bleeding and increased uterine contractility. About 50 percent of the patients have started to bleed at the time of prostaglandin treatment, and almost all within four hours thereafter. The mean duration of bleeding in the French study was eight days. In 89.7 percent of the women, bleeding lasted for 12 days or less. In 9 percent of the women the bleeding was described as severe or excessive in the 24 to 48 hours after mifepristone and 4 hours after prostaglandin therapy (U.K. Multicentre Trial, 1990). In 93 percent of the women there was little or no change in the hemoglobin value (up to 1 g/dl) before treatment and seven days after gemeprost administration, while in 1 percent of the patients there was a decrease of 2–4 g/dl. The frequency of hemostatic surgical procedures was between 0.4 and 1.0 in the two studies (Table B5.1).
Uterine pain is also common, especially during the first hours following prostaglandin treatment: 2 percent of the women reported severe pain after mifepristone administration, a frequency that increased to 21 percent two hours after gemeprost administration. The frequency had decreased to 10 percent by four hours in the British study. Clinical signs of infection were rare, or less than 1 percent. Nausea and vomiting were not uncommon during the first hours after prostaglandin therapy.
In the French study, which included more than 16,000 women, serious cardiovascular side effects were reported in four cases after sulprostone injection. They consisted of one acute myocardial infarction attributed to a coronary spasm and marked hypertension in the other three women. Since the drug has been marketed and used in more than 60,000 cases in France, two additional myocardial infarctions have occurred, one of them being fatal. The frequency of severe cardiac complications after sulprostone is thus approximately 1/20,000 cases. As a consequence, smoking more than 10 cigarettes per day, age over 35 years, and any suspicion of cardiovascular disease must be considered relative contraindications to the method. It is noteworthy that so far, no myocardial infarction has occurred when gemeprost has been used, and most likely smaller doses of prostaglandin may still be sufficient and associated with a lower frequency of this type of complication.
As indicated previously, pregnancy might continue in spite of treatment, as occurred in 0.4 to 1.5 percent of clinical trial subjects. Mifepristone is known to cross the placenta (Frydman et al., 1985), and concern has been expressed that the drug may have teratogenic potential. Since surgical uterine evacuation is recommended when the treatment fails, data on infants exposed to mifepristone in utero are rare. However, the information that does exist suggests that the antiprogestin does not cause fetal malformations (Lim et al., 1990; Pom et al., 1991).