Heterogeneous CAFs exist widely in the matrix of PDAC. They reshape the microenvironment by secreting extracellular matrix and produce soluble molecules and extracellular vesicles. These CAFs positively regulate immunosuppression through various mechanisms. CAFs secrete multitudinous cytokines and chemokines that promote the differentiation of macrophages into an immunosuppressive phenotype, weakening the function of immune cells and hindering the efficacy of immunosuppressive checkpoint inhibitors (35). CAFs have long been taken into account as the cells for building tumor immune escape shelters, and targeting them may be a promising therapeutic strategy. However, in recent years, many reports have pointed out that inhibiting CAFs in mouse pancreatic cancer models may lead to an increase in poorly differentiated pancreatic cancer cells and a worse prognosis. For example, genetic depletion of CAFs with high expression α-smooth muscle actin (αSMA) is faster tumor progression and a lower survival rates (36). This finding serves as an early warning that some cells in the complex CAFs family have cancer-suppressive properties. An indiscriminate attack on CAFs will promote PDAC. Single-cell sequencing was used to reveal the divergence mechanism and measure the tumor immunity of different types of branches in the CAFs group.

It is not completely sure that the origin of CAFs is PSCs, but this is the current mainstream view. Under the function of soluble molecules secreted by tumor cells, PSCs differentiate into two main subgroups with diverse distinguishing: myofibroblastic CAFs (myCAFs) with high expression of αSMA and inflammatory CAFs (iCAFs) in contrast (37). Studies have shown that myCAFs are mainly distributed around cancer cells as the main source of ECM, which has potential anti-tumor properties. On the contrary, iCAFs is far away from cancer cells and can induce inflammation in the microenvironment, which is beneficial to immunosuppression (38). Some research has identified a new type of CAFs that express MHC class II and CD74, which are called antigen-presenting CAFs (apCAFs). It has been speculated that apCAFs function in antigen presentation and bind to corresponding receptors on T cells. However, they cannot effectively stimulate T cells proliferation due to the lack of classical costimulatory molecules, resulting in immunosuppression (39). Some research have reported a kind of CAFs that have an anti-tumor effect and highly express the stem cell marker Meflin. The degree of infiltration of these CAFs is positively correlated with a better prognosis. However, the number of these cells gradually decreases as cancer progresses, leading to an increase in the immunosuppression of PDAC (40). In PDAC, the content of CAFs positive for fibroblast activation protein A (FAP) is increased, which may cause the development of cancer. FAP is a serine protease that is specific to in CAFs and assists in PDAC’s evasion of the immune system. FAP is being closely studied as a potential biomarker for diagnosis and immunotherapy target (41). Numerous cytokines stimulate the differentiation of CAFs and they can interact with each other. TGF-β relies on the Smads pathway to transmit signals that differentiate into myCAFs. Meanwhile, Interleukin-1α (IL-1α) utilizes the NF-κb pathway to increase the level of leukemia inhibitory factor (LIF), which stimulates the JAK-STAT3 signal to enhance the differentiation of iCAFs. TGF-β can inhibit the pathway that promotes iCAFs differentiation, resulting in the multiplication of myCAFs (42). CAFs differentiate into myCAFs upon receiving hedgehog signal (Hh) and into iCAFs upon receiving CXC chemokine receptor 2 signals (43, 44). Various subgroups of CAFs can convert into each other under certain conditions. P53 may play a key role in remodeling CAFs. CAFs that were originally co-cultured with P53 deficient cancer cells will transform into the dominant CAFs after being co-cultured with gain-of-function (GOF) mutant p53 (45). ICAFs and apCAFs can be transformed into myCAFs through technical cultivation. Discriminating between CAFs and other cells, as well as identifying different subtypes of CAFs, remains challenging. CAFs share common markers with certain normal cells, such as pericytes and microvascular endothelial cells. Numerous promising markers specific to CAFs have been identified as overexpressed in multiple subpopulations (46). Although there has been a breakthrough in understanding the heterogeneous of CAFs, it is difficult to elaborate on all mechanisms. Activated CAFs can recruit immunosuppressive cells. Inducing CAFs into a quiescent state may be a key strategy to improve the effectiveness of immunotherapy and reverse the immunosuppression of PDAC. Clinical trials have been conducted using drugs such as all-trans-retinoic acid (ATRA) based on this theory, in combination with chemotherapy drugs, to obtain safer evaluations (47).

Single-cell sequencing technology asserts that different CAFs play contradictory roles in PDAC. Completely inhibiting CAFs without discrimination could lead to counterproductive treatment. Ideally, we aim to support CAFs for cancer suppression and exterminate carcinogenic CAFs to obtain greater clinical benefits. Because of the complexity and mutual transformation of CAFs, the lack of biomarkers to accurately distinguish CAFs subgroups can result in to targeted drugs mistakenly injuring anti-tumor cells. It is extremely difficult to design drugs targeting CAFs. Under more arduous research, this method shows promise as a treatment for immunosuppression. It involves inducing CAFs to differentiate into tumor suppressors through reciprocal transformation mechanisms within the CAFs family.

