The mean age of menarche is 12.8 years. Normally, this event marks the completion of puberty. The onset of regular cyclicity in the menstrual cycle is determined by the duration of the maturation process of the HPO axis, which is quite variable. As a result, anovulation can occur in 50% to 80% of girls 2 years after menarche, and in more than 20% of girls, it can persist until 5 years after menarche. This period of time can be frustrating for many girls and their parents. The evaluation should include a clinical examination and reassurance. If positive findings are noted, limited, pertinent diagnostic tests (see later) are indicated.
Amenorrhea can be defined as either the absence of menarche by age 16 or no menses for more than three cycles in an individual who has previously had cyclic menses. The definition, although arbitrary, nonetheless gives a general guideline to the clinician for further evaluation. Although amenorrhea does not cause harm, in the absence of pregnancy, it may be a sign of genetic, endocrine, and/or anatomic abnormalities. If the outflow tract is intact, amenorrhea is most likely the result of disruption in the HPO axis. These aberrations can affect any level of control in the menstrual cycle and, thus, result in menstrual abnormalities.
Amenorrhea was formerly classified as primary or secondary depending on whether the individual had experienced menses in the past. This classification may lead to misdiagnosis of the cause of amenorrhea. Although primary amenorrhea is more often associated with genetic and anatomic abnormalities, each individual should be assessed by means of the history and clinical findings, including the presence or absence of secondary sexual characteristics (Table 13–3). The causes of amenorrhea are grouped according to the level of involvement in the regulatory systems that govern normal menstrual activity (ie, hypothalamic, pituitary, ovarian, and uterine). An algorithm for the workup of amenorrhea or oligomenorrhea in the presence of secondary sexual characteristics is illustrated in Figure 13–9.
Table 13–3 Assessment of Patients with Amenorrhea. ||Download (.pdf)
Table 13–3 Assessment of Patients with Amenorrhea.
|I. Absent breast development; uterus present|
|A. Gonadal failure|
|1. Gonadal agenesis|
|2. Gonadal dysgenesis|
|a. 45,X (Turner syndrome)|
|b. 46,X abnormal X (eg, short- or long-arm deletion)|
|c. Mosaicism (eg, X/XX, X/XX/XXX)|
|d. 46,XX or 46,XY (Swyer syndrome) gonadal dysgenesis|
|B. Defects in estrogen biosynthesis (46, XX)|
|1. 17,20-Lyase deficiency|
|2. CYP17α deficiency|
|C. Hypothalamic failure secondary to inadequate GnRH release|
|1. Insufficient GnRH secretion|
|b. Anorexia nervosa and bulimia|
|c. CNS neoplasm (craniopharyngioma, gliomas)|
|d. Excessive exercise|
|e. Constitutional delay|
|2. Inadequate GnRH synthesis (Kallmann syndrome)|
|3. Developmental anatomic abnormalities in central nervous system|
|D. Pituitary failure|
|1. Isolated gonadotropin insufficiency|
|2. GnRH resistance|
|3. Pituitary tumors (hyperprolactinemia)|
|4. Pituitary insufficiency|
|a. Infections (mumps, encephalitis)|
|b. Newborn kernicterus|
|5. Prepubertal hypothyroidism|
|II. Breast development; uterus absent|
|A. Androgen resistance (androgen insensitivity syndrome)|
|B. Congenital absence of uterus (utero-vaginal agenesis)|
|III. Absent breast development; uterus absent|
|A. Defects in testosterone biosynthesis (46,XY)|
|1. 17,20-Lyase deficiency|
|2. CYP17α deficiency|
|3. 17β-Hydroxysteroid dehydrogenase deficiency|
|B. Testicular regression syndrome (46,XY)|
|IV. Breast development; uterus present|
|B. Hypothalamic etiology|
|2. Anorexia nervosa and bulimia|
|3. Psychogenic (depression)|
|4. CNS neoplasm|
|5. Chronic disease|
|C. Pituitary etiology|
|1. Pituitary tumors (hyperprolactinemia)|
|2. Pituitary insufficiency|
|a. Hypotensive event (Sheehan syndrome)|
|c. Autoimmune destruction|
|d. Iatrogenic (surgery, radiation)|
|D. Ovarian etiology|
|a. Mosaicism (46,XX/XO,XX/XY)|
|b. Autoimmune destruction|
|c. Iatrogenic (radiation, chemotherapy)|
|d. Fragile X syndrome|
|2. Resistant ovarian syndrome (Savage syndrome)|
|E. Chronic estrogenized anovulation|
|b. Non classical congenital adrenal hyperplasia|
|c. Cushing syndrome|
|d. Androgen secreting tumors|
|a. Adrenal insufficiency|
|b. Thyroid disorders|
|F. Outflow tract|
|1. Congenital abnormalities|
|a. Transvaginal septum|
|b. Imperforate hymen|
|2. Asherman syndrome|
An algorithm for women who have developed secondary sexual characteristics, have experienced menarche, and are presenting with amenorrhea or oligomenorrhea. A negative human chorionic gonadotropin test documents the nonpregnant state. All women should be screened for known causes of oligomenorrhea, whether or not hyperandrogenism is present. Hyperandrogenism suggests either polycystic ovarian syndrome (PCOS) or late-onset congenital adrenal hyperplasia (CAH). If rapid in onset, progressive virilization requires an evaluation for an androgen-secreting tumor. Management of late-onset CAH and PCOS is similar. Clinical history and physical findings suggesting hypoestrogenism, stress, or thyroid dysfunction are indications for measurement of estradiol (E2), luteinizing hormone (LH), and free thyroid hormone (FT4) levels in plasma (APS, autoimmune polyglandular syndrome). *As indicated (refer to text).
The hypothalamus is the source of GnRH, which directs the synthesis and secretion of pituitary gonadotropins. Dysfunction at this level leads to hypogonadotropic hypogonadism or eugonadotropic hypogonadism. Disorders of GnRH production can result in a wide range of clinical manifestations. The individual's appearance is dependent on the age at onset and the degree of dysfunction.
Isolated GnRH deficiency results in hypogonadotropic hypogonadism. Female patients present with amenorrhea, and females and males present with absent or incomplete pubertal development secondary to absent or diminished sex steroids (estradiol in females, testosterone in males). They have normal stature with a eunuchoidal body habitus. Because the adrenal glands are unaffected by the absence of GnRH, body hair distribution is not affected.
Several genetic lesions associated with GnRH deficiency have been described. The best characterized form of GnRH deficiency is Kallmann syndrome, which involves the Kal-1 gene as well as a number of other genes. The Kal-1 gene normally codes for anosmin, an adhesion molecule that appears to be involved in the migration of GnRH and olfactory neurons from the olfactory placode to the hypothalamus. The Kal-1 gene is located on the short arm of the X chromosome. Most cases of Kallmann syndrome are sporadic, although the disorder has also been observed to have a familial pattern, and most often it is transmitted by X-linked recessive inheritance. Autosomal recessive and dominant patterns have been reported but are much less common. When mutations exist in the Kal-1 gene, there may be associated defects, including anosmia and, less frequently, midline facial defects, renal anomalies, and neurologic deficiency. The disorder affects both sexes, but because of the X-linked inheritance pattern it is more common in boys. Unlike males, the specific genetic mutations in the Kal-1 gene in females with hypogonadotropic hypogonadism have not been identified, suggesting there are other genetic mutations that cause this disorder. Recently, it was determined that heterozygous mutations in fibroblast growth factor receptor 1 (FGFR1) gene, known as the Kal-2, were identified in cases of Kallmann syndrome. FGFR1 and Kal-1 are coexpressed at the same sites during development and are thought to have a functional interaction. Mutations in the Kiss-1 derived peptide receptor, GPR54, have also been identified in these patients. Several mechanisms of how GPR54 may regulate gonadotropin production via GnRH function or production have been proposed. Several studies have shown that females with presumed Kallmann syndrome demonstrate variable responses to exogenous GnRH administration, which suggests a GnRH receptor defect. In fact, mutations in the GnRH receptor have been identified in both sexes and are inherited in an autosomal recessive fashion.
Management of hypogonadotropic hypogonadism involves scheduled hormone replacement therapy (HRT) to stimulate the development of secondary sexual characteristics and increase bone mineral density. If pregnancy is desired, treatment involves the administration of pulsatile GnRH or gonadotropin treatment.
Functional hypothalamic amenorrhea—Functional hypothalamic amenorrhea is one of the most common types of amenorrhea and accounts for 15% to 35% of cases. It is an endocrine disorder, although the exact mechanism has not been definitively determined. It is characterized by a reduced GnRH drive (decrease in pulse frequency and amplitude), leading to low or low-normal serum levels of FSH and LH and resulting in anovulation. The ratio of serum FSH to LH in these patients is often equivalent to that of a prepubertal female with a relative FSH dominance.
The adipocyte hormone, leptin, has been implicated in the development of this disorder. Leptin is an important nutritional satiety factor, but it is also necessary for maturation of the reproductive system. The potential link to the reproductive system is thought to be through leptin receptors, which have been identified in the hypothalamus and gonadotropes. This is supported by the observation that leptin can stimulate GnRH pulsatility and gonadotropin secretion. Several studies suggest that women with functional hypothalamic amenorrhea have lower serum leptin levels in comparison with eumenorrheic controls. This relative deficiency may lead to dysfunctional release of GnRH and subsequent development of functional hypothalamic amenorrhea.
Abnormal, and often subtle, activation of the hypothalamic-pituitary-adrenal axis is associated with functional hypothalamic amenorrhea. The inciting event may be excessive production of corticotropin-releasing hormone (CRH), which has been shown to decrease the pulse frequency of GnRH and increase cortisol levels in vivo. In contrast, another study suggests that although acute elevations of CRH can suppress GnRH release, this suppression cannot be maintained with CRH alone. The cause of functional hypothalamic amenorrhea often remains unclear, but the associated hypercortisolemia suggests that it is preceded by psychologic stress, strenuous exercise, or poor nutrition. There is support for the concept that these factors may act synergistically to further suppress GnRH drive. In fact, patients with functional hypothalamic amenorrhea resulting from psychologic stress are usually high achievers who have dysfunctional coping mechanisms when dealing with daily stress. The severity of hypothalamic suppression is reflected by the clinical manifestations. The significant interpatient variability in the degree of psychologic or metabolic stress required to induce a menstrual disturbance explains the heterogeneity of clinical presentations, ranging from luteal phase defects to anovulation with erratic bleeding to amenorrhea.
Functional hypothalamic amenorrhea is reversible. Interestingly, the factors that have predicted the rate of recovery are body mass index and basal cortisol levels. When patients recover, ovulation is preceded by return of cortisol levels to baseline. Some experts have shown that cognitive behavioral therapy, teaching the patient how to cope with stress—and nutritional consultation—reverse this condition. Complete reversal may be less likely if the functional insult occurs during the period of peripubertal maturation of the HPO axis.
The diagnosis of functional hypothalamic amenorrhea can be made if the FSH-LH ratio is greater than 1 in the presence of hypoestrogenemia. However, a minor disturbance of hypothalamic dysfunction may present with normal laboratory findings, a clinical history that coincides with stress, and a negative evaluation for other causes of anovulation. Interestingly, most of these patients, despite hypoestrogenism, do not have symptoms. The estrogen status should still be evaluated given the strong correlation between hypoestrogenemia and the development of osteoporosis. Estrogen status can be determined by means of the progesterone withdrawal test or by measurement of serum estradiol (<50 pg/mL). If there is no withdrawal or the estradiol is less than 50 pg/mL, HRT with combination contraceptive hormones (or traditional HRT) should be instituted. If withdrawal bleeding occurs, any cyclic progestin-containing therapy is adequate to combat unopposed estrogen and the development of endometrial hyperplasia.
Amenorrhea in the female athlete—Hypothalamic dysfunction has been observed in female athletes. Competitors in events such as gymnastics, ballet, marathon running, and diving can show menstrual irregularities ranging from luteal phase defects to amenorrhea. The female athletic triad as defined by the American College of Sports Medicine is characterized by disordered eating, amenorrhea, and osteoporosis. The associated nutritional deficiencies can lead to impaired growth and delayed sexual maturation. The neuroendocrine abnormalities are similar to those of women with functional hypothalamic amenorrhea.
These patients have very low body fat, often below the tenth percentile. There is evidence that a negative correlation between body fat and menstrual irregularities exists. In addition, there appears to be a critical body fat level that must be present in order to have a functioning reproductive system. Several studies have shown that these amenorrheic athletes have significantly lower serum leptin levels, which further supports leptin's role as a mediator between nutritional status and the reproductive system. The strenuous exercise these athletes engage in amplifies the effects of the associated nutritional deficiency. This synergism causes severe suppression of GnRH, leading to the low estradiol levels.
Amenorrhea alone is not harmful. However, low serum estradiol over a period of time may lead to osteoporosis and delayed puberty. An analysis of estrogen status may be obtained with measurement of serum estradiol levels or with the progestin withdrawal test (see earlier). If estrogen is low, bone mineral density (BMD) should be assessed by dual-energy x-ray absorptiometry (DXA) scan. All patients diagnosed with female athletic triad need combination contraceptive therapy or HRT.
Amenorrhea associated with eating disorders—Anorexia nervosa is a disorder characterized by relentless dieting in pursuit of a thin body habitus. Approximately 95% of cases occur in females, and the onset is chiefly in adolescence. The clinical features include extreme weight loss leading to a body weight less than 85% of normal for age and height, a distorted body image, and intense fear of gaining weight. These patients usually have a preoccupation with food and are hyperactive, with an obsessive-compulsive personality. The associated symptoms include hypothermia, mild bradycardia, dry skin, constipation, and symptoms of hypoestrogenemia. Furthermore, as part of the diagnostic criteria, they must experience at least 3 months of no menses.
The dysfunction in the neuroendocrine system is similar to but often more severe than that described in association with functional hypothalamic amenorrhea. The severe reduction in GnRH pulsatility leads to suppression of FSH and LH secretion, possibly to undetectable levels, and results in anovulation and low serum estradiol levels. Given the severe psychologic and metabolic stress experienced by these individuals, the hypothalamic-pituitary-adrenal axis is activated. The circadian rhythm of adrenal secretion is maintained, but both cortisol production and plasma cortisol levels are persistently elevated secondary to increased pituitary secretion of adrenocorticotropic hormone (ACTH). Serum leptin levels in these individuals are significantly lower than normal healthy controls and correlate with percentage of body fat and body weight. A rise in leptin levels in response to dietary treatment is associated with a subsequent rise in gonadotropin levels. This further suggests leptin's role as a potential link between energy stores and the reproductive system.
The self-induced starvation state associated with anorexia nervosa leads to additional endocrine abnormalities not observed in other causes of hypothalamic amenorrhea. For instance, thyroid hormone metabolism is altered. TSH and T4 levels are in the low-normal range, but T3 levels are usually below normal. This is attributable to decreased peripheral conversion of T4 to T3 and increased conversion of T4 to the metabolically inactive thyroid hormone, reverse T3—a change that often resembles other states of starvation. This may be a protective mechanism in that the relative hypothyroid state attempts to reduce basal metabolic function in response to a highly catabolic state.
Bulimia occurs in about half of patients with anorexia and is defined as binge eating followed by self-induced purging. Not all bulimics have low body weight—in fact, normal-weight bulimic individuals are much more common. These patients also have a variety of neuroendocrine aberrations—often to a lesser degree than those with anorexia—which also lead to menstrual disturbances. Leptin levels are lower than in matched controls but not as low as in individuals with anorexia nervosa. They also have neurotransmitter abnormalities—notably low serotonin levels—which might help explain the often coexisting psychologic difficulties.
Anorexia nervosa is a life-threatening illness with a significant mortality rate due to its metabolic consequences. Anorexic patients should be considered for inpatient therapy and management with a multidisciplinary approach that includes nutritional counseling and psychotherapy. Force-feeding may be necessary in some patients. If weight gain cannot be achieved with oral intake, meals may need to be supplemented by enteral or parenteral feeding. Because anorexia nervosa is a hypoestrogenic state and there is a high potential for the development of osteoporosis, all patients should receive hormone therapy either in the form of HRT or combination contraceptive pills. It is important to note that many patients with anorexia will continue to have reduced bone size regardless of sex steroid administration presumably due to the concurrent malnutrition and metabolic compromise. Weight rehabilitation and recovery is associated with bone accretion; however, the loss of bone may not be completely reversible.
In summary, the hypothalamic amenorrhea endocrine syndromes are probably a continuum of disordered eating and nutritional deficiencies resulting in increasingly severe abnormalities in the reproductive system. Furthermore, the age at onset affects the potential complications of these disorders. If low estradiol levels are present before age 20, bone mineralization may be profoundly affected, because this period is critical for building peak bone mass. In addition, if these conditions occur prior to puberty, it may result in stunted growth and delayed development of secondary sexual characteristics.
Numerous anatomic abnormalities within the central nervous system can result in menstrual disturbances. These include developmental defects, brain tumors, and infiltrative disorders. The most common anatomic lesion associated with delayed puberty and amenorrhea is a craniopharyngioma. It is derived from Rathke pouch and extends into the hypothalamus, pituitary, and third ventricle. The symptoms include headaches, visual loss, and hypoestrogenism.
Infiltrative disorders that involve the hypothalamus are uncommon but can result from systemic diseases, including sarcoidosis, histiocytosis, hemochromatosis, and lymphoma. These diseases do not initially present with amenorrhea. However, in the presence of these diseases, the hypothalamus may be affected, so they should be part of the differential diagnosis of amenorrhea.
There are a few genetic mutations affecting the pituitary that cause amenorrhea. Rare autosomal recessive mutations may cause deficiencies in FSH, LH, TSH, prolactin (PRL), and growth hormone (GH). The clinical manifestations may include delayed puberty, a hypoestrogenic state, and infertility.
A deficiency in FSH and LH may be a result of GnRH receptor gene mutations. Such mutations are primarily compound heterozygous mutations that affect GnRH receptor-dependent signal transduction. The phenotype of these individuals is similar to that of those with isolated GnRH deficiency. In fact, some investigators speculate that these receptor mutations may be the cause of isolated GnRH deficiency in women, given that no mutations have yet been identified in the ligand or Kal-1 gene. The estimated prevalence of GnRH receptor mutations in women with hypothalamic amenorrhea is 2%. In a family with other affected females, the prevalence is 7%.
Other rare genetic defects have been associated with amenorrheic women. Mutations in the FSHβ gene have been reported. These have an autosomal recessive pattern of inheritance and lead to low serum FSH and estradiol levels and high plasma LH levels. The clinical features include minimal development of secondary sexual characteristics and amenorrhea with no history of menses. Combined hormone deficiencies have also been described. Mutations in Prop-1, a pituitary transcription factor for Pit-1, lead to deficiencies in gonadotropins, TSH, PRL, and GH. These patients present with stunted growth, hypothyroidism, and delayed puberty in addition to amenorrhea.
Hyperprolactinemia is one of the most common causes of amenorrhea, accounting for 15% to 30% of cases. In the absence of pregnancy or postpartum lactation, persistently elevated PRL is almost always associated with a hypothalamic-pituitary disorder. Normal PRL secretion is regulated by several stimulatory and inhibitory factors (see Chapter 4). PRL secretion is primarily under tonic inhibition by dopamine, so that any interference with dopamine synthesis or transport from the hypothalamus may result in elevated PRL levels. In addition to menstrual disturbances, individuals with hyperprolactinemia may present with galactorrhea. In fact, hyperprolactinemia is a common cause of galactorrhea, and up to 80% of patients with amenorrhea and galactorrhea have elevated PRL levels. Other associated symptoms include headaches, visual field defects, infertility, and osteopenia.
The mechanism whereby hyperprolactinemia causes amenorrhea is not completely known. Studies have shown that PRL can affect the reproductive system in several ways. PRL receptors have been identified on GnRH neurons and may directly suppress GnRH secretion. Others have postulated that elevated PRL levels inhibit GnRH pulsatility indirectly by increasing other neuromodulators such as endogenous opioids. There is also evidence that GnRH receptors on the pituitary may be downregulated in the presence of hyperprolactinemia. Furthermore, PRL may affect the ovaries by altering ovarian progesterone secretion and estrogen synthesis. The best data now available suggest that hyperprolactinemia causes amenorrhea primarily by suppression of GnRH secretion.
Approximately half of patients with elevated PRL levels have radiologic evidence of a pituitary tumor. The most common type is a PRL-secreting tumor (prolactinoma), accounting for 40% to 50% of pituitary tumors. Prolactinomas are composed mainly of lactotrophs, and they secrete PRL. Occasionally these tumors secrete both GH and PRL.
The diagnosis of a pituitary adenoma is usually made by examination of the pituitary with magnetic resonance imaging (MRI). These tumors are categorized into two groups based on their dimensions—microadenomas are those less than 10 mm in diameter and macroadenomas are larger than 10 mm. These tumors are usually located in the lateral wings of the anterior pituitary. Microadenomas are typically wholly contained within the confines of the pituitary gland. Very rarely, a microadenoma infiltrates the surrounding tissue, including the dura, cavernous sinus, or adjacent skull base. A macroadenoma may expand farther and grow out of the sella to impinge on surrounding structures, such as the optic chiasm or may extend into the sphenoid sinus. As a result, macroadenomas are more frequently associated with severe headaches, visual field defects, and ophthalmoplegia. The incidence of a microadenoma progressing to a macroadenoma is relatively low, only 3% to 7%. During pregnancy the risk of a microprolactinoma enlarging is also low, but in the presence of a macroprolactinoma, the chance of tumor growth is up to 25%.
