Metabolic dysfunction-associated steatotic liver disease (MASLD) represents a burgeoning burden of liver disease, reaching a global prevalence of more than 30% [1,2]. This challenging landscape has been driven predominantly by the escalating prevalence of metabolic risk factors, such as overweight status, obesity, and type 2 diabetes mellitus (T2DM) [3-7]. MASLD encompasses a broad spectrum of clinical phenotypes ranging from isolated steatosis, different degrees of inflammation and fibrosis, to cirrhosis [7]. Although simple steatosis has been generally considered “benign”, a recent nationwide cohort study suggested that even isolated steatosis is associated with higher mortality [8]. It is difficult to estimate the economic burden of MASLD due to its underdiagnosis and complex relationship with other comorbidities [5,9]. However, the estimated healthcare costs of MASLD patients are nearly twice as high compared to age-matched counterparts without the disease, and are highest among those with advanced fibrosis and end-stage liver disease [10].

The human gastrointestinal tract harbors a complex and vast population of microorganisms composed of bacteria, fungi, viruses and archaea, outnumbering human host cell counts [11]. Knowledge of microbiota characteristics in metabolic disorders has substantially increased in the last few years, indicating that dysbiosis alteration in composition and functional capacities of the microbiota can impact the metabolic health of the human host [12-15]. The crosstalk between the gut and liver is increasingly recognized in the literature, strengthened by the prevalence of liver diseases and gastrointestinal/immune disorders. In particular, notable changes in microbiome composition can be observed in the natural history of MASLD (Table 1). For example, a study of 279 biopsy-proven MASLD patients evidenced that beta diversity was different according to metabolic dysfunction-associated steatohepatitis (MASH) presence, the relative abundance of F. prausnitzii was higher in controls, and the abundance of S. parasanguinis and S. salivarius was greater in the MASH cohort [16]. In addition, a recent systematic review suggests that the alpha diversity is reduced in MASLD, whereas Coprococcus, Faecalibacterium, and Ruminococcus decrease and Escherichia increase compared to healthy controls [17].

Due to the bidirectional relationship between the gut and its microbiota and the significant changes observed in the natural history of MASLD, therapies targeting the gut and microbiome could contribute to modulating steatohepatitis and fibrosis progression. Therefore, this review aims to critically evaluate the current landscape of microbiome-centered therapies in the management of MASLD, exploring the scientific rationale, clinical evidence, biomarkers, and potential mechanisms by which these interventions influence the disease trajectory. By integrating findings from recent clinical trials and experimental studies, we provide a comprehensive overview of the potential of these therapies, including prebiotics, probiotics, synbiotics, fecal microbiota transplantation, as well as promising therapies such as engineered bacteria, bacteriophage therapy, and other innovative strategies targeting the mycobiome.

Over 100 years ago, Hoefert et al. [18] first illustrated the concept of the gut-liver axis through identification of alterations in gut microbe composition in patients with chronic liver disease. With the subsequent recognition of impaired gut motility, intestinal permeability and bacterial translocation in patients with chronic liver disease [19-21], coupled with the role of dysbiosis in modulating metabolism, the need to investigate the role of altered gut microbiota in the development and progression of MASLD was appreciated [22,23].

Through a complex and deeply symbiotic relationship, the gut microbiome plays an important role in maintaining the immune and metabolic functions of the liver. Microbe-derived products such as D-lactic acid facilitate the elimination of circulating pathogens by Kupffer cells. Other metabolites, including lipopolysaccharide (LPS), support immune tolerance and xenobiotic metabolism, while sphingolipids derived from commensals can induce beta-oxidation, thereby reducing hepatic lipid accumulation. Murine models have revealed that the gut microbiome induces intense transcriptome changes in the liver sinusoidal endothelium, ultimately affecting expression of genes involved in metabolism, angiogenesis, and sphingolipid metabolism. As an example, upon sensing microorganisms, the sinusoidal endothelium signals for Kupffer cells and natural killer T cells to home towards the periportal regions. This targeted homing of the immune system towards the point of entry in the liver, protects against bacterial translocation.

