Physiology and Mechanisms of Action of Steroids

Steroid hormones are produced in the adrenal glands, ovaries, and testes, and regulate a wide range of physiologic functions. This review covers the action of sex steroids, with emphasis on the estrogens and progestins. A review of the action of sex steroid hormones, both natural and synthetic, should begin with a description of endogenous sources and a definition of such substances.

Estrogens display one common biologic activity, the ability to stimulate growth and maintain the female sex characteristics. As mediators of growth and differentiation of the reproductive tract in the nonpregnant woman, estrogens are produced predominantly by the ovary and in a clearly defined pattern that is related to changes in both physiology and morphology attuned to the menstrual cycle.

Estrogens are the most important steroids produced by the developing ovarian follicle. As the follicle reaches preovulatory size, estrogen synthesis reaches a maximum. Serum levels increase dramatically before ovulation, and are responsible for the positive feedback signal that triggers the release of luteinizing hormone (LH) from the pituitary gland, which in turn induces ovulation. In the second half of the menstrual cycle, after ovulation, estrogen is produced by the corpus luteum, although to a lesser degree and of secondary importance compared with progesterone production. In the first half of the cycle, estrogens are responsible for the growth of the uterine endometrium (proliferative phase); in the second half, progesterone action results in the secretory endometrium characteristic of the luteal phase of the menstrual cycle.¹ The most important estrogens produced by the ovary are estradiol and estrone, and the predominant estrogen in the circulation of nonpregnant women is estrone sulfate.² Estriol, the major estrogen in pregnant women, is produced by the placenta by conversion of steroid precurors supplied by the fetus. For these reasons, estriol is almost exclusively an estrogen of pregnancy.^3,4

The actions of estrogen are not limited to the reproductive tract and to purely endocrine interrelationships; there are many systemic effects beyond the scope of this review.⁵

Progestogens, as indicated by the name, are hormones whose major function is the maintenance of pregnancy. They have been specifically defined as agents that induce secretory changes in the proliferative endometrium and maintain pregnancy after ovariectomy in laboratory rodents.⁶ Although there are several natural progestogens, progesterone is the only natural progestogen of major biologic significance. This hormone does not display the multiplicity of systemic effects characteristic of the estrogens; its action is limited predominantly to the reproductive organs. The elevation of the basal body temperature after ovulation is due to the thermogenic effect of progesterone at the level of the hypothalamus.

Progesterone is produced by all tissues that have steroidogenic activity, and is a major component in the pathway of synthesis of many steroid hormones; therefore, low levels of progesterone are always produced by steroidogenic tissues. The weak progestogen 17-hydroxy progesterone also is produced by the ovary in a predictable pattern in relation to follicular and luteal function.⁷ The only dramatic increase in nonpregnant women occurs after ovulation, when the corpus luteum produces large amounts of progesterone, which in turn bring about the development of the secretory endometrium. The placenta assumes the function of progesterone production between the fifth and seventh weeks of gestation, and it produces increasing amounts during the course of pregnancy. The maintenance of pregnancy is dependent on the continued production of progesterone, first from the corpus luteum and, after a gradual shift, from the placenta for the remainder of pregnancy.⁸

Endogenous and exogenous steroid hormones are transported through the circulatory system to the target organs where they exert their specific hormonal actions. Steroids bind to a variety of macromolecules in the circulation. The affinity of these macromolecules for steroids can vary from weak (Kd = 10^-3 M) to strong (Kd = 10^-10 to 10^-8 M), and can be present in significant concentrations. The measurement of serum or plasma steroid concentration generally is a measure of the total steroid present and does not reflect the equilibrium that exists between steroid hormone bound to these macromolecules and those free in the blood. The bound and free forms of the steroid are in an equilibrium that is controlled by many physiologic factors. It is important to remember that generally only the free hormone can leave the circulation and enter the target cells, where it can bind to specific intracellular receptors to initiate the biochemical expression of specific sex steroids.^9,10,11

Examples of these specific steroid-binding macromolecules include a testosterone-estrogen-binding globulin, also known as sex hormone-binding globulin, which binds both testosterone and estradiol with high affinity (Kd = 10^-10 M). Its production is stimulated by increasing estrogen concentration and decreased by increasing testosterone concentration.¹² A second specific binding globulin is the corticoid-binding globulin, also known as transcortin, whose main function is to bind glucocorticoids, such as cortisol, but which also binds progesterone with a high affinity. The plasma concentration of this binding globulin is increased by estrogen; consequently, it is increased dramatically during pregnancy.¹²

