Background:  Conflicting data have been published about mother-to-child transmission (MTCT) and if maternal characteristics and timing of transmission have distinct barriers to or mechanisms of transmission. The purpose of this study was to correlate subtype C env gene diversity in and between mothers and their infants with the presence and timing of HIV-1 transmission or the absence of transmission. 

Methods:  Blood samples were obtained from a prospective cohort study designed to determine the effect of malaria on MTCT. The maternal sample was taken at time of hospital admission during labor (all mothers received single-dose nevaparine [sdNVP] after the sample was taken but before delivery), and infant samples were collected at birth and at 6 and 12 weeks of age. Transmission was characterized as intrauterine if the infant sample was HIV RNA positive at birth, and as intrapartum if the infant sample was HIV RNA negative at birth and positive at 6 weeks. We amplified the highly variable V1/V2 region of the HIV env gene using plasma viral RNA as template for real-time polymerase chain reaction (RT-PCR) then examined diversity of the viral population using a heteroduplex tracking assay.

Results:  Analysis was completed on the samples from 36 non-transmitting mothers along with 11 intrauterine- and 21 intrapartum-transmitting mother/infant pairs. Infants had significantly fewer variants than the transmitting mothers (3.3 vs 6.5, respectively, p <0.0001), indicative of a bottleneck during transmission. The intrauterine-infected infants had fewer variants than the intrapartum-infected infants (2.21 vs 3.95, p <0.002); the most common number of variants in intrauterine -infected infants was 1, while infants infected intrapartum most commonly had 3 to 4 variants and none with only 1 variant. Finally, infants infected intrauterine had variants that were readily detected in the mother, while infants infected intrapartum often harbored variants not detected in the mother’s blood. Non-transmitting mothers had fewer variants than transmitting mothers (3.3 vs 6.4, p <0.0001). There was no correlation in transmitting mothers between timing and maternal CD4 count or HIV RNA load.

Conclusions:  Intrauterine and intrapartum appear to represent distinct mechanisms of transmission. Intrauterine transmission represents a greater bottleneck but involves transmission of the predominantly circulating virus. Intrapartum transmission represents less of a bottleneck but often involves the transmission of rare variants that are either compartmentalized in the mother or highly selected.