Precipitants of Acute-on-Chronic Liver Failure: An Opportunity for Preventative Measures to Improve Outcomes

Abstract

Acute-on-chronic liver failure (ACLF) is a feared complication that can develop at any stage of chronic liver disease. The incidence of ACLF is increasing, leading to a significant burden to both the affected individual and health care systems. To date, our understanding of ACLF suggests that it may be initiated by precipitants such as systemic infection, alcohol use, or viral hepatitis. The prevalence of these vary significantly by geography and underlying liver disease, and these precipitants have a varying impact on patient prognosis. Herein, we present a review of our current understanding of the precipitants of ACLF, including gaps in current data and opportunities for meaningful intervention and areas of future research.

The Impact of ACLF

Acute-on-chronic liver failure (ACLF) represents the summation of chronic liver disease and the associated extrahepatic organ failure. It is a syndrome distinct from the natural stepwise progression of liver disease from compensated to decompensated cirrhosis, wherein there is a rapid deterioration of liver function with the associated hepatic and systemic inflammation and consequent extrahepatic organ failure (Fig. 1). Although this pattern underscores all forms of ACLF, the current operationalized definitions are heterogenous with varying inclusion, exclusion, and organ-failure criteria, leading to a lack of uniformity (Fig. 2).^(1–4) Regardless of this heterogeneity, ACLF is an ominous, acute portrayal of hepatic decompensation and portends both a poor prognosis and a great burden to the affected individual (Fig. 3).^(5–8) Although recent studies have demonstrated that the frequency and impact of ACLF are rising, a paucity of data regarding the precipitants of ACLF has limited efforts at targeted intervention. Herein, we review the current understanding of ACLF precipitants in an effort to highlight opportunities for meaningful intervention and guide future research.

FIG. 1.

The role of precipitating factors in ACLF.

FIG. 2.

Characteristics of major definitions of ACLF.^(1–4) High-risk organ failure includes renal failure or other organ failure with either renal or cerebral dysfunction.

FIG. 3.

The 1-month mortality by grade of ACLF.^(2,3)

Our current understanding of the clinical and pathophysiologic features of ACLF, regardless of precipitant, follows a similar paradigm.⁽⁹⁾ Specifically, cirrhosis and its associated portal hypertension, intestinal dysbiosis, and permeable gut barrier allow for bacterial translocation that, in turn, drives local and systemic inflammation, culminating in intrahepatic and extrahepatic organ failure.⁽⁹⁾ The precipitating events, or precipitants, represent bursts of inflammation that serve as the nidus of the vicious cycles that drive ACLF.^(10,11) To date, the precipitants of ACLF identified in the literature represent both extrahepatic (ie, infection, bleeding) and intrahepatic (ie, alcohol intake, viral hepatitis) insults. The prevalence of these precipitants varies by geography, etiology of liver disease, demographics, and definition of ACLF used (Table 1).^(12–15) This review focuses not only on the currently available data regarding the precipitants of ACLF, but it also highlights the preventative strategies that have been described.

TABLE 1.

Precipitants of ACLF with Percentages Reported in Large Databases

	Europe	North America	Asia
Precipitating factors  Extrahepatic
Infection	33	—	28
GI hemorrhage	13	—	10
Procedures (TIPS, LVP, ERCP)	9–11	—	2
Intrahepatic
Viral hepatitis	<5	—	36
Alcohol	25	—	6
DILI	—	—	11
Autoimmune disease	—	—	3
Not identifiable	44	—	20
More than 1 factor	14	—	9
Underlying etiology of cirrhosis
Viral	18	52	63
Alcohol	67	42	30
Cryptogenic	3	15	7
Miscellaneous	8	18	4

NOTE: Data are given as percentages and are from Moreau et al.,⁽²⁾ Bajaj et al.,⁽⁸⁾ Cholongitas et al.,⁽¹³⁾ Wehler et al.,⁽¹⁴⁾ Xia et al.,⁽¹⁵⁾ and Shi et al.⁽¹⁶⁾