In addition to the binding of steroids by these high-affinity, low-capacity macromolecules, albumin is a sex steroid binder with low affinity but tremendous capacity in the circulation. The concentration of albumin in the blood can be as high as 4%. At such a high concentration, it can alter the equilibrium distribution of steroids in blood, and therefore is an important consideration in the circulating economy of steroid hormones. Because of the greater concentration of albumin and the rapid dissociation of steroids from this low-affinity binder, albumin may serve a more important regulatory role than do the high-affinity, low-capacity binding proteins.¹²

Steroid hormones in the circulation (e.g., estrogen, progesterone) are transported bound to proteins, but also exist in free form in a complex equilibrium. The free hormones cross cell membranes in all tissues in the body, but only those tissues that are specific targets for steroids have receptors that bind and retain the steroids, leading to an intracellular accumulation that eventually allows the biologic expression that is characteristic of the particular steroid. These receptors are markedly different from gonadotropin receptors, which are localized in the cell membrane and have several second-messenger systems as mediators of receptor binding. Steroid receptors were identified many years ago, when radioactively labeled steroids became available. The dogma that persisted for many years was of the presence of an unoccupied cytoplasmic receptor that would bind the steroids that crossed the cell membrane into the cytoplasm. After binding, this unit was activated, and it crossed the nuclear membrane to bind within the nucleus and cause the expression of specific genes related to specific steroid actions. In the last few years, several studies changed the interpretation of this process. In particular, the availability of specific antibodies against the estrogen receptor showed that these steroid hormone receptors, occupied or unoccupied, are localized primarily in the nucleus. Earlier studies required tissue homogenization and centrifugation to separate the cytosolic and nuclear fractions. Because the unoccupied receptor in the nucleus is not tightly bound to nuclear proteins, it was easily translocated into the cytosolic fraction. On the other hand, the occupied receptor is specifically bound to structural components in the nucleus and, consequently, remained localized in that fraction. Therefore, the early suggestion of an unoccupied cytosolic receptor and an occupied nuclear receptor were artifacts of the homogenization and separation procedure. The availability of specific antibodies for estrogen receptors and their use in immunochemistry has allowed visualization of both free and occupied receptors in the nucleus. Although these studies were done initially with the estrogen receptor, other steroid-receptor interactions also have been investigated, with similar results. In addition, other methods that have allowed the separation of cytoplasmic and nuclear fractions before the trauma of homogenization and centrifugation have confirmed these localizations. The retention of the steroid-receptor complex in the nucleus is an important function in the mechanism by which steroids exert their biochemical actions. For example, estrogenic compounds that do not cause nuclear accumulation for 6 hours or longer (e.g., estriol) do not cause true uterine growth. Thus, not only production of RNA but also retention time in the nucleus is necessary for true expression of steroid activity. The intranuclear steroid-receptor complex binds to the target cell genome as the next step in the sequence of events in the expression of steroid activity. The binding of the steroid-receptor complex to chromatin of the target cell is highly specific. Little binding occurs to nontarget chromatin or with free hormone alone. This association initiates the production of specific messenger ribonucleic acids in the target cell. This evidence suggests that steroid hormones regulate gene expression primarily at a transcriptional level. These mRNAs are then exported to the cytoplasm, where protein synthesis takes place, resulting in alterations in cell growth or physiology that are characteristic of the steroid hormone for that target issue.^9,10,11

The concentration of steroid receptors in target cells is not constant, but varies with the stage of the menstrual cycle or, more importantly, with the recent history of hormonal exposure. Animal studies show that the concentration of sex steroid receptors changes after castration; the mechanism of regulation of receptor content is not the same for each steroid receptor. For example, estrogen treatment not only stimulates the production of estrogen receptors, but also results in the production of new progesterone receptors.¹³ This latter effect allows expression of the sequential nature of estrogen-progesterone action, which is seen so often in nature and probably is the most important aspect of the priming requirement for estrogen in progesterone action. Conversely, progesterone treatment results in a decrease in both progestin and estrogen cytoplasmic receptors.^13,14 This decrease in estrogen receptor may be the major reason why progesterone can exert an antiestrogenic activity.^13,15 The mechanisms by which estrogens and progestogens regulate steroid hormone receptors are not fully understood, and research continues in this area, although it is presumed that nuclear mechanisms are involved. The concentration of steroid receptors in target tissues is not constant, and these receptors must be present for tissues to respond to specific sex hormone treatment. Therefore, the response to any administered sex steroid may vary with the stage of the menstrual cycle and its varying steroidal milieu or with the duration and nature of any previous exposure to steroid treatment.

