Modulatory role of Akkermansia muciniphila in cancer pathophysiology: a systematic review and meta-analysis of preclinical studies
- Open Access
- 16.01.2026
- Review
Erschienen in:
Abstract
Introduction
The gut microbiota, comprising diverse microorganisms such as bacteria, fungi, archaea, and viruses, is crucial for maintaining host health. It impacts various physiological functions, including metabolism, immune regulation, and disease vulnerability [1]. The microbiome actively maintains homeostasis, generates metabolites, protects the intestinal barrier, and modulates immune responses [2]. An imbalance in the gut microbiota, known as dysbiosis, contributes to the onset of various diseases, including metabolic disorders, cardiovascular conditions, and cancer [3]. The growing recognition of microbiota’s involvement in disease pathogenesis has increased interest in understanding how specific microorganisms influence health outcomes and whether they can be harnessed for therapeutic purposes.
Akkermansia muciniphila (A. muciniphila) has become a potential therapeutic microorganism. This mucin-degrading bacterium, naturally found in the mucus layer of the intestinal epithelium, is beneficial for maintaining gut barrier integrity, regulating immune responses, and metabolic functions [4]. In particular, A. muciniphila has been shown to improve conditions such as metabolic and inflammation-related disorders by restoring gut barrier function and reducing systemic inflammation [5, 6]. It is also considered a next-generation probiotic due to its ability to regulate the gut microbiota and enhance host health [7].
Anzeige
Recent research has expanded the understanding of A. muciniphila’s effects, with increasing evidence indicating that this bacterium may also play crucial roles in cancer modulation [8]. Studies have examined its impact on various cancer types, including colorectal, gastric, and liver cancers, yielding results that are both promising and contradictory [8]. Some studies suggest that A. muciniphila protects against cancer by enhancing immune responses, promoting apoptosis in tumor cells, and reducing inflammation [9, 10]. Conversely, other studies have raised concerns that A. muciniphila may exacerbate tumor progression in specific contexts, potentially through its effects on the immune system or by promoting a favorable environment for tumor growth [11].
The complexity of the connection between A. muciniphila and cancer highlights the need for a comprehensive evaluation of its role across different types of tumors. While some evidence supports its potential as a therapeutic agent in cancer prevention and treatment, other findings suggest a more complex or context-dependent relationship. To address these uncertainties, this study aims to synthesize preclinical evidence from animal models to better understand how A. muciniphila modulates cancer progression and to assess whether A. muciniphila could be a potential therapeutic target in cancer management.
Materials and methods
This study adhered to the standards established by the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) and was registered in PROSPERO (CRD42024623508).
Search strategy
Relevant studies were identified by searching publications accessible as of July 2025 across the following databases: Embase, ScienceDirect, Web of Science, and PubMed. The search aimed at identifying studies assessing the impact of A. muciniphila on cancer in preclinical models, with a focus on mice, using the following keywords: (Akkermansia OR Akkermansia muciniphila OR A. muciniphila OR Akk) AND (Cancer OR Carcinoma OR Tumorigenesis OR Tumor) AND (Mice OR Mouse). Only articles written in English were included.
Anzeige
Inclusion and exclusion criteria
The following criteria were used for inclusion: (1) studies conducted in mice, (2) studies assessing cancer metrics, and (3) studies in which animals were supplemented with A. muciniphila, including live or heat-killed bacteria or derivatives like EVs, purified Amuc proteins, or its metabolites. Studies were excluded from the analysis if they contained insufficient information on cancer metrics or experimental design, or if they involved studies in which A. muciniphila was combined with other bacteria. Articles not presenting original research, non-research articles, reviews and meta-analyses, clinical trials, and books were also excluded.
Characteristics of included studies
The preliminary search produced 8,910 articles (published from 2000 to 2025). Following the article’s assessment based on the inclusion and exclusion criteria, 274 articles were selected for further evaluation. Of these, 30 articles investigated the impact of A. muciniphila on cancer in various mouse models. However, after a detailed review of the abstract and result sections, eight were excluded due to insufficient data, yielding 22 studies eligible for data extraction (Figure S1).
Table S1 presents the characteristics of the included studies, including the sex and age of the mice at the beginning of the study, the A. muciniphila strain used, intervention (dose and duration), and the overall findings. The included studies evaluated the impact of A. muciniphila on mouse models of intestinal tumors (ileum, cecum, and colorectum) [10, 12‐17], hepatocellular carcinoma (HCC) [18, 19], gastric cancer [20, 21], prostate cancer [22, 23], ovarian cancer [24], lung cancer [25, 26], melanoma [27] and breast cancer [28].
Heterogeneity was noted in many of the factors analyzed, which may be attributed to the small number of studies included. The Egger and Begg tests revealed little to no evidence of publication bias (Table S2). Sensitivity results detected outliers in the CI for some outcomes, which were removed from further analyses. Statistical analyses, including and excluding the outliers, are shown in Table S3.
A detailed description of the study’s characteristics, statistical, gut microbiome analyses, and the PRISMA checklist is provided in the supplementary data.