Systemic Infection

Despite geographic variations in prevalence, infections are the most common precipitants of ACLF. In the European CANONIC study, infection was estimated to cause nearly 40% of ACLF cases.⁽²⁾ Similar rates are found throughout the world, including 35% of ACLF in a Chinese cohort.⁽¹⁶⁾ These infections tend to be bacterial, specifically gram-negative rods (ie, klebsiella spp., pseudomonas, Escherichia coli). However, in recent years with the evolution of health care environments, increased utilization of antibiotics, and the emergence of multidrug-resistant bacteria, the distribution of infections as precipitants of ACLF are evolving to include more gram-positive and multidrug-resistant organisms (ie, enterococcus faecium, vancomycin-resistant enterococcus spp., staphylococcus aureus).^(17‘19) Fungal infections represent a small (<5%) infectious precipitant of ACLF, but they are associated with worse outcomes.^(20,21) The most likely source of infection varies by the study but includes urinary tract infections, spontaneous bacterial peritonitis, pneumonia, secondary peritonitis, spontaneous bacteremia, and skin/soft tissue infections.^(17,18,21)

Regardless of the source or type of infection, infection-precipitated ACLF is associated with a poor prognosis. In fact, some studies estimate that 30-day mortality rates associated with infection-precipitated ACLF are as high as 50%.^{(17,18,20–22)} Moreover, it is estimated that nearly one-third of patients with cirrhosis hospitalized with an infection will die within a month of admission and nearly another one-third will die within 1 year.⁽²³⁾ Despite these staggering numbers, the survival in patients with cirrhosis and severe sepsis seems to have improved little in the last 20 years.^(23,24) These findings highlight the need for prevention and appropriate treatment of infection in patients with ACLF.

Patients with cirrhosis are prone to increased infectious insults, as a consequence of the so-called “cirrhosis-associated immune deficiency syndrome,” which is the relative inefficacy of the innate and adaptive immune systems in patients with cirrhosis.^(25,26) This inefficacy coupled with increased gut permeability predisposes patients to not only the development of infectious complications, but it also exposes them to a number of proinflammatory mediators, driving an aberrant systemic inflammatory response, and leading to further intrahepatic and extrahepatic organ failure.⁽²⁵⁾

Treatment of infectious precipitants of ACLF centers around both prophylaxis and early rational antibiotic treatment when infection is suspected or confirmed.⁽²⁷⁾ Antibiotic prophylaxis can effectively reduce the risk of infection and consequently infection-precipitated ACLF. This is well supported in primary prophylaxis for certain patients with ascites, secondary prophylaxis in patients with ascites with prior spontaneous bacterial peritonitis, and in patients with upper gastrointestinal (GI) bleeding.^(28–30) Beyond antibiotics, there are few modalities established to prevent infection-precipitated ACLF. G-CSF in small randomized studies has been shown to decrease the probability of sepsis by roughly 30% during follow-up.^(31,32) In contrast, a large multicenter German study (GRAFT) was recently stopped after the interim analysis for lack of efficacy.⁽³³⁾ Most importantly, in infection-precipitated ACLF, there appears to be a window of opportunity where the prompt implementation of appropriate antibiotics may halt the progression of ACLF. In fact, although infection-precipitated ACLF is associated with a worse prognosis (51% infection-precipitated versus 38% noninfection-precipitated 90-day transplant-free mortality), if appropriate antibiotics were given, then on multivariate analysis only the appropriate use of antibiotics was significantly associated with 90-day transplant-free survival (hazard ratio, 0.40; 95% confidence interval, 0.26–0.62; P < 0.001).^(21,34) Furthermore, there appear to be 2 alternatives to mitigate the impact of infection on mortality in ACLF:

1. The use of albumin to limit the burst of systemic inflammation associated with ACLF.⁽³⁵⁾

2. The administration of prophylactic antibiotics in patients with a high risk of infection.⁽³⁴⁾

Nevertheless, still more data are necessary to consolidate these hypotheses.

PREVENTATIVE STRATEGY

The prevention of infection-precipitated ACLF is centered on prophylaxis and early rational antibiotic treatment. Both modalities have been associated with improved survival among patients with cirrhosis.

