CHAPTER 15: Reproductive Endocrinology

Reproductive endocrinology is the study of hormones and neuroendocrine factors that are produced by and/or affect reproductive tissues. These tissues include the hypothalamus, anterior pituitary gland, ovary, endometrium, and placenta. A hormone is classically described as a cell product that is secreted into the peripheral circulation and that exerts its effects in distant target tissues (Fig. 15-1). This is termed endocrine secretion. Additional forms of cell-to-cell communication exist in reproductive physiology. Paracrine communication, common within the ovary, refers to chemical signaling between neighboring cells. Autocrine communication occurs when a cell releases substances that influence its own function. Production of a substance within a cell that affects that cell before secretion is termed an intracrine effect.

A neurotransmitter, in classic neural pathways, crosses a small extracellular space called a synaptic junction and binds to dendrites of a second neuron (Fig. 15-2). Alternatively, these factors are secreted into the vascular system and are transported to other tissues where they exert their effects in a process termed neuroendocrine secretion or neuroendocrine signaling. One example is gonadotropin-releasing hormone (GnRH) secretion into the portal vasculature with effects on the gonadotropes within the anterior pituitary gland.

Normal reproductive function requires precise quantitative and temporal regulation of the hypothalamic-pituitary-ovarian axis (Fig. 15-3). Within the hypothalamus, specific centers or nuclei release GnRH in pulses. This decapeptide binds to surface receptors on the gonadotrope subpopulation of the anterior pituitary gland. In response, gonadotropes secrete glycoprotein gonadotropins, luteinizing hormone (LH) and follicle-stimulating hormone (FSH), into the peripheral circulation. Within the ovary, LH and FSH bind to the theca and granulosa cells to stimulate folliculogenesis and ovarian production of steroid hormones (estrogens, progesterone, and androgens), gonadal peptides (activin, inhibin, and follistatin), and growth factors. Among other functions, these ovarian-derived factors feed back to the hypothalamus and pituitary gland to inhibit or, at the midcycle surge, to augment GnRH and gonadotropin secretion. The ovarian steroids are also critical for preparing the endometrium for placental implantation if pregnancy ensues.

Hormones can be broadly classified as either steroids or peptides, each with their own mode of biosynthesis and mechanism of action. The receptors for these hormones can be divided into two groups: (1) those present on the cell surface, which in general interact with hormones that are water soluble, namely peptides, and (2) those that are primarily intracellular and interact with lipophilic hormones such as steroids. Hormones are normally present in serum and tissues in very low concentrations. Therefore, receptors must have both high affinity and high specificity for their ligand to produce the correct biologic response.

Peptide Hormones: LH, FSH, and hCG

The gonadotropins LH and FSH are biosynthesized and secreted by the gonadotrope subpopulation of the anterior pituitary gland. These hormones play a critical role in stimulating ovarian steroidogenesis, follicular development, and ovulation. The closely related peptide human chorionic gonadotropin (hCG) is produced by placental trophoblast and is important for maintenance of pregnancy.

LH, FSH, and hCG are heterodimers consisting of a common glycoprotein α-subunit linked to a unique β-subunit, which provides functional specificity. Although glycoprotein α- and β-subunits can be found in their unassociated form in the circulation, these “free” subunits are not known to have biologic activity. Nevertheless, their measurement may be useful in screening tests for conditions such as pituitary adenomas and pregnancy.

The LH and hCG β-subunits are encoded by two separate genes within a gene grouping called the LH/CG cluster. The amino acid sequence of the human LH and CG β-subunits demonstrates approximately 80-percent similarity. However, the hCG β-subunit contains an additional 24-amino-acid extension on the carboxy terminus. The presence of these additional amino acids has allowed the development of highly specific assays that can distinguish LH from hCG.

In pituitary thyrotropes, the shared glycoprotein α-subunit also interacts with the thyroid-stimulating hormone β-subunit to form thyroid-stimulating hormone (TSH). This similarity between TSH and hCG can have clinical sequelae. For example, molar pregnancies frequently produce very high levels of hCG, which can bind to TSH receptors, resulting in hyperthyroidism (Walkington, 2011).

Human Chorionic Gonadotropin

This glycosylated peptide hormone is produced by the placental syncytiotrophoblast. With this molecule, the degree and type of glycosylated moieties attached to the peptide frame is variable and may indicate pregnancy stage, placental function, or pathology (Fournier, 2015). One example is the hyperglycosylated hCG that is found more commonly in gestational trophoblastic neoplasia.

hCG can be detected in serum as early as 7 to 9 days after the LH surge. In early pregnancy, hCG levels increase rapidly, doubling approximately every 2 days. Levels of this peptide hormone peak at approximately 100,000 mIU/mL during the first trimester of pregnancy. This is followed by a relatively sharp decline in the early second-trimester concentrations and then maintenance at lower levels throughout the remainder of pregnancy.

hCG binds to LH/CG receptors on corpus luteum cells and stimulates steroidogenesis in the ovary. To maintain endometrial integrity and uterine quiescence, hCG levels are critical. Namely, hCG supports corpus luteum steroid production during early pregnancy before the placenta attains adequate steroidogenic capability. The transition in production of estrogens and progesterone from the ovary to the placenta is often called the “luteal-placental shift.” In addition to effects on ovarian function, hCG exerts autocrine/paracrine effects in the placenta, promoting syncytiotrophoblast formation, trophoblast invasion, and angiogenesis.

As the placenta is the primary source for hCG production, measurement of plasma hCG levels has proved to be an effective screening tool for pregnancies with altered placental mass or function. Relatively elevated levels of hCG are observed in multifetal gestations and fetuses with Down syndrome. Lower hCG levels are observed in cases of poor placentation including ectopic pregnancy or spontaneous miscarriage. Serial hCG measurements can be very helpful to monitor these latter conditions as the doubling time is relatively reliable. Markedly abnormal elevations in hCG levels are most often observed in the presence of gestational trophoblastic disease, discussed in Chapter 37.

Human CG is also secreted by nontrophoblastic neoplasias and can serve as a useful tumor marker. Ectopic (nonplacental) production of hCG, either the intact dimer or the β-subunit, is frequently associated with germ cell tumors and has been reported in various tumors arising from the mucosal epithelium of the cervix, bladder, lung, gastrointestinal tract, and nasopharynx. It has been postulated that hCG inhibits apoptosis in these tumors, thereby allowing rapid growth.

In addition to secretion by placental syncytiotrophoblast, hCG is produced by nonneoplastic cell types and presumably serves other functions (Cole, 2010). For example, cytotrophoblasts secrete a hyperglycosylated variant of hCG that may prove to be a sensitive marker of early pregnancy (Chuan, 2014). The pituitary gonadotropes also make small amounts of hCG. These concentrations rise in postmenopausal women and may be a rare cause of erroneously positive hCG testing in this age group (Cole, 2008).

Steroid Hormones

Classification

Sex steroids are divided into three groups based on the number of carbon atoms that they contain. Each carbon in this structure is assigned a number identifier, and each ring is assigned a letter (Fig. 15-4). The 21-carbon series includes progestins, glucocorticoids, and mineralocorticoids. Androgens contain 19 carbons, whereas estrogens have 18.

Steroids are given scientific names according to a generally accepted convention in which functional groups below the plane of the molecule are preceded by the α symbol and those above the plane of the molecule are indicated by a β symbol. A Δ symbol indicates a double bond. Those steroids with a double bond between carbon atoms 5 and 6 are called Δ⁵ steroids and include pregnenolone, 17-hydroxypregnenolone, and dehydroepiandrosterone. Those with a double bond between carbons 4 and 5 are termed Δ⁴ steroids and include progesterone, 17-hydroxyprogesterone, androstenedione, testosterone, mineralocorticoids, and glucocorticoids.

Steroidogenesis

Sex steroid hormones are synthesized in the gonads, adrenal gland, and placenta. Cholesterol is the primary building block. All steroid-producing tissues, except the placenta, are capable of synthesizing cholesterol from the two-carbon precursor, acetate. Steroid hormone production, which involves at least 17 enzymes, primarily occurs in the mitochondria and the abundant smooth endoplasmic reticulum found in steroidogenic cells (Mason, 2002). These enzymes are members of the cytochrome P450 superfamily. As such, genes that encode these enzymes begin with CYP.

Steroidogenic enzymes catalyze four basic modifications of the steroid structure: (1) side-chain cleavage (desmolase reaction), (2) conversion of hydroxyl groups to ketones (dehydrogenase reactions), (3) addition of a hydroxyl group (hydroxylation reaction), and (4) removal or addition of hydrogen to create or reduce a double bond (Table 15-1). The steroid biosynthesis pathway is shown in simplified form in Figure 15-5. This pathway is identical in all steroidogenic tissues, but the distribution of products synthesized by each tissue is determined by the presence of requisite enzymes. For example, the ovary is deficient in 21-hydroxylase and 11β-hydroxylase and thus is unable to produce corticosteroids. Of note, many steroidogenic enzymes exist as multiple isoforms, each with different precursor preferences and directional activities. As a result, specific steroids may be produced via multiple pathways in addition to the classic pathway shown in Figure 15-5 (Auchus, 2009).

Estrogens are synthesized by aromatization of C19 androgens by aromatase. The aromatase enzyme is a cytochrome P450 enzyme encoded by the gene CYP19. In addition to the ovary, aromatase is expressed in significant levels in adipose tissue, skin, and brain (Boon, 2010). Importantly, sufficient estrogen can be derived from peripheral aromatization to produce endometrial bleeding in postmenopausal women, especially those who are overweight or obese.

Circulating estrogens in the reproductive-aged female include estrone (E₁), estradiol (E₂), and estriol (E₃). Estradiol is the primary estrogen produced by the ovary during reproductive years. Levels are derived both from direct synthesis in the granulosa cells of developing follicles and through conversion of the less potent estrone. Estrone, the primary estrogen during menopause, is secreted primarily by the ovary. Estriol, the predominant estrogen during pregnancy, is primarily secreted from the placenta. However, both estrone and estriol can be converted from androstenedione in the periphery.

The ovary also produces androgens in response to LH stimulation of theca cell function. The primary products are the relatively weak androgens androstenedione and dehydroepiandrosterone (DHEA), although smaller amounts of testosterone are also secreted. Although the adrenal cortex primarily produces mineralocorticoids and glucocorticoids, it also contributes to approximately one half of the daily production of androstenedione and DHEA and essentially all of the sulfated form of DHEA (DHEAS). In women, 25 percent of circulating testosterone is secreted by the ovary, 25 percent is secreted by the adrenal gland, and the remaining 50 percent is produced by peripheral conversion of androstenedione to testosterone (Fig. 15-6) (Silva, 1987).

The adult adrenal gland is composed of three zones. Each of these zones expresses a different complement of steroidogenic enzymes and as a result synthesizes different products. The zona glomerulosa lacks 17α-hydroxylase activity but contains large amounts of aldosterone synthase (P450aldo) and therefore produces mineralocorticoids. The zona fasciculata and zona reticularis, both of which express the 17α-hydroxylase gene, synthesize glucocorticoids and androgens, respectively.

Within androgen synthesis, the 5α-reductase enzyme converts testosterone to dihydrotestosterone (DHT), a more potent androgen. DHT promotes transformation of vellus hair to terminal hair. Thus, medications that antagonize 5α-reductase are often effective in the treatment of hirsutism (Stout, 2010). This enzyme exists in two forms, each encoded by a separate gene. The type 1 enzyme is found in the skin, brain, liver, and kidneys. In contrast, the type 2 enzyme is predominantly expressed in male genitalia (Russell, 1994).

Steroid Hormone Transport in the Circulation

Most steroids in the peripheral circulation are bound to carrier proteins. These proteins may be either specific proteins, such as sex hormone-binding globulin (SHBG), thyroid-binding globulin, or corticosteroid-binding globulin, or nonspecific proteins such as albumin. Only 1 to 2 percent of androgens and estrogens are unbound or free.

Only the unbound steroid fraction is believed to be biologically active, although albumin’s low affinity for sex steroids likely allows steroids bound to this protein to exert some effect. The amount of free hormone is in equilibrium with the amount of bound. In other words, the amount of free, biologically active hormone is inversely related to the amount of bound hormone, and the amount of bound hormone is a direct reflection of the levels of carrier protein. As a result, small changes in carrier protein expression can produce substantial alterations in steroid effect.