Results
A. muciniphila improves tumor metrics in preclinical models
The SMD and 95% CI of the tumor metrics, including tumor number, volume, weight, and size, are shown in Table S3. A. muciniphila supplementation significantly reduced tumor metrics in mice models of colon, intestine, cecum, stomach, liver, prostate, lung, breast, and ovarian cancers (Fig. 1). For tumor number, three studies showed an increase in A. muciniphila-treated mice, which did not alter the significance of the overall analysis. Additionally, significance was obtained regardless of whether the outliers were included for tumor number, volume, and weight, except for tumor size (Table S3).
Subgroup analysis based on supplement type (1 = live A. muciniphila, 2 = non-viable forms), dose (1 = ≤ 10⁸ CFU, 2 = ≥ 10⁹ CFU), and cancer type (1 = CRC, 2 = other cancers) revealed distinct patterns across tumor-related outcomes. For tumor number (Figure S2), significant reductions were observed in both subgroups receiving live and non-viable A. muciniphila, as well as in low-dose interventions (≤ 10⁸ CFU) and CRC. Non-statistically significant differences were observed for interventions using a high dose and for other cancer types. For tumor volume (Figure S3) and tumor weight (Figure S4), all examined subgroups showed significant improvements, indicating that the anti-tumor effects of A. muciniphila-based interventions are broadly effective across supplement types, dosage levels, and cancer types. For tumor size (Figure S5), significant effects were detected in both supplement types, with notable reductions particularly in the low-dose subgroup and in studies targeting non-CRC cancers. This suggests that tumor size may be more responsive to interventions in other cancer types beyond CRC.
Fig. 1
Effect of A. muciniphila on tumor metrics. Forest plot of included studies in mice receiving A. muciniphila for tumor metrics (number, volume, weight, and size). The size of the square represents the effect size for individual studies, with the horizontal lines denoting the 95% CI. Colors distinguish different tumor types, including CRC, gastric, liver, lung, breast, ovarian, prostate, and melanoma models. Overall pooled estimates are indicated by solid green diamonds, demonstrating significant reductions in tumor metrics (p < 0.05). These findings suggest that A. muciniphila consistently decreases tumor burden across diverse tumor models. Outliers were excluded for clarity of presentation
Anzeige
A. muciniphila enhances immune responses
Included studies measured immune responses related to cytotoxic CD8+ T cells, M1 macrophages, and natural killer T (NKT) cells in the tumor microenvironment (TME). Activated CD8+ T cells secrete IFNγ and the serine protease Granzyme B (GZMB), which are used as markers for this cell lineage. M1 macrophages exhibit a pro-inflammatory phenotype, secreting TNFα, IL-6, IL-1β, IL-12, reactive oxygen species (ROS), and nitric oxide (NO), all of which possess tumor-killing properties [29]. They also express surface markers like inducible NO synthase (iNOS), CD80, and CD86. NKT cells have suppressive and activating roles in tumor progression. For example, NKT cells expressing the receptor CXCR6 are attracted to the TME by binding to its ligand CXCL16. This interaction activates NKT cells to produce INFγ and recruit CD8+ T cells [30]. Measures included infiltration of these cells into the TME, expression of cytotoxic and pro-inflammatory cytokines in the TME, and serum.
The SMD and 95% CI of immune response factors are shown in Table S3. A. muciniphila supplementation exhibited nonsignificant increases in CD8+ T cell counts and IL-23, as well as significant increases in IFNγ and TNFα levels in the TME (Fig. 2). The intervention also resulted in a non-significant decrease in the M2 macrophage-associated cytokine IL-10. Eleven studies also measured the expression of the proliferation markers Ki-67 and PCNA, showing a non-significant reduction in the combined markers in A. muciniphila-treated mice (Fig. 2).
Subgroup analysis showed that CD8⁺ T cell infiltration was significantly increased in studies using non-viable A. muciniphila interventions, low-dose supplementation (≤ 10⁸ CFU), and in models of CRC (Figure S6), indicating enhanced cytotoxic immune responses. IFNγ expression was significantly upregulated in both live and non-viable interventions and CRC models (Figure S7). IL-10 levels were modulated significantly by non-viable interventions and in CRC models (Figure S8). TNFα expression was significantly affected by non-viable interventions, low-dose supplementation, and in non-CRC cancer models (Figure S9). Tumor cell proliferation markers were significantly reduced by non-viable interventions at a high dose (≥ 10⁹ CFU), demonstrating a dose-dependent anti-proliferative effect (Figure S10). Reduction for these markers showed a non-significant decrease in all cancers.
Fig. 2
Effect of A. muciniphila on immune responses in the TME. Forest plots summarize data from mouse studies evaluating the impact of A. muciniphila on immune cell activity and inflammatory mediators within the TME. The size of the square represents the effect size for individual studies, with the horizontal lines denoting the 95% CI. Different tumor types are color-coded (colorectal, liver, lung, gastric, breast, prostate). The pooled estimates, shown as solid green diamonds, indicate significant increases in IFNγ and TNFα (p < 0.05), whereas effects on CD8⁺ T cells, IL-23, IL-10, and proliferation were not statistically significant. Outliers were excluded for clarity
Anzeige
A. muciniphila modulates systemic inflammation
IL-6 and IFNγ serum levels, as well as spleen weight, indicated systemic inflammation. The SMDs and 95% CI are shown in Table S3. Treatments exhibited significant reductions in spleen weight (Fig. 3), while no effects were seen for IL-6 and IFNγ levels.
