater Sci. 2025;13(2):457–65. 10.1039/D4BM01305K.

195. Rakhshandehroo T, Mantri SR, Moravej H, Louis BBV, Salehi Farid A, Munaretto L, et al. A CAR enhancer increases the activity and persistence of CAR T cells. Nat Biotechnol. 2025;43(6):948–59. 10.1038/s41587-024-02339-4.

196. Agliardi G, Liuzzi AR, Hotblack A, De Feo D, Núñez N, Stowe CL, et al. Intratumoral IL-12 delivery empowers CAR-T cell immunotherapy in a pre-clinical model of glioblastoma. Nat Commun. 2021;12(1):444. 10.1038/s41467-020-20599-x.

197. Bhatia V, Kamat NV, Pariva TE, Wu L, Tsao A, Sasaki K, et al. Targeting advanced prostate cancer with STEAP1 chimeric antigen receptor T cell and tumor-localized IL-12 immunotherapy. Nat Commun. 2023;14(1):2041. 10.1038/s41467-023-37874-2.

198. Misawa K, Bhat H, Adusumilli PS, Hou Z. Combinational CAR T-cell therapy for solid tumors: requisites, rationales, and trials. Pharmacol Ther. 2025;266:108763. 10.1016/j.pharmthera.2024.108763.

199. Pinto IS, Cordeiro RA, Faneca H. Polymer- and lipid-based gene delivery technology for CAR T cell therapy. J Control Release. 2023;353:196–215. 10.1016/j.jconrel.2022.11.038.

200. Michels KR, Sheih A, Hernandez SA, Brandes AH, Parrilla D, Irwin B, et al. Preclinical proof of concept for VivoVec, a lentiviral-based platform for in vivo CAR T-cell engineering. J Immunother Cancer. 2023;11(3):e 6292. 10.1136/jitc-2022-006292.

201. Chen Z, Shu J, Hu Y, Mei H. Synergistic integration of mRNA-LNP with CAR-engineered immune cells: pioneering progress in immunotherapy. Mol Ther. 2024;32(11):3772–92. 10.1016/j.ymthe.2024.09.019.

202. Bimbo JF, van Diest E, Murphy DE, Ashoti A, Evers M, Narayanavari SA, et al. T cell-specific non-viral DNA delivery and in vivo CAR-T generation using targeted lipid nanoparticles. J Immunother Cancer. 2025;13(7):13. 10.1136/jitc-2025-011759.

203. Xiong R, Sauvage F, Fraire JC, Huang C, De Smedt SC, Braeckmans K. Photothermal nanomaterial-mediated photoporation. Acc Chem Res. 2023;56(6):631–43. 10.1021/acs.accounts.2c00770.

204. Shokouhi AR, Chen Y, Yoh HZ, Brenker J, Alan T, Murayama T, et al. Engineering efficient CAR-T cells via Electroactive nanoinjection. Adv Mater. 2023;35(44):e2304122. 10.1002/adma.202304122.

205. Abdulla A, Abudureheman T, Chen K, Ghalandari B, Yan H, Zhou H, et al. Efficient capture and traceless release of functional CD8 + T cells with a microfluidic chip for enhanced in vitro and in vivo CD4-CAR transduction. Anal Chem. 2024. 10.1021/acs.analchem.4c03135.

206. Jeon H, Perez CR, Kyung T, Birnbaum ME, Han J. Separation of activated T cells using multidimensional Double spiral (MDDS) inertial microfluidics for high-efficiency CAR T cell manufacturing. Anal Chem. 2024;96(26):10780–90. 10.1021/acs.analchem.4c01981.

207. Li Z, Tian Z, Belling JN, Rich JT, Zhu H, Ma Z, et al. Acoustofluidics-based intracellular Nanoparticle delivery. Eng (Beijing). 2025;47:130–38. 10.1016/j.eng.2024.11.030.

208. Ghorashian S, Kramer AM, Onuoha S, Wright G, Bartram J, Richardson R, et al. Enhanced CAR T cell expansion and prolonged persistence in pediatric patients with ALL treated with a low-affinity CD19 CAR. Nat Med. 2019;25(9):1408–14. 10.1038/s41591-019-0549-5.

209. Giardino Torchia ML, Gilbreth R, Merlino A, Sult E, Monks N, Chesebrough J, et al. Rational design of chimeric antigen receptor T cells against glypican 3 decouples toxicity from therapeutic efficacy. Cytotherapy. 2022;24(7):720–32. 10.1016/j.jcyt.2022.03.008.

