Sander S. Shapiro
Table Of Contents
Sander S. Shapiro, MD
DIAGNOSIS AND TREATMENT
Failure of the female gonads to discharge ova in a monthly, repetitive manner is referred to as oligo-ovulation. Failure to discharge ova over a prolonged period, usually defined as longer than 3 months, is called anovulation. Anovulation is the normal state for both young women and those beyond their reproductive years. Between 16 and 40 years of age it is, in the absence of pregnancy, an abnormal state and a common gynecologic problem. When experienced over a long period, this condition may be associated with endometrial carcinoma and premature osteoporosis. Among infertile women, anovulation can be found approximately 30% of the time. Most anovulatory women experience either irregular vaginal bleeding or amenorrhea. However, anovulation and amenorrhea are not mutually inclusive states. The intermittent shedding of endometrium (i.e., menses) requires that there be sufficient circulating estrogen to promote tissue proliferation. Ovulation is not a requisite precondition for the production of this estrogen. Thus, although it is rare, eumenorrheic anovulation can occur. Antithetically, ovulation may be experienced by women who are amenorrheic. This happens when the uterus is incapable of responding to the cyclic stimulation and withdrawal of sex steroid hormones. Amenorrhea should not be taken as a certain sign of anovulation.
Rigor in the confirmation of ovulation requires the identification of an ovum outside the ovary, a feat that, short of pregnancy, is too involved in terms of cost and potential hazard to be justified. In most clinical situations, then, the diagnosis of ovulation is imprecise or based on circumstantial evidence. In contrast, anovulation can be determined with a greater sense of certainty.
Ovulation in women usually begins to occur in a cyclic fashion approximately 1 year after menarche. Barring disease, it continues into the 4th decade of life.1 Although ovulation itself is a brief event, the hormonal perturbations and physiologic changes that precede it are complex and lengthy. Much is now known about these alterations, although certain subtleties remain to be discovered.2 The events that lead to ovulation begin several cycles before ovulation with what is thought to be a nonhormonal event: the conversion of a group of primordial follicles into gonadotropin-sensitive primary follicles. These conversions occur continuously in both prepubertal girls and adult women.3 They also occur during pregnancy, while taking birth control pills, and after hypophysectomy, further establishing their independence from the influence of circulating gonadotropins and sex steroids. Follicles that have completed this first step at the very end of a menstrual cycle are boosted toward a secondary follicular status by a rising serum follicle-stimulating hormone (FSH) level. Those that begin the process at other times in the cycle are exposed to an inadequate milieu, fail to progress, and undergo atresia. Only those follicles that are temporally positioned for further stimulation continue to mature. Rising concentrations of FSH seen in the first few days of the menstrual cycle are a pituitary-hypothalamic response to the fall in steroid levels that follow corpus luteal failure. FSH along with low levels of luteinizing hormone (LH) induces the production of a steroid aromatase in primary follicular granulosa cells. The resultant estrogen production makes maturing follicles more sensitive to FSH and at the same time inhibits FSH secretion by feedback inhibition on the pituitary-hypothalamic axis. Thus, a mechanism is established that stimulates the most advanced follicle while removing maturational impetus from those secondary follicles that are lagging developmentally behind the dominant one.
As the dominant follicle outpaces the other follicles that began to mature at about the same time, it produces increasing amounts of 17 β-estradiol. Rising serum levels of this estrogen feed back on the pituitary to induce the midcycle LH surge and ovulation. Studies in pituitary stalk-sectioned monkeys have shown that this preovulatory surge is initiated by estrogen feedback directly on the pituitary gland rather than the hypothalamus. FSH and LH are synthesized by specific cells (gonadotropes) within the anterior pituitary gland. To activate gene expression, plus several other steps in gonadotropin production, these cells require intermittent exposure to gonadotropin-releasing hormone (GnRH). This is provided by hypothalamic neuronal cells that release the hormone in a pulsatile fashion into the pituitary portal circulation. Changes in GnRH pulse height and GnRH frequency are not required for initiation of the preovulatory LH surge.2 A continuous pulsatile stimulus from the hypothalamus, with a fixed interval and pulse height, is sufficient to facilitate the pituitary response. Nevertheless, there are characteristic changes in GnRH secretory patterns over the course of a normal menstrual cycle, and well-documented deviations from this pattern have been shown to occur in most anovulatory situations.
Hypothalamic stimulation of anterior pituitary gonadotropes occurs through pulsatile release of GnRH. Pulse height, duration, and frequency undergo small but perceptible variation throughout the menstrual cycle. Without the drumbeat of GnRH on the pituitary, significant quantities of FSH and LH are not produced and secreted into the circulation. The major alteration in LH secretion at midcycle is the result of a pituitary response to rising estradiol levels.4 Conversely, the midfollicular phase and luteal phase fall in FSH may result from feedback directly onto the hypothalamus. Hypothalamic GnRH secretion is also affected by dopamine, norepinephrine, and endogenous opioid peptides (β-endorphin, enkephalins, dynorphin), but the role of these agents in the ovulatory mechanism has not been established with clarity.
Teleologically, the problems of vertebrate ovulation are those of specificity and timing. To cope with seasonal and other environmental constraints, ovulation in various animals became dependent on specific sensory organs, and thus an association between the ovary and the central nervous system developed.5 Although human ovulation seems to be only marginally dependent on or directed by the external environment, a dynamic association between central nervous system and the ovary has been maintained. Furthermore, as in other mammals, this association has become a two-way system in which feedback has elevated the relationship to a complex level of integrated function.
In some mammals (e.g., lagomorphs), the endocrine status of the female is relatively steady during most of her adult life, with perturbations occurring only in the immediate preovulatory period. In others, including humans, the normal state of the adult female is one of cyclic reproductive alteration, both biochemically and morphologically. The impelling events in this involve an interplay between secretions and cells of the ovary, pituitary, and hypothalamus. In most cases, anovulation results from malfunction in one or more of these areas. Often, a single defect or primary cause interrupts the dynamic equilibrium between ovary and higher centers, resulting in a steady state in the system itself. Occasionally, systemic disease may impair the interaction of the hypothalamic-pituitary-ovarian axis. In addition, multiple subunits of the overall mechanism may fail, with a consequent breakdown in the dynamic feedback, leading to a steady state and anovulation.
Anovulation, then, is not a disease but rather a manifestation of many diverse disease entities. The defects that lead to anovulation are not completely known, even as the extensive interrelationship of the ovulatory system components are incompletely perceived. Lack of a comprehensive understanding of this system inhibits our ability to identify fundamental causes of certain anovulatory states. In addition, the subtle onset and chronicity of some diseases further confound attempts to positively identify first factors. In the infertile woman, anovulation is usually a symptom of great concern, but under other circumstances (e.g., hirsutism, dysfunctional uterine bleeding), the anovulatory state may be of secondary importance and recognizable only by the trained clinician.
