Cancer is a group of more than 100 diseases acquired by cellular defects, resulting in several hallmark features, such as uncontrolled cell growth, resistance to apoptosis, immune evasion, as well as metabolic dysregulation and metastasis (1). Approximately 90% of all cancers form tumors, which are composed of cancer cells, stromal cells and tissue-resident and infiltrating immune cells (2). The composition of the tumor, which includes all cellular and acellular factors is referred to as the tumor microenvironment (TME) (3–5). Cancers can arise in virtually all tissues of the body and is driven by a wide range of different inherited and acquired mutations, resulting in immense heterogeneity (6–8). At their core cancers are characterized by the loss of proliferative control, typically caused by the inactivation of tumor suppressor genes or the hyperactivation of oncogenes. Due to their role in regulating growth and cell cycle progression loss-of-function mutations in tumor suppressor genes results in loss of proliferative control, whereas activation of oncogenes promote uncontrolled growth (9–11). Most commonly, cancer is caused by mutations in genes encoding p53, PIK3CA, FAT4 and KRAS, with p53 mutations being observed in more than 50% of cancers (12, 13). Moreover, single-cell sequencing and spatial transcriptomics have further revealed that heterogeneity exists within the tumor itself (14–16).

The rapid rate of proliferation of cancer cells has been exploited therapeutically for decades. Chemotherapeutic drugs and radiotherapy preferentially target rapidly proliferating cells by either inducing DNA damage or by blocking central pathways involved in DNA replication (17). The rapid growth coupled with dysfunctional DNA repair boosts the accumulation of mutations, and thus accelerates the rate of cancer evolution. This makes the treatment of cancer especially challenging, as cancer cells not only develop resistance to chemotherapeutic drugs but can also adapt to harsh environmental conditions in the TME, such as hypoxia and nutrient deprivation, further enhancing their survival (18, 19). Although sequencing of cancer cells is increasingly used to predict treatment efficacy, primary and acquired therapy resistance still prevents efficient treatment of cancer in many patients (20). Moreover, as most cancer drugs work by targeting rapid proliferation, this also affects healthy cells with a high proliferation rate, often resulting in hair loss, gastrointestinal distress and immune suppression, among others (17). Together, this highlights the need for additional cancer targeting strategies. The ability to acquire resistance to treatment indicates that a single target is most likely insufficient to efficiently to adequately treat cancer patients. The identification of non-redundant pathways may result in a synergistic effect, thus requiring lower doses and potentially reducing adverse effects. This has been proven safe and effective in treatment of hypertension, and in vivo models suggest this also has the potential in cancer immunotherapies (21–23).

The immune system is a complex, interactive network of defense and surveillance mechanisms, comprising physical barriers such as the skin and mucosal surfaces, as well as specialized lymphoid organs, immune cells and molecules. Together, these components work in concert to protect the host from non-self, including pathogens. The active immune system is divided into two core responses, namely the innate and the adaptive immune systems, which are fundamentally differentiated by their speed, precision, and capacity to resolve infections in addition to differ in their ability to develop immunological memory (24, 25). The innate immune system serves as the first line of defense, continuously surveilling the body for general signs of infection or damage by recognizing conserved molecular patterns - pathogen association molecular patterns (PAMPs) or damage-associated molecular patterns (DAMPs) (26). Upon detection of foreign or damaged self, cells of the innate immune system are rapidly activated and recruited from the circulation within minutes to hours, eliciting a response that can be sustained for several days (25). Although the innate immune system lacks immunological memory, professional antigen-presenting cells (APCs) can also induce activation of the adaptive immune system, composed of B and T lymphocytes that can maintain memory to specific pathogens lasting decades (25, 27–29). This task is mainly performed by type 1 macrophages (M1), dendritic cells (DCs), and B lymphocytes. APCs capture antigens, process them, and present resulting antigenic peptide fragments via their major histocompatibility complex class II (MHC II) molecules, expressed on their surface to activate adaptive immune cells in lymphoid tissues (25, 30, 31). While MHC II molecules are exclusively expressed by APCs, all nucleated cells express MHC class I (MHC I) molecules, which present endogenous antigenic peptides to enable immune surveillance and thereby the elimination of infected or abnormal cells (30, 31). For T cells, antigenic peptide-MHC complexes are recognized by the T cell antigen receptor (TCR), which can discriminate between self- and non-self-molecules with remarkable specificity (31–33). Forming an essential part of the TCR is the CD3 complex, functioning as an intracellular signaling hub that translates extracellular antigen recognition into downstream signaling events that drive T cell activation to differentiation, effector function, and clonal expansion (33).

