Abstract
Objective
To investigate the influence of zidovudine (ZDV) prophylaxis duration in mothers and infants on the age at which infection becomes detectable by DNA polymerase chain reaction (PCR) in non-breastfed infants.
Methods
Blood samples were collected sequentially from birth to 6 months in a Thailand perinatal HIV prevention trial in which 98 transmissions occurred. The proportions of infections detectable at birth and the Turnbull distributions of age at which infection became detectable after birth were compared according to actual ZDV treatment duration (mothers: no more than 7.5 weeks versus more; infants: 3 days versus at least 4 weeks), provided an adherence greater than 75%.
Results
Detectable infection at birth was less frequent in children whose mothers received a long treatment as compared to a short treatment (27 vs 50%, P=0.04). When mothers received a long treatment, infant ZDV treatment duration did not influence the distribution of age at which infection became detectable after birth (median 24 days). However, when mothers received a short treatment, this distribution was shifted to the right when infants received a long treatment (median 43 days, P<0.0001), and to the left when infants received a short treatment (median 11 days, P<0.0001).
Conclusions
When mothers receive a short treatment, the proportion of infections detectable at birth is higher and the time at which infection becomes detectable after birth depends on the infant treatment duration. In the study conditions, a PCR result after 2 months could be used to define infection status.