Inhibition of progesterone secretion with trilostane for mid-trimester termination of pregnancy: randomized controlled trials

P.A. le Roux,1, S.K. Tregoning, P.M. Zinn and Z.M. van der Spuy

Reproductive Medicine Unit, Department of Obstetrics and Gynaecology, University of Cape Town/Groote Schuur Hospital, Cape Town, South Africa


    Abstract
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 Acknowledgements
 References
 
BACKGROUND: Progesterone is central to the maintenance of pregnancy, and is thus the ideal target for fertility regulation. Two mechanisms by which progesterone can be targeted are: receptor blockade and reduction of progesterone production through enzyme inhibition. Mifepristone, a receptor blocker, is usually given as `pretreatment' prior to prostaglandin administration in mid-trimester termination of pregnancy (TOP). Unfortunately, there are difficulties accessing mifepristone in developing countries, and TOP is therefore performed using prostaglandins alone, which results in unacceptably long induction-to-abortion intervals. Trilostane is a 3ß-hydroxysteroid dehydrogenase inhibitor which reduces progesterone production. In these mid-trimester studies it is evaluated as a method of pretreatment prior to misoprostol administration. METHODS: Three consecutive randomized controlled trials comparing different trilostane regimens for pretreatment were performed. In study 1, trilostane was compared with placebo; in study 2, two doses of trilostane were compared (1080 mg and 720 mg); in study 3, the effect of adding danazol to trilostane as combination therapy was evaluated. The primary outcome in all the studies was the induction-to-abortion interval. Serum progesterone, estradiol and cortisol were measured serially during treatment. RESULTS: In study 1, 48 women were randomized. The median induction-to-abortion interval was 9 h in the trilostane group and 18.5 h in the placebo group (P < 0.0001). Progesterone and estradiol production was significantly reduced in the women receiving trilostane, with maintenance of diurnal cortisol variation. Twenty-eight women were randomized in study 2, which demonstrated that there was no significant difference in the induction-to-abortion interval using 1080 mg and 720 mg trilostane when compared with the higher doses used in study 1. Study 3, in which 40 women were included, failed to show any additional benefit using combination therapy with danazol and trilostane. CONCLUSIONS: Trilostane is an effective pretreatment agent in mid-trimester TOP.

Key words: mid-trimester/misoprostol/pregnancy termination/trilostane


    Introduction
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 Acknowledgements
 References
 
The need for safe, effective forms of fertility regulation is demonstrated by the estimated 20 million unsafe abortions that are performed worldwide each year. It is calculated that a third of known pregnancies end in induced abortion (Tietze, 1983Go). Progesterone is central to the maintenance of pregnancy, and is thus an ideal target for fertility regulation. The role of progesterone in supporting early pregnancy has been demonstrated in elegant clinical experiments (Csapo and Pulkkinen, 1978Go) which showed that lutectomy (removal of the corpus luteum) in early pregnancy led to spontaneous abortion, which could be prevented by the administration of exogenous progesterone but not by estrogen replacement.

The introduction of mifepristone (a progesterone receptor antagonist) into clinical practice in Europe, and more recently the USA, has provided a method of safe medical termination of pregnancy (TOP) with low failure rates. This has not been available in the developing world where arguably the need for safe abortion is far greater.

In South Africa, late presentation is common and 50% of patients requesting TOP at Groote Schuur Hospital are already in the second trimester of pregnancy. This is due in part to a lack of awareness of the availability of services and in part to difficulty in accessing care. The Choice of Termination of Pregnancy Act (1997) allows TOP on request up to 12 weeks, and in consultation with two health workers up to 20 weeks gestation. It is relatively recent legislation in South Africa, and there are still problems with effective implementation of TOP services (Harrison et al., 2000Go). In practice, mid-trimester TOP is a problematic procedure. Patients usually receive misoprostol therapy alone, without anti-progestin `pretreatment', and this may result in protracted induction-to-abortion intervals and failure of medical therapy. This is distressing to both patients and health care staff and further prejudices the implementation of safe abortion services. Optimal regimens for mid-trimester TOP are essential if the morbidity and mortality of unsafe abortion are to be reduced.

