Any analysis of ZIKV vaccine development and the needs and interests of pregnant women must take account of the complex and rapidly evolving approach to maternal immunizations, the dangers of delaying accrual of an evidence base for biomedical interventions during pregnancy, and emerging consensus on ethical principles governing research with pregnant women.
Maternal immunization can offer significant benefits in a variety of ways.[ref] Some vaccines primarily serve to protect the pregnant woman from serious morbidity or mortality. This includes cases where pregnant women are one among many at-risk populations facing exposure to a virulent pathogen (e.g., yellow fever), as well as cases where they face higher morbidity and mortality than other population groups (e.g., influenza).[ref] In both instances, offspring also benefit. Preventing disease in a pregnant woman protects the fetus from harms of maternal illness and in utero exposures. Other maternal immunizations are aimed at preventing disease (e.g., pertussis) in newborns who are too young to receive vaccinations directly. Vaccinating pregnant women leads to conferred or passive immunity where maternal antibodies induced by the vaccine pass to the fetus. ZIKV vaccines will occupy a middle ground. Their primary purpose is to protect the next generation, but the target population is not exclusively pregnant women, and the vaccines will offer direct benefits to adults, such as protection against virus-related risks of Guillain-Barré Syndrome (GBS).
Despite the important role that maternal immunizations can play in preventing disease, there has historically been resistance to vaccinating women during pregnancy. Past public vaccination programs have tried to get around administering vaccines in pregnancy by focusing efforts on adolescents and children who are not yet of reproductive age or by interrupting viral circulation among young children to reduce exposure of pregnant woman, as in the case of rubella. Yet another strategy, called “cocooning”,[ref] focuses on immunizing non-pregnant adults within the household. There is accumulating evidence, however, that these strategies often fail to adequately protect pregnant women and their offspring.[ref] Paradoxically, efforts to protect them from risks of vaccination have had the unintended consequence of exposing pregnant women and their offspring to greater risks from infection and illness.[ref]
The critical importance of maternal immunizations is now increasingly recognized. In recent years, several vaccines developed for use by the general adult population have been endorsed for use in pregnancy. Multiple professional bodies recommend vaccination with inactivated influenza and tetanus, diphtheria, acellular pertussis (Tdap) for all pregnant women.[ref] The WHO now recommends the use of the yellow fever vaccine during pregnancy in outbreak contexts, even though it is a live attenuated vaccine with precautions issued for use in pregnancy.[ref] Other vaccines have been endorsed in pregnancy when there is a threat of exposure (e.g., hepatitis A and B, meningococcus, Japanese encephalitis) or as a post-exposure prophylaxis (e.g., anthrax, rabies, smallpox).[ref]
There has also been a push to develop vaccines specifically targeted to pregnant women. This new vanguard of vaccines is aimed primarily at preventing illness in the newborn through conferred immunity from the mother. Research and development efforts are targeting pathogens to which infants are highly vulnerable and that are prevalent among newborns, such as respiratory syncytial virus (RSV) and group B streptococcus.[ref] Because pregnant women are the only targets for these vaccines, the pathways to development and licensure necessarily include research with pregnant women and require the generation of evidence specific to their use in pregnancy.[ref]
The Evidence Gap for Pregnant Women
Most preventives and treatments developed for the general population lack evidence to guide decisions about their use in pregnancy. This problem has been particularly well characterized in the context of drug treatment in the US: data are insufficient to determine teratogenic risk for more than 98% of drugs approved by the US Food and Drug Administration (FDA since 2000, and 91% of drugs approved since 1980.[ref] For nearly three-quarters of drugs approved since 2000, there are no human pregnancy data whatsoever. Similarly, information to guide drug dosing is sorely lacking: more than 98% of pharmacokinetic studies done provide no data specific to use in pregnancy.[ref]
The dearth of evidence is due to many factors. One is the common practice of waiting to conduct reproductive toxicology, mutagenicity, and related studies until late in the R&D process when it is likely that the drug or biologic will proceed to licensure. This practice is an effective cost-management strategy but results in unintended downstream delays in understanding how the intervention works in pregnancy. Preclinical data are often critical to determinations of likely research-related risks and benefits of the intervention, required if pregnant women are to participate in clinical trials. These data also help to identify areas of potential concern or interest that should be pursued in research to further assess safety in pregnancy.[ref]
In large part, though, the lack of evidence to inform the use of preventives and treatments during pregnancy stems from a historical reticence to conduct interventional biomedical research with pregnant women. Helpful data on safety and efficacy can sometimes be gathered without involving pregnant women in interventional studies. However, there are many occasions when prospective enrollment of pregnant women, whether in small numbers to test pharmacokinetics or immunogenicity or in larger numbers to test safety and efficacy, can be critical to establishing an adequate evidence base.