While recent MASLD research has allowed us to identify implicative microbes and their metabolites, the following important questions remain unanswered:

1. How do microbial profiles compare in MASH versus MASH cirrhosis and beyond?

2.Does gut dysbiosis facilitate the development and progression of MASLD or is it simply a consequence of the disease, i.e., is it the chicken or the egg? How does this affect our ability to identify and interpret microbiome-related biomarkers?

3.Does microbial production of ethanol contribute to the development and progression of MASLD?

4.What is the contribution of host factors, including genetics, lifestyle, and concomitant disorders on microbiota composition, MASLD pathophysiology and efficacy of microbiome-centered therapies?

Patients with MASLD progressing to MASH and advanced fibrosis have been shown to have a different microbiota composition and function compared to those who are earlier on in the disease process. This is characterized by differential changes in alpha and beta-diversity, bacteria, fungal and viral composition, as well as functionality as assessed by bile acid profiles in the stool and circulation (Table 1). Moreover, the overlap between patients with and without a contribution of external alcohol on these disorders has unique impacts on the microbiota. As disease progresses to cirrhosis and its associated complications, the changes in microbiome become even more stark. With advancing disease, there is evidence of higher potential pathobionts (Enterobacteriaceae, Enterococcaceae), and lower relative abundance of commensal taxa. There are changes in the virome with lower diversity as disease progresses and mycobiome changes that track liver disease severity [24-31]. In addition to the microbial structure, specific functional aspects, including bile acids also change with liver disease progression. This includes differences in ratios of primary and secondary bile acids in the stool as well as the serum (Fig. 1) [32-34].

The tracking of liver disease severity with microbiome raises questions regarding the “chicken or the egg”. Current evidence regarding liver disease as a whole and MASH specifically, with respect to causation of disease, is not clear. In germ-free studies, depending on the model used, presence and quality of microbiome transferred can either enhance or inhibit liver injury progression [35-38].

In addition to liver disease severity, host genetics, immune health, lifestyle, socioeconomic, and geographical factors, as well as comorbid conditions, can affect the composition and function of the gut microbiome and plausibly, the efficacy of microbiome-centered therapies. Twin studies have shown higher similarity of gut microbiota between monozygotic twins versus dizygotic twins, illustrating the role of genetics on microbiota composition [39]. Furthermore, genetic variations in genes encoding innate immunity players such as C-type lectins, vitamin D receptor, NOD2, and TLRs (TLR2, TLR4, and TLR9 particularly) have been associated with gut microbiome composition change and altered function, including bile acid metabolism, gut permeability as well as development of steatosis in the liver [40,41]. Understanding the role of genetic variations and the superimposed effect of lifestyle factors such as diet, which independently facilitate gut microbiome alterations and liver fat deposition, is important to promote efficacy of microbiome-centered therapies. Microbiota-protecting diets, i.e., diets enriched in fibers that promote SCFA-producing bacteria and low in processed foods, can provide an ideal environment for therapeutic success. As an example, fecal microbiota transplantation (FMT) supplemented with propionate has been posited to help reduce weight gain and adiposity in obese patients and may be applicable to those with associated metabolic disorders, such as MASH [40]. Comorbid conditions (e.g., diabetes), and medications used to treat them, can also contribute to microbiome alterations. Specifically with metformin use, there is a higher relative abundance of beneficial Akkermansia, whereas SGLT2 inhibitors reduce the Firmicutes/ Bacteroidetes ratio, while pioglitazone reduces the relative abundance of pathobionts such as Escherichia and Salmonella. GLP-1 receptor agonists have a multimodal impact on gut microbiome with higher short-chain fatty acid (SCFA)-producing bacteria, higher Akkermansia and lower Firmicutes/Bacteroidetes ratios [42]. These can make the interpretation of the gut microbiome as biomarkers difficult to assess and translate into practice.

