Cancer cells produce lactic acid in large amounts with the consumption of glucose, even though there is enough oxygen, and it accumulates in cells prior transportation into the extracellular environment where it ultimately establishes and acidic TME [20]. A more aggressive type of cancer cells have been observed in acidic TME and it can also promote cancer progression by suppressing immune anti-tumor response [21]. This sums up with the fact that lactic acid levels increase lineally with the tumor burden [22] and it can be used as a nutrient source by cancer cells [23]. Lactic acid in the TME can suppress cancer immunosurveillance mechanisms and promote immune escape of cancer cells: 1. lactic acid affects NK cells and reduces interferon gamma (IFN-γ) secretion promoting apoptosis [24, 25]; 2. lactate in the TME activates the phosphorylation the transcription factor via mTOR pathway inhibiting vacuolar ATPase subunit in macrophages [26]; 3. lactate concentration in the TME has shown a negative correlation in survival of patients with cervical cancer [27].

When there is a fast oxygen consumption and an altered angiogenesis the hypoxia-inducible factor (HIF) is activated by hypoxia and it upregulates programmed death-ligand 1 (PD-L1) expression [10]. The accumulation of lactate interrupts lactate homeostasis by the influence of HIF-1, c-myc and PI3K/AKT oncogenes [28]. T-cell exhaustion occurs by the PD-1/PD-L1 pathway when the IFN-γ secretion is increased by lactate levels [10]. In addition, lactate blocks the transporter MCT-1 at the same time as it enhances Tregs by nuclear factor of activated T-cells 1 (NFAT1) translocation into nucleus and increases lactated absorption by upregulating MCT-1 and the expression of lactate dehydrogenase A which in all disrupts the anti-tumor effects of CTLs [29]. Tregs suppress CD8⁺ T-cells function of secreting perforin and cytotoxic granzymes. The tricarboxylic acid (TCA) cycle is also supported by lactate when there is an increase in IL-10 and TGF-β which enables proliferation in acidic and hypoglycemic conditions [30, 31]. Anti-tumor responses are also negatively influenced by the enhancement of M2 polarization by inducing arginase-1 (ARG-1), transcriptional repressors, HIF-1α and IL-6 via the upregulation of MCT-4 expression and glycolytic rate promoted by lactate [32]. The impact of lactate on TME tumor evasion has also been observed with HLA-DR- and CD86-high macrophages have a glycolytic phenotype that suppresses the secretion of anti-tumor IL-12 p70 via PKM2/HIF-1α axis [33, 34].

OXPHOS takes place in mitochondria to efficiently produce ATP with substrates like glucose, fatty acids (FAs) and amino acids. Cancer cells can prefer OXPHOS to generate ATP in less glucose-rich or more oxygenated tissues in order to support redox balance and biosynthesis. In more aggressive, rapidly growing tumors or hypoxic regions, tumor cells may rely on glycolysis. In well-vascularized regions with oxygen they can use OXPHOS, sometimes both mechanisms coexist and it can vary across tumor types [35].

Itaconate is synthesized from the decarboxylation of TCA-cycle derived cis-aconitate regulated by the mitochondrial enzyme immune-responsive gene 1 (IRG1) [36, 37]. Itaconate induces CD8⁺ T exhaustion, attenuate CD8⁺ T-cell proliferation and function and can suppress myeloid immune cells [36]. Itaconate could be a potential immunometabolite target in the TME [37, 38]: 1. high levels of IRG1 and itaconate increases ROS generation in macrophages via MAPK pathway; 2. inhibition of succinate dehydrogenase (SDH), glycolysis mediation, Nrf2 transcription factors and ATF3 activation associated with the modulation of exhaustion markers TIM-3 and PD-1 [37, 38].

