Pharmacologic Research in Pregnant Women — Time to Get It Right

Approximately 6 million women in the United States become pregnant every year, and 4 million of them will have a live birth. Many pregnant women (70 to 80%) take at least one prescription medication during the first trimester of pregnancy, and about 90% take at least one medication at any point during pregnancy.¹ Pregnant and postpartum women are often excluded from clinical drug trials, however, which leaves obstetrical care providers without adequate information regarding the safety, efficacy, and proper doses of many commonly used medications during these periods.

For more than a decade, several organizations, including the American College of Obstetricians and Gynecologists and the Society for Maternal-Fetal Medicine, have called on policymakers and researchers to address these gaps in knowledge and to facilitate the inclusion of pregnant women in clinical research. The 21st Century Cures Act established the Task Force on Research Specific to Pregnant Women and Lactating Women (PRGLAC), of which one of us is a nongovernmental member. The group held a series of meetings over the course of a year and reviewed detailed reports of research on medication use in pregnancy. This work culminated in a comprehensive report to the secretary of health and human services and Congress in September 2018.²

The report outlines strategies for identifying and addressing knowledge and research gaps regarding drug use during pregnancy and for improving the development of safe and effective therapies for pregnant and lactating women. It also includes a discussion of the practical and ethical issues surrounding the inclusion of pregnant and lactating women in clinical trials and recommendations for facilitating informed decision making by women and physicians about drug use during pregnancy and lactation.

In addition to these important topics, there are several issues the report does not address in great detail that we believe should be a focus of discussions about pharmacologic research in pregnant and lactating women. First, there is a need to develop clinical trial designs and trial sequencing strategies that are tailored to pregnant women and could generate data to support rational drug dosing guidelines in pregnancy. The PRGLAC report suggests studying therapeutic products that are used off label for treating pregnant and lactating women, using the 2002 Best Pharmaceuticals for Children Act as a model. This law, in combination with the 2003 Pediatric Research Equity Act, led to substantial improvement in the number and timely completion of studies in the pediatric population and subsequent FDA approval of drugs for pediatric use.³ A similar strategy could be used for prioritizing research on therapeutics for pregnancy indications or for use by pregnant women, although such medications would not have to go through a separate approval pathway, as they do for pediatric indications.

An area of particular importance is the inclusion of pregnant women in phase 2 and 3 trials. Two approaches for including pregnant women in drug trials have been proposed by Françoise Baylis and Scott Halperin.⁴ The first approach involves conducting stand-alone phase 1 trials (designed to study pharmacokinetics and safety) in pregnant women that are initiated at the same time as phase 3 trials in the general population. The second strategy involves conducting enhanced monitoring, including pharmacokinetics and safety evaluations — similar to what would be performed in stand-alone phase 1 trials — in pregnant women who enroll in late phase 2 or phase 3 trials.⁴ Baylis and Halperin argue that with the stand-alone, “staggered” approach, pregnant women and fetuses would not be exposed to any compounds that failed in general phase 1 and 2 trials, whereas the “embedding” approach would allow pregnant women to enroll in larger trials that include nonpregnant adults and still receive enhanced safety monitoring. An important consideration is that drug trials in pregnant women are subject to Department of Health and Human Services regulations that relate to research involving pregnant women, fetuses, and neonates, such as the requirement that nonclinical reproductive toxicity studies have not shown a risk of teratogenicity.

A third option is an opportunistic approach, which involves following women who become pregnant while taking medication as part of standard care or in the context of a clinical trial; safety and pharmacokinetics can then be assessed during pregnancy and post partum. An example of this approach is the inclusion of pregnant women in phase 2 and 3 trials (originally designed for nonpregnant women) examining the pharmacokinetics of a long-acting intramuscular form of cabotegravir for preexposure prophylaxis (PrEP) for HIV as part of the ongoing HPTN-084/IMPAACT-2026 trial (ClinicalTrials.gov number, NCT03164564). The advantages and disadvantages of these trial designs are described in the table.

