Antiretroviral therapy regimens initiated during pregnancy and risks of adverse perinatal outcomes

OBJECTIVE: Antiretroviral therapy prevents mother-to-child HIV transmission during pregnancy. We estimated risks of adverse perinatal outcomes following postconception initiation of DTG/FTC/TAF, EFV/FTC/TDF, DTG/FTC/TDF, and DTG/3TC/TDF. DESIGN: Observational emulation of a target trial using population-based birth outcomes surveillance data from Botswana. METHODS: We emulated the IMPAACT 2010/VESTED trial protocol in Botswana's surveillance data and compared the risk of a composite outcome - stillbirth, preterm birth, or small-for-gestational-age birth - among individuals initiating DTG/FTC/TAF, EFV/FTC/TDF, or DTG/FTC/TDF at 14-28 weeks' gestation. After benchmarking results against the trial, we added DTG/3TC/TDF and evaluated individual outcomes (stillbirth, preterm birth, small-for-gestational-age birth, very preterm birth, very-small-for-gestational-age birth, low birthweight, high birthweight, neonatal death). RESULTS: Adjusted risk of the composite outcome was 30.3% (unadjusted n = 83/278) for DTG/FTC/TAF, compared with 34.3% for EFV/FTC/TDF (1378/4042; risk difference -4.1% [95% confidence interval (CI) -10.6 to 2.4] and 34% for DTG/FTC/TDF [741/2,173; -3.7% (-10.3 to 2.9)]. All three risks were slightly higher than estimated in the trial. DTG/3TC/TDF showed a similar adjusted risk [34.7%; 1,008/2,883; -4.8% (-11.1 to 1.6)]. Risks of individual outcomes were generally comparable across regimens. CONCLUSION: Our observational study findings for the composite outcome were comparable to the IMPAACT 2010/VESTED trial in a real-world programmatic setting and enabled evaluation of additional regimen and broader set of outcomes. In settings where randomized trial data are limited, well designed observational emulations of trials using nonrandomized data can provide informative evidence. As Botswana expands use of TAF in pregnancy, these results provide reassuring evidence supporting its safety.