摘要
Acute-on-chronic liver failure’s (ACLF) complex pathophysiology involves systemic inflammation, immune-mediated tissue damage, and mitochondrial dysfunction that results in organ failure.1–3 Despite global variation in diagnostic criteria, AASLD guidance stresses the need for acute onset and rapid deterioration, new liver failure in a patient with chronic liver disease, and at least 1 extrahepatic organ failure.4 Consensus global diagnostic standards are needed to move the field forward and start developing ACLF-specific treatments, or better yet, preventions. Toward the goal of ACLF-specific biomarkers and therapeutic development, some have encouraged the addition of the possibility of recovery into any future definition to focus the field on early diagnosis when therapeutic intervention may alter the outcome. Feio-Azevedo and colleagues, in this issue of Hepatology, aim to understand the immunophenotype and function, metabolic function, and pathobiology associated with ACLF recovery (ACLF-R, defined as resolution of ACLF or a decrease of ≥2 grades by EASL-CLIF criteria at day 28) versus nonrecovery (ACLF-NR, defined as progression of >1 stage of ACLF by EASL-CLIF criteria or death).2,5 They should be commended for their innovative approach; they utilized single-cell transcriptional profiling in a discovery cohort of healthy patients, patients with acute decompensation, and patients with ACLF, and validated their findings in a separate cohort with bulk RNA-sequencing, metabolomics, and functional assays to better understand the immune-metabolic complexities of ACLF-R versus ACLF-NR. A key strength of the study is the identification of divergent classical monocyte (cMon) populations associated with ACLF-R versus ACLF-NR (Figure 1). Indeed, cMons may be the principal players and regulators of pro- versus anti-inflammation. ACLF results from an exaggerated proinflammatory response to a known (infection, ischemia, bleeding, surgery, etc.) or unknown precipitant(s). However, the crucial elements needed to predict and, more importantly, to promote recovery following the onset of ACLF remain elusive.FIGURE 1: Differences in the classical monocytes (c Mon) of patients who recover (ACLF-R) vs. do not recover (ACLF-NR) from ACLF.The authors show that ACLF-R cMons are characterized by large increases in the expression of resistin (RETN, which prevents excess inflammation to lipopolysacchaside) and alarmins (S100A8/9/11, which make monocytes less responsive to TLR4 stimulation). cMons associated with ACLF-R also have increased stress tolerance that downregulates inflammation. In contrast, ACLF-NR cMons have increased inflammatory thromboxanes and inflammasome activators—all factors promoting further inflammation. In addition, there are increased metabolic transcripts involved in the pentose phosphate pathway, an essential component of the proinflammatory and oxidative burst reactions in macrophages. A recent study revealed a 364-fold upregulation in the pentose phosphate pathway in patients with ACLF compared to healthy controls, emphasizing its important association with ACLF.6 Functionally, ACLF-R cMons exhibit a reduced phagocytic capacity, antigen presentation, and increased IL-10 production in response to lipopolysacchaside, along with increased ATP synthesis and protein metabolism. In contrast, ACLF-NR cMons demonstrate increased antigen processing and presentation, leukocyte activation, cell migration, and inflammatory responses. The functional assays that elucidated these results shed light on the cellular mechanisms that mitigate inflammation in ACLF-R and maintain systemic hyper-inflammation in ACLF-NR. These distinct cellular states reveal immunometabolic differences that correlate with patient outcomes. ACLF-R cMons demonstrate robust metabolic activity in crucial pathways, whereas ACLF-NR cMons display metabolic deficiencies. Specifically, ACLF-R cMons have upregulation in glutathione-related genes, oxidative phosphorylation, and glycolysis/gluconeogenesis and downregulation in arachidonic acid metabolism and the nonoxidative pentose phosphate pathway. In addition, ACLF-R cMons show higher levels of metabolites like glutamine, glutamate, and α-KG, suggesting a more effective tricarboxylic acid cycle cycle function and potential inflammatory response moderation through glutaminolysis and α-KG’s suppression that can promote endotoxin tolerance. In contrast, ACLF-NR cMons showed diminished activity in metabolic pathways linked to amino acid and nucleoside (specifically adenosine and uridine) biosynthesis. When analyzing T-cell subsets, patients with ACLF-R had fewer activated CD8 T cytolytic cells (CD8_CTL), and the CD8_CTL cells present had reduced function, as evidenced by less activation, migration, and cytokine transcriptional activity. In addition, B cells and plasma cells of patients with ACLF-R were no different in number but had reduced activation compared to patients with ACLF-NR. NK cell subsets were increased in all patients with ACLF, but both major subtypes had reduced