A Randomized Trial of Double Verses Single-Dose Etonogestrel Implant to Overcome the Interaction with Efavirenz-Based Antiretroviral Therapy

Concomitant use of efavirenz-based antiretroviral therapy and a standard-dose etonogestrel contraceptive implant resulted in 82% lower etonogestrel exposure compared to women not receiving antiretroviral therapy. The clinical impact of this reduced exposure is supported by retrospective cohort evaluations which demonstrated higher rates of unintended pregnancies when contraceptive implants were combined with efavirenz. We hypothesized that placement of two etonogestrel implants in those taking efavirenz-based antiretroviral therapy could increase etonogestrel exposure and improve measures of contraceptive efficacy. This study compared the rate of ovulation and etonogestrel pharmacokinetics in women on efavirenz-based antiretroviral therapy receiving two etonogestrel implants (136mg, DoublET group) versus one etonogestrel implant (68mg, Control group). This randomized, open-label study enrolled Ugandan women with regular menstrual periods who were receiving efavirenz-based antiretroviral therapy for treatment of HIV. Participants were randomized 1:1 to the DoublET or Control group, and the etonogestrel implant(s) were placed in the same arm at enrollment. All participants used a copper intrauterine device to prevent pregnancy. Ovulation was evaluated by weekly serum progesterone concentrations measured over 4 consecutive weeks at months 3 (weeks 9-12), 6 (weeks 21-24), and 12 (weeks 45-48). Progesterone concentrations >3ng/mL were interpreted as ovulation. The ovulation rate in each group was compared using Fisher's exact test by month and generalized estimating equations over 48 weeks. Plasma was collected at day 3 and weeks 1, 4, 12, 24, 36 and 48 after implant placement, and analyzed by a validated LC-MS/MS method for etonogestrel. Etonogestrel concentrations were summarized as median (interquartile range, IQR) and compared between groups by geometric mean ratio (GMR) with 90% Confidence Interval (CI). All participants (n=72) were cisgender Ugandan women, median age 31 (IQR 29, 36), and 36 participants were enrolled in each study group. Two participants in the Control group were discontinued; one at week 1 due to undetected pregnancy at entry and another at week 45 due to clinically significant depression. There were 47 ovulations over 104 person-months (45%) in 25 of 34 participants in the Control Group, and 2 ovulations over 108 person-months (2%) in 2 of 36 participants in the DoublET group [Month 3: 11 (31%) vs 0 (0%); Month 6: 17 (49%) vs 0 (0%); Mo