Effect of long-term treatment with low-dose mifepristone on the endometrium

D.T. Baird1,4, A. Brown1, H.O.D. Critchley1, A.R. Williams2, S. Lin3 and L. Cheng3

1 Contraceptive Development Network, Centre for Reproductive Biology, University of Edinburgh, Academic Centre, 49 Little France Crescent, Old Dalkeith Road, Edinburgh EH16 4SB, 2 Medical School, Teviot Place, Edinburgh EH8 9AG, UK and 3 Shanghai Institute of Family Planning Technical Instruction, Shanghai, People’s Republic of China 4 To whom correspondence should be addressed. e-mail: dtbaird{at}ed.ac.uk


    Abstract
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
BACKGROUND: Mifepristone in low daily doses has contraceptive potential by inhibiting ovulation and menstruation. Because follicular development is maintained, the endometrium is exposed to estrogen for prolonged periods unopposed by progesterone. METHODS: Endometrial biopsies were collected from 90 women in Edinburgh and Shanghai before (late proliferative) and 60 and 120 days after taking 2 or 5 mg mifepristone per day for 120 days. RESULTS: Ovulation and menstruation were inhibited in >90% of cycles and estrogen production was similar to that observed during the follicular phase of the control cycle. By 120 days, endometrial thickness increased significantly in women in Edinburgh but decreased in Shanghai. Endometrial histology showed inactive proliferative or cystic changes with dense stroma. There was a significant decrease in markers of proliferation, i.e. mitotic index and Ki67 staining. There were no pregnancies in a total of 200 women-months in 50 sexually active women who used no other method of contraception. CONCLUSIONS: We confirm that ovulation and menstruation were suppressed in the majority of cycles and there was asynchrony between ovarian activity and endometrial histology, which showed no signs of hyperplasia or atypia. These preliminary data suggest that daily low-dose mifepristone is potentially a safe estrogen-free contraceptive pill which has the added health benefit of amenorrhoea.

Key words: antigestogen/antiproliferation/contraception/mifepristone


    Introduction
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
Since it was marketed over 40 years ago, the combined oral contraceptive (COC) has proved to be a popular, highly effective method of hormonal contraception (Baird and Glasier, 1999Go). The combination of an orally active synthetic estrogen and gestogen suppresses gonadotrophins and hence ovulation. The pattern of menstrual bleeding in the majority of cycles is highly predictable and is one of the reasons for its continuing popularity worldwide. However, the COC is associated, in some women, with a number of side-effects such as acne, weight gain and menstrual irregularity, which have led to the search for newer gestogens with more favourable metabolic profiles and fewer side-effects. Rare, more serious, side-effects such as deep venous thrombosis and pulmonary embolism are thought to be associated with the estrogen component of some gestogens (Skegg, 2000Go) and have led to the development of progestogen-only pills (POP), implants and injections. The major drawback to all gestogen-only methods is a high incidence of unpredictable breakthrough bleeding which leads to relatively low continuation rates (Fraser, 2000Go).

Mifepristone is a synthetic C19 norsteroid which is a potent antagonist of progesterone (Ulmann, 2000Go) and is licensed in most European countries for induction of abortion in early pregnancy (Sitruk-Ware, 2000Go). Published data suggest that it is a highly effective emergency contraceptive and has potential as a once-a-month pill (Baird, 2001Go). In daily doses between 2 and 10 mg, ovulation is suppressed although variable degree of follicular development continues (Ledger et al., 1992Go; Croxatto et al., 1993Go, Cameron et al., 1995Go). We have previously reported that 90% of women who received 2 or 5 mg mifepristone per day were anovulatory and amenorrhoeic in spite of the fact that the levels of estradiol (E2) remained within the mid follicular range (Brown et al., 2002Go). Concern has been expressed that the endometrium would undergo hyperplastic or even malignant changes due to the continued exposure to unopposed estrogen (Murphy et al., 1995Go). However, studies in monkeys with mifepristone and other antigestogens have shown no evidence of endometrial hyperplasia (Van Vem et al., 1989Go; Ishwad et al., 1993Go). Rather there is a dose-related decrease in endometrial thickness associated with a reduction in mitotic index (Slayden et al., 1998Go). In a previous study, in which women were treated with 2 mg mifepristone per day for 30 days, we demonstrated a persistent proliferative endometrium which displayed a reduction in mitotic activity in spite of prominent markers of cell divisions such as Ki67 (Cameron et al., 1996Go). The molecular basis for this antiproliferative action of antigestogens is not clear but it has been suggested that there is an arrest of the cell cycle (Neulen et al., 1995Go; Chwalisz et al., 2000Go).

