Lymphokine-activated killer cells (LAK cells) represent a significant breakthrough in the field of adoptive immunotherapy, first developed by Steven Rosenberg’s team at the National Cancer Institute in 1982 (8). The core process involves isolating peripheral blood leukocytes (PBL) from the patient’s blood, then activating and expanding them in vitro with high-dose recombinant interleukin-2 (IL-2) (9, 10). IL-2-stimulated LAK cells exhibit unique biological properties: they recognize multiple tumor cells through non-MHC-restricted mechanisms, exerting broad-spectrum antitumor effects without relying on antigen presentation (9). This characteristic made them a vital tool in early tumor immunotherapy.

The antitumor mechanism of LAK cells primarily relies on the perforin and granzyme system they secrete. When LAK cells come into contact with tumor cells, perforin forms transmembrane pores in the target cell membrane, facilitating the entry of granzyme B into the cell and activating the caspase cascade reaction, ultimately inducing programmed cell death in tumor cells (11). However, this killing mechanism lacks tumor-specific recognition, posing a potential risk of damage to normal tissues. This limitation has led to its declining use in contemporary tumor immunotherapy.

IL-2 plays a dual role in LAK cell therapy: on one hand, it promotes the in vitro expansion and activation of LAK cells by binding to the high-affinity IL-2 receptor (CD25) on effector cell surfaces; on the other hand, IL-2 also enhances the cytotoxic activity and cytokine secretion capacity of LAK cells. However, clinical practice indicates that the high doses of IL-2 required for treatment (typically exceeding 6×10^5 IU/kg) induce severe systemic toxicities, including hypotension, capillary leak syndrome, and multiple organ dysfunction (11). These dose-limiting toxicities significantly constrain the clinical applicability of this therapy.

In early clinical studies, LAK cell therapy demonstrated certain antitumor activity in patients with metastatic melanoma and renal cell carcinoma (11). However, this therapy has two critical limitations: First, the high doses of IL-2 required for treatment induce severe dose-limiting toxicities such as capillary leak syndrome. Second, LAK cells lack tumor-specific recognition capabilities, resulting in limited clinical efficacy and poor safety profiles. These limitations prompted researchers in the late 1980s to shift focus toward more targeted immunotherapy approaches. This shift in research direction holds significant scientific importance: the shortcomings of LAK therapy directly propelled a strategic transition in tumor immunotherapy from “nonspecific killing” to “specific recognition.” Researchers began attempting to isolate naturally occurring tumor-reactive lymphocytes from tumor tissues. This innovative approach laid a crucial foundation for the subsequent development of TIL cell therapy.

TTILs are a type of immune cell isolated from the tumor microenvironment, representing a heterogeneous population rich in tumor-specific cytotoxic T lymphocytes and NK cells.

In 1988, Rosenberg first systematically reported TIL therapy in Science (12). The core technical approach of this therapy involves: surgically obtaining tumor tissue samples from patients, isolating TILs from these samples, selectively expanding them in vitro through IL-2 stimulation, and finally reinfusing large quantities of activated TILs back into the patient’s body (13). Compared to LAK cells, TILs demonstrate significantly enhanced tumor antigen recognition and targeted killing efficiency, primarily attributed to their TCRs having undergone prior selection and activation by tumor antigens (12).

In recent years, with the continuous advancement of tumor immunology research, TIL therapy has achieved significant breakthroughs in clinical translation. Multiple clinical studies have confirmed that TIL therapy demonstrates remarkable antitumor activity in patients with advanced melanoma (14, 15). For instance, a meta-analysis incorporating 13 clinical trials demonstrated that the combination of TIL ACT with IL-2 achieved an objective response rate (ORR) of 44% in melanoma patients, with some patients even achieving long-term progression-free survival (15). Notably, lifileucel (Amtagvi)—the world’s first TIL therapy granted accelerated approval by the FDA in February 2024—demonstrated a 65.2% objective response rate in PD-1-resistant advanced melanoma patients during pivotal trials, with a complete response (CR) rate of 31.8% (16, 17). Beyond melanoma, TIL therapy has demonstrated remarkable clinical potential in HPV-associated cervical cancer. A 2019 study revealed that isolating and expanding HPV-specific TILs from patient tumor tissues, followed by reinfusion, induced complete tumor regression in some patients and significantly prolonged survival (18).