210. Li D, Wang W, Xie S, Ge M, Wang R, Xu Q, et al. A T-cell independent universal cellular therapy strategy through antigen depletion. Theranostics. 2022;12(3):1148–60. 10.7150/thno.66832.

211. Zhang A, Sun Y, Du J, Dong Y, Pang H, Ma L, et al. Reducing hinge flexibility of CAR-T cells prolongs survival in vivo with low cytokines release. Front Immunol. 2021;12:724211. 10.3389/fimmu.2021.724211.

212. Wang P, Wang Y, Zhao X, Zheng R, Zhang Y, Meng R, et al. Chimeric antigen receptor with novel intracellular modules improves antitumor performance of T cells. Sig Transduct Target Ther. 2025;10(1):20. 10.1038/s41392-024-02096-5.

213. Wu L, Chen J, Cai R, Wang X, Liu Y, Zheng Q, et al. Difference in efficacy and safety of anti-CD19 chimeric antigen receptor T-Cell therapy containing 4-1BB and CD28 Co-stimulatory domains for B-Cell acute lymphoblastic leukemia. Cancers (Basel). 2023;15(10):2767. 10.3390/cancers15102767.

214. Jacoby E, Ghorashian S, Vormoor B, De Moerloose B, Bodmer N, Molostova O, et al. CD19 CAR T-cells for pediatric relapsed acute lymphoblastic leukemia with active CNS involvement: a retrospective international study. Leukemia. 2022;36(6):1525–32. 10.1038/s41375-022-01546-9.

215. Zhu Y, Liu K, Rosen ST, Liu W, Zhu H. Treatment-related adverse events of chimeric antigen receptor-T therapies for cancers in clinical trials: a systematic review and meta-analysis. EClinicalMedicine. 2025;84:103267. 10.1016/j.eclinm.2025.103267.

216. Bove C, Arcangeli S, Falcone L, Camisa B, El Khoury R, Greco B, et al. CD4 CAR-T cells targeting CD19 play a key role in exacerbating cytokine release syndrome, while maintaining long-term responses. J Immunother Cancer. 2023;11(1):e 5878. 10.1136/jitc-2022-005878.

217. Abramson JS, Palomba ML, Gordon LI, Lunning MA, Wang M, Arnason J, et al. Lisocabtagene maraleucel for patients with relapsed or refractory large B-cell lymphomas (TRANSCEND NHL 001): a multicentre seamless design study. Lancet. 2020;396(10254):839–52. 10.1016/S0140-6736(20)31366-0.

218. Abramson JS, Siddiqi T, Garcia J, Dehner C, Kim Y, Nguyen A, et al. Cytokine release syndrome and neurological event costs in lisocabtagene maraleucel–treated patients in the TRANSCEND NHL 001 trial. Blood Adv. 2021;5(6):1695–705. 10.1182/bloodadvances.2020003531.

219. Ying Z, Huang XF, Xiang X, Liu Y, Kang X, Song Y, et al. A safe and potent anti-CD19 CAR T cell therapy. Nat Med. 2019;25(6):947–53. 10.1038/s41591-019-0421-7.

220. Roddie C, Dias J, O’Reilly MA, Abbasian M, Cadinanos-Garai A, Vispute K, et al. Durable responses and low toxicity after fast off-rate CD19 chimeric antigen receptor-T therapy in Adults with relapsed or refractory B-Cell acute lymphoblastic leukemia. JCO. 2021;39(30):3352–63. 10.1200/JCO.21.00917.

221. Schroeder T, Martens T, Fransecky L, Valerius T, Schub N, Pott C, et al. Management of chimeric antigen receptor T (CAR-T) cell-associated toxicities. Intensive Care Med. 2024;50(9):1459–69. 10.1007/s00134-024-07576-4.

222. Zhou Q, An Y, Zhang X, Xiao X, Bai X, Liu P, et al. Efficacy and safety of tocilizumab in managing cytokine release syndrome after CD19 CAR-T therapy for relapsed or refractory B-cell acute lymphoblastic leukemia. Front Immunol. 2025;16:1530623. 10.3389/fimmu.2025.1530623.