It has been demonstrated that mifepristone, as a single 200 mg dose given prior to prostaglandin therapy, is effective in facilitating the action of subsequent prostaglandin administration in mid-trimester TOP (Webster et al., 1996Go). Mifepristone has recently been registered by the South African Medicines Control Council for use in medical TOP but its cost may limit its availability in the public sector. Alternative medical methods for TOP therefore need to be investigated in order to improve Women's Health in South Africa and other developing countries which face similar problems.

Trilostane is a competitive inhibitor of the 3ß-hydroxysteroid dehydrogenase (3ß-HSD) enzyme system with no intrinsic endocrine action, and inhibits the production of progesterone resulting in reduced plasma progesterone concentrations. A previous study has shown that trilostane is effective in reducing progesterone secretion in the first trimester and may facilitate TOP (van der Spuy et al., 1983Go; van der Spuy, 1984Go). Studies using epostane, a compound similar to trilostane, demonstrated its effectiveness in TOP (Webster et al., 1985Go; Birgerson and Odlind, 1987Go; Crooij et al., 1988Go), but this drug is no longer available for clinical use.

This study was undertaken to assess whether trilostane is effective adjuvant therapy in mid-trimester TOP and whether it will improve the outcome of treatment with prostaglandins. Trilostane was compared with placebo as pretreatment in women having mid-trimester TOP prior to misoprostol administration. Different dosages of trilostane were compared with establish the lowest effective dosage necessary for the treatment, and the addition of danazol was evaluated to see whether there is benefit in combination therapy with two agents. Danazol has previously been shown to reduce the effect of progesterone by inhibiting steroidogenesis (Barbieri et al., 1977Go; Stillman et al., 1980Go; Steingold et al., 1986Go), and blocking the progesterone receptor (Barbieri and Ryan, 1981Go).


    Materials and methods
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 Acknowledgements
 References
 
Protocol
Women were recruited from the Termination of Pregnancy Clinic at Groote Schuur Hospital (a university teaching hospital) for the three studies. Demographic data, including age, gravidity, parity, gestational age and indication for TOP, were recorded. The inclusion and exclusion criteria were the same for all three studies. Women in the mid-trimester of pregnancy (13–19 weeks) who requested TOP were included. Women with previous uterine surgery or severe medical problems and those under 18 years of age, and therefore unable to sign consent for a clinical trial, were excluded.

Approval to perform this research was granted by the Ethics Committee of the Faculty of Health Sciences (University of Cape Town). The South African Medicines Control Council gave permission to use trilostane (unregistered in South Africa) in these studies. All participants gave written informed consent.

The patients were randomized to compare the effect of different pretreatment regimens prior to prostaglandin administration. During pretreatment, 12 hourly blood samples were taken to measure steroid levels. After pretreatment, all women received 200 µg misoprostol vaginally every 4 h to induce abortion. The first dose of misoprostol was administered by the medical attendant, and subsequent doses were dispensed by the nursing staff and self-administered by the patients. Uterine evacuation was performed after delivery of the fetus, as was routine in our clinical practice following mid-trimester TOP.

Morphine was administered i.m. for analgesia as required by the patients. The patients' satisfaction with treatment was recorded by using a visual analogue scale. They were asked to rate their experience from 1 (very good) to 5 (very bad) on a visual analogue scale, to describe their global satisfaction with their care and the degree of pain experienced during the TOP procedure.

The women were randomized to the following pretreatment regimens: study 1: trilostane or placebo over 48–72 h (see Table IGo for trilostane dosage breakdown used in this study); study 2: 720 mg trilostane (240 mg day 1, 480 mg day 2) or 1080 mg trilostane (360 mg day 1, 720 mg day 2) over 48 h; study 3: 720 mg trilostane or 720 mg trilostane plus 800 mg danazol (400 mg day 1 and 2). Trilostane and danazol were administered in divided doses (two or three times daily) by the investigators conducting the study.


