Marions, Lena MD, PhD; Hultenby, Kjell PhD; Lindell, Ingrid; Sun, Xiaoxi MD; Ståbi, Berit; Danielsson, Kristina Gemzell MD, PhD
Emergency contraception is any drug or device used after unprotected sexual intercourse to prevent pregnancy. Any women at risk of an unwanted pregnancy may need these methods occasionally. Millions of unwanted pregnancies could be averted if emergency contraception was widely accessible.1 There are two hormonal methods available today. The Yuzpe method consists of two tablets of ethinylestradiol (0.1 mg) and levonorgestrel (0.5 mg) oral contraceptive given twice, 12 hours apart. The other is 0.75 mg of levonorgestrel alone, also administered twice with an interval of 12 hours. Levonorgestrel alone has become the method of choice because it has been proven to be more effective and has fewer side effects than the Yuzpe method.2,3
The antiprogestin mifepristone is a possible future alternative. A single dose of 600 mg of mifepristone is at least as effective as the Yuzpe method, and the treatment is associated with significantly lower side effects.4,5 A reduction of the dose from 600 mg to 10 mg did not influence efficacy, but the proportion of women with menses delay of more than 7 days decreased from 36% to 18%.6
To be effective, postcoital treatment should inhibit or delay ovulation or inhibit implantation. If sexual intercourse takes place shortly before or at the time of ovulation, and treatment is given after ovulation, an inhibitory effect on embryo development or implantation is necessary to prevent pregnancy. How levonorgestrel and mifepristone prevent pregnancy in the doses used for emergency contraception is not known. Levonorgestrel, 0.75 mg, in repeated doses has been shown to suppress proliferative activity of the endometrium when administered during the follicular phase. It suppresses ovarian function when administered in the late follicular phase and around the time of ovulation. No significant endometrial changes were detected when levonorgestrel was administered in the secretory phase.7
The effect of mifepristone varies depending on when the drug is administered and the dose given. Administration of mifepristone during the follicular phase of the menstrual cycle delays the estrogen rise and the luteinizing hormone (LH) surge, but has no effect on endometrial development during the implantation period. The situation is quite different when mifepristone is given immediately after ovulation. In this case, a sufficient dose of mifepristone will significantly inhibit endometrial development.8 It will also inhibit the expression of endometrial markers of endometrial receptivity such as leukema inhibitory factor, integrins, and cyclooxygenase.9–11
The purpose of the present study was to evaluate the effect of two regimens used for emergency contraception 0.75 mg of levonorgestrel given twice and 10 mg of mifepristone when given just before or after ovulation. We studied the effect on ovarian function, endometrial development, and markers of endometrial receptivity.
MATERIALS AND METHODS
Twelve healthy fertile women volunteered for the study during a 12‐month enrollment period. None had used steroidal contraception or an intrauterine device for a minimum of 3 months before the study. One woman was sterilized, and the others were advised to use barrier methods for contraception during the study. All participants were asked to keep daily records of side effects and bleeding.
The study included one control and two treatment cycles with a “wash‐out” cycle in between. The participants were consecutively allocated to one of two treatment groups without randomization (for demographics data, see Table 1). They received either 10 mg of mifepristone as a single dose (group A, n = 6), or 0.75 mg of levonorgestrel twice, 12 hours apart (group B, n = 6), before ovulation in the first treatment cycle and after ovulation in the second treatment cycle. Both participants and investigators were aware of the treatment given. In all control cycles and in the women treated before ovulation, the size of the growing follicle was measured by transvaginal ultrasound every day or every other day starting on cycle day 10 until the expected time of ovulation. A 7.5‐MHz vaginal probe (SonoLine; Siemens Medical Systems Inc., Issaquah, WA) was used, and at least two diameters of the follicle were measured to give a mean value. In the women who obtained treatment before ovulation, the treatment was given when the follicle size corresponded to that found 2 days before the LH peak in the control cycle. When the day of the LH peak measured with enzyme immunoassay was compared with the estimated LH peak determined by the self‐test, it was discovered in two cycles that the treatment with mifepristone had been given on the day of the LH peak instead of 2 days before. In these two women, new treatment cycles were performed.
