797    

Background:  Transmission through breastfeeding is an important cause of infant HIV-1 infections in developing countries; however, its mechanism remains largely unknown. We have explored the association between cell-free virus (CFV) and cell-associated virus (CAV) levels in breast milk, as reflected by viral RNA and proviral DNA, respectively, and the risk of infant HIV-1 infection after 6 weeks postpartum.

Methods:  We matched 61 HIV⁺ mothers who transmitted HIV-1 by breast milk to 61 HIV⁺ non-transmitting mothers, based on infant’s age at sample collection. CFV and CAV were quantified in a single breast milk specimen per mother, preceding the infant’s first HIV⁺ result and close to the estimated time of transmission. The C2-C5 env locus was amplified from the first positive infant blood sample and from breast milk HIV-1 CFV and CAV. Phylogenetic analysis was used to identify any preferential clustering of the transmitted infant viral sequences with either CAV or CFV in maternal milk. The role of N-linked glycosylation of the C2-C5 env region for transmission of either CFV or CAV through breastfeeding was also examined.

Results:  CAV and CFV loads in breast milk were positively correlated (Spearman’s rho = 0.46, p < 0.00001). In a conditional logistic regression model adjusting for maternal CD4⁺ cell counts and HIV disease stage, each 10-fold increase in CFV or CAV load was significantly associated with about a 3-fold increase in breast milk transmission (OR = 3.24, CI 95% 1.55 to 6.76 and OR = 2.83, CI 95% 1.43 to 5.58, respectively). While CAV load was predictive of breast milk transmission both before and after 9 months postpartum (OR = 5.0, p = 0.03 and OR = 5.0, p = 0.01, respectively), CFV was a significant predictor of transmission occurring only after 9 months (OR = 2.75, p = 0.08 and OR = 5.33, p = 0.008, respectively). Phylogenetic analyses of the C2-C5 env region showed that 85% (11 of 13) of infants harboring viruses that clustered with CFV in their mother's milk were infected after 9 months postpartum. A significantly larger number of milk HIV-1 CFV sequences had low N-linked glycosylation density in cases with phylogenetic evidence of CFV transmission (11 of 13) compared with cases transmitting CAV (6 of 16), Fisher’s exact test p = 0.02.

Conclusions:  A reduction in milk CAV and CFV loads might significantly decrease HIV-1 transmission by breastfeeding. Transmission of CFV seems to be facilitated at later lactation stages and might be associated with lower envelope glycosylation.

Keywords: HIV-1; mother-to-child transmission; breastfeeding