View this table:
[in this window]
[in a new window]
 
Table I. Dosage regimen used in study 1
 
The primary outcomes studied were the incidence of abortion within 24 h and the induction to delivery time after the administration of misoprostol. Secondary outcomes included the total number of misoprostol doses administered, the number of patients requiring morphine and patient satisfaction with the procedure as assessed by the visual analogue scale.

If the patient were undelivered after 24 h this was considered a failure of therapy. She then received further doses of misoprostol and, if no response was achieved within 72 h, extra-amniotic prostaglandin F2{alpha} was administered. Adverse events were carefully noted, even if they were not related to the study medication. If a patient experienced a serious adverse event during treatment she would be withdrawn, but included in intention-to-treat analysis regarding side-effect outcomes.

Because premature termination of these studies would be necessary if unexpected serious adverse events occurred during treatment, patients were randomized in blocks of four, facilitating interim analysis if the study were not completed.

Statistical analysis
Sample size calculations showed that 20 patients were needed in the trilostane and placebo groups to detect a difference of 40% in the number of patients delivered at 24 h. It was hypothesized that relatively few women in the placebo group (receiving low dose misoprostol only) would deliver in the first 24 h. The group-specific abortion rates in the first 24 h of 90 and 50% in the trilostane and placebo groups were used to make this calculation. The {alpha} value was set at 0.05 and the power at 80%.

In study 2, the same sample sizes were chosen which had resulted in a significant difference in efficacy (with a specific dose of trilostane) in study 1. Fourteen patients were needed in each dosage group. In study 3, a sample size of 40 patients was needed to detect a possible difference of 40% in the number of patients delivered within 8 h. The null hypothesis was that there would be no difference in the induction to delivery interval with trilostane alone compared with combination therapy. The alternative hypothesis stated that >90% of patients treated with combination therapy would deliver within 8 h compared with 50% treated with trilostane alone. The reason 8 h was chosen as a point of reference was that the first study comparing trilostane with placebo demonstrated that the median induction-to-abortion time with trilostane was 9 h. It was hypothesized that the combination therapy with danazol would further reduce this interval.

The computer packages Microsoft Excel (Microsoft, V&A Waterfront 8001, South Africa) and GraphPad Prism (GraphPad Software, Inc., San Diego, CA, USA) were used for data analysis. Statistical analyses on continuous variables were performed using unpaired Student's t-test or analysis of variance on normally distributed data and Mann–Whitney U-test and Kruskal–Wallis test on data that were not normally distributed. Paired data were assessed using the paired Student's t-test or the Wilcoxon rank test. Categorical variables were analysed using relative risks with corresponding 95% confidence intervals, {chi}2-test or Fisher's exact test. Significance levels were set at 0.05.

Assignment and blinding
The patients were randomized in blocks of four using computerized randomization techniques, to compare the effect of different pretreatment regimens. This randomization was done by a clinician not involved in recruiting the patients for the studies. The unit of randomization was the individual patient who was allocated to one of two treatment groups, as described in the protocol. The allocation list was not available to the investigators recruiting patients.

Trilostane capsules (120 mg), and identical placebo capsules, were donated by Stegram Pharmaceuticals (Billinghurst, Sussex, UK), and imported for the purpose of the study. Placebo and trilostane capsules were packaged by the clinician who performed the randomization. Study 1 was a double-blind, randomized, placebo-controlled trial. Study 2 was randomized but not blinded. In study 3, although the study was randomized, there were no placebo tablets available for danazol and it was therefore not possible to do a blinded study.


    Results
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 Acknowledgements
 References
 
Participant flow and follow-up
The women were recruited from 1997 to 2000 for the three consecutive studies. All eligible women were randomized and there were no refusals to participate in the studies. The interventions were allocated as follows: study 1: 48 women randomized to receive either trilostane or placebo; study 2: 28 women randomized to receive 1080 mg or 720 mg trilostane; study 3: 40 women randomized to receive trilostane alone or trilostane plus danazol.