Classification of the many entities that may result in failure of ovulation can take numerous forms, depending on the perspective of the clinician or research investigator. Attempts at classification are based either on a primary entity's effect on the normal physiologic mechanism of ovulation or on the entity's relationship to a series of diagnostic tests framed by a specific protocol. On the basis of the physiologic mechanisms described elsewhere in these volumes, etiologic factors can be grouped into five major categories: hypothalamic anovulation, pituitary anovulation, ovarian anovulation, integrative anovulation, and systemic anovulation. The use of such categories and the placement of specific disease states within them should not be considered immutable. As our understanding of the complex interrelationships that lead to ovulation improves, these categories and their constituents will change.
Active participation of the hypothalamus in the initiation and maintenance of ovulation has been amply shown through animal studies involving traumatic ablation of specific nuclei and the histochemical localization of GnRH-secreting neurons in the diencephalon. The extent to which other cell types and portions of the hypothalamus interact with GnRH- secreting cells and these nuclei is only partially understood. Histologic studies have shown that GnRH-secreting neurons are in close proximity to axons of a number of synaptic transmitter-type cells. Moreover, there is pharmacologic evidence that these neurotransmitters are involved in the control of GnRH release.6 Serotonin, dopamine, and norepinephrine levels in the rat hypothalamus vary with the estrous cycle, and they also vary in the peripheral circulation of women during the menstrual cycle.7 These data suggest that catecholamines are involved in the control of GnRH release and may be the link between higher centers and the hypothalamic cycling centers to which estrogen feeds back. Disease in these hypothalamic centers, in higher centers, or along the course of the efferent neurons from these centers can alter the normal cycling mechanism and cause anovulation. In addition, the hypothalamus may fail to respond to changes, in serum steroid concentrations, as a result of intrinsic or higher center defects. Those etiologic factors that affect the hypothalamus directly or indirectly through their influence on other portions of the brain are collectively categorized as the hypothalamic causes of anovulation.
It has long been recognized that subtle environmental influences may alter the menstrual cycle (e.g., the tendency toward synchronization of cycles among women living in close quarters).8 It is not surprising, then, that minor alterations in place or activity may lead to anovulation, as in the syndrome of schoolgirl or summer camp amenorrhea. Most likely, these conditions are caused by the influence of higher centers on the hypothalamus. More severe forms of psychic trauma are also apt to result in anovulation, with a frequency approaching 54%, as in the concentration camps of World War II9 ;little documentation of the hormonal status in such cases is available. In women with outright emotional disturbances that temporarily relate to their amenorrhea, normal concentrations of pituitary gonadotropins and normal prolactin levels have been found, serum estrogen levels have been in the low-to-normal range, and both cortisol and growth hormone levels have been elevated.10 The anterior pituitary responds in a generally normal fashion to releasing factors, and the ovary is capable of responding to pituitary gonadotropins. Pituitary pulsatile secretion of LH is, however, diminished, reflecting a decrease in GnRH neuronal activity. Mounting evidence suggests that elevated corticotropin-releasing factor (CRF) output is the suppressor of GnRH pacemaker activity, which, in turn, prevents pituitary stimulation of folliculogenesis.11,12 Exceptions to the general rule of hypogonadotropism in psychogenic amenorrhea have been reported. Patients with a history of temporally related hypergonadotropic anovulation, vasomotor symptoms, and acute stress have subsequently returned to cyclic menstruation and even pregnancy.13
Anorexia nervosa, a psychiatric-endocrine disease, presents as extreme weight loss, amenorrhea, and abnormal ideation concerning body image and food intake. Women with this condition consider themselves to be overweight and instigate drastic efforts at weight reduction.14 Amenorrhea often begins at the same time that weight reduction starts and occasionally precedes attempts at dieting. The entity may affect as many as 5% of adolescent women.15 When severe, the disease has a significant mortality rate. It is separated categorically from other traumatic forms of psychogenic anovulation because it is not ordinarily initiated by a specific temporal event and because its characteristic picture can be directly related to the metabolic alterations accompanying severe weight loss rather than to initiation of the abnormal behavior patterns. More than one third of the women experiencing anorexia manifest signs of an affective disorder later in their adult lives.16
Patients with anorexia nervosa have low serum gonadotropin levels, with an increase in the FSH/LH ratio. Serum estrogen levels are low, and prolactin levels are low to normal. The decreased estrogens prevent withdrawal bleeding after a progestin challenge and contribute to the development of osteoporosis. Thyroxine levels are normal, but diminished conversion of this hormone to triiodothyronine can result in the development of hypothyroid signs (e.g., bradychardia, dry skin, hypercarotenemia, slow relaxation reflexes). Growth hormone levels are frequently elevated during periods of inanition. Pituitary gonadotropin release patterns regress from the characteristic adult pulsations through the pubertal pattern and become identical to that seen in childhood.17 The pituitary response after a bolus of GnRH is variable, depending on the seriousness of the disease.18 Pulsatile administration of GnRH stimulates pituitary gonadotropin secretion and ovulation, suggesting that the major hormonal defect in anorexia, as in other psychological stress-induced states, lies above the pituitary gland. Pathologic hypothalamic function is also seen by loss of the normal paradoxic response to heat and cold and by the frequent development of diabetes insipidus.16 Coincidentally, CRF levels are elevated in this state.12 Thus, the degree to which anovulation is caused by CRF suppression of GnRH pulsatility or mechanisms related to body fat content may vary.19 Bulimic women with weight loss may present a pathophysiologic picture similar to that of anorexia nervosa. However, many maintain a normal weight while experiencing anovulation and amenorrhea. Provocative testing suggests that their cyclic dysfunction is at the level of the hypothalamus or pituitary.20
Amenorrhea, galactorrhea, and abdominal distention mimicking pregnancy can be induced by nonfactitious, psychogenic means. These simulated pregnancies are sometimes accompanied by a persistent corpus luteum. Contradictory hormonal pictures have been described in these patients, including both elevated and depressed serum LH concentrations.21 An exaggerated LH response to GnRH and an abnormal dexamethasone suppression test have also been found.22 Prolactin levels are usually elevated, and FSH levels are within the normal range.23 Estrogen and progesterone levels resemble those found during the early luteal phase or in Halban's syndrome (persistent corpus luteum). The endometrium may have a secretory histologic appearance.
Psychogenic stress can be manifest through several seemingly different pathways, and the resultant anovulatory state may be associated with either active or quiescent ovaries and a spectrum of gonadotropin levels.13 Most likely, the underlying neuropharmacology is similar in each of the states described here.21,24 The derangement in hypothalamic-pituitary function manifest in pseudocyesis may be the result of an impairment in dopaminergic activity with malfunction at a hypothalamic or higher level. Resolution of the underlying psychological conflict is usually necessary for the resumption of ovulation. When resolution occurs, hypothalamic-pituitary stimulation test responses rapidly return to normal.
