Misoprostol, a synthetic analogue of prostaglandin E1, is approved and marketed in oral tablet form as a prophylactic treatment against chronic nonsteroidal anti-inflammatory drug–induced gastric ulcers. Administered orally or vaginally, off-label use of misoprostol has been shown to be an effective agent for cervical ripening and labor induction.1,2 There are, however, disadvantages of administering misoprostol tablets, because intravaginal administration of tablet fragments does not provide accurate dose strengths of misoprostol. There are also concerns over the potent uterotonic effect if the patient is inadvertently given higher doses, and this effect may be exacerbated by the bolus dosing regimen.3,4 Use of the oral tablet fragments also makes removal of residual intravaginal misoprostol cumbersome in the event of an adverse reaction. A well-controlled, removable mode of administration of misoprostol could, therefore, improve this compound’s safety profile.
Due to these safety concerns, Controlled Therapeutics (Scotland) Ltd., East Kilbride, Scotland, United Kingdom has developed a controlled-release misoprostol vaginal insert (Fig. 1). The vaginal insert is similar to the Controlled Therapeutics–marketed dinoprostone product (Cervidil/Propess), but contains misoprostol.
Misoprostol administered locally to the target tissues in a controlled-release manner may have key advantages over other agents used for cervical ripening and labor induction. A vaginal application will provide a constant supply of misoprostol with low maximal systemic levels, thus reducing the incidence of adverse events in other body systems. The vaginal insert’s retrieval tape allows easy, rapid removal of the product, immediately stopping delivery of additional drug. In addition, the number of patient examinations and manipulations are reduced due to a single placement over a 24-hour period.
This dose-ranging study investigated the controlled-release misoprostol vaginal insert in pregnant parous women at dose reservoir strengths of 25, 50, 100, and 200 μg. These dose reservoir strengths were designed to represent a range between the lowest Cytotec dose recommended by the American College of Obstetricians and Gynecologists, a single 25-μg tablet fragment intravaginally, and the highest dose, 50 μg every 6 hours.5 This study was performed with parous women due to expected lower variation in the time to vaginal delivery.
MATERIALS AND METHODS
This randomized, double-blind, dose-ranging, multicenter study was conducted at six sites in the United Kingdom. Ethics approval was obtained before study start from the multicenter research ethics committee and also each local research ethics committee. The UK Regulatory Agency (Medicines and Healthcare products Regulatory Agency) also granted approval.
The primary outcome measure was the time from insertion of the misoprostol vaginal insert to vaginal delivery of the neonate. This was assessed for 25-, 50-, 100-, and 200-μg dose reservoirs of misoprostol administered intravaginally by controlled release over 24 hours. Secondary objectives included additional measures of efficacy and safety. Assessments of safety included incidence of uterine hyperstimulation syndrome defined as fetal compromise in the presence of excessive uterine activity; nonreassuring fetal heart rate and other maternal, fetal, and neonatal adverse events. Additional efficacy measures included change in modified Bishop score at 12 hours, time to onset of active labor and the determination of the frequency, and amount of oxytocin use.
All women included in the study provided written informed consent. Eligible pregnant women were aged 18 years or older, 37 to 42 weeks of gestation, requiring cervical ripening and labor induction, and with at least one previous delivery with gestation of 37 weeks or more. The current pregnancy had to be a singleton fetus with cephalic presentation, and the woman was judged to have had an uncomplicated pregnancy before going beyond term. A modified Bishop score of 6 or less was required at baseline.
Due to the increased risk of uterine rupture,3 women were excluded who had four or more previous full-term deliveries and those with a cesarean delivery. Other exclusion criteria were: spontaneous labor; tocolytic agents within 7 days before induction; any other cervical-ripening or labor-inducing agent before enrollment; suspected cephalopelvic disproportion; evidence or suggestion of fetal distress; nonsteroidal anti-inflammatory drugs (including aspirin) within 4 hours of the study drug treatment; pyrexia; unexplained genital bleeding after 24 weeks of this pregnancy; pelvic inflammatory disease; placenta previa; known or suspected allergy to misoprostol or other prostaglandins; prior serious adverse event related to prostaglandin administered by any route for any indication, and women unable to comply with the requirements of the protocol. Women were free to withdraw from the study at any time and for any reason.
The nonbiodegradable misoprostol vaginal insert has three components: a hydrogel polymer base (measuring approximately 30 mm in length, 10 mm in width and 0.8 mm in thickness), an absorbable dose “reservoir” of misoprostol and a retrieval tape. The retrieval tape consists of inert woven nonwicking polyester into which the polymer insert is placed.
