Dr O. S. Tang, Department of Obstetrics and Gynaecology, University of Hong Kong, Queen Mary Hospital, 6/F, Professorial Block, 102, Pokfulam Road, Hong Kong, SAR, China.
Objective To compare the efficacy, side effects and acceptability of sublingual and vaginal misoprostol for second trimester medical abortion.
Design Prospective randomised controlled trial.
Setting Tertiary referral unit and a teaching hospital.
Population Two hundred and twenty-four women at 12 to 20 weeks of gestation.
Methods The women were randomised to receive either sublingual or vaginal misoprostol 400 μg every 3 hours for a maximum of five doses. The course of misoprostol was repeated if the woman did not abort within 24 hours.
Main outcome measures The success rate at 48 hours, induction-to-abortion interval and the side effects.
Results There was no significant difference in the success rate at 48 hours (sublingual: 91%; vaginal: 95%). However, the success rate at 24 hours was significantly higher in the vaginal group (85%) compared with the sublingual group (64%). There was no difference in the median induction-to-abortion interval (sublingual: 13.8 hours; vaginal: 12.0 hours). Significantly more women in the sublingual group preferred the route to which they were assigned when compared with the vaginal group. The incidence of fever was also less in the sublingual group.
Conclusion The use of vaginal misoprostol for second trimester medical abortion resulted in a higher success rate than sublingual misoprostol at 24 hours but the abortion rate was similar at 48 hours. Vaginal misoprostol should be the regimen of choice but sublingual misoprostol is also an effective alternative.
Prostaglandin analogue is commonly used for termination of pregnancy in the second trimester. Misoprostol is the prostaglandin analogue commonly used, as it is cheap and stable at room temperature. It has been shown to be effective for second trimester termination of pregnancy with or without mifepristone.1–3 Although misoprostol is licensed for oral use, vaginal administration was used in most studies because it has been shown to be more effective.4 Recently, we have explored the use of sublingual misoprostol for medical abortion. A pharmacokinetic study has demonstrated that sublingual administration could achieve the peak concentration in the shortest time and had the highest bioavailability.5 Previous studies have shown that sublingual misoprostol is effective in first trimester medical abortion.6,7 A pilot study has demonstrated that it was feasible for second trimester medical abortion.8 It was the aim of this randomised study to compare the sublingual route to the vaginal route in the administration of misoprostol for medical abortion in the second trimester.
A total of 224 healthy women requesting legal termination of pregnancy at 12–20 weeks of gestation were recruited from October 2000 to July 2002 at Queen Mary Hospital in Hong Kong. Women who were using prescription drugs regularly, women with an intrauterine contraceptive device in utero, nursing mothers, multiple pregnancies and heavy smokers were excluded. An ultrasound examination was conducted to confirm the gestation of the pregnancy. An informed written consent was obtained. The study was approved by the ethics committee of the University of Hong Kong.
The women were randomly allocated to receive 400 μg of misoprostol (Cytotec; Searle, Crows Nest, NSW, Australia) either sublingually or vaginally every 3 hours for a maximum of five doses in 24 hours. The randomisation was done by computer-generated random numbers. The group assignments were put into sealed envelopes, which were opened when the women were recruited. This was an open randomised study and both the investigators and the subjects knew the routes that the women were assigned. If the women did not abort after the first course, the same regimen was repeated 24 hours after the start of the first dose of misoprostol. The blood pressure, pulse rate and temperature were monitored every 3 hours. Side effects were recorded every 3 hours by the nursing staff. Pethidine hydrochloride 50 mg (Pethidine; Antigen Pharmaceuticals, Roscrea, Ireland) was given for pain relief if necessary. After abortion, the products of conception were examined and if it was incomplete, evacuation of the uterus was performed. The decision for evacuation of the uterus was based on the clinical findings and ultrasound examination of the uterus was not done. The woman was discharged 24 hours after the abortion if there were no complications. Those who failed to abort after two courses of misoprostol took a rest for 24 hours and if they did not abort within these 24 hours, vaginal gemeprost was given.