The women were treated as inpatients during the studies and there was therefore no loss to follow-up prior to recording outcomes. In study 1, two patients were withdrawn because of severe nausea and vomiting and another two due to protocol violations (medication not taken at the correct dosage intervals), and in study 2, one patient was withdrawn because of side-effects and another because of protocol violation. There were no patients withdrawn from study 3.

There were 23 patients analysed in the placebo group and 21 patients analysed in the trilostane group in study 1. All withdrawals were included in an intention-to-treat analysis for the evaluation of side-effects. In addition, an intention-to-treat analysis was done for the primary outcome of the number of patients delivered in 24 h even though the withdrawn patients did not receive the full treatment protocol.

In study 2, one patient was withdrawn from each group. There were 13 patients suitable for analysis in each dosage regimen. In study 3, all patients were analysed.

Analysis
Study 1: trilostane versus placebo
There was no difference in age, gravidity, parity or gestational age in the trilostane and placebo groups (Table IIGo). All the patients in the study had TOP performed [as per section 2(1)(c)(i) of the Choice of Termination of Pregnancy Act (1996)] for socio-economic risk.


View this table:
[in this window]
[in a new window]
 
Table II. Demographic details
 
The median induction to delivery times in the trilostane and placebo groups were 9.0 h and 18.5 h respectively (P < 0.0001). There were no patients who had not aborted after 24 h in the trilostane group compared with six in the placebo group (P < 0.05) (Table IIIGo). The intention-to-treat analysis, including patients who had been withdrawn for side-effects or protocol violations, showed that there were still no patients who had not aborted in the trilostane group and seven patients not yet aborted in the placebo group after 24 h (P = 0.009).


View this table:
[in this window]
[in a new window]
 
Table III. Outcomes of therapy
 
Significantly fewer patients required i.m. morphine for analgesia in the trilostane group (19%) compared with the placebo group (70%). There was a statistically lower pain score recorded in the trilostane group compared with placebo after treatment, with median scores of 2 and 3 respectively on a visual analogue scale of 1–5 (P = 0.02). The women in the trilostane group had a significantly lower score (i.e. more satisfied) with the whole TOP experience (P = 0.02).

There was no significant difference in the mean drop in haemoglobin in the two groups, and no patients required blood transfusion. No patients required prostaglandin F2{alpha} or hysterotomy and there were no ruptured uteri.

The side-effects associated with trilostane were mild and self-limiting, and patients were able to tolerate therapy. It was more common to experience side-effects with a larger dose of trilostane on the second day of therapy (Table IVGo). The most common adverse effects noted were nausea/vomiting (30%) and a burning, itching or flushing feeling (26%) in the patient's face with trilostane therapy.


View this table:
[in this window]
[in a new window]
 
Table IV. Side-effects of therapy
 
The production of progesterone was inhibited in all patients on trilostane therapy. There was a mean decrease in serum progesterone of 79% from 198.8 to 58.6 nmol/l after 48 h treatment with trilostane (P < 0.0001) (Figure 1Go). The production of estradiol was also inhibited and the estradiol levels decreased from a mean of 13 643 to 5670 pmol/l after 48 h treatment (Figure 2Go). There was no significant change in serum cortisol (where normal diurnal variation was maintained; Figure 3Go). The statistically significant fall in the sex steroid levels was present from the first sample taken after treatment at 12 h and maintained during the duration of treatment. The comparison of the levels at 0 and 12 h is shown in Table VGo.



View larger version (13K):
[in this window]
[in a new window]
 
Figure 1. Serum progesterone levels in women given trilostane (n = 21), an inhibitor of the 3ß-hydroxysteroid dehydrogenase enzyme system, and women given placebo (n = 23) for mid-trimester termination of pregnancy. Serial sampling was done at12 hourly intervals for 48 h. Suppression of serum progesterone in the women treated with trilostane is demonstrated.

