Background:  Antiretroviral pre-exposure prophyaxis (PrEP) is a promising new approach for preventing HIV infection. Fertility rates are high in many settings where HIV is prevalent. Thus, the effect of PrEP on pregnancy rates and outcomes is important for PrEP implementation.

Methods:  We conducted a randomized, 3-arm trial of oral PrEP among 4758 HIV serodiscordant heterosexual couples from Kenya and Uganda. Seronegative partners were randomly assigned to once-daily tenofovir (TDF), emtricitabine/tenofovir (FTC/TDF), or placebo and followed monthly for as long as 36 months, including monthly HIV-1 testing and, for HIV-uninfected women, monthly pregnancy testing. HIV-uninfected women were not pregnant at the time of study enrollment; if they became pregnant, the study medication was discontinued for the duration of pregnancy and lactation. For the 1785 (37.6%) couples in which the HIV-uninfected partner was female, we assessed pregnancy incidence and birth outcomes. On July 10, 2011, the study Data and Safety Monitoring Board recommended discontinuation of the placebo arm of the trial because of demonstration of PrEP efficacy for HIV-1 protection.

Results:  As of July 10, 2011, 288 pregnancies had occurred during the study (incidence 10.3/100 woman-years, 95%CI 9.1 to 11.5). Pregnancy incidence was not statistically different across the 3 study arms:  TDF (112 pregnancies, incidence 11.9/100 woman-years, p = 0.2 vs placebo), FTC/TDF (80 pregnancies, incidence 8.8/100 woman-years, p = 0.4 vs placebo),  placebo (96 pregnancies, incidence 10.0/100 woman-years); 178 pregnancies (61.8% of 288) had been completed as of July 10, 2011, including 90 live births (42 TDF, 23 FTC/TDF, 25 placebo); 88 pregnancies (30.6% of the total 288) had ended in spontaneous (n = 70) or induced losses (n = 18); 65 of 70 (92.9%) spontaneous losses occurred at <20 weeks’ gestation.

Conclusions:  Pregnancy rates were similar for women receiving PrEP vs placebo. Early pregnancy losses in PrEP clinical trials reflect monthly pregnancy testing and measurement of “chemical pregnancies” that would otherwise not be detected without frequent pregnancy testing. Follow-up of pregnancy and infant outcomes in this cohort is ongoing. Given the high efficacy for HIV protection observed in the Partners PrEP Study, PrEP may offer a method for HIV-uninfected women with HIV-infected partners to reduce HIV risk during conception, which warrants further evaluation.