A defining feature of the AML TME is the production of excessive LA, creating a highly acidic environment (18). A study reported that LA inhibits T cell growth and proliferation, and its accumulation in the TME disrupts LA efflux from T cells, impairing their metabolism, function and antitumor immunity (28). Additionally, LA downregulates perforin and granzyme B, which are essential for T cell-mediated tumor cell killing and proliferation (29,30). Dichloroacetic acid (DCA) targets pyruvate dehydrogenase kinase (PDK), an enzyme that drives glucose metabolism toward glycolysis instead of oxidative phosphorylation (OXPHOS) in tumor cells. By inhibiting PDK, DCA offers a potential strategy to modulate glucose metabolism in the AML TME (31). DCA can inhibit the shift toward glycolysis, reducing LA levels and notably enhancing T cell proliferation, and cytokine production and function whilst decreasing apoptosis (32). Tregs use monocarboxylate transporter 1 to uptake LA, which activates signaling pathways that promote nuclear factor of activated T cells 1 translocation to the nucleus, upregulating PD-1 expression (Fig. 1) (33). Therefore, high LA levels increase PD-1 expression in Tregs, amplifying their immunosuppressive effects. Additionally, LA accumulation and the resulting acidic pH in the TME inhibit CD4⁺ T cell activity, particularly the secretion of key cytokines such as IL-2 and IFN-γ. These cytokines are crucial for immune responses, and their reduction weakens CD4⁺ T cell proliferation and helper function, further promoting immunosuppression (34). In summary, high LA levels in the AML TME are a major contributor to T cell dysfunction, impairing antitumor immunity through metabolic disruption, cytotoxic inhibition and induction of exhaustion.

Hypoxia is a defining feature of the bone marrow TME, markedly impacting AML cell growth, metabolic reprogramming and immune interactions. Studies have reported that patients with higher hypoxia risk scores tend to have shorter overall survival rates, linking hypoxia to poor prognosis in AML. Elevated hypoxia risk scores are also strongly associated with disease progression and the immunosuppressive TME (35,36). Under hypoxic conditions, hypoxia-inducible factor 1α (HIF-1α) is activated and directly upregulates programmed death-ligand 1 (PD-L1) expression. High PD-L1 levels bind to PD-1 on T cells, inhibiting their activation and signaling, further suppressing antitumor immunity (37).

AML tumor cells and immune cells compete for essential nutrients such as glucose and amino acids, which are vital for both rapid tumor growth and normal immune function. Tumor cells, with their high metabolic demands, prioritize the uptake of glucose, glutamine and other substrates for energy and biosynthesis, depriving immune cells of these resources. For instance, AML cells depend on glutamine for OXPHOS, and the inhibition of glutaminase-1 has been reported to suppress AML development in mouse models (38). Glutamine is also critical for T cell activation, proliferation and cytokine production (39). Glutamine deficiency in the TME promotes the generation of Tregs, further exacerbating immunosuppression (40). Glutamine deficiency not only restricts effector T cell proliferation and function, but also supports the metabolic adaptation of Tregs, enhancing their immunosuppressive activity. Studies have highlighted the antisense non-coding RNA at the INK4 locus (ANRIL) as a key regulator of glucose metabolism in AML, where it is notably upregulated. ANRIL modulates glucose metabolism via the AMP-activated protein kinase (AMPK)/ sirtuin 1 pathway, promoting AML cell survival. Its knockdown reduces glucose uptake and inhibits AML cell maintenance (41,42). Under oxygen-sufficient conditions, AML cells rely heavily on glucose as their primary metabolic substrate, rapidly converting it to LA through glycolysis to meet energy demands (43). A study by Cunningham and Kohno (44) using 18FDG labeling in 124 patients reported consistently high glucose uptake in AML bone marrow, highlighting the glucose dependence of AML cells. This excessive glucose consumption by AML cells suppresses T cell activation, induces exhaustion and drives leukemia progression.