Because the dynamic relation between various diencephalic centers and between the hypothalamus and the pituitary is partially governed by the biogenic amines, it is reasonable to expect that drugs affecting their metabolism might alter the ovulatory process.25 Agents such as phenothiazines (e.g., chlorpromazine, perphenazine, thioridazine), tricyclic antidepressants (e.g., imipramine, amitriptyline), reserpine, morphine, cimetidine, metoclopramide, dextroamphetamine, α-methyldopa, and verapamil can cause anovulation. Serum LH and prolactin levels are usually elevated, and FSH levels are normal in patients receiving long-term drug therapy. Discontinuation of an offending agent results in a rapid return of prolactin to normal levels and resumption of ovulatory cycles.26 Inhibitors of prostaglandin synthesis, such as indomethacin, may also affect ovulation. These inhibitors have a direct effect on the central nervous system as well as on the ovary.
Tumors and inflammatory or degenerative lesions in or adjacent to the hypothalamus can alter its functional capacity. These defects tend to affect gonadotropin secretion more frequently and at an earlier time than they do the other pituitary hormones. Both hypogonadotropic and normogonadotropic states can result.
Craniopharyngiomas are the most common tumor at this site, but gliomas, dermoid cysts, meningiomas, and germinomas have been found. Inflammatory lesions, including the reticuloendothelioses and tuberculosis, have also been found in the hypothalamus. In children, such lesions have been associated with precocious puberty; in the adult female, they are more often associated with hypogonadotropic anovulation.27 As in women with psychogenic anovulation, women with such lesions may also have deficits in pituitary function, which can be defined by their response to GnRH stimulation.28 These defects, however, have been found in patients who have already undergone conservative surgery and may be the result of the surgery and not of the primary lesion.
Familial hypogonadism that results from a hypothalamic defect has traditionally been linked with two syndromes: Kallmann's and Laurence-Moon-Biedl. In the former, a heterogeneous group of inheritance patterns have been implicated in a complex involving hypogonadotropic hypogonadism and anosmia.29 In addition to low circulating levels of LH and FSH, these patients have low levels of human pituitary prolactin (PRL) that fail to rise significantly after thyroid-releasing factor injection. An anatomic defect of the olfactory lobes in these patients is associated with migratory failure to the GnRH-secreting cells, with a subsequent lack of stimulation to the gonadotropin-producing cells within the pituitary gland. In males, the deletion of a gene (KAL) on the X chromosome that encodes for a neuronal migratory protein (anosmin) has been implicated in this defect.30 Although the syndrome has been described in and ascribed primarily to males, anosmia with hypogonadotropic hypogonadism also exists in females.31 Both the histology of the ovaries (follicles are all at the primary stage of development) and their response to exogenous gonadotropins indicate that the primary defect is above the ovarian level. The response to GnRH in women with Kallmann's syndrome has been variable, with successful ovulation induction occurring most often when menotropin therapy has been used.32
Laurence-Moon syndrome (polydactyly, obesity, mental retardation, retinitis pigmentosa, and hypogonadism) has been described most frequently in males and has an autosomal-recessive inheritance pattern. Patients have been described as having either hypogonadotropic hypogonadism, presumably because of hypothalamic malfunction, or primary hypogonadism.33 In women with the very similar Bardet-Biedl syndrome, reproductive dysfunction is common, although hypogonadism is infrequent.34
More recently, the application of molecular biologic techniques has identified a cluster of genetic defects among women who experience hypogonadotropic hypogonadism. This group exhibits delayed pubertal development, infertility, and low or absent serum levels of pituitary gonadotropins. Mutations in the FSH-β subunit, PROP-1, GnRH receptor, congenital adrenal hypoplasia, leptin, and leptin receptor genes, in addition to that of KAL, have been implicated.35 In each case, the defect is manifest by a failure to instigate sufficient gonadal activity to produce pubertal development and subsequent reproductive competence.
Anovulatory women who fail to shed endometrium after progestin stimulation and are without a recognizable cause of their condition are categorized as having functional hypothalamic amenorrhea. Although this is one of the most common forms of anovulation, its diagnosis is made by exclusion. When optimally evaluated, a significant proportion of such individuals show a subclinical eating disorder characterized by low fat and eucaloric intake.36 As a group, they are hypoestrogenic and mildly hypercortisolemic, with low or normal circulating levels of pituitary gonadotropins. When given GnRH, they usually respond normally. However, their LH secretory pulse frequency is prolonged, suggesting an abnormality in hypothalamic GnRH output.37 This presentation is essentially identical to that seen in women with non-anorexic weight reduction. Minor abnormalities in the control and secretion of other pituitary hormones(prolactin, growth hormone, thyroid hormone) have also been documented, leading to the suggestion that this clinical presentation is an adaptive contraceptive mechanism in response to an unfavorable psychic or physical environment.38 A high percentage of these women eventually undergo spontaneous return of ovulatory function.
The anterior hypophysis, once considered the “master gland,” now appears to be an intermediary between the brain and peripheral endocrine organs. It secretes a number of proteins, including FSH and LH, the primary hormonal stimuli to follicular maturation. Control in the output of these hormones appears to reside in the hypothalamus and ovary. Steroid feedback to the anterior pituitary acts as a modifier for the primary effects of GnRH. The ability of the anterior hypophysis to respond to GnRH may be altered or prevented by various diseases that affect the gland directly. Steroid hormones may alter the biologic activity of LH secreted by the gland. In addition, the gland may be altered by inherent defects that impair its capacity to secrete biologically active gonadotropins.
The frequent association of anterior pituitary tumors with amenorrhea has been recognized for more than 50 years. In the past, large tumors (more than 10 mm in diameter) usually came to attention because of debilitating symptoms such as a visual field loss or severe headaches. Amenorrhea, which frequently preceded these symptoms by 2 or more years, was often ignored. Today, with women seeking assistance for increasingly subtle symptoms and with the availability of sophisticated diagnostic aids, the diagnosis of pituitary tumor-induced anovulation is being made more frequently.
Large tumors of the anterior pituitary were originally thought to invoke amenorrhea through changes brought about during their growth. The limited space available for expansion within the sella turcica is presumed to cause pressure atrophy of the nontumorous, normal gland, with loss of trophic hormone secretion. Involution of the ovaries without simultaneous thyroid or adrenal atrophy was explained by the existence of a pattern of hierarchical functional loss within the gland. Evidence to support such a concept came from experimental pituitary ablation studies in dogs. These showed that when fractions of the gland were removed serially, the first secretory product to be lost was growth hormone. Subsequently, with excision of large fractions, the gonadotropins were lost. The last secretory products to be found after removal of major gland segments were thyrotropin and corticotropin. The role played by small tumors (less than 10 mm in diameter) in the development of anovulatory states was not resolved through these studies because it was difficult to conceive of their inducing pressure atrophy or a portal circulatory disturbance. The existence of small, nonprolactin-secreting tumors in patients with normal ovulatory function suggests that microadenomas associated with anovulation act by means other than either pressure atrophy of the gland or interruption of portal communication with the hypothalamus.39
When PRL was identified and methods of assay were developed, a new appreciation evolved for the role played by small tumors in the induction of anovulation. High circulating levels of PRL were frequently shown to be associated with anovulation or amenorrhea, and almost all of the small adenomas found in anovulatory women produced the hormone.