Before daily mifepristone is developed as a novel estrogen-free contraceptive pill is it important to demonstrate that it has no potentially harmful effects on the endometrium. In a previous paper, describing the endocrine profiles, we reported no obvious hyperplastic or malignant changes in the histological examination of endometrial biopsies from 90 women who had taken either 2 or 5 mg mifepristone per day for 120 days (Brown et al., 2002). The present paper reports detailed histological appearances of the endometrium in these women and relates the findings to the endocrine environment.


    Materials and methods
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
Ninety-eight healthy volunteers with regular menstrual cycles (25–35 days) aged 18–40 years were recruited for the study (58 women in Edinburgh and 40 women in Shanghai). Studies were approved by local ethical committees (Institutional Review Board) at both centres. All women gave written informed consent before being enrolled and were screened before entering the study. Screening included routine physical and gynaecological examination and measurement of height, weight, blood pressure and pulse. Blood samples were collected for measurement of full blood count, urea and electrolytes, liver function tests, random glucose, prolactin, lipids and cortisol. hCG was also measured to exclude pregnancy before entering the trial.

Subjects were studied for one pre-treatment cycle, for four cycles (120 days) of treatment, and for one post-treatment cycle. Subjects were randomly allocated to receive 2 or 5 mg of mifepristone daily for the 120 treatment days. The randomization was achieved from a table of random numbers and stratified by dose in groups of 10. The daily doses were issued in pre-packed identical bottles containing, either 2x1 mg mifepristone plus placebo, or 1x5 mg mifepristone plus two placebos. The investigator and subjects were both blinded to the dose that each subject was taking.

All 40 subjects in Shanghai and 10 women in Edinburgh used this method as the sole contraceptive. The remaining 40 women in Edinburgh were either not sexually active, using barrier contraceptives or were surgically sterilized.

Assessment of endometrial development
Endometrial biopsies were collected using a Pipelle endometrial sampling device (Prodimed, Neuilly-en-Thelle, France). Samples were collected in the late follicular phase of the pre-treatment cycle (day 12), after 60 days of mifepristone treatment, and after 120 days of treatment. Specimens were fixed in 100% neutral formalin, processed and embedded in paraffin wax. Sections (5 µm) were cut and then stained with haematoxylin and eosin. Histological assessment and semi-quantative mitotic scores of both glands and stroma were performed by two observers, who were blinded to sample timing and dose group. Mitotic score was assessed semiquantitatively as follows.

Glands: 0 = no mitoses seen per 20 intact gland profiles; 1 = 1–2 mitoses per 20 glands; 2 = 3–5 mitoses per 20 glands; 3 = >5 mitoses per 20 glands. Mitoses were only counted in intact glands, not in surface epithelium or disrupted strips. In specimens with <20 gland profiles, serial sections were assessed, or 10 glands assessed and the result doubled. Specimens with <10 glands were considered inadequate for assessment.

Stroma: 0 = no mitoses seen in five high power microscope fields (HPF) (x400); 1 = 1 or 2 mitoses per 5 HPF; 2 = 3–5 mitoses per 5HPF; 3 = >5 mitoses per five HPF.

Ki67 immunostaining
Ki67 immunostaining was detected using a mouse monoclonal primary antibody (NCL-Ki67-MM1, Novo Castra, Newcastle, UK) with an avidin–biotin–horse-radish peroxidase (ABC) detection system. Antigenic sites were unmasked using the microwave method of retrieval. Endogenous peroxidase was blocked by placing the slides in 3% hydrogen peroxide for 10 min at room temperature. A non-immune block was then performed by incubating the slides with normal horse serum for 20 min. The sections were then incubated with primary antibody (dilution 1:100) for 60 min at 37°C. The biotinylated secondary antibody and the ABC complex were then each applied for 30 min at room temperature. Diaminobenzidine tetrachloride was used to visualize the reaction, prior to counterstaining with haematoxylin. The number of positively stained cells (staining intensity) in glands and stroma viewed at x40 magnification were scored semi-quantitatively by two observers who were blinded to sample timing and dose group (score: 0 = no staining; 1 = mild staining; 2 = moderate staining; 3 = intense staining).

Transvaginal sonography
Endometrial thickness was measured by transvaginal sonography at the time of endometrial biopsy on day 12 of the pre-treatment cycle and after 60 and 120 days of treatment as previously reported (Brown et al., 2002Go). Measurement was performed in the sagittal plane, from one basal layer to the other, the diameter of any luminal fluid being subtracted from the measurement as previously reported.