Second, we believe that it will be important to address questions about the utility of efficacy trials as compared with pragmatic trials for assessing the use of pharmaceuticals during pregnancy. Pregnant women are different from nonpregnant women in many ways; pragmatic trials may therefore be preferable to efficacy trials when they are feasible, especially when pragmatism doesn’t compromise trial quality or the ability to answer clinical questions relevant to pregnant women. Pragmatic trials can include women with distinct medical and obstetrical conditions (such as multiple pregnancies) and various levels of adherence to medication, as well as women in various stages of pregnancy, thereby reflecting the real-life situations that clinicians typically encounter. These trials can be used to assess variability in pharmacokinetics, as well as drug effects, throughout pregnancy. Pragmatic approaches are increasingly used in clinical trials funded by the National Institutes of Health (NIH), which encourages investigators to incorporate questions from stakeholders, to include diverse, representative populations from heterogeneous settings, to study multiple outcomes that are important to guideline authors and policymakers, and to compare treatment options to real-world alternatives, rather than to placebo or no treatment. We believe that these issues deserve greater attention.

Third, developing an implementation-science framework for drug trials in pregnant women could help bridge the gap between trials that demonstrate the effectiveness of medications and the rational use of such drugs in clinical practice. Although implementation-science frameworks are rare in the obstetric-pharmacology literature, results have suggested that a multicomponent implementation-science strategy for pharmacist-led interventions can significantly increase the use of certain drugs.⁵ Experiences in pharmaceutical sciences have identified key challenges to the implementation-science process, including inadequate communication between researchers regarding the steps necessary to translate evidence into practice, limited knowledge of strategies for tackling obstacles to change at different levels, and the need for standardized pathways from evidence to practice. Implementing research findings in clinical practice is a multistep process, and diffusion, dissemination, and implementation strategies are needed to promote tangible change.

Finally, it will be beneficial to train more clinician-scientists in clinical pharmacology so that investigators can safely and efficiently design and conduct drug trials in pregnant women. To achieve these objectives, the NIH could expand its network of Obstetric-Fetal Pharmacology Research Centers (OPRCs) to include institutions that have established clinical pharmacology training programs. None of the three OPRCs in the United States currently has a clinical pharmacology training program that is accredited by the American Board of Clinical Pharmacology (ABCP). Marrying the infrastructure of the OPRC program with clinical pharmacology expertise could enable researchers to conduct safe, technically sophisticated, and complex studies that help clinicians and pregnant women make informed decisions about pharmacotherapies. Interested clinicians can also undergo fellowship training in these centers and become board certified by the ABCP.

The PRGLAC report and recommendations are a necessary step toward expanding research on medication use during pregnancy. We believe there is a tremendous opportunity to conduct well-designed pragmatic trials and expand the use and quality of implementation science in the field of obstetrical pharmacology. Clinicians, policymakers, and researchers can take steps to facilitate inclusion of pregnant women in clinical research to help answer important questions about the effects of medication use during pregnancy and the ways in which pregnancy alters pharmacokinetics and drug effects.

PERSPECTIVE.

Drug Trial Designs Suitable for Pregnant Women.
Trial Design	Advantages	Disadvantages
Staggered trial design — Conduct stand-alone phase 1 trials simultaneously with phase 3 trials in the general population.*	Trials will generate phase 1 outcome data in pregnant women; women in later stages of pregnancy can be enrolled before women in the first trimester.	Studies cannot begin until phase 2 trials in the general population are complete.
Embedded trial design — Embed phase 1 trials for pregnant women into late phase 2 or phase 3 trials.*	Researchers can provide pregnancy-specific data sooner than would be possible with stand-alone trials because the subgroup of pregnant women can be given priority in the analysis stage.	Studies are logistically challenging; the availability of data from the overall population can be delayed if pregnant women are enrolled at a slower rate than nonpregnant adults.
Opportunistic study design — Enroll women who become pregnant while taking a drug as part of their clinical care or as part of a study.	Investigators would already be familiar with the study protocol, since they would have participated in earlier research phases with nonpregnant women; start-up costs and monitoring requirements are lower than in other types of trials in pregnant women.	Enrollment is likely to be slower than in other types of trials in pregnant women.

^*The options are discussed in detail by Baylis and Halperin.⁴