The exact mechanism that causes PRL-producing microadenomas to form is unknown.40 Both an anomalous blood supply and the presence of specific oncogenes have been suggested as inciting factors in prolactinoma development. Generally, however, it is thought that high levels of PRL feed back on the hypothalamus to invoke an increase in dopamine production. High levels of hypothalamic dopamine then inhibit GnRH production, and anovulation results. A direct effect of hyperprolactinemia on folliculogenesis has also been postulated but is less well substantiated by experimental data. Small adenomas rarely grow into macroadenomas; they seldom cause debilitation beyond menstrual dysfunction or galactorrhea. Approximately 80% of large anterior pituitary tumors associated with anovulation secrete excess amounts of prolactin.41 It is probable that some large tumors actively inhibit pituitary function by a pressure effect, because a significant minority (20%) of these lesions do not secrete prolactin. However, many so-called functionless tumors have been shown to secrete gonadotropins in a nonphysiologic manner, thereby instigating aberrant folliculogenesis.42 In both forms of pituitary tumor (microadenomas and macroadenomas), the patient's hypogonadotropic status is what ultimately leads to the acyclic state. The mass effect that is produced by macroadenomas of the anterior hypophysis may also result from expansion of sarcoid, tuberculous, cystic, aneurysmal, or other space-occupying lesions.
Destruction of the secretory cells within the pituitary gland may also result from ischemia. Necrosis of this sort, as seen in Sheehan's syndrome and Simmond's disease, is usually global. Whereas an enlarging tumor may sequentially inhibit the secretion of growth hormone, gonadotropin, thyrotropin, and adrenocorticotropin by the gland, ischemic necrosis most often results in near-total loss of secretory capacity. The same may be said of other forms of pituitary destruction, such as infection. In the empty-sella syndrome, conversely, dysfunction is usually partial and anovulation infrequent.43
The stimulation of primary follicles in preparation for ovulation depends on adequate amounts of biologically active gonadotropin secretion. Defects in the pituitary secretory cells that result in gonadotropin deficiency have been shown to account for anovulation in at least a few patients. These defects are in one of the subunits of the hormones, usually the β-subunit. A two-base pair deletion in the FSH-β gene has been found in at least two of these women.44 Thus, in the patient with an FSH-β subunit deficiency, both biologically active LH and FSH-α subunit may be found in the circulation.45
The capacity of pituitary cells to vary the biologic activity of their secretory products is potentially more important to ovulation than isolated defects in gonadotropin synthesis. Evidence is accumulating to suggest that the serum LH level measured by immunologic methods does not vary directly with the serum LH level measured by bioassay. The implications of these findings could be great, both for the role of LH in normal ovarian stimulation and for its role in abnormal states such as polycystic ovarian disease.46 A similar discontinuity has not been observed for FSH.47
The normal ovary is a relatively inactive structure until stimulated. Advanced follicular maturation, ovulation, and steroidogenesis do not proceed in the absence of pituitary gonadotropins. In patients who have never been exposed to endogenous gonadotropins, follicles rarely develop beyond the primary stage. Rising titers of FSH and LH stimulate specific sequential events in the course of follicular maturation. The rapidly increasing serum estrogen concentrations that result are the signal responsible for hypothalamic/pituitary induction of the midcycle LH surge. Hypothetically, then, the ovarian factors responsible for anovulation might include absence of primordial follicles, resistance to gonadotropin stimulation, and failure of steroidogenesis, resulting in inadequate feedback to the pituitary and hypothalamus. In fact, examples of each of these ovarian causes for anovulation have been documented. Intrinsic defects compromising ovarian hormone production cause an increase in serum gonadotropin levels. This is primarily because low levels of circulating estrogens fail to suppress pituitary FSH and LH production. However, ovarian inhibin is also involved in the natural modulation of FSH secretion. Thus, estrogen therapy does not completely suppress serum gonadotropin levels in women with ovarian failure.
Failure to develop or maintain normal numbers of oocytes in a state of developmental arrest until puberty usually results from a genetic defect. The major gene loci involved in normal ovarian development are located on the arms of the X chromosome.48 In addition, case reports provide evidence for the existence of autosomal loci that contribute to normal ovarian development.
The most common karyotypic defect associated with abnormal ovarian development is the 45,X pattern of Turner's syndrome. Variants, including mosaics, make up 20% of the syndromic population. In addition, there are a number of 46,XX females with streak ovaries who do not exhibit Turner's stigmata. Two thirds of these women have a genetic (autosomal-recessive)basis for their streak ovaries, whereas the remainder are thought to result from infarction, infection, or other as-yet undetermined events.49 Although women in this latter group may not be recognized in childhood, lack of pubertal development eventually brings them to the attention of a physician.
Patients with ovarian dysgenesis have the capacity for cyclic hypothalamic and pituitary function. Their menopausal levels of pituitary gonadotropins in adulthood are presumably the result of a hypothalamic response to very low levels of serum estrogen and a lack of ovarian inhibin production. Even in the prepubertal state, however, these women have FSH levels that are elevated for their chronologic age.
PREMATURE OVARIAN FAILURE.
Menopause is a natural event having a mean onset at approximately 50 years of age. It occurs when the ovaries have become depleted of all functionally competent follicles and is accompanied by high serum levels of pituitary gonadotropins and low levels of estrogen. In contrast to menarche, which has begun, over the decades at progressively earlier ages, its temporal appearance has not changed during the past 200 years. When menopause occurs before age 40, it is referred to as premature. So few women (1%) have amenorrhea with hypergonadotropic hypoestrogenism develop at this relatively young age that the presence of a pathologic mechanism is suspected.50 The syndrome presents with either of two histologic findings in the ovaries. Most frequently, follicles are absent. Less often (less than 1%), there are abundant primordial follicles that show no evidence of maturation beyond that of primary follicles. Despite the presence of follicles in only one subset, this clinical presentation is generally referred to as premature ovarian failure. The history of cyclic menstruation serves to distinguish this entity from primary ovarian failure, a clinical presentation that may involve some overlapping etiologies. Although permanent in most cases, on rare occasions, menopause can reverse spontaneously, with return of ovulation and even pregnancy.
