Intratumoral microbiota and metabolites: dual roles in cancer progression and therapeutic opportunities.

Introduction
About one out of every five people will experience cancer at some point in their lives. It is projected that by 2050, new cancer case is expected to climb to 35 million. Further research into the mechanisms, diagnosis, prevention, and treatment of cancer is urgently needed, which could save many lives worldwide and bring significant economic and social benefits to countries in the coming decades [1]. With the deepening of tumor research, extensive studies have shown the inseparable relationship between cancer and microbiota. The human microbiota consists of one hundred trillion organisms including bacteria, fungi, viruses, and protozoa. These microbial components indwell every part of human body. Currently, the association of the gut microbiota with tumors has been extensively studied [2–4]. Microorganisms are also present in tumors, which is different from what was previously thought that the tumor is a sterile environment [5]. However, the intratumoral microbiota, another type of microbiota that is equally important for tumors, has not been sufficiently investigated. The gut microbiota mainly affects tumor development by regulating the body’s immune system and metabolic processes, and influences non-gastrointestinal tumors through axes [6–10], while the intratumoral microbiota directly acts on tumor cells to influence tumor progression [5]. Recent investigations highlight the involvement of intratumoral microbial component in various cancer types, thus providing new directions and targets for cancer treatment [5, 11–16]. Specific resident microorganisms have been recognized across diverse cancers, like lung, breast, colon, and oral cancers [17, 18]. Microorganisms in specific tumors can either promote or suppress tumor development and are linked to tumor prognosis. Boosting the presence of advantageous microorganisms and diminishing the presence of detrimental ones in tumors can augment the potency of tumor immunotherapy, chemotherapy and radiotherapy [19]. The ability of microorganisms to colonize tumors can also be exploited to deliver drugs directly to the tumor, enabling targeted killing or inhibition of cancer cells [5, 20–23].
Intratumoral microbiota-derived metabolites serve as a crucial link between intratumoral microorganisms and host tumor, mainly by regulating tumor metabolism and immune microenvironment. For example, bacterial metabolites can trigger endoplasmic reticulum stress in cancer cells, thereby promoting apoptosis, and simultaneously enhancing the host’s antitumor immune response to modulate tumor progression. Thus, intratumoral bacterial metabolites may hold significant implications for tumor therapy [5, 24].
In this review, we extensively explore the significance of the intratumoral microbiota and metabolites in diverse tumor types, as well as their implications for tumor diagnosis, prognosis, and treatment. This review synthesizes current evidence on the relationship between intratumoral microbiota and metabolites and tumors and provide novel insights into the future application of intratumoral microbiota and their metabolites in the diagnostic and therapeutic field of tumors.

Methods

Literature search strategy
To identify all relevant publications for this comprehensive review, a systematic literature search was performed. The primary electronic databases utilized were PubMed/MEDLINE, Cochrane, Embase and Google Scholar. The search was conducted for articles published from 1999 to 2025, to capture the modern era of cancer microbiology and metabolomics research. The search strategy employed a combination of keywords and Medical Subject Headings (MeSH) terms related to the core concepts of the review. The key search strings were constructed around the following themes:
Intratumoral Microbiota: (“intratumoral microbiota” OR “intratumoral microbiome” OR “tumor microbiota” OR “tumor microbiome” OR “cancer microbiota” OR “tumor-associated bacteria” OR “intratumoral bacteria”).
Tumor Microenvironment and Metabolites: (“metabolom” OR “metabolit” OR “oncometabolite” OR “tumor microenvironment” OR “TME”).
Cancer Progression and Therapy: (“neoplasms” OR “cancer” OR “tumor” OR “oncology”) AND (“progression” OR “evolution” OR “advancement” OR “development” OR “metastasis” OR “therapy” OR “treatment” OR “immunotherapy” OR “immune” OR “immunity”).

Inclusion criteria
Original research articles and high-impact review articles; studies that directly investigate the presence, composition, or function of intratumoral microbiota (bacteria, fungi, viruses) in human solid tumors or validated preclinical models; studies focusing on the interplay between intratumoral microbiota and metabolites and TME, including the production or modulation of specific metabolites; research elucidating the mechanistic roles of microbiota or their metabolites in cancer initiation, progression, metastasis, or therapy response; research providing value for tumor diagnosis and prognosis; articles published in the English language.

Exclusion criteria
Studies focusing solely on gut microbiota without direct measurement or implication of intratumoral microbes; articles not peer-reviewed (e.g., editorials, conference abstracts without full data, or opinion pieces), unless they provided critical foundational knowledge; studies published only in a language other than English; case reports with a sample size of fewer than five patients, unless they presented a unique and seminal finding; articles with outdated perspectives or limited influence.

Scope of the review
Databases: The synthesis is primarily based on literature retrieved from the comprehensive searches of PubMed/MEDLINE, Cochrane, Embase and Google Scholar.
Date Range: The review covers the period from 1999 to 2025, with a focus on the significant advancements made in the last five years.
Article Types: The analysis is grounded in evidence from original research articles using in vitro, in vivo, and clinical studies. Seminal review articles are also incorporated to provide context and summarize established paradigms.

Relationship between intratumoral microbiota and tumor

Origin and distribution of the microbiota within tumor
There exist some ideas about the source of the microorganisms within the tumor. On the one hand, intratumoral microorganisms may originate from microbes that colonize the mucosa and invade the tumor upon injury to the mucosal membrane [5]. To illustrate, non-mucosal organs (pancreas) and mucosal organs (esophagus, lung, colon, and cervix) harbor the same microbial species [9]. This suggested that microbes may disseminate through damaged mucosa. Research indicated that microorganisms invade from the broken mucous membranes of mucosal organs and then disseminate to the pancreatic duct, subsequently entering the pancreas. After that, TME in the pancreas is reshaped and provide suitable conditions for tumor growth [25, 26]. Another research showed that the microbial communities that migrate through the mucosa may be the primary source of microbiota in gastrointestinal tumor [27]. On the other hand, microbes can translocate into tumors via hematogenous dissemination. Findings demonstrated that gastrointestinal microbiota and their metabolites travel through the bloodstream and colonize extra-intestinal organs. The long-term action of translocated microbes on the TME induces tumorigenesis and influences tumor progression [28]. It was discovered that melanoma patients exhibit high levels of clostridial bacteria and low levels of Gardnerella vaginalis in their tumors. Interestingly, a similar phenomenon of high clostridial bacteria and low Gardnerella vaginalis was likewise detected in the intestines of melanoma patients. This suggested that microbes in the tumor may initially reside in the gut and then colonize the tumor via the bloodstream [29].
There are several contentions regarding the distribution of intratumoral microbes. From the perspective of tumor types favored by microbial colonization, intratumoral microbes often distributes within tumors that occur in mucosal-related organs such as lung, esophageal, pancreatic, colorectal, and cervical cancers [30–34]. From the perspective of cell types in tumor tissue, intratumoral microbes are mostly distributed in the nucleus of tumor and immune cells [5, 35].