Treatment after ovulation was given on day LH + 2. The LH surge in the urine was detected using a self‐test (Clear Plan; Searle Unipath, Bedford, UK) twice daily from cycle day 10. In control and treatment cycles, participants were also asked to collect daily morning samples for later analyses of LH, estrone, and pregnanediol glucuronide. In the control and treatment cycles, an endometrial biopsy was taken 6–8 days after the LH peak. In the women treated before ovulation, the day of biopsy was calculated based on follicular size and day of treatment. The biopsies were obtained from the uterine fundus using a Randall curette (Dimeda, Duttlingen, Germany) without prior cervical dilatation.
Two‐way analysis of variance (ANOVA) was used for the testing of differences between the means of LH and glucuronide values obtained in individual study groups. The differences between women as the second variable of ANOVA were not further evaluated. Differences in morphometric data and immunohistochemistry staining were evaluated by using the two‐tailed Wilcoxon signed rank test. Results were considered significant at P < .05. The study was approved by the ethics committee at the Karolinska Hospital. Informed consent was obtained from all the participants before they were included in the study.
Daily morning urine samples were analyzed for estrone, pregnanediol glucuronide, and LH using enzyme immunoassay.12 The hormones were expressed in nmol per mmol of creatinine for estrone and pregnanediol and in mIU per mmol of creatinine for LH.13 For creatinine analysis, a commercial kit (Sigma Diagnostics, St. Louis, MO) was used.
One part of the biopsy was immediately fixed in Bouin solution and used for light microscopic examination after embedding in paraffin and staining with hematoxylin. Morphometric analyses were performed measuring the number of glands per mm2, number of glandular and stromal mitosis per 1000 glandular or stromal cells, respectively, glandular diameter (μm), glandular epithelial height (μm), cells with basal vacuolization per 1000 glandular cells, as well as the number of pseudostratified cells and the degree of stromal edema.14
Part of the biopsies were rinsed in physiologic sodium chloride followed by immersion into 2% glutaraldehyde + 0.5% paraformaldehyde in 0.1 M of sodium cacodylate buffer + 0.1 M of sucrose (pH 7.4) at room temperature and refrigerated until further processing. Biopsies were washed in distilled water and placed in 70% ethanol for 30 minutes, 95% ethanol for 30 minutes, and absolute ethanol for 30 minutes, and placed in acetone. Biopsies were then dried in a critical point dryer (CPD 010; Balzers Union AG, Balzers, Liechtenstein) with carbon dioxide. After drying, the specimens were mounted on aluminium stubs and coated with 20‐nm platinum (Polaron, Watford, England). The specimens were analyzed in a Jeol JSM‐820 scanning electron microscope (Joel, Tokyo, Japan) at 15 kV. Digital images were captured and processed in a Kontron electronic imaging system KS 400 (Kontron AG, Eching, Germany).
Immunostaining was performed on cryostat sections of the endometrial biopsies. The expression of cyclooxygenase‐1 (COX‐1) and ‐2 (COX‐2) and of integrin α4 and β3 were measured using polyclonal antibodies (Scandinavian Diagnostic Services, Falkenberg, Sweden, and Chemicon International, Temecula, CA). The methods used have been described previously.10,11 Progesterone receptor concentration was measured by using the Abbot PgR‐ICA Monoclonal assay system (Abbot Laboratories, Downers Grove, IL). Positive and negative control labeling was performed using Abbot control slides.
The secretory components of the endometrium were detected by lectin histochemistry using biotinylated Dolichos biflorus agglutinin lectin at a concentration of 5 μg/mL and the Vectastain Elite ABC immunoperoxidase detection system (Vector Laboratories, Burlingame, CA). As a negative control, Dolichos biflorus agglutinin was coincubated with the corresponding carbohydrate ligand at 200‐nM concentration, which completely inhibited the binding.
