Bayesian Forecasting Tool to Predict the Need for Antidote in Acute Acetaminophen Overdose

Area of Science:

• Pharmacokinetics
• Toxicology
• Hepatology

Background:

• Acetaminophen (APAP) overdose is a leading cause of acute liver injury in the US.
• Current risk stratification using the Rumack-Matthew nomogram has shown inaccuracies in predicting the need for N-acetylcysteine (NAC) treatment.
• Accurate prediction of APAP concentrations is crucial for timely hepatoprotective intervention.

Purpose of the Study:

• To develop a population pharmacokinetic (PK) model for APAP following acute overdose.
• To evaluate the utility of Bayesian forecasting, based on the population PK model, for guiding NAC administration decisions.
• To assess the accuracy of predicted APAP concentrations and their impact on treatment decisions.

Main Methods:

• A population PK model was developed using retrospective data from acute APAP overdose cases.
• The model incorporated factors like APAP product type and activated charcoal administration.
• Bayesian forecasting was employed to predict individual APAP concentration-time profiles using one or two plasma acetaminophen concentrations (PACs).

Main Results:

• A one-compartment model effectively described APAP pharmacokinetics, with covariates influencing absorption and bioavailability.
• Bayesian forecasting demonstrated acceptable bias (6.2-9.8%) and accuracy (40.5-41.9%) in predicting APAP concentrations with one or two PACs.
• Using one PAC, Bayesian forecasted NAC decisions showed a sensitivity of 84% and a negative predictive value of 92.6%.

Conclusions:

• The developed population PK model provides a reliable platform for predicting individual APAP concentration-time profiles after overdose.
• This model, with at least one PAC, can aid in making early and informed N-acetylcysteine administration decisions.
• Population PK modeling offers a promising approach to enhance the management of acute acetaminophen poisoning.