The role of systemic inflammation in hepatic encephalopathy: advances in inflammatory mechanisms, prevention and treatment research

Abstract

Objectives

This review synthesizes current evidence establishing systemic inflammation as a key pathogenic driver in hepatic encephalopathy (HE) beyond hyperammonemia. It does not replace the ammonia hypothesis but rather acts as a critical synergistic factor, modulating and amplifying ammonia neurotoxicity. It further evaluates the mechanisms linking inflammation to HE and the therapeutic advances in inflammation-targeted prevention and treatment strategies.

Methods

A comprehensive narrative literature review was conducted, analyzing relevant preclinical models and clinical studies. The search and synthesis focused on inflammatory mechanisms in chronic liver disease, gut-liver-brain axis dysfunction, cirrhosis-associated immune dysfunction (CAID), and resulting neuroinflammatory pathways.

Results

Systemic inflammation, driven by gut dysbiosis, barrier failure, and CAID, amplifies ammonia neurotoxicity and independently contributes to neuroinflammation, blood-brain barrier disruption, and cerebral metabolic dysfunction in HE. Key inflammatory markers, such as IL-6, correlate with disease severity. Therapies targeting inflammation – particularly gut microbiota modulation with rifaximin and fecal microbiota transplantation (FMT) – demonstrate significant efficacy in reducing HE recurrence, lowering systemic inflammation, and improving cognitive outcomes. Other approaches, including albumin infusion, also show promise.

Conclusions

Systemic inflammation is a pivotal and synergistic factor in HE pathogenesis. Combining anti-inflammatory strategies that target the gut-liver-brain axis with traditional ammonia-lowering therapies offers a more comprehensive and effective treatment paradigm. Future research should prioritize protocol optimization, long-term safety assessment, and the development of personalized treatment approaches.

Graphical abstract

Key Messages

1. Systemic inflammation acts as a critical synergistic factor in hepatic encephalopathy pathogenesis, modulating and amplifying ammonia neurotoxicity rather than replacing the ammonia hypothesis.

2. Dysfunction of the gut-liver-brain axis – driven by gut dysbiosis, bacterial translocation, and cirrhosis-associated immune dysfunction – serves as the central mechanism linking systemic inflammation to neuroinflammation and cognitive impairment in HE.

3. Targeting inflammation through gut microbiota modulation (e.g. rifaximin, fecal microbiota transplantation) and systemic anti-inflammatory agents (e.g. albumin) offers a promising therapeutic paradigm that complements traditional ammonia-lowering strategies.

1. Introduction

Hepatic encephalopathy (HE) refers to a range of neuropsychiatric impairments resulting from liver dysfunction or portosystemic shunts. Its clinical presentation is marked by disturbances in cognition, behavior, and motor coordination, which can advance to altered consciousness and, in severe cases, progress to coma. Etiologically, HE is categorized into three distinct types. Type A occurs subsequent to acute liver failure (ALF), a condition characterized by coagulopathy, cerebral edema, elevated intracranial pressure, and a high risk of mortality. Type B arises from portosystemic shunting without underlying liver pathology. Lastly, Type C is linked to chronic liver disease (CLD) and is commonly differentiated into ‘covert’ and ‘overt’ phases, with the latter historically graded by the West Haven Criteria from 1 to 4. In the context of cirrhosis, the onset of HE signifies a critical event of acute decompensation and is linked to substantially worsened patient outcomes [1]. For decades, gut-derived ammonia has been assigned a pivotal role in the pathogenesis of HE, a concept extensively supported by both in vivo and in vitro research [2,3]. Nonetheless, clinical observations indicate that hyperammonemia is an imperfect biomarker, as its levels do not consistently correlate with the manifestation or seriousness of HE symptoms [4–6]. Emerging evidence suggests that systemic inflammation plays a critical role in the development and progression of HE. Importantly, systemic inflammation does not supplant the ammonia hypothesis but rather acts as a critical synergistic factor, modulating and amplifying the neurotoxic effects of ammonia. This nuanced interplay is the central focus of this review.

The release of pro-inflammatory cytokines, a hallmark of systemic inflammation, can exacerbate the neurotoxic effects of ammonia. The elevated levels of cytokines such as interleukin (IL)-6 correlate with the severity of HE in cirrhotic patients [7,8]. During ALF or CLD, key drivers of systemic inflammation include hepatic inflammation, bacterial translocation from the gut into the bloodstream, and superimposed secondary infections [9]. This inflammatory response-induced neuroinflammation contributes to the cognitive and psychomotor deficits of HE. The interplay between hyperammonemia and systemic inflammation exacerbates neuronal dysfunction. Pro-inflammatory cytokines can impair the blood-brain barrier (BBB), allowing harmful substances to enter the central nervous system (CNS) and further aggravate neuroinflammation [1]. Alterations in the gut-liver-brain axis – including gut dysbiosis and increased intestinal permeability that facilitates bacterial translocation into the systemic circulation – promote systemic inflammation [1]. Currently, treatment strategies have been applied to improve both ammonia levels and systemic inflammation. Lactulose and rifaximin are commonly used to decrease ammonia production and absorption in the gut, while also exhibiting anti-inflammatory properties [10,11]. Additionally, human albumin infusion has the potential to mitigate systemic inflammation and improve outcomes in patients with decompensated cirrhosis and HE [12]. Notably, fecal microbiota transplantation (FMT) has emerged as a promising therapeutic approach in this context [13].

In this review, we will critically synthesize and evaluate the inflammatory pathways and inflammation-target therapies based on current animal experiments and clinical trials. We also discuss the directions of future research for HE.

2. Mechanisms of systemic inflammation driving HE

2.1. Inflammatory characteristics in CLD: distinguishing type A and type C HE

Systemic inflammation is a complex biological response involving the activation of the immune system, leading to the release of pro-inflammatory cytokines and other mediators throughout the body. Systemic inflammation can be triggered by various stimuli, including infections, autoimmune diseases, and chronic conditions such as obesity and diabetes. Numerous diseases, including cardiovascular diseases, metabolic syndrome, and liver dysfunction, highlight the critical role of systemic inflammation in disease progression and patient outcomes [14]. Cytokines coordinate antimicrobial effector cells and provide regulatory signals that direct, amplify, and resolve the immune response. At increased levels, cytokines can cause collateral damage to vital organ systems [15]. Patients with liver diseases often exhibit distinct inflammatory responses that differ from those observed in other diseases. In cirrhosis, the altered hepatic architecture leads to the release of inflammatory mediators into the systemic circulation, a process that further exacerbates liver dysfunction and contributes to multi-organ failure [2].