223. Costa BA, Flynn J, Nishimura N, Devlin SM, Farzana T, Rajeeve S, et al. Prognostic impact of corticosteroid and tocilizumab use following chimeric antigen receptor T-cell therapy for multiple myeloma. Blood Cancer J. 2024;14(1):84. 10.1038/s41408-024-01048-0.

224. Wang X, Zhang B, Zhang Q, Zhou H, Sun Q, Zhou Y, et al. Impact of tocilizumab on anti-CD19 chimeric antigen receptor T-cell therapy in B-cell acute lymphoblastic leukemia. Cancer. 2024;130(15):2660–69. 10.1002/cncr.35316.

225. Caimi PF, Pacheco Sanchez G, Sharma A, Otegbeye F, Ahmed N, Rojas P, et al. Prophylactic tocilizumab prior to anti-CD19 CAR-T cell therapy for Non-Hodgkin Lymphoma. Front Immunol. 2021;12:745320. 10.3389/fimmu.2021.745320.

226. Le RQ, Li L, Yuan W, Shord SS, Nie L, Habtemariam BA, et al. FDA approval summary: tocilizumab for treatment of chimeric antigen receptor T cell-induced severe or life-threatening cytokine release syndrome. The Oncologist. 2018;23(8):943–47. 10.1634/theoncologist.2018-0028.

227. Lin M, Nam E, Shih RM, Shafer A, Bouren A, Ayala Ceja M, et al. Self-regulating CAR-T cells modulate cytokine release syndrome in adoptive T-cell therapy. J Exp Med. 2024;221(6):e20221988. 10.1084/jem.20221988.

228. Galán-Gómez V, González-Martínez B, Alonso-Saladrigues A, Rives S, Herrero B, Kwon M, et al. Siltuximab for the treatment of early complications after chimeric antigen receptor T-cell therapy for acute lymphoblastic leukemia in children, adolescents, and young adults. Exp Hematol Oncol. 2025;14(1):49. 10.1186/s40164-025-00638-3.

229. Bajwa A, Zhao Q, Geer M, Lin C, Westholder J, Maakaron J, et al. Siltuximab for chimeric antigen receptor T-cell therapy–related CRS and ICANS: a multicenter retrospective analysis. Blood Adv. 2025;9(1):170–75. 10.1182/bloodadvances.2024013688.

230. Caballero-Bellón M, Alonso-Saladrigues A, Bobillo-Perez S, Faura A, Arqués L, Rivera C, et al. Risk factors and outcome of chimeric antigen receptor T-Cell patients admitted to pediatric Intensive care unit: CART-PICU study. Front Immunol. 2023;14:1219289. 10.3389/fimmu.2023.1219289.

231. Zhou J, Sun L, Jia Y, Wang Z, Luo T, Tan J, et al. Lipid nanoparticles produce chimeric antigen receptor T cells with interleukin-6 knockdown in vivo. J Control Release. 2022;350:298–307. 10.1016/j.jconrel.2022.08.033.

232. Zhang H, Lv X, Kong Q, Tan Y. IL-6/IFN-γ double knockdown CAR-T cells reduce the release of multiple cytokines from PBMCs in vitro. Hum Vaccines Immunotherapeutics. 2022;18(1):1–14. 10.1080/21645515.2021.2016005.

233. Ma JF, Yan JW, Liu MJ, Yan CL, Tang XW, Qiu HY, et al. Safe and potent anti-CD19 CAR T-cells with shRNA-IL-6 gene silencing element in patients with refractory or relapsed B-cell acute lymphoblastic leukemia. Hemasphere. 2024;8(10):e70007. 10.1002/hem3.70007.

234. Yang Y, Zhang Y, Xing X, Xu G, Lin X, Wang Y, et al. IL-6 translation is a therapeutic target of human cytokine release syndrome. J Exp Med. 2023;220(11):220. 10.1084/jem.20230577.

235. Becher B, Tugues S, Greter M. GM-CSF: from growth Factor to central mediator of tissue inflammation. Immunity. 2016;45(5):963–73. 10.1016/j.immuni.2016.10.026.

236. Hao Z, Li R, Meng L, Han Z, Hong Z. Macrophage, the potential key mediator in CAR-T related CRS. Exp Hematol Oncol. 2020;9(1):15. 10.1186/s40164-020-00171-5.

