The present invention relates to the field of pharmacology and, in particular, to the design and use of various forms of contraception. BACKGROUND ART

Progesterone plays a critical role in mammalian reproduction because of its role in maintenance of the endometrium. It is also essential for the initiation and maintensuice of pregnancy. After the discovery of the progesterone receptor in 1970, researchers realized that a progesterone receptor antagonist would have a major impact on female reproductive he.alth. In 1981 , Philibert, Deraedt, and Teutsch, from the French ph.armaceutic.al company, Roussel Uclaf, reported on a newly synthesized glucocorticoid receptor antagonist known as RU 38486. It soon became evident that this antiglucocorticoid also displayed marked antiprogestin activity. RU 38486 was subsequently abbreviated to RU 486 and is now currently known by the generic name mifepristone. The original studies show that mifepristone has a relative binding affinity five times greater than that of progesterone and three times greater than that of dexamethasone, at their respective receptors. When compared to testosterone, it has a relative binding affinity of 25 % for the androgen receptor, but did not bind to either estrogen or mineralocorticoid receptors.

When antiprogestins bind to progesterone receptors, they prevent receptor initiated transcription and therefore, at the normal concentration in the body, progesterone cannot activate gene expression in its target cells. As a consequence, receptor-mediated progesterone actions are suppressed.

As a group, antiprogestins are characterized by substitutions at the l lβ and 17α positions of the steroid ring system and bind strongly to both progesterone and glucocorticoid receptors. Although they function

predominantly as antiprogestins and antiglucocorticoids, on occasion they display progestin antagonistic and even antiestrogenic properties. The most common clinical use of mifepristone is to induce a medical abortion in the early stages of pregnancy. However, progesterone maintains the endometrium and transforms it from a proliferative to a secretory state. It also facilitates the lutenizing hormone surge which initiates ovulation. As a consequence, antiprogestins are also suggested to have contraceptive potential. Although antiprogestins are known to delay or inhibit ovulation, reports thus far conclude that this effect is inconsistent, unless high doses are given. When they are used alone to suppress ovulation, the antiprogestin effect is likely to be associated with unopposed estrogen action on the endometrium.

Previous studies have shown that oral administration of mifepristone, at a dose of 1 mg/day or more for one menstrual cycle, interferes with ovarian follicular development, ovulation and ovarian hormone production, as well as with endometrial maturation. However, even oral administration of low doses of antiprogestins creates problems. The relative specificity and selectivity of antiprogestins taken orally can be quite low. The entire body is indiscriminately exposed to the antiprogestin, which is usually a modified steroid; and we are now just beginning to learn about some of the adverse consequences of such exposure. In addition, as with any low dose oral contraceptive, absolute compliance is both mandatory for efficiency and nearly impossible to achieve. Other hormonal contraceptive methods that act preferentially but not necessarily exclusively on the endometrium are the progesterone- releasing intrauterine device known as PROGESTASERT, produced by Alza Corp. and the levonorgestrel-releasing intrauterine device (IUD) known as MIRENA, produced by Leiras Oy. The progestins are released

directly into the uterine cavity by these devices and primarily act as an antiestrogen, which opposes the proliferative action of estrogen on the endometrium and desynchronizes its development from the very beginning of the endometrial cycle. As a consequence, the mechanisms involved in initiating and stopping endometrial bleeding are profoundly disturbed; therefore, the normal menstrual pattern is often replaced by intermenstrual bleeding and spotting and periods of amenorrhea. Therefore, these contraceptive methods are associated with these undesirable side affect which negatively affects the initial acceptability of the devices, as well as the continuation of their use by the women who use them.

