Acute-on-chronic liver failure: Definitions, pathophysiology and principles of treatment

Summary

The term acute-on-chronic liver failure (ACLF) defines an abrupt and life-threatening worsening of clinical conditions in patients with cirrhosis or chronic liver disease. In recent years, different definitions and diagnostic criteria for the syndrome have been proposed by the major international scientific societies. The main controversies relate to the type of acute insult (specifically hepatic or also extrahepatic), the stage of underlying liver disease (cirrhosis or chronic hepatitis) and the concomitant extrahepatic organ failure(s) that should be considered in the definition of ACLF. Therefore, different severity criteria and prognostic scores have been proposed and validated. Current evidence shows that the pathophysiology of ACLF is closely associated with an intense systemic inflammation sustained by circulating pathogen-associated molecular patterns and damage-associated molecular patterns. The development of organ failures may be a result of a combination of tissue hypoperfusion, direct immune-mediated damage and mitochondrial dysfunction. Management of ACLF is currently based on the supportive treatment of organ failures, mainly in an intensive care setting. For selected patients, liver transplantation is an effective treatment that offers a good long-term prognosis. Future studies on potential mechanistic treatments that improve patient survival are eagerly awaited.

Key points.

• •Different definitions of ACLF have been proposed by the major international scientific societies, depending on the types of precipitants and organ failures included.
• •According to the definition used, the clinical phenotypes of patients with ACLF vary in terms of aetiology of underlying liver diseases, nature of precipitating event and patient prognosis.
• •In ACLF, pathogen-associated molecular patterns and damage-associated molecular patterns are the drivers of an intense systemic inflammation which is also associated with features of immunosuppression.
• •Systemic inflammation is involved in the development of organ failures through tissue hypoperfusion, immune-mediated tissue damage and mitochondrial dysfunction.
• •Current management of ACLF relies on supportive therapy for organ failures.
• •Given the high rate of sepsis-related ACLF, an empirical antibiotic therapy tailored to the suspected site of the infection and the local ecology should be rapidly initiated.
• •Albumin treatment may have beneficial effects on systemic inflammation and infusion is recommended after high volume paracentesis, in case of spontaneous bacterial peritonitis, and in patients with AKI KDIGO stage 2-3.
• •Patients with ACLF should be promptly assessed for liver transplantation.

Background

It is universally accepted that the term “acutely decompensated cirrhosis” defines patients with cirrhosis who are non-electively admitted to the hospital for recent onset ascites, gastrointestinal haemorrhage, newly developed hepatic encephalopathy, bacterial infections, or any combination of these disorders.[1], [2], [3], [4], [5] Recently, the term acute-on-chronic liver failure (ACLF) has been used to define a syndrome which was observed among patients with acutely decompensated cirrhosis and characterised by high 28-day mortality.² The other characteristics of ACLF included its association with an intense systemic inflammatory response, its frequent and close association with a precipitating condition (infections, alcoholic hepatitis), and its association with single- or multiple organ failures (OFs). However, there is not yet a universally recognised definition of ACLF. Herein, we summarise the current knowledge and controversies in ACLF.

Definitions of ACLF

In the last decade, different definitions of ACLF have been developed by international consortia.[2], [3], [4], [5], [6], [7] These definitions are summarised in Table 1.

Table 1.

Characteristics of definitions of ACLF developed by 4 different consortia.