function in patients with ACLF-R compared to patients with ACLF-NR. These contrasting functional profiles provide a deeper understanding of the mechanisms influencing ACLF outcomes. The exaggerated immune response to the ACLF precipitant must be shut off; cMons are certainly associated with but may play a pivotal role in modulating lymphocyte activation and cytokine secretion that impacts the function of T cells, B cells, and NK cells to facilitate ACLF-R versus ACLF-NR. Ultimately, the study uncovers a better stress-adapted and less reactive state in both cMon and other immune cells of patients with ACLF-R. Given the general role antigen-presenting cells play in effector cell education, it is possible this is driven by cMon leaders and requires adequate anti-inflammatory cytokine and antioxidant presence, nutritional support, and mitochondrial function to achieve ACLF-R. The mitigation of oxidative stress in ACLF through the lens of immunometabolism reveals that patients with ACLF-R have increased activity in oxidative phosphorylation and glutathione metabolism—key anti-inflammatory and redox-regulating pathways. These findings highlight the therapeutic potential of targeting antioxidant pathways to counteract- inflammation and oxidative stress in ACLF and may be targets for future therapeutic intervention. However, the key will be immune modulation without clinical immunosuppression. The study is not without limitations. Its major limitations include the sourcing of all clinical samples from a single center, the small total sample size in both the discovery and validation cohorts, and the lack of validation for each study method utilized. The focus on alcohol-associated liver disease is both a strength (uniform population) and a weakness (may not apply to other disease etiologies). In addition, the human liver has a vast and complex array of monocytes and subsets of lymphocytes (during inflammation), and we assume that the peripheral blood analyses done in this paper reflect the important subsets present in the liver. The manuscript provides an in-depth analysis at one time point; however, the trajectory of (1) the number, activation state, and functionality of cell types, (2) pro- and anti-inflammatory cytokines, (3) immunometabolism alterations, and (4) the presence or absence of exhaustion were not analyzed over time and remains of great interest. Lastly, many associations are documented, which do not translate into the determination of causation. This study reminds us of the importance of nutrition in hospitalized patients with liver disease. Not only do lower glycogen stores in patients with cirrhosis result in overnight fasts, increasing muscle breakdown and the risk for frailly, but our sickest patients with ACLF are often left without nutrition for days, waiting for enteral tube placement or other procedures. This can lead to villous mass reduction and increase bacterial translocation.7,8 It is imperative to remember that even patients without robust gut function can often tolerate trickle tube feeds, which promote tight junctions and decrease bacterial translocation. Nutrition as a therapeutic modality to improve mortality is not new; in alcohol-associated hepatitis, nutritional optimization or supplementation has long been the standard of care to decrease mortality.9 However, could tailored dietary intervention with specific amino acids, antioxidants, and anti-inflammatory preparations impact the phenotype and function of cMon or other immune cells? Would this approach be most effective alone or in combination with microbiome-altering interventions to eliminate pathobionts and increase bacteria-producing anti-inflammatory products? In addition, it remains unclear how plastic the phenotype of cMon and other immune cells are, and whether disrupted intracellular metabolic pathways can be restored (and if so, how quickly). Ultimately, there may be a point of no return that needs to be identified. Hopefully, future research will elucidate the answer to these and other questions. In conclusion, the immunologic paradox couples chronic immune cell activation with local and systemic impaired immune cell function and predisposes patients with cirrhosis to infections and ACLF. Cells present and metabolites produced before or during ACLF can serve as biomarkers, aiding in patient risk stratification and providing insight into disease severity and recovery potential.3,10 Exploring monocyte tolerance mechanisms, such as genetic, epigenetic, and transcriptional regulation, may guide future interventions to alter immune function.11 Ultimately, effective treatments hinge on early diagnosis, risk stratification, and a nuanced understanding of ACLF’s immune-metabolic aspects. Given the high risk of death in all patients, once ACLF occurs, the long-term goal should be to develop biomarkers to predict ACLF and then develop interventions to prevent it. However, for now, Feio-Azevedo and colleagues have taught us that our first step toward understanding ACLF-R versus ACLF-NR requires asking the monocyte.