Therefore, there remains a significant need for a non-invasive way to administer an antiprogestin as a contraceptive in an effective manner, without the negative consequences of oral administration. That need is addressed by the present invention. SUMMARY OF THE INVENTION

In accordance with the present invention, there is provided a method of contraception. The method includes the steps of inserting into either the vagina or the uterus of a female, preferably a human female capable of conception, a source of at least one antiprogestin. The antiprogestin is provided in an amount which is equal in contraceptive efficacy to between about 0.1 and about 2 mg of mifepristone each day for a period of time which is sufficient to prevent pregnancy. For the purposes of illustration only, mifepristone can be administered in the amount of between 0.1 to about 2 mg per day for at least one day mid- cycle between about day 10 and about day 24, of the normal cycle of a human female. Alternatively, mifepristone could be administered in the same amount each day for up to 7 days during that same period of time.

In another embodiment, mifepristone could be provided continuously, each day, in an amount of between about 0.1 to about 2 mg per day throughout at least 28 days and, most preferably, during the entire cycle. More preferable, in accordance with the present invention, a method of contraception is provided which includes the steps of inserting into either the vagina or the uterus of a human female, a device which is capable of releasing a drug associated therewith. Drug is released in a controlled manner such that a generally equivalent amount of antiprogestin is released each day over a pre-selected period of days. For example, mifepristone could be administered in an amount of between about 0.1 to about 2 mg per day, each day while the device remains within the woman's vagina (1 day to 1 year). Mifepristone is provided in an amount which is sufficient, when administered over the pre-selected number of days, to prevent pregnancy.

Finally, in accordance with the present invention, there is provided a contraceptive device for administering a low dose of antiprogestin to the vagina of a female. The device is vaginally insertable and is capable of the controlled release of a drug associated therewith. Preferably, on average, the same amount of drug will be administered each day. Of course, some variation is to be expected, as is known in the pharmaceutical industry. As long as the minimal dose is consistently reached, the device is functional for its intended purpose. Preferably, the device will contain an amount of drug which is sufficient to provide for the administration of an antiprogestin in an amount which is equivalent in contraceptive effect to between about 0.1 and about 2 mg of mifepristone each day for the intended number of days. The device contains sufficient capacity to allow for the administration of that amount of antiprogestin for each day of its intended useful life. Therefore, if the device is intended to

remain within the vagina for 31 days, an amount of antiprogestin must be provided which is sufficient to allow for the administration of the intended dose each day for 31 days. In the case of mifepristone, the amount of drug necessary would, at least, range from between about 3.1 to about 62 mg. If a dose of 1 mg per day is administered for one year, at least 365 mg will need to be stored in the device. However, it is often useful to load a delivery device with an excess of drug.

The use of vaginal administration of low doses of antiprogestins during at least a portion of the normal menstrual cycle provides several significant and unexpected advantages over other routes of administration and higher dosages. Mifepristone, for example, was originally tested for both oral and vaginal efficacy. However, it was found at the time that vaginal applications of mifepristone were relatively inefficient to obtain the desired blood levels of drug thought necessary. See Heilάnhelmo et al. , "Intravaginal administration of RU 486 in humans and rats: inadequate absorption in humans," Hum. Repro. Vol. 2, no. 8, (1987) pp. 645-648. These findings clearly suggest that vaginal administration of large doses of mifepristone and other antiprogestins are contraindicated and that oral and other routes of administration should be used. Moreover, the finding that high doses of vaginally administered mifepristone did not yield sufficient blood levels of drug also argues against the use of low dose of antiprogestins, particularly vaginally.

These conclusions are bolstered by subsequent studies showing that the oral administration of even relatively low doses of mifepristone continuously throughout one menstrual cycle prevents normal ovulation in many instances during treatment. At the same, these dosing protocols allow full follicular growth and normal estrogen levels to be obtained in women. See Ledger W.L., Sweeting V.M., Hillier H., Baird D. T., (1992): "Inhibition of ovulation by low dose mifepristone (RU

486)," Hum. Reprod. , 7:6-11 and Croxatto H.B. , Salvatierra A.M., Croxatto H.D., Fuentealba A., (1993): "Effects of continuous treatment of low dose mifepristone throughout one menstrual cycle," Hum. Reprod. , 201-7. The apparent success of the oral administration of mifepristone only serves to validate the inevitable conclusions drawn from prior studies comparing higher dose vaginal and oral administration routes.