Assessment of menstrual bleeding pattern
Each subject kept a menstrual diary for the duration of the study. Each day was classified as no bleeding, spotting, normal bleeding, or heavy bleeding.

Assessment of ovarian function
Estrone glucuronide (E1G) and pregnanediol glucuronide were measured in samples of early morning urine and the results expressed per mol of creatinine (Cr) as previously reported (Brown et al., 2002Go). E2 and progesterone were measured by radioimmunoassay in samples of plasma collected at the time of endometrial biopsies.

Ovarian follicular activity during treatment was compared to the baseline ovarian activity shown in the pre-treatment cycle, and each subject was categorized according to the following criteria: (i) total suppression, E1G throughout treatment is <50% above the mean baseline; (ii) partial suppression, E1G raised >=50% above the mean baseline on 1 or 2 consecutive weeks; (iii) continued follicular activity, E1G raised >=50% above the mean baseline on at least two separate occasions. Ovulation was deemed to occur if the excretion of pregnanediol glucuronide was >0.5 mmol/mol Cr and was >=3-fold higher than that in the preceding week.

Statistical methods
Statistical analysis was carried out using SPSS (SPSS, Inc., Chicago, IL, USA) and Excel 97 (Microsoft Corporation). Wilcoxon signed rank, Kruskal–Wallis and paired t-test were used to compare hormonal assays, endometrial thickness and endometrial scores at various time-points.


    Results
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
In Edinburgh a total of 58 women were recruited to the study. Eight women withdrew prior to completion of the study. Two women withdrew due to gastric upset, one due to irregular bleeding, and one was withdrawn by the investigator after she became hypertensive. Four women withdrew due to personal reasons. Thus 50 women (mean ± SEM age: 30.3 ± 0.7 years; height: 163.3 ± 1.1 cm) in Edinburgh completed the study and their results are available for analysis (26 receiving 2 mg daily and 24 receiving 5 mg daily). In Shanghai, all 40 recruited women (age: 32.0 ± 0.5 years; height: 164.5 ± 0.6 cm) completed the study. Further details have been published previously (Brown et al., 2002Go).

Ovarian function
Subjects in Edinburgh and Shanghai differed in the endocrine response to treatment and the results are therefore considered separately. We have previously reported the details of hormone profiles in blood and urine (Brown et al., 2002Go). In the present paper the mean excretion of E1G was calculated for two treatment phases, that is days 0–60 and days 60–120 of treatment, to give an indication of estrogen exposure in the 2 months preceding the endometrial biopsies. The mean excretion of E1G was compared to that excreted in the follicular (i.e. weeks 1 and 2) phase of the pre-treatment cycle. In Edinburgh women the mean excretion of E1G during both treatment phases was not different from that excreted in the follicular phase of the pre-treatment cycle (Figure 1). Women in Edinburgh showed varying levels of follicular activity, with most women showing only partial suppression of ovarian activity or continued follicular activity. Most episodes of follicular activity, however, did not appear to continue through to ovulation, with only 10 ovulatory episodes identified by biochemical assay in the 104 months of treatment with 2 mg daily, and five ovulatory episodes in the 96 months of treatment with 5 mg daily.



View larger version (21K):
[in this window]
[in a new window]
 
Figure 1. Urinary excretion of estrone glucuronide (E1G) in women in Edinburgh and Shanghai before, during, and after taking mifepristone 2 or 5 mg/day. Results are expressed as mean E1G /creatinine ratio ± SEM. There are no significant differences between the E1G /Cr ratio after treatment in either Edinburgh or Shanghai.

 
The E1G excretion in Shanghai women was significantly lower than in Edinburgh. In both dose groups, mean excretion during both phases of treatment was similar to that in the pre-treatment follicular phase (Figure 1). In contrast to the Edinburgh subjects, follicular activity was suppressed in the majority of women in Shanghai and ovulatory episodes were fewer with two episodes out of 80 months of treatment with 2 mg daily, and one ovulation in the 80 months of treatment with 5 mg daily.

Menstrual bleeding patterns
In Edinburgh, 17/26 women (65%) receiving 2 mg daily were amenorrhoeic, as were 21/24 (88%) receiving 5 mg. The mean number of days bleeding during treatment was 4.4 days and 0.6 days for the 2 and 5 mg groups respectively. In Shanghai, 18/20 women (90%) in both dose groups were amenorrhoeic during treatment, with mean number of days bleeding of 0.4 and 0.7 respectively.

