1155

Chapter 24 - The Endocrine System

Anirban Maitra MBBS Abul K. Abbas MBBS

• Chapter 24 - The Endocrine System– Pituitary Gland

• Normal• Pathology

– Clinical Manifestations of Pituitary Disease– Pituitary Adenomas and Hyperpituitarism

• Morphology.• Clinical Course.

» PROLACTINOMAS» GROWTH HORMONE (SOMATOTROPH CELL)

ADENOMAS» CORTICOTROPH CELL ADENOMAS» OTHER ANTERIOR PITUITARY ADENOMAS

– Hypopituitarism– Posterior Pituitary Syndromes– Hypothalamic Suprasellar Tumors

» Morphology.– Thyroid Gland

• Normal• Pathology

– Hyperthyroidism» Clinical Course.

– Hypothyroidism» CRETINISM» MYXEDEMA

– Thyroiditis» HASHIMOTO THYROIDITIS

• Pathogenesis.• Morphology.• Clinical Course.

» SUBACUTE (GRANULOMATOUS) THYROIDITIS

• Pathogenesis.• Morphology.• Clinical Course.

» SUBACUTE LYMPHOCYTIC (PAINLESS) THYROIDITIS

• Morphology.• Clinical Course.

– Graves Disease• Pathogenesis.• Morphology.• Clinical Course.

– Diffuse and Multinodular Goiters» DIFFUSE NONTOXIC (SIMPLE) GOITER

• Morphology.• Clinical Course.

» MULTINODULAR GOITER• Morphology.• Clinical Course.

– Neoplasms of the Thyroid» ADENOMAS

• Pathogenesis.• Morphology.• Clinical Features.

» OTHER BENIGN TUMORS» CARCINOMAS

• Pathogenesis• Genetic Factors.• Follicular Thyroid Carcinomas.• Papillary Thyroid Carcinomas.• Medullary Thyroid Carcinomas.• Anaplastic Carcinomas.• Environmental Factors.• Papillary Carcinoma

• Morphology.• Clinical Course.

• Follicular Carcinoma• Morphology.• Clinical Course.

• Medullary Carcinoma• Morphology.• Clinical Course.

• Anaplastic Carcinoma• Morphology.• Clinical Course.

– Congenital Anomalies– Parathyroid Glands

• Normal• Pathology

– Hyperparathyroidism» PRIMARY HYPERPARATHYROIDISM

• Morphology.

• Clinical Course.• Asymptomatic Hyperparathyroidism.• Symptomatic Primary Hyperparathyroidism.

» SECONDARY HYPERPARATHYROIDISM• Morphology.• Clinical Course.

– Hypoparathyroidism– Pseudohypoparathyroidism

– The Endocrine Pancreas• Normal• Pathology

– Diabetes Mellitus» DIAGNOSIS» CLASSIFICATION» NORMAL INSULIN PHYSIOLOGY

• Regulation of Insulin Release• Insulin Action and Insulin Signaling

Pathways» PATHOGENESIS OF TYPE 1 DIABETES

MELLITUS• Mechanisms of β Cell Destruction• Genetic Susceptibility• The MHC Locus.• Non-MHC Genes.• Environmental Factors

» PATHOGENESIS OF TYPE 2 DIABETES MELLITUS

• Insulin Resistance• Genetic Defects of the Insulin Receptor and

Insulin Signaling Pathway.• Obesity and Insulin Resistance.• β-Cell Dysfunction

» MONOGENIC FORMS OF DIABETES• Maturity-Onset Diabetes of the Young

(MODY).• Mitochondrial Diabetes.• Diabetes Associated with Insulin Gene or

Insulin Receptor Mutations.» PATHOGENESIS OF THE COMPLICATIONS

OF DIABETES• Formation of Advanced Glycation End

Products.• Activation of Protein Kinase C.• Intracellular Hyperglycemia with

Disturbances in Polyol Pathways.

» MORPHOLOGY OF DIABETES AND ITS LATE COMPLICATIONS

• Morphology.• Pancreas.• Diabetic Macrovascular Disease.• Diabetic Microangiopathy.• Diabetic Nephropathy.• Diabetic Ocular Complications.• Diabetic Neuropathy.

» CLINICAL FEATURES OF DIABETES– Pancreatic Endocrine Neoplasms

» HYPERINSULINISM (INSULINOMA)• Morphology.