Subgroup analysis showed non-significant effects for serum IL-6 based on dose and cancer type (Figure S11). In contrast, spleen weight was significantly reduced by non-viable interventions, low-dose supplementation (≤ 10⁸ CFU), and in CRC models (Figure S12). These findings suggest that non-viable forms of A. muciniphila, particularly at lower doses, may contribute to systemic immune modulation reflected by reduced splenic hypertrophy, especially in CRC settings.
Fig. 3
Effect of A. muciniphila on systemic inflammation. Forest plots summarize preclinical studies evaluating systemic inflammatory markers and spleen weight in mice treated with A. muciniphila. Different tumor types are color-coded (colorectal, liver, lung, and breast cancer). Each square represents the effect size of an individual study, with the square size proportional to the study’s weight, and horizontal lines indicating 95% confidence intervals. The pooled effect estimates are represented by green diamonds, showing no significant effect on IL-6 or IFNγ. At the same time, the overall spleen weight is significantly reduced, as indicated by the solid green diamond (p < 0.05). Outliers were excluded for clarity
A. muciniphila abundance in colon cancer in humans and mice
Human and mouse studies with no intervention
A few included studies presented clinical data regarding the abundance of A. muciniphila in cancer patients (Table S4). Fan et al. [13] showed reduced A. muciniphila’s abundance in patients with colon adenomas and adenocarcinomas compared with healthy controls, which correlated with downregulation of toll-like receptor 2 (TLR2) and NLRP3. Similarly, Zhang et al. [14]. reported reduced A. muciniphila’s abundance in patients with adenomas and adenocarcinomas compared with patients with polyps. Similar tendencies were seen by Wang et al. [12] for ulcerative colitis, adenomas, and CRC, and by Li et al. [18] for NAFLD and HCC. Yu et al. [22] compared A. muciniphila’s abundance between castration-resistant prostate cancer (CRPC) patients who are resistant (Res) or sensitive (Sen) to androgen deprivation therapy (ADT). Importantly, ADT Sen patients have a better prognosis than Res patients [31]. This study showed that ADT Sen patients have a higher abundance of A. muciniphila than Res patients. Wang et al. [24] showed that ovarian cancer patients exhibit reduced A. muciniphila’s abundance compared with patients with benign tumors.
Similar reductions in A. muciniphila’s abundance were also observed in mouse models of CRC [12, 15, 16], HCC [18], and CRPC ADT Sen and Res [22], and ovarian cancer [24] compared to control mice (Table S4, mice microbiome). Jiang et al. [16] showed a reduction in A. muciniphila abundance during cancer progression (8 weeks) in the AMO/DSS model. The abundance of Romboutsia and Lachnospiraceae was upregulated, while Bacteroides, Prevotellaceae, and Fissicatena were reduced. Wang et al. [12] demonstrated that microbiome diversity and richness were reduced in the DSS colitis model, with over 50% reduction in A. muciniphila abundance compared with control mice. Ma et al. [15] also reported a reduction in A. muciniphila abundance in the AMO/DSS model, which was associated with a decrease in Muribaculaceae, Helicobacter, and Dubosiella, and an increase in Bacteroides and Parasutterella compared to controls.
Mouse studies using A. muciniphila intervention reporting Microbiome data
Several included studies reported microbiome analysis, confirming the upregulation of A. muciniphila by the intervention and reporting changes in microbiome diversity (Table S5). Zhang et al. [14] demonstrated that A. muciniphila in the AMO/DSS model increased the abundance of A. muciniphila in the feces and remodeled the microbiome. At the genus level A. muciniphila increased Bacteroides, Prevotellaceae_NK3B31_group and Prevotellaceae_UCG-001 and reduced Turicibacter, Desulfovibrio, Alloprevotella, norank_f_Erysipelotrichaceae and Pareprevotella. Interestingly, Wang et al. [12]. demonstrated that Amuc_100, but not live A. muciniphila, increased the abundance of A. muciniphila in both control mice and the DSS model.
Anzeige
Cancer-associated microbial dysbiosis and functional shifts in the gut Microbiome
Utilizing publicly available microbiome data from included preclinical studies, we compared the gut microbiome profiles of cancer and control cohorts. Three studies were included in the comparison: one on NASH-associated HCC and two on CRC (Table S6) [12, 24, 32]. Substantial differences in the gut microbiome were observed between the cancer and control groups, particularly in terms of diversity and taxonomic composition. Microbial diversity, assessed by the Shannon index, was significantly lower in the cancer cohort, whereas no significant difference was observed with the Chao1 index. This suggests that while both groups harbored a similar number of microbial taxa, the cancer cohort exhibited a community structure dominated by fewer taxa (Fig. 4a). The microbial composition was profiled and compared between groups to further characterize these differences. The cancer cohort displayed a dysbiotic microbiome characterized by a significant enrichment of opportunistic pathogens such as Escherichia-Shigella and a notable depletion of beneficial commensals, including Alloprevotella, Alistipes, Odoribacter, and members of the Lachnospiraceae family (Fig. 4b, c). Although Akkermansia abundance was lower in the cancer cohort, the difference did not reach statistical significance.