The polymer insert absorbs fluid from the vagina. As it hydrates and swells, the resulting concentration gradient leads to the continuous, controlled release of misoprostol over the dosing period, which can last up to 24 hours. Each insert was loaded with one of the four dose reservoirs of misoprostol being investigated: 25 μg, 50 μg, 100 μg, and 200 μg. The misoprostol vaginal insert is designed to release misoprostol at a rate of approximately 1, 2, 4, and 8 μg per hour from the 25-, 50-, 100-, and 200-μg dose reservoirs respectively.
Cardiotocographic monitoring was conducted for at least 15 minutes before study drug insertion to confirm a reassuring fetal heart rate pattern and absence of contractions. A digital examination was performed to assess the modified Bishop score6 and confirm eligibility immediately before insertion of the randomized misoprostol vaginal insert.
The misoprostol vaginal inserts were stored at room temperature (not exceeding 25°C). Women were stratified at each center by membrane status (intact or ruptured membranes) at the time of randomization. Women were randomly assigned to receive treatment according to a predetermined schedule generated by Controlled Therapeutics (Scotland) Ltd. The randomization provided for blocks of four containing one of each treatment group. Randomization codes were computer generated. Each site was provided with consecutively numbered treatment packs and provided with a list of all the treatment pack numbers. All members of the study team were unaware of which dose reservoir each woman received.
A single misoprostol vaginal insert was administered high into the posterior vaginal fornix and positioned transversely behind the cervix. A minimal quantity of water-soluble lubricant could be used to aid placement. The inserts were not prewetted or preswelled before insertion, and petroleum-based obstetric cream was not used, because this may have affected the release of drug from the polymer.
Cardiotocographic monitoring was conducted for at least 30 minutes after insertion of the misoprostol vaginal insert to ensure that there were no signs of fetal distress and thereafter according to the hospital protocol or clinical judgment. Vital signs (blood pressure, heart rate, temperature, and respiration rate) were recorded at scheduled intervals during the study. Modified Bishop scores were assessed at 6-hour intervals, at onset of active labor, and at removal of the study drug where applicable.
The misoprostol vaginal insert was to remain in place for up to 24 hours. The indications for removal were onset of active labor; evidence of maternal complications; evidence of fetal distress (nonreassuring fetal heart rate or fetal acidosis); oxytocin planned in the following 30 minutes; or if the woman was to undergo a cesarean delivery. Nonreassuring fetal heart rate pattern was defined as any fetal heart rate pattern that prompted clinical intervention, including one or more of the following: fetal bradycardia (baseline fetal heart rate less than 110 beats per minute [bpm]), fetal tachycardia (baseline fetal heart rate more than 160 bpm), unexplained absence of normal variability (6–25 bpm), recurrent late decelerations, moderate or severe variable decelerations, and other fetal heart patterns which prompted intervention.
Oxytocin was administered according to the local departmental protocol at each site. At least a 30-minute interval between removing the study drug and commencing oxytocin was required.
When indicated, the misoprostol vaginal insert was removed by pulling the retrieval tape. In addition to duration of insertion and reason for removal, the time of delivery, Apgar scores at 1 and 5 minutes, and mode of delivery (spontaneous vaginal, instrument or cesarean) were also recorded. Adverse events and concomitant medications were recorded throughout the study period. If delivery was not achieved within 24 hours, the misoprostol vaginal insert was removed and further management undertaken according to departmental protocol and clinical judgement.
Sample sizes were calculated using time-to-delivery data for parous women based on previous dinoprostone (Cervidil/Propess) clinical trials. Based on Cervidil studies conducted by the manufacturer, it was assumed that the difference in time to delivery between the lowest (25 μg) and highest (200 μg) dose of misoprostol would be similar to the differences observed between placebo and dinoprostone-treated patients. Sample size estimates were calculated to provide 80% power and an alpha of 5%, with an expected median time to delivery of 17.0 hours or less for the active treatments (50, 100 and 200 μg) and 27.7 hours for patients treated with 25-μg insert (similar to the placebo-treated patients in the previous dinoprostone trials). A sample size of 25 women per group was judged to be adequate to demonstrate a statistically significant difference for time to vaginal delivery.
