Acute‐on‐Chronic Liver Failure: A Distinct Clinical Syndrome That Has Reclassified Cirrhosis

http://aasldpubs.onlinelibrary.wiley.com/hub/journal/10.1002/(ISSN)2046-2484/video/14-5-reading-perricone a video presentation of this article

http://aasldpubs.onlinelibrary.wiley.com/hub/journal/10.1002/(ISSN)2046-2484/video/14-5-interview-jalan the interview with the author

Abbreviations

ACLF

acute‐on‐chronic liver failure

ACLF‐1

ACLF grade 1

ACLF‐2

ACLF grade 2

ACLF‐3

ACLF grade 3

AD

acute decompensation

d3‐7 ACLF

ACLF grade at days 3 to 7 after diagnosis

EASL‐CLIF

European Association for the Study of the Liver–Chronic Liver Failure

FiO₂

fraction of inspired oxygen

HE

hepatic encephalopathy

INR

international normalized ratio

LT

liver transplantation

MAP

mean arterial pressure

MELD

Model for End‐Stage Liver Disease

OF

organ failure

PaO₂

partial pressure of arterial oxygen

SpO₂

pulse oximetric saturation

UNOS

United Network for Organ Sharing

Conventionally, the severity of cirrhosis is classified into compensated, decompensated, and late decompensation, which defines prognosis. Acute‐on‐chronic liver failure (ACLF) adds substantially to this classification by identifying a subgroup of patients with cirrhosis who may progress rapidly after acute decompensation (AD) to development of organ failures (OFs) and high short‐term mortality.

Definition of ACLF

More than 13 distinct definitions of ACLF, largely based on personal experience or consensus agreements, have been proposed,1 but the only one that was specifically developed to define the diagnostic criteria for ACLF was by the European Association for the Study of the Liver–Chronic Liver Failure (EASL‐CLIF) Consortium. In 2009, the consortium started a prospective, multicenter, European observational study that included 1343 patients hospitalized for AD of cirrhosis (the CANONIC study).2 This study aimed to define ACLF in cirrhosis, to propose diagnostic criteria, to assess the prevalence and clinical course of the syndrome, and to develop new prognostic scores. The current review is largely based on this investigation. According to the EASL‐CLIF Consortium definition, ACLF is a specific syndrome characterized by AD of cirrhosis, OF(s), and high short‐term mortality. AD means development of ascites, hepatic encephalopathy (HE), gastrointestinal hemorrhage, and/or bacterial infections; ACLF may develop in patients with or without a prior history of AD. OFs (liver, kidney, brain, coagulation, respiration, circulation) are defined by the original CLIF‐SOFA score (the Sequential Organ Failure Assessment Scale adapted for liver patients) or its simplified version, CLIF Organ Failure score3 (Table 1). High short‐term mortality means a 28‐day mortality rate ≥15%.

Table 1.

The CLIF‐Organ Failure Score System

Organ/System	Subscore = 1	Subscore = 2	Subscore = 3
Liver	Bilirubin <6 mg/dL	Bilirubin ≥6 and <12 mg/dL	Bilirubin ≥12 mg/dL*
Kidney	Creatinine <2 mg/dL	Creatinine ≥2 and <3.5 mg/dL*	Creatinine ≥3.5 mg/dL or renal replacement*
Brain (West‐Haven grade for HE)	Grade 0	Grades 1‐2	Grades 3‐4* , †
Coagulation	INR <2	INR ≥2.0 and <2.5	INR ≥2.5*
Circulatory	MAP ≥70 mm Hg	MAP <70 mm Hg	Use of vasopressors*
Respiratory
PaO2/FiO2	>300	>200 and ≤300	≤200* , ‡
SpO2/FiO2	>357	>214 and ≤357	≤214* , ‡

Adapted with permission from Journal of Hepatology.3 Copyright 2014, Elsevier.

^*Criteria for diagnosing OFs.

^†Patients submitted to mechanical ventilation because of HE and not because of respiratory failure were considered as presenting a cerebral failure (cerebral subscore = 3).