Heterotypic macrophages are a prominent formation of immune cells in the microenvironment of pancreatic cancer. They have dissimilar phenotypes and functions that affect cancer progression and are considered a prognostic symbol. It is deemed that some macrophages in TME originate from bone marrow-derived classical mononuclear cells in circulation. These cells have high expression of CD14 and deletion of CD16, and they also express the immunosuppressive factor C-C Motif Chemokine Receptor 2 (CCR2) and bind to its corresponding ligand 2 (CCL2). The up-regulation of classic monocytes that can easily differentiate into tumor-promoting macrophages is associated with a worse prognosis. The macrophages of the pancreas are also derived from resident tissue macrophages generated by primitive hematopoiesis (48). In the TME, macrophages are typically classified as M1-like or M2-like based on their phenotype and biomarkers, although recent proposals suggest that this classification is too broad. M1-like macrophages express CD80, CD86, and inducible nitric oxide synthase (iNOS) on their cell surface, which has an anti-tumor effect and enhances Th1 to stimulate an immune response. After being activated by IFN-γ and TLR ligands, M1-like macrophages overexpress IL-12, IL-23, and MHC-II to control tumor growth by inducing nitric oxide synthase (49). TAMs, the major component of M2-like macrophages expressing CD160, CD206, and Arginase (ARG) -1 on the cell surface, have an immunosuppressive effect and assist Th2 in completing immune escape (50). TAMs increased gradually with the occurrence and development of malignant tumors linked to the resistance of PDAC (51). A high level of infiltration of TAMs indicates a lower overall survival rate. TAMs support tumor cell proliferation and metastasis through multiple mediators. Through a variety of cytokines and chemokines, TAMs act on tumor cells to play their own immunosuppressive role. TAMs also play an immunosuppressive role by inhibiting the function of anti-tumor macrophages. On the other hand, these cells will produce IL-10 to enhance the response of Tregs and promote Th2 to release IL-4, forming a positive feedback loop that maintains the level of TAMs (52). Inhibiting the IL-10-dependent pathway can promote the production of IL-12 by dendritic cells and alter the immune cell infiltration of PDAC to improve the prognosis (53). TAMs indirectly reduce T cell infiltration and attenuate T cell activation signal transmission through the PD-L1 site and Dectin1/galectin-9 axis (54). Additionally, Phosphatidylinositol 3-kinase γ (PI3Kγ), which is similar to the immune program switch, plays a significant role in the formation and immunosuppression of TAMs. Selectively inactivating it upregulates MHC-II and decreases IL-10 and arginase, which may become a potential method to regulate the immune response of TME to halt tumor progression (55). There is likewise a macrophage that does not express CCR2 but highly expresses endothelial receptors, which represents a worse prognosis. These cells can form new blood vessels and maintain hypoxia by combining with angiopoietin ligands (especially ANG1 and ANG2) that increase in the hypoxic microenvironment (56). According to existing cognition, macrophages are first recruited into the microenvironment and located around the vascular system. They then migrate to the hypoxic area depending on the influence of CCL2 or ANG. The infiltration of macrophages into hypoxic areas may enhance patient survival (57). TAMs in TME are frequently found surrounding the tumor vasculature, tumor cells, and other areas of malignant inflammation. On the contrary, these areas are often far away from M1-like macrophages. However, the spatial distribution of macrophages is not stable and changes as the disease progresses (49). It should also be pointed out that TAMs can be remodeled into M1-like macrophages in the presence of inflammatory factors in the microenvironment, indicating that TAMs are immune cells with remarkable plasticity potential (58). This suggests that utilizing the plasticity of TAMs to reverse the immunosuppression of PDAC could be an effective treatment. However, M1-like macrophages are not completely harmless, as they maintain a chronic inflammatory state that can promote malignant transformation (59).