Time to vaginal delivery is presented as a Kaplan-Meier plot stratified by dose and by baseline membrane status. Women delivering by cesarean were regarded as censored at the time of cesarean. Kaplan-Meier estimates of the median time to delivery for stratification level, along with 95% confidence intervals are presented (Table 1). Pair-wise comparisons of survivor function for the three higher-dose reservoirs were made with the survivor function of the lowest-dose reservoir (25 μg) using the log rank test. Time to delivery of the neonate was analyzed using a Cox proportional hazards model including the following covariates: treatment group, study center, parity, gestational age, baseline vaginal pH, baseline membrane status, and baseline modified Bishop score.
Although the original goal was to recruit 140 women, a lower than expected rate of cesarean delivery allowed recruitment to be halted before reaching this target. A total of 124 women were enrolled and received treatment: 33, 29, 32, and 30 women were randomly assigned to the 25-, 50-, 100-, and 200-μg misoprostol vaginal insert dose groups, respectively (Fig. 2). The first subject was enrolled in June 2003 and the last subject completed in March 2004. Each subject participated in the study for approximately 2 days.
The main reason for induction was postterm pregnancy (n=85, 69%). Baseline modified Bishop score (mean ± standard deviation) was 3.5±1.40. The treatment groups were well matched for baseline demographic and pregnancy characteristics (Table 2).
A total of 118 of the 124 women delivered vaginally (95%). Kaplan-Meier estimates of median times to vaginal delivery for each treatment group are shown in Figure 3 (times to cesarean deliveries are not included). Median time to vaginal delivery was significantly lower in the 100-μg (13.1 hour; P=.003) and 200-μg (10.6 hours; P<.001) dose reservoir groups compared with the 25-μg dose reservoir group (27.5 hours) (Table 1). Baseline membrane status did not affect time to vaginal delivery, although the number of patients with premature rupture of membranes at baseline was too low to make a definitive conclusion in this regard.
Successful cervical ripening was assessed in women with baseline modified Bishop score of 5 or less and was defined as achievement of one or more of the following endpoints by 12 hours: modified Bishop score of 6 or more; increase in modified Bishop score of 3 or more from baseline; vaginal delivery by 12 hours. The percentage of subjects with modified Bishop score of 5 or less at baseline who achieved the composite endpoint by 12 hours was 59%, 70%, 80%, and 86% for the 25-, 50-, 100-, and 200-μg dose groups.
The use of oxytocin was not standardized in this study; however, it was observed that low-dose regimens were used. The total amount of oxytocin administered ranged from 0.09–20 units per dose reservoir (median 2.61 units across all groups).
Cesarean deliveries were performed for 2 of 33, 1 of 29, 0 of 32, and 3 of 30 of women in the 25- to 200-μg groups, respectively. The reason for cesarean delivery included one each of obstructed labor and transverse presentation in the 25-μg group; one nonreassuring fetal heart rate pattern in the 50-μg group; two cases of cephalopelvic disproportion and one case of nonreassuring fetal heart rate pattern in the 200-μg group. There were no cesarean deliveries in the 100-μg group.
Adverse events are summarized in Table 3. Because it was not possible to determine whether an event occurring during labor could be strictly categorized as maternal or fetal, all maternal and fetal events were grouped as occurring to the “maternal–fetal unit.” Neonatal adverse events were categorized separately as “neonatal” adverse events.
A total of 78 women experienced 147 maternal–fetal events; 19 neonates experienced 23 adverse events. The most common maternal–fetal adverse event was nonreassuring fetal heart rate pattern, which occurred in 24%, 14%, 31%, and 27% in the 25-, 50-, 100-, and 200-μg groups, respectively. This event occurred in 12%, 7%, 13%, and 17% of women in the 25-, 50-, 100-, and 200-μg dose reservoir groups while the misoprostol vaginal insert was in place. Other common maternal–fetal adverse events included perineal laceration (9%, 28%, 22%, and 23%), postpartum hemorrhage (9%, 10%, 13%, and 10%) and vomiting (6%, 17%, 6%, and 13%). None of these events appeared to be dose-related. Uterine tachysystole (more than 5 contractions within 10 minutes) occurred in 3%, 0%, 6%, and 13% of the respective dose groups. There were no reports of uterine hypertonus (single contraction of more than 2 minutes duration). Uterine hyperstimulation syndrome, the combination of nonreassuring fetal heart rate pattern and uterine hyperstimulation (tachysystole or hypertonus) occurred in 3%, 0%, 6%, and 10% of women in the 25-, 50-, 100-, and 200-μg groups, respectively.