All subjects reported a menstrual bleed within 3 weeks of stopping mifepristone. There was a trend for this bleed to be longer than the normal pre-treatment bleed, but this increase was only significant in the group in Edinburgh treated with 2 mg daily (5.2 versus 10 days, P < 0.05).

Endometrial development
Transvaginal sonography
In Edinburgh women, there was a trend for the endometrial thickness to increase during treatment with mifepristone (Figure 2). This increase was significant by the end of 120 days of treatment in the 2 mg group (P < 0.01) and by 60 days of treatment in the 5 mg group (P = 0.015). Plasma E2 was measured at the time of endometrial assessment, and showed that normal pre-menopausal levels of E2 were maintained during treatment (2 mg, 436 ± 56 and 403 ± 61 pmol/l days 60 and 120; 5 mg, 438 ± 82 and 327 ± 56 pmol/l).



View larger version (25K):
[in this window]
[in a new window]
 
Figure 2. Endometrial thickness as measured by ultrasound in women before (control day 12), and 60 and 120 days after treatment with mifepristone 2 or 5 mg per day (mean ± SEM). Significantly different from control, *P < 0.02, **P < 0.001.

 
In contrast to the Edinburgh subjects, the Shanghai women showed a significant thinning of the endometrium, which was significant by 60 days of treatment in both dose groups (P > 0.001) (Figure 2). Nonetheless, although mean E2 levels were lower in the Shanghai subjects, levels within the pre-menopausal range were maintained (2 mg, 112 ± 6 and 127 ± 3 pmol/l days 60 and 120; 5 mg, 119 ± 5 and 137 ± 5 pmol/l).

In neither centre was there any correlation between endometrial thickness and plasma E2 level.

Histological appearances
Pre-treatment endometrial biopsies were collected on day 12 of the pre-treatment cycle. Samples were collected from all subjects, but three samples in the Edinburgh women (two in the 2 mg group, and one in the 5 mg group) were too small for assessment. The remaining 47 pre-treatment samples in Edinburgh all showed normal proliferative phase endometrium, as expected (Figure 3a).



View larger version (125K):
[in this window]
[in a new window]
 
Figure 3. Sections of endometrium before (control) and after treatment (up to 120 days) with mifepristone, stained with haemotoxylin and eosin; x10 (inserts x40). (a) Day 12 proliferative control. (b) After 120 days of treatment with mifepristone 2 mg/day showing small atrophic glands in compacted stroma. (c) After 120 days of treatment with mifepristone 5 mg/day (subject 332). Note cystic dilatation of inactive endometrial glands with minimal stratification (insert: glandular epithelium at higher power). (d) After 60 days of treatment with mifepristone 5 mg/day showing complex hyperplasia (subject 332 same as c). (e) Well-marked secretory changes after 120 days of treatment with mifepristone 5 mg/day in a subject who showed biochemical evidence of ovulation (plasma progesterone 88.8 nmol/l on day of biopsy). (f) After 60 days of treatment with mifepristone 2 mg/day in subject 132 who showed biochemical evidence of ovulation. Note absence of secretory changes.

 
Ten pre-treatment samples in the Shanghai women (six and four subjects in each dose group) were unassessable due to an inadequate sample. The remaining samples were consistent with proliferative phase endometrium, other than two samples which showed some features of early secretory activity.

All subjects had further endometrial samples collected on days 60 and 120 of treatment. A few samples were unassessable due to inadequate sample size. In Edinburgh women, the most common appearance at both treatment time-points, in both dose groups, was of proliferative phase type endometrium, with straight or coiled glands in a compact stroma (Table I, Figure 3b).


View this table:
[in this window]
[in a new window]
 
Table I. Summary of histological appearances
 
Around one-third of the treatment samples in Edinburgh women showed dilated but inactive glands, with minimal stratification (Figure 3c). In these samples mitoses were infrequent, and there was no cytological atypia. One Edinburgh subject in the 5 mg dose group showed some features of complex hyperplasia in the sample collected on day 60 with crowded ‘back-to-back’ glands and nuclear stratification, but no cytological atypia (Figure 3d). Reassuringly, by day 120 of treatment, this subject’s endometrium showed an inactive appearance (Figure 3c).

One subject showed a secretory type endometrium on the day 120 sample (Figure 3e) correlating with biochemical evidence of preceding ovulation, with a rise in urinary pregnanediol glucuronide to 0.57 mmol/mol Cr, and an elevated plasma progesterone of 88.8 nmol/l. Two subjects in the 2 mg group were also in the ‘luteal’ phase following biochemical evidence of ovulation, but did not show any secretory change, suggesting asynchrony between the endocrine cycle and endometrial development (Figure 3f).