A random forest classifier was constructed using genus-level abundance data, comprising 1,000 decision trees, to evaluate whether microbial features could distinguish between cancer and control samples. The model achieved an overall accuracy of 85.3%, exceeding the 50% baseline. Both cancer and control samples were accurately classified (AUC = 0.88 for both), with a clear separation evident in the class probability histogram (Fig. 4d and e). Among the top discriminative taxa identified by feature importance analysis, Alloprevotella ranked highest, followed by Dorea, an unclassified genus within the Atopobiaceae family, and Clostridium_sensu_stricto_1 (Fig. 4f). These results highlight pronounced alterations in the gut microbiome associated with cancer development, suggesting potential links between microbial dysbiosis, disease progression, and therapeutic opportunities to restore microbial homeostasis.
Computational functional predictions were performed to infer potential functional consequences of these taxonomic shifts (Fig. 4g). The cancer cohort demonstrated higher probabilities of Escherichia coli infection and bacterial invasion of epithelial cells, consistent with the observed enrichment of pathogenic taxa and loss of short-chain fatty acid (SCFA)-producing bacteria. Additionally, pathways related to genetic information processing were reduced in the cancer cohort. In contrast, catabolic metabolic pathways were enriched, unlike the predominance of anabolic pathways in the control group.
Fig. 4
Cancer-associated microbial dysbiosis and functional shifts in the gut microbiome. (a) Microbial alpha diversity. (b) Microbial composition at the phylum and genus levels. (c) Differential bacterial taxa between groups were identified using the linear discriminant analysis (LDA) effect size (LEfSe) algorithm (LDA > 2.0, p < 0.05). (d) Receiver operating characteristic (ROC) curves showing the predictive performance of the random forest classifier for control and cancer cohorts. Classification accuracy was evaluated by the area under the curve (AUC). The dashed line indicates the point of no discrimination. (e) Random forest class probability histograms for control (gray) and cancer (blue) samples. (f) Bar plots showing the top 20 genera most predictive of classification, ranked by relative importance scores. The right panel displays the mean relative abundance of these genera in each group, with error bars representing the standard error of the mean. (g) Differentially predicted KEGG pathways between control and cancer cohorts, identified by LEfSe analysis (LDA > 2.0, p < 0.05)
Anzeige
A. muciniphila treatment moderately mitigates cancer-associated gut dysbiosis
To evaluate the effects of A. muciniphila treatment on restoring gut microbiome dysbiosis associated with cancer progression, we compared the gut microbiome profiles of untreated cancer cohorts (Ctrl) and A. muciniphila-treated cancer cohorts (AKK) using publicly available preclinical microbiome data. Three studies were included in this analysis, involving CRC, ovarian, and liver cancer. One study administered pasteurized A. muciniphila, while the other two used live A. muciniphila (Table S6, intervention).
Microbial diversity was not altered by A. muciniphila treatment (Fig. 5a). The overall microbial composition differed markedly from that observed in the control versus cancer comparison; however, Proteobacteria remained the dominant phylum in the cancer cohorts, consistent with a persistent dysbiotic state (Fig. 5b). At the genus level, the most abundant taxa included an unclassified genus of the Xanthobacteraceae family, Vibrionimonas, and Rhodanobacter, all uncommon gut inhabitants that typically proliferate under oxidative and inflammatory conditions, further indicating persistent dysbiosis in the cancer group. A. muciniphila treatment induced only modest shifts in microbial composition while maintaining levels of Akkermansia. Changes at the genus level involved the enrichment and depletion of SCFA-producing bacteria, suggesting compensatory microbial adjustments rather than complete restoration of balance (Fig. 5b, c). Functional predictions revealed subtle metabolic shifts, characterized by reduced activity in riboflavin metabolism, bacterial secretion systems, and Vibrio cholerae pathogenic cycle pathways (Fig. 5d). These functional shifts likely reflect a reduction in oxidative metabolism and pathogenic activity, indicating partial restoration of gut microbial homeostasis following A. muciniphila treatment, although full recovery of the microbial ecosystem was not achieved.
Fig. 5
A. muciniphila treatment moderately mitigates cancer-associated gut dysbiosis. (a) Microbial alpha diversity. (b) Microbial composition at the phylum and genus levels. (c) Differential bacterial taxa between groups were identified using the linear discriminant analysis (LDA) effect size (LEfSe) algorithm (LDA > 2.0, p < 0.05). (d) Differentially predicted KEGG pathways between control and cancer cohorts, identified by LEfSe analysis (LDA > 2.0, p < 0.05)
Discussion
Despite the diversity of cancer models and study designs, interventions targeting A. muciniphila reduced tumor metrics. These reductions were associated with downregulation of proliferation markers (Ki-67 and PCNA in CRC, liver, prostate, stomach, and lung), increased infiltration of CD8+ T cells (liver, lung, colon stomach, prostate and ovaries), M1 macrophages (colon, stomach and prostate), and NKT cells (liver) and secretion of cytotoxic cytokines (IFNγ, TNFα) in the TME (Fig. 6). The interventions were also effective in improving systemic inflammation.