Furthermore, the past practice in research oversight policies of categorizing pregnant women as “vulnerable” encouraged the view that the proper ethical stance towards research with pregnant women was exclusion, rather than careful and thoughtful inclusion. [ref] Other causes for this reticence include misinterpretations or overly cautious interpretations of what is allowed under research regulations and international norms, as well as concerns about legal liability.[ref] There are a range of cultural norms surrounding pregnancy and gender dynamics that complicate the involvement of pregnant women in research in various contexts. Pharmaceutical companies face disincentives relating to liability exposure, not only for trial-related risks but also post-approval liability that can be triggered if an indication is sought for use of an intervention in pregnancy.[ref] Finally, there are a number of risk distortions that have been noted with pregnancy, including, critically, the tendency to overweight the potential research-related risks to the fetus while ignoring the risks to the offspring of not allowing the pregnant woman into a study.[ref]
For all these reasons, pregnant women have been treated differently and, we have argued, unfairly in the development of new drugs and biologics.[ref] In contrast to other adults, little if any evidence about safety and efficacy for pregnant women is available at the time of licensure. It is only well after licensure that evidence is usually generated, typically from clinical experience or passive surveillance systems.[ref]
Reliance on registries and other passive post-marketing systems is problematic. Selection biases in passive surveillance favor reporting of negative outcomes, and reports of adverse events may be incomplete.[ref] Although these systems are designed only to surface safety signals requiring further investigation, not to draw scientific conclusions, signals are sometimes over-interpreted as definitive evidence that a drug or biologic causes an adverse outcome.[ref] Perhaps most critically, relying on passive systems can lead to long delays in safety determination. In the US, it is estimated that the mean time it takes to assign a pregnancy-specific risk level to drugs with undetermined risk at the time of FDA approval is 27 years.[ref]
An increasing number of organizations, including the WHO, Pan American Health Organization (PAHO), Council for International Organizations of Medical Sciences (CIOMS), American College of Obstetrics and Gynecology (ACOG), and National Institutes of Health (NIH) Office of Research on Women’s Health (ORWH), now recognize the importance, both scientifically and ethically, of involving pregnant women in research.[ref] They call for a shift in the presumption from exclusion to inclusion, while recognizing that research with pregnant women poses unique ethical complexities because of risks and potential benefits to future offspring who cannot consent for themselves. These organizations point out the analogy with and lessons from research with children: the need to include their distinct needs in the research agenda; the fact that there can be pathways to responsible inclusion; that access to trials involving the prospect of direct benefit can be important as a matter of justice; and the imperative to protect groups through research, not just from research.
Ethical Principles for Pregnant Women and Biomedical Research
As the importance of including pregnant women more adequately in the biomedical research agenda has solidified, four principles guiding research ethics for pregnancy have emerged as a growing consensus.
1. Pregnant women deserve an evidence base for the prevention and treatment of their illnesses equal to others as a matter of justice.