T cells are broadly classified into two helper and cytotoxic T cells, designated CD4+ helper (Th) and CD8+ cytotoxic T cells (CTLs), respectively (24). CD4+ T cells recognize antigens presented on MHC II molecules and play a central role in regulating and coordinating immune responses. The CD4+ T cells can be broadly categorized into CD4+ effector cells, orTh cells, and regulatory T cells (Tregs). While CD4+ effector T cells are crucial for clearing infection and repairing tissue damage, Tregs are responsible for preventing excessive tissue damage and autoimmunity, striking a balance to maintain a functional and healthy immune environment (34–37). In contrast to CD4+ cells, CD8+ CTLs recognize antigens presented on MHC I molecules, and are responsible for directly eliminating target cells through the release of effector cytokines or cytotoxic granules (38). Although most effector T cells, including both CD4+ and CD8+ subsets, have a transient lifespan, a small fraction differentiates into memory T cells, thereby ensuring long-term immune surveillance against the same antigen (39).

The immune system’s intrinsic role in defending against non-self has fueled the longstanding hypothesis that immune cells can recognize cancer cells as foreign and thereby be weaponized to eliminate them. Indeed, the antigenic composition of tumors differs significantly from that of their non-transformed tissue counterparts, a distinction largely driven by their genetic instability – a core hallmark of cancer (40–42). The concept of leveraging the immune system to combat cancer, now known as immunotherapy, was first systematically introduced in the 1890s by William B. Coley, who documented several cases of spontaneous remission after administering a cocktail of killed bacteria and their products to stimulate the immune system in patients with inoperable cancer (43). What began as a foundational discovery led to decades of rigorous research in cancer immunology, ultimately positioning immunotherapy as the fourth cornerstone of cancer treatment alongside surgery, radiotherapy and chemotherapy (44). Immunotherapy now encompasses a wide variety of treatments, including immune checkpoint therapy (ICT), which will be the focus of this review, adoptive cell transfer (ACT), such as chimeric antigen receptor (CAR) T cells, as well as engineered antibodies (reviewed extensively in (45)).

The concept of immunotherapy builds on the ability of controlling aberrant antigen-induced immune cell activation (46). T cell stimulation through the TCR/CD3 complex requires concurrent perturbation of co-receptors. Whereas the initial interaction of the TCR with the MHC molecule secure antigen-specificity, co-receptor signaling is essential for tuning the activation process. This tuning depends on a balance between activating and inhibitory signals induced by perturbation of cell surface receptors with distinct functions. This is exemplified by the interplay between CD28 and the cytotoxic T-lymphocyte-associated protein 4 (CTLA-4) molecules, which compete for binding of CD80 and CD86 expressed on APCs (33). Whereas stimulation of the CD28 molecule provides a positive signal, CTLA-4 engagement delivers an inhibitory signal, thereby modulating the magnitude of initial T cell activation (33, 47). Additionally, programmed cell death receptor 1 (PD-1), also referred to as CD279, is activated by PD ligand-1 or 2 (PD-L1/2) and downregulates T cell activity. In line with this, PD-1 stimulation has an important role in regulating immunological tolerance and is vital in preventing autoimmunity and collateral tissue damage (48–51). Additionally, inflammatory cytokines elicited by the inflammatory process are essential signaling molecules that drive a productive T cell response and facilitate memory formation by shaping the activation of specific differentiation pathways within the cell (52–54).

It is now well established that cancer cells can trigger immune responses. This phenomenon is demonstrated by the utilization of tumor-infiltrating lymphocyte (TIL) therapy, where T cells are isolated from resected tumors,expanded ex vivo, then transferred back to the patient (55, 56). However, they also evolve mechanisms enabling them to evade immune detection and destruction, making immune evasion a defining hallmark of cancer (57–59). During cancer evolution, these mechanisms are continuously sculpted under the selective pressure of immunosurveillance, a phenomenon known as immunoediting (59). In this process, as patrolling immune cells selectively eliminate highly immunogenic cancer cells, they simultaneously impose a selection pressure that favors the survival and expansion of rare subclones with immune-evasive traits. Over time, these subclones can adapt, proliferate, and ultimately become immune-resistant (59). Consequently, from the earliest stages of tumor development, the immune system edits tumor immunogenicity, leading to the emergence of an immunoedited tumor dominated by cancer cell variants that have successfully evaded immune control.

Immune evasion can occur through multiple, non-mutually exclusive mechanisms (reviewed in (59) and (60)), with loss of tumor antigen presentation being one of the most well-characterized, allowing cancer cells to effectively hide in plain sight. This can result from defects in the machinery responsible for antigen processing and presentation, such as downregulation or loss of MHC I molecules on the cell surface, a phenomenon observed in 40-90% of cancers which shields cancer cells from recognition and elimination by CTLs (61, 62). In blood cancers, this “invisibility cloak” has been successfully targeted through the development of monoclonal antibodies or CAR T cells, another form of ACT, that recognize and bind tumor-specific surface antigens independently of the MHC I receptor (63). However, these approaches have shown limited efficacy against solid tumors, which, as mentioned, account for approximately 90% of all cancers (2).