GI Hemorrhage

Depending on the society, variceal or GI bleeding is not universally considered a precipitant of ACLF. In fact, Asian Pacific Association for the Study of the Liver only considers variceal bleeding a precipitant if it results in liver failure, whereas the European Association for the Study of the Liver and the American Association for the Study of Liver Diseases consider it as a precipitant in all cases.⁽²⁵⁾ Either way, it is suspected that acute variceal bleeding precipitates ACLF mainly due to hepatic ischemia, increased bacterial translocation from the gut, and subsequent bacterial infections.⁽³⁶⁾ The prevalence of variceal bleeding as the precipitant of ACLF ranges from 13% to 28% of patients depending on the study.^(2,25,37,38) The treatment of variceal bleeding in ACLF follows all established guidelines.^(39–41) These interventions include endoscopic therapy, vasoactive infusions, nonselective beta-blockers (NSBBs), and transjugular intrahepatic portosystemic shunt (TIPS) in selected candidates. NSBBs have been shown to decrease splanchnic vascular resistance, to increase gut motility, and to prevent bacterial translocation. Therefore, there may be an added role of NSBBs in the prevention of ACLF beyond just decreasing variceal hemorrhage.^(42–45) Also, preemptive TIPS in patients with Child C (≤13 points) seems to have a survival benefit.^(46–48) The mechanism of this protective effect is not clear, but given that ACLF is the strongest negative predictor for survival among patients with a GI bleed, there is suspicion that TIPS may prevent the development of ACLF.⁽⁴⁹⁾ In summary, these interventions have led to a significant reduction in the mortality associated with variceal bleeding.⁽⁵⁰⁾

PREVENTATIVE STRATEGY

The prevention of GI bleeding–precipitated ACLF is focused on preventing hepatic ischemia and bacterial translocation, which is accomplished through appropriate transfusion parameters, the use of vasoactive medications, early endoscopic evaluation, utilization of decompressive procedures, and early application of prophylactic antibiotics.

Invasive Procedures

The data regarding the role of invasive procedures as a precipitating event of ACLF are sparse. In the CANONIC study, large-volume paracentesis (LVP), TIPS, and major surgery precipitated 9% of all cases of ACLF.⁽²⁾ In an analysis from China, these represented 2%.⁽¹⁶⁾ Treatments for hepatocellular carcinoma (HCC; eg, transarterial chemoembolization and radiofrequency ablation) have been associated with acute deterioration and ACLF in patients with Child-Pugh stage B or C cirrhosis with a decreased hepatic reserve. The exact prevalence of HCC treatments as a precipitant of ACLF is not known.⁽²⁵⁾ A recent study investigating the safety of endoscopic retrograde cholangiopancreatography (ERCP) demonstrated that the rate of post-ERCP ACLF was 11%. However, in those who developed an adverse event after an ERCP, the rate was 26%. The risk of ACLF was significantly higher in those with greater hepatic decompensation (Model for End-Stage Liver Disease score ≥15) at baseline. The occurrence of ACLF after ERCP was driven by adverse events, particularly cholangitis.⁽⁵¹⁾ Regardless of the invasive procedure, it appears that diligent assessment of hepatic reserve should be completed prior to any treatment, in an effort to prevent the development of ACLF.

PREVENTATIVE STRATEGY

There are limited data on the prevention of ACLF after invasive procedures (eg, LVP, TIPS, major surgery, local regional therapy, ERCP). However, it appears that the development of ACLF is most linked to the degree of underlying liver disease and complications of the invasive procedure. Therefore, diligent assessment of hepatic reserve and the limitation of complications are critical in preventing ACLF after invasive procedures.

Viral Hepatitis

Viral hepatitis is a common precipitant of ACLF in Eastern countries.^(52,53) Although, it remains infrequent in Western countries, rates of hepatitis B virus (HBV) and hepatitis C virus (HCV) infections are rising due to increased intravenous drug use.^(2,54) ACLF due to viral hepatitis is severe with high mortality rates, and despite the availability of vaccines for hepatitis A virus (HAV), HBV, and hepatitis E virus (HEV) and better HCV treatments, ACLF due to viral hepatitis remains an ongoing issue with opportunities for intervention.

