(SLC-018) Enhancing Pregnancy Drug Safety Studies: Expert Consensus on Methods and Definitions

Background: Since the thalidomide catastrophe of the 1960s, the need for observational research to inform regulatory and clinical decision-making on drug safety during pregnancy has significantly increased. Numerous guidelines have been developed over the decades; however, a lack of consensus has complicated the comparability between studies due to differences in methodological approaches and definitions of key study variables. As the perinatal pharmacoepidemiology field is rapidly evolving, the ISPE MiPaL SIG established a working group to identify misalignment and gaps in existing guidance. The group established consensus on the definition of key variables and outlined challenges in the assessment of key endpoints for the generation of real-world evidence.

Objectives: This project aimed to define key challenges and considerations to provide recommendations for pregnancy anchors (i.e., etiologically relevant pregnancy milestones such as last menstrual period and trimester dates) and selected safety endpoints including congenital malformations and miscarriages.

Methods: A working group was led by members of the ISPE MiPAL SIG and consisted of representatives from academia, pharmaceutical companies, regulators, and contract research organizations. Experience in perinatal research ranged from mid-career to advanced career professionals. Existing guidance was collated and compared to identify consensus and gaps within existing guidance as well as drivers of misalignment. Findings were presented in an ISPE Annual Meeting 2024 workshop and participants voted and added comments through an online participation tool.

Results: Based on the degree of potential impact on study interpretation, the working group identified three areas of importance for alignment and identification of key safety endpoints; 1) pregnancy anchors including trimester boundaries, 2) considerations in estimating prevalence of congenital malformations including challenges integral to the malformation estimation process and, 3) misalignment in definitions of miscarriages and challenges with the identification and dating of miscarriages in existing data sources. For these three areas, specific recommendations for conducting pharmacoepidemiologic research were provided, focusing on the reduction of exposure and outcome misclassification in etiologically relevant time windows.

Conclusions: Definitions of pregnancy trimesters, malformations prevalence estimates, and miscarriages are three areas within perinatal pharmacoepidemiology with gaps in existing guidance and misalignment was identified. Agreed recommendations and rationale for these aspects of drug safety studies are provided to guide future research.