Interactions between intratumoral microbiota and tumor
Cellular and non-cellular components in tumor tissues are diverse, which are called host intermediates, including immune cells, untransformed cells, stromal cells, endothelial cells, and extracellular matrix (ECM). These host intermediates interact with the microorganisms in the tumor to make up the intricate TME [36, 37]. Intratumoral microbes have a regulatory effect on the TME, while microbiota dysbiosis induces the imbalance of TME [38]. The hypoxic conditions and abundant nutrient supplements in TME offer favorable niches for microbial habitation [39].
Intratumoral microbes directly interact with surface or internal receptors of tumor cells, modulating signaling pathways that relating to transcription of tumor proliferation-related genes. Intratumoral bacteria, such as F. nucleatum, directly interacts with toll-like receptor (TLR) 4 on the surface of tumor cells, leading to the upregulation of myeloid differentiation primary response gene 88 (MyD88) transcription and increases of microRNA-21 expression, which suppresses RAS protein activator like 1 and contributes to tumor proliferation [5]. Secondly, intratumoral microbiota interact with immune cells within or outside TME to modulate tumor development. This immune-driven interplay and coordinated feedback mechanism is termed as immuno-oncology microbiome axis (IOM) [40, 41]. Program the tumor microenvironment, converting immunologically ‘cold’ tumors into ‘hot’ phenotypes can enhance anti-tumor immune response. Subsequently, introducing specific microbes into the tumor can augment the effectiveness of immunotherapy. Similarly, engineered bacteria carrying immune checkpoint inhibitors (ICIs) can combine photothermal stimulation to produce an immunoactivating effect within the tumor [42, 43]. As elucidated in pan-cancer studies, the interplay between the fungal, bacterial and immune compartments indicates that fungi drive a “mycobacterial-like” signature, which is characterized by distinct immune responses and subsequently influences patient survival rates [44]. Moreover, except for direct interactions, the dialogue between intratumoral microbes and tumor cells usually occurs through microbiota-derived metabolites. Take an example, metabolites generated by intratumoral microbes can activate β-catenin signaling within tumor cells, thereby enhancing the expression of c-Myc, which is a transcription factor. The significant upregulation of c-Myc levels stimulates transcriptional and translational activities within the endoplasmic reticulum. The process leads to an accumulation of a substantial amount of unfolded proteins, which activates the Protein kinase R-like endoplasmic reticulum kinase pathway. Ultimately, this cascade of complex biological processes facilitates pyroptosis in tumor cells mediated by gasdermine [5, 24, 45, 46].

Tumor enriched microorganisms
Patients with shorter survival periods harbor a “dysbiotic” tumor microbiome, and interestingly a substantial discrepancy can be detected in microbial richness across different cancer types [16, 39, 47]. Among all tumor types, the most common bacterial phyla are Proteobacteria and Firmicutes, with 50–60% of tumors enriched in Proteobacteria, while Actinobacteria predominates in non-gastrointestinal malignancies, including breast cancer (BC). In BC, the richness of microbial communities such as Actinobacteria and Firmicutes has been found to be higher than in normal breast tissue of healthy subjects, and the bacterial load in adjacent breast tissue is intermediate between breast tumors and normal samples [48]. The most common bacteria in breast tumor tissue include Actinomyces, Bartonella, Brevundimonas, Corynebacterium, Bacteroides, Mycobacterium, Rickettsia, and Sphingomonas, while common fungi include Aspergillus, Candida, Blastomyces, Cunninghamella, Geomyces, Late Bacteria, and Red Yeast [49]. Furthermore, evidence suggests that the abundance of Fusobacterium in oral cancer is higher than that in healthy tissues [50]. In gastric cancer (GC), levels of Helicobacter pylori (Hp), Porphyromonas gingivalis (Pg), and oral microbiota (Fusobacterium and Streptococcus) are higher in tumor sites than in adjacent healthy tissues. Similarly, in esophageal tumors, the abundance of Pg, Fusobacterium, and Streptococcus is also higher within the tumor than in adjacent healthy tissues [17, 51]. Analysis of the microbiota in pancreatic ductal adenocarcinoma (PDAC) reveals that Proteobacteria accounts for 50% of all bacterial phyla in cancerous tissue [52]. Moreover, in lung adenocarcinoma (ADC), the abundance of Acinetobacter, Propionibacterium, Trueperella, Mycoplasma, and Staphylococcus is upregulated. Besides, in lung squamous cell carcinoma (SCC), the level of Enterobacteriaceae, Serratia, Morganella, Klebsiella, and Chryseobacterium is increased [53]. Concurrently, Fusobacterium is found to be abundant in SCC of the head and neck [54]. Fungi exhibit a unique distribution across various types of cancers. In the white infiltrating skin cancer subtype, Ascomycota and Basidiomycota dominate the tumor microbiome, with yeasts (belonging to Ascomycota) being the most abundant in colorectal cancer [44]. The presence of these tumor-associated microbial communities is significantly connected to the emergence of tumors. They induce inflammatory responses or immunosuppression in the host, which is predictive of a poor prognosis in cancer patients [55] (Table 1).

