Clinical trials and observational studies

This section provides an overview of the key considerations for the conduct of pregnancy safety studies and a list of materials and resources used in clinical trials and observational studies involving pregnant and breastfeeding people.

Introduction

This section focuses on studies of pregnancy safety (and sometimes efficacy) in larger numbers of women than are traditionally included in pregnancy pharmacokinetics (PK)-only studies, which generally provide preliminary and short-term safety data on small numbers of pregnant women. These studies may be clinical trials that enrol pregnant women (e.g. for drugs that may be used by large numbers of women who are or may become pregnant); clinical trials that enrol people who are not pregnant but that gather relevant pregnancy data in women who become pregnant on-study; and observational studies that enrol pregnant women (or women who become pregnant during follow-up). Examples of materials (protocols, informed consent forms, case report forms, and manuals of procedures) from each of these types of studies are provided in this section.

Ideally, for new drugs that are expected to be used by pregnant women, pregnancy PK, dosing and short-term safety data from at least a small number of women should be generated prior to drug licensure unless non-clinical or early clinical data suggest reproductive toxicity (see the PK section for more detail on pregnancy PK studies). In many cases, these data can be generated by gathering pregnancy PK data in consenting women who become pregnant during pre-licensure trials.

For drugs that will be used by adolescent and adult women, clinical studies evaluating the safety in pregnancy of a new drug should ideally be planned (or even underway) by the time of licensure or shortly thereafter, once pregnancy PK data are available to guide dosing in pregnancy, and as long as no reproductive or pregnancy toxicity concerns arise from non-clinical or early clinical data. Figure 1 shows a potential framework for the type and timing of pregnancy studies in relation to drug development stages (see also Abrams et al, 2021).

Figure 1. Framework for accelerated inclusion of pregnant women in pre-licensure clinical trials
(From: WHO, Approaches to enhance and accelerate study of new drugs for HIV and associated infections, 2021).

This graph shows a potential framework for the type and timing of pregnancy studies in relation to drug development stages
WHO
© Credits

Key considerations for conducting studies involving pregnant and breastfeeding women

Randomized controlled trials (RCTs)

Multiple factors make it challenging to fully avoid bias and measure/control for all confounders when comparing birth outcomes in an observational pregnancy study of a new agent to outcomes in a non-randomized control group. RCTs reduce selection, allocation, and other biases and unobserved confounding. While it is not possible to conduct pregnancy RCTs of all drugs, RCTs are ideal for evaluating the safety of high-priority drugs that are (or are likely to be) used by large numbers of pregnant women. 

Observational studies

It is not feasible (and in some instances not ethically acceptable) to perform a properly powered randomized clinical safety study in pregnant people, making observational studies and post-licensure surveillance important sources of pregnancy safety data. Observational studies (in particular, post-licensure studies or surveillance) are also generally the only way to fully evaluate both rare pregnancy safety outcomes and congenital anomalies (the first trimester is the period of greatest sensitivity to possible teratogens, but randomized trials during the first trimester are not usually feasible).

Background rates of birth outcomes can vary widely due to many factors, such as outcome definition and ascertainment method; the age, socio-economic and nutritional status of the relevant population; location; quality of local care accessible to the population; and even season of the year (see Outcome measures / Endpoint prevalence ranges). Variable gestational age at first drug exposure in comparator groups (and biases such as immortal time bias) can plague observational pregnancy analyses. In addition, predictors of birth outcomes are not always well known or comprehensively collected, resulting in the potential for unobserved confounding. It is hence important to:

  • compare pregnancy safety outcomes to the most relevant local comparator group possible;
  • optimize and harmonize outcome ascertainment methods and definitions (and consider these in comparative data analyses);
  • collect the highest-quality data possible for key exposure/confounder and outcome data (see Outcome measures / Endpoints and predictor definitions) (note: routine health data often have limitations such as incomplete or inaccurate data and varying outcome definitions and ascertainment methods);
  • account for gestational age of first drug exposure and potential immortal time bias in analyses; and
  • exercise caution when interpreting pregnancy safety data from observational studies.

Target trial emulation that uses observational data to emulate a hypothetical randomized controlled trial (RCT) is a promising approach to avoiding some of the biases that can affect observational studies, provided a large dataset that includes multiple key potential confounders is available (see Hernán et al, 2016). 

Efficacy

As a general principle, drugs that have established therapeutic or preventive efficacy in nonpregnant adults may be inferred to be efficacious in pregnancy, if drug exposure (PK) in pregnancy is equivalent or similar to levels in nonpregnant adults. In HIV, this is true for virologic treatment efficacy, vertical transmission prevention, and HIV prevention efficacy. However, drug exposure is frequently lower in pregnancy (particularly in the second and third trimesters), and the determination of whether pregnancy efficacy data are required will depend on multiple factors. 