Characteristics	European Association for the Study of the Liver - Chronic Liver Failure (EASL-CLIF) Consortium∗	North American Consortium for the Study of End-stage Liver Disease (NACSELD)	Chinese Group on the Study of Severe Hepatitis B (COSSH)	Asian Pacific Association for the Study of the Liver (APASL) ACLF Research Consortium (AARC)∗∗
Category of article(s) defining ACLF	Original article reporting the results of the CANONIC study, which is a prospective, observational study performed in 1,343 patients with cirrhosis non-electively admitted to 29 liver units in 12 European countries2	Original article reporting results of an analysis of 507 patients with cirrhosis whose data were prospectively collected in the NACSELD database, which includes non-electively hospitalised patients in 18 liver units across the USA and Canada3	Original article reporting the results of the COSSH study, which is a prospective, observational study performed in 1,322 patients with cirrhosis or severe liver injury due to chronic hepatitis B, non-electively hospitalised in 13 liver centres in China4	Consensus document involving international experts from the APASL, published in 20096 and updated in 20147 and 2019,5 in the context of AARC; the last 2 updates used internally reviewed data from 1,402 patients, and 3,300 patients, respectively
Patients considered in the definition	Patients with acutely decompensated cirrhosis, with or without prior episode(s) of decompensation	Patients with acutely decompensated cirrhosis, with or without prior episode(s) of decompensation	Patients with acute decompensation of HBV-related chronic liver disease, with or without cirrhosis	Patients with compensated cirrhosis (diagnosed or undiagnosed) or non-cirrhotic chronic liver disease, who had a first episode of acute liver deterioration due to an acute insult directed to the liver
Precipitating disorders	Intrahepatic (alcoholic hepatitis), extrahepatic (infection, gastrointestinal haemorrhage), or both	Extrahepatic (infection)	Intrahepatic (HBV reactivation), extrahepatic (bacterial infection) or both	Intrahepatic
Major organ systems considered for the definition	There are 6: liver, kidney, brain, coagulation, circulation and respiration (see Fig. 1A)	There are 4: kidney, brain, circulation and respiration (see Fig. 1B). Liver and coagulation are not considered	There are 6: Liver, kidney, brain, coagulation, circulation and respiration (see Fig. 2B)	Liver dysfunction is central to the definition; hepatic encephalopathy may be present, as a consequence
Basis of the definition	The definition of ACLF is based on the existence of the failure of 1 of the 6 major organ systems. The failure of each organ system is assessed using the CLIF-C Organ Failure scale∗	The definition of ACLF is based on the existence of 2 organ system failures or more (maximum 4) (see Fig. 1B)	The definition of ACLF is based on the failure of 1 of the 6 major organ systems. The failure of each organ system is assessed using the CLIF-C Organ Failure scale∗	The definition of ACLF is based on the presence of liver dysfunction. Extrahepatic organ failures may subsequently develop but are not included in the definition
Definition and stratification of ACLF	ACLF is divided into 3 grades of increasing severity. • ACLF grade 1 includes 3 subgroups: − patients with single kidney failure − patients with single liver, coagulation, circulatory or lung failure that is associated with creatinine levels ranging from 1.5 mg/dl to 1.9 mg/dl or hepatic encephalopathy grade 1 or grade 2, or both − patients with single brain failure with creatinine levels ranging from 1.5 mg/dl to 1.9 mg/dl •ACLF grade 2 includes patients with 2 organ failures •ACLF grade 3 includes patients with 3 organ failures or more had ACLF grade 3	Patients are stratified according to the number of organ failures 2, 3, or all 4 organ failures, respectively	ACLF is divided into 3 grades of increasing severity. • ACLF grade 1 includes 4 subgroups: − patients with single kidney failure − patients with single liver failure and either an INR of 1.5 or more, creatinine levels ranging from 1.5 mg/dl to 1.9 mg/dl, hepatic encephalopathy grade I or II, or any combination of these alterations − patients with single type of organ failure of the coagulation, circulatory or respiratory systems and either creatinine levels ranging from 1.5 mg/dl to 1.9 mg/dl, hepatic encephalopathy grade I or II, or both − patients with cerebral failure alone plus creatinine levels ranging from 1.5 mg/dl to 1.9 mg/dl •ACLF grade 2 includes patients with 2 organ failures •ACLF grade 3 includes patients with 3 organ failures or more	Acute hepatic insult manifesting as jaundice (total bilirubin levels of 5 mg/dl or more) and coagulopathy (INR of 1.5 or more, or prothrombin activity of less than 40%) complicated within 4 weeks by clinical ascites, encephalopathy, or both. The severity of ACLF is assessed using the AARC score∗∗ (see Fig. 2C). The grading system, defines Grade 1 by scores of 5–7, Grade 2 by scores 8–10 and Grade 3 for 11–15 (see Fig. 2D)
Short-term mortality rate of ACLF according to stratification	By 28 days: Grade 1: 20% Grade 2: 30% Grade 3: 80%	By 30 days: 2 organ failures: 49% 3 organ failures: 64% 4 organ failures: 77%	By 28 days: Grade 1: 23% Grade 2: 61% Grade 3: 93%	By 28 days: Grade 1: 13% Grade 2: 45% Grade 3: 86%

^∗The CLIF-C OF score includes sub-scores ranging from 1 to 3 for each of 6 components (liver, kidneys, brain, coagulation, circulation, and lungs), with higher scores indicating more severe organ system impairment. Aggregated scores range from 6 to 18 and provide information on overall severity. See also Fig. 1A and ref.8.

^∗∗The AARC score includes sub-scores ranging from 1 to 3 for each of 5 components (total bilirubin, hepatic encephalopathy grade, INR, creatinine levels, blood lactate levels). Aggregated scores range from 5 to 15, with higher scores indicating more severe ACLF. See also Fig. 2C, 2D and ref.5. ACLF, acute-on-chronic liver failure; INR, international normalised ratio.

Definitions according to the European, North American and Chinese consortia

All these definitions account for intra- and extrahepatic precipitants of ACLF and consider both liver and extrahepatic OFs. Moreover, patients who have had prior episode(s) of liver disease decompensation are not excluded by these definitions.

In 2013, the European Association for the Study of the Liver – Chronic Liver Failure (EASL-CLIF) Consortium proposed a definition based on the results of the large (1,343 patients), prospective, observational CANONIC study.² This definition applies to patients non-electively hospitalised for acutely decompensated cirrhosis. The diagnosis of organ failures is based on the Chronic Liver Failure-Consortium (CLIF-C) OF (CLIF-C OF) scoring system which assesses 6 organ systems (liver, kidney, brain, coagulation, circulation, and respiration) (Fig. 1A).⁸ The European definition of ACLF includes patients with a high risk of short-term death (including patients with single kidney failure; those with single “non-kidney” organ failure if it is associated with kidney of brain dysfunction; and those with ≥2 OFs) (Table 1). Accordingly, 4 groups of patients with acutely decompensated cirrhosis are defined: 1 group of patients without ACLF and 3 groups of patients with increasing severity of ACLF (grade 1, grade 2 and grade 3) on the basis of the type and number of OF(s).²

Fig. 1.

