Until 2013, there was no shared, established, evidence-based definition of ACLF, and the only published definitions were based on expert opinion. Moreover, the used definitions of ACLF differed between Eastern and Western countries. In Asia, the following liver-centered definition has been suggested: an acute hepatic insult manifesting as jaundice (serum bilirubin level > 5 mg/dL) and coagulopathy (international normalized ratio [INR] >1.5) complicated within 4 weeks by ascites and/or encephalopathy in a patient with previously diagnosed or undiagnosed chronic liver disease [2]. In Europe and the USA, a different definition of ACLF was used identifying ACLF as an acute deterioration of liver function in patients with cirrhosis which usually is associated with a precipitating event and results in the failure of one or more organs and high short-term mortality rates [3]. Finally, the sequential organ failure assessment (SOFA) score was also used to diagnose organ failures in patients with cirrhosis admitted to the intensive care unit [4].

Since diagnostic criteria of ACLF were based in both definitions on personal expert opinions rather than on objective data, in 2009, a group of European investigators decided to create the Chronic Liver Failure (CLIF) Consortium with the objective of stimulating research on complications of cirrhosis. The Consortium was endorsed by the European Association for the Study of the Liver (EASL) resulting in the EASL-CLIF Consortium. One of the first decisions by the Steering Committee of the Consortium was to perform a multicenter, prospective, observational study in patients with an acute decompensation of cirrhosis. This study was named CLIF Acute-on-Chronic Liver Failure in Cirrhosis (CANONIC) study and aimed at assessing the prevalence, diagnostic criteria, precipitating events, natural course, and prognosis of ACLF. The CANONIC study prospectively enrolled 1343 patients with cirrhosis hospitalized in 29 liver units from 8 European countries between February and September 2011. Enrolled patients were hospitalized for at least 1 day and had an acute development of large ascites, hepatic encephalopathy, gastrointestinal hemorrhage, bacterial infections, or any combination of these [5]. For the diagnosis of organ failures, investigators used a modified SOFA scale, called the CLIF-SOFA scale, which had been designed specifically by the Writing Committee of the CANONIC study before the onset of this study. The CLIF-SOFA scale assesses the function of six organ systems (liver, kidneys, brain, coagulation, circulation, and lungs) but also takes into consideration some specificities of cirrhosis. Each organ system receives a subscore ranging from zero (normal) to four (most abnormal). A total CLIF-SOFA score ranging from 0 to 24 can thus be calculated. Notably, all variables included in the CLIF-SOFA are easy to obtain in every hospital.

In the CANONIC study, three types of risk factors obtained from the CLIF-SOFA score at enrollment were found to be related to high 28-day mortality rate: (1) the presence of two organ failures or more, (2) the presence of one organ failure when the organ that failed was the kidney, and (3) the coexistence of a single “non-kidney” organ failure with kidney dysfunction (i.e., serum creatinine level ranging from 1.5 to 1.9 mg/dL) and/or mild to moderate hepatic encephalopathy [5].

Based on the findings from the CANONIC study, four stages of ACLF can be nowadays recognized:

These results show that ACLF is a new clinical entity that is distinct from decompensated cirrhosis.

The prevalence of ACLF in the CANONIC Study was 30% (20% at admission and 10% during hospitalization), and the overall 28-day and 90-day mortality rates were 33% and 51%, respectively. Mortality rates in patients without ACLF were low (28-day, 1.9%; 90-day, 10%). The prevalence and 28-day and 90-day mortality rates associated with the different grades of ACLF were 15.8%, 22%, and 41%, respectively, in ACLF-1; 10.9%, 32%, and 55% in ACLF-2; and 4.4%, 73%, and 78% in ACLF-3 [5].

Patients with ACLF were significantly younger than those without ACLF, and the main etiologies were alcoholism (60%), hepatitis C (13%), and alcoholism plus hepatitis C (10%). In only 5% of patients, the main etiology was cirrhosis associated with hepatitis B virus (HBV) infection. The commonest organ failure in patients with ACLF was renal failure (56%) with liver coagulation and cerebral, circulatory, and respiratory failures (44%, 28%, 24%, 17%, and 9%, respectively) also very frequent. The prevalence of circulatory and respiratory failure was significant only in patients with ACLF-3. Most importantly, patients with ACLF showed systemic inflammation (high count of C-reactive protein and leukocyte concentration) which was independent on the presence or absence of recognized bacterial infections. Patients with no history of decompensated cirrhosis developed a more severe form of ACLF than patients with previous episodes of decompensation (28-day mortality of 42% vs. 29%).