Despite demonstrating breakthrough efficacy in solid tumor treatment, TIL therapy faces multiple challenges in clinical application: First, the therapy requires surgical acquisition of tumor tissue, limiting its applicability for patients inoperable for biopsy; Second, the traditional TIL preparation cycle spans 4–6 weeks, potentially delaying treatment for patients with rapidly progressing disease. Additionally, immunosuppressive characteristics of the tumor microenvironment—such as regulatory T cell infiltration and high PD-L1 expression—may compromise the persistence and antitumor activity of reinfused TILs (19). In the future, it may be necessary to combine TIL therapy with genetic modifications (such as PD-1 knockout), cytokine engineering, or combination therapies (such as immune checkpoint inhibitors, oncolytic viruses, etc.) to enhance its persistence and antitumor effects in vivo.

In 1991, Schmidt-Wolf and colleagues first reported a novel type of immune effector cell—CIK cells (20). This cell population was generated by stimulating peripheral blood mononuclear cells (PBMCs) in vitro using anti-CD3 monoclonal antibodies combined with cytokines such as IL-2. Studies indicate that CIK cells possess unique biological characteristics: they exhibit both the antigen-specific recognition capability of T cells and the MHC-independent killing function of NK cells. This dual nature holds significant promise for their application in the field of tumor immunotherapy.

Studies indicate that adding IL-2 during short-term in vitro culture of human peripheral blood lymphocytes (PBLs) significantly promotes the proliferation and differentiation of effector NK cells and non-specific T cells, endowing them with LAK cell activity (21, 22). These IL-2-activated LAK cells exhibit unique antitumor properties: on one hand, they can effectively lyse fresh tumor cells in vitro through non-MHC-restricted mechanisms (23); on the other hand, in vivo, these cells demonstrate selective killing of tumor tissues while maintaining relative safety toward normal tissues. Furthermore, in vitro, CIK cells demonstrate enhanced proliferative potential. Experimental data indicate that under optimized culture conditions, CIK cells can expand over 1000-fold (20, 24). This characteristic provides an ample cellular source for clinical applications, representing a significant advantage for adoptive immunotherapy.

CIK cell therapy has emerged as a significant approach in tumor immunotherapy, garnering widespread attention due to its low toxicity and favorable safety profile. Extensive clinical studies demonstrate that this therapy has achieved remarkable progress in treating both hematologic malignancies and solid tumors (25) (26–28). In hematologic malignancies, CIK cells demonstrate potent cytotoxic activity against multiple myeloma and leukemia cells. When combined with stem cell transplantation, this approach effectively reduces disease recurrence rates (25). Clinical studies on non-Hodgkin lymphoma (NHL) have also confirmed the favorable tolerability and therapeutic efficacy of CIK cell therapy (26, 27).

In the field of solid tumor therapy, CIK cell therapy demonstrates broader application value. Advanced gastric cancer patients undergoing CIK cell therapy experienced significant improvements in immune function, reduced tumor burden, and enhanced quality of life. Combining this therapy with chemotherapy produced synergistic effects, prolonging patient survival (29, 30). For hepatocellular carcinoma (HCC) patients unresponsive to conventional treatments, minimally invasive procedures combined with CIK cell immunotherapy significantly extended recurrence-free survival (31, 32). Adjuvant CIK cell therapy after breast cancer surgery markedly improves patient prognosis and extends overall survival (33). Furthermore, in nasopharyngeal carcinoma treatment, the median overall survival in the chemotherapy plus CIK cell therapy group (32 months) significantly outperformed the chemotherapy-only group (9 months) (34).

However, CIK cell therapy still faces several pressing challenges. Its non-specific killing mechanism results in insufficient targeting, and issues such as high cellular heterogeneity after in vitro expansion constrain the standardized implementation and clinical promotion of this therapy. Future research should focus on optimizing culture systems to enhance cellular homogeneity and exploring combination strategies with other immunotherapy approaches to further improve the clinical efficacy of CIK cell therapy (28). Currently, DC-CIK combination therapy shows promising prospects, where dendritic cells (DCs) effectively present tumor antigens, significantly enhancing the tumor-specific killing capacity of CIK cells (35).