In addition to avoiding detection from the immune system, cancers can also suppress the effector functions of anti-tumorigenic immune cells. A key example is the co-option of the PD-1 pathway by cancer cells (48, 49, 51). In addition to activated T cells, PD-1 expression is also detected on B cells and natural killer (NK) cells. In all three cell types, PD-1 expression is normally decreased when an inflammatory response is resolved during acute antigen clearance (50, 64). However, in cases of persistent antigen exposure, such as cancer and chronic infections, PD-1 expression remains elevated, which contributes to T cell exhaustion (48, 51, 65, 66). The two PD-1 ligands, PD-L1 and PD-L2, exhibit distinct expression patterns. While PD-L2 is predominantly expressed on APCs in lymphoid tissues, PD-L1 is broadly expressed across hematopoietic (e.g., T cells, B cells, macrophages) and non-hematopoietic cells (e.g., endothelial cells) (51, 67). PD-L1 is upregulated by pro-inflammatory cytokines, particularly interferon-γ (IFN-γ), as a feedback mechanism to tune down immune activity (50, 68). However, in many solid tumors, PD-L1 expression is aberrantly elevated within the TME due to constitutive oncogenic signaling or as an adaptive response to inflammatory cues (51, 65).

To restore T cell activity in anti-tumor immune responses, antibodies targeting co-inhibitory receptors and their ligands have been developed (48, 51, 69, 70). These include, but are not limited to, immune checkpoint inhibitors targeting the CTLA-4 and PD-1 pathways, which have demonstrated clinical efficacy in certain cancers and patient subsets (70). However, as many as 60-80% of patients with solid tumors either fail to respond or experience only transient benefits from ICT, highlighting the diverse mechanisms tumors employ to evade the immune system (71, 72). In solid tumors, many of these mechanisms are driven by the TME, a complex and dynamic ecosystem encompassing cellular, physical and chemical components that are continuously restructured and manipulated throughout tumor progression. In addition to cancer cells, the cellular constituents of the TME include diverse non-malignant stromal cells along with tissue-resident and infiltrating immune cell populations that can be reprogrammed to support tumor growth (5, 73–75). In addition to cellular components, non-cellular components – such as the extracellular matrix (ECM), metabolites, soluble signaling molecules, and the surrounding hypoxic and acidic milieu – play important roles in tumor progression and therapy resistance (5, 73–75). Targeting the TME and its diverse components to enhance immunotherapy holds significant therapeutic promise, as many of its defining features are conserved across a range of tumor types (5, 75). In this review, we will provide a brief overview of key immunosuppressive hallmarks of the TME and discuss their potential as therapeutic targets to enhance immunotherapy in cancer, with a particular focus on ICT.

Over the last decades it has become clear that extracellular metabolites are crucial for optimal T cell activation. Nutrients and their metabolites exhibit significant interplay with the three core activation signals while also independently influencing the functional polarization of T cells. Consequently, metabolic inputs have been proposed as a novel dimension necessary for licensing the T cell immune response (reviewed in (76) and (77)). The connection between cellular function and metabolic phenotype in health and disease is therefore becoming increasingly evident across multiple fields, including immunology, cancer, and tumor angiogenesis (41, 52, 78–83). While energy and biomass production remain critical, recent insights explain how metabolism affects diverse processes such as activation, proliferation, migration and differentiation.

Although metabolic reprograming has only recently emerged as important in these processes, it has been known in cancer cells for close to a century. As early as in the 1920’s, Otto Warburg discovered that cancer cells, despite the presence of oxygen, preferentially rely on glycolysis and lactate production rather than mitochondrial respiration, a metabolic reprogramming now known as “The Warburg effect” ( Figure 1 ) (84, 85). Cancer cells frequently upregulate glucose transporters and key enzymes in the glycolytic pathway, including Glucose transporter type 1 (GLUT1), Hexokinase 2 (HK2), 6-phosphofructokinase 2/fructose 2,6-bisphosphatase 3 (PFKFB3), pyruvate kinase muscle form 2 (PKM2) and lactate dehydrogenase A (LDHA) (86). Together, these proteins facilitate a rapid turnover of glucose and a subsequent increase in lactate production. Although this metabolic phenotype was originally believed to be caused by a mitochondrial defect in cancer cells, it is now recognized that the mitochondria remain functional and possess remarkable metabolic plasticity. This flexibility enables them to dynamically utilize a wide range of substrates, including glutamine and fatty acids to fuel diverse cellular processes ( Figure 1 ) (87–89).