» ZOLLINGER-ELLISON SYNDROME (GASTRINOMAS)

• Morphology.» OTHER RARE PANCREATIC ENDOCRINE

NEOPLASMS– Adrenal Glands

– Adrenal Cortex• Normal• Pathology

» ADRENOCORTICAL HYPERFUNCTION (HYPERADRENALISM)

• Hypercortisolism (Cushing Syndrome)• Pathogenesis.• Morphology.• Clinical Course.

• Primary Hyperaldosteronism• Morphology.• Clinical Course.

• Adrenogenital Syndromes• 21-Hydroxylase Deficiency.

• Morphology.• Clinical Course.

» ADRENAL INSUFFICIENCY• Primary Acute Adrenocortical Insufficiency• Waterhouse-Friderichsen Syndrome• Primary Chronic Adrenocortical

Insufficiency (Addison Disease)• Pathogenesis.• Morphology.• Clinical Course.

• Secondary Adrenocortical Insufficiency• Morphology.

» ADRENOCORTICAL NEOPLASMS

• Morphology.» OTHER LESIONS OF THE ADRENAL

– Adrenal Medulla• Normal• Pathology

» PHEOCHROMOCYTOMA• Morphology.• Clinical Course.

» TUMORS OF EXTRA-ADRENAL PARAGANGLIA

» NEUROBLASTOMA– Multiple Endocrine Neoplasia Syndromes

» MULTIPLE ENDOCRINE NEOPLASIA, TYPE 1» MULTIPLE ENDOCRINE NEOPLASIA, TYPE 2

– Pineal Gland• Normal• Pathology

» PINEALOMAS• Morphology.

1156

The endocrine system contains a highly integrated and widely distributed group of organs that orchestrates a state of metabolic equilibrium, or homeostasis, among the various organs of the body. Signaling by extracellular secreted molecules can be classified into three types—autocrine, paracrine, or endocrine—on the basis of the distance over which the signal acts. In endocrine signaling, the secreted molecules, which are frequently called hormones, act on target cells that are distant from their site of synthesis. An endocrine hormone is frequently carried by the blood from its site of release to its target. Increased activity of the target tissue often down-regulates the activity of the gland that secretes the stimulating hormone, a process known as feedback inhibition.

Hormones can be classified into several broad categories on the basis of the nature of their receptors. Cellular receptors and signaling pathways were discussed in Chapter 3 , and only a few comments about signaling by hormone receptors follow:

• Hormones that trigger biochemical signals upon interacting with cell-surface receptors: This large class of compounds is composed of two groups: (1) peptide hormones, such as growth hormone and insulin, and (2) small molecules, such as epinephrine. Binding of these hormones to cell-surface receptors leads to an increase in intracellular signaling molecules, termed second messengers, such as cyclic adenosine monophosphate (cAMP); production of mediators from membrane phospholipids, such as inositol 1,4,5-trisphosphate or IP3 ; and shifts in

the intracellular levels of ionized calcium. The elevated levels of one or more of these can control proliferation, differentiation, survival, and functional activity of cells, mainly by regulating the expression of specific genes. • Hormones that diffuse across the plasma membrane and interact with intracellular receptors: Many lipid-soluble hormones diffuse across the plasma membrane and interact with receptors in the cytosol or the nucleus. The resulting hormone-receptor complexes bind specifically to recognition elements in DNA, thereby affecting the expression of specific target genes. Hormones of this type include the steroids (e.g., estrogen, progesterone, and glucocorticoids), and thyroxine.

A number of processes can disturb the normal activity of the endocrine system, including impaired synthesis or release of hormones, abnormal interactions between hormones and their target tissues, and abnormal responses of target organs. Endocrine diseases can be generally classified as (1) diseases of underproduction or overproduction of hormones and their resulting biochemical and clinical consequences and (2) diseases associated with the development of mass lesions. Such lesions might be nonfunctional, or they might be associated with overproduction or underproduction of hormones. The study of endocrine diseases requires integration of morphologic findings with biochemical measurements of the levels of hormones, their regulators, and other metabolites.