SHBG circulates as a homodimer that binds a single steroid molecule. This binding protein is primarily synthesized in the liver, although it has also been detected in the brain, placenta, endometrium, and testes. SHBG levels are increased by hyperthyroidism, pregnancy, and estrogen administration. In contrast, androgens, progestins, growth hormone (GH), insulin, and corticoids decrease SHBG levels. An increase in weight, particularly central body fat, can significantly blunt SHBG expression. In turn, this decreases bound hormone levels and increases active hormone levels (Hammond, 2012).

Clinically, unbound hormone can be technically difficult to measure, and results should be interpreted with caution. Free testosterone levels are the most commonly ordered free steroid hormone tests, but the most accurate assays are performed by only a few commercial laboratories (Rosner, 2007). The more available calculated free levels are relatively inaccurate. Moreover, free testosterone measurement is rarely necessary for clinical diagnosis in the female and is unlikely to add more information than the total testosterone level. For example, measurement of testosterone levels in patients with presumed polycystic ovarian syndrome (PCOS) is important to exclude an androgen-producing tumor, which will produce markedly elevated total testosterone levels. In contrast, normal or high-normal levels of total testosterone are consistent with the diagnosis of PCOS. Because testosterone lowers SHBG levels, patients with normal total testosterone levels, but with clinical evidence of hyperandrogenism (hirsutism and/or acne), invariably have either increased free testosterone levels or increased sensitivity of the hair follicle and sebaceous glands.

Ultimately, steroids are metabolized mainly in the liver and to a lesser extent in the kidney and intestinal mucosa. Hydroxylation of estradiol results in production of estrone or catechol estrogens. These estrogens are then conjugated to glucuronides or sulfates to form water-soluble compounds for excretion in the urine. Accordingly, administration of certain pharmacologic steroid hormones may be contraindicated in those with active liver or renal disease.

Steroid hormones and peptide factors differ in their specific receptor-mediated actions, yet both eventually lead to DNA transcription and protein production in the target cell.

G-Protein Coupled Receptors

These are cell-membrane-associated receptors that bind peptide factors. These receptors consist of a hydrophilic extracellular domain, an intracellular domain, and a hydrophobic transmembrane domain that spans the cell membrane seven times. When bound to hormone, these receptors undergo a conformational change, activate intracellular signaling pathways, and, through a series of phosphorylation events, ultimately modulate transcription of multiple genes within the target cell.

The gonadotropin-releasing hormone receptor (GnRH-R) is a G-protein-coupled receptor that has been identified in the ovary, testes, hypothalamus, prostate, breast, and placenta (Yu, 2011). Although data are still preliminary, GnRH and its receptor may form an autocrine/paracrine regulatory network in reproductive tissues including the ovaries and placenta in addition to the classic neuroendocrine hypothalamic-pituitary system (Kim, 2007; Lee, 2010).

Both LH and hCG bind to the same G-protein-coupled receptor known as the LH/CG receptor. Relative to LH, hCG has a slightly higher affinity for the receptor and has a longer half-life. In contrast, FSH binds to a unique G-protein-coupled receptor located on the granulosa cell membrane.

Within the ovary, the LH/CG receptor is expressed on thecal cells, interstitial cells, and luteal cells. In the granulosa cells of preantral follicles, LH/CG receptor mRNA is nearly undetectable. However, in the differentiated granulosa cells found during follicular maturation, high levels of this receptor are observed. In addition to the ovary, LH/CG and FSH receptors have also been identified in endometrium, myometrium, and placenta (Stilley, 2014; Ziecik, 2007. The function of the receptor in these extraovarian tissues is poorly understood.

Steroid Hormone Receptors

Classification and Structure

The nuclear receptor superfamily consists of three receptor groups: (1) those that bind steroidal ligands, (2) those that have affinity for nonsteroidal ligands such as thyroid hormone, and (3) orphan receptors. By definition, orphan nuclear receptors do not have an identified ligand. These are believed to be constitutively active, that is, they exhibit basal or intrinsic activity. Despite their structural similarities, estrogens, progestins, androgens, mineralocorticoids, and glucocorticoids all interact with unique members of the nuclear hormone receptor family.

Free steroids diffuse into cells and combine with specific receptors (Fig. 15-7A). Members of this receptor superfamily exhibit a modular structure of distinct domains (Fig. 15-8). Each region contributes distinct activities required for full receptor function. In general, nuclear receptors have two regions that are critical for gene activation, termed activation function 1 (AF1) and activation function 2 (AF2). AF1 is located in the A/B domain and is usually ligand independent. AF2 is in the ligand-binding domain (E) and is often hormone-dependent. The highly conserved DNA-binding region (C) inserts into the DNA helix. Subsequently, steroid receptors enhance or repress gene transcription through interactions with specific DNA sequences, called hormone response elements, in the promoter region of target genes (Klinge, 2001).

Estrogen, Progesterone, and Androgen Receptors

As a a general rule, nuclear hormone receptors are localized to the cytoplasm. Following ligand binding, they then are translocated to the nucleus to exert their effects.

Two isoforms of estrogen receptors, ERα and ERβ, are encoded by separate genes (Kuiper, 1997). These receptors are differentially expressed in tissues and appear to serve distinct functions. For example, both ERα and ERβ are required for normal ovarian function. However, mice lacking ERα are anovulatory and accumulate cystic follicles, whereas ovaries missing ERβ are normal histologically despite impaired ovulation (Couse, 2000).

The progesterone receptor also exists in multiple isoforms. Encoded from a single gene, PRA and PRB are identical except for an additional 164 amino acids at the amino terminus (Conneely, 2002). Similar to estrogen receptors, the PR isoforms are not interchangeable. For example, PRA is required for normal ovarian and uterine functions but is expendable in the breast (Lydon, 1996). In contrast to the estrogen and progesterone receptor situation, only one form of the androgen receptor has been identified.

Nongenomic Actions of Steroids

Recent studies have introduced the concept that a subset of steroids, including estrogens and progestins, may alter cell function via nongenomic effects, that is, independent of the classic nuclear hormone receptors (see Fig. 15-7C). These nongenomic effects occur rapidly and may be mediated via cell-surface receptors (Kowalik, 2013; Revelli, 1998). Pharmacologic agents under development specifically target these nongenomic effects to allow more precise therapy for steroid-sensitive disorders.

Receptor Expression and Desensitization

Many influences alter cellular response to sex steroids and peptide factors. The number of receptors within a cell or on the cell membrane is critical to attain maximum hormonal response. Importantly, the number of receptors on a cell can be modified through gene transcription and receptor protein degradation.

Hormonally induced negative feedback of receptors is termed homologous downregulation or desensitization. Desensitization limits the duration of a hormonal response by decreasing the cell’s sensitivity to a constant, prolonged level of hormone. Within the reproductive system, desensitization is best understood for the GnRH receptor and is used clinically to produce a hypoestrogenic state. Pharmacologic agonists of GnRH, such as leuprolide acetate (Lupron), initially stimulate receptors on pituitary gonadotropes to cause a supraphysiologic release of both LH and FSH. Over a period of hours, agonists downregulate GnRH receptor sensitivity and number, thus preventing further GnRH stimulation. Correspondingly, decreased gonadotropin secretion leads to suppressed estrogen and progesterone levels 1 to 2 weeks after initial GnRH agonist administration.

Immunoassays

These tests use antibodies to detect most polypeptide, steroid, and thyroid hormones. They are sensitive and easily automated. Hormone concentration is usually reported as international units per volume rather than mass per volume (Table 15-2). When interpreting immunoassays, several concepts must be understood. These include reference standards, the “hook effect,” normal ranges, and supplementary hormone levels.

First, to minimize assay-to-assay variability, a reference material is needed to standardize assays. Reference standards serve as anchors that can provide comparability across time and methods. Such reference preparations are produced by the World Health Organization (WHO) and the National Institutes of Health (NIH). More than 20 assay standards are available to measure LH, FSH, prolactin (PRL), and hCG. Thus, knowing which reference standard is used by a specific assay is essential, as results may differ significantly. Clinically, this can become an issue in patients with possible ectopic pregnancies when serial β-hCG levels are obtained at different health care facilities.

Second, the “hook effect” can alter immunoassay result interpretation. With this effect, significantly elevated hormone levels saturate the assay’s targeting antibody and create a falsely low reading. Moreover, the amount of hormone present in a sample does not necessarily correlate with the biological activity of that hormone. For example, PRL exists in multiple isoforms, many of which are immunologically detectable but not biologically active. Similarly, varying glycosylation patterns of gonadotropins at different times during the reproductive life span are believed to alter their biologic activity.

Another caveat is a result that lies in the “normal range.” For many hormones, a stated normal range is often broad. As such, the hormone level of an individual may double, but remain within the normal range even though the result is actually abnormal for that individual.

Last, the addition of other hormone levels may be necessary to define the significance of a result. In the context of the pituitary gland and its target endocrine glands, it may be adequate to measure the pituitary hormone alone. For example, high levels of circulating gonadotropins are almost invariably due to ovarian failure and loss of negative feedback. This is because pituitary overproduction of functional dimer is rare. Conversely, low gonadotropin levels can be attributed confidently to hypothalamic-pituitary dysfunction. Thus, the measurement of ovarian-derived products such as estrogen may be helpful to confirm the diagnosis but are not critical.

In other clinical scenarios, the measurement of both pituitary and target hormone levels may be indicated. For example, in many laboratories, an abnormal TSH value will lead to “reflex,” that is, automatic testing for thyroid hormone levels. Low levels of both a stimulating-hormone and target hormone indicate an abnormality in either hypothalamic or pituitary function. High levels of a target-gland hormone coupled with low levels of its stimulating pituitary hormone suggest autonomous secretion by the target organ such as occurs in the hyperthyroidism of Graves disease.

Stimulation Tests

These tests may be useful when hypofunction of an endocrine organ is suspected. These tests use an endogenous stimulating hormone to assess the reserve capacity of the tissue of interest. The trophic hormone used may be a hypothalamic releasing factor such as GnRH or thyrotropin-releasing hormone (TRH). Alternatively, a substitute pituitary hormone may be used, such as hCG as a substitute for LH or leuprolide acetate for GnRH. The ability of the target gland to respond is measured by an increase in the appropriate hormone’s plasma level. One example, the leuprolide stimulation test, may be used to evaluate abnormal pubertal development and is described in Chapter 14. Leuprolide substitutes for GnRH because clinical-grade GnRH is often unavailable (Rosenfield, 2013).

Suppression Tests

These tests may be performed when endocrine hyperfunction is suspected. For example, a dexamethasone suppression test may be given to a patient with suspected hypercortisolism (Cushing disease or syndrome). Described in full in Chapter 17, this test gauges the ability of dexamethasone to inhibit adrenocorticotropic hormone (ACTH) secretion and thus cortisol production by the adrenal. The failure of glucocorticoid treatment to suppress cortisol production would be consistent with primary hyperadrenalism.

Anatomy

The hypothalamus consists of nuclei located at the base of the brain, just superior to the optic chiasm. Neurons within the hypothalamus form synaptic connections with other neurons throughout the central nervous system (CNS). A subset of the hypothalamic neurons within the arcuate, ventromedial, and paraventricular nuclei project to the median eminence. In the median eminence, a dense network of capillaries arises from the superior hypophyseal arteries. These capillaries drain into portal vessels that traverse the pituitary stalk and then form a capillary network within the anterior pituitary gland (adenohypophysis). The primary direction of this hypophyseal portal system is from hypothalamus to pituitary. However, retrograde flow also exists. This creates an ultrashort feedback loop between the pituitary gland and hypothalamic neurons. The hypothalamus is thus a critical locus for integration of information from the environment, nervous system, and other organ systems.

The anterior pituitary gland consists of endocrine cells and is derived from an invagination of Rathke pouch in the roof of the embryonic oral cavity. In contrast, the posterior pituitary gland (neurohypophysis) is neural tissue and consists of the axon terminals of magnocellular neurons arising in the supraoptic and paraventricular nuclei of the hypothalamus (Fig. 15-9).

