Oral preexposure prophylaxis use and the risk of bacterial sexually transmitted infections and HIV among African women: A prospective observational cohort study.

Abstract

Oral preexposure prophylaxis (PrEP) effectively reduces HIV incidence when used with sufficient adherence, but does not protect against bacterial sexually transmitted infections (STIs). Several studies have documented high rates of bacterial STIs among individuals initiating and using PrEP. We evaluated the association between PrEP use and the risk of STI among African women accessing family planning clinics. We conducted a prospective cohort study nested within a large pragmatic stepped-wedge cluster randomized trial of PrEP delivery in Kenyan family planning clinics, with participant enrollment from June 18, 2021, to May 18, 2023, and follow-up through February 02, 2024 (ClinicalTrials.gov: NCT04666792). The study population included sexually active HIV-negative women aged ≥15 years at elevated HIV risk per Kenyan PrEP guidelines. Participants were offered standard-of-care oral PrEP with the option to decline and followed quarterly for 12 months with assessments of HIV status, sexual behavior, and PrEP use. Urine samples were batch tested for Neisseria gonorrhoeae and Chlamydia trachomatis using the GeneXpert CT/NG real-time polymerase chain reaction nucleic acid amplification test assay. The primary exposure was self-reported PrEP initiation and PrEP use consistency through 6 months, categorized as never used PrEP, inconsistently on PrEP, or consistently on PrEP. Multivariable modified Poisson generalized estimating equation (GEE) models with robust standard errors were used to estimate associations between PrEP use and incident STI; clinic-level intracluster correlation coefficients were negligible. The secondary outcomes were incident HIV infection and sexual behaviors, which included condomless sex at last sex, sex with any new partners in the past 3 months, and multiple sex partners. HIV testing was performed at each scheduled visit, at enrollment, and 1, 3, 6, 9, and 12 months following the Kenya national HIV testing algorithm, using Determine HIV-1/2 and Fast Response test kits. All models for the primary outcome were adjusted for baseline covariates determined apriori as potential confounders, which included age, STI diagnosis at enrollment, any contraceptive use, number of sexual partners (categorized as any more than one sexual partner), education status, marital status, last partner HIV status, any transactional sex in 3 months pre-enrollment, and clinic site. Among 650 women enrolled, 60.0% (389) initiated PrEP at baseline and 14.6% (38/261) initiated post-enrollment. Median age was 26 years (IQR 23-30), 40% (262/650) were aged ≤24 years, and 67% (436/648) did not know their primary partner's HIV status. At baseline, 11% (74/650) had an STI, including 9.9% (23/232) of consistent PrEP users, 9.2% (13/141) of inconsistent users, and 14.0% (38/277) of women who declined PrEP. During follow-up, 19.1% (114/597) had at least one STI diagnosis, with similar risk among women who initiated PrEP at baseline compared with those who declined (19.2% [68/354] versus 18.9% [46/244]; aRR 1.11, 95% CI 0.76-1.61; p = 0.580). Compared with non-PrEP users (12.7% [25/195]), STI risk was 6.0% (12/200) among consistent PrEP users (aRR 0.56, 95% CI 0.27-1.19; p = 0.130) and 16.5% (17/103) among inconsistent users (aRR 1.43, 95% CI 0.73-2.77; p = 0.290). Chlamydia accounted for 87.7% (100/114) of STI diagnoses. STI risk was higher among women aged ≤24 years (aRR 1.47, 95% CI 1.04-2.07; p = 0.029) and those with a baseline STI (aRR 2.96, 95% CI 2.12-4.14; p < 0.001). Four HIV infections occurred over 594 person-years (incidence 0.67, 95% CI 0.18-1.72 per 100 person-years), including three among women who declined PrEP (incidence 1.25, 95% CI 0.26-3.66 per 100 person-years). The main study limitation was oral PrEP use was assessed based on client self-report and not objectively through drug levels testing. In this prospective cohort study among African women at elevated risk for HIV, 60% initiated PrEP at baseline and 14.6% (38/261) post-enrollment. PrEP use was not associated with increased risk for STI diagnosis through one year of follow-up. HIV incidence was low overall, consistent with expanded PrEP availability in similar populations.