Yi-Min Yang1,2†
Jun-Long Zhao- 1State Key Laboratory of Holistic Integrative Management of Gastrointestinal Cancers, Department of Medical Genetics and Developmental Biology, Fourth Military Medical University, Xi’an, China
- 2Basic Medical College, Fourth Military Medical University, Xi’an, China
Chimeric antigen receptor (CAR) cellular immunotherapy has emerged as a revolutionary modality in cancer treatment. CAR-T cell therapy has demonstrated remarkable efficacy against hematological malignancies; however, its application in solid tumors is significantly constrained by inadequate tumor infiltration, a profoundly immunosuppressive tumor microenvironment (TME), and pervasive antigen heterogeneity. Conversely, macrophages — innate immune cells inherently poised within tissues — exhibit superior tumor-tropic migration, potent phagocytic capability, and a unique capacity to remodel the TME, establishing CAR-engineered macrophages (CAR-M) as a highly promising next-generation therapeutic platform. Despite this considerable promise, the clinical translation of CAR-M faces several critical bottlenecks, including heterogeneity in cell sources, challenges in manufacturing standardization, risks of on-target/off-tumor toxicity, and the dynamic, immunosuppressive nature of the TME. This review offers a systematic and in-depth analysis of the current research landscape and engineering advances in CAR-M therapy. It comprehensively details the molecular evolution of CAR-M designs, spanning from early constructs to sophisticated logic-gated circuits and innovative in vivo generation strategies utilizing lipid nanoparticles (LNPs). We critically evaluate the applicability and limitations of various cellular sources, such as peripheral blood mononuclear cells (PBMCs), induced pluripotent stem cells (iPSCs), and the THP-1 cell line. Furthermore, the review elucidates the multimodal antitumor mechanisms of CAR-M, including the direct “phagocytosis-presentation-activation” cascade, synergistic potential with immune checkpoint blockade, and deep reprogramming of the immunosuppressive TME. By synthesizing the latest preclinical and emerging clinical evidence, this article underscores the distinctive advantages and delineates a translational roadmap for CAR-M development. It is intended to serve as an authoritative reference for the field, providing strategic insights into intelligent receptor design, precision biomanufacturing, and rational combination therapies aimed at overcoming the enduring barriers in solid tumor immunotherapy.
Graphical Abstract. In this review, we summarize the antitumor advantages of CAR-M cells, the molecular design strategies, and the cell sources and adaptations used in the preparation process. By constructing the closed-loop logic framework of “targeting demand-molecular design-cell adaptation”, we propose its prospects for the construction of intelligent dual-environment recognition CAR-M engineering platforms on the basis of logic gating signals in the future.
1 Introduction
Cellular immunotherapy is a therapeutic strategy to reconstitute or enhance the antitumor immune response of patients by modifying and infusing functional immune cells in vitro (1). In recent years, on the basis of its personalized characteristics, controllable toxicity spectrum and relatively durable therapeutic effect, this therapy has become an important pillar of precision cancer medicine. With the breakthrough development of gene editing technologies (such as CRISPR/Cas9) (2, 3) and synthetic biology technologies (4), chimeric antigen receptor (CAR)-engineered immune cells have shown great potential in the treatment of malignant tumors. For example, the clinical efficacy of CAR-T cells in B-cell leukemia and lymphoma has been verified. At present, six CAR-T-cell therapies have been approved by the US Food and Drug Administration (FDA) (5). However, CAR-T cells face multiple biological challenges in the treatment of solid tumors, including T-cell exhaustion caused by persistent antigen exposure (characterized by abnormal expression of inhibitory receptors such as PD-1 and TIM-3 and ID3/SOX2 transcription factors) (6), as well as the synergistic inhibitory effect of immunosuppressive cell populations in the TME (such as regulatory T cells and tumor-associated macrophages) (7, 8). Together, these mechanisms lead to limited infiltration, functional inactivation and a low clinical response rate to CAR-T cells in solid tumors.
In view of this, macrophages, which naturally reside in the TME and account for up to 50% of the total population (9), have returned to the focus of research because of their unique biological characteristics. Macrophages are the core effector cells of innate immunity and can clear pathogens or malignant cells through nonspecific phagocytosis mediated by pattern recognition receptors (PRRs); they can also activate adaptive immune responses through antigen presentation, proinflammatory cytokine secretion (such as IL-12 and TNF-α) and costimulatory signal transduction (such as CD80/CD86), forming a multidimensional antitumor immune regulatory network (10–12). Thus, chimeric antigen receptor-modified macrophages (CAR-Ms) have emerged. It integrates the targeting specificity of CARs with the inherent functions of macrophages (phagocytosis, antigen presentation ability and proinflammatory cytokine secretion characteristics) to achieve efficient recognition, precise clearance of tumor cells and remodeling of the immune microenvironment (13). However, current research on CAR-Ms still faces key limitations, including the immunogenicity of allogeneic cell sources (14–16), the stability of in vitro polarization efficiency, the heterogeneity of TAMs (17, 18), the acclimatization of the immunosuppressive tumor microenvironment (19), and functional maintenance after cryopreservation and resuscitation (20).
This study systematically analyzes the core advantages of CAR-M therapy, focusing on its molecular design strategies (such as costimulatory domain optimization and suicide switch embedding), cell source selection (autologous vs. induced pluripotent stem cell differentiation) and engineering platforms (such as CRISPR screening-driven function enhancement) to construct a closed-loop logical framework of “targeted demand-molecular design-cell adaptation”. Moreover, this paper discusses the obstacles of TAM heterogeneity and the tumor immunosuppressive microenvironment to the practical application of CAR-M, proposes artificial intelligence-assisted intelligent gene circuit design (such as a logic-gated signaling system) and multiomics-driven precise regulation strategies, aiming to provide an innovative theoretical paradigm and a technical path for the clinical translation of CAR-M therapy.
2 The core advantages of CAR-M cells in tumor immunotherapy
Compared with traditional CAR-T-cell therapy, CAR-M-cell therapy has potential multidimensional biological advantages in the treatment of solid tumors by integrating the innate immune effector function of macrophages and the targeting specificity of the CAR.
2.1 Antigen presentation and multimodal immune activation
CAR-Ms can not only engulf tumor cells directly through Fcγ receptor (FcγR)-dependent and FcγR-independent pathways but also present unique antigen-presenting functions such as tumor-associated antigens (TAAs) to CD8+ T cells and CD4+ T cells through MHC-I/II molecules, thereby activating adaptive immune responses (21, 22). The study by Klichinsky et al. (16) is important: using the chimeric adenovirus vector Ad5f35, they engineered macrophages expressing the NY-ESO1 antigen and HLA-A2*01 and confirmed that Ad5F35-transduced CAR-M significantly induced CD8+ T-cell proliferation in coculture experiments with autologous specific T cells. Moreover, the activated T cells presented a 3-fold increase in IFN-γ secretion compared with that in the control group. These findings suggest that CAR-Ms can synergistically affect innate and adaptive immunity through the “phagocytosis, presentation, and activation” cascade to achieve deep immune surveillance of tumors.
Pierini et al. (23) An important breakthrough in the preclinical study of HER2-positive solid tumors was made: in the CT26-HER2 model with limited efficacy of PD-1 inhibitor monotherapy (only 14.3%), the complete response rate after combined CAR-M therapy significantly increased to 77.8%. The number of dendritic cells was increased approximately 6-fold in the combined treatment group, and the infiltration of CD4+ T cells, CD8+ T cells and NK cells was also significantly increased, as shown by multiple MIBI imaging. T-cell receptor beta chain (TCRB) sequencing further revealed an approximately 4-fold increase in the Morisita overlap index and a 42% decrease in the Simpson clonality index, indicating the expansion of tumor-reactive T cells in vivo and an increase in the diversity of the T-cell receptor repertoire. This study revealed that CAR-Ms can activate adaptive immune cells through multimodal immune activation and thus have synergistic antitumor effects with immune checkpoint inhibitors, which provides a new idea for the combined treatment of immune “cold” tumors.
2.2 Remodeling the immunosuppressive tumor microenvironment
TAMs usually present an immunosuppressive phenotype in the TME, which mainly drives tumor immune escape by secreting immunosuppressive factors such as IL-10 and TGF-β and promoting angiogenesis (24–26). Studies have shown that CAR-M cells can induce the reprogramming of immunosuppressive TAM subsets to a proinflammatory phenotype by secreting proinflammatory factors (such as IL-12 and TNF-α) (27, 28) and chemokines (such as CCL5 and CXCL9) (16, 29), thereby improving the immunosuppressive properties of the TME. The key evidence came from a study in HER2-positive solid tumors (16): HER2-specific CAR-M cells expressing the CD3ζ signal domain significantly reduced the proportion of immunosuppressive TAM subsets in the TME (from 71% to 14%) while increasing the expression of the proinflammatory phenotypic markers iNOS and CD86 in a humanized mouse model. Cytokine microarray analysis revealed that the secretion of the proinflammatory factor GM-CSF was significantly increased, while that of the immunosuppressive factor IL-10 was significantly decreased, and the overall transcriptional state was skewed toward the proinflammatory phenotype (30). This study confirmed the remodeling effect of CAR-M cells on the function of TAMs at the phenotypic and molecular levels. In addition, CAR-M cells can also indirectly weaken the immunosuppressive effect of regulatory T cells (Tregs) by eliminating the chemokine CCL22 from regulatory T (Treg) cells and jointly form a “two-way” microenvironmental regulatory network.
2.3 Potential for a universal therapeutic platform
The clinical application of CAR-T-cell therapy is limited by the need for human leukocyte antigen (HLA) matching and the risk of graft-versus-host disease (GvHD) (31, 32), whereas the low immunogenicity of macrophages provides a theoretical basis for the development of universal CAR-M (33)-cell therapy. Takata et al. (34) demonstrated for the first time that macrophages (iMACs) differentiated from induced pluripotent stem cells (iPSCs) could home to tumor tissues and differentiate into functional macrophage subsets in vivo. Zhang’s team (35) successfully constructed an ipSC-derived CAR-M (CAR-iMac) in 2020 and reported that it significantly inhibited the growth of ovarian cancer xenografts in an NSG mouse model. This technology solves the problems of low in vitro expansion efficiency and large batch heterogeneity of primary macrophages via the standardized iPSC differentiation process and lays the foundation for the large-scale production of “off-the-shelf” CAR-M. However, allogeneic CAR-M may trigger a mild host immune response (16, 36), and its long-term safety needs to be further verified by optimizing HLA matching strategies or gene editing to knock out immunogenic antigens (37, 38).
3 Engineering strategy for CAR-M cells and optimization of antitumor mechanisms
3.1 Molecular design strategy for CARs
The CAR structure is central to the function of CAR-M-cell therapy, which determines the ability of macrophages to recognize tumors, activate signals, polarize anti-inflammatory phenotypes, and achieve efficacy (39, 40). Its design has undergone an evolutionary process from directly learning from CAR-T-cell technology (16, 41) to exclusive optimization according to the biological characteristics of macrophages (42–44). The current CAR structure aims to improve the survival persistence of CAR-M cells in vivo, improve the TME, and further enhance the phagocytosis, tumor killing and antigen presentation ability of macrophages (42).
A CAR is a protein molecule that is artificially designed and synthesized via genetic engineering. Its core function is to give immune cells a new ability to recognize and attack specific targets, such as tumor cells (45). It consists of three major functional components (46): a single-chain variable fragment (scFv) that specifically recognizes target antigens on the surface of tumor cells, a hinge region and a transmembrane domain that links the extracellular and intracellular domains, and an intracellular signaling domain that triggers antitumor effects. The initial CAR-M initially activated only the phagocytic or proinflammatory ability of macrophages toward tumors: the use of CD3ζ or the ITAM motif of the structurally homologous FcϵRI-γ chain to transmit signals effectively initiates antigen-dependent phagocytosis, which is the earliest structural basis to verify the feasibility of CAR-M (16); subsequent introduction of the CD147 intracellular domain of macrophages itself to activate the expression of matrix metalloproteinases (MMPs) (47) or the intracellular domain (TIR) of Toll-like receptors such as TLR4 to mimic pathogen recognition signals (28, 42) initiates only a single basal signal. The basic feasibility of first-generation CAR-M cells was demonstrated in different ways.
Second-generation CAR-Ms can enhance multiple functions, such as phagocytosis, polarization, and proliferation, by integrating dual signals. Zhang et al. (42) realized the dual increase in CAR-M proinflammatory activity and phagocytic intensity by connecting the phagocytosis signal (CD3ζ) and the proinflammatory polarization signal (TIR) in series; Yan et al. (48) expressed Trop2 antigen-targeting second-generation CAR containing CD3ζ and costimulatory signaling molecule 4-1BB in macrophages to achieve the dual effects of phagocytosis enhancement and pro-inflammatory polarization; Ning et al. (43), on the one hand, used the highly expressed FcγRI signal domain of macrophages to enhance phagocytosis, and at the same time introduced α1β1 integrin signal domain to significantly promote the proliferation and survival of CAR-M itself. This “proliferation + phagocytosis” dual enhancement design shows better efficacy and safety than the traditional CD3ζ-CAR.
Third-generation CAR-Ms not only optimize the antitumor function of macrophages but also emphasize the modification of the TME by cells to improve their durability and comprehensive efficacy (49–51). On the basis of second-generation CAR-M, Kang et al. (52) used CD3ζ as the intracellular signal transduction domain and CD28 as the intracellular costimulatory domain and simultaneously expressed IFNγ in tandem, which not only strengthened the proinflammatory phenotype but also reshaped the TME and inhibited immunosuppressive cells (such as Tregs). It also recruits and activates other immune cells, such as T cells, to form synergistic antitumor immunity (Figure 1).