Hypothalamic Neuroendocrinology

The list of known neurotransmitters continues to expand as does our understanding of their anatomic distribution, mode of regulation, and mechanism of action. Neurotransmitters can be classified as: (1) biogenic amines (dopamine, epinephrine, norepinephrine, serotonin, histamine), (2) neuropeptides, (3) acetylcholine, (4) excitatory amino neurotransmitters (glutamate, glycine, aspartic acid), (5) the inhibitory amino acid gamma-aminobutyric acid (GABA), (6) gaseous transmitters (nitric oxide, carbon monoxide), and (7) miscellaneous factors (cytokines, growth factors).

The most significant neurotransmitters in reproductive neuroendocrinology are the three monoamines: dopamine, norepinephrine, and serotonin. Clinically important neuropeptides within the reproductive axis include the endogenous opiates, kisspeptin, neuropeptide Y, galanin, and pituitary adenylate cyclase-activating peptide.

Endogenous Opiates

Central opioidergic neurons are important mediators of hypothalamic-pituitary function. Depending on the precursor peptide from which they are derived, these neuropeptides can be categorized into three classes: endorphins, enkephalins, and dynorphins. Of these, endorphins (endogenous morphines) are cleavage products of the proopiomelanocortin POMC gene, which also yields ACTH and α-melanocyte stimulating hormone (α-MSH) (Taylor, 1997). The endorphins serve a wide range of physiologic functions that include regulation of temperature, cardiovascular and respiratory systems, pain perception, mood, and reproduction.

Proopiomelanocortin is produced in highest concentration in the anterior pituitary gland but is also expressed in the brain, sympathetic nervous system, gonads, placenta, gastrointestinal tract, and lungs. The primary peptide synthesized from this pathway depends on the tissue source. For example, the predominant products in the brain are the opiates, whereas pituitary biosynthesis results principally in ACTH production.

Opioids in the brain play a central role in menstrual cyclicity by tonically suppressing the hypothalamic release of GnRH (Funabashi, 1994). Estrogen promotes endorphin secretion, and this is increased further with the addition of progesterone (Cetel, 1985). Thus, endorphin levels increase during the follicular phase, peak during the luteal phase, and drop markedly during menses. This pattern suggests that both opioid tone and progesterone decrease GnRH pulse frequency in the luteal phase, thus stimulating FSH secretion. For reasons that are not fully understood, opioid suppression of GnRH is relieved at the time of ovulation (King, 1984). In addition, functional hypothalamic amenorrhea due to eating disorders, intensive exercise, and stress is correlated with an increase in endogenous opiate concentrations (Chap. 16).

Other Hypothalamic Neuropeptides

Hypothalamic kisspeptin neurons play a critical role in sexual differentiation, puberty initiation, and adult reproductive function. These neurons are part of the KNDy neuronal system, named for the coexpression of kisspeptin with neurokinin B and dynorphin. This system likely provides an important link between energy homeostasis and reproductive function. This link may stem in part from the action of the adipose-derived factor leptin, which regulates kisspeptin expression (Chehab, 2014).

Kisspeptin neurons send processes to GnRH neurons, allowing direct control of GnRH secretion. Interestingly, one group of kisspeptin neurons may mediate negative steroid feedback, whereas another is responsible for the positive feedback observed before ovulation (Lehman, 2010; Millar, 2014; Skorupskaite, 2014).

Other neurotransmitters, neuropeptide Y (NPY) and galanin, are expressed by neurons located throughout the hypothalamus and project to kisspeptin neurons, to GnRH neurons, and to other areas of the CNS that have roles in reproductive function. NPY and galanin secretion varies in response to changes in energy level as seen in anorexia and obesity. Both of these neuropeptides alter GnRH pulsatility and potentiate GnRH-induced gonadotrope secretion (Lawrence, 2011; Peters, 2009).

Pituitary adenylate cyclase-activating peptide (PACAP) is a hypothalamic peptide secreted into the pituitary portal system. It binds to receptors on anterior pituitary cells and stimulates hormone secretion including gonadotropin secretion, albeit more weakly than GnRH. Gonadotropes themselves also secrete PACAP, suggesting an autocrine/paracrine role for this hormone within the pituitary. PACAP modulates GnRH-receptor expression and, conversely, GnRH alters PACAP-receptor expression on the gonadotrope cell surface. Furthermore, pituitary PACAP gene expression is markedly increased by GnRH (Halvorson, 2014). Thus, these two important neuropeptides are functionally linked at the level of the anterior pituitary.

Anterior Pituitary Hormones

The anterior pituitary gland contains five hormone-producing cell types and their products. These include: (1) gonadotropes (which produce LH and FSH), (2) lactotropes (PRL), (3) somatotropes (GH), (4) thyrotropes (TSH), and (5) adrenocorticotropes (ACTH). Of these, gonadotropes comprise approximately 10 to 15 percent of all hormonally active cells in the anterior pituitary (Childs, 1983).

With the exception of PRL, which is under tonic inhibition, pituitary hormones are stimulated by hypothalamic neuroendocrine secretion. Both of the gonadotropins, LH and FSH, are regulated by a single releasing peptide, GnRH, which acts on the anterior pituitary’s gonadotrope subpopulation. Most gonadotropes contain secretory granules that contain both LH and FSH, although a significant number of cells are monohormonal, that is, secrete only LH or only FSH.

Of the other pituitary-releasing hormones, corticotropin-releasing hormone stimulates biosynthesis and secretion of ACTH by the pituitary adrenocorticotropes. Thyrotropin-releasing hormone increases thyrotrope secretion of TSH, also known as thyrotropin. Various hypothalamic secretagogues regulate expression of somatotrope-derived growth hormone. Last, PRL expression is primarily under inhibitory regulation by dopamine. As a consequence of these regulatory mechanisms, damage to the pituitary stalk results in hypopituitarism for LH, FSH, GH, ACTH, and TSH, but an associated increase in PRL secretion.

Hypothalamic Releasing Peptides

These peptides have characteristics that are important for both their biologic function and clinical use. First, they are small peptides with short half-lives of a few minutes due to their rapid degradation. Second, hypothalamic releasing peptides are released in minute quantities and are highly diluted in the peripheral circulation. Therefore, biologically active concentrations of these factors are locally restricted to the anterior pituitary gland. Clinically, the extremely low concentrations of these hormones render them essentially undetectable in serum. Thus, levels of their corresponding pituitary factors are measured as surrogate markers.

Gonadotropin-releasing Hormone

GnRH is a decapeptide with a half-life of less than 10 minutes. Amino acid modifications generate receptor antagonists or agonists with a prolonged half-life (Fig. 15-10) (Padula, 2005). Pulsatile GnRH input is required for activation and maintenance of GnRH receptors. This characteristic is exploited clinically by administering long-acting GnRH agonists to treat steroid-dependent conditions such as endometriosis, leiomyomas, precocious puberty, breast cancer, and prostate cancer. These agonists compete with endogenous pulsatile GnRH at the receptor, depressing gonadotropin secretion and thereby decreasing serum ovarian sex steroid levels.

Humans express two forms of GnRH termed GnRH I and GnRH II (Cheng, 2005). By convention, GnRH I is the classically described hypothalamic GnRH. The GnRH II peptide has been identified in peripheral tissues and differs in receptor activation (Neill, 2002). Further research is needed to determine the overlapping and divergent functions of these two forms.

Migration of the Gonadotropin-releasing Hormone Neurons

Many hypothalamic neurons arise within the CNS, but GnRH-containing neurons have a unique embryologic origin. Progenitor GnRH neurons originate in the medial olfactory placode and migrate along the vomeronasal nerve into the hypothalamus (Fig. 16-5). A series of soluble factors regulate GnRH neuronal migration at specific locations along their migratory route. These factors include secreted signaling molecules such as GABA, adhesion molecules, and growth factors (Wierman, 2011). Failure of normal migration may stem from various genetic defects in these signaling molecules and can lead to Kallmann syndrome, which is discussed in Chapter 16, and other forms of hypogonadotropic hypogonadism.

GnRH cell bodies are primarily located within the arcuate nucleus. From these neuronal cell bodies, GnRH is axonally transported along the tuberoinfundibular tract to the median eminence. GnRH is then secreted into the portal system that drains directly to the anterior pituitary gland and stimulates gonadotropin biosynthesis and secretion. The number of GnRH neurons in the adult is strikingly low, with only a few thousand cells dispersed within the arcuate nucleus.

The olfactory origin of GnRH neurons and nasal epithelial cells suggest a link between reproduction and olfactory signals. Compounds released by one individual that affect other members of the same species are known as pheromones. Pheromones obtained from the axillary secretions of women in the late follicular phase accelerate the LH surge and shorten menstrual cycles of women exposed to these chemicals. Secretions from women in the luteal phase have the opposite effects. Thus, pheromones may be one mechanism by which women who are together frequently often exhibit synchronous menstrual cycles (Stern, 1998).

A subset of GnRH neurons sends projections into other areas of the CNS, including the limbic system. These projections are not required for gonadotropin secretion, but they may play a role in modulation of reproductive behavior (Nakai, 1978; Silverman, 1987).

Pulsatile Gonadotropin-releasing Hormone Secretion

In elegant experiments, Knobil (1974) demonstrated that pulsatile delivery of GnRH to the pituitary gonadotropes is required to achieve sustained gonadotropin secretion. As shown in Figure 15-11, continuous infusion with GnRH rapidly decreases both LH and FSH secretion, an effect that is easily reversed with a return to pulsatile stimulation.

Compared with the luteal phase, follicular phase GnRH pulsatility is characterized by increased frequency and decreased amplitude. Higher pulse frequency preferentially stimulates LH, whereas lower frequency favors FSH secretion (Thompson, 2014). Therefore, changes in GnRH pulse frequency affect the absolute levels and the ratio of LH to FSH release.

Pulsatile activity is currently believed to be an intrinsic property of GnRH neurons. Other hormones and neurotransmitters provide modulatory effects (Clayton, 1981; Yen, 1985). In animal models, estrogen increases GnRH pulse frequency and therefore leads to an increase in LH levels relative to FSH levels. In contrast, progesterone decreases GnRH pulsatility. The increase in progesterone during the luteal phase may explain the preferential stimulation of FSH observed toward the end of this phase. This rise in FSH is critical for the initiation of follicular recruitment.

Other Hypothalamic-Pituitary Axes

Dopamine and Prolactin

In contrast to the other anterior pituitary hormones, PRL release is primarily regulated via inhibition, specifically by dopamine. These dopamine-containing fibers arise chiefly in the hypothalamic arcuate nucleus and project to the median eminence, where dopamine enters the portal vessels (Table 15-3). Prolactin-releasing factors, although less potent, include TRH, vasopressin, vasoactive intestinal peptide (VIP), endogenous opioids, and acetylcholine.

There are five forms of the dopamine receptor divided into two groups, D₁ and D₂. Cells in the anterior pituitary gland primarily express the D₂ subtypes. The medical treatment of prolactinomas has been improved in terms of both effectiveness and patient tolerance by the development of the D₂-specific ligands. For example, the dopamine agonist cabergoline is a D₂-specific ligand, whereas bromocriptine is nonspecific.

Thyrotropin-releasing Hormone

As indicated by its name, thyrotropin-releasing hormone stimulates secretion of thyroid-stimulating hormone from the anterior pituitary gland’s thyrotrope subpopulation. Of note, TRH is also a potent prolactin-releasing factor and results in a clinical link between hypothyroidism and secondary hyperprolactinemia (Messini, 2010).

TSH binds to specific receptors on the plasma cell membrane of thyroid gland cells. This stimulates thyroid hormone biosynthesis. Thyroid hormone exerts negative feedback on TRH- and TSH-releasing cells.

Corticotropin-releasing Hormone

This is the primary hypothalamic factor that stimulates synthesis and secretion of ACTH. Corticotropin-releasing hormone (CRH) is distributed in multiple locations within the hypothalamus and other CNS areas. Release of CRH is stimulated by catecholaminergic input from other brain pathways and inhibited by endogenous opioids.

Corticotropin-releasing hormone binds to CRH receptors in the anterior pituitary to stimulate ACTH biosynthesis and secretion. In turn, ACTH stimulates glucocorticoid production by the adrenal’s zona fasciculata and androgen production by its zona reticularis. CRH secretion is under negative-feedback regulation by circulating cortisol produced in the adrenal gland. In contrast, mineralocorticoid production by the zona glomerulosa is primarily regulated by the renin-angiotensin system. As a result, abnormalities in the CRH–ACTH pathway do not result in electrolyte disturbances.