Concomitant alcohol use is a concern because it can affect the microbiome’s ability as a biomarker and can signal MetALD rather than MASH, and further drive progression of liver disease. However, a bigger pathophysiological issue is the endogenous production of alcohol by microbiota that can drive the progression of MASH (Fig. 1). Several studies have shown in animal models, children, and adults with MASH that several different microbiome constituents are associated with this endogenous alcohol production [43-48]. The wide variety of microbes that are capable of producing alcohol to be relevant in pathogenesis of hepatic steatosis makes specific targeting difficult [49].

Microbiome-related biomarkers have been used to (a) predict advanced fibrosis, (b) predict cirrhosis-related outcomes and (c) determine transplant-related recovery [50-55]. Some microbial biomarkers such as bile acid profiles and bacterial constituents also predict the success of therapies in MASH and cirrhosis [56,57]. However, none of the biomarkers currently have reached the stage of clinically translatable biomarkers and more work is needed to standardize these to be useful.

Prebiotics are non-digestible food ingredients that can be selectively fermented, resulting in specific changes in the activity or growth of the bacteria in the gut [58]. Their potential role in MASLD relies on the capacity to modulate: (1) gut microbiota composition; (2) bile acid metabolism, which in turn impacts hepatic lipid and glucose homeostasis; (3) inflammatory pathways through reduction in the gut permeability and endotoxemia, down-regulation of pro-inflammatory cytokines such as TNF-a, IL-6, IL-1B, or the expression of Toll-like receptors (TLRs); and (4) SCFAs, which carry the potential to reduce oxidative stress and improve insulin sensitivity [59,60]. In humans, the main prebiotics of interest are fructo- and galacto-oligosaccharides, β-glucans, resistant starch and inulin. However, evidence in the MASLD population remains scarce, with only a few randomized controlled trials being carried out to date. A pilot study with seven MASLD patients found fructo-oligosaccharides to reduce AST levels [61]. More recently, another study showed that patients with MASH treated with prebiotics had greater reduction in liver steatosis than placebo [61,62]. While these studies are limited by small sample sizes and their pilot design, their potential in MASLD treatment is promising. New prebiotics, such as 2-fucosyllactose, warrant specific trials to demonstrate their utility in the MASLD treatment landscape [63].

Probiotics are safe, live bacteria that promote health in their host [64]. The rationale for their use in MASLD relies on similar proposed mechanisms as prebiotics. Although there is extensive preclinical evidence with respect to the effectiveness of a wide variety of probiotics [65], clinical results remain controversial. Few studies have demonstrated improvement in aminotransferase levels, inflammatory markers, and lipid profiles after probiotic supplementation in MASLD, with the combination of Lactobacillus + Bifidobacterium + Streptococcus being the most effective [66]. However, other studies have failed to replicate these findings, and the heterogeneity in probiotics used, duration of treatment, and patient populations studied make the interpretation of results difficult [67].

Synbiotics are combinations of live microorganisms and substrates that confer a benefit to host health with potentially synergistic effects [68]. Although a 7-week synbiotic intervention revealed improvement in ALT in metabolically healthy individuals, a one-year randomized controlled trial evaluating synbiotics in patients with MASLD did not reveal significant improvements in hepatic steatosis or fibrosis [69,70].

Postbiotics are preparations of non-viable microorganisms and/or their components, soluble products or metabolic byproducts, that confer health benefits to their host, without potential for colonization. Highly effective postbiotics, such as peptidoglycans, teichoic acids, cell-free supernatants, bacteriocins, exopolysaccharides and SCFAs, have been obtained from species belonging to Lactobacillus, Bifidobacterium, Saccharomyces, Bacillus, Streptococcus and Faecalibacterium genera [71]. Although there is no evidence in patients with MASLD, postbiotics such as pasteurized A. muciniphila have shown anti-obesogenic, and anti-inflammatory effects with improvement of insulin resistance [72].