Intratumoral microbiota and metabolites in tumor development

Mechanisms of intratumoral microbiota in tumor development

Intratumoral microbiota induce DNA damage
Inducing DNA mutations contributes to the carcinogenic process of microbiota [5]. Genotoxic microbiota can promote tumorigenesis. A significantly elevated proportion of mucosa-associated polyketone synthase-positive (pks+) Escherichia coli (E. coli) was found in colorectal cancer (CRC) patients, while deletion of pks genotoxic islands in E. coli NC101 can reduce tumor diversity and aggressiveness [88]. Pks island, which is genomic region, encodes the biosynthetic machinery to produce colibactin. The genotoxin colibactin is capable of alkylating DNA, thereby precipitating double-stranded breaks and interstream cross-links at genomic sites that coincide with somatic mutation hotspots observed in the genomes of CRC. The disruption of DNA structure leads to a reduction in the genetic homeostasis of human intestinal epithelial cells, ultimately culminating in the onset of CRC [89, 90]. Likewise, the aberrant DNA methylation is the chief pathway for Hp invasion to prompt stomach adenocarcinoma (STAD) [17]. Additionally, it has come to light that Bacteroides fragilis (B. fragilis) exerts a similar effect comparable to that of E. coli and Hp, as it resides within the epithelial cells of the intestinal mucosa in patients with familial adenomatous polyposis, and causes substantial DNA damage within these cells, which further aids in the development of CRC [91]. Furthermore, microorganisms can promote reactive oxygen species (ROS) production in bone marrow cells, and ROS further induces DNA damage, which in turn downregulates NAD+, and then leads to aging of M1-like macrophages and weakens host anti-tumor immunity [92]. For example, B. fragilis toxins and Campylobacter jejuni’s cell-lethal bulking toxin both contribute to the upregulation of spermine oxidase, thereby inducing ROS to damage DNA [93]. Furthermore, Salmonella typhi secretes a variety of pathogenic elements that induce DNA damage and induce the development of biliary tract cancer [94].

Intratumoral microbiota modulate signaling pathways
The intratumoral microbiota modulates various signaling pathways to affect the onset and progression of tumors, a process often associated with immune responses [5]. The surface receptors of T cells have a complex impact on tumorigenesis induced by intratumoral microbes. Lipopolysaccharides of Gram-bacteria recognize TLR4 and further activate the NF-κB pathway, eliciting the discharge of inflammatory factors (IL) (such as IL-1β, IL-6, IL-8, TNF-α, etc.), thereby boosting anti-tumor immune responses. For instance, Fusarium species bind to TLR4 on CRC cells, activating NF-κB signaling, which increases inflammatory cytokines and downstream ERK signaling, leading to an increase in 12,13-EpOME, collectively enhancing anti-tumor immunity [95]. Bifidobacteria that migrate and colonize the CRC site can activate dendritic cells (DCs) by enhancing their cross-priming ability and augmenting the CD47-based immune response in a STING-mediated fashion, thereby bolstering anti-tumor immunity [96]. These results highlight the capacity of intratumoral microbiota to augment anti-tumor immune responses and thereby suppressing tumor progression by mediating signaling pathways.
However, another intratumoral microbes can stimulate tumor development by mediating specific intratumoral signaling cascades. For instance, Pg within tumor cells promotes the activation of gingipains, thus activating mitogens and protein kinases, engaging signal transduction pathways involving protein kinases that enhance the proliferation of PDAC and CRC cells [97, 98]. Moreover, B. fragilis-produced enterotoxins can activate mTOR pathways regulated by the long non-coding RNA BFAL1, accelerating the growth of CRC [99]. Moreover, Fusarium induces the activation of ALPK1, a novel pattern recognition receptor (PRR), which further triggers the NF-κB signaling cascade, driving up the expression of ICAM1 and enhancing metastatic potential of CRC cells [100]. Furthermore, Hp, recognized as a potent carcinogen, facilitates its colonization of the gastric mucosa by increasing the levels of matrix metalloproteinase-10, subsequently secreting the oncoprotein cytotoxin-associated gene A, stimulating the Hippo pathway, causing prolonged inflammatory responses and changing the nature of the gastric mucosa, ultimately leading to tumorigenesis [101]. Parallelly, Streptococcus sanguinis interacts directly with gastric cells through the TMPC-ANXA2-MAPK signaling pathway, promoting gastric tumorigenesis [68]. Collectively, these findings demonstrate that several intratumoral microbes can facilitate tumor progression by modulating intratumoral signaling networks. An intriguing revelation is that some microbe-mediated signaling pathways and molecules are specifically activated and enriched within specific tumor tissues. For instance, the MetaCyc PWY-5159 pathway is activated by bacteria within tumors for the degradation of hydroxyproline, specifically in bone tumors. Furthermore, pathways mediated by bacteria for the degradation of harmful substances in smoke are found to be activated within lung cancer tumor tissues. This finding suggests that bacteria participate in and mediate signaling pathways responsible for the degradation of certain metabolites enriched within tumors [48].