The AML TME shapes T cell metabolism and function by secreting chemokines and cytokines, influencing tumor progression (45,46). Studies have reported that chemokines such as chemokine CCL3 and CXCL12 in the AML microenvironment promote Treg accumulation, which competitively inhibits Teffs and indirectly affects their metabolic activity (47,48). Whilst research does not directly address whether chemokines regulate T cell metabolic pathways, it highlights the potential of blocking Treg migration to delay disease progression. Inhibiting these chemokines has been reported to slow AML progression in mouse models (45).

Potassium ions (K⁺), abundant in intracellular fluid, are essential electrolytes that regulate immune cell function and several cellular processes (49,50). Tumor cell necrosis releases large amounts of K⁺ into the extracellular fluid of the TME. Elevated extracellular K⁺ concentrations impair T cell receptor (TCR)-mediated Akt-mTOR phosphorylation, hindering effector T cell activation and function (51). Conversely, higher K⁺ levels can enhance T cell stemness, maintaining their undifferentiated state (52). This indicates that K⁺ dynamics in the TME influence both immediate T cell effector functions and their long-term survival and potential. However, research on the role of K⁺ in the AML TME remains limited.

Before antigen exposure, naive T cells remain in a quiescent state maintained by IL-7. As they do not require clonal expansion or high cytokine production, their reliance on anabolic pathways for DNA, protein and molecule synthesis is minimal. Instead, they generate ATP primarily through mitochondrial OXPHOS (53). By contrast, tumor cells undergo a notable metabolic shift, favoring glycolysis over OXPHOS despite its lower ATP yield. This adaptation supports their rapid proliferation and survival by quickly meeting energy and metabolic intermediate demands (54). Tregs derived from CD4⁺ T cells serve a dual role: They maintain immune homeostasis and prevent autoimmune diseases; however, during AML progression, they suppress CTL activity, weakening antitumor immunity and promoting immune evasion (55). Studies have reported that AML blasts promote T cell differentiation into a Treg phenotype by expressing inducible T cell co-stimulator ligand, markedly expanding the Treg population (56,57). At the AML disease site, Tregs suppress CTL activity, limiting their proliferation and hindering the expansion of adoptively transferred CTLs in vivo. This suppression weakens CTL-mediated antitumor effects, further compromising immune responses against AML (58). Research indicates that mTOR and glucose transporter-1 (GLUT-1) regulate CD4⁺ T cell activation by influencing glycolysis. Increased glycolysis in CD4⁺ T cells enhances their activation, proliferation and survival whilst promoting effector T cell differentiation and inhibiting Treg development, which suppresses immune responses (59). Similarly, the differentiation of CD8⁺ T cells from naive to effector states requires upregulated glucose metabolism as glycolysis provides the energy needed for their immune effector functions (60). In mouse CD8⁺ T cells, branched-chain amino acid (BCAA) accumulation boosts glucose transporter 1 (GLUT1) levels via a forkhead box protein O1-dependent mechanism, enhancing glycolysis and OXPHOS to strengthen antitumor immunity. BCAA supplementation also improves the efficacy of PD-1 immunotherapy in tumors (61). Thus, increased glycolysis serves a crucial role in enhancing antitumor immune responses.

In AML, leukemia cells rely heavily on glycolysis (Warburg effect) for energy, resulting in the accumulation of glycolytic byproducts such as methylglyoxal. These reactive compounds react non-enzymatically with proteins, lipids and DNA, forming advanced glycosylation end products (AGEs) (62). AGEs bind to the receptor for AGEs (RAGE), activating signaling pathways such as NF-κB, MAPK and phosphoinositide 3-kinase (PI3K)/Akt, which drive pro-inflammatory and pro-survival responses, influencing cellular functions and disease progression (63,64). AGE-RAGE signaling promotes the proliferation of AML cell lines, such as HL60 and HEL, by inhibiting apoptosis and autophagy, enhancing cancer cell survival and invasiveness. This mechanism contributes to the progression of several tumors, including AML (64). Consequently, the AGE-RAGE axis represents both a hallmark of AML metabolic reprogramming and a potential diagnostic and therapeutic target.