Figure 1. Evolution of CAR structure in CAR-T versus CAR-M cells. (a) Progression of CAR-T cells. First-generation CARs were characterized by the inclusion of only the CD3ζ signaling domain and the absence of costimulatory elements. The second generation introduced one costimulatory domain, while the third generation incorporated two. Fourth-generation CAR-T cells are designed to secrete immunomodulatory factors (e.g., IL-12, IL-15) or express checkpoint inhibitors (e.g., anti-PD-1 single-chain variable fragments) to remodel the immunosuppressive tumor microenvironment. Fifth-generation CAR-T cells integrate cytokine receptor signaling domains—such as a truncated IL-2 receptor beta chain (IL-2Rβ) or JAK-STAT module—along with costimulatory domains. These enhancements enable autonomous cytokine-driven proliferation and incorporate logic-gating capabilities. (b) Progression of CAR-M cells. First-generation CAR-Ms consisted of a single intracellular domain (e.g. CD3ζ or FcγRI) to trigger phagocytosis. The second generation contains two intracellular domains: one responsible for phagocytosis and the other serving as a costimulatory domain. The third generation further incorporates a cytokine inducer coupling domain.
Of note, most current CAR-M constructs still utilize the CD3ζ chain derived from conventional CAR-T designs as the core signaling module. This preference stems from the presence of ITAM motifs in CD3ζ, which serve as a well-established signaling hub capable of efficiently initiating lymphocyte activation and macrophage phagocytosis (53). Such a “borrowing” strategy is both rational and efficient during the conceptual validation phase of first-generation CAR-M designs. However, macrophages also possess their own unique signaling receptors, such as MERTK and DAP12 (54, 55). Therefore, the next generation of CAR-M designs should not be confined to the T-cell paradigm, but instead explore signaling domains intrinsic to macrophages themselves. For instance, combining the TIR domain of TLR4 (48) or the intracellular domain of integrins (49) with CD3ζ may allow more precise regulation of macrophage polarization, survival, and phagocytic activity. Future research directions include: 1) screening for the most efficient combinations of intracellular signaling domains to activate macrophage-specific functions, such as M1 polarization, matrix degradation, and antigen presentation; and 2) developing novel synthetic signaling components entirely based on myeloid cell biology, to achieve native-like optimization and precise control of CAR-M function.
3.2 Increased solid tumor homing and stromal penetration
The limitations of CAR-T cells in the treatment of solid tumors can be attributed in part to insufficient infiltration due to the loss of chemokine gradients, such as low CXCL9/CXCL10 expression, and physical barriers formed by the extracellular matrix (ECM) (7, 56). In contrast, CAR-Ms can respond to gradient signals of chemokines such as CCL2 and CX3CL1 in the TME by highly expressing chemokine receptors (such as CCR2 and CX3CR1) and achieve efficient directional migration (Table 1) (58, 59). In addition, CAR-Ms can secrete MMPs (such as MMP-3, MMP-9, and MMP-14), directly degrade collagen and laminin in the ECM, and break through the physical barrier of tumors (60).
Table 1. Tumor infiltration and therapeutic efficiency of CAR-M.
A breakthrough study by Zhang et al. (47) provided key evidence for this finding: HER2-targeting CAR-M (CAR-147) significantly upregulated the expression of MMP3, MMP11, MMP13 and MMP14 when cocultured with HER2 + 4T1 tumor cells in vitro. In the orthotopically transplanted breast cancer model, the tumor ECM density of the mice in the CAR-147 treatment group was reduced, and the proportion of tumor-infiltrating T cells was increased by 4-fold compared with that in the control group, indicating that CAR-M could synergistically enhance immune cell infiltration through the dual mechanism of “chemochemical degradation” to reshape TME permeability.
Zhang’s team (57) designed a dual-target quadrivalent CAR structure by simultaneously targeting the GPC3 antigen of liver cancer cells and the FAP marker of tumor-associated fibroblasts (CAFs). Immunofluorescence staining revealed that the CAF layer was thinner, the fibrotic area was reduced, and the infiltration depth of CD8+ T cells was significantly greater than that in the LNP-Control group. Moreover, the dual-effect design of FAP-CAR-△TGFβRII significantly reduced the p-SMAD2/3 signal in the macrophages of the LNP-GF CAR mRNA group, thereby inhibiting its profibrotic effect. These results indicate that CAR-M cells not only have the ability to penetrate but also cooperate with endogenous immune cells to achieve deep infiltration and long-lasting antitumor effects by systematically removing the fiber barrier and fiber-promoting signals.
3.3 Synergistic effects of the multimodal inflammatory response and immunity
CAR-M cells exert anti-tumor effects that extend well beyond direct phagocytic activity. Functioning as a critical nexus between innate and adaptive immunity, CAR-M cells orchestrate multimodal inflammatory responses and promote immune synergy through the secretion of a distinct array of cytokines and chemokines (Table 2). As the core immunosuppressive component of the TME, TAMs drive the functional exhaustion of CAR-T cells by secreting immune checkpoint molecules such as IL-10 and TGF-β and expressing PD-L1 (61). The unique advantage of CAR-M cells is that they can remodel the TME through dual mechanisms. On the one hand, CAR-Ms secrete proinflammatory factors such as IL-12 and TNF-α and directly induce the reprogramming of immunosuppressive TAM subsets to a proinflammatory phenotype. On the other hand, it indirectly weakens Treg-mediated immunosuppression by removing the Treg chemokine CCL22 (62). This allows CAR-M cells to not only exert antitumor effects through direct phagocytosis and antigen presentation but also activate systemic immune responses due to their strong proinflammatory properties (63).
Table 2. The key effector molecules secreted by CAR-M cells after activation and their functions.
The study by Klichinsky et al. (16) revealed the molecular regulatory network of CAR-M: transcriptome analysis of anti-CD19 CAR-M (CD3ζ signaling domain) cells revealed that it significantly upregulated the expression of interferon signaling pathway, pattern recognition receptor pathway and proinflammatory phenotype TAM markers (inducible nitric oxide synthase (iNOS) and CD86). In vitro coculture experiments further confirmed that CAR-Ms could increase the secretion of IL-12 and downregulate the expression of arginase-1 (Arg1) in proinflammatory TAM subsets.
Zhang’s team (44) blocked the “do not eat me” signal (CD47-SIRPα axis) by silencing the SIRPα gene of CAR-M, increased the secretion of IL-1β by approximately 2–3-fold, and significantly increased the levels of nitric oxide (NO) and reactive oxygen species (ROS). Transcriptome analysis revealed that the expression of NADPH oxidase complex genes (such as NOX2 and RAC1) was upregulated in SIRPα-silenced CAR-Ms, confirming that SIRPα-silenced CAR-Ms enhanced tumor cell death through the oxidative stress pathway.
In the innovative study of Gu et al. (64) CD3ζ/TLR4 dual-signal domain CAR-Ms were constructed through an mRNA-LNP delivery system, which activated the TLR4-Nf-κb pathway in a peritoneal metastasis model, increased the M1/M2 ratio from 1:1 to 3:1, and significantly reduced Treg infiltration (by 10%). Single-cell sequencing further revealed that CAR-M cells could increase the proliferation ability of TCF1+PD-1+ exhausted CD8+ T cells (Tpex) and reconstruct the exhausted phenotype of T cells, which provides a theoretical basis for combined treatment with immune checkpoint inhibitors.
CAR-M cells exert a potent anti-tumor immune cascade by releasing a series of inflammatory cytokines and chemokines, which primarily function through three interconnected mechanisms: 1) Direct reprogramming of the TME: cytokines such as IL-12 and TNF-α polarize neighboring M2-type tumor-associated macrophages (TAMs) toward a pro-inflammatory M1-like phenotype, fostering a self-amplifying anti-tumor environment; 2) Initiation of adaptive immunity: after phagocytosing tumor cells, CAR-M cells cross-present tumor antigens to T cells via MHC-I/II pathways while simultaneously providing critical signals like IL-12 that act as the “third signal” for T cell activation, driving their clonal expansion and differentiation into effector and memory cells; 3) Immune memory formation: this coordinated process ultimately establishes long-term host immune surveillance against tumor neoantigens, constituting a durable in situ vaccine effect.
3.4 Proliferation and controllability design and safety optimization
At present, the proliferation capacity and controllability of CAR-M cells remain under exploration. A major challenge hindering clinical translation is the limited proliferative ability of terminally differentiated mature tissue macrophages, which leads to insufficient persistence in vivo and thus constrains long-term efficacy. This biological limitation highlights the need for a strategic shift in next-generation CAR-M development—from merely mitigating risks to proactively enhancing proliferation capacity and establishing sophisticated safety control systems.
3.4.1 Proactive enhancement of proliferative capacity
CAR-M cells are generally considered to exhibit “transient in vivo survival” (65, 66), meaning that their short post-infusion lifespan limits durable antitumor efficacy. To overcome the restricted proliferative capacity of these cells, current research aims to artificially enhance their expansion ability—even toward long-term proliferation—using synthetic biology approaches. For instance, CRISPR activation (CRISPRa) can be applied to upregulate endogenous cell cycle–regulating genes (e.g., c-MYC), or lentiviral transduction can be used to introduce exogenous immortalizing gene combinations (such as Raf and Myc, as validated in iBMDM models) (67, 68). Theoretically, such “long-acting CAR-M” or “immortalized CAR-M” could achieve sustained expansion and residence in vivo, enabling continuous tumor immune surveillance and potentially breaking through current therapeutic limitations (Table 3).
Table 3. Tumor killing efficiency of CAR-M in preclinical studies.
3.4.2 Safety control, exhaustion prevention, and dynamic balance strategies
Although terminally differentiated mature tissue macrophages possess limited proliferative capacity, genetic engineering may inadvertently introduce risks associated with uncontrolled activation (16, 65, 69). Sustained CAR signaling could trigger a threefold hazard: excessive inflammation or tissue damage (16, 35, 70), CAR-M cell exhaustion, and functional impairment of tumor-infiltrating T cells (53, 71, 72). More importantly, the sustained secretion of inflammatory factors by CAR-M cells can maintain tumor-infiltrating T cells in a state of prolonged exposure to high antigen levels and co-stimulatory signals. This persistent activation environment may accelerate their differentiation toward a terminally exhausted phenotype, thereby undermining the establishment of long-term adaptive immune memory (73, 74).
Future research should focus on integrating precise control strategies with dynamic equilibrium regulation. On one hand, spatiotemporal precision in CAR-M activity can be achieved by incorporating “suicide switches” (e.g., the iCasp9 system) (75, 76) or designing conditionally activated CARs (such as hypoxia-responsive promoters) (77). CRISPR-Cas9 technology may further be employed to knockout proliferation-related genes such as c-MYC (78, 79) or to restrict their division cycles through epigenetic editing (80, 81). On the other hand, safety designs for next-generation CAR-M should evolve from simple “emergency clearance” toward “dynamic equilibrium regulation.” Specific approaches include developing closed-loop feedback systems in which cytokines (e.g., IL-12) expressed by CAR-M are subject to self-inhibition or linked to metabolite levels (e.g., low arginine) for autonomous regulation (82). Intermittent activation switches can also be constructed to control CAR expression or signaling via exogenous small-molecule drugs, enabling “on-demand activation” and allowing recovery time for the immune system (83–85). Additionally, CAR-M may be engineered to secrete cytokines such as IL-21 or IL-7 under tumor microenvironment regulation to support immune partner cells and delay their functional exhaustion (86, 87).
The core of these strategies lies in shifting from merely pursuing “maximal persistence” to actively designing “optimal therapeutic kinetics.” This approach would allow CAR-M to operate efficiently during tumor clearance phases while remaining silent or being eliminated during phases of potential risk, ultimately achieving an optimal balance between efficacy and safety.
3.4.3 Prospective safety and ethical considerations
While safety measures such as gene editing and suicide switches show promise, their clinical translation requires careful evaluation of potential risks. These include possible off-target effects of CRISPR-Cas9 (88) and immunogenicity associated with exogenous suicide-switch proteins like iCasp9 (76, 89). Ethical concerns must also be addressed, particularly regarding technologies involving embryonic stem cells or heritable genome modifications. Furthermore, the complexity of these engineering strategies is likely to increase production costs, posing challenges to accessibility and future commercialization. Therefore, rigorous safety assessments, thorough ethical review, and cost-reducing innovations must advance in parallel with technological development. In the future, combining in vivo delivery systems—such as LNP—based platforms—with modular mRNA engineering may help lower application costs and mitigate ethical concerns, thereby facilitating broader clinical adoption.
4 Study of the heterogeneity of cell origins and their suitability for CAR-M technology
4.1 Peripheral blood mononuclear cells
As an expanding innovation in CAR-T-cell therapy, the selection and optimization of the cell source of CAR-M-cell technology is the core challenge of clinical translation. Several studies have confirmed that the antitumor activity of PBMC-derived CAR-modified cells is potentially feasible. Xie et al. (90) constructed a second-generation CAR molecule targeting HER2 and reported that PBMC-activated CAR-HER2-T cells significantly inhibited tumor growth and prolonged survival in a tumor-bearing mouse model. Miranda et al. (91) Furthermore, there was no significant difference in the activated phenotype, cytokine secretion or in vitro cytotoxicity of CAR-T cells derived from frozen PBMCs and fresh samples, suggesting that frozen PBMCs can be used as a standardized raw material for CAR-T-cell therapy to support global clinical promotion. However, as primary cells, PBMCs have limited proliferation ability in vitro, significant phenotypic heterogeneity, and are prone to genetic drift or microbial contamination during culture (92), which severely limits their large-scale clinical application. Therefore, current studies are more inclined to explore other cell sources that can be standardized for expansion.