Some investigators have found a correlation between pituitary adenoma size and serum PRL levels. If the serum PRL is less than 100 ng/mL, a microprolactinoma is more likely, whereas if it is greater than 100 ng/mL, a macroprolactinoma is often present. Although this correlation has been reported, the evidence supporting it is not strong. In fact, low PRL levels may also be associated with nonfunctioning macroadenomas. This is thought to reflect impingement of the adenoma on the pituitary stalk leading to reduced dopamine secretion and loss of the tonic suppression of prolactin secretion. These nonfunctioning tumors may synthesize glycoproteins such as FSH, LH, or their free alpha subunits. Rarely, functioning tumors may arise from other pituitary cells, resulting in excessive hormone secretion. If a macroadenoma is present, measurement of IGF-I, alpha subunit, TSH, and 24-hour urinary cortisol will begin to exclude other functioning adenomas.
Other tumors of nonpituitary origin may also result in delayed puberty and amenorrhea. The most common of these is a craniopharyngioma. Although craniopharyngiomas are commonly located in the suprasellar region, they can also involve the sella. These tumors have not been shown to produce hormones, but because they may compress the infundibulum, they can interfere with the tonic inhibition of PRL secretion and result in mildly elevated PRL levels.
Hyperprolactinemia in a patient with amenorrhea is defined as a PRL level greater than 20 ng/mL, although the limit of normal may vary between laboratories. Normal PRL release follows a sleep-circadian rhythm, but PRL may also be secreted in response to stress, physical exercise, breast stimulation, or a meal. Therefore, PRL should be measured in the midmorning hours and in the fasting state. Other causes of mildly elevated PRL include medications such as oral contraceptives, neuroleptics, tricyclic antidepressants, metoclopramide, methyldopa, and verapamil. Hyperprolactinemia has also been observed in several chronic diseases, including cirrhosis and renal disease. Furthermore, inflammatory diseases such as sarcoidosis and histiocytosis can infiltrate the hypothalamus or pituitary and result in hyperprolactinemia. Elevated PRL levels may be a physiologic response. During pregnancy, PRL levels may be two to four times baseline. With postpartum breast-feeding, the PRL level should be below 100 ng/mL after 7 days and below 50 ng/mL after 3 months. If a woman is not breast-feeding, PRL levels should return to baseline by 7 days postpartum.
Persistently elevated PRL levels may also be present in primary hypothyroidism. Approximately 40% of patients with primary hypothyroidism present with a minimal increase in prolactin (25-30 ng/mL), and 10% present with even higher serum levels. Individuals with primary hypothyroidism have an increase in thyrotropin-releasing hormone (TRH) from the hypothalamus, which stimulates TSH and PRL release and leads to hyperprolactinemia. Patients with long-standing primary hypothyroidism may eventually manifest profound pituitary enlargement due to hypertrophy of thyrotrophs. This mass effect with elevated PRL levels mimics a prolactinoma. Therefore, all patients with hyperprolactinemia should have their thyroid function investigated to exclude hypothyroidism as the cause.
Prolactinomas are the most common cause of persistent hyperprolactinemia. All patients with elevated PRL levels should have the test repeated for verification. In addition to blood tests, a careful clinical and pharmacologic history and physical examination should be performed to exclude other causes of hyperprolactinemia. If elevated PRL levels persist or if any measurement is found to be above 100 ng/mL, MRI of the hypothalamic-pituitary region should be performed. If a microadenoma is observed, the diagnosis of microprolactinoma can be made. If a macroadenoma is observed, other pituitary hormones should be measured to exclude other functioning adenomas or hypopituitarism. All patients diagnosed with macroadenoma should have a visual field examination.
The treatment of choice for prolactinoma is dopamine agonist therapy. These drugs (bromocriptine, cabergoline, pergolide, quinagolide) are very effective at lowering PRL levels, resolving symptoms, and causing tumor shrinkage. Treatment results in a rapid reduction in PRL levels in 60% to 100% of cases. Following reduction of PRL levels, 60% to 100% of women resume ovulatory menses within 6 weeks, and galactorrhea disappears within 1 to 3 months after starting treatment. Reduction in tumor size is usually evident after 2 to 3 months of drug therapy, but it may occur within days after initiation of treatment. The extrasellar portion of the tumor appears to be particularly sensitive to drug therapy, which explains the improvement in symptoms such as visual impairment or ophthalmoplegia with drug therapy. In patients diagnosed with a microadenoma that is manifested only as a menstrual disturbance, observation should only be considered if the patient does not desire conception. These individuals should be offered oral contraceptives to control the bleeding pattern and to protect bone from estrogen deficiency. However, the long-term sequelae of persistently elevated PRL levels are unknown. If dopamine agonist treatment is initiated for a microadenoma, therapy may be continued long term. If the tumor responds, the dose may be tapered and stopped after menopause. Many individuals with a macroadenoma take dopamine agonist therapy indefinitely. All patients diagnosed with a prolactinoma should have follow-up imaging, determination of serum PRL levels, and visual field examinations. Alternatively, patients with a micro- or macroprolactinoma may be offered discontinuation of therapy. Dopamine agonist therapy may be tapered and ultimately discontinued if follow-up PRL levels are normal, and follow-up MRI scans show no evidence of tumor (or tumor reduction of at least 50%, with the tumor at a distance of more than 5 mm from the optic chiasm and no evidence for tumor invasion of adjacent structures), and if follow-up after withdrawal is assured.
An alternative to medical management of pituitary tumors is transsphenoidal surgery, after which resolution of symptoms may be immediate. However, the success and recurrence rates vary and are dependent on the size of the tumor and the depth of invasion. The larger and more invasive the tumor is, the less chance there is for complete resection and the greater the chance of recurrence. In general, the success rate of surgery for a microadenoma in some centers may be 70% or higher; surgical cure rates for macroadenomas are generally less than 40%. Overall, the recurrence rate following surgery is approximately 50%. Surgery is a good alternative for resistant tumors or for patients intolerant of medical treatment. Because non-PRL-secreting pituitary tumors often do not respond well to medical therapy, operation is the treatment of choice for these tumors as well. The risks of surgery include infection, diabetes insipidus, and panhypopituitarism. Complete pituitary testing should be performed prior to surgery.
Women with hyperprolactinemia who desire to conceive should be administered dopamine agonist therapy. Elevated PRL levels can lead to anovulation and hormonal disturbances that can result in difficulty in achieving pregnancy. Once pregnant, patients with microadenomas may have their medication discontinued. If a macroadenoma is present, thought should be given to continuing dopamine agonist therapy throughout pregnancy. Women with prolactinomas may breastfeed since there is no evidence that it stimulates tumor growth. If breastfeeding, dopamine agonist therapy should be discontinued since these agents will interfere with lactation.
Pituitary destruction—Amenorrhea may be the result of not only pituitary neoplasms but also pituitary destruction. Infiltrative disorders that involve the hypothalamus can affect the pituitary. Other disorders that cause pituitary insufficiency may be situational. A rare autoimmune disease, lymphocytic hypophysitis, can cause pituitary destruction during the puerperium and ultimately results in panhypopituitarism. Pituitary necrosis may also occur secondary to a hypotensive event. Typically, 80% to 90% of the pituitary must be damaged before pituitary failure ensues, and the robust blood supply to the pituitary makes this an uncommon event. If pituitary ischemia and necrosis are related to postpartum hemorrhage, it is known as Sheehan syndrome; otherwise, it is called Simmond disease. Because this type of injury typically affects the entire pituitary, most often more than one or all the pituitary hormones may be deficient. Observation has suggested that hormone loss follows a pattern, starting with the gonadotropins, followed by GH and PRL; however, this is highly variable.
The physiologic period in a woman's life when there is permanent cessation of menstruation and regression of ovarian function is known as menopause (see later). The cause of ovarian failure is thought to be depletion of ovarian follicles. The median age at menopause is 51.1 years. Premature menopause is defined as ovarian failure prior to age 40, which is reported to occur in 1% of the population. The etiology of premature ovarian failure (POF) may have a genetic basis. Several mutations that affect gonadal function have been identified and include defects in hormone receptors and steroid synthesis. Other potential causes include autoimmune ovarian destruction, iatrogenic ovarian injury, and idiopathic ovarian failure. The most severe form of POF presents with absent secondary sexual characteristics and is most often due to gonadal agenesis or dysgenesis (see Chapter 14). Less severe forms may result only in diminished reproductive capacity.
The other ovarian cause of amenorrhea is repetitive ovulation failure or anovulation. Other than menopause, it is the most common cause of amenorrhea. Chronic anovulation may be secondary to disorders of the hypothalamic-pituitary axis and has been previously discussed. Anovulation may also be due to systemic disorders. The causes of ovarian failure and anovulation due to peripheral disorders will be discussed later.
Ovarian failure is diagnosed based on the clinical picture of amenorrhea and the demonstration of elevated FSH (>40 IU/L). This may occur at any time from embryonic development onward. If it occurs prior to age 40, it is called POF. The presence or absence of secondary sexual characteristics defines whether ovarian activity was present in the past. The most common cause of hypergonadotropic amenorrhea, in the absence of secondary sexual characteristics, is abnormal gonadal development, which occurs in more than half of these individuals. When the gonad fails to develop, this is known as gonadal agenesis. The karyotype of these individuals is 46,XX, and the cause of failure is usually unknown. If streak gonads are present, this indicates at least partial gonadal development and is called gonadal dysgenesis. The karyotype of these individuals may be normal, but it is more likely that there are alterations in sex chromosomes (see Chapter 14).
Premature Ovarian Failure
Two intact X chromosomes are necessary for the maintenance of oocytes during embryogenesis, and the loss of or any alteration in the sex chromosome leads to accelerated follicular loss. This implies that two intact alleles are required for the normal function of some genes on the X chromosome. Turner syndrome is a classic example of complete absence of one X chromosome. It manifests as short stature, sexual infantilism, amenorrhea, and ovarian dysgenesis. This is a well-recognized condition that occurs in 1:2000 to 1:5000 females at birth. Turner syndrome is associated with a number of other phenotypic abnormalities, including a webbed neck, broad chest, low hairline, and cardiovascular and renal defects. It is interesting that fewer than half of patients with Turner syndrome have a single cell line with the karyotype 45,X. The majority of patients actually present with a mosaic karyotype such as 45,X/46,XX. These patients have varying degrees of the Turner syndrome phenotype and may display some secondary sexual development or may have a history of menstrual function. Some pregnancies have been reported.
The term mixed gonadal dysgenesis is used to describe chromosomal aneuploidies with a Y-chromosome component. The most common type is a mosaic pattern that has been associated with Turner syndrome (45,X/46,XY). These patients may have some functional testicular tissue and present with varying degrees of genital ambiguity. If enough testicular tissue is present to produce AMH, these patients may also present with abnormalities of the internal genitalia. Individuals who have both ovarian tissue and testicular tissue along with Wolffian and Müllerian structures internally are termed true hermaphrodites. Interestingly, most true hermaphrodites have 46,XX karyotypes (see Chapter 14) but others may have a 46,XY or 46,XX/46,XY chimerism.
Patients with gonadal dysgenesis may be phenotypically normal, and the abnormality may be manifest only as delayed pubertal development and amenorrhea. They probably have normal Müllerian structures and streak gonads. These individuals can display an array of karyotypes, including 46,XY (Swyer syndrome). Patients with a male karyotype but a female phenotype presumably underwent testicular failure prior to internal or external genitalia differentiation. If a dysgenetic gonad contains a Y chromosome or a fragment of the Y chromosome, there is a 10% to 30% risk for future gonadal malignancy, and the risk is higher if a mutation in the SRY gene is present. These tumors may be hormonally active. The onset of spontaneous pubertal development in girls with mixed gonadal dysgenesis may be a clinical marker of tumor development. Gonadal extirpation is extirpation at the time of diagnosis.
POF is defined as ovarian failure before age 40 but after puberty. Because complete absence of an X chromosome results in a dysgenetic gonad, candidate genes for POF are probably those that escape X inactivation. In mammals, X inactivation occurs in all cells in order to provide dosage compensation for X-linked genes between males and females (Lyon hypothesis). Further observation has illustrated that terminal deletions in Xp lead to the classic stigmata of Turner syndrome, whereas other deletions in Xp or Xq present with varying degrees of early reproductive failure. Most of the genes involved in folliculogenesis appear to be located on the long arm of the X chromosome. Several regions on the X chromosome, including POF 1 and POF 2, have been evaluated with knockout models in animals and have shown varying effects on ovarian development (Figure 13–10).
Candidate genes on the X chromosome for premature ovarian failure (POF).
(Reproduced, with permission, from Davison RM, Davis CJ, Conway GS. The X chromosome and ovarian failure. Clin Endocrinol. 1999;51:673.)
Limited observations have found that deletions occurring closer to the centromere of the X chromosome manifest a more severe phenotype that includes disruption of pubertal development. In contrast, deletions that occur in the distal regions tend to present with early reproductive aging and infertility. An example of a distal mutation on the long arm of the X chromosome is that in the FMR1 gene (fragile X gene). An association has been described between the FMR1 permutation state and POF. The prevalence of FMR1 gene permutations accounts for approximately 2% to 3% of patients who present with sporadic POF and may be as high as 15% in familial cases. This permutation is also associated with the occurrence of a late-onset neurological disorder in male carriers designated as fragile-X tremor ataxia syndrome (FXTAS). Although a number of genes on the X chromosome have demonstrated involvement in ovarian physiology, the majority of patients with POF have no identifiable mutations on the X chromosome.
Autosomal recessive genes that have been shown to contribute to POF are very rare. FSH receptor mutations have been identified in humans with POF. These individuals present with a phenotype that ranges from absent secondary sexual development to normal development and early reproductive failure. The prevalence of FSH receptor mutations varies but is most common in the Finnish population (1% carriers). This mutation has not been observed in North America. An inactivated LH receptor has been identified in patients with normal puberty and amenorrhea but is quite rare. Mutations in genes involved in steroidogenesis have also been associated with POF. These enzymes include CYP17α and aromatase. Patients with CYP17α mutations may have a 46,XX or 46,XY karyotype. They have a similar phenotype except that those with 46,XY have absent Müllerian structures because AMH is produced from their testes. Individuals with aromatase deficiency present with sexual ambiguity and clitoromegaly. Several other autosomal mutations have been discovered that may have a role in ovarian physiology. However, at this time, most cases of POF with normal pubertal development have not been associated with any specific mutation (see Chapter 14).
Autoimmune ovarian destruction is another potential cause of POF. This diagnosis is difficult to make unless it presents with one of the autoimmune polyglandular syndromes (see Chapter 2). The circumstantial evidence supporting the diagnosis is found in the high incidence of concomitant autoimmune disease—20% or more in patients with POF. The strongest association is with autoimmune thyroid disease. In addition, 10% to 20% of individuals with autoimmune adrenal disease experience POF. Conversely, 2% to 10% of patients with idiopathic POF develop adrenal insufficiency.
Autoimmune-associated POF is often diagnosed based on the presence of another autoimmune disease or detection of autoantibodies. Thyroid antibodies are most frequently screened. However, these antibodies are present in 15% to 20% of reproductive-aged women. Others have suggested measuring ovarian-specific autoantibodies. There is, however, significant inconsistency in the testing methodology, and this is not recommended. Although the diagnosis is imprecise, all patients suspected of having POF should be screened regularly for thyroid dysfunction. In addition, these patients should be evaluated for the autoimmune polyglandular syndrome (APS) and particularly tested for adrenal insufficiency, especially if there is no identifiable cause of POF. Testing for Addison disease can be performed using provocative ACTH testing or alternatively adrenal antibodies to identify patients at risk. There are commercially available tests for both antiadrenal antibodies and antibodies to 21-hydroxylase. The sensitivity of these tests for the diagnosis of adrenal insufficiency is 100%, the specificity 98%, and the PPV is 67%. If a patient has either of these antibodies they have a 25% to 50% risk of developing adrenal insufficiency within 5 to 10 years.
Iatrogenic causes of POF include radiation therapy, chemotherapy, and ovarian insults resulting from torsion or surgery. The risk of POF following radiation and chemotherapy is proportionate to the patient's age. If the radiation dose is higher than 800 Gy, all women experience ovarian failure. Displacing the ovaries out of the radiation field (ovarian transposition) has shown to be very effective at preserving ovarian function in patients receiving pelvic radiation. Therefore, patients wishing to preserve their fertility should be offered this procedure. Chemotherapy alone, particularly with alkylating agents, may induce temporary or permanent ovarian failure. Studies have shown that the likelihood of experiencing amenorrhea during treatment is 50% to 100%, and on average 40% of patients experience irreversible menopause. In general, younger individuals with chemotherapy-induced ovarian injury are more likely to recover. In women older than age 40, the chance of recovery is less 10%. Modifications of chemotherapy drugs, duration of treatment, and dosing are modifiers of permanent ovarian injury.
A rare cause of hypergonadotropic amenorrhea associated with numerous unstimulated ovarian follicles is the resistant ovary syndrome. These patients classically have no history of ovulatory dysfunction and present with secondary sexual characteristics and symptoms suggestive of estrogen deficiency. This diagnosis was established in an era when ovarian biopsy was used to determine the cause of menstrual disturbances. However, the definition of resistant ovary syndrome is not universally accepted. In fact, in the original series, cases were included in which patients had demonstrated ovulatory function in the past but later developed a clinical picture suggestive of ovarian resistance. This pattern is more typical of ovarian aging and follicular depletion.
The cause is not known. Histologic features of ovarian biopsy demonstrate that there is no plasma cell or lymphocytic infiltration, indicating that it is not caused by autoimmune destruction. The presence of numerous follicles indicates that POF is not due to follicular depletion. Several studies have looked at gonadotropins, FSH receptors, and antibodies that serve as blockers to the gonadotropin receptors, and the literature to date is inconclusive about the cause.
Diagnosis can be established with certainty only with ovarian biopsy. However, current recommendations for management of amenorrhea do not include surgery to make a diagnosis. The diagnosis is, therefore, one of exclusion. In the absence of autoimmune disease and of any history of ovulation, karyotyping should be performed to exclude chromosomal abnormalities. In patients with normal karyotypes, the diagnosis of POF and resistant ovary syndrome is difficult to make without biopsy. Improvements in ultrasound technology may make it possible to differentiate these entities by the measurement of ovarian volumes and antral follicle counts.
In more than half of patients with POF, no specific cause can be identified. The age defining POF is somewhat arbitrary. By definition, menopause is preceded by reproductive failure. It is thought that the time interval between menopause and the end of fertility may be approximately 10 years, and we know that approximately 10% of women reach menopause by 46 years of age and 1% by 40 years. Therefore, women who experience menopause at 45 years of age probably encounter a decline in reproductive potential or even reproductive failure at 35 years of age. This has obvious implications for women who are delaying childbearing. Several studies have described a significant association between the menopausal ages of mothers and daughters, twins, and sisters. A number of studies have identified new genes that are involved in ovarian physiology. It is hoped that these investigations will help in the treatment of subfertility and result in a reduction in infertility. At a minimum, they may allow better prospective individual prediction of reproductive risk.
Women diagnosed with POF that is not readily explained must have an evaluation that consists of a karyotype, testing for fragile-X (FMR) permutation, and evaluation for APS. If a patient has an FMR permutation, one should offer screening to her father. If there is no identifiable cause of POF, screening for adrenal insufficiency is obligatory. All patients with POF should be considered for combination HRT. The increased risks associated with breast cancer are likely no greater than the age-adjusted risk of women with intact ovaries. Patients with absent secondary sexual characteristics should initially be given estrogen-only therapy in low doses and titrated every 3 to 6 months (see Chapters 14 and 15 and see below for management of menopause).
Chronic anovulation may be defined as repetitive ovulation failure, which differs from ovarian failure in that viable oocytes remain in the ovary. Anovulation is the most common cause of amenorrhea during the reproductive years. There are several causes; those associated with hypothalamic and pituitary disorders have previously been mentioned and will not be considered in this section. Other conditions that cause anovulation include the peripheral endocrinopathies. These disorders result in a hormonal imbalance—mainly elevated androgens or estrogens—and lead to inappropriate feedback mechanisms and ovulatory failure. The peripheral endocrine disorders will be discussed later in greater detail.