The disparity of results has been evaluated by numerous authors in the past few years through meta-analyses. Vakilpour et al. determined that pro-, pre-, and synbiotics had a significant effect on fasting insulin and HOMA-IR, which contrasts with other papers published previously [73-75]. Another meta-analysis revealed improved lipid profiles (reduced triglycerides, total cholesterol and LDL), in agreement with a previous one that included 22 RCTs (total 1,301 patients) [76,77]. Nevertheless, extensive evidence exists with regards to the effect of pre-, pro-, and synbiotics on reduction in liver enzymes, while the effect on liver fibrosis and liver steatosis is not always observed or reported [78-81].

The current evidence, although promising, is not conclusive, and additional studies, preferably randomized controlled trials, carried out after standardization of doses, strains and duration of treatments are required to determine long-term efficacy and establish the precise mechanisms of action.

Recent research into the dynamic role of the gut microbiome and its metabolites on the pathogenesis of MASLD has allowed FMT to be explored as a plausible therapeutic option [64].

Multiple mechanisms have been implicated in the potential therapeutic effect of FMT on MASLD pathophysiology. These include: reduction of endotoxin-producing bacteria [82]; repair of mechanical and immune mucosal barriers with improved intestinal permeability (“leaky gut”) [83,84]; downregulation of receptors of DAMPs [83]; upregulation of insulin receptors in the liver and improved insulin sensitivity [85,86]; modulation of the gut phageome [87]; activation of the bile acid-FXR-FGF-19 pathway; epigenetic regulation of genes involved in metabolism; downstream production of microbe-associated metabolites, e.g., secondary bile acids [88]; among others.

Animal studies involving transplantation of feces from healthy rodents into high-fat diet (HFD) rodents, and vice versa, as well as interspecies FMT between humans and mice laid the groundwork for human clinical trials evaluating FMT in patients with MASLD [83,89-91]. Craven et al. found that, relative to autologous FMT, single allogeneic FMT from lean, healthy donors improved small intestinal permeability but did not significantly improve body mass index (BMI), insulin resistance or steatosis based on MRI-PDFF. However, the study was limited by small sample size, low dose of donor stool, as well as follow-up periods that were too early or too late to see a sustained response in insulin resistance or MRI-PDFF, respectively [84]. Comparing FMTs from lean vegan donors to autologous FMTs in obese patients with MASH, Witjes et al. demonstrated a trend towards improved necroinflammatory scores, gut microbial composition as well as favorable changes in plasma metabolites and expression of liver genes involved in inflammation and lipid metabolism. However, this study was underpowered and thus findings did not reach statistical significance [92]. Xue et al. compared FMT with oral probiotics, both combined with lifestyle interventions, in lean and obese patients with MASH. The FMT group had significantly greater improvements in hepatic controlled attenuation parameter (CAP) scores as well as gut microbial abundance and diversity [93].

Human trials have demonstrated the safety of FMT in patients with MASLD and potential efficacy in improving gut permeability and hepatic necroinflammatory activity via modulation of gut microbial composition. However, not all FMTs will induce microbial composition changes or favorable metabolic profiles [94]. Factors including donor metabolic profile, recipient microbial composition, diet, and comorbidities likely influence response to donor FMT [92,95]. Future, larger studies are needed to evaluate the therapeutic potential of FMT in MASLD with specific focus on the role of varying donor types, route of administration, dose, frequency, as well as combination therapy with MASH/weight loss drugs.

Conventional therapies typically require accretion of doses to physiologically relevant quantities for efficacy, but are often associated with adverse reactions. Engineered bacteria offer an exciting opportunity to deliver a replicating and potentially low-cost therapeutic exactly when and where needed for the host. This would decrease adverse reactions, as targeted gene expression via these systems can be potentially controlled by environmental cues or inducing agents, such as food-grade peptides (nisin) [96]. Dysbiotic microbiotas are characterized by the collapse of areas of the microbial network that provide important bioenergetic components, nutrients, and vitamins [97]. Engineered bacteria can support the expression and delivery of beneficial peptides and molecules to the host, in addition to modifying and restoring microbiota function. Downstream effects of these changes include antimicrobial activity, pathogen control, immunomodulation, and metabolism of toxins [64].

