Microbiome dysbiosis and immune checkpoint inhibitors: Dual targets in Hepatocellular carcinoma management

Abstract

Hepatocellular carcinoma (HCC), a primary malignancy of the liver and leading cause of cancer-related mortality worldwide, poses substantial therapeutic challenges, particularly in advanced and unresectable stages. Immune checkpoint inhibitors (ICIs) have emerged as critical therapeutic agents, targeting immune checkpoint pathways to restore antitumor immune responses. Combinations such as atezolizumab (anti-programmed cell death ligand 1 with bevacizumab (anti-vascular endothelial growth factor), as well as antibodies directed against cytotoxic T-lymphocyte associated protein 4 and programmed cell death protein 1 (e.g., ipilimumab and nivolumab), have demonstrated improved clinical outcome in selected patients. However, the overall efficacy of ICIs remains hindered by variable response rate and primary or acquired resistance. Recent evidence suggests that the gut microbiome plays a pivotal role in modulating host immune responses and may significantly influence the therapeutic efficacy of ICIs. Dysbiosis within the gut-liver axis has been implicated not only in pathogenesis and progression of HCC but also diminishing immunotherapy effectiveness. Emerging studies highlight the potential of microbiome-targeted interventions including dietary modulation, prebiotics, probiotics, and fecal microbiota transplantation to enhance ICIs responsiveness. This review explores the evolving interplay between the gut microbiota and immunotherapy in HCC, with a focus on microbiome-based strategies aimed at optimizing clinical outcomes.

Key Words: Hepatocellular carcinoma; Immune checkpoint inhibitors; Targeted therapy; Gut microbiome; Intervention-focused modulation

Core Tip: Combination of immune checkpoint inhibitor (ICI) therapy has improved efficacy and survival in patients with unresectable hepatocellular carcinoma (HCC). However, a substantial proportion still fail to respond due to resistance driven by the tumor microenvironment. Emerging strategies targeting gut microbiota offer promising avenues to overcome this barrier. Certain microbial taxa have been associated with enhanced T-cell infiltration and improved ICI responses. Additionally, microbial metabolites like butyrate and desaminotyrosine exert immunomodulatory effects that may restore sensitivity to ICIs. Dual-target approaches combining ICIs and microbiota modulation hold potential to improve progression-free and overall survival in HCC.

INTRODUCTION

Hepatocellular carcinoma (HCC) is the most common primary liver malignancy, accounting for 80%–90% of all liver cancer cases, and ranks as the third leading cause of cancer-related mortality worldwide[1,2]. Its global burden is particularly pronounced in Asia, where countries such as Thailand, Vietnam, and Cambodia report incidence rates of 22-24 cases per 100000, while Mongolia has the highest incidence globally, exceeding 80 per 100000[3]. Additionally, recent epidemiological shifts have shown rising incidence in previously low-burden regions, including Iran, Iraq, and Nepal[4].

Immune checkpoint inhibitors (ICIs) have emerged as a promising therapeutic strategy for HCC by enhancing antitumor immune responses[5]. While ICIs have improved outcomes in subsets of patients, their overall efficacy remains limited, with response rate (RR) ranging between 20%-30% and modest long-term survival benefits[6-8]. Combination regimens with other ICIs have demonstrated improved efficacy compared to monotherapy, but a significant proportion of patients still fail to respond[9,10]. This variability in response underscores the need for predictive biomarkers and novel therapeutic targets.

Recent studies have proposed the gut microbiome as a key modulator of ICIs efficacy, given its role in shaping systemic immunity and its association with cancer progression and therapeutic outcomes[11,12]. Alterations in microbial composition, particularly within the microbiome, represents a potential target for enhancing immunotherapeutic efficacy in HCC. This review examines the evolving relationship between gut microbiome and immunotherapy in HCC, with an emphasis on microbiome-informed strategies to improve clinical outcomes.

ICIs IN HCC

Patients with Barcelona-clinic liver cancer stage B or C HCC with sufficient liver function and performance status but can no longer undergo liver replacement therapy due to disease progression or extrahepatic spread, systemic therapy should be considered[13]. HCC demonstrates resistance to conventional cytotoxic chemotherapy through multiple intricate molecular mechanisms. These include autophagy activation, apoptosis evasion, upregulation of drug efflux pumps, heightened intracellular drug metabolism, and enhanced deoxyribonucleic acid (DNA) repair processes, among others[14]. As of 2023, first-line treatment for HCC consists of targeted therapies, including multikinase inhibitors, anti-vascular endothelial growth factor (VEGF) agents, ICIs, or their combinations[15].

Immune checkpoints are membrane-bound proteins found on various cell types, including natural killer cells, dendritic cells (DCs), monocytes, tumor-associated macrophages, myeloid-derived suppressor cells, as well as B and T-lymphocytes[16]. ICIs have exhibited prolonged effectiveness in multiple solid tumors, including HCC, where the tumor's immunological profile makes it a suitable target for immune-based therapies. Evidence suggests that the infiltration of lymphocytes within HCC tumors is positively correlated with clinical outcomes, highlighting the pivotal role of immune mechanisms in the therapeutic management of HCC. The strength of the immune response and the activation of cytotoxic immunity are governed by the balance between costimulatory signals and immune checkpoints[17,18]. These checkpoints play a critical role in regulating the immune response, making them key targets in therapies designed to enhance antitumor immunity. T-cells are the main focus of immune checkpoint therapy for three fundamental reasons: Their selective recognition of peptide antigens from proteins in different cellular compartments, their ability to destroy antigen-expressing cells through cytotoxic CD8+ T-cells, and their crucial role in regulating immune responses via CD4+ helper T-cells[19]. In individuals with chronic liver inflammation, intrahepatic lymphocytes show increased programmed cell death 1 (PD-1) expression, whereas its ligands, programmed death-ligand (PD-L) 1 and PD-L2, are highly expressed in Kupffer cells, liver sinusoidal endothelial cells, and leukocytes[20,21]. Currently, two main classes of ICIs are utilized in clinical practice for advanced HCC: Cytotoxic T-lymphocyte-associated protein 4 (CTLA-4) inhibitors and PD-1/PD-L1 inhibitors. These agents are typically recommended as second-line treatments for HCC in patients who do not respond to first-line therapy with sorafenib[22] (Table 1).