This metabolic plasticity may also provide resistance to metabolic inhibitors, as demonstrated by Boudreau et al. (90), where the glycolytic pancreatic cancer cell line MIA PaCa-2 adopted an oxidative phenotype after long-term exposure to an inhibitor of lactate production. In line with this, glutamine reliance, a phenomenon sometimes referred to as “glutamine addiction”, is observed in several cancer cell lines (83, 87, 88, 91, 92). Although most cells are capable of synthesizing glutamine, rapidly proliferating cells rely on additional extracellular sources, so it is sometimes referred to as a conditionally essential amino acid (88, 91, 93). Interestingly, glutamine deprivation has been shown to reduce the rate of glycolysis by regulating both the expression and activity of key glycolytic enzymes (94, 95). Furthermore, the expression of the enzyme glutaminase 1 (GLS1), which is responsible for the deamidation of glutamine to glutamate, is upregulated in many cancer cell lines and correlates with decreased survival in patients (96). Indeed, Reinfeld et al. (97) showed that myeloid cells and T cells have a higher capacity for glucose uptake than cancer cells, while cancer cells have a higher capacity for glutamine uptake. The enzymes isocitrate dehydrogenase (IDH) 1 and 2 are also important in the TCA cycle, as they catalyze the formation of α-ketoglutarate from isocitrate (98). This supports redox homeostasis and lipid synthesis, along with providing α-ketoglutarate which can regulate epigenetics or be used as a backbone for glutamate and glutamine synthesis (98–100). In cancers, mutations in IDH1 or IDH2 can result in the formation of D-2-hydroxyglutarate (2-HG) (98). Some cancer cells also rely on fatty acid oxidation, a trait which is associated with upregulated expression of the mitochondrial fatty acid transporter carnitine palmitoyl transferase 1 (CPT1) in certain tumors (89).

Metabolic plasticity and reprogramming are important for both cancer and immune cells. An increasing body of evidence demonstrates that, in some contexts, healthy cells - including immune cells and endothelial cells - can adopt similar metabolic phenotypes (101–103). Indeed, immune cells, including macrophages and T cells, undergo metabolic reprogramming upon activation (81). Macrophages, which are traditionally classified into pro-inflammatory M1 or anti-inflammatory, wound-healing M2 subtypes, adapt to a glycolytic and oxidative metabolic program, respectively ( Figure 2 ) (104–109). Similar to cancer cells, M1 macrophages upregulate PFKFB3 to boost the rate of glycolysis (104). In line with this, deletion or inhibition of PFKFB3 in macrophages is shown to result in reduced secretion of pro-inflammatory cytokines, including IL-1β, IL-6 and tumor necrosis factor following stimulation with lipopolysaccharide. Moreover, mice with a myeloid-specific PFKFB3 deficiency exhibit increased survival in a murine sepsis model, as well as increased lymphangiogenesis following myocardial infarction (110, 111). Additionally, the tricarboxylic acid (TCA) cycle is used by M1 macrophages to produce succinate, which stabilizes hypoxia-inducible factor 1 α (HIF-1α), and citrate, which serves as a precursor for fatty acid synthesis and the antimicrobial metabolite itaconate (112, 113). In contrast, M2 macrophages utilize glutamine and fatty acids as substrates for the TCA cycle, fueling adenosine triphosphate (ATP) production through oxidative phosphorylation (114). Additionally, they metabolize tryptophane via the enzyme Indoleamine 2,3-dioxygenase (IDO) to generate the anti-inflammatory metabolite kynurenine (115, 116). IDO1 activity suppresses T-cell activity by depleting tryptophan, which is required for Th1 and CD8+ T-cell-mediated immunity, while kynurenine binds to the aryl hydrocarbon receptor (AHR), which directly activates Treg differentiation and activity, resulting in reduced anti-tumor responses (116–118).

Metabolic reprogramming is also important for T cells, with different T cell subsets adopting distinct metabolic adaptations ( Figure 3 ) (103, 119, 120). In their naïve state, T cells primarily rely on oxidative phosphorylation to produce ATP, maintaining a relatively low metabolic rate prior to activation. However, upon activation through perturbations of the TCR/CD3 complex in conjunction with CD28, their metabolic rate rapidly increases (121, 122). This process is aided by the presence of abundant mRNA encoding key glycolytic enzymes, particularly HK2, along with the availability of idle ribosomes, enabling the rapid production of proteins upon upregulation (121). Additionally, activation of Akt, also known as protein kinase B, quickly upregulates the cytosolic localization of GLUT1 required to increase glucose uptake (123). Chang et al. (124) demonstrated the metabolic plasticity of T cells by replacing glucose with galactose. Despite their inability to metabolize galactose through glycolysis, T-cell proliferation is reportedly unaffected by replacing glucose with galactose, however glucose deprivation resulted in decreased production of the cytokine interferon-γ (IFN-γ), as idle glyceraldehyde 3-phosphate dehydrogenase (GAPDH) bound the IFN-γ mRNA and prevented its translation (124) In addition to glucose, T cells also require glutamine to become fully activated, and glutamine deprivation or inhibition of GLS1 reduces both proliferation and cytokine secretion in CD4+ T cells (125–128). In fact, we recently demonstrated that, similar to cancer cells, glutamine deprivation in CD4+ T cells also regulate glycolysis (129). Furthermore, CD4+ T cell subsets also adapt to distinct metabolic profiles ( Figure 3 ), as previously reviewed in detail (130)), highlighting the close relationship between metabolism and functionality in T cells.