Clinical safety endpoint (outcome) measures

Safety evaluations of drugs in pregnancy and lactation should focus on outcomes that are uniquely important to mothers and infants – for example pregnancy and birth outcomes, neonatal mortality, infant growth, maternal mortality, and other specific maternal adverse events that may differ between the pregnant (or postpartum) and nonpregnant state (such as gestational diabetes, eclampsia/pre-eclampsia, hepatotoxicity or maternal neuropsychiatric concerns).

It is critical to the field of pregnancy research to collect a common set of key endpoints, use standardized definitions of these endpoints, and ascertain endpoints with the greatest accuracy possible (e.g. use fetal ultrasound for pregnancy data where possible). See Outcome measures / Endpoint and predictor definitions for details about suggested endpoints of interest and their definitions.

Prospective observational follow-up for collection of pregnancy safety data where possible is desirable to enhance data quality and completeness (e.g. collection of key predictor/confounder data, gestational age data, and more accurate outcomes). In addition, whenever possible, fetal ultrasound (biometry) for dating (estimating gestational age) should be performed, as estimated gestational age is a component of several key pregnancy outcome definitions (preterm birth, small for gestational age, miscarriage and stillbirth). 

Sample size

See Outcome measures / sample size calculations for discussion and additional information regarding study sample size.

 

 

Resources

Clinical trials enrolling pregnant and/or breastfeeding women 

HIV TREATMENT

IMPAACT 2010 / VESTED

IMPAACT 2010 was a phase 3 open-label trial that studied the virologic efficacy and safety of three ART regimens started in pregnancy after the first trimester.

IMPAACT 2010 trial materials are included because this trial enrolled women during pregnancy (between 14- and 28-weeks’ gestation) and randomized participants to start one of three study drug regimens, and because the trial was designed and powered to evaluate both virologic efficacy and safety endpoints (birth/pregnancy outcomes, and maternal and child adverse events and other relevant clinical outcomes).

 

DolPHIN 2

Like IMPAACT 2010, the DolPHIN 2 trial enrolled pregnant women at or after 28 weeks’ gestation, randomized them to start different ART regimens (one of two regimens in DolPHIN 2), and evaluated pregnancy efficacy, PK and safety outcomes. This trial is included as another example of an RCT that enrolled pregnant women, and because it enrolled women late in pregnancy (gestational age ≥28 weeks) and included PK, efficacy and safety components.

 

IMPAACT 1077BF / PROMISE trial (breastfeeding study)

IMPAACT 1077 was a factorial RCT that randomized pregnant women with HIV to different antepartum (after the first trimester) and postpartum treatment strategies, and babies to different antiretroviral prophylaxis strategies. The trial enrolled both a breastfeeding cohort (“1077BF”) and a formula-feeding cohort (“1077FF”); materials from the former are included here.

IMPAACT 1077BF materials are included because this is a unique example of a very large pregnancy RCT with a factorial design (sequential randomization), which included both safety and efficacy endpoints (with vertical transmission as the primary efficacy endpoint).

 

HIV PREVENTION

IMPAACT 2009

IMPAACT 2009 studies the PK, feasibility, acceptability, and safety of oral emtricitabine + tenofovir disoproxil fumarate (FTC/TDF) pre-exposure prophylaxis (PrEP) for primary HIV prevention during pregnancy and postpartum in adolescents and young women and their infants.

IMPAACT 2009 trial materials are included because they provide an example of a protocol that combines both pregnancy PK and pregnancy safety studies in one protocol; includes a single-arm intervention; and studies HIV prevention (PrEP) rather than HIV treatment (ART).

 

MTN-042 / DELIVER

MTN-042 is a Phase 3b, open-label randomized trial designed to assess the safety, adherence, and acceptability profiles of two different HIV prevention/PrEP interventions during pregnancy: the dapivirine vaginal ring and FTC/TDF oral tablet. This trial is included because it takes a novel approach to enrolling women in sequential cohorts from later to earlier in pregnancy by estimated gestational age (EGA), with pauses between cohorts to review interim safety data. The first cohort enrolled women 36+0/7 weeks to 37+6/7 weeks EGA, the second cohort enrols women 30+0/7 weeks to 35+6/7 weeks EGA, and the third and final cohort enrols women 12+0/7 weeks to 29+6/7 weeks EGA. In addition, adverse pregnancy outcomes and complications are compared with rates of these outcomes (obtained through chart review of routine patient records) in communities in which the study is being conducted.


We acknowledge the Microbicide Trials Network for development of these materials.

 

MTN-043 / B-PROTECTED

MTN-043 is a Phase 3b, open-label randomized trial designed to assess the maternal and infant safety (and PK) of two different HIV prevention/PrEP interventions during breastfeeding: the dapivirine vaginal ring and FTC/TDF oral tablet. This trial is included as an example of a safety and acceptability study conducted during breastfeeding.


We acknowledge the Microbicide Trials Network for development of these materials.

