Leveraging beneficial microbiome-immune interactions via probiotic use in cancer immunotherapy

The gut microbiome exerts its influence on both local and systemic immune responses through a sophisticated interplay shaped by various environmental factors (33, 34, 39). The gastrointestinal tract is lined with gut-associated lymphoid tissue (GALT), comprising a vast network of immune cells primed to respond to microbes. Unlike peripheral lymph nodes which are most of the time sterile and quiescent, GALT is constantly exposed to foreign antigens from both commensal microbiota and infectious pathogens (40). Mucosal addressin cell adhesion molecule-1 (MAdCAM-1) is a key endothelial adhesion molecule that directs lymphocyte trafficking into the gut by binding to the α4β7 integrin on T cells (41). Its expression on high endothelial venules is critical for Treg homing to the intestinal mucosa, maintaining immune tolerance and mucosal immune balance (41). Gut dysbiosis or antibiotic-induced downregulation of MAdCAM-1 increases systemic migration of Tregs, driving their accumulation in the TME and impairing responses to immune checkpoint blockade (42, 43). Restoration of gut MAdCAM-1, either by microbial or molecular intervention, preserves local Treg populations, reduces tumor infiltration, and enhances the efficacy of immune checkpoint inhibitors in preclinical models (43).

Through interactions among microorganisms themselves and between microorganisms and host immune cells, numerous metabolites and cellular components are produced that regulate immunity at both the local and systemic levels. At the local level, the microbiome plays a crucial role in maintaining the gut epithelial barrier by promoting mucus production, enhancing tight junction integrity, and preventing pathogenic bacterial overgrowth (33, 34, 39). It also influences systemic immunity through key metabolites, short chain fatty acids (SCFAs), tryptophan derivatives, and bile acids, which regulate both innate and adaptive immune responses (34, 44, 45). They have been shown to support gut barrier integrity and exert anti-inflammatory effects (33, 34, 39). Additionally, microbial structural components known as microorganism-associated molecular patterns (MAMPs), such as lipopolysaccharide (LPS), formyl peptides, and peptidoglycan, engage host pattern recognition receptors (PRRs), initiating immune signaling cascades that contribute to both immune activation and immune regulation (34). Beyond the local effects in the gut, microbial metabolites exert systemic immunomodulatory effects through interactions in the GALT which supports the diversification, propagation, and possibly selection of systemic B cells (40).

SCFAs, such as acetate, propionate, and butyrate, are produced via bacterial fermentation of complex carbohydrates (e.g. dietary fiber) in the colon. These metabolites bind to G protein-coupled receptors (GPCRs) including GPR41, GPR43, and GPR109A, expressed on epithelial and immune cells (45). Activation of these receptors triggers intracellular signaling cascades that regulate immune cell activation, cytokine secretion, and inflammatory responses in the colon. SCFA-GPCR signaling supports Treg expansion while suppressing the differentiation of pro-inflammatory Th17 cells within the intestinal immune system (46). Beyond receptor-mediated signaling, SCFAs, particularly butyrate, also inhibit histone deacetylase (HDAC) activity, leading to increased histone acetylation, chromatin relaxation, and altered gene transcription (47–49). These epigenetic modifications further promote Treg differentiation through upregulation of FoxP3, a master transcription factor for Tregs (45, 50). In murine melanoma and pancreatic cancer models, microbial SCFAs and butyrate through HDAC inhibition, enhance the anti-tumor activity of CD8+ cytotoxic T lymphocytes leading to increased effector cytokine production and proliferation (51).

Tryptophan metabolites also play a central role in immune regulation, primarily through activation of the aryl hydrocarbon receptor (AhR) on the dendritic or T cells of the GALT. AhR signaling enhances the differentiation and proliferation of Tregs while simultaneously inhibiting pro-inflammatory Th17 cell development (52). In addition, AhR activation in dendritic cells promotes a tolerogenic phenotype, limiting their capacity to initiate inflammatory responses and instead supporting Treg expansion. Tryptophan metabolites also modulate cytokine production, downregulating IL-6, TNF-α, and IL-17, and upregulating IL-10 and TGF-β, thereby reinforcing anti-inflammatory immune responses (52–54).

