AIM: To analyze characteristics and outcome of patients with acute-on-chronic liver failure (ACLF) according to the severity of underlying liver disease.

METHODS: One hundred and sixty-seven adult patients with chronic liver disease and acute deteriorated liver function, defined by jaundice and coagulopathy, were analyzed. Predisposition, type of injury, response, organ failure, and survival were analyzed and compared between patients with non-cirrhosis (type A), cirrhosis (type B) and cirrhosis with previous decompensation (type C).

RESULTS: The predisposition was mostly hepatitis B in type A, while it was alcoholic liver disease in types B and C. Injury was mostly hepatic in type A, but was non-hepatic in type C. Liver failure, defined by CLIF-SOFA, was more frequent in types A and B, and circulatory failure was more frequent in type C. The 30-d overall survival rate (85.3%, 81.1% and 83.7% for types A, B and C, respectively, P = 0.31) and the 30-d transplant-free survival rate (55.9%, 65.5% and 62.5% for types A, B and C, respectively P = 0.33) were not different by ACLF subtype, but 1-year overall survival rate were different (85.3%, 71.7% and 58.7% for types A, B and C, respectively, P = 0.02).

CONCLUSION: There were clear differences in predisposition, type of injury, accompanying organ failure and long-term mortality according to spectrum of chronic liver disease, implying classifying subtype according to the severity of underlying liver disease is useful for defining, clarifying and comparing ACLF.

Patients with either diagnosed or undiagnosed chronic liver disease occasionally present with an acute deterioration of liver function caused by direct or indirect insults to the liver. This event, in general, is termed acute-on-chronic liver failure (ACLF)[1]. However, because of varied etiology and manifestations, there has been a great heterogeneity in defining the disease and, in fact, a recent systematic review has found more than a dozen definitions for ACLF[2]. Among these, there are 2 mainstream definitions most widely used in clinical settings and research. The Asian Pacific Association for the Study of the Liver (APASL) consensus introduced an ACLF definition as “an acute hepatic insult manifesting as jaundice (serum bilirubin ≥ 5 mg/dL) and coagulopathy (INR ≥ 1.5), complicated within 4 wk by ascites and/or encephalopathy in patients with previously diagnosed or undiagnosed chronic liver disease/cirrhosis, and is associated with a high 28-d mortality”[3]. The European Association for the Study of the Liver-chronic liver failure (EASL-CLIF) Consortium defined ACLF as acute decompensation of cirrhosis in the form of one or more major complications of liver disease, including ascites, hepatic encephalopathy, gastrointestinal bleeding, and bacterial infection, associated with at least two organ failures with one being renal failure (serum creatinine ≥ 1.5 mg/dL) and high 28-d mortality of greater than 15%[4]. They share a common idea that ACLF is a spectrum of disease with varying severity, characterized by multiple organ failure and high mortality[5-8], with liver transplantation (LT) as the only definitive curative option[9-11]. However, the discordant details between the 2 definitions results in confusion rather than clarification of the situation. In specific, the 2 groups suggest different explanations for the duration of illness, severity of the underlying liver disease, and the type of precipitating events. However, these factors are what essentially define ACLF, because each factor determines the acuteness of event, chronicity of liver disease, and the type of insult, respectively.

To embrace previously suggested definitions and better clarify this condition, Jalan et al[12] recently proposed a new definition and classification for ACLF. They defined ACLF as “a syndrome in patients with chronic liver disease with or without previously diagnosed cirrhosis which is characterized by acute hepatic decompensation resulting in liver failure (jaundice and prolongation of the INR) and one or more extrahepatic organ failures that is associated with increased mortality within a period of 28 d and up to 3 mo from onset”[12]. In their definition, they included any chronic liver disease, regardless of the presence of cirrhosis, and a wide range of precipitating events that could damage liver function either directly or indirectly. Patients with chronic liver disease but without cirrhotic features were categorized as type A. Type B ACLF was a group with previously well-compensated cirrhosis, while those with a history of jaundice and/or complications of portal hypertension were included in type C. However, they admitted that this is a working definition which is only to identify patients from whom to collect data to ultimately reach a validated definition.

Therefore, the aims of this study were to investigate and compare clinical characteristics, including predisposition (etiologies of chronic liver disease), injury (precipitating events), response (manifestations), short and long-term outcomes according to spectrum of chronic liver disease (non-cirrhosis, cirrhosis and cirrhosis with previous decompensation), thereby, to determine relevance of subtyping ACLF by severity of underlying disease for future practice and research.

ACLF, by in the simplest term, is “abrupt hepatic decompensation in patients with chronic liver disease”[12]. Moreover, ACLF is usually defined as a condition wherein patients are at significantly increased risk for mortality, with improvement in survival with LT[12]. There is controversy regarding what constitutes a chronic liver disease. APASL includes non-cirrhotic chronic liver disease/cirrhosis, but not decompensated cirrhosis to define chronic liver disease[3,15] whereas EASL-AASLD includes only cirrhosis, either compensated or decompensated[4], The newly proposed ACLF definition includes all spectrum of chronic liver disease and categorizes them as type A (non-cirrhotic), type B (cirrhosis), and type C (decompensated cirrhosis)[12]. The present study included all stages of chronic liver disease patients with acute deteriorated liver function, defined by jaundice (serum bilirubin ≥ 5 mg/dL) and coagulopathy (PT INR ≥ 1.5). All types of patients showed comparably high short-term mortality, and improved survival with LT. The overall survival was significantly better for patients who received LT, and the survival benefit by LT was most prominent in type C (1-year survival rate: 80.0% vs 42.1% for with and without LT, P = 0.007). Thus, our finding support that ACLF can be defined for all spectrum of chronic liver disease (i.e., non-cirrhotic to decompensated cirrhosis).

