Gut microbiota drives cancer evolution and therapy resistance

Abstract

Initially recognized for its role in digestion, the gut microbiota is now acknowledged as a critical regulator of the host immune system. It influences not only the initiation and progression of cancer but also key therapeutic mechanisms through metabolic reprogramming, production of virulence factors, and remodeling of the immune microenvironment. Despite significant advancements in patient outcomes attributed to novel multi-target inhibitors and immune checkpoint blockers, several clinical challenges persist, including therapy resistance, low response rates, and considerable toxicity. The increasing recognition of the role of microbiota in oncology has positioned its modulation as a pivotal strategy influencing the efficacy and overall success of treatment. This review summarizes how alterations in the gut microbiota influence tumorigenesis, shape immune responses, and mediate resistance to cancer therapies. This study also explores therapeutic strategies designed to modulate the microbiota by restoring ecological balance, enhancing anti-cancer immunity, or rectifying metabolic dysregulation to overcome therapy resistance and improve clinical outcomes. These approaches present a novel paradigm for addressing the challenge of drug resistance in cancer therapy.

Introduction

The human gastrointestinal tract hosts an extensive community of approximately 38 trillion microorganisms, primarily bacteria, which achieve their highest density within the colon [1]. This collective constitutes a complex ecosystem, often termed the "second genome," due to its significant impact on host health, which arises from its intricate functions [2]. Gut microbiota (GM), through their extensive metabolic capabilities, engages in a symbiotic relationship with the host, playing a crucial role in regulating nutrient metabolism, immune function, and the integrity of the intestinal barrier [3, 4]. The microbiota performs essential functions both within the intestine and systemically. For instance, microbe-derived short-chain fatty acids (SCFAs), such as butyrate, contribute to the maintenance of the intestinal barrier by promoting mucus secretion and modulating pH levels. Additionally, SCFAs like butyrate directly and indirectly regulate immune homeostasis, thereby significantly enhancing the functional capacity of the host [5, 6].

Recent studies highlight the significant role of GM dysbiosis in promoting carcinogenesis. Colibactin, a genotoxin produced by adherent-invasive Escherichia coli (E. coli), induces DNA damage and contributes to colorectal carcinogenesis [7], while the virulence factor CagA from Helicobacter pylori (H. pylori) facilitates gastric cancer by disrupting Wnt-mediated planar cell polarity signaling, thereby dysregulating stem cell dynamics [8]. In addition to direct infection, GM and their metabolites exert a significant influence on cancer by reprogramming tumor cell metabolism and enhancing stemness. These metabolites activate oncogenic pathways and contribute to the formation of a pro-tumorigenic microenvironment that supports cancer stem cells and their aggressive phenotypes [9, 10]. For example, The colorectal cancer (CRC)-associated bacterium Fusobacterium nucleatum (F. nucleatum) produces formate, which maintains the stemness of CRC [11]. In a mouse melanoma model, butyrate was shown to promote both a FOXO1-driven stemness program and the differentiation of tumor-specific CD127⁺CD8⁺ T cells [12]. Malignant progression is driven by the accumulation of microbial metabolites or the crosstalk between tumour and immune cells within the tumour microenvironment (TME) [13]. Notably, distinct microbial signatures have emerged as potent, non-invasive biomarkers. For instance, the prevalence of Streptococcus anginosus and Streptococcus constellatus in stool samples demonstrates high sensitivity and accuracy in the detection of early-stage gastric cancer [14]. We have summarized the GMs of these potential tumor biomarkers in Table 1.

Table 1.

Gut microbiota-derived biomarkers in cancer diagnostics

Cancer type	Research design	Marker microbiota	Biomarker functions or clinical significance	References(PMID)
CRC	Retrospective cohort study	Neisseria oralis, Campylobacter gracilis, Treponema medium	Potential combined biomarkers for predicting disease progression in CRC	40325090
CRC	Prospective cohort study	Fusobacterium, Parvimonas, Bacteroides, Faecalibacterium	Fecal biomarkers for early diagnosis of CRC	38366793
CRC	Retrospective study	Porphyromonas endodontalis, Ruminococcus torques, Odoribacter splanchnicus, Fusobacterium nucleatum, Parvimonas micra	Associated with CRC staging and progressive loss of microbial diversity during disease progression	36992683
CRC	Retrospective study	Streptococcus thermophilus TH1435	Key biomarkers for distinguishing adenoma from CRC, contributing to early diagnosis	34031391
CRC	Retrospective study	Eubacterium rectale, Faecalibacterium prausnitzii	High-accuracy prediction of CRC based on gut microbial single-nucleotide variation profiles	33430705
CRC	Prospective cohort study	Ratio of Fusobacterium nucleatum to Faecalibacterium prausnitzii and Bifidobacterium spp.	A valuable biomarker for early-stage CRC screening	29914865
CRC	Prospective cohort study	Fusobacterium nucleatum	An independent predictor of poor prognosis	26311717
CRC	Prospective study	Alistipes indistinctus	Non-invasive indicator of lymphovascular invasion status in CRC	37249979
CRC	Multicenter cross-sectional study	Fusobacterium spp.	Progressive enrichment of Fusobacterium with advancing CRC stage	32370168
GC	Prospective study	Roseburia spp.	A biomarker predicting peritoneal metastasis in GC	39109683
GC	Retrospective	Helicobacter pylori	Useful for prognostic stratification and as a predictor of chemoresistance	40519938
GC	Retrospective case–control study	GC microbiome index (Lacticaseibacillus, Haemophilus, Campylobacter)	Independent predictors of GC risk	38509701
HCC	Prospective study	Odoribacter splanchnicus, Ruminococcus bicirculans	Synergistic use with serum metabolites enhances diagnostic efficacy for HCC	37260707
HCC	Retrospective study	Bacteroides, Lachnospiraceae incertae sedis, Clostridium XIVa	Prognostic biomarkers for adverse outcomes in HCC	33225985
HCC	Prospective study	Ruminococcus, Faecalibacterium, Coprococcus	Diagnostic model establishment for early diagnosis	30045880
Small bowel adenocarcinoma	Retrospective cohort study	Proteobacteria, Actinobacteria	Integration of gene mutations for postoperative recurrence risk prediction	34449929
Esophageal squamous cell carcinoma	Prospective cohort study	Prevotella spp.	Integration of Bacteroides as a biomarker for cancer diagnosis	38497715
Breast cancer	Retrospective study	Clostridium XIVa, Bacteroides, PrevotellaPrevotella	Biomarkers for early diagnosis of breast cancer	36583106

CRC colorectal cancer, GC gastric cancer, HCC hepatocellular carcinoma

Consequently, the GM is now acknowledged as a pivotal regulator across the entire cancer continuum, encompassing initiation, progression, and treatment response, rather than merely a passive participant. Beyond its role in promoting tumorigenesis, the microbiota contributes to therapy resistance by modulating mechanisms such as DNA repair and cell death pathways, thereby diminishing drug efficacy [15]. Drug resistance, a defining characteristic of cancer stemness, represents a significant clinical challenge that frequently results in treatment failure. Current research endeavors to elucidate the multifaceted mechanisms underlying resistance, which include tumor burden, driver gene mutations, and modifications in the TME [16]. The effectiveness of immunotherapy is significantly associated with the presence of specific bacterial strains [17].

In the field of oncology, the gut microbiome presents threefold potential: as a predictive biomarker for treatment response, a modulator of resistance, and a target for personalized intervention. Emerging technologies, including nanodrug delivery systems and engineered bacteria developed through synthetic biology, present promising tools for the precise modulation of the microbiome. Nonetheless, substantial challenges persist, including the spatiotemporal dynamics of host-microbiota interactions, significant interindividual variability, and concerns regarding long-term safety. The integration of personalized microbiome-based strategies into precision cancer therapy constitutes a critical challenge in the field of translational medicine.

Gut microbiota modulates chronic inflammation and remodels the immune microenvironment

Through the release of metabolites and direct translocation, the GM can continuously drive both local and systemic chronic inflammation. Chronic inflammation promotes tumorigenesis and immune evasion by inducing genomic instability, activating oncogenic pathways, and creating an immunosuppressive microenvironment [18, 19]. Healthy and dysbiotic GM exhibit contrasting roles in shaping the tumor immune microenvironment (Fig. 1).

Fig. 1.

Differential effects of healthy and dysbiotic gut microbiota on tumorigenesis

Promotion of chronic inflammation

The GM is essential for maintaining intestinal homeostasis and plays a significant role in the pathogenesis of inflammatory bowel disease and CRC [20, 21]. Inflammatory responses impede the restoration of the intestinal barrier by enhancing permeability and promoting the expression of pro-inflammatory factors. For instance, lipopolysaccharide (LPS), a constituent of the gram-negative bacterial cell wall, activates TLR4 on the surface of intestinal epithelial cells, thereby initiating an inflammatory response that leads to increased intestinal permeability [22]. This persistent inflammatory condition establishes a critical foundation for tumor initiation and progression [23]. Furthermore, a chronic inflammatory microenvironment may develop into an immunosuppressive niche. In primary CRC with liver metastasis, F. nucleatum increases the levels of pro-inflammatory cytokines, including TNF-α, IL-6, and IL-17A. This not only elevates the risk of metastatic dissemination but also diminishes the cytotoxic function of CD8⁺ T cells, thereby undermining anti-cancer immunity [24]. Recent research has demonstrated that Bacteroides fragilis toxin (BFT) activates IL-17R, NF-κB, and STAT3 signaling pathways in colon epithelial cells, thereby initiating a pro-inflammatory cascade that promotes cancer progression. This pro-carcinogenic signaling further results in the selective activation of NF-κB in the distal colon, leading to the secretion of various chemokines, such as CXCL1, which subsequently triggers myelocyte-dependent tumorigenesis at distal sites [25]. Furthermore, empirical evidence suggests that the surface protein PCWBR2 of Peptostreptococcus anaerobius activates the PI3K-AKT signaling pathway through its interaction with integrin α2/β1. This interaction results in the activation of NF-κB, which subsequently initiates a pro-inflammatory response. This response significantly increases the population of myeloid-derived suppressor cells (MDSCs), tumor-associated macrophages (TAMs), and tumor-associated neutrophils, thereby facilitating the progression of CRC [26].

Chronic inflammation contributes to tumorigenesis by disrupting electron transport within the respiratory chain and enhancing the production of reactive oxygen species (ROS). During electron transport, electron leakage can interact with oxygen, resulting in the formation of ROS, such as O₂⁻, H₂O₂, and OH⁻. These ROS induce oxidative damage to DNA, lipids, and proteins [27]. Enterococcus faecalis, a colonizer of the gastrointestinal tract, generates extracellular superoxide and hydroxyl radicals, which act as significant sources of oxidative stress in colonic epithelial cells and lead to DNA damage [28]. Furthermore, inflammation leads to the production of respiratory electron acceptors, such as nitrate and tetrathionate [29, 30], which bacteria like E. coli and Salmonella can exploit to gain a growth advantage, thereby intensifying intestinal inflammation.

Modulation of immune cells

Extensive evidence has identified MDSCs as pivotal mediators of resistance to cancer therapy, particularly immunotherapy, underscoring their substantial potential as therapeutic targets [31]. MDSCs inhibit the function of T cells, Natural Killer cells (NK cells), and other immune cells critical for anti-cancer responses through mechanisms such as amino acid depletion and the secretion of immunosuppressive factors, including programmed cell death ligand 1 (PD-L1) [32, 33]. Dietary factors further influence this process. Specifically, high-fat diet modifies the GM, resulting in the release of substantial leucine, which activates the mTORC1 pathway in myeloid progenitors. This activation promotes the differentiation of polymorphonuclear myeloid-derived suppressor cells (PMN-MDSCs) and accelerates tumor progression [34]. Research utilizing murine models has elucidated that GM induces immunosuppression in the liver through the expansion of PMN-MDSCs, thereby promoting hepatocellular carcinogenesis [35]. Furthermore, commensal bacteria that are dependent on TLR5 increase systemic IL-6 levels, which in turn promote the expansion of MDSCs, inhibit anti-cancer immunity, and expedite malignant progression [36].

