Objective: To measure the release rate of prostaglandin E2 (PGE2) in vivo from a controlled-release vaginal insert used for cervical ripening and induction of labour at term in women with intact membranes or pre-labour rupture of membranes (PROM).
Design: Open-label, single centre study.
Population: Women at term (> or = 37 gestational weeks) with unripe cervices (Bishop score < or = 6) scheduled for labour induction for mainly medical reasons.
Methods: sixty-eight women (47 with intact membranes and 21 with PROM) had the PGE2 vaginal insert placed in the posterior fornix of the vagina. Each insert was removed from the women at a predetermined time interval between 0.5 h and 24 h, or earlier if labour was induced, fetal distress was detected or maternal complications occurred. After removal, the vaginal insert was frozen and stored for subsequent assay of residual PGE2. Blood samples were collected immediately before insertion and at 4-hour intervals until removal of the vaginal insert to determine plasma concentrations of PGE2 and the major PGE metabolite, 15-Keto-13, 14-dihydro-PGF2alpha (PGEm). Vaginal pH was measured immediately before insertion and directly after removal of the vaginal insert. Bishop score was assessed before induction, after 8 h, 12 h and immediately after removal of the vaginal insert.
Results: There was a positive linear relationship between the amount of PGE2 released from the insert and the duration of treatment in women with intact membranes (rp = 0.95, P = 0.0001), with a calculated PGE2 release rate of 0.52 +/- 0.33 mg/h over 24 h. The PGE2 release rate in women with PROM was not linear. The PGE2 release rate was dependent on vaginal pH, with a faster release rate at higher vaginal pH. Forty-seven women (69.1%) had the insert removed due to the successful induction of labour and consequently discontinued study treatment before their allocated time period. At vaginal delivery, the released amount of PGE2 at onset of labour was 4.0 +/- 3.0 mg and 2.4 +/- 2.1 mg for nulliparous women with PROM and intact membranes, respectively (P = 0.1). In multiparous women, the equivalent mean released amount was 3.2 +/- 2.6 mg and 1.9 +/- 1.4 mg, respectively (P = 0.14). In women with intact membranes, the mean plasma concentrations of PGE2 and PGEm after treatment were not statistically different to those women with PROM (P = 0.27 and 0.64, respectively). In women who were delivered vaginally, the median induction to delivery time interval was 17.0 h (range 4-42) in nulliparous women and 8.7 h (range 5-19) in multiparous women (P = 0.003). Ten (14.7%) women, who were all nulliparous, were delivered by caesarean section.
Conclusions: In women with intact membranes, the PGE2 release rate was linear over 24 hours. There was a positive linear relationship between vaginal pH and PGE2 release rate. The metabolite analysis revealed no evidence of dose dumping neither in women with intact membranes or in women with PROM.