In the Shanghai women, the endometrial appearances were similar to the Edinburgh women, with the majority of women showing a proliferative type endometrium (Table I). The inactive cystic dilatation appearance was less commonly seen in the Shanghai women. Four subjects in Shanghai (one at each time-point in each dose group) showed some features of secretory change, but none of these subjects had biochemical evidence of preceding ovulation.

There was no relationship between histological appearance and endometrial thickness or degree of ovarian activity. Both proliferative and cystic changes were represented at all categories of ovarian activity. Three of the four subjects who showed secretory features were classified as showing total suppression of ovarian follicular activity.

Markers of proliferation
Mitotic figures were counted in sections of both control and treatment biopsies (Figure 4a and b). By day 60 there was a significant reduction in mitotic index in both dose groups and this was maintained after 120 days of treatment (Table II). This reduction in mitosis was confirmed by analysis of the Ki67 immunostaining which was performed on a sub-group of 40 subjects (20 in each dose group) in Edinburgh (Table II, Figure 4c and d). The samples from Shanghai subjects were too scanty to allow assessment of proliferation markers.



View larger version (136K):
[in this window]
[in a new window]
 
Figure 4. Sections (x40) of endometrium showing markers of proliferation. (a) Day 12 (proliferative) control cycle stained with haematoxylin and eosin (H&E). Note mitosis in gland (arrow). (b) H&E section of endometrial biopsy after 120 days of treatment with 5 mg mifepristone per day showing occasional mitosis in atrophic glands and densely packed stroma. (c) Day 12 (proliferative) control cycle showing Ki67 nuclear staining (brown) in glands. (d) Section of endometrial biopsy after 60 days of treatment with 5 mg mifepristone per day showing infrequent scattered Ki67 nuclear staining.

 

View this table:
[in this window]
[in a new window]
 
Table II. Markers of endometrial proliferation in Edinburgh subjects
 
Contraceptive efficacy
All 40 women in Shanghai and 10 of the women in Edinburgh used mifepristone as the sole contraceptive method during the 4 months of treatment, giving a total of 200 months of exposure to the risk of pregnancy. As previously reported, no pregnancies occurred (Brown et al., 2002Go).


    Discussion
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
This study confirms that in addition to a profound effect on ovarian function, mifepristone in low daily doses has significant suppressive effects on the endometrium. The commonest histological appearance of proliferation was not unexpected, as most of the women failed to ovulate throughout the study. A significant number of biopsies showed an unusual cystic dilation of the glands surrounded by compact dense stroma. The appearance superficially resembled cystic glandular hyperplasia which occurs as a consequence of prolonged exposure of the endometrium to unopposed estrogen (Kurman and Norris, 1994Go). However, in contrast to the latter condition, the glands were mostly lined by a single layer of atrophic glandular epithelium. A single subject (332), after 60 days of treatment with 5 mg per day, showed some features of hyperplasia that had reverted to inactive cystic glands after 120 days (Figure 3c and d). The cause of this unusual cystic change in the endometrium is unknown. It is unlikely to be solely due to the effect of high levels of unopposed estrogen because it occurred in some women, e.g. in Shanghai where there was complete suppression of ovarian follicular development and estrogen levels were very low. As discussed in our previous paper, the levels of estrogen both before and after mifepristone were lower in women in Shanghai (Brown et al., 2002Go).

A striking feature of the endometrium obtained from women after treatment was the reduced indices of proliferation. There was a significant reduction in both mitotic index and Ki67 immunostaining on treatment as compared to endometrium on day 12 of control cycles. In a previous study we reported a reduction in mitotic index after 21 days of treatment with mifepristone although the Ki67 immunostaining was increased (Cameron et al., 1996Go). We suggested that initially mifepristone may block completion of the cell cycle so that Ki67 protein persists for some time after cell division has been arrested. Our present observations are similar to those reported in the experimental primate in which it was shown that mifepristone and other antigestogens inhibit the ability of estrogen to induce proliferation of the endometrium (Ishwad et al., 1993Go; Heikinheimo et al., 1996Go; Slayden et al., 1998Go). The molecular basis for this ‘antiproliferative’ action of antigestogens is not clear, but evidence from transfected cells suggests that it is mediated through the progesterone receptor (Chwalisz et al., 2000Go).

It is of interest that in a recent study (Slayden et al., 2001Go) the androgen receptor was elevated by antiprogestin treatment, particularly in the glandular epithelium. Such an elevation of androgen receptor could result in enhanced androgen action in the endometrium (Brenner et al., 2002Go). Androgens suppress estrogen-dependent endometrial proliferation. Thus elevated expression of the androgen receptor following antiprogestin administration is a further candidate for mediation of the proliferative effects of antiprogestin treatment.