Metabolism in endothelial cells has also been extensively studied due to their critical role in angiogenesis. Importantly, deciphering metabolism in these cells has led to the identification of therapeutic targets with the potential to regulate angiogenesis (reviewed in (131)). It is known that glycolysis may account for up to 85% of the ATP production in endothelial cells despite sufficient levels of circulating oxygen (102). Notably, the rate of glycolysis can be upregulated in response to pro-angiogenic and inflammatory factors (102, 132–135). In fact, upregulation of glycolysis through increased expression of PFKFB3 is a strong driver of tip cell formation in vessel sprouting (102, 135). However, the proliferating tip and stalk cells, which are the building blocks of sprouting angiogenesis, rely not only on glycolysis, but also on fatty acid oxidation to produce dNTPs for DNA synthesis, and on glutamine metabolism to fuel the TCA cycle, thereby driving vessel propagation ( Figure 4 ) (93, 101, 136).

Despite the differences in metabolic profiles, most of the regulatory mechanisms are shared and conserved across cell types. Among these are mammalian target of rapamycin complex 1 and 2 (mTORC 1 and 2), HIF-1α and c-MYC (137–140). mTOR is a serine/threonine kinase that functions as part of two distinct complexes: mTORC1 and mTORC2. mTORC1 is an important regulator of anabolic metabolism, including protein and lipid synthesis, while also supporting catabolic processes by enhancing glycolysis through stabilization of HIF-1α and inducing enzymes responsible for glutaminolysis via the transcriptional activity of c-MYC (141). mTORC2 is associated with cell survival and fine-tuning of metabolic activity. It promotes fatty acid oxidation by regulating the transcription factor Forkhead box 01 (FOXO1) and enhance glycolysis through activation of Akt (142).

Stabilization of HIF-1α induces glycolytic metabolism through upregulating expression of GLUT1, HK, PFKFB3, PKM2, LDHA and Monocarboxylate transporter 4 (MCT4) (104, 143–146). Although HIF-1α is primarily stabilized by the absence of oxygen, several mechanisms can also promote its stabilization in the presence of oxygen, including the regulation by mTORC1 (141, 142). In cancer, HIF-1α stabilization is associated with a dismal prognosis for patients (147, 148). However, it also plays an important role in the functional activation of immune cells, highlighting its dual role in both promoting cancer cell growth and supporting anti-tumor immunity [reviewed in detail (149)].

The TME is generally poorly vascularized, with dysfunctional and leaky blood vessels resulting in decreased availability of nutrients, hypoxia and acidification that collectively contribute to immune suppression through several mechanisms ( Figure 5 ). Whereas hypoxia is associated with increasing tumor mass, acidification is associated with reprogrammed cancer cells producing lactate. Tumor hypoxia induces stabilization of HIF-1α which, as mentioned, drives glycolysis and is hence responsible for the lactate production (4, 5, 74, 75, 97, 150). Lactate accumulation is known to restrict T-cell proliferation by disrupting the redox homeostasis and inhibiting GAPDH activity (151). Interestingly, T cells produce acidic niches within lymph nodes to restrict their own effector functions to avoid hyperactivation, demonstrating the physiological importance of lactate (152). In addition to acidification, lactate is known to induce histone modification referred to as lactylation, which is associated with enhanced polarization of M2 cells, suppressed T cell effector functions and increased Treg differentiation, thereby supporting an anti-immune and pro-tumorigenic phenotype (153–155). In line with this, reduced lactylation has been associated with a favorable outcome in patients with solid tumors (153, 156). In addition to regulating glycolysis, HIF-1α stabilization can also act immunosuppressive by inducing the expression of PD-L1 in DCs, macrophages and myeloid derived suppressor cells (149). The lactate-induced acidification has also been shown to reduce the efficacy of immune checkpoint inhibition by influencing the binding properties of antibodies targeting PD-L1 (157). The hypoxia-lactate axis also contributes to the formation of dysfunctional blood vessels, together favoring tumor metastasis and repression of immune cell infiltration (158, 159).