Anzeige
Several mechanisms were explored for the anti-cancer effect of A. muciniphila, including infiltration of cytotoxic T cells, macrophage polarization, apoptosis, activation of the NLRP3 inflammasome, improved intestinal health, and regulation of energy metabolism. The following sections discuss the proposed mechanisms for A. muciniphila’s actions.
Colorectal cancer
Conflicting data exist on the effects of A. muciniphila on CRC in preclinical studies, with some showing increased tumor growth [10, 33, 34] despite the reduced abundance of this bacterium in CRC patients. Differences in study design, including the animal model, antibiotic use, DSS concentration, and A. muciniphila treatment, explain these contradictory effects.
For example, the study by Dingemanse et al. [10]. was the only included study using an intestine-specific Apc mouse model that found treatment with A. muciniphila or Helicobacter typhlonius increased tumor numbers in the ileum, cecum, and colon [10]. The tumor area, however, was not different from that of the controls. This study was also unique in its use of clarithromycin and omeprazole in addition to amoxicillin and metronidazole. Other studies reporting elevated tumorigenesis used the AMO/DSS model and different dosing strategies. The study by Wang, F. et al. [33] used a higher DSS dose (3%), compared with others using 1–2%, and administered only four doses of A. muciniphila at the beginning of the intervention, which raises questions about the stability of colonization. Wang, K. et al. [34] reported that the tumor diameter was elevated, but not tumor number, with A. muciniphila administered throughout the entire intervention period, except during the weeks when DSS was administered. However, A. muciniphila abundance dropped sharply by week 12. In contrast, studies showing reduced tumorigenesis treated animals with A. muciniphila and its derivatives from the beginning to euthanasia [12, 13]. Bacterial pre-treatment and sustained treatment during cancer development appear optimal for stable bacterial colonization and anti-tumor effects.
A. muciniphila’s tumor-suppressive effects in CRC (Figure S13) involved tumor cytotoxic immune responses, increased apoptosis, and improved intestinal barrier.
Hsp70-induced activation of CD8+ T cells
Jiang et al. [16] identified the active tumor-suppressive component of EVs as Amuc_2172, a general control non-repressed 5 protein (GCN5) acetyltransferase. Amuc_2172 acetylates Lysine 14 in histone 3 (H3K14Ac) in the heat-shock 70 (Hsp70) gene, leading to Hsp70 upregulation. Hsp70 then activates CD8+ T cells, which secrete IFNγ and GZMB. The tumor-suppressive effects of Amuc_2172 were inhibited by preventing Amuc_2172 acetyltransferase activity, downregulating Hsp70, or inhibiting CD8+ T cell function.
NLRP3 inflammasome activation
Fan et al. [13] reported that live bacteria’s tumor-suppressive effects were mediated by NLRP3 inflammasome activation, which polarized tumor-associated macrophages (TAMs) towards a cytotoxic M1 phenotype (iNOS+), secreting TNFα, IL-23, IL-27, and IL-6. A. muciniphila binding (likely via Amuc_100) to TLR2 activates the NF-kB/inflammasome pathway. Tumor load reduction was linked to PCNA downregulation and decreased spleen weight. Interestingly, M1-like monocytes were upregulated in the bone marrow and circulation, suggesting that A. muciniphila induced the M1 switch in the bone marrow. Patient data revealed a positive correlation between A. muciniphila abundance and the expression of TLR2 and NLRP3 (Table S4, human microbiome).
Tryptophan/Kynurenine pathway reduction
Zhang et al. [14] identify the tryptophan/kynurenine pathways, which mediate tumor immune evasion by inhibiting T-cell activation [35], as a mediator of A. muciniphila effects. Kynurenine metabolites bind to the aryl hydrocarbon receptor (AhR), a ligand-activated transcription factor, that induces the expression of genes that inhibit T cell function and activate Treg cells. AhR can also activate β-catenin, regulating proliferation and tumorigenesis [36]. A. muciniphila was shown to reduce tryptophan and the kynurenine/AhR pathway in tumors.
Regarding A. muciniphila interventions, the outer membrane protein Amuc_1100 was sufficient to reproduce the effects of the live bacteria, suggesting that TLR2-mediated signaling is the primary mechanism by which A. muciniphila exerts its effects. For example, Wang et al. [12] reported that pasteurized A. muciniphila or Amuc_1100 reduced colitis symptoms and tumor growth by increasing CD8+ T cell recruitment and elevating TNFα expression while lowering the checkpoint programmed cell death 1 (PD-1), which helps tumor cells evade the immune system [37]. Both interventions also reduced the DNA damage/senescence marker H2AX, as well as the proliferation marker Ki-67, while promoting apoptosis. Amuc_100 was also shown to mediate its effects via TLR2, likely in dendritic cells, which activate cytotoxic T cells [27].
