Development of a Pharmacokinetic Model of Transplacental Transfer of Metformin to Predict In Vivo Fetal Exposure.

Two types of systems are used in ex vivo human placental perfusion studies to predict fetal drug exposures, that is, closed systems with recirculation of the maternal and fetal buffer and open systems using a single-pass mode without recirculation. The in vivo fetal/maternal (F:M) ratio of metformin, a cationic drug that crosses the placenta, is consistent with that reported in an open system ex vivo but not with that in a closed system. In the present study, we aimed to develop a pharmacokinetic (PK) model of transplacental transfer of metformin to predict in vivo fetal exposure to metformin and to resolve the apparent inconsistency between open and closed ex vivo systems. The developed model shows that the difference between open and closed systems is due to the difference in the time required to achieve the steady state. The model-predicted F:M ratio (approx. 0.88) is consistent with reported in vivo values [mean (95% confidence interval): 1.10 (0.69-1.51)]. The model incorporates bidirectional transport via organic cation transporter 3 (OCT3) at the basal plasma membrane, and simulations indicate that the use of trimethoprim (an OCT3 inhibitor) to prevent microbial growth in the placenta ex vivo has a negligible effect on the overall maternal-to-fetal and fetal-to-maternal clearances. The model could successfully predict in vivo fetal exposure using ex vivo human placental perfusion data from both closed and open systems. This transplacental PK modeling approach is expected to be useful for evaluating human fetal exposures to other poorly permeable compounds, besides metformin. SIGNIFICANCE STATEMENT: We developed a pharmacokinetic model of transplacental transfer of metformin, used to treat gestational diabetes mellitus, in order to predict in vivo fetal exposure and resolve the discrepancy between reported findings in open and closed ex vivo perfusion systems. The discrepancy is due to a difference in the time required to reach the steady state. The model can predict in vivo fetal exposure using data from both closed and open systems.