Adherence thresholds for emtricitabine-tenofovir disoproxil fumarate preexposure prophylaxis against HIV acquisition in cisgender women: A randomized directly observed dosing study

Abstract

Background Oral emtricitabine/tenofovir disoproxil fumarate (F/TDF) preexposure prophylaxis (PrEP) effectiveness against HIV acquisition highly depends on adherence. For men who have sex with men, a dosing study in the United States (US) population defined clinically meaningful tenofovir diphosphate (TFV-DP) thresholds in dried blood spots (DBS) based on the rounded 25th percentile for 2, 4, and 7 doses/week as 350, 700, and 1,250 fmol/punch. However, divergent efficacy results in the first generation randomized clinical trials of F/TDF PrEP among African women led to several hypotheses to question whether the pharmacology and adherence requirement for oral F/TDF PrE may be different in cisgender women compared to what is already established for men. Methods and findings We conducted an open-label, parallel, randomized study of directly observed oral F/TDF PrEP among women without HIV who were not pregnant or breastfeeding in Kenya. Participants were randomly assigned to 2, 4, or 7 DOD doses/week for 8 weeks. Blood was collected weekly, and TFV-DP and emtricitabine triphosphate (FTC-TP) concentrations in DBS and peripheral blood mononuclear cells (PBMCs) were quantified using validated liquid chromatography-tandem mass spectrometry. For DBS, concentrations were quantified from a 3-mm punch using the 70% methanol/30% water (70:30) extraction method as the primary process—the same method used for the original TFV-DP benchmarks derived in US adults, and additionally with 50% methanol/50% water (50:50) extraction using punches from the same DBS spot to compare the extraction performance of 70:30 versus 50:50 methods. The primary outcome was the steady-state fitted concentrations of TFV-DP and dose proportionality in DBS and the observed PBMC TFV-DP levels by study dosing groups. Secondary outcomes included the quantitative concentrations of FTC-TP in DBS, TFV-DP half-life in DBS, and the relative TFV-DP recovery from DBS using the 70:30 versus 50:50 extraction method. One-compartment population pharmacokinetic models were fit to estimate steady-state DBS concentrations. Descriptive statistics were summarized as range, means, and medians with interquartile range (IQR) for continuous outcomes and proportions for categorical variables. Fifty-four women were enrolled and randomized. Median age was 22 (IQR, 20–25) years. The observed median (IQR) week 8 TFV-DP concentrations in DBS were 359 (266–464), 749 (596–923), and 1,389 (1,151–1,551) fmol/punch after 2, 4, and 7 doses/week, respectively. At week 8, FTC-TP was quantifiable in 71%, 19%, and 0% DBS samples for 7, 4, and 2 doses/week groups, respectively. Fitted median (IQR) steady-state DBS TFV-DP concentrations were 416 (316, 516), 832 (631, 1,033), and 1,457 (1,106, 1,808) fmol/ punch for 2, 4, and 7 doses/week, respectively, similar to previous estimates in US adults. TFV-DP exhibited a mean half-life of 17.5 days (95%CI: 16.7, 18.4) in DBS and steady-state TFV-DP concentrations varied in direct proportion to the dosing frequency [slope: 1.02 90% confidence interval 0.84, 1.20]. The 50:50 DBS extraction method yielded 1.27 (95% CI 1.25, 1.28) higher TFV-DP concentrations compared to the 70:30 method. When the 1.27 conversion factor was applied to the original 70:30 method-derived TFV-DP thresholds, the updated TFV-DP adherence interpretation benchmarks based on the 50:50 extraction were: <450 for <2 dose/week, 450–899 for 2–3 doses/week, 900–1,599 for 4–6 doses/week, and ≥1,600 fmol/punch for 7 doses/ week. The observed mean (standard deviation) steady-state PBMC TFV-DP concentrations was 11.99 ± 8.47, 31.81 ± 15.66, and 63.1 ± 28.97 fmol/106 cells after 2, 4, and 7 doses/week, respectively. Overall, oral F/TDF PrEP was well tolerated. No grade 3 or higher adverse events were observed during the dosing phase. The primary study limitation was dosing for 8 weeks, but population pharmacokinetic modeling enabled steady-state estimates. Conclusions Steady-state DBS TFV-DP concentrations from directly observed F/TDF PrEP dosing in African cisgender women participants are similar to previous estimates defined from US-based participants. These data demonstrate that cisgender women achieve similar DBS and PBMC TFV-DP concentrations as men for the same adherence level and validate the original TFV-DP benchmarks to interpret F/TDF adherence in HIV prevention studies and PrEP programs among cisgender women.