It has now been surprisingly found that the vaginal administration of low dose antiprogestins produces comparable, if not superior, contraceptive effects, when compared to an oral administration. In addition, antiprogestins such as mifepristone, are active on the endometrijd portion of a woman's uterus. Vaginal or uterine introduction of the antiprogestin allows for more efficient dosing of the target organ, the uterus, and therefore increases the efficiency of the dosing regime. Accordingly, it is possible to use much lower doses of antiprogestin in accordance with the present invention when compared to oral doses and still obtain comparable contraceptive results. Vaginal administration of antiprogestins also avoids both first pass metabolism and unnecessarily high levels of steroid in a woman's blood stream. Reducing the systemic exposure to steroids is a significant advantage in terms of the overall health of the woman and it can dramatically reduce potential long and short term side effects she may experience. In addition, not only has it been found that vaginal administration is more target organ specific, it is also surprisingly more organ selective. Unlike oral administration, there is less of an effect on the ovaries of a woman, while at the same time, sufficient effect is realized in the uterus to allow for effective contraception. Again, this constitutes a significant and unexpected advantage over oral administration.

In a particularly preferred aspect of the present invention, low doses of antiprogestins are administered vaginally or in the uterus,

continuously throughout the women's menstrual cycle. This protocol exhibits other significant advantages worthy of note. With antiprogestin pills, it is possible that a woman will take a non-lethal overdose of same in an effort to abort a pregnancy. Whether or not such action would actually cause an abortion is dependent on many factors. However, if unsuccessful, there is at least some risk that the effects of such overdoses can be profound on the physiology of both the mother and the embryo. Even if successful, the effects on the woman's physiology can be significant. If an antiprogestin is administered continuously by a vaginal ring or another vaginal or in-utero device which is replaced on a weekly, monthly or longer term basis, there is little chance of overdosing abuse. Thus, a slow releasing device is less amenable to abuse when compared to an oral form. This type of delivery of mifepristone also constitutes a significant advantage in terms of convenience. Oral mifepristone must be taken at the same time everyday in order to be truly effective. With a long-term vaginal or in-utero device, there are fewer compliance concerns. This is particularly important when low doses of antiprogestins are being used because insufficient compliance may result in unintended pregnancies. That danger is less when higher dosages are used and is essentially non- existent in accordance with the present invention.

Finally, when most drugs, including antiprogestins, are taken orally, there is an initial burst effect where relatively high levels of the drug are circulating within the system shortly after ingestion. This is even true when using oral timed release technology. Thereafter, it is anticipated that the release profile of the drug will begin to approximate zero order, i.e. a more constant rate of bioavailability. Toward the end of the dose, however, under-dosing may occur. The effects of these peaks and valleys in bioavailability may not be profound in terms of contraception, although they may be profound in terms of symptoms for the woman. By the use of

the present invention, however, a burst effect is realized only at the beginning of administration. With a once-a-month device, such effects would only be a possibility on the first day and last day of the dose. BRIEF DESCRIPTION OF THE DRAWINGS FIGURE 1 - Illustrates the follicular growth profile, as well as the serum E ₂ and progesterone levels in the four women discussed in Example 1.

FIGURE 2 - A photograph of the endometrium biopsy at the equivalent of day 21 of a standard cycle for the first of two women treated leading ovarian follicle is 14 ± 2 cm as provided in Example 1.

FIGURE 3 - A photograph of the endometrium biopsy at the equivalent of day 21 of a standard cycle for the second of two women treated leading ovarian follicle is 14 ± 2 cm as provided in Example 1.

FIGURE 4 - Photograph of endometrium biopsy at the equivalent of day 21 of the standard cycle in the first of two women treated on the day of follicular collapse.