It generally declares that neutrophils exist in acute inflammation to perform their functions. However, in the microenvironment of PDAC, neutrophils are not only responsible for causing inflammation but also contribute to tumor proliferation and metastasis as immunosuppressive components. Tumor-associated neutrophils (TANs) are a diverse group of cells in pancreatic cancer (60). Different phenotypic characteristics represent various abilities. The heterogeneity of neutrophils plays two opposing roles in PDAC. The active molecules in TME can regulate the differentiation balance between anti-tumor subgroups and tumor-promoting subgroups. Neutrophils respond to low-dose IFN-β and transform into anti-tumor subsets that secrete CCL3, CXCL9, and other chemokines, promoting immunity by recruiting CD8⁺T cells. They also increase TNF-α and ROS expression to amplify cytotoxicity (61). Furthermore, antineoplastic neutrophils mediate the release of IL-1 and hepatocyte growth factor (HGF) to exert antibacterial activity and increase the production of iNOS to kill tumor cells, respectively (61). TGF-β induced TANs differentiation of tumor subpopulations (62). TANs interact with other cells in the microenvironment, contributing to the chronic inflammation of tumors and strengthening the immune dilemma (63). Chemokines produced by TANs, such as CCL2 and CCL17, increase the ratio of Tregs and TAMs, decrease the ratio of IL-17, and regulate the therapeutic effect of sorafenib (64). TANs generate a variety of vascular growth-promoting molecules, such as VEGF and MMP, which have a synergistic effect on angiogenesis. However, these molecules can also be targeted for inhibition as a potential therapeutic strategy. Neutrophils have been observed to increase the expression neutrophil gelatinase-associated lipocalin (NGAL) in a variety of cancers, which can inhibit angiogenesis (65).

TANs can induce tumor cell invasion and migration through various ways, in which IL-1β occupies an eminent position. In PDAC, abnormal adipocytes secrete IL-1β, which activates PSCs to recruit TANs. TANs, in turn, release IL-1β to activate PSCs, forming a positive feedback loop that promotes tumor growth (66). P2RX1 is an ion channel purinergic receptor that activates inflammation (67). In cases of PDAC with distant metastasis in the liver region, a large number of neutrophils are present with either no or low expression of P2RX1 and increased MMP-9 levels. These neutrophils activate neutrophil transcription factor NF-E2 p45-related factor 2 (NRF2) to regulate redox reaction and affect cell metabolism (68). NRF2 also induces CD8⁺T cell apoptosis through PD-L1 (69). Neutrophils regulate the content of elastase, which can destroy the E-cadherin (E-cad) of tumor cells, leading to tumor proliferation (70). CXCL5 acts as a CXCR2 ligand depends, which is dependent on NF-ĸB’s instructions during the recruitment process of TANs. Blocking this signal pathway reduces the quantity of neutrophils that induce immunosuppression (71). IL-17 increases the infiltration of TANs and inactivates therapeutic immune cells by increasing the activity of chemokines, leading to resistance to immunotherapy (72).

Neutrophil extracellular traps (NETs) can shield circulating tumor cells from immune cell cytotoxicity. These traps are secreted by neutrophils (73). NETs also mediate the epithelial-to-mesenchymal transition (EMT) (74). NETs may promote distant metastasis of tumors and could serve as a potential prognostic marker for patients (75). NETs are mainly composed of histones, which also promote angiogenesis (76).

Studying the influence and mechanism of TANs in emerging treatment methods can benefit patients. In addition to inhibiting chemokine CXCR2 to weaken the recruitment and activation of TANs, there are many new explorations (77). For instance, BL-8040, an inhibitor of CXCR4, can improve the infiltration of CD8⁺T cells and more securely alleviate the patient’s condition when combined with pembrolizumab (78). Clinical trials are being conducted on targeted inhibition of TGF-β and IL-17, which promote neutrophils to differentiate into tumor-promoting functions and collect TANs, respectively (79, 80). Targeted therapy combined with other regimens may improve tolerance and therapeutic effect. There are two main solutions to suppress NETs: inhibiting peptide arginine deiminase 4 (PADI4) to reduce its formation or exploiting DNase and chloroquine to destroy its structure (61). Thrombomodulin inhibits the transformation of epithelial cells into mesenchymal cells induced by NETs (70).

The Neutrophil/Lymphocyte Ratio (NLR) can serve as a biomarker to predict the overall survival rate of cancer, stratify the severity of patients and formulate treatment plans. This index is simple, cost-effective, and safe to operate (81). An elevated NLR level is positively correlated with a poor prognosis.