4.2 Human immortalized mononuclear cell line (THP-1)
In 1980, Shigeru TSUCHIY et al. (93) THP-1 cells were isolated and cultured from the blood of a boy with acute monocytic leukemia to clarify the immune function of the monocyte sequence, and an immortalized human monocytic cell line that retains the characteristics of monocytes, such as immune function, was established. As a classic model for the study of monocyte function, the THP-1 cell line has been innovatively applied to the development of CAR-M cells in recent years. This cell line can be induced by a combination of PMA and 1,25-dihydroxyvitamin D3 (VD3) to differentiate into a phenotype characteristic of macrophages, including mitochondrial expansion, upregulation of the antiapoptotic protein Mcl-1, and enhanced resistance to apoptosis (94).
Research team of Yang et al. (95) designed CD147 and CD3ζ intracellular fragments into CARs and screened the HER2-CAR structural design on the basis of CAR-CD3ζ-CD147 through the FLAG tag test of CARs and the antibody-dependent cellular phagocytosis (ADCP) ability test of cells. Moreover, to avoid cytokine storms that can occur late in CAR-M treatment, researchers have stably transfected suicide gene-induced caspase-9 (iCasp9) into THP-1 cells (95) and treated them with different concentrations of AP1903 and different numbers of apoptotic cells. The expression of iCasp9 in THP-1 cells and its ability to rapidly induce THP-1 cell death were detected. Interestingly, the investigators reported that activated CAR-THP-1 cells exhibited an M1 phenotype, which promoted a shift from a suppressive to a proinflammatory immune microenvironment. Owing to its ultrafast average doubling rate of 30–50 hours, low homogeneous genetic background and wide availability, THP-1 cells greatly compensate for the limitations of PBMCs in terms of proliferation and passage ability and the high variability in cell performance caused by diverse genetic backgrounds (96). However, although THP-1-derived monocytes have typical macrophage characteristics, they cannot fully mimic all the properties (97) of macrophages. For example, LPS-stimulated THP-1-derived monocytes produced significantly less IL-6 than did PBMCs and monocytes under the same conditions. However, the secretion of IL-8 in THP-1 cell-conditioned medium was greater than that in THP-1 cell-conditioned medium (98). This finding also suggests that the cell source of CAR-M cells needs more stable and efficient activation in clinical protocols.
4.3 Human pluripotent stem cells
To overcome the technical bottleneck of the cell source, hPSCs have become a potential ideal source of CAR-M cells because of their unlimited proliferation potential and genetic editing ability. Zhang et al. (35) developed induced pluripotent stem cells (iPSCs) by reprogramming peripheral blood PBMCs and successfully differentiated them into functional CAR-Ms, which verified the feasibility of hPSC-derived CAR-Ms. Shen et al. (33) developed a monolayer culture system without a feeding layer, combined with the serum-free culture conditions established by Wilgenburg’s team (99), which significantly improved the differentiation efficiency of CD14+CD16lowCD163+ macrophages, laying the foundation for large-scale production. On the basis of the easy editing and immune compatibility of hPSCs, Wang et al. (100) used CRISPR-Cas9 to knock out the ACOD1 gene to increase the reactive oxygen species (ROS) generation ability of hPSC-derived CAR-M (CAR-iMac) and promote its polarization to the M1 type. Furthermore, it has a stronger tumor inhibitory effect in ovarian cancer and pancreatic cancer models. In addition, Lei et al. (42) further optimized the design of CARs and constructed second-generation CAR-Ms with a CD3ζ–TIR dual signal domain, which significantly improved the phagocytic activity and antitumor function of CAR-iMacs. However, the interindividual differences in hPSC differentiation efficiency, phenotypic instability and high production cost are still the main obstacles to their clinical translation.
4.4 Future developments in macrophage sources
THP-1 cells, PBMCs and hPSCs have their own advantages and limitations. Although PBMCs are clinically relevant, their expansion ability is insufficient. hPSCs have significant potential, but their differentiation efficiency and cost limit their application. The controllability and function of THP-1 cells are balanced but need to be further optimized to mimic the characteristics of natural macrophages. Future studies need to combine gene editing technology with novel differentiation strategies to overcome existing bottlenecks and promote the clinical translation of CAR-M therapy. Moreover, a study of mouse long-lived macrophages (LT-BMDMs) revealed that mouse bone marrow macrophages can be immortalized via the transduction of functional genes (101). Blasi et al. first named these cells immortalized bone marrow macrophages (iBMDMs) (101). Subsequent studies revealed that iBMDMs exhibited the same functions as primary bone marrow cells in terms of metabolic status, polarization phenotype, and response to external stimuli. Even in a mouse tumor model, it showed stronger antitumor activity than primary macrophages (102) did. A safety evaluation in the latest literature confirmed that iBMDMs could survive for 35 days in vitro and maintain vigorous proliferation activity for 4 weeks. The body has a life span of three weeks (102). Compared with those of primary BMDMs, such proliferation activity and survival ability are great breakthroughs, and the safety hazard caused by immortalization is overcome; thus, the authors named these cells long-lived macrophages (LT-BMDMs). These mouse LT-BMDMs have proliferative activity, editing ability, and powerful antitumor functions, suggesting the possibility of a new cell source for macrophage therapy in the future. Is it possible to induce and construct long-lived macrophages (hLT-BMDMs) from human primary bone marrow macrophages? Min Peng successfully screened the key regulatory genes of immortalized CAR-T cells via CRISPR-Screen technology. Although its safety and clinical application still need to be further verified and evaluated, it also provides a new idea for the construction of hLT-BMDMs.
5 Discussion: the bottleneck of CAR-M-cell therapy
5.1 TAM heterogeneity and the domestication of the immunosuppressive tumor microenvironment
However, preclinical studies and early clinical data suggest that CAR-M cells exhibit unique advantages over CAR-T cells in terms of phagocytosis, antigen presentation, and remodeling of the TME (13, 103, 104). These features make CAR-M cells promising immunotherapy agents for T-cell-refractory solid tumors. However, at the same time, the functional and spatial heterogeneity of TAMs, as well as the “domestication-depletion” effect of the TME, pose challenges for the further clinical translation of CAR-M cells.
Recent advances in spatial transcriptomics and single-cell technologies have further revealed unprecedented functional and spatial heterogeneity within TAM compartments (105, 106). Instead of having a uniform distribution, TAMs are organized into specific “ecotypes,” with distinct isoforms distributed in necrotic zones, hypoxic zones, invasion fronts, and perivascular spaces, each exhibiting a unique transcriptional program and role in immune regulation (67, 107, 108). This complex TME dynamic presents a new set of challenges for CAR-M therapy. First, different organ-specific microenvironments (e.g., liver, lung, and brain) cultivate resident macrophages in unique ways, which means that CAR-M cells may need to be tailored for different organ situations to overcome the local immunosuppressive network (68, 108–111). Second, owing to the regional specificity of different tumor sites in the same organ (such as the tumor core hypoxic zone, invasive marginal zone, and tertiary lymphatic structure), resident TAMs can also show different functional phenotypes (105, 112, 113), which means that different CAR-M cells may also need to be customized for different specific regions of tumors in the same organ. Finally, the long-term functional stability of CAR-M cells is under constant threat from highly specialized local signals, such as IL-10 and TGF-β in the TME, as well as from tumor-associated fibroblasts and the extracellular matrix, which may reprogram the phenotype of CAR-M cells and eliminate their antitumor activity (114, 115).
Overcoming these spatial and environmental barriers will require next-generation engineering strategies. Owing to their high permeability and retention effect on solid tumors (116), the recent in vivo editing of macrophages by LNPS may provide new ideas for solving the problems of different “ecotypes” and functional exhaustion of TAMs (117–119). On the basis of the dual delivery system of LNP-mRNA, Yang et al. (120) successfully constructed a bifunctional CAR-M that simultaneously targeted GPC3 and CD24-Siglec-G inhibitory signals. Targeted delivery to HCC-associated macrophages is achieved by the adsorption of specific plasma proteins. In vitro, CAR-M coexpressing GPC3-CAR and Siglec-GΔITIMs significantly increased the phagocytic capacity of GPC3+ Hepa1–6 cells by approximately 10-fold. Moreover, the expression of Siglec-GΔITIM can block the immunosuppressive signal mediated by CD24 and further enhance the antitumor function of CAR-M cells. This effectively avoids the heterogeneity problem of TAMs caused by dynamic changes in the TME, and endogenous CAR-M cells may show stronger natural tolerance to the local TME due to their “local advantage” (121–123).
Zhang et al. (57) used the LNP-mRNA system to edit macrophages in situ in vivo and constructed a quadrivalent CAR-M with GPC3-CAR-Super IL-2 and FAP-CAR-△TGFβRII. The experimental data revealed that after intravenous injection of LNP-GF CAR mRNA, approximately 70% of DID-labeled LNPs were taken up by liver macrophages, confirming its efficient targeting to endogenous macrophages. In vitro experiments further revealed that CAR-M increased the phagocytosis index of GPC3+ Hepa1–6 cells to approximately 11% and almost completely inhibited tumor growth at a high E/T ratio (10:1). More importantly, despite their transient expression, CAR-Ms successfully induced long-term T-cell immune memory through antigen presentation and local interactions with Super IL-2, which not only eliminated GPC3-positive tumors but also neutralized antigen-negative tumors through antigen diffusion, thus effectively preventing immunosuppressive reprogramming of the TME and tumor recurrence. This strategy provides a dual mechanism of immediate killing and long-term protection for the treatment of solid tumors, highlighting the potential of in vivo editing of CAR-M cells to overcome tumor heterogeneity and dynamic changes in the TME.
The LNP-mRNA in vivo delivery strategy offers a novel approach to overcoming TAM heterogeneity and reprogramming the immunosuppressive TME. By directly editing endogenous macrophages in situ across different TME niches, the resulting CAR-M cells—already present within the TME and pre-adapted to local conditions such as hypoxic or perivascular regions—naturally circumvent the “acclimatization” issues often faced by in vitro-modified CAR-M cells during homing and local adaptation. This “in situ transformation” strategy allows CAR-M to cover various functional regions inside the tumor, thereby addressing the challenge posed by spatial and functional heterogeneity of TAMs in achieving uniform therapeutic efficacy.
5.2 Tumor antigen heterogeneity
Conventional CAR cell therapy, which operates on the principle of specific “antigen–antibody” binding, has long faced the challenge of tumor antigen heterogeneity. Its high specificity often leads to immune escape due to antigen loss or downregulation, while on-target/off-tumor toxicity frequently forces the selection of heterogenous target antigens—a persistent therapeutic dilemma (124–126).
CAR-M therapy, however, built on a unique macrophage chassis, has reshaped this landscape. Unlike CAR-T cells, which rapidly induce target cell apoptosis via perforin and granzyme B within hours (127, 128), the core mechanism of CAR-M revolves around the phagosome–lysosome degradation pathway (16). Upon binding of the CAR single-chain variable fragment (scFv) to a tumor antigen, intracellular signaling triggers actin cytoskeleton reorganization, leading to the formation of phagocytic cups that engulf the target cell. The internalized target is enclosed within a phagosome (35), which subsequently fuses with lysosomes. Within the acidic environment of the phagolysosome, tumor cells are thoroughly degraded by hydrolases, reactive oxygen species, and reactive nitrogen species (129). This “root-and-branch” elimination not only physically eradicates cancer cells but also generates abundant tumor antigen fragments. These fragments serve as ideal substrates for efficient cross-presentation to T cells, thereby potently activating the adaptive immune system and establishing a long-term “in situ vaccine” effect (Figure 2).
Figure 2. Differences in anti-tumor mechanisms between CAR-T cells and CAR-M. (A) Anti-tumor mechanisms of CAR-M: (a) Direct phagocytosis and cytotoxicity: CAR-M directly engulfs tumor cells, accompanied by the secretion of reactive oxygen species (ROS) and other cytotoxic factors. (b) Antigen presentation and T-cell activation: promotion of T-cell activation and proliferation through efficient antigen presentation. (c) Remodeling of the immune microenvironment: the secretion of proinflammatory cytokines such as IL-6, IL-12, TNF-α, IFN-γ and other immune cells can activate T cells, NK cells and other immune cells, and induce the polarization of M2 macrophages to M1 macrophages, and finally reshape the tumor immune microenvironment. (d) Efficient directional migration: with the help of highly expressed chemokine receptors (such as CCR2 and CX3CR1), it responds to the concentration gradient of chemokines such as CCL2 and CX3CL1 in the tumor microenvironment to achieve efficient directional invasion. (B) Anti-tumor mechanisms of CAR-T cells: (a) Direct cytotoxicity: by releasing perforin and granzyme B, CAR-T cells directly lyse tumor cells expressing target antigens. (b) Immune activation and bystander effect: it promotes the activation and proliferation of bystander T lymphocytes in itself and the tumor microenvironment by secreting cytokines such as IFN-γ.(c) Antigen-independent killing: tumor heterogeneity can be overcome by inducing apoptosis of tumor cells that do not express target antigens through the Fas/FasL signaling pathway.
In a key study, Klichinsky et al. (16) demonstrated that CAR-M could phagocytose tumor cells expressing NY-ESO-1 antigen but lacking the cognate HLA-A2 molecule. When these CAR-M cells were co-cultured with NY-ESO-1–specific T cells, they robustly activated T cell responses, as indicated by significant upregulation of CD69 and high levels of IFN-γ secretion. This finding confirmed that CAR-M processes phagocytosed tumor antigens and cross-presents them via the MHC-I pathway to CD8+ T cells, effectively initiating tumor-specific T cell immunity. This process lies at the heart of the “in situ vaccine” effect: local immune reprogramming within the tumor stimulates systemic adaptive immunity, establishing durable immune surveillance and protection.