Hyperandrogenism and Anovulation
Polycystic Ovarian Syndrome
Hyperandrogenic anovulation accounts for more than 30% of cases of amenorrhea and up to 75% of all cases of anovulation. Most often it is due to polycystic ovarian syndrome (PCOS). The reported prevalence of PCOS depends on the criteria used to define it. Although there is considerable controversy over the definition, most investigators, until recently, have focused on the 1990 National Institutes of Health-National Institute for Child and Human Development (NIH-NICHD) diagnostic criteria (Table 13–4). This definition puts forth ovarian hyperandrogenism and anovulation as the cardinal features of PCOS. The criteria include ovulatory dysfunction with evidence of hyperandrogenism, either clinically or by laboratory testing, in the absence of identifiable causes of hyperandrogenism. Using these criteria, the prevalence of unexplained hyperandrogenic chronic anovulation approximates 4% to 6%, and it is considered the most common endocrine disorder in women of reproductive age. However, in 2003 the diagnostic criteria were revised by the Rotterdam consensus conference (see Table 13–4). The Rotterdam criteria consider a broader spectrum of ovarian dysfunction than previous definitions. Some experts criticize the new definition, because it encompasses many more women, notably those with hyperandrogenism and ovulatory cycles, in addition to women with anovulation without androgen excess. This is contrast to the 1990 NIH-NICHD definition, which required both androgen excess and irregular cycles. Most recently, the Androgen Excess Society task force proposed the criteria should be defined by the presence of hyperandrogenism and ovarian dysfunction (see Table 13–4). This definition expands upon the 1990 NIH-NICHD criteria and requires that all patients with PCOS are hyperandrogenic at some level. Although the revised criteria can include women with ovulatory cycles, the remaining discussion on PCOS focuses on women who have anovulatory cycles.
Table 13–4 Revised Diagnostic Criteria for Polycystic Ovary Syndrome. ||Download (.pdf)
Table 13–4 Revised Diagnostic Criteria for Polycystic Ovary Syndrome.
|1990 criteria (both 1 and 2)|
|1. Chronic anovulation|
|2. Clinical and/or biochemical signs of hyperandrogenism and exclusion of other etiologies|
|Revised 2003 criteria (2 of 3)|
|1. Oligo- or anovulation|
|2. Clinical and/or biochemical signs of hyperandrogenism|
|3. Polycystic ovaries and exclusion of other etiologies (congenital adrenal hyperplasia, androgen-secreting tumors, Cushing syndrome)|
Approximately 50% of women diagnosed with PCOS are obese, and most have polycystic ovaries present on sonography (see later). Underlying these features are numerous biochemical abnormalities that have been associated with this syndrome, including elevated circulating total testosterone, free testosterone, DHEAS, and insulin as well as decreased SHBG and an elevated LH-FSH ratio. However, these abnormalities are not present in all PCOS patients. In fact, 40% of women who present with only hirsutism have elevated total testosterone levels, and 30% to 70% have elevated DHEAS levels. Similarly, the evaluation of increased LH pulsatility, in association with low-normal FSH levels (LH-FSH ratio), is not a reliable diagnostic test. Although elevated LH-FSH ratios are common findings in thin women, in obese patients with PCOS, the ratio is within the normal range about half of the time. The short half-life of LH (∼20 minutes) is likely another major contributor to the inaccuracy of LH testing. Hyperinsulinemia has recently been hypothesized to play a major role in the pathogenesis of PCOS (see later). The prevalence of insulin resistance may approximate 50% to 60%, compared with 10% to 25% observed in the general population. However, insulin resistance is difficult to measure in the clinical setting. Part of the difficulty is that there is no universally agreed on definition of insulin resistance, and the laboratory tests are not standardized. Furthermore, baseline insulin levels vary depending on population and body weight. For example, up to 60% of ovulatory obese patients have demonstrated some form of insulin resistance. Nonetheless, there is good evidence that a subset of normal-weight women and obese women with PCOS have a greater degree of insulin resistance and compensatory hyperinsulinemia compared with weight-matched controls.
There is increasing evidence for a strong genetic component in the etiology of PCOS. Several candidate genes have been investigated, including genes involved in steroidogenesis and carbohydrate metabolism, but none has been conclusively linked with the disease. PCOS is heterogeneous clinically, raising the possibility of different genetic causes and a variable environmental contribution to the syndrome.
Diagnosis of PCOS—The diagnosis of PCOS is typically based on clinical features (irregular menstrual cycles, acne, hirsutism), although additional information may be obtained with biochemical testing and sonographic examination. Known causes of hyperandrogenism and anovulation should be excluded in all patients (ie, androgen-secreting tumors, thyroid and adrenal gland dysfunction, hyperprolactinemia).
In most situations, the manifestations of PCOS emerge in the peripubertal years, occasionally with premature pubarche, and likely irregular cycles that persist through much of reproductive life. The diagnosis may be challenging, however, because adolescent girls commonly display irregular cycles during the 5 years following menarche. Ultrasonography may help to solidify the diagnosis of PCOS in children given that the likelihood of restoration of normal ovarian morphology is small as long as irregular menses persist.
Polycystic ovaries tend to be enlarged and have been defined in multiple ways: (1) the presence of 10 or more cystic follicles that are between 2 and 8 mm in diameter and arranged along the subcapsular edge of the ovary in a string of pearls fashion; and (2) 12 or more follicles 2 to 9 mm in either ovary and/or an ovarian volume of 10 cm3 or greater. The increased ovarian volume and displacement of follicles toward the periphery may be explained by the hyperplastic stroma. The stroma contains hilar cells and secondary interstitial cells (theca cells) that are capable of androgen synthesis.
Polycystic-appearing ovaries are very common in women with clinical features suggestive of hyperandrogenism, independent of menstrual disturbances. This finding suggests that ovarian hyperandrogenism can occur in ovulatory states and should be considered in the spectrum of PCOS, if polycystic ovaries are present. However, the finding of polycystic ovaries alone does not establish the diagnosis of PCOS. In fact, based on the Rotterdam definition, more than 30% of normal women have this ovarian morphologic feature. The prevalence is significantly impacted by age. Approximately 60% of women with regular cycles between the ages of 25 to 30 have morphologic characteristics consistent with polycystic ovaries compared to 7% of women between the ages of 41 to 45. Furthermore, the morphogenesis of polycystic ovaries is not unique to PCOS as it has been observed in other scenarios (ie, late-onset congenital adrenal hyperplasia, HIV, epilepsy).
Hyperandrogenism—The origin of sex steroids is from two sources: the gonads and the adrenal gland. The adrenal gland is responsible for most of the precursor hormones in the general circulation, which serve as reservoirs for the more potent androgens and estrogens. The ovary produces bioactive androgens and estrogens as well as precursor hormones. Figure 13–11 illustrates the relative contribution of each organ to the pool of circulating sex steroids. DHEAS is the most abundant steroid in the circulation, and serves as a precursor to the more potent androgens and estrogens. Over 98% of DHEAS is secreted from the adrenal gland. Approximately half of the circulating levels of androstenedione are equally derived from the adrenal gland and gonads and the remaining half from conversion of DHEAS (DHEA) in the periphery. DHEAS also serves as a substrate for the ovary to produce bioactive sex steroids. It is evident that the majority of bioactive androgen is derived from precursor steroids that undergo peripheral conversion. It is suggested over 60% of the testosterone is derived from circulating androstenedione, and the remaining is by direct secretion. The contribution of testosterone from the ovary and the adrenal gland is not completely known. The fact that the various methods to obtain secretion rates are difficult and that assays used to measure testosterone can be unreliable with resultant variability contributes to the uncertain interpretation.
A common clinical sign of hyperandrogenism is hirsutism. Hirsutism affects approximately 5% to 15% of the population. The prevalence depends on the population and the method used to establish the diagnosis. The most common scoring method used to diagnose hirsutism is based on the Ferriman and Gallwey system. It is based on evaluating 9 to 11 body areas and assigning a score of 0 to 4 based on the density of hair. This method is primarily used for research purposes. Clinically, it is often diagnosed by history and the presence of excess hair growth in centrally located regions, not commonly found in women. For example, those areas most affected are the face (sideburns, mustache, and beard), chest, linea alba, or the inner aspect of the thighs. The characteristics and distribution of body hair may be influenced by ethnic or racial factors. Alternatively, hyperandrogenism may present as acne or alopecia.
The diagnostic evaluation of hirsutism should include a thorough history and physical examination (Figure 13–12). The etiology can be divided between nonandrogenic and androgenic causes. Organic causes of hyperandrogenism must be excluded. Nonandrogenic causes include chronic skin irritation, anabolic medications, and rarely acromegaly. The most common androgenic cause is PCOS which affects at least 70% of hirsute women. Less common is late-onset congenital adrenal hyperplasia, which affects approximately 5% of patients. PCOS is a diagnosis of exclusion—measurements of serum androgen levels should play only a limited role in the evaluation. Most patients who have hyperandrogenemia present with obvious clinical manifestations, and the presence of normal androgen levels in a patient with hirsutism or acne does not exclude the diagnosis of PCOS. However, there are subsets of amenorrheic patients (ie, Asians) who are hyperandrogenemic without clinical manifestations, most likely as a result of their relative insensitivity to circulating androgens. It is in these patients that assessing androgen levels may be of value in determining the cause of amenorrhea. More commonly, androgens such as DHEAS and testosterone are measured to exclude other causes of hyperandrogenic anovulation such as nonclassic adrenal hyperplasia and androgen-secreting tumors (see later).
Another cause of hirsutism, made only after excluding other diagnoses, is characterized by regular menstrual cycles, normal androgen levels, and the absence of polycystic-appearing ovaries; it is termed idiopathic hirsutism. This common disorder occurs more frequently in certain ethnic populations, particularly those of Mediterranean descent. It is thought that the etiology of hirsutism in these patients is related to higher 5α-reductase activity (increased sensitivity) within the pilosebaceous unit.
Mechanism of anovulation—The mechanism of anovulation in PCOS remains unclear. It is evident that the population of antral follicles is increased and that follicular development is arrested. It is also known that the development of preantral follicles is not primarily under hormonal control. Evidence supports the components of the intraovarian network as regulators of antral follicle development. It is known that many of the accumulated follicles in PCOS remain steroidogenically competent and are capable of producing estrogen and progesterone. In fact, it is interesting that women with PCOS produce both androgens and estrogen (estrone) in excess.
One of the most frequently described characteristics of PCOS is the functional derangement of LH secretion. Numerous studies have shown that frequency, amplitude, and mean levels of LH are increased. The aberration in LH secretion may be a result of heightened pituitary responsiveness and increased hypothalamic GnRH activity. Under normal conditions, follicles respond to LH after they reach approximately 10 mm in diameter. However, polycystic ovarian follicles acquire responsiveness to LH at a much smaller diameter, which may lead to inappropriate terminal differentiation of granulosa cells and result in disorganized follicular development. It has been suggested that the theca cells increase their expression of steroidogenic enzymes following stimulation by LH, whereas the granulosa display resistance to FSH (see later). The elevated LH levels and relative hyperinsulinemia that exist in some PCOS patients may synergistically potentiate disordered folliculogenesis. Although hyperandrogenism is part of the diagnostic criteria for PCOS, its direct impact on folliculogenesis is not clear. It is conceivable that androgens contribute to the effects of LH and insulin on follicular maturation. It is also possible that the excess estrogens may result in a negative-feedback loop to inhibit FSH release and prevent further follicular development.
Most experts are of the opinion that excess androgen production is a fundamental abnormality in women with PCOS. Androgens within the ovary are produced mainly by the thecal interstitial cells that surround the follicle and to a lesser extent by the secondary interstitial cells located in the stroma (see earlier). The CYP17α complex is thought to be the key enzyme in biosynthesis of ovarian androgens. Under normal conditions, a large proportion of the androgens produced by the theca cells diffuses into the granulosa cell layer of the follicle where they are rapidly converted to estrogen as shown in Figures 13–7 and 13–11 (two-cell theory). The intrinsic control of androgen production in the ovary is modulated by intraovarian factors and hormones (see section on ovarian steroidogenesis at the beginning of this chapter). It is the dysregulation of hormone production that is most likely responsible for PCOS.
Several studies have shown that women with PCOS have an exaggerated ovarian androgen response to various stimuli. To illustrate, hyperstimulation of 17-hydroxyprogesterone levels was noted when women diagnosed with PCOS were given a GnRH agonist or hCG, suggesting increased CYP17α activity. This study is supported by in vitro studies in which measurement of steroids in cultured human theca cells from polycystic ovaries revealed concentrations of androstenedione, 17α-hydroxyprogesterone, and progesterone that were respectively 20-fold, 10-fold, and 5-fold higher than levels in control cells. Additional studies have found increased expression of the genes encoding CYP17α hydroxylase, P450scc, the LH receptor, and StAR. These findings reflect a global enhancement of steroidogenesis. This situation is compounded by the hypertrophy of theca cells that is present in women with PCOS.
Several studies have evaluated intraovarian modulators as participants in the pathogenesis of PCOS (Figure 13–13). IGF-binding proteins (IGFBPs), especially IGFBP-2 and IGFBP-4, are found to be increased in the follicular fluid of polycystic ovaries. They may act locally to decrease free IGF-2 and, thus, decrease the effects of FSH on the oocyte and granulosa cells. Alternatively, the granulosa cells may downregulate their insulin receptors as a result of hyperinsulinemia. This would deprive the granulosa cells of their co-gonadotropin, IGF-2, and account for the relative FSH insensitivity. Inhibin is also a likely candidate, because a large proportion of women with PCOS have relative FSH suppression (Figure 13–13). However, studies have not shown consistent results, suggesting that if inhibin is involved, the effect is minimal. Follistatin, the activin-binding protein, was in the past thought to play an important role in the development of PCOS. It was initially implicated because activin functions to inhibit androgen production and enhance FSH expression. However, current studies do not demonstrate a significant association between abnormalities in follistatin and PCOS.
The adrenal gland may be significantly involved in the pathogenesis of some cases of PCOS. The connection seems plausible because adrenal androgens can be converted to more potent androgens in the ovary. Furthermore, a significant portion of women with congenital adrenal hyperplasia have polycystic ovaries (see later). Several studies have shown that DHEAS is elevated in 25% to 60% of patients with PCOS. However, ACTH levels are normal in PCOS women. Interestingly, it has been reported that there is an increased response of androstenedione and 17α-hydroxyprogesterone to exogenous ACTH. These findings suggest an underlying abnormality in the CYP17α expressed in the adrenal gland as well as in the ovary. However, minimal data support CYP17α dysfunction in the adrenal gland. It has also been shown that ovarian steroids can stimulate adrenal androgen production; however, additional findings suggest that the ovary is not the primary cause of adrenal hyperresponsiveness. The critical role played by adrenal androgens during the pubertal transition has not been fully investigated as a potential contributor to the development of PCOS.
Hyperinsulinemia and PCOS—A relationship between insulin and hyperandrogenism has been postulated based on several observations. This association may be demonstrated in the peripubertal years with premature pubarche, which is more commonly associated with insulin resistance than with congenital adrenal hyperplasia or androgen excess. Various case reports have shown that acanthosis nigricans—hyperpigmentation of skin in the intertriginous areas—is associated with severe insulin resistance. A number of these patients also presented with hyperandrogenism and anovulation. The relationship was substantiated when it was observed that the degree of hyperinsulinemia was correlated with the degree of hyperandrogenism. Further studies revealed that hyperinsulinemia is frequently identified in women with PCOS.
It has been shown that the cause of hyperinsulinemia is insulin resistance and that the defect lies in the postreceptor signaling pathway. The frequency and degree of hyperinsulinemia in women with PCOS is amplified in the presence of obesity. Although many women with PCOS exhibit insulin resistance, some do not. However, insulin resistance is also observed in some thin PCOS patients.
Insulin may cause hyperandrogenism in several different ways, although the exact mechanism has not been well defined. It has been suggested that insulin has a stimulatory effect on CYP17α. There is evidence from in vitro models that insulin may act directly on the ovary (see Figure 13–13). It has been shown that the ovary possesses insulin receptors and IGF receptors. In addition, several studies have reported that insulin stimulates ovarian estrogen, androgen, and progesterone secretion and that its effects are greatly enhanced by the addition of gonadotropins. Administration of an insulin-sensitizing agent (eg, metformin or a thiazolidinedione) to obese women with PCOS leads to a substantial reduction in 17α-hydroxyprogesterone levels, reflecting decreased CYP17α activity. However, clinical studies, in which insulin infusions were administered to normal women, failed to demonstrate increased testosterone production, and there were no changes in androgen levels when normal women were given insulin-sensitizing agents. These observations suggest that insulin has more of a modifying rather than a predisposing effect on androgen production.
The relationship between insulin and adrenal androgen production is less clear. Some studies have shown that insulin increases secretion of 17α-hydroxyprogesterone and DHEAS in response to ACTH. Other studies have shown that DHEAS decreases after acute insulin infusions are administered to normal men and women. Furthermore, when insulin–sensitizing agents were administered to women with PCOS, a decrease in DHEAS was observed. Although there is less evidence to support the association of insulin and adrenal androgen production, if there is an insulin effect, it is as a modulator of adrenal secretory activity.
Insulin may indirectly affect androgen levels. Several studies have reported that insulin directly inhibits SHBG production. There is an inverse correlation between insulin levels and SHBG, so that decreasing insulin levels would decrease the circulating bioavailable androgen level (via increases in SHBG). It has also been shown that insulin decreases IGFBP-1. This would increase free IGF-1, which could modulate ovarian androgen production in a fashion similar to insulin (see Figure 13–13). Although these indirect mechanisms may play a role, the literature suggests that insulin acts directly to augment androgen production (see Figure 13–13). However, it appears that a dysregulation in steroidogenesis must also exist in order for insulin to cause hyperandrogenism.
Increasing attention is being directed to the possibility that PCOS begins before adolescence. In fact, the initial insult may begin in utero, where there is ample exposure to androgens derived from the fetal adrenal gland and ovary. This hormonal environment may reprogram the ovary and alter steroidogenesis in a manner that predisposes to PCOS. The phenotypic expression of PCOS would then be determined by environmental factors such as diet and exercise.
The various biochemical abnormalities associated with PCOS have led to studies investigating metabolic sequelae of this syndrome. Long-term health problems such as the development of cardiovascular disease and diabetes have been linked to PCOS. In fact, several studies have suggested that the metabolic syndrome is significantly more prevalent in women with PCOS. However, studies looking at whether women with PCOS actually experience increased cardiovascular events are limited. Several observational studies have demonstrated that women with PCOS have alterations in their lipid profiles, including increased triglycerides and LDL and decreased high-density lipoprotein (HDL) compared with weight-matched controls. Furthermore, the degree of dyslipidemia has been correlated with the magnitude of insulin resistance. The fact that insulin resistance occurs with greater frequency in women with PCOS suggests that they are at higher risk for the development of diabetes mellitus. It is known that up to 30% to 40% of women with PCOS have impaired glucose tolerance, although it is most often seen in patients who are obese. Limited retrospective studies have suggested that women with PCOS have an increased frequency of developing type 2 diabetes. In summary, as demonstrated by surrogate markers, the available data show that these patients have increased risk factors for cardiovascular events and diabetes. Long-term prospective studies will be required to determine if these patients are actually at risk for increased mortality and morbidity.
The fact that hyperinsulinemia underlies many of the potential adverse sequelae raises a question about whether insulin resistance should be assessed in all patients diagnosed with PCOS. There are several methods of measuring insulin resistance (Table 13–5). However, the potential impact of hyperinsulinemia is unknown. Furthermore, there is no universal laboratory criterion or standardization for establishing the diagnosis of insulin resistance, which raises doubts about whether any potential adverse sequelae can be prevented if insulin resistance is identified. Further data are necessary, prior to widespread use of medications such as metformin or a thiazolidinedione to reduce insulin resistance in patients with PCOS to determine (1) if these patients are indeed at risk for cardiovascular events, (2) if hyperinsulinemia is an independent risk factor, and (3) whether long-term treatment with any insulin-sensitizing agent will decrease the potential sequelae.
Additional risks associated with PCOS—Another concern is that women who are anovulatory do not produce a significant amount of progesterone. This leads to a situation in which the uterine lining is stimulated by unopposed estrogens, which is a significant risk factor for development of endometrial cancer. In fact, an association has been found between endometrial cancer and PCOS. There is also evidence to suggest an association of PCOS with both breast cancer and ovarian cancer, but PCOS has not been conclusively shown to be an independent risk factor for either disease.
Management of PCOS—The management of PCOS should be dictated by the patients' risk factors for cardiovascular disease, diabetes, malignant sequelae, and symptomatology (see also "Infertility"). All patients should have a lipid profile measured. In addition, oral glucose tolerance tests should be performed—at least in obese women. In patients who have a long history of irregular cycles (>1 year), an endometrial biopsy should be considered.
The origin of circulating sex steriods: the relative contribution of each organ to the pool of circulating sex steroids.
An algorithm for women with hirsutism. All women should be screened for known causes of hyperandrogenism. Hyperandrogenism suggests either PCOS or late-onset congenital adrenal hyperplasia (CAH). If hirsutism is rapid in onset or rapidly progressive in nature or the patient is frankly virilized, an evaluation for an androgen-secreting tumor is required. Management of hirsutism in late-onset CAH or PCOS is similar. *Late-onset CAH can present with regular cycles. Depending on the specific criteria (see Table 13–4), PCOS can present with regular cycles. Serum hormones are not required for the diagnosis of idiopathic hirsutism or PCOS.
Regulation of sex steroid production in the theca and granulosa cells of the ovary. Several autocrine and paracrine factors such as insulin and insulin-like growth factors (IGFs) work synergistically with luteinizing hormone (LH) to stimulate androgen production through increased 17α-hydroxylase activity.
(Redrawn with permission from Ehrmann DA, Barnes RB, Rosenfield RL. Polycystic ovary syndrome as a form of functional ovarian hyperandrogenism due to dysregulation of androgen secretion. Endocr Rev. 1995;16:322.)