Arachidonic acid is a class of eicosapentaenoic acid and it is a substrate for prostaglandins synthesis associated with FAs in tumor cells [12]. PGE2 is an immunosuppressive factor [40] necessary for cell growth as well as a regulation factor in the cell that mediates inflammatory response [41]. PGE2 could play an active role in tumor progression by promoting cancer cell invasion, stimulating angiogenesis and suppressing apoptosis which ultimately enhances cell proliferation [42]. The immune response could also be altered by PGE2 by autocrine and paracrine methods. PGE2 accumulation can increase M2 macrophages that induce cancer by transforming M1 macrophages, that harbor a tumor suppressing effect [43]. Anti-inflammatory effects of PGE2 have also been observed in NK cells, neutrophils and monocytes [44, 45]. Furthermore, PGE2 can promote immune scape mechanisms and tumorigenesis: 1. it can stimulate bone marrow-like cells to promote CXCL1, IL-6 and granulocyte-colony-stimulating factor (G-CSF); 2. inhibit the production of IL-12 and TNF-α in myelocid cells stimulated by lipopolysaccharide (LPS).

In the TME, PGE2 can promote an immunosuppressive state by both preventing the pro-inflammatory response in immune cells, reprogram TME and also by directly upregulating tumor cells [46].

Plasma membrane phospholipids and molecules for signaling are generated via the anabolic pathways used by tumor cells which require an increase in the synthesis of de novo FAs [12]. On the other hand, immune cells require FAs for cell membranes and lipidic cell structures [47]: 1. effector immune cells that harbor a rapid growth require FAs for cell membranes; 2. memory immune cells grow slower and doesn’t need that amount of biosynthesis so the main way is by the FA oxidation (FAO) [48]. The metabolic reprogramming associated with the lipid metabolism can lead to an immunosuppressive state by the abnormal accumulation of short-chain FA, cholesterol, long chain FA, etc., in myeloid cells that are tumor-infiltrating [49].

The impact of FA on T-cells depends on the amount of FA: low concentration of FA can impact proliferation while an increase in FA concentration can lead to apoptosis because of lipotoxicity [50] and, therefore, FA catabolism can upregulate immune response. In the TME, dysregulated FAO and the modulated carnitine palmitoyl-transferase (CPT), that transports acyl-CoA into mitochondria, in cells act directly in the metabolic remodeling [51]. The receptor-interacting protein kinase 3 (RIPK3)-ROS-Caspase-PPAR pathway promotes FAO and M2 polarization of tumor-associated macrophages (TAMs): 1. inositol-requiring protein 1α (IRE-1α)-mediated endoplasmic reticulum (ER) stress which obtains immune inefficiency in DC and M2 polarization by inducing lipid peroxidation; 2. Tregs have higher FAS and FAO while TILs evade oxidative stress and show T-cell disability [52, 53]; 3. RIPK3 exacerbates M2 polarization and is downregulated in TAMs; 4. the shortage of RIPK3 is associated with decreased ROS and the inhibition of caspase-1mediated PPAR to ultimately promote FA metabolism [54].

In metastatic cells, FAO provides acetyl-CoA and are a substrate for OXPHOS, important to contribute to the high energy demand [55]. In organs where metabolic conditions favor the use of lipids, like liver and bone, the reliance of FAO is higher [56]. FAs synthesis plays an important role in cancer to induce energy production, membrane biosynthesis and are also involved in signaling molecules. Tumors such as prostate, ovarian, breast, colorectal and lung cancer are dependent on FA biosynthesis.

Cholesterol is a part of the cell membrane and tumor cells with rapid proliferation mechanisms require more synthesis. Tumor cells that have a high expression of cholesterol can evade immune surveillance mechanisms and treatment strategies [57]. High levels of cholesterol in immune cells have been associated with an increase in the expression of immune detection points as LAG-3, TIM-3 and PD-1, however, researchers have observed that there is more concentration of cholesterol in tumor cells compared to the levels in immune cells [58]. In this context, high concentrations of cholesterol can downregulate anti-tumor effect of T cells [59].