A critical distinction must be made between the inflammatory phenotypes of Type A and Type C HE. In Type A HE, systemic inflammation often manifests as a fulminant ‘cytokine storm’ – a rapid, overwhelming release of pro-inflammatory mediators driven by massive hepatocyte necrosis and secondary infections. This acute response rapidly disrupts the BBB, leading to vasogenic cerebral edema and life-threatening intracranial hypertension. Conversely, in Type C HE, inflammation is typically chronic, low-grade, and driven by cirrhosis-associated immune dysfunction (CAID). This creates a permissive background of neuroinflammation upon which acute precipitating events – such as infection, bleeding, or constipation – can trigger episodic worsening. This distinction is clinically critical, as the targets and urgency of anti-inflammatory interventions differ substantially between these contexts.

2.1.1. Identification of inflammatory markers in CLD

Systemic inflammation in liver disease is marked by increased levels of acute-phase proteins, proinflammatory cytokines and receptors, endothelial activation markers, signs of macrophage stimulation, systemic oxidative stress, and heightened expression of immune cell surface activation antigens [16]. Hematologic alterations are also common, including leukocytosis, leukopenia, anemia, thrombocytopenia, as well as increased ferritin and d-dimer levels. These shifts in cellular blood profiles likely arise from a multifaceted interplay involving cytokine-mediated modulation of production and bone marrow mobilization, immune-driven cell destruction, and chemokine-directed trafficking. Universal increases in nonspecific acute-phase reactants such as C-reactive protein (CRP) are frequently observed and often correlate with disease severity. Additionally, significant rises are commonly seen in serum levels of inflammatory mediators including interferon-γ (along with its induced chemokines CXCL9 and CXCL10), IL-6, IL-18, IL-1β, and soluble interleukin-2 receptor alpha – a reflector of T-cell activation [15]. The characteristics of systemic inflammation in liver diseases is summarized in Table 1.

Table 1.

Studies reporting the characteristics of systemic inflammation in acute liver failure and chronic liver disease.

Study	Experimental setting	Experimental results	Clinical results
Yaiza et al. 2024 [17]	rats with hyperammonemia and MHE	Increased levels of IL-17 and membrane expression of its receptor were observed in the rat cerebellum, along with enhanced activation of the IL-17 receptor in microglia and elevated levels of both IL-17 and TNFα.
Elina Manzhalii et al. 2019 [7]	Wistar rats with CCL4-induced cirrhosis and HE	↑ IFN-γ, IL-1β, IL-6	The degree of HE and depression correlated with a proinflammatory cytokine pattern.
	Patients with cirrhosis and HE	↑ IFN-γ, IL-1β, IL-6 correlated to the degree of HE and depression,↑ anti-inflammatory IL-4 as a physiologic counterbalance mechanism
Mangas-Losada et al. 2017 [18]	Patients with liver cirrhosis (125 without and 62 with MHE)	↑Levels of IL-6, IL-21, IL-17, IL-10, IL-18, CCL20, TNFα, CXCL13, IL-15, and CX3CL1 (fractalkine) in patients with MHE compared to those without; ↑IL-12, IL-1β, and IL-22 only in the MHE group.	These changes may promote immune cell infiltration into the brain and neuroinflammation, contributing to cognitive impairment in MHE.
Simon Johannes Gairing et al. 2022 [19]	Liver cirrhosis patients	IL-6 levels ≥ 8 pg/mL discriminated best between patients with and without MHE.	Higher IL‐6 levels remained independently associated with the presence of MHE.
Shawcross et al. 2011 [20]	100 patients with cirrhosis and HE (59% Grade 3; 41% Grade 4)	↑ SIRS score and SOFA score in patients with Grade 4 HE	The severity of HE (grades 3–4) and patient prognosis are linked to infection and systemic inflammation, rather than serum ammonia levels.
Nishi et al. 2025 [21]	Patients with HBV-ALF	TFS patients: ↑ IL-12p70, TNF-α, IFN-γ, G-CSF, RANTES, IL-6, IL-8, and fractalkine,↑IL-10 levels	A persistent inflammatory storm in non-TFS patients; reduced hepatic inflammation and improved survival in TFS patients; anti-inflammatory response in both groups.
		TFS Acute HBV-ALF: ↓IL-1α, IL-2, IL-6, and IL-8,↑IL-10 levels
Shuyao Li et al. 2025 [22]	18 cirrhotic patients (divided into HPA and LPA groups) and 3 healthy individuals	↑In the abundance of the phylum Proteobacteria (especially aerobic Enterobacteriaceae) in the HPA group, ↑SCFA-producing bacteria (e.g. Roseburia, Blautia, Coprococcus) in the LPA group	Longer reaction times in the Stroop test and more severe cognitive impairment in the HPA group

HE: hepatic encephalopathy; SIRS: systemic inflammatory response syndrome; SOFA: sequential organ failure assessment; MHE: minimal hepatic encephalopathy; IL: interleukin; TNF: tumor necrosis factor; CXCL13: C-X-C motif chemokine ligand 13; CX3CL1: C-X3-C motif chemokine ligand 1; CCl4: carbon tetrachloride; IFN: interferon; HBV-ALF: hepatitis B virus-related acute liver failure; TFS: non transplant-free survival; HPA: high plasma ammonia; LPA: low plasma ammonia; SCFA: short-chain fatty acid.

However, it is crucial to interpret these findings with caution. While elevated cytokines correlate with HE severity, this does not establish causation. The heterogeneity of patient populations makes it challenging to define universal inflammatory thresholds. Furthermore, most biomarkers are not specific to HE, limiting their utility in isolation. A critical gap remains in understanding whether these inflammatory signatures are drivers of neuroinflammation or merely epiphenomena of advanced liver disease and systemic illness. In this review, we aim to systematically label mechanisms as established (supported by robust human data), emerging (supported by preclinical or preliminary clinical data), or hypothesis-generating (requiring further validation).

