Current progress in the pathophysiology and potential therapeutic targets of acute-on-chronic liver failure through multiomics approaches

Abstract

Acute-on-chronic liver failure (ACLF) is a severe clinical syndrome that develops in patients with chronic hepatitis or liver cirrhosis. It is characterized by a high mortality rate, rapid progression, and complex pathological mechanisms. As it often involves multiorgan failure (e.g., of the liver, kidneys, and brain), traditional clinical indicators have limited predictive value and cannot support accurate, personalized treatment. However, the rapid development of omics sequencing technologies in recent years (e.g., genomics, transcriptomics, proteomics, metabolomics, and microbiomics) has provided unprecedented opportunities to analyze the molecular mechanisms of ACLF, discover novel biomarkers, and develop precise intervention strategies. Multiomics technology is an approach that systematically studies complex biological mechanisms of ACLF by integrating data from multiple omics disciplines. This approach overcomes the limitations of single-omics techniques, and offering a more comprehensive view of disease mechanisms, biomarker discovery, and precision medicine. In this review, we summarize studies on whole-genome sequencing, single-cell ribonucleic acid (RNA) sequencing, proteomics, lipidomics, and metabolomics in patients with ACLF and discuss associations between expression patterns, disease progression and prognosis, and the interconnections among multiomics.

1. Introduction

Acute-on-chronic liver failure (ACLF) is a severe syndrome characterized by acute impact in patients with chronic liver disease.^[1] The common etiopathogenesis of acute liver injury includes hepatitis B virus (HBV) infection, acute viral hepatitis, alcohol consumption, hepatotoxic drugs, and upper gastrointestinal bleeding.^[2,3,4] A meta-analysis of 30 cohort studies revealed that the overall incidence of ACLF was approximately 35%, with the highest incidence in South Asia (65%). Infection is the most frequent precipitating factor, and kidney dysfunction is the most common organ failure. The global 28-day mortality rate is approximately 45%, ranging from 28% in North America to 63% in South America. The worldwide 90-day mortality rate is 58%, with the highest reported in South America (73%) and the lowest in South Asia (68%).^[4]

Currently, the pathophysiological mechanisms of ACLF are not fully understood. A widely accepted theory is that local acute liver injury triggers innate immunity, leading to the release of proinflammatory cytokines and chemokines. These then “spill over” to recruit and activate systemic immune cells and induce an exaggerated systemic inflammatory response syndrome (SIRS). Subsequently, circulating anti-inflammatory factors are overproduced to regulate inflammatory homeostasis, a state known as compensatory anti-inflammatory response syndrome (CARS). CARS can further impair immune function, increasing the risk of multiple organ dysfunction syndrome (MODS) and death.^[5,6,7,8,9]

As the mechanisms are not yet clearly defined, the current management of ACLF is mainly symptomatic treatment and aimed at controlling complications. Liver transplantation is the only treatment that can improve prognosis, but its application is limited by donor shortages and timing constraints.^[10] Treatments involving cytokines, stem cells, and stem cell mobilizers have shown early promise, but they are still in the exploratory stage.^[11]

Early and accurate assessment, as well as halting progression, are crucial for ACLF patients. Currently, ideal early warning assessment methods and targeted treatment modalities are lacking. The rapid development of biotechnology and multiomics has become an important tool in modern disease research and offers scientists unparalleled capabilities to investigate the molecular mechanisms of diseases, cellular heterogeneity, and complex biological processes.^[12,13] Therefore, this review summarizes and discusses findings from multiomics studies on ACLF patients with the aim of constructing a network model of the mechanisms to explicit the occurrence, development, and prognosis of ACLF. The objective is to provide new prospects for the early diagnosis, risk stratification, prognostic assessment, and treatment of ACLF, offering opportunities for the development of precision-based and personalized medicine.

2. Whole-genome sequencing (WGS): Susceptibility loci and risk of ACLF

WGS enables the comprehensive analysis of genetic sequences, which can help identify genetic variations related to disease, known as susceptibility loci. These loci are important for understanding the genetic basis of ACLF, predicting individual risk, and developing targeted treatments.

