Application of in vitro, in silico, and in vivo methodologies to quantitatively study maternal-fetal disposition of xenobiotics

Abstract

Ninety-seven percent of pregnant women in the US use at least one medication during pregnancy, yet maternal-fetal disposition of xenobiotics is under-studied. Maternal-fetal disposition of xenobitoics is determined by the maternal disposition, fetal disposition, and transplacental distribution processes which depend on the physicochemical properties of the xenobiotics and the physiological factors of the mother and the fetus. However, due to ethical concerns and to avoid unnecessary risks pose to the developing fetus, only limited maternal-fetal disposition information can be ascertained through clinical studies. Therefore, alternative experimental approaches including in vitro, in silico, and in vivo methodologies are required to study the maternal-fetal disposition of xenobiotics. In this dissertation, the maternal-fetal disposition of domoic acid (DA), oxycodone, and fentanyl were quantitatively studied using in vitro, in silico, and in vivo methodologies. In chapter 2 and 3, the maternal-fetal disposition of DA, a hydrophilic algal toxin, was studied in vivo using cynomolgus monkeys as the preclinical model species. The toxicokinetics (TK) following repeated oral doses of DA before, during, and after pregnancy was measured and compared and the fetal disposition at term was described using a maternal-fetal TK model. The study showed that the renal clearance (CLr) of DA was increased by 30-90% during pregnancy, similar to the increase in creatinine clearance which suggested that the increase in CLr is likely mediated by the increase in glomerular filtration rate (GFR). The fetal-to-maternal plasma concentration (F/M) ratio at birth ranged between 0.3 to 0.6 and changed as a function of time. Using the maternal-fetal TK model, placental transport and recirculation of DA between the fetus and amniotic fluid were suggested to be the major determining factors of the maternal-fetal disposition of DA. In chapter 4, the fetal hepatic clearance of oxycodone, a CYP3A4 and CYP2D6 substrate, was studied in vitro using fetal liver microsomes (FLM) extracted from individual livers (n=18). The results of this study demonstrated that CYP3A7 metabolizes oxycodone to noroxycodone in the fetal liver, similar to the reaction mediated by CYP3A4 in the adult liver. The CYP3A7 expression in the FLMs was measured by HPLC-MS/MS to be 191-409 pmol/mg and the intersystem extrapolation factor (ISEF) was estimated to be 0.016-0.066 using 6β-OH-testosterone formation as the probe reaction. The noroxycodone formation clearance (CLint,FLM) predicted using the recombinant CYP3A7 activity together with the CYP3A7 expression and the ISEF of each FLM successfully predicted the observed CLint,FLM (0.15-1.13 μL/min/mg protein) with an average fold-error of 1.24-fold. To quantitatively predict the fetal hepatic clearance (CLh), the observed CLint,FLM was extrapolated using in vitro-to-in vivo extrapolation (IVIVE). The prediction suggested that the fetal liver plays a minimal role in maternal and fetal disposition of oxycodone. In chapter 5, the maternal-fetal disposition of fentanyl, a lipophilic opioid analgesic, following epidural dosing to parturient women was studied in silico using a maternal-fetal physiologically-based pharmacokinetic (mf-PBPK) model. To capture the disposition of fentanyl following epidural dosing, an epidural dosing site model was developed based on the physiology of the epidural space and verified using alfentanil as a model compound. Since fentanyl is predominantly metabolized by CYP3A4 in the adult, the fetal liver metabolism of fentanyl was measured in FLM to be 0.20 ± 0.05 μL/min/mg protein and extrapolated to predict the whole fetal liver intrinsic clearance (CLint,u) using IVIVE as established in chapter 4. The mf-PBPK model of fentanyl successfully predicted the maternal venous, umbilical venous and umbilical arterial plasma concentrations in parturient women and newborns following epidural dosing during labor and delivery with the average absolute fold errors (AAFEs) within 2-fold of the observed plasma concentrations. Using the verified mf-PBPK model, the impact of maternal, fetal, and transplacental distribution kinetics on the F/M ratio of fentanyl was demonstrated, highlighting the importance of these distribution kinetics on the interpretation of maternal-fetal disposition of xenobiotics. In conclusion, this dissertation demonstrated that the maternal-fetal disposition of xenobiotics can be quantitatively studied using in vitro, in silico, and in vivo methodologies and a framework was established to study the maternal-fetal disposition of other xenobiotics using these methodologies.