INTRODUCTION
The tumor microenvironment (TME) is a complex biological milieu in which malignant cells interact with stromal, vascular, and immune cells [1,2]. It comprises extracellular matrix (ECM), cancer-associated fibroblasts (CAFs), endothelial cells, pericytes, and infiltrating immune cells, which provide structural support, affect nutrient availability and metabolic conditions, and generate regulatory signals that control tumor growth, metastasis, and therapeutic responses [3,4]. Tumor-infiltrating T cells, dendritic cells, macrophages, neutrophils, mast cells, and natural killer (NK) cells together constitute the tumor immune microenvironment (TIME) and govern the initiation and maintenance of antitumor immunity [4,5].
Recognition of the immune system as a key regulator of cancer progression has led to substantial advances in cancer immunotherapy. Immune checkpoint inhibitors (ICIs) targeting programmed cell death protein 1 (PD-1)/PD-1 ligand 1 (PD-L1) and cytotoxic T-lymphocyte-associated protein 4 (CTLA-4), along with adoptive cellular therapies and other immune-based approaches, have achieved long-lasting clinical responses in certain patients across various malignancies [6-8]. However, many patients exhibit primary resistance to immunotherapy, while others develop acquired resistance following an initial response [9].
These limitations arise from both tumor-intrinsic mechanisms and immunosuppressive mechanisms within the TIME, including the accumulation of suppressive immune cells and persistent tumor antigen exposure that drive T-cell dysfunction and exhaustion [10-12]. Further, Immune cell function within tumors is also governed by nutrient availability and metabolic conditions, which modulate cellular metabolism, redox homeostasis, signaling pathways, and inflammatory responses [13,14]. Importantly, these metabolic conditions are determined not only by local tumor-associated factor but also by whole-body nutritional status and dietary exposures, suggesting a potential mechanistic connection between diet and antitumor immunity [15,16].
Diet represents a modifiable factor that influences host metabolism, inflammation, and immune function. Epidemiological and experimental studies indicate that dietary patterns and nutritional exposures are differentially associated with cancer risk, with some dietary patterns associated with increased risk and others with reduced risk [17]. High-fat and Western-style diets promote obesity, metabolic dysfunction, and chronic inflammation, and enhance tumorigenesis in experimental models of colorectal and breast cancer [18,19].
In contrast, diets rich in fruits, vegetables, whole grains, legumes, and other plant-based foods are associated with a reduced risk of certain cancers and provide bioactive compounds capable of modulating inflammatory and immune signaling pathways [20-22].
Beyond their direct effects on tumor cells, dietary factors can potentially alter whole-body metabolism, circulating metabolites, intestinal microbiota, and nutrient availability, thereby affecting immune cell recruitment, differentiation, activation, persistence, and effector function within the TIME [23]. For example, ω-3 polyunsaturated fatty acids (PUFAs) restrain inflammatory signaling and promote the resolution of inflammation [24]. However, the immunomodulatory effects of dietary components are often context-dependent and vary according to dose, duration and time of exposure, host metabolic status, tumor type, and concurrent anticancer treatment.
In this review, we discuss dietary components, including macronutrients, micronutrients, food-derived bioactive compounds, and metabolites generated through host or microbial metabolism, as well as dietary patterns that determine their availability. We focus on the molecular and cellular mechanisms by which these dietary factors modulate the TIME and regulate responses to cancer immunotherapy. In addition, we critically evaluate evidence from mechanistic studies, preclinical models, and clinical investigations, highlighting current limitations and future directions for translating dietary modulation into complementary strategies for cancer immunotherapy.
BARRIERS TO EFFECTIVE CANCER IMMUNOTHERAPY
The efficacy of ICIs, adoptive cellular therapies, and other immune-based approaches is limited by inadequate immune cell infiltration, progressive T-cell dysfunction, immunosuppressive cell populations, hypoxia, and metabolic constraints within the tumor [13,25]. These barriers can be broadly classified into four mechanisms. Because the same suppressive cell populations inhibit antitumor immunity through several mechanisms, we discuss each mechanism separately and identify the major cellular sources, rather than describing individual cell types independently.
Several of these barriers are closely linked to whole-body metabolic conditions and nutritional factors, providing a rationale for dietary modulation of the TIME [26].
T-cell exclusion and exhaustion
CD8+ T cells are major effectors of antitumor immunity and recognize tumor antigens through T-cell receptor (TCR) and kill malignant cells by releasing cytotoxic molecules, including granzymes and perforin [27]. The abundance and spatial distribution of tumor-infiltrating CD8+ T cells are associated with clinical outcomes and are incorporated into the Immunoscore as a prognostic measure in certain cancers [28,29]. However, effective antitumor immunity requires more than the presence of CD8+ T cells, as they must first enter tumors, proliferate, and maintain metabolic fitness to effectively execute their effector functions [27,30].
Limited T-cell infiltration is a feature of immunologically “cold” tumors and is associated with poor responses to immunotherapy.
The TME restricts T-cell infiltration, proliferation, metabolic fitness, and effector function at several levels. A dense and remodeled ECM, abnormal tumor vasculature, and stromal-cell-derived signals can limit T-cell trafficking and promote immune exclusion [31]. Persistent antigen exposure and chronic inflammatory signaling can then drive CD8+ T cells toward exhaustion, a dysfunctional state characterized by impaired cytokine production, reduced proliferative capacity, and increased expression of inhibitory receptors such as PD-1, T-cell immunoglobulin and mucin domain-containing protein 3 (TIM3), lymphocyte activating 3 (LAG3), and T cell immunoreceptor with Ig and ITIM domains (TIGIT).
T-cell exhaustion is a heterogeneous state in which progenitor-like exhausted cells retain proliferative and self-renewal capacity, whereas terminally exhausted populations exhibit more profound functional impairment [32,33]. Transcriptional regulators such as transcription factor 7 (TCF-7, also known as TCF-1), nuclear factor of activated T cells (NFAT), nuclear receptor subfamily 4 group A (NR4A) family members, thymocyte selection-associated high mobility group box protein (TOX), and B lymphocyte-induced maturation protein-1 (BLIMP-1, also known as PRDM1) contribute to the establishment and maintenance of exhausted T-cell states [33,34]. Metabolic stress is another driver of exhaustion and is discussed in “Hypoxia and nutrient competition” section.
Immunosuppressive signaling and cytokine competition
Immunosuppressive cells, including regulatory T cells (Tregs), tumor-associated macrophages (TAMs), and myeloid-derived suppressor cells (MDSCs), inhibit effector T-cell activity in the TIME through overlapping signaling mechanisms. These cells suppress antitumor immunity through the expression of inhibitory molecules, secretion of suppressive cytokines, disruption of antigen presentation, and competition for T-cell supporting cytokines such as IL-2.