Bile acids further exemplify the immunoregulatory capacity of microbial metabolites. Primary bile acids, synthesized from cholesterol in the liver, are transformed by gut microbiota into secondary bile acids (53). These metabolites bind to nuclear and membrane receptors, including Farnesoid X Receptor (FXR), Vitamin D Receptor (VDR) and Retinoid-Related Orphan Receptor gamma t (RORγt) in the lamina propria and Peyer’s patches, and G protein-coupled bile acid receptor 1 (GPBAR1 or TGR5) on gut macrophages (53, 55). Activation of FXR and VDR leads to suppression of pro-inflammatory cytokines and inhibition of NF-κB signaling. Simultaneously, inhibition of RORγt reduces Th17 cell differentiation, while stimulation of GPBAR1 promotes a shift in macrophage polarization toward an anti-inflammatory phenotype characterized by decreased expression of TNF-α, IFN-γ, IL-1β, IL-6, and CCL2 (55, 56). In murine colitis models, activation of GPBAR1 not only reprogrammed intestinal macrophages but also reduced the recruitment of circulating monocytes from the blood into the gut, demonstrating a direct systemic immune effect alongside local mucosal regulation (55).

Together, these pathways demonstrate that microbial metabolites influence the immune landscape through both receptor-mediated signaling and epigenetic regulation. By modulating the balance between pro- and anti-inflammatory cell populations and altering gene expression, the gut microbiome exerts both direct local effects and, via immune and cytokine trafficking, potentially far-reaching influence on systemic immunity and tumor sites. (Figure 1)

Over the past decade, a growing body of evidence has underscored the critical role of the gut microbiome in modulating the efficacy of ICIs (57–61). Notably, differences in gut microbial composition have been observed between patients who respond to ICIs and those who do not. Early studies reported that greater baseline gut microbiota richness and diversity were associated with improved responses to ICIs, including anti-CTLA-4 and anti-PD-1 therapies in melanoma and epithelial cancers, while antibiotic use was linked to diminished clinical benefit (58–61).

Further research has identified specific microbial taxa associated with favorable treatment outcomes, including bacteria from the Ruminococcaceae family, Akkermansia mucinophila, Alistipes indistinctus, Bifidobacterium longum, and Enterococcus faecium (58–61). Conversely, the presence of certain bacteria or an overall loss of microbial diversity (dysbiosis) has been linked to poor therapeutic responses (59, 60). For example, Enterocloster spp can modulate the migration of immunosuppressive T cell subsets into tumors via downregulation of MAdCAM-1 where they suppress anti-tumor immunity and promote resistance to immunotherapy (43). These findings underscore the promise of microbiome-targeted strategies to enhance cancer immunotherapy. Interventions such as probiotics, prebiotics, postbiotics, and FMT (Figure 2) are gaining traction for their potential to modulate the gut microbiome and improve treatment outcomes in cancer.