DCs are the most potent antigen-presenting cells in the body, playing a pivotal role in initiating and regulating both innate and adaptive immune responses. DC cell therapy primarily leverages their potent antigen uptake, processing, and presentation capabilities. By isolating DC precursor cells from a patient’s peripheral blood and differentiating them into mature DCs in vitro, while simultaneously exposing them to tumor-marked tissue or synthetic antigen peptides, these DCs can convey tumor markers to T lymphocytes. Upon reinfusion, this process initiates the body’s specific immune response against tumors (36). Following antigen uptake and processing, mature dendritic cells activate naive T cells (Th0) and further induce their proliferation and differentiation into effector T cells, including CD4+ helper T cells and CD8+ cytotoxic T cells, thereby initiating the body’s adaptive immune response (37–39). It can be said that dendritic cells play a central role in initiating, regulating, and sustaining immune responses.

Additionally, DCs direct the immune system to attack tumor cells. After recognizing and internalizing tumor-associated antigens via their surface pattern recognition receptors (PRRs), DCs process these antigens to form MHC-antigen peptide complexes expressed on the cell surface (36). Mature DCs then migrate to secondary lymphoid organs, where they activate T cell immune responses through antigen presentation and co-stimulatory signals (37). Specifically, the MHC class I-antigen peptide complexes on DC surfaces bind to the TCR of CD8+ T cells. Under the influence of co-stimulatory molecules (CD80/CD86) and cytokines (e.g., IL-12) (40), this interaction induces CD8+ T cells to differentiate into cytotoxic T lymphocytes (CTLs) with tumor-killing activity. These CTLs eliminate tumor cells through the following mechanisms: releasing perforin and granzyme to form pores in the tumor cell membrane and induce apoptosis; triggering death receptor-mediated apoptosis via the Fas/FasL pathway; and secreting cytokines such as IFN-γ to suppress tumor growth (36, 41). Concurrently, DC-activated CD4+ helper T cells further amplify the immune response by secreting cytokines (e.g., IL-2, IFN-γ) and help maintain the function of CTLs and memory T cells (36, 42). This DC-initiated, multi-layered immune response constitutes the core mechanism of the body’s antitumor immune defense.

In addition, after inducing DC maturation in vitro, they are mixed with the patient’s own lymphocytes to enhance the stimulation effect, prompting the patient’s T cells to generate more cytotoxic T lymphocytes. These enhanced cytotoxic T lymphocytes are then reinfused into the patient to directly kill tumor cells. However, overall, DC cell therapy remains in a phase of ongoing exploration and refinement.

With the continuous advancement of cellular medicine, DC-based immunotherapy has achieved rapid progress within just a few years. The first clinical trial of a DC vaccine commenced in 1996, utilizing tumor antigen-loaded peptides to activate patients’ T cells for postoperative immune reconstruction in prostate cancer (43). This pioneering study laid the foundation for subsequent clinical applications of DC vaccines. In 2010, the autologous prostate cancer vaccine Sipuleucel-T (Provenge), based on DC technology, received FDA approval for commercialization, becoming the world’s first therapeutic tumor vaccine utilizing dendritic cells. This therapy successfully activated specific anti-tumor T-cell responses by ex vivo conjugating patient dendritic cells with prostate acid phosphatase (PAP) antigen. Key clinical trial data demonstrated that Sipuleucel-T significantly extended overall survival in patients with metastatic castration-resistant prostate cancer, with a median survival increase of 4.1 months. This breakthrough achievement marked the dawn of a new era in cancer immunotherapy (44).

In recent years, the development of DC vaccines has shown rapid growth. Statistical data indicates that 34 DC vaccine studies were published between 2017 and 2019, with 59 studies ongoing during the same period. From 2015 to 2020, a total of 42 DC vaccine products were launched globally (45). These advancements demonstrate that DC vaccines have emerged as a highly promising personalized cancer treatment platform, with their combination with existing immunotherapies expected to further enhance clinical efficacy. Notably, DC vaccines demonstrate exceptional efficacy in melanoma treatment. Results from a Phase III clinical trial (NCT02993315) published in 2024 revealed that autologous DC vaccines achieved a 2-year overall survival rate of 84.7% in patients with stage III B/C melanoma, significantly outperforming the control group (46). Furthermore, recent studies have demonstrated that adenovirus-mediated delivery of transcription factors PU.1, IRF8, and BATF3 (PIB) can directly reprogram tumor cells in vivo into antigen-presenting DC-like cells (47). This in situ reprogramming technique not only overcomes the limitations of traditional DC therapies—specifically cell collection and in vitro expansion—but also significantly enhances antitumor immunity.

