The workflow for constructing and translating a PBPK model to pregnant women (Fig. 1 ) has recently been described [12 (link)] and is summarized here, as it is relevant to this analysis. PBPK models for acetaminophen were developed for a non-pregnant population for IV administration and then, keeping distribution- and clearance-related parameters unchanged, for oral administration [13 (link)]. The demographic measures of the virtual subjects matched those of the in vivo study group, if the latter were reported. The non-pregnant PBPK model incorporated the PK-Sim® standard model structure comprising 18 compartments [14 (link)]. The model was evaluated by comparing the pharmacokinetic simulation with observed in vivo pharmacokinetic data taken from literature. If needed, drug-specific parameters were refined in the non-pregnant model by fitting the simulated plasma concentration–time curve to the observed data using Monte Carlo algorithm implemented in the software’s Parameter Identification toolbox. Thereafter, all drug-specific parameters, except the fraction unbound (fu), were fixed and the model was extrapolated to pregnant women by substituting the standard model structure with the pregnancy model structure, which includes nine additional compartments [2 (link), 12 (link)]. Pharmacokinetic predictions were performed in a population of pregnant women that matched anthropometric measures of the in vivo study group of pregnant women. If not reported, the mean gestational age-specific demographic measures available in PK-Sim® were used. Finally, pharmacokinetic predictions were evaluated by comparing the results with in vivo pharmacokinetic data taken from literature.![]()
Schematic workflow of pregnancy physiologically based pharmacokinetic model development and validation. Css,avg average concentration at steady state, IV intravenous, NAPQI N-acetyl-p-benzoquinone imine, PBPK physiologically based pharmacokinetic, phys-chem physicochemical, PK pharmacokinetic