While the types of microbes used as probiotics are rapidly expanding, a limited number of microbial genera, particularly Lactobacillus and Bifidobacterium, have been extensively harnessed as safe, probiotic workhorses. Probiotics harbor the potential to improve pathophysiological factors associated with MASLD. These include improvements in liver chemistries, steatosis, blood glucose levels, insulin levels and lipids in MASLD patients taking probiotics [74]. The mechanisms likely include improvement of intestinal permeability through enhanced junctional connections (zonulin and claudin) between epithelial cells and reduction of inflammation. Engineered bacteria, based on these probiotics, are promising platforms for therapeutic delivery of beneficial peptides and epitopes in the intestinal tract. Furthermore, their favorable safety profile and ability to be stabilized in commercially-produced processes render them an attractive therapeutic intervention.

The first bacteria that offered synthetically incorporated mammalian genes expressing cytokines was Lactococcus lactis. This species is ubiquitously used in the food industry, particularly in dairy food fermentation. While initially deemed challenging, eventually IL-2 and IL-10 were engineered in L. lactis [98,99]. While these were developed with a focus on intestinal autoimmune diseases such as inflammatory bowel disease, they were eventually considered applicable to other inflammatory and metabolic conditions, including liver disease. In a mouse model, IL-22, a member of the IL-10 cytokine family, was administered via engineered L. reuteri into the intestinal environment and subsequently expressed antimicrobial molecules such as regenerating islet-derived 3-gamma. This resulted in decreased bacterial translocation from the intestinal lumen and reduced ethanol-induced liver injury, steatosis and inflammation [100]. Apart from cytokines, beneficial peptides such as glucagon-like peptide-1 (GLP-1) can also be delivered via engineered bacteria systems such as Escherichia coli Nissle 1917 and Lactobacillus gasseri, to facilitate increased insulin levels, weight loss, as well as improved lipid profiles, liver chemistries and liver histology [101,102]. Moving forward, engineered strains expressing one or multiple peptides may be able to deal with multifactorial issues concomitantly.

In the future, it may be ideal to have members of the host's indigenous microbiota produce missing components directly. This is technically possible with systems such as CRISPR-CAS, which allow specific DNA to be inserted at targeted genomic sites. Direct microbiota genetic modification, or alternatives that control gene expression with mRNA, will likely replace lab-engineered bacteria, though they are still very experimental. In this, foreign genes of interest will be directly and specifically inserted into indigenous microbiota members, with no carrier bacteria required. Researchers have already shown that it is possible to insert conjugative systems to undertake targeted antimicrobial actions via plasmid-based delivery of Cas9-derived antimicrobial agents [103].

Engineered bacteria have the potential to be used at scale to deliver treatments for local benefit in the intestinal tract and liver. This will potentially allow treatments to be low-cost and have fewer side effects. Even more exciting is the prospect of ameliorating dysbiotic microbiota directly to replenish missing components. Beyond engineered bacteria, nucleic acid technologies such as RNA-based, and CRISPR technologies are promising for management of gut dysbiosis and downstream conditions such as MASLD, however, public and regulatory confidence in utilizing these technologies for such indications, is still wanting.

The gut virome is an integral part of the microbiome and carries the potential to influence the development and progression of MASLD. The human virome is predominantly composed of bacteriophages, the Microviridae family as well as families of the Caudovirales order [104,105]. Decreased viral diversity has been demonstrated in MASLD, specifically in individuals with NAFLD Activity Scores (NAS) of 5-8 and F2-F4 fibrosis. Relative abundance of several bacteriophages, such as those specific to Streptococcus, Escherichia, Enterobacteria and Lactobacillus, is associated with higher NAS scores. Bacteriophages to Leuconostoc and Lactococcus are associated with lower BMI, blood glucose levels and HbA1c levels [27]. Additionally, emerging evidence suggests stark differences in the gut virome of patients with lean MASLD compared to their non-lean counterparts [106]. In non-invasive predictive models, this variation in fecal viral diversity, alongside other clinical data, has been shown to greatly improve the prediction of histologic fibrosis severity [27].