Immuno‐ and lectin histochemic staining of the frozen sections was evaluated blindly by two independent persons by using a semiquantitative subjective scoring system and a Zeiss light microscope at 200 × magnification. To compensate for uneven staining, observations were made in various fields of the whole section and included luminal, glandular, stromal, and perivascular cells. Cells were assigned a score of 0–3 based on the number of cells specifically stained as follows: 0 (0% stained cells), 1 = weak (less than 25% positive cells), 2 = moderate (25–75% positive cells), or 3 = strong (more than 75% positive cells).
In the control cycles, all women but one showed a normal ovulatory cycle (28–33 days) with a midcycle urinary LH peak. The rise in pregnanediol glucuronide levels in luteal phase failed to occur in this participant, and she was therefore excluded (levonorgestrel group). In all the other control cycles, the differences between the LH levels were highly significant; F values were 19.2 (mifepristone) and 9.7 (levonorgestrel), the probabilities were 2.2 E‐19 and 7.8 E‐11, respectively. The diameter of the leading follicle immediately before ovulation varied between 15 and 21 mm. There was a good agreement (less than 1 day) between the calculated day based on the LH peak in urine determined by enzyme immunoassay and the biologic dating based on morphometric analyses. Positive staining for COX‐1 was found in all biopsies in luminal, glandular, and endothelial cells, whereas in stromal cells the staining was weaker. The COX‐2 immunostaining was expressed in luminal cells, superficial glandular, and endothelial cells but not in stromal cells. Integrin α4 was strongly expressed in the luminal epithelium and to some extent in the glandular and endothelial cells. β3 was strongly expressed in luminal and epithelial cells.
Dolichos biflorus agglutinin lectin staining was very strong in luminal, glandular, and endothelial cells, whereas staining for progesterone receptors was almost negative in glandular cells and weak to moderate in stromal and luminal epithelial cells. Pinopodes in varying developmental stages were found in almost all participants.
When treated before ovulation, two women (out of six) showed considerable elevations of LH levels that were postponed by 2 and 5 days, respectively (Figure 1), in comparison with the ultrasound assessment. (The length of the cycles were prolonged by the same number of days.) However, when testing the differences between LH means on all the cycle days, no significance was found at the 95% significance level as established by an ANOVA (F = 0.67, P = .77). In the group of women who were given the drug 2 days after ovulation, the pattern of LH levels was similar to that seen in the control cycles (F = 12.3, P = 1.6 E‐9). The urinary estrone and pregnanediol glucuronide concentrations in both pre‐ and postovulatory treated cycles did not differ from those in the control cycle (Figure 2A,B). After preovulatory treatment with mifepristone, two women showed suppressed endometrial development with decreased glandular size corresponding to endometrium from the early secretory phase. The expression of COX‐1, COX‐2, integrin α4, progesterone receptors, and Dolichos biflorus agglutinin lectin staining was the same as in the control cycles. In two women, the expression of β3 was decreased in glandular epithelial cells.
When mifepristone was administered after ovulation, early secretory phase endometrium was found in four women. In one woman, the endometrium showed advanced maturation corresponding to LH + 9–10, and in another woman the endometrium did not differ from that expected on the day of biopsy. The downregulation of the progesterone receptor concentration was inhibited in five out of six participants. There was increased staining of epithelial cells from 0 in controls to 2.2 (1–3) after treatment (P < .05, Figure 3), and only the woman with abnormal endometrial development was unaffected. The Dolichos biflorus agglutinin lectin staining was reduced in four women, and the expression of integrin α4, β3, and COX‐2, in two. In the remaining markers, no difference could be detected, but none of the biopsies were found to be completely unaffected by treatment. Pinopodes were found to the same degree as they were in the control cycles after treatment with mifepristone both before and after ovulation.