ACLF shows susceptibility to inflammation-related genes. In a recent article mentioned, six single-nucleotide polymorphisms (SNPs) in such genes were genotyped to examine whether genetic factors influence systemic immune responses. It was found that the interleukin 1 (IL-1) gene cluster was associated with the inflammatory process in ACLF, with rs1143623 and rs4251961 identified as protective SNPs.^[14] Another report from the CANONIC study linked the NOD2-G908R genetic variant to a high risk of mortality, while the MBL_Yx and MASP2_371 variants were associated with an increased risk of bacterial infections.^[15] A genome-wide association study (GWAS) further examined the causal relationship between cytokines and ACLF and found that higher levels of interleukin-13 (IL-13) and lower levels of stem cell factors (SCF) and basic fibroblast growth factors (bFGF) were associated with increased ACLF risk.^[16]

In addition, the genotype characteristics of the major histocompatibility complex (MHC) are associated with HBV-related ACLF (HBV-ACLF). Human leukocyte antigen (HLA)-DRB1*12:02 and rs3129859 were identified as the top susceptible HLA alleles. Further analysis revealed that the rs3129859 allele was also associated with prolonged prothrombin time, faster ascites progression, and a higher 28-day mortality rate.^[17] HLA-DRB1 is primarily expressed on antigen-presenting cells, including dendritic cells, B cells, and monocytes, and activates adaptive immune cells through the presentation of peptide fragments.^[18,19] Therefore, targeting HLA-DRB1 could substantially affect immune system homeostasis. However, the potential risks of such targeted therapies are still unknown and require careful consideration.

3. Transcriptomics: Immune dysregulation as an important link

Although WGS has identified genetic susceptibility to ACLF and suggested associations with immune cell dysfunction, especially monocytes, the characteristics of immune dysregulation cannot be demonstrated from genomic results alone. In contrast, transcriptomics would provide a comprehensive view of the immune profile of patients with ACLF.

Gene expression profiles in ACLF differ significantly from those observed in other chronic liver diseases. They show high expression of genes related to inflammation, fibrosis, apoptosis, senescence, and tumorigenesis but low expression in oxidative stress pathways.^[20] In HBV-ACLF patients, the gene expression profile in peripheral blood mononuclear cells (PBMCs) resembles that observed in liver tissue. This includes high expression of pathways related to proinflammatory factors, impaired energy metabolism, and upregulation of innate immune responses, as well as downregulation of adaptive immune responses.^{[20,21,22,23]} Semaphorin-6B (SEMA6B) was highly expressed in PBMCs and has been proposed as a predictor of HBV-ACLF onset; it is also highly expressed in macrophages and hepatocytes, particularly in patients who die.^[24]

Transcriptional profiling of peripheral blood monocytes from patients with ACLF suggests an upregulation of expression of immunosuppression parameters as well as impaired antibacterial activity and antigen presentation.^[25] Ribonucleic acid (RNA) sequencing of neutrophils has identified a specific phenotype characterized by induced glycolysis and granule genes, and downregulated cell-migration and cell-cycle genes.^[26,27] RNA sequencing of natural killer (NK) cells showed that HBV-ACLF patients with high expression of caspase-3, marker of proliferation Ki-67 (Ki-67), C-X-C motif chemokine receptor 3 (CXCR3), killer cell lectin-like receptor K1 (KLRK1), natural cytotoxicity triggering receptor 1 (NCR1), CD107A, and colony-stimulating factor 2 (CSF2).^[28]

The above studies provide theoretical support for elucidating the detailed molecular basis and mechanisms of ACLF occurrence, development, and prognosis. However, the major shortcomings are that most research has centered on peripheral blood immune cells, with fewer reports on transcriptional changes in the profile of the liver, mainly due to the difficulty in obtaining samples. Therefore, establishing animal models is important for the transcriptomic characterization of ACLF at different disease stages, although current studies in this area are limited.