HBV is the most common virus infection to precipitate ACLF.⁽¹⁶⁾ In Asia, HBV accounts for approximately 35% of ACLF cases, whereas in Europe and the United States, HBV accounts for <10% of cases.^(2,16) HBV can cause ACLF in acute infection or in the setting of HBV reactivation in the setting of treatment cessation, HBV resistance, chemotherapy, or immunosuppression.^(16,55) Excess innate immune responses to viral antigens mediated by pathogen-associated molecular patterns and damage-associated molecular patterns (DAMPs) may play a role in the development of ACLF.^(2,56) Host and virus genetic factors have also been shown to predispose certain individuals to ACLF due to HBV.^(57,58) However, our ability to predict ACLF due to HBV still remains poor. The presentation of ACLF due to HBV is typically acute and severe resulting in an up to 50% 28-day mortality rate, and if the patient is listed for liver transplant, there is a high wait-list mortality.⁽⁵⁹⁾ Results differ on the estimated efficacy of nucleos(t)ide analogues in improving mortality in ACLF. However, studies have estimated that viral suppression with nucleos(t)ide analogues can lead to an absolute risk reduction ranging from 29% to 42% in 90-day mortality.^(59–61) That being said, because the mortality associated with HBV-precipitated ACLF still remains markedly elevated, the definitive treatment of ACLF due to HBV is liver transplantation, which provides a >80% 1-year survival.⁽⁶²⁾ However, liver transplant is not as readily available in endemic countries resulting in high inpatient mortality. The role of hepatitis D virus (HDV) in precipitating ACLF is not as well established. Given that ACLF due to HDV is difficult to treat with antiviral therapy, liver transplantation is the treatment of choice.⁽⁶³⁾ Because of the key contribution of HBV to HDV infection and a lack of an HDV vaccine, prevention of HDV infection can be best achieved by vaccination against HBV.

HAV and HEV are typically self-limiting illnesses that are the most common cause of acute hepatitis in Asia; however, they cause ACLF in approximately 6%–19% of cases.^(16,64) Individuals in endemic regions are often exposed to HAV at an early age, resulting in adult immunity, but in nonendemic regions, HAV can cause ACLF if individuals are not vaccinated, which has prompted increased vaccination efforts in Western countries.⁽⁵³⁾ In contrast to HAV, HEV exposure tends to occur in adults in endemic areas and the duration of HEV immunity may be shorter than HAV.⁽⁶⁵⁾ In endemic regions, rates of immunoglobin G antibodies to HAV are much higher in chronic liver disease patients than to HEV, predisposing adults to ACLF due to HEV.⁽⁵³⁾ Currently, a recombinant vaccine against HEV has been approved in China, which provides immunity in roughly 90% of patients for at least 4.5 years, but this vaccine is not yet widely available.⁽⁶⁶⁾ ACLF due to HAV and due to HEV is severe, resulting in high mortality. However, little is known about host and virus factors that predispose patients to ACLF, and no treatments exist except liver transplantation.

ACLF precipitated by HCV is a rare occurrence. There are currently no data available to estimate the role of acute HCV in the development of ACLF. Of the cases of ACLF in the setting of HCV described, these have often been related to HBV reactivation.⁽⁶⁷⁾

PREVENTATIVE STRATEGY

Particularly in endemic areas, the prevention of viral hepatitis (eg, primary infection, reactivation, and progression) is critical. The data support prevention, close monitoring, and use of antiviral medications in patients with cirrhosis at risk for ACLF. Vaccination, monitoring, and treatment strategies should follow guidelines for patients with cirrhosis.

Alcohol

Active alcohol use represents a frequent intrahepatic precipitant of ACLF.⁽⁶⁸⁾ The prevalence of alcohol as a precipitating event varies geographically. In the CANONIC European observational cohort study, active alcohol consumption (defined by more than 14 units per week in women and 21 units per week in men) within 3 months of hospitalization represented nearly 25% of precipitating events in patients with ACLF.⁽²⁾ In studies from the Asian Pacific Association for the Study of the Liver–Acute-on-Chronic Liver Failure Research Consortium (AARC), alcohol represented nearly 50% of intrahepatic precipitants of ACLF.⁽⁶⁹⁾ Therefore, although this burden varies geographically, it remains substantial universally.