How to most effectively leverage the growing understanding of the microbiome’s effect on immunity in the design of therapeutics remains a subject of intense research. Several strategies are being explored, ranging from administering live beneficial bacteria (probiotics) and dietary components that selectively nourish them (prebiotics), to delivering bioactive microbial metabolites (postbiotics) and performing fecal microbiota transplantation (FMT) using material from ICI responders or healthy donors. The Food and Agriculture Organization (FAO) and the World Health Organization (WHO) define probiotics as “live microorganisms which, when administered in adequate amounts, confer a health benefit on the host.” (62) These beneficial microbes are sometimes also delivered through fermented foods like yogurt and cheese, as well as fermented fruits, and vegetables (63, 64). Frequently used probiotic strains include Bifidobacterium spp. and Lactobacillus spp., although other Gram-positive genera such as Streptococcus, Bacillus, Enterococcus, and various yeasts are also utilized (65). Besides treatment with live bacteria to alter the microbiome, other studies focus on supplementing dietary content of molecules that are metabolized by microbes (prebiotics) to produce downstream metabolites previously linked to a desired effects. Recent studies have demonstrated the role of inulin, a naturally occurring prebiotic fiber that can reinforce cancer immunosurveillance by activating γT cells, which play a crucial role in the anti-tumor immune response (66). Furthermore, inulin supplementation has been associated with increased production of SCFAs by microbes in the gut, which have in turn been shown to promote a favorable immune environment and improve systemic immune responses (67). Inulin supplementation was shown to inhibit tumor growth in melanoma and colorectal cancer mouse models by shifting gut microbiota taxa towards those that increase anti-tumor immune activity (68). Additionally, inulin has been shown to enhance the effects of various cytotoxic drugs. It can modulate the microbiota, resulting in increased fecal SCFA levels and improved systemic antitumor immunity in mouse models treated with anti-PD-1 antibodies (69, 70). In contrast, postbiotics are non-viable microbial cells or their components including structural fragments, peptides, and metabolites (SCFA, tryptophan, Bile acids) that can exert health benefits despite lacking viability. The International Scientific Association for Probiotics and Prebiotics (ISAPP) defines postbiotics as “a preparation of inanimate microorganisms and/or their components that confer a health benefit on the host.” (63, 65)

FMT, which involves transferring fecal material from healthy or ICI-responsive donors to patients, offers a comprehensive approach to restoring gut microbial diversity and function. Preclinical studies have shown that FMT from ICI responders enhances antitumor efficacy in germ-free mice and non-responder patients (59, 60, 71). Gopalakrishnan et al. reported improved outcomes in mice receiving FMT from melanoma responders, whereas transplants from non-responders failed to confer benefit (59). Similarly, Sivan et al. demonstrated that oral administration of Bifidobacterium spp. in melanoma and bladder cancer murine models improved tumor control by enhancing dendritic cell function and CD8+ T cell priming and infiltration in the tumor microenvironment (71). Early-phase clinical trials by Baruch, Davar, Routy and Kim have shown that FMT followed by anti–PD-1 therapy can overcome resistance in refractory melanoma, with sustained microbial changes and clinical benefit (20, 61, 72, 73).

Among these microbiome-targeted strategies, probiotics warrant particular attention because they represent a practical, scalable, and non-invasive approach to modulating gut microbial composition. Probiotics can be standardized, administered orally, and integrated into routine clinical care with relative ease.

Probiotics can modulate both innate and adaptive immune responses, notably by enhancing dendritic cell (DC) function, antigen presentation, cytokine production, and T cell activation within the tumor microenvironment. A key mechanism involves probiotic-induced activation and maturation of DCs, which improves MHC-I-mediated antigen presentation in lymphoid tissues like Peyer’s patches and lymph nodes, thereby increasing tumor-specific CD8+ T cells (71, 86–88). In a murine melanoma model, oral Bifidobacterium spp. administration led to expanded T cell populations in tumor-draining lymph nodes and increased IFN-γ production, likely driven by DC activation. This was accompanied by upregulation of genes involved in MHC-I presentation (Tapbp), DC maturation (Relb), chemokine signaling (Cxcl9), and type I interferon pathways (Irf1), promoting robust CD8+ T cell-mediated antitumor responses in the TME (71). Supporting this, another study identified 11 bacterial strains that enhanced T cell activation via DC–MHC I interaction (89).