Table 1 Immune checkpoint inhibitors efficacy in hepatocellular carcinoma.

Study name	Regimen	Line of therapy	No. of patients	ORR (%)	mPFS (months)	mOS (months)	Findings
ICI single therapy
CheckMate 459[7,36]	Nivolumab vs Sorafenib	1st line	743	15	3.7 vs 3.8	16.4 vs 14.7	HR death 0.85 (95%CI: 0.72-1.00; P = 0.0522)
Keynote-224[9]	Pembrolizumab	2nd line	104	17	4.9	13.2	Durable anti-tumour activity and improvement in BOR. CR increased vs the primary analysis (3.8% vs 1.0%)
Keynote-240[8]	Pembrolizumab vs Placebo	2nd line	413	18 vs 4	3.0 vs 2.8	13.9 vs 10.6	Did not met the threshold HR of 0.781 (95%CI: 0.611 to 0.998; P = 0.0238) and 0.775 (95%CI: 0.609-0.987; P = 0.0186) for OS and PFS
Keynote-394[37]	Pembrolizumab vs Placebo	2nd line (Asian)	453	13.9 vs 1.3	2.6 vs 2.3	14.6 vs 13.0	Significance OS/PFS benefit (HR = 0.79; 95%CI: 0.63-0.99; P = 0.018)
HIMALAYA[38]	STRIDE (Durva + Tremeli) vs Sorafenib	1st line	1171	20.1	3.78 vs 4.07	16.4 vs 13.8	Significance OS (HR = 0.78; 96%CI: 0.65–0.92; P = 0.0035)
RATIONALE-301[39]	Tislelizumab vs Sorafenib	1st line	674	14.3 vs 5.4	2.3 vs 3.3	15.9 vs 14.1	OS non-inferior (HR = 0.85; 95%CI: 0.712-1.019)
Sangro et al[40] 2013	Tremelimumab	2nd line	21	17.6	NA	8.2	Median TTP 6.48 months (95%CI: 3.95–9.14)
ICI combination therapies
CheckMate 040[47]	Nivolumab + Ipilimumab	2nd line	148	32-31	2.96-4.0	22.8-12.5	Arm A, the 12-mOS rate was 61% (95%CI: 0.46-0.73)
CheckMate 9DW[46]	Nivolumab + Ipilimumab vs Sorafenib/Lenvatinib	1st line	1084	36 vs 13	9.1 vs 9.2	23.7 vs 20.6	Significantly improve OS (HR = 0.79, 95%CI: 0.65-0.96; P = 0.018)
IMbrave150[43]	Atezolizumab + Bevacizumab vs Sorafenib	1st line	336 vs 165	27.3 vs 11.9	6.8 vs 4.3	19.1 vs 13.4	HR death 0.58 (95%CI 0.42–0.79; P < 0.001)
AMETHISTA[44]	Atezolizumab + Bevacizumab (Single arm)	1st line	152	26.9	8.51	18.23	TEAEs in 28.9%
COSMIC-312[45]	Atezolizumab Cabozantinib vs sorafenib	1st line	837	NA	6.8 vs 4.2	15.4 vs 15.5	HR death 0.63 (95%CI: 0.44-0.91, P = 0.0012)

HR: Hazard risk; BOR: Best overall response; CR: Complete response; OS: Overall survival; PFS: Progression-free survival; TTP: Time to progression; TEAE: Treatment-emergent adverse events.

CTLA-4 inhibitor

CTLA-4 is an intracellular protein found in resting T-cells and is constitutively expressed on regulatory T-cells (Tregs). Its activation occurs exclusively within lymph nodes, where it plays a crucial role in modulating immune responses[23]. On Tregs, CTLA-4 is essential for inhibiting effector T-cell activity through multiple mechanisms. Furthermore, CTLA-4 plays an important role in regulating T-cell activation under normal physiological conditions, helping to prevent excessive immune responses[24]. CTLA-4 contributes to immunosuppression within the tumor microenvironment by improving the activity and differentiation of Tregs and disrupting the DC function[25]. Upon activation of the T-cell receptor by CD28, CTLA-4 is transported to the cell surface. Once expressed on T-cells, it binds to CD80 and CD86, thereby inhibiting their interaction with CD28, which is essential for the transmission of costimulatory signals. This interaction mediates inhibitory signals to T-cells, leading to the suppression of their proliferation and activation[22,26,27]. CTLA-4 signaling promotes tumor progression by inhibiting antigen presentation by antigen-presenting cells and stimulating the production of transforming growth factor-β, a critical immunosuppressive cytokine[28]. CTLA-4 inhibitors, such as ipilimumab and tremelimumab, have demonstrated that targeting this inhibitory pathway with antibodies can effectively restore T-cell activation and proliferation, thereby enhancing the immune response against tumors[29]. CTLA-4 inhibitor therapy has demonstrated significant clinical activity in advanced metastatic melanoma, with an objective RR (ORR) exceeding 15% and durable responses lasting over 10 years, even after treatment cessation[26,30,31]. In HCC, various immune response dysregulations and heightened inflammatory activity have been observed, including the overexpression of CTLA-4 and PD-L1, alongside an increase in immunosuppressive cytokines. This suggests that enhancing the immune response could potentially be a therapeutic strategy for HCC[32]. Anti CTLA-4 antibodies may enhance the anti-tumor immune response by targeting Tregs, which exhibit high CTLA-4 expression. This interaction potentially reduces Tregs within the tumor microenvironment, thereby fostering a more robust immune response against the tumor[24].

