Pregnancy and contraceptive use among participants of childbearing potential in the HVTN 705 HIV vaccine trial in Southern Africa.

Background: HIV vaccine trial participants include sexually active cisgender females who agree to avoid pregnancy during the active vaccination period. Nevertheless, some pregnancies occur in almost all studies. We examined contraceptive use, pregnancy incidence, and the relationship between pregnancy and HIV seroconversion in one HIV vaccine trial.

Methods: We performed an exploratory analysis of data collected for HVTN 705/HPX2008, a phase IIb HIV vaccine trial enrolling cisgender women across 23 sites in five southern African countries. Baseline characteristics and contraceptive use were assessed among participants who became pregnant and those who did not during the active vaccination phase (months 0-15). Pregnancy incidence rates were calculated for this phase and the duration of follow up (36 months). Cox regression analysis was used to assess factors associated with incident pregnancy.

Results: There were 2,636 participants who received at least one vaccine or placebo dose (mean age: 23 years, standard deviation: 3 years). At enrolment, when contraception was required, 62.9% reported using injectable contraceptives. Overall pregnancy rate was 2.95 per 100 person-years (95% CI: 2.40, 3.58), with 101 pregnancies reported by month 15. Cumulative incidence of pregnancy at month 15 was similar between trial arms (log-rank p = 0.688). Each additional year of age was associated with an 8% decrease in pregnancy incidence (p = 0.014). Women aged 31-35 years had the lowest pregnancy incidence [1.75 (0.48, 4.48) per 100 person-years]. In a Cox regression analysis covering months 0-15, all contraceptive methods significantly reduced the incidence of pregnancy compared to no contraceptive use. Oral contraception was associated with the least reduction in pregnancy risk; implants were associated with the most reduction in pregnancy risk (p < 0.001).

Conclusions: In HVTN 705/HPX2008, higher incidence of pregnancy was associated with younger age and oral contraception (compared to other methods). These data may inform future designs of HIV prevention or vaccine trials.