^‡Other patients enrolled in the study with mechanical ventilation were considered as presenting a respiratory failure (respiratory subscore = 3).

Diagnostic Criteria and ACLF Grades

Mortality rate of the patients in the CANONIC study was clearly related to the presence and number of OFs. Also, renal dysfunction (as defined by a serum creatinine concentration of 1.5–1.9 mg/dL) and/or cerebral dysfunction (grade 1–2 HE), when associated with a single OF, were found to predict prognosis. Based on the presence of, renal and/or cerebral dysfunction, and short‐term mortality rate, the following groups of patients were proposed to have ACLF or no ACLF:

1. No ACLF: no OF or a single nonrenal OF without renal dysfunction and cerebral dysfunction

2. ACLF grade 1 (ACLF‐1): single renal failure and single nonrenal OF that is associated with renal dysfunction and/or cerebral dysfunction

3. ACLF grade 2 (ACLF‐2): two OFs of any combination

4. ACLF grade 3 (ACLF‐3): three or more OFs of any combination

Among the different OFs in ACLF, the most frequently affected organs were the kidneys (56% of patients), followed by the liver (44%), coagulation (28%), brain (24%), circulation (17%), and lungs (9%). Kidney failure is the most prevalent OF in ACLF‐1. For ACLF‐2, liver failure is the most prevalent OF followed by kidney, brain, and coagulation failure. For ACLF‐3, the prevalence of all OFs is high.

Epidemiology, Health Burden, and Mortality

ACLF is a major worldwide medical problem, with prevalence rates in at‐risk populations in the region of 20% to 35%. The worldwide reported mortality rate according to the EASL‐CLIF Consortium definition ranges between 30% and 50% and correlates closely with the number of OFs. In Europe, most of the prevalence and natural history data come from the CANONIC study.2 Approximately 23% of patients admitted to the hospital for an AD of the disease had ACLF at admission. Furthermore, 11% of the patients without ACLF at enrolment experienced development of the syndrome during hospitalization, which gives a total prevalence rate of ACLF in patients admitted to the hospital with AD of 31%. Among patients with ACLF, 51% had ACLF‐1, 35% ACLF‐2, and 13% ACLF‐3. The average 28‐day and 90‐day mortality rates without liver transplantation (LT) were 1.9% and 10% in patients with AD without ACLF and 33% and 51% in patients with ACLF (Table 2). The health care burden of ACLF and cirrhosis is associated with extremely high costs, exceeding the yearly costs of inpatient management of more common medical conditions.

Table 2.

ACLF Grades and Mortality Without LT

Category	28‐Day Mortality Without LT (%)	90‐Day Mortality Without LT (%)
No ACLF	1.9	10
ACLF (total)	33	51
ACLF‐1	23	41
ACLF‐2	31	55
ACLF‐3	74	78

Adapted with permission from Gastroenterology.2 Copyright 2013, Elsevier.

Precipitating Illness

In most cases, the development of ACLF is associated with a precipitating factor. The most common precipitating events are bacterial infections, active alcoholism, and reactivation of HBV, particularly in patients with underlying hepatitis B virus infection in the East. However, in up to 40% of patients, no precipitating factor can be identified. The potential role of drug‐induced liver injury as a precipitating event in ACLF has been insufficiently explored in both the East and the West. Data from the CANONIC study2 showed that mortality was independent of the type of precipitating factor, and that it was mainly related to other factors such as the type and number of OFs, the intensity of inflammatory response, and the early clinical course of the syndrome. Nearly 40% of patients with ACLF had a bacterial infection as a precipitating event. Nosocomial infections may be associated with a higher risk for ACLF compared with that of community‐acquired infections. Bacterial infections tend to cause ACLF more frequently in patients without a previous history of decompensation compared with patients with previous decompensation. Severity of infection also increases the risk for ACLF. A recent study evaluated the prevalence and characteristics of bacterial and fungal infections causing and complicating ACLF, the predictors of new bacterial infections, and the impact of bacterial infections on survival.4 The main findings were: (1) patients with ACLF are at high risk for acquiring new bacterial infections; (2) severe infections (spontaneous bacterial peritonitis, pneumonia, severe sepsis/shock, nosocomial infections, and infections caused by multiresistant organisms) are more prevalent in patients with ACLF; (3) bacterial infections, either at diagnosis or during follow‐up, are key prognostic determinants; (4) bacterial infections are independent predictors of 90‐day mortality in patients with ACLF‐1 and ACLF‐2; and (5) inappropriate empirical antibiotic strategies increase 90‐day mortality.