Definitions of organ system failures used by European and Chinese investigators and North American investigators for defining ACLF.

(A) Chronic Liver Failure-Consortium Organ Failure (known as CLIF-C OF) scale used by investigators from Europe (European Association for the Study of the Liver – Chronic Liver Failure Consortium)) and China (Chinese Group on the Study of Severe Hepatitis B).^4,6 The red and yellow boxes indicate the thresholds for organ system failure and organ dysfunction, respectively. (B) Definitions of organ system failures used by the investigators of the North American Consortium for the Study of End-stage Liver Disease (known as NACSELD).³ E, epinephrine; FiO₂, fraction of inspired oxygen; HE, hepatic encephalopathy; INR, international normalised ratio; MAP, mean arterial pressure; NE, norepinephrine; PaO₂, partial pressure of arterial oxygen; RRT, renal replacement therapy; SpO₂, oxygen saturation as measured by pulse oximetry.

The definition by the North American Consortium for the Study of End-stage Liver Disease (NACSELD) is based on observational data from 507 patients with acutely decompensated cirrhosis non-electively hospitalised for infection (Table 1).³ The North American definition uses standard definitions of shock, the need for mechanical ventilation, the need for renal replacement therapy and West Haven grade III or IV hepatic encephalopathy for the diagnosis of extrahepatic OFs (Fig. 1B). This definition does not include changes in liver function and coagulation. ACLF is defined by the presence of ≥2 extrahepatic OFs.³ A second study by the NACSELD has validated the definition of ACLF in a large cohort of patients with acutely decompensated cirrhosis, precipitated or not by infection.⁹

The Chinese Group on the Study of Severe Hepatitis B (COSSH) developed a definition for HBV-related ACLF.⁴ This definition applies to patients with acutely decompensated HBV-related chronic liver disease (with or without cirrhosis). The Chinese investigators use the CLIF-C OF scoring system for the diagnosis of OFs and distinguish 3 grades of ACLF which are very similar to those defined by the European investigators. However, ACLF grade 1 in the Chinese classification includes an additional subgroup comprising patients with single liver failure who have an international normalised ratio (INR) of ≥1.5 (Table 1).⁴

Definition according to the Asian Pacific Association for the Study of the Liver ACLF Research Consortium

Based on expert opinion, the Asian Pacific Association for the Study of the Liver (APASL) published a definition in 2009,⁶ which was subsequently updated by the APASL ACLF Research Consortium (AARC) in 2014⁷ and 2019.⁵ This definition considers patients with compensated cirrhosis (diagnosed or undiagnosed) and those with non-cirrhotic chronic liver disease, who have a first episode of acute liver deterioration due to an acute insult directed to the liver. The acute hepatic insult is defined by jaundice (total bilirubin levels of ≥5 mg/dl) and coagulopathy (INR of ≥1.5, or prothrombin activity of <40%) complicated within 4 weeks by clinical ascites, hepatic encephalopathy, or both.⁵ Patients who have extrahepatic precipitants and those with kidney, circulatory, or respiratory failures are excluded from this definition (Table 1).

Clinical phenotypes

The clinical phenotypes of patients with ACLF differ according to the definition used.

ACLF phenotype according to the EASL-CLIF definition

In the CANONIC study, the primary causes of cirrhosis were alcohol-related liver disease, followed by chronic hepatitis C and a combination of both.² The most frequent precipitating disorders for acutely decompensated cirrhosis, with or without ACLF, were both hepatic (alcohol-induced liver injury) and extrahepatic (bacterial infections or gastrointestinal haemorrhage). However, in a significant proportion of cases (up to 30-40%), no apparent precipitating event was found.² The most prevalent OFs were kidney failure (55.8% of patients with ACLF) and liver failure (43.6%), followed by coagulation (27.7%) and cerebral (24.1%) failures. Cardiovascular and respiratory failures were less frequent (16.8 and 9.2% respectively).² At presentation, the prevalence of ACLF grade 1, 2, and 3 was 49%, 35%, and 16%, respectively. The 28-day transplant-free mortality was 32.8% in patients with ACLF compared to 1.9% in patients without ACLF.² Among patients with ACLF, the observed 28-day transplant-free mortality was 23%, 31% and 74% for grade 1, grade 2 and grade 3 ACLF, respectively.²

Among patients with ACLF, the phenotype differed according to the precipitating disorder. Compared to patients with ACLF unrelated to infection, those with infection-related ACLF more often had cerebral failure (31% vs. 17% in non-infected patients), circulatory failure (34% vs. 18%), and respiratory failure (20% vs. 10%). More importantly, patients with infection-related ACLF had more intense systemic inflammation and a higher 90-day mortality rate (51% vs. 38%).¹⁰

Several studies have investigated risk factors and predictors of ACLF development.[11], [12], [13] Data from a large retrospective cohort from the U.S. Veterans Health Administration reported the lowest incidence of ACLF among patients with underlying chronic HCV infection or non-alcoholic fatty liver disease. Conversely, patients with alcoholic cirrhosis or concomitant alcohol and HCV infection had the highest incidence rates.¹¹ In a large single-centre Italian prospective cohort of outpatients with cirrhosis, ascites, higher model for end-stage liver disease (MELD) score, lower mean arterial pressure and lower haemoglobin levels were identified as independent predictors for ACLF development.¹² Interestingly, grade III obesity (BMI >40 kg/m²) was identified as a risk factor for ACLF in a large retrospective population of patients with cirrhosis waitlisted for liver transplantation.¹³