The proposed mechanisms for all cancer types are summarized in Fig. 6 and Table S7.
Fig. 6
Anti-tumor mechanisms of A. muciniphila across different cancers. A. muciniphila and its components (Amuc_1100, Amuc_2172, PEA, acetate, inosine) act on tumor microenvironments to promote CD8⁺ T cell activity, reduce Treg cells, and downregulate proliferation markers (Ki-67, PCNA). In colorectal, gastric, lung, prostate, liver, and ovarian cancers, these effects include modulation of NF-κB and AhR signaling, enhanced M1 macrophage activity, reduced glycolysis, and decreased tumor growth, collectively supporting anti-tumor immunity. The Figure was created using BioRender
Gastric cancer
Glycolysis Inhibition
Xu et al. [38] demonstrated that live, but not heat-killed, A. muciniphila enhanced the efficacy of oxaliplatin, a drug used in patients with advanced gastric cancer, by reducing Ki-67 and PCNA in tumor cells. The active metabolite pentadecanoic acid (PEA) binds to far upstream element-binding protein 1 (FUBP1), inhibiting its effects on glycolysis. FUBP1 is overexpressed in several cancers, promoting cell survival by upregulating hexokinase (HK) 1 and 2, and thereby enhancing glycolysis in tumoral cells [39]. Critically, oxaliplatin respondents had higher A. muciniphila and PEA and lower FUBP1 than non-respondents.
Liver cancer
Natural killer cell infiltration and bile acid metabolism
Li et al. [18] demonstrated that an A. muciniphila strain from human breast milk (AM06) reduced NASH scores and HCC progression, unlike a fecal strain (AM02). AM06’s mechanism involved increased infiltration of CXCR6 + NKT cells, enhanced CXCL16 expression, and reduced macrophage levels in the liver. CXCL16 is a ligand for the CXCR receptor and mediates NKT cell infiltration into the liver. A. muciniphila also altered hepatic bile acid metabolism, increasing primary and reducing secondary bile acids. Primary bile acids, such as chemo-deoxycholic acid (CDCA), can promote CXCL16 expression and NKT cell infiltration, which is required to reduce HCC [40].
While a limited number of studies have compared the effects of specific A. muciniphil strains on tumor metrics in preclinical models, other research has demonstrated strain-specific effects regarding inflammation, which are key drivers of tumorigenesis. For example, Liu et al. [41] compared the effects of three human strains (FSDLZ39M14, FSDLZ36M5, and FSDLZ20M4) with the ATCC BAA-835 strain in a mouse model of DSS-induced colitis. Interestingly, FSDLZ36M5 reduced colitis scores and intestinal inflammation, while FSDLZ20M4 exacerbated these outcomes. FSDLZ39M14 and BAA-835 showed no effects. Authors speculated that a set of genes mainly expressed by the FSDLZ36M5 strain could mediate the protective effects seen in the study.
CD8+ T cell infiltration and bile acid metabolism
Lan et al. [19] used A. muciniphila in combination with PD-1 antibodies, which are commonly used to treat various types of cancer, and found reduced tumor weight, PD-1 expression, and Ki-67, as well as increased CD8+ T cell infiltration and IFNγ, compared to PD-1 antibodies alone. A. muciniphila increased taurine-conjugated bile acids, such as taurorsodeoxycholic acid (TUDCA). Since this molecule has shown HCC tumor-suppressive properties, the authors speculated that TUDCA mediated the A. muciniphila’s anti-tumor effects.
Prostate cancer
M1 macrophage upregulation
The inosine, LPS, NF-kB, AR pathway
In a castration-resistant prostate cancer model with a compromised intestinal barrier, Yu et al. [22] demonstrated that live, not heat-killed, A. muciniphila reduced tumor metrics and Ki-67 levels and upregulated tight junction proteins. The effect was mediated by the metabolite inosine, which improved barrier function, preventing LPS transport into the prostate tissue and subsequent NF-κB activation and androgen receptor (AR) upregulation. AR activation by testosterone is critical for the initiation and progression of tumorigenesis. Thus, AR is a target for therapeutic interventions [42]. Low A. muciniphila increases the likelihood of progression to ADT resistance.
Lung cancer
Immune tolerance evasion (IL-10, Treg cells) and proliferation index
Chen et al. [26] combined A. muciniphila with cisplatin, showing a downward trend in tumor metrics in reducing Ki67, p53, and IL-10 in the tumor, while increasing IL-6, TNFα, and IFNγ in the serum. Although a reduction in tumor growth was not evident, the data are promising because survival rates are higher in patients with low Ki-67 and p53 [43]. IL-10 has immunosuppressive effects, altering the function of antigen-presenting cells; thus, in tumor cells, elevated IL-10 promotes immune tolerance [44]. The combined treatment also reduced CD4 + CD25 + Foxp3 + Treg cells. These cells play essential roles in regulating immune responses by preventing excessive immune system activation and the development of autoimmune diseases. These cells maintain immune activity by reducing cytotoxic CD4 + and CD8+ T cells. The treatment promoted cytokine-cytokine receptor interaction pathways and the JAK-STAT signaling pathway. This study was one of the few done in female mice. Yufen et al. [25] demonstrated that the protein Amuc_1100 inhibited tumor growth by recruiting and activating CD8+ T cells, as well as its secreted molecules (IFNγ, GZMB, and CCL5), an effect mediated by JAK/STAT inhibition.
