Anti-tumour immunity is the ability of the body’s immune system to recognise and eliminate tumour cells. This phenomenon has the potential to cure cancer even if cells are widely disseminated through multiple metastatic sites and has been harnessed to develop different immunotherapy drugs. With increased understanding of immune surveillance process by innate immune cells and discovery of T cell immune checkpoints, such as PD-1, PD-L1, and CTLA-4; cancer immunotherapy has significantly improved patient survival and quality of life (1–5). Treatments aim to promote successful infiltration and activation of antigen presenting cells and boost T-cells cytotoxic activity to promote anti-tumour immunity. However, despite promising results, not all tumour types or patients respond equally to immunotherapy (6–8). The major reasons for failure of immunotherapy are (1) reduced antigenicity (9–11) and (2) immunosuppressive tumour microenvironment (TME) (12–15). The TME is highly heterogeneous; consisting of tumour cells, stromal cells, extracellular matrix (ECM) and immune cell types including macrophages, dendritic cells, T and B lymphocytes, Natural killer (NK) cells, mast cells, myeloid derived suppressor cells (MDSCs) and neutrophils (16–20). The anti-tumour immune response relies on the antigen presenting cells (APCs) to prime naïve T cells. Tissue resident macrophages can activate T cells locally in the tumour; whereas dendritic cells (DCs), the professional APCs, are thought to migrate into the tumour draining lymph nodes (TDLNs) to prime T cells (21). However, immune surveillance by APCs and T-cell infiltration can be impaired by dynamic changes within the tumour microenvironment such as induction of chemokines, cytokines, growth factors, inflammation, ECM modulators and immune checkpoint proteins (22–27). This review focuses on the immunosuppressive properties of the TME and how these mechanisms alter activation, maturation and trafficking of dendritic cells to enable immune escape and tumour progression.

TLS are established at sites of chronic inflammation and can structurally and functionally resemble secondary lymphoid organs (120–122). Recent studies on murine models of TLS have shown the role of PDPN+ FAP+ immunofibroblasts in driving the development and expansion of TLSs (123, 124). These form part of the TME and can benefit from quick surveillance and locally primed immune response against tumour antigens ( Figure 2 ). Occurrence of TLS correlated with high number of mature DCs, strong T-cell infiltration and long-term survival in human primary lung, breast, colorectal, melanoma and other tumours (120, 125–128). However, factors such as TLS location, tumour stage, tumour mutations, treatment history can affect immune cell infiltration and anti-tumour response (128, 129). The cells residing in TLS in tumours are known to express Th1, CD4, CD8, CD31, CD23, FOXP3, chemokines (CCL19, CCL21) and clusters of DC-Lamp⁺ mature dendritic cells (120, 130, 131) providing an immune-supportive niche (132–134). Typically, TLS at the periphery of the tumour have more organised and distinct DC/T-cell and B-cell zones than intratumoral TLS which contain mostly B cells (133). Future research understanding the immunological features of extratumoral versus intratumoral TLS will be useful to predict responsiveness to immunotherapy and overall survival.

The other group of molecules responsible for causing dysfunction in tumour-infiltrating DCs are immune checkpoint proteins PD-L1, PD-1, ILT2, CTLA4, TIM3 expressed by tumour cells or other immune cells (135–141). As mentioned before, expression of these inhibitory molecules is variable among DC subsets. For example, PD-1 and TIM-3 are mostly expressed on cDC1s; PD-1 expression specifically has been shown to inhibit NF-kB activation which is critical for DC functions including costimulatory molecule expression, antigen presentation and cytokine release leading to T cell inactivation (78, 135, 137, 139, 140). On contrary, ILT2 is expressed on pDCs and cDC2s, but not on cDC1s (78). The central goal of immunotherapies is inhibition of immune checkpoint genes and the expansion of mature cDCs and cytotoxic CD8+ T cells within tumours. It is associated with positive patient outcomes in multiple cancer types when combined with chemotherapy or radiotherapy treatments (28, 135, 142, 143). Despite this, many patients still fail to respond to immune checkpoint blockade. A better understanding of the role of inflammatory mediators in determining tumour progression will also provide therapeutic avenues to improve immunotherapy outcomes (144–147).

Different labs have reported direct inhibition of pro-tumourigenic inflammation in combination with immune checkpoint blockade as a powerful strategy to improve the patient survival rates (27, 148–150). One such example is the use of aspirin that blocks the COX-2/PGE2 pathway and has shown promising results in preclinical melanoma models (27, 149). Prostaglandin E2 (PGE2), catalysed by the enzyme COX-2 is elevated in many tumours (151) and plays a role in tumour evasion by directly inhibiting cytotoxic immune responses and subsequently mediates expression of other inflammatory molecules such as CXCL9, CXCL10, CXCR4, CXCL12, IDO1 and interferon (IFN)-γ (27, 144, 148, 150, 152–154). Induction of CXCL12, CXCR4 and IDO1 in tumours have been associated with accumulation of myeloid derived suppressor cells (90, 155). Moreover, direct interaction of EP2/EP4 receptors (present on DCs) with the available PGE2 can modulate DC maturation, metalloprotease-driven DC motility, and immune response in tumours (27, 149, 152, 156–158). Thus, targeting the inflammatory environment of the tumour is important to restore DC function to harvest the full potential of immunotherapy.