Concern has been expressed previously that long-term use of antigestogens may lead to the risk of endometrial hyperplasia and possible malignancy due to the exposure of the endometrium to the effects of unopposed estrogen (Murphy et al., 1995Go). In a study in which women with pelvic endometriosis were treated with 50 mg mifepristone per day for 6 months, there was evidence of endometrial hyperplasia and numerous mitotic features. In a case report of a young woman treated with 400 mg mifepristone per day for 6 months because of inoperable Cushings’ syndrome, massive simple hyperplasia of the endometrium developed, which reversed after stopping therapy (Newfield et al., 2001Go).

In both these reports, the daily dose of mifepristone was considerably higher than in our study and was sufficient to suppress ovarian function and ovarian E2 secretion. The antiglucocorticoid action of mifepristone at this dose was sufficient to stimulate an increased secretion of adrenal androgens such as androstenedione, which could be a precursor for extraglandular synthesis of estrogen in peripheral tissues including the endometrium (Kettel et al., 1991Go; Kettel et al.,1996Go). Thus the absence of endometrial hyperplasia in our study and others could be due to the absence of significant effect on the pituitary–adrenal axis of daily doses of <10 mg.

Although ovulation occurred in a minority of cycles, there appeared to be asynchrony between the ovarian cycle and endometrial development. Of the 15 women in Edinburgh who showed biochemical evidence of ovulation at least once, only four reported a menstrual bleed subsequent to the rise in pregnanediol. In two subjects who had an endometrial sample collected (fortuitously) in the apparent luteal phase following an episode of presumed ovulation, there was no evidence of secretory change (Figure 3e). Conversely, four subjects from Shanghai who showed some evidence of secretory endometrium did not have biochemical evidence of ovulation. It has been previously reported in a study using 1 mg mifepristone per day that secretory changes can occur without preceding ovulation. It has been suggested that when endogenous progesterone levels are low, as in post-menopausal women, mifepristone can exert an agonist action (Gravanis et al., 1985Go). These results demonstrate that mifepristone, in addition to disturbing ovarian function, has a direct effect on the endometrium which makes it unlikely that a pregnancy could become established even if ovulation occurred.

The significant increase in the thickness of the endometrium as measured by pelvic ultrasound of women in Edinburgh is similar to that reported after treatment of women in Chile with mifepristone at a dose of 1 mg/day (Croxatto et al., 1998Go). It is likely that much of this apparent thickness is due to the collection of fluid within the lumen of dilated glands. However some women who showed thickening of the endometrium on ultrasound had no evidence of cystic glandular formation in the biopsy. It seems unlikely therefore that the apparent thickening of the endometrium is due to cystic glands in all women. In Shanghai, where there was an overall reduction in endometrial thickness during treatment, ovarian function and estrogen excretion were profoundly suppressed. The fact that there was no correlation between individual plasma E2 levels and endometrial thickness makes it unlikely that hormonal factors alone are responsible for the variation in endometrial structure. Rather the interaction between dose, effect on ovarian cyclicity, variation in individual and possible genetic sensitivity determine the response of the endometrium to mifepristone.

The basal excretion of E1G and concentration of E2 were lower in women in Shanghai than in Edinburgh. As previously discussed in detail, this difference is unlikely to be due to methodological factors and reflects a genuine difference in secretion and/or metabolism of estrogens between Chinese and Caucasian women (Key et al., 1990Go; Aldercreutz et al., 1994Go; Brown et al., 2002Go). Ovarian function and endometrial development were more easily suppressed by low doses of mifepristone in Shanghai women, suggesting that there may be ethnic differences in the metabolism of steroid hormones such as that suggested in men (Anderson and Wu, 1996Go). Whether these differences are genetic or related to diet or other aspects of life style is not known.

In conclusion, the present study demonstrates that mifepristone in a dose of 2 or 5 mg per day for 120 days has effects that are likely to be contraceptive. There is suppression of ovulation and menstruation in the majority of cycles and asynchrony between endometrial and ovarian activity. After 120 days of treatment there are no features of endometrial hyperplasia and markers of endometrial proliferation are suppressed. Preliminary data confirm that in 50 women the dose is contraceptive. By inducing amenorrhoea, the method could have considerable health benefits by reducing the considerable morbidity associated with menstruation. Recent surveys suggest that a pill that abolished periods would be popular with a significant number of women in Western Europe (Den Tonkelaar and Oddens, 1999Go) and some developing countries (A.F.Glasier et al., unpublished data). It is now time to mount a long-term efficacy and safety study with the prospect of realizing the potential of mifepristone as a novel ‘estrogen-free’ daily contraceptive pill.