FIGURE 5 - Photograph of endometrium biopsy at the equivalent of day 21 of the standard cycle in the second of two women treated on the day of follicular collapse. FIGURE 6 - Illustrates the in vitro release of mifepristone from vaginal ring.

FIGURE 7 - An elevational view of a vaginal ring device useful for administering low doses of antiprogestin.

FIGURE 8 - A cross-sectional view of the vaginal ring of FIGURE 7 taken along line AA.

FIGURE 9 - A cross-sectional view of the vaginal ring of FIGURE 7 taken along line AA.

FIGURE 10 - A cross-sectional view of the vaginal ring of FIGURE 7 taken along line AA.

BEST MODE OF CARRYING OUT INVENTION

Vaginal administration of very low doses of mifepristone using a slow releasing device has been found to interfere selectively with endometrial changes induced by progesterone without affecting ovarian function. The fact that this treatment allows the ovary to perform its cyclic functions such as hormone production represents an advantage for many women who prefer to perceive the subjective changes associated with the normal menstrual cycle. In addition, when cyclic ovarian function is preserved, there is no risk of continuous unopposed estrogen effect on the endometrium.

The ability to administer an antiprogestin, for example, mifepristone, in such a way that it interferes selectively with endometrial actions of progesterone without disturbing ovarian hormone production constitute a clear innovation and advantage over existing hormonal contraceptive methods. The administration of very low doses of mifepristone and/or other antiprogestins from a slow releasing device worn in the vagina, such as a vaginal ring, will interfere only with the second phase of the endometri.al cycle which is normally dependent on progesterone action. Thus, estrogen-dependent endometrial bleeding will be preserved and luteolysis, the process that allows the onset of the new cycle, will continue to take place regularly because ovarian function is unaffected by this treatment. Thus, the treatments of the present invention will not interfere with the occurrence of a regular menstrual pattern.

The chemical name of mifepristone is 11 -[4- (dimethylamino)phenyl- 17-hydroxy- 17-( 1 -propynly)-( 11 β , 17bβ)-estra-4,9- dien-3-one. It is a derivative of norethindrone and possesses an additional (4-dimethylamino)phenyl group at the llβ position and 1-propynyl chain at the 17α position. The conjugated C9-C10 double bond in mifepristone should also be noted. Since the initial report of mifepristone, over 400

additional antiprogestins have been synthesized. See Hodgen G.D.,(1991): "Antiprogestins: the political chemistry of RU 486," Fertil. Steril. 56:394- 95, the text of which is hereby incorporated by reference, as is U.S. Patent No. 4,978,657. The 17α substitution is responsible for promoting higher binding affinity to the receptor. The vast majority of these antagonists possess a substituent at the l lβ position. It is likely that the l lβ substituent is responsible for the antagonistic action. A series of l lβ- substituted compounds such as RTI 3021-022 and RTI 3021-020 have been developed with 16α-ethyl-17β-acetyl substitutions in the D ring. RTI 3021-022 and RTI 3021-020 act as potent progestins and possess no antagonistic activity. In addition, antiprogestins such as ORG 31167 and ORG 31343 have been developed in which the dimethylaminophenyl group is situated at C18. Thus both the nature and position of substitutions in the steroid structure appear to be critical for the antagonistic activity. Although a selective antiprogestin that does not bind to glucocorticoid receptors (GRs) has not been reported, recently, a number of compounds were described with minimum antiglucocortoid activity. One such example is RU 46556.

In addition to its ability to maintain endometrial integrity after implantation of the embryo, progesterone acts at several discrete anatomical sites to regulate a number of steps in the female reproductive process that precede implantation. Thus, besides disrupting an ongoing implantation, blockade of progesterone action may conceivably prevent pregnancy by other mechanisms. Rather than reviewing all known progesterone actions, the discussion focuses only on those that have been blocked by antiprogestins in women.