The most common precipitating events were bacterial infections and active alcoholism. In patients with ACLF the prevalence of alcoholic cirrhosis (60%) was higher than the prevalence of active alcoholism, indicating that alcoholic hepatitis accounts for only part of cases of ACLF associated with alcoholic cirrhosis. There was a small proportion of other precipitating events. As a trigger, gastrointestinal hemorrhage was less frequent in patients with ACLF than in patients without ACLF, suggesting that hemorrhage, if not associated to other complications (i.e., active drinking and/or bacterial infections), is not related to ACLF development. Finally, a significant proportion of patients developing ACLF did so in the absence of any identifiable trigger. Mortality was independent of the presence and type of precipitating events, indicating that although triggers are important in the development of ACLF, mortality depends on other factors, such as the clinical course and number of organ failure [5, 6].

Regarding the prognostic relevance of ACLF, some findings have been reported as significant in determining patient outcomes [5]:

Another very important finding from the CANONIC study needs to be outlined: the fact that the presence or absence or the type of precipitating event is not related to the severity of ACLF and the short-term mortality rate [5]. Therefore, precipitating events are important in the occurrence of the syndrome but once it develops the prognosis depends on the number of organ failures. This observation indicates that the severity of ACLF probably depends more on the individual response to the precipitating event [6]. Finally, in almost half of patients enrolled in the CANONIC study with ACLF, the syndrome develops in the absence of a prior history of decompensation or has developed within a few weeks after the first episode of decompensation. This finding outlines that ACLF is not a terminal event in a long-lasting history of decompensated cirrhosis.

The CANONIC study results clearly show that white cell count and plasma C-reactive protein (CRP) levels are higher in patients with ACLF than in those without indicating higher degree of systemic inflammation in the former patients [5]. Furthermore, the higher white cell count or CRP levels the higher the number of failing organs. All in all, these findings suggest that organ failures may result from an excessive inflammatory response, for which the term immunopathology was proposed [7]. Although the precise mechanisms involved in ACLF have yet to be clarified, the immune system seems to play a predominant role in the setting of cirrhosis. The homeostatic role of the liver in the systemic immune response is already well known, [8, 9] and the definition of “cirrhosis-associated immune dysfunction” which includes the main syndromic abnormalities of immune function, immunodeficiency and systemic inflammation, well depicts the key role of the immune system in this setting [10]. The immune dysfunction in cirrhosis is a dynamic condition which leads to oscillation from predominantly pro-inflammatory to predominantly immunodeficient situations, is multifactorial, and reflects a complex interaction between many systems predisposing these patients to infections [10]. It is thought that this susceptibility is not due to an only sole responsible factor but rather to the concomitant presence of various facilitating mechanisms such as portal hypertension with portosystemic shunting (thus impairing detoxification and reticuloendothelial system phagocytic activity), increased gut permeability and bacterial overgrowth (all of them increases the risk of bacteremia and the occurrence of endotoxemia), albumin and lipoprotein dysfunction, or aberrant toll-like receptor expression in hepatic Kupffer cells [1]. Moreover, comparing septic patients to ACLF patients, Wasmuth et al. formulated the concept of “sepsis- like immune paralysis” based on a profoundly decreased production of TNF-α and low monocyte HLA-DR expression in both groups. They also postulated that this cellular immune impairment could contribute to increased mortality [11]. Endotoxins have also been proposed to play a role in mediating the full activation of neutrophils, which paradoxically would render them unable to act against the insult. The role that cytokines play in ACLF remains a key point in the pathogenesis of the inflammatory response. Elevated serum levels of many cytokines including TNF-α, sTNF-αR1, sTNF-αR2, interleukin (IL)-2, IL-2R, IL-4, IL-6, IL-8, IL-10, and interferon-α has been described. In particular IL-6 and TNF-α had been proposed to have a dual action, producing hepatocyte death and also enhancing hepatocyte proliferation through a complex interplay with Kupffer cells and hepatocytes [1]. This entire cascade eventually leads to hepatocyte death and liver dysfunction. It has also been outlined that hepatocytes apoptosis rather than necrosis can be the predominant mode of cell death in ACLF, as high levels of some apoptosis markers occurs in ACLF patients [12].