Fatty acid uptake and metabolism are essential for AML cell proliferation, providing energy and metabolic intermediates whilst inhibiting apoptosis and conferring resistance to cytotoxic drugs. AML cells thus depend on fatty acid metabolism for survival (65). Lipid metabolism also serves a key role in T cell metabolic reprogramming, supporting membrane expansion through the production of phospholipids and cholesterol. Naive CD8⁺ T cells in the lymphatic system primarily rely on FAO for energy (66). The differentiation of Teffs compared with memory T cells depends on the strength of signals from co-stimulatory molecules, cytokines and antigen presentation (67). Strong stimulation leads to the generation of short-lived terminally differentiated effector cells, whilst weaker stimulation promotes the differentiation of memory precursor cells, which further transform into long-lived memory cells, providing protection against re-infection (68). Certain memory T cells also arise from Teffs that survive apoptosis at the end of an immune response (69). In the AML TME, memory T cells exhibit markedly reduced metabolic adaptability and persistence. Research indicates that these cells fail to sustain critical metabolic pathways such as FAO and mitochondrial OXPHOS resulting in energy deficiency that compromises their survival and long-term effector function (6,70). AML cells exacerbate this dysfunction by competing for essential nutrients such as glucose and glutamine, further suppressing T cell metabolism. Prolonged nutrient deprivation not only impairs memory T cell function, but also promotes their exhaustion, characterized by upregulated expression of exhaustion markers such as PD-1 and thymocyte selection-associated high mobility group box protein, and diminished cytotoxic capacity (56,71). Notably, memory T cells also serve a role in modulating the response of hematological malignancies to PD-1 blockade therapy (72). Studies have reported that TNF receptor-associated factor 6 (TRAF6) influences CD8⁺ memory T cells through lipid metabolism regulation. Mice with T cell-specific TRAF6 deficiency exhibit strong effector T cell responses but fail to form memory T cells effectively (73,74). IL-15 regulates mitochondrial spare respiratory capacity and oxidative metabolism by modulating mitochondrial biogenesis and the key enzyme carnitine O-palmitoyl transferase 1 (CPT1)a. CPT1a is critical for the rate-limiting step in mitochondrial FAO (75). Enhancing FAO through AMPK activation or mTOR inhibition can notably increase memory T cell numbers (73,76). These findings emphasize the pivotal role of lipid metabolism in T cell metabolic reprogramming.

Studies have reported that amino acids regulate immune responses by influencing T cell activation, cytokine production and other immune functions (77). Glutamine, the most abundant amino acid in serum, is vital for maintaining metabolic balance and cell function. Its absence in culture medium markedly impairs naive T cell activation, proliferation and cytokine production (39). In the hypoxic TME, glutamine acts as a primary carbon source, supporting the energy and metabolic needs of tumor cells (78). Amino acids and their metabolites are essential in regulating both tumor and immune cell proliferation within the TME. For instance, glutamine and its metabolic pathways are crucial for tumor cell glycolysis. Using glutamine antagonists can effectively inhibit glycolysis in tumor cells, suppressing their growth and enhancing the antitumor immune response by altering the immunosuppressive TME, thereby overcoming immune evasion (79). Besides glutamine, the metabolism of arginine and tryptophan also serves a pivotal role in immune regulation within the TME. Studies have reported that monocytes, under macrophage-stimulating factor influence, rapidly degrade tryptophan through increased IDO activity, thereby suppressing T cell proliferation. This mechanism aids tumor cells and macrophages in immune evasion. Furthermore, extracellular arginine availability in the TME directly influences T cell function and antitumor responses (80,81). Arginine deprivation leads T cells to initiate autophagy, downregulate the CD3ζ chain and ultimately undergo apoptosis. In AML, this effect is more pronounced, as AML blasts express and secrete arginase II, a key enzyme for arginine metabolism, whilst arginase I is typically low and only detectable under specific conditions. This metabolic regulation further impairs T cell function, aiding AML cells in evading immune surveillance (70). These findings highlight the crucial association between amino acid metabolism in the TME and T cell-mediated antitumor immunity.