Data obtained from the CANONIC cohort showed that ACLF is a very dynamic syndrome that may evolve to resolution, improvement or worsening in a short period of time.^2,14 Its clinical course after 3-7 days from diagnosis is a better predictor of outcome than its initial severity. A reassessment of ACLF grade and CLIF-C OF score within this timeframe reliably predicted patient prognosis, enabling the authors to stratify patients by severity and to monitor their response to treatment.¹⁴ Patients with grade 3 ACLF 3-7 days after diagnosis showed the worst prognosis. However, among these severely ill patients, the prognosis differed according to the number of OFs. Indeed, those with 3 OFs had lower 28-day transplant-free mortality than those with ≥4 OFs (53% vs. >90%, respectively). For patients with ≥4 OFs, discontinuation of intensive support should be considered when liver transplantation is contraindicated or not available.¹⁴ These findings may have implications for clinical decision making in patients with ACLF.

Of note, the EASL-CLIF definition of ACLF has been applied in different cohorts worldwide.^4,[15], [16], [17], [18] In a Chinese population of patients with cirrhosis associated with hepatitis B, patients with and without ACLF showed an average 28-day mortality rate of 44% and 2.6%, respectively. Moreover, the observed 28-day mortality for patients with grade 1, grade 2 and grade 3 ACLF was 23.6%, 40.8%, 60.2%, respectively.¹⁵ In a large retrospective U.S. cohort derived from 127 Veteran Affairs hospitals, patients without ACLF had a 28-day mortality rate of 10.4% compared to 25.5% in patients with ACLF (16.9%, 26.8% and 53.3% for grade 1, grade 2 and grade 3 ACLF, respectively).¹⁷ Similar results were also obtained when applying EASL-CLIF criteria to a prospective cohort in India.¹⁸

Another Chinese study retrospectively applied the EASL-CLIF criteria to a prospective cohort of patients with acutely decompensated cirrhosis, in whom 2 forms of ACLF were identified: one precipitated by hepatic insults (characterised by liver and coagulation failures) and the other by extrahepatic insults (characterised by extrahepatic OFs).¹⁹

ACLF phenotype according to the NACSELD definition

The epidemiology of liver diseases, the main causes of cirrhosis and potential precipitating events for acutely decompensated cirrhosis are similar between Europe and North America.^2,3,20 In the first study conducted in the context of NACSELD, the most prevalent organ system failures were brain (36%), followed by circulatory (16%), kidney (13%), and respiratory (9%).³ The observed 30-day mortality rate progressively increased from 4% in patients without any OF, to 27%, 49%, 64% and 77% in patients with 1, 2, 3, or 4 OFs, respectively.³ In a large series of infected and non-infected patients with acutely decompensated cirrhosis, the 30-day mortality rate was 41% among patients with NACSELD-ACLF compared to 7% among patients without NACSELD-ACLF.⁹

Of note, when compared with the EASL-CLIF definition of ACLF, the NACSELD definition considered only very severe and high-risk patients. The comparison of the 2 definitions in a large North American population showed that less than 40% of patients with EASL-CLIF ACLF were captured by NACSELD criteria.¹⁷ However, the 28- and 90-day mortality rates were still substantial among North American patients who did not fulfil the NACSELD criteria of ACLF.¹⁷

ACLF phenotype according to the COSSH definition

Seventy percent of patients with HBV-related ACLF, as defined by COSSH criteria, had cirrhosis. The most common precipitating disorder was, as expected in China, hepatic insult due to HBV reactivation alone (59% of cases), followed by a combination of a hepatic insult (most often HBV reactivation) and extrahepatic insult (bacterial infection) in 14% of cases. Of note, bacterial infection alone was a precipitating disorder in 9% of cases.⁴ Among patients with HBV-related ACLF, the most common failing organ systems were the liver (95%), coagulation (70%), kidney (13%), and brain (7%). There were very few cases of circulatory and respiratory failures. At presentation, the prevalence of ACLF grade 1, 2, and 3 was 60.6%, 33%, and 6.4%, respectively. The 28-day mortality rate for ACLF grade 1, grade 2, and 3 was 23%, 61%, and 93%, respectively.⁴ Together these findings indicate that although the European and Chinese definitions of ACLF were very similar, the clinical phenotypes of the syndrome differed between the 2 continents. For example, compared to European patients, Chinese patients more frequently had liver and coagulation failures and less frequently kidney and brain failures. Moreover, the mortality associated with ACLF, particularly the mortality associated with grade 2 ACLF, was greater in the Chinese cohort (Table 1).