 

HPTN-084 / LIFE

HPTN 084 is a double-blind placebo-controlled randomized trial that evaluated the safety and efficacy of long-acting injectable cabotegravir vs daily oral TDF/FTC PrEP to prevent HIV acquisition in women at high risk for HIV. HPTN 084 is included in these materials because the open-label extension includes a study examining the safety and PK of long-acting cabotegravir (CAB LA) in women who become pregnant (and during breastfeeding) and who consent to staying on study drug and on study, compared with participants originally randomized to TDF/FTC who stay on this regimen.

 

Observational studies, including implementation science

HIV TREATMENT

IMPAACT 2026

IMPAACT 2026 is a Phase IV, prospective “opportunistic” observational PK study of selected antiretroviral (ARV) and anti-tuberculosis (TB) drugs during pregnancy and postpartum.

This study is included because it provides an example of an intensive “opportunistic” study of pregnancy and breastmilk PK (with safety and efficacy data in small numbers of participants) among women who are taking a drug (or drug combination) of interest as part of routine clinical care. The protocol may be periodically updated to add new drugs to study. IMPAACT 2026 materials are also included in the PK section of the toolkit.

 

ILANA (Implementing Long-Acting Novel Antiretrovirals)

ILANA is a pragmatic real-world trial examining the implementation of long-acting injectable CAB + rilpivirine in clinics and decentralised community-based settings in the UK. Women becoming pregnant in this study can remain on the study drug following a risk-based discussion, and pregnancy outcomes are recorded.

 

HIV PREVENTION

CATALYST

Catalyzing Access to New Prevention Products to Stop HIV (CATALYST) is a product introduction study within the MOSAIC umbrella (Maximizing Options to Advance Informed Choice for HIV Prevention). The study will evaluate an enhanced HIV prevention service delivery package that includes three PrEP modalities, including oral PrEP, dapivirine ring, and injectable cabotegravir in Kenya, Lesotho, South Africa, Uganda and Zimbabwe. This study is included because it is an observational study that will collect pregnancy safety data on women who become pregnant on various forms of PrEP, using primarily routine health data related to pregnancy-related outcomes.

 

MATRIX-007: Safety evaluation following exposure to cabotegravir-, dapivirine-, and tenofovir-based PrEP during Pregnancy (CARE PrEP)

A prospective observational cohort of study participants in the CATALYST PrEP implementation study exposed to antiretroviral-based HIV prevention methods (long-acting cabotegravir, dapivirine ring, or tenofovir-based oral PrEP) during pregnancy and their infants, with 6 monthly follow-up throughout pregnancy and to 6 months postpartum for mother/infant. Primary objectives are to describe outcomes of pregnancies with exposure to one or more ARV-based PrEP and major birth defects in the infants.

 

Tshireletso

Tshireletso is a hybrid safety/implementation study aiming to evaluate whether CAB LA for HIV prevention (PrEP) is acceptable, feasible and safe in a post-partum breastfeeding population in Botswana. This study is included because it enrolls breastfeeding women and builds in breastfeeding PK and safety aims.

 

Supporting Oral and Long-acting PrEP Decision Making Among Pregnant Women in Lilongwe, Malawi

These pilot studies aim to evaluate the feasibility, acceptability, and appropriateness of a shared decision-making intervention to support personally appropriate decision making about pre-exposure prophylaxis (PrEP) use during pregnancy and breastfeeding. The study intervention, My Choice for HIV Prevention (MyChoice), is a counselor-delivered shared decision-making approach for pregnant women considering PrEP.

 

HEPATITIS TREATMENT

STORC

The ‘Safety, Tolerability, and Outcomes of Velpatasvir/SofosbuviR in Treatment of Chronic Hepatitis C Virus [HCV] during Pregnancy’ (STORC) study is a Phase IV, single-arm, open label cohort study of treatment for chronic HCV infection during pregnancy. This study provides an example of a clinical trial in pregnant women receiving treatment for HCV.

 

OTHER TREATMENT

Pharmacokinetics and Safety of Commonly Used Drugs in Lactating Women and Breastfed Infants (CUDDLE)

This study is conducted by the Pediatric Trials Network and aims to assess the safety of a number of commonly used off-patent medications, such as ciprofloxacin, doxycycline and amoxicillin, given to breastfeeding mothers. The study will include collection of breast milk and maternal and infant blood samples, to determine passage into breast milk and safe dosing.

Study website

 

 Additional resources

PrEPWatch

PrEPWatch provides a package of materials designed for use by policy makers, programme implementers, and healthcare workers to facilitate provision of PrEP to pregnant and breastfeeding people.

Disclaimer

The provision of study materials and links from the toolkit to other websites is provided for your convenience and does not indicate endorsement of those materials or sites by WHO. WHO accepts no responsibility for the validity or accuracy of their content. The mention of specific companies or of certain manufacturers' products does not imply that they are endorsed or recommended by WHO in preference to others of a similar nature that are not mentioned. Errors and omissions excepted, the names of proprietary products are distinguished by initial capital letters.