Intratumoral microbiota regulate anti-tumor immune response
In tumor regions with high bacterial diversity, high expression profiles of PD-1 and CTLA-4 reflect an inseparable relationship between intratumoral bacteria and tumor immunity [55, 102]. The intratumoral microbiota influence the immune system through various mechanisms, playing different and even opposing roles in the development and progression of tumors. While intratumoral microbes can enhance anti-tumor immunity to inhibit tumor development, they can also reduce immune defense against tumors and drive cancer advancement [5].
Intratumoral microbiota can boost anti-tumor immune defense and thereby inhibit tumor development via mechanisms like activation of STING pathway, recruitment and activation of T and natural killer (NK) cells, formation of tertiary lymphoid structures (TLS), and presentation of tumor-derived antigens [5]. Bifidobacteria migrate and colonize in CRC sites, activating DCs and promoting immune reactions [96]. Helicobacter hepaticus (Hhep) colonizes CRC sites, encouraging the production of T follicular helper cells in colon, supporting the maturation of TLS near Hhep+ tumors, thereby promoting immune responses [103]. The overall survival (OS) rate of PDAC patients is remarkably positively correlated with the density of granzyme B+ (Gzm B+) cells in tumor tissue. Interestingly, there is a strong positive correlation between the tissue density of CD8+ and Gzm B+ cells and the diversity of the intratumoral microbiome [104]. Additionally, hippurate produced by Bifidobacterium can boost the activity of NK cells, leading to the regression of melanoma [105]. Furthermore, microbes such as Fusobacterium nucleatum (Fn), Epstein-Barr virus, Hepatitis B virus, and Merkel cell polyomavirus can trigger cytokine synthesis, including interferon-γ (IFN-γ), and increase CD8+ T cell penetration, so as to facilitate anti-tumor immune responses in patients with cutaneous melanoma [106–109]. In addition, peptides produced by intratumoral bacteria can be presented by DCs, thereby activating T cell responses [5]. Thus, intratumoral microbiota can suppress tumor development by enhancing anti-tumor immune reactions.
Conversely, intratumoral microbiota can also contribute to malignant progression by upregulating ROS, inducing immune acceptance and downregulating the number and activity of immune cells [5]. B. fragilis promotes the progression of CRC by producing ROS [110]. Similarly, Fusobacterium promotes the progression of gastrointestinal tumors by upregulating ROS [52]. Additionally, in CRC tissues, the number of Fn+ cells is negatively correlated with the concentration of CD3+ T cells. Fn may downregulate T cell-mediated adaptive immunity to suppress anti-tumor effects, thereby enhancing the progression of CRC [111]. Certain microbiomes in PDAC induce an immune tolerance protocol by stimulating selective TLRs in monocytes, thereby promoting immunosuppression. Hence, certain intratumoral microbes can promote tumor progression by inducing immunosuppression.
Microbial microbiota also affect the complement system and provoke the complement cascade. Glycans of wall of Malassezia bind to mannose-binding lectin (MBL) in TME, activating C3 invertase and increasing C3a. C3a targets C3a receptor (C3aR) on tumor cells, enhancing proliferation, motility, and invasiveness of tumor. Following that, it was found that low expression of MBL or knockdown of C3aR can slow tumor growth, leading to the conclusion that Fungi promote the development of PDAC by initiating the complement cascade via MBL binding [5, 112–114] (Fig. 1).

Intratumoral microbiota facilitate chronic inflammation
Chronic inflammation significantly contributes to tumorigenesis, progression, metastasis and recurrence by causing genetic chaos, gene mutations or epigenetic modifications [115]. Nevertheless, intratumoral microbiota can cause local chronic inflammation by activating inflammatory signaling pathways and releasing inflammatory factors, thereby promoting cancer occurrence and development [116].
The intratumoral microbiome can initiate inflammatory signaling and elicit cascades by engaging PRRs in the TME and then acting on TLRs, such as Fn and Campylobacter conisus. [16, 117] PRRs can identify microbial pathogens by detecting pathogen-associated molecular patterns, including nucleic acids from microbes, lipopolysaccharides (LPS), and α-mannans. While PRRs recognize pathogens, they also trigger and attract immune cells (neutrophils, monocytes and lymphocytes) to infection or injury, leading to localized chronic inflammation that accelerates tumorigenesis [118]. Studies have found that mice with selective defects in PRR signaling pathways (including Mincle, TLR4, TLR7 and TLR9) experience a slower advancement of PDAC [119]. In addition, Fn can affect TLR4 and enhance the IL-6/p-STAT3/c-MYC pathway, which leads to the polarization of M2-like macrophages and subsequently correlates with the progression of CRC in mice [117]. In addition, C. conisus can lead to the elevated expression of PRRs and the accumulation of the IFN-inducible protein 16 (IFI16) mediated inflammasome activation, which could be the cause of esophageal cancer induced by C. conisus [120]. Intratumoral microbiota can also augment inflammatory pathways. Autophagy inhibition mediated by Fn in CRC cells promotes the accumulation of ROS, stimulating the release of TNF-α, IL-8 and IL-1β. Interestingly, in a CRC mouse model, the depletion of neutrophils is observed to facilitate the proliferation of Akkermansia within the tumor, subsequently boosting IL-17 synthesis and augmenting B cell infiltration, thus facilitating inflammation and tumor growth [121]. Besides, enterotoxigenic Bacteroides fragilis toxin prompts a STAT3-NF-κB signaling pathway-specific pro-inflammatory signaling cascade, releasing inflammatory factors, which can attract tumorigenic-supporting myeloid cells and promote the occurrence of distal CRC [122]. Furthermore, outer membrane protein of Hp, namely HopQ, interacts with immune cell surface antigen CEACAM1 and translocates the pathogenicity factor CagA into host cells to enhance the expression of inflammation-promoting substances like IL-8, which contributes to the progression of GC [65]. An increasing body of evidence posits a positive correlation between Propionibacterium acnes and both acute and chronic inflammation. Studies reported that Propionibacterium acnes stimulate the secretion of inflammatory markers in prostate epithelial cells to promote the progression of prostatitis and prostate cancer. Furthermore, preclinical models demonstrated that the administration of Propionibacterium acnes isolated from prostate tumors to mice induces prostatic inflammation and tumorigenesis [123, 124]. The above studies indicate that intratumoral microorganisms can induce local chronic inflammation to the onset and progression of tumors.

Intratumoral microbiota modulate tumor metastasis
Different intratumoral microbiota play different even completely opposite roles in tumor metastasis. Intratumoral microorganisms trigger tumor metastasis by disrupting the mucosal barrier at the primary tumor site and causing premetastatic niche (PMN). E. coli C17 in rectal cancer patients disrupts the gut vascular barrier (GVB), facilitating tumor cells to spread to liver and form a PMN in the liver [125]. Study indicated that microbial metabolites can be involved in the induction of epithelial-mesenchymal transition (EMT), which is favorable for the establishment of a PMN [46]. For example, Fn can induce CRC cell multiplication and migration. There is a chance that the mechanism is that Fn increases intracellular Cytochrome P450 Family 2 Subfamily J Member 2v expression and linoleic acid production by activating TLR4 signaling, leading to EMT, which in turn promotes CRC formation and metastasis [126]. Fn can also colonize CRC tissues with its lectin Fap2 and inhibit the aggregation of T cells infiltrating the tumor to drive development of tumors and metastatic advance [127]. Moreover, Fusobacterium sclerotinia promotes the emission of C-X-C motif chemokine receptor 1 and IL-8 and contributes to the multiplication and dispersal of HCT116 cells, a type of CRC cell [128]. Additionally, in PDAC, Malassezia is considerably elevated, and the glycans in its wall interact with MBL in TME to facilitate tumor proliferation and metastasis [114]. It should be noted that intratumoral microbiota is crucial in promoting the dissemination of tumor cells and in reducing the damage to tumor cells during the metastatic process. In BC, staphylococcus and streptococci can inhibit the RhoA/ROCK pathway to remodel the cytoskeleton, thus aiding tumor cells in withstanding vascular mechanical stress and avoiding damage when metastasizing from primary tumors [129, 130].
Interestingly, in contrast, Fn is a type of microorganism within oral squamous cell carcinoma and research showed it is bound up with a lower frequency of lymph node infiltration and fewer distant recurrences. Compared to patients with Fn−, those with Fn+ show significantly prolonged survival outcomes [130–132] (Fig. 2).