Tregs are frequently enriched in cancers, particularly in solid tumors, resulting in an increased proportion of Tregs among tumor-infiltrating CD4+ T cells. An increased Treg-to-effector T-cell ratio has been associated with an immunosuppressive tumor milieu and, in several cancer types, with poorer clinical outcomes [35]. Tregs express inhibitory molecules, such as CTLA-4, that bind to CD80 and CD86 on antigen presenting cells (APCs) and reduce the costimulatory signals required for T-cell activation.
They also produce immunosuppressive cytokines, including interleukin (IL)-10 and TGF-β, which can directly inhibit effector T-cell function and indirectly impair T-cell activation by inhibiting APC function and antigen presentation. In addition, Tregs express higher levels of CD25, the α-chain of the IL-2 receptor, than conventional CD4+ and CD8+ T cells. By more efficiently consuming extracellular IL-2, they can limit IL-2 availability to conventional T cells and subsequently restrict their proliferation and survival.
Tregs can further suppress effector T cells through CD39/CD73-mediated production of extracellular adenosine, which inhibits T-cell activation and cytokine production [35,36].
TAMs are particularly abundant in many solid tumors and can suppress antitumor immune responses through multiple mechanisms [37]. TAMs highly express inhibitory ligands, including PD-L1 and PD-L2, which engage PD-1 on T cells and attenuate T-cells activation. They also secrete IL-10 and TGF-β, like those produced by Tregs, which inhibit effector T-cell function and promote an immunosuppressive environment.
In addition, TAMs drive the recruitment or expansion of Tregs and contribute to the accumulation of other immunosuppressive cell populations within tumors [37]. TAMs can also facilitate tumor progression by supporting angiogenesis, ECM remodeling, and cell invasion [38]. Notably, TAM consumption of glucose, lipids, and amino acids can reduce the availability of these metabolic substrates for T-cell proliferation and effector function.
Furthermore, TAM-associated metabolic pathways can generate immunosuppressive metabolites such as adenosine through CD39/CD73 activity and kynurenine through indoleamine 2,3-dioxygenase 1 (IDO1)-mediated tryptophan catabolism, thereby inhibiting antitumor responses [39,40].
MDSCs represent another important immunosuppressive population within the TIME and can limit antitumor T-cell responses [41]. MDSCs produce IL-10 and TGF-β, which inhibit effector T-cell function and contribute to a suppressive tumor milieu. MDSCs can deplete extracellular L-arginine required for T-cell proliferation and function through arginase 1 (ARG1) or metabolize L-arginine to NO through inducible iNOS, with NO and peroxynitrite impairing T-cell signaling and antigen recognition.
In addition, these cells can reduce extracellular tryptophan through IDO1, resulting in tryptophan depletion and kynurenine accumulation. Kynurenine can activate aryl hydrocarbon receptor (AhR) signaling and promote immunosuppressive responses. MDSCs also generate reactive oxygen species that can disrupt T-cell receptor signaling and further impair T-cell function.
Hypoxia and nutrient competition
Rapid tumor cell proliferation and structurally abnormal vasculature create heterogeneous regions of hypoxia and inadequate nutrient delivery. Tumor vessels are irregularly branched and poorly covered by pericytes, which increases vascular permeability and reduces perfusion. The resulting hypoxia and nutrient gradients create metabolic stress that impairs immune cell function, while poor perfusion can limit the delivery of immune cells and therapeutic agents [42,43].
Accordingly, normalizing the vasculature, for example by modulating VEGF signaling, has been explored as a strategy to improve immunotherapy [42].
Hypoxia activates hypoxia-inducible factor (HIF)-dependent transcriptional programs that promote tumor-cell survival while also impairing immune cell function. In T cells, hypoxia alters proliferation, cytokine production, cytotoxicity, and differentiation, and impaired mitochondrial function under hypoxia is linked to T cell dysfunction [44]. Under hypoxic conditions, HIF-1α signaling induces forkhead box P3 (FOXP3) expression and supports Treg differentiation and function [45,46], while hypoxia-induced CCL28 promote Tregs recruitment into tumors [47].
Hypoxia can also upregulate PD-L1 expression on myeloid cells [48], promotes recruitment and immunosuppressive polarization of TAMs and MDSCs [49,50], and increases CD39/CD73-mediated extracellular adenosine accumulation [51]. The immunosuppressive functions of these populations are described in the “Immunosuppressive signaling and cytokine competition” section.
Tumor cells and immune cells also compete for nutrients. Highly glycolytic tumor cells consume large amounts of glucose, and glucose restriction constrains T-cell glycolysis and impairs effector T-cell function (details in the “Carbohydrates and glucose metabolism” section) [52]. Depletion of amino acids can similarly compromise T-cell responses.
Arginine is required for T-cell survival and antitumor activity [53]. As discussed in the “Immunosuppressive signaling and cytokine competition” section, MDSC- and TAM-mediated arginine depletion can impair T cells function. Tumor cells can also outcompete T cells for methionine and glutamine, potentially compromising T-cell metabolic fitness and effector activity [54,55].
Similarly, IDO1-mediated tryptophan depletion and kynurenine accumulation can suppress T-cell proliferation and promote AhR-dependent immunosuppressive signaling [56,57]. In addition, lactate production and acidification of the extracellular space impair T-cell effector function and promote immunosuppressive macrophage polarization [58]. Tregs are comparatively tolerant of lactate-rich environment and can utilize lactate as a metabolic substrate [59], providing them with a metabolic advantage over effector T cells.
Microbiota-dependent immune regulation
The intestinal microbiota regulates antitumor immunity locally in the gut and systemically [60,61]. Commensal microorganisms contribute to the development, activation, and function of several immune cell populations involved in tumor control, and microbiota-derived metabolites regulate T-cell differentiation, myeloid-cell polarization, inflammatory signaling, and immune tolerance [60,61].
The microbiota has also emerged as an important factor determining responses to cancer immunotherapy. In preclinical studies, specific commensal bacterial species have been shown to enhance antitumor immunity and the efficacy of immune checkpoint blockade (ICB) [62]. Clinical studies have demonstrated an association between gut microbiota composition and responses to ICIs, and fecal microbiota transplantation from responding patients improved tumor control in recipient mice [63,64].
Conversely, antibiotic-induced disruption of the microbiota has been associated with poorer clinical outcomes in patients receiving ICIs [64,65]. Differences in microbial composition and function may therefore contribute to variation in treatment responses among patients.