Central CRH pathways are believed to mediate many stress responses (Kalantaridou, 2004). Clinically, in women with hypothalamic amenorrhea, CRH levels have been found to be elevated. Increased levels of CRH inhibit hypothalamic GnRH secretion by direct action and by augmenting central opioid concentrations (Fig. 16-6). This functional pathway may explain the association between hypercortisolism and menstrual abnormalities.

Growth Hormone–releasing Hormone

Growth hormone secretion by pituitary somatotropes is stimulated by hypothalamic growth hormone-releasing hormone (GHRH) and inhibited by somatostatin. GHRH is primarily secreted by the hypothalamus, but small quantities are released by placental and immune cells. In contrast, somatostatin is widely distributed in the CNS and in the placenta, pancreas, and gastrointestinal tract.

As with GnRH, GHRH depends on pulsatile secretion to exert a physiologic effect. Exercise, stress, sleep, and hypoglycemia stimulate GH release, whereas free fatty acids and other factors related to adiposity blunt GH release. Estrogen, testosterone, and thyroid hormone also play a role in increased GH secretion.

GH stimulates skeletal and muscle growth, regulates lipolysis, and promotes the cellular uptake of amino acids. This hormone induces insulin resistance, and thus GH excess may be associated with new-onset diabetes mellitus. Most of the growth effects of GH are mediated via the insulin-like growth factors, IGF-I and IGF-II. These growth factors are produced in high quantities in the liver. Many of the target tissues in which they exert local effects also synthesize IGFs. Within the ovary, IGF-I and IGF-II stimulate granulosa cell proliferation and steroidogenesis during folliculogenesis (Silva, 2009). IGFs also suppress GH secretion through negative feedback mechanisms.

Posterior Pituitary Peptides

Neurons projecting to the posterior pituitary synthesize and secrete the nine-amino-acid cyclic peptides oxytocin and arginine vasopressin. Precursors for these peptides are produced in the neuronal cell body and transported down the axon in secretory granules. During transport, precursors are cleaved into mature peptides and a carrier protein—neurophysin (Verbalis, 1983). Activation of these neurons generates an axon potential that results in calcium influx and secretion of granule contents into the perivascular space. These secreted peptides then enter adjacent blood vessels for transport throughout the peripheral circulation.

Of these two peptides, oxytocin has significant roles in reproduction, specifically parturition and lactation (Kiss, 2005). The role of oxytocin in labor initiation is disputed as serum oxytocin levels are constant until the expulsive portion of labor (Blanks, 2003). Nevertheless, an increase in myometrial and decidual oxytocin-receptor expression has been noted near term, primarily due to an increase in estrogen levels.

Once labor is initiated, oxytocin is the primary mediator of myometrial contractility. Cervical and vaginal stimulation results in an acute release of oxytocin from the posterior pituitary in a process known as the Ferguson reflex. Clinically, oxytocin’s ability to induce uterine contractions is exploited to induce or augment labor.

Vaginal distention, such as occurs with coitus, also increases oxytocin release. Based on this observation, oxytocin may be responsible for the rhythmic uterine and tubal contractions that aid sperm delivery to the oocyte. Oxytocin may also play a role in orgasm and ejaculation.

During lactation, PRL is critical for milk production in breast alveoli. The glandular cells of the alveoli are surrounded by a mesh of myoepithelial cells. Suckling triggers nerve impulses from mechanoreceptors in the nipple and areola that increase hypothalamic neuronal activity. Subsequent oxytocin release prompts the myoepithelial cells to contract and thereby express milk from the alveoli into the ducts and sinuses (Crowley, 1992). Other conditioned stimuli, such as the sight, sound, or smell of a baby or sexual arousal, will have similar effects.

Oxytocin expression has also been detected in the anterior pituitary, placenta, fallopian tubes, and gonads, with high expression in the corpus luteum (Williams, 1990). Its function in these tissues is unknown.

The “typical” menstrual cycle is 28 ± 7 days with menstrual flow lasting 4 ± 2 days and blood loss averaging 20 to 60 mL. By convention, the first day of vaginal bleeding is considered day 1 of the menstrual cycle. Menstrual cycle intervals vary among women and often for an individual woman at different times during her reproductive life. In a study of more than 2700 women, menstrual cycle intervals were found to be most irregular in the 2 years following menarche and the 3 years preceding menopause (Treloar, 1967). Specifically, a trend toward shorter intervals followed by interval lengthening is common during the menopausal transition. The menstrual cycle is least variable between the ages of 20 and 40 years.

When viewed from a perspective of ovarian function, the menstrual cycle can be defined as a preovulatory follicular phase and postovulatory luteal phase (Fig. 15-12). Corresponding phases in the endometrium are termed the proliferative and secretory phases (Table 15-4). For most women, the luteal phase of the menstrual cycle is stable, lasting 13 to 14 days. Thus, variations in normal cycle length generally result from variable duration of the follicular phase (Ferin, 1974).

The Ovary

Ovarian Morphology

The adult human ovary is oval with a length of 2 to 5 cm, a width of 1.5 to 3 cm, and a thickness of 0.5 to 1.5 cm. During the reproductive years, the ovary weighs between 5 and 10 g. It is composed of three parts: an outer cortical region, which contains both the germinal epithelium and the follicles; a medullary region, which consists of connective tissue, myoid-like contractile cells, and interstitial cells; and a hilum, which contains blood vessels, lymphatics, and nerves that enter the ovary (Fig. 15-13).

Ovaries have two interrelated functions: the generation of mature oocytes and the production of steroid and peptide hormones that create an environment in which fertilization and subsequent implantation in the endometrium can occur. Within each cycle, endocrine functions of the ovary correlate closely to the morphologic appearance and disappearance of follicles and corpus luteum.

Embryology of the Ovary

The ovary develops from three major cellular sources: (1) primordial germ cells, which arise from the endoderm of the yolk sac and differentiate into the primary oogonia; (2) coelomic epithelial cells, which develop into granulosa cells; and (3) mesenchymal cells from the gonadal ridge, which become the ovarian stroma. Additional information regarding gonadal differentiation is found in Chapter 18.

Primordial germ cells can be seen in the yolk sac as early as the third week of gestation (Gosden, 2013). These cells begin their migration into the gonadal ridge during the sixth week of gestation and generate primary sex cords. The ovary and testes are indistinguishable by histologic criteria until approximately 10 to 11 weeks of fetal life.

After the primordial cells reach the gonad, they continue to multiply through successive mitotic divisions. Starting at 12 weeks’ gestation, a subset of oogonia will enter meiosis to become primary oocytes. Primary oocytes are surrounded by a single layer of flattened granulosa cells, creating a primordial follicle.

Oocyte Loss with Aging

All oogonia either develop into primary oocytes or become atretic. Classical teaching states that additional oocytes cannot be generated postnatally. This differs markedly from the male, in whom sperm are produced continuously throughout adulthood. Exciting recent studies suggest that ovarian stem cells may be able to generate mature oocytes, providing hope for significant advances in female fertility preservation. Currently, these results remain preliminary and somewhat controversial (Notarianni, 2011; Virant-Klun, 2015).

The maximal number of oogonia is achieved at the 20th week of gestation, at which time 6 to 7 million oogonia are present in the ovary (Baker, 1963). Approximately 1 to 2 million oogonia are present at birth. Fewer than 400,000 are present at the initiation of puberty, of which fewer than 500 are destined to ovulate. Therefore, most germ cells are lost through atresia (Hsueh, 1996). Follicular atresia is thought not to be a passive, necrotic process, but rather a precisely controlled active process, namely apoptosis, which is under hormonal control. Apoptosis begins in utero and continues throughout reproductive life.

Oocyte Maturation

As previously mentioned, primary oogonia enter meiosis in utero to become primary oocytes. These oocytes are arrested in development at prophase during the first meiotic division. Meiotic progression resumes each month in a cohort of follicles. Meiosis I is completed in the oocyte destined for ovulation in response to the LH surge. Meiosis II begins, and the process is arrested, this time in the second meiotic metaphase. Meiosis II is completed only if the ovum is fertilized (Fig. 15-14).

Normal oocyte development requires cytoplasmic modifications in addition to meiotic maturation. Changes in microtubules and actin filaments enable rearrangement of cellular organelles to allow for successful polar body extrusion and fertilization (Coticchio, 2015). The cumulus cells modulate maturation both by cell-to-cell contact via gap junctions and by secretion of paracrine factors. Our growing understanding of these factors and processes is improving in vitro maturation protocols to aid fertility preservation and infertility treatments.

Stromal Cells

Ovarian stroma contains interstitial cells, contractile cells, and connective tissue cells. These last cells provide structural support to the ovary. The group of interstitial cells that surround a developing follicle differentiates into theca cells. Under gonadotropin stimulation, these cells increase in size and develop lipid stores, characteristic of steroid-producing cells (Saxena, 1972).

Another group of interstitial cells in the ovarian hilum are known as hilus cells. These closely resemble testicular Leydig cells, and hyperplasia or neoplastic changes in hilar cells may result in excess testosterone secretion and virilization. The normal role of these cells is unknown, but their intimate association with blood vessels and neurons suggest that they may convey systemic signals to the remainder of the ovary.

Ovarian Hormone Production

The normal functioning ovary synthesizes and secretes estrogens, androgens, and progesterone in a precisely controlled pattern determined, in part, by the pituitary gonadotropins, FSH and LH. The most important secretory products of ovarian steroid biosynthesis are progesterone and estradiol. However, the ovary also secretes quantities of estrone, androstenedione, testosterone, and 17α-hydroxyprogesterone. Sex steroid hormones play an important role in the menstrual cycle by preparing the uterus for implantation of a fertilized ovum. If implantation does not occur, ovarian steroidogenesis declines, the endometrium degenerates, and menstruation ensues.

Two-Cell Theory

Ovarian estrogen biosynthesis requires the combined action of two gonadotropins (LH and FSH) on two cell types (theca and granulosa cells). This concept is known as the two-cell theory of ovarian steroidogenesis (Fig. 15-15) (Peters, 1980). Until the late antral stage of follicular development, LH-receptor expression is limited to the thecal compartment, and FSH-receptor expression is limited to the granulosa cells.

Theca cells express all of the enzymes needed to produce androstenedione. This includes high levels of CYP17 gene expression, whose enzyme product catalyzes 17-hydroxylation. This is the rate-limiting step in the conversion of progesterones to androgens (Sasano, 1989). This enzyme is absent in the granulosa cells, so they are incapable of producing the androgenic precursors needed to produce estrogens. Granulosa cells therefore rely on the theca cells. Namely, in response to LH stimulation, theca cells synthesize the androgens androstenedione and testosterone. These androgens are secreted into the extracellular fluid and diffuse across the basement membrane to the granulosa cells to provide precursors for estrogen production. In contrast to theca cells, granulosa cells have high levels of aromatase activity in response to FSH stimulation. Thus, these cells efficiently convert androgens to estrogens, primarily the potent estrogen estradiol. In sum, ovarian steroidogenesis is dependent on the effects of LH and FSH acting independently on the theca cells and granulosa cells, respectively.

Steroidogenesis across the Life Span

Circulating levels of the gonadotropins LH and FSH vary markedly at different ages of a woman’s life. In utero, the fetal human ovary has the capacity to produce estrogens by 8 weeks’ gestation. However, a minimal amount of steroid is actually synthesized during fetal development (Miller, 1988). During the second trimester, the plasma levels of gonadotropins rise to levels similar to those observed in menopause (Temeli, 1985). The fetal hypothalamic-pituitary axis continues to mature during this time, becoming more sensitive to the high circulating levels of estrogen and progesterone secreted by the placenta (Kaplan, 1976). Prior to birth and in response to these high steroid levels, fetal gonadotropins fall to low levels.

After delivery, gonadotropin levels in the neonate rise abruptly due to separation from the placenta and subsequent freedom from placental steroid inhibition (Winter, 1976). The elevated gonadotropin levels persist for the first few months of life and then decline to low levels in early childhood (Schmidt, 2000). There may be multiple etiologies for the low gonadotropin levels during this period of life. The hypothalamic-pituitary axis has increased sensitivity to negative feedback, even by the low circulating levels of gonadal steroids at this stage. There may be a direct CNS role in maintaining low gonadotropin levels. In support of this mechanism, low levels of LH and FSH are found even in children with gonadal dysgenesis who lack negative feedback by gonadal steroids.