ACLF phenotype according to the AARC definition

Studies using AARC criteria have shown that, as expected, HBV reactivation was the most frequent trigger of ACLF.^5,20 Other potential precipitating disorders were HEV infection and drug-induced liver injury.^5,20 Various bacterial, parasitic and fungal infections (directly and primarily affecting the liver) are also listed among triggers of ACLF in Asia.⁵ Using the AARC criteria, up to 95% of cases of ACLF present with an identifiable precipitating event,⁵ in contrast to only 60% of cases in Western countries. A study has investigated the course of patients with ACLF defined by AARC criteria.²¹ Almost 80% of patients had complications, including bacterial or fungal infection in 32% of patients, hepatorenal syndrome in 15%, and gastrointestinal haemorrhage in 9%. The 28-day transplantation-free mortality rate was 28%.

Prognostic scoring systems

Prognostic score systems are shown in Box 1 and Fig. 2A-D.^4,5,8,21,22

Box 1. Prognostic scoring systems used in patients with ACLF.

AARC, Asian Pacific Association for the Study of the Liver ACLF Research Consortium; ACLF, acute-on-chronic liver failure; CLIF-C, Chronic Liver Failure-Consortium; COSSH, Chinese Group on the Study of Severe Hepatitis B; HE, hepatic encephalopathy; ICU, intensive care unit; MELD, model for end-stage liver disease; MELD-Na, MELD-sodium; OF, organ failure; SOFA, sequential organ failure assessment.

Fig. 2.

Scores developed by different consortia to assess the prognosis of ACLF.

(A) Scores developed by European (CLIF-C ACLF score) and Chinese (COSSH-ACLF score) groups.^4,6 (B) The organ system assessment with the HBV-SOFA scale enables calculation of the HBV-SOFA score.⁶ (C) Components of the AARC scoring system.⁵ (D) Grading of ACLF according to AARC scores.⁵ AARC, Asian Pacific Association for the Study of the Liver ACLF Research Consortium; ACLF, acute-on-chronic liver failure; CLIF-C, Chronic Liver Failure-Consortium; COSSH, Chinese Group on the Study of Severe Hepatitis B; FiO₂, fraction of inspired oxygen; HE, hepatic encephalopathy; INR, international normalised ratio; MAP, mean arterial pressure; PaO₂, partial pressure of arterial oxygen; RRT, renal replacement therapy; SOFA, sequential organ failure assessment; SpO₂, oxygen saturation as measured by pulse oximetry.

Pathophysiology of ACLF

Inducers of systemic inflammation

Systemic inflammation can be induced by the presence in body fluids of pathogen-associated molecular patterns (PAMPs) and damage-associated molecular patterns (DAMPs) (Table 2; Fig. 3).[23], [24], [25], [26] PAMPs – expressed by microbes – are unique molecular structures that are recognised by dedicated receptors called pattern-recognition receptors (PRRs), an example being Toll-like receptors (TLRs), which are expressed in innate myeloid cells (i.e., monocytes and neutrophils) and other cells of the innate immune system.[23], [24], [25], [26] PRR engagement drives intracellular signalling cascades, ultimately leading to the transcription and synthesis of inflammatory mediators. A classical paradigm of these mechanisms is the engagement of TLR4 by lipopolysaccharide, a PAMP derived from the cell wall of gram-negative bacteria, resulting in the downstream transcription and activation of multiple inflammatory mediators and cytokines.^24,26 High levels of circulating PAMPs, which are unrelated to ongoing bacterial infections but mostly related to translocation of bacterial products from the intestinal lumen may contribute to cases of ACLF without any identified precipitating disorder.^1,27 These translocated PAMPs are the final result of intestinal bacterial overgrowth, increased permeability of the intestinal mucosa, and impaired function of the intestinal innate immune system.²⁸ Of note, bacterial virulence factors can induce inflammation, not through their direct recognition by PRRs but through functional effects they induce in cells; for example, pore-forming toxins induce a K⁺ efflux through the cell membrane that contributes to the activation of the NLRP3 (NLR family pyrin domain containing 3) inflammasome.²⁴

Table 2.

Exogenous and endogenous inducers of inflammation and their receptors.

