FDA Postmarketing Requirements for Pregnancy Registries, Lactation Studies, and PLLR Labeling in Recent Approvals

FDA postmarketing requirements for pregnancy registries, lactation studies, and PLLR labeling in recent approvals

FDA often attaches a similar package of maternal-health obligations to many new drugs and biologics that women of reproductive potential are expected to use. Approval letters from 2023 to 2026 show three recurring elements, two of them postmarketing requirements and one a labeling requirement:

1. A pregnancy exposure registry, paired with a complementary pregnancy outcomes study of a different design.
2. A lactation study, usually milk-only, that measures drug concentrations in breast milk.
3. Section 8 labeling written in Pregnancy and Lactation Labeling Rule (PLLR) format (a labeling requirement under 21 CFR 201.57, not a PMR), built to absorb the data those studies generate.

This article covers how FDA words these requirements, the reasons reviewers give for them, how the designs have changed, and how the results are meant to reach labeling.

Key takeaways

• Legal basis. These obligations are usually issued as FDAAA section 505(o)(3) postmarketing requirements (PMRs). Before imposing one, FDA must find that spontaneous adverse event reporting and its active surveillance system would not be enough ³⁴¹³⁴². Approval reviews document that finding for each product ⁴⁵⁶⁴⁵⁷.
• Two-study pregnancy design. The usual pregnancy package is a prospective registry plus a second study of a different design. The second study is typically a claims or EHR cohort, or a case-control study. FDA required this pairing for Zepbound, Wegovy, Omvoh, Bimzelx, Ebglyss, Journavx, Simtriyo, and other recent approvals ^898458629282.
• Rare diseases. For small populations, FDA substitutes a worldwide descriptive pregnancy (and sometimes lactation) exposure study for a comparative registry. Voyxact, Rystiggo, Ultomiris, Elfabrio, and Trutakna are examples ^171518725.
• Lactation studies. These PMRs now commonly apply to small molecules, antibiotics, peptides, and monoclonal antibodies. They specify a validated milk assay and, increasingly, measurement of major metabolites ^169171177.
• Labeling. PLLR labeling is mandatory for applications submitted on or after June 30, 2015 ³¹⁵. FDA continues to push older labels into PLLR format through labeling supplements ^481482484485.

Regulatory basis for pregnancy and lactation PMRs

Under section 505(o)(3), FDA may require a postmarketing study or trial at approval for any of three purposes: to assess a known serious risk, to assess a signal of serious risk, or to identify an unexpected serious risk when available data indicate the potential for one ³⁴¹. FDA can also impose a requirement after approval when it becomes aware of new safety information. That includes pregnancy safety studies addressing serious risk to the fetus or pregnant woman ³⁴¹. Before imposing a requirement, FDA must find that adverse event reporting and active risk-identification systems are insufficient. Before requiring a trial rather than an observational study, it must also find that studies would be insufficient ³⁴².

FDA's final May 2026 guidance on postapproval pregnancy safety studies makes clear that not every drug needs a pregnancy registry ³⁴². The guidance describes three approaches: case reports and series, prospective pregnancy registries, and complementary data sources. It presents these as contributing differently to the evidence, not as a hierarchy ³³⁹³⁴⁰.

For lactation, FDA's 2019 draft clinical lactation guidance states that FDA has required lactation studies under 505(o)(3) in some circumstances and is considering others. Sponsors may also run these studies voluntarily ³⁷⁷³⁶⁴. A separate draft guidance ties a lactation PMR to the same statutory triggers, such as a known or potential serious risk related to lactating women ³⁷⁸.

Pregnancy PMRs: the registry plus complementary study model

Why FDA requires two studies

The guidance explains the logic. Registries collect detailed prospective patient-level data and can include a comparator group, but enrollment problems mean they often cannot assess pregnancy safety on their own ³⁴³. Complementary observational studies address registry weaknesses such as low statistical power and selection bias. They typically use retrospective claims, EHR data, or population-based surveillance in cohort or case-control designs ³⁴³³⁴⁴.

The Zepbound (tirzepatide) review illustrates the product-specific reasoning:

• FDA expected the drug to be used by females of reproductive potential, including some exposed in the first trimester before they recognize a pregnancy. The chronic weight management indication was expected to increase unintended exposures ⁴⁵².
• Animal studies showed reduced fetal growth and malformations, while human pregnancy data were too limited to characterize risk ^453454455.
• FDA therefore required a registry and a complementary study, after finding that spontaneous reports and the active surveillance system would not be enough ⁴⁵⁶⁴⁵⁷.

Recent examples

Design trends in the PMR language

Common template. Recent registry PMRs share a standard set of terms:

• a preference for a disease-based multiproduct registry;
• alignment with ex-US protocols to reach the target sample size;
• capture of major and minor malformations, spontaneous abortion, stillbirth, preterm birth, and small-for-gestational-age births;
• infant follow-up through at least the first year of life.