Further reinforcing this concept, Zhang et al. (57) developed a quadravalent CAR-M targeting GPC3 and FAP and elucidated its role as an “in situ vaccine.” Mice cured by this therapy completely rejected subsequent challenges with GPC3-negative tumors—a protective effect abolished upon T cell depletion, indicating dependence on long-term T cell memory. Using an antigen spreading model, the authors showed that after engulfing GPC3+ tumor cells, CAR-M cross-presented the internal model antigen OVA to antigen-specific CD8+ T cells, thereby triggering T cell responses against non-CAR target antigens. This study confirmed that, although CAR-M persistence may be transient, its ability to mediate antigen cross-presentation and local secretion of super IL-12 can establish a relatively sustained “in situ vaccine” effect capable of overcoming antigen heterogeneity—offering a promising strategy for preventing tumor recurrence.
Through this unique mechanism of stimulating systemic antitumor immunity, CAR-M has the potential to fundamentally dismantle the heterogeneous defense network of tumors, thereby enabling more effective control of solid cancers.
5.3 Target-specific challenges and risks of on-target/off-tumor toxicity
While CAR-M therapy holds the theoretical potential to overcome intratumoral antigenic heterogeneity through an “in situ vaccine” effect, its therapeutic window fundamentally relies on the specific recognition of target antigens by the CAR structure—a principle that shares the same risk of “on-target/off-tumor” toxicity observed in CAR-T therapy (125). This risk stems from the fact that many desirable tumor-associated antigens (TAAs) are also expressed, even if at low levels or in a temporally restricted manner, in normal tissues (126). Clinical experience with CAR-T cells has clearly demonstrated this challenge: for instance, ERBB2-targeted CAR-T cells induced fatal pulmonary toxicity due to low-level target expression in the lung (130), while CAIX-directed CAR-T cells caused reversible liver injury by attacking bile duct epithelial cell (131). As a therapeutic approach built on the same targeting logic, researchers in the CAR-M field have explicitly acknowledged from the outset that CAR-M cells targeting TAAs may similarly inflict “on-target, off-tumor” toxicity on normal tissues expressing the same antigen (16).
It should be emphasized, however, that the off-tumor toxicity profile of CAR-M may differ from that of CAR-T. Owing to the potent phagocytic capacity, robust inflammatory secretion, and tissue-resident properties of macrophages, their attack on normal tissues could lead to more widespread and sustained inflammatory damage, potentially driving chronic fibrosis (16, 70). This concern is grounded in solid pathophysiological rationale: persistently activated macrophages are established key drivers of tissue fibrosis, promoting excessive extracellular matrix deposition (132). Furthermore, under conditions of immune hyperactivation, macrophages have been shown to serve as primary effector cells mediating extensive inflammatory tissue injury and subsequent fibrotic remodeling (133). For example, HER2-directed CAR-M cells might not only target HER2+ tumor cells but also affect normal tissues such as the heart where cardiomyocytes exhibit low HER2 expression, lungs, or skin, potentially causing myocarditis, pneumonitis, or dermatitis (130, 134). The injury pattern in such cases may skew more toward disordered tissue remodeling and chronic fibrosis rather than acute necrosis. Additionally, the strong antigen-presentation and immune-activating functions of CAR-M represent a double-edged sword. Theoretically, normal tissue debris phagocytosed by CAR-M could be processed and presented to T cells, thereby activating responses against self-antigens and potentially triggering autoimmune-like adverse effects. While direct evidence remains to be accumulated, the secondary autoimmune phenomena already observed in CAR-T therapy following on-target damage to normal tissues (135) underscore the need for high vigilance regarding this risk.
To strictly confine CAR-M activity in time and space, next-generation engineering strategies are focusing on novel synthetic biology designs and logic-gate optimizations. These include, for example, TME-responsive logic-gate systems or tissue-restricted promoters engineered to silence CAR expression in specific normal tissues (124). Preclinical validation can be enhanced through humanized mouse models that simulate the systemic distribution and organ-specific responses of CAR-M in a humanized immune context (136). Furthermore, patient-derived organoid co-culture systems can directly assess the potential aggressiveness of CAR-M against normal human tissues expressing the target antigen in a near-physiological setting, providing indispensable personalized safety predictions (137). In summary, advancing CAR-M toward successful clinical translation will require a dual focus: maximizing anti-tumor potency while proactively mitigating off-tumor risks through refined and intelligent engineering design.
5.4 Challenges in manufacturing and phenotype standardization
The clinical translation of CAR-M therapy depends not only on innovative receptor design but also faces a more pronounced bottleneck than CAR-T: the fundamental tension between the high intrinsic plasticity of macrophages and the batch-to-batch consistency required for clinical-grade manufacturing (138, 139). Macrophage polarization is not a simple M1/M2 dichotomy but a continuous functional spectrum finely tuned by microenvironmental signals. For instance, GM-CSF primarily drives M1-like polarization with high expression of IL-12 and inducible iNOS, while M-CSF promotes an M2-like phenotype characterized by elevated IL-10 and arginase 1 (139). Single-cell sequencing further reveals that even monocytes from the same donor, cultured under standardized conditions, can differentiate into transcriptionally and functionally heterogeneous subsets, with phagocytic capacity varying by up to threefold across subpopulations (140). This inherent plasticity makes the standardization of CAR-M phenotype and function a multidimensional engineering challenge.
Current production processes are highly sensitive to microenvironmental variables. First, minor fluctuations in medium composition can have significant functional consequences. For example, exosomal miRNAs (such as miR-21) present in different serum lots can modulate macrophage activity via the TLR4/NF-κB pathway, directly affecting inflammatory responsiveness (141). Second, physicochemical parameters play a decisive role. Variations in oxygen tension (5% vs. 21% O2) can reprogram the metabolome of M2 macrophages, altering levels of TCA cycle intermediates, amino acids, and fatty acids (142). Similarly, substrate stiffness (1 kPa vs. 50 kPa) profoundly influences polarization and function through mechanisms such as YAP/TAZ pathway activation (143); Furthermore, conventional surface-marker-based assays (CD86/CD206) are insufficient to capture functional diversity. PD-L1+ macrophages, for example, exhibit distinct immunosuppressive activity in the tumor microenvironment, requiring integrated multi-omics profiling—transcriptomic, metabolomic, and epigenomic—for accurate characterization (144, 145); Finally, source-dependent heterogeneity complicates standardization. Even under identical culture conditions, CD14+ monocytes from healthy donors display significant functional variation (146). While iPSC-derived macrophages offer a theoretically uniform starting cell source, their differentiation—especially from hematopoietic progenitors to monocytes—often suffers from low efficiency (<30%) and poor synchrony, leading to heterogeneous final products (147).
To overcome these barriers, emerging strategies integrate process control, cellular engineering, and intelligent monitoring. Innovations in bioprocessing include dynamic perfusion culture, which enhances M1 phenotype stability by 27% compared to static systems (148), and chemically defined media supplemented with metabolites such as butyrate (0.5 mM) to maintain M2 phenotype for over seven days (149). A more transformative approach lies in synthetic biology-based “hardwiring” of cell function. For instance, overexpression of the master transcription factor MAFB in the THP-1 cell line increased expression of the M2 marker CD163 eightfold, indicating that epigenetic engineering can enable phenotypic locking (149).
Looking forward, standardized CAR-M manufacturing must converge with intelligent, precision-platform technologies. Priority directions include microfluidic single-cell culture systems for real-time, non-invasive tracking of phenotypic drift; machine learning-assisted medium optimization leveraging metabolomic and other omics datasets to predict ideal cytokine and nutrient combinations; and improved targeted cryopreservation methods to minimize post-thaw functional loss. In summary, achieving phenotype standardization in CAR-M necessitates interdisciplinary integration across biology, engineering, and data science. Only through the convergence of precision bioprocessing, deep cellular engineering, and intelligent quality control can this promising cell therapy evolve into a reliable, consistent, and clinically deployable “off-the-shelf” anticancer agent.
6 New engineering strategy and logic-gated optimization for CAR-M
By integrating the innate tumor-homing ability, antigen-presenting function and engineered killing module of macrophages, CAR-M therapy achieves a dual breakthrough in the infiltration of the solid tumor TME and immune remodeling, which represents an innovation in the tumor immunotherapy paradigm (Table 4). However, the clinical translation of this technology still faces multiple challenges: immune remodeling may trigger systemic cytokine release syndrome (CRS) (153), and long-term TME pressure easily induces clonal escape mutations in CAR-M cells, leading to attenuated targeting (60, 153). Moreover, the conventional manufacturing process remains complex, involving isolation, genetic modification, and expansion of cells in vitro before reinfusion—a procedure that is both costly and time-consuming. Therefore, moving beyond the traditional framework of in vitro CAR design, generating or editing CAR-M cells in vivo, and developing responsive logic-gate systems tailored to the TME have emerged as pivotal directions for optimizing CAR-M functionality and enabling its clinical translation (Figure 3).
Table 4. Advances in CAR-M for cancer research.
Figure 3. New engineering strategy and logic-gated optimization for CAR-M. (a) The in vivo in situ generation of CAR-M using LNP technology has emerged as a pivotal strategy for optimizing and advancing the clinical translation of this therapeutic approach. (b) Conventional CAR designs typically target a single tumor marker; however, prolonged exposure to tumor microenvironment (TME) pressures can readily induce clonal escape mutations in CAR-M, leading to a progressive loss of targeting efficacy. The distinct physicochemical properties of the TME provide a molecular foundation for developing novel CAR logic-gated systems: (c) In the acidic TME, incorporating pHLIP into the hinge or transmembrane domain of CAR may enable controlled activation of CAR-M specifically under acidic conditions; (d) whereas in hypoxic regions, leveraging CD147—a protein highly upregulated via HIF-1α stabilization—allows for specific CAR-M activation exclusively within hypoxic microenvironments.
Breaking through the traditional in vitro preparation framework and generating CAR-M cells in vivo has emerged as a key strategy to advance the clinical translation of this therapy. Among various approaches, lipid nanoparticle (LNP)-based delivery systems show particular promise. Billingsley et al. (154) systematically validated the potential of LNPs as efficient delivery vehicles by constructing a library of 24 ionizable lipids and identifying the optimal formulation, C14-4, for delivering mRNA into primary human T cells and monocytes/macrophages, thereby inducing the expression of functional CD19-CAR protein. Building on this foundation, Zhou et al. (48) achieved a critical transition from a general platform to CAR-M-specific therapy. They ingeniously designed LNPs loaded with Trop2-CAR plasmid DNA and successfully reprogrammed macrophages into CAR-M cells in situ in tumor-bearing mice through either intravenous or intratumoral injection. Together, these studies outline a clear developmental trajectory for in vivo CAR-M generation using LNP technology. However, several challenges remain in current LNP research, including how to prevent off-target cell modification due to LNP leakage (48, 155, 156) and how to enhance the targeting efficiency toward specific myeloid cell subsets. Beyond LNPs, emerging strategies involve implantable biological scaffolds loaded with chemokines, cytokines, or genetic materials that can recruit host immune cells and “educate” them in situ to become CAR-M cells, potentially enabling long-term CAR expression in vivo. For instance, Gao’s team (157) successfully generated CAR-M cells in the brain tumor region through local injection of gene nanocapsules, while Liu’s group (158) developed multifunctional nanopolymers that simultaneously enable CAR expression and macrophage phenotype modulation, demonstrating the precision and synergistic potential of in vivo editing strategies. Nevertheless, several aspects of these approaches require further optimization, including the renewal pathway after loaded materials are depleted (159), the biocompatibility of carrier materials, and the controllability of release kinetics (160). These in vivo strategies collectively offer innovative solutions to challenges related to cell sources, heterogeneity, and manufacturing costs in the clinical development of cellular immunotherapies.
While pursuing in situ editing strategies, researchers are also moving beyond the conventional “antigen–antibody” CAR design paradigm by leveraging distinctive features of the tumor metabolic microenvironment to achieve breakthroughs in targeted immunotherapy. The unique physical and chemical properties of the TME provide a molecular basis for the design of new CAR logic gates. In the hypoxic microenvironment, prolyl hydroxylase (PHD) activity is inhibited, leading to the stable accumulation of hypoxia-inducible factor-1α (HIF-1α) and the activation of downstream prosurvival signaling pathways (161). Notably, this pathway upregulates several tumor-specific or highly expressed surface proteins, such as carbonic anhydrase IX and CD147 (162–164). This observation led us to hypothesize that a CAR could be engineered to target HIF-1α-induced surface antigens. By leveraging this strategy, CAR-M cells can utilize their scFv domain to accurately identify and eradicate tumor cells actively surviving in the hypoxic microenvironment, thus conferring highly specific targeting and minimized off-target toxicity; pH-low insertion peptide (pHLIP) is a 35-amino acid pH-sensitive peptide that enables the transport of conjugated nanocarriers into cells to facilitate phiven active targeting (165). Tang et al. (166) coupled pHLIP to the nanodrug delivery system, which significantly improved the efficiency of drug accumulation at the tumor site. Therefore, the integration of pHLIP into the hinge region or transmembrane domain of CAR is expected to achieve the controlled activation of CAR-M in the acidic TME and reduce the risk of CRS caused by nonspecific immune activation. The introduction of artificial intelligence (AI) technology has further expanded the innovation dimension of CAR design. Through high-throughput data mining, deep learning algorithms can predict the spatial conformation and affinity of the CAR domain and optimize the synergistic effect between the antigen recognition module and the costimulatory signal domain.
Notably, CAR-M therapy demonstrates inherent advantages over conventional CAR-T in terms of CRS safety. Studies indicate that CAR-NK cells exhibit a more favorable safety profile with regard to CRS and neurotoxicity; similarly, CAR-M, as an innate immune cell-based platform, has also shown promising safety outcomes (167). Furthermore, logic-gating designs help mitigate CRS risk by raising the activation threshold of CAR-M. Since activation requires the simultaneous fulfillment of multiple conditions, logic-gated CAR-M remain inactive in normal tissues, thereby avoiding unintended immune activation and cytokine release (168, 169). Additionally, preventive strategies against CRS align with and complement the logic-gating concept. For instance, neutralization or genetic inactivation of granulocyte-macrophage colony-stimulating factor (GM-CSF) has been shown to significantly reduce the secretion of monocyte-dependent CRS biomarkers—including MCP-1, IL-6, and IL-8—without compromising the antitumor efficacy of CAR-T cells (170). Similar approaches could be applied to CAR-M therapy to further diminish the risk of CRS.