Inducers of inflammation	Receptor(s)
Exogenous: pathogen-associated molecular patterns
Triacyl lipoprotein	Toll-like receptor (TLR)1
Lipoprotein	TLR2
Double-stranded RNA	TLR3
Lipopolysaccharide	TLR4, caspases 4/5
Flagellin	TLR5, NLRC4 (IPAF), NAIP5 and NAIP6
Diacyl lipoprotein	TLR6
Single-stranded RNA	TLR7, TLR8
Unmethylated DNA with CpG motifs	TLR9
Unknown	TLR10
Profilin-like molecule	TLR11∗
Profilin	TLR12∗
23S ribosomal RNA	TLR13∗
γ-D-glutamyl-mesodiaminopimelic acid (iE-DAP)	Nucleotide-binding oligomerisation domain (NOD) 1
Muramyl dipeptide (MDP)	NOD2, hexokinase
RNA (vita-PAMP)	NLRP3
Short double-stranded RNA, 5'triphosphate double-stranded RNA	Retinoic acid-inducible gene (RIG)-I
Long double-stranded RNA	Melanoma differentiation-associated protein 5 (MDA5)
β-Glucan	Dectin-1, dectin-2 (also known as C-type lectin domain containing 6A, CLEC6A)
Double-stranded DNA	Absent in melanoma (AIM)-2, interferon gamma-inducible protein 16 (IFI16), Z-DNA binding protein 1 (ZBP1), cyclic GMP-AMP synthase (cGAS)
Double-stranded DNA, single-stranded DNA	High mobility group box (HMGB) proteins, LRRFIP1, Leucine rich repeat (in FLII) interacting protein 1 (LRRFIP1)
Cyclic diadenosine monophosphate (c-diAMP; vita-PAMP∗∗)	Stimulator of interferon genes protein (STING)
Endogenous inducers: damage-associated molecular patterns (DAMPs)
Released by dying cells
Double-stranded DNA	TLR9, AIM2
High mobility group box 1 (HMBG1)	TLR2, TLR4, TLR9, advanced glycation end products receptor (AGER; alias, RAGE), CD24
Histones	TLR2, TLR4
Sin3A-associated protein (SAP) 130	C-type lectin domain family 4 member E (CLEC4E; alias MINCLE)
Mitochondrial DNA	TLR9, NLRP3
Mitochondrial N-formyl peptides	Formyl peptide receptor (FRP)-1
Cytochrome c	Unknown
ATP	Purinoceptors
N-myc and STAT interactor (NMI)	TLR4
Interferon-induced protein 35 (IFP35)	TLR4
S100 calcium-binding proteins
S100A8, S100A9	TLR4
S100A12	AGER
K+ ions	K+ channels
Cold-inducible RNA binding protein (CIRBP)	TLR4-MD2 complex
Peroxiredoxins	TLR2, TLR4
Heat shock proteins (HSP60, HSP70, HSP90, GP96)	TLR2, TLR4, CD14, CD40, CD24
Calreticulin	LDL Receptor Related Protein 1 (LRP1, alias CD91), C1Q
Defensins	CCR6, TLR4
Galectins	CD2
Interleukin (IL)-1α	IL-1R1 (specific receptor), IL-1R3 (co-receptor)
IL-33	IL-1R4 (specific receptor), IL-1R3 (co-receptor)
Extracellular DAMPs
Short-fragment hyaluronan	CD44-TLR4-Lymphocyte antigen 96 (MD2) complex, TLR2
Biglycan	TLR2, TLR4
Versican	TLR4
Heparan sulfate	TLR4
Extracellular matrix fragments from collagen, elastin, laminin	CD14, TLR4, Serum hyaluronan-associated protein (SHAP)

NOTE: TLRs form homodimers except TLR2 that heterodimerises with TLR1 or TLR6. NLRC4, NLR family, CARD domain containing 4; IPAF, ice protease-activating factor; NAIP, nucleotide-binding oligomerisation domain (NOD)-like receptor (NLR) family, apoptosis inhibitory protein; NLRP3, NLR family, pyrin domain containing 3.

^∗Expressed in mice but not humans.

^∗∗A vita-PAMP is a PAMP which indicates the presence of a living microbe.

Fig. 3.

Pathophysiology of ACLF.

Schematic of induction of systemic inflammation and its role in the development of organ failures. ACLF, acute-on-chronic liver failure; AKI, acute kidney injury; DAMPs, damage-associated molecular patterns; PAMPs, pathogen-associated molecular patterns; PRRs, pattern-recognition receptors; iNOS, inducible nitric oxide synthase; OxPhos, oxidative phosphorylation; RAAS, renin-angiotensin-aldosterone system; SNS, sympathetic nervous system.

Systemic inflammation can also occur in the absence of infection.²⁵ This sterile inflammation is due to the release of circulating DAMPs by dying or damaged host cells that bind to and activate specific PRRs (Table 2).²⁵ DAMPs consist of intracellular components from various compartments.²⁹ Several forms of liver injury are well-known causes of DAMP release. In alcoholic hepatitis, alcohol-induced hepatocyte apoptosis has been shown to be triggered by endoplasmic reticulum (ER) stress involving the ER-resident adaptor STING, a cytosolic PRR for cytosolic DNA.³⁰ Ischemia-reperfusion liver injury is characterised by the release of high mobility group box 1 (HMGB1) from hepatocytes exposed to hypoxia and oxidative stress.³¹ Moreover, HMGB1 can induce cytokine production and promote chemotaxis by binding to several receptors (Table 2). Also, the submassive hepatic necrosis that characterises patients with HBV-associated ACLF may give rise to the release of DAMPs and high levels of inflammatory cytokines.³²

Outcomes of systemic inflammation

Tissue hypoperfusion

PAMPs and inflammatory mediators can induce inducible nitric oxide (NO) synthase in splanchnic arteriolar walls. The resulting NO overproduction causes splanchnic vasodilation which decreases effective arterial blood volume, triggering homeostatic overactivation of the endogenous neurohumoral vasoconstrictor system (Fig. 3). Neurohumoral mediators then cause intense vasoconstriction, particularly in the renal circulation, resulting in kidney hypoperfusion, decreased glomerular filtration rate and acute kidney injury (AKI).¹