The mechanism by which alcohol use precipitates ACLF is not clear. Although, alcohol as a precipitant of ACLF is suspected to follow the same paradigm as seen in all precipitants of ACLF, studies have alluded to several key differences. First, the systemic inflammation and immune dysfunction, as manifested by white blood cell count and C-reactive protein, are noted to be more elevated in all patients with alcohol-related ACLF.^(2,70,71) This suggests that alcohol-related liver injury may lead to a more potentiated inflammatory and immune reaction, as compared with the other precipitants of ACLF. This variation in the potency of the inflammatory and immune response may be related to a variation in the cytokines circulating in alcohol-related ACLF. Specifically, interleukin (IL) 8 tended to be more elevated in patients with alcohol-related ACLF as compared with IL6 with infection-related ACLF.⁽⁷²⁾ Second, there may be a difference in the mechanism of cell death in alcohol-related ACLF. Cell death is a key driver of ACLF, regardless of the precipitant. That being said, it is now established that DAMPs have immunologic properties that drive the inflammatory response. Recent evidence suggests that alcohol-related cell death is driven through nonapoptotic mechanisms, unlike nonalcohol-related cell death, which leads to a different profile of DAMPs being released that consequently alters the immunologic response.⁽⁷³⁾ Together, these findings suggest that alcohol-precipitated ACLF may cause a specific immunologic response that is distinct and possibly more potentiated than other precipitants of ACLF. These findings highlight both the importance of better modalities to improve alcohol cessation and precipitant-specific treatments of the downstream inflammatory and immunologic response.

PREVENTATIVE STRATEGY

Alcohol is a common and severe trigger of ACLF. Given its burden on patients with cirrhosis, preventative strategies should focus on interventions that promote abstinence. These include screening for use, counseling to prevent recurrence, and preventative strategies (eg, Alcoholics Anonymous, pharmacotherapy, and behavior modification).

Drug-Induced Liver Injury and Hepatotoxic Exposures

The data are limited regarding other intrahepatic precipitants of ACLF, such as drug-induced liver injury (DILI), hepatotoxic substances, and ischemic hepatitis. This is likely a consequence of an overall low prevalence of these conditions in the consortiums designed to study ACLF and the difficulty in confirming these diagnoses. However, a recent manuscript from the AARC database demonstrated that among 3132 patients in the cohort, 329 (11%) had a drug that was implicated as a precipitant to their ACLF. There was a significantly higher 90-day mortality among drug-induced ACLF as compared with nondrug-induced ACLF (47% versus 39%; P = 0.007).⁽⁷⁴⁾ These data highlight the need to quantitate the prevalence of these precipitants to determine if large-scale intervention to prevent their occurrence is warranted.

PREVENTATIVE STRATEGY

Efforts are needed to accurately assess the burden of DILI and hepatotoxin-associated ACLF. Preventative strategies should focus on patient and physician education regarding medication/supplement safety among patients with cirrhosis.

Autoimmune-Related Liver Disease

It is estimated that roughly 20% of patients with autoimmune hepatitis (AIH) have liver failure at initial presentation; although, how many of these patients qualified as ACLF remains unknown.⁽¹⁹⁾ To date, a single study has commented on the role of AIH flare in ACLF. This study demonstrated that roughly 3% of patients in the AARC database had an AIH flare as a precipitating event. Furthermore, their work suggests that those who received corticosteroid therapy have improved 90-day survival.⁽⁷⁵⁾ The prevalence and efficacy of corticosteroids for AIH-ACLF in non-Asian countries remains unknown. A recent article from Mexico with the majority of patients having AIH and/or primary biliary cirrhosis revealed that mitochondrial dysfunction may be associated with the development of ACLF in these patients.⁽⁷⁶⁾ There is a need to determine the role of uncontrolled autoimmune-related disease in the development of ACLF.