Exxua, Velsipity, and Litfulo all use this template ^114162164.

Biologic-specific outcomes. For immunomodulatory biologics, FDA adds neonatal deaths and infant infections to the outcome set. Ebglyss, Omvoh, Bimzelx, and Tremfya are examples ^13339138121.

Comparator choices. For migraine products, FDA has required two unexposed control groups: women with migraine and women without migraine. This applied to Zavzpret and Symbravo ¹³¹¹¹⁷. Briumvi's registry also uses two unexposed control populations ⁴⁴⁴⁵. Cobenfy's registry compares women with schizophrenia who were exposed with unexposed women with schizophrenia ¹⁶¹.

Population scope. The registry population varies by product:

• Indication-agnostic ("regardless of indication"): Omvoh, Velsipity, Nereus, Simtriyo ^39162152123
• Disease-specific: alopecia areata for Litfulo and Leqselvi; ulcerative colitis for Tremfya ^164158120
• Indication-specific within a molecule: chronic weight management for Zepbound and Foundayo ¹⁴⁵¹⁰³

More weight on database studies. PMRs from 2025 and 2026 increasingly call for "adequately powered" retrospective claims/EHR cohorts "with medical chart validation." Examples include Journavx, Nereus, Simtriyo, and Orzeyful ^157153125110. For Orzeyful (oveporexton, Aug 2026), the retrospective cohort is listed as the first pregnancy PMR (5020-1) and the registry second (5020-2) ¹¹⁰¹¹³.

Long timelines. Registry final reports are generally due about 10 to 15 years or more after approval. Examples include Exxua (March 2036), Symbravo (September 2040), and Orzeyful's cohort study (April 2039) ^115118112.

New indications and existing registries

Pregnancy PMRs are also issued with efficacy supplements:

• Tremfya. The September 2024 ulcerative colitis supplement carried a new IBD-focused registry (4696-4) and a retrospective study covering guselkumab exposure regardless of indication (4696-5) ¹²⁰¹²¹.
• Bimzelx. For the November 2024 hidradenitis suppurativa supplement, FDA allowed the registry already required under PMR 4268-3 to be used to meet the new requirement (4704-2) ⁶⁹⁷⁰.

Rare diseases: worldwide descriptive studies

When a comparative registry is not feasible, FDA has required a "worldwide descriptive study" that collects prospective and retrospective pregnancy and/or lactation exposure data, with infant follow-up through at least one year. Examples include:

• Ultomiris (4252-2, 4449-1) ^64656667
• Rystiggo (4427-1) ⁴⁷⁴⁸⁴⁹
• Elfabrio (3972-2) ⁴²⁴³
• Voyxact (4907-4) ⁶¹⁶²⁶³
• Trutakna (4970-3) ⁷¹⁷²

The Voyxact review states the rationale directly. IgA nephropathy is uncommon and the exposed population would likely be small, so systematically collecting outcomes worldwide was preferable to a registry that would be hard to populate ⁵²¹. Fetal exposure was possible and no human pregnancy data were available ⁵²².

Lactation PMRs

Wording of the requirement

Recent lactation PMRs follow a standard formula. The sponsor must perform a lactation study in lactating women who have received the drug, measure concentrations in breast milk using a validated assay, and "assess the effects on the breastfed infant, if available, based on study population." The required design falls into one of three categories.

Some requirements go beyond the parent drug:

• Crenessity, Voyxact, Exdensur, Uplizna, and Nereus require measurement of "major metabolites" ^{196203211218205}.
• Utebzi (tebipenem pivoxil, June 2026) requires measurement of tebipenem, its major metabolites, and pivalate ¹⁹⁹.

Lactation milestones are much shorter than pregnancy milestones. Final reports typically fall three to five years after approval. Examples include Crenessity (June 2028), Exdensur (June 2029), and Utebzi (January 2030) ^197212200.

Reviewer rationale

Reviews repeat the same core reasoning. Use during lactation is foreseeable, but no data show how much drug reaches milk or what that exposure means for the infant. A milk-only study comes first, and a mother-infant study may follow if transfer is found ⁵¹⁰⁵¹¹.

Monoclonal antibodies. Reviewers note that endogenous IgG and monoclonal antibodies are present in human milk, so transfer is expected even without drug-specific data. FDA applied this reasoning to Ebglyss, Piasky, and Voyxact ^512515516. For Ebglyss, the milk-only PMR was based on "the lack of available lactation data and the anticipated use of Ebglyss in females of reproductive potential, including lactating women" ⁵¹². The Bimzelx review states that a mother-infant pair study "may be required" if transfer is found ⁵¹³⁵¹⁴.

Small molecules. For Myqorzo, findings in newborn rats and the drug's properties suggested milk transfer ⁵¹¹. For Wayrilz, concern about serious infections in infants added to the case for a milk-only study ⁵¹⁸.