Implications for dietary modulation
These findings highlight potential mechanisms through which dietary factors may modulate inflammatory and checkpoint signaling (“Immunosuppressive signaling and cytokine competition” section), alter substrate availability for effector and immunosuppressive immune cells (“Hypoxia and nutrient competition” section), and shape gut microbial composition and microbial metabolite production (“Microbiota-dependent immune regulation” section). In “DIETARY MODULATORS OF TIME” section, we discuss how specific dietary components and dietary patterns modulate these pathways and, in turn, antitumor immune responses.
DIETARY MODULATORS OF TIME
Dietary factors that regulate antitumor immunity include macronutrients, micronutrients, bioactive food-derived compounds, and broader dietary patterns and interventions. Their effects on the TIME may arise from direct modulation of immune cell metabolism and signaling, as well as indirect mechanisms involving whole-body metabolism, inflammation, the gut microbiota, and circulating metabolites. Importantly, dietary components rarely act through a single pathway, and their effects may vary depending on dose, duration of intervention, host metabolic status, tumor type, and cancer therapy [66].
Mechanistically, dietary factors may regulate the TIME through three major pathways (Fig. 1). First, direct mechanisms involve nutrients and metabolites reaching the TIME and interacting with immune cells through nutrient transporters, nutrient-sensing kinases (mTOR, AMP-activated protein kinase [AMPK], general control nonderepressible 2 [GCN2]), nuclear receptors (vitamin D receptor [VDR], retinoic acid receptor/retinoid X receptor [RAR/RXR], PPARs), and G protein-coupled receptors (GPCRs) including GPR41, GPR43, and GPR109A.
These interactions can regulate glycolytic and oxidative metabolism, epigenetic modifications, and effector-gene expression. Second, whole-body metabolism involve diet-induced changes in circulating glucose, insulin/insulin-like growth factor 1 (IGF-1), lipids, and inflammatory mediators, which influence nutrient availability and signaling pathways in T cells and myeloid cells. Third, microbiota-mediated mechanisms involve dietary modulation of gut microbial composition and the production of metabolites, including short-chain fatty acids (SCFAs), bile acids, tryptophan derivatives, which can influence immune-cell priming and function both locally and systemically.
The following sections discuss the major classes of dietary components, highlighting their underlying molecular mechanisms and potential effects on antitumor immunity. The major dietary components and their immunomodulatory effects are summarized in Table 1.
Macronutrients and metabolic substrates
Macronutrients, including carbohydrates, lipids, and proteins, provide energy and metabolic substrates that support cellular biosynthesis and immune function. Their availability and metabolism can regulate immune cell activation, differentiation and effector function [67]. Digestible carbohydrates provide glucose, whereas nondigestible carbohydrates such as dietary fiber modulate immune responses through gut microbial metabolism.
Dietary lipids provide fatty acids that regulate cellular metabolism and signaling pathways, and dietary proteins supply amino acids required for protein biosynthesis, metabolic reprogramming, and immune cell functions [67]. Within the TIME, tumor and immune cells compete for the same metabolic substrates, including glucose, fatty acids, and amino acids, and nutrient availability shapes CD8+ T-cell differentiation and effector function [68]. Consequently, nutrient competition and metabolic constraints within the TIME, as discussed in “Hypoxia and nutrient competition” section, may contribute to impaired antitumor immunity and resistance to cancer immunotherapy [68,69].
Carbohydrates and glucose metabolism
Carbohydrates are a major source of glucose and therefore affect whole-body glucose availability. Glucose is a major metabolic substrate for both malignant and immune cells. Many tumor cells exhibit elevated glycolytic rates, which makes dietary carbohydrate restriction a potential strategy to limit tumor growth [70].
However, T-cell activation and effector function also depend heavily on glucose metabolism [71,72]. Upon activation, T cells increase aerobic glycolysis to support the biosynthetic and energetic demands of rapid proliferation and effector function. Continuous NAD+ regeneration during lactate production is required to sustain glycolytic flux and thereby support effector cytokine production [72-74].
High glucose consumption by tumor cells can reduce glucose availability in the TIME and impair the function of tumor-infiltrating lymphocytes (TILs) [25,52]. Chang et al. demonstrated that tumor glucose consumption suppresses T-cell glycolytic capacity, mTOR activity, and IFN-γ production, and that ICB can restore glucose availability and T-cell glycolysis within the tumor [52]. Mechanistically, when glycolysis is active, GAPDH is predominantly engaged in the glycolytic flux rather than binding the 3′UTR of IFNG mRNA, thereby relieving translational repression and subsequently increasing IFN-γ production [71].
In addition, the glycolytic intermediate phosphoenolpyruvate (PEP) sustains TCR-triggered Ca²+-NFAT signaling by inhibiting sarco/endoplasmic reticulum Ca²+-ATPase (SERCA)-mediated Ca²+ re-uptake. Consequently, glucose deprivation can reduce intracellular PEP levels and impair NFAT-mediated T-cell activation [75]. Together, these mechanisms explain how glucose limitation impairs T-cell effector function in the TIME.
These findings also reveal a metabolic paradox: restricting glucose may constrain highly glycolytic tumor cells but also impair glucose-dependent antitumor T cells. Conversely, selectively enhancing glucose uptake and utilization by effector T cells may improve antitumor immunity. Supporting this concept, metabolic engineering of tumor-specific T cells to increase glucose transport and glycolytic capacity has been shown to enhance cytokine production and antitumor activity in preclinical models [76].
Dietary fiber and microbial metabolites
Dietary fiber is a distinct class of carbohydrates that resists digestion in the small intestine and can modulate host immunity primarily through the gut microbiota and microbial metabolism. Other poorly digestible carbohydrates, such as resistant starch, can act similarly. Fermentation of fiber and resistant starch by commensal bacteria generates SCFAs, principally acetate, propionate, and butyrate [77,78].
Resistant-starch supplementation has been demonstrated to increase SCFA production in many, although not all, human studies [79]. SCFAs function through GPCRs including GPR41, GPR43, and GPR109A and regulate cellular functions through mechanisms that include histone deacetylases (HDACs) inhibition [78,80]. However, the immunological effects of SCFAs are highly cell-type-dependent.
In human intestinal CD4+ T cells, butyrate and, to a lesser extent, propionate inhibited class I HDACs and increased histone acetylation at effector genes, yet suppressed T-cell activation, proliferation, and effector cytokine production [80]. In contrast, HDAC inhibition in naïve CD4+ T cells increased acetylation at the Foxp3 locus and promoted Treg differentiation [81]. In cytotoxic T cells, butyrate and pentanoate enhanced metabolic and epigenetic reprogramming, including elevated mTOR activity, and subsequently increased IFN-γ and TNF-α production and antitumor immunity in experimental mouse models [82].