    Acknowledgements
 
We are grateful to Mrs Theresa Henderson and Dr Nitish Narvekar for help in histology and preparing photomicrograph plates. Mrs Ann Mayo and Dr Karen Smith in Edinburgh helped in running and co-ordinating the study. The study was supported by a grant to the Contraceptive Development Network from the Department for International Development and the Medical Research Council, UK (G9523250). The mifepristone was supplied through WHO Special Programme of Research, Development and Research Training in Human Reproduction (Project No. 96503).


    References
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
Aldercreutz, H., Gorbach, S.L., Goldin, B.R., Woods, M.N., Dwyer, J.J. and Hamalainen, E. (1994) Estrogen metabolism and excretion in Oriental and Caucasian women. J. Natl. Cancer Inst., 86, 1076–1082.[Abstract]

Anderson, R.A. and Wu, F.C.W. (1996) Comparison between testosterone enanthate-induced azoospermia and oligospermia in a male contraceptive study III: higher 5{alpha}-reductase activity in oligospermic responders during testosterone enanthate administration. Clin. Endocrinol., 81, 902–908.

Baird, D.T. (2001) Antigestogens: the holy grail of contraception. Reprod. Fertil. Dev., 13, 1–6.[ISI][Medline]

Baird, D.T. and Glasier, A.F. (1999) Science, medicine and the future: contraception. Br. Med. J., 319, 969–972.[Free Full Text]

Brenner, R.M., Slayden, O.D. and Critchley, H.O.D. (2002) Antiproliferative effects of progesterone antagonists in the primate endometrium; a potential role for the androgen receptor. Reproduction, 124, 167–172.[Abstract/Free Full Text]

Brown, A., Cheng, L., Lin, S. and Baird, D.T. (2002) Daily low dose mifepristone has contraceptive potential by suppressing ovulation and menstruation: a double blind randomised control trial of 2 and 5mg per day for 120 days. J. Clin. Endocrinol. Metab., 87, 63–70.[Abstract/Free Full Text]

Cameron, S.T., Thong, K.J. and Baird, D.T. (1995) Effect of daily low dose mifepristone on the ovarian cycle and on the dynamics of follicle growth. Clin. Endocrinol., 43, 407–414.[ISI][Medline]

Cameron, S.T., Critchley, H.O.D., Thong, K.J., Buckley, C.H., Williams, A.R. and Baird, D.T. (1996) Effects of daily low dose mifepristone on endometrial maturation and proliferation. Hum. Reprod., 11, 2518–2526.[Abstract]

Chwalisz, K., Brenner, R.M., Fuhrmann, U.U., Hess-Stump, H. and Elger, W. (2000) Antiproliferative effects of progesterone antagonists and progesterone receptor modulators on the endometrium. Steroids, 65, 741–751.[CrossRef][ISI][Medline]

Croxatto, H.B., Salvatierra, A.M., Croxatto, H.D. and Fuentealba, B. (1993) Effects of continuous treatment with low dose mifepristone throughout one menstrual cycle. Hum. Reprod., 8, 201–207.[Abstract]

Croxatto, H.B., Kovacs, L., Massai, R., Resch, B.A., Fuentealba, B., Salvatierra, A.M., Croxatto, H.D., Zalanyi, S., Viski, S. and Krenacs, L. (1998) Effects of long term low-dose mifepristone on reproductive function in women. Hum. Reprod., 13, 793–798.[Abstract]

Den Tonkelaar, I. and Oddens, B.J. (1999) Preferred frequency and characteristics of menstrual bleeding in relation to reproductive status, oral contraceptive use and hormone replacement therapy use. Contraception, 59, 357–362.[CrossRef][ISI][Medline]

Fraser, I.S. (2000) Progestogen-only contraception. In Glasier, A.F. and Gebbie, A. (eds), Handbook of Family Planning and Reproductive Healthcare. Churchill Livingstone, Edinburgh, pp. 77–103.