Pathological factors can cause immature bone marrow cells to differentiate into MDSCs along with bone marrow progenitor cells. MDSCs, like other immune cells, are heterogeneous and cause immunosuppression by regulating various mechanisms (82). These cells are mainly divided into monocyte subsets (M-MDSC) and polymorphonuclear cell subsets (PMN-MDSC) as they are composed of immature macrophages and immature granulocytes. The proportion of the latter is higher than that of the former, but the immunosuppressive function of the former is more important (83). Tumor cells regulate the aggregation of MDSCs to tumor sites by releasing chemokines CCL2 and CCL5 (84). Growth factors such as GM-CSF and G-CSF also participate in the recruitment of MDSCs (85). TNF-α、IL-6 and IL-1β contribute to the recruitment of MDSCs by acting on other immune cells. Prostaglandins PGE2 and IL-17 can increase the number of MDSCs and promote them to produce more ARG1 to maintain immunosuppression (86). Sphingosine-1-phosphate (S1P) is expressed more in the hypoxic microenvironment of PDAC and recruits MDSCs (14). The content of MDSCs in pancreatic cancer changes with disease progression. MDSCs are activated by VEGF from stromal cells and can also secrete VEGF to further enhance their activity (87). MDSCs play an extremely prominent role in the immunosuppression of PDAC and are related to a variety of tumor-promoting cells and bioactive molecules. Because of the vascular collapse of the matrix and the high-pressure environment, the nutrients available for immune cells supplied by TME are quite scarce. MDSCs inhibit the proliferation and function of surface receptors of CD8⁺ T cells by producing ARG1 and iNOS, which consume nutrients by utilizing L-arginine (88). Arg1, iNOS, and NADPH oxidase 2 (NOX2) have oxidative activity that inhibits the tumor immune response by enhancing the content of NO and ROS (86). MDSCs can adjust the content of indoleamine 2,3-dioxygenase (IDO), which degrades tryptophan. With the increase of IDO content, there are more toxic metabolites in the microenvironment. MDSCs also resist tumor immunity by enhancing the expression of PD-L1 (89). Similarly, MDSCs release TGF-β and IL-10, which block the effect of natural killer cells. The mechanisms of immunosuppressive effects of MDSCs can induce T cells to differentiate into Tregs, indirectly assisting tumor immune escape by increasing Tregs invasion (86). MDSCs are also associated with the decrease of anti-tumor T cell content in precancerous lesions is also associated with the role of MDSCs. MDSCs can also improve the differentiation ratio of TAMs (14). Cooperating with other immune cells, MDSCs produce VEGF and MMP9, which are conducive to tumor angiogenesis (90). MDSCs promote tumor cells to invade and migrate outward by affecting the pre-metastasis niche.

MDSCs can cause immunosuppression through various mechanisms, but these effects may not be universal among patients. Despite this, the significant infiltration of MDSCs in PDAC patients leads to resistance to various treatment strategies and a worse prognosis. In recent years, many therapeutic approaches targeting MDSCs have been put into animal models. One potential strategy is to encourage the differentiation and maturation of MDSCs. Trans-retinoic acid is proposed for use due to its ability to induce differentiation. Animal models and patient studies have shown that this drug can effectively reduce MDSCs (91). The immunosuppressive pathway targeting MDSCs is also available. Inhibition of PGE2 can restore the tumor immune response (92). These therapeutic methods that inhibit MDSCs are conducive to enhancing the therapeutic effect of immunotherapy targeting PD-1 (93). Lactic acid can regulate the tumor-promoting function of MDSCs and alleviate the immunosuppression of TME under radiation stimulation (94). The combined treatment of GM-CSF, which blocks the recruitment of MDSCs and gemcitabine reduces the differentiation ratio of MDSCs decreases and improves the function of anti-tumor immune cells (95). We speculate that inhibiting MDSCs will effectively alleviate the immunosuppressive characteristics of PDAC and promote the effect of immunotherapy.

The TME contains complex T cells, including CD8⁺T cells and CD4⁺T cells. CD8⁺T cells play a cytotoxic role and are also known as cytotoxic T cells, while CD4⁺T cells can be further classified as Th1, Th2, Th17, and Tregs (96). Gene mutations in TP53, CDKN2A, and SMAD4 occur during the progression of pancreatic precancerous lesions to cancer. Mutated genes that cause chronic inflammation of TME aggravate tissue loss by activating CAFs, increasing VEGF release, and mediating the NF-ĸB pathway. These genes are also associated with the inhibition of cytotoxic lymphocytes (CTL) and an increase in Tregs (97).

CD8⁺T cells are predominantly distributed in the microenvironment far away from tumors. Moreover, under the effects of various immunosuppressive cells, the number of activated cells and the infiltration of CD8⁺T cells are reduced. PDAC is typically classified as a cold tumor that is difficult to elicit an immune response (98). However, studies have shown that surgical treatment can increase infiltration of CD8⁺T cells in PDAC patients, indicating their potential as an anti-tumor agent (99). The distribution and degree of T cell infiltration have been linked to different phases of pancreatic cancer. The general trend is that CD8⁺T cells decrease gradually and are mainly distributed in the tumor margin and dense matrix, while Tregs increase gradually and are mainly distributed around the tumor cells (100). The dense matrix restricts CD8⁺T cells from approaching and clearing the tumor. CXCL12 can mediate the migration of CD8⁺T cells out of the marginal region (101). The prognosis of patients improves with an increase in CD8⁺T cells in close proximity to tumor cells. CD8⁺T cells eliminate tumor cells through the release of granzyme and perforin. The infiltration of granzyme B⁺ CD8⁺ T cells represents longer survival time. CD8⁺T cells release anti-tumor mediators such as TNF-α, and IFN-γ to destroy abnormal cells (14). The formation of massive lymphocyte aggregation to form tertiary lymphoid structure (TLS) represents a better prognosis (102). The low mutation rate of PDAC results in a lack of CD8⁺T cells and few new antigens to induce an effective immune response (103). The immunosuppressive programmed death receptor (PD-1) and corresponding ligand (PD-L1) are up-regulated on the surface of CD8⁺T cells, leading to their weakening primarily through the Fas-mediated apoptosis pathway (104). The quality of T cells is more important than their quantity. Under these mechanisms, the reduced number of CD8⁺T cells is difficult to activate and presents a depletion phenotype.