However, DC immunotherapy still faces numerous challenges, such as limited migration efficiency of DCs within the tumor microenvironment and impaired antigen presentation efficiency due to immunosuppressive microenvironments. The latest research by Lin Yuan and Liang Jiankai’s team at Sun Yat-sen University revealed that the oncolytic virus M1 (OVM1) can significantly enhance the antitumor efficacy of DC vaccines by downregulating SIRPα on DC surfaces and CD47 on tumor cell surfaces, thereby relieving their immunosuppressive effects on DC vaccines (48). This discovery provides new insights for optimizing DC vaccines. In the future, strategies such as in situ reprogramming technology, combined immunotherapy, and genetically engineered DCs may significantly enhance efficacy and accessibility. However, their safety, long-term effects, and clinical feasibility still require systematic validation. Therefore, while DC immunotherapy holds great promise, it necessitates further optimization in standardization, efficiency, and efficacy enhancement.

Since the 21st century, breakthroughs in genetic engineering technology have propelled the development of TCR-T therapy. TCR-T therapy employs genetic modification to engineer patients’ T cells to express TCRs that specifically recognize tumor antigens, enabling precise targeting and elimination of cancer cells (49). Compared to non-specific cell therapies like LAK and CIK, TCR-T possesses precise antigen recognition capabilities, allowing it to specifically target tumor-associated antigens (49). The core of this technology lies in the TCR molecule, which acts as the “gatekeeper” of T cell function. Its interaction mechanism with the MHC determines the intensity and quality of the antitumor immune response (50).

TCR-T cell therapy demonstrates unique biological advantages and clinical potential in the treatment of solid tumors. From a molecular perspective, the TCR molecule—a key member of the immunoglobulin superfamily—forms a heterodimer composed of α and β chains. This heterodimer precisely couples with the CD3 complex (comprising four signaling chains: γ, δ, ϵ, and ζ) to create the complete TCR-CD3 complex. This structural feature endows TCR-T cells with highly specific recognition of tumor antigens (50). Regarding mechanism of action, TCR-T cells retain the innate T cell activation pathway. Through specific binding of TCR to MHC-antigen peptide complexes (particularly recognition of MHC class II molecules), they not only directly activate the cytotoxic function of CD8+ T cells but also establish a more comprehensive and durable antitumor immune response through the synergistic action of CD4+ helper T cells (51, 52). Compared to CAR-T cells, which primarily recognize surface-specific antigens on tumors (53), the greatest advantage of TCR-T cells lies in their ability to recognize intracellular tumor antigens presented by MHC molecules. This includes tumor-specific mutated antigens (such as neoantigens) and virus-associated antigens (e.g., HPV E6/E7 proteins), significantly broadening the spectrum of targetable antigens (54). This unique antigen recognition characteristic enables TCR-T cells to demonstrate superior tissue infiltration and persistence within the solid tumor microenvironment.

In recent years, TCR-T cell therapy has achieved significant progress in clinical translation and commercialization. The landmark research conducted by the Rosenberg team first demonstrated the therapeutic potential of TCR-T cells in melanoma patients (6). TCR-T therapy targeting the NY-ESO-1 antigen showed substantial objective response rates in patients with synovial sarcoma and melanoma, with some achieving complete remission (55). Furthermore, a TCR-T cell therapy targeting the HPV-16 E7 protein (NCT02858310) demonstrated objective responses in 6 out of 12 patients in a Phase 1 clinical trial for metastatic HPV-associated epithelial carcinoma, confirming the clinical feasibility of viral antigen-targeting strategies (56). In 2022, Kimmtrak (tebentafusp), the world’s first TCR-T therapy, received FDA approval for treating HLA-A*02:01-positive metastatic uveal melanoma patients. Phase III clinical trial data showed a median overall survival (OS) of 21.7 months, significantly longer than the 16.0 months observed in the control group (57). In August 2024, Adaptimmune’s TCR-T product Tecelra (afami-cel) received FDA accelerated approval, becoming the first TCR-T therapy for synovial sarcoma. Its SPEARHEAD-1 trial demonstrated an objective response rate (ORR) of 43%, a complete response rate of 4.5%, and sustained responses exceeding 12 months in 39% of responders (58). In China, Xiangxue Pharmaceutical’s TAEST16001 injection has been designated as a breakthrough therapy. Its Phase I clinical trial for advanced soft tissue sarcoma demonstrated an ORR of 41.7%, and the company is currently advancing Phase II studies (59).

