Inhibition of anti-tumor immunity was reported to be the main pathogenic mechanism of TAM. TAM could downregulate the release of the immunostimulatory factor IL-12, which can trigger the tumor-killing effect of natural killer (NK) cells and cytotoxic CD4+ T cells (23). In addition, many immunosuppressive factors produced by TAM could also mediate cancer development, such as IL-10, transforming growth factor–β, and prostaglandin E2 (10, 24, 25).

TAM can also directly inhibit the function of T cells through specific enzyme activities, such as arginase 1 (ARG1), which is a hydrolase that controls the catabolism of L-arginine. ARG1 is induced by multiple signaling pathways mediated by IL-4, IL-10, and hypoxia and affects T-cell function by limiting the activity of the semi-essential amino acid L-arginine (25). TAM can also promote T-cell apoptosis by inhibiting the expression of programmed cell death ligand 1 (PD-L1) and B7 homolog 1 on T cells (12, 25).

The function of M2 macrophages in promoting tumor development depends on the proinflammatory cytokines, such as TNF-α, IL-6, and IL-11, which can activate the nuclear factor–κB (NF-κB) and signal transduction and activator of transcription 3 (STAT3) pathway in cancer cells (10, 12, 13, 18, 25). In addition, M2 TAM promoted tumor progression by promoting angiogenesis and lymphangiogenesis by increasing the expression of VEGF-A and VEGF-C (18, 20, 25).

The PD-1/PD-L1 pathway was abnormally activated in various cancers (6, 26), and anti–PD-1/PD-L1 immunotherapy has been widely used or tried in clinical trials in many solid tumors, such as lung cancer, advanced metastatic melanoma, esophagus cancer, and colorectal cancer (27, 28). However, the efficacy of PD-1 inhibitors was still dismal in some patients with high expression of PD-L1, and the concrete mechanisms remain largely unknown.

Previous studies have demonstrated that TAMs can regulate the expression of PD-1/PD-L1 through the activation of different signaling pathways (Figure 1), which, in turn, affects the efficacy of PD-1/PD-L1 inhibitors. CD163+ TAMs in the tumor microenvironment are reported to be positively correlated to PD-L1 expression in various cancers, including pancreatic cancer and liver cancer. Multiple cytokines released by TAM, including IL-6 and TNF-α, can upregulate PD-L1 expression by activating Janus kinase (JAK)/STAT3, phosphoinositide 3-kinase (PI3K)/AKT, NF-κB, or Extracellular signal-regulated kinase (ERK) 1 and 2 signaling pathways (29, 30). In addition, PD-L1 protein expression could also be upregulated by TNF-α through post-translational regulation (29).

In addition to the PD-L1 expression on tumor cells, the tumor microenvironment was also a key factor associated with anti–PD-1 resistance. As mentioned above, cytokines released by TAMs could regulate PD-L1 protein expression, which was reported to be an important predictor for anti–PD≥1/PD≥L1 therapy. In recent years, multiple immune cells have been identified in TME, and the cancer ecosystem has evolved over time, which plays a complex role in cancer development (31, 32). The interaction between macrophages and other immune cells was explored and demonstrated to be correlated to the response to immunotherapy (31). Single-cell and spatial analysis showed that interaction between FAP⁺ fibroblasts and SPP1⁺ macrophages could promote the formation of immune-excluded desmoplasic structures and restrict T-cell which reduces the efficacy of immunotherapy (31). In triple-negative breast cancer, high levels of CXCL13⁺ T cells are associated with the proinflammatory features of macrophages and can predict the clinical benefit of checkpoint inhibitors (32).

Exosomes are small extracellular vesicles that play a crucial role in various cell activities in cancer. Recent studies have reported that macrophage-derived exosomes may promote the formation of a pre-metastatic niche that facilitates tumor growth and metastasis. M2 macrophage–derived EVs can drive anti–PD≥1/PD≥L1 therapy resistance, promote the expression of drug-resistant genes in tumor cells, or affect the immune cell spectrum in TME (33, 34). Therefore, the interaction between TAMs and TME may contribute to anti–PD≥1 therapy resistance in cancer, providing a theoretical basis for the combination use of targeting macrophages and anti–PD≥1/PD≥L1 therapy.

Previous studies have shown that PD-1 inhibitors have an impact on TME in various cancers (35). In non–small cell lung cancer, single-cell RNA sequencing demonstrated that the tumor microenvironment was remodeled after neoadjuvant PD-1 blockade combined with chemotherapy, and TAMs were transformed into a neutral type instead of an anti-tumor phenotype (36). Furthermore, anti–PD-L1 therapy can inhibit tumor growth by reducing PD-L1 expression and promoting the expression of the co-stimulatory molecules CD86 and major histocompatibility complex class II (MHC-II) (37). In addition, the phagocytic ability and immune function of macrophages were also enhanced by anti–PD-L1 therapy, which activates T cells in the TME and eradicates cancer cells (37). Therefore, anti–PD-L1 therapy may repolarize macrophages, enhance the phagocytic ability of macrophages, and ameliorate the tumor microenvironment in some patients.

