Reviews/Analyses Medical approaches to termination of early pregnancy P.F.A. Van Look1 & M. Bygdeman2 Termination of early pregnancy by vacuum aspiration is a well-established technique. The procedure is highly effective and has a low complication rate, particularly when performed during the first few weeks after missed menses. However, this parasurgical method requires skilled health care personnel, and attempts have therefore been made to find medical alternatives. Prostaglandins were the first agents to be used successfully for this purpose, but the high incidence of mainly gastrointestinal side-effects has limited their widespread use. More recently, attention has focused on antiprogestational steroids, i.e., compounds that prevent the action of progesterone either by inhibiting its biosynthesis or by preventing it from binding to the cellular receptors in target organs. Current experience indicates that these com- pounds, particularly when used in combination with a low dose of a synthetic prostaglandin analogue, have the potential to become a safe, effective, and acceptable medical method for the termination of early pregnancy. Introduction efficacy and acceptability, as the currently employed parasurgical procedure, vacuum aspiration. Prior to The total number of legal abortions carried out examining the medical alternatives it is therefore globally each year is uncertain, but is likely to be appropriate to review briefly the performance of the 30-40 million (1). During the first 12 weeks of preg- parasurgical method. For this purpose, the dis- nancy, the uterus can be evacuated effectively by cussion of vacuum aspiration will be restricted to vacuum aspiration, and the rate of complications procedures carried out during the first 3-4 weeks that arise using this procedure is very low, particu- after missed menses (corresponding to 7-8 weeks' larly during the first few weeks after missed menses. amenorrhoea), since termination using medical However, parasurgical termination of pregnancy methods can only be used successfully during this requires skilled health care personnel, and attempts period (Fig. 1). have therefore been made to develop medical alter- For a detailed account of vacuum aspiration natives. during early pregnancy (a technique also known as Here, we review the experience gained with the mini-suction, mini-aspiration, or menstrual extrac- two medical approaches currently available: use of tion) see, for example, the review by van der Vlugt & prostaglandins and antiprogestational compounds. Piotrow (3). The latter comprise two distinct classes of drugs that have different modes of action: steroid biosynthesis inhibitors, such as epostane, and progesterone recep- Vacuum aspiration tor blockers, such as mifepristone (RU 486). Vacuum aspiration is a well-established technique In order to be clinically acceptable, it is manda- and will not be described in detail here. During the tory that any medical approach to termination of first 3-4 weeks after missed menses, it is usually per- early pregnancy is at least as good, in terms of safety, formed using a modified 50-ml syringe and a flexible plastic cannula of the type that was designed and 1 Medical Officer, Special Programme of Research, Development popularized by Karman in the late 1960s. Although and Research Training in Human Reproduction, World Health the exact design varies between manufacturers, the Organization, 1211 Geneva 27, Switzerland. Requests for reprints should be sent to cannulae generally have a blunt tip with two offset this author. apertures. Instead of a Karman syringe a number of 2 Chairman, Department of Obstetrics and Gynaecology, Karol- inska Hospital, Stockholm, Sweden. alternative suction sources can be used, such as a pre-evacuated bottle or a mechanical or electrical Reprint No. 5016. pump. In theory, neither anaesthesia nor dilatation Bulletin of the World Health Organization, 67 (5): 567-575 (1989) © World Health Organization 1989 567 P.F.A. Van Look & M. Bygdeman Fig. 1. Plot of the ability of prostaglandins to Induce Table 1: Distribution of significant Immediate com- complete abortion, by duration of amenorrhoea. plications among 12888 women who underwent mens- lW0, trual regulation by vacuum aspiratlon' 97 A\ 90 Complication No. of cases Uterine perforation Definite 3 (0.02)b so 81 \ Suspected 1 (0.01) Cervical laceration Sutured 2 (0.02) 70 199 Not sutured 5 (0.04) \ Blood loss (>100 ml) 51 (0.40) Shock 25 (0.19) ;ai 60 \I c Fever (> 38°C) 19 (0.15) 0 .t Pelvic infection 1 (0.01) -850 l Apnoea 2 (0.01) I CL Total 109 (0.85) E 40 8 39 See reference 5. b Figures in parentheses are percentages. 