Tumor hypoxia also induces expression of CD39 and CD73, which together catalyzes the formation of adenosine from adenosine monophosphate (AMP) and ATP. Adenosine exerts its effects through ligation of the 4 subtypes of purinergic adenosine receptors (A1, A2A, A2B, and A3) (160, 161). The ARs differ in their affinity for adenosine, but are all coupled to various cellular signaling pathways through G-protein coupled receptors (GPCRs). To this end, A2AR and A2BR are upregulated in response to hypoxia and anergic signaling in T cells and induce immune suppression by activating adenylate cyclases (ACs), that initiate the synthesis of intracellular cyclic AMP (cAMP), which in turn activates the cAMP-dependent protein kinase A (PKA) (161). PKA is a holoenzyme consisting of a regulatory (R) subunit dimer and two catalytic (C) subunits. While the R subunits are encoded by four separate genes (PRKAR1A and -B, PRKAR2A and -B), the catalytic subunits are primarily encoded by two major genes, PRKACA and PRKACB, which give rise to several tissue- and cell-specific splice variants, including immune cell-specific Cβ2 (162, 163). In addition to Cβ2, the splice variants Cα1 and Cβ1 are expressed in immune cells. The prevailing dogma is that PKA activation suppresses both early and late phases of T cell activation, including proliferation and clonal expansion, the latter mediated by downregulation of IL-2 production (164, 165). In line with its inhibitory role, PKA also promotes differentiation into Tregs and Th2, which favors the tumor, while repressing effector functions of Th1, Th17 and CD8+ T cells (166, 167). Additionally, Tregs can directly induce PKA activation by transporting cAMP into target T cells through the gap-junction protein connexin 43 (Cx43) (168). As a result, in the TME, PKA activation drives T cell exhaustion, in conjunction with upregulation of PD-1 and CTLA-4 expression (72). Recently, we showed that deletion of the immune-specific PKA Cβ2 significantly suppressed tumor growth and enhanced survival in a murine metastatic cancer model (169). This was associated with increased infiltration of pro-inflammatory Th1, Th9 and Th17 cells into the tumors (169). Moreover, in 2020, Na et al. demonstrated that knockout of PKA Cβ in macrophages prevents M2 polarization and that liposomal delivery of PKA inhibitors to tumor-infiltrating macrophages enhances the therapeutic efficacy of anti-CTLA-4 antibodies, effectively counteracting breast cancer tumor growth and metastatic potential in mice (170). Together this suggest that Cβ may convey signals supporting a proinflammatory phenotype. In support of this, mice that are ablated for Cβ2 are prone to develop autoimmunity, a phenotype also reflected in upregulation of proinflammatory immune cells (171).

In addition to the hypoxia-lactate-adenosine axis, some cancer mutations result in metabolic phenotypes that contribute to an immunosuppressive TME. Cancer cells with IDH1 or IDH2 mutations cause accumulation of 2-HG, which supports the cancer cells by maintaining a stem-like phenotype, while suppressing T-cell- and macrophage-mediated immune activity (172). 2-HG is taken up by CD8+ T cells, where it destabilizes HIF-1α and acts as an inhibitor of LDH, resulting in reduced chemotaxis, cytotoxic activity and production of IFN-γ (172–174).

As the TME exerts broad immunosuppressive characteristics, targeting the TME also offers the potential for new therapies. As mentioned, the hypoxia-lactate axis and the hypoxia-adenosine-PKA axis are known to inhibit the immune system. Thus, targeting hypoxia directly by hyperbaric oxygen (HBO) treatment has been proposed (175). Although this approach is reported to enhance immune activity and reduce growth of pulmonary tumors in a mouse model (176) a recent meta-analysis found overall weak evidence that HBO treatment alone improves long-term survival (176) and its use in humans remains limited due to the lack of high-quality studies (177). Although HBO treatment might not be effective as a cancer treatment, targeting the downstream effects of hypoxia on metabolism, PKA activity and angiogenesis may still have therapeutic benefits. Moreover, targeting IDH1 and IDH2 have been explored. This is due to the fact that the cancer-specific mutation of IDH has potential to be targeted without affecting the non-mutated IDH isoforms in healthy cells (178).

Developing a drug is a lengthy and complex process typically involving several stages that include early drug identification and optimization followed by preclinical development and application for regulatory approval to initiate clinical trials (232). Once regulatory approval is granted, the compound enters clinical testing in humans, which is conducted in at least three phases. Phase I focuses on evaluating safety and determining appropriate dosage; Phase II assesses efficacy and monitors for adverse side effects; and Phase III involves large-scale trials to confirm the efficacy and safety in a broader patient population and compares the new treatment to current standard-of-care therapies (232). During this process more than 90% of preclinical drug candidates are disqualified for further development (233). Due to the complexity of metabolism, redundancies in metabolic pathways and the fact that various cells share vital metabolic features, most drugs developed to target metabolic enzymes show low efficacy or will have side effects. An example is the glutamine antagonist 6−diazo−5−oxo−L−norleucine (DON). DON, which is a non-proteogenic amino acid that blocks glutamine metabolism in all cells, resulting in severe adverse effects in patients due to distinct roles of glutamine in different cells and tissues (234, 235). Because of this, research on novel glutaminase inhibitors have led to the identification of several compounds including inhibitors BPTES which is an allosteric inhibitor with a low µM IC₅₀ (236) and CB839, which is effective in the low nM range (237). Both BPTES and CB839 are glutaminase isoform type 1 specific with different mechanisms of action where CB839 appears to be well-tolerated with few side effects by patients and it is currently in several clinical trials for the treatment of various diseases including different cancers (238). ( Table 1 , https://clinicaltrials.gov/).