These findings highlight the context-dependent immunomodulatory effects of SCFAs, which may either suppress or enhance T-cell responses depending on the cellular environment [80-82].
Beyond these direct effects on immune cell function, dietary fiber may influence cancer immunotherapy by remodeling the gut microbiota and its metabolic output. In patients with melanoma receiving ICB, higher dietary fiber intake was associated with improved progression-free survival, with the most pronounced association observed among patients reporting sufficient fiber intake without probiotic supplementation [83]. In mouse experiments, a low-fiber diet impaired responses to anti-PD-1 therapy and reduced the frequency of IFN-γ-positive cytotoxic T cells in the TIME [83].
These findings identify dietary fiber as a potentially modifiable nutritional factor, but the clinical studies primarily demonstrate an association rather than a causal benefit. However, a recent study across multiple murine tumor models found that isolated dietary fiber had limited and inconsistent effects on anti-PD-1 efficacy, while the overall dietary context affected the microbiome and metabolome more strongly than fiber content alone [84]. Prospective dietary intervention trials are therefore required to determine whether increasing dietary fiber intake directly improves cancer immunotherapy outcomes.
Dietary lipids and fatty acids
Dietary lipids provide fatty acids that serve as energy substrates, membrane components, and signaling molecules, thereby influencing both whole-body metabolism and immune cell function [85]. Lipid metabolism is frequently altered in tumors, and changes in lipid availability, uptake, synthesis, and oxidation affect both malignant and immune cells [86]. Excess lipid accumulation within the TIME can impair CD8+ T-cell function and contribute to an immunosuppressive metabolic environment [87].
However, fatty acid metabolism can also support T-cell adaptation to the metabolically restrictive TIME. In a mouse melanoma model, CD8+ TILs exposed to hypoglycemia and hypoxia exhibited increased PPAR-α signaling and fatty acid catabolism, which partially preserved effector function. Further enhancing fatty acid catabolism improved tumor control and synergized with PD-1 blockade [88].
Mechanistically, PPAR-α activation (e.g., by fenofibrate) up-regulates genes involved in fatty acid oxidation, including carnitine palmitoyltransferase 1A (CPT1A), thereby supporting ATP production and effector cytokine secretion in TILs under low glucose condition [88]. In contrast, CD36-mediated uptake of oxidized lipids can promote overall lipid peroxidation and ferroptosis in CD8+ T cells, which impairs their antitumor function [89]. More recent work has also shown that fatty acid metabolism supports the fitness and function of tumor-resident CD8+ T cells and improves anti-PD-1 treatment through a mechanism involving Scm polycomb group protein like 4 (SCML4) [90].
Thus, the effects of lipids on antitumor immunity depend not simply on lipid abundance alone but also on lipid composition, uptake pathways, and the metabolic fate.
PUFAs are precursors of bioactive lipid mediators. The ω-3 PUFAs eicosapentaenoic acid (EPA) and docosahexaenoic acid (DHA) are converted to specialized pro-resolving mediators (SPMs), including resolvins, protectins, and maresins, which actively resolve inflammation [91,92]. SPMs signal through GPCRs on macrophages and neutrophils, including lipoxin A4 receptor (ALX)/formyl peptide receptor 2 (FPR2), chemokine-like receptor 1 (ChemR23/CMKLR1), and GPR32, to inhibit NF-κB-driven cytokine release, limit neutrophil infiltration, and enhance phagocytic clearance of cell debris [93,94].
Because chronic unresolved inflammation can promote carcinogenesis, SPMs have been investigated for their potential chemopreventive and therapeutic effects [95]. In established tumors, SPMs may further modulate the TIME, particularly macrophage-mediated inflammation, and promote the resolution of therapy-induced inflammatory responses [96]. Sulciner et al. demonstrated that debris from tumor cells killed by cytotoxic or targeted therapies could stimulate tumor growth through macrophage-mediated inflammatory responses.
The resolvins RvD1, RvD2, and RvE1 inhibited debris-stimulated tumor progression, enhanced macrophage clearance of debris, and reduced the production of inflammatory cytokines and chemokines [97].
However, the resolution of inflammation does not necessarily enhance antitumor immunity, although it may prevent chronic inflammation and limit inflammation-associated cancer development. Efferocytosis, a process promoted by pro-resolving pathways, clears dying cells and suppress debris-driven inflammation, but in some tumor contexts it may also facilitate immune escape [98]. Efferocytosis is mediated by receptors such as MerTK, whose activation can promote an immunosuppressive macrophage polarization characterized by increased IL-10 and TGF-β production [99].
MerTK on tumor-associated leukocytes has been implicated in promoting tumor growth and metastasis [100], and in ICB-resistant hepatocellular carcinoma, macrophage efferocytosis was linked to fatty acid recycling to tumor cells and therapeutic resistance [101]. Thus, although resolution of chronic inflammation may help limit cancer development, activation of pro-resolving pathways in established tumors should not be assumed to uniformly enhance antitumor immunity.
Different fatty acid classes exert distinct effects on inflammatory and immune pathways, and PUFAs have been investigated as potential modulators of the responses to cancer immunotherapy [85,102]. However, evidence supporting specific dietary fatty-acid interventions remains less established than the mechanistic evidence linking cellular lipid metabolism to antitumor immunity. Therefore, despite promising preclinical and some clinical evidence, the translation of fatty acid-based nutritional interventions into effective cancer immunotherapy strategies remains challenging [102].
Protein and amino-acid availability
Dietary protein supplies amino acids, which serve as substrates for protein synthesis, metabolic intermediates, and regulators of cellular signaling. Amino acid metabolism is extensively reprogrammed within tumors, and competition between malignant and immune cells affects antitumor immunity [103,104]. Among these, arginine, tryptophan, glutamine, and methionine are particularly relevant to T cell proliferation, survival and function (“Hypoxia and nutrient competition” section).
Arginine depletion by arginase-expressing immunosuppressive populations and tumor cells contributes to T-cell dysfunction [105]. Mechanistically, ARG1-mediated arginine depletion downregulates CD3ζ expression and induces T-cell cell-cycle arrest at the G0–G1 phase [106]. Conversely, increased arginine availability can promote oxidative metabolism and memory-like T-cell differentiation, thereby enhancing antitumor activity [53].