Guidance expectations for study design

FDA's draft guidance sets out these design expectations:

• Default design. Milk-only studies are recommended unless there is a reason for another design. Collecting maternal plasma alongside milk adds pharmacokinetic information ³⁵⁷.
• Mother-infant pair studies. These are suggested when milk-transfer data already exist and infant absorption is likely ³⁷⁹.
• Sampling. Generally, the full milk volume from both breasts is collected over 24 hours, at steady state for chronic dosing ³⁵³³⁵⁴.
• Infant dose estimates. These use 150 mL/kg/day, with 200 mL/kg/day considered in early infancy ³⁸⁴.
• Newer approaches. The ICH E21 draft also describes maternal-only pump-and-discard studies and opportunistic enrollment of patients already taking a marketed drug ³⁸⁰³⁸¹.

PLLR labeling: where the data are meant to land

Required structure

Under the PLLR (effective June 30, 2015), pregnancy letter categories A, B, C, D, and X must be removed, and subsections 8.1 to 8.3 are restructured ²⁶⁸²⁶⁹:

• 8.1 Pregnancy. Headings: Pregnancy Exposure Registry (if applicable), Risk Summary (required), Clinical Considerations, and Data. Human data are presented before animal data ^257270271.
• Registry statement. If a scientifically acceptable registry exists, 8.1 must state: "There is a pregnancy exposure registry that monitors pregnancy outcomes in women exposed to (name of drug) during pregnancy." Enrollment contact information follows, with a cross-reference in Patient Counseling Information ²⁷².
• 8.2 Lactation. Uses a required Risk Summary. For systemically absorbed drugs it ends with the standard breastfeeding benefit-risk statement, unless breastfeeding is contraindicated ²⁷³²⁷⁴.
• 8.3 Females and Males of Reproductive Potential. Covers pregnancy testing, contraception, and infertility, as applicable. The subsection is omitted entirely when none applies ²⁷⁷.

Registry statements in current labels

Recent labels show the registry statement in use with several types of registry:

• Product-specific registries: Vyepti, Xeljanz ⁴⁴¹⁴⁴³
• MotherToBaby studies: Adbry ⁴⁴⁰
• Disease registries: the Gaucher registry for Cerezyme ⁴³⁹
• Class registries: the North American Antiepileptic Drug Pregnancy Registry for Lamictal ⁴⁴²

This matches the PMR preference for disease-based multiproduct registries.

How reviewers apply PLLR

Division of Pediatric and Maternal Health (DPMH) reviews revise 8.1 to 8.3 for PLLR compliance as a routine consult task ³¹⁴. The reviews show that 8.3 is included only when the data support it:

• Nuzolvence: reviewers recommended contraception for males with female partners for at least three months after the single dose, but did not recommend contraception for exposed females ⁴¹⁵⁴¹⁶.
• Sincalide and Rhapsido: reviewers recommended omitting 8.3 because no relevant fertility or embryo-fetal concern was identified ⁴¹⁸⁴¹⁹.
• Rhapsido 8.2: reviewers recommended the standard benefit-risk language despite the absence of milk-transfer data ⁴¹⁴.

Converting legacy labels

FDA continues to convert older labels to PLR/PLLR format, sometimes on its own initiative:

• Isovue: FDA requested a prior-approval supplement on May 17, 2023, and approved the conversion on December 6, 2024 ⁴⁸⁵⁴⁸⁶.
• Zyloprim: FDA proposed the conversion during review of unrelated supplements, and it was approved in September 2023 ^487488489.
• Other recent conversions: Adzenys XR-ODT (April 2025), Valium (September 2026), and Matulane (September 2026) ^484482483481.

Closing the loop: PMR fulfillment

FDA has recently released several long-running pregnancy registry PMRs:

The documents reviewed for this article did not show a recent lactation PMR whose results had been added to 8.2 ⁵⁵¹⁵⁵². Many of the lactation studies required in 2023 and 2024 have final reports due between 2026 and 2028 ^237217225. The next few years should show whether these requirements produce labeling changes.

Implications for sponsors

• Plan for both pregnancy studies. Expect a pregnancy PMR pair whenever a product is likely to be used by females of reproductive potential and human pregnancy data are limited. Plan both the registry and a database study with chart validation, and expect registry milestones to run about a decade.
• Line up registry participation early. FDA's preference for disease-based multiproduct registries and ex-US protocol alignment makes early participation in existing programs (for example, MotherToBaby or disease registries) a practical consideration ⁴⁴⁰.
• Rare diseases. For small populations, be prepared to justify a worldwide descriptive study instead of a comparative registry on feasibility grounds, as FDA did for Voyxact ⁵²¹.
• Lactation PMRs. Expect them frequently for systemically absorbed products without human milk data, including monoclonal antibodies. Build in metabolite assays and a possible mother-infant follow-up ^512516513.
• Draft Section 8 early. Write 8.1 to 8.3 in PLLR format from the start, include the registry statement once a registry is operational, and include 8.3 only when the data support it ^272277418.