Furthermore, it has become clear that efficacy of drugs targeting metabolism may be limited due to the inherent flexibility, compensation and redundancies in metabolic pathways (90). As a result, even inhibitors with nanomolar affinity for their target may show limited therapeutic effect when used as monotherapies. Several examples illustrate this challenge. For instance, drugs targeting glycolytic enzymes downstream of hexokinase can be bypassed through the pentose phosphate pathway (PPP) where glucose 6-phosphate is shunted into the PPP and re-enters glycolysis as intermediates, effectively circumventing vital metabolic steps in the glycolytic pathway (239). Similarly, pyruvate from glycolysis can enter the TCA cycle via different routes depending on oxygen levels, allowing cells to maintain energy production under both aerobic and anaerobic conditions. Another example comes from the fact that pyruvate can enter the TCA cycle via different routes depending on oxygen levels in both an oxidative and energy-dependent fashion (240). In addition to this, drugs targeting LDH may simply shift metabolism towards a more oxidative phenotype and in that way be inefficient in inhibiting energy extraction in rapidly proliferating cells (90). Moreover, inhibiting glutaminase in glutamine consuming cells, may be compensated for by increased combustion of glucose, during which lack of carbons from glutamine is compensated for by glucose (88, 241). Finally, drug sensitive cancer cell lines can also develop resistance to metabolic inhibitors by rewiring their metabolic programs (90, 242–246). Together, these examples highlights the limitations of metabolic inhibitors when used alone and underscore the need for combination therapies. this stresses the complexity of metabolic inhibitors used as mono therapeutics. Due to this the use of metabolic inhibitors in conjunction with other treatments is emerging.

The rationale behind combinatorial strategies in cancer treatment includes targeting more than one metabolic process/pathway simultaneously reducing the required dose of individual agents and minimizing the risk of tumor immune evasion. Using the combination of drugs targeting metabolic pathways that support an immunosuppressive TME with ICT has been shown to further boost the anti-tumor responses of the immune system in preclinical models. In line with this, the IDH inhibitor ivosidenib is currently being tested as a candidate for combination therapy across several clinical trials in combination with PD-1 blockade ( Table 1 ). There are also several non-metabolic drug candidates targeting cell surface receptors such as A2A and A2B, which have entered clinical trials in combination with PD-1 blockade, and while early reports indicate some adverse effects, including autoimmunity, these are considered manageable (247). The evidence supporting the use of inhibitors targeting glycolysis or glutamine metabolism in combination with ICT in patients is currently lacking, and hence need further exploration. At present, LDH inhibitors such as FX11, GNE-140, NCI-737, Galloflavin have also been postulated to be used to prevent tumor immune evasion when used in conjunction with PD-1 blockade (248–251). The same, but less convincing, is the case with CB839 in conjunction with PD-1 (96). In a mouse melanoma models, CB839 on its own has little effect, but when combined with anti−PD−1 but also anti−CTLA−4, it significantly suppressed tumor growth and increased infiltration of CD4⁺ and CD8⁺ T−cells (252). However, a phase I/II study of the safety and efficacy of CB839 in combination with the PD-1 inhibitor nivolumab in patients with metastatic melanoma, renal cell carcinoma, and non-small-cell lung cancer was well tolerated, but did not show increased efficacy (253). The reason for this is not known. However, the patients were not stratified based on metabolic phenotyping. As the metabolic landscape is highly variable across patients and tumors (254–256), future clinical trials should attempt to incorporate metabolic profiling to determine whether specific metabolic phenotypes correlate with improved outcomes of combination therapies.