Knowledge of tissue receptor concentrations may serve a valuable diagnostic function. Perhaps the most important example of the use of estrogen (E_r) and progesterone receptor (P_r) concentration is in breast tumor biopsy specimens taken to determine the potential value of endocrine versus nonendocrine therapy. It was known for many years that some patients with breast cancer would respond to endocrine ablative therapy, whereas others would not. With the knowledge and techniques for E_r and P_r measurement in small tumor biopsy specimens, a positive correlation was found between tumors with positive steroid receptor content and application of endocrine treatments. It is interesting that the emerging predominant treatment for these E_r tumors is tamoxifen, a specific estrogen antagonist that competes with natural estrogens for the receptors in these tumors, thereby obviating the necessity of removing the gonads or adrenals.^16,17

Estrogens

The natural estrogens (estradiol, estrone, and estriol) were isolated in the late 1920s and 1930s. Originally, it was thought that these steroids were orally inactive. More recently, it was learned that micronization of estradiol-17β crystals substantially enhances their oral activity. The first orally active estrogen was the nonsteroidal compound diethylstilbestrol. The first orally active steroidal estrogen, ethynyl estradiol, was developed in 1938 by attaching an ethynyl group at C-17 of the estradiol molecule.¹⁸ Subsequently, mestranol was synthesized by attaching a methyl group to the hydroxyl group at C-3. These two compounds remain the only orally active synthetic estrogens used in birth control formulations today. Mestranol may be less potent under certain circumstances than ethinyl estradiol because mestranol first must be converted to ethinyl estradiol to be active. Mestranol will not bind to the cytoplasmic estrogen receptor; therefore, ethinyl estradiol is the active estrogen for both of these synthetic compounds.¹⁹

Progestogens

The discovery that ethinyl substitution leads to oral potency led to the preparation of ethisterone, an orally active derivative of testosterone. In 1951, it was found that removal of the carbon-19 from ethisterone to form norethindrone did not destroy the oral activity and, most importantly, changed the major hormonal effect from that of an androgen to that of a progestogen. Accordingly, the progestational derivatives of testosterone were designated 19-nortestosterones. The androgenic properties of these compounds, however, were not completely eliminated, and minimal anabolic and androgenic activity remains. Examples of this class of progestogens include norethindrone, norethynodrel, ethynodiol diacetate, and some other related compounds not used in the United States. The second group of 19-nor compounds are gonanes, which have an 18-ethyl instead of an 18-methyl group. They include racemic norgestrel, levonorgestrel, and three newer compounds: gestodene, desogestrel (a pro-drug that must be converted to 3-ketodesogestrel to be biologically active) and norgestimate, which is the 17-acetyl-3-oxime derivative of norgestrel, into which it is rapidly metabolized.

A second group of progestogens became available for use when it was discovered that acetylation of the 17-hydroxy group of 17-hydroxyprogesterone produced oral potency. Acetylation of the 17 position gives oral potency, but an addition at the 6 position is necessary to give sufficient progestional strength for human use, probably by inhibiting degradative metabolism. The chief examples of this class are medroxyprogesterone acetate (MPA), megestrol acetate, and chlormadinone acetate (CMA).¹⁹

These compounds, with the exception of depot medroxyprogesterone acetate and progestogen-only minipills, are used clinically in combination with an ethynyl estrogen. Therefore, it is inappropriate to extrapolate the various biologic activities of the progestogens as determined in animal assays of the pure compound to a clinical situation where a particular ratio of estrogen to progestogen has been chosen to maximize contraceptive efficacy and endometrial control.

The natural sex steroids exert their influence at all levels of reproductive function, including the hypothalamic-pituitary-gonadal endocrine axis as well as the physiologic regulation of the reproductive system, so it is not surprising that the synthetic estrogens and progestogens can exert some regulatory function at many sites in the body. The following sections discuss some of these modes of action.