Macrophages play a crucial role in the initiation and progression of cancer by maintaining inflammatory environments, facilitating the migration of tumor cells, and impairing anti-cancer immune responses [37, 38]. The GM plays a significant role in modulating the functions of TAMs [39]. In CRC, F. nucleatum facilitates the infiltration of macrophages through the activation of CCL20 and induces the polarization of macrophages towards the M2 phenotype, thereby augmenting its metastatic potential [40]. In the context of pancreatic cancer, Bifidobacterium pseudolongum activates macrophages via TLR2/TLR5, resulting in the induction of an immunosuppressive phenotype [41]. Metabolites of tryptophan produced by GM activate the aryl hydrocarbon receptor in TAMs, thereby altering the immunosuppressive microenvironment characteristic of pancreatic ductal adenocarcinoma [42]. In gastric adenocarcinoma, H. pylori circumvents immune surveillance through the IDO1-mediated metabolism of tryptophan, which facilitates the polarization of TAMs and the differentiation of regulatory T cells (Tregs) [43].

NK cells are integral to cancer immune surveillance; however, the presence of gut pathogens can significantly impair their anti-cancer efficacy, thereby facilitating immune evasion [44]. For instance, F. nucleatum impedes NK cell-mediated tumor cytotoxicity through its Fap2 protein, which interacts with the inhibitory receptor TIGIT present on NK cells and various T cells [45]. GM facilitates the progression of pancreatic ductal adenocarcinoma by inhibiting NK cell activity and reducing intratumoral infiltration [46]. Furthermore, the microbiota-derived metabolite polyamine inhibits lymphocyte proliferation and IL-2 production, while also impairing macrophage-mediated phagocytosis, neutrophil motility, and NK cell activity [47, 48].

Microbial metabolites also participate in the regulation of T-cell function and differentiation [49]. There is substantial evidence indicating that GM modulates the abundance and function of T cells [50, 51]. In CRC, deoxycholic acid (DCA) impairs CD8⁺ T cell activity by targeting the plasma membrane Ca²⁺ ATPase and inhibiting the Ca²⁺-NFAT2 signaling pathway [52]. Succinate produced by F. nucleatum inhibits the cGAS–interferon-β pathway, thereby restricting the infiltration of CD8⁺ T cells into the TME and consequently diminishing their anti-cancer efficacy [53]. Tregs generally sustain immune homeostasis within the gut; however, colonization by enterotoxigenic Bacteroides fragilis (ETBF) facilitates the accumulation of Tregs, thereby inducing IL-17-mediated pro-carcinogenic inflammation [54]. The hydroxylated derivative of DCA, 3β-hydroxydeoxycholic acid, diminishes the immunogenicity of dendritic cells (DCs) by upregulating the expression of Foxp3, a critical transcription factor for Tregs, thereby facilitating tumor immune evasion [55]. The intricate interplay between immune cells, facilitated by microbial metabolites, further refines tumor immunity. For example, Th17 cells secrete inflammatory cytokines, such as IL-17, which stimulate Paneth cells to release antimicrobial peptides and attract circulating neutrophils [56]. Thus, elucidating the interaction networks between the GM and the TME not only uncovers novel mechanisms of immune evasion but also proposes promising therapeutic strategies for cancer treatment.

Remodeling of the tumor microenvironment

The GM represents a fundamental component of the CRC microenvironment, influencing it through various mechanisms, including the induction of inflammation, epigenetic remodeling, and the regulation of immune cells [44, 57, 58]. In the initial stages of gastric adenocarcinoma, H. pylori DNA, LPS, and other bioactive molecules activate pathogen-associated molecular patterns that facilitate the recruitment of immunosuppressive cells while activating TNF-α, NF-κB, and IL-6-JAK-STAT3 signaling pathways, thereby reshaping the immune microenvironment of gastric adenocarcinoma [43]. In CRC, F. nucleatum selectively recruits various immunosuppressive myeloid-derived cells, particularly MDSCs [59, 60]. Considering that DCs and tumor-associated neutrophils facilitate tumor progression and angiogenesis [61], it is plausible that F. nucleatum promotes tumor progression by inhibiting anti-cancer immunity through the accumulation of MDSCs [24]. The microbial metabolite 4-hydroxyphenylacetic acid further contributes to immunosuppression by facilitating the infiltration of PMN-MDSCs into the TME. Mechanistically, 4-hydroxyphenylacetic acid activates the JAK2/STAT3 signaling pathway, upregulates the transcription of CXCL3, and recruits PMN-MDSCs to CRC. This process remodels the TME and inhibits the tumoricidal capacity of CD8⁺ T-cells [62].

In a similar manner, lactate metabolism is intricately linked to tumor growth, TME remodeling, and resistance to chemotherapy [63, 64]. The GM plays a significant role in lactate production through their metabolic processes, with the accumulation of colonic lactate serving as an indicator of microbial dysbiosis [65]. The extracellular accumulation of lactate by tumor cells, along with the acidification of the TME, constitutes a fundamental mechanism that facilitates tumor progression [66]. Lactate attracts myeloid cells and Tregs while inducing protein lactylation. These processes establish a self-reinforcing positive feedback loop that accelerates malignant progression [67]. Other GM metabolites also contribute to TME remodeling. Lipoteichoic acid, a constituent of Gram-positive bacteria, acts synergistically with obesity-induced DCA to augment the expression of senescence-associated secretory phenotype (SASP) factors and COX-2 in hepatic stellate cells. The COX-2-mediated synthesis of prostaglandin E₂ subsequently suppresses anti-cancer immunity through signaling via PTGER4 [68]. Secondary bile acids exhibit immunosuppressive properties. They facilitate tumor immune evasion by diminishing the hepatic accumulation of natural killer T cells and promoting the expansion of Tregs [55, 69].

Gut microbiota metabolites drive the progression from tumorigenesis to therapy resistance

Metabolites produced by GM play a significant role in the initiation and progression of cancer through various mechanisms, including metabolic reprogramming, the induction of genomic instability, and the activation of oncogenic signaling pathways (Fig. 2). These metabolites impact oncogenesis through direct cellular interactions and by orchestrating complex processes, such as the modulation of signal transduction, remodeling of the immune microenvironment, and metabolic reprogramming [57, 70]. Individual metabolites are known to exhibit pleiotropic effects. A pertinent example is the colonization by Helicobacter hepaticus, which induces T follicular helper cells. These cells facilitate the maturation of tertiary lymphoid structures and the release of cytotoxic molecules, thereby establishing a tumor-suppressive microenvironment that inhibits the progression of CRC in this specific context [71]. Table 2 provides a summary of the various mechanisms and functional roles of GM metabolites in the progression of cancer.

Fig. 2.

Multifaceted mechanisms of gut microbiota metabolites in driving cancer pathogenesis and therapy resistance

Table 2.

The role and implications of gut microbial metabolites in cancer progression

Cancer Type	Microorganisms	Metabolites	Relevant Findings	Mechanism	References(PMID)
CRC	Enterotoxigenic Bacteroides fragilis	B. fragilis toxin	ETBF infection significantly increases colonic tumor burden in mouse models	ETBF upregulates spermine oxidase in intestinal epithelial cells, promoting DNA damage	21876161
	Clostridia groups	N1, N12-diacetylspermine	N1, N12-diacetylspermine is significantly upregulated in tumor tissues, promoting biofilm formation	Promotes cancer cell proliferation and enhances biofilm stability	25959674
	CRC-enriched microbiota	Isovalerate	Isovalerate drives self-renewal and tumorigenesis of colorectal cancer stem cells	Increased 5-HT synthesis activates the Wnt/β-catenin pathway through HTR1B/1D/1F receptors	35550066
	Desulfovibrio fairfieldensis, Rhodococcus erythropolis, Brucella abortus, Chlamydia muridarum	Secondary bile acids	Squalene epoxidase overexpression induces gut microbiota dysbiosis and promotes inflammation	Activates cholesterol synthesis pathway and inhibits apoptosis proteins	35232776
	Enterococcus, Streptococcus, Lachnoclostridium	DCA	DCA promotes vasculogenic mimicry formation and epithelial-mesenchymal transition in CRC	Activates VEGFR2, leading to upregulation of ZEB1/2 and induction of EMT	34811848
	Lactobacillus plantarum, Bacillus subtilis	Gallic acid	Gallic acid accumulates in the distal gut and promotes tumorigenesis	Restores TCF4 binding to chromatin, reactivates WNT signaling, and promotes β-catenin accumulation	32728212
	Sulfate-Reducing Bacteria	H₂S	Elevated luminal H₂S concentrations are associated with increased CRC risk	Induces DNA damage and genomic instability via free radical-mediated oxidative stress	17475672 34745023
HCC	Gram-positive bacteria	DCA, LTA	DCA and LTA synergistically suppress antitumor immunity and accelerate HCC progression	Translocation to liver activates NF-κB, upregulates COX-2, and impairs CD8⁺ T cell function	28202625
Primary liver tumors and liver metastasis	Clostridium scindens	Secondary bile acids	Clostridium scindens promotes tumor progression by converting primary bile acids to secondary bile acids	Downregulates CXCL16 in liver sinusoidal endothelial cells, reducing NKT cell recruitment	29798856
Breast cancer and Melanoma	Desulfovibrio spp.	Leucine	Elevated leucine induces immunosuppression and accelerates tumor growth	Activates mTORC1 in myeloid progenitor cells, promoting PMN-MDSC differentiation	38709933
Breast cancer	Firmicutes, Bacteroidetes	Estrogen	β-Glucuronidase hydrolyzes estrogen glucuronides, releasing active estrogens	Microbial β-glucuronidase mediates estrogen activation, promoting breast cancer progression	34458248
Bladder cancer	Escherichia coli	BBN	BBN is converted to BCPN, which accumulates in the bladder and drives tumorigenesis	Bacterial oxidation of BBN to BCPN induces DNA adduct formation in urothelium	39085612

CRC colorectal cancer, ETBF enterotoxigenic Bacteroides fragilis, HCC hepatocellular carcinoma, DCA deoxycholic acid, LTA lipoteichoic acid, PMN-MDSC polymorphonuclear myeloid-derived suppressor cell, BBN N-butyl-N-(4-hydroxybutyl)-nitrosamine, BCPN N-butyl-N-(3-carboxypropyl)-nitrosamine

It is crucial to emphasize that the impact of GM metabolites on carcinogenesis is not unidirectional. A delicate balance exists between metabolites with tumor-promoting properties. Disruption of this equilibrium towards a dominance of pro-carcinogenic metabolites, often termed “dysbiotic metabolism”, is a hallmark of cancer-favoring microenvironments. It is important to note that GM metabolites can be involved in the process from carcinogenesis to therapy resistance.

Compromising genomic stability

Metabolites produced by the GM have the potential to induce genomic instability in the host, which constitutes a fundamental mechanism in the initiation of cancer [72]. Recent research has underscored a significant correlation between specific gut microbial metabolites and the development of CRC. For example, the prevalence of Bilophila wadsworthia, a microorganism that synthesizes hydrogen sulfide (H₂S) through dissimilatory sulfate reduction, shows a positive correlation with levels of dietary fat and bile acids [73]. This genotoxic gas causes DNA damage in colonic epithelial cells, thereby elevating the risk of CRC [74]. H₂S plays a regulatory role in essential cellular processes, including autophagy, ferroptosis, and apoptosis [75, 76]. Although there is no direct evidence, one study confirmed that H₂S can reduce the oxidative stress caused by antibiotics by neutralizing reactive oxygen species, thereby protecting cells [77]. However, the mechanism by which anthracyclines inhibit tumors is the production of large amounts of ROS, which cause DNA damage. The findings underscore the potential of targeting H₂S production as a strategy for the prevention of CRC.

Polyamines, as critical signaling molecules, play a regulatory role in fundamental cellular processes, including proliferation, autophagy, and DNA maintenance [78]. The GM facilitates tumorigenesis and influences cancer stemness primarily through complex interactions with polyamine metabolic pathways [79, 80]. The stemness of tumors is one of the key factors for therapy resistance. In strains of E. coli that contain the pks island responsible for the production of the genotoxin colibactin, the polyamine spermidine is crucial for achieving their complete genotoxic potential [81]. GM harboring the pks genomic island, which enables production of colibactin, can induce DNA damage in colorectal cancer cells. Prolonged exposure to colibactin selects for a subpopulation of cancer cells with restored homologous recombination function, leading to cross-resistance to irinotecan and its active metabolite SN38 [82]. This reveals the critical role of microbiota metabolite-driven DNA repair remodeling in mediating chemotherapy resistance. Spermidine facilitates colorectal tumorigenesis by inducing colibactin-mediated DNA damage. Additionally, polyamines serve as immunomodulatory agents that suppress immune cell activity, thereby diminishing anti-cancer immunity [48]. Certain metabolic byproducts of polyamines exhibit direct carcinogenic properties. SMOX facilitates the oxidation of spermine to spermidine, resulting in the production of hydrogen peroxide. This reaction generates ROS, which are implicated in DNA damage and chronic inflammation [83]. Significantly, the virulence factor CagA of H. pylori induces the upregulation of SMOX expression in gastric epithelial cells, which is associated with an elevated risk of gastric cancer [84]. Beyond intestinal malignancies, the metabolism of GM plays a role in carcinogenesis in other organs. For example, nitrosamine compounds derived from tobacco smoke, such as N-butyl-N-(4-hydroxybutyl)-nitrosamine, can be metabolized in the gut into compounds like N-butyl-N-(3-carboxypropyl)-nitrosamine, which induces DNA damage in the bladder. This finding suggests a potential mechanism by which GM may promote bladder carcinogenesis [85]. In metabolomic analyses of antibiotic-treated mice, decreased levels of indole-3-propionic acid and indoxyl sulfate were observed. Concurrently, tumor tissues exhibited suppressed expression of 53BP1 along with increased expression of BRCA1, indicating enhanced DNA damage repair. This altered repair response may represent a key mechanism underlying cisplatin resistance [86].