Probiotics also regulate cytokine production. Commensal bacteria interact with DCs and epithelial cells in the gut through pattern recognition receptors, leading to the release of cytokines such as TNF-α, TGF-β, IL-12, and IL-10 locally (90). This gut-derived cytokine milieu induces maturation and activation of innate and adaptive immune cells, which can lead to systemic effects, including priming and recruitment of effector T cells in secondary lymphoid organs and ultimately the tumor microenvironment in murine models (86, 90, 91) Macrophage polarization represents another important axis of probiotic-driven immune modulation. Tumor-associated macrophages typically exhibit immunosuppressive (“M2-like”) polarization. Orally administered Bifidobacterium spp. and Bacteroides fragilis in mouse colon adenocarcinoma models, were showed to accumulate in the TME using bacterial tracing and cultures (92, 93). Furthermore, in the EG7 lymphoma mode, Bifidobacterium selectively accumulated within the hypoxic regions of the TME following systemic administration (93). Within the tumor, these bacteria are sensed by infiltrating macrophages and dendritic cells, which activate the STING pathway in response to bacterial DNA and signals. This activation triggers type I interferon production and polarizes macrophages toward the pro-inflammatory, antitumor (M1-like) phenotype (93–96). (Figure 3)

Probiotics also shape adaptive immunity through multiple mechanisms. Microbial-derived pathogen- and microbe-associated molecular patterns (PAMPs/MAMPs) can cross the gut mucosal barrier, reach lymphoid organs or tumor sites, and prime T cells for antitumor responses via molecular mimicry. Probiotic metabolites further drive immune activation; for instance, inosine produced by Bifidobacterium pseudolongum which was administered as a probiotic in murine cancer models, enhanced MHC-I expression and activates IFN-γ and TNF signaling, leading to increased CD8+ T cell proliferation and reduced PD-1 expression, thereby limiting T cell exhaustion in TME (97).

However, our understanding of how the potential effects of probiotics on the immune system translates to beneficial reprogramming of the tumor microenvironment and enhancement of ICI response remains incomplete. Most mechanistic insights are derived from murine models, which may not fully recapitulate human immune-microbiome interactions. In humans, evidence largely consists of correlative studies or small pilot trials, leaving uncertainty about causality, optimal strains, and dosing strategies. Key questions remain regarding the identification of the most effective probiotic strains, strategies to account for patient-specific microbiome variability, and the safety of administering live microorganisms in immunocompromised populations.

Some studies have suggested that probiotic supplementation can help restore microbiome composition, preserve gut barrier integrity, and prevent systemic immunosuppression, potentially maintaining or improving response to ICI therapy (61, 98–100). It is found that while ICIs can “release the brakes” on T cells, the presence of certain commensal bacteria appears to intensify this T-cell mediated attack. Commensal bacteria act as natural adjuvants, priming innate and adaptive immunity in the gut through microbial products that activate Toll-like receptors and inflammasome pathways, enhancing DC maturation and cytokine production to support T cell activation (61),.

Similarly, Bifidobacterium breve and Bifidobacterium longum have been shown to facilitate DC activation and T-cell priming. In mouse models of melanoma, oral administration of Bifidobacterium spp. significantly enhanced antitumor immunity to a degree comparable to PD-L1 blockade. When combined, the two treatments nearly eradicated tumors, suggesting a synergistic effect likely mediated by increased CD8+ T cell infiltration (71). Further mechanistic insights were provided by Vetizou et al., who found that the antitumor activity of CTLA-4 inhibition depends on specific Bacteroides species, including B. thetaiotaomicron and B. fragilis. Oral supplementation with B. fragilis in murine models induced Th1 immune responses in tumor-draining lymph nodes and promoted DC maturation within tumors. Even in germ-free mice, this intervention enhanced CTLA-4 blockade efficacy via IL-12–dependent Th1 responses (101). In human pancreatic cancer, T cells were shown to recognize both bacterial antigens and homologous tumor neoantigens (102), suggesting that if matching antigens were delivered through probiotic administration, it could potentially be used to prime or boost antitumor T cell responses via molecular mimicry. In the Bacteroides–CTLA-4 model, T cell responses against B. thetaiotaomicron and B. fragilis were correlated with tumor regression (101).