Nucleotides, as essential components of genetic material, are critical for highly proliferating cells, particularly purine and pyrimidine nucleotides. Consequently, nucleotide metabolism presents a potential target for cancer therapy (82,83). Drugs such as methotrexate, which target nucleotide metabolism, have been reported to be effective in treating acute lymphoblastic leukemia (ALL). However, non-specific targeting of nucleotide metabolism can inhibit normal cell processes, leading to severe side effects (84,85). In the AML TME, high concentrations of adenosine act as an immunosuppressive metabolite. Elevated adenosine levels suppress T cell activity by inhibiting activation, proliferation and cytokine secretion through adenosine receptor binding such as A2A receptors (56). The A2A adenosine receptor signaling pathway markedly inhibits T lymphocyte proliferation, activation and cytokine production. Additionally, this pathway activates immunosuppressive cells, such as Tregs and MDSCs, further impairing effector immune cell function (86). Activation of A2A receptors not only suppresses effector T cell activity, but also enhances Treg cell immunoregulatory function by upregulating key molecules and metabolic pathways, thus promoting immune suppression (Fig. 2) (87). Furthermore, studies have reported that LA treatment reduces nucleotide abundance in T cells, impairing proliferation and cell cycle activity. LA also disrupts several metabolic pathways, including amino acid biosynthesis and pyrimidine metabolism. NaBi itself can serve as a substrate for multiple carboxylase reactions such as pyrimidine metabolism (88,89). The application of NaBi can reverse these changes, pyrimidine metabolism increased in T cells rescued with NaBi (19).

mTOR, a serine/threonine kinase, serves a pivotal role in several cellular processes, including metabolism regulation. It is a key regulator of cell metabolism and serves as the catalytic subunit of both mTORC1 and mTORC2 complexes. mTORC1 supports effector T cell function by promoting glycolysis and protein synthesis, whilst mTORC2 regulates T cell differentiation and survival through the cytoskeleton and lipid metabolism (90). mTOR enhances glucose uptake and glycogen synthesis by modulating the insulin receptor substrate 1 (IRS1)/PI3K/Akt pathway, thereby boosting glycolysis. Inhibition of mTOR activation or its downregulation in CD4⁺ T cells reduce glycolysis, impairing their activation (31). mTOR activation also increases GLUT1 expression, promoting T cell proliferation and cytokine production (Fig. 3) (91). Additionally, mTOR is a crucial regulator of memory CD8⁺ T cell differentiation, with the mTOR-specific inhibitor rapamycin, an immunosuppressive drug, demonstrating an immunostimulatory effect on memory CD8⁺ T cells (87).

In the AML TME, AMPK, as a key energy sensor, serves a crucial role in regulating the metabolic state of immune cells, enabling them to effectively maintain their activity and function. It also serves an important role in T cell metabolic reprogramming. A previous study reported that AMPK signaling promotes lipid metabolism to generate functional memory CD8⁺ T cells (92). Furthermore, as an upstream inhibitor of mTOR activity, AMPK can inhibit mTOR through the AMPK activator metformin, which helps reduce glycolysis in T cells (Fig. 4). This, in turn, promotes the generation of Tregs by suppressing Th1 and Th17 cells (93). Metformin also activates AMPK and inhibits the proliferation of AML cell lines and primary AML cells (94). However, future research needs to further explore the specific molecular mechanisms of AMPK activation and its potential for clinical translation.