Anti-tumour immunity relies on cross-presentation of tumour antigens by DCs to elicit a CD8+ T cell response. Among various DC subsets, cDC1s (XCR1+, CD103+) play a critical role in anti-tumour immunity. CLEC9A, (also known as DNGR1) is highly expressed on cDC1s and binds necrotic cell debris and promotes antigen processing in tumours (159–161). One of the reasons for checkpoint blockade failure is poor antigen presentation due to absence of co-stimulatory molecules and therefore modulation of DC function could increase responses to these therapies. One method to address this issue is the development of DC vaccines for cancer treatment, bypassing the need to activate and mature DCs within the tumour. DC-based cancer vaccines work by recruiting ex-vivo generated dendritic cells (or monocyte derived patient DCs) that are genetically engineered, matured, and loaded with tumour-specific antigens (162–164) or by reprogramming endogenous DCs by injecting biomaterial-based scaffolds providing favourable microenvironment for the recruitment of activated DCs (165, 166). An ideal DC vaccine must be able to increase cross-presentation by DCs, express high levels of co-stimulatory molecules, induce tumour-specific T cells with high migratory and cytolytic capabilities. Furthermore, the use of dendritic growth factor Flt3L in combination with checkpoint inhibitors or DC vaccines has improved number of activated intratumoural cDC1s and enhanced anti-tumour immunity to BRAF and checkpoint blockade in preclinical models (167–170).

Presence of co-inhibitory signals (e.g., IL-10, IL-6, PGE2, TGF-β) or absence of co-stimulatory molecules (e.g. CD80 and CD86) can result in inefficient antigen presentation by DCs and poor induction of antigen-specific CD8+T cells. Therefore, inflammatory cytokines secreted by tumour cells and tumour-associated stroma have been identified as promising candidates to potentiate current immunotherapies including immune checkpoint blockade and CAR-T therapy (149, 171–173). Stroma present around most tumours can also magnify inflammation and impede DC phenotype (174–177) and hence manipulating stroma/DC crosstalk in the TME could help improve DC function.

It is now established that tumours can exploit their surroundings to create an immunosuppressive microenvironment to control DC function within both the TME and TDLNs (178, 179). These signals including cytokines, chemokines, prostaglandins, growth factors, immune checkpoint genes, etc., may target different DC subsets infiltrating tumours and influence DC maturation, antigen uptake and DC migration (53, 180). Although the success of immunotherapy relies on enhanced T cell activity, activation of tumour-specific T cells cannot be achieved without prior antigen presentation by professional DCs. To overcome immunosuppressive signals, personalized vaccines loaded with patient-derived engineered DCs or delivery of innate stimulus such as TLR3 ligand or a STING agonist to DCs at the tumour site are being developed and have shown promising results (181, 182). Repurposing of existing anti-inflammatory drugs such as aspirin along with DC vaccines or immunotherapies has also been successfully tested in pre-clinical models (149).

This review also addresses the importance of local versus TDLN priming of anti-tumoural T cell responses. Tissue resident memory CD103+ CD8⁺ T cells residing in the non-lymphoid tissues have shown to provide local immunosurveillance and enhanced immune responses in melanoma, lung and breast tumours (183–187). Moreover, melanoma patients with higher resident T cell population responded better to anti-PD-1 immunotherapy with improved survival (188, 189). However, what is still unclear is how are tissue resident memory CD8⁺ T cells primed ( Figure 2 ) and whether there is a distinct population of DCs required to activate them. Although the exact regulatory mechanisms remain to be explored further, it is hypothesized that crosstalk between tissue resident memory T cells, tumour cells, stromal cells and DCs within the TME potentiate secondary T-cell responses against tumours ( Figure 2 ). This also opens discussion on the role of tumour associated tertiary lymphoid structures (TLSs) in intra-tumoural DC maturation; and sourcing T cells and B cells to the tumour (190). Although TLS has been positively correlated with anti-tumour responses, there are still many questions remain to be answered such as TLS composition and TLS induction at tumour site before TLS can be adopted as a predictive tool or as a therapeutic option. Our discussion demonstrates the importance of site of antigen presentation in DC maturation and trafficking which must be exploited therapeutically to enhance immune response against cancer.

YG and SA planned the concept and design of the review. YG and AK collected previous literature on the topic and drafted the article. YG made the figures. SA performed critical revision of the article. YG and SA edited the final version of the article.

This work is funded by Cancer Research UK (Career development fellowship CRUK-A19763 to SA) and Medical Research Council (MC-U12266B).

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.