Tryptophan metabolism represents another prominent mechanism of metabolic immune suppression. IDO- and tryptophan 2,3-dioxygenase (TDO)-catalyzed tryptophan catabolism reduces local tryptophan availability and promotes kynurenine accumulation, which suppresses T-cell responses and favors immune escape [107]. Tryptophan depletion activates the stress kinase GCN2 in T cells, potentially contributing to proliferative arrest, while kynurenine acts as an AhR ligand that can promote Treg differentiation and PD-1 expression on CD8+ T cells [56,57,108].
Methionine availability also influences T-cell epigenetic regulation. Tumor cells can compete with CD8+ T cells for methionine through SLC43A2 transporter, reducing H3K79me2 and STAT5 expression in T cells, subsequently impairing their effector function [54]. Glutamine metabolism is similarly important for rapidly proliferating tumor and immune cells, making glutamine availability and utilization potential target for modulating TIME [109].
In preclinical models, pharmacological glutamine blockade has been shown to impair tumor cell metabolism while promoting metabolic adaptation and memory-like phenotype in T cells [109,110].
These observations have stimulated interest in targeting amino acid metabolism to improve immunotherapy [103,104]. However, local amino-acid competition within tumors should not be equated with dietary amino acid restriction or supplementation. Because the same amino acids can support both tumor-cell growth and immune cell function, dietary manipulation of protein or individual amino acids may produce divergent effects depending on cell type, immune status, and metabolic context.
Thus, evidence supporting specific modifications of dietary protein or individual amino acids as established strategies to improve immunotherapy remains limited.
Micronutrients
Micronutrients, including vitamins and trace elements, are required in relatively small amounts but play significant roles in regulating immune cell differentiation, signaling, redox homeostasis, and metabolism. Their availability may therefore influence both malignant and immune cells in the TIME. Their effects depend on nutritional status, dose, and tumor type, and current evidence is insufficient to support micronutrient supplementation as an established strategy to improve immunotherapy outcomes [111].
Vitamin D
Vitamin D has received particular attention because vitamin D receptor signaling regulates both innate and adaptive immune responses and may modulate immune cell function within the TIME. In the prospective PROVIDENCE study, systematic vitamin D supplementation in patients receiving ICIs was associated with improved clinical outcomes and a lower incidence of thyroid-related immune adverse events [112]. Vitamin D acts through the nuclear VDR/RXR complex, which binds to vitamin D response elements to regulate gene transcription in T cells, dendritic cells, and macrophages.
It generally promotes tolerogenic dendritic cells and Tregs while attenuating Th1/Th17 inflammation [113,114]. These immunoregulatory effects contribute to the reduced incidence of immune-related adverse events, although they do not necessarily explain the observed improvement in clinical outcomes [112]. In mice, vitamin D acting on intestinal epithelial cells reshaped the gut microbiota, including Bacteroides fragilis, toward a composition that support antitumor immunity and enhanced ICB responses [115].
However, these findings do not establish vitamin D supplementation as a standard strategy to improve immunotherapy efficacy and require confirmation in controlled trials.
Vitamin A/retinoids
Vitamin A and its bioactive metabolite retinoic acid (RA) regulate immune cell differentiation, trafficking, and function [116]. Within tumors, RA signaling appears highly context dependent. Experimental studies have demonstrated increased RA signaling in tumor-associated CD8+ T cells and a requirement for intrinsic RA signaling for the survival of tumor antigen-specific CD8+ T-cell [117], while other studies have investigated T-cell-intrinsic vitamin A metabolism as a potential target for memory T-cell-based immunotherapy [118].
RA signals through RAR–RXR heterodimers that bind retinoic acid response elements and differentially regulates T-cell survival, differentiation, and the expression of gut-homing receptors including α4β7 and CCR9, across T-cell subsets [116]. The same pathway promotes Treg differentiation together with TGF-β, which is one reason why vitamin A effects depend on cell type and context. Thus, the immunological effects of vitamin A should not be assumed to translate directly into benefits from dietary vitamin A supplementation.
Vitamin C
Vitamin C has antioxidant and metabolic functions, but pharmacological high-dose vitamin C can have biological effects that differ substantially from those of ordinary dietary intake. In preclinical breast, colorectal, melanoma, and pancreatic tumor models, high-dose vitamin C altered immune cell infiltration, delayed tumor growth in a T-cell-dependent manner, enhanced the cytotoxicity of adoptively transferred CD8+ T cells, and showed enhanced antitumor effects when combined with ICB [119]. Ascorbate is a cofactor for Fe²+/α-ketoglutarate-dependent dioxygenases, including ten-eleven translocation (TET) enzymes involved in DNA demethylation, can thereby regulate DNA methylation and gene expression [120].
At pharmacological concentrations, it can also act as a pro-oxidant and promote extracellular H2O2 generation, leading to oxidative damage in susceptible tumor cells [121]. However, the extent to which these mechanisms directly contribute to the effects of high-dose vitamin C on tumor-infiltrating T cells remains unclear. In mouse tumor models, high-dose vitamin C increased CD4+ and CD8+ T-cell infiltration and activation, and its antitumor effect required T cells; antioxidant co-treatment did not abolish it [119].
In experimental non-small cell lung cancer (NSCLC) model, high dose ascorbate enhanced CD8+ T-cell cytotoxicity and improved the antitumor effects of anti-PD-1 treatment [122]. These findings justify further investigation but do not establish that increasing dietary vitamin C intake enhances immunotherapy efficacy.
Trace elements
Zinc and selenium contribute to enzyme activity, redox regulation, cellular signaling, and immune function, but their effects on antitumor immunity are context dependent. In preclinical tumor models, high dietary zinc increased Foxp3+ Treg frequency and promoted tumor progression, whereas zinc chelation enhanced anti-PD-1 responses in melanoma models [123]. Zinc serves as both a structural component and a signaling ion for T cells.
For example, it contributes to the tethering of the Src-family tyrosine kinase Lck to the CD4 and CD8 coreceptors [124]. In melanoma-bearing mice, high dietary zinc intake increased Foxp3+ Tregs frequency and accelerated tumor progression, potentially through sirtuin 1 (SIRT1)–FOXO1 signaling involving TGF-β contributing. Genetic deletion of FOXO1 in CD4+ T cells attenuated these effects, whereas zinc chelation with clioquinol enhanced anti-PD-1 therapy [123].
Selenium acts many of its biological effects through selenoproteins including glutathione peroxidase 4 (GPX4) [125], which protects T cells from lipid peroxidation and ferroptosis [126]. More broadly, evidence linking trace-element manipulation to ICB remains predominantly preclinical, and direct human evidence for clinically actionable interventions is limited [127]. Therefore, changes in trace-element intake cannot currently be assumed to improve immunotherapy outcomes.