Nervous System, Hypothalamus, and Pituitary

Sex hormones affect the central nervous system in regions other than the hypothalamopituitary system. For example, the amygdala and cerebellum participate in the feedback effects of progesterone.²⁰ Both estrogen and progesterone inhibit the uptake of norepinephrine by synaptosomes from rat brain.²¹ Both estrogen and progesterone alter monoamine metabolism in the forebrain, midbrain, and hindbrain of gonadectomized mice.²² Kanematsu and Sawyer and Kawakami and Sawyer showed a direct central action of progestogens by their effect on the electroencephalogram (EEG) afterreaction threshold.^23,24 EEG and other studies indicate that progestogens have an effect on the nervous system, cerebral cortical function, and human behavior. EEG evidence points toward a mode of action similar to that of the minor tranquilizers. Large intravenous doses have an anesthetic effect in humans.

The relationship of the corpus luteum hormone to infertility and inhibition of ovulation was under investigation in the early part of this century. By 1966, the biphasic effect of progesterone on ovulation in the rat had been recognized.²⁵

Lesions in the suprachiasmatic region block the progesterone-stimulated release of LH but not that of follicle-stimulating hormone. This finding appears to indicate that progesterone, and probably other progestational compounds as well, can exert qualitative as well as quantitative influences on gonadotropin release. The prevailing level of estrogen is important in the response of the hypothalamopituitary system to progestogens,²⁶ which may explain why intracerebral implants of progesterone can induce or inhibit ovulation in rats, depending on the time of the cycle at which the experiment is performed.²⁷ Conversely, progestogen pretreatment may alter the response to estrogens, perhaps by interfering with the induction of estrogen receptor, and possibly by other mechanisms.²⁸

Experiments attempting to localize the action of hormones to the hypothalamus versus the anterior pituitary are complicated by blood flow through the portal system from the median eminence to the pituitary on the one hand and by technical difficulties causing drugs to travel in the reverse direction on the other.²⁹

Experiments in immature female rats given radioactive progesterone showed little retention of radioactivity in the hypothalamus, but selective and prolonged pituitary uptake.³⁰ More recently, use of the synthetic progestin R 5020 (17α, 21-dimethyl-19-nor-pregn-4,9-diene-3,20-dione) has made possible the identification of 7S progestogen-binding compounds in the cytosol of both hypothalamus and anterior pituitary; such binding was greatly increased by pretreatment of immature rats with estrogen. There have been reports that the 5α-reduced metabolites of progesterone play an important role.³⁰

Clinical observations of the effect of continuous administration of synthetic progestational compounds agree with the findings in laboratory animals. Using 0.4 mg ethynodiol diacetate daily, Mishell and Odell³¹ may have observed an LH surge at the beginning of treatment that is compatible with the positive-feedback effect of progesterone noted by others. Depending on the potency and dose of the administered progestogen, the first inhibitory effect to be observed is a diminution³² or suppression of the midcycle LH surge.³³ This effect has been observed with various 19-norsteroids as well as with CMA and MPA. Elimination of the gonadotropin peak is not always accompanied by an absence of luteal-phase progesterone secretion.^34,35 Occasional shortened luteal phases with lowered levels of plasma progesterone or urinary pregnanediol may be observed.^36,37 With the use of 0.3 to 0.5 mg norethindrone (NET) daily, estrogen excretion was increased, although it did not have the usual biphasic character.^36,38 In many instances, this increasing plasma estrogen level did not trigger corpus luteum formation and secretion³⁹; if it did succeed, defective corpus luteum function was likely to follow. The standard regimen of medroxyprogesterone acetate (Depo-Provera, Upjohn, Kalamazoo, MI) 150 mg intramuscularly every 3 months stops the gonadotropin surge⁴⁰ and prevents the process of ovulation and corpus luteum formation. Large doses of progestogens may depress basal levels of FSH and LH as well. However, even with depressed basal secretion, the response of FSH and LH to an acute stimulus is unaltered.⁴¹ The low levels of progestogen that are released continuously by the norgestrel-containing polymeric silicone (Silastic) implant* permit a considerable degree of gonadotropic activity, as evidenced by sonographic measurements of follicular growth. In some cycles, there is evidence of progesterone secretion; however, in view of the excellent contraceptive efficacy, this effect probably is due to luteinized unruptured follicles, and not to a postovulatory corpus luteum.

It was established during the early clinical trials of oral contraceptives that initiation of treatment early in the cycle was essential for consistent inhibition of ovulation. Matsumoto and colleagues⁴² found by laparotomy that starting treatment by cycle day 6 was a requirement for consistent ovulation inhibition.