4. Single-cell sequencing: Exhausted T cells and activated monocytes as important contributors

Single-cell RNA sequencing (ScRNA-seq) is frequently employed to map the immune landscape of PBMCs and liver tissue and to investigate underlying immunological mechanisms. For example, analyses of data from tumor tissues have identified cell subpopulations linked to the development and progression of liver cancer. Clarifying the functions and molecular mechanisms of these subpopulations can aid in the development and optimization of treatment strategies.^[29,30]

In HBV-ACLF, immune exhaustion following excessive activation is associated with disease progression. Pseudotime analysis of scRNA-seq data revealed that T cells transition from naïve to effect states through an intermediate stage, ultimately shifting exhausted T cells, such as CD4+TIGIT+ and CD8+LAG3+. These exhausted T cells exhibit a reduced capacity to produce interleukin-2 (IL-2), interferon-γ (IFN-γ), and tumor necrosis factor-α (TNF-α).^[31]

Compared to adaptive immune cells, innate immune cells respond rapidly to acute injury during the early stages of ACLF, resulting in a significant systemic and organ-specific inflammatory response. Analysis of PBMC scRNA-seq data showed increased expression of gene sets associated with inflammatory response and cell metabolism in ACLF patients, with thrombospondin-1 (THBS1)-mono—which upregulates proinflammatory cytokines and chemokines—being enriched in nonsurvivors.^[32] Further ScRNA-seq analysis of PBMCs from patients who recovered or did not recover from ACLF revealed that classical monocytes (cMons) had two distinct states: a stress-induced tolerant state with an oxidative phosphorylation program and blunted activation among lymphoid populations in patients with ACLF-recovery, and elevated expression of inflammatory and stress response genes in nonrecovery cases.^[33]

ScRNA-seq of liver tissue revealed monocyte and macrophage infiltration, which induces apoptosis and dysfunction of liver lymphoendothelial cells (LyECs) through secretion of secreted phosphoprotein-1 (SPP1), contributing to the progression of HBV-ACLF.^[34] TREM2+ mono has anti-inflammatory characteristics and may be related to immunosuppression triggered by inflammation. It represents the endpoint of monocyte transformation, with upregulated lipid metabolism genes.^[35,36] Neutrophils in liver tissue, which show high expression of C-C chemokine ligand 4 (CCL4), are enriched in HBV-ACLF patients and exhibit strong inflammatory and chemokine-driven responses while relying on oxidative phosphorylation for energy supply.^[36]

Dynamic profiling of peripheral immunity in patients with HBV-ACLF indicates early inflammatory activation of myeloid cells, induction of hepatocyte apoptosis, and eventual imbalance between proinflammatory and anti-inflammatory responses, leading to an immunosuppressive microenvironment and poor prognosis. Restoration of T cell cytotoxicity is associated with favorable outcomes, and alleviating the immunosuppressive effects of CXCR2+ neutrophils contributes to the functional recovery of T cells.^[37] These findings suggest that the early and late stages of ACLF exhibit distinct immunological and pathological characteristics and may require phase-specific treatment strategies. For instance, suppressing excessive immune activation may be beneficial in the early stages, whereas reversing immunosuppression may be more important in later stages. However, current methods for evaluating patients’ pathological status, such as single-cell sequencing, remain overly complex. There is therefore a need to develop simpler and more efficient assessment approaches through integration with clinical data.

Previous studies have reported that during the early stages of HBV-ACLF, monocytes are predominantly the M1 type and secrete proinflammatory cytokines. In the late stages, they shift toward the M2 type and secrete anti-inflammatory cytokines.^[36,38] This change may be related to the pathophysiological course of ACLF, which increases the risk of recurrent infection and death. Interventions that inhibit the transformation of monocytes at different disease stages may improve outcomes. For instance, mesenchymal stem cells (MSCs) and cytokine agents are used to improve treatment outcomes by modulating immunity.^[31,39,40] In our study on treating decompensated cirrhosis, MSCs were found to transform the function of MX1+ monocytes from mediating proinflammatory chemotaxis to antigen presentation, thereby coordinating effector T-cell overactivation and modifying B-cell function.^[41] Therefore, in ACLF treatment, reversing the function of inflammation-activated monocytes may be a key factor in achieving a good prognosis.

5. Proteomics: Constructing diagnostic and prognostic models for ACLF

Proteomics is the large-scale study of protein features, including their expression levels, posttranslational modifications, and protein–protein interactions (PPIs), with the goal of obtaining a comprehensive understanding of disease.