2.1.2. The gut–liver–brain axis and immunity in CLD

Cirrhosis induces profound alterations in the gut–liver–brain axis, characterized by gut dysbiosis, increased intestinal permeability, and bacterial translocation [23]. Intestinal hemodynamics are altered due to portal hypertension, leading to mucosal congestion, reduced blood flow, and hypoxia. The mechanical barrier is compromised by downregulation of tight junction proteins (e.g. occludin, ZO-1), increasing intestinal permeability. Immunologically, activated innate immune cells release pro-inflammatory cytokines such as tumor necrosis factor-α (TNF-α) and IL-6, further disrupting barrier integrity. Biologically, gut dysbiosis occurs, characterized by a decrease in beneficial taxa (e.g. Bifidobacteria, Lachnospiraceae) and an overgrowth of potentially pathogenic bacteria, including ammonia-producing Enterobacteriaceae and Staphylococcus species [24,25]. This microbial shift promotes bacterial translocation, allowing live bacteria and their products (e.g. endotoxins) to cross the intestinal barrier into the portal circulation. In liver cirrhosis, impaired detoxification capacity of the liver enables these gut-derived toxins to bypass the hepatic filter and enter the systemic circulation [25].

Once in the systemic circulation, microbial products such as lipopolysaccharides (LPS) act as pathogen-associated molecular patterns (PAMPs) that persistently stimulate Kupffer cells and monocytes via Toll-like receptors (TLR), driving systemic inflammation [24]. Concurrently, CAID is now recognized as a clinical entity that manifests as a dual defect: innate immune hyperactivation coexists with exhaustion. Innate immune exhaustion is evidenced by monocyte dysfunction (e.g. reduced Human Leukocyte Antigen-DR isotype (HLA-DR) expression, impaired cytokine responses), while adaptive immune failure involves T-cell exhaustion (increased Programmed Death-1 (PD-1), impaired cytotoxicity) and B-cell dysregulation, creating a state of immunodeficiency despite ongoing inflammation [26,27]. This dysfunction accelerates progression to complications such as acute-on-chronic liver failure (ACLF) [24].

Among the gut-derived toxins reaching the systemic circulation, ammonia remains the key neurotoxin in HE. Ammonia crosses the blood-brain barrier (BBB) and triggers astrocyte swelling and cerebral edema, a hallmark of HE, through disruption of osmotic balance and mitochondrial function, leading to oxidative stress [28]. Furthermore, bacterial translocation leads to the accumulation of other neurotoxins, including manganese, mercaptans, and inflammatory mediators, which activate microglia, the brain’s resident immune cells [29]. This sustained neuroinflammation, in synergy with systemic inflammation and ammonia, compromises BBB integrity, exacerbates neuronal dysfunction and synaptic deficits, ultimately contributing to cognitive impairment [23,28,29]. The understanding of HE pathogenesis has thus evolved from a solely ammonia-centric view to a multifaceted model where ammonia cooperates with systemic inflammation and other gut-derived toxins to drive neurotoxicity and astrocytic dysfunction [29].

2.1.3. Systemic inflammation impairs liver function

Hepatic macrophages, or Kupffer cells, are key players in this interaction, responding to systemic inflammatory signals and releasing pro-inflammatory cytokines that can further propagate inflammation [30]. Systemic inflammation can lead to hepatocyte injury and exacerbate liver dysfunction through various mechanisms, including oxidative stress, apoptosis, and fibrosis [14]. Elevated levels of inflammatory cytokines such as IL-6 and TNF-α have been associated with the progression of liver disease and the development of complications like HE [8]. Moreover, systemic inflammation can disrupt the regulatory functions of the liver, impairing its ability to detoxify harmful substances and synthesize essential proteins, which are critical for maintaining homeostasis [16]. The presence of systemic inflammation has been shown to correlate with increased mortality rates in patients with liver disease, highlighting its role as a prognostic marker [31]. Additionally, systemic inflammation is implicated in the pathogenesis of complications such as ascites and hepatorenal syndrome, which further complicate the clinical management of patients with advanced liver diseases [32].

2.2. Systemic inflammation exacerbates HE

The interplay between elevated ammonia levels, inflammatory mediators, and the disruption of BBB highlights the intricate relationship between liver diseases and neurological dysfunction.

2.2.1. Inflammation-mediated neurotoxicity

The pro-inflammatory cytokines can disrupt neuronal function and contribute to cognitive impairment. Cirrhotic patients exhibit elevated levels of cytokines such as IL-6, TNF-α, and IL-1β [7]. These cytokines can induce neuronal apoptosis, alter neurotransmitter metabolism, and promote oxidative stress, exacerbating the neurotoxic environment in the brain. For instance, elevated IL-6 level could predict the onset of overt HE in cirrhotic patients [8]. Importantly, emerging evidence from preclinical studies is beginning to elucidate the specific molecular mechanisms by which inflammation modulates ammonia neurotoxicity. First, it has been hypothesized that inflammation alters the metabolic phenotype of brain cells, rendering them more vulnerable to ammonia. Animal studies have demonstrated that inflammatory cytokines such as TNF-α and IL-1β inhibit key enzymes of the mitochondrial electron transport chain [33], impairing ATP production. This is critical because astrocytic glutamine synthesis – the primary cerebral ammonia detoxification pathway – is an energy-intensive process. By inducing an energy deficit, inflammation reduces the brain’s capacity to handle ammonia, amplifying its neurotoxic effects. Second, preclinical evidence suggests that inflammation directly facilitates ammonia entry into the brain. In our ALF mouse models, we observed that TNF-α promoted the transport of ammonia from the blood to brain tissues by upregulating the ammonia transporter Rh glycoprotein C (Rhcg) [34]. This represents a potential molecular synergy that warrants further investigation in humans.