Intratumoral microbiota modulate tumor metabolism
Intratumoral microorganisms modulate tumor metabolism, hence serving a dual purpose in tumor development. In BC, Bacteroides are implicated in metabolic pathways to induce carcinogenic pathways and tumor development [133]. On the contrary, P. fungorum upregulates metabolism of alanine, aspartate, and glutamate by acting on their metabolic pathways, thereby inhibiting the growth of intrahepatic cholangiocarcinoma [134]. Similarly, interference with Clostridium butyricum (C. butyricum) or its metabolite, butyrate, triggers intracellular lipid accumulation in tumor cells and oxidative stress, thereby enhancing their vulnerability to ferroptosis and thus suppressing the development of PDAC [74].

Intratumoral microbiota-derived metabolites influence cancer development and treatment
Microbiota-derived metabolites are important mediators of intratumoral microbes and host’s metabolic environment [135, 136]. For instance, one metabolite that can enhance immune efficacy is inosine, produced by Bifidobacterium. For one thing, it can act as a substitute carbon source for CD8+ T cells, so that providing energy for immune cells, for another thing, it can inhibit UBA6 in tumor cells so as to enhance tumor immunogenicity [137]. In a melanoma mouse model, the concurrent use of inosine and anti-PD-1 antibody treatment retards tumor expansion and increases patients’ OS [138]. Similarly, indole-3-aldehyde (I3A), an agonist of aromatic hydrocarbon receptor (AhR), released by Lactobacillus reuteri, effectively enhances tumor-targeting immunity and promotes treatment reaction to ICIs in melanoma preclinical models [87, 139]. Besides, in triple-negative BC, it was found that trimethylamine N-oxide (TMAO) derived from Clostridiales can promote anti-tumor immunity [24]. It is worth mentioning that C. butyricum generates butyrate to enhance oxidative stress and lipid accumulation, finally triggering ferroptosis in cancer cells [74]. Additionally, butyrate has the capacity to reconfigure the metabolic pathways in CRC cells by activating pyruvate kinase isozyme M2, thereby exerting anti-cancer effects [140]. Sodium butyrate can also suppress upregulation of essential adhesion-related proteins, including RadD and FadA, so as to inhibit the proliferation of Fn and diminish its colonization and invasive capabilities [141]. Additionally, bile salt hydrolase produced by Bacteroides, Clostridium, and Lactobacillus species can convert bile acids (BA) into secondary bile acids (SBA), like lithocholic acid (LCA), glycochenodeoxycholic acid, and ursodeoxycholic acid. SBA modulates T-cell differentiation and promotes anti-tumor immune response by binding to a range of receptors including muscarinic receptors, TGR5, FXR, and GPCRs. SBA even can predict early remission in inflammatory bowel disease (IBD) patients receiving anti-cytokine therapy and the prognosis of CRC patients [46, 142, 143]. It is particularly worth mentioning that LCA can not only induce oxidative phosphorylation and tricarboxylic acid (TCA) cycles, inhibit the VEGF production and EMT to inhibit tumor advancement, but also reverse lipid metabolism, inhibit adipogenesis to induce apoptosis of BC cells [144, 145]. These findings highlight the anti-tumor possibilities of intratumoral metabolites and pave the way for new approaches for cancer therapy strategies.
However, it is crucial to highlight that other subsets of microbial metabolites are positively associated with tumor progression by restraining immunity, inducing gene mutation, facilitating tumor invasion and promoting inflammation. Many bacteria that reside within human tumors ferment to produce lactate, such as phyla Bacteroidetes, Firmicutes, and Proteobacteria. Increased concentration of lactate can promote tumor progression like pancreas carcinoma [48, 146]. The mechanism by which lactate works is to inhibit the anti-tumor immunity effect and promote cell growth and dissemination. Lactate accumulation in TME accelerates macrophage-M2 conversion, transforms MDSCs into M-MDSCs rather than normal macrophages, and restrains T cells activation and NK cells cytotoxicity [147–150]. Lactate may also function by increasing the VEGF expression mediated by hypoxia-inducible factor 1α (HIF1α) and enabling tumor-associated macrophages (TAM) to differentiate into M2 [151]. Thus, metabolic reprogramming and immune remodeling through intratumoral lactate depletion provides a new therapeutic direction [152]. Furthermore, Pg, a bacterium known for its secretion of peptidylarginine deiminase, augments the mutation rates of TP53 and KRAS, thus playing a role in the emergence of PDAC [153]. Likewise, another research indicated that succinic acid derived from Fn may promote tumor development. Succinic acid can inhibit the cGAS-interferon-β pathway, thereby reducing the number of CD8+ T cells in TME to inhibit anti-tumor responses [154]. Fn in GC can secrete endotoxins to suppress immune functions and alter the tumor inflammatory microenvironment [66]. Surprisingly, SBA can sometimes facilitate tumor progression. LCA, produced by Clostridioides difficile, stimulates the urokinase plasminogen activator receptor, thereby inducing tumor infiltration and metastatic expansion. Additionally, Clostridioides difficile harbors an active 7-dehydroxylase capable of transforming BA into deoxycholic acid (DCA), which is associated with increased Cyclo-oxygenase 2 (COX-2) activation and prostaglandin production, resulting in DNA damage, inflammation, and fibrosis, thus promoting oncogenic neoplasms’ advancement, such as ovarian, colorectal, and pancreatic malignancies [46, 155]. Notably, certain microbial metabolites can interact with anti-tumor drugs, decreasing the cytotoxic effects of the drugs on tumors. For instance, cytidine deaminase from Mycoplasma and Proteus species can degrade gemcitabine into inactive compound 2’,2’-difluorodeoxycytidine, thereby conferring tumor cells resistance to the drug. This resistance is eliminated when bacteria are eradicated by ciprofloxacin [156]. These discoveries elucidate the detailed link between intratumoral microbial metabolites and tumor progression, providing novel insights for cancer diagnosis and treatment in the future (Fig. 3).