Most women who present with premature ovarian failure have either lost their complement of follicles through an active process or failed to produce sufficient numbers during embryogenesis. The relative likelihood of each of these alternative possibilities is unknown. With the exception of a very few Turner's variant women who have had cyclic menstruation, patients experiencing premature ovarian failure have a normal 46,XX karyotype.51
There are only a few documented pedigrees in which multiple members experience premature ovarian failure and have a normal karyotype. Several genes residing on the X chromosome have been proposed as causative agents in such cases but, as yet, none have been compellingly shown to be a culprit.52 Perhaps the most likely suspect for such a role is the fragile X syndrome gene: FMR-1. This is because carriers of the fragile X mutation frequently (13% to 25%) experience premature ovarian failure. However, the mechanism by which ovarian failure is induced remains obscure since expression of the FMR-1 protein appears to be the same as in unaffected women with the normal number of X chromosome trinucleotide repeats. An abnormality of at least one autosomal gene (FOXL2 on chromosome 3) has also been implicated as a possible inducer of premature ovarian failure.53
The relative infrequency of familial premature ovarian failure suggests that the syndrome seldom has a primary genetic basis.54 A more solidly founded explanation for the early exhaustion of follicular activity is an autoimmune reaction. Both systemic autoimmune diseases (e.g., juvenile rheumatoid arthritis) and those involving specific organs (e.g., thyroid [Hashimoto's, Graves], adrenal [Addison's], hypoparathyroidism, juvenile-onset diabetes mellitus, pernicious anemia, idiopathic thrombocytopenic purpura, myasthenia gravis) have been linked with premature ovarian failure. This type of association has been noted in 18% of the women experiencing premature ovarian failure.55 Circulating antibodies to ovarian proteins are sometimes found in patients with premature ovarian failure.56 More often (up to 92%), a less-specific autoimmune response can be shown in these women.57
Failure of other endocrine tissues in women with premature menopause is seen with sufficient frequency to justify a general survey of endocrine tissue function in each newly diagnosed patient. Adrenal insufficiency is the most commonly found associated disorder. Attempts to reverse ovarian dysfunction with glucocorticoids have produced resumption of menses in a few women, but pregnancy has not been reported after this form of therapy.
Other known causes of afollicular hypergonadotropic hypogonadism include mumps oophoritis, and β-galactosemia.58 Hypergonadotropic hypogonadism in the presence of unstimulated follicles(Savage or Gonadotropin-Resistant Ovary Syndrome)has been shown to be caused by at least two different pathophysiologic conditions: a mutation in the β-subunit of FSH59 and a mutation in the FSH receptor gene.60 The former defect has been identified in only two homozygous women, so its incidence remains unknown. In one of these women, conception was achieved through the administration of FSH. In the latter anomaly, also autosomal recessive, an incidence of one in 8399 females was found among a Finnish population. Because the follicles that are present in this condition fail to respond to the high levels of endogenous FSH and LH by undergoing maturation, it has been called the resistant ovary syndrome. Both primary resistance and acquired resistance have been reported.61 The differential diagnosis between this state and that of follicular exhaustion is an important one when pregnancy is desired, because some women who have follicles ovulate in response to estrogen or menotropin administration.62
Anomalous follicular luteinization may result from a failure of ovulation followed by luteinization of the unruptured follicle or from ovulation with inadequate progesterone production. In the luteinized unruptured follicle (LUF) syndrome, basal body temperatures show an elevation consistent with ovulation, serum progesterone levels are elevated over those of the follicular phase, and the endometrium takes on a secretory histologic appearance. Serial ultrasound examinations show the dominant follicle to be intact during the early luteal phase, and laparoscopy fails to show an ovulatory stigma. LUF is thought to occur in 10% of cycles. In some women, it can be repetitive, causing infertility.63 The underlying cause of this defect is unknown. However, because follicular rupture requires the activation of collagenase, a prostaglandin-dependent event, anything that perturbs the cyclo-oxygenase system could be at fault. It is, therefore, not surprising that intake of nonsteroidal anti-inflammatory drugs has been associated with LUF.64 Cessation of these agents has resulted in a return to normal ultrasound-appearing ovulation. In spontaneous LUF, administration of clomiphene citrate has produced ovulation. The degree to which this syndrome produces a low serum progesterone level that is not sufficient to develop a normal secretory endometrium is unknown, and the extent to which the condition overlaps with the inadequate luteal phase syndrome has not been determined.65
Tumors of the ovary may disrupt the ovulatory cycle by destruction of the ovarian parenchyma, stimulation of adjacent ovarian stroma to hormone production, or direct production of steroid hormones. The stimulation of stroma by either a primary or a metastatic tumor is the least well understood of these three mechanisms. A postulate is that the tumor secretes an inducer substance that, in turn, stimulates the ovarian stroma to produce androgens. The androgens may then be converted to estrogens at peripheral tissue sites.66 In at least one case, the tumor secretagogue was shown to be a chorionic-like gonadotropin.67 The androgens and estrogens that result from ovarian tumor growth disrupt hypothalamic cycling activity and cause the patient to become anovulatory.
Ovulation is the end product of a highly integrated set of events involving the hypothalamus, pituitary, and ovary. Perturbations of this axis have not been fully delineated. Events that seem on the surface to be remote from those of ovulation can have a strong influence on the cyclic function of the axis. In some of the entities cited below, functional hypothalamic defects can be shown, but in none of these has a primary mechanistic defect been identified. They have in common a general malfunction of the integrative functions between the involved organs. Although these entities have traditionally been categorized as having hypothalamic etiology, they are placed in a separate section here to emphasize the importance of integrative functions in their induction of anovulation.
POLYCYSTIC OVARIAN SYNDROME.
An illdefined but seemingly ubiquitous condition, polycystic ovarian syndrome includes several metabolic defects, hirsutism, and occasionally obesity in association with anovulation.68 It is sometimes referred to as a disease, but because a means of specific definition is lacking, the presentation is more often called a syndrome. An elevation in circulating androgens, an inversion in the ratio of estrone to estradiol, insulin resistance, a normal-to-elevated serum LH level, and a normal serum FSH level are the common biochemical findings of the disorder.69 Obese women with the syndrome frequently have acanthosis nigricans develop that is concomitant to their insulin resistance. A hallmark of the syndrome is the presence of polycystic ovaries, which may be identified clinically by palpation or by ultrasound.70 However, because there are several well-established conditions (e.g., Cushing's disease, congenital adrenal hyperplasia, and hyperprolactinemic amenorrhea) that can also produce a polycystic ovarian morphology, a determination of polycystic ovarian syndrome must be made by exclusion. Polycystic ovaries are sometimes (7%) found in otherwise healthy women. More often, this ovarian morphology is seen with anovulation (87%), hirsutism (54%), an abnormal LH/FSH ratio, and an elevated serum testosterone level (85%).70 Specific criteria for making a diagnosis of polycystic ovarian syndrome may vary but usually include clinical, ultrasonographic, and biochemical parameters.