Beyond these direct effects, recent research has highlighted the crucial role of neutrophils in mediating HE-associated neuroinflammation. In cirrhosis, systemic inflammation promotes neutrophil activation, characterized by increased surface expression of adhesion molecules (e.g. CD11b/CD18) and the release of Neutrophil Extracellular Traps (NETs). These web-like structures, composed of chromatin, histones, and granular proteins (e.g. myeloperoxidase), while designed to trap pathogens, can activate TLRs in sterile inflammation, exacerbating endothelial damage and pro-inflammatory cascades [35]. Activated neutrophils have been shown in preclinical models to demonstrate enhanced transendothelial migration capacity, releasing matrix metalloproteinases (MMP) (e.g. MMP-9) that degrade the basement membrane and compromise blood-brain barrier integrity. Once within the CNS, neutrophils amplify neuroinflammation through release of reactive oxygen species, elastase, and pro-inflammatory cytokines, directly activating astrocytes and microglia, ultimately contributing to synaptic dysfunction and cognitive impairment [36]. Clinical studies have confirmed that plasma levels of NETs markers such as alpha-defensins, human neutrophil peptides (HNP1-3) are significantly elevated in alcohol-related liver cirrhosis patients and HNP1-3 can predict the development of HE [37]. While NETs are known to compromise BBB integrity via MMP-9, emerging evidence from other neuroinflammatory conditions suggests they may contribute to neurological dysfunction through additional mechanisms. In conditions such as stroke and traumatic brain injury, NETs have been implicated in promoting microthrombosis and impairing cerebral microcirculation, leading to tissue ischemia [36]. Whether NETs induce similar microvascular pathology in the HE brain – contributing to the cerebral energy crisis – remains an important, hypothesis-generating question for future research.

Oxidative stress and metabolic dysregulation are increasingly recognized as essential contributors to the pathogenesis of HE. Systemic inflammation may affect the neuronal metabolism, energy supply and oxidative stress, which further exacerbates neuronal injury and dysfunction [38]. Importantly, emerging evidence suggests that systemic inflammation profoundly disrupts cerebral hemodynamics and energy metabolism. Hyperammonemia and inflammatory cytokines (particularly TNF-α and IL-1β) impair neurovascular coupling – the precise matching of cerebral blood flow to neuronal activity. This occurs through inflammation-induced cerebrovascular endothelial dysfunction, characterized by impaired endothelium-dependent vasodilation and dysregulated nitric oxide metabolism, ultimately compromising cerebral autoregulation [9]. The resulting cerebral blood flow dysregulation coincides with a cerebral energy crisis. Both ammonia detoxification and inflammatory responses are energy-intensive processes. Glutamine synthesis in astrocytes consumes substantial ATP, while inflammatory cytokines inhibit mitochondrial electron transport chain function, reducing ATP production [33]. Neuroimaging studies reveal a paradoxical state of relative cerebral hyperperfusion with metabolic hypoactivity in cirrhotic patients, where maintained blood flow fails to support efficient substrate utilization, creating an energy deficit that impairs cognitive function [39].

The liver function impairment can lead to insufficient neuronal energy supply, neuronal damage and death [16]. Additionally, the metabolic demands of the inflammatory response can divert nutrients away from peripheral organs, including the brain, resulting in energy deficits that impair neuronal function [32]. Abnormal brain metabolism in patients with liver insufficiency are closely associated with the increase of blood ammonia level and the release of inflammatory factors [14]. Emerging evidence from rodent models suggests that extracellular vesicles (EVs) may represent a novel mechanism for transmitting peripheral inflammation to the brain, leading to neuroinflammation and subsequent motor and cognitive deficits in rats [40]. These findings require validation in human studies.

As detailed in Section 2.1.2, gut-derived bacterial translocation and systemic inflammation synergize with ammonia to drive neuroinflammation. A landmark study by He et al. in 2025 has provided groundbreaking evidence that may revolutionize our understanding of gut-brain communication in HE, moving beyond the concept of ‘generalized dysbiosis’ to the analysis of ‘specific functional microbial genes’ [41]. Using metagenomic sequencing, the authors demonstrated that the phenylalanine decarboxylase (PDC) gene from Ruminococcus gnavus is significantly enriched in the gut microbiome of cirrhotic patients. This gene encodes an enzyme that converts dietary phenylalanine to phenylethylamine (PEA). PEA was found to accumulate in the brain, where it directly drives neurological symptoms. This study provides a causal, functional link between a specific microbial gene and HE, identifying a novel therapeutic target (PDC inhibition) and representing a paradigm shift in the field.

2.2.2. Inflammatory response disrupt blood–brain barrier

The integrity of the BBB is crucial for maintaining brain homeostasis and protecting the CNS from peripheral inflammatory signals. In the context of HE, systemic inflammation can lead to BBB dysfunction, allowing the infiltration of immune cells and pro-inflammatory cytokines into the brain [23]. This disruption of BBB and entry of neurotoxins and inflammatory mediators contribute to the neurocognitive deficits of HE. Elevated levels of inflammatory cytokines can compromise the tight junctions between endothelial cells in the BBB, leading to increased permeability and subsequent neuroinflammation [23,34]. For example, proinflammatory cytokines like TNF-α and IL-1β modulate endothelial cell function by triggering the expression of adhesion molecules (E-selectin, Vascular Cell Adhesion Molecule-1, and Intercellular Adhesion Molecule-1), thereby facilitating leukocyte migration into the brain [16]. This process not only exacerbates the inflammatory response but also contributes to neuronal injury and cognitive decline. Furthermore, the BBB’s response to systemic inflammation is not uniform; it can vary based on the severity and duration of the inflammatory stimulus. Acute inflammation may lead to transient BBB disruption, while chronic inflammation can result in more sustained alterations in BBB integrity [32]. The complex interplay between gut-derived factors, systemic inflammation, and neuroinflammation is illustrated in Figure 1.

Figure 1.