However, TCR-T therapies still face challenges such as HLA restriction and off-target toxicity. For example, a TCR-T therapy targeting MAGE-A3 caused severe neurotoxicity due to cross-reactivity (60). Future development of TCR-T therapies should focus on discovering novel targets (such as personalized neoantigens), combining with immune checkpoint inhibitors (such as PD-1/PD-L1 blockers), and developing universal TCR-T products (60).

TCR-T therapy, constrained by HLA restriction and potential off-target toxicity, has spurred the development of more flexible targeting strategies. Among these, CAR-T cells have rapidly emerged as a research hotspot due to their unique advantage of bypassing MHC dependence and directly recognizing tumor surface antigens. The CAR structure has undergone multiple generations of iterative optimization: the extracellular domain employs a single-chain variable fragment (scFv) for antigen-specific recognition, the transmembrane domain typically originates from CD8 or CD28 (61), and the intracellular signaling domain incorporates CD3ζ along with one or more co-stimulatory molecules (e.g., CD28, 4-1BB), collectively forming a complete T-cell activation pathway (62). Currently, second-generation CAR-T cells dominate clinical applications, demonstrating significant efficacy in hematologic malignancies (62). With advancing research, multifunctional CAR design has gained prominence. By fusing multiple co-stimulatory factors (CD28, 4-1BB, OX40) within the intracellular signaling domain, or adding cytokine expression modules (IL-12, IL-15), and incorporating logic gates (AND, OR, NOT) to enable multi-antigen recognition, these designs significantly enhance the persistence and precision of antitumor responses (63). For instance, “armored CAR-T” enhances local immune effects through autocrine cytokine secretion, while “bispecific CAR” mitigates the risk of immune escape triggered by single-target mutations (64).

In recent years, CAR-T cell therapy has made significant advances in the treatment of B-cell malignancies. Multiple clinical trials have demonstrated its efficacy in refractory/relapsed B-cell acute lymphoblastic leukemia (B-ALL) and NHL. refractory B-ALL and NHL. These studies primarily target CD19 (65, 66), CD20 (67), or CD30 (68). Among these, CD19-targeted CAR-T cells achieved CR rates of 70%-94% in B-ALL patients (65, 69). In 2017, the world’s first CAR-T product received regulatory approval for market launch. Subsequently, China also approved Axicabtagene Ciloleucel and Rixotumab Tigolizumab for the treatment of large B-cell lymphoma (70–72). Furthermore, CAR-T therapy has demonstrated preliminary efficacy in other B-cell malignancies such as mantle cell lymphoma and marginal zone lymphoma (7, 73, 74).

However, the application of CAR-T therapy in solid tumors still faces significant challenges, including tumor heterogeneity, immunosuppressive microenvironments, and insufficient CAR-T cell infiltration (74–76). Despite these hurdles, researchers are actively exploring novel targets such as CLDN18, CD276, and KRAS. Research on CLDN18 targeting has surged by 400% since 2021, demonstrating CAR-T’s potential for solid tumors (77–79). Solid tumors are also emerging as another frontier for CAR-T cell therapies, particularly demonstrating novel therapeutic potential in melanoma. A research team at the University of California, Los Angeles developed a novel CAR-T cell therapy targeting the TYRP1 protein, which is highly expressed in approximately 30% of cutaneous melanomas and up to 90% of rare subtypes such as uveal melanoma (80). Preclinical studies demonstrate that these CAR-T cells effectively eradicate cancer cells without causing severe side effects, and clinical trials are currently planned (80).

Therefore, future CAR-T therapy research should focus on exploring new targets with high expression levels and low expression in normal tissues, optimizing CAR structural design, and integrating microenvironment modification strategies. Overall, while its potential is immense, overcoming microenvironment barriers and safety challenges remains essential.