With puberty, one early sign is a sleep-associated increase in LH secretion (Fig. 15-16). Over time, increased gonadotropin secretion is noted throughout the day. An increased FSH to LH ratio is typical in the premenarchal girl and postmenopausal woman. During the reproductive years, LH exceeds FSH levels, inverting this ratio. Increased gonadotropin levels stimulate ovarian estradiol production. The rise in estrogen levels prompts the growth spurt, maturation of the female internal and external genitalia, and development of a female habitus including pubertal breast enlargement, which is termed thelarche. Activation of the pituitary-adrenal axis results in an increase in adrenal androgen production and the associated pubertal development of axillary and pubic hair, termed adrenarche or pubarche. Increased gonadotropin levels ultimately lead to ovulation and subsequent menses. The first menstrual period defines menarche. This developmental process takes approximately 3 to 4 years and is discussed further in Chapter 14.

Following menopause, the postmenopausal ovary contains only a few follicles. As a result, plasma estrogen and inhibin levels decrease markedly after cessation of ovulatory cycles. Through loss of this negative feedback, LH and FSH levels are strikingly elevated. Elevated LH levels can stimulate production of C-19 steroids (mainly androstenedione) in ovarian stromal cells. This ovarian-derived androstenedione and adrenal androgens can be converted by peripheral tissues to estrone, the principal serum estrogen in the postmenopausal women. The major site for the conversion of androstenedione to estrone is adipose tissue. Peripheral conversion of circulating androstenedione to estrone is directly correlated to body weight. For a given body weight, conversion is higher in postmenopausal women than in premenopausal women. These low circulating estrogen levels are usually not adequate to protect against bone loss.

Gonadal Peptides and the Menstrual Cycle

Of the multiple gonadal peptides, three—inhibin, activin, and follistatin—modulate gonadotrope activity in addition to effects within the ovary (de Kretser, 2002). As suggested by their names, inhibin decreases and activin stimulates gonadotrope function. Follistatin suppresses FSHβ gene expression, most likely by binding to and thereby preventing the interaction of activin with its receptor (Xia, 2009).

Inhibin and activin are closely related peptides. Inhibin consists of an α-subunit (unrelated to the LH and FSH glycoprotein α-subunit) linked by a disulfide bridge to one of two highly homologous β-subunits to form inhibin A (αβ_A) or inhibin B (αβ_B). Activin is composed of homodimers (β_Aβ_A, β_Bβ_B) or heterodimers (β_Aβ_B) of the same β-subunits as inhibin (Bilezikjian, 2012). In contrast, follistatin is structurally unrelated to either inhibin or activin.

Although originally isolated from follicular fluid, these “gonadal” peptides are expressed in the pituitary, ovary, testes, and placenta and in the brain, adrenal, liver, kidney, and bone marrow to provide diverse tissue-specific functions (Muttukrishna, 2004). Activin and follistatin most likely act as autocrine/paracrine factors in the tissues in which they are expressed, including the ovary.

In contrast, ovarian-derived inhibins circulate in significant concentrations and are believed to be critical for negative feedback of gonadotropin gene expression. Specifically, during the early follicular phase, FSH stimulates the secretion of inhibin B by the granulosa cells (Fig. 15-17) (Buckler, 1989). However, increasing levels of circulating inhibin B blunt later FSH secretion in the follicular phase. During the luteal phase, regulation of inhibin production comes under the control of LH and switches from inhibin B to inhibin A (McLachlan, 1989). Inhibin B levels peak with the LH surge, whereas inhibin A levels peak a few days later, in the midluteal phase. All inhibin levels decline with the loss of luteal function and remain low during the luteal-follicular transition and early follicular phase. The inverse relationship between circulating inhibin levels and FSH secretion is consistent with a negative-feedback role for inhibin in regulating FSH secretion.

Distinct from these three peptides, insulin-like growth factors also mediate ovarian function. Only IGF-II is involved in primordial follicle development, but both IGF-I and IGF-II stimulate growth of secondary follicles. Gonadotropins stimulate IGF-II production in theca cells, granulosa cells, and luteinized granulosa cells. Receptors for IGF are expressed on the theca and granulosa cells, supporting an autocrine/paracrine action in the follicle. FSH also mediates expression of IGF-binding proteins. This system, although complex, allows additional fine-tuning of intrafollicular activity (Silva, 2009).

Follicular Development

Follicle Stages

Development begins with primordial follicles that were generated during fetal life (see Fig. 15-14). These follicles consist of an oocyte arrested in the first meiotic division surrounded by a single layer of flattened granulosa cells. The follicles are separated from the stroma by a thin basement membrane. Preovulatory follicles are avascular. As such, they are critically dependent on diffusion and on the later development of gap junctions for obtaining nutrients and clearing metabolic waste. Diffusion also allows passage of steroid precursors from the theca to the granulosa cell layer.

In the primary follicle stage, the granulosa cells of developing follicles become cuboidal and increase in number to form a pseudostratified layer. Intercellular gap junctions develop between adjacent granulosa cells and between granulosa cells and the developing oocyte (Albertini, 1974). These connections allow the passage of nutrients, ions, and regulatory factors between cells. Gap junctions also allow cells without gonadotropin receptors to receive signals from cells with receptor expression. As a result, hormone-mediated effects can be transmitted throughout the follicle.

During this stage, the oocyte begins to secrete products to form an acellular coat known as the zona pellucida. The human zona pellucida contains at least three proteins, named ZP1, ZP2, and ZP3. In current physiologic models, receptors on the acrosome head of the sperm recognize ZP3. This interaction releases acrosomal contents that permit penetration of the zona pellucida and ovum fertilization. Enzymes released from the acrosome induce alterations in ZP2 that result in hardening of the coat. This prevents fertilization of the oocyte by more than one sperm (Gupta, 2015).

Development of a secondary, or preantral, follicle includes final growth of the oocyte and a further increase in granulosa cell number. The stroma around the granulosa cell layer differentiates into the theca interna and the theca externa (Eppig, 1979).

Tertiary follicles, also called antral follicles, form from ongoing development in selected oocytes. In these, follicular fluid collects between the granulosa cells, ultimately producing a fluid-filled space known as the antrum. Granulosa cells in the antral follicle are histologically and functionally divided into two groups. The granulosa cells surrounding the oocyte form the cumulus oophorus, whereas the granulosa cells surrounding the antrum are known as mural granulosa cells. Antral fluid consists of a plasma filtrate and factors secreted by the granulosa cells. These locally produced factors, which include estrogen and growth factors, are present in substantially higher concentrations in follicular fluid than in the circulation and are likely critical for successful follicular maturation (Asimakopoulos, 2006; Silva, 2009). Further accumulation of antral fluid results in a rapid increase in follicular size and development of a preovulatory, or graafian, follicle (Hennet, 2012).

During this process, early stages of development (up to the secondary follicle) do not require gonadotropin stimulation and thus are said to be “gonadotropin-independent.” Final follicular maturation requires adequate amounts of circulating LH and FSH and is therefore said to be “gonadotropin-dependent” (Butt, 1970). Of note, data suggest that progression from gonadotropin-independent to dependent stages is not as discrete as previously believed.

Concept of a Selection Window

Follicular development is a multistep process, which proceeds over at least 3 months and culminates in ovulation from a single follicle. Each month, a group of follicles known as a cohort begins a phase of semisynchronous growth. The size of this cohort appears to be proportional to the number of inactive primordial follicles within the ovaries and has been estimated at 3 to 11 follicles per ovary in young women (Hodgen, 1982; Pache, 1990). Importantly, the ovulatory follicle is recruited from a cohort that began development two to three cycles prior to the ovulatory cycle. During this time, most follicles will die as they will not be at an appropriate stage of development during the selection window.

During the luteal-follicular transition, a small increase in FSH levels is responsible for selection of the single dominant follicle that will ultimately ovulate (Schipper, 1998). As previously described, theca cells produce androgens, which are converted to estrogens by the granulosa cells. Estrogen levels increase with increased follicular size, enhance the effects of FSH on granulosa cells, and create a feed-forward action on follicles that produce estrogens.

Intrafollicular levels of the insulin-like growth factors are believed to synergize with FSH to help select the dominant follicle (Son, 2011). Additional studies have also demonstrated elevated levels of vascular endothelial growth factor (VEGF) around the follicle that will be selected. This follicle would presumably be exposed to higher levels of circulating factors such as FSH (Ravindranath, 1992).

Granulosa cells also produce inhibin B, which passes from the follicle into the plasma and specifically inhibits the release of FSH, but not of LH, by the anterior pituitary. The combined production of estradiol and inhibin B by the dominant follicle results in the decline of follicular-phase FSH levels and may be responsible at least in part for the failure of the other follicles to reach preovulatory status during any one cycle.

Estrogen-dominant Microenvironment

Ongoing follicular maturation requires the successful conversion from an “androgen-dominant” microenvironment to an “estrogen-dominant” one. At low concentrations, androgens stimulate aromatization and contribute to estrogen production. However, intrafollicular androgen levels will rise if aromatization in the granulosa cells lags behind androgen production by the thecal layer. At higher concentrations, androgens are converted to the more potent 5α-androgens, such as dihydrotestosterone. These androgens inhibit aromatase activity, cannot be aromatized to estrogens, and inhibit FSH induction of LH-receptor expression on the granulosa cells (Gervásio, 2014).

This model predicts that follicles that lack adequate FSH receptor and granulosa cell number will remain primarily androgenic and will therefore become atretic. An increased androgen-to-estrogen ratio is found in the follicular fluid of atretic follicles, and several studies have demonstrated that high estrogen levels prevent apoptosis.

IGF-I also has apoptosis-suppressing activity and is produced by granulosa cells. This action of IGF-I is suppressed by certain IGF-binding proteins that are present in the follicular fluid of atretic follicles. The action of FSH to prevent atresia may therefore result, in part, from its ability to stimulate IGF-I synthesis and suppress the synthesis of the IGF-binding proteins.

Menstrual Cycle Phases

Follicular Phase

During the end of a previous cycle, estrogen, progesterone, and inhibin levels decrease abruptly with a corresponding increase in circulating FSH levels (Hodgen, 1982). As just described, this increase in FSH level is responsible for recruitment of the cohort of follicles that contains the follicle destined for ovulation. Despite general belief, sonographic studies in women have demonstrated that ovulation does not alternate sides, but occurs randomly from either ovary (Baird, 1987).

In women with waning ovarian function, the FSH level at this time of the cycle is elevated relative to that of younger women, presumably due to a loss of ovarian inhibin production in the previous luteal phase. As a result, measurement of early follicular or cycle day 3 FSH and estradiol levels is frequently performed in infertility clinics. The accelerated increase in serum FSH levels results in more robust recruitment of follicles and may explain both the shortened follicular phase observed in these older reproductive-aged women and the increased incidence of spontaneous twinning.

During the midfollicular phase, follicles produce increased amounts of estrogen and inhibin, resulting in a decline in FSH levels through negative feedback. This drop in FSH levels is believed to contribute to selection of the follicle destined to ovulate, termed the dominant follicle. Based on this theory, nondominant follicles express decreased numbers of FSH receptors and therefore are unable to respond adequately to declining FSH levels.

During most of follicular development, granulosa cell responses to FSH stimulation include an increase in granulosa cell number, an increase in aromatase expression, and, in the presence of estradiol, expression of LH receptors on the granulosa cells. With the development of LH-receptor expression during the late follicular phase, granulosa cells begin to produce small amounts of progesterone. This progesterone decreases granulosa cell proliferation, thereby slowing follicular growth (Chaffkin, 1992).

Ovulation

Toward the end of the follicular phase, estradiol levels increase dramatically. For reasons that are not completely understood but perhaps relate to changes at the kisspeptin neurons, the rapid estradiol level increase triggers a change from negative to positive feedback at both the hypothalamus and anterior pituitary gland to generate a surge in LH levels. Estradiol concentrations of 200 pg/mL for 50 hours are necessary to initiate this surge (Young, 1976). A small preovulatory increase in progesterone concentrations generates an FSH level surge, which occurs in tandem with the LH surge (McNatty, 1979). Progesterone may also augment the ability of estradiol to trigger the LH surge. These effects may explain the occasional induction of ovulation in anovulatory amenorrheic women when given progesterone to induce menses.

