(312) Sources and Characteristics of Quantitative Human Pregnancy Safety Data in U.S. FDA Approved Product Labeling, 2015-2021

Background: As part of ongoing U.S. FDA efforts to optimize pregnancy safety studies, insight is needed on factors that promote incorporation of postmarketing studies into product labeling.

Objectives: To examine sources and characteristics of quantitative human pregnancy safety data in U.S. product labeling.

Methods: This cross-sectional study evaluated “quantitative” human pregnancy safety data in product labeling. Pregnancy labelings approved or updated between July 2015 and December 2021 were reviewed for inclusion in the study, a timeframe after implementation of the “Pregnancy and Lactation Labeling Rule” (PLLR), which requires narrative explanations of risks and supporting data. Quantitative data was defined as measured numeric values describing drug exposures or study results. For each pregnancy study supporting these statements, a standardized extraction sheet captured study type (pregnancy exposure registry, database study with prespecified outcome, database study with no prespecified outcome, clinical trial, case report/series, or other), product therapeutic class, study objective, geographic location, publication status, and the presence of a post-marketing requirement (PMR) or commitment (PMC).

Results: From 2015 to 2021, the U.S. FDA approved 1,795 PLLR compliant product labelings. Most labelings (n=987) did not contain human safety data in the pregnancy subsection. After exclusion of labelings in which the human safety data were described as insufficient or classified as “qualitative” (n=318), 145 unique product labelings (supported by 177 unique studies) were identified for this analysis. The most frequently identified study type supporting quantitative safety data in PLLR-compliant labeling was pregnancy exposure registries (66 of 145, 46%), followed by database studies with a prespecified outcome (51 of 145, 35%), clinical trials (37 of 145, 26%), expert opinions, guidelines, or literature reviews (29 of 145, 20%), and case reports/case series (15 of 145, 10%). More than half of pregnancy exposure registries in labeling (38 of 51, 75%) were from the Antiretroviral Pregnancy Registry. Other common data sources included Scandinavian birth registers (N=21) and Medicaid (N=11).

Conclusions: For the majority of PLLR labelings from 2015-2021, no quantitative human pregnancy safety data were available. Gaining insight into drug safety in pregnancy remains a challenge. One of the PDUFA VII Pregnancy Safety Commitments is to develop a framework describing how to use data from postapproval pregnancy studies and timely incorporate that information into product labeling.