TIMING OF ON-STUDY PREGNANCIES WITH A VAGINAL pH MODULATOR IN THE PHASE 3 AMPOWER TRIAL

Abstract

The Vaginal pH Modulator (VPM; Phexxi®) is a non-hormonal vaginal gel that maintains the pH of the vagina in the presence of semen in order to immobilize sperm and prevent pregnancy. In the phase 3 AMPOWER study, VPM resulted in a 7-cycle typical-use cumulative pregnancy percentage of 13.7%. The objective of this post-hoc analysis was to analyze timing of on-study pregnancies. AMPOWER was a multicenter, single-arm, 7-cycle study conducted in the US in healthy women age 18-35 years (NCT03243305). Women were directed to administer VPM intravaginally ≤1 hour before each act of vaginal intercourse. The primary endpoint was the 7-cycle cumulative pregnancy percentage in the typical-use population calculated using Kaplan–Meier methodology; cycles were only included in the efficacy analysis if they were 21-35 days in length, had at least one recorded act of vaginal intercourse, and had no recorded use of concomitant contraception. Over the course of the study, 101 pregnancies occurred in 1183 women out of 24,289 acts of vaginal intercourse. Approximately half (50% [51/101]) of on-study pregnancies occurred in the first 2 cycles; whereas less than one quarter (24% [24/101]) occurred in the last 3 cycles (Table). When baseline characteristics were compared between women who had pregnancies early (defined as Cycles 1-2) vs late (Cycles 3-7), African American women had fewer pregnancies in Cycles 1-2 (43% [9/21]) vs Cycles 3-7 (57% [12/21]), whereas White women had the same number of pregnancies in both groups (50% [37/74] for Cycles 1-2 and 3-7). Hispanic/Latino women had fewer pregnancies in Cycles 1-2 (36% [8/22]) vs Cycles 3-7 (64% [14/22]), whereas non-Hispanic/Latino women had slightly more pregnancies in early vs late cycles (Cycles 1-2, 54% [43/79]; Cycles 3-7, 46% [36/79]). Women who had ≥2 prior pregnancies were more likely to become pregnant later in the study (Cycles 1-2, 42% [22/53]; Cycles 3-7, 58% [31/53]) as compared to women who had 0-1 prior pregnancies who were more likely to become pregnant early in the study (Cycles 1-2, 60% [29/48]; Cycles 3-7, 40% [19/48]). Compliance rates were similar between the two pregnancy groups (Cycles 1-2, mean 88%; Cycles 3-7, mean 86%). The majority of on-study pregnancies occurred in the first 3 cycles of VPM use. Some baseline characteristics differed between women who had early vs late on-study pregnancies.