Proteomic studies in patients with ACLF aim to screen proteins associated with disease onset and prognosis and to construct diagnostic or predictive models.^[42,43,44] For example, a model (P4) based on apolipoprotein C3 (APOC3), histidine-rich glycoprotein (HRG), tissue factor (TF), and kallikrein B1 (KLKB1) could effectively differentiate between ACLF and chronic hepatitis B (CHB). Another model (P8), based on vitamin D binding protein (GC), HRG, haptoglobin-related protein (HPR), serpin family a member 6 (SERPINA6), along with age, neutrophil (NEU), international normalized ratio (INR), and total protein, could predict 28- and 96-day mortality rates in ACLF patients.^[42] A model based on plasminogen (PLG) levels, P5, differentiated ACLF severity and predicted mortality more accurately than the Child–Pugh score and the model for end-stage liver disease (MELD) score.^[43]

In the above study, the researchers also identified differentially expressed proteins (DEPs). Gene Ontology (GO) enrichment analysis showed that DEPs were primarily enriched in pathways of fatty acid metabolism and transport, immune response, complement and coagulation systems, protein metabolism, energy pathways, and negative regulation of transporter and hydrolase activity.^[42,43,44] We conducted GO and Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analyses of 91 proteins (shown in Table S1, http://links.lww.com/IDI/A80) using RStudio. The results indicated enriched clusters in complement activation, humoral immunity, complement and coagulation cascade, cholesterol metabolism, and other processes (Figure 1A, 1B). Protein–protein interaction (PPI) network construction using the Search Tool for the Retrieval of Interaction Gene/Proteins (STRING) database (https://cn.string-db.org/) showed that 76 proteins interacted, among which PLG, fibrinogen alpha chain (FGA), serpin family F member 2 (SERPINF2), serpin family C member 1 (SERPINC1), and vitronectin (VTN) were centrally located and involved in various pathways (Figure 1C).

Figure 1.

GO pathway enrichment (A), pathway enrichment (B), and PPI network (C) of 91 proteins. GO: Gene ontology; KEGG: Kyoto Encyclopedia of Genes and Genomes; PPI: Protein–protein interaction; 91 proteins obtained from three articles published by Sun Z^[42], Wu D^[43], and Acharya P^[44] respectively.

Proteomic results suggest that coagulation abnormalities and abnormal lipid metabolism are the main causes of ACLF progression and a poor prognosis. Therefore, therapies targeting relevant proteins, such as PLG, may improve patient outcomes. A search of the ChEMBL database (https://www.ebi.ac.uk/chembl/) revealed that PLG inhibitors, including aminocaproic acid, tranexamic acid, and aprotinin, are commonly used in hemorrhagic disorders; whether they can be used in the treatment of ACLF requires further investigation.

Recent studies of coronavirus disease 2019 (COVID-19) patients have shown that platelet activation leading to increased monocyte TF expression is associated with disease severity and high mortality.^[45] However, abnormalities in the coagulation system and monocytes also coexist in ACLF patients, which suggests that further studies are needed to determine whether the same pathogenesis promotes disease progression in ACLF.

6. Lipidomics: Interaction with cell injury and immune response imbalance

Lipids function as signaling molecules for intercellular interactions and homeostasis maintenance.^[46,47] Alterations in lipid composition can interfere with immune regulation, leading to immune metabolism dysregulation.^[48] As the liver is a major source of lipid production, lipid abnormalities are common in ACLF patients. However, the role of lipid mediators in the development of systemic inflammation and organ failure remains unexplored.

Untargeted lipidomics showed that declining lipid levels parallel disease severity, that cholesteryl ester (CE) levels best differentiate between acutely decompensated cirrhosis and ACLF, and that liver failure and infections are major causes of abnormal CE levels.^[49] Targeted lipidomics results suggested that leukotriene E4 (LET4) levels positively correlate with ACLF severity; moreover, increased 12,13-epoxy-9-keto-10(trans)octadecenoic acid (EKODE) and LET4, together with reduced lipoxin A5 (LXA5), were significantly associated with 28-day mortality in ACLF.^[50]

The lipids (shown in Table S1, http://links.lww.com/IDI/A80) were analyzed using MetaboAnalyst (https://www.metaboanalyst.ca/), and the enrichment analysis was performed using the metabolic pathway relationship database. The results suggested that the pathways of glycerophospholipid catabolism, linoleic acid metabolism affected by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), glycerophospholipid biosynthesis, and phosphatidyl inositol (PI) metabolism were enriched (Figure 2A). Therefore, abnormal glycerophospholipid metabolism may be the primary cause of ACLF, according to the results of enrichment. Glycerophospholipids are major components of cell and organelle membranes, and hepatocyte injury associated with acute stimulation and immune activation may be the main cause of their abnormal levels, though the exact mechanism requires further investigation.