The cause or causes of polycystic ovarian syndrome are not positively known.71 In some instances, there is a clear indication that inheritable factors are involved.72 Kindred studies are consistent with an autosomal-dominant form of inheritance. A high incidence of associated insulin resistance (50%) suggests that elevated levels of circulating insulin may stimulate insulin-like growth factor receptors within the ovary, raising steroidogenic enzyme activity and thereby effecting an increase in androgen production.73,74 This thesis, together with the determination that both positive and negative feedback mechanisms are intact, suggests that the syndrome may have its origins directly within the ovary.
Whatever its origin or origins, a relatively effective means of treating its anovulatory component exists with the use of clomiphene citrate and insulin sensitizing agents.75 Less satisfactory, and therefore more controversial, are the means for treating the hirsutism and endometrial hyperplasia that often present in this syndrome.
NONPSYCHOGENIC WEIGHT DISTURBANCES.
Weight loss to below the 10th percentile for body height is usually accompanied by anovulation and amenorrhea. Although the primary defect is unknown, most of the metabolic and physiologic disturbances found in such patients resemble those of anorexia nervosa.16 The major differences seem to be in the degree of hypothalamic dysfunction and the absence of severe psychiatric debilitation. These women usually have low-to-normal levels of circulating gonadotropins and low levels of serum estrogens. Pituitary responsiveness to GnRH is normal, but gonadotropin pulse frequency is depressed, causing some investigators to categorize this entity as a form of functional hypothalamic amenorrhea.38 A gain in weight to above the 10th percentile usually results in return of ovulatory function. Decreased total body fat with a concomitant decrease in aromatization of adrenal androgens has been suggested as the functional defect in this syndrome.
An analogous condition is frequently found in competitive athletes and other women who undertake intense physical training. Ballet dancers, rowers, and runners who cover more than 35 miles per week often experience the luteal phase defect or anovulation.76 They have low gonadotropin and estrogen levels and high serum endorphins, and they show an abnormal adrenocorticotrophic hormone-cortisol response to CRF.77 Despite high caloric intake, their fractional body fat content is low and they experience relative hypoglycemia/hypoinsulinemia associated with glucoregulatory adaptations including hypercortisolemia.37 These alterations may be the underlying reason for their long-observed hypothalamic- pituitary malfunction.78 When hypoestrogenism is prolonged, osteoporosis becomes a problem. Cessation of intense physical activity usually results in fairly prompt resumption of menstruation.79 However, because amenorrhea occurs more frequently in those athletes who experienced menstrual dysfunction before beginning training, cessation of intense exercise does not guarantee initiation of regular ovulation.
In a converse situation, morbid obesity, anovulation has been related to high circulating levels of androgens. These women have normal serum estradiol concentrations. Serum FSH levels are usually depressed and LH levels are slightly elevated.80 Dietary weight loss has been followed by resumption of cyclic menses.
ACUTE AND CHRONIC DISEASE.
Acute, critical illness can induce temporary hypogonadotropic hypogonadism. Severe, chronic debilitating disease of almost any type, such as renal failure, lymphomas, untreated tuberculosis, and cystic fibrosis, can also result in anovulation.81 Adrenal and thyroid disease, although usually not so debilitating, frequently present as anovulation or amenorrhea. In Cushing's syndrome of primary adrenal origin, the etiology of the anovulation is unknown. In those patients with bilateral hyperplasia, a primary abnormality of pituitary or hypothalamic function can usually be shown to accompany the failure to cycle. Normal serum estrogen and gonadotropin levels are present in these women. Adrenal failure can also be accompanied by anovulation. In untreated cases, this is usually of the hypogonadotropic type that accompanies debilitation, but a number of cases with an autoimmune etiology are known. These have been accompanied by high levels of circulating gonadotropins.
Congenital adrenal hyperplasia patients with a 21- or 11-hydroxylase deficiency are frequently anovulatory during adulthood. In part, this is due to less-than-optimal glucocorticoid therapy. The longer these patients remain untreated, the more difficult it becomes to induce ovulation. In the rare instance when congenital adrenal hyperplasia results from a deficiency of 17-hydroxylase, hypoestrogenism and anovulation will continue despite optimal gluco- and mineralocorticoid replacement therapy.82 Ovulation and pregnancy are, nevertheless, possible using GnRH agonists and menotropins plus estrogen therapy to overcome the physiologic abnormalities caused by the block in follicular estrogen production.83
Hyperthyroidism is usually accompanied by an elevated serum estrogen concentration, which is due to altered steroid metabolism and clearance. With higher levels of estrogen, there is a failure of the hypothalamic feedback mechanism similar to that seen in polycystic ovarian syndrome, with normal FSH and elevated LH levels. Because hepatic disease can impair the metabolism of estrogens and obesity can enhance conversion of androgens to estrogens, both conditions can induce a similar picture. Hypothyroidism may be accompanied by elevated serum estrogens and a feedback problem. Anovulation may also occur as a result of the hyperprolactinemia that often accompanies primary hypothyroidism and is a consequence of an increased output of thyroid-releasing factor.
Anovulation is the rule for women taking combination oral contraceptive pills. For most users, there is only a slight delay in the return of ovulation after cessation of therapy and a minimal time lapse between the initial attempt to conceive and pregnancy. However, 0.2% to 3.4% of women discontinuing contraceptive pills fail to ovulate. Although this rate is approximately the same as the rate of secondary amenorrhea among nonpill users,84 there is a persistent suspicion that oral contraceptives leave a residual of anovulation. A retrospective study suggests that postpill anovulation may occur in an epidemiologically different group than does spontaneous anovulation.85 The demonstration that oral contraceptives are or are not an etiologic factor in anovulation would not diminish the need for thorough evaluation of all postpill patients who fail to menstruate, because a number of these women are found to have mild abnormalities of their hypothalamic feedback mechanisms86 and others may have a concomitant disease such as a pituitary tumor.87
|DIAGNOSIS AND TREATMENT|
Anovulation in the young adult is normal only during pregnancy or postpartum lactation. Persistent anovulation unrelated to childbearing is abnormal. Once pregnancy-related states are eliminated from the differential diagnosis, all else is pathology, with a more or less serious prognosis. In the case of adolescent anovulation, the age of the woman, the extent to which pubertal changes have occurred, and evidence of previous menses must be considered in determining the extent to which she should be evaluated. In adults, the complexity of the protocol depends, to a degree, on the patient's expectations. Many women found to be anovulatory during investigation of an infertility problem are eumenorrheic or oligomenorrheic and would not have recognized themselves as being anovulatory. Anovulatory-amenorrheic women, conversely, usually seek medical attention out of concern for what they perceive as an abnormal condition. These women may not desire pregnancy, but they do desire assurance concerning the cause of their condition and its prognosis. Both situations demand attention but with slightly different approaches and emphasis.