 


View larger version (13K):
[in this window]
[in a new window]
 
Figure 2. Serum estradiol levels in women given trilostane(n = 21), an inhibitor of the 3ß-hydroxysteroid dehydrogenase enzyme system, and women given placebo (n = 23) for mid-trimester termination of pregnancy. Serial sampling was done at12 hourly intervals for 48 h. Suppression of serum estradiol in the women treated with trilostane is demonstrated.

 


View larger version (14K):
[in this window]
[in a new window]
 
Figure 3. Serum cortisol levels in women given trilostane(n = 21), an inhibitor of the 3ß-hydroxysteroid dehydrogenase enzyme system, and women given placebo (n = 23) for mid-trimester termination of pregnancy. Serial sampling was done at12 hourly intervals for 48 h. Maintenance of the normal diurnal variation in serum cortisol is demonstrated in both groups.

 

View this table:
[in this window]
[in a new window]
 
Table V. Change in serum progesterone and estradiol
 
Study 2
There were no demographic differences between the two groups studied (see Table VIGo). The mean induction-to-abortion times were 9.78 h in the group receiving 720 mg trilostane and 8.91 h in the group receiving 1080 mg trilostane (P = not significant). The severity of pain as recorded by visual analogue scale was significantly more in the group receiving the larger dose of 1080 mg trilostane (P = 0.005). The lowest dose of 720 mg was as effective as larger doses and resulted in a less painful TOP. No differences in the incidence of side-effects were observed.


View this table:
[in this window]
[in a new window]
 
Table VI. Dose comparison study (1080 mg versus 720 mg trilostane pretreatment)
 
Study 3
There were no demographic differences between the patients randomized to the trilostane or combination groups (Table VIIGo). There was no apparent benefit in combination therapy; the induction-to-abortion times were 10.1 and 11.35 h with trilostane and combination therapy respectively. There was also no significant difference found in failure of the regimen or side-effect profile in the two groups. The fall in serum progesterone was similar even with combination therapy, and danazol did not alter the pattern of cortisol production.


View this table:
[in this window]
[in a new window]
 
Table VII. Trilostane versus trilostane + danazol
 

    Discussion
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 Acknowledgements
 References
 
Trilostane is effective pretreatment in TOP prior to prostaglandin administration. It resulted in a 50% reduction of the induction-to-abortion interval and significantly fewer failures of therapy 24 h after prostaglandin administration.

The role of progesterone in pregnancy is central to understanding any intervention to terminate pregnancy. Progesterone relaxes uterine smooth muscle and allows pregnancy to continue uninterrupted (Csapo, 1970Go). Experiments in which lutectomy was performed in early pregnancy showed that with `surgical' removal of progesterone, uterine contractions would ensue, aborting the fetus (Csapo and Pulkkinen, 1978Go). By 7 weeks gestation, placental production of progesterone obviated the effect of ovarian lutectomy, preventing spontaneous abortion. It is also postulated that the onset of spontaneous labour at term is brought about by an alteration of the estrogen:progesterone ratio (Challis 1971Go; Liggins et al., 1973Go; Darne et al., 1987Go). The mechanism by which a drop in progesterone facilitates smooth muscle contraction is not entirely clear, although it is probably by activation of the endogenous prostaglandin pathway (Csapo et al., 1973Go). Progesterone enhances stability of lysosomal membranes, and reduction of progesterone levels may facilitate the release of lysosomal phospholipase A2, which initiates prostaglandin production (Kelly et al., 1986Go). Progesterone also inhibits the formation of estradiol receptors and gap junctions which are important for co-ordinated uterine activity (Bygdeman and Swahn, 1990Go), and leads to an inverse increase in the number of oxytocin receptors in the myometrium (Csapo et al., 1973Go). The only practical methods available to reduce progesterone are by reduction of production through enzyme inhibition, or reducing effect by receptor blockade.