The PPAR family includes PPARα, PPARδ and PPARγ. The nuclear receptor PPARγ serves an essential role in adipogenesis, immune responses and the metabolism of lipids and carbohydrates. Fatty acids can also act as ligands for PPARγ (95,96). Study have reported that in chronic lymphocytic leukemia (CLL), high doses of glucocorticoids induce the activation of PPARα and downstream FAO, leading to drug resistance (97). Moreover, it has been reported that activation of PPAR promotes the proliferation of CD8⁺ T cells, increasing the number of functional Teffs. The activation of the PPAR pathway can also rescue PD-1 blockade-induced T cell apoptosis by upregulating anti-apoptotic proteins such as Bcl2, baculoviral IAP repeat containing 3 and apoptosis inhibitor 5. Additionally, PPAR activation can reprogram CTL energy metabolism and overcome the reduction in the number of functional Teffs associated with PD-1 blockade by reducing apoptosis or increasing proliferation (98). Therefore, targeting the PPAR signaling pathway, such as by using PPAR agonists, may serve as a potential therapeutic target for AML.

HIF-1α is a central transcription factor in hypoxic cells and a hallmark of TME. It is also a downstream target of GLUT-1 (99). It facilitates Treg migration by promoting glycolysis and FAO. Elevated glucose uptake by cancer cells stabilizes HIF-1α, thereby suppressing antitumor immunity (100,101). Moreover, HIF-1α-driven transcription enhances glycolysis in T cells, supporting Th17 differentiation whilst inhibiting Tregs (91). Mitochondrial dysfunction and HIF-1α-mediated metabolic reprogramming contribute to T cell exhaustion, a process reversible through glycolysis inhibition (102). Treatment with digoxin or acriflavine, both inhibitors of HIF-1 expression and function, in subcutaneous tumor mice has been reported to limit tumor growth (103). Therefore, targeting HIF-1 in the TME may be an effective therapeutic strategy for AML.

ICs are molecular mechanisms that regulate immune system activity, comprising co-stimulatory receptors such as CD40 and CD80 and inhibitory receptors such as cytotoxic T-lymphocyte-associated protein 4 and PD-1. These checkpoints maintain immune tolerance, protect normal tissues from excessive immune responses, or, in certain cases, enable cancer cells to evade immune surveillance (34). The advent of IC inhibitors has improved the prognosis for numerous solid tumors and certain lymphomas by blocking inhibitory signals such as the PD-1/PD-L1 pathway, thereby enhancing antitumor immunity (104). However, their efficacy in AML remains limited, particularly with PD-1/PD-L1 inhibitors (105), for reasons that are not yet fully understood. In the AML TME, competition for nutrients between AML cells and T cells restricts T cell access to glucose and glutamine, impairing their metabolic function and antitumor response. Consequently, IC inhibitors fail to enhance T cell-mediated antitumor effects (22). Furthermore, the PD-1/PD-L1 interaction serves a critical immunosuppressive role in the TME, promoting regulatory T cell function and inhibiting the activation and proliferation of Teffs, thereby further dampening the antitumor immune response (106).

ACT enhances immune responses against tumors or infections by modifying or expanding autologous or donor-derived T cells ex vivo. This includes engineering T cells with chimeric antigen receptors (CARs) or TCRs to recognize specific tumor antigens. Following expansion, these Teffs are re-infused to mediate targeted immune responses (107,108). However, unlike in ALL, CAR-T cell therapy shows limited efficacy in AML, largely due to the immunosuppressive TME (29). Lymphodepleting chemotherapy prior to CAR-T cell infusion can enhance therapeutic outcomes by reducing Tregs in the TME and alleviating their suppressive effects, thereby improving CAR-T cell proliferation and persistence in vivo (109). In AML, expanded Tregs secrete immunosuppressive cytokines such as IL-10 and TGF-β, which impair CAR-T cell function (57). Additionally, AML cells secrete arginase II, disrupting T cell metabolism and promoting immune evasion (41). Inhibition of arginine metabolism has been reported to enhance the efficacy of CD33-CAR T cells in preclinical AML models (110). Similarly, blocking the adenosine A2A receptor (A2AR), a downstream mediator of adenosine signaling, improves CAR-T cell efficacy in solid tumors (111). Adenosine suppresses T cell proliferation, activation and effector function via A2AR, whilst promoting Treg expansion, thereby dampening antitumor immunity (112). However, whether this mechanism extends to hematologic malignancies such as AML and CLL remains unclear.