237. Sterner RM, Sakemura R, Cox MJ, Yang N, Khadka RH, Forsman CL, et al. GM-CSF inhibition reduces cytokine release syndrome and neuroinflammation but enhances CAR-T cell function in xenografts. Blood. 2019;133(7):697–709. 10.1182/blood-2018-10-881722.

238. Manni S, Del Bufalo F, Merli P, Silvestris DA, Guercio M, Caruso S, et al. Neutralizing IFNγ improves safety without compromising efficacy of CAR-T cell therapy in B-cell malignancies. Nat Commun. 2023;14(1):3423. 10.1038/s41467-023-38723-y.

239. Bailey SR, Vatsa S, Larson RC, Bouffard AA, Scarfò I, Kann MC, et al. Blockade or deletion of IFNγ reduces macrophage activation without compromising CAR T-cell function in hematologic malignancies. Blood Cancer Discov. 2022;3(2):136–53. 10.1158/2643-3230.BCD-21-0181.

240. Ruemmele T, Macedo R, Stein MN, Chan HT, Mapara MY, Jacquemont CF, et al. Emapalumab for severe cytokine release syndrome in solid tumor CAR-T: a case report. Front Oncol. 2025;15:1543622. 10.3389/fonc.2025.1543622.

241. Cai W, Lu Y, He H, Li J, Liu S, Geng H, et al. Efficacy of emapalumab in the management of anti‑CD19 chimeric antigen receptor T‑cell therapy‑associated cytokine release syndrome: a report of two cases. Oncol Lett. 2024;29(2):71. 10.3892/ol.2024.14817.

242. Manghisi B, Cotilli G, Fedele M, Perfetti P, Terruzzi E, Verga L, et al. Case report: successful use of emapalumab in adult B-cell acute lymphoblastic leukemia experiencing severe neurotoxicity and hemophagocytic lymphohistiocytosis-like features after CAR-T cell therapy. Front Immunol. 2025;16:1563736. 10.3389/fimmu.2025.1563736.

243. Chen Y, Li R, Shang S, Yang X, Li L, Wang W, et al. Therapeutic potential of TNFα and IL1β blockade for CRS/ICANS in CAR-T therapy via ameliorating Endothelial activation. Front Immunol. 2021;12:623610. 10.3389/fimmu.2021.623610.

244. Jain MD, Smith M, Shah NN. How I treat refractory CRS and ICANS following CAR T-cell therapy. Blood. 2023;141:2022017414. 10.1182/blood.2022017414.

245. Frigault MJ, Yao N, Berger TR, Wehrli M, Gallagher KME, Horick N, et al. Single-cell dynamics of breakthrough toxicities after anakinra prophylaxis for axicabtagene ciloleucel in lymphoma. Blood Adv. 2025;9(9):2122–35. 10.1182/bloodadvances.2024015161.

246. Nouveau L, Buatois V, Cons L, Chatel L, Pontini G, Pleche N, et al. Immunological analysis of the murine anti-CD3-induced cytokine release syndrome model and therapeutic efficacy of anti-cytokine antibodies. Eur J Immunol. 2021;51(8):2074–85. 10.1002/eji.202149181.

247. Leclercq G, Steinhoff N, Haegel H, De Marco D, Bacac M, Klein C. Novel strategies for the mitigation of cytokine release syndrome induced by T cell engaging therapies with a focus on the use of kinase inhibitors. Oncoimmunology. 2022;11(1):2083479. 10.1080/2162402X.2022.2083479.

248. Huarte E, O’Connor RS, Peel MT, Nunez-Cruz S, Leferovich J, Juvekar A, et al. Itacitinib (INCB039110), a JAK1 inhibitor, reduces cytokines associated with cytokine release syndrome induced by CAR T-cell therapy. Clin Cancer Res. 2020;26(23):6299–309. 10.1158/1078-0432.CCR-20-1739.

249. Zi FM, Ye LL, Zheng JF, Cheng J, Wang QM. Using JAK inhibitor to treat cytokine release syndrome developed after chimeric antigen receptor T cell therapy for patients with refractory acute lymphoblastic leukemia. Med (Baltim). 2021;100(19):e25786. 10.1097/MD.0000000000025786.

250. Wang Y, Ding L, Wang Y, Li H, Wu M, Li S, et al. Tetrandrine alleviates macrophage activation syndrome after CAR-T cell therapy. Phytomedicine. 2025;139:156483. 10.1016/j.phymed.2025.156483.