Clinical investigations using releasing hormones, gonadotropins, and estrogen competitors in combination with sensitive and specific assays for protein and steroid hormones have given us a great deal of insight into the pathophysiology of the various anovulatory states. These investigations, while showing the potential of various tests of the hypothalamic-pituitary-ovarian axis, have left the clinician in a quandary. He or she must decide to what extent to apply these tests in each case; that is, when a general determination of cause is sufficient and when a thorough application of these tests to establish the exact defect is necessary. The many different stimulation and suppression protocols that have been devised can be exceedingly expensive and demanding on the capacity of the laboratory. In determining the extent and type of protocol to follow in evaluating an anovulatory patient, both the need to know and the ability to determine the exact cause must be considered.
From the clinician's perspective, the problem can be stated by asking four questions: Is the primary defect potentially dangerous to the patient? Can the primary defect be successfully corrected? Will the primary defect alter the patient's capacity to respond to ovulation induction? Will the condition result in long-term morbidity? These questions usually can be answered without resorting to intensive, involved suppression and stimulation tests.
A number of fairly similar protocols have been devised, in general after a classification scheme recommended by the World Health Organization.88 The clinical investigation hinges on determining the patient's levels of estrogen and gonadotropins, which can be done simply by one bioassay (progestin withdrawal test) and one immunoassay (serum FSH). The presence of estrogens indicates an intact ovary. Absence of the hormones results from either an ovarian defect or a lack of ovarian stimulation by pituitary gonadotropins. High levels of gonadotropins point to an ovarian problem. Low or normal levels of gonadotropins suggest that the primary defect does not originate in the ovary.
The initial evaluation of anovulatory women (Fig. 1) should include a thorough history and physical examination, which may well provide the clues that lead to a correct diagnosis. Suspicions of systemic or psychological disease will arise from this first meeting. An initial serum prolactin and thyroid function study (thyroid-stimulating hormone and T4) will direct attention to areas that might not be specifically suspect from the physical examination. Hyperprolactinemia is accompanied by galactorrhea in only 30%of women, and mild hyperthyroid or hypothyroid states frequently go unrecognized unless laboratory tests are undertaken. Although these tests might be delayed until a second visit, they are eventually necessary in all patients other than those with ovarian failure. Should hyperprolactinemia be discerned, an investigation designed to identify or exclude each of its potential causes is undertaken. The systematic elimination of each possible cause is especially important when a drug that cannot be discontinued is a prime suspect.
The anovulatory amenorrheic woman is then given a course of a progestin: one intramuscular injection of 100 mg progesterone in oil, or oral administration of 10 mg medroxyprogesterone acetate for 5 days. The patient who does not experience withdrawal bleeding can be assumed to be hypoestrogenic because the alternative possibility, pregnancy, has already been eliminated by the physical examination or a chemical pregnancy assay.89 Generally, patients with sustained estradiol concentrations less than 30 pg/mL do not experience bleeding after a progestin challenge; the amount of circulating estrogen is simply insufficient to promote endometrial proliferation. As an alternative to the progestin challenge, the blood estradiol level itself can be measured to assess the patient's level of ovarian function. Both methods are generally reliable, but each has potential pitfalls. Occasionally, patients with sustained low estrogen levels experience bleeding after a progestin challenge, although it is generally very light and brief in duration, and their hypoestrogenism may not be immediately recognized. Conversely, estrogen levels may sporadically rise to higher concentrations in some patients who are otherwise profoundly hypoestrogenic. The progestin challenge is simple to perform, less costly than measuring the estradiol level, and also serves to show a functional endometrium and patent outflow tract. Those who do not experience withdrawal bleeding require further evaluation. The demonstration of elevated gonadotropin levels (both FSH and LH) in the menopausal range then points to a diagnosis of ovarian failure(see Table 1).89 In the rare instance when only FSH is in the menopausal range, a pituitary tumor should be suspected.46
FSH, follicle-stimulating hormone.
The need to investigate ovarian failure to determine its exact cause has been questioned on the grounds that a specific diagnosis may not directly benefit the patient. One exception is the infrequent mosaic X,XY, which may present as a female phenotype with gonadal dysgenesis and carries a predisposition to gonadal malignancy.90 The incidence of gonadal malignancy is sufficiently high to justify a determination of karyotype in all young women (younger than 25 to 30 years) who are hypergonadotropic. Neither 45,X nor 46,XX karyotypes predispose to ovarian malignancy, and other associated medical problems can be identified by directed testing. The relatively high rate of thyroid disease, cardiovascular abnormalities, and renal collecting system abnormalities that appear in patients with Turner's syndrome demands investigation. When a normal karyotype accompanies hypergonadotropic hypogonadism in a woman who has not reached menopause, ovarian resistance to pituitary gonadotropins must be considered. So few women with resistant ovaries have been identified and induced to ovulation that the value of laparoscopic ovarian biopsy to differentiate these patients from those with true ovarian failure must be questioned. Women desiring pregnancy may, at this juncture, undergo estrogen stimulation therapy to induce ovulation.62 Histologic demonstration of primordial oocytes by open biopsy should be considered before attempting induction of ovulation using high-dose menotropin therapy. Ovarian histology may also, in rare instances, suggest the existence of an active immune process amenable to glucocorticoid therapy. When a phenotypically normal patient has previously borne children or does not desire pregnancy, evaluation beyond the determination of a postmenopausal level of gonadotropins is, with one exception, not justified. A general survey of endocrine function (especially adrenocorticoid activity) is indicated because of the possibility that autoimmune glandular compromise may be present.
Selective testing in the woman with a low gonadotropin level should be individualized as to the need (see Table 2). If she is of normal weight, does not have a history of drug use, and is normoprolactinemic, it is unnecessary to obtain detailed studies of the sella turcica or hypothalamus. A lateral skull film rules out large nonprolactin-secreting tumors of the pituitary. Microadenomas that do not secrete either prolactin or adrenocorticotropin have not been shown to alter, significantly, a patient's health or capacity for ovulation induction. In patients with hyperprolactinemia, a computed tomographic or nuclear magnetic resonance scan may be obtained if the demonstration of a microadenoma will lead to surgical extirpation. However, when surgery is not contemplated, a lateral skull film will suffice. In this way, the presence of a macroadenoma can be determined at a considerable cost saving. The use of suppression and stimulation tests to determine the existence of a pituitary tumor in hyperprolactinemic patients has proved of only marginal value. The various modes of therapy for microadenomas can be considered if the radiographic study suggests such a diagnosis.
Severe psychic trauma
GnRH, gonadotropin-releasing hormone; FSH, follicle-stimulating hormone.