Pituitary Gland

Normal

The pituitary is a small bean-shaped organ that measures about 1 cm in greatest diameter and weighs about 0.5 gm, although it enlarges during pregnancy. Its small size belies its great functional significance. It is located at the base of the brain, where it lies nestled within the confines of the sella turcica in close proximity to the optic chiasm and the cavernous sinuses. The pituitary is attached to the hypothalamus by the pituitary stalk, which passes out of the sella through an opening in the dura mater surrounding the brain. Along with the hypothalamus, the pituitary gland plays a critical role in

1157

Figure 24-1 Hormones released by the anterior pituitary. The adenohypophysis (anterior pituitary) releases five hormones that are in turn under the control of various stimulatory and inhibitory hypothalamic releasing factors. TSH, thyroid-stimulating hormone (thyrotropin); PRL, prolactin; ACTH, adrenocorticotrophic hormone (corticotropin); GH, growth hormone (somatotropin); FSH, follicle-stimulating hormone; LH, luteinizing hormone. The stimulatory releasing factors are TRH (thyrotropin-releasing factor), CRH (corticotropin-releasing factor), GHRH (growth hormone-releasing factor), GnRH (gonadotropin-releasing factor). The inhibitory hypothalamic influences are comprised of PIF (prolactin inhibitory factor or dopamine) and growth hormone inhibitory factor (GIH or somatostatin).

the regulation of most of the other endocrine glands. The pituitary is composed of two morphologically and functionally distinct components: The anterior lobe (adenohypophysis) and the posterior lobe (neurohypophysis).

The anterior pituitary, or adenohypophysis, constitutes about 80% of the gland. It is derived embryologically from Rathke pouch, which is an extension of the developing oral cavity. It is eventually cut off from its origins by the growth of the sphenoid bone, which creates a saddle-like depression, the sella turcica. The anterior pituitary has a portal vascular system that is the conduit for the transport of hypothalamic releasing hormones from the hypothalamus to the pituitary. Hypothalamic neurons have terminals in the median eminence where the hormones are released into the portal system, from where they traverse the pituitary stalk and enter the anterior pituitary gland. The production of most pituitary hormones is controlled predominantly by positive-acting releasing factors from the hypothalamus ( Fig. 24-1 ). Prolactin is the major exception, since its primary hypothalamic control is inhibitory, through the action of dopamine, while pituitary

Figure 24-2 A, Photomicrograph of normal pituitary. The gland is populated by several distinct cell populations containing a variety of stimulating (trophic) hormones. B, Each of the hormones has different staining characteristics, resulting in a mixture of cell types in routine histologic preparations. Immunostain for human growth hormone.

growth hormone receives both stimulatory and inhibitory influences via the hypothalamus. In routine histologic sections of the anterior pituitary, a colorful array of cells is present that contain eosinophilic cytoplasm (acidophil), basophilic cytoplasm (basophil), or poorly staining cytoplasm (chromophobe) cells ( Fig. 24-2 ). Specific antibodies against the pituitary hormones identify five cell types:

1. Somatotrophs, producing growth hormone (GH): These acidophilic cells constitute half of all the hormone-producing cells in the anterior pituitary.

2. Lactotrophs (mammotrophs), producing prolactin: These acidophilic cells secrete prolactin, which is essential for lactation.

3. Corticotrophs: These basophilic cells produce adrenocorticotropic hormone (ACTH), pro-opiomelanocortin (POMC), melanocyte-stimulating hormone (MSH), endorphins, and lipotropin.

4. Thyrotrophs: These pale basophilic cells produce thyroid-stimulating hormone (TSH).

1158

5. Gonadotrophs: These basophilic cells produce both follicle-stimulating hormone (FSH) and luteinizing hormone (LH). FSH stimulates the formation of graafian follicles in the ovary, and LH induces ovulation and the formation of corpora lutea in the ovary.

The posterior pituitary, or neurohypophysis, consists of modified glial cells (termed pituicytes) and axonal processes extending from nerve cell bodies in the supraoptic and paraventricular nuclei of the hypothalamus, through the pituitary stalk to the posterior lobe. These neurons produce two peptide hormones, anti-diuretic hormone (ADH, also called vasopressin) and oxytocin. The hormones are stored in axon terminals in the

posterior pituitary and are released into the circulation in response to appropriate stimuli. Oxytocin stimulates contraction of the smooth muscle cells in the gravid uterus and cells surrounding the lactiferous ducts of the mammary glands. ADH is a nonapeptide hormone synthesized predominantly in the supraoptic nucleus. In response to a number of different stimuli, including increased plasma osmotic pressure, left atrial distention, exercise, and certain emotional states, ADH is released from the axon terminals in the neurohypophysis into the general circulation. The posterior pituitary is derived embryologically from an outpouching of the floor of the third ventricle, which grows downward alongside the anterior lobe. In contrast to the anterior lobe, the posterior lobe of the pituitary is supplied by an artery and drains into a vein, where its hormones are released directly into the systemic circulation. Thus, the pituitary has a dual circulation, composed of arteries and veins and a portal venous system linking the hypothalamus and the anterior lobe.