Figure 2.

Pathway enrichment of 28 lipids, pathway and network enrichment and of 125 metabolites, as well as the protein–metabolite interaction network pathway enrichment dot plot of 28 lipids^[49,50] (A), pathway enrichment dot plot of 125 metabolites^{[52,53,54,55]} (B), the network of pathway enrichment of 125 metabolites^{[52,53,54,55]} (C) constructed by the RaMP-DB database respectively; (D) protein–metabolite interaction network based on 91 proteins and 125 metabolites constructed the STITCH database^{[42,43,44,52,53,54,55]}.

An interaction between monocyte macrophages and lipid metabolism has been suggested.^[51] In ACLF, abnormalities in monocytes, lipid metabolism-related proteins, and lipid levels have been found, but no study has constructed a network of these interactions to investigate the underlying mechanism. This may be a promising direction for future research.

7. Metabolomics: Terminal biomarkers reflecting abnormalities in cytomics, proteomics, and lipidomics

Compared to proteomics and lipidomics, metabolomics lies downstream. Metabolites are the end products of gene expression; therefore, metabolic changes more directly reflect the pathophysiological characteristics of ACLF.

Untargeted metabolomics of plasma showed that metabolite levels were significantly elevated in patients with HBV-ACLF and were associated with renal failure but not with infection.^[52] Categorization of altered metabolites revealed that ACLF was associated with increased protein and lipid hydrolysis, intracellular glucose metabolism, and oxidative stress as well as decreased mitochondrial beta-oxidation.^{[52,53,54,55,56]} In addition, increased levels of inflammatory factors and estrogen, alongside decreased levels of thyroid and steroid hormones, were associated with the development of ALCF.^[53,55] These studies also found significantly higher levels of metabolites derived from intestinal microbes.^[53,55] Targeted lipidomics analysis revealed that the lysophosphatidylcholine/autotaxin/lysophosphatidic acid axis plays an important role in ACLF progression by affecting monocyte activation and increasing proinflammatory cytokine production.^[57]

Researchers have constructed models to predict the occurrence and mortality of ACLF based on metabolomic changes. Weiss et al^[54] constructed prediction models (CLIF-C MET models 1 and 2) using 4-hydroxy-3-methoxyphenylglycol sulfate, hexanoylcarnitine, D-galacturonic acid, age, bilirubin, INR, and MET model 2 (shown in Table S1, http://links.lww.com/IDI/A80). The results suggested that the CLIF-C MET model had a better predictive value for ACLF prognosis than the MELD-Na score. Zhang et al^[53] used pipecolate and gamma-carboxyethyl hydroxychroman (γ-CEHC) with total bilirubin, INR, and the presence of cirrhosis to construct a model for predicting the occurrence of ACLF, and used pipecolate, N-acetyl-aspartyl-glutamate (NAAG), and ureidopropionate with INR, neutrophil-to-lymphocyte ratio, creatinine, sodium, albumin, and hepatic encephalopathy grade to construct a model for predicting 90-day mortality. The results showed that the occurrence prediction model was superior to MELD-Na score and that the prognostic prediction model for ACLF was superior to CLIF-C ACLF.

Enrichment analysis revealed that metabolites were enriched in pathways of amino acid metabolism and derivatives, solute carrier (SLC)-mediated transmembrane transport, amino acid transport across the plasma membrane, and transport of bile salts, organic acids, metal ions, and amine compounds, among others (Figure 2B). An enrichment pathway interaction network was constructed, revealing two general parts: lipid transport–related pathways and lipid metabolism–related pathways (Figure 2C). Protein−metabolite interaction analysis demonstrated that PLG, F2, KNG1, CPB2, L-lysine, glycine, and citrulline formed the interaction network, with L-lysine being a central node (Figure 2D). These results may provide a novel perspective for identifying new targets in the treatment of ACLF; however, given that the results integrate data from multiple studies, a validation study is required.