NK cells, as key effector cells of the innate immune system, play an irreplaceable role in tumor immune surveillance and antitumor immune responses. Unlike T cells, NK cells possess unique MHC-independent killing properties, enabling them to recognize and eliminate tumor cells without prior antigen sensitization. This characteristic makes them a significant focus of research in the field of tumor immunotherapy (81). In recent years, advances in genetic engineering technologies have demonstrated the broad potential of engineered NK cells (such as CAR-NK) in targeted tumor therapy.

The antitumor mechanisms of NK cells exhibit multidimensional characteristics. First, through antibody-dependent cell-mediated cytotoxicity (ADCC), the FcγRIII (CD16) on the surface of NK cells can bind to tumor-specific antibodies, enabling specific killing of tumor cells. This mechanism has become a crucial foundation for various antibody drug therapies (81). Second, NK cells directly kill target cells by releasing effector molecules such as perforin and granzyme. Notably, this killing exhibits high selectivity, sparing the NK cells themselves from damage—a property that ensures safety for clinical applications (82). Third, TRAIL and Fas ligand expressed on NK cell surfaces can induce apoptosis in tumor cells expressing their respective receptors (83). This death receptor pathway, together with the perforin pathway, constitutes the NK cell killing network. Concurrently, NK cells can traverse the blood-brain barrier to infiltrate brain tumor tissues (84). Furthermore, NK cells can interact with other immune cells (such as DCs, T cells, and macrophages) by secreting various chemokines, growth factors, and cytokines, thereby activating the adaptive immune response and inhibiting tumor progression (82).

NK cell therapy can be further categorized into autologous NK cell therapy, allogeneic NK cell therapy, CAR-NK cell therapy, and monoclonal antibody-based NK cell therapy. Autologous NK cell therapy utilizes the patient’s own NK cells, thereby avoiding immune rejection risks. However, NK cells from cancer patients are often functionally suppressed, which compromises therapeutic efficacy (85). Allogeneic NK cell therapy employs NK cells from healthy donors, overcoming functional deficiencies while enhancing antitumor activity through KIR-ligand mismatch mechanisms and exhibiting lower GVHD risk (85). The most groundbreaking approach is CAR-NK cell therapy (86), which combines CAR engineering with NK cell characteristics to demonstrate unique advantages: First, its safety profile significantly outperforms CAR-T therapy. NK cells do not secrete key inflammatory mediators like IL-6, drastically reducing CRS risk. Additionally, they are not HLA-restricted, resulting in extremely low GVHD incidence (87). Second, NK cells possess multiple recognition mechanisms. Beyond CAR structures, they can identify tumor cells through innate receptors such as DNAM-1 and NKG2D, enhancing targeting precision (88, 89). Finally, NK cells are widely accessible, obtainable from peripheral blood, umbilical cord blood, or NK cell lines, facilitating large-scale production (88, 89). Recent studies have also extended CAR-NK cell survival in vivo through genetic modification, enhancing their infiltration capacity into solid tumors, and engineered enhanced CAR-NK cells resistant to immunosuppressive microenvironments (90). Furthermore, NK and T cell activity is regulated by multiple surface inhibitory receptors, including NKG2A, LAG3, TIM-3, and TIGIT (91). Currently, several monoclonal antibodies are undergoing clinical evaluation to block these immune checkpoints. Research data indicates that blocking these receptors may unleash NK cells, potentially yielding positive therapeutic effects (91).

In recent years, NK cell therapy has achieved breakthrough progress in the field of tumor immunotherapy, with its application expanding from hematologic malignancies to solid tumors and infectious diseases. In hematologic malignancies, allogeneic NK cell infusion has demonstrated durable antitumor responses in patients with acute myeloid leukemia (AML) (92), while CAR-NK cell therapy has shown favorable safety profiles and preliminary efficacy in clinical studies for multiple myeloma (MM) (93). For hepatocellular carcinoma treatment, research from Chonnam National University in South Korea confirmed that locally administered high-dose autologous NK cells combined with hepatic arterial infusion chemotherapy (HAIC) significantly extended progression-free survival (PFS) and overall survival (OS) in patients, achieving a disease control rate as high as 80% (94). Regarding infectious diseases, studies indicate NK cell therapy may positively impact chronic hepatitis B and HIV infection (95, 96). Research on SARS-CoV-2 infection also suggests NK cells hold potential as a key component of COVID-19 immunotherapy (97). Notably, In 2025, research published by Academician Cao Xuetao’s team in Signal Transduction and Targeted Therapy revealed that Neo-2/15-modified CAR-NK cells significantly enhance antitumor activity in solid tumors such as pancreatic and ovarian cancers by activating the c-Myc/NRF1 signaling pathway (98), offering a novel strategy to overcome immune suppression in the tumor microenvironment.