Table 13–5 Measurements of Insulin Resistance. ||Download (.pdf)
Table 13–5 Measurements of Insulin Resistance.
|Frequently sampled intravenous glucose tolerance test (FSIGT)a|
|Euglycemic insulin clampa|
|Insulin tolerance test (ITT) |
|Fasting insulin levels|
|Oral glucose tolerance test (OGTT) + insulin levels and calculating area under the curve|
|Glucose-insulin ratio |
|Derivatives of glucose-insulin ratiob|
|Insulin resistance index = glucose × insulin/25|
|Homeostasis model assessment of insulin resistance HOMA-IR = 22.5 × 18/glucose × or log (HOMA-IR)|
|Quantitative insulin sensitivity check index (QUICKD = 1/log (glucose) + log (insulin)|
Cardiovascular risk factors, weight loss, and progression to diabetes can be improved with diet and exercise. In the subset of patients with glucose intolerance, administration of insulin-sensitizing agents should be considered. Metformin, a biguanide, is the most widely studied. Several studies have shown modest transitory weight loss, and possibly a decreased incidence of diabetes in patients with PCOS due to its use. Metformin is not effective for treatment of hyperandrogenic symptomatology. The use of metformin is not recommended unless there is evidence of metabolic derangements that are not fixed with diet and exercise alone.
Irregular bleeding can be improved by the administration of oral contraceptives which induce scheduled withdrawal bleeding. However, another benefit of oral contraceptives has been the significant reduction in endometrial cancer in the general population. It is rational to expect that a similar benefit would accrue to women with PCOS, because the ovarian stimulation is minimized and the progestin would counteract the estrogenic environment.
Acne and hirsutism in women diagnosed with PCOS or of idiopathic origin can be treated with oral contraceptives. The mechanism is not completely known, but oral contraceptives decrease the amount of bioavailable androgens by increased SHBG production and by ovarian suppression. It has also been shown that progestins can inhibit 5α-reductase activity, which further decreases the production of dihydrotestosterone, the major androgen that stimulates hair growth. The maximal effect is evident after 6 months of treatment (the hair cycle length is estimated to be 4 months). If hirsutism is severe or if oral contraceptives alone are not effective, the addition of spironolactone may be beneficial. Spironolactone is an antimineralocorticoid agent that inhibits androgen biosynthesis in the adrenal gland and ovary, inhibits 5α-reductase, and is a competitive inhibitor of the androgen receptor. Side-effects are minimal and include diuresis in the first few days; dyspepsia; nausea; skin hypersensitivity; breast tenderness; and abnormal bleeding, which can be alleviated by increasing the dose to desired amount over 3 weeks. Spironolactone is used concomitantly with oral contraceptives. Because oral contraceptives and spironolactone act by different mechanisms, the combined effect is synergistic.
If oral contraceptives are not desired or have not improved the excess hair growth, there are several other methods that can be used in conjuction with oral contraceptives or administered alone. The only topical agent approved for hirustism by the FDA is Vaniqa (Eflornithine) cream, which is an inhibitor of L-ornithine decarboxylase. It has been shown to be effective for control of facial hair. Long-acting GnRH agonists and 5α-reductase inhibitors (finasteride) have been used for refractory cases with some success. Shaving, bleaching, chemical depilation, plucking, or waxing are temporary measures to control unwanted hair growth. However, several of these methods can cause skin irritation and result in progressive hair growth. Permanent techniques such as electrolysis or laser depilation have shown promising results.
Congenital Adrenal Hyperplasia
Congenital adrenal hyperplasia is another disorder that may cause hyperandrogenism. It presents in a wide range of clinical forms, ranging from severe—which may be classified as classic, salt-wasting, or simple virilizing—to milder forms known as acquired, adult–onset, nonclassic, or late-onset congenital hyperplasia. The clinical manifestations reflect the severity of the enzymatic defect. Severe or classic forms are discussed in Chapter 14 and will not be considered here. This section will discuss the nonclassic forms, which affect 1% to 10% of the population, depending on the ethnicity of the patient. The clinical features are similar to those of patients diagnosed with PCOS and include menstrual irregularities, hyperandrogenism, and infertility, and up to 50% have polycystic ovaries.
The adrenal gland consists of a cortex and a medulla. The cortex is divided into three functional zones based on location and the principal hormone secreted (see Chapter 9). The zona glomerulosa is the outermost zone and lies adjacent to the adrenal capsule. It is primarily responsible for aldosterone production. The zona fasciculata lies immediately below the glomerulosa. It principally secretes glucocorticoids, although it is capable of producing androgens. The zona reticularis is located beneath the zona fasciculata and overlies the adrenal medulla. It is this zone that principally secretes androgens. Both the zona fasciculata and the zona reticularis are regulated by ACTH. It is the secretory activity of these two zones that results in nonclassic adrenal hyperplasia.
Congenital adrenal hyperplasia is an autosomal recessive disorder that is caused by mutations of genes involved with adrenal steroidogenesis. The mutations mostly occur in the 21–hydroxylase gene (P450c21B) and rarely in the 3β-hydroxysteroid dehydrogenase gene or 11β-hydroxylase genes (P450c11B and P450c11AS). In classic forms, the enzymatic defects are severe, affect both alleles, and result in cortisol deficiency diagnosed at birth. Owing to the cortisol deficiency, there is ACTH excess and hyperstimulation of the adrenal gland. The adrenal precursors produced proximal to the enzymatic defect accumulate and promote the synthesis of DHEA and androstenedione through the androgen synthetic pathway. These are converted in the periphery to more potent androgens, resulting in symptoms of hyperandrogenism. Approximately 50% of patients diagnosed with nonclassic congenital adrenal hyperplasia are compound heterozygotes with one of the mutant alleles coding for a severe defect. Typically, patients who are carriers have a normal phenotype. Therefore, most patients with nonclassic adrenal hyperplasia do not demonstrate deficient cortisol production or excess ACTH. It is suggested that most of the androgen excess in nonclassic adrenal hyperplasia arises as a consequence of subtle alterations in enzyme kinetics. Furthermore, some studies report a generalized adrenocortical hyperactivity rather than deficient enzyme activity.
The diagnosis can be established by measuring early morning basal 17-hydroxyprogesterone levels. Levels greater than 800 ng/dL (24.24 pmol/L) are diagnostic of 21-hydroxylase deficiency. However, the elevation in 17-hydroxyprogesterone is often not impressive and does not differ from that observed in PCOS. If the basal 17-hydroxyprogesterone levels are greater than 200 ng/dL (6.06 pmol/L) and less than 800 ng/dL (24.2 pmol/L), a provocative test with ACTH (250 μg intravenously) should be performed. If 17-hydroxyprogesterone levels are greater than 1000 ng/dL (30.30 pmol/L) 1 hour after administration of ACTH, the diagnosis of 21-hydroxylase deficiency can be made. Elimination of a false elevation in 17-hydroxyprogesterone due to ovulation must be excluded by simultaneous measurement of progesterone. A 17-hydroxyprogesterone screening level greater than 200 ng/dL has 100% sensitivity, but only 7% positive predictive value for the diagnosis of nonclassic adrenal hyperplasia. The other rare enzymatic defects that result in nonclassic adrenal hyperplasia can be assessed with measurements of steroid products proximal to the blockade following provocative testing.
Treatment of nonclassic adrenal hyperplasia is similar to that of PCOS, raising a question about whether an etiologic diagnosis is necessary. It is certainly an expense, leading to no change in management except with regard to infertility treatment and management of subsequent pregnancies (see later). Establishing a specific diagnosis may be advisable to facilitate genetic counseling in the woman with adrenal hyperplasia who intends future childbearing and to prepare for in utero treatment should an affected fetus be identified by amniocentesis. Although some experts suggest dexamethasone treatment for symptoms of hyperandrogenism, studies show inconsistent results, and there is concern about the consequent adrenal suppression.
Chronic glucocorticoid excess, whatever its cause, leads to the constellation of symptoms and physical features known as Cushing syndrome. The most common cause is iatrogenic as a result of glucocorticoid treatment. However, an ACTH-secreting microadenoma (Cushing disease) accounts for more than 70% of cases of endogenous hypercortisolism. Less common causes include primary adrenal disease (tumors or hyperplasia) and ectopic (not hypothalamic-pituitary) ACTH-producing or CRH-producing tumors. Patients with Cushing syndrome have a range of clinical manifestations that vary with age at onset and etiology. This section will discuss the adult clinical presentation briefly. For a more detailed discussion, see Chapter 9.
Cushing syndrome (noniatrogenic) is rare and occurs in approximately 2.6 patients per million individuals in the population. It is responsible for less than 1% of those individuals who present with hirsutism. Although Cushing syndrome is uncommon, it presents in a manner similar to PCOS and congenital adrenal hyperplasia and needs to be considered in the differential diagnosis of hyperandrogenism and anovulation. Patients with ACTH excess typically have additional clinical features suggestive of glucocorticoid or mineralocorticoid hypersecretion. The most common features include obesity with increased centripetal fat, moon facies, muscle weakness, and striae. Other manifestations may include diabetes, hypertension, and osteoporosis. Women with primary tumors tend to have a rapid onset of symptoms and often manifest with severe hyperandrogenism (frank virilization) that includes male pattern baldness, deepening voice, clitoromegaly, and defeminization.
Hirsutism or acne is present in about 60% to 70% of women with Cushing syndrome. However, the exact mechanism of hyperandrogenic effects is not completely known. It is evident that excess ACTH causes hyperstimulation of the zona fasciculata and zona reticularis and results in hypersecretion of cortisol and androgens. It is also known that adrenal tumors may selectively overproduce androgens.
Menstrual irregularities occur in more than 80% of patients with Cushing syndrome. The exact cause of anovulation is unclear. It has already been observed that hyperandrogenemia may have a significant impact on ovulation. However, several studies have shown that glucocorticoids can also suppress the hypothalamic-pituitary axis. Thus, the elevated glucocorticoids may be an additional factor in the pathophysiology of anovulation associated with this syndrome (see Chapters 4 and 9 for the diagnosis and treatment of Cushing syndrome).
If there is a rapid onset of androgenic symptoms, an androgen-secreting adrenal tumor should be suspected. Elevated testosterone (>200 ng/dL; 6.9 nmol/L) and DHEAS (>700 ng/mL; 19 μmol/L) levels should raise the suspicion of a tumor. However, more than 50% of adrenal androgen-secreting tumors have testosterone levels below 200 ng/dL (6.9 nmol/L). Furthermore, the majority of patients with high testosterone levels do not have tumors. Testosterone levels above 250 ng/dL are only 10% predictive for androgen-secreting tumor. Measurement of DHEAS levels in these patients yields similar inconsistencies. This suggests that laboratory tests have limited value in screening for androgen-secreting tumors, and a clinical history and physical examination are better predictors. The presence of systemic symptoms such as weight loss, anorexia, bloating, and back pain favor an androgen-secreting tumor. If suspicion is high, an abdominal computed tomography (CT) scan confirms the diagnosis. The treatment involves surgical resection, mitotane (adrenolytic), and steroid synthesis inhibitors.
Androgen-secreting tumors can also originate from the ovary. The incidence approximates 1:500 to 1:1000 hyperandrogenic patients. Testosterone levels >200 ng/dL (6.9 nmol/L) arouse suspicion, although in 20% of patients with ovarian androgen–producing tumors, testosterone levels are below this value. Again, the best screening procedures are the clinical history and physical examination. In the absence of Cushingoid features, adrenal and ovarian tumors present similarly. Ovarian tumors often have unilateral ovarian enlargement that can be palpated on pelvic examination. Ultrasonography often confirms the diagnosis. In selected cases, selective venous sampling may be performed if CT or sonography cannot identify the source of androgen production.
Obesity is the most prevalent chronic disease in the United States. Adverse health events are dramatically increased in obese subjects (see Chapter 20). These include cardiovascular disease, diabetes, joint disease, respiratory dysfunction, and colon, endometrial, and ovarian cancers. Obesity is often also associated with menstrual irregularities, and the relationship between the two is strengthened the earlier the onset of the obesity. Alterations in sex steroid metabolism are clearly evident in obese females, and the consequences are attenuated release of gonadotropins, which manifests as anovulation. Obese patients have increased production rates and MCRs of androgens; therefore, their serum levels of androgens are largely within the normal range. However, free testosterone levels tend to be in the high-normal range due to the decreased SHBG levels. A portion of androgen metabolism occurs in body fat. The excess adipose tissue aromatizes these androgens and increases the amount of circulating estrone causing a state of functional hyperestrogenism. In fact, studies have shown that the rate of peripheral conversion of androstenedione to estrone is correlated with body weight. Other studies have illustrated that the conversion of estrone to estradiol in adipose tissue is higher in visceral fat than subcutaneous fat. The increased visceral fat associated with obesity is also associated with hyperinsulinemia, which may have independent effects on ovarian function.
The management of obesity includes diet and exercise (see Chapter 20). In patients who have irregular uterine bleeding, an endometrial biopsy must be performed to rule out endometrial cancer. Oral contraceptives are effective at treating the irregular bleeding and reduce the incidence of endometrial cancer. Weight loss reestablishes normal menstrual cycles in the majority of these women.
Anovulation Unrelated to Excess Sex Steroid Production
Adrenocortical insufficiency may be categorized as primary or secondary. Primary adrenal insufficiency (Addison disease) is caused by destruction of adrenal cortical tissue. Secondary adrenal insufficiency is due to defects in the hypothalamic-pituitary axis resulting in a deficiency in ACTH. Both types lead to cortisol deficiency, which is life threatening.
The main cause of primary adrenal failure is autoimmune destruction of the adrenal cortex. In fact, in the industrialized world, an autoimmune pathogenesis accounts for more than 60% of cases of primary adrenocortical deficiency. The symptoms are typically those of chronic insufficiency and include weakness, fatigue, menstrual disturbances, and gastrointestinal symptoms such as nausea, abdominal pain, and diarrhea. Additional signs may include weight loss, hypotension, and hyperpigmentation of the skin and mucous membranes. These symptoms may appear insidiously with a mean duration of approximately 3 years. Symptoms usually wax and wane until there is complete decompensation.
Autoimmune adrenal failure may occur as an isolated event, but estimates link more than 70% of cases to an autoimmune polyglandular syndrome (APS) of two subtypes (type 1 and 2). APS-1 usually presents in childhood with hypoparathyroidism, chronic mucocutaneous candidiasis, adrenal insufficiency, and other autoimmune disorders such as celiac disease and ovarian failure. It is caused by mutations in the autoimmune regulator gene (AIRE) and has an autosomal recessive inheritance. APS-2 (Schmidt syndrome) has its onset in adulthood (usually the third decade). Common manifestations include type 1 diabetes mellitus, myasthenia gravis, Hashimoto thyroiditis, ovarian failure, and adrenal insufficiency. Susceptibility to this disorder seems to be inherited as a dominant trait in linkage dysequilibrium with the HLA-B region of chromosome 6. Autoimmune adrenal insufficiency is associated with autoantibodies that are directed toward enzymes involved with steroidogenesis. Several studies suggest that the antigens in this disorder include 17α-hydroxylase, 21-hydroxylase, and P450scc. Detection of antibodies directed at these enzymes is helpful in making the diagnosis of autoimmune adrenal failure (see Chapters 2 and 9).
Worldwide, infection—especially tuberculosis—is the most common cause of primary adrenal failure. The adrenal cortex and medulla are involved and may be completely replaced by caseating granulomas. This phenomenon is always associated with other evidence of a tuberculosis infection. Fungal, viral, and bacterial pathogens are less common causes.
The most common cause of secondary adrenal insufficiency is adrenal suppression after exogenous glucocorticoid administration. Adrenal insufficiency is also seen following treatment of Cushing disease and with a variety of hypothalamic-pituitary lesions that result in hypopituitarism. These patients more commonly present with symptoms suggestive of acute adrenal insufficiency. The clinical features include abdominal pain, hypotension, fever, severe volume depletion, and possibly profound shock.
Menstrual disturbances are a frequent presentation in patients with adrenal insufficiency. Autoimmune adrenal insufficiency is often accompanied by gonadal failure. In the two forms of APS, POF develops in APS-1 in 50% of patients and in APS-2 in 10% of patients. It has been shown that antibodies—particularly to CYP17α and P450scc—are associated with POF. Other causes of adrenal failure are associated with menstrual disorders more than 25% of the time. The cause of anovulation is not known with certainty, but chronic illness itself is probably responsible.
Diagnosis and treatment are discussed in Chapter 9. The screening process includes blood chemistries and basal cortisol levels. The diagnosis is confirmed with provocative tests using exogenous ACTH.
The prevalence of overt thyroid dysfunction is 1% to 2% in women of reproductive age. Thyroid disorders can develop secondary to an insult in the hypothalamus, pituitary, or thyroid, the latter being most common. In order to understand the pathophysiology of these disorders, it is important to be familiar with the normal physiologic regulation (see Chapter 7). This section will briefly review the common causes of hyperthyroidism and hypothyroidism, their manifestations after the onset of puberty, and their impact on reproductive function.
Hyperthyroidism—Hyperthyroidism is the clinical syndrome associated with excessive thyroid hormone activity. The clinical presentation of hyperthyroidism (thyrotoxicosis) depends on the age at onset and the degree of thyrotoxicosis. The clinical manifestations can involve most organ systems, and the presentation ranges from asymptomatic to thyroid storm. Typical features include nervousness, malaise, palpitations, heat intolerance, weight loss, and inability to concentrate. Additional features may involve the eyes and include lid lag, proptosis, and ophthalmoplegia. The reproductive abnormalities include menstrual abnormalities, infertility, and spontaneous abortions.
The most common cause of hyperthyroidism is an autoimmune process that affects thyroid hormone production (Graves disease). Antibodies bind to the TSH receptor and stimulate the thyroid gland to secrete increased amounts of thyroid hormone. Less common causes include subacute thyroiditis, toxic multinodular goiter, and struma ovarii.
Excess thyroid hormone has an impact on sex steroids. Thyroid hormones stimulate hepatic production of SHBG. As a result, total serum estradiol, estrone, testosterone, and dihydrotestosterone are increased, yet free levels of these hormones remain within the normal range. The metabolic clearance pathways appear to be altered, which can be explained in part by the increased binding. The conversion rates of androstenedione to estrogen and testosterone are increased. The significance of the alterations in metabolism has not been determined.
Menstrual irregularities frequently occur in hyperthyroid states. The exact mechanism is unclear. Altered levels of TRH and TSH do not appear to have a significant impact on the HPO axis. The LH surge may be impaired, although studies have shown that hyperthyroid patients have normal FSH and LH responses to exogenous GnRH. It is possible that the weight loss and psychologic disturbances associated with this disease may contribute to the menstrual abnormalities. It is interesting to note that endometrial biopsies of many amenorrheic patients with hyperthyroidism have demonstrated a secretory endometrium, indicating that many of these women remain ovulatory. Menstrual abnormalities return to normal with treatment.
Hypothyroidism—A more common disorder than hyperthyroidism in women of reproductive age, hypothyroidism results from inadequate thyroid hormone production. Manifestations can involve almost any organ system, and the presentation can range from asymptomatic to myxedema coma. Common symptoms include lethargy, memory defects, cold intolerance, dry skin, hair loss or occasionally excess hair growth, deepening of the voice, nausea, and constipation. Physical findings include somnolence; bradycardia; mild hypertension; dry skin; periorbital puffiness; nonpitting edema of the hands, face, and ankles; and decreased tendon reflexes. Reproductive abnormalities include menstrual disorders, infertility, and spontaneous abortions.
The most common cause is autoimmune destruction of the thyroid (Hashimoto thyroiditis). It is mediated by humoral and cell-mediated processes. The antibodies are directed toward thyroglobulin (anti-Tg) and thyroid peroxidase (anti-TPO antibody). Histologic specimens show lymphocytic infiltration. Other causes include ablative therapy of the thyroid gland (postsurgery or radioactive iodine), end-stage Graves disease, and transient thyroiditis (viral, drug-induced, or postpartum).
Inadequate thyroid hormone levels influence the metabolism of sex steroids. The production of SHBG is decreased. As a result, serum estradiol and testosterone concentrations are decreased, but free hormone levels remain within the normal range. However, the metabolism of these steroids is altered and differs from that found in individuals with hyperthyroidism.
The mechanism underlying menstrual abnormalities in hypothyroidism is incompletely understood. Because primary hypothyroidism is associated with elevated serum PRL in up to one-third of patients, it is plausible that hyperprolactinemia is a contributing factor (see Chapter 4). However, menstrual abnormalities are also observed in the absence of elevated PRL. Alterations in FSH and LH levels have been investigated, and studies have shown inconclusive results, although several studies suggest that the midcycle surge is absent. Menstrual function returns to normal with thyroid hormone replacement.
Thyroid dysfunction often presents with nonspecific symptoms, which delays the diagnosis. If only menstrual abnormalities are present, it is prudent to screen for thyroid abnormalities. Most cases are detected by TSH assays. Confirmation is obtained with a repeat TSH level and serum thyroid hormone levels.
The true prevalence of Müllerian tract abnormalities is not known. It is reported in as many as 45% of women. The reproductive consequences depend on the type of abnormality identified. A septate uterus is the most common defect described. These patients often present with infertility or obstetric complications. Other abnormalities include unicornous, bicornous, and didelphic uteri. These abnormalities present most commonly with reproductive or obstetric complications. Müllerian agenesis, androgen insensitivity syndrome, and congenital outflow obstruction defects are abnormalities that present with primary amenorrhea with no history of menses. In the following section these two abnormalities are discussed in greater detail.