The inflammatory gut-liver-brain axis in hepatic encephalopathy. This integrative figure illustrates the complex interplay between gut-derived factors, systemic inflammation, and neuroinflammation in the pathogenesis of hepatic encephalopathy (HE), as well as therapeutic interventions targeting each level of this axis. ① Gut Pathology: Portal hypertension in cirrhosis and impaired liver function lead to mucosal congestion, reduced blood flow, and hypoxia. These hemodynamic alterations, combined with gut dysbiosis (↓ beneficial taxa, ↑ Enterobacteriaceae), compromise intestinal barrier integrity through downregulation of tight junction proteins. This results in increased intestinal permeability and enhanced bacterial translocation. ② Bacterial Translocation and Systemic Inflammation: Live bacteria and pathogen-associated molecular patterns (PAMPs, e.g. LPS) translocate into the portal and systemic circulation. This drives cirrhosis-associated immune dysfunction (CAID), characterized by a dual defect: (a) hyperactivation of Kupffer cells and monocytes with excessive cytokine release, and (b) immune exhaustion evidenced by reduced HLA-DR expression on monocytes. Systemic levels of pro-inflammatory cytokines (TNF-α, IL-6, IL-1β) are elevated. ③ Therapeutic Interventions: Interventions are positioned at their primary sites of action: (a) Gut-targeted therapies include lactulose, rifaximin, fecal microbiota transplantation (FMT), dietary interventions, and probiotics/prebiotics, which aim to restore eubiosis, reduce ammonia production, and enhance barrier function. (b) Systemically, albumin and interleukin inhibitors target circulating inflammatory mediators. (c) Conventional ammonia-lowering therapies complement these approaches. ④ Blood–Brain Barrier (BBB) Interaction: Systemic cytokines, particularly TNF-α, upregulate the ammonia transporter Rhcg on brain endothelial cells, facilitating ammonia entry into the central nervous system. This represents a key molecular mechanism by which systemic inflammation synergistically amplifies ammonia neurotoxicity. ⑤ Neuroinflammation and Cerebral Consequences: Within the CNS, infiltrating neutrophils release neutrophil extracellular traps (NETs) and matrix metalloproteinase-9 (MMP-9), degrading the basement membrane and compromising BBB integrity. Microglia are activated by cytokines and PAMPs, releasing further inflammatory mediators. This neuroinflammatory milieu, combined with ammonia-induced astrocyte swelling, leads to an energy crisis characterized by impaired mitochondrial ATP production, neuronal dysfunction, and synaptic deficits, ultimately manifesting as cognitive impairment. This figure integrates the key pathogenic mechanisms discussed in Sections 2.1.2 and 2.2.1, and the therapeutic interventions detailed in Section 3.

3. Targeted interventions for systemic inflammation

The management of HE primarily involves addressing precipitating factors, including infections, gastrointestinal bleeding, excessive diuresis or dehydration, electrolyte disturbances, and constipation. Conventional therapies such as lactulose and rifaximin continue to form the foundation of HE treatment by effectively lowering ammonia levels and reducing the risk of recurrence. Emerging strategies also offer potential for ameliorating systemic inflammation. Therapeutic approaches targeting the gut microbiome encompass prebiotics, probiotics, FMT, antibiotics, and synbiotics [42]. Adequate intake of probiotics can enhance host health through modulation of the intestinal microbiota, while prebiotics promote the growth of beneficial gut bacteria [42,43]. Synbiotics, which combine prebiotics and probiotics, exert synergistic effects. A meta-analysis incorporating 9 randomized controlled trials confirmed the benefits of prebiotics and probiotics in cases of minimal HE (MHE) [43]. This section discusses the clinical applications of FMT, the role of rifaximin, and other potential anti-inflammatory therapies.

3.1. Fecal microbiota transplantation

FMT has gained attention as a promising therapeutic strategy for conditions associated with gut microbiota dysbiosis, particularly for its potential role in modulating systemic inflammation. FMT can be administered through multiple routes, such as colonoscopy, enema, nasoenteric tubing, or oral capsules [25]. Its efficacy depends on several variables, including donor selection, stool processing protocols, and the administration pathway. The establishment of standardized stool banks has increased the use of unrelated donor samples, while autologous FMT has also emerged as a viable option. Delivery methods vary by target site: upper gastrointestinal transfer typically employs esophagogastroduodenoscopy, nasoenteric tubes, or oral capsules, whereas colonoscopy or enema is used for lower gastrointestinal delivery. Among these, oral capsules constitute a more recent advancement in FMT methodology. FMT has demonstrated efficacy in restoring gut microbial balance, especially in recurrent Clostridioides difficile infection (CDI) and ulcerative colitis [44–46]. Its therapeutic mechanisms include immunomodulatory effects and the promotion of short-chain fatty acid (SCFA) production by commensal bacteria. SCFAs contribute to immune regulation by enhancing regulatory T-cell activity and suppressing pro-inflammatory cytokine release [47]. Furthermore, FMT has shown potential in ameliorating metabolic disorders, such as obesity and diabetes, through functional and compositional reshaping of the gut microbiome [48–50]. Emerging evidence also suggests that FMT may influence neurological health via the gut–brain axis, through modulation of neuroactive metabolites and immune-mediated pathways [23,31]. Beyond digestive diseases, FMT is being explored for its potential benefits in metabolic syndrome, cardiovascular disorders, and certain neurodegenerative diseases [51].

Cirrhosis, especially when complicated by HE, is frequently linked to gut dysbiosis [52]. Restoring a balanced gut microbiota (eubiosis) may help attenuate systemic inflammation, enhance intestinal barrier integrity, limit bacterial translocation, and support normal bile acid metabolism. Preclinical studies in rat models of HE have shown that FMT alleviates intestinal edema, mitigates mucosal injury, and reduces inflammatory cell infiltration [53]. Additionally, FMT has been found to lower blood ammonia concentrations and reduce proinflammatory cytokines, including IL-1β, IL-6, and TNF-α [53]. In a rat model of D-galactosamine-induced liver injury, administration of FMT supplemented with B. adolescentis shifted gut microbial composition by reducing the abundance of pathogenic Proteus and increasing beneficial taxa involved in lipid and amino acid metabolism, such as Coriobacteriaceae, Bacteroidales, and Allobaculum [54]. Following the first case of treating HE with FMT [55] and the first randomized clinical trial conducted in 2017 [56], multiple clinical studies and case series have investigated its therapeutic potential [57–62]. Oral FMT capsules were shown to enhance cognitive function in HE patients, with treatment efficacy influenced by specific donor and recipient characteristics [57]. Responders to FMT exhibited higher baseline and sustained abundances of Bifidobacterium and other beneficial microbes. Conversely, recipients of FMT from donors with low fecal SCFA levels demonstrated limited cognitive improvement [57]. A systematic review reported significant neurocognitive enhancement in HE patients receiving FMT, with success rates reaching 86.7% in certain trials [13]. A recent meta-analysis [63] found that FMT significantly reversed MHE, relieved overt HE, decreased ammonia, neurocalcitonin level and hospitalization rates. A systematic review about FMT in treating HE and CDI [64] proved a persistent improvement in the cognitive performance.

