BACKGROUND:
No randomized trial has directly compared the efficacy of prolonged infant antiretroviral prophylaxis versus maternal antiretroviral therapy (mART) for prevention of mother-to-child transmission throughout the breastfeeding period.

SETTING:
Fourteen sites in Sub-Saharan Africa and India.

METHODS:
A randomized, open-label strategy trial was conducted in HIV-1-infected women with CD4 counts ≥350 cells/mm (or ≥country-specific ART threshold if higher) and their breastfeeding HIV-1-uninfected newborns. Randomization at 6-14 days postpartum was to mART or infant nevirapine (iNVP) prophylaxis continued until 18 months after delivery or breastfeeding cessation, infant HIV-1 infection, or toxicity, whichever occurred first. The primary efficacy outcome was confirmed infant HIV-1 infection. Efficacy analyses included all randomized mother-infant pairs except those with infant HIV-1 infection at entry.

RESULTS:
Between June 2011 and October 2014, 2431 mother-infant pairs were enrolled; 97% of women were World Health Organization Clinical Stage I, median screening CD4 count 686 cells/mm. Median infant gestational age/birth weight was 39 weeks/2.9 kilograms. Seven of 1219 (0.57%) and 7 of 1211 (0.58%) analyzed infants in the mART and iNVP arms, respectively, were HIV-infected (hazard ratio 1.0, 96% repeated confidence interval 0.3-3.1); infant HIV-free survival was high (97.1%, mART and 97.7%, iNVP, at 24 months). There were no significant differences between arms in median time to breastfeeding cessation (16 months) or incidence of severe, life-threatening, or fatal adverse events for mothers or infants (14 and 42 per 100 person-years, respectively).

CONCLUSIONS:
Both mART and iNVP prophylaxis strategies were safe and associated with very low breastfeeding HIV-1 transmission and high infant HIV-1-free survival at 24 months.