Müllerian agenesis (Mayer-Rokitansky-Küster-Hauser syndrome) is the second most common cause of primary amenorrhea. It is a congenital condition that occurs in one in 5000 female births. These individuals have normal ovarian development, normal endocrine function, and normal female sexual development. The physical findings are a shortened or absent vagina in addition to absence of the uterus, although small masses resembling a rudimentary uterus may be noted (see the section "Embryology and Anatomy" at the beginning of this chapter). About one-third of patients have renal abnormalities, and several have had bone abnormalities and eighth nerve deafness. These individuals have a 46,XX karyotype.
The exact cause has not been identified. It is known that regression of Müllerian structures in males is controlled by AMH, which is secreted by the Sertoli cells of the testis. One hypothesis assigns the underlying defect to an activating mutation of either the AMH gene or its receptor. The genes for both AMH and AMH receptor have been investigated, but no mutations have yet been identified in patients with this syndrome.
Androgen Insensitivity Syndrome
Androgen insensitivity syndrome (AIS) presents in somewhat the same way as Müllerian agenesis. The presentation differs in that individuals with complete AIS have minimal sexual hair. These patients have a male karyotype with a mutation of the androgen receptor on the X chromosome. They have normal testicular development and endocrine function. However, because the internal and external male sexual structures need testosterone for development, they are absent. This results in a female phenotype. Because the testis still secretes AMH, Müllerian regression does occur. Secondary sexual characteristics (female) develop as a result of peripheral conversion of testosterone to estradiol, effectively resulting in unopposed estrogen stimulation.
The diagnosis of either disorder is entertained when pelvic examination reveals a short or absent vagina and no uterus on rectal examination. Confirmation of absent uterus can be obtained with ultrasound, MRI, or laparoscopy. These two disorders can usually be differentiated based on physical examination; patients with AIS have no pubic hair. However, the differential diagnosis becomes more difficult when patients have incomplete AIS. A testosterone level and karyotype can easily differentiate the two syndromes. This disorder is further discussed in Chapter 14.
Congenital Outflow Obstruction
Transvaginal septum and imperforate hymen are typical obstructive abnormalities. These patients usually present with cyclic lower abdominal pain and amenorrhea. The physical findings are a shortened or absent vagina. However, this syndrome differs from Müllerian agenesis in that the pelvic organs are present. Behind either defect is old blood that has not escaped with menses. The differential diagnosis is sometimes difficult, although bulging of the introitus suggests imperforate hymen, because the defect is thinner than a transvaginal septum.
The embryologic formations of the transvaginal septum and imperforate hymen are similar but not identical. Transvaginal septum is due to failure of complete canalization of the vaginal plate (see the section "Embryology and Anatomy" at the beginning of this chapter). The septum can vary in thickness and can be located at any level in the vagina. The hymen represents the junction of the sinovaginal bulbs and urogenital sinus. Typically, the hymen is perforated during fetal development. The hymen is thin and is always at the junction of the vestibule and vagina. It is important to distinguish these defects because the surgical correction procedures are different and require different levels of expertise.
Intrauterine adhesions or synechiae (Asherman syndrome) are an acquired condition that may obliterate the endometrial cavity. These patients usually present with a range of menstrual disturbances, infertility, and recurrent spontaneous abortions. The most frequent symptom is amenorrhea.
Intrauterine adhesions result from damage to the endometrial basal layer. A common antecedent factor is a surgical procedure within the uterine cavity, and most often it is endometrial curettage that occurs shortly after pregnancy. The concurrent presence of infection or heavy bleeding increases the risk. Endometrial tuberculosis and septic abortion are rare causes.
The diagnosis is entertained after demonstrating no withdrawal bleeding after administration of estrogen and progesterone. Confirmation is made with a hysterosalpingogram, saline sonogram, or hysteroscopy. Treatment involves lysis of adhesions and hormonal therapy.
The ovary is unique in that the age at which it ceases to function in women appears to have remained constant despite the increase in longevity experienced by women over the last century. Because the loss of ovarian function has a profound impact on the hormonal milieu in women and the subsequent risk of the development of disease resulting from the loss of estrogen production, improving our understanding of reproductive aging is critical for optimal female health.
Human follicles begin development in the fourth gestational month. Approximately 1000 to 2000 germ cells migrate to the gonadal ridge and multiply, reaching a total of 6 to 7 million around the fifth month of intrauterine life. At this point, multiplication stops and follicle loss begins, declining to approximately 1 million by birth. In the human male, the germ cells become quiescent and maintain their stem cell identity. In contrast, in the human female, between weeks 12 and 18, the germ cells enter meiosis and differentiate. Thus, in the female, all germ stem cells have differentiated prior to birth. In the adult woman, the germ cells may remain quiescent, may be recruited for further development and ovulation, or may be destroyed by apoptosis. Over time, the population of oocytes is depleted (without regeneration) through recruitment and apoptosis until less than a thousand oocytes remain and menopause ensues. Approximately 90% of women experience menopause at a mean age of 51.2 years (range, 46-55). The remainder experiences menopause prior to age 46 (often termed early menopause), with 1% of women experiencing menopause before age 40 years (POF, as discussed above).
Understanding ovarian aging has been difficult. The variability in definitions has made comparisons from study to study difficult. The participants in the Stages of Reproductive Aging Workshop (STRAW) developed criteria for staging female reproductive aging. They utilized menstrual cyclicity and early follicular FSH levels as the primary determinants for this staging system. Five stages precede the final menstrual period and two stages follow it. Stages −5 to −3 include the reproductive interval; stages −2 to −1 are termed the menopausal transition; and stages +1 and +2 are the postmenopause (Figure 13–14). The menopausal transition begins with increased variability in menstrual cyclicity (>7 days) in women with elevated FSH levels. This stage ends with the final menstrual period, which cannot be recognized until after 12 months of amenorrhea. Early postmenopause is defined as the first 5 years following the final menstrual period. Late postmenopause is variable in length, beginning 5 years after the final menstrual period and continuing until death.
Stages of reproductive aging.
(Reproduced, with permission, from Soules MR, et al. Stages of Reproductive Aging Workshop [STRAW]. J Women's Health Gend Based Med. 2001;10:843.)
This system is said to include endocrinologic aspects of ovarian aging, but it still depends largely on menstrual cyclicity as a key indicator of ovarian age. The system includes measurement of FSH; however, by the time FSH is elevated, even in the face of cyclic menstrual cycles, oocyte depletion has already proceeded to such an extent that fertility (as a marker of reproductive aging) is significantly diminished. Evidence suggests that genetic and environmental factors influence both age at menopause and the decline in fertility, although the specific nature of these relationships is poorly characterized. Premature menopause can be due to failure to attain adequate follicle numbers in utero or to accelerated depletion thereafter. Potentially, either of these causes could be affected by genetic and environmental factors. The timing of menopause has a consistent impact on overall health with respect to osteoporosis, cardiovascular disease, and cancer risk. Over the next decade, it is estimated that more than 40 million women in the United States will enter menopause.
Endocrine System Changes with Aging
The entire endocrine system changes with advancing age. The somatotrophic axis begins to decline in the fourth decade, prior to the decline in ovarian function. This decline is accelerated in the face of ovarian failure and may act to accelerate the decline in ovarian function. However, pituitary concentrations of GH as well as ACTH and TSH remain constant into the ninth decade. While the thyroid gland undergoes progressive fibrosis with age and concentrations of T3 decline by 25% to 40%, elderly patients still remain euthyroid. Beta cell function also undergoes degeneration with aging such that by age 65 years, 50% of subjects have abnormal glucose tolerance tests. Frank diabetes is less frequent, however, occurring in approximately 7% of the population. The female reproductive system, on the other hand, undergoes complete failure at a relatively early age.
Secretion of reproductive hormones during the menopausal transition (MT) was previously thought to decline progressively in a linear fashion, but hormone levels have since been shown to fluctuate widely. Studies in large cohorts of women have demonstrated that circulating FSH concentrations rise progressively during the MT. The initial monotropic rise in FSH is attributed to a decrease in ovarian inhibin secretion rather than to a decrease in estradiol production.
Inhibin and activin are proteins produced by the granulosa cells and have been shown to play major roles during the MT. Inhibin consists of a covalently bound dimer with an α subunit and one of two different β subunits, designated as βA and βB; the resulting heterodimers are known as inhibin A and inhibin B. Inhibin A is secreted by the corpus luteum and inhibin B by antral and dominant follicles. Consequently, inhibin A levels increase during the luteal phase, and inhibin B concentrations rise during the follicular phase. Both inhibins inhibit pituitary FSH secretion. Activins are a related class of proteins that stimulate pituitary FSH release. The activin molecule is a homodimer composed of two covalently linked inhibin β subunits, designated as activin A (βAβA) and activin B (βBβB).
AMH is also secreted by the granulosa cells of secondary and preantral follicles. Circulating concentrations remain relatively stable across the menstrual cycle and correlate with the number of early antral follicles. Levels of AMH decrease markedly and progressively across the MT to a time point approximately 5 years prior to the final period when levels are below the limit of detection.
During the late reproductive stage (stage −3), follicular phase inhibin B levels decrease as FSH concentrations rise. As the MT progresses, luteal phase inhibin A levels also decline. Activin A concentrations also are elevated in perimenopausal women. Whereas activins clearly play a local role in regulating pituitary FSH secretion, their ability to act as endocrine factors to influence the production of FSH has not been established. Thus, a decrease in secretion of inhibin A and inhibin B, and a corresponding increase in activin production may favor increased FSH secretion in the absence of any decrease (and perhaps an increase) in estradiol production.
The main circulating estrogen during the premenopausal years is 17β-estradiol. Levels of this hormone are controlled by the developing follicle and resultant corpus luteum. The fact that oophorectomy reduces peripheral estradiol levels from 120 to 18 pg/mL confirms that over 85% of circulating estradiol is derived from the ovary. In perimenopausal women, estradiol production fluctuates with FSH levels and can reach higher concentrations than those observed in young women under age 35. Estradiol levels generally do not decrease significantly until late in the MT. Estradiol levels may be quite variable, with chaotic patterns and occasionally very high or very low levels. This dramatic variability may lead to an increase in symptomatology during the perimenopausal years (stages −2 to −1). As peripheral gonadotropins rise, LH pulsatile patterns become abnormal. There is an increase in pulse frequency, with a decrease in GnRH inhibition by opioids. Despite continuing regular cyclic menstruation, progesterone levels during the early MT are lower than in women of mid-reproductive-age and vary inversely with body mass index. Women in the late MT exhibit impaired folliculogenesis and an increasing incidence of anovulation, compared to mid-reproductive-aged women. Essentially all estradiol in postmenopausal women is derived from peripheral conversion from estrone.
The predominant estrogen in the postmenopausal woman is estrone, with a biologic potency approximately one-third that of estradiol. The circulating levels of estrone (and estrone sulfate) in older women are approximately one-third to one-half of the concentration observed in women of reproductive age (Table 13–6). This is due to estrone production, resulting largely from peripheral aromatization of androstenedione (see earlier). This aromatase activity increases with aging by two- to fourfold and is further amplified by the increased adiposity that typically accompanies the aging process (see earlier). Estrone and estradiol production rates during the postmenopausal years are 40 and 6 μg/d, respectively. This compares with 80 to 500 μg/d for estradiol during the reproductive years.
Table 13–6 Plasma Concentrations of Sex Steroids in Premenopausal and Postmenopausal Women. ||Download (.pdf)
Table 13–6 Plasma Concentrations of Sex Steroids in Premenopausal and Postmenopausal Women.
|Estrone sulfate (pg/mL)||1000-1800a||350|
|Testosterone (ng/mL)||<1||< 1|
In contrast to estrogens, circulating androgens (DHEAS, androstenedione, and testosterone) decrease less dramatically after physiologic menopause. DHEAS undergoes a linear reduction with aging in both men and women (∼2% per year starting at age 30), but there is no specific decline associated with menopause. Changes in DHEAS levels have been associated with alterations in body composition with aging. Androstenedione production similarly decreases with aging, but the circulating levels are affected less because ovarian secretion is maintained, albeit at a reduced rate (see Table 13–6). Testosterone levels do not vary appreciably during the MT. Testosterone levels decrease after menopause. Historically it was believed the postmenopausal ovary produced a larger percentage of testosterone (50%) than did the premenopausal ovary. However, the literature is conflicting on the origin of androgens in postmenopausal women. The conventional view was that the adrenals, ovaries, and periphery all contributed to circulating androgen levels in postmenopausal women. Contemporary data suggest that androgen production after menopause is largely derived from adrenal precursors. This is supported by studies that have shown that postmenopausal women receiving dexamethasone suppression, or who have endogenous adrenal insufficiency, have undetectable levels of circulating androgens. Investigators have also demonstrated that the postmenopausal ovary has no appreciable enzymatic activities capable of generating sex steroids. The fact that women of reproductive age who undergo a bilateral oophorectomy have less testosterone and androstenedione than menopausal women with intact ovaries challenges this concept. However, the conflict may be explained by the hypothesis that the postmenopausal ovary produces androgens for a limited time.
Significant additional information regarding the hormonal changes during the menopausal transition is being developed in the multiethnic, community-based, longitudinal study of perimenopausal women at seven sites throughout the United States, The Study of Women's Health Across the Nation (SWAN). While most attention has been devoted to the role of the ovarian and oocyte decline in the onset of menopause, new evidence suggests potential alterations in the hypothalamic-pituitary feedback system. In evaluating the ability of estradiol to stimulate an effective LH surge, data from SWAN suggest the MT is characterized by three distinct hormonal patterns: (1) estrogen rise with rise in LH but anovulation; (2) estrogen rise without a concomitant rise in LH; or (3) failure of rise in either estradiol or FSH. Additionally, there may be ethnic differences in the sensitivity of the pituitary to negative feedback.
Given the endocrinologic changes associated with aging, many symptoms appearing in the aging female may be due to estrogen deficiency or diminished androgen or GH secretion. Disorders that are definitely due to estrogen deprivation include vasomotor symptoms and urogenital atrophy. Osteoporosis is also thought to be due largely to estrogen deficiency, and it may be exacerbated by the relative decline in GH levels. The same may be said for the hormone-related increase in the prevalence of atherosclerotic cardiovascular disease and psychosocial symptoms, including insomnia, fatigue, short-term memory changes, and possibly depression. Both DHEAS and GH may impact these phenomena as well. Most women who are symptomatic during the MT present with frequent or excessive bleeding or with hot flashes and other symptoms of estrogen deficiency. Other common symptoms during the MT include decreased libido, forgetfulness, vaginal dryness, and urinary incontinence.
Mood disorders also are increased during the MT. Community-based surveys have shown that perimenopausal women report significantly more psychological distress and have an increased risk for significant depression, compared with premenopausal or postmenopausal women. In an 8-year longitudinal study of premenopausal women having no prior history of depression, a depressive disorder was more likely to be diagnosed during the MT than in the premenopausal years (OR, 2.5; 95% CI, 1.25-5.02, p = 0.01).
Vasomotor symptoms (hot flushes) are experienced with greatest frequency during stages −1 and +1, with about 75% to 85% of women complaining of this symptom. Although 80% of US women have symptoms lasting for at least 1 year, only 25% of women are still symptomatic at 5 years after the final menstrual period. Studies of hot flushes with external monitoring of skin temperature and resistance have shown a frequency of approximately 54 ± 10 minutes. In sleep studies, hot flush frequency has been shown to interrupt rapid eye movement sleep and may contribute to some of the psychosocial complaints. Hot flushes are temporarily correlated with pulses of LH, but exogenous LH does not induce a flush, suggesting that there is some central mediator leading to both the flush and the elevation in LH.
Sleep disturbances, most likely the result of vasomotor activity, also are very common during the MT. In longitudinal studies of perimenopausal women, the prevalence of sleep disturbances has ranged from 32% to 40% in the early MT and from 38% to 46% in the late MT.
Vasomotor symptoms during the MT can be treated with hormone therapy (HT) using estrogen or progestin alone or in combination, neuroactive agents, or other nonhormonal alternatives. Estrogen therapy provides the best treatment for severe vasomotor symptoms, reducing their frequency and severity in 80% to 85% of women over 12 weeks compared with 30% of women receiving placebo. Hormone therapy also may help in the management of depression associated with the MT. In a randomized, placebo-controlled, 12-week trial involving perimenopausal women ages 40 to 55, symptoms of depression were improved in 68% of women receiving unopposed estrogen treatment (0.1 mg transdermal estradiol patch), compared with 20% of those receiving placebo (p = 0.001).
Other drugs have been tried in women for whom estrogen is contraindicated, although none have the efficacy associated with ERT. These alternatives include transdermal clonidine, ergot alkaloids, and, more recently, selective serotonin reuptake inhibitors and gabapentin. High-dose progestins may also produce some relief.
The vagina, vulva, urethra, and bladder trigone not only share embryonic proximity but all contain estrogen receptors. Atrophy begins in stage −2 to −1. The most common symptoms include itching and vaginal thinning, with decreased distensibility and reduced secretions, leading to vaginal dryness and pain with intercourse. This and the change in pH with resultant changes in vaginal flora increase the incidence of vaginal and urinary tract infections. Estrogen is the treatment of choice, and treatment must continue for at least 1 to 3 months for symptomatic improvement to be noted. The systemic dosage necessary for vaginal protection is somewhat higher than that needed for bone protection (see later), and local therapy by means of creams or vaginal rings may thus be advisable to limit systemic absorption. It should be noted, however, that vaginal absorption of steroids is quite efficient once estrogenization and revascularization have occurred. If the goal is to limit systemic absorption, slow-release rings may be superior to estrogen creams.
Vaginal estrogen frequently improves symptoms of urinary frequency, dysuria, urgency, and postvoid dribbling. Its direct effect to improve stress incontinence is less clear.
Osteoporosis is a condition in which bone loss is sufficient to allow fracture with minimal trauma. Major risk factors for the development of osteoporosis are the peak bone density attained in the late teens and early twenties (stressing the importance of bone building in the young) and the rate of loss (accelerated with estrogen deficiency). Primary or senile osteoporosis usually affects women between the ages of 55 and 70 years. The most common sites include the vertebrae and the long bones of the arms and legs. Secondary osteoporosis is caused by a specific disease (such as hyperparathyroidism) or medication usage (such as glucocorticoids) (see Chapter 8).
Menopausal bone loss begins before the final menstrual period during stage −1. Postmenopausal osteoporosis causes over 1.3 million fractures annually in the United States. Most of the more than 250,000 hip fractures are due to primary osteoporosis, and—given that 15% of patients die within a year after a hip fracture and 75% of patients lose their independence—the social costs, not to mention the financial costs, are great.
Bone loss following natural menopause is approximately 1% to 2% per year compared with 3.9% per year following oophorectomy. A woman's genetic background, lifestyle, dietary habits, and coexisting disease also influence the development of osteoporosis. Cigarette smoking, caffeine usage, and alcohol consumption also negatively affect bone loss, whereas weight-bearing activity appears to have a positive influence. See Chapter 8 for a detailed discussion of bone and mineral metabolism and osteoporosis.
Estrogen therapy—Estrogen therapy acts via the inhibition of bone resorption. Both BMD and fracture rate are improved with estrogen therapy. However, with cessation of estrogen therapy, there is a rapid and progressive loss of bone mineral content. By 4 years after therapy, bone density is no different from that of patients who were never treated with estrogen. Estrogen is approved for prevention of osteoporosis, and there is also some support for its usage as a treatment modality in established disease. Dosages of 0.625 mg of conjugated estrogens orally daily—and, more recently, as low as 0.3 mg—have been shown to slow bone loss and provide adequate protection against the development of osteoporosis. Higher dosages may be required to treat existing disease.
Alternative therapies for osteoporosis
Calcitonin—Calcitonin is a hormone normally secreted by the thyroid gland. Calcitonin (salmon) is available as a nasal spray specifically developed to decrease local side effects caused by subcutaneous injection. Although few studies have been performed and no data are available regarding reduction in hip fracture, it does seem to be especially beneficial for women with a recent and still painful vertebral fracture. Intranasal calcitonin has also been shown to improve spinal bone density and decrease the vertebral fracture rate in established osteoporosis. The increase in bone density appears to peak in as little as 12 to 18 months. The waning effects of calcitonin therapy over time may be due to downregulation of calcitonin receptors on osteoclasts and/or the development of neutralizing antibodies.
Bisphosphonates—These compounds are analogs of pyrophosphates and have a high affinity for hydroxyapatite in bone matrix. The basic structure of bisphosphonates allows a large number of manipulations of the basic molecule, producing different types of bisphosphonates that vary considerably in their potency on bone. In order of increasing potency are pamidronate, alendronate, risedronate, ibandronate, and zoledronic acid.
Alendronate has been evaluated extensively in patients with osteoporosis. Alendronate has been shown to inhibit markers of bone remodeling and increase BMD at the lumbar spine, hip, and total body and reduces fracture risk. Alendronate is taken orally; the recommended doses are 70 mg once weekly or 10 mg daily; alendronate must be taken according to a strict dosing schedule (in the morning on an empty stomach, with the patient required to remain upright for 30 minutes thereafter). The medication has very poor bioavailability (approximately 1%), and for that reason these instructions must be meticulously obeyed. Alendronate also has a propensity for causing irritation of the esophagus and stomach, especially in women with preexisting esophageal reflux, gastric or duodenal disease. Risedronate is similarly effective in the dosage of 35 mg weekly or 5 mg once daily, and the same dosing regimen is recommended. Ibandronate has the easiest schedule for administration 150 mg once monthly.