Mifepristone is a progesterone receptor blocker with additional anti-glucocorticoid receptor activity (Mahajan and London, 1997Go). Misoprostol, used alone in the second trimester, has success rates varying from 73 to 92% (Scheepers et al., 1999Go), with only 80% of TOP completed by 18 h (Bulgalho et al., 1993Go). Patients who are refractory to misoprostol therapy often end up with either prostaglandin F2{alpha} therapy or surgical TOP, which are both procedures with inherent risks. Two-thirds of all major complications in TOP and half of the maternal mortality is related to mid-trimester TOP (Tietze, 1981Go). Once mifepristone is added as pretreatment to misoprostol therapy, success rates climb to 97% (El Refaey and Templeton, 1995Go), and mifepristone has thus become the `gold standard' pretreatment in second trimester TOP. It makes medical TOP in the second trimester a safer and more effective procedure. Unfortunately, its availability and cost limit its use in developing countries and other agents for TOP need to be studied. The estimated cost of mifepristone in South Africa is ~R1000.00 compared with R96.00 for a course of trilostane.

The other major strategy to block the effect of progesterone is to reduce its production by inhibition of steriodogenesis. Enzyme inhibition is certainly not a new concept, and well-known examples of inhibitors include metyrapone for control of corticosteroid excess in Cushing's syndrome and aminoglutethimide to reduce estrogens and androgens in hormone-dependent malignancies. Although aminoglutethimide reduces progesterone production it was not shown to cause abortion in women or baboons (Glasser et al., 1972Go; Steinetz et al., 1975Go). Azastene, an isoxazole structurally similar to pregnenolone, caused abortion in rats by enzyme inhibition but was not successful in women (Csapo et al., 1979Go). Both danazol and oxymethalone have also been shown to have some inhibitory effect on 3ß-HSD activity but failed to cause abortion in women when used alone (Brenner et al., 1977Go; Wentz and Sapp, 1978Go; Asch et al., 1980Go). This may be due to the fact that these two drugs were limited in their ability to overcome the ß-HCG support of the corpus luteum. Danazol was used in this study in an effort to augment the effect of trilostane by further reducing serum progesterone. No additive effect with the danazol was evident|endocrinologically or clinically.

Trilostane and epostane, derived from testosterone and 17{alpha}-methyl testosterone respectively, both inhibit 3ß-HSD and overcome the effect of ß-HCG in pregnancy. They were developed in the 1970s, but the drugs were subsequently withdrawn from the market. Another pharmaceutical company has now acquired the patent for trilostane, and the role of trilostane in gynaecology has been re-examined in these studies. These are the only studies in women where TOP has been studied using enzyme inhibition by trilostane followed by prostaglandin therapy. The Cochrane database does not contain any systematic reviews on misoprostol, mifepristone or trilostane therapy for mid-trimester TOP (Cochrane Library, 2001Go). The only randomized studies with trilostane are recorded in the Cochrane Register of Randomised Controlled Trials which contains one trial evaluating trilostane therapy for incomplete abortion and 5 trials comparing trilostane and aminoglutethimide therapy in the treatment of post-menopausal breast carcinoma. There are no randomized trials listed in the Cochrane database that evaluate trilostane therapy in medical TOP.

Trilostane inhibits the conversion of pregnenolone to progesterone, which resulted in a mean fall of 79% in serum progesterone in these studies. Trilostane itself has no inherent steroidal activity, and although the precursor pregnenolone rises due to enzyme blockade, it is biologically inert. At the doses used in this study, there was minimal adrenal suppression of cortisol production and normal diurnal variation was maintained. Obviously higher doses of trilostane can be used to inhibit cortisol production, and the drug is registered in the UK to treat Cushing's syndrome by this mechanism.

The production of estradiol is also inhibited, and although this is apparently unimportant in TOP, the possible therapeutic effect should be investigated in conditions where hypo-estrogenism is beneficial, such as breast carcinoma and endometriosis.

The use of trilostane for pretreatment prior to misoprostol therapy is an entirely new approach to TOP. In these studies, we have shown that it is effective and acceptable to women. The women taking part is these trials were from several ethnic groups, and these results should be equally valid in other population groups.

