Women who started antiretroviral therapy during pregnancy had an eightfold higher risk of mother-to-child transmission compared to women who became pregnant whilst already on ART, according to a four-year study of women in Johannesburg, South Africa, published in the advance online edition of Journal of Acquired Immune Deficiency Syndromes.
For those who began ART whilst pregnant each additional week of treatment reduced the odds of transmission by eight percent (95% CI: 0.87 to 0.99, p=0.02).
Access and treatment regimens for HIV-positive pregnant women determine mother-to-child transmission (MTCT) rates. In some parts of the world these range from 1 to 2%.
In South Africa, antenatal HIV prevalence is 29.3%. In the public sector, antiretroviral treatment for pregnant HIV-positive women has been provided at a CD4 cell count below 200 cells/mm3 when immunosuppression is already well advanced. (South African guidelines changed in December 2009 to promote earlier treatment, at a CD4 count below 350, for pregnant women.)
In this setting, HIV testing at antenatal clinics is a common entry point into care. This means that many HIV-positive women are only identified when they are already pregnant and at an advanced stage of immunosuppression. So they often then start treatment late in their pregnancy with resulting high rates of mother-to-child transmission (MTCT). Few data exist on the effects and length of treatment regimens on MTCT rates for this group.
Separate delivery of HIV care and antenatal care has made it especially difficult for women of reproductive age to get appropriate care, notably when they are HIV-positive.
Previously the authors had reported on an innovative way of delivering health care: the integration of antenatal and antiretroviral clinics at two hospitals in Johannesburg. The goal was to get pregnant women with advanced HIV disease onto treatment sooner, so improving maternal and infant health.
A combined prospective cohort from these sites for the period January 2004 to August 2008 was undertaken to look at the effects of type and length of treatment on MTCT rates.
The two ANC-ARV clinics at Charlotte Maxeke Johannesburg Academic Hospital and Rahima Moosa Mother and Child Hospital receive referrals from approximately 28 clinics in Johannesburg city and the surrounding suburbs.
Women followed at the clinics were pregnant, tested positive for HIV, and had a baseline CD4 cell count of less than 250 cells/mm3 or other indication for antiretroviral treatment.
Study eligibility criteria included women attending the clinic from January 2004 to August 2008, those who had previously been given single-dose nevirapine, those who became pregnant while on ART, and those who started ART during pregnancy. The cohorts were followed throughout pregnancy until four to six weeks after having given birth, when infant HIV testing was done. Once the infant’s status was determined, mother and child were then referred to HIV clinics in the community for long-term follow-up.
Of the 1142 women followed, data were available for 873 infants. Of the other pregnancies, either the mother and/or mother-infant pair was lost to follow-up or the pregnancy resulted in a stillbirth.
The median time on treatment at delivery for the 85% of women who started ART during their pregnancy was 10.7 weeks, compared to 93.4 weeks for those who became pregnant on ART. 43 of the 874 (4.9%) infants tested positive with no differences seen among treatment regimens.
The transmission rate in women who became pregnant while on ART was significantly lower than for those starting ART while pregnant (0.7% compared to 5.7%; p=0.01)
Transmission rates for this cohort with low CD4 cell counts are high, the authors note, especially when compared to other countries. For example, the authors cite a recent study in West Africa of women who started ART while pregnant with a comparable treatment duration (mean of 11.7 weeks) but who showed an MTCT rate of 2.3% at twelve months, half that of the South African cohort at four to six weeks.
The authors are uncertain of the reasons for the difference but suggest the shorter time on treatment and greater proportion of women with very low CD4 cell counts may explain the higher transmission rate in South Africa.
The authors note that women with twelve weeks of treatment or less had the same transmission rate of 7.4% as those who only received single-dose nevirapine. Yet, for women who delivered without any kind of treatment at all, the rate was more than twice as high (17.4%).
They add that length of time on treatment served as a surrogate for viral load at the time of delivery. As they note, maximal suppression of viral load takes place 10 to 16 weeks after the start of ART, and the level of viral load is a primary risk factor for MTCT.
Given the impracticalities of viral load testing in resource-limited settings they stress that finding other means to identify and categorise risk among pregnant women is key.
The authors highlight that their study lends further support to the findings that regardless of the setting, the earlier ART is started and the longer the mother is on treatment, the better the outcomes for the mother’s health and that of her child.
Yet, the authors note that, while in this setting integration of antenatal care and HIV services have been successful, women with HIV are still being identified late. So “further efforts are needed to address social and health service barriers that may contribute to late identification of HIV-infected women in South Africa.”
Hoffman RM et al. Effects of highly active antiretroviral therapy duration and regimen on risk for mother-to-child transmission of HIV in Johannesburg, South Africa. Journal of Acquired Immune Deficiency Syndromes, advance online edition, March 2010.