CD8⁺T cells are activated by Th1 cells through the secretion of IFN-γ to facilitate tumor immunity (105). The immune infiltration of Th2 cells can induce the secretion of cytokines that inhibit immunity and promote the differentiation of TAMs (106). Th17 cells produce IL-17 and also play an immunosuppressive role (98). During the early stages of tumor occurrence, Tregs infiltrate both the tumor and circulation (101). Tregs overexpression of CTL-associated antigen 4 (CTLA-4), which inhibits antigen-presenting cells from expressing costimulatory ligands CD80 and CD86 for activated T cells (107). Similarly, some CAFs bind to CD8⁺T cells without activation. Tregs also mediate a variety of signal pathways, such as IL-10 and TGF-β, to prevent the immune function of CD8⁺T cells (14).

Continuous exposure to tumor antigens can cause T cells to become exhausted, reducing their proliferation ability and increasing the expression of inhibitory receptors such as PD-1 and CTLA-4 on their surface, transforming into exhausted T cells. These exhausted T cells can be divided into two subgroups: the incompetence subgroup and the aging subgroup (108). Cytokines like TGF-β and IL-2 can stimulate the increase of PD-1. PD-L1 and PD-L2 ligands of PD-1 come from a variety of cells, including tumor cells, TAMs, and MDSCs. Their contents are positively regulated by IFN-γ. The PD-1 pathway impedes the dephosphorylation of the T cell surface receptor CD28, leading to its inactivation (109). CTLA-4 interferes with costimulatory receptors, thereby restraining CD28 (110).

At present, extensive research has been carried out to restore and utilize the immune function of CD8⁺T cells. But PDAC does not benefit as much as other cancers. Various research directions exist based on the mechanism of T cells in tumors, including ICB, chimeric antigen receptor (CAR) T cells, and tumor vaccines. At present, the targets of immunotherapy for pancreatic cancer are PD-1 and CTLA-4. However, due to the presence of rare mutations and new antigens in PDAC, as well as the complex relationship between internal immunosuppression is complex, these treatments have not shown significant therapeutic gains (111). Gemcitabine is the first choice chemotherapy drug for PDAC. It has been shown to inhibit the immunosuppressive effect of Tregs (112). Combining multiple therapies that target the mechanism of T cell immunosuppression may be a promising direction for future research. The mechanism by which T cells induce immune escape and weaken the anti-cancer effect requires further consideration and exploration.

Dendritic cells (DCs)play a crucial role in innate and adaptive immunity by recognizing antigens, transmitting signals to T cells, and activating the immune response of T cells. As antigen-presenting cells (APCs), DCs exist in tumors as the first step to capture and process tumor cells (113). DCs are a heterogeneous cell population that can be classified into conventional DCs (cDCs) and plasma cell-like DCs (pDCs) according to different classification methods. The former can be subdivided into two types of dendritic cells: BDCA3⁺cDC1s, which activate CD8⁺T cells, and BDCA1⁺cDC2s, which transmit signals to CD4⁺T cells. Ideally, BDCA3⁺cDC1s can induce a T cell immune response in the microenvironment under the influence of CCL2 (114). But in PDAC, immunosuppressive components exist widely in the microenvironment. The hypoxic and nutrient-deficient microenvironment makes it difficult for DCs to play a strong immune function. Immunosuppressive cells and their products hinder the proliferation and complete activation of DCs. VEGF and IL-6 from tumor cells impair the function of DCs (115). TAMs and Tregs overexpress IL-10, which down-regulates the content of IL-12 released by DCs, thereby limiting the stimulation of the tumor immune response (107). TGF-β and PD-1 strengthen the inhibition of effector T cells by limiting the function of DCs. Immunosuppression-controlled DCs promote a reduction in Th1 differentiation and an increase in Th2 differentiation among CD4⁺T cells (116). This hinders the smooth execution of the initial stage of the tumor immune response and affects the entire anti-tumor mechanism network. Therefore, DCs have more functions in promoting immunosuppression and drug resistance. DCs can also be separated into CD11-positive myeloid DCs and CD11-negative lymphoid DCs. The medullary DCs’ high degree of infiltration results in a longer survival time (117). Treatment strategies based on improving DCs are also being tried.