The foundational justification for this principle rests on the recognition that, because pregnant women are the moral equivalents of all other human beings and have equal moral standing, their interests and needs deserve to be treated fairly in the public investment in research. This principle has been reaffirmed in multiple international contexts, most recently by CIOMS in its explication of what just access to the benefits of research entails: “Equity in the distribution of the benefits of research requires that research not disproportionately focus on the health needs of a limited class of people, but instead aims to address diverse health needs across different classes or groups. … Since information about the management of diseases is considered a benefit to society, it is unjust to intentionally deprive specific groups of that benefit”.[ref] CIOMS explicitly includes pregnant women as such a group.[ref]
Just allocation of research investments to the health needs of pregnant women also accords with a core commitment of public health ethics to prioritize the needs of overly burdened and disadvantaged groups and to narrow unfair health disparities.[ref] Pregnancy often brings increased risk of illness and death and an often doubled health burden, to the woman and future child, especially in low- and middle-income countries and for poor women, globally.[ref]
2. Pregnant women should not be categorized as a “vulnerable population” for purposes of human subjects research review.
Until recently, pregnant women had been categorized as a “vulnerable population” for purposes of research regulations and guidance. This included, influentially, the US Federal Policy for the Protection of Human Subjects, which designated pregnant women as vulnerable alongside those whose capacity to make valid decisions about research participation is compromised, such as children and adults of limited cognitive ability.[ref] It was increasingly realized that such a designation was problematic, tacitly suggesting that pregnant women are incapable of offering valid consent.[ref] Further, the designation had unintended consequences of increasing health burdens: rather than safeguarding pregnant women and their future children from risk, it is now widely recognized that the categorization had the perverse result of adding risk to them by limiting the possibility of responsible research into their potentially distinctive health needs.
Both CIOMS and the Federal Policy for the Protection of Human Subjects have been recently updated to acknowledge that pregnancy itself does not make a woman “vulnerable” in the context of research participation. The revised 2016 CIOMS guidelines explicitly state that “pregnant women must not be considered vulnerable simply because they are pregnant,”[ref] and the recently adopted updates to the Federal Policy for the Protection of Human Subjects confirm “the final rule no longer includes pregnant women… as examples of populations that are potentially vulnerable to coercion or undue influence,” effective January 2018.[ref] While various factors can make specific pregnant women vulnerable, pregnant women as a group should not be characterized as a vulnerable population for purposes of human subjects research review.
3. It is ethically permissible to conduct research with pregnant women that meets specific risk standards.
Like any research involving human subjects, research with pregnant women must meet all standard research protections: risk must be the least needed for scientific purposes, for instance, and appropriate informed consent must be obtained before research proceeds. Because it involves implications for potential offspring, there is widespread agreement that responsible research with pregnant women also requires added levels of distinct oversight for it to proceed.[ref] Most centrally are specific standards of what research-related risk is acceptable, especially to the fetus and future child, who cannot consent to those risks.
There are two different standards, depending on whether the trial in question offers the prospect of direct benefit to participants or offspring (see box below).
Prospect vs. No Prospect of Direct Benefit
Trials involving the prospect of direct benefit—sometimes called “therapeutic research”—are those in which the study intervention may directly benefit the research participant. There is only a prospect of direct benefit, both because there is not yet confirmation of efficacy (that being one of the points of clinical research), and because, for trials with control arms, a given participant may not receive the experimental treatment being studied or an alternative intervention of proven benefit.
In contrast, studies with no prospect of direct benefit are those in which the possibility of benefit cannot reasonably be attributed. These studies include many early phase trials in which researchers have intentionally minimized the study intervention dose as a strategy to answer specific questions about safety, trials marked by too little evidence to reach a threshold of any reasonable prospect of benefit (even if benefits do accrue during the study), and studies whose focus is to better understand a point of biology rather than to test a potential preventive or therapeutic intervention. With studies that have no prospect of direct benefit, enrollment is purely for the value of advancing biomedical knowledge to the potential benefit of future populations and patients.
For trials that involve no prospect of direct benefit to either the woman or the future child, research-related risks to the future child are capped at a low risk threshold. In general, trials that do not carry any prospect of direct benefit to either the fetus or the pregnant woman can pose no more than “minimal risk” to the fetus, a standard commonly understood as comparing the probability and magnitude of anticipated harms with those ordinarily encountered in daily life or during the performance of routine physical or psychological examinations or tests.[ref] Exceptions are given for research involving particularly compelling needs for the population of pregnant women and their infants: CIOMS allows a “minor increase over minimal risk”[ref] and the Department of Health and Human Services (HHS) regulations carry a provision of increased risk under special HHS Secretarial review.[ref] While research involving no prospect of direct benefit to woman or future child can be important, it is not generally at issue in pregnancy and ZIKV vaccine research.