Gravanis, A., Schaison, G., George, M., DeBrux, J., Satyaswaroop, G., Baulieu, E-E. and Rubel, P. (1985) Endometrial and pituitary responses to the steroid antiprogestogen RU486 in post menopausal women. J. Clin. Endocrinol. Metab., 60, 156–163.[Abstract]

Heikinheimo, O.J.G., Hsiu, J.G., Gordon, K., Kim, S., Williams, R.F., Gibbons, W.E. and Hodgen, G.D. (1996) Endometrial effects of RU486 in primates—antiproliferative action despite signs of oestrogen action and increased cyclin-B expression. J. Steroid Biochem. Mol. Biol., 59, 179–190.[CrossRef][ISI][Medline]

Ishwad, P.C., KatKam, R.R., Hinduta, I.N., Chwalisz, K., Elger, W. and Puri, C.D. (1993) Treatment with a progesterone antagonist ZK 98,299 delays endometrial development without blanking ovulation in Bonnet Monkeys. Contraception, 48, 57–70.[CrossRef][ISI][Medline]

Kettel, L.M., Murphy, A.A., Mortola, J.F., Liu, J.H., Ulmann, A. and Yen, S.S.C. (1991) Endocrine responses to long-term administration of the antiprogesterone RU486 in patients with pelvic endometriosis. Fertil. Steril., 56, 402–407.[ISI][Medline]

Kettel, L.M., Murphy, A.A., Morales, A.J., Baulieu, E-E. and Yen, S.S.C. (1996) Treatment of endometriosis with antiprogesterone mifepristone (RU486). Fertil. Steril., 65, 23–28.[ISI][Medline]

Key, T.J.A., Chen, J., Wang, D.Y., Pike, M.C. and Boreham, J. (1990) Sex hormones in women in rural China and in Britain. Br. J. Cancer, 62, 631–636.[ISI][Medline]

Kurman, R. and Norris, H.J. (1994) Endometrial hyperplasia and related cellular changes. Blaustein’s Pathology of the Female Genital Tract, 4th edn. Springer-Verlag, New York, p. 412.

Ledger, W.L., Sweeting, V.M., Hillier, H. and Baird, D.T. (1992) Inhibition of ovulation by low dose mifepristone (RU486). Hum. Reprod., 7, 945–950.[Abstract]

Murphy, A.A., Kettel, L.M., Morales, A.J., Roberts, V., Parmley, T. and Yen, S.S.C. (1995) Endometrial effects of long-term low-dose administration of RU486. Fertil. Steril., 63, 761–766.[ISI][Medline]

Neulen, J., Williams, R.F., Breckwoldt, M., Chwalisz, K., Baulieu, E-E. and Hodgen, G.D. (1995) Non-competitive anti-oestrogen actions of progesterone antagonists in primate endometrium: enhancement of oestrogen and progesterone receptors with blockade of post-receptor proliferative mechanisms. Hum. Reprod., 11, 1533–1537.[Abstract/Free Full Text]

Newfield, R.S., Spitz, I.M., Isacson, C. and New, M.I. (2001) Long-term mifepristone (RU486) therapy resulting in massive benign endometrial hyperplasia. Clin. Endocrinol. (Oxf.), 54, 399–404.[CrossRef][ISI][Medline]

Sitruk-Ware, R. (2000) Approval of mifepristone (RU486) in Europe. Zentrabl. Gynakol., 122, 241–247.[Medline]

Skegg, D.C.G. (2000) Third generation oral contraceptives. Br. Med. J., 7255, 190–191.[CrossRef]

Slayden, O.D., Zelinski-Wooten, M.B., Chwalisz, K., Stouffer, R.L. and Brenner, R.M. (1998) Chronic treatment of cycling rhesus monkeys with low doses of the antiprogestin ZK137316: morphometric analysis of the uterus and oviduct. Hum. Reprod., 13, 269–277.[CrossRef][Medline]

Slayden, O.D., Nayak, N.R., Burton, N.R., Chwalisz, K., Cameron, S.T., Critchley, H.O.D., Baird, D.T. and Brenner, R.M. (2001) Progesterone antagonists increase androgen receptor expression in the rhesus macaque and human endometrium. J. Clin. Endocrinol. Metab., 86, 2668–2679.[Abstract/Free Full Text]

Ulmann, A. (2000) The development of mifepristone: a pharmaceutical drama in three acts. J. Am. Med. Women’s Assoc., 55 (Suppl. 3), 117–120.

Van Vem, J.F.H.M., Hsiu, J.G., Chillik, C.F., Danforth, D.R., Ulmann, A., Baulieu, E-E. and Hodgen, G.D. (1989) Contraceptive potential of RU486 by ovulation inhibition: Pituitary versus ovarian action with blockade of estrogen induced endometrial proliferation. Contraception, 40, 171–183.[ISI][Medline]

Submitted on July 1, 2002; resubmitted on August 27, 2002. accepted on October 1, 2002