The main targets for targeted metabolite therapy are upstream of the pathway, such as metabolite sources and metabolism-related enzyme proteins. Therefore, constructing a metabolic network based on multiomics analysis would help identify critical metabolic targets for more precise ACLF therapy. However, no such study has been reported, making this area worthy of further investigation.

8. Gut microbiota: Declining abundance induces ACLF progression

The liver–gut axis exhibits bidirectional crosstalk to maintain homeostasis.^[58] However, in the presence of liver disease, impairment of the intestinal mucosal barrier facilitates the translocation of intestinal flora and their products, thereby promoting disease progression through the induction of inflammatory responses.^[59]

Several studies have found that gut microbial abundance continues to decline as liver disease progresses.^[60,61,62] Stratified analyses based on the presence or absence of complications found a further reduction in the abundance of microbiota in patients with decompensated cirrhosis compared to those with compensated cirrhosis.^[62] Species analysis of gut flora revealed an increase in the abundance of Enterococcus, Streptococcus digestans, Streptococcus, Thickettsia, Aspergillus, and Actinobacteria in HBV-ACLF patients but a decrease in the abundance of genera such as Bacteroidetes, Roseburia, and Firmicutes.^[60,61,62] Changes in the abundance of specific gut microbiota can be used to predict the occurrence and prognosis of ACLF. For example, incorporating microbiota such as Enterobacteriaceae and Enterococcaceae in models can improve the predictive utility for ACLF occurrence.^[62] Higher abundances of Enterococcus, Streptococcus thermophilus, and Ruminococcus lactaris can predict disease progression and death, while higher abundances of Paraprevotella clara, Bacteroides salyersiae, Clostridium sp., Roseburia hominis, and Faecalibacterium are usually associated with a good prognosis.^[61,62] A combined metabolomics and gut microbiology study found that sex steroids and benzoate compounds associated with Enterococcaceae are also linked to high mortality rates in ACLF.^[55]

Enterococcaceae are a class of conditionally pathogenic bacteria that are universally resistant to commonly used antibiotics and are a major cause of nosocomial infections.^[63] When Enterococcaceae colonize in the intestinal tract, the use of antibiotics can allow them to become the dominant strain, promoting the progression of ACLF through the secretion of virulence factors and cellular adhesion. Currently, dalbavancin, oritavancin, tedizolide, and telavancin are used to combat multidrug resistance; however, whether they can be used to treat secondary infections in patients with ACLF to improve prognosis requires further investigation.^[64,65]

9. Multi-omics cross-talk: A driving force in ACLF pathogenesis

Integrated multiomics analyses reveal extensive cross-talk among immune components, proteins, lipids, gut microbiota, and metabolites, all of which are closely associated with ACLF progression and prognosis. Patients with ACLF exhibit a distinct immune profile characterized by hyperinflammatory monocyte activation and concurrent exhaustion of energy metabolism. Proteomic pathway enrichment analysis indicates dysregulation in complement activation, coagulation cascades, and lipid metabolism pathways. Lipidomics profiling demonstrates predominant alterations in sex steroids, CE, and glycerophospholipids, while metabolomic analysis identifies significant upregulation of proteolysis, lipolysis, and glycolysis. Furthermore, gut microbiota contribute to the generation of specific abnormal metabolites.

Collectively, these findings delineate a pathophysiological vicious cycle: severe hepatocyte injury in ACLF impairs coagulation and reduces the synthesis of proteins, lipids, and glycogen. To compensate for energy deficits, proteolysis and lipolysis are markedly increased. This metabolic reprogramming shifts monocytes toward anaerobic glycolysis. Simultaneously, lipids and their metabolites directly exacerbate monocytic inflammatory activation, thereby collectively promoting the initiation and progression of ACLF.

PPI network analysis revealed that PLG is centrally located and interacts with other DEPs. Protein−metabolite interaction network analysis also suggested that PLG is involved in regulating differential metabolites. These results indicate that PLG plays an important role in ACLF. Some researchers have found that proinflammatory macrophages highly express PLG receptors and that their migration is highly dependent on PLG levels. PLG can promote macrophage polarization toward the M2 phenotype and alleviate inflammatory response.^[66,67] However, it is currently unknown whether similar pathophysiological changes exist in ACLF patients. In addition, most current multiomics studies are cross-sectional, and it is unclear whether PLG and monocyte-macrophages change with disease progression. Nevertheless, these findings imply that drugs such as PLG activator inhibitor-1 could play a role in ACLF treatment by inhibiting M1 polarization of macrophages.