Pathology

Clinical Manifestations of Pituitary Disease

The manifestations of pituitary disorders are as follows:

• Hyperpituitarism: Arising from excess secretion of trophic hormones. The causes of hyperpituitarism include pituitary adenoma, hyperplasia and carcinomas of the anterior pituitary, secretion of hormones by nonpituitary tumors, and certain hypothalamic disorders. The symptoms of hyperpituitarism are discussed in the context of individual tumors below. • Hypopituitarism: Arising from deficiency of trophic hormones. This may be caused by destructive processes, including ischemic injury, surgery or radiation, and inflammatory reactions. In addition, nonfunctional pituitary adenomas may encroach upon and destroy adjacent normal anterior pituitary parenchyma and cause hypopituitarism. • Local mass effects: Among the earliest changes referable to mass effect are radiographic abnormalities of the sella turcica, including sellar expansion, bony erosion, and disruption of the diaphragma sella. Because of the close proximity of the optic nerves and chiasm to the sella, expanding pituitary lesions often compress decussating fibers in the optic chiasm. This gives rise to visual field abnormalities, classically in the form of defects in the lateral (temporal) visual fields, so-called bitemporal hemianopsia. In addition, a variety of other visual field abnormalities may be caused by asymmetric growth of many tumors. Like any expanding intracranial mass, pituitary adenomas can produce signs and symptoms of elevated intracranial pressure, including headache, nausea, and

vomiting. On occasion, acute hemorrhage into an adenoma is associated with clinical evidence of rapid enlargement of the lesion, a situation appropriately termed pituitary apoplexy. Acute pituitary apoplexy is a neurosurgical emergency, since it can cause sudden death (see below). • Diseases of the posterior pituitary often come to clinical attention because of increased or decreased secretion of one of its products, ADH.

Pituitary Adenomas and Hyperpituitarism

The most common cause of hyperpituitarism is an adenoma arising in the anterior lobe. Other, less common, causes include hyperplasia and carcinomas of the anterior pituitary, secretion of hormones by some extrapituitary tumors, and certain hypothalamic disorders. Pituitary adenomas can be functional (i.e., associated with hormone excess and clinical manifestations thereof) or silent (i.e., immunohistochemical and/or ultrastructural demonstration of hormone production at the tissue level only, without clinical symptoms of hormone excess). Both functional and silent pituitary adenomas are usually composed of a single cell type and produce a single predominant hormone, although exceptions are known to occur. Pituitary adenomas are classified on the basis of hormone(s) produced by the neoplastic cells detected by immunohistochemical stains performed on tissue sections ( Table 24-1 ). Some pituitary adenomas can secrete two hormones (GH and prolactin being the most common combination), and rarely, pituitary adenomas are plurihormonal. Finally, pituitary adenomas may be hormone-negative, based on absence of immunohistochemical reactivity and ultrastructural

TABLE 24-1 -- Classification of Pituitary Adenomas

Prolactin cell (lactotroph) adenoma

Growth hormone cell (somatotroph) adenoma

Densely granulated GH cell   adenoma

Sparsely granulated GH cell adenoma   with fibrous bodies

Thyroid-stimulating hormone cell (thyrotroph) adenomas

ACTH cell (corticotroph) adenomas

Gonadotroph cell adenomas

Silent gonadotroph adenomas include   most so-called null cell and oncocytic adenomas

Mixed growth hormone-prolactin cell (mammosomatotroph) adenomas

Other plurihormonal adenomas

Hormone-negative adenomas

ACTH, adrenocorticotropic hormone.

1159

demonstration of lineage-specific differentiation. Both silent and hormone-negative pituitary adenomas may cause hypopituitarism as they encroach on and destroy adjacent anterior pituitary parenchyma.