At present, several clinical trials of CAR-M cells have entered the early clinical validation stage, and the preliminary results support its further development (Table 5). Recently, the world’s first CAR-M therapy (CT-0508, NCT04660929) (103) used the chimeric adenovirus vector AD5F35 to modify macrophages and target the HER2 antigen. Phase I clinical trial data revealed that the number of tumor-infiltrating lymphocytes (TILs) increased approximately 2.1-fold after infusion. The expansion and activation of T-cell clones were significantly enhanced, and no CRS events above grade 3 were observed, which initially verified the safety and feasibility of CAR-M cells and their great advantages for future clinical promotion. In addition, intraperitoneal infusion of autologous HER2-directed CAR-M is planned to evaluate local antitumor effects in patients with HER2-positive advanced gastric cancer and peritoneal metastasis (NCT06224738) (151). Another study (NCT05007379) aimed to systematically evaluate the killing activity of CAR-M cells in breast cancer with different HER2 expression levels by using patient-derived organoid models to provide a basis for individualized treatment.
Table 5. Progress in CAR-M related clinical trials.
In terms of target expansion, a recent preclinical study on pancreatic cancer revealed the great potential of c-MeT-targeted CAR-M (CAR-m-c-Met) (30). Through bioinformatics and tissue microarray analysis, they confirmed that c-MET was highly expressed in pancreatic cancer tissues and was associated with poor prognosis. In vitro and orthotopic mouse pancreatic cancer model experiments revealed that CAR-M-c-MET could specifically recognize and efficiently phagocytose c-MeT-positive tumor cells and secrete GM-CSF and other immune activators to reshape the tumor microenvironment. In addition, no obvious toxicity or side effects were observed in the mouse model, providing solid data support for future clinical research.
Although the field of CAR-M has broad prospects, its development process is also accompanied by strategy adjustment and experience accumulation. For example, earlier studies, such as MY-M11 (NCT03608618), which targets mesothelin (MSLN), and TAK-103 (NCT05164666), developed by Takeda, have been discontinued. An in-depth analysis of these cases can provide a valuable reference for follow-up research. The strategy of MY-M11 using peripheral blood mononuclear cells (PBMCs) to generate CAR-Ms in vitro and undergo intraperitoneal infusion is safe in the early stage but may not yield sufficient therapeutic advantages because of insufficient antitumor efficacy or limited cell persistence (152). The termination of these early explorations reflects the challenges of CAR-M technology in terms of cell preparation, in vivo persistence, and target selection. It also prompts the field to focus resources on more promising candidates, such as CT-0508, and motivates the development of next-generation engineering strategies (e.g., logic gating and in situ generation).
7 Conclusion and perspectives
In the field of cellular immunotherapy, CAR-M, an emerging treatment method, has gradually shown great potential in tumor treatment. By integrating the innate immune characteristics of macrophages with synthetic biological modification strategies, the immunosuppressive TME can be reprogrammed, which provides a new paradigm for the treatment of solid tumors. Current studies have revealed multiple antitumor mechanisms involving the secretion of proinflammatory factors (such as TNF-α and IL-12), the induction of M1 polarization, and the enhancement of T-cell cross-presentation. However, the heterogeneity of cell sources, insufficient persistence in vivo and bottlenecks associated with large-scale production still need to be overcome.
Future research on CAR-M cells is expected to focus on multiple dimensions. 1) Multimodal gene editing technology: CRISPR-Cas9 and base editing technology are used to target the metabolic pathway of CAR-M or integrate the suicide switch (iCasp9) to balance efficacy and safety; 2) microenvironment-responsive intelligent design: develop light-controlled, tissue-responsive, pH- or oxygen concentration-dependent CAR logic gate systems to achieve space-time specific activation; 3) interdisciplinary technology integration: combine single-cell transcriptome sequencing and an AI-driven CAR optimization platform to analyze the dynamic functional phenotype of CAR-M and predict treatment response; 4) clinical translation verification: multicenter clinical trials were conducted to evaluate the universality of CAR-M in heterogeneous tumors, and a standardized production quality control system was established. With the coordinated development of synthetic biology, computational immunology and precision medicine, CAR-Ms are expected to overcome the barriers associated with solid tumor treatment and become the core pillar of the next generation of tumor immunotherapy.
Author contributions
Y-MY: Writing – original draft, Conceptualization, Supervision, Writing – review & editing. Y-FD: Conceptualization, Supervision, Writing – original draft. Y-YH: Conceptualization, Writing – original draft. FF: Funding acquisition, Supervision, Writing – original draft. J-LZ: Writing – original draft, Funding acquisition, Conceptualization, Supervision.
Funding
The author(s) declared that financial support was received for this work and/or its publication. This work was supported by grants from the National Natural Science Foundation of China (82173082 and 82373270), the Natural Science Foundation of Shaanxi (2024SF-ZDCYL-03–28 and 2025SYS-SYSZD-090), and the State Key Laboratory of Holistic Integrative Management of Gastrointestinal Cancers Project (2025GTEP013).
Conflict of interest
The authors declared that this work was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
Generative AI statement
The author(s) declared that generative AI was used in the creation of this manuscript. In the preparation of this document, ChatGTP 4-o was utilized solely to increase readability and refine the language. All the scientific content was carefully reviewed and verified for accuracy by all the authors following each edit, ensuring that the integrity and accuracy of the research were maintained.
Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.
Publisher’s note
All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.
Glossary
ACOD1: Aconitate Decarboxylase 1
ADCP: Antibody-Dependent Cellular Phagocytosis
Ad5F35: Adenovirus 5-Fiber 35
AI: Artificial Intelligence
Arg1: Arginase 1
BMDM: Bone Marrow-Derived Macrophages
CAR: Chimeric Antigen Receptor
CCL2: C-C Motif Chemokine Ligand 2
CCL5: C-C Motif Chemokine Ligand 5
CCL22: C-C Motif Chemokine Ligand 22
CD: Cluster Of Differentiation
CRISPR: Clustered Regularly Interspaced Short Palindromic Repeats
CRS: Cytokine Release Syndrome
CX3CL1: C-X3-C Motif Chemokine Ligand 1
CXCL9: C-X-C Motif Chemokine Ligand 9
CXCL10: C-X-C Motif Chemokine Ligand 10
c-MYC: Myeloid Cytomatosis Oncogene Homolog
ECM: Extracellular Matrix
FcγR: Fcγ
FDA: Food And Drug Administration
GvHD: Graft-Versus-Host Disease
HER2: Human Epidermal Growth Factor Receptor 2
HIF-1α: Hypoxia-Inducible Factor-1α
HLA: Human Leukocyte Antigen
hLT-BMDM: Human Long-Lived Bone Marrow-Derived Macrophages
hPSC: Human Pluripotent Stem Cells
iBMDM: Immortalized Bone Marrow-Derived Macrophages
iCasp9: Inducible Caspase 9
ID3: Inhibitor Of DNA Binding 3
IFN-γ: Interferon-γ
IL-1β: Interleukin 1β
IL-10: Interleukin 10
IL-12: Interleukin 12
iMacs: Induced Macrophages
iNOS: Inducible Nitric Oxide Synthase
iPSCs: Induced Pluripotent Stem Cells
LPS: Lipopolysaccharides
LT-BMDM: Long-Lived Bone Marrow-Derived Macrophages
LNP: Lipid Nanoparticle
Mcl-1: Myeloid Cell Leukemia-1
MHC: Major Histocompatibility Complex
MMPs: Matrix Metalloproteinases
mRNA: Messenger RNA
NADPH: Nicotinamide Adenine Dinucleotide Phosphate
NF-κB: Nuclear Factor Kappa-B
NO: Nitric Oxide
NOX2: NADPH Oxidase 2
PBMC: Peripheral Blood Mononuclear Cells
PD-1: Programmed Death 1
PD-L1: Programmed Cell Death Ligand 1
PHD: Prolyl Hydroxylase
pHLIP: pH-Low Insertion Peptide
PMA: Phorbol-12-Myristate-13-Acetate
PRRs: Pattern Recognition Receptors
RAC1: Ras-Related C3 Botulinum Toxin Substrate 1
ROS: Reactive Oxygen Species
SIRPα: Signal Regulatory Protein α
TAAs: Tumor-Associated Antigens
TAMs: Tumor-Associated Macrophages
TCF1: T-Cell Factor 1
TGF-β: Transforming Growth Factor Beta
THP-1: Human Monocytic Leukemia Cells
TILs: Tumor-Infiltrating Lymphocytes
TIM-3: T-Cell Immunoglobulin Domain And Mucin Domain-3
TLR4: Toll-Like Receptor 4
TME: Tumor Microenvironment
TNF-α: Tumor Necrosis Factor α
Tpex: Precursors Of Exhausted T Cells
Treg: Regulatory T Cells
VD3: 1,25-Dihydroxyvitamin D3
References
1. Mishra AK and Malonia SK. Advancing cellular immunotherapy with macrophages. Life Sci. (2023) 328:121857. doi: 10.1016/j.lfs.2023.121857
2. Ma Y, Zhang L, and Huang X. Genome modification by CRISPR/cas9. FEBS J. (2014) 281:5186–93. doi: 10.1111/febs.13110
3. Wang S-W, Gao C, Zheng Y-M, Yi L, Lu J-C, Huang X-Y, et al. Current applications and future perspective of CRISPR/Cas9 gene editing in cancer. Mol Cancer. (2022) 21:57. doi: 10.1186/s12943-022-01518-8
4. Andrianantoandro E. Manifesting synthetic biology. Trends Biotechnol. (2015) 33:55–6. doi: 10.1016/j.tibtech.2014.12.002
5. Chohan KL, Siegler EL, and Kenderian SS. CAR-T cell therapy: the efficacy and toxicity balance. Curr Hematol Malig Rep. (2023) 18:9–18. doi: 10.1007/s11899-023-00687-7
6. Good CR, Aznar MA, Kuramitsu S, Samareh P, Agarwal S, Donahue G, et al. An NK-like CAR T cell transition in CAR T cell dysfunction. Cell. (2021) 184:6081–6100.e26. doi: 10.1016/j.cell.2021.11.016
7. Xia X, Yang Z, Lu Q, Liu Z, Wang L, Du J, et al. Reshaping the tumor immune microenvironment to improve CAR-T cell-based cancer immunotherapy. Mol Cancer. (2024) 23:175. doi: 10.1186/s12943-024-02079-8
8. Zhang X, Zhu L, Zhang H, Chen S, and Xiao Y. CAR-T cell therapy in hematological Malignancies: current opportunities and challenges. Front Immunol. (2022) 13:927153. doi: 10.3389/fimmu.2022.927153
9. Jing L, An Y, Cai T, Xiang J, Li B, Guo J, et al. A subpopulation of CD146+ macrophages enhances antitumor immunity by activating the NLRP3 inflammasome. Cell Mol Immunol. (2023) 20:908–23. doi: 10.1038/s41423-023-01047-4
10. Cutolo M, Campitiello R, Gotelli E, and Soldano S. The role of M1/M2 macrophage polarization in rheumatoid arthritis synovitis. Front Immunol. (2022) 13:867260. doi: 10.3389/fimmu.2022.867260
11. Shapouri-Moghaddam A, Mohammadian S, Vazini H, Taghadosi M, Esmaeili S-A, Mardani F, et al. Macrophage plasticity, polarization, and function in health and disease. J Cell Physiol. (2018) 233:6425–40. doi: 10.1002/jcp.26429
12. Liu L, Stokes JV, Tan W, and Pruett SB. An optimized flow cytometry panel for classifying macrophage polarization. J Immunol Methods. (2022) 511:113378. doi: 10.1016/j.jim.2022.113378
13. Maalej KM, Merhi M, Inchakalody VP, Mestiri S, Alam M, Maccalli C, et al. CAR-cell therapy in the era of solid tumor treatment: current challenges and emerging therapeutic advances. Mol Cancer. (2023) 22:20. doi: 10.1186/s12943-023-01723-z
14. The complement system and innate immunity - Immunobiology - NCBI Bookshelf. Available online at: https://www.ncbi.nlm.nih.gov/books/NBK27100/ (Accessed September 27, 2025).
15. June CH and Sadelain M. Chimeric antigen receptor therapy. N Engl J Med. (2018) 379:64–73. doi: 10.1056/NEJMra1706169
16. Klichinsky M, Ruella M, Shestova O, Lu XM, Best A, Zeeman M, et al. Human chimeric antigen receptor macrophages for cancer immunotherapy. Nat Biotechnol. (2020) 38:947–53. doi: 10.1038/s41587-020-0462-y
17. Lower EE and Baughman RP. The effect of cancer and chemotherapy on monocyte function. J Clin Lab Immunol. (1990) 31:121–5.
