780    

Background:  To optimize antiretroviral prevention strategies, determining the time in which mother-to-child HIV transmission (MTCT) occurs is of major importance. In the absence of breastfeeding, the timing of transmission must be deduced through a limited number of viral detection test results performed in the infant during the first weeks of life. We analyzed the influence of the duration of zidovudine (ZDV) prophylaxis on the timing of MTCT.

Methods:  Data were collected for PHPT-1, a clinical trial comparing the efficacy of different ZDV treatment durations in mothers and infants to prevent HIV-1 transmission. Mothers were given ZDV 300 mg twice daily starting at 28 or 35 weeks of pregnancy, followed by a 300-mg dose of ZDV every 3 hours (oral loading dose) from onset of labor until delivery. Infants were given ZDV 2 mg/kg every 6 hours for 3 days or 6 weeks. Mothers and infants were classified according to actual treatment duration (mothers: long > 7.5 weeks, short ≤ 7.5 weeks; infants: long ≥ 4 weeks, short 3 days, provided adherence > 75%). The studied sample consisted of 1321 mother-infant pairs and 93 infected infants. A hierarchical model in a Bayesian framework (Markov process for viral marker evolution, timing of transmission as a mixture of 3 distributions associated with early, late in utero and intrapartum transmissions) was developed to back calculate the time of transmission according to ZDV treatment duration. The decrease in transmission rate during a given period from 1 treatment condition to another, i.e. during the last 2 weeks before birth from mother short to long treatment duration, when infants received a short treatment, was estimated as avoided transmission rate.

Results:  When mother treatment was long as compared to short, 1 to 2% transmission occurring more than 2 weeks before birth were avoided, independently of infant’s treatment duration; 4 to 5% transmission occurring during the last 2 weeks of pregnancy were avoided by mother long treatment as compared to short treatment, when infants received a short treatment, whereas no effect of mother treatment duration was found when infants received a long one; 3 to 4% transmissions occurring during the last 3 days of pregnancy were avoided by infant long treatment as compared to short, but only when mother had received a short treatment.

Conclusions:  These results may help optimize perinatal HIV prevention strategies, especially in women whose ZDV treatment is initiated late in pregnancy.

Keywords: Mother-to-child HIV-1 transmission; Timing of transmission; ZDV treatment/prophylaxis