Ovarian cancer
Acetate and intestinal barrier integrity
Wang et al. [24] found that ovarian cancer patients have 4-fold lower A. muciniphila levels compared with patients with benign tumors, along with reduced Blautia and Subdoligranulum, and increased Escherichia-Shigella, and Klebsiella. They generated an ID8 tumor-bearing mouse model treated with feces from patients with ovarian cancer in which A. muciniphila reduced tumor metrics, increased colon mucus thickness, and the expression of Muc2, Occludin, and Tjp1. The effect was primarily modulated by the metabolite acetate, which reduced tumor metrics and increased CD8+ T cell levels. A. muciniphila promotes mucus production via a positive feedback loop involving SCFA production (acetate), which stimulates goblet cells.
In addition to its direct anti-inflammatory effects, A. muciniphila reshapes the gut microbiota composition, further contributing to reduced inflammation. Ma et al. revealed that A. muciniphila significantly decreased the abundance of proinflammatory and potentially carcinogenic bacteria, such as Prevotella and Tannerellaceae, while increasing the abundance of health-promoting bacteria, including Akkermansia and Muribaculaceae [45]. The rise in Muribaculaceae abundance is crucial for enhancing intestinal barrier integrity, promoting mucus layer production, and activating innate immunity [46]. Interestingly, Muribaculaceae, a prominent bacterial family in mice, is inversely associated with epithelial cell apoptosis and IL-6 levels [47].
Microbial and functional signatures of cancer-associated dysbiosis and therapeutic potential of A. muciniphila
Microbiome changes in cancer progression have been widely studied for gastric cancer and CRC due to the direct interaction between cancer cells and the gut microbiome. However, other cancers, including lung, liver, prostate, and ovarian cancers, have also been explored [48‐53]. Common changes in the gut microbiome across various cancers included an expansion of pathobionts (Escherichia-Shigella), a loss of SCFA-producing commensals (Alistipes and Alloprevotella), and barrier dysfunction leading to systemic exposure to microbe-associated molecular patterns (MAMPs) such as LPS. Notably, Alloprevotella stood out as the most distinctive taxon differentiating the cancer from the control cohort. Alloprevotella is an SCFA-producing commensal, particularly generating propionate, and its abundance increases following dietary fiber interventions [54, 55]. With these characteristics, it has generally been considered a beneficial commensal; however, studies have also suggested potential associations between Alloprevotella and certain cancers, including gastric and oral cancers [56, 57]. Considering that Alloprevotella is prevalent in both the oral cavity and the gut [58], it may exhibit site-dependent behaviors. Prevotella has often been described as a double-edged genus due to its widespread distribution throughout the body and its numerous species, which enable it to function as a context-dependent taxon [59]. Further studies focusing on the specific roles of these taxa in the gut environment and cancer progression are needed.
According to our functional prediction analysis, housekeeping functions, particularly those related to genetic information processing, were reduced in the cancer cohort. Generally, these fundamental pathways are considered more stable or reliably inferred pathways [60], and their reduction has been observed in dysbiosis due to the loss of fast-growing commensals and as part of stress responses. Previous studies in brain (glioblastoma) and gastric cancer have reported decreased levels of genetic information processing associated with dysbiosis [61, 62], which aligns with our results and underscores alterations in core microbial functionality in cancer-associated dysbiosis. Collectively, these microbial and functional shifts plausibly amplify tumor-promoting inflammation and metabolic reprogramming, while also yielding microbiome fingerprints with diagnostic potential.
A. muciniphila administration showed limited ability to restore overall dysbiosis. Its anti-cancer effects may stem from its components (e.g., Amuc_1100, Amuc_2172, PEA) rather than its broad modulation of microbial ecology. Furthermore, Amuc_1100 and Amuc_1409 proteins have demonstrated health-promoting effects by enhancing gut barrier integrity and promoting intestinal stem cell proliferation [12, 63]. This observation implies potential limitations of A. muciniphila as a therapeutic agent, as continuous supplementation may be required to maintain its health benefits. However, this also represents a promising finding, as it suggests that therapeutic effects can be achieved without the need to administer live bacteria, thereby minimizing potential side effects associated with consuming live microorganisms.
The Microbiome and sex hormones
The gut microbiome has a significant influence on sex hormone levels, which can subsequently increase the risk of certain types of cancer. The term “estrobolome” refers to a group of gut bacteria capable of modulating estrogen detoxification and excretion [64]. Under normal conditions, the liver deactivates estrogens by conjugating them with glucuronic acid, allowing their excretion in the bile. Certain bacteria within the estrobolome express the enzyme β-glucuronidase, which breaks down the glucuronide conjugate. This process reactivates estrogens, allowing them to be reabsorbed into the bloodstream [65]. This enhanced estrogen recycling is particularly detrimental for estrogen receptor-positive (ER+) breast cancer in women [66] and men [67]. Furthermore, the estrobolome is linked to a higher risk of other estrogen-dependent malignancies, including uterine, ovarian, and cervical cancer [68]. This mechanism of β-glucuronidase-mediated estrogen recycling has also been confirmed in murine models [64].