Immune-mediated tissue damage

Like sepsis in the general population, ACLF is commonly associated with blood leukocytosis, comprising activated immune cells that may migrate into tissues and cause immunopathology¹ (Fig. 3). There is some evidence for this hypothesis in the context of cirrhosis. For example, tumour necrosis factor-α and NF-κB-dependent signalling pathways may play a role in impaired left ventricular contractility,³³ in NO-mediated pulmonary dysfunction and macrophage accumulation in lung microvasculature,³⁴ and in hepatocyte apoptosis.³⁵ Like sepsis-induced AKI, ACLF-associated AKI may not only involve tissue hypoperfusion (see above) but also capillary leukocyte infiltration, vascular microthrombosis, and cell apoptosis.³⁶ Moreover, the direct inflammatory damage to tissues and cells leads to the release of a huge amount of circulating cellular products, which act as DAMPs on immune cell receptors. Therefore, a vicious cycle sustains and exacerbates inflammatory responses, providing the mechanistic link between systemic inflammation, cell injury and organ failure (Fig. 3).²³

Mitochondrial dysfunction

High-throughput blood metabolomics performed in a large cohort of patients with acutely decompensated cirrhosis (CANONIC cohort), revealed that, in ACLF, peripheral organs may have a marked decrease in mitochondrial fatty acid β-oxidation in peripheral organs, resulting in decreased oxidative phosphorylation and ATP production.³⁷ These findings suggest that defective energy production may play a role in the development of OFs in ACLF (Fig. 3).

Immunosuppression in ACLF

Investigators in Europe,¹⁰ North America^3,38 and Asia^21,39,40 all agree that secondary infections are common complications of ACLF. Moreover, among patients with ACLF, 90-day mortality was higher in those who develop secondary infection than in those who remain free of this complication during the entire period of follow-up,^10,41 indicating the extreme severity of secondary infection in this context. More importantly, the high risk of secondary infections indicates that patients with ACLF are immunosuppressed. Among patients with ACLF who are free of infections at presentation, higher blood levels of immunosuppressive molecules, including interleukin-10 and the tryptophan metabolite quinolinate are predictors of secondary infections.⁴²

There are findings suggesting that, in ACLF, some subsets of immune cells have defective antimicrobial functions that contribute to the high risk of secondary infection. Defective responses to PAMPs have been shown in macrophages derived from circulating monocytes obtained from patients with ACLF. Moreover, patients with ACLF have a higher frequency of CD14⁺ monocytes expressing the receptor tyrosine kinase MerTK and CD14⁺CD15^-HLA-DR^-myeloid-derived suppressor cells,^43,44 with both monocyte subsets suppressing innate responses to bacterial PAMPs. Another study found decreased frequencies of other myeloid mononuclear cells (conventional and plasmacytoid dendritic cells) in patients with severe alcoholic hepatitis, including patients who had ACLF.⁴⁵ Finally, studies have shown that neutrophils in patients with decompensated cirrhosis have a marked defect in both the production of antimicrobial superoxide anion and bactericidal activity.⁴⁶

Collectively, these “humoral” and immune-cell alterations may favour the development of serious infections that are frequent complications of ACLF. Of note, although the lymphoid lineage plays a major role in host resistance to infection, little is known about lymphocyte frequency and function in acutely decompensated cirrhosis. Future studies should investigate lymphocytes in patients with acutely decompensated cirrhosis with and without ACLF.

Management of ACLF

Principles of treatment of ACLF are provided in Fig. 4. Several recommendations are based on results of studies conducted in critically ill patients without cirrhosis.¹

Fig. 4.

Principles of treatment of organ failure in ACLF.

What should be done is shown in green boxes. What should be avoided is shown in red boxes. ACLF, acute-on-chronic liver failure; AKI, acute kidney injury; KDIGO, Kidney Disease Improving Global Outcomes; LT, liver transplantation; NSAID, non-steroidal anti-inflammatory drugs; PPI, proton pump inhibitors; RRT, renal replacement therapy.

Admission of patients with ACLF to the intensive care unit

The admission of patients with cirrhosis to the intensive care unit (ICU) should no longer be denied solely because of the existence of the underlying chronic liver disease. The proportion of patients with ACLF admitted to the ICU is therefore increasing and several studies have shown that the ICU prognosis of cirrhotic patients has improved in recent years (period effect independently associated with mortality).^47,48 Nevertheless, ICU mortality rates remain high in some patient groups and some factors can be used to guide the admission of these patients to ICUs. As for patients without cirrhosis, the prognosis of critically ill patients with cirrhosis admitted to the ICU largely relies on the presence of OF(s), graded using different scores.^22,49 CLIF-sequential organ failure assessment and CLIF-C ACLF scores perform better than general ICU scores, such as Acute Physiology And Chronic Health Evaluation (APACHE) II or Simplified Acute Physiology Score (SAPS) II scores, and liver-specific scores, such as MELD or Child-Pugh.^22,50 In addition, the reason for ICU admission should be considered; the prognosis of gastrointestinal haemorrhage being better than the prognosis of septic shock. Finally, some evidence suggests a better prognosis in case of early admission to the ICU.

Treating acute precipitants

Antimicrobial therapy

In a recent study, about 37% of patients with ACLF presented with a bacterial infection at diagnosis. Furthermore, 46% of the remaining patients with ACLF developed bacterial infections within the next 4 weeks.¹⁰ Multidrug-resistant (MDR) pathogens are involved in one-third of cases with differences in prevalence according to region.^10,51 A systematic search for infection, including microbiological and cytological examination of ascitic fluid, should therefore be systematically performed at admission. An empirical antibiotic therapy tailored to the suspected site of infection and the local ecology should be rapidly initiated.¹ Broad-spectrum molecules should be preferred in case of severe infection or in the presence of risk factors for MDR pathogens.