PDAC is a heterogeneous cell population that contains numerous B-lymphocytes. The relationship between B cells and tumors is very complex, containing both promotion and inhibition of tumors. B cells can destroy tumors by producing antitumor antibodies and presenting antigens to T cells. However, some subsets of B cells also promote tumor growth by performing immunosuppressive functions (118). B cells have been observed in pancreatic precancerous lesions. Stromal cells regulate B cells through the chemokine CXCL13, which leads to their accumulation in tumors as cancer progresses (119). The hypoxia-inducible factor HIF1a forms and maintains the anoxic state of TME. Inhibiting HIF1a activity can increase CXCL13 to recruit more B cells (120). Alleviating B cells exhaustion can mitigate the development of PDAC, further highlighting the significant role of B cells in the occurrence of PDAC. The spatial distribution of B cells may influence the prognosis. B cells distributed in the matrix have been associated with terrible tumor outcomes, potentially due to increased expression of PD-L1 and TGF-β. Conversely, B cells distributed in ectopic lymph node-like TLS have been related to the prolongation of survival times, possibly due to improved capability of B cell proliferation and activation (121). The poor immunogenicity of PDAC not only makes it difficult to activate cellular immunity but also affects humoral immunity. Anti-cancer autoantibodies secreted by B cells are only found in a minority of cases (approximately 30%) (122). This makes it difficult for the humoral immune response to achieve therapeutic objectives. B cells can promote the matrix formation of TME. B cells promote the fibrogenic properties of CAFs by secreting cytokines such as PDGF-β (123). Under the influence of B cells, CAFs increased collagen expression and enhanced their ability to remodel epithelial cells into stroma. Circulating plasmablasts, which are the best at expressing of this ability in heterogeneous B cell populations, not only activate CAFs but also secrete LOXL2 that acts on collagen (124). LOXL2 enhances the stretching ability of collagen and regulates the differentiation of fibroblasts into myCAFs (125). This type of cell population is rarely observed in healthy people. On the contrary, circulating plasmablasts are specifically present in the peripheral blood of PDAC patients, especially in untreated patients. These plasmablasts are continuously exposed to tumor antigens in circulation (126). Based on this unique existence, circulating plasmablasts may become a marker for the diagnosis and prognosis of PDAC. It is unclear whether B cells consistently promote or inhibit the formation of the fibrotic matrix. In B cells deficient PDAC mouse model, the fibrogenic property is significantly weakened, showing relatively slowing tumor progression (122).

Regulatory B-cell subsets (Bregs) in B cells mainly play a role in tumor promotion. THE increased infiltration of Bregs in TME leads to poor prognosis and low survival rates (127). Bregs cause immunosuppression and drug resistance by promoting Tregs production, inhibiting the production of Th1 and inflammatory factors, and inducing the anti-inflammatory transformation of antibodies (122). Bregs interact with PDAC cells, and tumor cells release IL-18 to promote the production of Bregs (128). Bregs overexpress PD-L1, bind to CD8⁺T cells in an inhibitory manner, and reduce the secretion of IFN-γ by CD8⁺T cells, thereby assisting PDAC in escaping immune surveillance (129). IL-10 and IL-35 released by Bregs decrease the level of IFN-γ and increase the number of Tregs and tumor cells (130). Bregs secrete Bruton tyrosine kinase (BTK), inducing the differentiation of immunosuppressive TAMs in TME (131). The immunosuppressive properties of B cells in PDAC stimulate the exploration of new therapeutic methods. Targeted therapy based on the cancer-promoting mechanism of B cells may be crucial in improving the survival rate for pancreatic cancer. Evidence suggests that heterogeneous B cells contain tumor-inhibiting subsets (132). Non-selective exclusion of all types of B cells is not feasible. Precisely localizing and selecting patients with indications has become a huge challenge for therapeutic methods targeting B cells. If the B cells that constitute the immunosuppressive properties of PDAC are precisely blocked, this will provide a better working environment for immunotherapy.

As a momentous member of the innate immune system, NK cells take advantage of cytotoxicity to fight against tumors in the early stages of tumor development. However, NK cells exhibit varying degrees of infiltration in the microenvironment depending on the type of tumor. In PDAC, low-invasive NK cells are confined to the tumor periphery and have limited contact with tumor cells. Nevertheless, it has not been determined whether there is a significant relationship between the infiltration degree of NK cells and the prognosis of patients (133). NK cells require mediation by cell surface receptors to exert their anti-tumor activity. If the TME of PDAC contains numerous ligands that activate NK cells, such as CCL2, it can attract NK cells to infiltrate the tumor. The infiltration of NK cells can affect the prognosis of tumors through this mechanism. NK cells have the ability to eliminate tumor cells and identify and kill tumor stem cells. The anti-tumor function of NK cells is magnified when chemotherapy drugs are applied, resulting in a decrease in the recurrence rate of PDAC mice (134). In PDAC, a suppressed IFN-γ pathway can affect the recruitment of NK cells and lead to a poor prognosis (135). Like CD8⁺T cells, NK cells also nutrient-deficient and suppressed in their anti-tumor function. Therefore, exploring immunotherapy to enhance the function of NK cells is a focus.