PREVENTATIVE STRATEGY

There are limited data regarding the burden of autoimmune-related disease as a trigger for ACLF. Presumably, improved disease control would prevent ACLF, but data supporting this are limited.

Unknown

A precipitating event is not always identified in patients who present with ACLF. In fact, the rates of unidentified precipitants are significant. In the CANONIC study, the rate was >40% and, in the evaluation of the AARC database, the rate was 20%.^(2,4,11,16) However, how much of this outcome is related to diagnostics that lack the precision to diagnose low-level infection or alternative pathways is not clear. For instance, the circulating bacterial DNA of specific pathogens, but not bacterial DNA in ascites, was associated with systemic inflammation in decompensated patients, suggesting that bacterial translocation may be an important precipitant in patients with no identified precipitant.^(77,78) The significant proportion of unidentified precipitants highlights an opportunity for future studies to either better evaluate the known precipitating factors (ie, unreported alcohol use or hepatotoxic medications) or to investigate alternative precipitants of ACLF (ie, bacteremia, pathogen-associated molecular patterns, or DAMPs).

PREVENTATIVE STRATEGY

The significant proportion of patients with cirrhosis and ACLF who have no identified precipitant highlights the need for improved diagnostics so that preventative strategies can be developed.

Future Directions: Areas for Further Research

This review of the precipitants of ACLF highlights several of the key gaps that should drive future research. First, we have yet to identify all the precipitants of ACLF, as upward of 40% of patients who develop ACLF have no known precipitant. This finding demands that we better characterize and diagnose the known precipitants (ie, infection, ongoing alcohol use, DILI, and others) and also explore alternative pathways (ie, PAMPs, DAMPs).^(2,16) Second, a significant proportion of cases of ACLF are precipitated by underlying liver disease and underlying systemic inflammation. Thus, a better understanding of the distinct profiles of systemic inflammation in patients with cirrhosis and acute decompensation is warranted. Third, infection, regardless of geography, is a major precipitant, and therefore, further investigations are needed into the expanded role of prophylactic antibiotic and nonantibiotic regimens in patients with chronic liver disease. Underlying all of these needs is a major weakness of current research in ACLF: the diversity of the definitions and designs of the studies of ACLF limit the generalizability and applicability of the findings. This highlights a need for future direction to improve the uniformity of current definitions and study design.

Conclusion: Moving Toward the Prevention of ACLF

In this review, we reported the precipitants of ACLF to highlight the opportunities for prevention. We demonstrated that there are significant regional variations in the precipitants of ACLF. We also described the impact of these precipitants on outcomes and reviewed the available treatments of those precipitants. Ultimately, we suspect that the most efficacious modalities to improve outcomes in patients with ACLF are those strategies focused on preventing its occurrence. In light of the burden that ACLF inflicts on the patient with cirrhosis, identifying these modalities to prevent the precipitants of ACLF are paramount because they have the greatest potential to improve outcomes.

Abbreviations:

AARC

Asian Pacific Association for the Study of the Liver–Acute-on-Chronic Liver Failure Research Consortium

ACLF

acute-on-chronic liver failure

AIH

autoimmune hepatitis

EASL-CLIF

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

DAMP

damage-associated molecular pattern

DILI

drug-induced liver injury

ERCP

endoscopic retrograde cholangiopancreatography

FiO₂

fraction of inspired oxygen

GI

gastrointestinal

HAV

hepatitis A virus

HBV

hepatitis B virus

HCC

hepatocellular carcinoma

HCV

hepatitis C virus

HDV

hepatitis D virus

HE

hepatic encephalopathy

HEV

hepatitis E virus

HIV

human immunodeficiency virus

IL

interleukin

INR

international normalized ratio

LVP

large-volume paracentesis

MAP

mean arterial pressure

MV

mechanical ventilation

NACSELD

North American Consortium for the Study of End-Stage Liver Disease

NSBB

nonselective beta-blocker

PaO₂

partial pressure of arterial oxygen

RRT

renal replacement therapy

TIPS

transjugular intrahepatic portosystemic shunt