Clinically diagnosed pituitary adenomas are responsible for about 10% of intracranial neoplasms; they are discovered incidentally in up to 25% of routine autopsies. In fact, using high-resolution computed tomography or magnetic resonance imaging suggest that approximately 20% of "normal" adult pituitary glands harbor an incidental lesion measuring 3 mm or more in diameter, usually a silent adenoma. [1] Pituitary adenomas are usually found in adults, with a peak incidence from the thirties to the fifties. Most pituitary adenomas occur as isolated lesions. In about 3% of cases, however, adenomas are associated with multiple endocrine neoplasia (MEN) type 1 (discussed later). Pituitary adenomas are designated, somewhat arbitrarily, microadenomas if they are less than 1 cm in diameter and macroadenomas if they exceed 1 cm in diameter. Silent and hormone-negative adenomas are likely to come to clinical attention at a later stage than those associated with endocrine abnormalities and are therefore more likely to be macroadenomas.

With recent advances in molecular techniques, substantial insight has been gained into the genetic abnormalities associated with pituitary adenomas:[2]

• The great majority of pituitary adenomas are monoclonal in origin, even those that are plurihormonal, suggesting that most arise from a single somatic cell. Some plurihormonal tumors may arise from clonal expansion of primitive stem cells, which then differentiate in several directions simultaneously. • G-protein mutations are possibly the best-characterized molecular abnormalities in pituitary adenomas. G-proteins are described in Chapter 3 ; here we will review their function in the context of endocrine neoplasms. G-proteins play a critical role in signal transduction, transmitting signals from cell-surface receptors (e.g., GHRH receptor) to intracellular effectors (e.g., adenyl cyclase), which then generate second messengers (e.g., cyclic AMP, cAMP). These are heterotrimeric proteins, composed of a specific α-subunit that binds guanine nucleotide and interacts with both cell surface receptors and intracellular effectors ( Fig. 24-3 ); the β- and γ-subunits are noncovalently bound to the specific α-subunit. Gs is a stimulatory G-protein that has a pivotal role in signal transduction in several endocrine organs, including the pituitary. The α-subunit of Gs (Gs α) is encoded by the GNAS1 gene, located on chromosome 20q13. In the basal state, Gs exists as an inactive protein, with GDP bound to the guanine nucleotide-binding site of the α-subunit of Gs . On interaction with the ligand-bound cell-surface receptor, GDP dissociates, and GTP binds to Gs α, activating the G-protein. The activation of Gs α results in the generation of cAMP, which acts as a potent mitogenic stimulus for

a variety of endocrine cell types (such as pituitary somatotrophs and corticotrophs, thyroid follicular cells, parathyroid cells), promoting cellular proliferation and hormone synthesis and secretion. The activation of Gs α, and resultant generation of cAMP, are transient because of an intrinsic GTPase activity in the α-subunit, which hydrolyzes GTP into GDP. A mutation in the α-subunit that interferes with its intrinsic GTPase activity will therefore result in constitutive activation of Gs α, persistent generation of cAMP, and unchecked cellular proliferation ( Fig. 24-3 ). Approximately 40% of somatotroph cell adenomas bear GNAS1 mutations that abrogate the GTPase activity of Gs α. The mutant form of GNAS1 is also known as the gsp oncogene because of its effects on tumorigenesis. In addition, GNAS1 mutations have also been described in a minority of corticotroph adenomas; in contrast, GNAS1 mutations are absent in thyrotroph, lactotroph, and gonadotroph adenomas, since their respective hypothalamic release hormones do not mediate their action via cAMP-dependent pathways. • Multiple endocrine neoplasia (MEN) syndrome (discussed in detail below) is a familial disorder associated with tumors and hyperplasias of multiple endocrine organs, including the pituitary. A subtype of MEN syndrome, known as MEN-1, is caused by germ line mutations of the gene MEN1, on chromosome 11q13. While MEN1 mutations are, by definition, present in pituitary adenomas arising in context of the MEN-1 syndrome, they are uncommon in sporadic pituitary adenomas. • Additional molecular abnormalities present in aggressive or advanced pituitary adenomas include activating mutations of the RAS oncogene and overexpression of the c-MYC oncogene, suggesting that these genetic events are linked to disease progression.[3]

Figure 24-3 The mechanism of G-protein mutations in endocrine neoplasia. Mutations in the G-protein-signaling pathway are seen in a variety of endocrine neoplasms, including pituitary, thyroid, and parathyroid adenomas. G-proteins play a critical role in signal transduction, transmitting signals from cell-surface receptors (GHRH, TSH, or PTH receptor) to intracellular effectors (e.g., adenyl cyclase), which then generate second messengers (cAMP).

1160

Morphology.