251. Dibas A, Rhiel M, Patel VB, Andrieux G, Boerries M, Cornu TI, et al. Cell-based models of ‘Cytokine release syndrome’ Endorse CD40L and Granulocyte–macrophage Colony-Stimulating Factor knockout in chimeric antigen receptor T cells as mitigation strategy. Cells. 2023;12(21):2581. 10.3390/cells12212581.

252. Miller IC, Zamat A, Sun L, Phuengkham H, Harris AM, Gamboa L, et al. Enhanced intratumoural activity of CAR T cells engineered to produce immunomodulators under photothermal control. Nat Biomed Eng. 2021;5(11):1348–59. 10.1038/s41551-021-00781-2.

253. Liu Y, An L, Wang X, Dai Y, Zhang C, Wen Q, et al. Engineering a controllable and reversible switch for CAR-based cellular immunotherapies via a genetic code expansion system. J Hematol Oncol. 2024;17(1):122. 10.1186/s13045-024-01648-0.

254. Del Bufalo F, De Angelis B, Caruana I, Del Baldo G, De Ioris MA, Serra A, et al. GD2-CART01 for relapsed or refractory high-risk neuroblastoma. N Engl J Med. 2023;388(14):1284–95. 10.1056/NEJMoa2210859.

255. Foster MC, Savoldo B, Lau W, Rubinos C, Grover N, Armistead P, et al. Utility of a safety switch to abrogate CD19.CAR T-cell–associated neurotoxicity. Blood. 2021;137(23):3306–09. 10.1182/blood.2021010784.

256. Li H, Wong NM, Tague E, Ngo JT, Khalil AS, Wong WW. High-performance multiplex drug-gated CAR circuits. Cancer Cell. 2022;40(11):1294–305.e4. 10.1016/j.ccell.2022.08.008.

257. Labanieh L, Majzner RG, Klysz D, Sotillo E, Fisher CJ, Vilches-Moure JG, et al. Enhanced safety and efficacy of protease-regulated CAR-T cell receptors. Cell. 2022;185(10):1745–63.e22. 10.1016/j.cell.2022.03.041.

258. Shirzadian M, Moori S, Rabbani R, Rahbarizadeh F. SynNotch CAR-T cell, when synthetic biology and immunology meet again. Front Immunol. 2025;16:1545270. 10.3389/fimmu.2025.1545270.

259. Reddy NR, Maachi H, Xiao Y, Simic MS, Yu W, Tonai Y, et al. Engineering synthetic suppressor T cells that execute locally targeted immunoprotective programs. Science. 2024;386(6726):eadl 4793. 10.1126/science.adl4793.

260. Nie J, Zhou L, Tian W, Liu X, Yang L, Yang X, et al. Deep insight into cytokine storm: from pathogenesis to treatment. Sig Transduct Target Ther. 2025;10(1):112. 10.1038/s41392-025-02178-y.

261. Santomasso BD, Nastoupil LJ, Adkins S, Lacchetti C, Schneider BJ, Anadkat M, et al. Management of immune-related adverse events in patients treated with chimeric antigen receptor T-Cell therapy: ASCO guideline. JCO. 2021;39(35):3978–92. 10.1200/JCO.21.01992.

262. Synnott D, O’Reilly D, De Freitas D, Naidoo J. Cytokine release syndrome in solid tumors. Cancer. 2025;131(17):e70069. 10.1002/cncr.70069.

263. Strati P, Ahmed S, Furqan F, Fayad LE, Lee HJ, Iyer SP, et al. Prognostic impact of corticosteroids on efficacy of chimeric antigen receptor T-cell therapy in large B-cell lymphoma. Blood. 2021;137(23):3272–76. 10.1182/blood.2020008865.

264. Poiret T, Vikberg S, Schoutrop E, Mattsson J, Magalhaes I. CAR T cells and T cells phenotype and function are impacted by glucocorticoid exposure with different magnitude. J Transl Med. 2024;22(1):273. 10.1186/s12967-024-05063-4.

265. Jain MD, Smith M, Shah NN. How I treat refractory CRS and ICANS following CAR T-cell therapy. Blood. 2023;141:2430–42. 10.1182/blood.2022017414.

266. Zhang J, Xie D, Tang M, Gong Q, Huang X, Li J, et al. Utilizing plasma exchange for severe cytokine release syndrome after CAR-T cell therapy: clinical experience and literature insights. Front Immunol. 2025;16:1597512. 10.3389/fimmu.2025.1597512.

