Acute liver failure: What is it?


  • Potential conflict of interest: Nothing to report.

To the Editor:

Quantitative synthesis of prognostic research findings in acute liver failure (ALF) requires uniformity in the definition of this disease. We wish to assess the variability and explicitness of the used ALF definitions in studies focusing on evaluating indicators of ALF outcome. We included studies published in HEPATOLOGY in the last decade that report original data from a clinical trial or observational study on adult patients with ALF and of which one of the main objectives was evaluating prognostic indicators for prediction of outcome, including adverse outcome.

Out of eight included studies, seven used five different ALF definitions (Table 1), and one did not report an explicit definition.

Table 1. Summary of Definitions
StudyNo. of Patients Evaluating Outcome PredictorsDefinition
Hepatic EncephalopathyWeeks From OnsetCoagulopathyNo Preexisting Liver DiseaseOther
  1. Abbreviations: INR, international normalized ratio; PT, prothrombin time.

  2. +, included in the definition.

Khandelwal etal.1309+26PT >15s or INR ≥1.5+ 
Kumar etal.270+4 + 
Volkmann etal.370+8INR >1.5+Jaundice
Parekh etal.4187+26PT >15s or INR ≥1.5+ 
Schmidt and Larsen51242-4    
Taylor etal.629+26PT >15s or INR ≥1.5+ 
Kremers et al.738828PT (INR) Asterixis, hyperbilirubinemia, or hypoglycemia
Antoniades etal.850Not reported

The differences in the ALF definitions can be characterized by four main components: (1) presence of hepatic encephalopathy with or without its grade of severity (reported in all studies, two of which reported its grade); (2) the interval between onset of disease and occurrence of hepatic encephalopathy (reported in six studies, three of which reported 26 weeks, two of which reported 8 weeks, and one of which reported 4 weeks); (3) presence of coagulopathy with or without its threshold value (reported in five studies, one of which operationalized it by international normalized ratio [INR], four of which operationalized it by prothrombin time or INR; four studies presented thresholds); and (4) presence of preexisting liver disease (reported in five studies). Some studies mentioned other elements in the definition, such as jaundice (one study) and asterixis, hyperbilirubinemia, or hypoglycemia (one study).

We found a large variation in the definition of ALF in eight studies in one decade in one journal. Although the exact implications of the differences in ALF definitions on performance of prognostic markers remain unknown, these differences hinder such quantitative analysis. There is room for improvement in the reporting of ALF definitions, and there is a need for a uniform accepted definition of ALF in future studies to evaluate prognostic indicators.

Kama A. Wlodzimirow*, Saeid Eslami*, Ameen Abu-Hanna*, Martin Nieuwoudt†, Robert A.F.M. Chamuleau‡, * Department of Medical Informatics, Academic Medical Center, University of Amsterdam, Amsterdam, The Netherlands, † South African DST/NRF Centre of Excellence in Epidemiological Modelling and Analysis, Stellenbosch, South Africa, ‡ Tytgat Institute for Liver and Intestinal Research, Academic Medical Center, University of Amsterdam, Amsterdam, The Netherlands.