The best-known functions of circulating progesterone before implantation include the following: (a) facilitation and reinforcement of the positive feedback of estrogen on the lutenizing hormone (LH) surge near

-l i¬ the end of the follicular phase, (b) synergism with estradiol during the luteal phase to maint n negative feedback control on gonadotropin secretion, (c) transformation of the endometrium from a proliferative to a secretory state for reception of the fertilized egg, and (d) maintenance of endometrial integrity. From a theoretical standpoint, antagonizing some or all of these actions of progesterone could result in a contraceptive effect. Antiprogestins could thus function as contraceptive agents by (a) inhibiting ovulation, thereby preventing fertilization; (b) interfering with the secretory transformation of the endometrium rendering it non-receptive for implantation, and (c) inducing shedding of the endometrium and dislodging the implanting embryo.

Administration of antiprogestins during the follicular phase of the menstrual cycle disrupts development of the leading follicle. This suggests that the presence of the low levels of circulating progesterone typical of the follicular phase or of the high levels of intrafoUicular progesterone or both are essential for the growth of the dominant follicle, the last stage in the development of a pre-ovulatory follicle in primates. There are two ways by which antiprogestins prevent ovulation: firstly, by interfering with the growth of the dominant follicle, the so-called antifolliculotropic or foUiculostatic effect; and secondly, by inhibiting the LH surge. This antifolliculotropic effect was suspected when investigators observed that administration of large doses of mifepristone (25-100 mg/day) during the mid to late follicular phase prolonged the follicular phase and delayed the LH surge. This occurred in .association with lower estradiol levels than were normally observed at corresponding times of the normal cycle. Ultrasonography clearly documented cessation of growth of the dominant foUicle.

This delay in the LH surge consequent to antiprogestin administration was also evident with considerably lower doses of

mifepristone and was observed following 1 mg given for 10 days during the follicular phase after the dominant follicle had achieved a size of 14-16 mm. Five mg of mifepristone administered for up to three days in the follicular phase impaired follicular development, as documented by ultrasound, and was associated with a reduction in serum estradiol and inhibin levels. This foUiculostatic effect of antiprogestins is exquisitely sensitive and a single 5-mg dose of mifepristone transiently arrests follicular growth, an effect that is evident 12 h after pill ingestion. The antifolliculotropic activity is also displayed by onapristone and it has also been observed with mifepristone, onapristone and lilopristone in non- human primates. This antifolliculotropic effect of antiprogestins is not evident during the earliest part of the menstrual cycle. Thus, administration of mifepristone during the first three days of the follicular phase, at a time when no dominant follicle is present, has no effect on follicular growth .and fails to alter subsequent events of the cycle. This effect of antiprogestins is best demonstrated when the dominant follicle reaches a diameter of 12 mm. This suggests that growth beyond 12 mm is .associated with functional changes in the follicle, such as acquisition of progesterone receptors by granulosa cells of the dominant follicle as it approaches maturity.

Continuous oral administration of 5 or 10 mg of mifepristone per day throughout one menstrual cycle prevented the leading follicle from achieving maturity and from producing adequate circulatory estradiol levels necessary to trigger the LH surge. With lower oral doses, e.g. 1 or 2 mg/day, full folUcular growth and normal estrogen levels were attained in several women, but normal ovulation failed to occur in many instances during treatment. With these low oral doses, the occurrence of unruptured leuteinized foUicle was inferred by elevation of serum progesterone without evidence of ovulation on ultrasound. Thus, with high doses of

mifepristone, ovulation suppression during treatment can be accounted for primarily by the absence of estradiol positive feedback. Even though adequate estrogen signaling does occur with lower doses, ovulation still fails due to either an inadequate LH surge or an inadequate follicular responsiveness.