Corticosteroids for alcoholic hepatitis

Corticosteroids remain the first-line treatment for severe alcoholic hepatitis. The response to corticosteroids can be assessed by calculating the Lille score after 7 days of treatment.^1,52,53 The prognosis of patients who do not respond is poor. The probability of response to corticosteroids depends on the presence or absence of ACLF at presentation. Indeed, the probability of response is lower in patients with ACLF compared to those without ACLF (38% and 77%, respectively).⁵⁴ Moreover, the probability of response to corticosteroids decreases with ACLF grade, being 52%, 42% and 8% for grade 1, grade 2 and grade 3, respectively.⁵⁴ Nevertheless, among responders, a beneficial effect of corticosteroids on patient survival has been shown.^54,55 Given the risk of bacterial infection, the risk to benefit ratio should be carefully evaluated before initiating corticosteroids in patients with ACLF and severe alcoholic hepatitis.

Acute variceal haemorrhage

The use of transjugular intrahepatic portosystemic shunting is summarised in Box 2.⁵⁶

Box 2. Use of albumin, use of TIPS for acute variceal haemorrhage, use of RRT, and prioritisation for LT.

ACLF, acute-on-chronic liver failure; AKI, acute kidney injury; CLIF-C, Chronic Liver Failure-Consortium; HRS, hepatorenal syndrome; KDIGO, Kidney Disease Improving Global Outcomes; LT, liver transplantation; MELD, model for end-stage liver disease; MELD-Na, MELD-sodium; RRT, renal replacement therapy; SBP, spontaneous bacterial peritonitis; TIPS, transjugular intrahepatic portosystemic shunt.

Organ support

Intravenous fluids

Fluid therapy should use crystalloids, while balanced salt solutions may limit the risk of hyperchloremic acidosis and subsequent adverse kidney events. Nevertheless, beneficial effects of albumin resuscitation have been demonstrated in patients with cirrhosis and may be related to more than mere volume expansion. Indeed, besides the overall decrease of albumin, the function of albumin is also impaired with alterations in its chemical structure, resulting in reduced binding capacity to bacterial products, reactive oxygen species, and other mediators involved in ACLF.⁵⁷ Some studies suggested that albumin may modulate systemic oxidative stress and inflammation^58,59 or restore immune defense.⁶⁰

The intravenous use of human albumin is discussed in Box 2.^52,61,62

Renal replacement therapy

The use of renal replacement therapy is discussed in Box 2.[63], [64], [65]

Extracorporeal liver support

To date, the best-known devices are based on the principle of albumin dialysis. Two multicentre randomised European trials in patients with acutely decompensated cirrhosis compared these systems with standard medical treatment. These studies showed an improvement of biological cholestasis and hepatic encephalopathy in patients treated with albumin dialysis but did not demonstrate any benefit on 28- and 90-day survival.^66,67 More recently, the use of an artificial liver support system was associated with improved short-term survival (14- and 28-day) in patients with ACLF and multiple OFs in a retrospective study⁶⁸ and a meta-analysis.⁶⁹ Therefore, these devices may be interesting as a bridge to liver transplantation or recovery. Finally, evidence points to the possible use of plasma exchange to remove endotoxins and inflammatory mediators and replace albumin. A randomised clinical trial (APACHE-trial, NCT03702920) is currently ongoing. Regenerative medicine using stem cell technology such as heterologous human adult liver-derived progenitor cells (HepaStem) is another strategy currently being developed, although safety issues remain to be solved.⁷⁰

Liver transplantation

Liver transplantation for critically ill patients with cirrhosis and extrahepatic OFs is becoming more and more frequent.^1,71 While post-transplant survival rates of patients with grade 1 and 2 ACLF seem similar to those of patients without ACLF, 1-year survival of patients with grade 3 ACLF greatly differs between studies and ranges from 44 to 83%.^72,73 Given the poor short-term prognosis of patients with ACLF without liver transplantation, these data strongly support the use of liver transplantation as a therapeutic strategy for patients with ACLF.⁷⁴ However, as discussed in Box 2, prioritisation of these patients for liver transplantation remains problematic.^71,[75], [76], [77], [78]

Conclusions

ACLF is a major medical problem worldwide and its occurrence is a challenging clinical event for hepatologists and intensivists, due to its acute presentation, rapid clinical course and associated high short-term mortality. Geographical differences exist regarding the definition of ACLF and its diagnostic criteria, resulting in varying clinical phenotypes. The therapeutic management of patients with ACLF is currently based on the treatment and support of different OFs. A better understanding of ACLF pathophysiology may help in the development of mechanistic treatments, either curative or preventive.

Financial support

The authors report no specific grants for this research.

Authors' contributions

Drafting of the manuscript (GZ, EW); critical revision of the manuscript for important intellectual content (RM).

Conflicts of interest

E. Weiss reports personal fees form Baxter, MSD France, Biomerieux and Akcea therapeutics, and travel reimbursements from MSD France, outside the present work. The other authors report no conflict of interest related to this work.

Please refer to the accompanying ICMJE disclosure forms for further details.