30 in the above-mentioned IFRP survey was 0.85% 20 (Table 1). Excessive bleeding, defined as a loss of more than 100 ml blood, accounted for slightly less 10 a than half of the immediate complication rate. i In the IFRP survey, delayed complications, . i.e., those that occurred more than 24 hours after 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 the evacuation but within a 6-week period, were re- Weeks since last menstrual period ported for 399 (3.53%) of the 11309 women who were followed up. Failed procedures, fever that of the cervix are required, although local anaesthesia required antibiotics, pelvic infection, or bleeding that is often used to minimize vasovagal symptoms, and required curettage (including incomplete abortions) mechanical dilatation of the cervix may be necessary were the most common delayed complications and in some women. The reported effectiveness of the procedure is occurred in about equal proportions (Table 2). A total of 508 significant, immediate or delayed com- high, with failure rates (i.e., the proportion of women plications were observed in the survey among the who require a repeat evacuation because of incom- 12888 women who underwent the procedure. This plete abortion or continuing pregnancy) of around represents a 3.94% overall complication rate if it is 2-6% (3-6). In a survey conducted by the Interna- assumed that the 1579 women (12.25%) who were tional Fertility Research Program (IFRP) between lost to follow-up did not experience a significant 1972 and 1976 in 61 clinics from 21 countries, the delayed complication. failure rate of vacuum aspiration among 11 309 women who were followed up was 1.2%, while a further 0.9% required curettage because of incom- Table 2: Distribution of significant delayed complications plete evacuation (5). In general, failure-to-terminate among 11309 women who were followed up after under- rates decrease with advancing duration of amenor- going termination of early pregnancy by vacuum aspira- rhoea and increasing experience of the individual tlon' performing the procedure. Complication No. of cases Complication rates for aspiration are low and are usually classified according to the time of their Failed proceduresb 111 (0.98)c occurrence. Immediate complications, i.e., those that Fever that required antibiotics 95 (0.84) Pelvic infection 90 (0.80) occur during the procedure itself or within 24 hours, Bleeding that required curettage 101 (0.89) include, inter alia, uterine perforation, cervical lacer- Undiagnosed ectopic pregnancy 2 (0.02) ation, excessive blood loss, shock, fever, pelvic infec- tion, and apnoea. For women who have no Total 399 (3.53) are not pre-existing medical conditions and who ' See reference 5. undergoing concurrent surgery or IUD insertion, the b Positive pregnancy test at follow-up visit. overall incidence of these immediate complications c Figures in parentheses are percentages. WHO Bulletin OMS. Vol. 67 1989 Medical approaches to termination of early pregnancy To the best of our knowledge, no systematic excessive frequency of mainly gastrointestinal side- studies have been conducted of the possible risk of effects, which precluded their routine use. The fre- late or long-term complications, i.e., those that occur quency of these side-effects can be reduced by 6 weeks or more after vacuum aspiration carried out intrauterine administration, but the clinical useful- during very early pregnancy. Such late complications ness of this approach is clearly very limited. Never- might include a tendency towards spontaneous abor- theless, these early studies were of importance since tion or prematurity in subsequent pregnancies they demonstrated the feasibility of using PGs and (which could be caused by damage to the cervix) as gave impetus to the search for analogues that have a well as acquired infertility, menstrual disorders, more selective effect on the myometrium and a rhesus immunization, or an increase in ectopic preg- greater resistance to in vivo enzymatic degradation nancies. Prospective studies, such as the investiga- by prostaglandin-15-dehydrogenase. This search has tion conducted jointly by the Royal College of been fairly successful and several analogues that are General Practitioners and the Royal College of effective when administered vaginally have been Obstetricians and Gynaecologists in the United developed (Table 3). Kingdom (7), indicate that subsequent pregnancies Among the first vaginally administered ana- are not adversely affected in women who have pre- logues to be tested clinically were 15-methyl-PGF2. viously undergone induced abortion. Since the methyl ester and 16,16-dimethyl-PGE2. For both of majority of women in these studies had their preg- these compounds, the frequency of complete abor- nancy terminated at more advanced stages of gesta- tions was satisfactory, but neither was appropriate tion, it seems reasonable to assume that vacuum for routine clinical use. In particular, for the PGF2. aspiration performed during the first 3-4 weeks after analogue the incidence of gastrointestinal side- missed menses is unlikely to have an adverse effect effects, although lower than that of the natural PGs, on the outcome of subsequent pregnancies. was still too high, whereas the PGE2 analogue was There is also no evidence that induced abortion, not stable in the suppository base. when properly performed, carries an increased risk More recently, PGE analogues have been devel- of secondary infertility.