Intratumoral microbiota in tumor diagnosis and prognosis
Intratumoral microbiota’s abundance is closely associated with tumor types, thereby endowing it with the capacity to function as a diagnostic instrument. An escalating volume of research reported diagnostic value of microbes within tumors [157, 158]. For instance, microbiota features within head and neck squamous cell carcinomas are correlated with clinical pathological tumor features. Fusobacterium and Treponema within tumor can evaluate tumor stage and histological grade [159]. Additionally, Pg and Aggregatibacter actinomycetemcomitans (Aa), as one of the etiological factors of pancreatic cancer, also holds certain diagnostic value [160]. Moreover, a molecular subtype of GC, microsatellite instability high (MSI-high), is enriched with unique microbes, including Bacteroides, Prevotella, Streptococcus, and Fusobacterium genera. These distinct microbes have an important impact on the pathogenesis and prognosis of GC [161]. Besides, Solicoccozyma is considered as a gastric fungal identifier for GC staging [162]. Aykut et al. found that PDAC commonly harbors DNA of Malassezia, the genus with the highest prevalence in duodenal fluid. They also uncovered the proof of disease-causing mechanisms of Malassezia, which means Malassezia sets off the complement cascade via interaction with MBL, consequently advancing PDAC. Anticipatory gathering of duodenal fluid before PDAC diagnosis can identify microbiome alterations correlated with the risk of PDAC [114, 163]. In esophageal squamous cell carcinoma (ESCC), intratumoral presence of Fusobacterium and Prevotella is significantly correlated with tumor staging, with a notably higher abundance of Fusobacterium in pathology (p) T3-4 stage compared to pT1-2 stage [164]. Within BC tumors, these five microbes- A. seifertii, Devosia spp. strain LEGU1, Achromobacter deleyi, Ancylobacter pratisalsi, Microcella alkaliphila- can help to distinguish between immuno-enriched and immuno-deficient phenotypes of BC, with the immuno-enriched phenotype being associated with better prognosis [165]. In liver cell carcinoma tissues with Child-Pugh scores of Child B and Child C, levels of Sphingobacterium, Aeromicrobium, and Erysipelotrichaceae exceed Child A [77]. Moreover, Bacillus and Corynebacterium, which are genera of bacteria, significantly influence the T stage of nasopharyngeal carcinoma. Notably, there is a greater bacterium abundance observed in T1 and T2 tumors, suggesting they may function as safeguarding elements in tumor progression. Consequently, these bacterium’s presence is positively correlated with a more favorable prognosis for patients [166].
Intratumoral microbiota can serve as diagnostic biomarkers of early-stage cancer. Compared with the microbiota in the gut and blood, intratumoral microbiota have a very close relationship with the living environment of tumor cells, which can provide a more accurate reflection of the actual status within tumor. Due to the difficulty of obtaining tissue samples, along with the susceptibility of specimens to contamination, the accuracy of the microorganisms detected is questionable. Therefore, it is possible to combine the hematological and spatial distribution of the microbiome, and then use multi-omics data analysis to improve the accuracy of cancer diagnosis [16].
Intratumoral microorganisms can also reflect tumor prognosis. In PDAC, tumor microbiome of patients with long-term survival had higher α-diversity, and higher abundance of Pseudoxanthellas, Streptomyces, Glucospora saccharopolysporus, and Bacillus claueri. Thus, intratumoral bacteria α diversity can serve as a prognostic indicator to predict existence outcomes of PDAC patients [104]. Anaerobic bacteria’s presence like Bacteroides, Gasophilus and Lactobacillus in PDAC is positively tied to shorter survival time [76]. Surprisingly, it was found that individuals with long-term survival of PDAC had a great number and quality of microbial epitopes’ neoantigens [167]. In STAD, Kytococcus sedentarius and Actinomyces oris can cause DNA methylation, which is a crucial element in the distant metastasis and deterioration of tumor cells [17]. Treponema and Prevotella genera occur more frequently in positive lymph nodes group of esophageal cancer. Streptococcus is more abundant in T3-4 than in T1-2. The abundance of Streptococcus and Prevotella is positively correlated with unfavorable survival rates and could function as an independent prognostic marker for ESCC [168]. Moreover, lactic acid bacteria (LAB) in ESCC were shown to significantly worsen patients’ outcomes. The amount of LAB had a positive correlation with the Shannon index, while the Shannon index exhibited a negative correlation with the proportion of NK cells within TME. This suggests that LAB may promote tumor development by suppressing the immune system [64]. On the contrary, high Streptococcus’s abundance may indicate extended disease-free survival (DFS) in ESCC [63]. Within ovarian cancer, tumors enriched with Phaeosphaeriaceae family or Phaeosphaeria genus are in relation to significantly decreased progression-free survival (PFS), with the median PFS decreasing from 498 days to 135 days [44]. However, when considering the intratumoral microbiota as cancer biomarkers, other clinical and pathological factors, like tumor subtype and age, must also be taken into account, considering their importance in the cancer diagnosis and prognosis [5].

Intratumoral microbiota and their metabolites on cancer therapy

Impact of intratumoral microbiota on tumor immunotherapy
There is a strong link between intratumoral microorganisms and the response to immunotherapy, especially to ICIs [136]. Melanoma patients who responded to ICI were found to have higher intratumoral Clostridium content and lower Gardnerella vaginalis compared with non-responders [48, 169]. In metastatic melanoma patients unresponsive to ICI therapy, there is a marked enrichment of bacteria such as Ruminococcus and Fusobacterium [44]. Moreover, Fn colonized in colorectal tumors can inhibit the TME, leading to immunosuppression and greatly reducing tumor response to ICI therapy. Following this discovery, the Fusobacterium membrane (FM) of Fn was fused with colistin-loaded liposomes to construct a nanomedicine, which achieved the selective killing of intratumor-colonized Fn and enhanced the response of colorectal tumors to immunotherapy [170]. Moreover, Clostridium species induce tumor resistance to ICIs. A significant enrichment of intratumoral Clostridium was found among a subgroup of CRC patients who were unresponsive to ICI treatment [171]. Furthermore, Bifidobacteria can colonize CRC sites and activate DCs through the STING signaling pathway to alter the response to CD47-based immunotherapy [96] (Fig. 4).