Fourteen women experienced 20 adverse events that led to early removal of the misoprostol vaginal insert (Table 4). Uterine hyperstimulation syndrome was the reason for removal in 3% of women (4/124). The remaining two women with uterine hyperstimulation syndrome experienced this event after removal of the misoprostol vaginal insert. One case of nonreassuring fetal heart rate pattern in the 200-μg dose reservoir group, which occurred while the insert was in situ, resulted in cesarean delivery.
With one exception, all neonates had a 5-minute Apgar score above 6. One neonate had a 5-minute Apgar score of 4 at 5 minutes. The mother of this neonate had received the 25-μg dose reservoir, which was removed 7 hours and 42 minutes before the onset of fetal bradycardia and meconium-stained amniotic fluid. The neonate required cardiac massage upon delivery and was intubated for 15 minutes. After an overnight stay in the neonatal intensive care unit, the neonate recovered fully, was returned to the mother, and went home on full feeds.
The results show that misoprostol vaginal inserts at reservoir doses of 100 and 200 μg resulted in a faster time to vaginal delivery compared with the 25-μg dose reservoir. Significantly more women delivered vaginally by 12 hours in the 100- and 200-μg dose reservoir groups compared with the 25-μg dose reservoir group, particularly in the subgroup of women with a baseline modified Bishop score of 3 or less (44% and 64%, respectively). Median time to delivery was also reduced in the 50-μg dose reservoir group, but this difference was not statistically significant (P=.07).
The 25-μg dose reservoir seemed to be similar to placebo inserts used in the clinical trials of dinoprostone, with a median time to vaginal delivery of 27.5 hours with the 25-μg misoprostol vaginal insert compared with 27.7 hours in the placebo arm of dinoprostone trials. A consistent dose response was observed for all efficacy parameters studied, with the higher-dose reservoirs (100- and 200-μg groups) resulting in statistically significant reductions in time to onset of active labor. Fewer women required oxytocin predelivery, and the total mean dose was lower in the higher-dose groups compared with the 25-μg dose group (P=.028), suggesting a potent uterotonic effect that obviates the need for oxytocin. The efficacy of the misoprostol vaginal insert plateaued between the 100- and 200-μg reservoir doses, as evidenced by minimal decrease in time to delivery or time to onset of active labor between these two groups.
There was a dose-dependent increase in uterine hyperstimulation syndrome, with 3%, 0%, 6%, and 10% of women in the 25-, 50-, 100-, and 200-μg groups, respectively, experiencing hyperstimulation. There was no evidence of a dose relationship between nonreassuring fetal heart rate patterns and dose reservoir of misoprostol. Nonreassuring fetal heart rate patterns were noted in 24% of women from the 25-μg dose group and 27% of those from the 200-μg dose group.
It is interesting to compare the safety results of the current study with those from a meta-analysis performed by Sanchez-Ramos et al.4 Although meta-analysis results must be viewed with caution due to the varying misoprostol and oxytocin regimens used in these studies, the safety profile of the misoprostol vaginal insert observed in the current study is encouraging by comparison. Uterine tachysystole was observed in 5.6% (7/124) of women across all doses in this misoprostol vaginal insert study as compared with 20.1% reported in the meta-analysis. Uterine hyperstimulation syndrome in this study occurred primarily in the two higher-dose reservoir groups, with 6%, and 10%, respectively, for the 100- and 200-μg dose reservoir groups as compared with 5.8% of women in the meta-analysis. Cesarean deliveries due to abnormal fetal heart rate patterns occurred in 1.6% (2/124) of women in the misoprostol vaginal insert study as compared with 6.7% reported in the meta-analysis.
The primary limitations of this study are that a relatively small number of women were tested at each reservoir dose, and no comparator product was included. The study was not controlled for baseline membrane status, and the low number of women with ruptured membranes at baseline precluded any meaningful analysis of the possible different outcomes in this subgroup. Oxytocin regimen was not specified in the protocol, and no analyses were conducted to determine if different regimens affected the outcome of the study, however, the maximum dose in all cases was below 20 units, indicating little reliance on oxytocin. The rate of nonreassuring fetal heart rate pattern and uterine hyperstimulation syndrome may have been underreported in this study because not all of the sites conducted continuous cardiotocographic monitoring. However, it is unlikely that any clinically significant uterine hyperstimulation would not have been recorded.
These early results in a relatively small study demonstrate that the misoprostol vaginal insert represents a promising method for a controlled-release, removable means of delivering misoprostol, a potent cervical ripening and labor induction agent. These early results are supported by a study conducted by Castañeda et al.7 Further investigation of the misoprostol vaginal insert is required in a large randomized, controlled study.