Increases in bone density with alendronate, risedronate, and ibandronate are greater than what is seen with calcitonin and similar to what is seen with HRT. The escape phenomenon seen with calcitonin is not seen with these oral bisphosphonates.
The final issue concerning long-term administration of bisphosphonates relates to their long half-lives in bone and their incorporation into the bone matrix. Although short-term fracture data appear favorable, the long-term effects of these agents on fracture incidence have never been assessed.
SERMs—Raloxifene is the first selective estrogen receptor modulator (SERM) approved for prevention and treatment of osteoporosis and for the prevention of breast cancer. SERMS act selectively as estrogen receptor agonists in some tissues (bone and heart) and antagonists in others (breast and uterus and possibly brain). Data with raloxifene show preservation of BMD, albeit less well than that seen with alendronate, risedronate, ibandronate, or HRT, and spinal fracture data support a protective effect.
It is believed that the differential effects of estrogens and antiestrogens are related to the transcriptional activation of specific estrogen response elements. Two different domains of the estrogen receptor (AF-1 and AF-2) are responsible for this transcriptional activation. Estrogens and antiestrogens appear to act via different domains, leading to their differential effects. Both act to maintain bone density—at least partially—via regulation of the gene for TGF-β.
Calcium and vitamin D—These are critical adjuvants for any type of antiresorptive therapy. Decreased ability to absorb calcium among older women is due in part to impaired vitamin D activation and effect. Older women may have limited exposure to sunlight, and their dietary vitamin D intake may be lower than that of younger women.
Daily intakes of 1000 to 1500 mg of calcium and 800 to 1200 IU of vitamin D are probably sufficient to reduce the risk of fragility fractures by 10% or more.
Anabolic therapy—The first anabolic therapy for the treatment of severe osteoporosis parathyroid hormone (1-34) or teriparatide has been approved for clinical use. Details about its efficacy and use in advanced postmenopausal osteoporosis are described in Chapter 8.
Atherosclerotic Cardiovascular Disease
Cardiovascular disease is the number one cause of mortality in both men and women in Western societies. This is largely attributed to age and lifestyle. Lifestyle modifications are known to decrease the incidence of atherosclerotic cardiovascular disease. For women, cardiovascular disease is largely a disease of the postmenopause. Women now spend more than a third of their lives in the postmenopausal years, and preventive measures are thus of paramount importance. Although a large body of observational evidence supported a protective effect of ERT on cardiovascular disease, observational data are limited by the confounding variables of patient self-selection. Animal and in vitro studies, as well as assessment of surrogate markers in women, have also shown a positive effect of ERT and HRT against cardiovascular disease development. However, several large randomized, controlled studies have failed to support a protective role for HRT in preventing cardiovascular events.
One of the first such trials was the Heart and Estrogen-Progestin Replacement Study (HERS), a secondary prevention trial that evaluated the use of daily HRT (0.625 mg conjugated estrogens plus 2.5 mg medroxyprogesterone acetate [MPA]) in 2763 postmenopausal women with a mean age of 66.7 years and preexisting vascular disease. The study failed to demonstrate any overall difference in vascular events. This occurred despite improvements in lipid parameters in those patients receiving HRT. The Estrogen Replacement and Atherosclerosis (ERA) Trial compared 3.2 years of treatment with estrogen, combined estrogen and progestin, and placebo in postmenopausal women aged 42 to 80 years. It also failed to demonstrate a significant difference in the rate of progression of coronary atherosclerosis between the three groups. The importance of this study was the inclusion of an estrogen-only arm.
The Women's Health Initiative (WHI) was the first large randomized study to look at primary prevention of cardiovascular disease. This study compared (1) the combination of conjugated equine estrogen (CEE) and MPA with placebo and (2) CEE to placebo. It was designed to assess the overall risks and benefits of HRT in a prospective randomized fashion.
The WHI demonstrated that there was an unacceptable risk profile for the combination HRT arm of the trial. There was an increase in the incidence in breast cancer (an increase of 8 cases per 10,000 women) with no cardiovascular protection (and potentially increased cardiovascular risk). There was, in fact, an increase in venous thromboembolism, strokes, and coronary heart disease. The risk of stroke and thromboembolism continued for the 5 years of study, whereas most of the coronary heart disease was limited to the first year of treatment. There were, however, documented decreases in the risk of fracture and colon cancer.
In the CEE-only arm, there was an increased risk of stroke and a decreased risk in hip fractures compared to placebo. This study showed that the use of ERT had no protective effect on coronary heart disease. Interestingly, there appeared to be a trend toward a reduction in breast cancer (0.77, CI 0.59-1.01) with use of ERT. More recent studies and reevaluation of large datasets have suggested that combined estrogen/progestin has a more profound increased risk on breast cancer, than estrogen alone or cyclic progestins, consistent with increase in mitogenic activity of the breast during the normal luteal phase when progesterone levels are high.
As noted earlier, long-term use of HT in older menopausal women has been associated with increased risks for venous thromboembolism, coronary events, stroke, and breast cancer. Although short-term treatment of symptomatic women during the MT likely poses significantly fewer risks, HT generally should be used in the lowest effective dose and for the shortest time required. Data suggest the use of HT does not increase the risk of breast cancer, on a yearly basis, more than would continuing spontaneous cycles for any given age. This suggests treatment of symptomatic women in their forties may not increase breast cancer risk over normal cycling. Low-dose estrogen regimens (conjugated equine estrogens, 0.3 mg daily, or its equivalent) can achieve as much as a 75% reduction in vasomotor symptoms over 12 weeks, approaching the efficacy of standard-dose HT regimens, and may be associated with fewer risks and side-effects. The decision to use HT should be made only after first carefully reviewing its risks and benefits for the individual.
The relative safety of HT during the MT has not been thoroughly investigated. The results of one observational study have suggested that women who start HT near menopause had a decreased risk of coronary heart disease when taking estrogen alone (relative risk [RR] = 0.66; 95% CI, 0.54-0.80) or in combination with progestin (RR = 0.72; 95% CI, 0.56-0.92). A secondary analysis conducted by the investigators involved in the WHI revealed that risk for coronary heart disease was not significantly increased in women under age 60 years of age or within 10 years of menopause. Further studies to evaluate the safety and efficacy of HT during the MT and the early postmenopausal years are ongoing.
Concerns about the risks of HT have increased interest in nonhormonal alternatives for the treatment of symptoms in the MT. In some women, vasomotor symptoms during the MT can be reduced by wearing layered clothing, avoiding caffeine and alcohol, and by keeping the ambient temperature a few degrees cooler. Herbal treatments such as black cohosh (Remifemin Menopause; Enzymatic Therapy, Green Bay, WI) have been shown to have marginal or no benefit in placebo- controlled trials. Neuroactive agents, including selective serotonin reuptake inhibitors (SSRIs), serotonin-norepinephrine reuptake inhibitors (SNRIs), alpha adrenergic agents, and others all have some efficacy in the treatment of vasomotor symptoms (Table 13–7). Both SSRIs and SNRIs may be effective because norepinephrine and serotonin appear to be involved in the hypothalamic regulation of temperature homeostasis and to play a role in the development of hot flashes. Randomized placebo-controlled trials have shown that SSRIs (citalopram, sertraline, paroxetine) and SNRIs (venlafaxine) can help to reduce the severity and frequency of hot flushes. Clonidine (an alpha adrenergic agonist) and gabapentin also have some efficacy. Gabapentin is the only agent compared head-to-head with estrogen suggesting equivalent efficacy when used in high doses.
Table 13–7 Treatment for Hot Flashes. ||Download (.pdf)
Table 13–7 Treatment for Hot Flashes.
|Alpha adrenergic agents||Clonidine||0.1|
|Gamma aminobutyric acid||Gabapentin||300-900|
The WHI study did not address the effect of hormone treatment on hot flushes and vaginal atrophy. Clearly, there are alternatives for the treatment of osteoporosis and cardiovascular disease that are superior if prevention of both conditions is the sole reason for HRT. Every woman should discuss with her caregiver the optimal management for her as an individual. This should take into account medical and family history as well as symptomatology. It can be uniformly recommended, however, that menopausal women maintain appropriate nutrition, weight reduction, and exercise along with moderation in alcohol and caffeine intake and cessation of smoking.
Infertility is defined as the inability of a couple to conceive after 1 year of frequent unprotected intercourse without contraception. This definition is based on observational data showing that approximately 85% of couples achieve pregnancy after 1 year of unprotected coitus. Using this definition, approximately 15% of reproductive-age couples experience infertility. However, the diagnosis of infertility does not mean that they cannot conceive—a more precise diagnosis term would be subfertility, or a diminished capacity to conceive. The actual probability of the fertility potential of a population may be better assessed with variables that can quantify a monthly cycle rate. The concepts that have been used for quantitative analysis are fecundability and fecundity. Fecundability is defined as the probability of achieving a pregnancy within one menstrual cycle, and in normal couples the chance of conception after 1 month is approximately 25%. Fecundity is a related concept that is defined as the ability to achieve a live birth within one menstrual cycle.
In the United States, demands for infertility treatment have dramatically increased. The 1995 National Survey of Family Growth reported that 9.3 million women received infertility treatment in their lifetime compared with 6.8 million in 1988. This rise in treatment is due not only to increased public awareness—it also reflects the significant demographic, societal, and economic changes in our society. These include the aging of the baby boom generation, which has increased the size of the reproductive-age population. Perhaps more important is the increased use of contraception and postponement of childbearing until the last two decades of a woman's reproductive life. Approximately 20% of women in the United States now have their first child after 35 years of age.
Age alone has a significant impact on fertility and affects a woman many years before the onset of menopause. One factor is the age-dependent loss of ovarian follicles (see later). A 38-year-old woman has 25% of the fecundability of a woman under 30 years of age. Another age-related subfertility factor is that the spontaneous abortion rate increases with advancing age. The overall incidence of clinical abortion increases from 10% in women under age 30 to more than 40% in women over 40. The increased pregnancy loss can be largely attributed to abnormalities in the aging oocyte; older follicles have an increased rate of meiotic dysfunction, resulting in higher rates of chromosomal abnormalities.
The main causes of female subfertility can be classified in the following way: (1) ovulatory defects, (2) pelvic disorders, and (3) male factors. These factors account for 80% to 85% of couples diagnosed with infertility. They are not mutually exclusive—about 15% of couples have more than one cause of subfertility. In approximately 20% of couples, the cause remains unknown and is classified as unexplained infertility. This section will briefly discuss the causes of subfertility and review the diagnosis and management.
Ovulatory disorders are responsible for 25% of cases of infertility. Ovulatory status can be obtained from the history. If a woman experiences cyclic, predictable menses at monthly intervals, ovulation can be predicted 98% of the time. This is not an invariable rule, however, and irregular menstrual cycles are not a sure sign of anovulation.
The only way to confirm ovulation is by achieving pregnancy. However, a variety of methods can indicate that ovulation has occurred. For example, a thermal shift occurs around the time of ovulation. Prior to ovulation, morning basal body temperature (BBT) is below 98°F (36.7°C), and after ovulation the temperature increases at least 0.4°F (0.2°C) for at least 1 to 3 days. This rise in temperature reflects the progesterone that is secreted from the corpus luteum, which consequently raises the hypothalamic set point for BBT. The temperature rise occurs approximately 2 days after ovulation because of the time and dose required for the progesterone effect at the hypothalamus. BBT can therefore not be utilized to prospectively predict ovulation. Measurements of midluteal serum progesterone concentration can also be performed to document the occurrence of ovulation. If this level is greater than 3 μg/L (9.5 nmol/L), it is a strong indication that ovulation has occurred. An endometrial biopsy can confirm ovulation. It is performed during the luteal phase and gives a qualitative assessment of ovulation, because the duration of progesterone exposure produces predictable endometrial histology. Lastly, a sonographic examination documenting a decrease in follicle size—or disappearance altogether of the previously developed follicle—is suggestive of ovulation. All of these methods indicate that ovulation has occurred.
There are only a few ways to predict that ovulation is going to occur. The most common way is to detect the LH surge. Ovulation typically occurs 34 to 36 hours after the onset of the LH surge.
Although ovulation may occur, some women may have a luteal-phase defect. This is characterized by an inadequate quantity or duration of progesterone secretion by the corpus luteum. There is a distinct window of time for implantation. The theory is that the progesterone deficiency desynchronizes ovulation (egg) and implantation (endometrium). However, the incidence of luteal phase defects is difficult to assess because the definition is not standardized. Typically, the diagnosis is established by luteal phase endometrial dating; if the histologic development of the endometrium lags more than 2 days beyond the day of the cycle, it is diagnostic of a luteal phase defect. However, up to 30% of women with normal cycles meet this criterion. Another method involves measuring midluteal progesterone levels. If progesterone is less than 10 ng/mL, it suggests a luteal phase defect. This is not reliable because progesterone is intermittently secreted, and the serum progesterone level can change from 1 hour to the next in the same individual. Furthermore, the lack of properly validated tests places the existence of luteal phase defects and their association with subfertility in question. Lastly, as empiric treatment for unexplained infertility has developed, delaying treatment (see later: superovulation with IUI) for exact diagnoses has less importance. More sophisticated testing of endometrial proteins required for implantation (eg, integrins, glycodelin) may revive enthusiasm for making a specific diagnosis in the future.
The cause of ovulatory dysfunction has been previously discussed. All anovulatory patients should have determination of PRL and TSH levels—and, if necessary, androgen levels—to identify the cause of the ovulatory disturbance. Treatment should be directed toward the cause, which is discussed in greater detail later. It could be argued that all patients should have evaluation of early follicular cycle (days 2-4) FSH, LH, and estradiol to assess ovarian reserve (age).
Pelvic disorders account for more than 30% of couples with the diagnosis of infertility. Uterine tube damage and adhesion formation are responsible for most pelvic pathologic processes causing infertility, whereas endometriosis is the primary pelvic disorder causing subfertility. The causes of tubal damage and adhesions include postinfectious state (pelvic inflammatory disease [PID]), endometriosis, and a history of pelvic surgery (especially surgery for ruptured appendicitis).
PID is defined as infection of upper genital tract structures and is usually caused by a sexually transmitted disease. The known initiating organisms are chlamydiae and Neisseria gonorrhoeae. The symptoms are variable but usually include lower abdominal pain, nausea, and vaginal discharge. However, in up to 30% of chlamydial infections, PID may be clinically inapparent and may remain undiagnosed until presenting with subfertility.
Endometriosis is the presence of endometrial glands and stroma outside of the uterus. There may be no manifestations other than infertility, or symptoms may progress to include severe pelvic pain, dysmenorrhea, and dyspareunia. The diagnosis is suspected if findings on surgical exploration show characteristic lesions. Lesions can be staged according to published criteria. Diagnosis is confirmed with biopsy of the peritoneal lesions. The disease occurs in approximately 3% to 10% of reproductive-age women and may be responsible for up to 25% to 35% of the female factors responsible for subfertility.
The pathogenesis of endometriosis is not completely known. A prominent theory (Sampson) involves retrograde menstruation. It is well established that menses can flow through the uterine tubes into the abdominal cavity. In fact, this phenomenon is thought to occur in almost all menstruating women. There is good evidence that the endometrial tissue subsequently invades and proliferates into the peritoneum. It is theorized that the immune system should normally dispose of the tissue and that altered immunity may result in implantation of this endometrial tissue outside the uterus in those women who subsequently develop endometriosis.
There is a strong association between adhesive disease and endometriosis. In these cases, the cause of subfertility is a result of distorted anatomy and consequently altered function. However, in mild cases of endometriosis, where only peritoneal lesions are identified and no anatomic distortion exists, the cause of infertility remains uncertain and controversial. There is evidence that the peritoneal fluid is altered in the presence of endometrial tissue with increased macrophages and inflammatory mediators. Several studies suggest that inflammatory changes result in adverse effects on folliculogenesis, ovum transport, fertilization, and implantation.
Tubal damage can be diagnosed with a hysterosalpingogram or surgical exploration. Hysterosalpingography involves the introduction of radiopaque contrast media into the pelvis through the cervix and then fluoroscopy, revealing an outline of contrast in the uterine cavity, uterine tubes, and peritoneal cavity. The diagnosis of pelvic endometriosis can only be made by surgery, and most often the surgical procedure is laparoscopy. Diagnostic laparoscopy is usually done when there is a high suspicion of endometriosis based on the clinical history or adhesive disease based on a history of PID or pelvic surgery. Laparoscopy may also be performed, if all other tests are normal and the couple continues not to achieve a pregnancy.
The male factor contributes in 40% to 50% of cases of diagnosed infertility, and all evaluations should include the male partner. The diagnostic test for male factor infertility is the semen analysis. Although this is a largely descriptive test (volume, sperm count, motility, and morphology), there is some correlation with pregnancy outcome. This should be used solely as a screening test. To understand the pathophysiology of male factors, it is important to review the physiology of spermatogenesis and the anatomy of the male reproductive tract as described in Chapter 12.
Approximately 15% to 20% of the couples diagnosed with infertility have no identifiable cause after a full investigation. The workup should include documentation of ovulation (regular menstrual cycles, progesterone levels, or LH testing), hysterosalpingography to verify patent fallopian tubes, and a semen analysis. The term unexplained implies that there is a potential explanation for the subfertility, but the cause has yet to be identified. The cause may be subtle abnormalities in folliculogenesis, sperm-ovum interactions, or defective implantation.
Several studies have evaluated the natural history of unexplained infertility. It is estimated that fecundity in younger couples (female partner under the age of 40) with unexplained infertility is 3% to 5% compared with 20% to 25% in the age-matched couples with normal fertility. Treatment involves methods that increase fecundability and are discussed later.
Management of the Infertile Couple
It is important to remember that in most couples, there is a chance for spontaneous conception. Recent studies estimate that the average probability for live birth without treatment is 25% to 40% during the 3 years after the first infertility consultation. This translates into a cycle fecundity rate of 0.7% to 1% per month. The presence of endometriosis, abnormal sperm, or tubal disease independently reduces the chance of spontaneous pregnancy and live birth by approximately 50% for each variable. Infertility for more than 3 years, female age over 30 years, and primary infertility were important negative prognostic factors.
Evaluation should focus on known causes of infertility or subfertility: ovulatory defects, pelvic disorders (tubal disease, endometriosis), and male factor issues.
Treatment should be diagnosis-specific, if possible. For the female, this means that the cause of any ovulatory defect should be determined and specific treatment then instituted. This enhances outcome and decreases the risk of complications (spontaneous abortion and multiple gestation). This treatment may include the use of dopamine agonists (hyperprolactinemia), thyroid replacement (hypothyroidism), pulsatile GnRH (hypogonadotropic hypogonadism), or clomiphene citrate (for PCOS). The most common cause of anovulation is inappropriate feedback such as in PCOS. Ovulation in patients with PCOS can be induced with clomiphene citrate, a nonsteroidal agonist-antagonist of estrogen that blocks the hypothalamic-pituitary axis from feedback by circulating estrogens. As a result, there is increased gonadotropin release to stimulate follicular recruitment and ovulation. In addition, because PCOS is associated with insulin resistance, and elevated insulin directly has an impact on the ovary, insulin sensitizers such as metformin have been used to enhance ovulatory response in women with PCOS. Recently, a large multicenter randomized trial compared clomiphene to metformin and found clomiphene to be significantly more effective for ovulation induction and achieving a live birth.
In general, adhesive tubal lesions should be treated surgically. The location and extent of disease should, however, be evaluated. Patients with distal tubal occlusion—unless it is very mild—are most often better served by assisted reproduction (in vitro fertilization [IVF]). Other possible causes of infertility should also be examined. If a patient has a history of documented tubal disease, in addition to other abnormalities (ovulatory dysfunction or male factor infertility), or if they are over 35, the likelihood for successful surgical management decreases by approximately 50%, and consideration for avoiding surgery and moving directly to assisted reproduction is paramount. The exception to this rule is documentation of hydrosalpinges on ultrasound. The presence of hydrosalpinges that retain fluid when nondistended (ie, not seen only with hysterosalpingography) leads to a significant reduction in outcome with assisted reproductive therapy (ART). Prior removal or proximal occlusion of the tube to prevent contamination of the uterine cavity should be performed before ART is offered.
There are conflicting data in the literature concerning the appropriate treatment for mild endometriosis. A well-designed randomized trial from Canada evaluated the effect of surgical treatment on pregnancy outcome for patients diagnosed with mild endometriosis without anatomic distortion. It showed that pregnancy rates at 9 months postlaparoscopy were 27% in the surgically treated group compared with 18% in the untreated group. Severe endometriosis (disease that alters the pelvic anatomy or involves the ovary with endometriomas) should be surgically treated to restore normal pelvic anatomy. There appears to be no advantage to medical therapy for endometriosis in women seeking fertility.
