Tamoxifen for the treatment of unfavorable bleeding patterns in etonogestrel contraceptive implant users
Awarded 2014
Complex Family Planning Fellowship Research
Katharine Simmons, MD
Oregon Health and Science University

Nearly all of the 3 million unintended pregnancies in the United States each year result from inconsistent or non-use of contraception. Increasing use of the most effective methods of contraception will reduce unintended pregnancies and their social, medical and economic consequences. The contraceptive etonogestrel implant (ENG implant) is 20 times more effective at pregnancy prevention than oral contraceptive pills, but it has bleeding side effects that make it unappealing for some women. Tamoxifen, a selective estrogen receptor modulator (SERM), has previously been shown to dramatically reduce bleeding in users of an older levonorgestrel-based contraceptive implant (Norplanttm). If tamoxifen could stop bleeding in users of the ENG implant, it would give patients and physicians a valuable option for management of progestin-induced irregular bleeding.
Statement of purpose
The primary objective of this study was to determine whether tamoxifen taken by users of the ENG implant on an as-needed basis for frequent or prolonged bleeding reduces the number of bleeding days by at least 40% over 180 days, when compared to placebo.  Secondary objectives were to determine whether tamoxifen: 1) improves satisfaction and continuation of the ENG implant, and 2) compromises the contraceptive properties of the ENG implant by disruption of ovulation suppression.
This was a double-blind, placebo controlled trial conducted at Oregon Health and Science University, USA, between March 2014 and August 2015. A sample size of 56 was estimated to detect a difference of six days of bleeding/spotting per 30 days with 80% power. Participants were randomized to tamoxifen 10mg or placebo twice daily for seven days, to be taken at the onset of bleeding. A total of 56 ENG implant users aged 15-38 with frequent or prolonged bleeding/spotting enrolled. Subjects completed a daily text-message bleeding diary. They initiated study drug after three consecutive days of bleeding/spotting, and could repeat treatment as needed up to three times over 180 days of participation. Satisfaction was assessed by visual analog scale after each use of study drug. Ovulation was monitored by urinary metabolites of progesterone for 14 days after first initiation of drug.
Though only 34 subjects completed the full 180 days of study, short-term results were available for 51 of 56 subjects. Subjects randomized to tamoxifen reported more baseline bleeding/spotting than placebo (22.2 vs 17.3 days respectively, p=0.02). Tamoxifen did not reduce the proportion of bleeding/spotting days over 180 days compared to placebo (41.9% vs 36.2%, respectively, mean difference 5.7 days, 95% CI -7.8, 19.2). However, tamoxifen subjects reported significantly fewer days of bleeding/spotting (mean difference 5.0, 95% CI 0.1, 9.9), and more continuous bleeding-free days (mean difference 15.2, 95% CI 2.8, 27.5) than placebo during the 30 days after first use of study drug. No ovulation was detected in either group.
Consistent with one prior report in the levonorgestrel implant, this study provides evidence that a single course of tamoxifen may be an effective short-term treatment for ENG implant-related bleeding. These findings should be generalizable to ENG implant users with bothersome bleeding at any time during implant use.