Predicting Human Fetal Drug Exposure to Placental P-glycoprotein Substrates by PBPK Modeling and Simulation

Abstract

More than 80% of pregnant women take at least one drug during pregnancy and about 50% during the first trimester when the fetus is the most vulnerable to drug toxicity (Scaffidi, Mol, & Keelan, 2017). To inform fetal risk and to optimize maternal-fetal drug therapy, it is important to measure or predict fetal drug exposure throughout pregnancy. Estimating fetal drug exposure by repeated sampling of fetal blood is not ethical or feasible. Obtaining a single plasma (blood) sample from the fetus is possible at term (e.g., from the umbilical vein). However, these sparse data are not sufficient to determine fetal drug exposure and hence alternative methodologies like physiologically-based pharmacokinetic (PBPK) modeling can be employed to predict (rather than measure) fetal drug exposure throughout pregnancy. We previously developed and verified a maternal-fetal (m-f) PBPK model that can successfully predict maternal and fetal exposure at term to drugs that passively diffuse across the placenta (Zhang et al., 2017; Zhang & Unadkat, 2017). Here, for the first time, we extended this model to predict fetal drug exposure to drugs that are effluxed by P-glycoprotein, a drug transporter highly abundant in the human placentae (Anoshchenko et al., 2020; Han, Gao, & Mao, 2018; Mathias, Hitti, & Unadkat, 2005). To do so, we used the efflux ratio-relative expression factor (ER-REF) approach to predict fetal drug exposure, Kp,uu, at term (i.e., the ratio of fetal and maternal unbound drug plasma AUC), of four placental P-gp substrates: dexamethasone, betamethasone, darunavir and lopinavir. The ER-REF approach relies on scaling the in vitro drug efflux ratio (ER) in a transporter-overexpressing cell monolayer (e.g., hMDR1-MDCKcP-gpKO cells where human P-gp was overexpressed and the endogenous canine P-gp was knocked-out) to in vivo Kp,uu using the relative expression factor (REF), the ratio of transporter abundances in placental tissue (Chapter 2) and in hMDR1-MDCKcP-gpKO cells (Chapter 3). To verify our predictions, we compared the ER-REF predicted fetal Kp,uu with the in vivo fetal Kp,uu, estimated from the observed UV/MP ratio data (in multiple maternal-fetal dyads) using our m-f PBPK model (Chapter 3). The predicted fetal Kp,uu for dexamethasone, betamethasone, darunavir and lopinavir were 0.63, 0.59, 0.17 and 0.08, respectively, and fell within the 90% confidence interval (CI90%) of their estimated observed fetal Kp,uu (0.30 – 0.66, 0.29 – 0.71, 0.11 – 0.22, 0.04 – 0.19, respectively), indicating success of our ER-REF approach. Using the above fetal Kp,uu data and our m-f PBPK model, we designed alternative dosing regimens for dexamethasone and betamethasone (Chapter 4) to address concerns regarding their maternal-fetal safety and efficacy (Vogel et al., 2017). To retain fetal efficacy, we propose maintaining dexamethasone total dose (24 mg) administered to the mother IM over 48 h. For betamethasone, we found that its dose (24 mg over 48 h) could be potentially decreased by up to 80%. These regimens illustrate the utility of our ER-REF approach and m-f PBPK model to dynamically predict fetal exposure to drugs and therefore alternative dosing regimens of drugs administered to pregnant women. Any proposed alternative dosing regimens should be rigorously tested in the clinic for efficacy and toxicity prior to implementation. The success of our ER-REF approach supports further utility of this approach together with our m-f PBPK model to estimate 1) fetal Kp,uu at term of other placental P-gp substrates; 2) fetal Kp,uu at term for substrates of other placental transporters or multiple transporters (e.g., BCRP or BCRP/P-gp), provided their placental abundance at term has been quantified (see Chapter 2 for such quantification); 3) fetal Kp,uu at earlier gestational ages for substrates of various/multiple transporters (provided placental transporter abundance at various gestational ages is quantified; see Chapter 2 for such quantification). This work emphasizes the need to predict fetal drug exposure to inform fetal drug efficacy and safety and to optimize drug dosing regimens for the maternal-fetal dyad.