The autoxidation of cholesterols results in metabolites called oxysterols, always associated with high excess of cholesterols. Oxysterols have functions that can alter the immune response and landscape. The 22-hydroxycholesterol (22-OHC) can engage CD11b^high and Gr1^high neutrophils in cancer cells and act as an immunosuppressor in the TME [60]. The 27-hydroxycholesterol (27-OHC) has been described in high concentrations in tumors of patients with estrogen receptor-positive breast cancer [61]. 27-OHC has been associated with CD8 T cells exhaustion that can promote tumor proliferation and metastasis [62]. Previous studies point out that lowering cholesterol concentration in the TME can alleviate T cell immune suppression [12].

Cancer cells have a rapid and non-controlled proliferation that requires an active nucleic acid metabolism in the TME [63]. The dysregulated nucleic acid metabolism collaborates with cell division promotion, genetic integrity and adaptation to hostile conditions. Cancer cells have a DNA and RNA altered metabolism that affects the ability of tumor cells to grow, metastasize and, ultimately, survive. Due to the increased demand for nucleids, cancer cells overexpress purine and pyrimidine biosynthesis pathways to increase nucleotide synthesis. Therefore, enzymes like thymidylate synthase, inosine monophosphatase dehydrogenase and AMP deaminase are upregulated to supply nucleotide synthesis.

In DNA replication, cancer cells upregulate and enhance the availability of nucleotides for synthesis. In DNA repair, pathways such as homologous recombination, nucleotide excision repair and base excision repair are increased although a dysregulated repair can contribute to resistance to therapy and genomic instability [64]. RNA metabolism often requires alterations in RNA polymerase activity to support the uncontrolled cell division and adapt to hypoxia and nutrient depravation. In this context, RNA-binding proteins or microRNA profiles can be altered to regulate gene expression [65]. Furthermore, alternative splicing can change proteins to evade immune surveillance, promote tumorigenesis and contribute to metastasis.

Glutamine is a key component of amino acid production, glycosylation, autophaghy, epigenetics, the synthesis of nucleotides and ECM production [66], therefore, it is important for the proliferation of both cancer cells and normal cells [67]. Macrophages, lymphocytes and neutrophils require glutamine for the development of cell fate and immune responses [68, 69]. The transformation of CD4⁺ T cells to inflammatory subtypes is upregulated by glutamine metabolism as well as the activation of immune cells [69, 70]. Glutamine is also required for macrophages to produce and secrete pro-inflammatory cytokines (IL-1, IL-6 and TNF) [12]. If there is a nutritional deficiency, cancer cells are able to destroy large molecules in order to obtain glutamine as it has been observed in the excessive activation of the oncogene RAS where tumor cells degrade extracellular proteins into amino acids which include glutamine in order to supply and provide more nutrients to maintain cancer cell proliferation [71]. Glutamine also plays an important role in fueling OXPHOS in metastatic tumor when glutamine is converted to glutamate and further alfa-ketoglutarate that enters the TCA cycle and sustains OXPHOS [72].

The argininosuccinate synthethase 1 (ASS1) is an enzyme envolved in the production of arginine and is lacking in the majority of cancer cells which lowers the capacity of synthetizing arginine and, therefore, tumor cells will use exogenous arginine [73]. As previously mentioned for glutamine, arginine also plays an important role in immune responses and T cell activation [12]: 1. arginine stimulates T cell and NK cell cytotoxicity as well as effector cytokine production; 2. stimulation of effector cytokine production in vitro and in treatment with anti-PD-L1, mice with osteosarcoma showed longer survival rates [74]. In previous studies, immunomodulatory cells that express ARG-1 (M2-TAMs, Treg cells and tolerogenic DCs) can increase the degradation of arginine and limit the source for T cells [75].