Gonadal Effects

Whether progestogens have a direct effect on the gonad or whether the effect is exerted exclusively through changes in gonadotropic control has been debated for several decades. The 19-norsteroids block the ovulatory effect of pregnant mare serum gonadotropin or LH in immature rats^43,44; others have found no such ovulation block with progesterone,⁴⁵ 19-norsteroids,^46,47 or 17-acetoxyprogestins.⁴⁸ In human subjects, Taymor and Rizkallah⁴⁹ found no interference by NET acetate 5 mg/day on ovulation induced by human menopausal gonadotropin/human chorionic gonadotropin (hMG/hCG), and Johansson⁵⁰ elicited rises in plasma progesterone with hCG in subjects given large postovulatory doses of 19-norprogestins or 17-acetoxyprogestins, indicating that the pre-hCG decline in plasma progesterone was not a luteolytic effect.

Anatomic and histologic changes in the ovaries of subjects using progestational steroids are receiving continuing attention. Corpora lutea commonly are found in women who are taking low-dose oral megestrol acetate,⁵¹ but many of these corpora are accompanied by low progesterone output.⁵² Similar observations were made with exposure to 75 μg norgestrel daily.⁵³ Tertiary and graafian follicles appeared normal.

With continued use of large doses (e.g., 1 mg ethynodiol diacetate), early follicular atresia is seen, with degeneration of the granulosa and diminished thecal proliferation. With the use of injectable progestogens, mature follicles disappear, antral follicles are found in about half of the cases, and secondary follicles are present consistently. Cysts or cystic follicles are conspicuous by their absence. With progressive diminution of the steroid dose in oral contraceptives and the introduction of triphasic regimens, more ovarian activity is seen. As might be predicted, more follicular activity is seen when a triphasic oral contraceptive is started on cycle day 5 than if it is started on cycle day 1.⁵⁴

Progestational compounds have been examined for their antifertility potential in male subjects. MPA can compete with testosterone for nuclear androgen receptor, and thus may interfere with important testosterone effects. Theoretically, progestogens could suppress testicular function by competition for androgen receptor, thus reducing the effective level of testosterone at the target tissue by affecting the prevailing gonadotropic environment or by altering androgen biosynthesis, transformation, or metabolic clearance. MPA, for example, can increase the metabolic clearance rate of testosterone. Medrogestone has a direct effect on testicular 17-hydroxylase, and MPA also has a direct effect on steroid biosynthesis at high concentrations, although the enzymes involved have not been identified.

In addition to such antiandrogenic effects (most clearly demonstrated by cyproterone acetate), both C19 and C21 progestogens have androgenic effects in laboratory animals. Synthetic progestogens can potentiate the effect of androgens on a variety of tissues (e.g., kidney, submaxillary gland, preputial gland), as shown by the striking effect on renal β-glucuronidase activity.⁵⁵ The relationship of these androgenic, synandrogenic, and antiandrogenic activities to fertility regulation by synthetic progestins is not well understood.

Vaginal and Cervical Factors

The effects of endogenous progesterone during the luteal phase or during pregnancy are well known. Under the influence of this hormone, the mucus is scanty, viscous, and cellular, with low spinnbarkeit and no ferning. The proteins, enzymes, and electrolytes, but not the pH or the trace elements, are altered, and sperm penetration is inhibited. A single 0.5-mg dose of megestrol acetate has an effect on the penetrability of mucus within 4 hours; the effect wears off after a day.⁵⁶ Comparable changes have been induced by 19-norsteroids^57,58,59 administered orally or from intravaginal or intracervical slow-release devices. With norgestrel treatment at 50 to 75 μg/day, postcoital specimens obtained from the human endocervix showed spermatozoa that were actively motile but incapable of entering the uterine cavity. This finding is consistent with observations made with high dose combination oral contraceptives (norethynodrel 10 mg + mestranol 150 μg) that motile spermatozoa occasionally could be observed within the cervical mucus, although relative impenetrability was the rule.

—The cervical factor generally is considered the chief site of action of low-dose progestogen-only contraceptives. Findings suggest that progestationalized mucus not only acts as a mechanical barrier but also may affect the spermatozoon itself.⁵⁹ On the other hand, contraceptive failures due to occasional omissions of medication (even with high-dose combination oral contraceptives) are well known. This occurrence suggests that the cervical effect of progestational compounds displays individual variation in dose response, or it may be that some populations of spermatozoa have extraordinary penetrating power.