There was no difference at the 95% significance level between the means of LH measurements during the entire cycle when 0.75 mg of levonorgestrel was administered twice before ovulation (Figure 4), as established by an ANOVA (F = 1.61, P =.10). The length of the menstrual cycle corresponded to that of the control cycles except in one woman in whom the cycle length was shortened by 7 days. The postovulatory treatment with levonorgestrel resulted in a cycle pattern that was similar to that of a control cycle. The urinary excretion of estrone and pregnanediol glucuronide in both pre‐ and postovulatory treated cycles was also similar to that of the control cycles. In three women, normal endometrial development was not different from the control cycle. The amount of tissue was insufficient in one woman, and one had started to bleed at the time of the biopsy.
Except for one participant, in whom a reduced staining of COX‐2 in glandular cells was found, no differences in markers for endometrial receptivity could be demonstrated. The endometrium was out of phase in two women and normal in three women in the group treated with levonorgestrel after ovulation. However, one biopsy out of phase had actually been obtained in the early secretory phase (LH + 5). The expression of progesterone receptors was not affected. Reduced expression of COX‐2 in glandular epithelial cells was found in one participant and in luminal epithelial cells in another participant. Among the remaining markers, no differences could be found. Pinopodes were found in the same degree as they were in the control cycles after treatment with levonorgestrel both before and after ovulation (Figure 5).
Both 0.75 mg of levonorgestrel given 12 hours apart and one dose of 10 mg of mifepristone are effective methods for emergency contraception. The proportion of pregnancies prevented with levonorgestrel compared with the expected number without treatment has been reported to be 85%.3 The same figure has been found for 10 mg of mifepristone even when treatment was extended from 72 to 120 hours after unprotected intercourse.6
Emergency contraception may be used on any day of the menstrual cycle, but it is only when sexual intercourse takes place during a 6‐day period that ends on the estimated day of ovulation that there is a risk of pregnancy.15 Because emergency contraception is administered up to 72 hours after sexual intercourse, the intercourse could occur at or just before ovulation and the treatment given just afterwards. To clarify how emergency contraception prevents pregnancy, we gave mifepristone and levonorgestrel just before and after ovulation in doses shown to be effective in clinical trials.
A number of factors mirror endometrial receptivity, and some were included in this study. Probably most important is the progesterone receptor concentration because many, if not all, endometrial cytokines are progesterone regulated. Dolichos biflorus agglutinin lectin histochemistry illustrates the secretory activity of the endometrium, and COX‐1 and COX‐2 the role of prostaglandins.11 Integrins are cell surface receptors for the extracellular cell adhesion proteins to cytoskeletal components and are involved in blastocyst adhesion.16
The effect of mifepristone on follicular development, LH surge, and ovulation is dependent on time of treatment and dose given. Repeated administration of 25–100 mg daily, for a limited time in mid‐ to late follicular phase delays the estrogen surge, the LH peak, and ovulation.17,18 After cessation of mifepristone treatment, there is a resumption of follicular growth or recruitment of a new follicle, and the following luteal phase is normal.18
Continuous administration of 5 mg or 10 mg of mifepristone per day throughout one menstrual cycle prevents the leading follicle from achieving maturity and from producing the circulating estradiol levels necessary to trigger the LH surge.19,20 A low dose of mifepristone (5 mg) administered for up to 3 days impairs follicular growth. Even a single 5‐mg dose of mifepristone transiently arrests follicular growth.21 The effect is most pronounced when the dominant follicle reaches a diameter of 12 mm, probably because of the acquisition of progesterone receptors by granulosa cells.21 With daily doses of 1 or 2 mg of mifepristone, full follicular growth is obtained, but ovulation fails to occur. Instead unruptured luteinized follicles appear to be associated with elevated serum progesterone levels.19,20,22
In the present study, 10 mg of mifepristone was given as a single dose when the leading follicle had reached a mature state (greater than 12 mm). The LH peak was either inhibited or depressed, whereas the progesterone levels were similar to those in the control cycles. Our data indicate that treatment either delayed the development of the follicle or inhibited ovulation. This was either through a direct effect on the follicle or inhibition of the positive feedback of preovulatory progesterone on the hypothalamicpituitary system.23 There was no effect of the luteinization of the unruptured follicle.