It has been observed, however, that the continuous oral administration of 2 or 5 mg of mifepristone daily for one month suppresses ovulation, whereas 1 mg daily fails to do so consistently. See Spitz, Croxatto & Robbins, (1996): "Antiprogestins Mechanism of Action and Contraceptive Potential," Annu. Rev. Pharmacol. Toxicol., 36: 47-81; Ledger W.L., Sweeting V.M., HiUier H., Baird D.T., (1992): "Inhibition of Ovulation by Low-Dose Mifepristone (RU 486)," Hum. Reprod., 7:6- 11; Croxatto H.B., Salvatierra A.M., Croxatto H.D., Fuentealba A., (1993): "Effects of Continuous Treatment With Low-Dose Mifepristone Throughout One Menstrual Cycle," Hum. Reprod., 8:201-7 and Batista M.C., Carteledge T.P., ZeUmer A.W., Merino M.J., Axiotios C, et al., (1992): "Delayed Endometrial Maturation Induced by Daily Administration of the Antiprogestin RU 486: A Potential New Contraceptive Strategy," Am. J. Obstet. Gynecol , 167:60-65. In accordance with the present invention, mifepristone is administered in the uterus or in the vagina. This can be accomplished by any means known in the art. Medicated suppositories, tampons, creams, gels, douches, foams, etc. may be used daUy. This type of administration may be useful for vaginal administration of low doses during a relatively few days during the cycle. For example, mifepristone could be administered for several days during the first half of the luteal phase to interfere with the secretory transformation of the endometrium. It could .also be administered for between about 1-7 days at mid-cycle, days 12-22, depending upon the specific need. Efficacy of any protocol can be

measured as described in Example 1. However, the preferred protocols in accordance with the present invention involve continuous administration throughout the entire 28+ day cycle. Creams, gels or tampons would be as inconvenient as tablets in such circumstances. While these protocols would still have benefits over oral administration, the compliance issues make them less desirable for long term continuous administration of low doses of antiprogestins

Alternatively, a patch which may be inserted and retained within the vagina, a vaginal ring, a vaginal or uterine implant or an IUD may be used. Such devices could either be erodable or could merely provide a release of the drug without any decomposition of the delivery device itself. Examples of preferred devices include a soft polymer based vaginal ring and a steroid-releasing IUD such as those sold under the trade names PROGESTASERT and MIRENA. Any antiprogestin which is capable of meeting the criteria of the present invention may be used. Such antiprogestins must be safe and efficacious for vaginal or uterine introduction. They must be capable of reUable and constant administration in an amount which is equivalent in effect to between about 0. 1 and about 2 mg of mifepristone from creams, foams, tampons or devices, such as vaginal rings, as discussed herein. Most preferably, antiprogestins in accordance with the present invention will provide contraceptive efficacy, without interrupting the normal menstrual cycle.

When the antiprogestin used is mifepristone, it may be provided in an amount of between about 0. 1 and about 2 mg per day. In certain instances it may be preferred to administer between about 1 and 2 mg/day. However, with specific women, it may be advantageous to administer between about 0. 1 and about 1 mg/day. Most preferably, the antiprogestin will be provided in an amount which is equivalent to between

0.25 to about 2 mg of mifepristone and more preferably, about 0.5 to about 2 mg. The equivalency of effect of various doses of antiprogestins can be compared to doses of mifepristone by using the test described in Example 1 and/or J Delettri, et. al., "Steroid Flexibility and Receptor Specificity," J. Steroid Biochem. , 13:45-59, 1979. Preferably, the antiprogestin should be administered continuously over the entire menstrual cycle. Most preferably, this is accomplished by using a single device, designed to administer antiprogestin on as consistent and even a basis as possible, throughout at least about 28-31 days. For example, it may be possible to accomplish this by using a series of four vaginal rings, which are replaced at the end of each week. The use of a plurality of devices in series is specifically contemplated; however, it would be preferred to administer all of the drug through a single monthly applied device.