Regulating oncogenic signaling pathways

The GM can contribute to tumorigenesis and therapy resistance through the secretion of specific toxins that activate key oncogenic signaling pathways. For instance, the BFT produced by ETBF directly targets colonic epithelial cells, activating the pro-inflammatory NF-κB and STAT3 signaling pathways. This activation results in the robust expression of CXC chemokines, which initiates a persistent inflammatory response and cultivates a tumor-promoting microenvironment, ultimately facilitating the progression of CRC [25]. Moreover, in addition to direct toxin secretion, another study has demonstrated that a hot and humid environment can decrease the abundance of probiotics, such as Lactobacillus murinus, resulting in the accumulation of lithocholic acid. This bile acid subsequently activates the PI3K/AKT/NF-κB signaling pathway [87]. GM dysbiosis can also contribute to tumor progression in organs beyond the digestive tract. For instance, the enrichment of Bacteroides and Prevotella species in patients with ovarian cancer induces systemic inflammation by activating the TLR4/MYD88/NF-κB pathway. This activation subsequently upregulates Hedgehog signaling, thereby promoting tumor proliferation and metastasis [88]. However, the activation of NF-κB leads to the upregulation of CCL20, which recruits Tregs and ultimately results in resistance to 5-fluorouracil (5-FU) in CRC [89]. Similarly, activation of STAT3 results in the release of CCL2, which recruits TAMs and suppresses the activity of cytotoxic T lymphocytes, thereby contributing to the development of resistance to immune checkpoint blockade (ICB) therapy [90]. Furthermore, the toxin TcdB, produced by Clostridium difficile, which is prevalent in mucosal samples from CRC patients, aberrantly activates the Wnt/β-catenin signaling pathway. This pathway is a critical driver of a tumorigenic phenotype in mouse models [91]. In a similar manner, the FadA adhesin from F. nucleatum and the effector protein AvrA from Salmonella have been demonstrated to dysregulate β-catenin signaling, thereby contributing to tumorigenesis [92, 93]. Parida et al. have demonstrated that ETBF can induce hyperplasia in mammary epithelial cells and significantly promote tumor growth and metastasis. Breast cancer cells exposed to the toxin develop a "BFT memory," mediated by the β-catenin and NOTCH1 signaling pathways, which persistently enhances tumor growth and metastasis long after the initial exposure [94]. In breast cancer, BFT-1 produced by ETBF interacts with the receptor NOD1 on breast cancer stem cells. This interaction recruits G-associated kinase, which phosphorylates and facilitates the degradation of the tumor suppressor NUMB. Consequently, this process diminishes NUMB's inhibition of the NOTCH1-HEY1 pathway, thereby enhancing stemness and chemoresistance in breast cancer stem cells [95]. The activation of Notch1 pathway can also regulate the stemness of a variety of tumors and lead to therapy resistance [96, 97]. The gut microbiota-derived metabolite urolithin A acts as an inhibitor of tumor cell proliferation by targeting the PI3K/AKT/mTOR signaling pathway [98]. However, mTOR activation confers resistance to a broad spectrum of targeted therapies, including PI3K, EGFR, and Hsp90 inhibitors [99].

Elucidating the mechanisms by which specific gut microbial metabolites modulate oncogenic signaling pathways offers novel insights into the etiology of cancer and establishes a theoretical basis for the development of innovative therapeutic strategies against cancer.

Driving metabolic reprogramming

Metabolic reprogramming are essential processes that facilitate tumor progression. A significant factor in this reprogramming is the dynamic interaction between intratumoral microbiota and the TME [100]. Trimethylamine N-oxide (TMAO), a well-established deleterious metabolite resulting from gut microbial metabolism, demonstrates circulating levels that are significantly associated with an elevated risk of colon cancer [101]. The mechanisms through which TMAO facilitates tumorigenesis are complex and multifaceted. It induces proangiogenic metabolic reprogramming by upregulating VEGF-A, thereby promoting the proliferation of CRC cells. Additionally, as a potent inducer of inflammation and oxidative stress, TMAO contributes to the initiation and progression of CRC by inducing chronic inflammation, oxidative damage, and genomic instability [102]. Interestingly, one study revealed that TMAO reprograms macrophages in a type I interferon‑dependent manner, thereby converting the immunosuppressive tumor microenvironment of pancreatic cancer into an immunologically active state, which enhances T‑cell function and improves the efficacy of ICB [103].

Dysregulated lipid metabolism is a characteristic feature of CRC. In colon cancer, the upregulation of squalene epoxidase, a rate-limiting enzyme in cholesterol biosynthesis, contributes to GM dysbiosis and facilitates tumor cell proliferation by inhibiting apoptosis [104]. The GM facilitates oncogenesis by influencing the host's lipid metabolism. Specifically, in cases of treatment-resistant prostate cancer, certain bacteria, such as Ruminococcus, augment phospholipid metabolism by increasing the availability of substrates for LPCAT1, resulting in its overexpression [105]. LPCAT1, a critical enzyme involved in lipid remodeling, when overexpressed, modifies the composition of cellular membranes, which may facilitate the activation of oncogenic signaling pathways [106]. In studies of therapeutic resistance, LPCAT1 contributes to drug resistance, including to tyrosine kinase inhibitors such as gefitinib, by regulating phospholipid metabolism and modulating the EGFR/PI3K/AKT pathway [107]. LPCAT1 drives therapeutic resistance in advanced clear cell renal cell carcinoma by upregulating IL-1β expression through AKT pathway activation while concurrently promoting inflammasome-mediated maturation of IL-1β. The resulting pro-inflammatory and immunosuppressive tumor microenvironment diminishes the efficacy of both tyrosine kinase inhibitors and immunotherapy [108].

SCFAs, with butyrate being particularly notable, represent some of the most advantageous metabolites synthesized by the GM [109]. Patients with CRC consistently demonstrate systemic reductions in butyrate levels compared to healthy individuals [110], a phenomenon similarly observed in breast cancer cohorts [111]. SCFAs are readily absorbed by intestinal epithelial cells and interact with G protein-coupled receptors on immune cells, thereby influencing immunoregulation and modulating inflammatory cytokines. Butyrate, in particular, accumulates in the nucleus and primarily functions as a histone deacetylase inhibitor, thereby suppressing cancer cell proliferation by inducing cell cycle arrest and apoptosis [112]. SCFAs can enhance the sensitivity of chemotherapy, including cisplatin and doxorubicin, by inhibiting glycolysis, promoting fatty acid oxidation, optimizing the metabolic adaptability of T cells, and increasing anti-cancer activity [113]. In contrast to these anti-cancer effects, other microbial metabolites facilitate oncogenesis. Isovalerate, a short-chain fatty acid derivative synthesized by Clostridium sporogenes, augments Tph2 gene expression and the biosynthesis of 5-hydroxytryptamine by inhibiting the accumulation of the NuRD complex on the Tph2 promoter [114]. Another study confirmed that Wnt/β-catenin signaling activates HIF-1α induced metabolic reprogramming, leading to the development of 5-FU resistance in CRC [115]. In studies concerning breast and lung cancer, methylmalonic acid, a byproduct of abnormal propionate metabolism, has been observed to accumulate, thereby promoting the migration and invasion of cancer cells [116]. These findings collectively elucidate a fundamental paradigm: the GM produces a diverse array of bioactive small molecules that drive tumor metabolic reprogramming through complex metabolic activities. This reprogramming endows tumor cells with enhanced proliferative capacity, stemness, invasiveness, and therapeutic resistance.

Mediating epigenetic modification

The GM serves as a crucial regulator of host epigenomes, exerting its influence through a variety of post-translational modifications that impact tumor initiation, progression, and therapeutic response. Investigations into the human gut microbiome have identified a wide array of protein modifications, including lysine acetylation, propionylation, and succinylation [117]. Studies have shown that infections with bacteria such as F. nucleatum and Hungatella hathewayi can induce promoter hypermethylation of tumor suppressor genes in colon epithelial cells, forming a basis for therapy resistance [118]. For example, microbes like F. nucleatum can induce hypermethylation of the MLH1 promoter, thereby promoting CRC progression [119]. Research utilizing murine models has revealed that bile acid derivatives synthesized by the GM facilitate the enrichment of histone H3K4 monomethylation at specific gene promoters. This epigenetic modification impairs the function of CD8⁺ T cells, thereby compromising anti-cancer immunity [120]. Indole-3-propionic acid, co-produced by Lactobacillus johnsonii and Clostridium sporogenes, enhances the anti-cancer activity of CD8⁺ T cells. This is achieved through the specific induction of H3K27 acetylation at the super-enhancer region of the key transcription factor gene Tcf7 [51]. A decrease in H3K27 acetylation is one of the mechanisms contributing to therapeutic resistance to the DNA methylation inhibitor 5-azacitidine in BRAF‑mutant CRC [121]. This underscores a beneficial cooperative mechanism in which commensal bacteria collaborate to generate advantageous metabolites that enhance immune surveillance through epigenetic pathways.

Elevated intestinal lactate concentrations enhance host lysine lactylation, a lactate-dependent epigenetic modification that plays a critical role in conferring resistance to tumor therapy [122]. In the context of glioblastoma, histone H3K9 lactylation facilitates the transcriptional activation of the LUC7L2 gene, resulting in the suppression of mismatch repair mechanisms and consequently imparting resistance to temozolomide [123]. In CRC, histone H3K18 lactylation enhances the expression of the autophagy enhancer protein RUBCNL, thereby promoting protective autophagy and resulting in resistance to bevacizumab treatment [124]. In addition to facilitating epigenetic modifications, microbial products can modulate gene expression by influencing key signaling pathways. For instance, microbially derived peptidoglycan fragments play a critical role in maintaining gut homeostasis. Specifically, N-acetylmuramic acid inhibits AKT1 activation and suppresses tumorigenesis by directly binding to and blocking the phosphorylation of AKT1 [125]. While this mechanism does not constitute a traditional epigenetic modification, it illustrates how microbial small molecules can directly influence the activity of key host signaling proteins in the gut.

Role of the gut microbiome in anti-cancer therapy resistance

Due to the close anatomical and physiological relationship between the GM and the digestive tract, its impact is most immediate and significant in malignancies originating within the gastrointestinal system. The direct exposure of the intestinal mucosa to microbial products and metabolites allows for well-defined mechanistic studies, resulting in a predominant research focus on the role of the GM in therapy resistance for CRC. Conversely, evidence elucidating its impact on therapy resistance in cancers outside the digestive system is less established. Current understanding in these areas often relies on preclinical models and foundational studies, underscoring the need for more extensive clinical validation and mechanistic exploration in non-digestive malignancies.

Digestive system malignancies

Patients with CRC demonstrate unique microbial profiles, wherein specific bacterial species function as diagnostic biomarkers and active modulators of immune and therapeutic responses [100, 126]. Consequently, the GM exerts a direct or indirect influence on cancer treatment outcomes through immunoregulatory mechanisms. F. nucleatum is among the most significantly enriched bacteria in patients with recurrent CRC [127]. Mechanistically, it induces autophagy through the TLR4/MYD88 signaling pathway, thereby conferring resistance to chemotherapy. In addition to promoting autophagy, F. nucleatum upregulates inhibitors of apoptosis proteins (IAPs), which directly suppress the caspase cascade, thereby inhibiting apoptosis and diminishing the sensitivity of CRC cells to 5-FU [128]. Furthermore, this bacterium activates the E-cadherin/β-catenin/TCF4 signaling pathway to enhance the expression of GPX4, thereby inhibiting oxaliplatin-induced ferroptosis and conferring chemoresistance in CRC [129]. Various species of GM contribute to chemotherapy resistance through distinct mechanisms. For example, Bacteroides vulgatus may provide nucleotides necessary for DNA repair, thereby safeguarding tumor cells from DNA-damaging agents and enhancing resistance to neoadjuvant chemoradiotherapy in cases of locally advanced rectal cancer [130]. In a similar manner, colonization by Bacteroides fragilis activates the NOTCH1 signaling pathway in colon cancer cells, thereby inducing EMT and stemness, which subsequently inhibits chemotherapy-induced apoptosis [131]. Citrobacter freundii contributes to increased resistance to oxaliplatin in CRC by downregulating NINJ2 [132].

