Epidemiological studies have long suggested that women using injectable depot medroxyprogesterone acetate (DMPA) may be at increased risk of HIV acquisition. However, due to the inherent potential biases in observational study designs, it remains unknown if the observed associations in some, but not all, studies are a result of a true underlying increased biological risk, or a result of residual confounding due to differences in sexual behavior between users of different contraceptive methods. Due to the limitations of the observational data, the Evidence for Contraceptive Options and HIV Outcomes (ECHO) trial was initiated in 2015 - the first clinical trial to randomize HIV-negative women to DMPA, the copper T intrauterine device (IUD) and a levonorgestrel (LNG) implant to compare HIV acquisition rates. The trial goal is to provide high-quality evidence to inform women, service providers and normative bodies (e.g., the World Health Organization (WHO)) on the comparative HIV risks associated with these three highly effective contraceptive methods. Due to the nature of the contraceptive methods under evaluation in the ECHO trial, blinding was not possible. Thus, while randomization will provide balance of potential confounders at baseline, it is possible that participant risk behavior could differ by randomized method during follow-up, resulting in differential exposure to HIV, and biasing overall trial results. Such post-randomization sexual behavior changes may influence the ECHO trial intention-to-treat and/or causal analyses. Understanding these potential effects on the internal and external validity of trial results is imperative. Our overall goal is to evaluate objective biological markers of HIV risk by randomized contraceptive method within the ECHO trial to inform interpretation of the primary trial results. With support from ECHO trial leadership, we will test stored genital specimens from a subset of ECHO trial participants who provided consent for future research for prostate-specific antigen (PSA) and sexually transmitted infections (STI). These data will be combined with ECHO trial demographic, behavioral, laboratory and clinical data to accomplish the specific aims which are to: 1) determine whether women randomized to DMPA have more or less frequent condomless vaginal sex than women randomized to the IUD or implant during trial follow-up by comparing vaginal PSA, a marker of recent semen exposure, between randomized contraceptive groups across multiple ECHO sites, and secondarily, the prevalence of non-viral STIs; and 2) evaluate whether women randomized to DMPA are more or less likely to misreport condomless sex than women randomized to the IUD or implant during trial follow-up by evaluating the concordance of self-reported condom use or abstinence during trial follow-up with PSA detection overall, and by randomized contraceptive group, to inform use of self-reported data in planned trial causal modelling.
Epidemiological data have suggested an increased risk of HIV acquisition among depot medroxyprogesterone users for over two decades, and the Evidence for Contraceptive Options and HIV Outcomes (ECHO) trial is the pivotal randomized controlled trial to address this critical public health question. Modeling studies suggest that if injectable contraception was removed and replaced with an alternative contraceptive, countries with high injectable contraceptive use may experience substantial increases in the number of live births and in maternal mortality, whereas if DMPA is causally associated with HIV, it may be result in 27,000-130,000 new HIV cases per year. Our research will provide data ? objective markers of post-randomization sexual behavior change within the ECHO trial by randomized group - critical to the accurate interpretation of trial results, which are eagerly awaited by policymakers, providers and contraceptive users to minimize the risk of both unwanted reproductive outcomes and HIV among reproductive aged women worldwide.