However, several critical challenges must be acknowledged. First, the mechanisms of efficacy remain incompletely understood. At present, it is unclear whether benefits are mediated by specific keystone species (e.g. Bifidobacterium), a broader change in community metabolic function (e.g. short-chain fatty acid production), or bacteriophage transfer. These represent hypothesis-generating questions for future research. Second, significant regulatory and logistical barriers impede widespread clinical adoption. Unlike the well-defined, synthetic drug rifaximin, FMT is a complex, living biologic product. Issues of donor variability, stool processing standardization, and route of administration lead to significant heterogeneity in clinical outcomes. Responders to FMT exhibit higher baseline and sustained abundances of beneficial microbes, whereas recipients of FMT from donors with low fecal SCFA levels demonstrate limited cognitive improvement [57]. Third, the risk-benefit profile must be carefully weighed against established therapies. While the rate of serious adverse events (SAE) in cirrhotic cohorts is generally low, cases of extended-spectrum beta-lactamase (ESBL)-producing Escherichia coli bacteremia transmitted via FMT have been reported [65], underscoring the risk of transferring occult pathogens to an immunocompromised population. This contrasts with rifaximin’s well-established long-term safety profile [66].

Importantly, the strength of evidence must be interpreted with caution. The promising results discussed above – including findings from rodent models of HE [53,54] – are valuable for hypothesis generation. However, they may not fully recapitulate the complex, multi-year dysregulation of the gut-liver-brain axis in cirrhotic humans with multiple comorbidities. High-quality randomized controlled trials (RCTs) in humans [56,57,59] provide the strongest evidence, and these have been clearly demarcated in our discussion. Future efforts should focus on standardizing FMT protocols, establishing rigorous donor screening criteria, and conducting long-term follow-up studies to optimize therapeutic outcomes and ensure safety.

3.2. Rifaximin

Traditionally, non-absorbable disaccharides, such as lactulose, serve to reduce ammonia levels in HE. However, some patients with HE exhibit intolerance, poor adherence [67], or an inadequate response to lactulose monotherapy [68]. In recent years, rifaximin, a poorly absorbed antibiotic, has emerged as a promising alternative or adjunct therapy for HE. Rifaximin acts as a broad-spectrum antibacterial agent, effective against both gram-positive and gram-negative bacteria, including aerobic and anaerobic strains [69]. Rifaximin inhibits bacterial urease activity, limiting ammonia generation from urea hydrolysis [70]. It also enhances ammonia clearance by improving intestinal barrier function, reducing bacterial translocation [71]. Following the pivotal RCT in 2010 ⁶⁶, subsequent studies and meta-analyses have demonstrated that rifaximin can reduce the risk of recurrent HE and hospitalization rates [68,72,73]. Numerous RCTs showed that rifaximin improved overt HE (OHE) by ameliorating neuropsychiatric symptoms and enhancing cognitive function [71,74–76]. Timely administration of rifaximin to MHE patients can significantly improve long-term outcomes, reduce hospitalization rates, and enhance quality of life [77–79]. Transjugular Intrahepatic Portosystemic Shunt (TIPS) is an effective treatment for portal hypertension-related complications. A recent study [80] demonstrated that TIPS-induced reduction in portal hypertension leads to significant attenuation of systemic inflammation and bacterial translocation over time. However, 35-50% of patients develop OHE after TIPS [81]. In a RCT study [82] and a meta-analysis [83], rifaximin (not lactulose) reduced the risk of HE after TIPS, which exemplifies a strategy targeting the chronic, low-grade inflammatory background characteristic of Type C HE. By modulating gut microbiota and reducing systemic inflammation, rifaximin decreases susceptibility to acute precipitating events that trigger episodic worsening in this context. Adding rifaximin to lactulose therapy reduced the hospitalization rate for HE from 41.6% to 22.2% [84]. This dual approach significantly reduces HE recurrence and mortality compared to monotherapy [75].

Furthermore, rifaximin can also reduce the recurrence of spontaneous bacterial peritonitis (SBP) [85,86]. In Huang et al.’s study [87], rifaximin therapy decreased the total liver-related score and serum CRP of cirrhosis patients and increased the rate of portal vein thrombosis (PVT) recanalization. In decompensated cirrhotic patients [88], the use of rifaximin was associated with lower infection rates, displaying an incidence rate ratio (IRR) of 0.64. Among cirrhotic patients in the intensive care unit (ICU) receiving broad-spectrum antibiotics, overt HE reversal rates were similar despite discontinuation of rifaximin [89,90].

When administered for HE, rifaximin demonstrates a favorable safety profile. Adverse events are typically mild and uncommon, with severe diarrhea necessitating therapy withdrawal in merely 1.6% of cases [66]. There were two cases of CDI reported in the pivotal study [91]. In a retrospective study [92], cirrhotic patients who are on chronic rifaximin have decreased rates of CDI compared with those not on this therapy. Notably, Turner et al. [93] showed that rifaximin could cause cross-resistance to daptomycin in vancomycin-resistant Enterococcus faecium (VREfm). Emerging evidence suggests that rifaximin’s therapeutic effects may extend beyond simple antimicrobial eradication. The term ‘eubiotic’ effect has been proposed based on recent mechanistic studies to describe its ability to promote a favorable microbial metabolic environment and enhance gut barrier function, rather than merely depleting bacterial biomass. Patel et al. in the RIFSYS randomized controlled trial, demonstrated that rifaximin-α not only reduced systemic inflammation and neutrophil TLR-4 expression but also enriched a TNF-α- and interleukin-17E-associated intestinal microenvironment and promoted gut barrier repair [76]. This suggests that rifaximin actively reshapes the host-microbe interface towards a more homeostatic state, a mechanism distinct from traditional antibiotics. Studies about the effects of therapies targeting the gut microbiota on systemic inflammation are summarized in Table 2.