PD-1/PD-L1 inhibitor

PD-1 is an immune inhibitory receptor expressed on T-cells, B-cells, and others. It regulates immune responses through a bidirectional signaling mechanism. PD-L1, also known as CD274 or B7-H1, is broadly expressed on various somatic cells in response to proinflammatory cytokines and primarily functions to inhibit T-cell activity[33]. PD-L2 (CD273 or B7-DC), on the other hand, is less frequently expressed on APC. The binding of PD-1 to its ligands triggers apoptosis in antigen-specific T-cells within lymph nodes while concurrently inhibiting apoptosis in Tregs, thereby modulating immune tolerance and suppression[22,34]. Prolonged engagement between PD-1 and PD-L1 can amplify inhibitory signaling, fostering an immunosuppressive microenvironment. Cancer cells have adapted to exploit this pathway by continuously expressing PD-L1 or PD-L2, which activates PD-1 in tumor-infiltrating lymphocytes (TILs), enabling immune evasion[24,35]. Therapeutic antibodies targeting PD-1, such as pembrolizumab and nivolumab, or PD-L1, including durvalumab and atezolizumab, block this interaction, thereby mitigating immunosuppression and restoring an immunostimulatory tumor microenvironment[32].

EFFICACY OF ICIs IN HCC

ICIs have transformed the therapeutic landscape of advanced HCC. Various clinical trials have evaluated PD-1, PD-L1, and CTLA-4 inhibitors either as monotherapy or in combination, with results summarized in Table 1. Monotherapy trials with nivolumab (CheckMate 459)[36], pembrolizumab (KEYNOTE-224[9], KEYNOTE-240[8], KEYNOTE-394[37]), durvalumab (HIMALAYA)[38], tislelizumab (RATIONALE-301)[39], and tremelimumab consistently demonstrated ORR ranging from 15% to 20%, with manageable toxicity profiles. However, only KEYNOTE-394[37] achieved its prespecified overall survival (OS) and progression-free survival (PFS) endpoints, particularly in Asian populations. Despite early promise, the confirmatory CheckMate 459[36] and KEYNOTE-240[8] trials failed to meet statistical significance for OS improvement, leading to withdrawal of nivolumab’s Food and Drug Administration approval for second-line therapy in 2021. Tremelimumab also showed high disease control rate, highlighting its potential as a promising therapeutic option for HCC[40-42].

Dual-agent regimens have shown greater efficacy. The STRIDE regimen durvalumab combined with a single priming dose of tremelimumab improved OS over sorafenib with sustained long-term benefit, as shown in the HIMALAYA trial[38]. Likewise, the combination of atezolizumab and bevacizumab (IMbrave150[43], AMETHISTA[44] trials) led to a paradigm shift, becoming a first-line standard due to superior survival outcomes compared to sorafenib. The COSMIC-312 trial showed the combination of atezolizumab and cabozantinib significantly improved PFS. However, interim results showed no substantial OS benefit compared to sorafenib, pending the final survival analysis[45]. The combination of ipilimumab and nivolumab, studied in CheckMate9DW and CheckMate 040, also showed promising results in first-line and second-line settings, with exploratory analyses suggesting that “inflammatory signatures” may predict better outcomes[46,47]. Overall, these findings support the use of ICIs either alone or in combination, particularly in immunologically enriched tumors.

Although some studies have shown superior efficacy of ICIs, clinical responses are still variable. The resistance is still a major challenge, affecting both monotherapy or combination regimen. Anti-PD-1 monotherapy reaches RR of 42%-45%, whereas combination with anti-CTLA-4 boosted the RR to 58%. Nonetheless, resistance to PD-1 inhibitor monotherapy developed in approximately 55% of patients, and in 40% of those receiving combination regimens. Furthermore, 25% of initial responders demonstrated acquired resistance within a two-year period[48]. Likewise, combination therapy is more at risk of developing higher rates of grade 3–4 treatment-related adverse events[49]. Mechanisms underlying this challenge are tumor-intrinsic factors, like B2M mutations, downregulation of MHC-1 which impairs antigen presentations, and interferon-gamma (IFN-γ) pathway malformation through JAK1/2 mutations[50]. Novel strategies are necessary to overcome resistance by combining ICIs treatment with therapies that are able to modulate disease microenvironment[51].