Impact of intratumoral microbiota and their metabolites on tumor radiotherapy and chemotherapy
Different intratumoral bacteria exert distinct effects on the tumor response to radiotherapy and chemotherapy. Current research on the effect of intratumoral microbiota on radiotherapy and chemotherapy for tumors mainly focuses on the fact that, during tumor treatment, the involvement of microbes can lead to treatment resistance in tumors [172, 173]. For example, some intratumoral bacteria suppress the tumor’s response to chemotherapy. A higher load of Fn may result in inadequate responses to neoadjuvant chemotherapy in ESCC patients, and antibiotic intervention targeting Fn could potentially improve the chemotherapeutic response and improve patient prognosis [63]. In colorectal tumors, certain bacteria, involving B. fragilis, E. coli, Bifidobacterium breve, and Micrococcus, were found to confer resistance to the toxicity of 5-fluorouracil (5-FU), reducing its bioavailability and potentially diminishing the efficacy of the drug [174]. Similarly, Fn modulates autophagy-related proteins ATG7 and LC3, and the production of autophagosomes, to enhance resistance to 5-FU, docetaxel and cisplatin [175]. Moreover, E. coli producing colibactin inhibits the tumor immune microenvironment (TIM), thereby promoting chemoresistance in CRC [70]. Klebsiella pneumoniae enhances lung tumor’s resistance to gemcitabine (Gem) therapy [176]. Etoposide (ETO) is a principal drug for lung cancer, however, pulmonary microbiota reduce intratumoral concentrations of ETO through microbial transformation, thereby diminishing the efficacy of ETO therapy [177]. In Cervical squamous cell carcinoma and endocervical adenocarcinoma (CESC), LAB induces metabolic reprogramming by generating L-lactate, which in turn fosters resistance to chemotherapy [86, 178]. While little researches demonstrate that some intratumoral bacteria enhance the tumor’s response to chemotherapy. As an illustration, patients with a high abundance of intratumoral streptococci have longer DFS after neoadjuvant chemoimmunotherapy (NACI) [63].
When it comes to radiotherapy, some intratumoral bacteria suppress response to radiotherapy. For example, in CESC, LAB can enhance radiotherapy resistance [86, 178]. In animal experiments, oral administration of Lachnospiraceae can lead to elevated levels of butyrate in systemic circulation and in tumor of mice, thereby diminishing the efficacy of radiotherapy in the mice [179]. However, other microbes make a positive influence in radiotherapy. In regard to certain bacteria, combination of bacteriological therapy and radiotherapy can achieve better outcomes than radiotherapy alone [180–182]. Research showed that Bifidobacterium and its particular monoclonal antibody can alter the immune microenvironment of tumors, converting “cold” tumors into “hot” tumors and specifically target tumors and improve the hypoxic condition of normal cells within the tumor, thereby serving as a sensitizer for tumor radiotherapy [182]. It is worth noting that Bifidobacterium and its specific monoclonal antibody can also enhance the abscopal effect of radiotherapy [183]. Cytolysin A is a cytotoxin produced by certain intestinal bacteria, such as Salmonella typhi, can boost the effectiveness of radiotherapy by arresting tumor cell cycle at radiation-sensitive G2/M phase [181]. Recent finding indicated that after determining the bacterial species and abundance in the tumors of locally advanced rectal cancer (LARC) patients who achieved complete response (CR) following neoadjuvant chemoradiotherapy (nCRT), it was found that Ruminococcus genera can predict CR after nCRT in LARC, while Fusobacterium, Porphyromonas, and Oscillibacter can predict non-CR after nCRT [184]. Based on further study on the effect of the intratumoral microbiota on conferring resistance and sensitization to radiotherapy and chemotherapy, targeting the microbiota while patients undergo these treatments could serve as a novel adjuvant strategy in anti-tumor therapy.

Microbiota serve as carriers for treating tumors
Bacteria can be used to produce and deliver anti-tumor drugs by simple genetic manipulation or complex synthetic bioengineering [185]. The symbiotic interaction between tumor-resident bacteria and cancer cells can be used for tumor targeted therapy by fusing liposomes with the outer membrane of intratumoral bacteria to form a novel delivery system [170]. Microorganisms can be used as carriers to deliver substances including cytotoxic agents, cytokines, small interfering RNAs, and prodrug invertases, which can be carried to tumor tissues or lymph nodes to suppress tumor amplification and metastasis [186].

