Pharmacometrics in obstetrics and maternal–fetal medicine research: Bridging gaps in maternal and fetal pharmacology

Although pharmacometric approaches play a critical role in modern drug development, their application in pregnancy is still limited, despite the widespread use of medications during gestation. Approximately 70%–80% of pregnant women use at least one prescription medication during the first trimester, and 90% take at least one medication during the course of their pregnancy²; yet, the effects of many of these drugs on pregnancy remain unknown. By leveraging complex mathematical models such as PBPK and PopPK approaches, researchers can simulate maternal and fetal drug exposure, optimize therapeutic regimens, and predict potential drug–drug interactions. The significant potential of pharmacometrics to address these critical issues in maternal and fetal pharmacology underscores the need for greater integration of these methodologies into clinical practice and research.

Pregnancy is a unique physiological state characterized by profound alterations in the absorption, distribution, metabolism, and elimination (ADME) of drugs.³ Pregnancy-induced physiological changes affect multiple organ systems, including the cardiovascular, renal, hepatic, and gastrointestinal systems. As gestation progresses, maternal blood volume increases, glomerular filtration rate (GFR) rises, and hepatic enzyme activity is altered, impacting bioavailability, drug metabolism, and clearance.³ For instance, in pregnancy, the activity of cytochrome P450 enzymes such as CYP3A4 increases while the activity of others like CYP1A2 decreases, leading to significantly greater variability in drug disposition.³ These changes can pose significant challenges in determining optimal dosing, efficacy, and safety profiles for medications used during pregnancy, raising concern for both under- and overtreatment. Notably, most knowledge regarding the pharmacokinetics and safety of medications used during pregnancy is typically acquired 6–8 years after initial drug licensure,⁴ highlighting the urgent need for advanced modeling approaches for earlier prediction of maternal and fetal drug exposure. Pharmacometrics provides an invaluable framework for addressing these challenges, making it indispensable in contemporary obstetrics and maternal–fetal-medicine research.

Pharmacometrics has shown utility in critical areas of obstetrics, particularly in predicting drug dosing and ensuring drug safety. For instance, PBPK models have effectively predicted maternal and fetal drug exposure for medications like nifedipine, allowing for safe management of preterm labor and pregnancy-induced hypertension.⁵ Additionally, PopPK approaches have been employed to optimize dosing and to identify key covariates affecting drug disposition for magnesium sulfate administration for seizure prophylaxis in pre-eclampsia, considering factors such as altered plasma protein binding, volume of distribution, and clearance.⁶ However, there is limited data on the application of pharmacometrics in many other pregnancy-specific diseases, including intrahepatic cholestasis of pregnancy, HELLP syndrome, hyperemesis gravidarum, placental abruption, placenta previa, and eclampsia. These conditions are complex and exhibit variable clinical presentations, making it challenging to accurately model drug pharmacokinetics and pharmacodynamics.

Maternal PBPK models have evolved significantly to describe gestational-dependent changes across trimesters, enhancing our understanding of how predicted maternal drug exposure correlates with observed outcomes and facilitating more precise and safer dosing regimens.⁷ However, further refinement is needed to address gaps in predicting drug interactions and to account for understudied metabolic pathways and elimination routes in pregnancy.

Fetal PBPK models are equally essential for predicting fetal drug exposure and often build on existing ex vivo placental perfusion studies to estimate drug transfer across the fetal–placental interface into the fetal circulation. Several fetal PBPK models have been developed in recent years. Zhang and Unadkat⁸ incorporated key fetal physiological characteristics, such as placental compartments, fetal hepatic metabolism, and renal excretion, enabling reliable predictions of fetal exposure to antiretroviral drugs (ART) like tenofovir and nevirapine in pregnant women with HIV. More recently, improvements in fetal PBPK models have included the integration of placental blood flow and transporter expression, further refining predictions of drug transfer between the pregnant woman and her fetus. Shenkoya et al.⁹ advanced the field by incorporating the lymphatic system into fetal PBPK models, allowing for the prediction of ART penetration into lymphoid tissues—a critical reservoir for HIV—thus enhancing our understanding of ART distribution and prevention of perinatal HIV transmission. Despite these advancements, existing fetal PBPK models require further adjustment to better account for factors like gestational age-dependent changes, particularly during the first trimester, as well as other potential improvements discussed in Table 1. Incorporating a wider range of physicochemical drug properties and clinical scenarios could improve these models and provide valuable data for guiding medication dosing during pregnancy.

Population pharmacokinetic (PopPK) studies offer a powerful approach for developing models that characterize drug concentration–time profiles and optimize drug dosing while accounting for variability both between and within individuals. In a pregnancy PopPK study by Eke et al.,¹⁰ the probability of target attainment—defined as the proportion exceeding the area under the concentration–time curve (AUC) of >1.99 μg h/mL, the 10th percentile of average tenofovir exposure in nonpregnant controls, was demonstrated to be 68%, 80%, 87%, and 93% for tenofovir disoproxil fumarate (TDF) doses of 300, 350, 400, and 450 mg, respectively during pregnancy, and 88%, 92%, 96%, and 98% above the target with same doses in postpartum women, confirming the fact that dose adjustment of TDF during pregnancy is not generally warranted, thus optimizing TDF therapy for pregnant women living with HIV.