In the study of the impact of microbiome dynamics on human disease, the implications of altered gut virome on the pathogenesis of MASLD is an evolving focus of interest. While the exact mechanisms remain unclear, gut virome alterations may affect hosts directly (e.g., immunomodulation, altered gut permeability, etc.) or by influencing bacterial microbiota [27]. Furthermore, infections by certain “obesogenic” viruses shift host metabolism towards a more adipogenic state, leading to chronic inflammation and altered lipid metabolism [107]. As an example, human adenovirus-36 (ADV36), has been implicated in stimulating enzymes and transcription factors that result in storage of triglycerides and maturation of pre-adipocytes into mature adipocytes [108]. Interestingly, ADV36 seropositive status has been shown to minimize hepatic steatosis, insulin resistance and facilitate weight loss [109].

The therapeutic potential of bacteriophages is currently being studied in pre-clinical models. Utilization of bacteriophages and archaeal viruses to target and modify microbial dysbiosis has been evaluated in animal models. Phage therapy directed against ethanol-producing Klebsiella pneumoniae attenuated the development of steatohepatitis and inhibited the production of inflammatory cytokines, and lipogenic gene expression in mouse models [110].

The human virome in MASLD is a developing area of research. The exact pathophysiological role of virome dynamics on MASLD severity and progression are yet to be fully elucidated. Significant challenges remain in complete identification of the human virome owing to limitations in wide-spread availability of metagenomic sequencing. Furthermore, only a small fraction of human viral taxa is available in public databanks, and much remains to be discovered regarding bacteriophage-bacteria interactions.

The human mycobiome comprises only 0.03% to 2% of the overall gut microbiota [111,112]. While Candida spp. and Saccharomyces cerevisiae are the most prevalent fungal organisms in human gut microbiota, deployment of culture-independent techniques such as 18S rRNA, internal transcribed spacer (ITS) and whole-genome shotgun sequencing has made it possible to extend the analysis to connote the presence of all four major phyla: Ascomycota, Basidiomycota, Chytridiomycota and Zygomycota, in healthy human subjects [113-115].

Alteration in the fecal mycobiome has been reported in persons with MASLD. Comparison of the mycobiome composition of patients with MASLD with matched healthy cohorts revealed significant taxonomic changes in gut mycobiota. Patients with MASLD had relative abundance of several fungal genera, predominantly Talaromyces, Parahaeosphaeria, Lycoperdon, Curvilaria and Sordaria, and depletion of Leptosphaeria, Fusicolla and Pseudopithomyces. Furthermore, Cladosporium and Paecilomyces were associated with development of fibrosis [116]. Significant differences in beta diversity were observed in patients with lean MASLD compared to their obese counterparts, with additional differences in the composition of the mycobiome noted in the setting of advanced fibrosis. Patients with MASH were found to have an abundance of Candida albicans, Mucor spp., Cyberlindnera jadinii and Penicillium spp., amongst others. Additionally, an augmented immune response to Candida albicans was exhibited in patients with advanced fibrosis, suggested by increased plasma levels of Candida albicans–specific IgG. Adminstration of antifungals in germ-free mice models was found to be protective against development of steatohepatitis [28]. Lastly, a direct relationship between the type of diet and composition of the mycobiome has been discovered in animal models, with Heisel et al. illustrating increased abundance of the pro-inflammatory Candida albicans and decreased abundance of non-obesogenic Saccharomyces boulardii in mice fed an HFD [117-119]. A compensatory increased abundance of cholesterol-lowering, lovastatin-producing Aspergillus terreus has also been noted in MASLD [120]. Furthermore, a complex syntrophic relationship between the mycobiome and microbiome has also been postulated to be influenced by a carbohydrate-rich diet [121].