Food-derived bioactive compounds
Plant-derived foods contain numerous bioactive compounds, commonly referred to as phytochemicals, that can regulate cellular processes relevant to cancer development and progression [128]. Among them, polyphenols have received particular attention because of their effects on signaling pathways involved in tumor progression and antitumor immunity. Curcumin (from turmeric), resveratrol (from grapes and other plants), and epigallocatechin-3-gallate (EGCG, from green tea) are representative polyphenols that modulate inflammatory and oncogenic signaling pathways, including NF-κB, PI3K/Akt/mTOR, and STAT3 [129].
More recently, their potential effects on immune-checkpoint signaling and antitumor T-cell responses have generated interest in their possible relevance to the TIME and immunotherapy [130]. The following sections discuss representative polyphenols and related compounds with experimental evidence supporting their their immunomodulatory effects and potential interactions with immune checkpoint pathways.
Curcumin
Curcumin, a polyphenol from turmeric, has been extensively investigated for its anticancer and immunomodulatory properties. In addition to its direct effects on tumor-cell signaling, it can modulate immune-checkpoint signaling, T-cell function, and TAM phenotypes in experimental models [131,132].
In head and neck squamous cell carcinoma (HNSCC) models, curcumin modulated multiple inhibitory immune-checkpoint pathways. Liu et al. examined checkpoint ligands in tumor cells and PD-1 and TIM3 expression on CD4+, CD8+, and Treg populations, and evaluated the effects of curcumin combined with PD-L1 blockade using T-cell cytotoxicity and cytokine assays [131]. In tumor-bearing mice, curcumin preserved effector and memory T-cell populations, attenuated suppression of T-cell proliferation, reduced Treg-mediated immunosuppression, and enhanced effector T-cell-mediated tumor cell killing [133].
Similarly, in a preclinical adoptive T-cell therapy model, curcumin increased tumor accumulation and function of CD8+ T cells and enhanced their cytotoxicity, accompanied by reduction in TGF-β, IDO, and Tregs [134]. In patients with lung cancer, curcumin treatment was associated with decreased peripheral Treg frequency and increased Th1 cells, and in vitro experiments showed that curcumin promoted the conversion of patient-derived FOXP3+ Tregs toward a Th1-like phenotype, with reduced FOXP3 transcription and increased IFN-γ expression [135]. This human evidence is noteworthy, although it does not establish that curcumin improves clinical outcome following cancer immunotherapy.
Curcumin may also affect the myeloid compartment of the TIME. In an experimental oral cancer system, curcumin reprogrammed M2-like TAMs toward a more proinflammatory phenotype, decreasing TGF-β, ARG1, IL-10, and CD206 while increasing TNF-α, iNOS, IL-6, and CD86. This effect was associated with inhibition of monoamine oxidase A (MAO-A)/STAT6 signaling [132].
Dendrosomal curcumin shifted macrophage-associated markers toward an M1-like profile in a mouse model of metastatic breast cancer [136]. Similarly, in glioblastoma models, curcumin altered tumor-associated microglia, suppressing ARG1-high/iNOS-low tumor-promoting phenotype while promoting an ARG1-low/iNOS-high phenotype, with activation of NF-κB and STAT1 [137].
More recently, curcumin has been shown to modulate the PD-1/PD-L1 axis and CD8+ T-cell responses. In anaplastic thyroid carcinoma models, curcumin reduced tumor cell PD-L1 expression by inhibiting AKT/mTORC1/STAT3 signaling and increased IFN-γ-producing CD8+ T cells. In a syngeneic mouse model, curcumin increased CD8+ T-cell infiltration, and combination with anti-PD-1 further enhanced CD8+ T-cell function and antitumor immunity compared with either treatment alone [138].
Studies in colorectal cancer models have also suggested that curcumin can potentiate PD-1/PD-L1 blockade through gasdermin E (GSDME)-dependent pyroptosis, involving increased GSDME expression and caspase-3-mediated GSDME cleavage, accompanied by changes in tumor-infiltrating immune populations [139].
Together, these findings indicate that curcumin can modulate effector T-cell, Tregs, TAMs, and tumor-associated microglia, in addition to regulating immune-checkpoint signaling. However, most of this evidence derives from preclinical models, often using purified or formulated curcumin rather than dietary turmeric. Thus, these findings provide mechanistic support for curcumin as an investigational immunomodulatory compound but do not establish that dietary curcumin intake improves responses to cancer immunotherapy.
Resveratrol
Resveratrol is a stilbene found in grapes, berries, peanuts, and other plant-derived foods. In addition to its direct effects on malignant cells, resveratrol has been investigated for effects on TAMs, CD8+ T cells, and immune-checkpoint signaling [140,141]. In a murine lung adenocarcinoma model, resveratrol treatment inhibited tumor growth, decreased the proportion of M2-like TAMs, and increased the activation of tumor-infiltrating CD8+ T cells.
Coculture and antibody-neutralization experiments further suggested that macrophage-derived IL-18 contributed to CD8+ T-cell activation [141]. Verdura et al. demonstrated that resveratrol interfered with PD-L1 N-linked glycosylation by inhibiting enzymes involved in PD-L1-processing, resulting in accumulation of an abnormally glycosylated form of PD-L1 in the endoplasmic reticulum and impaired trafficking of PD-L1 to the tumor-cell surface [142]. Resveratrol has also been investigated for its effects on CAR T-cell metabolism.
In an in vitro CD123-specific CAR-T-cell model, resveratrol increased mitochondrial respiratory capacity and enhanced antitumor function in T cells derived from cytomegalovirus (CMV)-seropositive donors, suggesting that its effects may depend on the donors' prior immune history [143].
However, resveratrol does not uniformly reduce tumor-cell PD-L1 expression. In breast and colorectal cancer cells, resveratrol increased surface PD-L1 expression through NF-κB signaling [144]. Similarly, in lung cancer cells, resveratrol induced PD-L1 expression through Snail-mediated activation of Wnt signaling [145].
These apparently divergent observations suggest resveratrol can regulate PD-L1 at multiple levels, including transcriptional, post-translational glycosylation, and intracellular trafficking. Therefore, changes in total PD-L1 expression should not necessarily be interpreted as equivalent to changes in cell surface PD-L1 abundance or its immunoregulatory activity.
It has been proposed that resveratrol-induced PD-L1 expression might, under selected experimental conditions, make tumors with low baseline PD-L1 expression more susceptible to PD-L1 blockades, but it remains a hypothetical therapeutic approach rather than an established clinical benefit [146]. Furthermore, most available evidence derives from preclinical studies using purified resveratrol, and whether dietary resveratrol intake can meaningfully modulate the TIME or improve immunotherapy outcomes remains uncertain.