Improve clinical translation of intratumoral microbiota and metabolites
Intratumoral microbes and metabolites have a potential dual role in future therapeutic strategies, and further discovery is still needed [187]. Microbes are crucial in tumor treatment. We can enhance anti-tumor efficacy, especially immunotherapy sensitivity of tumors by regulating the variety and quantity of intratumoral microbiota. The use of bacteria-targeted therapy for tumors may be adjunctive to other anti-tumor therapies or even as a stand-alone treatment to treat tumors [185].
Although the chemotherapy drugs commonly used in the clinical treatment of tumors, they are easily metabolized with the blood flow or can not dwell in the tumor tissue over a lengthy span. If a “drug” can be made to work in a still area away from the blood vessels, it will focus on the tumor tissue to a greater extent, killing or inhibiting the tumor tissue without reducing the damage to normal tissue. While even if the microorganism can not be directly injected into the tumor, the therapeutic bacteria can spread into the tumor by intravenous injection [188]. The miraculous phenomenon was that the amount of bacteria that were initially delivered to the tumor and normal tissues by intravenous injection was similar [188]. Similarly, an interesting trend was that the bacteria present in the circulating and non-cancerous tissues were eliminated within hours and days, each of their own, but bacteria within tumor continued to multiply to much larger quantities than initial dose [189]. These studies suggest that some “therapeutic bacteria” have the property of selective colonization, possibly by the fact that tumor tissues have a unique immune microenvironment or a hypoxic and inflammatory environment [190]. Motile Salmonella is an excellent delivery system that can overcome the challenges of chemotherapy [191].
Currently, biomaterials derived from bacteria have been used in oncology therapeutics and diagnostics, the application of “engineered bacteria” is also under development [192]. Some studies suggest that “engineered bacteria” can be formed on the basis of natural bacteria according to simple genetic laws or modified by modern biotechnology. In addition to intrinsic anti-tumor effects, bacteria also induce immune responses within the TME [189, 193]. What’s more, “engineered bacteria " can accurately deliver anti-tumor drugs into tumors, activate anti-tumor immune as well as inducing apoptosis of tumor cells [194]. Compared with traditional chemoradiotherapy and immunotherapy, engineered bacteria act on the TIM, gather in TME and pinpoint the tumor without harming surrounding normal tissues. Moreover, engineered microorganisms have the advantage of better permeability in tumor tissues, which can better spread in tumor tissues and inhibit cell proliferation to a greater extent [195]. Another advantage of “engineered bacteria” is that they are not susceptible to attack by the human immune system after entering solid tumors, so they can continuously release anti-tumor drugs to achieve long-term inhibition of tumor progression.
Recently, it has been shown that imbalances in the microbial community can drive cancer development, and fecal microbiota transplantation (FMT) can reverse microbiota dysbiosis [196, 197]. In the days to come, the tumor microbiome could be sequenced to stratify patients before treatment. The potential for T cells that identify tumor neoantigens to also react with microbial antigens (a phenomenon known as mimicry) present within the tumor may open another avenue to explain the processes through which bacteria exert immune activation [104]. Delivering tumor-targeting bacteria and probiotics with innate anti-tumor properties into the TME in therapeutic effective doses is a fresh orientation of exploration for tumor treatment. To make a tight binding between bacteria and tumor cells, it is important to discover or design a tumor-specific binding protein [198].

Current challenges and future directions
With the popularization of chemoradiotherapy and the emergence of immunotherapy and targeted therapy, cancer therapy has achieved significant improvement, but there is still a lack of research on the effects of microorganisms and metabolites within tumors on cancer therapy. In addition, there are also some disadvantages of microbiota-based anti-tumor therapy itself, such as the high toxicity of microorganisms and their metabolites to non-cancerous cells, the insufficiency in combating tumors located deep within tissues, and the induction of anti-drug resilience within neoplastic cells [185]. Although the majority of bacteria employed in cancer therapy have been weakened and inactivated, clinical application still faces some problems and challenges, such as the potential risk of endogenous bacterial infection. Although there have been cases of tumor regression after bacterial infection in the 19th century, the use of bacteria to treat cancer has not been widely used due to the availability of patients and the limitations and side effects of bacterial therapy [199]. So far, weakened bacillus of Calmette and Guerin is the only internationally approved bacterium-based tumor treatment modality that can be applied clinically for individuals with bladder cancer who are at high risk. [200] Besides, low-biomass microbiome research, such as intratumoral microbiota, is fraught with technical challenges, primarily due to contamination risks and analytical artifacts. Environmental and reagent-derived DNA contamination - commonly referred to as the “kitome” - poses major technical hurdle [201]. Research emphasized the necessity of negative controls (e.g., blank extraction and PCR controls) to distinguish contaminants from true microbial DNA [202]. Formalin-fixed paraffin-embedded (FFPE) tissues may reduce microbial DNA yield and introduce environmental contaminants. Study showed that FFPE samples require specialized protocols to minimize false positives [203]. Meanwhile, bioinformatic decontamination tools have been developed to subtract background noise using negative controls. Thus, criteria are needed to establish causality in microbiome studies. Key criteria include concordant multi-omics evidence, orthogonal validation and functional rescue experiments. Techniques such as FISH (fluorescence in situ hybridization), qPCR, or single-cell RNA-seq can verify microbial presence and activity independently. Gnotobiotic animal models or microbial transplantation tests whether observed phenotypes are transferable. In conclusion, robust low-biomass microbiome research demands stringent contamination controls, transparent reporting, and causal inference frameworks. Microbiota within tumors contribute significantly to cancer occurrence and progression. In the future, they might serve as diagnostic and prognostic indicators for cancer in clinical settings and provide more targeted and efficient therapeutic approaches. The potential applications of intratumoral microbiota can be realized through the following means: First, further investigation into the interactions between intratumoral bacteria and their metabolites with tumors can help distinguish between pro-tumorigenic and anti-tumorigenic bacteria, as well as identify pathogenic bacteria enriched in different types of cancer, which can function to selectively alter the composition and plentifulness of intratumoral microbiota, enriching beneficial bacteria in tumors, screening out pro-tumorigenic bacteria and eliminating bacteria that induce drug resistance in tumor cells. Second, targeting the tumor-promoting or tumor-suppressing pathways mediated by intratumoral bacteria and their metabolites can achieve therapeutic effects against tumors. Currently, the integration of intratumoral bacteria with conventional radiotherapy, chemotherapy, and immunotherapy represents an emerging and feasible anti-tumor strategy. Particularly in the field of immunotherapy, the link between intratumoral microbiota and immune cells or pathways holds significant promise for cancer treatment. Moreover, intratumoral microbiota can be used as vectors or engineered strains to directly exert anti-tumor effects, either by directly implanting microorganisms into tumor tissues or by delivering drugs to tumor tissues via these microorganisms. Combining intratumoral bacteria with chemotherapeutic drugs to deliver them into specific tumor tissues, where they exert anti-tumor effects, provides a cutting-edge therapeutic approach for treating deep-seated tumors. Additionally, through genetic modification of engineered bacteria, the genes that are originally capable of harming normal cells are knocked out. This further enhances the safety of intratumoral bacterial therapy for tumor treatment. Intravenously infusing microorganisms capable of colonizing within tumors is a very promising research direction and the ability of microorganisms to disseminate through blood vessels and their efficiency in penetrating tumor tissues are important issues that need to be addressed [204]. In summary, intratumoral bacteria hold a very promising future in the diagnosis and treatment of tumors, offering renewed hope to cancer patients and warranting further exploration and investigation.

Supplementary Information