PrEP (pre-exposure prophylaxis) adherence among East African women

Maria Pyra | 2018

Advisor: Jared Baeten

Research Area(s): Global Health, Infectious Diseases


HIV incidence remains disproportionately high for women, particularly young women, in Sub-Saharan Africa; women are also at heightened risk while pregnant, which may account for a substantial portion of their adult lives. Although HIV pre-exposure prophylaxis (PrEP) in pill form is known to be efficacious for women, there remain unanswered questions about adherence in open-label and real-world settings, as well as regarding the effectiveness of PrEP during pregnancy. In the work presented in this dissertation, we first examined how women used PrEP in an open-label demonstration project and particularly, how adherence was related to HIV risk behaviors. Second, to better assess adherence, we evaluated the sensitivity and specificity of a biomarker among East African men and women using PrEP. Finally, we examined the effect of pregnancy on PrEP concentrations. First, we found that women in known serodiscordant relationships were able to take PrEP effectively; more than half took PrEP during their entire risk period, with ≥6 doses for most weeks when on PrEP. HIV incidence was reduced 93% (95% CI 77%-98%) for all women and 91% (95% CI 29%-99%) among women under 25 years old. In further analysis, we found evidence of four adherence trajectories and two risk behavior trajectories over the first six months of PrEP use. Women with a declining risk behavior trajectory were more likely to have a declining adherence trajectory, while women with steady risk were more likely to have high steady adherence; this supports the idea of prevention-effective adherence, which optimizes PrEP use. In the second aim, we found low sensitivities for the adherence biomarker tenofovir-diphosphate, using thresholds established in U.S. populations. Adherence counseling based on biomarkers should carefully consider the trade-offs between sensitivity and specificity. Finally, we found that concentrations of PrEP are significantly lower in pregnant women compared to non-pregnant women, as well as during pregnancy compared to pre-pregnancy, after adjusting for adherence. Additional pharmacology and epidemiology studies are needed to determine if PrEP dosing should be altered to sustain systemic levels of tenofovir during pregnancy.