RR AND ITS MECHANISMS OF ICIs IN HCC

Dysbiosis refers to an imbalanced state of the gut microbiome, often caused by antibiotics, diet, and infection[52]. This disruption may alter the immune signaling through the production of less-immunogenic lipopolysaccharide (LPS), thus disrupting inflammation pathways[53]. Studies have indicated the impact of dysbiosis on immunotherapy. A study using mice with antibiotic-associated dysbiosis failed to control tumor growth under anti-CTLA-4 immunotherapy[54]. In humans, a cohort study of advanced cancer patients treated with ICIs also revealed that administration of broad-spectrum antibiotics impacted both the RR and longer response time of ICIs[55]. In the context of HCC, dysbiosis may aggravate immune resistance by facilitating immune tolerance mechanisms. A key pathway involves Tregs differentiation which is recruited into the tumor microenvironment[56]. These Tregs can suppress cytotoxic T-cells activity, and impair anti-tumor immunity, contributing to immunotherapy resistance[57]. Moreover, Tregs enhance the expression of CTLA-4, a target for checkpoint blockage, which makes it susceptible to ICIs[58].

Many clinical trials about HCC revealed limited benefit and response to ICIs, highlighting HCC-specific resistance in many patients (Table 1). To date, ICIs-based therapy has indeed shown better outcomes compared to previous first-line standard of therapy, like sorafenib. However, long-term responses remain limited due to resistance mechanisms[59]. For instance, the combination of atezolizumab and bevacizumab showed an ORR of 34% in sorafenib-naïve patients, with only one patient gaining a complete response[60]. These suggest special resistance mechanisms in HCC which limit ICIs treatment. This includes the presence of cold tumor conditions characterized by low T-cell infiltration[61] and PD-L1 expression[62]. The cause of this condition is unclear, but it is thought to be linked to immune deficiency or innate immune response dysfunction that is related to T-lymphocyte rejection[63]. This immune dysfunction may create an unfavorable microenvironment that resists ICIs therapy. Given its role in immune regulation, gut microbiome modulations have shown their potential in improving ICIs efficacy. Microbiomes may convert cold into a hot tumor microenvironment by enhancing T-cell infiltration, thus reversing dysbiosis-associated resistance. This offers a novel strategy to reverse HCC-specific resistance in ICIs therapy[64].

GUT MICROBIOME, THE GUT-LIVER AXIS, AND ICIs IN HCC

The gastrointestinal tract constitutes the body's largest immune organ and is integral to maintaining physiological and immunological homeostasis. It harbors a diverse and complex community of microorganisms-comprising bacteria, archaea, eukaryotes, viruses, and parasites—collectively referred to as the gut microbiota[65,66]. This microbial ecosystem plays a critical role in modulating host immunity and has been increasingly implicated in liver health and disease progression through its bidirectional communication with the liver, a relationship defined as the gut-liver axis[67].

This axis is anatomically and functionally mediated by the portal vein, biliary system, and systemic circulation, allowing the liver to influence gut homeostasis via bile acids (BAs) and antimicrobial molecules, while gut-derived microbial metabolites and components, such as LPS, influence hepatic function[68,69]. Meta-analyses of 16s ribosomal-ribonucleic acid gene sequences have identified seven dominant phyla within the gut microbiota: Firmicutes, Bacteroidetes, Proteobacteria, Actinobacteria, Verucomicrobia, Fusobacteria, and Cyanobacteria[70,71].

Synthesized from cholesterol in the liver, BAs undergo extensive enterohepatic circulation and are modified by the gut microbiota through processes such as deconjugation, dehydroxylation, and epimerization[72]. This interplay regulates the composition of both BAs and microbial populations. However, dysregulated BA-metabolism-particularly the accumulation of hydrophobic BAs like cholic acid (CA) and chenodeoxycholic acid can trigger hepatocyte apoptosis through p38 mitogen-activated protein kinase signaling, promote chronic inflammation, and drive hepatic carcinogenesis[66,73] (Figure 1).

Figure 1 Gut-liver axis. The gut microbiota participates in various stages of bile acid metabolism, producing primary bile acids like cholic acid (CA) and chenodeoxycholic acid (CDCA), which are further transformed into secondary bile acids such as deoxycholic acid (DCA) and lithocholic acid. The buildup of hydrophobic bile acids, particularly CA and CDCA, can trigger apoptosis in hepatocytes by activating the p38 mitogen-activated protein kinase signaling pathway, further contributing to chronic inflammation. The farnesoid X receptor (FXR) is essential in safeguarding against liver cancer development. Pathological conditions, such as cholestasis, can lower FXR expression, intensifying the inflammatory response. Additionally, less-immunogenic lipopolysaccharide, produced by gut microbes, interact with toll-like receptor 4 on hepatic stellate cells (HSCs). This interaction activates HSCs, promoting liver fibrosis and facilitating the advancement of hepatocellular carcinoma. DCA is known to cause DNA damage by activating the nuclear factor kappa B signaling pathway, leading to the release of inflammatory cytokines, including interleukin-1β, interleukin-6, tumor necrosis factor-alpha, interferon-gamma, interleukin-8, and reactive oxygen species. These processes exacerbate liver damage and contribute to the progression of Hepatocellular carcinoma. CA: Cholic acid; CDCA: Chenodeoxycholic acid; DCA: Deoxycholic acid; FXR: Farnesoid X receptor; TNF-α: Tumor necrosis factor-alpha; IL: Interleukin; IFN-γ: Interferon-gamma.

Microbial dysbiosis further contributes to hepatocarcinogenesis through immune activation and the release of pro-inflammatory metabolites. Elevated LPS levels activate toll-like receptor 4 (TLR4) on hepatic stellate cells, stimulating fibrogenesis and accelerating tumor progression[74]. Additionally, secondary BAs such as deoxycholic acid (DCA) can induce DNA damage and activate pro-inflammatory pathways, including nuclear factor kappa B, thereby enhancing the secretion of cytokines such as interleukin-1β, interleukin-6, tumor necrosis factor-alpha, IFN-γ, interleukin-8, and reactive oxygen species (ROS)[69]. These processes contribute to a tumor-promoting hepatic microenvironment, particularly in obesity-associated HCC.