Functional and anatomical bi-directional interactions between the mycobiome and liver are vital in modulating immune regulation and homeostasis [122]. Deranged gut-liver axis and increased intestinal permeability have been implicated in the pathogenesis of MASLD [123,124]. Resultant fungal translocation into the hepatic cells, specifically exposure to fungal polysaccharide β-glucan results in augmented production of proinflammatory cytokines, consequently leading to hepatocellular injury [125,126].

Understanding the association between mycobiome dynamics and pathogenesis of MASLD offers various therapeutic targets. Administration of fungal prebiotics, such as Saccharomyces boulardii, to obese and diabetic mice led to reduced hepatic steatosis and inflammation, increased liver function and mitigated progression of hepatic fibrosis [127]. Additionally, oral Aureobasidium pullulans-derived β-glucan, a known immune stimulator, prevented development of fatty liver in mice fed HFDs [128]. Lastly, oral antifungals, such as amphotericin B, reduced the risk of developing steatohepatitis in fecal microbiome humanized mice by virtue of honing fungal overgrowth and mitigating hepatic injury [28]. Deployment of oral amphotericin B is generally considered to be safe and has a favorable side effect profile. While these strategies certainly hold promise of a potential therapeutic avenue, ultimately, it would be pertinent to determine which subgroups of MASLD patients would benefit from specific interventional therapy given the significant heterogeneity and diversity of mycobiome in humans.

As mentioned in the previous sections, the use of microbiome-related biomarkers may help achieve a precise diagnosis and prognostic assessment of MASLD patients, allowing better identification of at-risk populations [129,130]. However, transitioning to clinically translatable biomarkers has proven difficult, and more data is needed in this regard. While the therapeutic potential of microbiome modulation in MASLD is promising, more information is currently needed to determine which approach would be the most effective for a given patient. MASLD is a heterogeneous disease, and its sub-phenotypes are consequential to a dynamic and complex interaction of several factors (e.g., sex, genetic background, comorbidities, alcohol consumption) including diverse microbiome composition and/or the presence of dysbiosis or altered intestinal permeability [131]. Additionally, the disease stage may influence the microbiome status, thereby impacting the potentially beneficial effects derived from microbiome-centered therapies [132]. Therefore, it remains uncertain which MASLD patients might benefit from microbiome modulation.

Head-to-head comparisons of microbiome-centered therapies, particularly engineered bacteria, bacteriophages, and antifungal therapies, are lacking in human subjects. Furthermore, direct comparisons of microbiome-centered therapies to current and emerging weight management and MASH therapies, e.g., GLP-1 receptor agonists and thyroid hormone receptor-beta agonists, are also awaited in human subjects with MASH. Nevertheless, microbiome-centered therapies, including postbiotics, FMTs and engineered bacteria can facilitate increased endogenous GLP-1 release, and GLP-1 delivery, respectively, making them ideal supplements to GLP-1 receptor agonists [133]. Similarly, recent evaluations in mouse models have revealed that the effect of thyroid hormone-beta receptor agonists on MASH may be mediated by modulation of glucosylceramide synthase (GCS)-catalyzed monoglucosylation of gut microbial sphingolipids, potentially identifying a new target for microbiome-centered therapies [134]. While an early meta-analysis showed that probiotics and synbiotics may improve ALT, liver stiffness measurements, and hepatic steatosis, many studies have not replicated these findings [135]. FMT holds promise in the management of MASLD through potential improvements of gut permeability and necroinflammatory scores on liver histology, as well as reduced expression of genes involved in inflammation and lipid metabolism [84,92]. Compared to oral probiotics, FMT may be superior with respect to improving CAP scores as well as gut microbial abundance and diversity [93]. Theoretically, assessing microbiome changes combined with host genetics, comorbidities, medications and lifestyle choices, could help delineate a therapeutic approach (Fig. 2).