Akkermansia muciniphila has also been shown to strengthen the gut lining and help DC maturation, which leads to more CD8+ T cells entering tumors (25, 61, 84, 103). Lactobacillus rhamnosus and Lactobacillus reuteri can also influence T-cell differentiation and cytokine production, potentially reversing immune exhaustion in patients resistant to ICI (84). In a melanoma mouse model, daily oral administration of Lactobacillus reuteri alongside PD-L1 and CTLA-4 inhibitors significantly improved antitumor activity and survival (104). Lactobacillus reuteri was found to reach tumors outside the gut and reshape the TME into a more immune-activating state, characterized by IFN-γ production from CD4+ Th1 and CD8+ Tc1 cells. This was accompanied with increased immune cell proliferation and upregulation of CCL5 and CCL4 in CD8+ T cells, enhancing their recruitment into the tumor. These T cells also showed higher Granzyme B expression, reflecting greater cytotoxic potential (104).

In mice with colorectal cancer, supplementation with Lacticaseibacillus rhamnosus enhanced the effectiveness of ICI. This effect was driven by increased production of SCFAs particularly butyric acid, which promoted CD8+ T cell infiltration and activity while suppressing Tregs in the TME (105). Notably, butyrate generated by Faecalibacterium prausnitzii in murine colorectal cancer models, acts as a histone deacetylase (HDAC) inhibitor, leading to acetylation of histone proteins at the suppressor of cytokine signaling 1 (SOCS1) promoter (106). This upregulates SOCS1, dampens the JAK-STAT signaling pathway, and mediates tumor-suppressive and immunomodulatory effects in vivo in the TME (25).

Together, these preclinical studies (Table 1) demonstrate that probiotics may augment ICI responses through diverse mechanisms, including improved antigen presentation, modulation of cytokine signaling, restoration of microbial homeostasis, and production of immunoregulatory metabolites. However, while murine models provide critical mechanistic insights and proof-of-concept data, their predictive value for human efficacy is limited due to differences in microbiome composition, immune architecture, and reliance on germ-free or antibiotic-treated conditions. Consequently, translating these findings into clinical benefit requires rigorous validation through human correlative studies and well-designed clinical trials.

Several bacterial strains have demonstrated immunostimulatory properties that may potentiate the effects of cancer immunotherapy. These findings have led to the development of rationally designed probiotics, combining strains with complementary immunomodulatory effects to optimize immunotherapy outcomes. Several retrospective studies from Japan between 2020 and 2022 explored the impact of various probiotics in patients with non-small cell lung cancer (NSCLC). One of the earliest studies evaluated 118 patients with advanced NSCLC, 39 of whom received the Clostridium butyricum MIYARI 588 strain before or after ICI therapy. Probiotic use was associated with increased Bifidobacterium spp. abundance, reduced gut epithelial damage, and prolonged median progression-free survival (PFS) (250 vs. 111 days) and median overall survival (OS) (not reached vs. 361 days), even among those who had received antibiotics (107).

In a larger retrospective cohort, Morita et al. assessed 927 patients with advanced NSCLC. Among those treated with ICI monotherapy, commercially available multi-strain and single-strain probiotic use significantly improved PFS (7.9 vs. 2.9 months) and OS (not attained vs. 13.1 months). In patients receiving ICIs combined with chemotherapy, probiotics improved OS (not attained vs. 22.6 months) but had no significant effect on PFS (8.8 vs. 8.6 months) (108). Similarly, Takada et al. al analyzed 294 NSCLC patients receiving PD-1 inhibitors and found that probiotics (including Bifidobacterium spp., C. butyricum, and antibiotic-resistant lactic acid bacteria) were associated with improved disease control and longer PFS (HR 1.73 [1.42–2.11]), although OS was not significantly affected (HR 1.40 [1.13–1.74]). Subgroup analyses suggested greater benefit in never-smokers and patients with poor ECOG performance status (109).