The LH surge acts rapidly on both the granulosa and theca cells of the preovulatory follicle to terminate the genes involved in follicular expression and turn on the genes necessary for ovulation and luteinization. In addition, the LH surge initiates the reentry of the oocyte into meiosis, expansion of the cumulus oophorus, synthesis of prostaglandins, and luteinization of granulosa cells. The mean duration of the LH surge is 48 hours, and ovulation occurs approximately 36 to 40 hours after the onset of the LH surge (Hoff, 1983; Lemarchand-Beraud, 1982). Abrupt termination of the surge is postulated to follow acutely increased steroid and inhibin secretion by the corpus luteum. Alternatively, the secretion of a gonadotropin surge-inhibiting/attenuating factor (GnSIF/AF) by either the ovary or hypothalamus is also postulated. However, the identity of this factor remains unknown (Vega, 2015).

The granulosa cells surrounding the oocyte, unlike mural granulosa cells, do not express LH receptors or synthesize progesterone. These cumulus oophorus granulosa cells develop tight gap junctions between themselves and with the oocyte. The cumulus mass that accompanies the ovulating oocyte is believed to provide a rough surface and increased size to improve oocyte “pick-up” by the tubal fimbria.

It has recently been found that amphiregulin, epiregulin, and beta-cellulin, which are epidermal growth factor-like factors, can be substituted to elicit the morphologic and biochemical events triggered by LH (Hsieh, 2009). Thus, these growth factors are part of the downstream cascade that begins with LH binding to its receptor and ends with ovulation.

Based on sonographic surveillance, extrusion of the oocyte lasts only a few minutes (Fig. 15-18) (Knobil, 1994). The exact mechanism of this expulsion is poorly defined but is not due to an increase in follicular pressure (Espey, 1974). The presence of proteolytic enzymes in the follicle, including plasmin and collagenase, suggests that these enzymes are responsible for follicular wall thinning (Beers, 1975). The preovulatory gonadotropin surge stimulates expression of tissue plasminogen activator by the granulosa and theca cells. The surge also decreases expression of plasminogen inhibitor, resulting in a marked increase in plasminogen activity (Piquette, 1993).

Prostaglandins also reach a peak concentration in follicular fluid during the preovulatory gonadotropin surge (Lumsden, 1986). Prostaglandins may stimulate smooth muscle contraction in the ovary, thereby contributing to ovulation (Yoshimura, 1987). Women undergoing infertility treatment are advised to avoid prostaglandin synthetase inhibitors in the preovulatory period to avoid luteinized unruptured follicle syndrome (LUFS) (Smith, 1996). Controversy exists as to whether LUFS should be considered pathologic or simply a sporadic event (Kerin, 1983).

Luteal Phase

Following ovulation, the remaining follicular cells differentiate into the corpus luteum, literally yellow body (Corner, 1956). This process, which requires LH stimulation, includes both morphologic and functional changes known as luteinization. The granulosa and theca cells proliferate and undergo hypertrophy to form granulosa-lutein cells and smaller theca-lutein cells, respectively (Patton, 1991).

During corpus luteum formation, the basement membrane that separates granulosa cells from theca cells degenerates and allows vascularization of previously avascular granulosa cells. Capillary invasion begins 2 days after ovulation and reaches the center of the corpus luteum by the fourth day. This increase in perfusion provides these luteal cells with access to circulating low-density lipoprotein (LDL), which is used to provide precursor cholesterol for steroid biosynthesis. This marked increase in blood supply can have clinical implications, as pain from a hemorrhagic corpus luteum cyst is a relatively frequent presentation to emergency rooms.

Adequate steroidogenesis in the corpus luteum depends on serum LH levels, LH receptors on luteal cells, and a sufficient number of luteal cells. Thus, it is critical that LH receptor expression on granulosa cells was appropriately induced during the prior follicular phase. Furthermore, blunted serum LH concentrations have been correlated with a shortened luteal phase. Luteal function is also influenced by gonadotropin levels from the preceding follicular phase. A reduction in LH or FSH secretion is correlated with poor luteal function. Presumably, a lack of FSH leads to a decrease in the total number of granulosa cells. Furthermore, luteal cells in these suboptimal cycles will have a decreased number of FSH-induced LH receptors and thus will be less responsive to LH stimulation.

Based on the corpus luteum’s steroidogenic products, the luteal phase is considered progesterone dominant, which contrasts with the estrogen dominance of the follicular phase. Increased vascularization, cellular hypertrophy, and an increased number of intracellular organelles transform the corpus luteum into the most active steroidogenic tissue in the body. Maximal levels of progesterone production are observed in the midluteal phase and have been estimated at an impressive 40 mg of progesterone per day. Ovulation can be safely assumed to have occurred if the progesterone level exceeds 3 ng/mL on cycle day 21.

Although progesterone is the most abundant ovarian steroid during the luteal phase, estradiol is also produced in significant quantities. Estradiol levels drop transiently immediately after the LH surge. This decline may explain the midcycle spotting noticed by some women. The reason for this decrease is not known, but it may result from a direct inhibition of granulosa cell growth by increasing progesterone levels (Hoff, 1983). The decline in estradiol levels is followed by a steady increase to reach a maximum during the midluteal phase.

The corpus luteum produces large quantities of inhibin A. This coincides with a decrease in circulating FSH levels in the luteal phase. If inhibin A levels decline at the end of the luteal phase, FSH levels rise once more to begin selection of a oocyte cohort for the next menstrual cycle.

If pregnancy does not occur, the corpus luteum regresses through a process called luteolysis. The mechanism for luteolysis is poorly understood, but luteal regression is presumed to be tightly regulated as luteal cycle length varies minimally among women. Following luteolysis, the blood supply to the corpus luteum diminishes, progesterone and estrogen secretion drop precipitously, and the luteal cells undergo apoptosis and become fibrotic. This creates the corpus albicans (white body).

If pregnancy occurs, hCG produced by the early gestation “rescues” the corpus luteum from atresia by binding to and activating the LH receptor on luteal cells. hCG stimulation of corpus luteum steroidogenesis maintains endometrial stability until placental steroid production is adequate to assume this function late in the first trimester. For this reason, surgical removal of the corpus luteum during pregnancy should be followed by progesterone replacement as outlined in Chapter 9 until approximately 10 weeks’ gestation.

Histology across the Menstrual Cycle

The endometrium consists of two layers: the basalis layer, which lies against the myometrium, and the functionalis layer, which is apposed to the uterine lumen. The basalis layer, which does not change significantly across the menstrual cycle, serves as the reserve for endometrium regeneration following menstrual sloughing. The functionalis layer is further subdivided into the more superficial stratum compactum, a thin layer of gland necks and dense stroma, and the underlying stratum spongiosum containing glands and large amounts of loosely organized stroma and interstitial tissue.

After menstruation, the endometrium is 1 to 2 mm thick. Under the influence of estrogen, the glandular and stromal cells of the functionalis layer proliferate rapidly following menses (Fig. 15-19). This period of rapid growth, the proliferative phase, corresponds to the ovary’s follicular phase. As this phase progresses, glands become more tortuous and cells lining the glandular lumen undergo pseudostratification. The stroma remains compact. Endometrial thickness approximates 12 mm at the time of the LH surge and does not increase significantly thereafter.

Following ovulation, the endometrium transforms into a secretory tissue. The period during and after this transformation is defined as the secretory phase of the endometrium and correlates to the ovary’s luteal phase. Glycogen-rich subnuclear vacuoles appear in cells lining the glands. Under further stimulation by progesterone, these vacuoles move from the glandular cells’ base toward their lumen and expel their contents. This secretory process peaks on approximately postovulatory day 6, coinciding with the day of implantation. Throughout the luteal phase, glands become increasingly tortuous, and the stroma becomes more edematous. In addition, spiral arteries that feed the endometrium increase their number and coiling.

If a blastocyst does not implant and the corpus luteum is not maintained by placental hCG, progesterone levels drop and endometrial glands begin to collapse. Polymorphonuclear leukocytes and monocytes from nearby vessels infiltrate the endometrium. The spiral arteries constrict, leading to local ischemia, and lysosomes release proteolytic enzymes that accelerate tissue destruction. Prostaglandins, particularly prostaglandin F_2α, are present in the endometrium and likely contribute to arteriolar vasospasm. Prostaglandin F_2α (PGF_2α) also induces myometrial contractions, which may aid in expelling the endometrial tissue.

The entire endometrial functionalis layer is thought to exfoliate with menstruation, leaving only the basalis layer to provide cells for endometrial regeneration. However, in studies, the amount of tissue shed from different levels of the endometrium varies widely. Following menstruation, reepithelialization of the desquamated endometrium is initiated within 2 to 3 days after the onset of menses and completed within 48 hours.

Regulation of Endometrial Function

Tissue Degradation and Hemorrhage

Within the endometrium, numerous proteins maintain a delicate balance between tissue integrity and the localized destruction required for menstrual sloughing or for trophoblast invasion during implantation. Cytokines, growth factors, and steroid hormones are believed to regulate the genes encoding these tissue proteins (Critchley, 2001). Of these proteins, tissue factor, a membrane-associated protein, activates the coagulation cascade upon contact with blood. In addition, urokinase and tissue plasminogen activator (TPA) increase the conversion of plasminogen to plasmin to activate tissue breakdown. TPA activity is blocked by plasminogen activator inhibitor 1, which is present in endometrial stroma (Lockwood, 1993; Schatz, 1995). Another key mediator group, matrix metalloproteinases (MMPs), is an enzyme family with overlapping substrate specificities for collagens and other extracellular matrix components. The composition of MMPs varies within different endometrial tissues and during the menstrual cycle. Endogenous MMP inhibitors are increased premenstrually and limit MMP degradative activity.

Vasoconstriction and Myometrial Contractility

Effective menstruation depends on appropriately timed endometrial vasoconstriction and myometrial contraction. Vasoconstriction produces ischemia, endometrial damage, and subsequent menstrual sloughing. Within the endometrium, epithelial and stromal cells secrete endothelin-1, a potent vasoconstrictor. Enkephalinase, an endothelin-degrading enzyme, is expressed at its highest levels in the midsecretory endometrium (Head, 1993). However, in the late luteal phase, the drop in serum progesterone leads to a loss of enkephalinase expression. This permits increased endothelin activity promoting a physiologic environment amenable to vasoconstriction.

In concert with endometrial sloughing, myometrial contractions control blood loss by compressing endometrial vasculature and expelling menstrual discharge. A fall in serum progesterone decreases an enzyme that degrades prostaglandins. This increases PGF_2α activity in the myometrium and triggers myometrial contractions (Casey, 1980).

Estrogens and Progestins

The expression of estrogen and progesterone receptors in the endometrium is highly regulated across the menstrual cycle. This provides an additional mechanism for controlling steroid effects on endometrial development and function. Estrogen receptors are expressed in the nuclei of epithelial, stromal, and myometrial cells, and concentrations peak during the proliferative phase. However, during the luteal phase, rising progesterone levels decrease estrogen receptor expression. Endometrial progesterone receptors peak at midcycle in response to rising estrogen levels. By midluteal phase, progesterone receptor expression in the glandular epithelium is nearly absent, although expression remains strong in the stromal compartment (Lessey, 1988).

The proliferation and differentiation of the uterine epithelium is under the control of estradiol, progesterone, and various growth factors. The importance of estrogens for endometrial development is demonstrated by the predominance of endometrial hyperplasia seen in women receiving unopposed estrogen therapy. Estrogen interacts directly with estrogen receptors but can also indirectly induce various growth factors that include IGF-I, transforming growth factor α, and epidermal growth factor (Beato, 1989; Dickson, 1987). The effects of progesterone on endometrial growth vary among endometrial layers. Progesterone is critical for the conversion of the functionalis layer from a proliferative to a secretory pattern. Within the basalis layer, progesterone appears to promote cellular proliferation.

Growth Factors and Cell Adhesion Molecules

Numerous growth factors and associated receptors act in the endometrium. Each factor has its own pattern of expression, making it difficult to determine which factor is most critical for endometrial function (Ohlsson, 1989; Sharkey, 1995).

In addition to growth factors, cell adhesion molecules play an important role in endometrial function. These molecules fall into four classes: integrins, cadherins, selectins, and members of the immunoglobulin superfamily. Each has been implicated in endometrial regeneration and embryo implantation, discussed next.

