Treatment decisions during pregnancy are a balancing act that should be informed by available data and expert input, notes an article in The New England Journal of Medicine. Shifting the default from exclusion of pregnant women to a careful consideration of benefits and risks can ensure that pregnant women and their foetuses receive the care they need.
More than 50 years ago, an epidemic of serious birth defects caused by pre-natal exposure to thalidomide shattered the prevailing notion that the placenta served as a barrier against damaging influences and led to recognition that exposures during pregnancy can result in harm to a developing foetus. Since that time, ensuring that a pregnant woman has access to potentially lifesaving treatments while safeguarding her foetus has become a delicate balancing act, one that requires a careful evaluation of risks and benefits to both the mother and her foetus.
Recent findings have again highlighted the need to carefully weigh both risks and benefits in developing recommendations for the treatment of pregnant women. In May of this year, the World Health Organisation, the US Food and Drug Administration (FDA), and other organisations released statements about a safety signal that suggested a possible link between peri-conceptional use of the antiretroviral medication dolutegravir and neural tube defects.
Most clinical trials of dolutegravir had excluded pregnant women, and dolutegravir treatment was discontinued in women who inadvertently became pregnant during trials, so there was a paucity of safety data on its use during pregnancy. The initial analysis of data from an observational study in Botswana did not reveal any cases of major birth defects among the 280 pregnancies in women with first-trimester dolutegravir exposure,1 but an updated interim analysis identified neural tube defects in 4 cases (approximately 0.9% of pregnancies) in which there had been peri-conceptional dolutegravir use, as compared with 0.1% of pregnancies with use of other antiretroviral medications.
Although this signal warrants further rapid evaluation, caution is needed before long-term changes in recommendations are made, given the small number and heterogeneity of the cases of neural tube defects identified and the potential implications for the care of reproductive-age and pregnant women.
Previous experiences with safety signals that led to changes in recommendations — some of which later failed to be confirmed — argue for the importance of a systematic approach. For example, concerns were raised about lamotrigine, a medication used for treatment of seizures and other neurologic disorders, when a pregnancy registry suggested a six-fold increase in the risk of cleft lip or cleft palate among infants born to mothers using the medication.
These findings led to an FDA alert to health professionals, noting the need for more information. Additional data failed to confirm this link, and in fact, lamotrigine now has a better safety profile than other anticonvulsants.
Similarly, an early signal, based on animal studies and case reports, suggested that the HIV medication efavirenz might be associated with an increased risk of neural tube defects. This initial signal led to recommendations from the FDA and the European Medicines Agency (EMA) to avoid first-trimester efavirenz exposure. Subsequent studies have not confirmed this finding,2 and efavirenz is now considered an alternative to dolutegravir. These examples emphasise the importance of a balanced response to preliminary signals.
Although other HIV treatment options are available, decisions about treatment of HIV-infected pregnant women are complicated, especially given the need to prevent perinatal transmission and treat maternal disease. Though the potential risk of teratogenicity is critical to these decisions, other factors must be considered and highlight the need for treatment options; such factors include the extent and timing of viral suppression, the risk of perinatal transmission, the likelihood of developing resistance, the presence of side effects that might affect treatment compliance, and other barriers to care, such as cost.
Because of the high frequency of unintended pregnancies in many settings, concerns about the safety of a drug during early pregnancy can have implications for access for all reproductive-age women. Though well-intentioned, resulting restrictions could raise serious issues of equity of care for women if the drug of concern is superior to alternatives.
For women to have access to potentially lifesaving treatments, a comprehensive evaluation of potential risks and benefits to pregnant women and their foetuses as well as consideration of the much larger population of women of reproductive age are needed before decisions regarding treatment are made, even when data on foetal risks are inadequate. Patients should be informed of both risks and benefits and should have a voice in final treatment decisions.
Unfortunately, situations in which there is insufficient information on which to base decisions about treatment of pregnant women are common. Because clinical trials examining treatment efficacy and safety nearly always exclude pregnant women, scant data are available to inform treatment options for many common conditions affecting reproductive-age and pregnant women, such as depression and asthma. A review of drugs approved by the FDA between 2000 and 2010 revealed that the teratogenic risk in human pregnancy was “undetermined” for more than 97% of these drugs.3
This dearth of data has led to public health dilemmas. For example, before the 2009 H1N1 influenza pandemic, limited information was available on the use of influenza medications in pregnant women. Data from human use were available for only 61 women who had been exposed to oseltamivir during pregnancy.4
However, after weighing the risks and benefits, including available data on morbidity and mortality among pregnant women during previous pandemics and seasonal influenza and on influenza medication use during pregnancy, the US Centres for Disease Control and Prevention recommended early treatment with oseltamivir for pregnant women.
Data from the 2009 H1N1 pandemic supported these recommendations: pregnant women with 2009 H1N1 influenza who were treated late (more than 4 days after symptom onset) were 50 times as likely to die and 7 times as likely to be admitted to an intensive care unit as women who were treated less than 2 days after symptom onset.5 Although the number of exposed pregnancies with data on outcomes remains low, no risks to the foetus related to the use of oseltamivir during pregnancy have been identified.
Urgent follow-up is needed to rapidly collect and assess additional data regarding first-trimester dolutegravir exposure in order to clarify potential risks, particularly given that this signal has already resulted in changes in recommendations for treatment of HIV-infected pregnant and reproductive-age women.
Including pregnant women in clinical trials and establishing post-marketing surveillance of pregnancy exposures can expand the evidence base to guide more informed clinical and public health decisions, not just for dolutegravir but for all medications.
Treatment decisions during pregnancy will continue to be a balancing act that should be informed by available data and expert input. However, shifting the default from exclusion of pregnant women to a careful consideration of benefits and risks can ensure that pregnant women and their foetuses receive the care they need.