Metabolic reprogramming facilitated by the GM constitutes a significant mechanism contributing to therapeutic resistance. Specifically, lipid metabolic reprogramming offers cancer cells alternative energy sources, thereby enhancing their survival during chemotherapy [133]. Elevated levels of glycerophospholipids are significantly associated with the development of an immunosuppressive microenvironment that promotes tumor progression and chemoresistance. Notably, colibactin-producing E. coli (CoPEC) colonizing tumor cells confer resistance to multiple chemotherapeutic agents, including oxaliplatin, a phenomenon closely linked to lipid metabolic reprogramming [134]. A study has demonstrated that the abnormal accumulation of lipid droplets in colon cancer cells with mutations in the KRAS and the BRAF is significantly associated with a poor response to erlotinib [135].

The GM profile is significantly correlated with resistance to immunotherapy [136]. Succinate derived from F. nucleatum inhibits the cGAS–STING–IFN-β signaling pathway, thereby impairing the migration of CD8⁺ T cells into the TME [53, 137]. An animal studies have corroborated that E. coli reduces the infiltration of CD3⁺ and CD8⁺ T cells in murine CRC and diminishes the efficacy of anti- programmed cell death protein 1 (PD-1) therapy [138]. Cheng et al. have identified arginine as a critical differential metabolite in gastrointestinal cancers associated with primary resistance. This essential amino acid modulates T cell metabolism through various transcriptional regulators, ultimately ultimately increasing treatment sensitivity [139]. In contrast, multi-omics analyses have revealed that asymmetric dimethylarginine (ADMA), a metabolite of arginine, exhibits a positive correlation with Tregs. ADMA hinders macrophage proliferation and phagocytosis, thereby attenuating immune responses [140]. Given the critical role of the GM in anabolic processes, targeting microbial communities may constitute a novel approach to overcoming therapy resistance induced by metabolic reprogramming in CRC. Integrated multi-omics analyses hold promise for elucidating resistance mechanisms in CRC.

The bidirectional interaction between the GM and the pancreas has gained increasing recognition in recent research. GM contributes to chemotherapy resistance in pancreatic cancer through various mechanisms. Due to their anatomical proximity, GM can translocate via the duodenum to colonize the pancreas, a process facilitated by inflammatory intestinal environments [141]. Intratumoural Gammaproteobacteria have the capacity to inactivate the chemotherapeutic agent gemcitabine by metabolizing it into 2′,2′-difluorodeoxycytidine, thereby contributing to chemoresistance in pancreatic ductal adenocarcinoma [142]. F. nucleatum, which is prevalent in pancreatic ductal adenocarcinoma [143], possesses the anti-apoptotic gene BIRC3. This gene inhibits chemotherapy-induced cell death in CRC [128]. Nevertheless, this mechanism necessitates validation within the context of pancreatic cancer. In oesophageal squamous cell carcinoma, F. nucleatum enhances the expression of NOD-, LRR-, and NLRP3 and increases the presence of MDSCs, thereby contributing to chemoresistance. [144]. Moreover, the colonization of senescent oesophageal squamous cell carcinoma cells by F. nucleatum exacerbates DNA damage, thereby augmenting the SASP and inducing chemoresistance in tumors [15].

Non-digestive system malignancies

When considering malignancies beyond the digestive system, research on GM-mediated therapy resistance has predominantly focused on immunotherapy. Studies have primarily investigated how the GM influences intratumoral immunity, particularly its role in creating immunosuppressive microenvironments that compromise the efficacy of immune checkpoint inhibitors. Although this focused investigation has enhanced our understanding of immune-related resistance mechanisms, the potential contribution of the GM to resistance against other major therapies, such as chemotherapy, targeted therapy, or radiotherapy, remains relatively underexplored in these malignancies.

Notably, the microbiome has significant immunomodulatory effects in the treatment of melanoma. The GM can mitigate microbiome-dependent resistance to PD-1 by downregulating interactions between PD-L2 and the RGMb [145]. In contrast, the administration of Bifidobacterium and Lactobacillus rhamnosus probiotics adversely affects the efficacy of anti-PD-1 therapy, as evidenced by a reduction in interferon-γ-producing cytotoxic T cells within the TME [146]. Moreover, the gut microbial metabolite butyrate reduces the effectiveness of anti-cytotoxic T-lymphocyte-associated protein 4 (CTLA-4) therapy. Mechanistically, butyrate inhibits DC maturation and CD28 signaling, which are typically enhanced by anti-CTLA-4 treatment, while also hindering the upregulation of the inducible T-cell costimulatory molecule on T cells. Collectively, these effects lead to a reduction in the accumulation of tumor-specific and memory T cells, thereby attenuating anti-cancer activity [147]. In patients with non-small cell lung cancer (NSCLC), H. pylori suppresses both innate and adaptive immune responses by altering dendritic cell cross-presentation activity, thereby impairing anti-cancer CD8⁺ T cell responses and diminishing the efficacy of anti-PD-1 therapy [148]. Recent research has demonstrated that individuals exhibiting a sustained response to immune checkpoint inhibitor (ICI) therapy are characterized by a gut microbiome with a heightened capacity for butyrate production, which facilitates a favorable therapeutic outcome [149]. Similarly, taurolithocholic acid has been shown to enhance PD-1 sensitivity and promote the proliferation of CD4⁺/CD8⁺ T cells [150].

Drug resistance constitutes a significant impediment to the successful treatment of multiple myeloma. Ammonium derived from Citrobacter freundii contributes to this resistance by augmenting the acetylation of NEK2 and inhibiting its ubiquitin-mediated degradation, thereby stabilizing NEK2 levels. This mechanism facilitates the proliferation of myeloma cells, suppresses apoptosis, and directly diminishes sensitivity to bortezomib [151]. Nonetheless, urolithins produced by the GM have exhibited a synergistic therapeutic effect in vitro, even in the presence of bortezomib resistance [152].

Breast cancer represents one of the most prevalent malignancies affecting women globally [153]. The diverse therapeutic approaches to breast cancer have significantly improved its prognosis. Nevertheless, GM can also induce treatment resistance by influencing chemotherapy sensitivity and other mechanisms. A study has demonstrated that species of the Genus Ruminococcus, which become enriched during chemotherapy, may activate resistance pathways dependent on SOX-8 through signals mediated by the gut-brain axis, thereby reducing chemosensitivity [154]. The GM can affect not only the outcomes of chemotherapy but also the effectiveness of endocrine therapy in breast cancer. In ER⁺ breast cancer, the estrogen-like cholesterol metabolite 27-hydroxycholesterol promotes ER-dependent tumor growth and may undermine the efficacy of tamoxifen [155]. Significantly, the GM affects the bioavailability of oral drugs through biotransformation and modified drug transport processes, factors that are frequently neglected in treatment regimens [156]. Temozolomide (TMZ) serves as the primary chemotherapeutic agent for glioma treatment; however, its inconsistent efficacy and high rates of resistance considerably restrict its clinical application [157]. The sensitivity to TMZ is associated with distinct gut microbial compositions: mice that are non-sensitive exhibit a higher abundance of Bacteroides and a lower abundance of Akkermansia, whereas sensitive mice demonstrate an enrichment of Desulfovibrio and Muribaculum. This microbial variation directly affects TMZ chemosensitivity by modulating the infiltration of macrophages and CD8α⁺ T cells, as well as the release of inflammatory factors [158]. Another study demonstrated that TMZ itself influences the GM, leading to an increased abundance of Verrucomicrobia, Anaerotruncus, Akkermansia, Bifidobacterium, Intestinimonas, and Clostridium XVIII. These microbial alterations facilitate anti-cancer responses through metabolic regulation and immune modulation [159], indicating that targeting the microbiota could potentially enhance TMZ sensitivity.

Unraveling the mechanisms of gut microbiota-induced therapy resistance in cancer

The development of resistance to therapy continues to pose a significant challenge in the field of oncology, stemming from a variety of mechanisms. Mutations that confer resistance may occur within the target proteins themselves or in alternative components of the same or parallel signaling pathways, allowing tumors to circumvent the inhibited oncogenic signals. This adaptive plasticity represents a major clinical obstacle [16, 160]. Species within the GM play a crucial role in this intricate landscape, exerting significant influence on cancer progression, immune modulation, and chemosensitivity through various mechanisms [161]. Understanding the role of microbial communities in influencing mechanisms of therapeutic resistance is a crucial area of research for developing innovative translational strategies to address treatment failure.

Gut microbiota-mediated drug inactivation

GM or its metabolites mediated drug inactivation is one of the common mechanisms of therapeutic resistance. Nevertheless, its effectiveness is often undermined by resistance mechanisms mediated by the GM. A prominent mechanism involves the metabolic alteration of chemotherapeutic agents by microbes colonizing the gut and tumors, which can transform active drugs into inactive metabolites. For example, β-glucuronidase produced by the GM hydrolyzes the inactive SN-38G back to the highly cytotoxic SN-38, thereby exacerbating intestinal toxicity. Concurrently, the majority of the regenerated SN-38 is swiftly adsorbed and inactivated by bacterial cells and luminal contents, ultimately diminishing its systemic bioavailability [162]. In a similar manner, Gammaproteobacteria deactivate gemcitabine through metabolic modification, converting it into the inactive compound 2′,2′-difluorodeoxyuridine, which results in chemoresistance in patients with pancreatic ductal adenocarcinoma [142]. Cytidine deaminase produced by Enterococcus faecium also inactivates the drug via this mechanism [163]. Raoultella planticola converts doxorubicin into 7-deoxydoxorubicinol and 7-deoxydoxorubicinolone via reductive deglycosylation. Furthermore, doxorubicin can also be degraded by E. coli BW25113 [164]. Various gram-negative bacteria, including E. coli, produce nicotinamidase, which converts nicotinamide into nicotinic acid. This conversion may activate the Preiss‑Handler pathway in cancer cells, thereby bypassing NAD⁺ synthesis inhibition by NAMPT inhibitors and potentially leading to drug resistance [165].

Gut microbiota in metabolic reprogramming

Emerging evidence underscores the pivotal role of the GM in fostering therapeutic resistance through metabolic reprogramming of tumor cells. This process involves diverse microbial mechanisms that rewire host energy metabolism and biosynthetic pathways, ultimately compromising treatment efficacy. Upregulated polyamine synthesis in drug-resistant cancer cells enhances EIF5A hypusination, which in turn promotes mitochondrial protein translation and oxidative phosphorylation, thereby driving drug resistance. This resistance can be reversed by using the polyamine inhibitor DFMO [166]. Additionally, intratumoural Lactobacillus iners (L. iners) contributes to chemoradiation resistance through lactate-induced metabolic reprogramming. In patients with cervical cancer, colonization by L. iners is associated with poor treatment response and an increased risk of recurrence. As an obligate L-lactate producer, L. iners upregulates glycolytic and tricarboxylic acid cycle activity, resulting in lactate accumulation and the activation of HIF-1α and ROS-related pathways. This metabolic shift modulates DNA damage repair mechanisms, ultimately promoting radiotherapy resistance [167]. In a similar manner, Bacteroides vulgatus facilitates microbial nucleotide synthesis, thereby enhancing the DNA repair capacity of tumors and diminishing the efficacy of neoadjuvant radiotherapy in rectal cancer [130]. Candida tropicalis promotes chemoresistance in colon cancer by modulating the mismatch repair system through increased lactate production [168]. Future research should focus on elucidating the precise molecular crosstalk and developing translational interventions aimed at the tumor-microbe-metabolism axis. Furthermore, enhanced lipid metabolism can modulate therapy resistance through multiple mechanisms [169]. For example, CoPEC can locally establish a glycerophospholipid‑rich microenvironment, thereby promoting tumor cell metabolism and contributing to therapy resistance [134].

