In discussions of mother-to-child transmission of HIV, the assumption is often made that the mothers themselves were already HIV-positive at or before the time of becoming pregnant. This assumption was challenged at a Tuesday morning session on mother and child health issues at the Sixteenth Conference on Retroviruses and Opportunistic Infections in Montreal. Presenter Lydia Lu of the Centers for Disease Control, on behalf of a US/Botswana research team, estimated that over 40% of infants who acquired HIV via mother-to-child transmission (MTCT) in 2007 may have been born to mothers who became HIV-infected during pregnancy or in their first post-partum year, when they were likely to be breastfeeding.
In Botswana, nearly all women who present at antenatal centres are tested for HIV only once, at a median of 22 weeks into their pregnancy. Prevention of mother-to-child transmission (PMTCT) programmes rarely re-test women after initial screening. This study therefore sought to find out how many initially HIV-negative pregnant women later contracted HIV during pregnancy or breastfeeding, and to estimate how many infants subsequently became infected as a result.
The study enrolled two groups of women who had tested HIV-negative at antenatal visits: 400 women on maternity wards, and another 244 women attending child immunisation clinics with infants aged 9 to 15 months. Of the 400 maternity ward patients who agreed to re-testing, five were found to be HIV-positive (1.3%; 95% confidence interval [CI], 0.5% – 3.1%). Their second HIV test took place a median of 17 weeks after the first. Seven of the 244 clinic attendees who agreed to re-testing were found to be HIV-positive (2.9%; 95% CI, 1.3% – 5.6%). The median time between tests for these women was 62 weeks.
The researchers then used these values to estimate the HIV incidence rate for mothers in their first post-partum year. Simply subtracting the 1.3% 17-week incidence rate from the 62-week rate of 2.9% yielded an estimated 1.6% incidence rate over 45 weeks, or 1.8% in the first post-partum year (52 weeks).
The next step was to estimate how many mother-to-child transmissions might be attributable to mothers who became HIV-infected during pregnancy or after delivery. Of 43,000 women who gave birth in Botswana in 2007, 13932 were diagnosed HIV-positive before or during antenatal care, and an estimated 620 infants became infected with HIV from these women (a transmission rate of 4.7%).
The estimated incidence rates found in this study would result in an additional 378 women acquiring HIV later during pregnancy, and 450 while breastfeeding during the first post-partum year. Since these women's HIV status would be presumed negative, the lack of prevention interventions would be expected to cause a disproportionately high number of new infant infections. Estimated MTCT rates of 73% and 36% (for infection during pregnancy and breastfeeding, respectively) would lead to an additional estimated 462 cases of MTCT in the 2007 calendar year.
Should all these estimates be accurate, then during 2007 in Botswana, there may actually have been approximately 1082 rather than 620 new infant infections, the additional 462 (43%) of which were due to women who themselves became HIV-infected only later during pregnancy or breastfeeding.
The researchers concluded that new and undetected infections in mothers in Botswana may result in nearly half of all infections in infants, and that this figure is "almost certainly" an underestimate. This represents an enormous opportunity for further testing and prevention interventions that are currently being missed. Their recommendations include routine re-testing for HIV-negative pregnant women and provision of "late PMTCT", and further studies of incidence to verify the estimates derived in this study. The best interventions for such late maternal HIV infections are also unknown, and need to be investigated.
Several points emerged during the question session, including the hazards of viewing pregnant women as not having sexual lives, the fact that women themselves may not feel the need for condom use if they are already pregnant, and the need to include women's HIV-positive male partners in prevention efforts. Lu noted that, while the uptake rate for couples testing was "much lower than [they] would like it to be", the re-testing of women done during this study did seem to encourage their partners to come forward for couples' testing.
Lu L et al. HIV incidence in pregnancy and the first post-partum year and implications for PMTCT programs, Francistown, Botswana, 2008. Sixteenth Conference on Retroviruses and Opportunistic Infections, Montreal, abstract 91, 2009.