Tryptophan is an essential amino acid required for protein synthesis and metabolism and it can degradate by inoleamine-2, 3-dioxygenase (IDO1) and tryptophan-2, 3-dioxygenase (TDO2) to tryptophan to kynurenic acid [76]. Previous studies have determined that in gastric adenocarcinoma patients who harbor high levels of IDO1 and TDO2, these enzymes can promote tumor progression and are associated with a worse prognosis [77]. The immune activity and function of T cells can be suppressed if there is low availability of tryptophan in the TME and it has been suggested that high levels of IDO1 and TDO2 in the TME are associated with lower values of tryptophan [12]. As it occurs for arginine, when tumor cells use tryptophan, it consequently associates a lack of tryptophan for T cells and it can trigger apoptosis in these immune cells [78, 79]. In the presence of an IDO1 inhibitor, tryptophan levels can be increased and the accumulation can induce tumor regression by upregulating the production of IL-12 and IFN-γ cytokines and T cells and neutrophils in mouse metastastic liver and blade tumor models [12].

CAFs are key regulators of the TME through the secretion of growth factors, cytokines, and ECM proteins that facilitate tumor proliferation, invasion, and angiogenesis [83]. In addition to providing structural support for tumor expansion, CAFs actively remodel the ECM, creating physical barriers that limit immune cell infiltration [84].

Hypoxia, a hallmark of solid tumors resulting from aberrant vasculature, exacerbates the complexity of the TME. Under hypoxic conditions, the stabilization of HIFs promotes angiogenesis, metabolic reprogramming, and epithelial-to-mesenchymal transition (EMT), thereby enhancing metastatic potential [85–87]. Moreover, the acidic and nutrient-depleted nature of the TME, driven by altered cancer cell metabolism, influences both stromal and immune cell function. The Warburg effect, characterized by increased aerobic glycolysis and lactate production, contributes to an immunosuppressive and acidic microenvironment that further supports tumor progression and immune evasion [88, 89].

TAMs are abundantly present in the TME and play a central role in tumor progression [90]. Chemokines such as colony-stimulating factor 1 (CSF1) and C-C motif chemokine ligand 2 (CCL2) recruit monocytes from peripheral circulation, where they differentiate into macrophages [91]. These macrophages can adopt either an M1 phenotype, which exhibits tumoricidal activity, or an M2 phenotype, which promotes tumor growth, metastasis, and immune suppression through inhibition of CD8⁺ T cells [91].

Glucose metabolism significantly influences TAM polarization toward a tumor-promoting state. While tumor cells are known to produce lactate via glycolysis, recent findings suggest that TAMs exhibit an enhanced capacity for glucose consumption and lactate production, thereby contributing to the acidic TME [92]. Lactate has been shown to drive TAM polarization toward the M2 phenotype, leading to increased expression of vascular endothelial growth factor (VEGF) and ARG-1. VEGF facilitates angiogenesis, whereas ARG-1 catalyzes polyamine synthesis, supporting cancer cell proliferation [93]. Furthermore, lactate suppresses ATP6V0d2 expression in macrophages, preventing lysosomal degradation of HIF-2α, thereby further promoting tumor progression [94, 95]. Additionally, lactate enhances PD-L1 expression through NF-κB, reinforcing TAM-mediated immunosuppression [96].

The tumor-supportive function of TAMs is further driven by amino acid and lipid metabolism. Methionine adenosyltransferase II α (MAT2A) elevates S-adenosylmethionine (SAM) levels in macrophages, facilitating TAM polarization via histone methylation [97]. Additionally, increased expression of lipid metabolism genes in TAMs highlights their metabolic adaptation within the tumor niche [98]. Notably, downregulation of monoacylglycerol lipase (MGLL) in TAMs supports tumor growth by altering the catabolism of 2-arachidonoylglycerol (2-AG), which activates cannabinoid receptor-2 (CB-2) and promotes M2 macrophage polarization [99].

The anatomical location of the tumor also influences its metabolic characteristics. Tumors exhibit organ-specific metabolic profiles shaped by local nutrient and oxygen availability, which significantly impact the behavior of tumor and immune cells [100, 101]. For example, liver tumors have distinct metabolic adaptations compared to brain tumors, highlighting the role of the local microenvironment in cancer progression.