Gut microbiota modulation of tumor stemness and cell death

Alterations in GM, particularly dysbiosis and the presence of specific pathogens such as Bacteroides fragilis, can impede apoptosis and enhance tumor stemness and chemoresistance. This occurs through mechanisms such as disruption of the intestinal barrier, activation of pro-inflammatory signaling pathways, or direct modulation of tumor cell pathways. Notably, dysbiosis leads to a reduction in SCFA production, which compromises the integrity of the intestinal barrier and facilitates the entry of LPS into systemic circulation. This process subsequently activates the NF-κB pathway and augments IL-6/STAT3 signaling, thereby promoting tumor proliferation and inducing resistance to docetaxel [170, 171]. Bacteroides fragilis colonizes CRC cells by binding its surface protein, SusD/RagB, to the NOTCH1 receptor on these cells. This interaction activates NOTCH1 signaling, induces epithelial–mesenchymal transition and tumor stemness, and consequently inhibits apoptosis induced by chemotherapy [131]. CoPEC induce cellular senescence in CRC. This senescent phenotype subsequently promotes EMT and the emergence of cancer stem cells, thereby enhancing resistance to chemotherapeutic agents, such as 5-FU, irinotecan, and oxaliplatin [172].

In addition to maintaining tumor stemness, genetic modifications further modulate cell death pathways, thereby contributing to chemotherapy resistance. F. nucleatum imparts resistance to apoptosis by modulating several signaling pathways, including TLR4/MYD88, inhibitor of IAPs, and E-cadherin/β-catenin/TCF4. This modulation diminishes the sensitivity of CRC cells to chemotherapeutic agents, such as 5-FU and oxaliplatin [127]. Mechanistic investigations have demonstrated that F. nucleatum activates specific microRNAs through the TLR4/MYD88 signaling pathway, thereby inducing protective autophagy and enhancing tumor cell survival during chemotherapy [127]. A similar study has demonstrated that F. nucleatum upregulates inhibitor of IAPs to inhibit caspase-dependent apoptosis, thereby reducing the sensitivity of CRC cells to 5-FU [128]. Furthermore, the adhesin FadA of F. nucleatum binds to E-cadherin, thereby activating the β-catenin/TCF4 signaling pathway. This activation enhances the binding of TCF4 to the GPX4 promoter, resulting in upregulation of GPX4 expression. Increased levels of GPX4 inhibit oxaliplatin-induced ferroptosis by mitigating lipid peroxidation and iron accumulation [129]. These findings collectively demonstrate that F. nucleatum facilitates chemoresistance in CRC by orchestrating the inhibition of distinct cell death pathways.

Gut microbiota-mediated remodeling of the tumor microenvironment

GM and its metabolites drive therapy resistance through the recruitment and reprogramming of immune cells, culminating in a remodeled, pro-tumor immune microenvironment. Polyamines, via TDG-mediated DNA demethylation of the PPARG promoter as an epigenetic alteration, upregulate PPARG expression, thereby driving macrophage polarisation toward an M2-like pro-tumour phenotype and subsequently suppressing CD8⁺ T cell function [173].

The GM plays a crucial role in influencing the effectiveness of immunotherapy by modulating immune cell function and altering the metabolic environment of the TME. ICIs are fundamental to immunotherapy. Monoclonal antibodies targeting PD-1, PD-L1, and CTLA-4 have shown substantial efficacy against various advanced or metastatic cancers, including melanoma and NSCLC [174]. Conversely, the microbiota has the capacity to alter the TME through metabolic reprogramming. The excessive proliferation of Clostridium species disrupts bile acid metabolism, downregulates MAdCAM-1, and facilitates the migration of RORγt⁺ T-reg 17 cells from the gut to the tumor, where they contribute to the establishment of an immunosuppressive niche [175]. In hepatocellular carcinoma, the accumulation of the microbial metabolite N1-acetylspermidine activates SRC signaling, which polarizes CCL1⁺ macrophages and recruits CCR8⁺ Tregs, thereby establishing an immunotherapy-resistant microenvironment [176]. Moreover, the dual role of arginine metabolism highlights the essential nature of metabolic reprogramming. Arginine enhances T cell function through the BAZ1B/PC4 and PSIP1/TSN axis, while its metabolite ADMA facilitates Treg proliferation and inhibits macrophage activity, collectively influencing sensitivity to immunotherapy [139, 140].

Gut microbiota-induced immunosuppression

The GM is instrumental in modulating anti-cancer immunity by influencing immune tolerance, pathogen clearance, and tumor immunogenicity, and it has been significantly associated with the outcomes of ICI treatment [177]. For example, certain members of the Ruminococcaceae family, such as Faecalibacterium prausnitzii, have been consistently linked to enhanced responses to ICIs across various studies [178–180]. On the one hand, the microbiota can undermine the efficacy of immunotherapy through direct immune suppression. Specifically, H. pylori impairs the cross-presentation capability of DCs, thereby reducing the activation of anti-cancer CD8⁺ T cells and diminishing the response to PD-1 inhibitors in patients with NSCLC [148]. Elevated systemic levels of butyrate can similarly inhibit the maturation of DCs induced by the CTLA-4 antibody ipilimumab, as well as the CD28/inducible T-cell costimulator signaling pathway, thereby hindering the formation of T cell memory [147]. Moreover, infection with E. coli can lead to a reduction in the infiltration of CD3⁺ and CD8 + T cells in murine CRC models, thereby diminishing the effectiveness of PD-1 blockade therapy [138]. In head and neck squamous cell carcinoma, F. nucleatum secretes outer membrane vesicles containing tryptophanase, which induces indole production in TAMs. This process activates the TDO2/AHR pathway, leading to the upregulation of immunosuppressive cytokines and checkpoint molecules, suppression of cytotoxic T lymphocytes, and ultimately contributes to resistance to immunotherapy [181].

Radiotherapy is a well-established and efficacious modality for cancer treatment. Recent evidence indicates a potential association between the GM and the efficacy of radiotherapy, including its capacity to modulate anti-cancer immune responses at sites distal to the intestine following treatment [182]. Nonetheless, the complex interplay between GM and resistance to radiotherapy remains insufficiently understood. Cui et al. have demonstrated that disruptions in circadian rhythm significantly modify the composition and function of the GM, which may unexpectedly enhance the tumoricidal efficacy of radiotherapy [183]. A study demonstrated that conventional mice with intact microbiota exhibited increased radiosensitivity compared to germ-free mice, an effect potentially mediated by FIAF [184]. Initially explored as an intestinal radioprotector, FIAF appears to fulfill a dual function: its production is enhanced by Bacteroides and Enterococcus faecalis, whereas it is suppressed by E. coli [185].

Therapeutic strategies and future clinical prospects

Numerous reports have accumulated regarding the role of GM in treatment resistance across various cancers. Translating these findings into clinical applications, however, requires further exploration. This necessitates continuous optimization across multiple stages, spanning from the precise identification of microbial constituents to individualized microbiota modulation. Although current evidence predominantly derives from preclinical studies with inherent limitations, understanding the role of GM in treatment resistance, specifically targeting gut microbes and their metabolites to enhance therapeutic sensitivity, and ultimately reversing resistance through GM modulation and remodeling remain promising research strategies (Table 3).

Table 3.

Microbial contributions to cancer therapy resistance and corresponding intervention strategies

Cancer type	Microorganisms and metabolites	Therapy resistance	Mechanism	Microbial therapy strategy	Limitations	References(PMID)
CRC	F. nucleatum	Chemotherapy resistance	It induces autophagy through the TLR4/MYD88 signaling pathway	N/A	Potential subtype-specific limitations; absence of prospective clinical intervention trial data	28753429
			Upregulates inhibitors of apoptosis proteins, which directly suppress the caspase cascade	Clear or reduce F. nucleatum in tumor tissues by using specific antibiotics	Observational design; absence of prospective interventional trial data	30630498
			Activates the E-cadherin/β-catenin/TCF4 signaling pathway to enhance the expression of GPX4, thereby inhibiting oxaliplatin-induced ferroptosis	Clear or reduce F. nucleatum in tumor tissues by using specific antibiotics	Limited clinical sample size	38657754
		Immunotherapy resistance	Succinate derived from F. nucleatum inhibits the cGAS–STING–IFN-β signaling pathway, thereby impairing the migration of CD8⁺ T cells into the TME	Metronidazole reduces F. nucleatum in the gut, lowers serum succinate levels, and restores CD8⁺ T cell infiltration and cytotoxicity	Limited clinical sample size and retrospective single-center design; absence of prospective clinical intervention trial data	37130518
	E. coli	Immunotherapy resistance	E. coli reduces the infiltration of CD3⁺ and CD8⁺ T cells in murine CRC and diminishes the efficacy of anti-PD-1 therapy	N/A	Limited clinical sample size and single-center design; absence of prospective clinical intervention trial data	32037530
	Colibactin-producing E. coli	Chemotherapy resistance	CoPEC locally establishes a high-glycerophospholipid microenvironment with lowered immunogenicity by inducing metabolic reprogramming in cancer cells	Blocking lipid droplet formation using Triacsin C	Observational design; absence of prospective interventional trial data	38417029
	Citrobacter freundii	Chemotherapy resistance	Reduced NINJ2 gene expression suppressed oxaliplatin-induced apoptosis and ROS generation	N/A	Absence of clinical trial data	40567024
	Bacteroides fragilis	Chemotherapy resistance	Binding its surface protein SusD/RagB to the NOTCH1 receptor on these cells induces EMT and tumor stemness, and consequently inhibits apoptosis	Phage VA7 completely reversed the Bacteroides fragilis-mediated chemoresistance	Observational design; absence of prospective interventional trial data	40446807
	Gammaproteobacteria	Chemotherapy resistance	Inactivate the chemotherapeutic agent gemcitabine by metabolizing it into 2′,2′-difluorodeoxycytidine	The combined use of ciprofloxacin, ampicillin, or chloromycin eliminated intratumoral Gammaproteobacteria	Absence of prospective interventional trial data	28912244
Rectal cancer	Bacteroides vulgatus	Neoadjuvant chemoradiotherapy resistance	Facilitates microbial nucleotide synthesis, thereby enhancing the DNA repair capacity of tumors	Treatment with the nucleoside transport inhibitor NBMPR inhibited tumor cell nucleoside uptake and reversed resistance	Absence of prospective interventional trial data	36563680
ESCC	F. nucleatum	Chemotherapy resistance	F. nucleatum induces MDSCs enrichment via activation the NLRP3 inflammosome in ESCC cells, leading to cisplatin resistance	N/A	Observational design; absence of prospective interventional trial data	35435776
			F. nucleatum exacerbates DNA damage, thereby augmenting the SASP and inducing chemoresistance in tumors	N/A	Observational design; absence of prospective interventional trial data	37017266
CRC and GC	Asymmetric dimethylarginine	Immunotherapy resistance	Asymmetric dimethylarginine hinders macrophage proliferation and phagocytosis, thereby attenuating immune responses	N/A	Limited clinical sample size	38194971
BC	Genus Ruminococcus	Chemotherapy resistance	May activate resistance pathways dependent on SOX-8 through signals mediated by the gut-brain axis	N/A	Limited clinical sample size; absence of longitudinal data for monitoring dynamic changes in GM	40460162
ER⁺ BC	27-hydroxycholesterol	Endocrine therapy resistance	Estrogen-like cholesterol metabolite 27-hydroxycholesterol promotes ER-dependent tumor growth and may undermine the efficacy of tamoxifen	Cholesterol-lowering drugs such as statins can improve endocrine therapy outcomes by reducing cholesterol levels	Absence of preclinical data and gut microbiota-related analyses	28380313
Multiple myeloma	Microbial systemic SCFA	Immunotherapy resistance	Butyrate inhibits DC maturation and CD28 signaling and hinders the upregulation of the inducible T-cell costimulatory molecule on T cells	N/A	Limited clinical sample size	32358520
	Citrobacter freundii	Targeted therapy resistance	Ammonium from Citrobacter freundii stabilizes NEK2 via acetylation-mediated inhibition of ubiquitin degradation, thereby promoting myeloma cell proliferation and suppressing apoptosis	Furosemide sodium, as a loop diuretic, inhibits ammonium uptake in multiple myeloma cells by downregulating SLC12A2	Limited clinical sample size and single-center design	38113887
Glioma	GM distribution	Chemotherapy resistance	Modulating the infiltration of macrophages and CD8α⁺ T cells, as well as the release of inflammatory factors	N/A	Absence of in vitro and clinical trial data	36927689
NSCLC	H. pylori	Immunotherapy resistance	Altering dendritic cell cross-presentation activity, thereby impairing antitumor CD8⁺ T cell responses	N/A	Limited clinical sample size and retrospective design	34253574
Head and neck squamous cell carcinoma	F. nucleatum	Immunotherapy resistance	F. nucleatum induces indole production in TAMs and activates the TDO2/AHR receptor pathway, leading to the upregulation of immunosuppressive cytokines and checkpoint molecules	The TDO2 inhibitor enhanced the immunotherapy of PD-1 monoclonal antibody in head and neck squamous cell carcinoma	Observational design; incompletely characterized off-target effects of TDO2 inhibitors	40241230
Cervical cancer	L. iners	Radiotherapy resistance	L. iners upregulates glycolysis and the TCA cycle, leading to lactate accumulation and consequent activation of HIF-1α and ROS signaling, ultimately impairing DNA damage repair mechanisms	The vaginal microbiome can be remodeled to reconstruct a beneficial Lactobacillus crispatus strain using topical metronidazole; LDH inhibitors can block the conversion of lactate to pyruvate	Absence of in vivo trial data; potential inter-institutional bias in the two-center prospective cohort	37863066