The aim of precision medicine is to provide an effective and safe treatment tailored to the individual. However, there are several limitations and challenges when integrating microbiome-based therapeutics into personalized medicine [136]. Firstly, interindividual variability in microbiome changes driven by gender, lifestyle, ethnicity, etc. preclude reproducibility of studies. Secondly, geographical and cultural factors may also influence microbiome variations. Thirdly, the causality of microbiome changes on MASLD severity or progression remains difficult to assess in available studies, and thus the current range of microbiome-centered therapies serves to be supportive rather than definitively resolutive in the management of MASLD. These factors pose significant barriers to precision microbiome-centered therapies in general and in the context of MASLD in particular [129].

A glimpse into the future may be found in recent attempts to develop personalized probiotics, microbiota transplantation, targeted bacterial elimination, and dietary interventions tailored to an individual’s unique microbiome profile (personalized nutrition). Next-generation probiotics (i.e., health-promoting commensals) may hold promise in MASLD if animal data translates into human benefits [137]. One example is Akkermansia muciniphila, which has been found in decreased abundance in mice and some patients with MASLD and seems to induce beneficial effects on metabolic parameters [138]. Recent publications suggest that this microorganism holds therapeutic potential as a next-generation probiotic for MASLD [139].

Additionally, FMT, as discussed above, would likely be effective in a subset of MASLD patients, and selection criteria for its use need to be developed. Targeted bacterial elimination using bacteriophages could be useful in selected patients (e.g., those harboring alcohol-producing Klebsiella pneumoniae) as shown by Gan et al. [110]. Alternative approaches include the use of predatory bacteria and RNA-guided nucleases utilizing the guide ribonucleic acid (gRNA) to silence specific genes (e.g., antibiotic resistance genes), thus increasing the effectiveness of antibiotic treatments [140]. Finally, supplementation with metabolites such as SCFAs may have beneficial effects on GLP-1 levels, impacting glucose metabolism and improving diabetes control. Of note, recent data from the LIBRE trial showed that SCFAs are key mediators of the favorable effects of the Mediterranean diet on intestinal barrier integrity [141]. The current goal is understanding the individual microbiota and host responses to diet, and with this data, design personalized diets and meal plans. Considering the high prevalence of sarcopenia in patients with MASLD and the shared pathway of gut dysbiosis, inflammation and altered gut permeability, diets can potentially be further supplemented with microbiome-centered therapies to improve muscle strength and function [142,143]. Extrapolating the literature in the elderly population, prebiotics may assist in improving muscle strength, while probiotics may also improve physical function [144]. Furthermore, animal studies have shown that FMT from non-sarcopenic mice to sarcopenic mice induces improvements in muscle metabolism and muscle fiber structure, similar to their donors [145]. Further prospective studies in human subjects with MASLD are needed to validate these hypotheses.

In summary, precision microbiome-centered therapies hold great promise for treating MASLD. However, further research is needed to fully understand the mechanisms, optimize therapeutic strategies, and ensure safety and efficacy across diverse patient populations. Personalized strategies could be tailored according to gut microbiome features, measurements of gut microbiome-blood metabolites, customized dietary interventions and lifestyle improvements. The use of artificial intelligence and machine learning approaches to target gut microbiota may also enhance the efficacy of microbiome-centered therapies for MASLD.

The gut microbiome plays a pivotal role in the pathogenesis and progression of MASLD. Microbiome-centered therapies offer a promising avenue for modulating the disease trajectory, potentially addressing the complex interplay between gut health and liver disease. This review highlights the various therapeutic approaches, including prebiotics, probiotics, synbiotics, fecal microbiota transplantation, and novel interventions like engineered bacteria and bacteriophage therapy. While the current evidence is encouraging, the transition from experimental studies to clinical applications necessitates a deeper understanding of the underlying mechanisms and personalized treatment strategies. Factors such as individual microbiome variability, disease stage, and comorbid conditions significantly influence therapeutic outcomes. Therefore, future research should focus on refining these therapies, developing precise biomarkers, and leveraging artificial intelligence and machine learning to tailor interventions to individual patient profiles. By advancing our knowledge and application of microbiome-centered therapies, we can potentially transform the management of MASLD, offering more effective and targeted treatments to improve patient outcomes.