Usually, people take into account the allergic properties of Mast cells (MCs) to cause diseases. However, recent studies have shown that MCs also exist in the occurrence and progression of tumors, providing support at different stages. These cells are often widely distributed in the skin, respiratory tract, and visceral mucosa. MCs migrate to TME under the action of chemokines such as VEGF in TME. MCs present in tumors have been given a new concept: tumor-associated mast cells (TAMCs) (136). The distribution of MCs in tumor tissue is inconsistent, with a notable absence at the center of the tumor tissue. The degree of MCs infiltration is negatively correlated with a good prognosis (137). MCs secrete soluble factors such as TNF-α, TGF-β, MMP, IL-8, IL-13, VEGF, histamine, and tryptase, which affect the proliferation, activation, and function of other cells in the microenvironment (138). Under the guidance of the above cytokines, MCs accelerate the expansion, diffusion and migration of tumor cells, promote the formation of blood vessels and lymphatic vessels (139). MCs can promote the proliferation of PSCs and CAFs to form the dense matrix of pancreatic cancer. MCs play an effective role in recruiting MDSCs and can block the therapeutic effect of chemotherapy drugs (140). Controlling the tumor-promoting mechanism of MCs can retard tumor progression. Some studies have indicated that MCs play an anti-tumor role by secreting IL-33 (141).

Cancer stem cells (CSCs) are a group of cells that can self-renew and differentiate like stem cells. They mix with tumor cells and exist in the microenvironment. CSCs of PDAC increase the carcinogenic effect, which is characterized by reaching the markers such as CD44 and CD24. Patients with high CD44⁺CSCs content have a worse prognosis. These cell surface markers can be stably expressed in the offspring (55). This mechanism maintains the immunosuppressive effect of CSCs. The proliferation, differentiation, and resistance therapy of CSCs benefit from anoxic acidic TME. Tumor cells located in an appropriate niche under hypoxia are prone to maintain their stemness, including an undifferentiated state, self-replication, and migration (142). CSCs with different phenotypes are subject to different degrees of hypoxia and chemical action (143). CSCs not only promote cancer cell proliferation but also metastasis through nestin, a protein marker. TGF-β increases the content of nestin under anoxic conditions and plays a positive role in regulating TGF-β (144). At present, treatment methods for CSCs are not yet mature.

Microbiota is often ignored in pancreatic cancer, but they can promote the occurrence and development of pancreatic cancer. By comparing normal pancreatic tissue, researchers found that bacterial DNA and 16S rRNA increased in PDAC tissue. Microbiota affects the TME of PDAC. The number of CD8⁺T cells and M1-like macrophages in TME of PDAC increased and the number of MDSCs decreased after inhibition of the microbiota by antibiotics. The expression of bacterial lipopolysaccharide was positively correlated with PD-L1. Oral antibiotics and anti-PD-L1 immunotherapy have synergistic effect. These effects ultimately lead to the slowing down of PDAC progression in the mouse model. However, the benefits of this combination therapy for patients with PDAC are controversial. Other tumor patients showed poor prognosis caused by antibiotics before immunotherapy. These effects ultimately lead to the slowing down of PDAC progression in the mouse model. However, the benefits of this combination therapy for patients with PDAC are controversial. Other tumor patients showed poor prognosis caused by antibiotics before immunotherapy. Other immunotherapies, such as cancer vaccines and adoptive cell therapy, may also involve microbiome responses. In PDAC, the relationship between microbiota and immunosuppression needs further exploration (145).