Beyond its role in liver carcinogenesis, the gut microbiome also plays a pivotal role in modulating the efficacy of ICIs. Specific microbial metabolites, including short-chain fatty acids (SCFAs) such as butyrate and desamino tyrosine (DAT), have demonstrated immunomodulatory effects that enhance ICIs responses. SCFAs can inhibit histone deacetylases, regulate mTOR-S6K signaling, and influence PD-1 ligand expression, thereby altering T-cell function and antitumor immunity[75-77]. DAT has been shown to potentiate type I interferon signaling, promote cytotoxic T-cell differentiation, and enhance the efficacy of anti-PD-1 and anti-CTLA-4 therapies[78,79].

Importantly, DAT also shifts microbial composition in favor of taxa such as Burkholderiales and Bacteroidales-bacterial orders associated with favorable ICIs responses[54,79]. In vivo studies confirm that DAT supplementation delays tumor growth and enhances immunotherapy outcomes, even in the presence of dysbiosis. Additionally, the concept of antigenic mimicry, wherein microbial peptides share structural homology with tumor antigens, has been proposed as a mechanism through which the microbiome enhances T-cell mediated antitumor responses. This has been demonstrated in both murine models and clinical studies[80].

Collectively, these findings position the gut microbiome as a critical determinant of immunotherapy responses. Its capacity to modulate host immunity, influence hepatic oncogenesis, and enhance ICIs efficacy underscores its potential as both a predictive biomarker and therapeutic target in HCC (Figure 2).

Figure 2 Mechanism of action of gut microbiomes and immune checkpoint inhibitors. Gut microbiomes will release metabolites, such as short-chain fatty acids (SCFA) and desamino tyrosine (DAT). SCFA works by entering T cells and inhibiting histone deacetylases (HDAC). Inhibition of HDAC can affect the mechanistic target of rapamycin-ribosomal S6 kinase pathway, which will express cytokines. It leads to the upregulation of programmed cell death 1 ligands and will enhance immunotherapy response. Furthermore, DAT expression may alter gut microbiome composition and balance dysbiosis. DAT also promotes type I interferon signaling, which will promote cytotoxic T-cell and cancer apoptosis, which eventually enhance immune checkpoint inhibitor effectiveness. SCFA: Short-chain fatty acids; DAT: Desamino tyrosine; PD-1: Programmed cell death 1; CTLA-4: Cytotoxic T-lymphocyte-associated protein 4.

ENHANCING ICIs OUTCOME

Microbiota dysbiosis plays a pivotal role in modulating the efficacy of ICIs. The presence of specific microbial taxa has been correlated with enhanced therapeutic responses to ICIs. Modulation of the gut microbiota can be achieved through various strategies, including fecal microbiota transplantation (FMT), administration of probiotics and prebiotics, dietary and lifestyle modifications, as well as the use of antibiotics (Table 2). Several studies have demonstrated the capacity of gut microbiota to enhance ICIs efficacy. For instance, Bifidobacterium longum, Collinsella aerofaciens, and Enterococcus faecium have been shown to exert immunomodulatory effects by downregulating Tregs, promoting the activation of Batf3 DC, and enhancing Th1 immune responses during PD-L1 inhibitor therapy[81]. Similarly, Bifidobacterium spp. has been reported to potentiate the local effects of anti-CD47 immunotherapy and to promote tumor-specific CD8+ T cell responses in the context of PD-L1 inhibition[82]. In addition, Bacteroides fragilis, Bacteroides thetaiotaomicron, and members of the Burkholderiales order have been implicated in driving interleukin-12–dependent Th1 responses during CTLA-4 inhibitor therapy[54]. Moreover, a combination of Lactobacillus rhamnosus and Escherichia coli Nissle 1917 has been shown to reshape the gut microbiome by increasing the abundance of beneficial bacteria such as Prevotella and Oscilibacter, which produce anti-inflammatory metabolites. These changes contribute to reduced Th17 polarization and promote the differentiation of anti-inflammatory Tregs/Tr1 cells within the gut[83].

Table 2 Gut microbiome modulation to enhanced immune checkpoint inhibitors outcome.