ACLF, by definition, also implies acute insult that leads to acute deterioration of liver function[3]. For example, superimposed viral hepatitis in chronic liver disease is a well-known injury to put patients at higher risk for acute deterioration of liver function[16]. Bacterial infection or sepsis is also a major risk factor for adverse event in cirrhosis patients[17,18]. Acute insults that lead to ACLF can be categorized as hepatotrophic insult (viral hepatitis, alcohol, drug, autoimmune hepatitis, Wilson’s disease, etc.) and non-hepatotrophic insults (infection, surgery, bleeding, etc.)[3]. Because ACLF is a potentially reversible disease[1], it is very important to identify acute insults, and have immediate intervention (e.g., antiviral therapy for hepatitis B, treatment for infection). Importantly, except for few causes where specific injury was uncertain, all type A injuries were hepatotrophic injury (Table 1). In contrast, type C mostly comprised non-hepatotrophic insults, although hepatotrophic insults were observed as well. Type B was in between type A and type C. Organ failure rate was similar in terms of renal, cerebral, coagulation and respiratory failure by ACLF type, but hepatic failure was less frequent and circulatory failure was more frequent in type C (Table 1). Higher rate of circulatory failure in decompensated cirrhosis could be explained by circulatory dysfunction observed in cirrhosis[19]. Differences in injury (precipitating insult) and response (organ failure rates) by ACLF types provide a rationale to categorize ACLF as type A, B and C, which is also helpful in searching potential injury and planning specific intervention in patients with ACLF. It is also noteworthy that type B and type C were still at risk for mortality after 30 d, leading to a decreased 1-year survival compared to type A. This also provides a rationale to categorize patients who experienced ACLF, as different long-term prognosis is expected.

The presence of “one or more extrahepatic organ failure” was suggested to define ACLF in the newly proposed ACLF definition[12]. However, question remains whether “extrahepatic organ failures” are mandatory to define ACLF. ACLF, by its simplest meaning, does not include “extrahepatic organ failures”. The purpose of including “extrahepatic organ failures” in the ACLF definition is to define the population at high risk for mortality. Our data did indeed show extremely poor short-term survival for patients with extrahepatic organ failures defined by CLIF-SOFA (55.9% and 48.9% for overall and transplant free 30-d survival), and good short-term survival for patients without any organ failures defined by CLIF-SOFA (90.0% and 78.7% for overall and transplant free 30-d survival). This is consistent with previous findings that extrahepatic organ failures are an important factor for survival in ACLF[20]. The association between organ failure and poorer prognosis is supported by numerous studies[4,11,21-26]. However, patients with only hepatic/coagulation failure also showed high short-term mortality (79.6% and 54.9% for 30-d overall and transplant-free survival). Furthermore, mortality rate exceeding 15% without LT (90.0% and 79.7% for 30-d overall and transplant-free survival) even for patients without any organ failures defined by CLIF-SOFA. Short-term mortality rate of 15% are used to define increased mortality in patients with ACLF[4]. In this sense, our findings suggest that patients without extrahepatic organ failures could be defined as ACLF as well, when they present with acutely deteriorated liver function, defined by jaundice and coagulopathy. It is likely that presence of “extrahepatic organ failures”may reflect severity of ACLF, however, it may not be a prerequisite for being defined as ACLF. In fact, the APASL definition does not require “extrahepatic organ failures” to define ACLF[3].

This study had some limitations. First, some of information on precipitating events could not be obtained regardless of thorough review of medical records because of the retrospective study design. We defined acute deteriorated liver function by jaundice and coagulopathy, as suggested by APASL[15]. However, in the CANONIC study, acute development of large ascites, hepatic encephalopathy, gastrointestinal hemorrhage, bacterial infection or any combination of above, are used to define acute deteriorated liver function[4]. Thus, patients who do not fulfill APASL criteria for liver failure, but present with acute deteriorated liver function, defined in the CANONIC study, are beyond the scope of this study. Lastly, sample size was relatively small, requiring larger scale, prospective study.

In conclusion, ACLF is a distinct syndrome that can be defined as acute deteriorated liver function, precipitated by either direct or indirect injury to the liver, resulting in high short-term mortality with LT as a curative option. The study showed that all spectrum of chronic liver patients with acute deteriorated liver function defined by jaundice (serum bilirubin ≥ 5 mg/dL) and coagulopathy (PT INR ≥ 1.5) can be defined as ACLF. The newly proposed ACLF classification was clinically relevant because different insults and response could be expected, which also suggest different pathophysiology, different management strategy, and different long-term outcome. This suggest that classifying ACLF subtype according to the severity of underlying liver disease is useful for defining, clarifying and comparing ACLF.