With the foregoing in mind, a device in accordance with the present invention is one which will release at least one antiprogestin in an amount desired on a daily basis, over a period of one or more days. For a once a day tampon intended to administer mifepristone, the tampon must be capable of delivering between about 0.1 and about 2 mg of mifepristone to the woman per day as desired. If a vaginal ring is intended for weekly administration, then the vaginal ring must be capable of delivering between about 0.1 to about 2 mg of mifepristone per day, each day over its entire seven day useful life. An IUD intended to be changed on a yearly basis can include mifepristone as weU. In that case, the device will have to be capable of delivering about 0.1 to about 2 mg of mifepristone each day throughout the entire year.

As shown in FIGURE 7, the sustained release device can be, for example, a vaginal ring 10. The ring 10 can be made of a polymer which is porous or which is bioerodable. The ring 10 will have the antiprogestin associated with it. "Associated" in this context means that

the antiprogestin can be dispersed uniformly throughout a drug releasing matrix. For example, FIGURE 8 shows vaginal ring 10 in cross-section. The antiprogestin 11 is shown dispersed uniformly in a matrix 12 which wUl control the release of the antiprogestin 11. FIGURE 9 shows another embodiment where the drug 11 is disposed in a layer 12 coated over a core 13. Additional rate controlling or protective layer(s) (not shown) can cover layer 12 as well. The antiprogestin could also be stored in a reservoir or hoUow lumin 14 within a ring 10. It can be stored as a liquid, solid, gel or emulsion. See FIGURE 10. When another antiprogestin is used, it may be provided in a higher or lower amount depending upon d e potency of that particular drug, as well as other necess.ary factors. However, an amount of that other antiprogestin should be provided such that the contraceptive efficacy is the same as mifepristone when provided on a daily basis of about between about 0.1 to about 2 mg/day. This can be measured by a progesterone receptor affinity assay. See, J. Delettre', supra.

Antiprogestins can also be evaluated by a second progesterone receptor binding assay for their ability to competitively inhibit binding of [ H]-ρromegestone (New England Nuclear Labs, Boston, Massachusetts, USA) to progesterone receptors in cytosol prepared from rat uteri. This test can also be used to determine the equivalency of doses of various antiprogestins to doses of mifepristone in terms of contraceptive effect. Female Sprague-Dawley rats weighing 125-150 g are ovariectomized. Seven days later, the rats are treated with 1.0 μ.g estradiol daily for 3 days. The rats are sacrificed 24 hours after the last injection, and the uteri removed, weighed, and minced with scissors. All subsequent steps are carried out at 4°C. The minced tissue is homogenized in 2 volumes of lOmM Tris-HCl buffer (pH 7.4) containing 1.5 mM

EDTA, 0.5 mM DTT, 10 mM Na ₂ Mo0 ₄ , 10% glycerol, and 1 mM phenylmethylsulfonyl fluoride. The homogenate is centrifuged at 1000 x g

for 30 minutes. The supernatant is separated and recentrifuged at 105,000 x g for 90 minutes, and the final supernatant (uterine cytosol) is used for competitive binding.

Aliquots of 100 μL of cytosol .are incubated overnight at 4°C with [ H]-promegestone in the presence or absence of synthetic progestin derivatives. Unbound steroids are removed by treatment with dextran- coated charcoal (0.5 mL containing 1.25% charcoal-Norit A) and 0.125% dextran T-70 in assay buffer. After 10 minutes of incubation at 4°C, the samples are centrifuged at 1000 x g for 10 minutes, and the radioactivity in the supernatant is determined in a Beckman scintillation counter.

