In utero exposure to antiretroviral drugs and pregnancy outcomes: Analysis of the French ANRS pharmacovigilance database


Aims: In 2018, 1.07 million pregnant women received antiretroviral drugs, raising whether this affects pregnancy outcomes. We assessed the adverse pregnancy outcomes associated with prenatal antiretroviral drug exposure, notified to the French ANRS pharmacovigilance system.

Methods: An exhaustive case report series has been performed using the ANRS pharmacovigilance database. All ANRS-sponsored HIV clinical research studies using antiretroviral drugs either in pregnant women or women of childbearing age were eligible from 2004 to 2019. We analysed the following pregnancy outcomes: abortion, ectopic pregnancy, stillbirth, prematurity (<37 weeks of gestational age), low birth weight (<2500 g) and congenital abnormalities. A logistic regression was performed to assess the odds ratio (OR) for each outcome separately (if occurrence >50) compared to the outcome observed when exposed to non-nucleoside-reverse-transcriptase-inhibitor (NNRTI)-based regimen as the reference.

Results: Among the 34 studies selected, 918 deliveries occurred, of whom 88% had pregnancy outcomes documented. Pregnant women were mainly exposed to PI (n = 387, 48.6%), NNRTI (n = 331, 41.5%) and INI-based combinations (n = 40, 5.0%, 18 on dolutegravir). Compared to NNRTI-based combinations, there was no significant association observed with exposure to other antiretroviral combination for spontaneous abortion, prematurity or low birth weight, except an increased risk of low birth weight in new-born exposed to exclusive nucleoside-reverse-transcriptase-inhibitor (NRTI) combinations (n = 4; OR 7.50 [1.49-37.83]).

Conclusions: Our study, mainly based on protease inhibitor (PI) and NNRTI-based regimens, is overall reassuring on the risk of adverse pregnancy outcomes, except for NRTI which should be interpreted cautiously (small number, indication bias). In this study, the number of integrase inhibitor (INI)-based combinations was too low to draw any conclusions.