Furthermore, an integrated analysis of multiomics data from public databases on HBV-ACLF has identified cell division cycle 20 (CDC20), C-X-C motif chemokine ligand 14 (CXCL14), Fc gamma receptor IIb (FCGR2B), hexokinase domain containing 1 (HKDC1), and G protein−coupled bile acid receptor 1 (GPBAR1) as biomarkers of M1-like macrophage polarization for diagnosing HBV-ACLF, with enrichment in tryptophan metabolism and cofactor biosynthesis pathways. Investigation of the long non-coding RNA (lncRNA)–microRNA (miRNA)–messenger RNA (mRNA) regulatory network confirmed the involvement of the X inactive specific transcript (XIST)/miR-296-3p/CXCL14 and small nucleolar RNA host gene 14 (SNHG14)/miR-510-5p/CXCL14 axes in disease progression. Bile acid and deoxycholic acid exhibited strong binding affinity to these five biomarkers, suggesting their potential as candidate therapeutic agents.^[68]

These insights point to the value of multiomics integration and the establishment of large-scale prospective cohorts as an essential step toward improving the robustness and clinical applicability of such findings.

10. Conclusion

The development of ACLF is influenced by multiple factors. Genomic studies have revealed genetic susceptibility and driver mutations; transcriptomic analyses have examined mechanisms related to immune dysregulation; proteomic studies have identified novel biomarkers and therapeutic targets; metabolomic investigations have clarified the direct function of energy metabolism disorders; and microbiomics analyses have confirmed the importance of the gut−liver axis.

The complexity and heterogeneity of ACLF mean that a single omics approach is limited; therefore, integrating multiomics analyses will be central to future research. Horizontal integration—such as the combination of genomics, transcriptomics, proteomics, and metabolomics—can be used to analyze how genetic variations affect gene expression and downstream molecules, thereby elucidating the specific mechanisms of ACLF. Through longitudinal integration, researchers can dynamically monitor multiomic changes in patients before and after treatment to identify key driver molecules related to ACLF progression and prognosis. However, several bottlenecks still remained. For example, research cohorts for HBV-ACLF primarily enroll Asian populations, whereas studies on non-HBV-ACLF predominantly include non-Asian populations. Although divergent findings have been observed across such studies, substantial ethnic disparities limit the direct comparability of findings. Consequently, establishing ethnically homogeneous research cohorts is necessary to characterize the multiomic signatures of various underlying causes. In addition, data integration across different omics platforms is hindered by technical barriers, resulting in a lack of standardization. As most studies are still in the discovery stage, prospective cohort studies are needed to verify the clinical utility of proposed biomarkers.

Nonetheless, ongoing technological innovation and interdisciplinary cooperation are paving the way for multi-omics analysis to achieve several significant goals:

(1) Early diagnosis: constructing a new biomarker panel to improve the accuracy of early ACLF prediction.

(2) Precise classification: guiding personalized, precise treatment strategies.

(3) Novel therapies: developing new interventions targeting immune regulation, microbial intervention, and metabolic regulation.

(4) Dynamic prognosis assessment: using liquid biopsy for real-time monitoring of treatment response to adjust intervention strategies.

Overall, the decreasing cost of sequencing and advances in computational biology mean that multiomics-driven precision medicine is expected to play a key role in managing ACLF and ultimately improving patient prognosis.

Author Contributions

Yingying Yu, Mengyao Li, Yunbo Xie, Fu-sheng Wang, Ming Shi made substantial contributions to the study concept and design, first drafting of the manuscript, and critical revision of the manuscript for important intellectual content, All authors read and approved the final manuscript.

Conflicts of Interest

None.

Editor note:

Fu-Sheng Wang is the editor of Infectious Diseases & Immunity. The article was subject to the journal’s standard procedures, with peer review handled independently by this editor and his research group.

Funding Statement

This work was supported by National Key Research and Development Program of China (2022YFC2304402).