Tumor-associated myeloid cells in cancer immunotherapy

Several molecules such as CCL2, CCL5, colony-stimulating factor 1 (CSF-1), transforming growth factor-beta (TGF-β), vascular endothelial growth factor (VEGF), and platelet-derived growth factor can recruit monocytes into TME and promote their differentiation into TAM [25, 26]. Macrophages are classified into two categories based on their surface receptor expression and secreted products: classically activated macrophages (M1) polarized by IFN-γ and lipopolysaccharide, and alternatively activated macrophages (M2) polarized by IL-4 and IL-13 [27]. Interestingly, the polarization state of TAM is plastic. High levels of proliferator receptor gamma (PPARγ), enhancement of the NF-κB signaling pathway, and a weakly acid or hypoxic TME promote the polarization of TAM from the M1 phenotype to the M2 phenotype. When these conditions are suppressed, M2 can be repolarized into M1 [28]. Functionally, M1 (F4/80⁺CD11b⁺CD86⁺) can kill tumor cells by expressing inducible nitric oxide synthase (iNOS), reactive oxygen species (ROS), and cytokines such as IL-1β, tumor necrosis factor-alpha (TNF-α), IL-12 and IFN-γ. Moreover, M1 also triggers the activity of natural killer (NK) cells and prime cytotoxic T-cells. Conversely, M2 (F4/80⁺CD11b⁺CD206⁺) inhibits the activity of T-cells and NK cells by expressing ARG-1, IL-10, and TGF-β, induces angiogenesis and promotes tumor growth and invasion by expressing VEGF and matrix metal protein (MMP), and can also generate chemokines such as CCL2 and CCL5 to recruit regulatory T-cells (Tregs). In the TME, the phenotypic change of TAM is characterized by a significant increase in the proportion of M2 with tumor progression, which is associated with poor prognosis in many malignancies [29]. M2-TAMs have been shown to be largely responsible for PD-1/PD-L1 blockade resistance by inducing T-cell exclusion within tumor sites and secreting indoleamine 2, 3-dioxygenase (IDO) to inhibit T-cell functions [30]. However, increasing evidence indicates that there are many macrophages with overlapping M1 and M2 characteristics between pro-inflammatory M1 and anti-inflammatory M2 cells [31]. Therefore, the heterogeneity of TAM in the TME is more complex than what we previously thought of as a binary classification. For example, Peng et al. used scRNA-seq to redefine five TAM subtypes based on gene expression profiles rather than on phenotypes. Genes involved in Class II antigen presentation (HLA and CD74), genes related to extracellular matrix (ECM) deposition and remodeling (COL1A), genes for the recruitment of myeloid cells (CCL2), chemokine genes responsible for cytotoxic T-cell infiltration (CXCL9 and CXCL10), and genes that generate pro-inflammatory cytokines (S100A8 and S100A9) were highly expressed in the five cell clusters [32].

Emerging evidence indicates that after site transition from the periphery to the TME, TAMs can still accumulate through tumor-trained proliferation. Recent studies using mouse models have indicated that the proliferative capacity of TAMs is higher than that of mammary tissue macrophages in breast cancer tissues, promoting tumor growth [33]. Furthermore, analysis of clinical liver cancer tissues has revealed a high level of TAM proliferation [34]. However, the precise mechanisms that regulate TAM self-renewal remain unclear.

Wang et al. found that adenosine, produced by adenosine monophosphate (AMP) and catalyzed by CD73 in hepatoma cells, is an essential element in the induction of TAM proliferation. Simultaneously, autocrine GM-CSF released by tumor-stimulated TAMs enhances A2A receptor expression in TAMs. Moreover, they demonstrated that GM-CSF functions synergistically with adenosine to activate the downstream AKT/ERK signaling pathway, inducing TAM proliferation and promoting tumor progression. Therefore, their findings further support the hypothesis that TAMs exhibit increased proliferation, which is an important mechanism contributing to TAMs accumulation in human HCC. These self-replicating TAMs (Ki67⁺ TAM) exhibit an immunosuppressive phenotype and are inversely correlated with the density of CD8⁺ T cells, playing a role in promoting tumors and predicting a poor prognosis. Kaplan–Meier analysis showed that local TAM replication was correlated with reduced overall survival (OS) and recurrence time in patients with HCC [34]. Expression levels of the adenosine receptor ADORA2B were negatively correlated with the OS of patients with HCC [35]. Therefore, through this mechanism, self-proliferation may be a potential target for future antitumor immunotherapy, providing a new strategy for inhibiting TAM accumulation in HCC.

G-CSF and GM-CSF are essential for various generations of TAMCs, contributing to the myeloid-biased differentiation of BM and splenic HSPCs, trans-differentiation of erythroid precursors into myeloid cells, and induction of A2A receptor expression to promote TAM self-proliferation. Many studies have shown that intervention strategies targeting GM-CSF can reverse the differentiation direction and restore the antitumor function of TAMCs of diverse origins, which has great potential to enhance the efficacy of antitumor immunotherapy. For example, the combination of GM-CSF signaling blockade and gemcitabine suppressed MDSC generation and reversed T-cell inhibition in pancreatic cancer [72]. In addition, anti-GM-CSF antibodies can enhance the efficacy of tumor-specific adoptive T-cell therapy as well as combination therapy with anti-PD-1 and anti-CTLA4 antibodies [73]. Moreover, the crucial role of anti-G-CSF antibodies in promoting protective antitumor immunity has been demonstrated. Therefore, the combination of anti-G-CSF antibodies has important prospects for improving immunotherapy [74].