Implantation Window

The embryo enters the uterine cavity 2 to 3 days after fertilization with implantation beginning approximately 4 days later. Studies have demonstrated that normal implantation and embryonic development require synchronous development of the endometrium and the embryo (Pope, 1988). The human blastocyst may have less stringent requirements for implantation than other species as ectopic implantation occurs relatively frequently.

Uterine receptivity can be defined as the temporal window of endometrial maturation during which trophectoderm attaches to endometrial epithelial cells with subsequent invasion into endometrial stroma. The window of implantation in the human is relatively broad, extending from day 20 through day 24 of the menstrual cycle. Precise determination of this temporal window is critical since only those factors expressed during this time act as direct functional mediators of uterine receptivity.

Endometrial receptivity is associated with loss of surface microvilli and ciliated cells and with development of cellular protrusions, called pinopods, on the apical surface of the endometrium. Pinopods are considered an important morphologic marker of periimplantation endometrium. Pinopod formation is known to be highly progesterone dependent (Yoshinaga, 1989).

Various factors are believed to be necessary for uterine receptivity, including cell adhesion molecules, immunoglobulins, and cytokines. Integrins have been a particularly well-studied factor in this regard (Casals, 2010). However, to date, no single integrin molecule has been determined to be the critical marker of the implantation window.

The term luteal phase defect describes dyssynchrony between endometrial development and menstrual cycle phase that leads to subsequent implantation failure and early pregnancy loss (Noyes, 1950; Olive, 1991). This term currently is of limited utility in clinical practice due to our inability to accurately diagnose or treat the disorder.

Following implantation, the endometrium undergoes essential remodeling by invading trophoblast. In addition, the maternal endocrine environment changes extensively because of altered maternal physiology and contributions by the placenta and fetus. A more detailed discussion of these changes can be found in Williams Obstetrics, 24th edition (Cunningham, 2014).

Knowledge of normal hypothalamic-pituitary axis function, described in the preceding sections, is essential to understand reproductive endocrine pathology. Classically, abnormalities in this axis result in low gonadotropin and resultant low sex steroid levels. Termed hypogonadotropic hypogonadism, this can be developmental or acquired. Developmental lesions due to inherited genetic defects include Kallmann syndrome and idiopathic hypogonadotropic hypogonadism. Acquired abnormalities include functional disorders (eating disorders, excessive exercise, stress) and hypothalamic-pituitary lesions due to tumor, infiltrative diseases, infarction, surgery, or radiation therapy. Information regarding hypothalamic disorders and other causes of hypogonadotropic hypogonadism can be found in Chapter 16. Hyperprolactinemia and pituitary adenomas are discussed here.

Hyperprolactinemia

Etiology

Elevated circulating PRL levels can be caused by various physiologic activities including pregnancy, sleep, eating, and coitus. Increased PRL levels may also be observed following chest wall stimulation such as occurs with suckling, breast examination, chest wall surgery, herpes zoster infection, or nipple piercing (Table 12-2). PRL secretion is primarily regulated through tonic inhibition by dopamine released by the hypothalamus. PRL secretion is increased by serotonin, norepinephrine, opioids, estrogen, and TRH. Therefore, medications that block dopamine-receptor action (phenothiazines) or deplete catecholamine levels (monoamine oxidase inhibitors) may increase PRL levels (Table 12-2). Moreover, hyperprolactinemia may be caused by tumor, radiation, or infiltrative diseases such as sarcoid and tuberculosis. These can damage the pituitary stalk and thereby prevent dopamine-mediated inhibition of PRL secretion.

Primary hypothyroidism is also associated with mild elevations in serum PRL levels. Specifically, low circulating thyroid hormone levels produce a reflex increase in hypothalamic TRH levels due to loss of feedback inhibition. TRH can bind directly to anterior pituitary lactotropes and stimulate PRL production. As a rule, thyroid function tests should be performed when confirming a diagnosis of hyperprolactinemia, as a patient may require thyroid replacement rather than further evaluation for pituitary adenoma (Hekimsoy, 2010).

Prolactin-secreting adenomas, also termed prolactinomas, are the most common pituitary adenoma and the most common adenomas to be diagnosed by gynecologists. Affected women typically present with microadenomas and signs of PRL excess such as galactorrhea and amenorrhea.

Diagnosis

Hyperprolactinemia is, by definition, present in any patient with an elevated serum PRL level. Optimally, PRL levels are drawn in the morning, that is, at the time of the PRL nadir. Prior to testing, breast examination is avoided to prevent false-positive results. If a mildly elevated PRL level is found, sampling is repeated because PRL levels vary throughout the day. Moreover, many factors including the stress of venipuncture may produce false elevations.

Normal PRL levels are typically <20 ng/mL in nonpregnant women, although the upper limit of normal varies by assay. Importantly, PRL levels rise nearly 10-fold during pregnancy and make detection of a prolactinoma difficult at this time. Occasionally, the reported PRL value will be falsely low due to a “hook effect” present in the assay (Frieze, 2002). As explained on page 341, very high levels of endogenous hormone oversaturate the test antibodies and thereby prevent required binding between a patient’s PRL and the assay PRL. This problem is overcome with dilution of a patient’s sample. Importantly, a mismatch between the adenoma size noted on magnetic resonance (MR) imaging and the degree of PRL level elevation should alert a clinician to either the possibility of an incorrect assay result or the likelihood that the macroadenoma is actually not primarily PRL secreting. Macroadenomas of any cell type may damage the pituitary stalk and prevent transfer of hypothalamic dopamine to the lactotropes.

Conversely, a patient may rarely have an elevated PRL level on assay despite a lack of clinical features of hyperprolactinemia. The hyperprolactinemia in these patients is thought to be secondary to alternate forms of PRL, including the so-called big or macroprolactin, which contains multimers of native PRL. Macroprolactin is not physiologically active but may be detected by PRL assays (Fahie-Wilson, 2005).

For all patients with confirmed hyperprolactinemia, MR imaging is advisable. Some advocate limiting imaging to women with a PRL level >100 ng/mL, as lower levels are most likely due to small microadenomas (Fig. 15-20). Although this is undoubtedly a safe approach in most women, mildly elevated PRL levels also may be due to pituitary stalk compression by a nonprolactin-secreting macroadenoma or a craniopharyngioma, diagnoses with severe potential consequences.

The availability of sensitive neuroimaging techniques now affords earlier diagnosis and intervention. In the past, pituitary adenomas were identified using a coned-down view of the sella turcica during standard head radiography. Although computed tomography (CT) scanning provides useful information on tumor size, bony artifacts may limit interpretation. Therefore, MR imaging, using both T1- and T2-weighted images, has become the preferred radiologic approach due to its high sensitivity and excellent spatial resolution (Ruscalleda, 2005). Frequently, MR imaging is performed with and without gadolinium infusion for maximum definition of tumor size and extension.

Associated Amenorrhea

The primary mechanism linking hyperprolactinemia and amenorrhea is believed to be a reflex increase in central dopamine levels. Stimulation of the dopaminergic receptors on the GnRH neurons alters GnRH pulsatility, thereby disrupting folliculogenesis. As dopamine receptors have also been identified in the ovaries, detrimental effects on folliculogenesis may also play a role. Additional mechanisms undoubtedly exist in view of the complexity of the interactions among the various hormones, peptides, and neurotransmitters that influence hypothalamic function.

Pituitary Adenomas

Classification

Pituitary adenomas are the most common cause of acquired pituitary dysfunction and comprise approximately 15 percent of all intracranial tumors (Melmed, 2015; Pekic, 2015). Clinically, symptoms of galactorrhea, menstrual disturbances, or infertility may lead to its diagnosis. Most tumors are benign, and only an estimated 0.1 percent of adenomas develop into frank carcinoma with metastasis (Kaltsas, 2005). Nevertheless, pituitary adenomas may cause striking abnormalities in both endocrine and nervous system function (Table 15-5).

Pituitary adenomas were historically classified as eosinophilic, basophilic, or chromophobic according to their hematoxylin and eosin staining characteristics. Tumors are now classified by their hormonal expression pattern as determined by immunohistochemistry (Fig. 15-21). Adenomas are further grouped by size into microadenomas (<10 mm in diameter) and macroadenomas (>10 mm in diameter).

The most common adenomas secrete PRL alone. However, adenomas may secrete any of the pituitary hormones either singly (monohormonal adenoma) or in combination (multihormonal adenoma). In the past, a subset of tumors was considered nonsecreting. However, with more sensitive assays, most have been determined to secrete the common α-subunit or the gonadotropin β-subunits and therefore are gonadotrope-derived. Rarely, both α- and β-subunits are secreted as functional dimeric hormone.

Symptoms

Pituitary adenomas may cause symptoms via excess hormone secretion and lead to clinical conditions such as hyperprolactinemia, acromegaly, or Cushing disease. Alternatively, adenomas may result in hormone deficiency due to damage of other pituitary cell types or the pituitary stalk by an expanding adenoma or following treatment of the primary lesion.

As might be predicted, pituitary microadenomas are typically diagnosed during evaluation of an endocrinopathy. Macroadenomas frequently present with patient symptoms from invasion of surrounding structures. The anterior pituitary gland neighbors both the optic chiasm and cavernous sinuses. Disruption of the optic chiasm by suprasellar growth of the pituitary mass may create bitemporal hemianopsia, in which the outer portion of the right and left visual fields is lost. The cavernous sinuses are a paired collection of thin-walled veins located on either side of the sella turcica. Pituitary tumor compression can lead to cavernous sinus syndrome. This constellation of symptoms includes headache, visual disturbances, and cranial nerve palsies, specifically of cranial nerves III, IV, and VI.

Any pituitary mass or infiltrate can lead to reproductive dysfunction that may include delayed puberty, anovulation, oligomenorrhea, and infertility. The exact mechanisms linking adenomas to menstrual dysfunction are not well understood for many adenoma subtypes. Macroadenomas likely affect reproductive function either by compressing the pituitary stalk, which results in hyperprolactinemia, or less commonly, by directly compressing gonadotropes.

Spontaneous hemorrhage into a pituitary adenoma, termed pituitary apoplexy, is a rare life-threatening medical emergency. Signs and symptoms include acute visual changes, severe headache, neck stiffness, hypotension, loss of consciousness, and coma. These symptoms result from: (1) leakage of blood and necrotic material into the subarachnoid space, (2) acute hypopituitarism, and (3) a rapidly expanding hemorrhagic intrasellar mass that compresses the optic chiasm, cranial nerves, or hypothalamus and internal carotid arteries. Apoplexy may lead to severe hypoglycemia, hypotension, CNS hemorrhage, and death. Nevertheless, with rapid diagnosis and management, the outcome of patients with pituitary apoplexy is excellent (Singh, 2015). Glucocorticoid replacement is a mainstay of treatment. Surgical decompression is frequently but not invariably required.

Pregnancy and Pituitary Adenomas

The pituitary gland enlarges during pregnancy, primarily due to hypertrophy and hyperplasia of the lactotropes in response to elevated serum estrogen levels. Although the tumor can enlarge during pregnancy, the risk of clinically significant growth is small.Tumor growth causing significant symptoms has been reported in approximately 2 percent of patients with microadenomas and 21 percent of those with macroadenomas without prior surgical or radiation treatment (Molitch, 2015). However, because significant expansion may lead to headaches or compression of the optic chiasm and blindness, visual field testing is considered in every trimester for women with macroadenomas. Although dopamine agonist therapy has been associated with an increased risk of spontaneous abortions, preterm deliveries, and congenital malformation, the data overall suggest that treatment is safe for most patients. Nevertheless, most experts advise that dopamine agonist therapy be discontinued during pregnancy when possible.

Treatment of Hyperprolactinemia and Pituitary Adenomas

Most pituitary tumors grow slowly, and many cease growth after attainment of a certain size. Thus, asymptomatic patients with a microprolactinoma may be managed conservatively with serial MR imaging and serum PRL levels every 1 to 2 years as the risk of progression to a macroadenoma is <10 percent (Schlechte, 1989). These women should be followed for even mild changes in menstrual cyclicity as they are at risk for developing hypoestrogenism.