CRC colorectal cancer, F. nucleatum Fusobacterium nucleatum, TME tumor microenvironment, E. coli Escherichia coli, PD-1 programmed cell death protein 1, CoPEC colibactin-producing E. coli, ROS reactive oxygen species, MDSCs myeloid-derived suppressor cells, ESCC esophageal squamous cell carcinoma, SASP senescence-aseociated secretory phenotype, GC gastric cancer, BC breast cancer, GM gut microbiota, ER estrogen receptor, DCs dendritic cells, SCFAs short-chain fatty acids, CTLA-4 cytotoxic T-lymphocyte-associated protein 4, NSCLC non-small cell lung cancer, H. pylori Helicobacter pylori, TAMs tumor-associated macrophages, L. iners Lactobacillus iners

Precise identification and analysis of the microbiome

Precise identification of the microbiome represents a crucial prerequisite for optimizing cancer treatment. The key lies in determining the microbial signatures across different patients and treatment regimens. Initial screening of the GM can be performed using 16S rRNA gene amplicon sequencing [186]. Algorithms such as LEfSe can be employed to identify characteristic bacterial taxa associated with specific clinical phenotypes [187]. Furthermore, shotgun metagenomic sequencing can be applied to analyze the genomic composition of the microbiota, thereby providing candidate targets for subsequent research [188]. Finally, PCR can be utilized for rapid validation and precise quantification [189]. For example, by specifically detecting genes such as clbB and clbN within the pathogenic pks island unique to CoPEC [134].

Following the determination of GM composition, RNAscope technology can be employed to visualize low-abundance bacteria through specific signal amplification, and to perform co-localization analysis with host gene expression. This approach directly links microbial distribution to local microenvironmental features [190]. Ultimately, establishing causal links between microbial signatures and tumor phenotypes, followed by functional validation, is essential. Quantitative analysis of microbiota-related metabolites via metabolomics can further elucidate their functional roles, which can be subsequently verified in in vivo models.

Reshaping the gut microbiota to overcome therapeutic resistance

Targeted elimination of antibiotic-resistant bacteria through antibiotics represents a direct and effective strategy, the feasibility of which has been demonstrated in numerous studies. For instance, in a colorectal cancer model, ciprofloxacin inhibited the growth of Mycoplasma hyorhinis and alleviated gemcitabine resistance mediated by this bacterium [142]. Similarly, corroborating previous findings, targeting tumor-promoting microbes has been shown to enhance the response to gemcitabine and extend survival in patients with advanced pancreatic cancer [191]. Specifically, F. nucleatum-derived succinate inhibits the cGAS–interferon-β pathway, thereby limiting the migration of CD8⁺ T cells into the TME. Treatment with metronidazole significantly reduces the intestinal abundance of F. nucleatum, which in turn lowers the serum succinate levels and restores sensitivity to immunotherapy [53]. Furthermore, studies have shown that F. nucleatum can accompany CRC cells to metastatic sites. Treatment of a colon cancer-bearing mouse model with metronidazole reduced the Fusobacterium load, inhibited tumor cell proliferation, and ultimately improved clinical outcomes in CRC patients [192]. In a related strategy, targeting intratumoral oncomicrobes within the immunosuppressive tumor microenvironment using antibiotic-loaded liposomes can induce the release of cancer-specific microbial neoantigens. This treatment promoted the infiltration of cytotoxic CD8⁺ T cells into the tumor. Moreover, the homologous epitopes demonstrated anti-cancer efficacy, potentially enabling the immune system to recognize both infected and uninfected tumor cells [193]. However, the use of antibiotics also has certain limitations. For example, in glioma, the application of broad-spectrum antibiotics has been shown to restrict the efficacy of TMZ and impair the recruitment of immune cells [158]. Additionally, eradication of H. pylori does not reverse the low responsiveness to cancer immunotherapy that is associated with H. pylori infection [148]. This may be due to the unintended elimination of immunostimulatory gut microbiota by broad-spectrum antibiotics, or to the persistence of a H. pylori-induced immune memory that remains even after the bacterium has been cleared [194]. In addition to antibiotics, other approaches can also reverse drug resistance by targeting the GM. For instance, in a mouse model of colorectal cancer, the phage VA7 completely reversed the resistance to 5-FU and oxaliplatin that was mediated by Bacteroides fragilis [131]. Moreover, in cases where radiotherapy and chemotherapy resistance in cervical cancer is induced by lactate metabolism of L. iners, remodeling the vaginal microbiome represents a potential strategy. In bacterial vaginosis, topical application of metronidazole followed by restoration with a beneficial strain, Lactobacillus crispatus CTV-05, resulted in vaginal colonization by Lactobacillus crispatus in nearly 80% of patients [195].

Targeting the metabolites of GM represents a more precise therapeutic strategy. For example, ammonium derived from Citrobacter freundii promotes the proliferation of myeloma cells, inhibits apoptosis, and reduces sensitivity to bortezomib by enhancing the acetylation of NEK2 and suppressing its degradation via the ubiquitination pathway. In contrast, furosemide sodium, a loop diuretic, inhibits ammonium uptake in multiple myeloma cells by downregulating SLC12A2 [151]. In CRC, Bacteroides vulgatus promotes the synthesis of microbial nucleotides, thereby enhancing tumor DNA repair capacity. Administration of the nucleoside transporter inhibitor NBMPR suppresses nucleoside uptake by tumor cells and reverses the associated resistance to neoadjuvant chemoradiotherapy [130]. In another study, CoPEC was shown to remodel the tumor microenvironment toward high glycerophospholipid content, reducing immunogenicity and leading to chemotherapy resistance. Treatment with Triacsin C, which blocks lipid droplet formation, significantly reversed the oxaliplatin resistance induced by CoPEC [134]. In cervical cancer, the development of resistance to chemoradiotherapy is associated with lactate metabolism mediated by L. iners. Beyond eliminating L. iners, targeting lactate metabolism represents a potential therapeutic strategy. Studies have shown a synergistic effect between gemcitabine and LDH inhibitors in cancer treatment [196], providing a rationale for the adjuvant use of LDH inhibitors.

Fecal microbiota transplantation (FMT), an innovative approach for modulating the gut microbiome, has attracted considerable attention in the field of oncology in recent years [197]. Emerging FMT techniques, such as fecal filtrate transplantation and fecal virome transplantation, enhance safety by eliminating potentially pathogenic elements. Transplanting GM from healthy individuals into patients can restore a healthy gut microbial community and help overcome therapy resistance [198]. In a cohort of patients with advanced melanoma resistant to PD-1 inhibitors, the combination of FMT and anti-PD-1 therapy modified the gut microbiome and reprogrammed the TME, ultimately overcoming resistance to anti-PD-1 therapy [199]. Mouse models have further demonstrated that the combination of FMT with PD-1 inhibition leads to an enrichment of Bacteroides vulgatus and Parabacteroides distasonis, thereby restoring treatment sensitivity in individuals who were previously non-responsive [200]. A Phase I clinical trial involving 10 patients with anti-PD-1-refractory melanoma demonstrated that FMT resulted in clinical responses in three participants, including one complete response and two partial responses [201]. A Phase I study involving 15 patients with refractory melanoma demonstrated that six patients experienced a reversal of PD-1 resistance following the administration of FMT in conjunction with anti-PD-1 therapy [199].

While there is currently no direct evidence establishing a link between FMT and radiosensitization, FMT has been demonstrated to be effective in alleviating the adverse effects induced by radiotherapy. A clinical study indicated that FMT significantly ameliorated gastrointestinal symptoms, such as diarrhea and rectal bleeding, in three out of five patients with gynecologic cancer who had undergone abdominal or pelvic irradiation [202]. A study utilizing an animal model further demonstrated that FMT effectively restored the gut microbiome and increased the abundance of metabolites, such as indole-3-propionic acid and valeric acid, thereby enhancing survival rates in rats [203]. FMT represents a promising direct approach for the manipulation of microbiota in cancer therapy. By restoring microbial diversity, FMT has the potential to enhance the efficacy of immunotherapy and mitigate the toxicity associated with it. As understanding of host-microbiome interactions in cancer treatment continues to advance, FMT is poised to become a valuable adjuvant strategy, thereby expanding the therapeutic options available to cancer patients.

To enhance the efficacy of cancer therapy, a range of microbial intervention strategies have been developed, encompassing lifestyle modifications, conventional microbial manipulations, and precision microbiome therapies (Fig. 3).

Fig. 3.

Intervention strategies targeting gut microbiota to overcome cancer therapy resistance

Dietary modulation reshapes gut microbiota

Dietary intervention represents the most accessible and well-tolerated strategy for modulating the human GM. Long-term dietary patterns play a significant role in shaping the microbial community structure [204], influencing not only the ecological balance of the intestine but also potentially affecting tumor development through immune and metabolic reprogramming. In a preclinical melanoma study, patients benefited most significantly from a high dietary fiber intake [146]. Dietary fibre is primarily fermented by the gut microbiota into SCFAs such as acetate, propionate, and butyrate. In animal models, SCFAs, particularly butyrate, augment the anti-cancer efficacy of CAR-T cells by facilitating metabolic and epigenetic modifications that enhance the expression of cytotoxic effector molecules in ROR1-specific CAR-T cells [205]. Moreover, SCFAs are closely linked to the long-term positive outcomes of ICI therapy in patients with NSCLC and melanoma [146, 206]. Interestingly, a mouse model study revealed that butyrate can compromise the efficacy of anti-CTLA-4 therapy. Specifically, it suppresses the therapy-induced upregulation of CD80/CD86 in dendritic cells [147]. A high-salt environment enhances anti-cancer immunity in mouse models by mediating the functional inactivation of MDSCs [207], suggesting a potential immunotherapeutic strategy for targeting MDSCs. Recent studies have demonstrated that a diet high in salt may compromise the effectiveness of FOLFOX chemotherapy in CRC by influencing macrophage immunomodulation through the metabolism of tryptophan by gut bacteria [208]. Thus, the role of salt in cancer therapy is contingent upon the specific context, demonstrating a dualistic effect that varies with the TME and treatment modalities. Furthermore, a ketogenic diet modifies the gut microbial composition, leading to an increased abundance of the commensal bacterium Eisenbergiella massiliensis and elevated serum levels of the ketone body 3-hydroxybutyrate. The presence of 3-hydroxybutyrate inhibits the upregulation of PD-L1 on myeloid cells during ICI treatment while promoting the expansion of CXCR3⁺ T cells, thereby enhancing anti-cancer immunity and improving the efficacy of ICI in melanoma and lung cancer models [209]. Inosine, a metabolite derived from GM, enhances the efficacy of immunotherapy through various mechanisms. Its ribose subunit supplies ATP and biosynthetic precursors to T cells, thereby supporting their expansion and functional requirements [210]. Inosine augments the immunogenicity of tumors by inhibiting the ubiquitin-activating enzyme UBA6 within tumor cells, thereby increasing their sensitivity to ICB therapy [211]. Additionally, it engages the adenosine A2A receptor on T cells to further enhance the sensitivity to ICI [212]. Consequently, the supplementation of dietary inosine or the transplantation of inosine-producing microbes into patients may constitute a safe approach to overcoming resistance to ICB.