When a high dose of mifepristone (200 mg) is given on day LH + 2, endometrial development is inhibited and endometrial function impaired.24 The effect is sufficient to inhibit implantation.25 In this study, 10 mg of mifepristone had a similar but much less marked effect. The endometrium was slightly out of phase, and the down‐regulation of progesterone receptor concentration was inhibited in five of the six women. The effect on markers for endometrial receptivity was less marked. It is difficult to know if these effects are sufficient to inhibit implantation. The efficacy of mifepristone when used for emergency contraception and the possibility to extend the interval between the unprotected sexual intercourse and treatment to 120 hours suggest that this may be the case. Low daily doses of mifepristone (0.5 mg) had a similar effect on endometrium without influencing ovulation and significantly reduced fertility.26
In this study, treatment with two doses of 0.75 mg of levonorgestrel 12 hours apart before ovulation inhibited the LH peak, but the levels of estrone and pregnanediol glucuronide were similar to those of the control cycle. A decreased midcycle LH peak with a normal pregnanediol excretion has also been reported by Kesseru et al27 and Spona et al28 after treatment with levonorgestrel in the follicular phase. When participants were treated with 0.75 mg of levonorgestrel on cycle days 9, 11, 13, and 15, three women showed follicular activity with complete lack of luteal function, seven women had incomplete luteal function, and six women had normal ovulatory function. When levonorgestrel was administered after ovulation, only minor effects were found. This is in accordance with previous results.7 Earlier studies on gestagens as emergency contraceptives have shown an effect on cervical mucus and uterine fluid in ways that make sperm transport difficult.27 To be of contraceptive importance, these effects require that treatment is started very early after unprotected sexual intercourse because there is experimental evidence for the very rapid transport of spermatozoa from the vagina to the fallopian tube.29
In conclusion, our data suggest that the ovulation process is the main target for emergency contraception with mifepristone and levonorgestrel. However, the effect on the endometrium and on the fallopian tube may also contribute to the contraceptive effectiveness of mifepristone.
1. Consensus statement of emergency contraception. Contraception 1995;52:211–3.
2. Ho PC, Kwan MSW. A prospective randomised comparison of levonorgestrel with the Yuzpe regimen in post-coital contraception. Hum Reprod 1993;8:389–92.
3. WHO Task Force on Postovulatory Methods of Fertility Regulation. Randomised controlled trial of levonorgestrel versus the Yuzpe regimen of combined oral contraceptives for emergency contraception. Lancet 1998;352:428–33.
4. Glasier A, Thong KJ, Dewar M, Mackie M, Baird DT. Mifepristone (RU 486) compared with high-dose estrogen and progesterone for emergency postcoital contraception. New Engl J Med 1992;327:1041–4.
5. Webb AMC, Russell J, Elstein M. Comparison of Yuzpe regimen, danazol, and mifepristone (RU 486) in oral post-coital contraception. Br Med J 1992;305:927–31.
6. WHO Task Force on Postovulatory Methods for Fertility Regulation. Comparison of three single doses of mifepristone as emergency contraception: A randomised trial. Lancet 1999;353:697–702.
7. Landgren BM, Johannisson E, Aedo AR, Kumar A, Yongen S. The effect of levonorgestrel administered in large doses at different stages of the cycle on ovarian function and endometrial morphology. Contraception 1989;39:275–89.
8. Bygdeman M, Gemzell Danielsson K, Marions L, Swahn M-L. Contraceptive use of anti progesterone. Eur J Contracept Reprod Health Care 1999;4:103–7.
9. Gemzell Danielsson K, Swahn ML, Bygdeman M. The effect of various doses of mifepristone on endometrial leukaemia inhibitory factor expression in the midluteal phase — an immunohistochemical study. Hum Reprod 1997;12:1293–7.
10. Marions L, Gemzell Danielsson K, Bygdeman M. The effect of antiprogestin on integrin expression in human endometrium—an immunohistochemical study. Molec Hum Reprod 1998;4:491–5.
11. Marions L, Gemzell Danielsson K. Expression of cyclooxygenase in human endometrium during the implantation period. Molec Hum Reprod 1999;10:961–5.
