Background:  Standard of care for prevention of maternal-to-child transmission (MTCT) of HIV is rapidly changing in resource-limited settings. In Mozambique, national policy provides potent ART for HIV-infected pregnant women with CD4 counts <350 cells/µL with continued ART post-partum, and many women breastfeed. Our objective was to describe HIV MTCT in the context of nevirapine (NVP) -based ART during late pregnancy and postpartum.

Methods:  This prospective observational study included 146 HIV-1-infected pregnant women at >16 weeks’ gestation with CD4 cell counts <350 cells/µL receiving prenatal care in Maputo, Mozambique, between August 2004 to June 2005, and their 149 infants.  Women started NVP-based ART in pregnancy and continued postpartum. Infant formula was available and mothers were free to choose breast or formula feeding. Infant HIV infection was defined by a positive HIV RNA polymerase chain reaction (PCR) from a single dried blood spot sample obtained after 2 months of age.

Results:  Of 149 infants, 17 (11%) died before HIV testing occurred and another 26 (17%) survived but were not tested.  Of 106 infants tested, 7 were HIV-infected (7%; 95%CI 4 to 10%). Maternal CD4 counts, ante-natal and post-partum ART duration and age at testing did not differ between infant groups (median and interquartile range presented). Maternal WHO HIV disease stage, weight, anemia, mode of delivery, infant birth weight, gestational age, and feeding strategy were also not predictive of infant HIV infection.

Conclusions:  The HIV MTCT rate, which is probably underestimated because of missing data and relatively early infant testing, appears to be unexpectedly high in this cohort of primarily breastfed infants whose mothers took potent ART in late pregnancy and post-partum. These data could reflect in utero HIV transmission. Alternatively, maternal post-partum ART may not be the optimal preventive MTCT strategy in the context of breastfeeding because of limited penetration of ART into breast milk or other factors.