The estrobolome is also implicated in prostate cancer, where an elevated estrogen-to-testosterone ratio serves as a key risk factor [69]. The prostate tissue expresses ERα, and its activation can trigger a glycolytic metabolic switch that promotes proliferation and carcinogenesis [70]. Consequently, the overlap signaling pathways of estrogen and testosterone are essential for the initiation and progression of prostate cancer. Beyond receptor activation, the oxidative metabolism of estrogen generates carcinogenic metabolites–specifically the catechol estrogen quinones–which damage the DNA by forming adducts and inducing mutations [71]. The production of these quinones is also linked to the generation of ROS, further exacerbating DNA damage.
In contrast to the proliferation effects seen in other tissues, estrogen signaling can be protective in certain contexts. For instance, in colorectal cancer, activation of ERβ by 17β-estradiol (E2) has been demonstrated to have tumor-suppressive effects, reducing inflammation and inhibiting the proliferation and migration of tumor cells [71].
Key bacterial genera within the estrobolom include several species of Bacteroides (e.g., B. Intestinalis, B. fragilis) and Clostridium (e.g., C. asparagiforme), as well as Escherichia coli, Alistipes, and Faecalibacterium prausnitzii [64, 65]. Although the included studies did not report hormone levels, the higher abundance of Escherichia coli observed in the cancer cohort in our analysis suggests that estrogen recycling may be upregulated. Consequently, further evaluation of how A. muciniphila might modulate these estrogen-related effects is warranted.
Regulatory status and authorization of A. muciniphila for human consumption
European union
The European Food Safety Authority (EFSA) published a Scientific Opinion in 2019, in which the EFSA Panel on Nutrition, Novel Foods, and Food Allergens evaluated the use of A. muciniphila as a novel food. The panel concluded that the pasteurized form of A. muciniphila, administered at a dose of 3.4 × 1010 cells per day, is safe as a food supplement in adults, excluding pregnant and lactating women [72]. This safety determination was given only for the pasteurized, not the live bacteria. The Panel also set a maximum dose of 1010 cells per day.
United States
In the U.S., the Food and Drug Administration (FDA) has determined that specific strains of pasteurized A. muciniphila are Generally Recognized as Safe (GRAS) and are safe for their intended use in food and supplements. The New Dietary Ingredient Notification (Nº 1326) was released on April 8, 2024. These safety determinations have enabled A. muciniphila to enter the U. S. market as medical foods or dietary supplements.
These regulatory approvals were supported by a 2019 pilot clinical study (Clinical Trial No. NCT02637115) that demonstrated oral supplementation with live or pasteurized A. muciniphila to be well-tolerated and safe in overweight and obese participants [73].
Conclusion
Our findings reinforce the notion that cancer-associated dysbiosis involves both taxonomic and functional alterations that favor the expansion of pathobionts and compromise commensal resilience. Our analysis suggests that bacterial components, rather than live colonization, may suffice to confer therapeutic benefits, pointing toward the potential of postbiotic strategies in cancer management. Future studies should move beyond taxonomic cataloging to integrate multi-omics and mechanistic analyses that unravel the causal pathways linking microbial metabolism, host immunity, and tumor biology. Translational efforts should focus on identifying bioactive molecules or structural components responsible for the effects of A. muciniphila and other commensals, paving the way for the development of next-generation postbiotic therapeutics. Ultimately, bridging microbial ecology with host molecular responses will be key to transforming these correlative observations into actionable microbiome-based interventions for cancer management. Although preclinical studies are promising, future clinical studies are needed to further confirm the safety, dosing, and long-term microbial colonization in the cancer ecosystem, as well as systemic effects.
Limitations
Not all the included studies confirmed the upregulation of A. muciniphila or assessed the intervention’s effects on microbiome composition. Confirmation of A. muciniphila levels in the gut in studies using the live bacterium was performed in only 7 studies [15, 19, 22, 24, 33, 34, 38], with one reporting a lack of bacterial colonization during the intervention [10]. The absence of statistical significance for some outcomes may be attributed to the limited number of studies. Another limitation is the limited number of studies involving females, which restricts the generalizability of the findings to both sexes and reduces the impact of our conclusions on females. Not all studies measured the full range of outcomes of interest, such as IL-1β in TME. Heterogeneity in study designs represents another limitation, as differences in mouse strains, tumor models, dosages, and duration of administration may have contributed to variability of results. Finally, the exclusive focus on preclinical models means that while these findings provide valuable mechanistic insights, their applicability to human populations remains uncertain. The relatively small number of included studies also limits statistical power, particularly for detecting smaller effect sizes or non-significant trends.
Acknowledgements
NA.
Declarations
Competing interests
The authors declare no competing interests.
Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/.
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
