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


Of a total of 10,553 pregnancies in HIV-1–infected women with an outcome ≥ 14 WG included in the EPF cohort between 2005 and 2015, 7079 were included in the more detailed component of the cohort (EFP CO01), which collects data from conception to delivery (Figure 1. Flow chart), one half of whom (N = 3574) were on antiretroviral therapy at conception.

Figure 1.

Flow chart. ART, antiretroviral therapy; VL, viral load (plasma HIV-1 RNA); WG, week gestation.

Overall, there was a treatment change before 14 WG because of safety concerns in 411 (23.2%) of the 1775 women who were on antiretroviral therapy at conception and had a viral load < 50 copies/mL in the first trimester and no documented intolerance.

Factors Associated With Changes

Among women with a viral load < 50 copies/mL in the first trimester who changed solely for safety concerns (N = 411), as expected, the incidence of ART switch was higher when their regimen was NR (not recommended) or A (alternative) compared with those receiving R (recommended) first-line treatment [48.7% (315/647), 9.9% (71/720), and 6.1% (25/408), respectively; P < 0.01]. Other factors associated with an ART switch were younger maternal age, increasing time between HIV diagnosis and pregnancy, being single versus married/cohabiting, unemployed versus employed/studying, birth in sub-Saharan Africa, primiparity, living in the Paris area, and a low CD4 count. Changing for safety concerns was not significantly associated with the timing of the first prenatal (booking) visit, the level of neonatal care in the institution, BMI, tobacco use or alcohol use, singleton versus twin pregnancy, mode of HIV acquisition, or assisted reproduction versus spontaneous conception (Table 1 and Table 2).

In the multivariate analysis, ART changes for safety concerns remained significantly associated with the type of treatment according to guidelines; adjusted ORs were 2.2 [1.3–3.7] for alternate ARTs and 23.1 [14.0–38.2] for NR ARTs, versus first-line therapy; P < 0.01 (Table 3). The geographical region, delivery period, marital status, time since HIV diagnosis, and number of previous pregnancies enrolled in the EPF cohort remained significantly associated with the probability of first-trimester treatment switching for safety concerns.

We established a propensity score for each patient from the final multivariate model to study association between ART change for safety concerns and virological and pregnancy outcomes (Table 3). The variables retained in the final model for the propensity score were the number of previous pregnancies in the cohort, geographical region, delivery period, marital status, and time since HIV diagnosis.

Impact of ART Switch on Viral Load and Pregnancy Outcomes

In women with viral suppression on ART at conception, changing ART in the first trimester for safety concerns was not associated with the time to virological rebound during pregnancy [Kaplan–Meier estimates: 19.3% in the switch group vs. 15.6%, HRa: 1.0 (0.7–1.4)], nor was it associated with the incidence of virological rebound, defined as a plasma viral load >50 copies/mL near delivery: 6.5% versus 4.6% ORa: 1.1 (0.6–2.0) (Table 4).

First-trimester ART changes were not associated with any adverse pregnancy outcomes (1.9% vs. 2.9%), mode of delivery, perinatal deaths, or HIV transmission to the child (Table 4).

The proportion of ART changes for any reason beyond 14 WG tended to be higher in women who had a first-trimester switch for safety concerns than in women who did not switch (23.1% vs. 14.0%), but the difference was not statistically significant (P = 0.4); the reasons for these changes did not differ between groups (P = 0.9). The reasons for subsequent changes in women with a first-trimester switch for safety concerns were intolerance in 30.6%, virological rebound in 25.8%, and other pregnancy-related concerns in 43.6%. The respective proportions for those who had no first-trimester switch for safety concerns were, respectively, 36%, 20%, and 44%.

In subgroup analyses, virological rebound was not associated with early change for safety concerns when the initial regimen was not recommended (see Table B, Supplemental Digital Content, When the initial regimen was considered as an alternative treatment, the proportion of virological rebound tended to be higher following changes for safety concerns [27.5% vs. 14.3%, HRa: 1.6 (0.9–2.9)], although the difference was not statistically significant (P = 0.2).

We performed the same analyses for women on efavirenz who had a first-trimester switch, in line with French guidelines throughout the study period. There was no significant difference in the proportion of VL > 50 copies/mL [16.8% in 143 pregnancies with a first-trimester switch from efavirenz vs. 15.6% in the overall group of 1364 pregnancies without a first-trimester switch, HRa: 1.1 (0.7–1.8), P = 0.7] (see Table C, Supplemental Digital Content,