Maternal Response to Antiretroviral Therapy in Johannesburg, South Africa: Adherence and Drug Toxicities

South Africa has one of the highest HIV prevalences in the world, with women of reproductive age disproportionately affected by the epidemic. Access to lifesaving highly active antiretroviral therapy (HAART) is expanding in the region, and many HIV-positive women are experiencing pregnancy after initiating lifelong treatment with HAART. The benefits of continued treatment with HAART during pregnancy include prevention of mother-to-child transmission of HIV, as well as maximization of maternal health. Optimal effectiveness of HAART, however, is dependent on a complex set of factors, most of which have not been adequately described in women established on HAART prior to pregnancy. Using high quality observational data from a large clinical HIV cohort in Johannesburg, South Africa, and robust epidemiologic methods, including inverse probability weighted marginal structural models, we examined maternal responses to HAART among women established on treatment at the time of pregnancy. An optimal adherence indicator was derived from routinely collected antiretroviral drug refill data from nearly 9,000 adult HIV-positive men and women, and evaluated based on ability to predict virological failure among the non-adherent. In our cohort of 7,510 HIV-positive women on treatment, pregnancy was common after HAART initiation, with 896 women experiencing at least one pregnancy during follow-up. Risk of non-adherence was similar among non-pregnant and pregnant women (weighted Risk Ratio (RR): 0.95, 95% confidence interval (CI): 0.78 1.17), while women in the postpartum period, defined as six months after birth, experienced an increased risk of non-adherence compared to non-pregnant women (weighted RR: 1.46, 95% CI: 1.17, 1.82). Among the women in our cohort, we also observed few serious adverse events of renal impairment related to use of tenofovir, a widely used first-line agent in HAART regimens, regardless of pregnancy exposure. Despite limitations of our pregnancy exposure data, our findings were robust to sensitivity analyses. In general, our results suggest that for women established on treatment prior to conceiving, continuation of HAART through pregnancy does not seem to increase maternal risks in respect to adherence or renal toxicity related to tenofovir use.Doctor of Philosoph