Uterine Effects

As the site of the critical process of nidation, the uterine environment has received a great deal of attention. Unfortunately, the process of nidation varies from species to species, so the relevance of experiments in laboratory animals to human reproduction is open to question.

The human intrauterine environment in the presence of a progesterone-releasing device has been studied. Uterine washings had an inhibitory effect on the oxygen uptake, glucose use,⁶⁰ and peptidase activity of human spermatozoa; this was more apparent at a progesterone release rate of 50 μg/day than at 30 μg/day. These washings also inhibited the in vitro capacitation of rabbit spermatozoa.

Extensive studies of the endometrial histology of women using progestogen-only contraception have been published. With continuous oral administration of CMA,^37,61 postcoital use of levonorgestrel or quingestanol,⁶² or subcutaneous implants of megestrol acetate⁶³ or NET acetate capsules,⁶⁴ a variety of patterns may be seen. A substantial percentage of patients show normal or irregular secretory changes, often with a disparity between the glandular, vascular, and stromal elements. The stroma often shows predecidual changes, patchy hemorrhages, or edema. Some endometria show more advanced progestational changes tending toward the involutional pattern, which is more common and more extreme with the injectable progestogens⁶⁵ (e.g., depo-medroxyprogesterone acetate) or with intrauterine devices releasing progesterone⁶⁶ or levonorgestrel.⁶⁷ Histochemical studies of the endometrium exposed to intrauterine progesterone showed a decrease in zinc, alkaline phosphatase, and β-glucuronidase activity, an increase in acid phosphatase activity, and an eightfold to tenfold increase in the concentration of endometrial progesterone.⁶⁰

More detailed studies of human endometrial epithelial cells⁶⁸ have shown that progesterone, CMA, and MPA induced nuclear differentiation, giant mitochondria, and glycogen accumulation, whereas NET, norethynodrel, dimethisterone, norgestrel, and ethynodiol diacetate did not cause nucleolar differentiation (basket formation), although enlarged mitochondria and glycogen accumulation were seen.

Daily treatment for 6 to 8 weeks with either NET (350 μg/day) or norgestrel (75 μg/day) produced similar endometrial changes as judged by light microscopy and scanning electron microscopy. Many secretory cells were partially or totally denuded, and degenerative changes similar to those seen in menopausal women were noted.⁵⁸

With the advent of locally acting progestogens for contraception, in vitro studies of potency have acquired a new dimension. Organ culture of human endometrium⁴⁹ showed that six progestogens (MPA, NET, NET acetate, norethynodrel, ethynodiol diacetate, norgestrel) induced an increase in tissue glycogen at lower concentrations than did progesterone; however, only MPA had a relative affinity greater than that of progesterone. The potencies and receptor affinities had the same relative order, but differed in relative magnitude. Uniyal and co-workers⁶⁸ studied human endometrial and myometrial norgestrel-binding receptor and found that NET and progesterone showed maximum competition, whereas CMA, testosterone, and corticosterone competed poorly. Norgestrel bound to two endometrial proteins and one myometrial cytosol protein. The relative binding affinities of progesterone, norgestrel, and NET were in the same proportion for endometrial and myometrial receptors. The concentration of binding sites for MPA does not exhibit the cycle variations of the progesterone binder, possibly because MPA has broader specificity and interacts with androgen and corticoid binders as well.⁶⁹

Effects on the Oviduct

Certain surgical procedures in which the oviducts are removed and the ovaries implanted in the uterus are associated with a pregnancy rate of nearly 14%.⁷⁰ Thus, normalcy of the oviduct is not a sine qua non for human fertility. Little work on the effect of synthetic progestogens on the oviducts of humans has been reported. In subjects receiving DMPA or NET enanthate for 18 to 36 months, the mucosa appeared static and markedly suppressed. Additionally, the columnar cells were reduced in height, and the cilia were inconspicuous and stunted. Secretory cells were decreased in number, and they appeared shrunken. Secretory activity, periodic acid-Schiff-reactive material, and vascularity were markedly diminished.⁷¹