PDAC has two fundamental signal pathways, PD-1/PD-L1 and CTLA-4/B7,that prevent T cells from performing their immune functions. These pathways serve as the immune checkpoint of classical immunotherapy. PD-1 is an immune cell surface receptor that is up-regulated in PDAC. It combines with PD-L1 from PDAC cells to obstruct the generation of an effective immune response (181). In order to restore the proliferation, migration, and anti-tumor activity of CD8⁺T cells, it is advantageous to repress the PD-1/PD-L1 signal pathway using drugs such as nivolumab, durvalumab, atezolizumab, avelumab, and pembrolizumab. The combined inhibition of PD-1/PD-L1 and chemokines CXCR4 or CXCL12 is also safe and achievable, which can increase the infiltration of T cells and the occurrence of cytotoxic reactions (182). The level of CCL5 is indirectly affected by the overexpression of PD-L1, and co-inhibition of both can reduce the recruitment of Tregs and increase the anti-tumor effect (183). There is further evidence that the combined inhibition of PD-L1 and BTK has not yet brought more significant benefits (184). Blocking the CTLA-4 pathway with ipilimumab and tremelimumab can reduce competition with CD28, allowing for more immunostimulatory molecules to combine on the surface of immune cells. This weakens the immunosuppression of Tregs and enables expansion and utilization of more anti-tumor-T cells (185). The TME of PDAC contains a variety of vasopromoting substances. The combination of anti-angiogenesis therapy, such as Bevacizumab and Axitinib, with ICB not only ameliorate the immune response but also perfect the microvascular structure. Factors that may influence the therapeutic effect of ICI include high lymphocyte infiltration in the tumor, increased of PD-L1 levels, and high tumor mutation rates (31, 186). In addition to positive effects, toxicity and adverse events of immunotherapy should also be considered. This will promote us to study safer treatment strategies.

DCs, as APCs, are unable to complete their necessary functions in PDAC, which hinders the activity of anti-tumor T cells. At present, research on tumor vaccines mainly focuses on DCs (187). DCs are obtained from PDAC patients through leukocyte isolation technology. Tumor antigens are sometimes added to strengthen the function of activating T cells before the DCs vaccine is returned to the patients. The DCs vaccine is designed to be used together with anti-angiogenic drugs, which may be instrumental in the long-term survival of patients (188). In addition, it includes whole-cell and DNA/peptide vaccines. GVAX, one of the tumor vaccine therapies for whole tumor cells, promotes the recognition of APCs and absorbs tumor antigens by expressing GM-CSF (189). The KRAS vaccine is a type of peptide vaccine. Although the specific vaccine treatment in clinical trials is still being adjusted, it is feasible to ameliorate the patient’s condition of PDAC (23). MRNA vaccine is a new and promising vector for the development of personalized vaccines against PDAC. The mRNA vaccine overcomes the heterogeneity of the tumor and meets the requirements of precise targeting, efficiency, safety, and economy in treatment(Personalized pancreatic cancer therapy: from the perspective of mRNA vaccine). Recently, the research on the combined application of tumor vaccines and ICB is under excavation.

CAR T cells, as a new treatment for pancreatic cancer, have attracted much attention. By modifying patients’ T cells, the newly generated CAR T cells carry receptors that specifically recognize and bind tumor cell antigens and can transmit signals to induce T cells activation. The latest CAR T cells also increase the expression of costimulatory receptors (190). However, it is complex and arduous for CAR T cells to precisely target specific antigens of tumor cells. Adoptive cell therapy targeting the cell surface marker CD19 has been demonstrated to carry a high risk of tumor recurrence and acute B cells leukemia or B cells lymphoma (191). In order to effectively apply this expensive treatment strategy, the durability of CAR T cells is required. They potentially interfere with normal cells. Therefore, CAR T cells with suicide genes were proposed and examined (192). The targets entering the clinical trial stage include mesothelin(MSLN), CD133, prostate stem cell antigen (PSCA), Claudin 18.2, HER2, MUC1, carcinoembryonic antigen (CEA), etc (193). The great challenge of adoptive cell therapy is not only to prevent CAR T cells from reacting with non-target cells but also to promote their penetration of the dense matrix environment and contact with cancer cells. CAR NK cells can be used as anticancer drugs in solid tumors. NK cells are easy to obtain. Them will not cause serious side effects in allogeneic injection. At present, most trials on CAR NK cells are in preclinical or early clinical trials. The development of CAR NK cells requires more complex and more optimized transfection techniques (194). TCR T cells need to extract individual peripheral blood mononuclear cell samples and screen out the samples that can selectively bind to TCR sequence of target antigen. After the T cells in the patient’s peripheral blood are then genetically engineered and injected to eliminate tumor cells. The fourth generation of TCR T cells is a highly specific cellular immunotherapy for new antigens, with a higher safety profile. Some cases have proved that PDAC patients experience objective regression after using TCR T cells (195). Compared to CAR T cells, TCR T cells have better prospects because they are easier to enter the interior of solid cancer. Despite poor persistence, significant adverse reactions, and high consumption, adoptive cell therapy is increasingly suitable for clinical trials as it undergoes continuous improvement. Combining adoptive cell therapy with other therapies may be a promising anti-tumor method.

CD40 exists on the surface of both tumor and immune cells. It indirectly activates anti-tumor T cells by stimulating DCs to release IL-12. CD40 agonist reduces the density of the PDAC matrix, remodels macrophages, amplifies the function of DCs, and contributes to the proliferation and activation of T cells (196). Chemotherapy prior to the application of CD40 agonist can further ameliorate the therapeutic efficacy (197).