267. McNerney KO, Diorio C, Annesley C, Gardner RA, Graham AK, Sanchez-Pinto LN, et al. Management practices of CAR T-cell-related inflammatory toxicities: a survey of pediatric CAR T-cell providers. Transplant Cellular Ther. 2026;32(2):158–68. 10.1016/j.jtct.2025.09.028.

268. Maude SL, Frey N, Shaw PA, Aplenc R, Barrett DM, Bunin NJ, et al. Chimeric antigen receptor T cells for sustained remissions in leukemia. N Engl J Med. 2014;371(16):1507–17. 10.1056/NEJMoa1407222.

269. Yoshikawa T, Ito Y, Wu Z, Kasuya H, Nakashima T, Okamoto S, et al. Development of a chimeric cytokine receptor that captures IL-6 and enhances the antitumor response of CAR-T cells. Cell Rep. Med. 2024;5(5):101526. 10.1016/j.xcrm.2024.101526.

270. Meister H, Look T, Roth P, Pascolo S, Sahin U, Lee S, et al. Multifunctional mRNA-based CAR T cells display promising antitumor activity against glioblastoma. Clin Cancer Res. 2022;28(21):4747–56. 10.1158/1078-0432.CCR-21-4384.

271. Pellegrino C, Favalli N, Volta L, Benz R, Puglioli S, Bassi G, et al. Peptide-guided adaptor-CAR T-Cell therapy for the treatment of SSTR2-expressing neuroendocrine tumors. Oncoimmunology. 2024;13(1):2412371. 10.1080/2162402X.2024.2412371.

272. Volta L, Myburgh R, Pellegrino C, Koch C, Maurer M, Manfredi F, et al. Efficient combinatorial adaptor-mediated targeting of acute myeloid leukemia with CAR T-cells. Leukemia. 2024;38(12):2598–613. 10.1038/s41375-024-02409-1.

273. Zhang H, Hu Y, Shao M, Teng X, Jiang P, Wang X, et al. Dasatinib enhances anti-leukemia efficacy of chimeric antigen receptor T cells by inhibiting cell differentiation and exhaustion. J Hematol Oncol. 2021;14(1):113. 10.1186/s13045-021-01117-y.

274. Zhang M, Fu S, Feng J, Hong R, Wei G, Zhao H, et al. Dasatinib and CAR T-Cell therapy in newly diagnosed Philadelphia chromosome–positive acute lymphoblastic leukemia. JAMA Oncol. 2025;11(6):625. 10.1001/jamaoncol.2025.0674.

275. Baur K, Heim D, Beerlage A, Poerings AS, Kopp B, Medinger M, et al. Dasatinib for treatment of CAR T-cell therapy-related complications. J Immunother Cancer. 2022;10(12):e 5956. 10.1136/jitc-2022-005956.

276. Vogt KC, Silberman PC, Lin Q, Han JE, Laflin A, Gellineau HA, et al. Microenvironment actuated CAR T cells improve solid tumor efficacy without toxicity. Sci Adv. 2025;11(4):eads 3403. 10.1126/sciadv.ads3403.

277. Jambon S, Sun J, Barman S, Muthugounder S, Bito XR, Shadfar A, et al. CD33–CD123 IF-THEN gating reduces toxicity while enhancing the specificity and memory phenotype of AML-Targeting CAR-T cells. Blood Cancer Discov. 2025;6(1):55–72. 10.1158/2643-3230.BCD-23-0258.

278. Zhu L, Man C, Harrison RES, Wu Z, Limsakul P, Peng Q, et al. Engineering a Programmed Death-ligand 1-targeting monobody via directed evolution for SynNotch-gated cell therapy. ACS Nano. 2024;18(11):8531–45. 10.1021/acsnano.4c01597.

279. Zheng R, Shen K, Liang S, Lyu Y, Zhang S, Dong H, et al. Specific ECM degradation potentiates the antitumor activity of CAR-T cells in solid tumors. Cell Mol Immunol. 2024;21(12):1491–504. 10.1038/s41423-024-01228-9.

280. Allen GM, Frankel NW, Reddy NR, Bhargava HK, Yoshida MA, Stark SR, et al. Synthetic cytokine circuits that drive T cells into immune-excluded tumors. Science. 2022;378(6625):eaba 1624. 10.1126/science.aba1624.