12. Cekan SZ, Becsac MS, Wang E, Shi S, Masironi B, Landgren BM, et al. The prediction and/or detection of ovulation by means of urinary and steroid assay. Contraception 1986;33:327–45.
13. Metcalf MC, Hunt EG. Calculation of estrogen excretion rates from urinary estrogen to creatinine ratios. Clin Biochem 1976;8:75–7.
14. Johannisson E, Landgren BM, Rohr HP, Diczfalusy E. Endometrial morphology and peripheral hormone levels in women with regular menstrual cycles. Fertil Steril 1987; 48:401–8.
15. Wilcox AJ, Weinberg CR, Baird DD. Timing of sexual intercourse in relation to ovulation. N Engl J Med 1995; 133:1517–21.
16. Lessey BA, Castelbaum AJ, Buck CA. Further characterisation of endometrial integrins during the menstrual cycle and in pregnancy. Fertil Steril 1994;62:497–506.
17. Schoupe D, Mishell DR, Page MA, Madkour H, Spitz IM, Lobo RA. Effects of the antiprogesterone RU 486 in normal women. II. Administration in the late follicular phase. Am J Obstet Gynecol 1987;157:1421–6.
18. Swahn ML, Johannisson E, Daniore V, De la Torre B, Bygdeman M. The effect of RU 486 administered during the proliferative and secretory phase of the cycle on the bleeding pattern, hormonal parameters and the endometrium. Hum Reprod 1988;3:915–21.
19. Ledger W, Sweeting V, Hillier H, Baird D. Inhibition of ovulation by low dose mifepristone (RU486). Hum Reprod 1992;7:945–50.
20. Croxatto HB, Salvatierra AM, Croxatto HD, Fuentealba B. Effects of continuous treatment with low dose mifepristone throughout one menstrual cycle. Hum Reprod 1993; 8:201–7.
21. Croxatto HB, Salvatierra AM, Fuentealba B, Leiva L. Follicle stimulating hormone-granulosa cell axis involvement in the antifolliculotropic effect of low dose mifepristone (RU486). Hum Reprod 1995;10:1987–91.
22. Baird DT, Thong KJ, Hall C, Cameron ST. Failure of oestrogen induced luteinizing hormone surge in women treated with mifepristone (RU 486) every day for 30 days. Hum Reprod 1995;10:2270–6.
23. Batista MC, Cartledge TP, Zellmer AW, Nieman LK, Merriam GR, Loriaux DL. Evidence for a critical role of progesterone in the regulation of the midcycle gonadotropin surge and ovulation. J Clin Endocrinol Metab 1992; 74:565–70.
24. Gemzell Danielsson K, Svalander P, Swahn ML, Johannisson E, Bygdeman M. Effects of a single post-ovulatory dose of RU486 on endometrial maturation in the implantation phase. Hum Reprod 1994;9:2398–404.
25. Gemzell Danielsson K, Swahn ML, Svalander P, Bygdeman M. Early luteal phase treatment with RU 486 for fertility regulation. Hum Reprod 1993;8:870–3.
26. Marions L, Viski S, Gemzell Danielsson K, Resch BA, Swahn ML, Bygdeman M, et al. Contraceptive efficacy of daily administration of 0.5 mg mifepristone. Hum Reprod 1999;14:2788–90.
27. Kesseru E, Garmendia F, Westphal N, Parada J. The hormonal and peripheral effects of dl-norgestrel in postcoital contraception. Contraception 1974;10:411–24.
28. Spona J, Matt K, Schneider WHF. Study of the action of D-norgestrel as a post-coital contraceptive agent. Contraception 1975;11:31–43.
29. Kuntz G, Beil D, Deininger H, Wildt L, Leyendecker G. The dynamics of rapid sperm transport through the female genital tract: Evidence from vaginal sonography of uterine peristalsis and hysterosalpingoscintigraphy. Hum Reprod 1996;11:627–32.
© 2002 by The American College of Obstetricians and Gynecologists.