The primary outcome measure was the success rate at 48 hours. Success rate was defined as abortion occurring without the need for further prostaglandin analogue or syntocinon. The side effects and the acceptability of the two regimens were also assessed. Differences in continuous variables were analysed with Student's t test for normally distributed data and the Mann–Whitney U tests for skewed data. Differences in discontinuous variables were analysed by χ2 test and the Fisher's exact test as appropriate.
The difference in success rate at 48 hours was used to calculate the sample size required. According to previous studies, the use of vaginal misoprostol would achieve a success rate of 91% at 48 hours.3 Sublingual misoprostol was of no clinical value if the success rate at 48 hours dropped below 75%, which was reported in a study using gemeprost given at 1 mg every 3 hours.9 A sample size of 224 has a power of 0.8 at 5% significance to detect a difference of 15% in success rate.
Three hundred and thirteen women requesting second trimester termination of pregnancy were screened. A total of 224 women at 12–20 weeks of gestation met the recruitment criteria and agreed to participate (Fig. 1). Three women from the sublingual and one woman from the vaginal were excluded from the study after randomisation. Two women from the sublingual group had withdrawn their consent after the first dose of misoprostol because of side effects. Medical abortion was continued with vaginal misoprostol, which was the standard regimen of our hospital. One woman from each group developed a rash after the first dose of misoprostol and was excluded and medical abortion was done by gemeprost. Therefore, there were 108 women in the sublingual group and 112 women in the vaginal group included in the final analysis. Table 1 showed the demographic characteristics of these 220 women. The success rate at 48 hours was 91% (95% CI: 84–95%) in the sublingual group and 95% (95% CI: 89–98%) in the vaginal group, respectively (Table 2). There was no statistically significant difference between the two groups. However, the success rate at 24 hours was significantly higher in the vaginal group (86%) compared with the sublingual group (72%). This was mainly due to the difference in success rate in the nulliparous women. No difference was found between the two routes of administration in the parous women. The median induction-to-abortion interval was similar in the two groups (sublingual: 12.2 hours and vaginal: 10.5 hours). Seventeen women (17%) in the sublingual group and 16 women (15%) in the vaginal group require surgical evacuation for incomplete abortion and this was not statistically significant. Two women from the vaginal group aborted after 48 hours without the need for further prostaglandin analogue or syntocinon. The remaining 14 women who failed to abort took a rest for 24 hours and all of them aborted with the use of intravaginal gemeprost.
Table 1. Demographic characteristic. Values are expressed as n (%) or mean [SD].
Sublingual group (n= 108)
Vaginal group (n= 112)
Gestational age (weeks)
Haemoglobin level (g/dL)
Subjects with previous abortion
Subjects with previous deliveries
Table 2. Treatment outcomes.
Sublingual group (n= 108)
Vaginal group (n= 112)
Median induction-to-abortion interval in hours (range)
Success rate in 24 hours (%)
Success rate in 48 hours (%)
Side effects were comparable in the two groups except that the incidence of fever was significantly higher in the vaginal group. And the number of women with breast tenderness was also higher (Table 3). No serious complication was reported in both groups.
The preference for the routes of administration was assessed by a questionnaire after the medical abortion before they were discharged (Table 4). Significantly more women (79%) in the sublingual group preferred the route of administration that they were assigned to when compared with the vaginal group (48%), even if the two women who withdrew their consent were included. The most common reason for choosing the sublingual routes was that it was more convenient, whereas the effectiveness and the taste of the drug were common reasons for choosing the vaginal route.
Table 4. Preference on the routes of administration.
Routes of administration preferred
Sublingual group (n= 107)
Vaginal group (n= 110)
More women in the sublingual group preferred the route that they were assigned, P= 0.0001.
Misoprostol with or without mifepristone has been investigated for medical abortion in the second trimester. Pretreatment with mifepristone resulted in a shorter induction-to-abortion interval compared with regimens without mifepristone.1–3,10 However, mifepristone is not available in many developing countries and it is expensive. The use of mifepristone will increase the cost of the medical abortion. Therefore, there is a need to develop an effective regimen without mifepristone.