In colorectal cancer, metabolic reprogramming leads to altered nutrient availability and a greater reliance on glycolysis, driving lactate accumulation and reinforcing immune suppression [102, 103]. These metabolic adaptations within the TME underscore their role in both cancer progression and resistance to immunotherapy. Consequently, targeting the metabolic flexibility of tumor cells and their interactions within the TME offers a promising avenue for improving therapeutic outcomes.

The impact of these metabolic alterations on the regulation of oncogenes, tumor suppressor genes, epigenetics, and microenvironmental factors is summarized in Figure 1.

Despite the promising potential of targeting tumor metabolism to overcome immunotherapy resistance, several challenges remain to be addressed. Tumor metabolic heterogeneity poses a significant barrier, as different tumors and even different regions within the same tumor may exhibit distinct metabolic profiles [47]. This heterogeneity complicates the development of universal metabolic-targeting therapies and necessitates personalized approaches based on tumor-specific metabolic signatures.

Many drugs that work in preclinical models have been evaluated, but unfortunately, they do not work for patients. Below, we summarize some of them and discuss the reasons why they may not have worked to date. Table 2 summarizes them.

Devimistat, an investigational anti-mitochondrial agent, targets cancer cell mitochondria by disrupting energy production, specifically the TCA cycle. Its efficacy was evaluated in the AVENGER 500 trial (ClinicalTrials.gov identifier: NCT03504423), a global, randomized phase III study [132]. The trial assessed the efficacy and safety of devimistat combined with modified FOLFIRINOX versus standard-dose FOLFIRINOX in treatment-naive patients with metastatic pancreatic cancer (mPC). The primary endpoints were overall survival (OS) and progression-free survival (PFS), with no significant differences observed between the control and experimental arms. No new toxicity signals were observed. These findings indicate that devimistat, when combined with chemotherapy, did not improve long- or short-term outcomes in mPC patients compared to standard chemotherapy. The lack of efficacy of devimistat may be attributed to several factors. Pancreatic cancer (PC) exhibits prognostically distinct metabolic subgroups based on glycolytic and cholesterogenic gene expression. Tumors with poor prognosis often show low expression of mitochondrial pyruvate carriers MPC1 and MPC2, potentially reducing the effectiveness of metabolism-targeting agents such as devimistat. However, due to resource limitations, the AVENGER 500 protocol did not mandate baseline tumor tissue collection for gene expression analysis, leaving potential imbalances in metabolic or basal subtypes across treatment arms undetermined. Additionally, the dosing and administration schedule used in the study may have contributed to the lack of observed efficacy.

Telaglenastat is an investigational, first-in-class, selective, oral GLS inhibitor that blocks glutamine utilization and downstream pathways. Preclinically, telaglenastat synergized with cabozantinib, a VEGFR2/MET/AXL inhibitor, in renal clear-cell cancer (RCC) models. CANTATA (ClinicalTrials.gov Identifier: NCT03428217) was a randomized, placebo-controlled, double-blind, pivotal trial in patients had metastatic RCC following progression on 1 to 2 prior lines of therapy [133]. Compare the efficacy and safety of telaglenastat plus cabozantinib versus placebo plus cabozantinib. The primary endpoint was PFS assessed by blinded independent radiology review. No differences were observed in terms of PFS, objective response rate (ORR), or grade 3/4 adverse events between the experimental arm and the control arm. Possible reasons for the lack of efficacy of telaglenastat in the CANTATA trial include insufficient drug exposure, although pharmacokinetic analyses suggest otherwise. Cabozantinib may not be the optimal partner for dual metabolic inhibition of glucose and glutamine. Additionally, telaglenastat may not sufficiently inhibit GLS in vivo, or GLS may not be an ideal target in RCC at the tested dose. Identifying biomarkers of response could clarify these mechanisms.