The percentage of radioligand bound in the presence of competitor compared to that bound in its absence is plotted against the concentrations of unlabeled competing steroid. The molar concentrations of steroid competitor that reduced radioligand binding by 50% (ED ₅₀ ) .are determined. The ratio of these concentrations, multiplied by 100 is termed the relative binding affinity (RBA). Example 1,

Four volunteer women, protected from pregnancy by surgical sterilization, with regular menstrual cycles, participated in this study. Each subject was studied during one treatment cycle. Treatment consisted of the introduction of a tampon containing mifepristone in the vagina. Based on in vitro studies, the tampon released approximately 1-2 mg of mifepristone per day. Treatment started when the leading ovarian follicle was 14 ± 2 cm in two volunteers, who kept the tampon in the vagina for three days. In the other two subjects, treatment started on the day folUcular collapse was detected and continued for six consecutive days. In these subjects, after using the tampon for three days, the tampons were removed and new tampons inserted for another three days.

Follicular growth was monitored daily by ultrasonography from cycle day 7 until ovulation was detected. A blood sample for the determination of serum levels of estradiol (E ) and progesterone were taken on the same days that echography were performed and also at 3 and 6 days after ovulation. On day 6 after ovulation, an endometrial biopsy was taken in all cases for histologial assessment.

FolUcular growth profile, as well as serum Eo and progesterone levels were consistent with normal mid-cycle phase in the four cases (Figure 1). However, the endometrium which was biopsied at the equivalent of day 21 of a standard cycle was classified as secretory endometrium day 17, in the two subjects treated during die follicular phase (Figs 2 and 3) and secretory endometrium day 16 and 17 (Figs 4 and 5) in the women treated after ovulation.

Since a woman cannot determine d e date of follicular collapses or the size of the leading follicle, the best practical solution is the continuous dosing of antiprogestin each day of the cycle. Because the present invention uses very low doses administered in the vagina or uterus, the chance of side effects from prolonged use is quite small. Very low systemic levels are realized. Example 2

Rings were made from Dow Corning Elastamer 382 which is a silastic polymer. This was accompUshed by taking the Elastamer and mixing it with a solution of stannous octoate catalyst in toluene (1 or 2 drops per gram of elastomer). The mixture was then placed into a ring- shaped dye at room temperature and allowed to solidify. The ring was then removed from the dye. The resulting ring had an 8.4 mm cross- section and a 58 mm total diameter. Next, the ring was cut and pulled through a solution which was a mixture of additional silastic polymer and mifepristone in a 20-80% weight ratio. A coating of approximately 0.3 mm

was added to the surface of the ring thereby. The ring was then dipped in solution of additional stannous octoate catalyst in toluene (10 drops: 35 mL). The ring was air-dried for 5 minutes and then heated to completely evaporate the solvent. Heating occurred for about 30 minutes at 100°C. The ring was weighed to determine the mifepristone content, which generally ranged from between about 71 to about 84 mg. Next, a length of Dow Corning MDX-4-4515 tubing (6.4 mm I.D. by 0.30 mm of wall thickness) was expanded in hexane and slipped over the cut ring. The ends of the tubing were trimmed and coated wit Dow Corning medical adhesive and then the two ends were connected. Through this operation, the tubing had stretched to a raw thickness with about 0.25 mm.

To test the release profile of the rings, a ring was incubated in a medium of benzalkonium chloride/aqueous solution in a ratio of 1 :750 (1 fluid ounce/127 fluid ounces). The ring was placed into a sufficient volume of solution to completely cover same and then incubated for 24 hrs. at 37°C. Each day, the medium was changed and assayed for the concentration of mifepristone by UV spectrophotmaty at 306 nanometers. Fresh benzal konium chloride was used as the blank. Concentration was run against a standard curve. The results are iUustrated in Figure 6. After an initi ly higher level of release, the release rate of the ring was between about 0.65 and about 0.70 milligrams per day throughout the remainder of the cycle. Figure 6 presents an average of the data from tests involving three separate rings made in accordance with this procedure.

As has been demonstrated herein, the application at low doses of antiprogestins in the vagina or uterus of a woman can be an effective method of birth control. INDUSTRIAL APPLICABILITY

The present invention is applicable to the pharmaceutical industry and to the design and use of devices for contraception. In

particular, it is relevant to the construction of low dose, multi-day contraceptive devices.