LAUSR

Objectives To assess the effect of the depot medroxyprogesterone acetate injectable (DMPA) and of the levonorgestrel (LNG) implant on genital HIV shedding among women receiving antiretroviral therapy (ART). Methods We randomized HIV-infected Malawian women to either DMPA or LNG implant from May 2014 to April 2015. HIV RNA was measured in cervicovaginal lavage (CVL) fluid and TearFlo Strips (TFS), and HIV DNA was measured in cells collected by CVL. We compared the frequency and magnitude of HIV genital shedding before and for 6 months after initiation of contraception and between arms among women receiving ART. We also compared genital HIV RNA levels obtained by sample type (TFS versus CVL). Results We analyzed data for 68 HIV-infected women receiving ART: 33 randomized to DMPA and 35 randomized to the LNG implant. Overall, HIV RNA was more often detectable and the quantity was higher on TFS compared with CVL. HIV DNA was detected very rarely in CVL cell samples (4 of 360 samples). The frequency of genital shedding and the genital HIV quantity did not increase after contraceptive initiation with either DMPA or LNG implant among women receiving ART. Conclusions HIV-infected women receiving ART initiating contraception with either DMPA or LNG implant did not have any increase in genital HIV shedding during the first 6 months of contraceptive use. These findings are consistent with growing evidence that progestin contraception is not associated with increased HIV transmission risk from such women to their male partners. Consistent with other studies, genital HIV RNA detection was higher in TFS than in CVL fluid. Implications In this randomized trial, neither DMPA nor the LNG implant, two of the most commonly used hormonal contraceptives among African women with HIV, was associated with increased genital HIV shedding in HIV-infected women receiving ART. These findings are reassuring and add to the currently limited information available for the highly effective contraceptive, LNG implant.