When tumors of any size are associated with amenorrhea or galactorrhea, therapy is considered (Fig. 15-22). Neurosurgical evaluation is mandatory when visual field defects or severe headaches are present. In general, first-line treatment is medical for both micro- and macroadenomas. Specifically, women should receive a dopamine agonist such as the nonspecific dopamine-receptor agonist bromocriptine (Parlodel) or the dopamine-receptor type 2 agonist cabergoline (Dostinex). Of note, the patient with hyperprolactinemia and normal imaging likely has an undetectable microadenoma and should be treated if symptomatic. The incidence of this occurring is decreasing with the advent of highly sensitive MR imaging.

Dopamine agonists decrease PRL secretion and shrink tumor size. However, bromocriptine treatment is associated with several common side effects, including headache, postural hypotension, blurry vision, drowsiness, and leg cramps. Most of these are attributable to activation of type 1 dopamine receptors. Due to its receptor specificity, cabergoline treatment is generally better tolerated than bromocriptine. Cabergoline also has a longer half-life, allowing once- or twice-weekly dosing compared with the multiple daily doses that may be required for bromocriptine. Typical initial cabergoline dosages are 0.25 mg orally twice weekly. Cabergoline has been found to be more effective than bromocriptine in normalizing PRL levels (dos Santos Nunes, 2011). Nevertheless, cabergoline can be prohibitively expensive. Most patients can tolerate bromocriptine if started at a low dose—½ tab or 0.125 mg—each night to minimize associated nausea and dizziness. This dose can be slowly increased to three times daily as tolerated. Reliable measurement of posttreatment serum PRL levels can usually be obtained 1 month following a steady medication dose.

Neurosurgery is required for refractory tumors or those causing acutely worsening symptoms. The pituitary is approached through a transsphenoidal route whenever possible (Fig. 15-23). Complications of surgery, although rare, include intraoperative hemorrhage, a cerebrospinal fluid leak (rhinorrhea), diabetes insipidus, damage to other pituitary cell types, and meningitis (Miller, 2014).

Radiation therapy may be used for patients with surgically nonresectable, persistent, or aggressive tumors. The radiation dose necessary to stop tumor growth is lower than the dose necessary to achieve normalization of hormonal hypersecretion. More precise stereotactic radiosurgical approaches such as the gamma knife have improved radiation-beam focus, which significantly decreases local tissue damage and improves patient tolerance. Risks include optic nerve damage and delayed development of hypopituitarism (Pashtan, 2014).

Clinical disorders may result from mutations in genes affecting hormone biosynthesis or the receptors that transduce the response. Both mechanisms have been described for pituitary peptide and steroid hormones.

First, numerous mutations have been identified in the genes that encode the LH/CG and FSH receptors. Most of the identified mutations are inactivating. These produce gonadotropin resistance of variable severity that ranges from primary amenorrhea to oligoamenorrhea and infertility (Latronico, 2013). Activating receptor mutations appear to be rare. However, a constitutively active LH/CG mutant receptor does cause a familial gonadotropin-independent precocious puberty that is limited to males (Ulloa-Aguirre, 2014). Although uncommon, mutations in the genes encoding the gonadotropin hormones themselves result in varying degrees of hypogonadism (Basciani, 2012; Kottler, 2010).

Second, gene mutation can lead to decreased function in steroid receptors. The best known of the latter is androgen insensitivity syndrome (AIS). With AIS, inactivating mutations in the androgen receptor impair the ability to respond to androgens. As a result, these 46,XY individuals have female external genitalia and scant to absent pubic and axillary hair. However, their testes do produce anti-müllerian hormone, their müllerian ducts fail to develop, and thus a blind-ending vaginal pouch without uterus or fallopian tubes results. Breasts develop in response to estrogens derived from aromatization of circulating testicular steroids (Tadokoro-Cucaro, 2014). Phenotypes of AIS are further described in Chapter 18.

Last, mutations in the enzymes necessary for steroidogenesis create broad clinical effects. Phenotypes depend on the location and severity of the resulting enzymatic deficiency within the steroidogenic pathway (Miller, 2011). The most common is congenital adrenal hyperplasia (CAH), typically due to a 21-hydroxylase deficiency. With severe enzymatic deficiency, affected patients have life-threatening salt wasting and female-to-male disorder of sexual differentiation (46,XX DSD). A less severe mutation may lead to “simple virilizing CAH” in which increased androgen levels are the primary hormonal abnormality (Auchus, 2015). The mildest abnormalities present as “nonclassic,” “late-onset,” or “adult-onset” CAH. In these patients, activation of the adrenal axis at puberty increases steroidogenesis and unmasks a mild 21-hydroxylase deficiency. Excess androgen provides negative feedback to GnRH receptors in the hypothalamus. These patients often present with hirsutism, acne, and anovulation. Thus, late-onset CAH may mimic PCOS (McCann-Crosby, 2014). Additional information regarding the CAH spectrum is found in Chapter 18.

In gynecology, estrogen and/or progestins are used for contraception and the treatment of abnormal uterine bleeding, endometriosis, leiomyomas, PCOS, and menopausal symptoms. Specific uses are found in respective chapters on these topics. Of the various estrogen and progesterone preparations, each differs in its biologic efficacy, and clinicians should understand the reasons behind some of these differences.

Estrogens

Classic estrogens are C-18 steroid compounds containing a phenolic ring (Fig. 15-24). This group contains the natural estrogens—estradiol, estrone, estriol—and their derivatives as well as conjugated equine estrogens (CEE). The predominant synthetic C-18 estrogen is ethinyl estradiol, the estrogen present in combination oral contraceptives. Synthetic nonsteroidal estrogens include diethylstilbestrol (DES) and selective estrogen-receptor modulators such as tamoxifen and clomiphene citrate. Despite their variation from the classic steroid ring shape, these nonsteroidal estrogens are still able to bind to the estrogen receptor.

Of the natural estrogens, 17β-estradiol is the most potent followed by estrone and then estriol. In comparing some pharmacologically used estrogens, ethinyl estradiol, a derivative of 17β-estradiol, has been estimated to be approximately 100 to 1000 times more potent on a per weight basis than either micronized estradiol or CEE in terms of increasing SHBG levels, which is one marker of estrogen potency (Kuhl, 2005; Mashchak, 1982).

Progestogens

Although there is no formal rule, progestogens include natural progesterone and synthetic progestogens called progestins. Only progesterone can maintain human pregnancy. Progestins can be classified as derivatives of either 19-norprogesterone or 19-nortestosterone (Kuhl, 2005). Of the 19-norprogesterones, the most commonly used are medroxyprogesterone acetate and megestrol acetate.

Most progestins used in contraceptives are derived from 19-nortestosterone. These are commonly described as first generation (norethindrone), second generation (levonorgestrel, norgestrel), or third generation (desogestrel, norgestimate). As described in Chapter 5, each generation has been designed to have progressively less androgenic effect. The fourth-generation progestin, drospirenone, is unique in that it is derived from spironolactone. Although it has no androgenic activity, drospirenone has an affinity for the mineralocorticoid receptor approximately five times that of aldosterone. This explains its diuretic action.

Selective Steroid-receptor Modulators

As indicated by their names, these synthetic compounds bind to their target receptors and exert tissue-specific effects, acting as agonists in some tissues and antagonists in others (Table 15-6). The best known of these are the selective estrogen-receptor modulators (SERMs) (Haskell, 2003). The variable activity among the SERMs can be attributed to differences at the molecular level. Each SERM binds to an estrogen receptor to generate a unique molecular conformation that affects the interaction of the complex with transcriptional cofactors and gene promoter regions. The response is also modified by the relative expression of ERα and ERβ receptors in the target tissue. The hormonal milieu may also be important in determining the agonist–antagonist profile of a specific SERM. For example, a SERM may act as an estrogen agonist in a low-estrogen state, such as menopause, but as a competitive antagonist in a patient with high circulating levels of the potent estrogen estradiol.

Development of new SERMs with specific agonist/antagonist profiles is an active area of current research. The SERM ospemifene (Osphena) may have favorable characteristics for long-term relief of genitourinary syndrome of menopause (Archer, 2015). Another SERM, bazedoxifene, is combined with CEE and marketed as Duavee. This coupled approach yields a drug group categorized as tissue-selective estrogen complexes (TSECs). The goal of TSECs is to craft agents with estrogen agonist/antagonist profiles that optimize clinical efficacy and safety.

More recently, selective progesterone-receptor modulators (SPRMs) have been developed to improve emergency contraception efficacy and expand treatment options for disorders including leiomyomas and endometriosis (Chwalisz, 2005). Selective androgen-receptor modulators (SARMs) are also under investigation for the treatment of osteopenia and decreased libido in women. Ideally, these will avoid the virilizing effects of testosterone treatment (Negro-Vilar, 1999).

As indicated by the preceding discussion, the agonist–antagonist effect of a steroid hormone is inextricably related to the clinical tissue of interest. Although this concept is most frequently discussed in terms of selective steroid modulators, in fact all steroid hormones within a class exert differences in their pattern of action across tissues. As a result, when a steroid is chosen for treatment, each clinical end point should be considered individually.

Steroid Hormone Potency

The efficacy of estrogen and progesterone treatments is altered by numerous factors such as: (1) receptor binding affinity, (2) formulation, (3) administration route, (4) metabolism, and (5) affinity for binding globulins. First, even small chemical modifications can substantially alter the biologic effects of steroid preparations. For example, the progestins in clinical use all exert progestogenic effects but may also act as weak androgens, antiandrogens, glucocorticoids, or antimineralocorticoids. These differences are likely explained by variations in binding affinity for each of these steroid receptors (Table 15-7).

Second, estrogens and progestins can be administered as oral, transdermal, vaginal, or intramuscular preparations, among others. The choice of carrier molecule affects hormone bioavailability. For example, although crystalline progesterone is poorly absorbed via the intestine, dispersion of the progesterone into small particles (micronization) markedly increases surface area and uptake.

Third, oral medications pass through the intestine and the liver prior to systemic dissemination. As these tissues are sites for steroid metabolism, oral medications and their levels may be significantly altered prior to reaching their target organs. As an example, the bioavailability of orally administered micronized progesterone is less than 10 percent and compares poorly with the estimated 50- to 70-percent bioavailability for norethindrone and 100 percent for levonorgestrel. This difference is due to a high level of “first pass” metabolism of micronized progesterone but not these modified progestins (Stanczyk, 2002). As another example, the half-life of ethinyl estradiol is greatly extended relative to that of unconjugated estradiol by the presence of the ethinyl group, which impairs metabolism.

Absorption and metabolism rates may differ between individuals due to inherited or acquired differences in liver, intestinal, and renal function (Kuhl, 2005). Local metabolism will also lower steroid efficacy and can include conversion between steroids (for example, androgens to estrogens by aromatase) or within a steroid type (for example, estradiol to the weaker estrone). Diet, alcohol consumption, cigarette smoking, exercise, and stress have all been postulated to alter steroid metabolism. Thyroid disease also affects drug metabolism rates. Medications that increase hepatic enzyme activity may increase estrogen metabolism. Best understood is the ability of some antiepileptic drugs to decrease estrogen-containing contraceptive efficacy (O’Brien, 2010).

Last, steroid potency depends on affinity for the various carrier proteins produced by the liver. Only unbound hormone and to a much lesser extent, the amount bound to albumin or cortisol-binding globulin (CBG) is functionally active. Steroid bound to SHBG is considered inactive. Ethinyl estradiol is bound nearly exclusively to albumin, and this increases its bioavailability (Barnes, 2007). As shown in Table 15-7, significant differences in carrier binding are also observed for progestogens (Wiegratz, 2004).

Importantly, hormonal status affects expression of carrier proteins. Specifically, estrogens and thyroid hormone stimulate SHBG while androgens blunt serum SHBG levels. To add further complexity, it is now believed that target cells can secrete SHBG, which then acts locally as a membrane receptor to stimulate cyclic adenosine monophosphate (cAMP) intracellular signaling pathways (Rosner, 2010).

Steroid Bioassays

A limited number of studies have used bioassays to evaluate the efficacy of estrogens in women using clinical, endocrinologic, and metabolic parameters (Table 15-8) (Kuhl, 2005). As seen in animal studies, different preparations vary markedly in their potency. Of note, estrogens also demonstrate differences in terms of their tissue specificity. For example, 17β-estradiol and CEE suppress pituitary FSH to a similar extent, whereas CEE is a more potent stimulator of liver SHBG production.