Male factor infertility is discussed in Chapter 12. Like female partner treatment, therapy should be targeted, if possible, toward the cause of subfertility. Obstructive disease may be treated surgically. A prominent varicocele with a stress pattern on semen analysis (decreased motility with an increase in abnormal morphology) may suggest a need for surgical repair. Any endocrinologic abnormalities (while less common in the male) should be treated (eg, prolactinoma). Unfortunately, beyond this point, most treatments require a combined approach very similar to that discussed below for unexplained infertility.
The treatment of unexplained infertility can be frustrating for the physician and the patients because the recommendations for therapy are not targeted toward a specific diagnosis. Although there are very limited evidence-based data to guide treatment, therapy should be directed toward increasing the fecundability rate. The two main treatments are superovulation plus intrauterine insemination (IUI) and IVF.
Superovulation methods are designed to qualitatively improve the cycle and to hyperstimulate the ovary with rescue of more follicles (quantitative improvement). Administration of clomiphene citrate or of gonadotropins alone can be used for superovulation. However, higher success rates are observed when superovulation is combined with IUI of washed sperm (Table 13–8). IVF provides a higher fecundability rate. However, these procedures are significantly more costly and more invasive, and they should only be used after a trial (three or four cycles) of superovulation and IUI has failed. With older patients, aggressive therapy should be considered earlier in the treatment effort.
Table 13–8 Pregnancy Rates after Treatment for Unexplained Infertility. Aggregate Data for Each Treatment. ||Download (.pdf)
Table 13–8 Pregnancy Rates after Treatment for Unexplained Infertility. Aggregate Data for Each Treatment.
|Treatment||No. of Studies||No. (%) of Pregnancies per Initiated Cycle||Percent of Quality-Adjusted Pregnancies per Initiated Cycle|
|Control groups||11||64/3,539 (1.8)||1.3|
|Control groups, randomized studies||6||23/597 (3.8)||4.1|
|Clomiphene citrate||3||37/617 (6)||5.6|
|Clomiphene citrate + IUI||5||21/315 (6.7)||8.3|
|hMG + IUI||14||207/1133 (18)||17.1|
In the United States, oral contraceptive pills are the most widely used method for contraception. There are two types of oral contraception: combination pills and progestin-only pills. The various hormones used in birth control pills are illustrated in Figure 13–17.
Oral contraceptive pill hormonal components. All except endogenous hormones are synthetic steroids.
The development of oral contraceptive agents began with the isolation of progesterone. However, progesterone was very expensive and difficult to isolate. Ethisterone, a derivative of an androgen, was found to have progestin activity and was much easier to isolate than progesterone. With removal of carbon 19, the progestational activity was increased, and the new compound was termed norethindrone. When this hormone was administered to women, ovulation was inhibited. During the process of norethindrone purification, an estrogen contaminant was found. When this contaminant was removed, women would experience breakthrough bleeding. The estrogen was added back, thereby creating the first-generation combination birth control pill, which was approved by the Food and Drug Administration (FDA) in 1960.
Oral contraceptives can be divided into generations based on dose and type of hormone. The first-generation birth control pills contained more than 50 μg of ethinyl estradiol or mestranol and a progestin. The adverse events associated with high-dose estrogen, such as coronary thrombosis, led to development of the second-generation pill, which contained less than 50 μg of ethinyl estradiol and progestins other than levonorgestrel derivatives. Next, attention was directed toward the progestin, which was thought to have adverse androgenic effects such as affecting lipid profiles and glucose tolerance. This led to the development of third-generation pills that contained both a lower dose of estrogen (20-30 μg of ethinyl estradiol) and newer progestins (gonanes: desogestrel or norgestimate). Indeed, studies have demonstrated a reduction in metabolic changes associated with these progestins, but limited data are available to show any actual reduction of cardiovascular events. Another recently developed progestin, drospirenone, has antimineralocorticoid and antiandrogenic activity in addition to its pharmacologic progestational effects. As an analog of spironolactone rather than androgen, it competitively binds to aldosterone receptors, and it may counteract the estrogen stimulation of the renin-angiotensin system, resulting in more weight stability and less water retention. A new combination oral contraceptive, Yasmin, which contains 3 mg drospirenone and 30 μg ethinyl estradiol, has been approved by the FDA and is prescribed for women with hyperandrogenism or other side-effects attributable to oral contraceptives. However, the relative antiandrogenic activity of drospirenone is small compared with cyproterone acetate or the therapeutic dose of spironolactone used for the treatment of hirsutism.
Contraceptives can be classified also based on formulas or schedules of administration. The theory behind phasic preparations was to further decrease the amount of total progestin administered in an attempt to reduce metabolic changes attributed to the progestin, thereby decreasing adverse effects. The traditional monophasic pill (eg, Loestrin) contains 30 μg of ethinyl estradiol and 1.5 mg of norethindrone. This dose is given every day for 3 weeks with a 1-week hormone-free interval. The progestin dose remains constant throughout the cycle. The second type is the biphasic pill (eg, Ortho-Novum 10/11), which contains 35 μg of ethinyl estradiol and either 0.5 or 1 mg of norethindrone. The 0.5 mg of norethindrone is administered in the first 10 days of the month, and the 1 mg is administered for the following 11 days. The last 7 days of the cycle are free of hormone. With this combination, there was a theoretical increase in breakthrough bleeding and an increased pregnancy rate. A meta-analysis revealed no difference, but limited data were available.
Because of concerns that this regimen might result in both breakthrough bleeding and pregnancies, another phasic formulation was developed. The triphasic pills (eg, Triphasil, Ortho-Novum 7/7/7) contain 0.5, 0.75, and 1 mg norethindrone combined with 35 μg ethinyl estradiol. Theoretically, this formulation improves cycle control. There are several other regimens, some of which alter estrogen doses to simulate the estrogen cyclic rhythm (Triphasil-30, Triphasil-40, Triphasil-30 μg ethinyl estradiol) and possibly decrease breakthrough bleeding. A meta-analysis comparing biphasic versus triphasic pills revealed that triphasic pills significantly improved cycle control. However, the progestins in each pill tested were different, and this could account for better cycle control rather than the phasic formulation. An additional meta–analysis was performed on triphasic versus monophasic pills to assess cycle control and metabolic effects. This analysis revealed no difference between the formulations. Therefore, there is little scientific rationale for prescribing phasic preparations in preference to the monophasic pill.
The pharmacologic activity of progestins is based on the progestational activity and bioavailability of each progestin as well as the dose. The relative potencies of the different progestins are levonorgestrel greater than norgestrel greater than norethindrone. The active estrogen component of oral contraceptives is ethinyl estradiol (even if mestranol is administered).
When the hormones are administered for 21 days of the cycle, there is enough progestin to inhibit rapid follicle growth for about 7 more days. Figure 13–18 demonstrates that during the steroid-free interval there is no rise in estrogen, indicating no follicular maturation. It is likely that pills missed after this time are responsible for some of the unintended pregnancies. Therefore, it is important that this interval not be extended.
Progestin effects on steroidogenesis and ovulation.
(Reproduced, with permission, from Brenner PF, et al. Serum levels of d-norgestrel, luteinizing hormone, follicle-stimulating hormone, estradiol and progesterone in women during and following ingestion of combination oral contraceptives containing dl-norgestrel. Am J Obstet Gynecol. 1997;129:133.)
Pharmacologic doses of progestin inhibit ovulation by suppressing GnRH pulsatility and possibly inhibiting release of pituitary LH. Progestins also impair implantation and produce thick, scanty cervical mucus that retards sperm penetration. These latter methods play a minor role in the mechanism of oral contraception.
Ethinyl estradiol helps prevent the selection of a dominant follicle by suppressing pituitary FSH. In addition to FSH suppression, ethinyl estradiol provides stability to the endometrium, decreasing breakthrough bleeding. It also upregulates the progesterone receptor and decreases clearance, thereby potentiating the activity of the progestin.
Traditionally, a pill is administered daily for 3 weeks out of 4, preferably at the same time each day (more critical with progestin-only pills). This regimen was designed to mimic the menstrual cycle with monthly withdrawal bleeding. The conventional start date is on the first Sunday after menses (first day of menses for triphasics). An alternative method is to start at the time of the clinic visit, regardless of the day of the menstrual cycle, with a backup method for 7 days (Quick-Start). This method has the advantage of immediate contraception without adverse bleeding events. In addition, with the Quick-Start method, women are more likely to start the second pack of oral contraceptives, suggesting increased compliance. During the 28-day regimen, there is a 1-week steroid-free interval. If the steroid-free interval is prolonged beyond the 7-day window, ovulation is possible. Therefore, this is a critical time not to neglect taking pills. On the other hand, a woman may continue to take the hormone pills and skip the steroid-free interval to avoid monthly bleeding. A randomized trial comparing continuous oral contraception with the traditional cyclic method revealed a significantly greater incidence of erratic bleeding with (overall) the same number of days of bleeding, albeit with a reduced volume of bleeding. New formulations have been developed for those who do not desire cyclic bleeding. This regimen involves 84 days of continuous hormone administration followed by a steroid-free interval of 1 week. This can easily be done with the traditionally packaged oral contraceptive pills.
A routine for daily administration improves compliance and contraceptive efficacy. Failure to take the pill at the same time every day and not understanding the package insert are associated with missing two or more pills during the cycle. To decrease failure rates, women should understand that if they forget to take the pill, they must use barrier prophylaxis.
Postpartum women who are not breast-feeding may begin combination oral contraceptives 3 weeks after delivery. For women who are breast-feeding, it is advised that institution of combination oral contraceptives be delayed until 3 months postpartum. The recommendation for this delay is due to decreased milk letdown secondary to estrogen but may be waived once lactation is well established.
Noncompliance increases the incidence of unwanted pregnancies. Appropriate use of birth control is achieved 32% to 85% of the time in the general population. Teenagers have at most a 50% continuation rate, and 25% of pill users discontinue the practice in the first year. The efficacy of the oral contraceptive under conditions of perfect use is 0.1 failures per 100 woman-years (0.1 per 100 users). With typical use, the failure rate is 3%, with first-year failure rates approaching 7.3% to 8.5%. Side-effects contribute to noncompliance. The most common side-effect is breakthrough bleeding. Other unwanted symptoms include bloating, breast tenderness, nausea, and possibly headaches, weight gain, and depression. Some studies suggest that altering estrogen doses may improve symptoms. Failure rates may also be associated with concomitant use of drugs (eg, rifampin, hydantoins) that accelerate hormone metabolism.
There are noncontraceptive benefits to the pill. These include reduced monthly blood loss (less iron deficiency) and less dysmenorrhea as well as reduced benign breast disease and mastalgia. Oral contraceptives also reduce the incidence of PID and ectopic pregnancies. Other significant benefits include a reduction in risk of ovarian cancer, endometrial cancer, and colorectal cancer. Oral contraceptive use also has cosmetic benefits; it can improve excess hair growth and acne (see PCOS). It is not unusual for the pill to be administered for noncontraceptive problems.
In general, oral contraceptives have proved to be safe for most women, but the possibility of adverse effects has received much attention. Unfortunately, the literature is full of conflicting reports. Data concerning controversial adverse effects will be discussed in the following section.
Although the estrogen in combination oral contraceptives tends to increase triglycerides and total cholesterol, these levels are still within the normal range, and the contraceptives appear to increase HDL and decrease LDL. Progestins attenuate these effects, which suggest an adverse metabolic milieu. Although it is known that low HDL-LDL ratios are associated with cardiovascular events, the importance of lipid changes associated with birth control pills is unknown. To date there is no strong evidence of an increased incidence of myocardial infarctions in healthy nonsmoking oral contraceptive users. In patients with other cardiovascular risk factors such as hypertension—at least in Europe—there is an up to 12-fold increased risk of cardiovascular events. Women who are over 35 years of age and who smoke are at increased risk of cardiovascular events. This risk is amplified with use of the birth control pills. If more than 15 cigarettes per day are smoked, there is a relative risk (RR) of 3.3 for a cardiovascular event, compared with an RR of 20.8 with concomitant use of oral contraceptives. However, if a woman smokes fewer than 15 cigarettes per day, there is an RR of 2.0 for a cardiovascular event, compared with an RR of 3.5 with concomitant contraceptive use. Former smokers after 1 year have no significant increased risk. First-generation oral contraceptives (>50 μg ethinyl estradiol) imposed an RR for stroke (ischemic or hemorrhagic) of 5.8. With low-dose agents (<50 μg ethinyl estradiol), there appears to be no significant increased risk of stroke among healthy normotensive nonsmoking women. The RR associated with hemorrhagic stroke for hypertensive women is 10.2 to 14.2. Because the potential exists for adverse outcomes, women taking oral contraceptives should be screened regularly for cardiovascular risk factors to ensure safe administration.
The risk of deep vein thrombosis and pulmonary embolism is increased twofold to threefold with administration of the pill. The mechanism by which oral contraceptives enhance venous thrombosis is unknown, but there may be estrogen-related changes in coagulation parameters. These include increased clotting factors and activation of platelets and a decrease in protein S and fibrinolytic activity. However, these changes in measured serum clotting factors do not predict the occurrence of deep vein thrombosis. Genetic thrombophilias increase the risk of venous thrombosis. The prevalence of factor V Leiden in the general population is 5%. The incidence of deep vein thrombosis among this population is 60 per 100,000 per year, and with use of oral contraceptives, the incidence approaches 280 to 300 per 100,000 per year. The baseline incidence of deep vein thrombosis in women is approximately 3 per 100,000 per year, and with current oral contraceptive uses it is 9.6 to 21.1 per 100,000 per year. For comparison, during pregnancy, the incidence of deep vein thrombosis is 60 per 100,000 per year. Older age (>40-44) increases the incidence two- to threefold but does not affect the RR. There is no evidence that smoking has an effect on the incidence of deep vein thrombosis with oral contraceptive use. At this time, universal screening for thrombophilias is not cost effective. However, any history of deep vein thrombosis warrants a workup for thrombophilia.
The association of oral contraceptives and cancer risk has been evaluated for breast, cervical, and liver cancer. Observational studies investigating a possible association between oral contraceptive use and breast cancer have reported conflicting results. The most recent information is that oral contraceptive usage (current or past use) has no impact on the incidence of breast cancer—RR 1.0 (CI 95% 0.8-1.3)—among women 35 to 64 years of age. Several observational studies have linked oral contraceptive use with invasive cervical cancer, although it is not clear if this association is causally related. A recent study investigating the association between oral contraception use and cervical cancer revealed a nearly threefold increased risk among human papillomavirus carriers with 5 to 9 years of oral contraceptive use (RR 2.82; 95% CI 1.46-5.42). This evidence suggests that women taking oral contraceptives should be screened yearly with Pap smears to prevent cervical cancer. In the 1980s, there was an association between hepatocellular carcinoma in women under 50 years of age and oral contraceptive use. With further investigation there appears to be no increased risk of hepatic cancer with the use of oral contraceptives.
After discontinuation of oral contraceptives, the activity of the HPO axis gradually returns to a precontraceptive state. After a 2- to 4-week prolongation of the follicular phase, the LH peak is observed, which suggests that the suppressive effects of the oral contraceptive have dissipated and that cyclic menses will resume.
Contraindications to oral contraceptive administration are summarized in Table 13–9.
Table 13–9 Contraindications to Combination Oral Contraceptive Use. ||Download (.pdf)
Table 13–9 Contraindications to Combination Oral Contraceptive Use.
|History of thromboembolic disease|
|Smoking and age ≥ 35 y|
|History of cerebrovascular or cardiovascular disease|
|Undiagnosed vaginal bleeding|
|First-degree relative with history of thromboembolic disease|
Progestin-only pills (Ortho Micronor, Nor-QD, 0.35 mg norethindrone; Ovrette, 0.075 mg levonorgestrel) are also available for contraception. The target population for administration of progestin-only contraception includes women with contraindications to estrogen, breast-feeding mothers, and older women.
The circulating levels of progestin following ingestion of progestin-only pills (minipill) are 25% to 50% of that following ingestion of estrogen-progestin oral contraceptive pills. Serum levels that peak 2 hours after administration are followed by rapid elimination (Figure 13–19). The peak levels of norethindrone and levonorgestrel vary (4-14 and 0.9-2 ng/mL, respectively), but 24 hours after pill ingestion, serum levels are 0.2 to 1.6 ng/mL and 0.2 to 0.5 ng/mL, respectively. Thus, there is no accumulation of progestin over time. Progestin-only administration results in lower steady state levels and a shorter half-life compared with concomitant administration with estrogen.
Serum levels attained with progestin-only pills.
[Reproduced, with permission, from McCann MF, Potter LS. Progestin-only oral contraception: a comprehensive review. Contraception. 1994;50(6 suppl 1):S1.]
Owing to the lower levels of progestin in the progestin-only pill, there is less influence on the inhibition of ovulation and more impact of thickening cervical mucus (hostile environment) inhibiting sperm penetration. Sperm that are able to penetrate have decreased mobility. Progestins also alter the endometrial lining (inhibition of progesterone receptor synthesis and reduction in endometrial glandular development, preventing implantation) and perhaps inhibit the motility (number and motility of the cilia) of the uterine tube. LH peaks—as well as FSH peaks—are suppressed compared with pretreatment levels. The change in cervical mucus takes place 2 to 4 hours after the first dose. However, after 24 hours, thinning of the cervical mucus is evident, allowing unimpaired sperm penetration, which is why it is critical to take the progestin-only pills at the same time every day.
The progestin-only pill should be started on the first day of menses. This pill should be taken at the same time every day. If administration is 3 hours late, a backup method should be used for 48 hours. If a pill is missed, a backup method should be used for 48 hours, and if two or more pills are missed, a backup method should be used for 48 hours due to rapid resumption of the cervical mucus effect. If no menses occur in 4 weeks, a pregnancy test is necessary. Progestin-only pills may be administered immediately postpartum.
The efficacy of progestin-only pills under conditions of perfect use is 0.3 to 3.1 failures per 100 woman-years (failure rates of 1.1%-9.6% in the first year) or 0.5 per 100 users. This efficacy rate is achieved only with careful compliance. The typical use is associated with a greater than 5% failure rate. Failure rates were lowest in women over 38½ years of age and those who were breast-feeding. The efficacy may also be influenced by body weight and by concomitant use of anticonvulsants. The major disadvantage is that the pill must be administered at the same time every day. As a result of even slight flexibility in the schedule, there is increased contraceptive failure.
The risks associated with progestin-only pills are minimal. Various studies have revealed no significant impact on lipids, carbohydrate metabolism, blood pressure, or the incidence of myocardial infarction and stroke. Furthermore, no adverse coagulation parameters have been associated with its use. There are almost no data on the association of progestin-only pills and endometrial, ovarian, cervical, or breast cancer. The major side-effect is breakthrough bleeding (40%-60%). Other side-effects include acne and persistent ovarian cysts. With discontinuation of the pill, menses resume with no impact on subsequent pregnancy rates or future fertility.
Contraception: Long-Acting Contraceptives
The high rate of unintended pregnancies has led to the development of long-acting reversible contraceptive modalities. Interest in long-acting methods is increasing because they offer convenience, obviate problems of compliance, and thus offer higher efficacy. Most long-acting systems contain either combination or progestin-only hormones. The effectiveness of these hormones is prolonged, mostly due to the sustained system that results in a gradual release. The modes of administration include injectables, transdermal patches, subdermal rods, vaginal rings, and intrauterine devices. The various types of long-acting contraceptives are discussed below.
Injectable progestins that contain medroxyprogesterone acetate (MPA, Depo-Provera) are beneficial when women have contraindications to estrogen, use antiepileptics, are mentally handicapped, or have poor compliance. Furthermore, there is good evidence that its use is safe in the presence of coronary artery disease, congestive heart failure, diabetes, tobacco use, and a history of venous thromboembolism. Other uses of MPA include treatment of metastatic endometrial or renal carcinoma.
Although most other long-acting contraceptives are sustained-release formulations, Depo-Provera (150 mg MPA) is provided as an aqueous microcrystalline suspension that gradually declines throughout the cycle (Figure 13–20). Pharmacologic levels (>0.5 ng/mL) are achieved within the first 24 hours and peak (at 2 ng/mL) within the first week after the injection. Serum concentrations are maintained at 1 ng/mL for approximately 3 months. Interestingly, estrogen concentration is in the early to midfollicular level (below 100 pg/mL) and persists for 4 months after the last injection. The serum concentration of MPA decreases to 0.2 ng/mL during the last 5 to 6 months (ovulation occurs when levels are <0.1 ng/mL). However, one study observed progesterone levels to rise after 3½ months.
MPA levels following injection of Depo-Provera. Open bars represent serum estradiol; closed bars represent serum progesterone concentrations.
(Reproduced, with permission, from Ortiz A, et al. Serum medroxyprogesterone acetate [MPA] and ovarian function following intramuscular injection of depo-MPA. J Clin Endocrinol Metab. 1977;44:32.)
The mechanism of action depends mainly on the ability of higher peaks of hormone to inhibit ovulation (LH surge). Like other progestins, MPA increases cervical mucus viscosity, alters the endometrium, and decreases the motility of the uterine tubes and uterus. FSH levels are minimally suppressed with Depo-Provera.
The manufacturer's recommendation is to administer the agent every 3 months, starting within 5 days of the onset of menses and not to exceed 1 week. The agent is injected deeply into the upper-outer quadrant of the buttock or deltoid without massage to ensure slow release. If the subject is postpartum and not breast-feeding, Depo-Provera should be given within 3 weeks after delivery and if lactating within 6 weeks (Table 13–10).
