Population pharmacokinetic–pharmacodynamic modelling to describe the effects of paracetamol and N‐acetylcysteine on the international normalized ratio

Paracetamol is one of the most common pharmaceutical agents taken in self‐poisonings, and can increase the prothrombin time (PT) through liver injury, and in overdose without hepatic injury by reducing functional factor VII. PT is a measure of hepatic injury used to predict and monitor hepatotoxicity, reported as the international normalized ratio (INR). The antidote for paracetamol poisoning, N‐acetylcysteine (NAC), has been reported to have an effect on the PT. This analysis included patients from a retrospective case series, a prospective inception cohort of paracetamol and psychotropic (control) overdoses, and a cross‐over clinical trial. A population pharmacokinetic–pharmacodynamic model describing the pharmacodynamic effects of paracetamol and NAC on the INR was developed in Phoenix NLME. The dataset included 172 patients; the median age was 22 years (range 13–71 years). A one‐compartment model with first‐order input and linear disposition best described paracetamol pharmacokinetics. The population mean estimate of the concentration that induced a response halfway between the baseline and maximal pharmacological effect of paracetamol was 1302 μmol/L (242), the maximum effect of paracetamol was 0.534 (202; from baseline) and the maximum effect of NAC was 0.325 (9.03; from baseline). Both paracetamol and NAC contributed a pharmacological effect to the elevation of INR. The estimated paracetamol concentration that induced a response halfway between the baseline and maximal pharmacological effect was within the range of plasma paracetamol values studied, fivefold greater than the maximum therapeutic concentration, suggesting that an elevated INR would not be expected within the therapeutic range. Simulated 24 and 48 g paracetamol overdoses with NAC administration produced INR values (50th percentile) that reached the upper limit of, or exceeded, the reference range.