Mechanisms of ACLF

The specific pathophysiological features of ACLF are systemic and hepatic inflammation.2, 5 It is not clear whether systemic inflammation, manifested by elevated white cell count and C‐reactive protein, represents an alteration of host response to injury or whether it is due to an inability to resolve inflammation. Another feature is the increase in the circulating cytokines; the changes in the pattern of cytokines are not consistent and depend on the severity of ACLF, the underlying cause of the liver disease and the precipitating event. These changes in circulating markers of inflammation are associated with changes in the functional characteristics of the circulating inflammatory cells. Clària and colleagues5 demonstrated that: (1) patients with ACLF have significantly higher levels of inflammatory cytokines, human nonmercaptalbumin‐2, and plasma renin concentrations than those without ACLF; (2) different cytokine profiles were identified according to the type of ACLF precipitating event; and (3) and there was a good correlation between the course of systemic inflammation and the clinical course of ACLF. It has also become clear that molecules released after cell death (damage‐associated molecular patterns) have immunogenic properties and can result in systemic inflammation. Recently, it has been shown that the predominant mechanism of cell death in ACLF is nonapoptotic,6 which may provide an explanation for the severity of systemic inflammation observed.

Clinical Course

ACLF is an extraordinarily dynamic syndrome that has potential for reversibility.7 Overall, ACLF resolves or improves in 49.5% of patients, followed by a steady or fluctuating course with unchanged final ACLF grade in 30.4% and worsened in 20.1%. Frequency of ACLF resolution was high in patients with initial ACLF‐1 and low in those with initial ACLF‐3. In contrast, the proportion of patients with final ACLF‐3 was low in patients with initial ACLF‐1 and very high in those with ACLF‐3.

Overall, the 28‐day transplant‐free mortality rate was low in patients with ACLF resolution (5.8%), moderate in those with final ACLF‐1 (18.2%), high in those with final ACLF‐2 (41.7%), and very high in those with final ACLF‐3 (91.8%), independently of whether they presented with ACLF‐1, ‐2, or ‐3 at diagnosis. The final ACLF grade was already defined at days 3 to 7 in 81% of patients. ACLF grade at days 3 to 7 after diagnosis predicted significantly better 28‐ and 90‐day mortality rates than ACLF grade at diagnosis. The probability of 28‐day transplant‐free survival was high for patients with no ACLF at days 3 to 7 and ACLF‐1 at days 3 to 7 (89.6% and 78.7%, respectively) and low to very low for patients with ACLF‐2 and ‐3 at 3 to 7 days (42.9% and 12.8%, respectively). These differences were maintained at 90 and 180 days.

Prognostic Score

To allow ongoing stratification of patients for intensive care, fast‐track listing for LT, early hospital discharge, or determination of futility of further intensive care, the CANONIC investigators developed and validated two prognostic scores: the CLIF‐C ACLF score is for patients with ACLF,3 and the CLIF‐C AD score is for patients with AD who did not fulfill criteria for the diagnosis of ACLF.8 These two scores were designed because a single score was insufficient to satisfactorily delineate the prognosis associated with AD and ACLF. The CLIF‐C ACLF and AD scores provided a significantly better estimate of the risk for death compared with the Model for End‐Stage Liver Disease (MELD) score, the MELD‐Sodium score, and the Child‐Pugh score. Organ allocation for LT using the MELD score seriously disadvantages the patient with ACLF. The performance of the CLIF‐C ACLF score improved over the follow‐up period, suggesting that it should be updated daily.3