Depletion of macrophages in the tumor microenvironment may be an effective treatment strategy for cancer, either alone or in combination with chemotherapy. Inhibition of the signal transduction axis of colony-stimulating factor-1/colony-stimulating factor-1 receptor (CSF1/CSF1R), which is necessary for macrophage survival, can induce apoptosis of macrophages. On the one hand, inhibition of CSF-1R combined with radiotherapy or chemotherapy can improve T-cell responses. Blockade of CSF1R signaling can effectively deplete the immunosuppressive TAM and then stimulate the CD8+ T-cell response, resulting in prolonged survival in glioblastoma brain tumors (49). At present, CSF1R inhibitors in combination with chemotherapy are being tested in clinical trials in some cancers, such as localized prostate cancer and orthotopic glioblastoma (49, 50). In addition, blocking CSF1/CSF1R can improve the efficacy of a variety of immunotherapies, including CD-40 agonists (51) and PD-1 inhibitors (52).

As TAM was transformed from monocytes, blocking the recruitment of monocytes in the circulation to the tumor site was another method to reduce TAM in the tumor microenvironment. Recruitment of monocytes from bone marrow to the tumor site is dependent on C-C motif ligand 2 (CCL2)-CC chemokine receptor 2 (CCR2) signal transduction (53). Inhibition of CCR2 causes monocyte retention in bone marrow and leads to depletion of monocytes in the peripheral circulation, reduction of monocyte recruitment to the primary tumor sites and metastatic foci, and consequent reduction of TAM number, resulting in tumor shrinkage and survival improvement (54–56).

Other pathways involved in macrophage recruitment include CXCL12-CXCR4 and the angiopoietin 2 (ANG2)–TIE2 axis (57–59). Therefore, depletion of TEM may cause vascular destruction, neutralization of ANG2 may improve the response to vascular VEGFA blockade, and inhibition of TEM recruitment may inhibit tumor growth (60).

As macrophages were the main phagocyte and antigen-presenting cell in the tumor, the immune stimulation function of macrophages was lost after the removal of TAMs. Therefore, reprogramming or repolarization of TAM to enhance its anti-tumor function and limit tumor-promoting properties is a more attractive strategy for cancer treatment. For example, in the mouse model of breast cancer, TAM represents the main source of IL-10 and inhibition of IL-10 signal transduction can significantly improve the efficacy of chemotherapy. The IL-10 secreted by TAM inhibits the IL-12 produced by APCs, thereby inhibiting the anti-tumor response of CD8+ T cells induced by paclitaxel and carboplatin (23). In addition, the repolarization of TAM makes it specifically express the proinflammatory cytokine IFN-α, which could activate NK cells and T cells in the tumor environment and significantly slow tumor growth in the mouse model (61). The epigenetic reprogramming of macrophages by inhibiting histone deacetylase (HDAC) can also trigger an immune response in T cells (62, 63). In the breast cancer model, selective class IIa HDAC inhibitor induces the anti-tumor macrophage phenotype, promotes the T-cell immune response, and increases the response to chemotherapy and immune checkpoint inhibitors (62). In addition, the activation of the PI3K signaling in macrophages can drive the immunosuppressive activity in TAM, whereas inhibition of the PI3K pathway can reprogram macrophages enhance T-cell responses (64, 65).

CAR-T cells are reported to be effective in hematological malignancies, whereas the efficacy of CAR-T therapy remains dismal in solid tumors, as the entry of T cells into tumors is restrained (66, 67). However, CAR-macrophages (CAR-M) overcome this disadvantage as the macrophages in the TME could be replenished by circulating monocytes. CAR expression could enhance the antigen-dependent functions of macrophages, such as the secretion of cytokines, polarization, enhanced phagocytic ability, and anti-cancer activity (68). CAR-M cells mediate phagocytosis, exhibit M1 functions in a relatively stable way, and exert anti-tumor effects in primary and metastatic tumors (69). Currently, several clinical trials are underway or being developed to evaluate the anti-cancer efficacy of CAR-M in different tumors.

The combination of targeting macrophages and anti–PD-1 therapy in cancer has been investigated in vitro and in vivo ( 37, 70–72). As we have noted above, repolarization of TAM was considered a promising strategy for cancer treatment, and this approach can potentiate anti–PD-1 therapy efficacy in hepatocellular carcinoma (72). Chemotherapy and radiotherapy may reset macrophages toward an M1 phenotype and improving the efficacy of immunotherapy in cancer (71). Vinblastine can drive the polarization of TAMs to the M1 phenotype by activating NF-κB, increasing CD8+ T-cell populations, and improving the survival outcome of malignant tumor immunotherapy (71). Bi-target treatment such as PD-1–IL-2 cytokine variant (IL2v), which employs anti–PD-1 as a target moiety that is fused into an immuno-stimulatory IL2v, can improve the therapeutic efficacy by reprogramming immunosuppressive TAMs (70). In conclusion, targeting macrophages combined with anti–PD–1 therapy may be a promising strategy to overcome drug resistance in patients with cancer.