18. Kiss M, Caro AA, Raes G, and Laoui D. Systemic reprogramming of monocytes in cancer. Front Oncol. (2020) 10:1399. doi: 10.3389/fonc.2020.01399
19. Pennisi G, Valeri F, Burattini B, Bruzzaniti P, Sturiale CL, Talacchi A, et al. Targeting macrophages in glioblastoma: current therapies and future directions. Cancers (Basel). (2025) 17:2687. doi: 10.3390/cancers17162687
20. Zhang J, Yin Z, Liang Z, Bai Y, Zhang T, Yang J, et al. Impacts of cryopreservation on phenotype and functionality of mononuclear cells in peripheral blood and ascites. J Transl Int Med. (2024) 12:51–63. doi: 10.2478/jtim-2023-0136
21. Li N, Geng S, Dong Z-Z, Jin Y, Ying H, Li H-W, et al. A new era of cancer immunotherapy: combining revolutionary technologies for enhanced CAR-M therapy. Mol Cancer. (2024) 23:117. doi: 10.1186/s12943-024-02032-9
22. Chen S, Saeed AFUH, Liu Q, Jiang Q, Xu H, Xiao GG, et al. Macrophages in immunoregulation and therapeutics. Signal Transduct Target Ther. (2023) 8:207. doi: 10.1038/s41392-023-01452-1
23. Pierini S, Gabbasov R, Oliveira-Nunes MC, Qureshi R, Worth A, Huang S, et al. Chimeric antigen receptor macrophages (CAR-M) sensitize HER2+ solid tumors to PD1 blockade in pre-clinical models. Nat Commun. (2025) 16:706. doi: 10.1038/s41467-024-55770-1
24. Li M-Y, Ye W, and Luo K-W. Immunotherapies targeting tumor-associated macrophages (TAMs) in cancer. Pharmaceutics. (2024) 16:865. doi: 10.3390/pharmaceutics16070865
25. Yerolatsite M, Torounidou N, Gogadis A, Kapoulitsa F, Ntellas P, Lampri E, et al. TAMs and PD-1 networking in gastric cancer: A review of the literature. Cancers. (2023) 16:196. doi: 10.3390/cancers16010196
26. Liang X, Bai R, Sun Y, Ma P, Lu X, Han T, et al. TAMs: Guardians or foes? Overcoming heterogeneity and delivery barriers for precision immunotherapy. Int Immunopharmacol. (2025) 163:115195. doi: 10.1016/j.intimp.2025.115195
27. Cao Q, Wang Y, Chen J, Wang R, Chen T, Gloss B, et al. Targeting inflammation with chimeric antigen receptor macrophages using a signal switch. Nat BioMed Eng. (2025) 9:1502–16. doi: 10.1038/s41551-025-01387-8
28. Duan Z, Li Z, Wang Z, Chen C, and Luo Y. Chimeric antigen receptor macrophages activated through TLR4 or IFN-γ receptors suppress breast cancer growth by targeting VEGFR2. Cancer Immunol Immunother. (2023) 72:3243–57. doi: 10.1007/s00262-023-03490-8
29. Arora L, Kalia M, and Pal D. Role of macrophages in cancer progression and targeted immunotherapies. Adv Protein Chem Struct Biol. (2023) 135:281–311. doi: 10.1016/bs.apcsb.2022.11.010
30. Zheng H, Yang X, Huang N, Yuan S, Li J, Liu X, et al. Chimeric antigen receptor macrophages targeting c-MET(CAR-M-c-MET) inhibit pancreatic cancer progression and improve cytotoxic chemotherapeutic efficacy. Mol Cancer. (2024) 23:270. doi: 10.1186/s12943-024-02184-8
31. Kong Y, Li J, Zhao X, Wu Y, and Chen L. CAR-T cell therapy: developments, challenges and expanded applications from cancer to autoimmunity. Front Immunol. (2024) 15:1519671. doi: 10.3389/fimmu.2024.1519671
32. Du B, Qin J, Lin B, Zhang J, Li D, and Liu M. CAR-T therapy in solid tumors. Cancer Cell. (2025) 43:665–79. doi: 10.1016/j.ccell.2025.03.019
33. Shen J, Lyu S, Xu Y, Zhang S, Li L, Li J, et al. Activating innate immune responses repolarizes hPSC-derived CAR macrophages to improve anti-tumor activity. Cell Stem Cell. (2024) 31:1003–1019.e9. doi: 10.1016/j.stem.2024.04.012
34. Takata K, Kozaki T, Lee CZW, Thion MS, Otsuka M, Lim S, et al. Induced-pluripotent-stem-cell-derived primitive macrophages provide a platform for modeling tissue-resident macrophage differentiation and function. Immunity. (2017) 47:183–198.e6. doi: 10.1016/j.immuni.2017.06.017
35. Zhang L, Tian L, Dai X, Yu H, Wang J, Lei A, et al. Pluripotent stem cell-derived CAR-macrophage cells with antigen-dependent anti-cancer cell functions. J Hematol Oncol. (2020) 13:153. doi: 10.1186/s13045-020-00983-2
36. Charlesworth CT, Hsu I, Wilkinson AC, and Nakauchi H. Immunological barriers to haematopoietic stem cell gene therapy. Nat Rev Immunol. (2022) 22:719–33. doi: 10.1038/s41577-022-00698-0
37. Poirot L, Philip B, Schiffer-Mannioui C, Le Clerre D, Chion-Sotinel I, Derniame S, et al. Multiplex genome-edited T-cell manufacturing platform for “Off-the-shelf” Adoptive T-cell immunotherapies. Cancer Res. (2015) 75:3853–64. doi: 10.1158/0008-5472.CAN-14-3321
38. Deuse T, Tediashvili G, Hu X, Gravina A, Tamenang A, Wang D, et al. Hypoimmune induced pluripotent stem cell-derived cell therapeutics treat cardiovascular and pulmonary diseases in immunocompetent allogeneic mice. Proc Natl Acad Sci U.S.A. (2021) 118:e2022091118. doi: 10.1073/pnas.2022091118
39. Sloas C, Gill S, and Klichinsky M. Engineered CAR-macrophages as adoptive immunotherapies for solid tumors. Front Immunol. (2021) 12:783305. doi: 10.3389/fimmu.2021.783305
40. Shin MH, Oh E, Kim Y, Nam D-H, Jeon SY, Yu JH, et al. Recent advances in CAR-based solid tumor immunotherapy. Cells. (2023) 12:1606. doi: 10.3390/cells12121606
41. Gatto L, Franceschi E, Di Nunno V, Maggio I, Lodi R, and Brandes AA. Engineered CAR-T and novel CAR-based therapies to fight the immune evasion of glioblastoma: gutta cavat lapidem. Expert Rev Anticancer Ther. (2021) 21:1333–53. doi: 10.1080/14737140.2021.1997599
42. Lei A, Yu H, Lu S, Lu H, Ding X, Tan T, et al. A second-generation M1-polarized CAR macrophage with antitumor efficacy. Nat Immunol. (2024) 25:102–16. doi: 10.1038/s41590-023-01687-8
43. Du F, Ye Z, He A, Yuan J, Su M, Jia Q, et al. An engineered α1β1 integrin-mediated FcγRI signaling component to control enhanced CAR macrophage activation and phagocytosis. J Control Release. (2025) 377:689–703. doi: 10.1016/j.jconrel.2024.11.064
44. Zhang H, Huo Y, Zheng W, Li P, Li H, Zhang L, et al. Silencing of SIRPα enhances the antitumor efficacy of CAR-M in solid tumors. Cell Mol Immunol. (2024) 21:1335–49. doi: 10.1038/s41423-024-01220-3
45. Labanieh L and Mackall CL. CAR immune cells: design principles, resistance and the next generation. Nature. (2023) 614:635–48. doi: 10.1038/s41586-023-05707-3
46. Sadelain M, Brentjens R, and Rivière I. The basic principles of chimeric antigen receptor design. Cancer Discov. (2013) 3:388–98. doi: 10.1158/2159-8290.CD-12-0548
47. Zhang W, Liu L, Su H, Liu Q, Shen J, Dai H, et al. Chimeric antigen receptor macrophage therapy for breast tumours mediated by targeting the tumour extracellular matrix. Br J Cancer. (2019) 121:837–45. doi: 10.1038/s41416-019-0578-3
48. Zhou J, Zhou Z, Wang Z, Sun L, Li F, Tang Y, et al. Lipid nanoparticles produce chimeric antigen receptor macrophages (CAR-M) in situ for the treatment of solid tumors. Nano Today. (2025) 61:102610. doi: 10.1016/j.nantod.2024.102610
49. Noy R and Pollard JW. Tumor-associated macrophages: from mechanisms to therapy. Immunity. (2014) 41:49–61. doi: 10.1016/j.immuni.2014.06.010
50. Kumar S, Mittal S, Gupta P, Singh M, Chaluvally-Raghavan P, and Pradeep S. Metabolic reprogramming in tumor-associated macrophages in the ovarian tumor microenvironment. Cancers (Basel). (2022) 14:5224. doi: 10.3390/cancers14215224
51. Kundu M, Butti R, Panda VK, Malhotra D, Das S, Mitra T, et al. Modulation of the tumor microenvironment and mechanism of immunotherapy-based drug resistance in breast cancer. Mol Cancer. (2024) 23:92. doi: 10.1186/s12943-024-01990-4
52. Kang M, Lee SH, Kwon M, Byun J, Kim D, Kim C, et al. Nanocomplex-mediated in vivo programming to chimeric antigen receptor-M1 macrophages for cancer therapy. Adv Mater. (2021) 33:e2103258. doi: 10.1002/adma.202103258
53. Morrissey MA, Williamson AP, Steinbach AM, Roberts EW, Kern N, Headley MB, et al. Chimeric antigen receptors that trigger phagocytosis. Elife. (2018) 7:e36688. doi: 10.7554/eLife.36688
54. Lanier LL. DAP10- and DAP12-associated receptors in innate immunity. Immunol Rev. (2009) 227:150–60. doi: 10.1111/j.1600-065X.2008.00720.x
55. Rothlin CV, Carrera-Silva EA, Bosurgi L, and Ghosh S. TAM receptor signaling in immune homeostasis. Annu Rev Immunol. (2015) 33:355–91. doi: 10.1146/annurev-immunol-032414-112103
56. Elewaut A, Estivill G, Bayerl F, Castillon L, Novatchkova M, Pottendorfer E, et al. Cancer cells impair monocyte-mediated T cell stimulation to evade immunity. Nature. (2025) 637:716–25. doi: 10.1038/s41586-024-08257-4
57. Zhang S, Lu H, Zhang H, Ding X, Wu Q, Lei A, et al. In vivo engineered quadrivalent CAR-macrophages break tumor barrier, remodel TME and overcome tumor heterogeneity through antigen spreading. (2024). doi: 10.1101/2024.11.06.621821
58. Lee H-W, Choi H-J, Ha S-J, Lee K-T, and Kwon Y-G. Recruitment of monocytes/macrophages in different tumor microenvironments. Biochim Biophys Acta. (2013) 1835:170–9. doi: 10.1016/j.bbcan.2012.12.007
59. Xu M, Wang Y, Xia R, Wei Y, and Wei X. Role of the CCL2-CCR2 signalling axis in cancer: Mechanisms and therapeutic targeting. Cell Prolif. (2021) 54:e13115. doi: 10.1111/cpr.13115
60. Chen K, Liu M-L, Wang J-C, and Fang S. CAR-macrophage versus CAR-T for solid tumors: The race between a rising star and a superstar. Biomol BioMed. (2024) 24:465–76. doi: 10.17305/bb.2023.9675
61. Mantovani A, Marchesi F, Malesci A, Laghi L, and Allavena P. Tumour-associated macrophages as treatment targets in oncology. Nat Rev Clin Oncol. (2017) 14:399–416. doi: 10.1038/nrclinonc.2016.217
62. Peranzoni E, Lemoine J, Vimeux L, Feuillet V, Barrin S, Kantari-Mimoun C, et al. Macrophages impede CD8 T cells from reaching tumor cells and limit the efficacy of anti-PD-1 treatment. Proc Natl Acad Sci U.S.A. (2018) 115:E4041–50. doi: 10.1073/pnas.1720948115
63. Brudno JN, Maus MV, and Hinrichs CS. CAR T cells and T-cell therapies for cancer: A translational science review. JAMA. (2024) 332:1924–35. doi: 10.1001/jama.2024.19462
64. Gu K, Liang T, Hu L, Zhao Y, Ying W, Zhang M, et al.Intraperitoneal programming of tailored CAR macrophages via mRNA-LNP to boost cancer immunotherapy. bioRxiv. (2025). doi: 10.1101/2024.07.30.605730v1
65. Amoozgar B, Bangolo A, Mansour C, Elias D, Mohamed A, Thor DC, et al. Engineering innate immunity: recent advances and future directions for CAR-NK and CAR-macrophage therapies in solid tumors. Cancers (Basel). (2025) 17:2397. doi: 10.3390/cancers17142397
66. Wei D, Wang L, Liu Y, Zuo X, Shen X, and Bresalier RS. CAR-macrophage cell therapy: A new era of hope for pancreatic cancer. Clin Cancer Res. (2025) 31:4018–31. doi: 10.1158/1078-0432.CCR-25-1201
67. Beltraminelli T and De Palma M. Biology and therapeutic targeting of tumour-associated macrophages. J Pathol. (2020) 250:573–92. doi: 10.1002/path.5403
68. Ginhoux F and Guilliams M. Tissue-resident macrophage ontogeny and homeostasis. Immunity. (2016) 44:439–49. doi: 10.1016/j.immuni.2016.02.024
69. Porter DL, Hwang W-T, Frey NV, Lacey SF, Shaw PA, Loren AW, et al. Chimeric antigen receptor T cells persist and induce sustained remissions in relapsed refractory chronic lymphocytic leukemia. Sci Transl Med. (2015) 7:303ra139. doi: 10.1126/scitranslmed.aac5415
70. Murray PJ and Wynn TA. Protective and pathogenic functions of macrophage subsets. Nat Rev Immunol. (2011) 11:723–37. doi: 10.1038/nri3073
71. Long AH, Haso WM, Shern JF, Wanhainen KM, Murgai M, Ingaramo M, et al. 4-1BB costimulation ameliorates T cell exhaustion induced by tonic signaling of chimeric antigen receptors. Nat Med. (2015) 21:581–90. doi: 10.1038/nm.3838
72. Ahn E, Araki K, Hashimoto M, Li W, Riley JL, Cheung J, et al. Role of PD-1 during effector CD8 T cell differentiation. Proc Natl Acad Sci U.S.A. (2018) 115:4749–54. doi: 10.1073/pnas.1718217115
73. Wherry EJ and Kurachi M. Molecular and cellular insights into T cell exhaustion. Nat Rev Immunol. (2015) 15:486–99. doi: 10.1038/nri3862
74. Ruffell B, Chang-Strachan D, Chan V, Rosenbusch A, Ho CMT, Pryer N, et al. Macrophage IL-10 blocks CD8+ T cell-dependent responses to chemotherapy by suppressing IL-12 expression in intratumoral dendritic cells. Cancer Cell. (2014) 26:623–37. doi: 10.1016/j.ccell.2014.09.006
75. Clinical trial: NCT03016377 - my cancer genome. Available online at: https://www.mycancergenome.org/content/clinical_trials/NCT03016377/ (Accessed September 28, 2025).
