Does Changing Antiretroviral Therapy in the First Trimester of Pregnancy for Safety Concerns Have an Impact on Viral Suppression?

Violaine Peyronnet, MD; Josiane Warszawski, MD, PhD; Jeanne Sibiude, MD, PhD; Olivia Dialla, MCS; Agnès Bourgeois-Moine, MD; Eida Bui, MD; Caroline Simon Toulza, MD; Delphine Peretti, MD; Cécile Brunet-Cartier, MD; Véronique Avettand-Fenoel, MD, PhD; Jérôme L. Chenadec, MD, PhD; Albert Faye, MD, PhD; Roland Tubiana, MD; Laurent Mandelbrot, MD; for the ANRS EPF-French Perinatal Cohort Study


J Acquir Immune Defic Syndr. 2019;80(5):574-584. 

In This Article


Our findings are reassuring regarding the decision to change antiretroviral therapy in early pregnancy to avoid exposure to medications with potential fetal risks. There was no increase in virological rebound above 50 copies/mL, compared with pregnancies in which the initial ART regimen was maintained. In all cases, the ART regimen could be changed again later in pregnancy in case of poor tolerance or inefficacy. Of concern, whatever the treatment group, the proportion of women who failed to maintain viral suppression throughout their pregnancy was on the average 15.6%. The proportion is, however, much lower than in some other high-resource settings.[32] The incidence of virological rebound in all groups was quite similar to that reported in the nonpregnancy literature.[33–35]

By contrast, Floridia et al[36] previously reported that women who had a treatment change in pregnancy were almost twice as likely to have a viral load higher than 400 copies/mL at the end of pregnancy compared to women with no such changes. However, the design of the study differed from ours because it included switches for toxicity or failure/resistance, patients who were not on ARV at the time of conception or with baseline viral loads above 400 copies/mL accounting for more than one-half of the population, as well as ARV switches in the second and third trimester. Nevertheless, as previously reported in the Italian cohort,[37] treatment interruptions during the first trimester, which were common at the time of the study in 2001–2008, were not associated with an increased incidence of detectable viral load before delivery.

In our cohort, the overall proportion of ART switches during pregnancy was 35%, of which more than one-half occurred in the first trimester, comparable with findings from a British study conducted in an earlier period.[38] The main reason for switching ART in the first trimester was to conform to the guidelines. Women were 23 times more likely to switch ART if it was not recommended, but decisions regarding ART regimens classified as alternative were less clear-cut, with a 2-fold increase in first-trimester ART switches, reflecting uncertainty about risks/benefits. In a small group of pregnancies, a first-line regimen was changed despite recorded efficacy and no reported intolerance; the reasons were not available. Efavirenz-based ART was singled out in French guidelines, as well as in US guidelines until recently, as requiring a first-trimester switch, whereas other guidelines recommended it throughout pregnancy.[22,25,39] We did not observe a higher proportion of viral load rebound in women who changed from efavirenz.

Nevertheless, nearly one-half of women with ART regimens that were not recommended at the time maintained their therapy in the first trimester. This may be due to previous resistance or tolerance issues, differences in the frequency and timing of clinic visits for HIV care, lack of knowledge regarding the latest pregnancy guidelines among some clinicians, and concern about jeopardizing treatment compliance. Treatment switches were less frequent for women who had a longer time since their HIV diagnosis, possibly in view of their treatment history limiting the therapeutic options. It is interesting to note that among women with NR regimens who did not switch in the first trimester, a high proportion ended up switching later in the pregnancy.

In our analysis, markers of social and personal vulnerability or deprivation were associated with changes in ART regimen later in pregnancy, which may reflect poorer engagement in care including family planning and treatment adaptation before conception.

This highlights the need for multidisciplinary care, particularly preconceptionally, to choose an ART regimen taking into account the desire for pregnancy with an optimal control of viral load, adherence, and tolerance, in view of the benefits/risks for the pregnancy as well as for the woman's long-term health and to avoid HIV transmission to the partner.[40]

The main strengths of this study are the large, multicenter enrollment and the prospective collection of detailed data, especially regarding the reasons for treatment change. Also, we carefully classified antiretroviral prescriptions and changes with regard to successive changes in guidelines. We chose to perform some of the survival analyses to account for the delay between first-trimester change and the possible occurrence of a complication or subsequent changes in ART because the gestational age may have an influence on the decisions as well as later outcomes. The observational design is a limitation of our study, which is only partially offset by use of a propensity score. The main limitation is the extent to which our results concerning French recommendations and practices are applicable in resource-limited settings, where close follow-up and viral load monitoring are not available.

In the decade we studied, the principal ART regimens that were discouraged in the French guidelines were efavirenz and triple NRTIs, as well as several antiretrovirals which are no longer in use, for instance, didanosine and stavudine. We defined a change in ART to include any addition, deletion, or change of at least one drug in the combination. Today, the main issue is whether to switch from single-tablet regimens without pregnancy safety data to regimens with several tablets or administered twice a day. Because efavirenz-based therapy was usually a single-tablet regimen, switching to a boosted PI-based regimen meant increasing the pill burden, and this did not lead to poorer virological outcomes. Some studies[41,42] have found that compliance is improved during pregnancy, which may suggest that women are willing to make special efforts to improve their chance of having a healthy child.

When recommending the avoidance of certain antiretrovirals, a fundamental issue is whether the potential risks justify this precaution. For most recent medications, there is simply not enough pregnancy data to conclude about their safety. Few antiretrovirals have documented risks for the fetus. Regarding efavirenz, preclinical primate data showed an increased risk of neural tube defects, and an increased incidence of malformations was reported after first-trimester exposure in some clinical studies including the French cohort.[15] This led the French expert panel to discourage efavirenz in the first trimester of pregnancy, whereas WHO guidelines[22] recommended efavirenz in pregnant women, based on reassuring data from other sources..[43]Although the safety of efavirenz now seems to be confirmed,[26] the recent alert concerning dolutegravir[26,27] is an illustration of the potential risk of treating pregnant women with medications that have not yet been evaluated for safety in pregnancy. In case of such controversy, approaches may differ between expert guidelines because of different analyses of the benefits and risks according to the populations, health systems, and resources.