Effects on Spermatozoa

Lee and Blandau⁷² studied the effect of progesterone on the swimming speed of spermatozoa by their laser light-scattering technique. At concentrations of 0.2 or 2 μg/ml, there was no effect, but when the progesterone concentration was three orders of magnitude greater than the physiologic level (i.e., approximately 40 μg/ml), a reduction in speed was observed. This finding might be relevant to antifertility mechanisms operative in women using a progesterone-releasing intrauterine device. Progesterone, lynestrenol, and norethynodrel reduce the motility of washed sperm.⁷³ Norethynodrel affects the oxidative metabolism of sperm at a concentration of 32 μg/ml and the glycolytic metabolism at 320 μg/ml. At this concentration, there may be a change in membrane permeability, with consequent loss of cofactors that are essential for sperm metabolism. The action of lynestrenol may involve binding to the plasma membrane, leading to a disruption of a membrane function associated with sperm motility. Progesterone released from an intrauterine device has an effect on sperm oxygen uptake and glucose use as well as on tetracycline binding and release processes.⁷⁴ These changes may represent a capacitation-inhibiting effect of uterine secretion in the presence of progesterone. Continuous low-dose or postcoital administration of levonorgestrel influences intrauterine pH and increases the rate of disappearance of spermatozoa from the uterine cavity.⁷⁵ It appears that progestational compounds have significant effects on sperm fertility, whether by way of systemic administration or by direct exposure in the intrauterine environment.

Postcoital Contraception

Postcoital contraception is designed to interfere with conception between the time of unprotected intercourse at midcycle and before implantation. Presumably, the events immediately after ovulation can be altered by steroid treatment. It was found more than 50 years ago that crude ovarian extracts, presumably showing an estrogenic action, would block nidation in rodents.⁷⁶ The interceptive action of postcoital estrogens in women was first reported by Morris and Van Wagenen⁷⁷ in 1966. Since then, many clinical studies have shown the efficacy of postcoital estrogen administration in the prevention of pregnancy. A variety of estrogens have been used in these trials, including diethylstilbestrol, conjugated equine estrogens, and ethinyl estradiol.^78,79 A significant drawback of these high-dose estrogen treatments has been the high incidence of side effects, in particular, nausea and vomiting.^77,80

Progestogens also have been used as postcoital oral contraceptives. The first used in humans was quingestanol acetate,⁸¹ but because of estrogenic activity of this steroid in rodents,⁸² it is difficult to ascribe any success only to its progestational activity. Other progestogens have been used, and the effectiveness of levonorgestrel⁸³ and norethisterone⁸⁴ as postcoital agents was inadequate. In another study, postcoital use of levonorgestrel had an effect on intrauterine pH and enhanced the disappearance of sperm from the uterine cavity.⁷⁵ These postcoital progestogen studies have been done with a limited number of patients.

More recently, the combination of racemic norgestrel and 50 μg ethinyl estradiol (Ovral, Wyeth-Ayerst, Philadelphia, PA) has been used extensively for postcoital contraception.⁸⁵ If given twice daily for 2 days within 72 hours of unprotected coitus or, preferably, if initiated within 24 hours, effectiveness rates of greater than 75% have been achieved, presumably on the basis of the antiendometrial and possibly tubal effects.⁸⁶

The mechanism of estrogen treatment as a postcoital contraceptive has not been clearly defined. This action of estrogen is postovulatory and not postcoital. In primates, estrogens will not interfere with fertilization and will not terminate pregnancy once implantation has been established. Estrogens interfere with ovum transport in laboratory rodents, but this effect does not seem to be a significant factor in women.^70,87 More importantly, estrogens are luteolytic agents in primates⁸⁷; consequently, treatment may undermine the progestational requirements for endometrial development before implantation. In addition, high-dose estrogen treatment may change the normal pattern of endometrial development, resulting in a tissue that is out of phase with ovum maturation and perhaps unsuitable for implantation.

Despite the improvements that result from the use of the combination estrogen and progestin regimen, side effects, such as nausea and vomiting, still make this regimen less than ideal. Recently, the use of mifepristone (RU 486) as a postcoital interceptive has been tested, and it appears to have advantages over the estrogen and progestin regimen.⁸⁷ Mifepristone is a progesterone antagonist, and because progesterone is essential in the differentiation of the endometrium to make it implantable, the ability to block the actions of progesterone would undermine these endometrial requirements. In a large group of women, RU 486 was as effective as or more effective than the standard steroid regimen and was associated with less nausea and vomiting. More studies are required to confirm these findings, but the excellent outlook for this compound as a substantial improvement over current modalities seems clear.