Risk prediction of hepatotoxicity in paracetamol poisoning

Anselm Wong,Andis Graudins
DOI: https://doi.org/10.1080/15563650.2017.1317349
Abstract:Context: Paracetamol (acetaminophen) poisoning is the most common cause of acute liver failure in the developed world. A paracetamol treatment nomogram has been used for over four decades to help determine whether patients will develop hepatotoxicity without acetylcysteine treatment, and thus indicates those needing treatment. Despite this, a small proportion of patients still develop hepatotoxicity. More accurate risk predictors would be useful to increase the early detection of patients with the potential to develop hepatotoxicity despite acetylcysteine treatment. Similarly, there would be benefit in early identification of those with a low likelihood of developing hepatotoxicity, as this group may be safely treated with an abbreviated acetylcysteine regimen. Aim: To review the current literature related to risk prediction tools that can be used to identify patients at increased risk of hepatotoxicity. Methods: A systematic literature review was conducted using the search terms: "paracetamol" OR "acetaminophen" AND "overdose" OR "toxicity" OR "risk prediction rules" OR "hepatotoxicity" OR "psi parameter" OR "multiplication product" OR "half-life" OR "prothrombin time" OR "AST/ALT (aspartate transaminase/alanine transaminase)" OR "dose" OR "biomarkers" OR "nomogram". The search was limited to human studies without language restrictions, of Medline (1946 to May 2016), PubMed and EMBASE. Original articles pertaining to the theme were identified from January 1974 to May 2016. Of the 13,975 articles identified, 60 were relevant to the review. Paracetamol treatment nomograms: Paracetamol treatment nomograms have been used for decades to help decide the need for acetylcysteine, but rarely used to determine the risk of hepatotoxicity with treatment. Reported paracetamol dose and concentration: A dose ingestion >12 g or serum paracetamol concentration above the treatment thresholds on the paracetamol nomogram are associated with a greater risk of hepatotoxicity. Paracetamol elimination half-life: Patients with more severe hepatotoxicity are more likely to have a longer paracetamol elimination half-life. While median elimination half-life increases in those developing hepatotoxicity, there is wide variation in half-life, making this an insensitive parameter to use as a negative risk prediction tool. Prothrombin time (PT): An initially normal PT is associated with a lower risk of developing hepatotoxicity, but cannot be used alone to identify patients not requiring acetylcysteine treatment. Hepatic aminotransferase activity: A normal ALT activity on presentation is associated with a high negative predictive value of hepatotoxicity following paracetamol-poisoning. Psi parameter: The psi parameter takes into account the time from ingestion, the serum paracetamol concentration and the time to initiation of acetylcysteine. A hepatotoxicity risk nomogram based on this parameter may be easier to use, but is limited to acute ingestions. Paracetamol-aminotransferase multiplication product: If a hepatotoxicity risk nomogram is not available, an alternate strategy may be to use the paracetamol-aminotransferase product (<1500 low risk, 1500-10,000 low to moderate risk, >10,000 mg/L × IU/L high risk) to define liver injury risk. Serial blood tests can be performed if patients present prior to 8 h post-overdose for ultimate specificity, or a single blood test can be taken if presenting more than 8 h post-overdose. Patients receiving acetylcysteine within 8 h of their overdose, with a product less than 10,000 mg/L × IU/L have a low likelihood of developing hepatotoxicity. Any clinical trials of intensified treatment (e.g., higher dose) to prevent fulminant hepatic failure might potentially use a product of >10,000 mg/L × IU/L as a criterion for inclusion. The paracetamol-aminotransferase product <1500 mg/L × IU/L may also identify those suitable for an abbreviated acetylcysteine regimen. Newer biomarkers: These show promise in the early identification of patients with a higher risk of developing hepatic injury. Point of care devices measuring paracetamol adducts need further trials. Conclusions: Risk prediction tools can stratify those that are more likely to develop hepatotoxicity. Currently, the paracetamol-aminotransferase multiplication product may be such a tool. Novel biomarkers show promise but need further validation and greater clinical availability. These tools may help inform clinical trials on modified acetylcysteine regimens.
What problem does this paper attempt to address?