EGCG
EGCG is the most abundant catechin in green tea and is a dietary polyphenol with anticancer and immunomodulatory properties. Experimental evidence suggests that EGCG can regulate tumor-cell immune checkpoint expression, T-cell function and immunosuppressive myeloid cells [147,148]. In melanoma models, EGCG inhibited IFN-γ-induced PD-L1 and PD-L2 expression by suppressing JAK/STAT signaling.
Mechanistically, EGCG reduced STAT1 expression and phosphorylation, resulting in decreased interferon regulatory factor 1 (IRF1) expression and subsequent downregulation of PD-L1 and PD-L2 [147].
The tumor growth-inhibitory effect of EGCG in vivo was dependent on CD8+ T cells [147]. In T cells, EGCG reduced PD-1 expression by inhibiting NF-κB phosphorylation and nuclear translocation [148]. In addition, EGCG attenuated MDSC-mediated immunosuppression in 4T1 murine breast cancer model, reducing MDSC accumulation and suppressive signaling associated with the ARG1/iNOS/Nox2/NF-κB/STAT3 pathway [149].
These preclinical findings indicate that EGCG can modulate the PD-L1/PD-L2–PD-1 axis, CD8+ T-cell responses, and MDSC-mediated immunosuppression [147-149]. However, the effect of purified EGCG should not be equated with those of green tea consumption, and evidence that dietary EGCG improves immunotherapy outcomes in patients remains insufficient.
Clinical translation of dietary polyphenols is limited by factors including poor absorption, rapid metabolism, uncertain activity of circulating metabolites, formulation variability, and potential drug interactions [129]. Human studies of curcumin, resveratrol, and EGCG have predominantly demonstrated biomarker modulation, while consistent benefits in terms of tumor regression, recurrence, progression-free survival, or overall survival have not been established [129]. Thus, EGCG should currently be regarded as an investigational immunomodulatory compound rather than an established nutritional intervention.
DIETARY PATTERNS AND INTERVENTIONS IN CANCER IMMUNOTHERAPY
Dietary patterns differ from individual nutrients because they simultaneously alter the intake of multiple macronutrients, micronutrients, and bioactive compounds and can broadly change in whole-body metabolism and the gut microbiota. Consequently, dietary interventions may affect antitumor immunity through multiple interconnected mechanisms rather than through a single nutrient-dependent pathway. Mediterranean dietary patterns, ketogenic diets, and fasting or fasting-mimicking interventions have received particular attention.
However, the strength of evidence varies considerably among these approaches, ranging from observational associations in patients to mechanistic studies conducted predominantly in experimental tumor models.
Mediterranean dietary patterns
The Mediterranean diet is characterized by relatively high consumption of whole grains, vegetables, fruits, nuts, legumes, fish, and unsaturated fats, with lower intake of highly processed foods and red and processed meats. It provides several components with potential immunomodulatory activity, including dietary fiber, unsaturated fatty acids, micronutrients, and plant-derived bioactive compounds. These components may act together on whole-body metabolism, inflammation, and the gut microbiota, providing potential mechanisms through which habitual diet could modify responses to cancer therapy.
In a multicenter cohort of 91 patients in the Netherlands and United Kingdom receiving ICB, habitual dietary intake was assessed before treatment. Greater adherence to a Mediterranean dietary pattern was positively associated with overall response rate and 12-month progression-free survival [150]. These findings support an association between overall dietary composition and ICB outcomes and complement evidence linking individual components of plant-rich diets, particularly dietary fiber, with immunotherapy responsiveness.
Although the underlying mechanisms were not directly investigated in this cohort, potential pathways include fiber fermentation to SCFAs (HDAC/GPR signaling, “Dietary fiber and microbial metabolites” section), ω-3 fatty acid-derived SPMs (“Dietary lipids and fatty acids” section), and polyphenol-mediated modulation of NF-κB/STAT3 and immune checkpoint signaling (“Food-derived bioactive compounds” section). However, the contribution of these pathways to the observed clinical associations remains hypothetical [150].
Because these findings are observational, they do not establish that adopting a Mediterranean diet after cancer diagnosis will improve ICB efficacy, nor do they identify which components of the dietary pattern are responsible for the observed relationship. Larger prospective studies and dietary intervention trials are therefore required to determine whether modifying habitual dietary patterns can causally enhance antitumor immunity or immunotherapy efficacy.
Ketogenic diets and ketone metabolism
Ketogenic diets involve marked carbohydrate restriction and increased reliance on fat-derived energy substrates, resulting in hepatic production and increased circulating concentrations of ketone bodies. These metabolic changes have raised interest in ketogenic diets as potential modifiers of tumor metabolism and antitumor immunity.
Ferrere et al. investigated the association between ketogenic metabolism and ICB in experimental tumor models. A ketogenic diet, or administration of 3-hydroxybutyrate (3HB), a major circulating ketone body, induced T-cell-dependent inhibition of tumor growth. In mice on a standard diet, anti-PD-1 alone or combined anti-PD-1/anti-CTLA-4 therapy was ineffective, but ketogenic feeding or oral 3HB restored therapeutic responses [151].
Mechanically, 3HB prevented ICB-associated PD-L1 upregulation on myeloid cells and promoted the expansion of CXCR3+ T cells. These effects involved signaling through GPR109A (HCAR2), a receptor of 3HB on myeloid cells [151]. Disrupting ketogenesis by adding sucrose to the diet or pharmacological inhibition of GPR109A, abolished the antitumor effect.
The ketogenic diet also altered the gut microbiota composition, suggesting that both metabolic and microbiota-dependent mechanisms may contribute [151].
Although these findings provide mechanistic evidence that ketogenic metabolism can modulate antitumor immune responses and ICB, the evidence supporting improved immunotherapy outcomes in patients remains limited. A ketogenic diet is also a substantial dietary intervention rather than simple carbohydrate restriction, and its effects on body weight, nutritional status, lipid metabolism, and treatment tolerance must be considered in patients with cancer. Thus, the enhancement of ICB responses by ketogenic diets or ketone bodies in preclinical models does not yet justify their routine use during cancer immunotherapy.
Caloric restriction, fasting, and fasting-mimicking diets
Fasting and fasting-mimicking diets (FMDs) modify whole-body nutrient availability and metabolic signaling. Unlike chronic caloric restriction, they involve defined periods of marked energy or nutrient restriction followed by refeeding. Experimental evidence indicates that they can affect both tumor-cell sensitivity to treatment and immune cell composition and function.