One of the few prospective studies evaluating probiotics in lung cancer included 253 patients with advanced disease, of whom 71 received commercially available multi-strain probiotic formulations, most often including Lactobacillus acidophilus, Bifidobacterium lactis, Streptococcus thermophilus, Bacillus subtilis, Enterococcus faecium, Lactobacillus rhamnosus, and Bacillus licheniformi. While no significant difference in median progression-free survival (PFS) was seen among NSCLC patients (16.5 vs. 12.3 months, p = 0.56), patients with small cell lung cancer (SCLC) showed significantly improved PFS with probiotic use (11.1 vs. 7.0 months, p = 0.049), indicating potential histology-specific benefits) (110). This discrepancy from prior studies may reflect differences in patient baseline gut microbiome diversity, probiotic strain and dosing heterogeneity, or the influence of cancer histotype on microbiome-immune reaction. Another prospective study from China assessed the impact of antibiotics and probiotics on immunotherapy outcomes in patients with advanced EGFR-mutant NSCLC. Probiotic use did not influence PFS, but antibiotic use was associated with shorter PFS, suggesting that gut microbiome disruption from antibiotics may impair clinical benefit from immunotherapy. Additionally, responders to immune checkpoint inhibitors exhibited distinct metabolic profiles, with higher levels of deoxycholic acid, glycerol, and quinolinic acid, whereas non-responders had elevated L-citrulline (111).

Beyond lung cancer, an open-label single-center study by Dizman et al. assessed Clostridium butyricum (CBM588) in treatment-naïve metastatic renal cell carcinoma (mRCC) patients. Those receiving nivolumab–ipilimumab plus CBM588 had significantly improved PFS (12.7 vs. 2.5 months) and a numerically higher ORR (58% vs. 20%) (112). In a follow-up study of treatment-naïve mRCC patients receiving cabozantinib and nivolumab, the CBM588 group again showed superior ORR (74% vs. 20%, p = 0.01) and higher 6-month PFS (84% vs. 60%), indicating enhanced clinical activity (113). However, no differences in Bifidobacterium spp. abundance or alpha diversity were observed between arms, which suggests that while probiotic supplementation appears to enhance ICI efficacy, the absence of consistent biomarker changes means that the underlying mechanisms remain poorly defined. In a retrospective study of 352 patients with advanced digestive tract cancers (hepatocellular, colorectal, and gastric cancers) those treated with PD-1 inhibitors plus antiangiogenic agents and probiotics had a higher ORR (OR 2.4, 95% CI: 1.2–4.7; p = 0.013) (114).

A meta-analysis of six studies involving 1,123 patients with NSCLC, mRCC, and melanoma showed that probiotic supplementation with various strains was associated with a significant improvement in OS (HR: 0.527) and ORR (OR: 2.831), with trends toward improved PFS and DCR (85). Subgroup analysis in NSCLC patients receiving ICIs and probiotics demonstrated significantly longer PFS (HR: 0.532; 95% CI: 0.354–0.798; p = 0.002), improved OS (HR: 0.528; 95% CI: 0.306–0.912; p = 0.022), higher ORR (OR: 2.552; 95% CI: 1.279–5.091; p = 0.008), and DCR (OR: 2.439; 95% CI: 1.534–3.878; p < 0.001). An updated meta-analysis including 8 retrospective and 4 prospective studies with 3,142 patients further confirmed that probiotics significantly prolonged OS (HR: 0.58 0.54–0.81; p < 0.001) and PFS (HR: 0.66 0.54–0.81; p < 0.001), and improved ORR (OR: 1.75; 1.27–2.40; p = 0.001) and DCR (OR: 1.93; 1.11–3.35; p = 0.002). Notably, in NSCLC patients exposed to antibiotics, probiotics mitigated the negative effects on OS (HR: 0.45; 0.34–0.59; p < 0.001) and PFS (HR: 0.48; 95% CI: 0.38–0.62; p < 0.001) (16).

While these findings suggest a potential benefit of probiotic supplementation in enhancing ICI efficacy, most studies to date are retrospective, with small sample sizes and heterogeneous study designs (Table 2). Additionally, it is important to note that majority of clinical studies evaluated a wide range of commercially available probiotics, often multi-strain formulas or single-strain agents, without standardization. For example, in the studies by Morita et al. and Takada et al., patients received a heterogeneous mix of probiotic products such as Clostridium butyricum, Bifidobacterium spp., and antibiotic-resistant lactic acid bacteria (108, 109). Similarly, in the prospective lung cancer study by Tong et al., patients used diverse multi-strain formulations including Lactobacillus acidophilus, Bifidobacterium lactis, Streptococcus thermophilus, Bacillus subtilis, Enterococcus faecium, Lactobacillus rhamnosus, and Bacillus licheniformi according to product instructions, starting with immunotherapy and continuing until progression (110). In contrast, some studies, such as that by Ebrahimi et al., used a defined dosing regimen with CBM588 at 80 mg twice daily (113).