Management, Futility of Intensive Care Support, and Role of LT

Currently, the accepted strategy for management of ACLF consists of early recognition and treatment of the precipitating event, and supportive care with intensive monitoring and support of failing organs. There is currently no evidence to justify alternative strategies for the management of OFs in patients with cirrhosis compared with other critically ill patients.9, 10 In case of contraindication of LT, the presence of ≥4 OFs or a CLIF‐C ACLF score >70 at days 3 to 7 after diagnosis could indicate the futility of care.7

LT represents the only definitive therapeutic option for patients with ACLF. There seems to be a clear agreement in the literature regarding outcomes of urgent LT in patients with ACLF, with acceptable to excellent 1‐ and 5‐year post‐LT survival reported in some studies. Some patients, particularly those with respiratory failure, have less successful results. Because patients with severe ACLF have a high mortality rate while on the waiting list, salvage LT is feasible and associated with a clear survival benefit in selected patients with ACLF‐311, 12, 13, 14 (Table 3). It is possible that many patients with ACLF are not listed for LT on the assumption that they are too ill to survive LT. There could also be a significant delay in listing for logistical reasons or because of indecisiveness about the utility or futility of LT in such a situation. In addition, there also may be center‐specific differences in listing for LT in the presence of multiple OFs. These factors introduce a dimension of selection bias in the studies published to date. Because a large proportion of patients with ACLF die while on the waiting list, a better rule for organ allocation is needed for this group. The specific scores for ACLF are more accurate for prediction of short‐term outcomes than the MELD score. The implementation of these scores could decrease the mortality rate of patients while on the waiting list, but they need further evaluation and validation. The limits defining when a patient should be considered too sick for transplantation and LT should be considered futile are currently largely unknown.

Table 3.

ACLF and LT

Study	Experience	Criteria for ACLF Diagnosis	No. of LTs	Survival Post‐LT	Notes
Gustot et al. (2015)7	CANONIC	CLIF‐C criteria	35 patients with initial ACLF: 25 patients with ACLF at LT, 10 patients with ACLF resolution at LT	1‐year: 75.3% (ACLF‐1, 80%; ACLF‐2, 71.6%; ACLF‐3, 77.8%) versus 90% for 10 patients with ACLF resolution before LT	LT within 28 days (median time between ACLF diagnosis and LT: 11 days)
					6‐Month probability of survival of d3‐7 ACLF‐2 or ‐3 patients undergoing LT compared with LT‐free survival probability in d3‐7 ACLF‐2 or ‐3 patients: 80.9% versus 10%
Levesque et al. (2017)11	France, one center	CLIF‐C criteria	140 patients with ACLF at LT	1‐year: 70% (ACLF‐1 or ‐2, 77.2%; ACLF‐3, 43.3%)	1‐Year survival post‐LT in patients without ACLF: 91.4%
Artru et al. (2017)12	France, three centers	CLIF‐C criteria	73 patients with ACLF‐3	1‐year: 83.6%	1‐Year survival of 119 non‐LT controls: 7.9%
					100% patients with ACLF‐3 experienced complications
Thuluvath et al. (2018)13	UNOS	CLIF‐C criteria	3556 patients with ≥3 OFs at LT; 677 patients with 5‐6 OFs at LT	1‐year: 3 OFs, 84%; 4 OFs, 81%; 5‐6 OFs, 81%	LT median time 4‐5 days
					Only 2% of patients with 5‐6 OFs remained on the list at 30 days
Sundaram et al. (2018)14	UNOS	CLIF‐C criteria	6680 patients with ACLF‐1; 6996 with ACLF‐2; 6010 with ACLF‐3	1‐Year: 81.1% in ACLF‐3 versus 88.4%‐91.7% in the other groups	1‐Year survival without LT for ACLF‐3: 23.5%

Conclusions

The accumulated data in more than 1000 articles after its initial description15 have confirmed that ACLF is clinically, prognostically, and pathophysiologically distinct from mere AD. Better clinical characterization and understanding of the pathophysiology of the syndrome has reclassified cirrhosis and proposes a new framework to develop new therapies for this syndrome, which has an unacceptably high risk for death.