Ref.	Country	Samples (n)	Microbiota enrichment or microbiota modulation	ICIs	Outcomes
Zheng et al[138] (2020)	China	HCC with BCLC C (n = 8)	Firmicutes, Bacteroidetes, Proteobacteria dominated both in responder and non-responder	Camrelizumab (Anti-PD-1)	Several beneficial lactic acid bacteria, including four Lactobacillus species (L. oris, L. vaginalis, L. mucosae, and L. gasseri), Streptococcus thermophilus, and Bifidobacterium dentium, were significantly enriched, contributing to the support of host metabolism and immune function. Additionally, commensal bacteria enriched in responders—particularly members of the Ruminococcaceae family and Akkermansia muciniphila—promoted host health by preserving intestinal barrier integrity and mitigating systemic immunosuppression
Mao et al[113] (2021)	China	Unresectable HCC (n = 35) or advanced biliary tract cancer (n = 30)	Bacteroidetes, Proteobacteria, and Firmicutes dominated in both the clinical benefit response (CBR) group and the non-clinical benefit response (NCB) group	Anti-PD-1 (patients that progressed from gemcitabine plus cisplatin and the first-line chemotherapy)	The CBR group had more Bacteroidetes (P = 0.028), while Proteobacteria trended higher in the NCB group (P = 0.067). In HCC patients, Veillonellaceae was enriched in the NCB group and linked to poorer outcomes. In contrast, Erysipelotrichaceae bacterium-GAM147 and Ruminococcus callidus were associated with longer progression-free survival. These findings suggest that gut microbiota composition influences ICI response and survival
Shen et al[114] (2021)	Taiwan	Advanced HCC (n = 36)	Bifidobacterium, coprococcus, acidaminococcus	Anti-PD-1/anti-PD-L1 monotherapy or in combination with an immunomodulatory agent	Responders showed higher levels of Succinivibrio and Tyzzerella subgroup 4, while nonresponders had more Akkermansia. However, in patients with disease control, significant enrichment was observed in Bifidobacterium, Alloprevotella, Blautia, Megasphaera, Succinatimonas, Lachnospira, Acidaminococcus, Tyzzerella subgroup 4, and Coprococcus subgroup 3. Notably, by week 8 of ICI therapy, the increased abundance of Bifidobacterium, Acidaminococcus, and Coprococcus was no longer present
Lee et al[140] (2022)	Taipei	Unresectable HCC	Firmicutes and bacteroides	Nivolumab and pembrolizumab	Lachnospiraceae and Veillonellaceae were enriched in responders, while Prevotellaceae and Enterobacteriaceae increased in progressive disease with reduced Lachnospiraceae and Veillonellaceae. Secondary bile acids (e.g., UDCA, MDCA, tauro-UDCA, UCA) were higher in responders and correlated positively with Lachnoclostridium and Ruminococcus gnavus, but negatively with Prevotella 9. High Lachnoclostridium and low Prevotella 9 were linked to improved overall survival (median OS: 22.8 months)
Ahmed et al[55] (2018)	New York	Advance cancer (n = 60), HCC (n = 5)	Systemic antibiotic. Broad spectrum antibiotic (including cephalosporin). Narrow spectrum antibiotic (including vancomycin)	Anti-PD-1 or anti-PDL-1	Broad-spectrum antibiotics were linked to poorer treatment responses compared to narrow-spectrum antibiotics. Using antibiotics shortly before or after starting ICI therapy was associated with shorter progression-free survival. Overall, patients who received antibiotics had worse overall survival than those who did not

BCLC: Barcelona-clinic liver cancer; HCC: Hepatocellular carcinoma; PD-1: Programmed cell death 1; PDL-1: Programmed death-ligand 1; ICI: Immune checkpoint inhibitor; CBR: Clinical benefit response; NCB: Non-clinical benefit.

Epigenetic modulation by probiotics has also been implicated in the prevention of HCC pathogenesis. A synergistic formulation comprising Saccharomyces cerevisiae and Lactobacillus acidophilus, in combination with selenium and glutathione, has been shown to prevent carbon tetrachloride (CCl₄)-induced liver fibrosis through activation of silent information regulator 1 (SIRT1) in hepatocytes. Activation of SIRT1 attenuates ROS, endoplasmic reticulum stress, and inflammation associated with CCl₄ exposure[84]. Furthermore, probiotics such as Streptococcus thermophilus, Lactobacillus rhamnosus, and Weissella cibaria have demonstrated the ability to regulate aflatoxin metabolite contamination. Aflatoxins, which are secondary metabolites produced by Aspergillus flavus and Aspergillus parasiticus, are well-established carcinogens[85]. Aflatoxicosis has been shown to be potentially fatal and represents a major risk factor in the development of HCC[86]. The identification of next-generation probiotics, including Akkermansia, Bacteroides, and Faecalibacterium, has further expanded the therapeutic potential of microbiota-based interventions[87]. Other beneficial species, such as Propionibacterium acidipropionici and Propionibacterium freudenreichii, produce SCFAs including propionate and acetate, which contribute to tumor suppression by inhibiting tumor proliferation, enhancing antitumor immunity, and inducing apoptosis in tumor cells. Additionally, Lactobacillus casei has been reported to produce ferrichrome metabolites capable of directly inducing tumor cell death[88,89].

Inulin and fructo-oligosaccharides (FOS), classified as prebiotics, have demonstrated antitumor effects primarily through the modulation of gut microbiota. These prebiotics stimulate the growth of Bifidobacterium species, which has been associated with enhanced efficacy of ICIs therapy in murine cancer models[90]. In particular, inulin contributes to the modulation of gut microbiota composition, promotes systemic memory T-cell responses, and enhances anti-PD-1 efficacy, thereby improving ICIs treatment outcomes[91]. FOS has also been shown to suppress colonization resistance by Clostridium difficile. Members of Clostridium clusters XIVa and XI are known to facilitate the transformation of primary to secondary BAs, a process that has been implicated in hepatic carcinogenesis[92].

Dietary interventions have also been linked to improved ICIs responsiveness. The ketogenic diet, characterized by high fat content as the primary source of calories, has been shown to stimulate the production of pro-inflammatory cytokines, enhance CD8+ T cell-mediated cytolysis, increase infiltration of CD4+ T cells, and improve T cell cytotoxic activity. Moreover, it's found to down regulate ICs expression (CTLA-4 and PD-1/PDL-1) on TILs, thereby mitigating immune escape mechanisms. The ketogenic diet also activates AMPK, which promotes PD-L1 degradation[93,94]. Additionally, adherence to a ketogenic dietary pattern has been associated with increased abundance of beneficial gut microbes such as Akkermansia muciniphila, which has been shown to restore responsiveness to ICIs[95]. Furthermore, diets rich in dietary fiber can be fermented by gut microbiota to produce SCFAs, which exhibit anti-inflammatory and antitumorigenic properties. High-fiber diets also promote greater microbial diversity and richness in the gut ecosystem[96]. This dietary pattern contributes to a lower fecal pH, which reduces the proliferation of carcinogenic bacteria involved in BAs metabolism[97].