Trilostane should be at the forefront of research in both first and second trimester TOP in the future. It will be of great clinical value in the difficult scenario of mid-trimester TOP where mifepristone is unavailable. Advantages in cost, efficacy or side-effect profile in future research may indeed make it preferable to mifepristone. Further studies should address the feasibility of using it on an outpatient basis and evaluate its success in conjunction with higher doses of misoprostol.


    Acknowledgements
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 Acknowledgements
 References
 
The authors would like to acknowledge financial assistance received from the Medical Research Council of South Africa, the Abortion Rights Action Group, the Nina Lipsitz Fund and Stegram Pharmaceuticals.


    Notes
 
1 To whom correspondence should be addressed at: Department of Obstetrics and Gynaecology, Old Main Building, Groote Schuur Hospital, Observatory, Cape Town 7925, South Africa. E-mail: pleroux{at}uctgsh1.uct.ac.za Back


    References
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 Acknowledgements
 References
 
Asch, R.H., Fernandez, E.O., Siler-Khodr, T.M. et al. (1980) Mechanism of induction of luteal phase defects by danazol. Am. J. Obstet. Gynaecol., 136, 932–940.[ISI][Medline]

Barbieri, R.L. and Ryan, K.J. (1981) Danazol: endocrine pharmacology and theraputic applications. Am. J. Obstet. Gynecol., 141, 453–463.[ISI][Medline]

Barbieri, R.L., Canick, J.A., Makris, A. et al. (1977) Danazol inhibits steriodogenesis. Fertil. Steril., 28, 809–813.[ISI][Medline]

Birgerson, L. and Odlind, V. (1987) Early pregnancy termination with antiprogestins: a comparative clinical study of RU486 given in two dose regimens and Epostane. Fertil. Steril., 48, 565–570.[ISI][Medline]

Brenner, P.F., Mishell, D.R., Stanczyk, F.Z. et al. (1977) Serum levels of d-norgestrel, luteinizing hormone, follicle stimulating hormone, estradiol and progesterone in women during and following ingestion of combined oral contraceptives containing dl-norgestrel. Am. J. Obstet. Gynecol., 129, 133–140.[ISI][Medline]

Bulgalho, A., Bique, C., Amieda, L. et al. (1993) Pregnancy interruption by vaginal misoprostol. Gynecol. Obstet. Invest., 34, 226–229.

Bygdeman, M. and Swahn, M.L. (1990) Uterine contractility during pregnancy and the effect of abortifacient drugs. Ballière's Clin. Obstet. Gynaecol., 4, 249–261.

Challis, J.R.G. (1971) Sharp increase in free circulating oestrogens immediately before parturition in sheep. Nature, 229, 208.[ISI][Medline]

Cochrane Library (2001) Issue 2. Update Software Inc., California.

Crooij, M.J., de Nooyer, C.C.A., Rao, B.R. et al. (1988) Termination of early pregnancy by the 3ß-hydroxysteroid dehydrogenase inhibitor epostane. New Engl. J. Med., 319, 813–817.[Abstract]

Csapo, A.I. (1970) The diagnostic significance of the intrauterine pressure. Obstet. Gynaecol. Surv., 25, 515–543.[Medline]

Csapo, A.I. and Pulkkinen, M. (1978) Indispensibility of the human corpus luteum in maintenance of early pregnancy. Lutectomy evidence. Obstet. Gynaecol. Surv., 33, 69–81.[Medline]

Csapo, A.I. and Pulkkinen, M. and Kaihola, H.L. (1973) The effect of lutectomy induced progesterone withdrawal on the oxytocin and prostaglandin response of the first trimester pregnant human uterus. Prostaglandins, 4, 421–429.[Medline]

Csapo, A.I., Resch, B., Csapo, E.F. et al. (1979) Effects of antiprogesterone on pregnancy. Am. J. Obstet. Gynecol., 133, 176–183.[ISI][Medline]