Misoprostol has been studied extensively in this regard, as it is cheap and stable at room temperature. A previous study using vaginal misoprostol 400 μg every 3 hours achieved a success rate of 90% in 48 hours. The median induction-to-abortion interval was 15.2 hours.3 In our study, both the vaginal and sublingual groups achieved a success rate of over 90% in 48 hours. The median induction-to-abortion interval was comparable for the two routes of administration (10–12 hours). This means that both vaginal and sublingual misoprostol are effective for medical abortion in the second trimester. However, significantly more women aborted in 24 hours if vaginal misoprostol was used. If the side effects and patient acceptability are to be considered as well, sublingual misoprostol is an acceptable alternative.
A pharmacokinetic study has shown that after a single dose of sublingual misoprostol, the peak concentration is achieved in a shorter time than with vaginal misoprostol. The peak concentration and bioavailability were also higher with sublingual administration. Therefore, it was postulated that sublingual route of administration might be the most effective for misoprostol.5 However, the clinical efficacy of vaginal misoprostol in this study appeared to be slightly superior to sublingual misoprostol, as shown by the success rate at 24 hours. It was observed clinically that the misoprostol tablets took a long time to dissolve in the vagina.11,12 As a result, although the sublingual absorption was very effective and can achieve the higher peak concentration and bioavailability, the misoprostol might be metabolised and disappear from the blood stream very quickly. On the other hand, the slow vaginal absorption may be able to achieve a sustained and long-lasting effect. The cumulative bioavailability may be even greater than that of sublingual administration. It was shown in the pharmacokinetic study that at the end of 6 hours, the serum levels of misoprostol acid in the vaginal groups were higher than those of the sublingual group. Therefore, the effect of misoprostol may linger on for more than 6 hours after a single dose. Although the clinical effect of that low serum level is difficult to ascertain, the serum level can accumulate if vaginal misoprostol is repeated at an interval shorter than 6 hours as in the present regimen. In addition, a direct vagina-to-uterus transport was described for progesterone absorption recently.13,14 A similar mechanism may exist for misoprostol absorption and this can explain the more favourable clinical effect with the vaginal administration when compared with sublingual administration.
Previous pilot studies suggested that sublingual misoprostol was associated with a higher incidence of side effects especially diarrhoea, fever and chills. This was not supported by the present study. In fact, vaginal misoprostol was associated with a slightly higher incidence of fever. This may again be explained by the higher bioavailability associated with repeated administration of vaginal misoprostol. The high incidence of breast tenderness was likely to be a chance finding. It has not been reported in the literature that breast tenderness is a significant side effect of misoprostol.
It appeared from the acceptability study that sublingual administration was a more preferable way of giving misoprostol. This was not a double-blinded randomised study; the subjects had only tried the route of administration that they were assigned. However, we believed that sublingual administration was acceptable to more women in this study as significantly fewer women preferred their own route of administration in the vaginal group compared with the sublingual group. Sublingual misoprostol was considered to be more convenient and less uncomfortable. It was probably because with the present dosing interval, the women in the vaginal group required vaginal administration every 3 hours. This might be inconvenient and it also made the abortion process less private. These findings agreed with a previous study which showed that women preferred to take the misoprostol tablets by mouth as this could avoid the uncomfortable vaginal examination and provide more privacy during medical abortion.4 Another advantage of sublingual misoprostol is that the absorption of the drug is not affected if the women start to bleed. This is especially true for second trimester medical abortion when repeated administration of misoprostol is often required. However, the questionnaire did not take into the consideration the efficacy of the two regimens as this was not known before the study. The patient acceptability may be modified if efficacy is also taken into consideration.
This is the first randomised study comparing sublingual to vaginal misoprostol for second trimester abortion. The dose and dosing interval of sublingual misoprostol was chosen because this was comparable to the most common regimen used for vaginal administration. This dose and dosing interval may not be optimal. Further studies are required to find out the optimal dose so that this more acceptable route of administration can be used with results comparable to that of vaginal misoprostol.
In conclusion, the use of vaginal misoprostol for second trimester medical abortion resulted in a higher success rate than sublingual misoprostol at 24 hours but the success rate was similar at 48 hours. Vaginal misoprostol should be the regimen of choice but sublingual misoprostol is also an effective alternative.
The work described in this article was supported by a grant from the Research Grants Council of the Hong Kong Special Administrative Region, China (Project No: HKU 7244/01M).