Notably, splenic EMH is a common upstream link between splenic HSPCs accumulation and the mass production of erythroid progenitor cells. Therefore, selective abrogation of tumor-promoting splenic EMH can inhibit EMH-triggered extramedullary myelopoiesis, which is particularly promising for restoring antitumor ability. For example, a low-dose c-Kit inhibitor can act on splenic HSPCs and reduce the expression of endogenous GM-CSF, thereby synergistically enhancing ICB efficacy [54]. Moreover, the CCL2/CCR2 signaling pathway plays a key role in selectively recruiting splenic HSPCs and triggering EMH-mediated myelopoiesis. Targeting CCL2/CCR2 may also be a powerful measure for inhibiting tumor-promoting myelopoiesis [54]. Other studies have shown that icaritin decreases the tumor infiltration of MDSCs and their immunosuppressive activity by blocking tumor-induced splenic myeloid-biased hematopoiesis in tumor-bearing mice, resulting in reduced tumor progression [75]. Hou et al. [76] reported that both local tumor-ionizing radiation and anti-PD-L1 treatment significantly decreased tumor-induced EPC abundance in the mouse spleen in an interferon- and CD8⁺ T cell-dependent manner.

Erythropoietin (EPO), an important indicator and trigger of EMH, is commonly used to treat anemia. Although it promotes red blood cell growth and differentiation, it does not improve the survival of patients with cancer or anemia. Studies have reported that Recombinant EPO promotes resistance to radiotherapy and anti-PD-L1 therapies by restoring the number of EPCs [76]. More importantly, EphB4, an EPO receptor, promotes tumor growth and progression [77]. Treatment with EPO-neutralizing antibodies has been found to significantly reduce tumor growth in B16-F10-bearing mice [78]. Therefore, the functions of EPO require further investigation.

A common approach to restoring antitumor immune function, regardless of TAMCs origin, is to mitigate their immunosuppressive ability. Recent studies have found that the nuclear factor erythroid 2-related factor 2 (Nrf2) pathway activator CDDO-Me reduces intracellular ROS production, decreasing tumor growth [89]. Moreover, the COX-2/PGE2/EP axis maintains suppressive functions of myeloid cells, such as ARG-1 expression [20, 90, 91]. Therefore, therapies targeting COX-2, such as celecoxib, are of great interest, and the relevant clinical trials are shown in Table 3 [92]. In preclinical mouse models, PDE-5 inhibitors, such as sildenafil and tadalafil, abrogate the immunosuppressive mechanisms of MDSCs by downregulating iNOS and ARG-1 activities [93‐95]. Clinical trials have also shown promising antitumor effects in patients with head and neck squamous cell carcinoma and metastatic melanoma (Table 3) [96, 97].

However, reducing the inhibitory power of TAMC alone is unlikely to completely eliminate tumors, and its combination with ICB therapy is a promising cancer treatment strategy. For example, the antitumor mechanism of histamine dihydrochloride (HDC) involves a decrease in the ability of MDSC to inhibit CD8⁺ T cell proliferation by reducing the formation of NOX2-derived ROS. HDC was particularly effective in combination with PD-1/PD-L1 inhibition compared to PD-1/PD-L1 inhibition alone [98]. In addition, olaparib at sub-IC50 concentrations can block the expression of ARG-1, iNOS, and COX-2 and reduce the inhibitory function of MDSC, which is more effective in blocking tumor progression when used in conjunction with anti-PD-1 [99]. Remarkably, several promising pharmacological targets exist, such as the myeloid receptor tyrosine kinases TYRO3, AXL, and MERTK. Studies have confirmed that inhibition of these targets can diminish the suppressive ability of MDSC, slow the growth of tumors, and increase the infiltration of CD8⁺ T cells. In particular, the inhibition of these targets has been shown to augment the efficacy of anti-PD-1 immunotherapy, offering promising applications [100].

Regaining the antitumor ability of TAMCs is more radical than simply reducing their immunosuppressive ability. For example, TLR agonists efficiently skew TAM from an M2-like state to an M1-like state, which has been evaluated in many clinical trials (Table 3) [101, 102]. The TLR7/8 agonists 3M-052 and NKTR-262, in combination with PD-1 blockade, can prolong therapeutic efficacy and duration [103]. Anti-CD40 mAbs promote TAM transformation to the antitumor phenotype [104], and synergistic treatment with anti-CSF-1R antibodies stimulates the antitumor activities of TAMs and assists tumor remission [105]. Additionally, several small molecules, such as ATRA, vitamin D3, and curcumin, can induce the differentiation of MDSCs and reestablish tumoricidal immunity [106‐110]. Interestingly, studies have shown that the blockade of PD-1-mediated signals, as a representative ICB therapy, can diminish the accumulation of immunosuppressive and tumor-promoting MDSCs, skewing the myeloid lineage fate to mature myeloid effector cells [111].

Car-M therapy, in which macrophages are modified with a specific chimeric antigen receptor (CAR), converts the TAM from M2 to M1. With its antigen-specific phagocytosis and tumor clearance abilities, Car-M can easily invade the TME and enhance the antitumor activity of T cells. Therefore, compared with CAR-T therapy, which is difficult to use in the TME, CAR-M therapy has shown great potential for treating solid tumors such as ovarian cancer and has entered clinical evaluation [112]. However, CAR-M therapy has several limitations, including possible cytokine release syndrome and weak antitumor ability in vivo. Therefore, further studies are needed to fully elucidate the structure of CAR and the mechanism of CAR-M therapy.

In addition, one of the serious obstacles to the clinical application of CAR-related therapies is their high manufacturing cost. In this context, the recently proposed universal CAR (UniCAR), which targets different epitopes in a modular design, can not only significantly reduce the cost of CAR-T therapy but also avoid antigen escape and use in relapsed or refractory cancers [113]. This research progress is of great value for the clinical application of Car-M therapy and may be a direction worth exploring in the future.