Darne, J., McGarrigle, H.H.G. and Lachelin, G.C.L. (1987) Saliva oestriol, estradiol, estrone and progesterone levels in pregnancy: spontaneous labour at term is preceded by a rise in the saliva estriol:progesterone ratio. Br. J. Obstet. Gynaecol., 94, 227–235.[ISI][Medline]

El Refaey, H. and Templeton, A. (1995) Induction of abortion in the second trimester by a combination of misoprostol and mifepristone: a randomized comparison between two misoprostol regimes. Hum. Reprod., 10, 475–478.[Abstract]

Glasser, S.R., Northcutt, R.C., Chytil, F. et al. (1972) The influence of an anti-steroidogenic drug (aminoglutethimide phosphate) on pregnancy maintenance. Endocrinology, 90, 1363–1370.[ISI][Medline]

Harrison, A., Montgomery, E.T., Lurie, M. et al. (2000) Barriers to implementing South Africa's Termination of Pregnancy Act in rural KwaZulu/Natal. Health Policy Plan., 15, 424–431.[Abstract]

Kelly, R.W., Healy, D.L., Cameron, M.J. et al. (1986) The stimulation of prostaglandin production by two anti-progesterone steroids in human endometrial cells. J. Clin. Endocrinol. Metab., 62, 1116–1123.[Abstract]

Liggins, G.C., Fairclough, R.G., Grieves, S.A. et al. (1973) The mechanism of parturition in the ewe. Recent Prog. Horm. Res., 29, 111–150.[Medline]

Mahajan, D.K. and London, S.N. (1997) Mifepristone (RU486): a review. Fertil. Steril., 68, 967–976.[ISI][Medline]

Scheepers, H.C.J., van Erp, E.J.M. and van den Bergh, A.S. (1999) Use of misoprostol in first and second trimester abortion: a review. Obstet. Gynecol. Surv., 54, 592–600.[Medline]

Steinetz, B.G., Niemann, W.H., Giannina, T., Duncan, D. and Chart, J.J. et al. (1975) Effect of aminoglutethimide on serum progesterone and estrogen in the pregnant baboon. Proc. Soc. Exp. Biol. Med., 148, 790–792.[Abstract]

Steingold, K.A., Lu, J.K.H., Judd, H.L. et al. (1986) Danazol inhibits steriodogenesis by the human ovary in vivo. Fertil. Steril., 45, 649–655.[ISI][Medline]

Stillman, R.J., Fencl, M.D, Schiff, I. et al. (1980) Inhibition of adrenal steriodogenesis by danazol in vivo. Fertil. Steril., 33, 401–406.[ISI][Medline]

Tietze, C. (1981) Second trimester abortion: a global view. In Berger, G.S., Brenner, W.E. and Keith, L.G. (eds), Second Trimester Abortion. John Wright, Boston, pp. 1–11.

Tietze, C. (1983) Induced Abortion. A World Review, 5th edn. The Population Council, New York.

Van der Spuy, Z.M. (1984) Inhibition of progesterone secretion in women. PhD thesis, University of London.

Van der Spuy, Z.M., Jones, D.L., Wright, C.S.W. et al. (1983) Inhibition of 3-beta-hydroxy steroid dehydrogenase activity in first trimester human pregnancy with trilostane and WIN 32729. Clin. Endocrinol., 19, 521–532.[ISI][Medline]

Webster, D., Penney, G.C. and Templeton, A.A. (1996) A comparison of 600 and 200 mg mifepristone prior to second trimester abortion with the prostaglandin misoprostol. Br. J. Obstet. Gynaecol., 103, 706–709.[ISI][Medline]

Webster, M.A., Phipps, S.L. and Gillmer, M.D.G. (1985) Interruption of first trimester human pregnancy following Epostane therapy. Effect of prostaglandin E2 pessaries. Br. J. Obstet. Gynaecol., 92, 963–968.[ISI][Medline]

Wentz, A.C. and Sapp, K.C. (1978) Danazol as a luteolytic agent. Fertil. Steril., 29, 23–25.[ISI][Medline]

Submitted on November 27, 2001; accepted on February 4, 2002.