IrAEs are common toxicities associated with ICIs and are thought to reflect immune system overactivation. Probiotics have emerged as a potential strategy to modulate these responses (Table 3) (115, 116). Certain microbial signatures have been associated with either increased or reduced irAE risk. Firmicutes have been linked to higher irAE rates, whereas Bacteroidetes spp., Bifidobacterium spp., and Desulfovibrio spp. may have protective effects (117). In murine models, Bifidobacterium spp. supplementation reduced ICI-induced colitis by increasing Tregs and enhancing Lactobacillus abundance, while Lactobacillus resolved colitis by reducing activation of innate lymphoid cells (116). Findings such as these highlight the importance of rigorous clinical and translational studies to better understand the optimal application of probiotics.

Clinical data on probiotics and irAEs remain mixed. In the large retrospective study by Morita et al., probiotic use was not associated with a higher incidence of irAEs in patients on ICI monotherapy (108). However, among those receiving combination therapy with chemotherapy, increased rates of colitis and hypophysitis were reported. No significant associations were observed for other irAEs such as hypothyroidism, pneumonitis, hepatitis, or nephritis, though small subgroup sizes may have limited the findings. Other microbes, such as Faecalibacterium prausnitzii, have also been studied for their anti-inflammatory properties. In a murine model, F. prausnitzii alleviated ICI-induced colitis while preserving or even enhancing antitumor immunity, marked by increased tumor-infiltrating T cells and minimal systemic inflammation. Notably, patients who developed ICI-induced colitis had lower baseline levels of F. prausnitzii (118). A prospective study of 95 patients with gastrointestinal cancers treated with ICIs found that higher levels of Lactobacillus spp. and Bifidobacterium spp. correlated with fewer irAEs in gastric cancer (119). Similar findings have been reported in NSCLC, where higher abundance of these genera was associated with reduced incidence or severity of irAEs (117, 120, 121).

Mechanistic insights into probiotic and ICI interactions can be initially explored through ex vivo immune assays that replicate tumor–microbe–immune cell interactions in controlled settings. These assays involve co-culturing patient-derived immune cells, such as peripheral blood mononuclear cells or T lymphocytes, with tumor cells, followed by the introduction of selected probiotic strains or their metabolites (125, 126). This approach allows direct evaluation of how probiotics influence immune activation, including cytokine secretion and tumor cell cytotoxicity, outside the complexity of a living organism. Outputs such as T cell proliferation, cytotoxic activity, and cytokine levels (e.g., IFN-γ, IL-2, TNF-α) can be measured to identify probiotic strains with the strongest anti-tumor effects (126, 127). These ex vivo assays offer a rapid and informative readout of probiotic-driven immune modulation and help guide the design of follow-up in vivo studies.

Findings from cell-based assays can be expanded using animal models, particularly germ-free or gnotobiotic mice. Germ-free mice offer a sterile gut environment that allows for controlled colonization with specific probiotic strains or defined microbial communities. Researchers can then evaluate tumor responses to ICIs within this tailored microbial landscape (128, 129). Seminal studies have demonstrated that commensal bacteria can significantly influence the efficacy of immunotherapy. Sivan et al. (2015) found that mice colonized with Bifidobacterium spp. exhibited enhanced tumor control during anti–PD-L1 treatment (71). Similarly, Vétizou et al. (2015) showed that Bacteroides species were essential for optimal responses to anti–CTLA-4 therapy (101). More recent work has confirmed that fecal microbiota transplants from ICI-responsive cancer patients into germ-free or antibiotic-treated mice produce stronger anti-tumor effects than transplants from non-responders, emphasizing the microbiome’s role in modulating immune responses. Expanding on these findings, germ-free mice can be colonized with candidate probiotic strains to assess their influence on tumor growth, ICI efficacy, and immune cell infiltration in vivo (72, 130, 131). These models allow for mechanistic end-point analyses, such as measuring changes in intratumoral T cell activation or myeloid cell phenotypes in the tumor microenvironment (132–134). Overall, the use of germ-free mice colonized with defined microbiota offers a powerful and tightly controlled in vivo system to study the therapeutic and immunological impacts of probiotics in cancer immunotherapy.