The common pituitary adenoma is a soft, well-circumscribed lesion that may be confined to the sella turcica. Larger lesions typically extend superiorly through the diaphragm sella into the suprasellar region, where they often compress the optic chiasm and adjacent structures, such as some of the cranial nerves ( Fig. 24-4 ). As these adenomas expand, they frequently erode the sella turcica and anterior clinoid processes. They may also extend locally into the cavernous and sphenoid sinuses. In up to 30% of cases, the adenomas are not grossly encapsulated and infiltrate adjacent bone, dura, and (rarely)

brain, but they do not demonstrate the ability for distant metastasis. Such lesions are termed invasive adenomas. Foci of hemorrhage and necrosis are common in larger adenomas.

Histologically, pituitary adenomas are composed of relatively uniform, polygonal cells arrayed in sheets or cords. Supporting connective tissue, or reticulin, is sparse, accounting for the soft, gelatinous consistency of many of these lesions. The nuclei of the neoplastic cells may be uniform or pleomorphic. Mitotic activity is usually modest. The cytoplasm of the constituent cells may be acidophilic, basophilic, or chromophobic, depending on the type and amount of secretory product within the cells, but it is generally uniform throughout the cytoplasm. This cellular monomorphism and the absence of a significant reticulin network distinguish pituitary adenomas from nonneoplastic anterior pituitary parenchyma ( Fig. 24-5 ). The functional status of the adenoma cannot be reliably predicted from its histologic appearance.

Clinical Course.

The signs and symptoms of pituitary adenomas include endocrine abnormalities and mass effects. The abnormalities associated with the secretion of excessive quantities of anterior pituitary hormones are mentioned below, when we describe the specific types of pituitary adenoma. Local mass effects may be encountered in any type of pituitary tumor and have been discussed previously under clinical manifestations of pituitary disease. Briefly, these include radiographic abnormalities of the sella turcica, visual

Figure 24-4 Pituitary adenoma. This massive, nonfunctional adenoma has grown far beyond the confines of the sella turcica and has distorted the overlying brain. Nonfunctional adenomas tend to be larger at the time of diagnosis than those that secrete a hormone.

Figure 24-5 Pituitary adenoma. The monomorphism of these cells contrasts markedly with the mixture of cells seen in the normal anterior pituitary. Note also the absence of reticulin network.

field abnormalities, signs and symptoms of elevated intracranial pressure, and occasionally hypopituitarism. Acute hemorrhage into an adenoma is sometimes associated with pituitary apoplexy, as was noted previously.

With this general introduction to pituitary adenomas, we proceed to a discussion of the individual types of tumors.

PROLACTINOMAS

Prolactinomas (lactotroph adenomas) are the most frequent type of hyperfunctioning pituitary adenoma, accounting for about 30% of all clinically recognized pituitary adenomas. These lesions range from small microadenomas to large, expansile tumors associated with substantial mass effect. Microscopically, the overwhelming majority of prolactinomas are composed of weakly acidophilic or chromophobic cells (sparsely granulated prolactinoma); rare prolactinomas are strongly acidophilic (densely granulated prolactinoma) ( Fig. 24-6 ). Prolactin can be demonstrated within the secretory granules in the cytoplasm of the cells using immunohistochemical approaches. Prolactinomas have a propensity to undergo dystrophic calcification, ranging from isolated psammoma bodies to extensive calcification of virtually the entire tumor mass ("pituitary stone"). Prolactin secretion by functioning adenomas is characterized by its efficiency—even microadenomas secrete sufficient prolactin to cause hyperprolactinemia—and by its proportionality, in that serum prolactin concentrations tend to correlate with the size of the adenoma.

Increased serum levels of prolactin, or prolactinemia, cause amenorrhea, galactorrhea, loss of libido, and infertility. The diagnosis of an adenoma is made more readily in women than in men, especially between the ages of 20 and 40 years, presumably because of the sensitivity of menses to disruption by hyperprolactinemia. This tumor underlies almost a quarter of cases of amenorrhea. In contrast, in men and older women, the

hormonal manifestations may be subtle, allowing the tumors to reach considerable size (macroadenomas) before being detected clinically.