Table 13–10 Scheduling for Injectable Contraceptives. ||Download (.pdf)
Table 13–10 Scheduling for Injectable Contraceptives.
|Spontaneous menstrual cycle||Within 5 d of menses onset||Within 5 d of menses onset|
|Spontaneous or elective first-trimester abortion||Within 7 d||Within 7 d|
|Term delivery||Within 3 wk postpartum if not lactating; within 6 wk postpartum if lactating||Between 21 and 28 d postpartum if not lactating|
|Switching from combination OCs||With administered active pills or within 7 d after administering the pill pack's last active tablet||With administered active pills or within 7 d after administering the pill pack's last active tablet|
|Switching from DMPA||—||Within 13 wk after last DMPA injection|
|Switching from MPA/E2C||Within 33 d after previous injection||—|
|Switching from levonorgestrel implant||Any time within 5 y of implant insertion; use of a condom back-up is recommended for 1 wk||Any time within 5 y of implant insertion; use of a condom backup is recommended for 1 wk|
|Switching from Copper T 380A IUD||First injection should occur before IUD removal and within 10 y after IUD insertion; a condom should be used as a back-up if the first injection is not administered within 5 d of menses onset||First injection should occur before IUD removal and within 10 y after IUD insertion condom should be used as a back-up if the first injection is not administered within 5 d of menses onset|
|Injection interval||Every 12 wk or 3 mo; earlier reinjections are acceptable||Every 28 d or 4 wK or monthly; reinjection earlier than 23 d may impair cycle|
|Grace period||2 wk (14 wk from last injection); after 1 wk manufacturer recommends pregnancy testing before repeat injection||Control ± 5 d (23-33 d from last injection); thereafter, pregnancy testing needed before repeat injection|
Because compliance is not an issue, the failure rate is minimal at 0 to 0.7 per 100 woman-years (0.3 per 100 users). Weight and use of concurrent medications do not affect the efficacy. However, continuation rates are poor at 50% to 60% because of the side-effect profile. The major dissatisfaction that leads to discontinuation is breakthrough bleeding, which approaches 50% to 70% in the first year of use. Other side-effects include weight gain (2.1 kg/y), dizziness, abdominal pain, anxiety, and possibly depression. Another disadvantage with the use of Depo-Provera is a delay in fertility after discontinuation. Ovulation returns when serum levels are less than 0.1 ng/mL. The time from discontinuation to ovulation is prolonged. Only 50% of patients ovulate at 6 months after discontinuing the medication, and although the agent does not cause infertility, achieving pregnancy may be delayed for more than 1 year. (The length of time for release at the injection site is unpredictable.) After the first year, 60% of women become amenorrheic, and at 5 years, the incidence of amenorrhea approaches 80%, which can be considered a potential benefit. Other benefits with use of MPA include prevention of iron deficiency anemia, ectopic pregnancy, PID, and endometrial cancer. In addition, Depo-Provera is a recommended contraceptive for women with sickle cell disease (decreased crisis) and seizure disorders (raises seizure threshold). Other therapeutic uses include dysmenorrhea and endometrial hyperplasia or cancer.
One major concern with use of Depo-Provera is the development of osteopenia, with possible advancement to osteoporosis later in life. Several observational studies have evaluated the potential impact on bone. A prospective trial revealed that current users, after 12 months of use, experience a mean BMD loss of 2.74%. However, on examining former users 30 months later, it was found that mean BMD was similar to that of nonusers, indicating that the loss is reversible and of minimal clinical importance. An ongoing multicenter study assessing BMD in users versus nonusers should clarify the impact of Depo-Provera on bone. BMD in adolescents has also been investigated because adolescence is a critical time in bone mineralization. A small prospective study revealed that BMD was decreased by 1.5% to 3.1% after 1 and 2 years of use in Norplant users compared to increases in BMD of 9.3% and 9.5% in controls. This is a potential concern and has also led to a prospective multicenter study investigating the use of Depo-Provera in adolescents. Although one possible cause is less exposure to estrogen, an alternative and perhaps not exclusive theory involves MPA-dependent glucocorticoid activity that impairs osteoblast differentiation. Other potential risks include an adverse lipid profile (increase in LDL, decrease in HDL) and a slightly increased risk of breast cancer. The association of breast cancer with use of Depo-Provera is minimal within the first 4 years of use, with no risk after 5 years of use. Paradoxically, MPA has been used for treatment of metastatic breast cancer.
The development of monthly combination injectables (Lunelle) has responded to the erratic bleeding associated with Depo-Provera (Figure 13–21). The cycle control is similar to what is achieved with combined oral contraceptives. The monthly withdrawal bleeding occurs 2 weeks after the injection. The target populations are adolescents and women who have difficulty with compliance. Lunelle is an aqueous solution containing 25 mg of MPA and 5 mg of estradiol cypionate per 0.5 mL. In women who receive repeated administration of Lunelle, peak estradiol levels occur approximately 2 days after the third injection and are 247 pg/mL (similar to peak ovulatory levels). The estradiol level returns to baseline 14 days after the last injection (100 pg/mL); the drop in estradiol is associated with menstrual bleeding (2-3 weeks after the last injection). Peak MPA levels (2.17 ng/mL) occur 3½ days after the third monthly injection. The mean MPA level is 1.25 ng/mL. The level at day 28 of the cycle is 0.44 to 0.47 ng/mL (level needed for contraceptive effect is 0.1-0.2 ng/mL). The earliest return of ovulation seen in women with multiple injections has been 60 days after the last dose. The mechanism of action is similar to that of combined oral contraceptives.
Serum MPA and estradiol levels following Lunelle injection.
[Reproduced, with permission, from Rahimy MH, Ryan KK, Hopkins NK. Lunelle monthly contraceptive injection (medroxyprogesterone acetate and estradiol cypionate injectable suspension): steady-state pharmacokinetics of MPA and E2 in surgically sterile women. Contraception. 1999;60:209.]
Lunelle is administered intramuscularly in the buttock or deltoid every month. The first injection should be given within the first 5 days of the menstrual cycle (Table 13–10). Even though pharmacokinetic analysis reveals a delay in ovulation, the manufacturer recommends a 5-day grace period. The failure rate is 0.1 per 100 woman-years. Neither body weight nor use of concomitant drugs appears to affect the efficacy. Although this contraceptive has the advantages of the oral contraceptives and is associated with better compliance, the continuation rate is only 55%. This may be due to its side-effect profile, which is similar to that of the combined oral contraceptives with the addition of monthly injections.
There are limited data on potential risks. The risk potential is probably similar to that of the combined oral contraceptives, with a potentially lower incidence of deep vein thrombosis secondary to the absence of the first-pass effect. On discontinuation of Lunelle, achieving pregnancy may be delayed for as long as to 3 to 10 months after the last injection.
Lunelle was approved for use in the United States in 2000. However, in October 2002, Lunelle was recalled from the market, due to plant manufacturing problems. Alternatives to Lunelle that are administered outside the United States are: Mesigyna, Perlutal, Yectames, and Chinese Injectable No 1.
The Norplant package consists of six capsules (34 mm in length, 2.4 mm in diameter), with each capsule providing 36 mg of levonorgestrel (total 216 mg). The target population is women who have contraindications to or adverse side-effects from estrogen, women who are postpartum or breast-feeding, and adolescent mothers. This method provides long-term continuous contraception (approved for 5 years) that is rapidly reversible. The advantages, side-effects, risks, and contraindications are similar to those of oral progestins. The major disadvantage—not present with use of oral progestins—is the surgical insertion and removal of the rods. A newer system, Norplant II, contains two rods (4 cm in length, 3.4 cm in diameter) and releases 50 μg/d of norgestrel (approved for 3 years). The two-rod system has the same mechanism of action and side-effect profile as its predecessor. However, the rods are much easier and faster to insert and remove than the capsules.
Within the first 24 hours, serum concentrations of levonorgestrel are 0.4 to 0.5 ng/mL. The capsules release 85 μg of levonorgestrel per 24 hours for the first year (equivalent to the daily dose of progestin-only pills) and then 50 μg for the remaining 5 years. The mean serum levels of progestin after the first 6 months are 0.25 to 0.6 ng/mL, slightly decreased at 5 years to 0.17 to 0.35 ng/mL. A levonorgestrel concentration below 0.2 ng/mL is associated with increased pregnancy rates. The site of implantation (leg, forearm, and arm) does not affect circulating progestin levels. Even though progestin levels are sufficient to prevent ovulation within the first 24 hours, the manufacturer recommends use of a backup method for 3 days after insertion. On removal, the progestin levels rapidly decline and undetectable serum levels are achieved after 96 hours. As a result, most women ovulate within 1 month after removal of the implants.
Norplant provides contraception in several ways. In the first 2 years, the levonorgestrel concentration is high enough to suppress the LH surge—most likely at the hypothalamic level—and thereby inhibits ovulation. However, given the low concentrations of progestin, there is no real effect on FSH. The estradiol levels approximate those in ovulatory women. In addition, there are irregular serum peaks (often prolonged) and declines in serum estrogen levels that may contribute to erratic bleeding. By 5 years, more than 50% of the cycles are ovulatory. However, ovulatory cycles while using Norplant have been associated with luteal phase insufficiency. Other mechanisms of contraception are similar to oral progestins and include thickening of the cervical mucus, alterations of the endometrium, and changes in tubal and uterine motility.
The failure rate is 0.2 to 2.1 failures per 100 woman-years (0.9 per 100 users). Like oral progestins, body weight affects circulating levels and may result in more failures in the fourth or fifth year of use. Similar to oral progestins, the incidence of ectopic pregnancy among failures is increased to 20% (overall incidence is 0.28-1.3 per 1000 woman-years). The continuation rate (discontinuation rate of 10%-15% per year) is age-dependent and ranges from 33% to 78%. Menstrual disturbances are the most frequent side-effect; they approach 40% to 80%, especially in the first 2 years. Although the incidence of abnormal uterine bleeding is similar to that of Depo-Provera, a significant difference between these methods is that Norplant provides only a 10% amenorrhea rate at 5 years. Other reported side-effects include headache (30% indication for removal) and possibly weight gain, mood changes, anxiety, and depression—as well as ovarian cyst formation (eightfold increase), breast tenderness, acne, galactorrhea (if insertion occurs on discontinuation of lactation), possible hair loss, and pain or other adverse reactions at the insertion site (0.8% of cases at discontinuation).
The transdermal patch (Ortho Evra) is another approach to contraception. The thin 20-cm2 patch is composed of a protective layer, a middle (medicated) layer, and a release liner that is removed prior to application. The system delivers 150 μg of norelgestromin (active metabolite of norgestimate) and 20 μg of ethinyl estradiol per day to the systemic circulation. The target population is similar to that described above for Lunelle. One advantage of this system over Lunelle is that there are no monthly injections, and as a result, there is greater autonomy for the patient. The patch is applied once a week for 3 consecutive weeks, followed by a patch-free week for monthly withdrawal bleeding. The patch should be changed on the same day each week. The mechanism of action, contraindications, and side effects are similar to what has been described in the section on oral contraceptives.
With use of the transdermal patch, the peak ethinyl estradiol and norelgestromin levels are 50 to 60 pg/mL and 0.7 to 0.8 ng/mL, respectively. Because of this unique delivery system, hormone levels achieve a steady state condition throughout the cycle (see later and Figure 13–22). After the seventh day of application, there are adequate hormone levels to inhibit ovulation for 2 more days. With each consecutive patch, there is minimal accumulation of norelgestromin or ethinyl estradiol. The amount of hormone delivered is not affected by the environment, activity, or site of application (abdomen, buttock, arm, torso). The adhesive is very reliable in a variety of conditions, including exercise, swimming, humidity, saunas, and bathing. Complete detachment occurs in 1.8% of cases and partial detachment in 2.9% of cases.
Comparative serum steroid (CSS) levels of norelgestromin (NGMN) and ethinyl estradiol (EE) following patch administration.
(Reproduced, with permission, from Abrams LS, et al. Multiple-dose pharmacokinetics of a contraceptive patch in healthy women participants. Contraception. 2001;64:287.)
The failure rate is 0.7 per 100 woman-years under conditions of perfect use. Body weight has not been shown to affect the efficacy. The compliance with perfect use ranges from 88.1% to 91% among all age groups. This is significantly different from what is achieved with oral contraceptives (67%-85%), especially with women under 20 years of age. The side-effect profile is similar to that of oral contraceptives, except that there is slightly more breakthrough bleeding with the transdermal patch in the first 1 to 2 months (up to 12.2% vs 8.1%) and less breast tenderness (6.1% vs 18.8%). The incidence of skin reaction was 17.4%, characterized as mild in 92%, resulting in discontinuation in under 2%.
Since the early 1900s, it has been recognized that the vagina is a place where steroids can be rapidly absorbed into the circulation. A study in the 1960s revealed that silicone rubber pessaries containing sex steroids would release the drug at a continuous rate. These studies led to the development of contraceptive vaginal rings.
As with oral contraceptives, there are combination and progestin-only formulations. Several progestin-only rings have been introduced since the 1970s. However, they were associated with significant menstrual disturbances. More recently, combination types have been developed. The most recent (2002) FDA-approved vaginal ring is a combination type called the NuvaRing.
The NuvaRing is made of ethylene vinyl acetate that provides 0.015 mg of ethinyl estradiol and 0.120 mg of etonogestrel per day. Maximum serum concentrations are achieved within 1 week after placement. The ring is designed to be used for 21 days and then removed for 1 full week to permit withdrawal bleeding. This device is capable of inhibiting ovulation within 3 days after insertion. After removal, the time to ovulation is 19 days. The mechanism of action, contraindications, and risks are similar to those of oral contraceptives. However, when assessing systemic exposure, use of the vaginal ring allows for 50% of the total exposure to ethinyl estradiol (15 μg in the ring compared with a 30-μg ethinyl estradiol-containing oral contraceptive).
The failure rate is similar to that reported with oral contraceptives. The continuation rate is 85.6% to 90%. Irregular bleeding is minimal (5.5%), and overall, the device is well tolerated, with an associated 2.5% discontinuation rate. Side-effects are similar to those of oral contraceptives, but cycle control appears to be improved. The reported incidence of vaginal discharge is 23% versus 14.5% with oral contraceptives. The ring does not appear to interfere with intercourse (1%-2% of partners reported discomfort); however, the device can be removed for 2 to 3 hours during intercourse without changing efficacy.
Intrauterine devices (IUDs) are another modality of contraception that has been used clinically since the 1960s. Historically, these devices were made of plastic (polyethylene) impregnated with barium sulfate to make them radiopaque. Several other devices were subsequently developed, including the Dalkon Shield. After the introduction of the Dalkon Shield, an increase in pelvic infections was observed secondary to its multifilament tail. Furthermore, tubal infertility and septic abortions were increasing, and massive litigation occurred as a result. Consequently, even though the modern IUD has negligible associated risk, the use of IUDs in the United States is minimal—less than 1% of married women.
Currently, two types of IUDs are used in the United States: the copper- and the hormone-containing devices. The most recent FDA-approved intrauterine system contains levonorgestrel (Mirena) and is approved for 5 years of use. Several studies have demonstrated that these devices are unlike the Dalkon Shield and are very safe and efficacious. The target population is women who desire highly effective contraception that is long-term and rapidly reversible.
The copper (TCu-380A) IUD is a T-shaped device. The mechanism of action is mostly spermicidal due to the sterile inflammatory reaction that is created secondary to a foreign body in the uterus. The abundance of white blood cells that are present as a result kills the spermatozoa by phagocytosis. The amount of dissolution of copper is less than the daily amount ingested in the diet. However, with release of copper, salts are created that alter the endometrium and cervical mucus. Sperm transport is significantly impaired, limiting access to the oviducts.
There are two hormone-containing intrauterine devices: the progesterone-releasing device (Progestasert) and the levonorgestrel-releasing device (Mirena). The Progestasert contains progesterone, which is released at a rate of 65 mg/d (approved for 1 year). This diffuses into the endometrial cavity, resulting in decidualization and atrophy of the endometrium. Serum progesterone levels do not change with the use of Progestasert. The main mechanism of action is to impair implantation. Mirena contains 52 mg of levonorgestrel, which is gradually released at a rate of 20 μg/d (approved for 5 years). Unlike Progestasert, systemic absorption of levonorgestrel inhibits ovulation about half the time. Although women may continue to have cyclic menses, over 40% have impaired follicular growth, with up to 23% developing luteinized unruptured follicles. Other mechanisms of action are similar to those described for Progestasert and the progestin-only pills. Mirena has the added advantage of significantly decreasing menstrual flow and has been used to treat menorrhagia.
The IUD should be placed within 7 days after onset of the menstrual cycle or at any time postpartum. Protection begins immediately after insertion. The failure rates after the first year of use are 0.5% to 0.8% for the TCu-380A, 1.3% to 1.6% for Progestasert, and 0.1% to 0.2% for Mirena. The expulsion rate is approximately 10%.
If a woman becomes pregnant with an IUD in place, the incidence of an ectopic pregnancy is 4.5% to 25%. The incidence of ectopic pregnancies with IUDs varies depending on the type of device. With Progestasert, the ectopic rate is slightly higher (6.80 per 1000 woman-years), most likely because its mechanism of action is limited to inhibiting implantation in the endometrium—in contrast to the copper or levonorgestrel IUD (0.2-0.4 per 1000 woman-years), both of which also interfere with conception.
The continuation rate range for the current IUDs is 40% to 66.2% (Mirena). The side-effects of the copper IUD include dysmenorrhea and menorrhagia. The most common adverse effect associated with the hormone-containing devices is erratic, albeit significantly less, bleeding. In fact, 40% of women experienced amenorrhea at 6 months and 50% at 12 months. The incidence of spotting in the first 6 months was 25% but decreased to 11% after 2 years. Other reported side-effects of levonorgestrel include depression, headaches, and acne. There is a tendency to develop ovarian cysts early after insertion with the levonorgestrel-containing device that resolves after 4 months of use.
The nominal risks associated with IUD use include pelvic infection (within 1 month after insertion), lost IUD (ie, perforation into the abdominal cavity; 1:3000), and miscarriage. There is no association between IUD use and uterine or cervical cancer. Contraindications to IUD use are active genital infection and unexplained bleeding.
Postcoital contraception is a method that may be used by a woman who believes her contraceptive method has failed or who has had unprotected intercourse and feels that she may be at risk for an unintended pregnancy. The first study to evaluate the efficacy of emergency contraception with hormones was in 1963. Subsequently, several studies have been performed with various contraceptives that paved the way for more widespread use. In 1997, the FDA approved the use of high-dose oral contraceptives for postcoital contraception. Since then, pharmaceutical companies have marketed specific packaging for the use of emergency contraception. Other methods, including mifepristone (RU-486) and the IUD have also been effective for postcoital contraception.
Similar to the formulas of oral contraception, there are both combination and progestin-only types of emergency contraception. The combination method (Yuzpe regimen) entails administration of two doses of two tablets (Ovral: 50 μg ethinyl estradiol, 0.5 mg norgestrel) 12 hours apart (total: 200 μg ethinyl estradiol, 2 mg norgestrel). Other oral contraceptives may be used with adjustment in the number of pills for equivalence (ie, two doses of four tablets: any second-generation oral contraceptive 12 hours apart). The specific medication (Preven) that is FDA approved and marketed for postcoital contraception contains two doses of four tablets (ethinyl estradiol 50 μg, levonorgestrel 0.25 mg) 12 hours apart. The progestin-only method involves two doses of 10 pills (Ovrette 0.075 mg) 12 hours apart (Plan A). The marketed form (Plan B) contains two doses of one tablet (levonorgestrel 0.75 mg) 12 hours apart.
Following a single oral dose of 0.75 mg of levonorgestrel, the serum concentration peak (5-10 ng/mL) was at 2 hours with a rapid decline during the first 24 hours. The mechanism of action is uncertain, but levonorgestrel most likely inhibits ovulation and alters the endometrium to prevent implantation. Studies have shown decreased sperm recovery from the uterine cavity, possibly due to thickened cervical mucus, or the alkalinization of the intrauterine environment. Others have shown that changes in other factors, such as integrins, can alter endometrial receptivity. The mode of action likely depends on the timing of intercourse relative to ovulation and to the administration of emergency contraception.
Maximum efficacy is achieved if the first dose is administered within 72 hours after intercourse and repeated in 12 hours. Combination formulas have a failure rate of 2% to 3%, and progestin-only preparations have a failure rate of 1%. Emergency contraception effectively reduces the rate of unintended pregnancies from 8% to 2%, a 75% reduction. However, with increasing time since unprotected intercourse, the efficacy changes from 0.4% to 1.2% to 2.7% for the first, second, or third 24-hour period after unprotected intercourse. For maximum efficacy, emergency contraception may be prescribed in advance so women will already have the correct dosing. No increase in risk-taking behavior has been noted with this strategy.
Significant nausea or emesis (51.7%) is associated with use of emergency contraception, although it is substantially less with progestin-only formulations. An antiemetic should be administered 1 hour before each treatment. If a patient vomits within 1 hour after ingestion, additional pills need to be administered.
Contraindications to emergency contraception with the combination regimen are possibly the same as those described for oral contraceptives; for progestin-only pills, there are no contraindications. Emergency contraception should be an optional function of the rape management protocol.