Epacadostat, an IDO1 selective inhibitor, was evaluated in a randomized, placebo-controlled, double-blind, phase 3 trial (ClinicalTrials.gov Identifier: NCT02752074) in patients with unresectable stage III or IV melanoma previously untreated with PD-1 or PD-L1 inhibitors [134]. Primary endpoints were PFS and OS. Epacadostat plus pembrolizumab did not improve PFS or OS compared to placebo plus pembrolizumab. No relevant differences were observed in grade 3/4 adverse events. The absence of benefit of adding epacadostat to pembrolizumab was surprising, given early-phase clinical data. In two open-label, phase 1–2 studies of patients with advanced melanoma, the combinations of epacadostat plus pembrolizumab (ECHO-202) [135] and epacadostat plus nivolumab (ECHO-204) [136] showed promising antitumour activity (56–65% of patients achieving an ORR) compared with historical data. Epacadostat was chosen based on preclinical and clinical data showing optimal activity and tolerability. However, higher doses might be needed for better target coverage. The role of IDO1 in melanoma is controversial due to variability in IDO expression. The high response rates and favorable PFS in early trials remain unexplained, despite similar patient and tumor characteristics to other studies.

Pegvorhyaluronidase alfa (PEGPH20) is a novel PEGylated recombinant human hyaluronidase developed as an anticancer therapy for use in combination with other systemic therapies to facilitate their delivery to the TME. Results from early clinical trials with PEGPH20 in patients with advanced solid tumors, including PC, were consistent with preclinical findings and supported additional clinical development [137, 138]. Unfortunately, the phase III HALO 109-301 clinical trial [139], which evaluates the efficacy and safety of PEGPH20 plus nab-paclitaxel/gemcitabine in patients with hyaluronan-high mPC, did not improve OS or PFS. However, it did show an increase in ORR (47% versus 36%). The safety profile of PEGPH20 was consistent with that found in previous studies. The study was well-conducted, with comparable progression rates and lower discontinuation rates than previous studies. However, the HALO-301 trial suggests the need to re-evaluate stroma remodeling strategies. PEGPH20 with modified FOLFIRINOX showed inferior survival, possibly due to dose reductions and unforeseen drug interactions. The complexity of tumor stroma and its role in PDA progression requires further investigation. Additionally, current methods for assessing hyaluronan status may be unreliable, and more data on metastatic sites are needed.

IACS-010759, a potent and selective OXPHOS inhibitor, was studied in phase I trials for relapsed/refractory acute myeloid leukemia (NCT02882321) and advanced solid tumors (NCT03291938) [140]. The drug had a narrow therapeutic index with dose-limiting toxicities, including elevated blood lactate and neurotoxicity, hindering target exposure. Modest target inhibition and limited antitumor activity were observed at tolerated doses, leading to trial discontinuation. Mouse studies showed IACS-010759 induced peripheral neuropathy, mitigated by coadministration of a HDAC6 inhibitor [141]. Further studies are needed to understand the link between OXPHOS inhibition and neurotoxicity, and caution is advised in developing complex I inhibitors as antitumor agents.

A phase Ib study in Japanese patients with advanced solid tumors investigated retifanlimab (anti-PD-1), INCB001158 (an oral ARG inhibitor), and their combination [142]. No fatal adverse events or discontinuations occurred. Retifanlimab showed a 33.3% overall response rate and 50% disease control rate. The combination treatment did not show additional benefit. Further research is needed to identify biomarkers predicting response to INCB001158 or the combination therapy.

Advances in metabolomics and transcriptomics are enabling the identification of tumor-specific metabolic vulnerabilities, paving the way for personalized metabolic interventions [105]. The integration of these technologies with existing immunotherapy strategies could improve patient stratification and therapeutic outcomes [106]. Among the possible strategies to overcome metabolic resistance to immunotherapy are:

1.

Combining metabolic inhibitors with standard-of-care treatments, such as chemotherapy, radiotherapy, and ICIs, is a promising strategy to enhance therapeutic efficacy. However, optimizing dosing regimens and managing potential toxicities associated with metabolic interventions remain key challenges.