The application of PopPK models, while increasingly used in obstetrics drug research, faces significant challenges. One major issue is the difficulty in integrating empiric “central” and “peripheral” compartment-based modeling with the unique anatomy and physiology of pregnancy, as well as the dynamic physiological changes that occur throughout gestation. Inter- and intra-individual variability among pregnant women further complicates model generalization. This variability arises from factors, such as genetic differences, pre-existing medical conditions, pregnancy-related diseases, and varied responses to physiological changes, which can significantly impact drug pharmacokinetics. Trimester-dependent weight changes lead to intrinsic inter-individual variability and further complicate the accurate prediction of drug clearance and distribution across different stages of pregnancy. Overall, the scarcity of high-quality pregnancy data, particularly from the first trimester, hampers the development of robust models. Traditional PopPK models rely on population-wide data that is often lacking or unavailable for pregnant women. Future data collection is limited by ethical concerns regarding maternal and fetal drug exposure. Thus, while PopPK models hold promise for personalizing drug dosing during pregnancy, their accurate implementation is limited by physiological, ethical, and data-related barriers.

In addition to the challenges already discussed, including in Table 1, several key barriers hinder the widespread implementation of pharmacometrics in obstetrics. One critical challenge is the paucity of trained clinician-scientists with expertise in clinical pharmacology and pharmacometrics. Obstetric pharmacology is highly specialized, requiring a deep understanding of both maternal–fetal physiology and advanced modeling techniques, yet the number of experts who can bridge these disciplines is limited. This shortage hampers the translation of pharmacometric insights into clinical practice, as few researchers possess the dual expertise to conduct studies that integrate pharmacometrics into maternal–fetal medicine. Compounding this issue is the lack of an established implementation–science framework specifically designed for pharmacometrics in pregnancy. Such a framework is essential to guide the systematic incorporation of pharmacometric models into obstetric clinical trials and care pathways, ensuring that these tools are validated, optimized, and applied effectively. Without a clear roadmap for implementation, the integration of pharmacometric findings into routine obstetric care remains slow and inconsistent. Moreover, limited funding for obstetric research further exacerbates these challenges. Obstetric pharmacology often competes with other medical fields for research grants, and given the complex ethical and logistical issues surrounding studies in pregnant women, funding agencies may be hesitant to invest in this area. The scarcity of financial resources restricts the scope of pharmacometric research, limiting the development of comprehensive models that could otherwise advance personalized drug therapy in pregnancy.

Evolving regulatory frameworks and growing awareness of the need for inclusive research continue to provide opportunities for pharmacometric modeling to bridge these gaps by integrating real-world and population-specific data from observational studies. Models that account for bidirectional drug movement and fetal metabolism in the highly complex fetal–placental interface are challenging but critical for accurate predictions. A hybrid modeling approach that combines PopPK and PBPK models can leverage the strengths of both methods to address scalability challenges (Table 1). This approach would integrate the detailed physiological representations of PBPK models, which account for organ-specific drug distribution and metabolism, with the large-scale variability captured by PopPK models, allowing for more accurate predictions of drug exposure across diverse pregnant populations while maintaining mechanistic insights into drug behavior in both maternal and fetal compartments. The successful application of pharmacometrics in obstetrics necessitates a multidisciplinary approach, requiring collaboration between obstetricians, perinatologists, pharmacologists, bioinformaticians, and statisticians. By integrating clinical data, advanced modeling techniques, and ethical frameworks, these teams can overcome barriers to implementing pharmacometrics in obstetric research. Coordinating efforts across these diverse fields to create biologically plausible and clinically relevant models are complex but critical for advancing pharmacometric research in obstetrics.

The intersection of pharmacogenomics and pharmacometrics opens exciting new avenues for personalized medicine in obstetrics and maternal–fetal medicine (Figure 1). Genetic polymorphisms in drug-metabolizing enzymes (e.g., CYP enzymes), transporters (e.g., P-glycoprotein), and drug targets can significantly influence drug response during pregnancy. For example, integrating pharmacogenomic data, such as polymorphisms in the gene encoding UDP-glucuronosyltransferase 1A1 (UGT1A1), which metabolizes certain ART, can affect drug clearance rates in pregnant women. Incorporating such data into pharmacometric models allows for more tailored dosing regimens that consider both genetic variability and pregnancy-related physiological changes. By combining pharmacogenomic insights with pharmacometric modeling, drug dosing strategies can be refined to enhance efficacy and minimize adverse effects. Additional opportunities for improving pharmacometric models in pregnancy are outlined in Table 1.

The future of pharmacometrics in obstetrics is promising, with several exciting advancements on the horizon.³ Advances in machine learning and artificial intelligence are poised to revolutionize pregnancy pharmacometric modeling by enabling the integration of larger and more complex datasets. These technologies can enhance the predictive power of models, making them more accurate in simulating pregnancy-specific drug responses and optimizing therapeutic regimens. Furthermore, ongoing efforts to establish pregnancy-specific pharmacokinetic/pharmacodynamic databases, coupled with the increased inclusion of pregnant women in clinical trials, will significantly strengthen the data available for pharmacometric modeling. Regulatory agencies such as the U.S. Food and Drug Administration (FDA) and the European Medicines Agency (EMA) are now acknowledging the importance of pharmacometric approaches in obstetrics, further encouraging their application in drug development and safety.

In conclusion, pharmacometrics offers a transformative approach to addressing the unique pharmacological challenges in obstetrics and maternal–fetal medicine. By leveraging advanced modeling techniques, it enables the personalization of drug therapy for pregnant women, ensuring optimal outcomes for the mother–infant dyad. As the field continues to evolve, the integration of pharmacometrics into routine obstetrics research will be instrumental in advancing maternal–fetal medicine and improving the safety and efficacy of drug therapies in pregnancy.

Overall support for this work was provided by the Eunice Kennedy Shriver National Institute of Child Health and Human Development (NICHD) of the National Institutes of Health (NIH) under the Award Number 1K23HD104517 and DP1HD115433 (Ahizechukwu Eke).

The content is solely the responsibility of the authors and does not necessarily represent the official views of the NIH. For the remaining authors, none were declared.