In patients with AML, elevated intracellular and plasma arginase activity markedly inhibits T cell proliferation, contributing to immune dysfunction. This effect is largely mediated by increased expression and activity of arginase II in AML cells, identifying it as a potential biomarker for immune status and disease progression. Inhibition of arginase II has been reported to restore T cell proliferation and enhance antitumor immunity (110,113). When PD-1 binds to its ligand PD-L1, activated T cells are unable to continue glycolysis and normal amino acid metabolism, which results in insufficient energy production to support their effector functions. In addition to inhibiting glycolysis and amino acid metabolism, PD-1 may also impair T cell oxidative detoxification capacity, reducing their ability to cope with oxidative stress (114). Thus, elucidating the PD-1 signaling axis is crucial for understanding T cell dysfunction and identifying novel therapeutic targets.

Although the role of T cell metabolism in the AML TME has been preliminarily elucidated, further in-depth exploration is needed in the following key areas.

In future research, individualized treatments targeting T cell metabolic reprogramming in the TME of patients with AML should focus on the impact of metabolic heterogeneity on treatment responses. Differences in metabolic characteristics between patients with AML may profoundly influence the metabolic state of T cells and their antitumor capabilities. For example, variations in glycolysis, FAO or amino acid metabolism across different patients could lead to notable differences in therapeutic efficacy. Utilizing metabolomics and single-cell analysis techniques could uncover individual differences in T cell metabolic reprogramming and provide a basis for precise metabolic interventions. However, this approach faces several challenges, such as the ways to integrate multidimensional data to accurately identify key metabolic nodes, implement personalized metabolic regulation of targets in clinical applications and minimize potential side effects of metabolic interventions on systemic metabolic homeostasis. In the future, combining advanced technological methods and large-scale clinical studies will be necessary to explore the feasibility of personalized metabolic interventions, with the goal of achieving precision treatment for patients with AML.

In future research on T cell metabolic reprogramming within the AML TME, emerging technologies will provide crucial support for uncovering metabolic regulatory mechanisms and therapeutic potential. The application of single-cell metabolomics and spatial metabolomics can capture the dynamic changes and spatial heterogeneity of T cell metabolism in the AML microenvironment with high resolution, offering a new perspective on the role of metabolic reprogramming in tumor immune evasion. Moreover, metabolic flux analysis can track the dynamic changes in key metabolic pathways within T cells in real-time, revealing the flow and regulation patterns of metabolites under different conditions. By contrast, CRISPR screening technology can precisely identify key genes and metabolic nodes involved in T cell metabolic reprogramming, providing specific targets for developing intervention strategies. The combination of these technologies will not only deepen the understanding of T cell metabolic regulation mechanisms but also advance the design of personalized metabolic treatment plans. However, the application of the aforementioned technologies in AML still faces challenges, such as integrating multidimensional data, high costs and unclear clinical translation pathways. Future multi-disciplinary collaborations will be required to further optimize their application.

There are still notable obstacles and challenges in translating the research on T cell metabolic reprogramming within the AML TME into clinical application. A key obstacle from basic research to clinical use is the way to simplify complex metabolic mechanism studies into clear clinical targets, whilst ensuring these targets have broad applicability across heterogeneous patient populations. Moreover, although certain preliminary progress has been made in clinical trials of metabolic interventions in AML, such as improving the immune microenvironment through the regulation of glycolysis, FAO or amino acid metabolism, issues such as individual variability in efficacy, long-term treatment side effects and resistance remain prominent. Metabolic interventions may affect systemic metabolic homeostasis leading to unpredictable toxic reactions, and the adaptive metabolic mechanisms of tumor cells may induce resistance. Therefore, future research needs more precise targeting strategies to achieve efficient regulation within specific metabolic pathways, whilst minimizing systemic effects. The combination of advanced technologies, such as metabolomics and single-cell analysis for personalized treatment design, along with the development of combination therapies to mitigate resistance, may be key approaches to overcoming these challenges.