Table 2.

Studies reporting on the effects of therapies targeting the gut microbiota on systemic inflammation.

Study	Agent studied	Experimental setting	Experimental results	Clinical results
Paula Izquierdo-Altarejos et al. 2024 [94]	MSC-EVs	Hyperammonemic rats	Ex vivo studies demonstrate that MSC-EVs reduce pro-inflammatory factors such as TNFα, suppress NF-κB activation, and inhibit key pathways responsible for motor incoordination.	MSC-EV administration suppressed the activation of cerebellar microglia and astrocytes, and ameliorated motor coordination deficits in a rat model of hyperammonemia.
Xueyan Lin et al. 2025 [95]	Rifaximin and lactulose	CCl4 induced rat model of MHE	↑Portal and serum LPS levels and circulatory pro-inflammatory mediators, including IL-1β and TNF-α, ↑Clostridiales and Spirochaetia in MHE rats, ↑Lactobacillus, Erysipelotrichia, Allobaculum, and Bifidobacterium in rifaximin group, ↑Proteobacteria and Bacteroides in lactulose group, ↓portal LPS levels,IL-1β and TNF-α in rifaximin and lactulose group	Rifaximin and lactulose↑cognitive performance in MHE rats
Keyvan Amirshahrokhi et al. 2025 [96]	Edaravone	HE mice induced by the injection of thioacetamide	↓the expression of p-NF-κB and iNOS. ↓the levels of NO, MPO and MMP-9 in the brain of mice. ↓the brain levels and expressions of inflammatory cytokines IL-1β, IL-6, TNF-α and IFN-γ.	↑the locomotor function and ↓brain histopathological changes in mice with HE
Hao Wu et al. 2016 [97]	STAT3	HCC patients	↑Serum levels of IL-1β, IL-6, IL-17a, IFNγ and IFNλ3 in patients with MHE. ↑STAT3 in circulating PBMCs of MHE.	STAT3 inhibitor treatment prevented cytokine-induced neuronal apoptosis in vitro.
Alba Mangas-Losada et al. 2019 [98]	Rifaximin	30 controls without liver disease, 30 cirrhotic patients without MHE and 22 patients with MHE. Patients with MHE were treated with rifaximin.	↓IL-17, CXCL13, CX3CL1 (fractalkine) and IL-22 to normal values in responders but not in non-responders.	Rifaximin improved MHE in 59% of patients with MHE.
Vishal C Patel et al. 2022 [76]	Rifaximin-α	38 cirrhosis patients with HE were randomised 1:1 to receive either rifaximin-α or placebo for 90 days.	Rifaximin-α↓circulating neutrophil TLR-4 expression and plasma TNF-α,↓oralisation of the gut, ↓the abundance of mucin-degrading and sialidase-rich species such as Streptococcus spp., Veillonella atypica, V. parvula, Akkermansia, and Hungatella. Rifaximin-α↑a TNF-α- and interleukin-17E-enriched intestinal microenvironment,↑antibacterial responses to invading pathobionts,↑gut barrier repair	Resolution of overt and covert HE, ↓the likelihood of infection, ↓oralisation of the gut and ↓systemic inflammation.
Arpan Jain et al. 2022 [99]	A combination of LOLA, lactulose, and rifaximin (n = 70) or placebo, lactulose, and rifaximin (n = 70).	Patients of cirrhosis with OHE grade III-IV	↓Levels of inflammatory markers. ↓levels of blood ammonia, IL-6, and TNF-α in the LOLA group.	LOLA combined with lactulose and rifaximin demonstrated superior efficacy compared to lactulose and rifaximin alone, resulting in greater improvement in HE grading, shorter recovery time from encephalopathy, and reduced 28-day mortality.
Bajaj et al. 2019 [59]	FMT	20 cirrhotic patients with recurrent HE and MELD < 17 receiving standard of care therapy.	↑ Gut microbiota diversity with ↑ abundance of Ruminococcaceae and Bifidobacteriacceae, ↓ Streptococcaceae and Veillonellaceae. ↑ E-cadherin and defensin alpha 5. ↓ pro-inflammatory cytokines.	↑ Cognition.
Andrew Fagan et al. 2023 [100]	Albumin infusions	Outpatients with cirrhosis and prior HE, MHE and hypoalbuminemia already on treatment for HE.	↓IL-1β and endothelial dysfunction markers was also observed in the albumin group.	↑Cognitive function and psychosocial QoL

STAT3: signal transducer and activator of transcription; HCC: hepatocellular carcinoma; IL: interleukin; IFN: interferon; PBMC: peripheral blood mononuclear cell; MHE: minimal hepatic encephalopathy; MSC-EVs: extracellular vesicles from mesenchymal stem cells; CCl4: carbon tetrachloride; LPS: lipopolysaccharide; TNF: tumor necrosis factor; HE: hepatic encephalopathy; NF-κB: nuclear factor kappa-light-chain-enhancer of activated B cells; iNOS: inducible nitric oxide synthase; NO: nitric oxide; MPO: myeloperoxidase; MMP-9: matrix metalloproteinase-9; CXCL13: C-X-C motif chemokine ligand 13; CX3CL1: C-X3-C motif chemokine ligand 1; TLR-4: Toll-Like receptor 4; LOLA: L-Ornithine L-Aspartate; OHE: overt hepatic encephalopathy; FMT: fecal microbiota Transplantation; MELD: model for end-stage liver disease; QoL: quality of life.