On the other hand, antibiotics also play a significant role in modulating the gut-liver axis and influencing HCC development. Vancomycin, a first-generation glycopeptide antibiotic, disrupts gram-positive bacterial cell wall synthesis[98]. It selectively depletes intestinal gram-positive bacteria, particularly members of the Lachnospiraceae, Ruminococcaceae, Bifidobacteria, and Clostridia families[99], thereby offering protection against HCC driven by secondary BAs. Secondary BAs have been implicated in establishing a pro-carcinogenic hepatic microenvironment[100]. Vancomycin also reduced liver cancer development by reducing bacteria expressing butyryl-CoA: Acetate transferase which are predominantly involved in SCFAs production in inulin-fed TRLR5- deficient mice. Butyrate, one of SCFA, is reported to have a double-sword effect, known as ‘butyrate paradox’. Under certain pathological conditions such as cholestasis, butyrate may exert pro-oncogenic effects and further exacerbate gut dysbiosis[101]. Moreover, vancomycin has been found to promote hematopoietic stem cells senescence by depleting Clostridium clusters XI and XIVa, which leads to a reduction in the concentration of DCA, a metabolite known to cause DNA damage and contribute to hepatocarcinogenesis[102]. In murine models overexpressing the MYC oncogene, the combined administration of vancomycin, primaxin, and neomycin significantly reduced both the number and size of HCC tumors[103]. Rifaximin is an antibiotic with no significant side effect on gut microbiome, acts improving intestinal permeability by inhibiting the LPS-TLR4 signalling pathway and suppressing portal endotoxin. Importantly, rifaximin has a low rate of antimicrobial resistance, making it a promising candidate for long-term therapeutic strategies in HCC management[104,105].

INTERVENTION-FOCUSED MODULATION

FMT

FMT is a therapeutic approach that involves transferring fecal material from a healthy donor to a recipient’s gastrointestinal tract, directly modifying the gut microbiota and potentially offering clinical benefits[106]. FMT can be administered orally through lyophilized or frozen capsules or directly via colonoscopy or gastroscopy[107,108]. Currently, no clinical trials have specifically examined the use of FMT in patients with HCC. However, preclinical studies using animal models have shown encouraging results. One study found that FMT from wild-type mice enhanced anticancer responses in HCC-bearing mice[109]. FMT also showed favorable effect to overcome ICIs-associated colitis and toxicity by modulation gut microbiota[110].

Probiotics

The gut microbiota plays a crucial role in tumor development. The administration of specific probiotics derived from gut microbiota has emerged as a strategy for microbiota manipulation[111]. Common probiotics, including Lactobacillus and Bifidobacterium species, are widely used to enhance the efficacy of immune ICIs, particularly in cases where antibiotics have disrupted gut microbiota. For example, Lactobacillus rhamnosus synergizes with ICIs, restoring microbial diversity and composition while increasing beneficial bacteria such as Bifidobacterium pseudolongum and Bacteroides[112]. Clinical studies on unresectable HCC patients undergoing anti-PD-1 therapy reveal that Ruminococcus calidus and Erysipelotrichaceae bacterium-GAM147 are more abundant in patients with prolonged PFS and OS. Conversely, higher Veillonellaceae abundance is associated with poorer outcomes[113]. Additionally, enrichment of Bifidobacterium, Coprococcus, and Acidaminococcus correlates with ICIs efficacy in HCC patients. Those gut microbiota were associated with disease control rather than objective response. The abundance of Bifidobacterium, Acidaminococcus, and Coprococcus significantly diminished within eight weeks following the initiation of ICIs therapy in patients without disease control. In contrast, patients not achieving disease control demonstrated a sustained abundance of these microbiota after immunotherapy[114].

Prebiotics

Prebiotics are a food source for microbiota. Fermentation of prebiotic can produce a favorable substance for preventing and reducing HCC. Specific dietary sources can provide prebiotics properties such as soluble fiber diets, including fructans, gums, pectins. Some plants like banana, garlic, chicory root, onions, asparagus, leeks, Jerusalem artichokes, rye, barley and wheat are non-digestible polysaccharide, highly contain both of FOS and inulin, which may have antitumor effect[102,115]. Non-digestible oligosaccharides found in nuts, tea, wine, vegetables, and fruit, contain prebiotic substances including flavonoids, phenolics acid and lignin. Non-digestible properties of these prebiotic allow them to reach the colon and ferment by colonic microbiota[116]. Plant polyphenols showed prebiotic effect against HCC by its immunomodulatory activity[117]. Study showed that a combination of FOS and raspberry polyphenol can enhance their effect on regulation of lipid metabolism in the liver[118].