76. Di Stasi A, Tey S-K, Dotti G, Fujita Y, Kennedy-Nasser A, Martinez C, et al. Inducible apoptosis as a safety switch for adoptive cell therapy. N Engl J Med. (2011) 365:1673–83. doi: 10.1056/NEJMoa1106152
77. Chinn HK, Gardell JL, Matsumoto LR, Labadie KP, Mihailovic TN, Lieberman NAP, et al. Hypoxia-inducible lentiviral gene expression in engineered human macrophages. J Immunother Cancer. (2022) 10:e003770. doi: 10.1136/jitc-2021-003770
78. Ma K, Cheng H, and Zhang L. cBAF and MYC: decision makers for T cell memory differentiation. Signal Transduction Targeted Ther. (2022) 7:389. doi: 10.1038/s41392-022-01238-x
79. Stine ZE, Walton ZE, Altman BJ, Hsieh AL, and Dang CV. MYC, metabolism, and cancer. Cancer Discov. (2015) 5:1024–39. doi: 10.1158/2159-8290.CD-15-0507
80. Liu XS, Wu H, Ji X, Stelzer Y, Wu X, Czauderna S, et al. Editing DNA methylation in the mammalian genome. Cell. (2016) 167:233–247.e17. doi: 10.1016/j.cell.2016.08.056
81. Konermann S, Brigham MD, Trevino AE, Joung J, Abudayyeh OO, Barcena C, et al. Genome-scale transcriptional activation by an engineered CRISPR-Cas9 complex. Nature. (2015) 517:583–8. doi: 10.1038/nature14136
82. Ball M, Kremp M, Qureshi R, Sonawane P, Schmierer M, VanDuzer J, et al. 315 Characterization of CT-0508, an anti-HER2 chimeric antigen receptor macrophage (CAR-M), manufactured from patients enrolled in the phase 1, first in human, clinical trial of CT-0508. J Immunother Cancer. (2022) 10:A315. doi: 10.1136/jitc-2022-SITC2022.0315
83. Weber EW, Lynn RC, Sotillo E, Lattin J, Xu P, and Mackall CL. Pharmacologic control of CAR-T cell function using dasatinib. Blood Adv. (2019) 3:711–7. doi: 10.1182/bloodadvances.2018028720
84. Wu C-Y, Roybal KT, Puchner EM, Onuffer J, and Lim WA. Remote control of therapeutic T cells through a small molecule-gated chimeric receptor. Science. (2015) 350:aab4077. doi: 10.1126/science.aab4077
85. Sakemura R, Terakura S, Watanabe K, Julamanee J, Takagi E, Miyao K, et al. A tet-on inducible system for controlling CD19-chimeric antigen receptor expression upon drug administration. Cancer Immunol Res. (2016) 4:658–68. doi: 10.1158/2326-6066.CIR-16-0043
86. Markley JC and Sadelain M. IL-7 and IL-21 are superior to IL-2 and IL-15 in promoting human T cell-mediated rejection of systemic lymphoma in immunodeficient mice. Blood. (2010) 115:3508–19. doi: 10.1182/blood-2009-09-241398
87. Pellegrini M, Calzascia T, Toe JG, Preston SP, Lin AE, Elford AR, et al. IL-7 engages multiple mechanisms to overcome chronic viral infection and limit organ pathology. Cell. (2011) 144:601–13. doi: 10.1016/j.cell.2011.01.011
88. Fu Y, Foden JA, Khayter C, Maeder ML, Reyon D, Joung JK, et al. High-frequency off-target mutagenesis induced by CRISPR-Cas nucleases in human cells. Nat Biotechnol. (2013) 31:822–6. doi: 10.1038/nbt.2623
89. Saengboon S, Ciurea S, Popat U, Ramdial J, Bashir Q, Alousi A, et al. Long-term outcomes after haploidentical stem cell transplantation for hematologic Malignancies. Blood Adv. (2024) 8:3237–45. doi: 10.1182/bloodadvances.2023010625
90. Xie X, Li X, Liu G, Zhao H, Zhou Z, and Xiong S. T cells expressing a HER2-specific chimeric antigen receptor as treatment for breast cancer. J Cancer Res Clin Oncol. (2023) 149:11561–70. doi: 10.1007/s00432-023-04996-5
91. Abraham-Miranda J, Menges M, Atkins R, Mattie M, Kanska J, Turner J, et al. CAR-T manufactured from frozen PBMC yield efficient function with prolonged in vitro production. Front Immunol. (2022) 13:1007042. doi: 10.3389/fimmu.2022.1007042
92. Yi P-C, Zhuo L, Lin J, Chang C, Goddard A, and Yoon OK. Impact of delayed PBMC processing on functional and genomic assays. J Immunol Methods. (2023) 519:113514. doi: 10.1016/j.jim.2023.113514
93. Tsuchiya S, Yamabe M, Yamaguchi Y, Kobayashi Y, Konno T, and Tada K. Establishment and characterization of a human acute monocytic leukemia cell line (THP-1). Int J Cancer. (1980) 26:171–6. doi: 10.1002/ijc.2910260208
94. Daigneault M, Preston JA, Marriott HM, Whyte MKB, and Dockrell DH. The identification of markers of macrophage differentiation in PMA-stimulated THP-1 cells and monocyte-derived macrophages. PloS One. (2010) 5:e8668. doi: 10.1371/journal.pone.0008668
95. Yang B, Wang X, Wei X, and Ma J. Development of a novel HER2-CAR monocyte cell therapy with controllable proliferation and enhanced anti-tumor efficacy. Chin Med J (Engl). (2024) 137:2590–602. doi: 10.1097/CM9.0000000000002944
96. Chanput W, Mes JJ, and Wichers HJ. THP-1 cell line: an in vitro cell model for immune modulation approach. Int Immunopharmacol. (2014) 23:37–45. doi: 10.1016/j.intimp.2014.08.002
97. Liu Q, Li J, Zheng H, Yang S, Hua Y, Huang N, et al. Adoptive cellular immunotherapy for solid neoplasms beyond CAR-T. Mol Cancer. (2023) 22:28. doi: 10.1186/s12943-023-01735-9
98. Schildberger A, Rossmanith E, Eichhorn T, Strassl K, and Weber V. Monocytes, peripheral blood mononuclear cells, and THP-1 cells exhibit different cytokine expression patterns following stimulation with lipopolysaccharide. Mediators Inflammation. (2013) 2013:697972. doi: 10.1155/2013/697972
99. van Wilgenburg B, Browne C, Vowles J, and Cowley SA. Efficient, long term production of monocyte-derived macrophages from human pluripotent stem cells under partly-defined and fully-defined conditions. PloS One. (2013) 8:e71098. doi: 10.1371/journal.pone.0071098
100. Wang X, Su S, Zhu Y, Cheng X, Cheng C, Chen L, et al. Metabolic Reprogramming via ACOD1 depletion enhances function of human induced pluripotent stem cell-derived CAR-macrophages in solid tumors. Nat Commun. (2023) 14:5778. doi: 10.1038/s41467-023-41470-9
101. Blasi E, Mathieson BJ, Varesio L, Cleveland JL, Borchert PA, and Rapp UR. Selective immortalization of murine macrophages from fresh bone marrow by a raf/myc recombinant murine retrovirus. Nature. (1985) 318:667–70. doi: 10.1038/318667a0
102. Xie D-K, Yao J, Li P-H, Zhu Y-W, Chen J-N, Cao X-L, et al. Phenotypic comparison and the potential antitumor function of immortalized bone marrow-derived macrophages (iBMDMs). Front Immunol. (2024) 15:1379853. doi: 10.3389/fimmu.2024.1379853
103. Reiss KA, Angelos MG, Dees EC, Yuan Y, Ueno NT, Pohlmann PR, et al. CAR-macrophage therapy for HER2-overexpressing advanced solid tumors: a phase 1 trial. Nat Med. (2025) 31:1171–82. doi: 10.1038/s41591-025-03495-z
104. Wang M, Qin Z, Bian X-W, and Shi Y. Harnessing chimeric antigen receptor macrophages against solid tumors. Cancer Commun (Lond). (2025) 45:1344–66. doi: 10.1002/cac2.70053
105. Wu SZ, Al-Eryani G, Roden DL, Junankar S, Harvey K, Andersson A, et al. A single-cell and spatially resolved atlas of human breast cancers. Nat Genet. (2021) 53:1334–47. doi: 10.1038/s41588-021-00911-1
106. Ji AL, Rubin AJ, Thrane K, Jiang S, Reynolds DL, Meyers RM, et al. Multimodal analysis of composition and spatial architecture in human squamous cell carcinoma. Cell. (2020) 182:497–514.e22. doi: 10.1016/j.cell.2020.05.039
107. Che L-H, Liu J-W, Huo J-P, Luo R, Xu R-M, He C, et al. A single-cell atlas of liver metastases of colorectal cancer reveals reprogramming of the tumor microenvironment in response to preoperative chemotherapy. Cell Discov. (2021) 7:80. doi: 10.1038/s41421-021-00312-y
108. Saqib J, Park B, Jin Y, Seo J, Mo J, and Kim J. Identification of niche-specific gene signatures between Malignant tumor microenvironments by integrating single cell and spatial transcriptomics data. Genes (Basel). (2023) 14:2033. doi: 10.3390/genes14112033
109. Zhou Z, Hu Y, Wu Y, Qi Q, Wang J, Chen L, et al. The immunosuppressive tumor microenvironment in hepatocellular carcinoma-current situation and outlook. Mol Immunol. (2022) 151:218–30. doi: 10.1016/j.molimm.2022.09.010
110. Zhou Y, Qian M, Li J, Ruan L, Wang Y, Cai C, et al. The role of tumor-associated macrophages in lung cancer: From mechanism to small molecule therapy. BioMed Pharmacother. (2024) 170:116014. doi: 10.1016/j.biopha.2023.116014
111. Quail DF and Joyce JA. The microenvironmental landscape of brain tumors. Cancer Cell. (2017) 31:326–41. doi: 10.1016/j.ccell.2017.02.009
112. Trehan R, Huang P, Zhu XB, Wang X, Soliman M, Strepay D, et al. SPP1 + macrophages cause exhaustion of tumor-specific T cells in liver metastases. Nat Commun. (2025) 16:4242. doi: 10.1038/s41467-025-59529-0
113. Cabrita R, Lauss M, Sanna A, Donia M, Skaarup Larsen M, Mitra S, et al. Tertiary lymphoid structures improve immunotherapy and survival in melanoma. Nature. (2020) 577:561–5. doi: 10.1038/s41586-019-1914-8
114. Peng C, Xu Y, Wu J, Wu D, Zhou L, and Xia X. TME-related biomimetic strategies against cancer. Int J Nanomedicine. (2024) 19:109–35. doi: 10.2147/IJN.S441135
115. Tiwari A, Trivedi R, and Lin S-Y. Tumor microenvironment: barrier or opportunity towards effective cancer therapy. J BioMed Sci. (2022) 29:83. doi: 10.1186/s12929-022-00866-3
116. Han Y, Lu H, Gu Z, Guan P, and Liu Z. In situ reprogramming of tumor-associated macrophages enhances cancer immunotherapy. Nano Today. (2024) 58:102420. doi: 10.1016/j.nantod.2024.102420
117. Kiaie SH, Majidi Zolbanin N, Ahmadi A, Bagherifar R, Valizadeh H, Kashanchi F, et al. Recent advances in mRNA-LNP therapeutics: immunological and pharmacological aspects. J Nanobiotechnology. (2022) 20:276. doi: 10.1186/s12951-022-01478-7
118. Back PI, Yu M, Modaresahmadi S, Hajimirzaei S, Zhang Q, Islam MR, et al. Immune implications of cholesterol-containing lipid nanoparticles. ACS Nano. (2024) 18:28480–501. doi: 10.1021/acsnano.4c06369
119. Pan X, Zhang Y-W-Q, Dai C, Zhang J, Zhang M, and Chen X. Applications of mRNA delivery in cancer immunotherapy. Int J Nanomedicine. (2025) 20:3339–61. doi: 10.2147/IJN.S500520
120. Yang Z, Liu Y, Zhao K, Jing W, Gao L, Dong X, et al. Dual mRNA co-delivery for in situ generation of phagocytosis-enhanced CAR macrophages augments hepatocellular carcinoma immunotherapy. J Control Release. (2023) 360:718–33. doi: 10.1016/j.jconrel.2023.07.021
121. Yakupova EI, Maleev GV, Krivtsov AV, and Plotnikov EY. Macrophage polarization in hypoxia and ischemia/reperfusion: Insights into the role of energetic metabolism. Exp Biol Med (Maywood). (2022) 247:958–71. doi: 10.1177/15353702221080130
122. Boutilier AJ and Elsawa SF. Macrophage polarization states in the tumor microenvironment. Int J Mol Sci. (2021) 22:6995. doi: 10.3390/ijms22136995
123. Miar A, Arnaiz E, Bridges E, Beedie S, Cribbs AP, Downes DJ, et al. Hypoxia induces transcriptional and translational downregulation of the type I IFN pathway in multiple cancer cell types. Cancer Res. (2020) 80:5245–56. doi: 10.1158/0008-5472.CAN-19-2306
124. Sterner RC and Sterner RM. CAR-T cell therapy: current limitations and potential strategies. Blood Cancer J. (2021) 11:69. doi: 10.1038/s41408-021-00459-7