Applications of probiotics, prebiotics, and next-generation probiotics

While dietary modulation is recognized for its safety, natural origin, and sustainability, its effects necessitate prolonged adherence and exhibit variability among individuals. Nonetheless, the uncertain efficacy, delayed onset, and potential risks associated with dietary modulation warrant further investigation. Probiotics, particularly those of the Bifidobacterium species, are crucial in maintaining gut homeostasis [213] and have demonstrated potential in inhibiting tumor progression and enhancing treatment outcomes [214]. Their potential to enhance the efficacy of immunotherapy, increase drug sensitivity, and reduce therapy-related toxicity is noteworthy [215]. For instance, supplementation with Bifidobacterium has been shown to synergize with anti-PD-L1 therapy by enhancing DC function and facilitating CD8⁺ T cell priming and tumor infiltration [216]. Extracellular vesicles originating from commensal Bifidobacterium are internalized by lung cancer cells through dynein-mediated endocytosis. This process leads to the upregulation of PD-L1 via TLR4-NF-κB signaling, thereby influencing anti-PD-1 responses in NSCLC [217]. Enterococcus species produce SagA, an NlpC/p60 peptidoglycan hydrolase, which augments the efficacy of anti-PD-L1 therapy [218]. Lactobacillus reuteri translocates to melanoma sites and releases indole-3-aldehyde, which activates the aryl hydrocarbon receptor in CD8⁺ T cells, thereby enhancing anti-cancer immunity and increasing sensitivity to ICI [219]. Table 4 provides a summary of the mechanisms mediated by microbiota that enhance cancer therapy. In comparison to traditional probiotics, these well-characterized bacterial formulations provide therapeutic advantages without the associated risks of pathogen transmission. For example, Tanoue et al. isolated a consortium of 11 bacterial strains, comprising seven Bacteroides and four non-Bacteroides species, from faecal samples of healthy human donors. This consortium spontaneously induces IFN-γ-producing CD8⁺ T cells in the intestine and enhances the efficacy of ICIs [220]. Similarly, a high abundance of Faecalibacterium prausnitzii is associated with an enhanced response to ICIs. The Faecalibacterium prausnitzii EXL01 strain has been shown to enhance T cell activation in the presence of ICIs in vitro. Oral administration of the EXL01 strain did not affect fecal microbial diversity or composition, indicating that it may directly influence immune responses in the small intestine [221]

Table 4.

Strategies for applying gut microbiota in cancer therapy

Cancer type	Microorganisms	Clinical Efficacy	Biological Functions	References(PMID)
CRC	Lactobacillus johnsonii La1	Promotes T lymphocyte proliferation and suppresses dendritic cell overactivation	Modulates the local immune microenvironment by inhibiting dendritic cell activation and enhancing T-cell responses	20066735
	11-strain human commensal consortium	Enhances efficacy of immune checkpoint inhibitors and suppresses tumor growth	Expands intestinal IFN-γ⁺ CD8⁺ T cells and activates tumor antigen-specific T cell responses	30675064
	Lactobacillus casei BL23	Significantly reduces tumor incidence in mouse models (0% vs. 67% in controls)	Downregulates the pro-tumor cytokine IL-22 and upregulates apoptosis-related gene expression	29209314
	Bifidobacterium lactis, Lactobacillus acidophilus	Promotes beneficial bacteria, suppresses procarcinogenic bacteria, and improves microbial composition	Butyrate inhibits proliferation, induces apoptosis, and upregulates tumor suppressor gene expression in colon cancer cells	28944067
Melanoma	Bifidobacterium spp.	Combined with anti-PD-L1, nearly abrogates tumor growth; increases tumor-infiltrating CD8⁺ T cells 2–threefold	Enhances dendritic cell function, promotes CD8⁺ T cell activation, and upregulates MHC class II, costimulatory molecules, and cytokines	26541606
	Enterococcus spp.	Enhances anti-PD-L1 efficacy, leading to significant tumor growth reduction in mice	SagA-derived muropeptides activate NOD2, promoting CD8⁺ T cell infiltration and activation	34446607
	Lactobacillus reuteri	Serum indole-3-carboxaldehyde levels correlate with improved ICI response and survival	Indole-3-carboxaldehyde activates AhR in CD8⁺ T cells, promoting Tc1 differentiation and IFN-γ secretion	37028428
Melanoma, NSCLC, Renal cell carcinoma	Eisenbergiella massiliensis, Akkermansia muciniphila	Improves response to immunotherapy and enhances ICI efficacy	Increases serum 3-hydroxybutyrate, inhibits PD-L1 upregulation on myeloid cells, and expands CXCR3⁺ T cells	33320838
Melanoma, B-cell lymphoma, Breast cancer	Engineered Escherichia coli	Suppresses tumor growth, reduces metastasis, and prolongs survival	Releases TLR agonists to activate macrophages/dendritic cells and enhance phagocytosis	31270504
Melanoma, Breast cancer	Bifidobacterium pseudolongum	Enhances anti-CTLA-4/anti-PD-L1 therapy and increases IFN-γ⁺ CD8⁺ T cell infiltration	Inosine activates adenosine A₂A receptor, promoting TH1 differentiation and IFN-γ secretion	32792462
NSCLC	Parabacteroides distasonis, Bacteroides vulgatus	Prolongs survival when combined with anti-PD-1 therapy	Depletes Tregs, activates effector T cells, and increases IFN-γ⁺ CD8⁺ T cells in the TME	34006584
B-cell lymphoma	Lactobacillus johnsonii	Delays lymphoma development and extends survival in mice	Reduces oxidative stress, inflammatory cytokines, and colonization of procarcinogenic bacteria	23860718
Cancers treated with radiotherapy	Bacteroides, Lactobacillus, Prevotella	Mitigates radiation-induced toxicity and aids hematopoietic recovery	Upregulates VEGF expression to promote angiogenesis in the small intestine	28242755
Colonic adenocarcinoma	Engineered Escherichia coli Nissle 1917	Enhances anti-PD-L1 efficacy, leading to significant tumor reduction	Increases L-arginine in the TME, promoting CD8⁺ T cell activation	34616044

CRC colorectal cancer, NSCLC non-small cell lung cancer, PD-L1 programmed cell death ligand 1, ICI immune checkpoint inhibitor, CTLA-4 cytotoxic T-lymphocyte-associated protein 4, PD-1 programmed cell death protein 1, TME tumor microenvironment

Prebiotics are substances selectively utilized by host microorganisms, conferring health benefits to the host. Prebiotics emulate the function of probiotics by facilitating digestion and maintaining gut homeostasis, and they are increasingly recognized for their potential roles in cancer therapy. Ginseng polysaccharides, which are the primary constituents of ginseng, enhance anti-PD-1 responses in NSCLC by augmenting CD8⁺ T cell function and suppressing Tregs [222, 223]. This prebiotic-immunotherapy combination reestablishes gut homeostasis, enhances the abundance of Lactobacillus and Bacteroides, and increases the SCFA levels. Additionally, ginseng polysaccharides decrease kynurenine levels and IDO activity, potentially improving the efficacy of PD-1 inhibitors [200]. In a similar vein, inulin has been documented to significantly increase the prevalence of Faecalibacterium and Bifidobacterium, which may enhance the effectiveness of immunotherapy [224]. Diosgenin, a steroidal saponin, augments anti-cancer immunity by promoting the infiltration of CD4⁺/CD8⁺ T cells and enhancing IFN-γ expression, while also serving as a sensitizer for anti-PD-1 therapy [225].

The advancement of whole-genome sequencing technology has facilitated the isolation and identification of novel beneficial microorganisms from the human microbiome, which show promise for development as next-generation probiotics (NGPs) [226]. Several studies have documented the contributions of NGPs in cancer therapy [227]. For instance, Akkermansia muciniphila enhances sensitivity to cisplatin in the treatment of lung cancer by modulating immune cells, promoting tumor cell apoptosis, activating key signaling pathways such as JAK-STAT and PI3K-Akt, and upregulating tumor suppressor genes including IFI2712 and IGFBP7 [228]. Limosilactobacillus fermentum enhances the sensitivity of cervical cancer to vincristine by inducing apoptosis and inhibiting the PI3K/AKT/mTOR survival pathway. Notably, its co-administration with vincristine allows for a significant dose reduction, thereby mitigating chemotherapy-related toxicity [229]. Additionally, Roseburia intestinalis restores intestinal barrier function and promotes butyrate production. Butyrate can directly bind to the TLR5 receptor on CD8⁺ T cells, activating them via NF-κB signaling, which in turn improves sensitivity to anti-PD-1 therapy [230]. While dietary, prebiotic, and NGP interventions hold promise for modulating immunity and enhancing cancer treatment, challenges related to efficacy, consistency, and long-term adherence impede their clinical adoption. Current trials are assessing their potential impact on cancer prognosis (Table 5). Future research should prioritize the development of precision prebiotics and probiotics with specific microbial targets to optimize their synergy and clinical utility in conjunction with immunotherapy.

Table 5.