Dietary intervention and lifestyle

Dietary intervention showed a favorable effect on HCC by preventing its risk factor. The most recently studied Mediterranean diet reported on reducing obesity, non-alcoholic fatty liver disease and type 2 diabetes mellitus. Furthermore, a population based study showed reduction in HCC incidence in liver cirrhosis patients who adhere to a mediterranean diet[119]. The Mediterranean diet contains low saturated fat and cholesterol, high monounsaturated fatty acid (MUFA) and polyunsaturated fatty acid (PUFA), complex carbohydrates, fibers, and polyphenols, increasing the number of Lactobacillus, Bifidobacterium, and Faecalibacterium[120-123]. However, high fiber contained in Mediterranean diet, particularly fermented fiber, might be unfavourable due to inducing dysbiosis, resulting in high BA production, and hyperbilirubinemia. This insight underscores the need for modification of the Mediterranean diet[123]. Diets containing MUFA and omega-3 PUFA have been found to reduce risk of HCC development[124-126] In animal study, high MUFA and PUFA diets gain a higher microbial diversity, including Bifidobacteria, which reportedly have protective effects against HCC[124,127]. In animal models with prostate and renal cell carcinoma, a low-protein diet in addition to anti-PD-1 amplifies the capacity of tumor-associated macrophages to eradicate tumor cells[95]. Compared with the low fiber diet, high fiber diet showed a higher effectiveness of anti-PD-1 on melanoma progression (OR = 5.3, 95%CI: 1.02-26.3)[128].

Antibiotics usage

Antibiotic administration in HCC patients undergoing ICIs therapy showing inconsistent results. A study of 414 HCC patients treated with anti-PD-1 monotherapy suggest that administration of beta-lactam or quinolones 30 days before or after ICIs initiation are associated with extended PFS[129]. In contrast, some studies reported that exposure to antibiotics in HCC with ICIs therapy demonstrates a worsen HCC outcome. A study consisting of 4100 patients with HCC found an unfavorable effect of antibiotic therapy[130]. Moreover, a study in Hong Kong involving 395 patients with HCC also reported higher mortality rate after antibiotic use during immunotherapy[131]. In 59 patients with advanced HCC received nivolumab suggest antibiotic administration shortening OS (P = 0.04) when compared to patients who did not receive antibiotics. Furthermore, the higher risk of death during ICIs therapy was linked with administration of antibiotic against anaerobes compared to antibiotic against aerobes[132,133].

CHALLENGES AND FUTURE DIRECTIONS

The rapid evolution of therapeutic options for HCC has created significant knowledge gaps, paving the way for future research[134]. One major challenge is determining the optimal management strategy following progression on ICIs-based therapies, now established as front-line treatments. Questions remain regarding the potential benefit of switching to alternative ICIs combinations for patients who fail to respond to initial regimens. Additionally, the availability of various front-line combinations highlights the urgent need for biomarker development to identify patients most likely to benefit from specific therapies[135]. Another critical area is the treatment of patients with advanced cirrhosis (Child-Pugh B or C), who constitute a large portion of the HCC population yet are often excluded from clinical trials[32]. While the phase II CheckMate 040 trial demonstrated that nivolumab was tolerable in patients with Child-Pugh B cirrhosis[7], future studies must clarify treatment approaches for HCC in this subgroup, where survival is more influenced by liver function than tumor burden[136].

Given the rising global incidence of HCC and the scarcity of actionable mutations, it is crucial to unravel host-intrinsic and host-extrinsic factors contributing to HCC and shaping therapeutic outcomes[1]. Predictive biomarkers that can guide treatment decisions remain elusive, making it unclear which patients will derive the most benefit from specific therapies. As the ICIs combination landscape expands, such as ICIs plus anti-VEGF or dual ICIs therapies, a deeper understanding of predictive biomarkers and precise patient stratification could significantly enhance treatment outcomes[32]. Recent biomarker analyses from trials of atezolizumab-bevacizumab have provided insights, showing that pre-existing immunity, VEGF receptor 2 expression, Tregs levels, and myeloid inflammation signatures correlate with better outcomes. Conversely, high Tregs/Teff ratios, glypican-3 expression, and elevated alpha-fetoprotein levels were associated with reduced benefit. These findings emphasize the complexity of biomarker-driven therapy in HCC[137].

Emerging areas of research offer promising avenues to improve HCC management. For instance, preliminary evidence suggests that the gut microbiota plays a critical role in modulating responses to ICIs, making it a promising target for enhancing the efficacy of immunotherapy in HCC[138]. The gut microbiota influences systemic immunity, with specific bacterial taxa associated with favorable responses to ICIs in preclinical and clinical studies. For instance, a higher abundance of Akkermansia muciniphila and Bifidobacterium species has been linked to enhanced anti-tumor immunity and better outcomes with ICIs. Combining ICIs with interventions targeting the gut microbiota, such as FMT, probiotics, prebiotics, antibiotics, or dietary intervention, may help optimize treatment responses. Recent studies have also explored the use of dietary modifications and microbiota-derived metabolites, such as SCFAs, to modulate the tumor immune microenvironment. However, further prospective studies are needed to better define the predictive role of the microbiota and validate its therapeutic potential in combination with ICIs for HCC[11,139,140].

CONCLUSION

The gut microbiome plays a pivotal role in regulating liver function, metabolic processes, and immune responses through its metabolites and bidirectional communication via the gut-liver axis. Certain microbial taxa, including Ruminococcaceae, Akkermansia muciniphila, Lactobacillus species, Streptococcus thermophilus, Bifidobacterium dentium, Erysipelotrichaceae bacterium, and Lachnoclostridium, contribute to maintaining intestinal barrier integrity, mitigating systemic immunosuppression, modulating responses to ICIs, and enhancing patient survival. Modulating the gut microbiome in conjunction with ICIs offers promising potential for improving HCC treatment. Integrating microbiome-targeted strategies with existing ICIs, such as anti-CTLA-4 and anti-PD-1 antibodies, may improve clinical outcomes.