The value of more specific hypothalamic-pituitary testing for an individual hypogonadotropic patient must be seriously questioned. The trial use of multiple tests for suppression and stimulation of the hypothalamic-pituitary axis has failed to show the efficacy of any one test to differentiate between hypothalamic and pituitary disease. Such tests, however, have given clinical investigators new insights into the workings of both the normal and the abnormal ovulatory axis. Although one may expect that continued investigations of this sort will yield further information, it is important for the practitioner to weigh the relative value of these tests before accepting them into his or her clinical practice. Correspondingly, although determination of LH pulse height and frequency in various pathologic conditions has furthered our pathophysiologic understanding, the routine recording of pituitary secretory patterns has not proved to be of general clinical benefit.4
If the patient desires pregnancy, induction of ovulation can be attempted without a deeper understanding of her specific abnormality, provided that her psychological or nutritional status is not significantly impaired. The choice of medications for induction will not be determined by the results of such tests. A systematic progression from low-dose clomiphene citrate to higher doses, through the addition of human chorionic gonadotropin, with eventual use of menotropins is necessary to establish the proper ovulation regimen for each patient. The use of ovulation-inducing drugs is not indicated in either the evaluation or treatment of anovulatory women who are not desirous of pregnancy. For the woman who is concerned about present or future infertility, extensive testing of hypothalamic function holds no specific benefits. However, all anovulatory patients for whom a definite, specific diagnosis cannot be made should be re-evaluated at periodic intervals; this may be done at the woman's yearly examination.
All amenorrheic women, assuming a normal, potentially responsive endometrium and an intact outflow tract, are anovulatory. Women with a history of intermittent menstruation, conversely, must be proved anovulatory. Because evidence for ovulation is, by and large, circumstantial, several different parameters are usually obtained before the determination is made. Basal body temperature graphs showing a biphasic pattern are usually seen in ovulatory women, but there are instances in which anovulation has coexisted with biphasic temperature patterns. Elevated progesterone levels in nonpregnant women usually indicate ovulation, but here again judgment should be reserved concerning the certainty of ovulation.91 Endometrial biopsy results that show a secretory histologic pattern are an accepted means of verifying ovulation, although evidence has presented that this method is not completely trustworthy in diagnosing the event. Dating a biopsy specimen by the method of Noyes and Hertig92 gives further assurance when endometrial histologic findings correlate with the temperature curve and onset of menses. Discrepancies between these parameters may suggest a failure of ovulation,93 and lack of evidence for ovulation in the eumenorrheic patient then indicates the need for an evaluation.
The fact that menses has occurred suggests that a good part of the hypothalamic-pituitary-ovarian axis is intact. The same is true for amenorrheic women who have withdrawal bleeding after progestin administration. The fact that they do bleed eliminates severe hypothalamic disease and large pituitary adenomas from further consideration. At this point, one may consider the need for further diagnostic testing. If the patient desires pregnancy, induction of ovulation may be undertaken. In the patient desiring future fertility, assurance can be given that pregnancy is attainable and that her anovulation is not because of a progressively worsening condition.
The patient who bleeds after progestin withdrawal or who has spontaneous menses without ovulation and does not desire pregnancy requires counseling concerning the risks she carries for endometrial carcinoma. Sustained endogenous estrogens unopposed by progesterone make her considerably more likely to have the disease develop. She should be made aware of the benefits of intermittent progestin therapy. Similarly, the woman who fails to bleed after a course of progestin should be counseled concerning the risks of osteoporosis, cardiovascular disease, and so forth and the various methods of prophylactic therapy that might be undertaken.
This protocol does not include the direct measurement of estrogens or androgens because the progestin withdrawal test gives at least as much clinically useful information as do the more expensive quantitative serum or urine assays for estrogen. Determination of androgen levels is of value when hirsutism is the primary patient problem and when there is concern about overlooking an androgen-secreting tumor or adrenal hyperplasia in a candidate for induction of ovulation. Both of these latter possibilities should already be suspected, however, because of virilizing signs found during the initial examination.
The protocol outlined in Figure 1 results in economy without significant loss of clinical effectiveness. There will, of course, be times when more extensive testing becomes necessary.
4. Veldhuis J, Rogol A, Samojlik E et al: Role of endogenous opiates in the expression of negative feedback actions of estrogen and androgen on pulsatile properties of luteinizing hormone secretion in man. J Clin Invest 74: 47, 1984
11. Judd S, Wong J, Salonekles S et al: The effect of alprazolam on serum cortisol and luteinizing hormone pulsatility in normal women and in women with stress-related anovulation. J Clin Endocrinol Metabol 80: 818, 1995
15. Rome E, Imrie R, Rybicki et al: Prevalence of abnormal eating attitudes and behaviors in hospital-based primary and tertiary care clinics: A window of opportunity? J Pediatr Adolesc Gynecol 9: 133, 1996
18. Cornelia J, Binsbergen V, Hergan J et al: A comparative and longitudinal study on endocrine changes related to ovarian function in patients with anorexia nervosa. J Clin Endocrinol Metabol 71: 705, 1990
29. Schwanzel-Fukuda M, Bick D, Plaff D: Luteinizing hormone-releasing hormone(LHRH)-expressing cells do not migrate normally in an inherited hypogonadal (Kallmann's) syndrome. Mol Brain Res 6: 311, 1989
31. Georgopoulos N, Prolong F, Seidman C et al: Genetic heterogeneity evidenced by low incidence of KAL-1 gene mutations in sporadic cases of gonadotropin- releasing hormone deficiency. J Clin Endocrinol Metab 82: 213, 1997
46. Suzinami H, Masaoka H, Korzumi Y et al: Biological and immunological characterization of human luteinizing hormone discharge by the stimulation of synthetic luteinizing hormone-releasing hormone in normal and ovulating women. Endocrinol Jpn 29: 523, 1982
56. Winquist O, Gebre-Medhin G, Gustaffson J et al: Identification of the main gonadal autoantigens in patients with adrenal insufficiency and associated ovarian failure. J Clin Endocrinol Metabol 80: 1717, 1995
64. Smith G, Roberts R, Hall C et al: Reversible ovulatory failure associated with the development of luteinized unruptured follicles in women with inflammatory arthritis taking nonsteroidal anti-inflammatory drugs. Br J Rheum 35: 458, 1996
74. Nestler J, Jakuowicz D: Decrease in ovarian cytochrome P450c 17a activity and serum free testosterone after reduction of insulin secretion in polycystic ovary syndrome. N Engl J Med 335: 617, 1996
83. Rabinovici J, Blankstein J, Goldman B et al: In vitro fertilization and primary embryonic cleavage are possible in 17a hydroxylase deficiency despite extremely low intrafollicular 17 β estradiol. J Clin Endocrinol Metabol 68: 693, 1989
86. Santen R, Friend J, Trojanowski E et al: Prolonged negative feedback suppression after estradiol administration: Proposed mechanism of eugonadal secondary amenorrhea. J Clin Endocrinol Metabol 47: 1220, 1978
91. Wathen N, Perry L, Lilford R et al: Interpretation of single progesterone measurements in diagnosis of anovulation and defective luteal phase. Br Med J 288: 7, 1984