125. O’Rourke DM, Nasrallah MP, Desai A, Melenhorst JJ, Mansfield K, Morrissette JJD, et al. A single dose of peripherally infused EGFRvIII-directed CAR T cells mediates antigen loss and induces adaptive resistance in patients with recurrent glioblastoma. Sci Transl Med. (2017) 9:eaaa0984. doi: 10.1126/scitranslmed.aaa0984
126. Labanieh L, Majzner RG, and Mackall CL. Programming CAR-T cells to kill cancer. Nat BioMed Eng. (2018) 2:377–91. doi: 10.1038/s41551-018-0235-9
127. D’Aloia MM, Zizzari IG, Sacchetti B, Pierelli L, and Alimandi M. CAR-T cells: the long and winding road to solid tumors. Cell Death Dis. (2018) 9:282. doi: 10.1038/s41419-018-0278-6
128. Stinchcombe JC, Bossi G, Booth S, and Griffiths GM. The immunological synapse of CTL contains a secretory domain and membrane bridges. Immunity. (2001) 15:751–61. doi: 10.1016/s1074-7613(01)00234-5
130. Morgan RA, Yang JC, Kitano M, Dudley ME, Laurencot CM, and Rosenberg SA. Case report of a serious adverse event following the administration of T cells transduced with a chimeric antigen receptor recognizing ERBB2. Mol Ther. (2010) 18:843–51. doi: 10.1038/mt.2010.24
131. Lamers CH, Sleijfer S, van Steenbergen S, van Elzakker P, van Krimpen B, Groot C, et al. Treatment of metastatic renal cell carcinoma with CAIX CAR-engineered T cells: clinical evaluation and management of on-target toxicity. Mol Ther. (2013) 21:904–12. doi: 10.1038/mt.2013.17
132. Misharin AV, Morales-Nebreda L, Reyfman PA, Cuda CM, Walter JM, McQuattie-Pimentel AC, et al. Monocyte-derived alveolar macrophages drive lung fibrosis and persist in the lung over the life span. J Exp Med. (2017) 214:2387–404. doi: 10.1084/jem.20162152
133. Channappanavar R and Perlman S. Pathogenic human coronavirus infections: causes and consequences of cytokine storm and immunopathology. Semin Immunopathol. (2017) 39:529–39. doi: 10.1007/s00281-017-0629-x
134. Iqbal N and Iqbal N. Human epidermal growth factor receptor 2 (HER2) in cancers: overexpression and therapeutic implications. Mol Biol Int. (2014) 2014:852748. doi: 10.1155/2014/852748
135. Larson RC and Maus MV. Recent advances and discoveries in the mechanisms and functions of CAR T cells. Nat Rev Cancer. (2021) 21:145–61. doi: 10.1038/s41568-020-00323-z
136. Wen H, Qu Z, Yan Y, Pu C, Wang C, Jiang H, et al. Preclinical safety evaluation of chimeric antigen receptor-modified T cells against CD19 in NSG mice. Ann Transl Med. (2019) 7:735. doi: 10.21037/atm.2019.12.03
137. Logun M, Wang X, Sun Y, Bagley SJ, Li N, Desai A, et al. Patient-derived glioblastoma organoids as real-time avatars for assessing responses to clinical CAR-T cell therapy. Cell Stem Cell. (2025) 32:181–190.e4. doi: 10.1016/j.stem.2024.11.010
138. Saeed S, Quintin J, Kerstens HHD, Rao NA, Aghajanirefah A, Matarese F, et al. Epigenetic programming of monocyte-to-macrophage differentiation and trained innate immunity. Science. (2014) 345:1251086. doi: 10.1126/science.1251086
139. Kelly B and O’Neill LAJ. Metabolic reprogramming in macrophages and dendritic cells in innate immunity. Cell Res. (2015) 25:771–84. doi: 10.1038/cr.2015.68
140. Ni D, Zhou H, Wang P, Xu F, and Li C. Visualizing macrophage phenotypes and polarization in diseases: from biomarkers to molecular probes. Phenomics. (2023) 3:613–38. doi: 10.1007/s43657-023-00129-7
141. Yang Z, Wu H, Wang Z, Bian E, and Zhao B. The role and application of small extracellular vesicles in glioma. Cancer Cell Int. (2024) 24:229. doi: 10.1186/s12935-024-03389-z
142. Huang SC-C, Smith AM, Everts B, Colonna M, Pearce EL, Schilling JD, et al. Metabolic reprogramming mediated by the mTORC2-IRF4 signaling axis is essential for macrophage alternative activation. Immunity. (2016) 45:817–30. doi: 10.1016/j.immuni.2016.09.016
143. Gong D, Zhen P, and Wang R. Role of macrophage polarization in periprosthetic osteolysis: its research status and advance. Chin J Tissue Eng Res. (2018) 22:1120–5. doi: 10.3969/j.issn.2095-4344.0127
144. Azizi E, Carr AJ, Plitas G, Cornish AE, Konopacki C, Prabhakaran S, et al. Single-cell map of diverse immune phenotypes in the breast tumor microenvironment. Cell. (2018) 174:1293–1308.e36. doi: 10.1016/j.cell.2018.05.060
145. Gordon SR, Maute RL, Dulken BW, Hutter G, George BM, McCracken MN, et al. PD-1 expression by tumour-associated macrophages inhibits phagocytosis and tumour immunity. Nature. (2017) 545:495–9. doi: 10.1038/nature22396
146. Van den Bossche J, O’Neill LA, and Menon D. Macrophage immunometabolism: where are we (Going)? Trends Immunol. (2017) 38:395–406. doi: 10.1016/j.it.2017.03.001
147. Liu Y, Tan H, Dai J, Lin J, Zhao K, Hu H, et al. Targeting macrophages in cancer immunotherapy: Frontiers and challenges. J Adv Res. (2025) 76:695–713. doi: 10.1016/j.jare.2024.12.043
148. Zheng RB and Xiao Y. Research progress of CAR-macrophages in Malignant hematological diseases. Zhonghua Xue Ye Xue Za Zhi. (2024) 45:413–6. doi: 10.3760/cma.j.cn121090-20231103-00248
149. Scheurlen KM, Snook DL, Littlefield AB, George JB, Parks MA, Beal RJ, et al. Anti-inflammatory mechanisms in cancer research: Characterization of a distinct M2-like macrophage model derived from the THP-1 cell line. Cancer Med. (2023) 12:21172–87. doi: 10.1002/cam4.6681
150. Gu K, Liang T, Hu L, Zhao Y, Ying W, Zhang M, et al. Intraperitoneal programming of tailored CAR macrophages via mRNA-LNP to boost cancer immunotherapy. Nature Communications (2025). doi: 10.1101/2024.07.30.605730
151. Efficacy evaluation of chimeric antigen receptor-modified human peritoneal macrophages in the treatment of gastric cancer. Available online at (Accessed September 25, 2025).
152. Annunziata C, Dansky-Ullmann C, Ghobadi A, Weng D, Vanas J, Ekwede I, et al. 1308TiP - A phase I study of intraperitoneal MCY-M11 Anti-mesothelin CAR for women with platinum resistant high grade serous adenocarcinoma of the ovary, primary peritoneum, or fallopian tube, or subjects with peritoneal mesothelioma with recurrence after prior chemotherapy. Ann Oncol. (2019) 30:v531. doi: 10.1093/annonc/mdz253.133
153. Chen Y, Yu Z, Tan X, Jiang H, Xu Z, Fang Y, et al. CAR-macrophage: A new immunotherapy candidate against solid tumors. BioMed Pharmacother. (2021) 139:111605. doi: 10.1016/j.biopha.2021.111605
154. Billingsley MM, Singh N, Ravikumar P, Zhang R, June CH, and Mitchell MJ. Ionizable lipid nanoparticle-mediated mRNA delivery for human CAR T cell engineering. Nano Lett. (2020) 20:1578–89. doi: 10.1021/acs.nanolett.9b04246
155. Cheng Q, Wei T, Farbiak L, Johnson LT, Dilliard SA, and Siegwart DJ. Selective organ targeting (SORT) nanoparticles for tissue-specific mRNA delivery and CRISPR-Cas gene editing. Nat Nanotechnol. (2020) 15:313–20. doi: 10.1038/s41565-020-0669-6
156. Akinc A, Maier MA, Manoharan M, Fitzgerald K, Jayaraman M, Barros S, et al. The Onpattro story and the clinical translation of nanomedicines containing nucleic acid-based drugs. Nat Nanotechnol. (2019) 14:1084–7. doi: 10.1038/s41565-019-0591-y
157. Gao L, Shi C, Yang Z, Jing W, Han M, Zhang J, et al. Convection-enhanced delivery of nanoencapsulated gene locoregionally yielding ErbB2/Her2-specific CAR-macrophages for brainstem glioma immunotherapy. J Nanobiotechnology. (2023) 21:56. doi: 10.1186/s12951-023-01810-9
158. Liu Y, Jing W, Zhang J, Chen C, Gao L, Shi C, et al. In situ MUC1-specific CAR engineering of tumor-supportive macrophages stimulates tumoricidal immunity against pancreatic adenocarcinoma. Nano Today. (2023) 49:101805. doi: 10.1016/j.nantod.2023.101805
159. Nie S, Qin Y, Ou L, Chen X, and Li L. In situ reprogramming of immune cells using synthetic nanomaterials. Adv Mater. (2024) 36:e2310168. doi: 10.1002/adma.202310168
160. Chen J, Zheng H, Zheng S, Zhang S, Xiao X, Chen S, et al. Near-infrared light-triggered drug delivery system based on aminated TiO2 nanotube arrays functionalized with black phosphorus composite hydrogel. J Drug Delivery Sci Technol. (2025) 107:106786. doi: 10.1016/j.jddst.2025.106786
161. Dai W, Guo R, Na X, Jiang S, Liang J, Guo C, et al. Hypoxia and the endometrium: An indispensable role for HIF-1α as therapeutic strategies. Redox Biol. (2024) 73:103205. doi: 10.1016/j.redox.2024.103205
162. Chen M, Liu Z, Zheng K, Hu C, and Peng P. The potential mechanism of HIF-1α and CD147 in the development of triple-negative breast cancer. Med (Baltimore). (2024) 103:e38434. doi: 10.1097/MD.0000000000038434
163. Cígerová V, Adamkov M, Grendár M, and Mešťanová V. Immunohistochemical expression and association of hypoxia-inducible factor 1a and carbonic anhydrase IX in colorectal cancer. Pol J Pathol. (2025) 76:16–24. doi: 10.5114/pjp.2025.149408
164. Hsin M-C, Hsieh Y-H, Hsiao Y-H, Chen P-N, Wang P-H, and Yang S-F. Carbonic anhydrase IX promotes human cervical cancer cell motility by regulating PFKFB4 expression. Cancers (Basel). (2021) 13:1174. doi: 10.3390/cancers13051174
165. Moshnikova A, DuPont M, Visca H, Engelman DM, Andreev OA, and Reshetnyak YK. Eradication of tumors and development of anti-cancer immunity using STINGa targeted by pHLIP. Front Oncol. (2022) 12:1023959. doi: 10.3389/fonc.2022.1023959
166. Tong F, Hu H, Xu Y, Zhou Y, Xie R, Lei T, et al. Hollow copper sulfide nanoparticles carrying ISRIB for the sensitized photothermal therapy of breast cancer and brain metastases through inhibiting stress granule formation and reprogramming tumor-associated macrophages. Acta Pharm Sin B. (2023) 13:3471–88. doi: 10.1016/j.apsb.2022.11.003
167. Zhang Y, Hu R, Xie X, and Li Y. Expanding the frontier of CAR therapy: comparative insights into CAR-T, CAR-NK, CAR-M, and CAR-DC approaches. Ann Hematol. (2025) 104:4305–17. doi: 10.1007/s00277-025-06538-0
168. Jiang C, Zhao Z, East AK, Bandyopadhyay S, Jiang Z, and Chan J. Logic-gated approach for targeted delivery and site-selective activation of photothermal agents in precision cancer treatment. Chem Sci. (2025) 16:5155–65. doi: 10.1039/d4sc08228a
169. Boucher JC and Davila ML. Chimeric antigen receptor design today and tomorrow. Cancer J. (2021) 27:92–7. doi: 10.1097/PPO.0000000000000514
Keywords: cancer immunotherapy, chimeric antigen receptor, logic gate, macrophage, synthetic biology
Citation: Yang Y-M, Ding Y-F, Hu Y-Y, Fan F and Zhao J-L (2026) CAR-M therapy in the era of tumor immunotherapy: current research progress and engineering strategies. Front. Immunol. 16:1723270. doi: 10.3389/fimmu.2025.1723270
Received: 12 October 2025; Accepted: 25 December 2025; Revised: 17 December 2025;
Published: 14 January 2026.
Edited by:
Alejandro Villagra, Georgetown University, United StatesReviewed by:
Shadan Modaresahmadi, Texas Tech University Health Sciences Center, United StatesYi-Lun Wang, Linkou Chang Gung Memorial Hospital, Taiwan
Copyright © 2026 Yang, Ding, Hu, Fan and Zhao. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
*Correspondence: Jun-Long Zhao, YmlvLWp1bmxvbmd6aGFvQDE2My5jb20=; Fan Fan, Y2FzZGlhbEBsaXZlLmNvbQ==
†These authors have contributed equally to this work