To the Editor:

Remien etal.1 have created an intriguing model that strives to account for multiple dynamic processes in the course of acetaminophen (APAP)-induced liver injury. They hope to devise a better instrument for predicting the need for liver transplant as early as possible. Hepatologists and clinical toxicologists know well that the patients who will die without transplant tend to do so with frightening speed. Unlike most other conditions warranting liver transplant, the time between diagnosis and a fatal outcome without transplant is only a few days.

Complexity undermines the model's utility. The model requires 10 simultaneous equations using 21 parameters (10 arbitrarily selected and 11 calculated) to solve for an estimated (not “predicted,” since the event was in the past) APAP dose and time since ingestion. From these two estimated values, one must then further estimate the probability of death. Unless the model can be distilled into a clever app for smartphones, it likely will not enjoy widespread use, even if the authors' results can be validated.

Some of the premises underpinning the model raise questions: that serum aspartate aminotransferase (AST) concentration is two times the serum alanine aminotransferase (ALT) concentration at baseline, that N-acetylcysteine (NAC) started after 24 hours is futile, and that the fraction of APAP metabolized to NAPQI is constant both among a population and within an individual throughout the course of injury. First, we know that serum AST and ALT concentrations are nearly equal in healthy people without liver disease or injury.2, 3 Second, we have known for two decades that NAC started after acute liver injury has occurred, and well after 24 hours since APAP overdose, actually reduces mortality by half.4, 5 Third, although about 4% of a therapeutic dose undergoes oxidation to form N-acetyl-p-benzoquinoneimine (NAPQI), there is no evidence that this proportion remains fixed in all patients regardless of ingested dose.

The authors assume APAP dose and time of exposure can reliably be calculated from subsequent transaminase concentrations and international normalized ratio (INR). There is no attempt to validate these estimates either by comparison with patient histories or by pharmacokinetic estimation based on measured APAP concentrations. There is no evidence that estimated APAP dose and time, if different from available history, have any prognostic value once acute liver failure (ALF) has occurred. The model still requires a post-hoc adjustment for serum creatinine concentration to improve its sensitivity.

The authors compare their work to the Rumack-Matthew nomogram6 and the psi parameter of Sivilotti etal.7 However, both of these determine need for NAC treatment in the first 24 hours (before liver injury occurs) and are not designed to predict death after liver injury appears.

The authors retrospectively apply the model to 53 patients but present data on 84 patients without explaining the discrepancy. Because most APAP-poisoned patients undergo repeated laboratory testing as the illness unfolds, the Model for Acetaminophen-induced Liver Damage (MALD) should be assessed with each new set of laboratory values. The risk assessment likely changes with each laboratory draw, and the earliest set may not be the single set with the best performance as a predictor.

Remien etal. conclude that the Kings College Criteria (KCC) are inferior to their MALD, while using only two parts of the KCC out of convenience. An incomplete assessment of the KCC will have poorer prognostic value than the complete KCC. Likewise, an incomplete assessment of the KCC will naturally perform more poorly than any other instrument that performs as well as the full KCC.

In our clinical practice, most patients with acute liver injury after APAP overdose infrequently meet the threshold of INR >6.5. This may be the least sensitive criterion in the KCC. We teach our residents to use an INR threshold of 2 when using the KCC.

For now, the MALD appears to be a pretender to the throne. Long live the King!


  1. Top of page
  • 1
    Remien CH, Adler FR, Waddoups L, Box TD, Sussman NL. Mathematical modeling of liver injury and dysfunction after acetaminophen overdose: early discrimination between survival and death. Hepatology 2012 Jul 6. DOI: 10.1002/hep.25656.
  • 2
    Piton A, Poynard T, Imbert-Bismut F, Khalil L, Delattre J, Pelissier E, etal. Factors associated with serum alanine transaminase activity in healthy subjects: consequences for the definition of normal values, for selection of blood donors, and for patients with chronic hepatitis C. Hepatology 1998; 27: 1213-1219.
  • 3
    Kunde SS, Lazenby AJ, Clements RH, Abrams GA. Spectrum of NAFLD and diagnostic implications of the proposed new normal range for serum ALT in obese women. Hepatology 2005; 42: 650-656.
  • 4
    Harrison PM, Keays R, Bray GP, Alexander GJ, Williams R. Improved outcome of paracetamol-induced fulminant hepatic failure by late administration of acetylcysteine. Lancet 1990; 335: 1572-1573.
  • 5
    Keays R, Harrison PM, Wendon JA, Forbes A, Gove C, Alexander GJ, etal. Intravenous acetylcysteine in paracetamol induced fulminant hepatic failure: a prospective controlled trial. Br Med J 1991; 303: 1026-1029.
  • 6
    Rumack BH, Matthew H. Acetaminophen poisoning and toxicity. Pediatrics 1975; 55: 871-876.
  • 7
    Sivilotti MLA, Good AM, Yarema MC, Juurlink DN, Johnson DW. A new predictor of toxicity following acetaminophen overdose based on pretreatment exposure. Clin Toxicol 2005; 43: 229-234.

Michael E. Mullins M.D.*, Evan Schwarz M.D.*, * Washington University School of Medicine, Emergency Medicine, St. Louis, MO.