Another obstacle in developing drugs targeting metabolism is that drugs may fail to reproduce the beneficial effect seen in preclinical models, and thus, be screened out in early clinical trials (232, 233). A striking example of this is the ECHO-301 trial, a phase III clinical trial where the IDO1 inhibitor epacadostat in combination with anti-PD1 treatment failed to provide a significantly improved patient outcome (257). However, Muller et al. (258) argues that there are several points that were inadequately discussed, which may explain the outcome, including uncertainty of whether IDO1 activity was sufficiently inhibited within the tumor, pathways bypassing IDO1 were not considered and the choice of immunotherapy over DNA damaging therapy, highlighting the need for increased understanding of metabolism within the TME. The lack of effect of a metabolic drug targeting the TME in clinical trials may be attributed to the fact that most inhibitors are screened in single cell cultures and homogenous tumor models. The latter may encompass human tumors in patient derived xenografts (PDX) animal models that may not encapsulate the complexity of tumors in the individual patient. In line with these tumors are frequently sequenced to determine patient-specific features to determine prognosis and treatment strategy. However, downstream of genetic mutations patient-specific metabolic profiles may require differential treatments despite that patients may harbor related tumors and identical oncogenic mutations. Because of this, it may be necessary to also determine metabolic phenotypes to better utilize metabolic inhibitors. Metabolic phenotypes in e.g. the TME have until now been difficult to determine. However, with extracellular flux analysis using Seahorse technology coupled with Flow cytometry has opened for more opportunities and more accurately in profiling tumor metabolic phenotypes from biopsies (259). Seahorse technology has been used to determine the metabolic phenotype of a wide array of cell types, mitochondria, 3D cell culture spheroids and now recently intact tissue biopsies (260). Seahorse profile analysis when combined with bioinformatics and artificial intelligence may in the future be useful and potential instrumental in determining combinations of drugs and treatment regimens for patient-specific targeting.

ICT is considered a game-changer in modern cancer treatment. However, favorable responses are observed in only 20-40% of patients with solid tumors (71, 72). Moreover, even when effective, current treatment strategies are often associated with a wide range of adverse effects, including liver, kidney and cardiovascular toxicity, and ICT may trigger autoimmune responses (261). These limitations highlight the need for additional therapeutic targets that can enhance anti-tumor efficacy while minimizing the side effects. Identifying targets capable of inducing synergistic or multifaceted responses might reduce the required treatment doses, thereby limit off-target toxicity while enhancing tumor clearance.

We have briefly summarized how hypoxia-driven metabolic processes in the TME contribute to the reprogramming of infiltrating immune cells and the development of a dysfunctional tumor vasculature- both of which aids cancer immune evasion and hinder effective drug delivery to the tumor. Although these factors currently pose a barrier to efficient cancer therapy, advancing our understanding of these mechanisms may enable the development of new treatment strategies for solid tumors. Given the central role of the tumor vasculature in the TME, anti-angiogenic drugs are being used in cancer therapy, and it is hypothesized that their combination with ICT may further enhance anti-tumor immune activity (262–264). However, although angiogenesis can be blocked by targeting the VEGF pathway, resistance to VEGF blockade is common (185, 187). Moreover, the combined blockade of VEGF and ICT is also correlated with adverse effects, including an increased risk of cardiovascular disease, highlighting the need for alternative therapeutic targets (265, 266). In this context, targeting metabolism in the TME might offer an alternative strategy for combination therapy. However, this approach requires the identification of metabolic targets - such as enzymes and pathways - that can be safely targeted, ideally offering synergistic effects when combined with existing therapies. Notably, the PFKFB3 inhibitor 3PO was shown to increase vessel integrity and enhance tumor perfusion, resulting in decreased hypoxia and increased drug delivery (132). As hypoxia results in upregulation of CD39 and CD73, increased tumor perfusion may also abrogate PKA-mediated immune suppression through reducing adenosine production in the TME (161). However, as PFKFB3 and glycolysis are also important in immune cell activation, it remains unclear if this would ultimately enhance or impair the efficacy of ICT (52, 104, 110, 111, 218). However, inhibition of LDH has been shown to reduce tumor growth in immunocompetent mice, but not in RAG knockout mice, indicating that inhibition of glycolysis may be beneficial in combination with ICT (249). Moreover, blocking lactate transport by targeting MCT1 and MCT4 has been associated with enhanced efficacy of PD-L1 blockade (210, 221). While this may partly stem from effects on reducing lactate-induced inhibition of T cells, reducing tumor acidification may also increase antibody affinity within the TME (210, 267). Glycolysis can further be targeted indirectly by disrupting glutamine metabolism via inhibition of the transcription factor MondoA (94). Although glutaminolysis is required for adequate T cell proliferation and cytokine secretion (126, 129), the GLS1 inhibitor CB839 - which is currently approved for phase 1B clinical trials - has shown minimal effects on CD4+ T cells at higher doses than those required for growth inhibition in cancer cells (125). Moreover, CB839 reduces endothelial cell proliferation without cytotoxic effects, as well as promoting M1-like macrophage polarization, suggesting its potential for combination with ICT to further boost anti-tumor immune responses (198). Additionally, these strategies may also be combined with drugs targeting cancer-specific mutations, including mutated IDH1 and IDH2, which are also known contributors of TME-induced immunosuppression.

This body of evidence suggests that targeting the metabolism of the TME might have synergistic effects by alleviating multiple aspects of TME-induced vascular dysfunction and immune suppression. Although this review has focused on ICT, there is evidence that these concepts are applicable to ACT as well. Future studies are needed to elucidate the synergistic potential for combining metabolic inhibition with ICT.