In mouse models of breast cancer and melanoma, cycles of an FMD combined with chemotherapy increased bone-marrow common lymphoid progenitors and cytotoxic CD8+ TILs, enhanced tumor-cell apoptosis, and delayed tumor progression [152]. In breast tumors, the increase in CD8+ T-cell infiltration was partially associated with downregulation of heme oxygenase-1 (HO-1), accompanied by reduced intratumoral Treg abundance. Although FMDs reduced circulating glucose and IGF-1, the study primarily implicated hematopoietic modulation and HO-1-associated changes in the intratumoral immune compartments [152].
In mouse colon cancer models, FMD reduced tumor-infiltrating IgA+ B cells by suppressing B-cell class switching to IgA, thereby relieving IgA+ B-cell-mediated suppression of CD8+ T cells [153]. Adoptive transfer of IgA+ B cells reversed this effect, and enhanced fatty acid oxidation in CD8+ T cells was also observed [153]. In patients with cancer, a study evaluating the feasibility of FMD reported reductions in circulating glucose, insulin, and IGF-1 levels, decreased immunosuppressive myeloid cell populations, and enhanced cytotoxic T-cell and NK-cell responses, and increased intratumoral CD8+ T-cell infiltration and activation [154].
These findings provide a translational rationale for further investigation but do not establish that fasting or FMDs improve clinical responses to ICB.
Clinical implementation requires caution. Patients with cancer may be vulnerable to weight loss and nutritional deterioration, and the metabolic demands and safety of dietary restriction are likely to vary according to tumor type, disease stage, treatment regimen, and baseline nutritional condition. Controlled clinical trials are therefore necessary to establish whether fasting interventions can improve immunotherapy efficacy without compromising nutritional status.
CONCLUSION AND PERSPECTIVES
Diet may affect antitumor immunity through direct effects on immune cell metabolism, whole-body metabolic and inflammatory changes, and the intestinal microbiota. Key molecular pathways implicated in experimental studies include mTOR-dependent regulation of T-cell glycolysis and effector function [52,71,75], HDAC-mediated epigenetic regulation by SCFAs and ketone bodies [80-82], GPCR signaling through receptors such as GPR109A and those involved in SPM receptor signaling [78,93,94,151], JAK/STAT and NF-κB regulation of immune checkpoints [138,144,147,148], and amino-acid-sensing pathways involving GCN2 and AhR [56,57,108].
Mechanistic studies have linked macronutrients and metabolic substrates, micronutrients, food-derived bioactive compounds, and dietary patterns to processes that can limit immunotherapy, including T-cell exhaustion, immunosuppressive signaling, hypoxia and nutrient competition, and microbiota-dependent immune regulation. However, most supporting evidence remains preclinical or observational. Notable clinical observations include associations between dietary fiber intake [83] or a Mediterranean-style dietary pattern [150] and ICB outcomes in patients with melanoma, together with changes in immune-related biomarkers or clinical outcomes reported in selected supplementation studies [112,129,135].
Nevertheless, these associations do not establish causality, and no dietary intervention reviewed here has been established as an effective strategy for improving cancer immunotherapy outcomes.
Several important limitations and challenges hinder the clinical translation of these mechanistic findings into clinical practice. First, the evidence base is uneven: most findings come from mouse models, cell lines, or observational clinical studies, whereas prospective dietary intervention trials in patients receiving immunotherapy remain limited. Observational studies are also susceptible to residual confounding and errors in dietary assessment, making it difficult to distinguish the independent effects of dietary exposures from those of overall lifestyle, nutritional status, and concurrent cancer treatments.
Second, many favorable experimental findings involve purified compounds at pharmacological doses, such as high-dose ascorbic acid, curcumin, resveratrol, and EGCG [119,129], which may not reflect exposure achievable through ordinary food intake. Poor absorption, rapid metabolism, uncertain activity of circulating metabolites, and variation among formulations further limit translation [129]. Moreover, isolated nutrients or bioactive compounds may not reproduce their effects when consumed as components of foods or dietary patterns.
Third, the direction and magnitude of immunological effects are highly context dependent and, in some cases, conflicting. Glucose restriction may constrain tumor cells yet simultaneously impair T cells [52,71,75]. SCFAs enhance cytotoxic T-cell function in some experimental systems and inhibit CD4+ T cells in others [80-82].
Similarly, pro-resolving lipid signaling may support resolution of inflammation [97] but may also contribute to efferocytosis-mediated immune suppression under certain tumor conditions [99-101]. Resveratrol has been reported to both reduce and induce PD-L1 depending on the experimental setting [142,144,145]. Such divergent effects likely reflect differences in dose, duration and timing, host metabolic status, tumor type, and concurrent therapy.
Comparisons across studies are further complicated by differences in experimental models, dietary regimens, and immunological endpoints. Importantly, changes in immune cell abundance, activation markers, or circulating metabolites do not necessarily translate into improved tumor control or clinical outcomes.
Fourth, individual differences in baseline diet, gut microbiota, metabolic status, nutritional state, and treatment history may contribute to heterogeneous responses and limit the reproducibility of reported findings. In addition, patients with cancer may be vulnerable to weight loss and nutritional deterioration, which can limit the safety of restrictive approaches such as ketogenic diets or fasting. Such interventions require careful patient selection and nutritional monitoring, particularly in patients with poor appetite or treatment-related gastrointestinal toxicity.
These diets must also be practical and acceptable to patients, and should be evaluated for adherence and potential interactions with cancer treatments [151,154].
Future work should therefore focus on translating mechanistic observations into clinical testable nutritional strategies. Research priorities include: (1) Prospective dietary intervention trials in patients receiving ICIs or other immunotherapies, with clinical outcomes, nutritional status, and treatment tolerance systematically assessed; (2) Standardized and quantitative dietary assessment to distinguish the effects of overall dietary patterns from those of individual nutrients or supplements; (3) Pharmacokinetic and tissue-exposure studies to determine whether dietary or supplemental intake produces biologically relevant concentrations of bioactive compounds and their metabolites; (4) Integrated biomarker analyses combining the gut microbiome, circulating metabolites, and tumor immune profiles to identify patients most likely to benefit from dietary interventions; (5) Human mechanistic studies, including longitudinal and paired tumor sampling where feasible, to determine whether mechanisms identified in preclinical models are also present in patients. Ultimately, well-controlled prospective trials integrating clinical, nutritional, metabolic, microbiome, and immune measurements will be essential to determine whether dietary interventions can safely and reproducibly enhance cancer immunotherapy and to identify the patients and treatment contexts most likely to be benefit.