Clinical trials applying gut microbiota modulation in cancer therapy

Identifier	Country/Region	Start date	Cancer type	Methodology	Phase	Primary objective	Primary outcome measures	Prognosis evaluation
NCT04163289	UK	2020–01	Renal cell carcinoma	Single-arm Open-label	Phase 1	To evaluate the safety of FMT combined with ipilimumab/nivolumab	Incidence of immune-related colitis	PFS/OS
NCT04645680	USA	2020–06	Melanoma	Two-arm Randomized Double-blind	Phase 2	To investigate the effect of diet on gut microbiome structure/function during immunotherapy	Change in gut microbiome composition	ORR/PFS
NCT04721041	China	2021–01	Malignant tumors	Single-arm Non-Randomized Open-label	N/A	To assess efficacy/safety of WMT for oncotherapy-related intestinal complications	Gastrointestinal symptom scores; stool frequency/consistency	N/A
NCT03819296	USA	2021–02	Melanoma, Genitourinary cancer	Single-arm Open-label	Phase 1	To study the role of gut microbiome and FMT efficacy for drug-induced GI complications	Change in stool microbiome pattern; incidence of FMT-related AEs	N/A
NCT04038619	USA	2021–02	Genitourinary cancer	Single-arm Open-label	Phase 1	To investigate FMT efficacy in managing ICI-induced diarrhea/colitis	Incidence of FMT-related AEs; clinical response/remission of colitis	N/A
NCT04729322	USA	2021–02	CRC	Non-randomized Open-label	Phase 2	To study FMT effect on anti-PD-1 response in previous non-responders	Objective response rate	ORR
NCT04131803	China	2021–10	CRC	Single-arm Randomized Single-blind	N/A	To evaluate efficacy/safety of probiotics combined with standard chemo/targeted therapy	Objective response rate	ORR
NCT04951583	Canada	2021–11	NSCLC, Melanoma	Multi-center Single-arm Open-label	Phase 2	To assess anti-tumor activity of FMT combined with ICI therapy	Objective response rate	PFS/OS
NCT04975217	USA	2021–12	Pancreatic ductal adenocarcinoma	Single-arm Open-label	Early Phase 1	To assess the safety, tolerability, and feasibility of FMT	Incidence of adverse events	N/A
NCT05286294	Norway	2022–06	Melanoma, head and neck squamous cell carcinoma	Single-arm Open-label	Phase 2	To evaluate safety, feasibility, and efficacy of FMT in ICI non-responders	Safety of FMT; tumor response evaluation	OS/ORR/PFS/DRR
NCT05251389	Netherlands	2022–08	Melanoma	Two-arm Randomized Double-blind	Phase 1/2	To investigate FMT potential to reverse ICI resistance in refractory metastatic melanoma	Efficacy assessment	PFS
NCT05533983	South Korea	2022–09	Solid carcinoma	Open-label	Phase 2	To evaluate efficacy/safety of FMT with nivolumab in advanced solid cancers progressed on anti-PD-(L)1	Overall response rate	ORR
NCT05592886	Hong Kong, China	2022–12	CRC	Two-arm Randomized Double-blind	N/A	To assess efficacy of synbiotic in reducing advanced adenoma recurrence post-resection	Incidence of metachronous advanced colorectal neoplasia at 1 year	N/A
NCT05462496	USA	2023–03	Pancreatic cancer	Multi-center Single-arm Open-label	Phase 2	To determine immune activation change in tumor tissue following treatment with antibiotics, pembrolizumab	Achievement of overall immune response	ORR/OS
NCT05502913	Israel	2023–09	Lung cancer	Two-arm Randomized Triple-blind	Phase 2	To evaluate safety/efficacy of FMT in altering response to immunotherapy	PFS	PFS/OS/ORR
NCT05690048	Germany	2024–01	HCC	Two-arm Randomized Single-blind	Phase 2	To assess safety and immunogenicity of FMT combined with standard immunotherapy	Immunogenicity assessment; safety of the combination therapy	PFS/OS
NCT06370884	USA	2024–02	Sigmoid colon cancer	Single-center Single-arm Open-label	Phase 1	To test safety/feasibility of IMT in patients undergoing colon resection	Safety of IMT; comparison of fecal microbiota pre/post-IMT	N/A
NCT06039644	Taiwan, China	2024–04	Breast cancer	Two-arm Randomized Double-blind	N/A	To evaluate efficacy of probiotics in improving/preventing chemotherapy side effects	Incidence of chemotherapy-associated side-effects	N/A
NCT06346093	China	2024–04	GC	Randomized Double-blind	N/A	To evaluate efficacy/safety of FMT capsules combined with chemo-immunotherapy	Objective response rate; Disease control rate	PFS/OS
NCT06205862	China	2024–04	Colorectal adenomas	Randomized Open-label	Phase 2	To investigate efficacy/safety of FMT in reducing CRA recurrence after endoscopic resection	CRA recurrence rate	N/A
NCT06405113	China	2024–06	GC	Randomized Double-blind	Phase 2	To explore efficacy/safety of FMT combined with SOX and Sintilimab in first-line advanced gastric cancer	2-year overall survival rate	OS/median PFS/ORR
NCT06403111	China	2024–06	NSCLC	Multi-center Single-arm Open-label	Phase 2	To evaluate efficacy of FMT combined with platinum-based chemotherapy and tislelizumab	12-month PFS	median PFS/ORR
NCT06349590	USA	2024–06	CRC	Single-center Single-arm Open-label	Phase 1/2	To evaluate impact of preoperative dietary modulation on preventing postoperative recurrence/metastasis	Change in dietary intervention lab values (collagenolytic potential)	N/A
NCT06486220	China	2024–07	Nasopharyngeal Carcinoma	Two-arm Randomized Open-label	Phase 3	To evaluate if FMT enhances efficacy and reduces side effects of low-dose 5-FU and immunotherapy	PFS	OS
NCT06793137	Brazil	2024–07	Rectal adenocarcinoma	Single-center Single-arm Open-label	Phase 2	To evaluate efficacy of oral antibiotics prior to neoadjuvant radiotherapy	Complete clinical response rate	RFS
NCT06428422	Turkey	2024–08	NSCLC	Two-arm Randomized Double-blind	N/A	To evaluate effect of probiotics on clinical effectiveness of immunotherapy	Clinical response; PFS; OS	PFS/OS
NCT06563947	China	2024–08	HCC	Single-center Single-arm Open-label	Phase 2	To evaluate efficacy/safety of oral enterobacterial capsules after treating with ICI and anti-angiogenesis therapy	PFS; OS; ORR	PFS/OS/ORR
NCT06563934	China	2024–08	HCC	Two-arm Randomized Double-blind	Phase 2	To evaluate additional efficacy/safety of oral enterobacterial capsules after treating with TKI and immunotherapy	PFS; OS; ORR	PFS/OS/ORR
NCT06551272	France	2024–12	HCC	Single-arm Open-label	Phase 2	To evaluate efficacy of EXL01 in reversing immunotherapy resistance	ORR at week12; Adverse events	PFS/OS
NCT06768931	China	2024–12	Breast cancer	Two-arm Randomized Open-label	Phase 2	To evaluate efficacy/safety of Biolosion combined with standard neoadjuvant therapy	Pathological complete remission	OS/DFS
NCT06393400	USA	2025–01	Pancreatic ductal adenocarcinoma	Single-arm Non-randomized Open-label	Phase 1	To confirm safety of combining oral FMT with gemcitabine/nab-paclitaxel as first-line treatment	Incidence of adverse events	PFS/OS/ORR
NCT05669846	USA	2025–01	NSCLC	Single-arm Open-label	Phase 2	To determine if FMT improves anti-tumor response in relapsed/refractory PD-L1 Positive NSCLC	ORR	PFS/OS/ORR
NCT06823323	China	2025–03	CRC	Two-arm Randomized Double-blind	N/A	To verify effectiveness/safety of L. johnsonii with CapeOX and Pembrolizumab in MSS/pMMR mCRC	PFS	PFS/ORR
NCT06030037	USA	2025–03	Melanoma	Two-arm Randomized Open-label	Phase 2	To evaluate efficacy of FMT combined with pembrolizumab and lenvatinib in advanced melanoma	ORR	PFS/OS
NCT06801665	China	2025–04	Colon cancer with liver metastasis	Multi-center Single-arm Open-label	Phase 2	To investigate efficacy of FMT combination therapy in colon cancer with liver metastasis	ORR	PFS/OS
NCT06623461	Canada	2025–04	Melanoma	Randomized Open-label	Phase 2	To evaluate efficacy of FMT (LND101) in combination with immune checkpoint blockade	PFS	PFS/OS/ORR
NCT06772090	USA	2025–04	Solid tumor	Two-arm randomized Single-blind	N/A	To evaluate feasibility of probiotics as an adjunct to cytotoxic chemotherapy	Number of participants completing the study	N/A

FMT fecal microbiota transplantation, WMT washed microbiota transplantation, GI gastrointestinal, AEs adverse events, ICI immune checkpoint inhibitor, CRC colorectal cancer, PD-1 programmed cell death protein 1, NSCLC non-small cell lung cancer, PD-L1 programmed cell death ligand 1, HCC hepatocellular carcinoma, IMT intestinal microbiota transplantation, GC gastric cancer, CRA colorectal adenoma

Novel technological strategies

Genetically engineered probiotics developed using gene-editing tools and synthetic biology approaches represent a significant strategy for expanding the current selection of probiotic strains and remodeling the complex gastrointestinal microbial environment. For instance, Chowdhury et al. developed a non-pathogenic strain of E. coli that specifically lyses within the TME to release a nanobody antagonist of CD47 (CD47nb). The localized delivery of CD47nb activates tumor-infiltrating T cells and induces rapid tumor regression. Furthermore, the local injection of these CD47nb-expressing bacteria provokes a systemic and tumor antigen-specific immune response, thereby aiding in the eradication of distant metastases and the suppression of untreated tumor growth [231]. In a further instance, a bioengineered strain of E. coli Nissle 1917 was developed to convert ammonia, a metabolic waste product present in tumors, into L-arginine, thereby enhancing the fitness and survival of T cells. This conversion results in an increased number of tumor-infiltrating T cells and demonstrates synergistic effects when combined with PD-L1 therapy, thereby promoting tumor clearance [232]. The engineered probiotic E. coli Nissle 1917is capable of precisely targeting and delivering a single-chain variable fragment of the anti-TREM2 to tumors. This process reverses immunosuppression and enhances the efficacy of αPD-L1 therapy [233]. The same strain exhibits the expression of surface-anchored adenosine deaminase under hypoxic conditions, facilitating the conversion of immunosuppressive adenosine into immunostimulatory inosine, thereby enhancing anti-cancer immunity [217]. The present findings underscore the potential of engineered probiotics to modulate the metabolism of the TME, thereby offering a novel strategy for augmenting cancer immunotherapy.

Nanomedicine presents a promising approach for the modulation of microbiota in oncology. In contrast to broad-spectrum antibiotics, nanodrugs are capable of selectively targeting specific microbial populations without disrupting commensal communities. Chen et al. developed a nanodrug by integrating the cytoplasmic membrane of F. nucleatum with colistin-loaded liposomes, facilitating the selective eradication of tumor-colonizing F. nucleatum [234]. Dong and colleagues identified a phage specific to F. nucleatum and utilized its capsid protein to construct silver nanoparticles for the precise targeting of intratumoral bacteria [235]. Ginger-derived exosome-like nanoparticles are preferentially internalized by Lactobacillaceae and Lachnospiraceae, facilitating the delivery of aly-miR159a-3p to enhance the accumulation of docosahexaenoic acid. The circulating docosahexaenoic acid subsequently binds to the PD-L1 promoter in tumor cells, leading to the downregulation of PD-L1 expression and augmenting the function of tumor-infiltrating CD8⁺ T cells. This process ultimately improves the efficacy of anti-PD-L1 therapy in melanoma [236]. These nanodrugs reduce off-target effects on beneficial microbiota while enhancing treatment precision.

In conclusion, the strategic manipulation of microbiota and their metabolites constitutes a promising advancement in cancer therapy. The selective utilization of specific microbes and their metabolic products can enhance the efficacy of immunotherapy while mitigating treatment-related adverse effects. The ongoing development of engineered bacteria and nanomedicines will further augment the repertoire of microbiota-based cancer interventions, thereby facilitating the emergence of more personalized and effective therapeutic strategies.

Challenges and future perspectives

While the GM has been implicated in contributing to therapeutic resistance, the translation of existing findings into clinical practice necessitates further investigation. Although studies have identified specific taxa or metabolites associated with adverse therapy outcomes, their predictive capacity is often confounded by factors such as individual variability, diet, medication use, among others. Much of the evidence is derived from single time-point analyses of fecal microbiota, which may not adequately capture the dynamic microbial shifts induced by therapy. Future research should prioritize longitudinal, multi-center cohorts with serial sampling of fecal, tumor tissue, and serum specimens, integrated with comprehensive multi-omics profiling. This approach will help elucidate whether microbial biomarkers are merely passengers of resistance or true drivers, and determine their utility in stratifying patients for tailored interventions.

While each of these strategies demonstrates potential, they are accompanied by specific limitations. Dietary and probiotic interventions, although generally safe, encounter challenges due to the ecological variability of the individual microbiome and diverse patient responses, leading to uncertain efficacy. Although antibiotics can enhance certain therapies by eliminating resistance-conferring pathogens, their broad-spectrum activity may cause collateral damage to commensal bacteria essential for therapeutic success, particularly in the context of immunotherapy. FMT, while potent, carries risks of pathogen transmission and currently lacks standardization in donor screening, bacterial preparation, and administration. Engineered probiotics and nanomedicine offer targeted action as precision strategies but remain in early-stage development; questions regarding long-term safety, immune responses to engineered vectors, and scalability remain unanswered. Crucially, the field must prioritize identifying which mechanistic pathways hold the greatest therapeutic potential in the setting of treatment resistance.

The advancement in overcoming microbiota-mediated therapy resistance is contingent upon transitioning from broad modulation to personalized oncology, which incorporates precise microbiota-targeting strategies. This shift requires moving beyond merely observing microbial composition to functionally understanding the mechanisms underlying individual tumor-gut microbiota-drug interactions. A particularly promising approach involves the development of synthetic microbial consortia or engineered bacterial strains capable of executing specific therapeutic functions, such as locally activating anti-cancer immunity or enhancing chemosensitivity, while allowing for controllable delivery and clearance. By integrating microbiota data with host immune status, tumor biology, and pharmacogenomics, predictive models can be constructed. Such a comprehensive understanding will facilitate the rational selection of microbiota-directed strategies to restore a therapy-favorable microenvironment, thereby disrupting the positive feedback loop that perpetuates therapy resistance.

Abbreviations

GM

Gut microbiota

SCFAs

Short-chain fatty acids

CRC

Colorectal cancer

TME

Tumor microenvironment

LPS

Lipopolysaccharide

BFT

Bacteroides fragilis Toxin

MDSCs

Myeloid-derived suppressor cells

TAMs

Tumor-associated macrophages

ROS

Reactive oxygen species

NK cells

Natural killer cells

PD-L1

Programmed cell death ligand 1

PMN-MDSCs

Polymorphonuclear myeloid-derived suppressor cells

Tregs

Regulatory T cells

DCA

Deoxycholic acid

ETBF

Enterotoxigenic Bacteroides fragilis

DCs

Dendritic cells

SASP

Senescence-associated secretory phenotype

H₂S

Hydrogen sulfide

5-FU

5-Fluorouracil

ICB

Immune checkpoint blockade

TMAO

Trimethylamine N-oxide

IAPs

Apoptosis proteins

CoPEC

Colibactin-producing Escherichia coli

PD-1

Programmed cell death protein 1

ADMA

Asymmetric dimethylarginine

CTLA-4

Cytotoxic T-lymphocyte-associated protein 4

NSCLC

Non-small cell lung cancer

ICI

Immune checkpoint inhibitor

TMZ

Temozolomide

FMT

Fecal microbiota transplantation

NGPs

Next-generation probiotics

CD47nb

Nanobody antagonist of CD47

E. coli

Escherichia coli

H. pylori

Helicobacter pylori

F. nucleatum

Fusobacterium nucleatum

L. iners

Lactobacillus iners

Authors’ contributions

All authors contributed to the conception, drafting, drawing, and final revision of the manuscript.

Funding

Liaoning Province Doctoral Scientific Research Start-up Foundation (2025-BS-0570).

Data availability

No datasets were generated or analysed during the current study.

Declarations

Competing interests

The authors declare no competing interests.