CAR-NK cell therapy, as a significant advancement in NK cell treatment, has achieved remarkable clinical outcomes in recent years. Currently, multiple CAR-NK products worldwide have entered clinical stages, such as Fate Therapeutics’ CD19 CAR-iNK cell therapy FT522 and China’s Zhongsheng Suyuan’s NCR-300 (99, 100). Regarding solid tumor treatment, a 2025 study revealed that knocking out the SMAD4 gene significantly enhances CAR-NK cells’ resistance to TGF-β inhibition, improving their infiltration and killing capacity within the tumor microenvironment (101). Additionally, off-the-shelf CAR-NK cells derived from induced pluripotent stem cells (iPSCs) are gaining prominence due to their scalability and ease of genetic modification, offering potential solutions to the high costs associated with personalized therapies (100).

Despite the promising prospects of CAR-NK therapy, several challenges remain. First, NK cells exhibit limited in vivo persistence, typically surviving only 1–4 weeks, which constrains their long-term efficacy. Second, immunosuppressive factors in the solid tumor microenvironment (such as TGF-β and adenosine) impair CAR-NK cell function (102). Lastly, CAR-NK transduction efficiency remains low, and large-scale production still faces technical bottlenecks (103). Future research should focus on optimizing CAR structural design (e.g., multi-target CARs), enhancing cellular metabolic adaptability (e.g., IL-15 co-expression), and developing universal CAR-NK products to further improve their clinical applicability. With breakthroughs in these technologies, CAR-NK therapy is poised to become a significant breakthrough in tumor immunotherapy following CAR-T.

Due to the complex tumor microenvironment, physical barriers, and immunosuppressive factors in solid tumors, CAR-T and CAR-NK cell therapies face numerous challenges in treating solid tumors (102). Consequently, researchers have begun focusing on other types of immune cells and attempting to engineer them using CAR technology to overcome these limitations. Macrophages, with their potent tumor infiltration capabilities, superior phagocytic functions, and antigen-presenting properties, have emerged as a promising candidate immune cell type (104). As progress in this field continues, CAR-M cells have been developed, offering new hope for the immunotherapy of solid tumors.

The basic structure of CAR-M is similar to CAR-T, comprising an antigen recognition domain (typically scFv), a transmembrane domain, and an intracellular signaling domain. However, CAR-M often incorporates signaling elements such as FcRγ, CD3ζ, CD28, or CD40 to activate phagocytosis and inflammatory cytokine release signaling pathways. For instance, the FcRγ and DAP12 adaptor proteins activate phagocytosis via Syk kinase; the PI3K/AKT and NF-κB pathways drive the release of inflammatory mediators (TNF-α, IL-12, etc.), thereby enhancing the remodeling of the tumor microenvironment (105–107). Furthermore, the CAR-M-induced M1 polarization state helps sustain the antitumor immune cycle (108, 109). Through these mechanisms, CAR-M cells play a crucial role within the tumor microenvironment, offering a novel immunotherapeutic approach for treating solid tumors.