3.3. Other potential anti-inflammatory therapies

Rifamycin SV MMX (RiVM), a non-absorbable rifampin derivative, acts primarily in the colon. A phase 2 randomized controlled trial [101] provided preliminary clinical evidence that RiVM was well tolerated in patients with MHE. Treatment led to shifts in gut microbiota composition and function, reduced ammonia levels, decreased systemic inflammation and cerebral oxidative stress, and improved parameters related to sarcopenia; however, no significant cognitive improvement was observed. In contrast, VE303 is a defined bacterial consortium composed of eight clonally purified strains that are known to produce metabolites potentially beneficial in HE. Bloom et al. [102] conducted an early-phase RCT of VE303 in recurrent overt HE patients, providing proof-of-concept data that require confirmation in larger trials. Of the 8 administered strains, 2 engrafted in >50% of recipients. Both strain engraftment and increased fecal butyrate showed trends toward improved psychometric HE score (PHES). Although short-term albumin use offers limited benefit in OHE, growing evidence supports extended (>2 weeks) or long-term (over months) albumin administration for enhancing cognitive function and reducing relapse rates in OHE patients [103]. Dietary interventions (e.g. Mediterranean diet), probiotics/prebiotics, and targeted anti-inflammatory agents (such as interleukin inhibitors) can reduce systemic inflammation by modulating gut microbiota, enhancing barrier function, or blocking inflammatory pathways, with demonstrated benefits in metabolic disorders, IBS, and cardiovascular disease [47,51,104,105]. However, their role in hepatic encephalopathy requires further investigation.

The anti-inflammatory role of human albumin in HE has been underestimated. Beyond volume expansion, albumin acts as a pleiotropic immunomodulator and endotoxin scavenger, binding circulating pathogen-associated molecular patterns (PAMPs) and damage-associated molecular patterns (DAMPs) to reduce systemic inflammation. The recent HEAL study – a double-blind randomized placebo-controlled trial in outpatients with cirrhosis and prior HE – provided robust evidence that long-term albumin administration significantly reduced pro-inflammatory cytokines (IL-1β) and endothelial dysfunction markers [100]. This reduction was associated with improved cognitive function and quality of life [103]. Notably, the divergent outcomes of short-term versus long-term albumin administration underscore the importance of context-specific anti-inflammatory strategies: while short-term albumin offers limited benefit in acute HE episodes, long-term therapy targets the chronic, low-grade inflammation characteristic of Type C HE, reducing susceptibility to acute exacerbations.

4. Prophylaxis of hepatic encephalopathy

Preventing hepatic encephalopathy is critical in managing cirrhosis, though the optimal approach differs fundamentally by clinical context. As discussed in Section 2.1, the distinction between Type A and Type C HE inflammatory phenotypes has direct implications for prophylactic strategies. For patients with cirrhosis (Type C HE), the goal of prophylaxis is to modulate the chronic, permissive inflammatory milieu, thereby reducing susceptibility to acute precipitating events. This is achieved through interventions that target gut-derived inflammation and ammonia. In contrast, Type A HE, which occurs in the setting of acute liver failure, is characterized by a fulminant cytokine storm requiring urgent, often intensive care-based interventions that fall outside the scope of chronic prophylactic strategies discussed here.

For patients at high risk of HE, the cornerstone of prophylaxis lies in reducing gut-derived ammonia and inflammation. Lactulose and rifaximin are the mainstay agents. Rifaximin’s prophylactic efficacy is increasingly attributed to its anti-inflammatory and eubiotic effects, as detailed in Section 3.2, rather than solely to ammonia reduction [106]. A network meta-analysis provided valuable evidence regarding the treatment of subclinical HE [107]. According to the analysis, the most effective combinations for preventing progression to overt HE were, in descending order: lactulose with rifaximin, followed by rifaximin with L-carnitine, and then lactulose combined with rifaximin and zinc. In terms of ammonia reduction, L-ornithine L-aspartate (LOLA) was most effective, followed by nitazoxanide and lactulose. The combination associated with the fewest adverse effects was lactulose and nitazoxanide, followed by rifaximin with L-carnitine, and then probiotics. However, the study did not yield statistically significant findings regarding mortality outcomes or changes in quality of life.

Beyond pharmacological ammonia-lowering strategies, comprehensive preventive care must address the identification and management of precipitating factors. Nutritional intervention plays a fundamental role that directly intersects with inflammation. Current guidelines emphasize adequate protein intake (1.2–1.5 g/kg/day) to prevent sarcopenia [108]. Skeletal muscle is not only an alternative site for ammonia detoxification but also an endocrine organ. Loss of muscle mass (sarcopenia) is associated with a pro-inflammatory state, characterized by elevated myokines that can perpetuate systemic inflammation. Therefore, preserving muscle mass through optimal nutrition represents an indirect but important anti-inflammatory prophylactic strategy. Finally, for patients with recurrent HE refractory to medical therapy, evaluation for portosystemic shunt embolization or liver transplantation represents the definitive preventive intervention, offering a potential cure for the underlying portal hypertension and liver dysfunction [109].

5. Conclusion

Systemic inflammation is now recognized as a synergistic mechanism in HE pathogenesis. It involves established pathways such as pro-inflammatory cytokines and gut dysbiosis, as well as emerging mechanisms including oxidative stress, metabolic dysregulation, and neutrophil extracellular traps (NETs). Addressing systemic inflammation through various interventions, including FMT, rifaximin, dietary modifications, and targeted pharmacological therapies, holds great promise for improving health outcomes in patients with chronic inflammatory conditions. Future research should prioritize: 1) Defining Causality: employing longitudinal human studies and more sophisticated animal models that better mimic the chronicity and complexity of human cirrhosis to move beyond associations and transform hypothesis-generating findings into established causal mechanisms; 2) Precision Medicine: identifying biomarkers that can predict which patients will respond to specific anti-inflammatory therapies, moving away from a one-size-fits-all approach; 3) Therapeutic Optimization: conducting large-scale, pragmatic clinical trials that not only assess efficacy but also address practical challenges like safety in a fragile population, cost-effectiveness, and long-term management strategies for chronic conditions like HE; and 4) Mechanistic Elucidation: deconvoluting the specific mechanisms by which interventions like FMT and rifaximin exert their effects, moving from emerging observations to established understanding, to develop more targeted and reproducible therapies.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data availability statement

This is a review article. No new data were created or analyzed in this work. All discussed information is based on previously published studies cited in the reference list.