281. Dewaker V, Morya VK, Kim YH, Park ST, Kim HS, Koh YH. Revolutionizing oncology: the role of Artificial Intelligence (AI) as an antibody design, and optimization tools. Biomark Res. 2025;13(1):52. 10.1186/s40364-025-00764-4.

282. Abdalla AME, Miao Y, Taha YA, Meng N, Ouyang C. Strategies of AI-Driven CAR-T cell therapy towards solid tumours. Immunology. 2026. 10.1111/imm.70124.

283. Chen Y, Ren R, Yan L, Zhou Y, Sun R, Song H, et al. From bench to bedside: emerging paradigms in CAR-T cell therapy for solid malignancies. Adv Sci. 2025;12(40):e 5822. 10.1002/advs.202505822.

284. Martarelli N, Capurro M, Mansour G, Jahromi RV, Stella A, Rossi R, et al. Artificial Intelligence-powered molecular docking and steered molecular dynamics for accurate scFv selection of anti-CD30 chimeric antigen receptors. IJMS. 2024;25(13):25. 10.3390/ijms25137231.

285. Qiu S, Chen J, Wu T, Li L, Wang G, Wu H, et al. CAR-Toner: an AI-driven approach for CAR tonic signaling prediction and optimization. Cell Res. 2024;34(5):386–88. 10.1038/s41422-024-00936-1.

286. Ansari MS, Chauhan V, Singh A, Akhtar A, Chaudhary N, Tyagi R, et al. AI-guided CAR designs and targeted pathway modulation to enhance multi-antigen CAR T cell durability and overcome antigen escape. Nat Commun. 2026;17(1). 10.1038/s41467-025-68272-5.

287. Baena JC, Victoria JS, Toro-Pedroza A, Aragón CC, Ortiz-Guzman J, Garcia-Robledo JE, et al. Smart CAR-T nanosymbionts: archetypes and proto-models. Front Immunol. 2025;16:1635159. 10.3389/fimmu.2025.1635159.

288. Wang M, Komanduri KV, Datta D, Patel A, Whitaker B, Belov A, et al. An AI model classifies risks of early relapse post–CAR T-cell therapy in a multicenter real-world population with DLBCL. Blood Adv. 2025;9(22):5837–52. 10.1182/bloodadvances.2025016375.

289. Gong Y, Fei P, Zhang Y, Xu Y, Wei J. From multi-omics to Visualization and beyond: bridging micro and macro insights in CAR-T cell therapy. Adv Sci. 2025;12(20):e2501095. 10.1002/advs.202501095.

290. Li G, Du R, Wang D, Zhang X, Wang L, Pu S, et al. Improved efficacy of triple-negative breast Cancer immunotherapy via hydrogel-based Co-delivery of CAR-T cells and mitophagy agonist. Adv Sci. 2025;12(14):e2409835. 10.1002/advs.202409835.

291. Zandaki DA, Selukar S, Bi Y, Li Y, Zinsky M, Bonifant CL, et al. EASIX and m-EASIX predict CRS and ICANS in pediatric and AYA patients after CD19-CAR T-cell therapy. Blood Adv. 2025;9(2):270–79. 10.1182/bloodadvances.2024014027.

292. Pennisi M, Sanchez-Escamilla M, Flynn JR, Shouval R, Alarcon Tomas A, Silverberg ML, et al. Modified EASIX predicts severe cytokine release syndrome and neurotoxicity after chimeric antigen receptor T cells. Blood Adv. 2021;5(17):3397–406. 10.1182/bloodadvances.2020003885.

293. Sevenler D, Toner M. High throughput intracellular delivery by viscoelastic mechanoporation. Nat Commun. 2024;15(1):115. 10.1038/s41467-023-44447-w.

294. Peng L, Sferruzza G, Yang L, Zhou L, Chen S. CAR-T and CAR-NK as cellular cancer immunotherapy for solid tumors. Cell Mol Immunol. 2024;21(10):1089–108. 10.1038/s41423-024-01207-0.

295. Jiang S, Li H, Zhang L, Mu W, Zhang Y, Chen T, et al. Generic Diagramming platform (GDP): a comprehensive database of high-quality biomedical graphics. Nucleic Acids Res. 2025;53(D1):D1670–76. 10.1093/nar/gkae973.

Associated Data

Data Availability Statement

Not applicable.