CAR-M therapy is an emerging immunotherapy technology designed to enhance immune responses against solid tumors through engineered macrophages. This approach combines CAR technology with the innate properties of macrophages (104), demonstrating potential in overcoming challenges in solid tumor treatment. Initial first-generation CAR-M cells primarily employed gene transfection techniques to express tumor-specific CAR structures on the macrophage surface, thereby enhancing their ability to recognize and phagocytose tumor cells (110). However, first-generation CAR-M cells exhibited limitations such as low in vitro expansion efficiency, suboptimal transfection rates, and limited in vivo persistence. With ongoing technological advancements, second-generation CAR-M cells have been optimized in multiple aspects. First, improvements in vector systems—such as the use of adenovirus vectors or electroporation methods—significantly enhanced cell engineering efficiency. Concurrently, the introduction of co-stimulatory signals (e.g., CD28 or 4-1BB) further boosted CAR-M functionality, particularly in activating and amplifying immune responses. Second-generation CAR-M not only improved cellular efficiency but also overcame several technical bottlenecks faced by first-generation CAR-M (110, 111). Recently, the emergence of third-generation CAR-M cells has brought further innovations. In CAR design, researchers have incorporated cytokine expression modules, such as Granulocyte-Macrophage Colony-Stimulating Factor (GM-CSF) or IL-12, along with autocrine activation signals (110). These enhancements have improved CAR-M cell survival, proliferation, and immune activation within the tumor microenvironment, offering new avenues for their application in complex solid tumors (112).

Research on CAR-M began in 2020, when Klichinsky et al. first reported the efficacy of humanized CAR-M against HER2-positive tumors in vitro and in animal models (109). In 2021, Carisma Therapeutics initiated the first clinical trial of HER2-targeted CAR-M, marking its entry into the clinical validation phase (89). In 2024, a team from Peking Union Medical College Hospital developed c-MET-targeted CAR-M cells, demonstrating significant efficacy in pancreatic cancer models and advancing into preclinical studies (113). By 2025, Carisma announced Phase I clinical results showing disease stabilization in some patients with favorable safety profiles (114). Concurrently, strategies for generating CAR-M cells via induced pluripotent stem cell (iPSC) differentiation gained traction, offering potential for standardized and scalable production (115).

By comparing CAR-M cell therapy with CAR-T and CAR-NK therapies ( Table 1 ), its unique advantages and limitations become more apparent. While CAR-T therapy has achieved breakthroughs in hematologic malignancies (116), its application in solid tumors remains constrained by difficulties in penetrating tissue barriers, the impact of immunosuppressive microenvironments, and the risk of cytokine storms (102). CAR-NK therapy exhibits lower immunotoxicity and natural killer activity but has limited persistence and expansion capacity in vivo (102, 117). In contrast, CAR-M cells can actively infiltrate tumor core regions and initiate T-cell responses through antigen presentation, with a lower incidence of CRS, making them theoretically more suitable for solid tumor immunotherapy (118, 119).

However, CAR-M therapy still faces several challenges. Macrophages may be induced to adopt a pro-tumor M2 phenotype within the tumor microenvironment (120), thereby diminishing their anti-tumor function. Furthermore, the preparation and expansion techniques for CAR-M cells require further optimization to achieve standardized and scalable production. Utilizing iPSCs as a uniform, stable cell source enables large-scale in vitro expansion and differentiation into NK cells, macrophages, or even T cells, followed by CAR modification. This approach facilitates the development of “off-the-shelf” immune cell therapies. Existing research demonstrates that iPSC-derived CAR-NK and CAR-M cells exhibit robust antitumor activity in vitro, coupled with enhanced controllability and product consistency, laying the groundwork for future clinical translation (121).

In the future, CAR-M therapy holds promise for combination use with other immunotherapy modalities (such as CAR-T, immune checkpoint inhibitors, etc.) (122), leveraging synergistic effects to further enhance treatment efficacy. Research indicates that cytokines from CAR-T cells, such as IFN-γ and GM-CSF, can convert macrophages into the M1 phenotype, thereby amplifying the cytotoxicity of CAR-M cells (123). This mechanism offers a novel therapeutic strategy for combining CAR-M and CAR-T cells, potentially improving overall efficacy in solid tumor treatment. Furthermore, CAR-M technology is being explored for non-neoplastic disease applications. For instance, in 2025, a research team from Huazhong University of Science and Technology developed microneedle-delivered CAR-M (CAR-eM) for treating intervertebral disc degeneration. This marked the first extension of CAR-M therapy into non-tumor disease areas, significantly broadening its application prospects (124). This advancement not only demonstrates the potential of CAR-M therapy but also paves the way for its application in more clinical diseases.

Throughout the development of ACT, while the overall strategy relies on modifying or expanding immune cells in vitro before reinfusion to achieve antitumor effects, different cellular platforms exhibit significant differences in biological characteristics, recognition mechanisms, and clinical application prospects ( Table 2 ).