Animal models can be complemented by advanced microphysiological systems that help bridge the gap between mouse and human biology. Gut-on-a-chip technology offers a high-fidelity in vitro platform that simulates the human intestinal ecosystem, enabling controlled investigation of microbiota and immune system interactions. These microfluidic devices culture living human intestinal epithelial cells under dynamic flow conditions and can support co-culture with live bacteria for extended periods (135). Gut-on-a-chip models replicate key aspects of the intestinal environment, including three-dimensional tissue architecture, mechanical forces similar to peristalsis, and oxygen gradients. This setup allows for long-term maintenance of both host cells and commensal microbes in close proximity (135). Incorporation of immune components, such as circulating or tissue-resident immune cells, enables the recreation of the intestinal mucosal immune interface. For example, placing peripheral immune cells in a basal channel to represent the lamina propria has demonstrated that exposure to bacterial molecules like lipopolysaccharide triggers cytokine responses (IL-8, IL-6, IL-1β, TNF-α) that closely mimic in vivo inflammation (135). This platform allows researchers to introduce specific probiotic strains or their metabolites to human gut epithelium under flow and directly observe resulting changes in immune cell activation, barrier function, and inflammatory signaling. Overall, gut-on-a-chip technology provides a physiologically relevant, human-based model to study probiotic and immune interactions with strong translational potential for clinical applications.

NGS has become an indispensable tool for uncovering molecular determinants that influence responses to ICIs.Beyond tumor genomics, NGS has also enabled detailed characterization of the gut microbiome to stratify patients based on ICI response. Shotgun metagenomic sequencing of fecal samples has revealed microbial composition and functional potential associated with therapeutic outcomes (136, 137). Notably, genome-resolved metagenomic approaches have uncovered baseline subspecies-level microbial signatures predictive of response, moving beyond strain-level analyses (137). Structural variants (SVs) within bacterial genomes offer an even more granular layer of microbiome heterogeneity, with recent analyses identifying SVs associated with ICI responsiveness, survival outcomes, and irAEs across seven clinical trials involving 996 patients (138). Building on these findings, integrative meta-analyses leveraging NGS data have led to the development of Gut OncoMicrobiome Signatures (GOMS), a standardized framework for identifying reproducible microbial biomarkers across diverse patient cohorts and therapeutic regimens (28). Together, these NGS-enabled approaches provide a comprehensive view of host and microbial factors that shape ICI efficacy, supporting their continued integration into precision oncology strategies.

Because gut microbiota and host immune interactions span multiple biological layers, a multi-omics approach is essential to fully understand the mechanisms underlying probiotic and ICI interplay. Integrating metagenomics (microbial composition and genes), metatranscriptomics (gene expression), metabolomics (small-molecule metabolites), and proteomics (protein signaling) enables comprehensive mapping of how probiotics influence the host–microbe ecosystem (139, 140). Recent multi-omic studies in immunotherapy patients have identified specific microbial taxa and metabolites linked to treatment outcomes. SCFAs and tryptophan derivatives, such as indoles, have emerged as candidate biomarkers associated with improved ICI responses (139). Combining microbiome sequencing with host immune and metabolic profiling can help pinpoint molecular mediators of probiotic effects. For instance, analyzing stool metagenomes alongside cytokine levels or immune transcriptomes from blood or tumor tissue has revealed associations between specific microbes and metabolic pathways that enhance T cell infiltration into tumors (141).