For trials offering the prospect of direct benefit to the pregnant woman, future child, or both, the standard of acceptable risk is importantly different. Rather than a specific threshold, acceptable risk is determined by the reasonability of the relation of research-related risks to the potential benefits offered by participation.[ref] The risk is justified by the anticipated benefits to the subjects. More specifically, the likelihood and importance of the potential benefits must be reasonably judged to outweigh the risks. These potential benefits must be at least as good as any available alternative preventive or therapeutic, as judged by a credible interpretation of available evidence, understanding that all such determinations will involve contexts of uncertainty.[ref]
There is no settled view about whether the prospect of benefit to the pregnant woman can justify an increment of research-related risk to the fetus. Important questions thus remain about how to proceed when interpreting acceptable fetal risk in research that carries the prospect of clinical benefit to the woman but none to the fetus. These questions are generally not relevant to ZIKV vaccine research, however, because both the pregnant woman, and especially her offspring, will experience the benefits if they materialize. In these kinds of cases, there is clear agreement that research that has a favorable potential risk-benefit balance (see box below) to the fetus can proceed so long as other protective regulatory standards are met.
Reasonable Judgments of Favorable Risk-Benefit Balance
Reasonable judgments of favorable risk-benefit balance entail credible interpretation of available evidence that the probability and magnitude of research-related risks is outweighed by the probability and magnitude of prospective benefit.
4. Justice requires that pregnant women have fair access to research that offers the prospect of direct benefit.
The distinction between research involving the prospect of direct benefit and those that do not is also key to understanding another implication of the demands of justice as a core principle of research ethics; the importance of fair access to participate in research involving the prospect of direct benefit.[ref] There is broad consensus that while biomedical research ethics includes the ethical imperative of protection from research harms and risks, it also includes the ethical imperative of fair opportunity to the benefits that participation in research can offer. Inclusion criteria for who is eligible for enrollment in research that offers a prospect of benefit must not unfairly exclude any group of persons or individual.
Fair opportunity to access the potential benefits of research participation stands as a critical ethical principle of justice that cannot be reduced to the scientific utility of a given population. Even in cases where it may not be scientifically necessary to include pregnant women in clinical trials to generate valid conclusions on the use of a product in pregnancy, pregnant women may still have compelling claims to participate in trials that offer the prospect of direct benefit to them or their offspring. This may be particularly true in the case of emerging infectious diseases and public health emergencies, when there are often few if any alternatives available for pregnant women to protect and preserve their health and that of their future offspring.
Fair access does not mean an automatic right to enrollment in all research involving the prospect of direct benefit. If a subpopulation does not meet the scientific eligibility requirements, or the risks of the trial are not in proportion to benefits for the group, then their exclusion is justified. Instead, fair access requires that a group must be judged eligible to participate so long as it meets general criteria of scientific relevance; that participation is otherwise allowable under applicable regulations and ethics guidance, including that there is a reasonable judgment of benefit favorable to risk; and that cost considerations do not suffice as a justification for exclusion.
Regulatory commentary and scholars in research ethics make clear that pregnant women are no exception to this principle.[ref] Pregnant women do not forfeit due consideration of how their health and interests could be advanced by participation in research simply because they are pregnant. More than that, in a great many cases, including ZIKV vaccine research, the benefits at stake with pregnant women’s inclusion are benefits that accrue to two entities, not just one: the woman herself, as well as her offspring. The greater the potential benefits at stake in participation, the more important it is not to exclude a class of persons who are otherwise eligible for inclusion.
Pregnant women are also entitled to treatment equal to other adults with regard to authorization of research participation. Fair access to research that offers a prospect of direct benefit requires that only the informed consent of the pregnant woman be solicited, and that her consent, alone, is sufficient to authorize research participation.
Learn more about the reasons to include pregnant women in research