Hyperprolactinemia may result from causes other than prolactin-secreting pituitary adenomas. Physiologic hyperprolactinemia occurs in pregnancy; serum prolactin levels

1161

Figure 24-6 Ultrastructural features of prolactinomas. A, Electron micrograph of a sparsely granulated prolactinoma. The tumor cells contain abundant granular endoplasmic reticulum (indicative of active protein synthesis) and small numbers of secretory granules (6000X). B, Electron micrograph of densely granulated growth hormone-secreting adenoma. The tumor cells are filled with large, membrane-bound secretory granules (6000X). (Courtesy of Dr. Eva Horvath, St. Michael's Hospital, Toronto, Ontario, Canada.)

increase throughout pregnancy, reaching a peak at delivery. Prolactin levels are also elevated by nipple stimulation, as occurs during suckling in lactating women, and as a response to many types of stress. Pathologic hyperprolactinemia can also result from lactotroph hyperplasia, such as when there is interference with normal dopamine inhibition of prolactin secretion. This may occur as a result of damage to the dopaminergic neurons of the hypothalamus, pituitary stalk section (e.g., owing to head trauma), or drugs that block dopamine receptors on lactotroph cells. Any mass in the suprasellar compartment may disturb the normal inhibitory influence of the hypothalamus on prolactin secretion, resulting in hyperprolactinemia, a phenomenon called the stalk effect. Therefore, a mild elevation in serum prolactin in a patient with a pituitary adenoma does not necessarily indicate a prolactin-secreting tumor. Several classes of drugs can cause hyperprolactinemia, including dopamine receptor antagonists such as the neuroleptic drugs (phenothiazines, haloperidol) and older antihypertensive drugs, such as reserpine, which inhibit dopamine storage. Other causes of hyperprolactinemia include estrogens, renal failure, and hypothyroidism. Prolactinomas are treated by surgery or,

more commonly, with bromocriptine, a dopamine receptor agonist, which causes the lesions to diminish in size.

GROWTH HORMONE (SOMATOTROPH CELL) ADENOMAS

GH-secreting tumors are the second most common type of functioning pituitary adenoma. As we have mentioned, 40% of somatotroph cell adenomas express a mutant GTPase-deficient α-subunit of the G-protein, Gs . Somatotroph cell adenomas may be quite large by the time they come to clinical attention because the manifestations of excessive GH may be subtle. Histologically, GH-containing adenomas are also classified into two subtypes: densely granulated and sparsely granulated. The densely granulated adenomas are composed of cells that are monomorphic and acidophilic in routine sections, retain strong cytoplasmic GH reactivity on immunohistochemistry, and demonstrate cytokeratin staining in a perinuclear distribution. In contrast, the sparsely granulated variants are composed of chromophobe cells with considerable nuclear and cytologic pleomorphism, and retain focal and weak GH reactivity.[4] Bihormonal mammosomatotroph adenomas that are reactive for both GH and prolactin are being increasingly recognized with the availability of better reagents for immunohistochemical analysis; morphologically, most bihormonal adenomas resemble the densely granulated pure somatotroph adenomas.

Persistent hypersecretion of GH stimulates the hepatic secretion of insulin-like growth factor I (IGF-I or somatomedin C), which causes many of the clinical manifestations. If a somatotrophic adenoma appears in children before the epiphyses have closed, the elevated levels of GH (and IGF-1) result in gigantism. This is characterized by a generalized increase in body size with disproportionately long arms and legs. If the increased levels of GH are present after closure of the epiphyses, patients develop acromegaly. In this condition, growth is most conspicuous in skin and soft tissues; viscera (thyroid, heart, liver, and adrenals); and bones of the face, hands, and feet. Bone density may be increased (hyperostosis) in both the spine and the hips. Enlargement of the jaw results in protrusion (prognathism) with broadening of the lower face. The hands and feet are enlarged with broad, sausage-like fingers. In most instances, gigantism is also accompanied by evidence of acromegaly. These changes develop for decades before being recognized, hence the opportunity for the adenomas to reach substantial size. GH excess is also correlated with a variety of other disturbances, including gonadal dysfunction, diabetes mellitus, generalized muscle weakness, hypertension, arthritis, congestive heart failure, and an increased risk of gastrointestinal cancers.

The diagnosis of pituitary GH excess relies on documentation of elevated serum GH and IGF-1 levels. In addition, failure to suppress GH production in response to an oral load of glucose is one of the most sensitive tests for acromegaly. The goals of treatment are to restore GH levels to normal and to decrease symptoms referable to a pituitary mass lesion while not causing hypopituitarism. To achieve these goals, the tumor can be removed surgically or destroyed by radiation therapy,

1162

or GH secretion can be reduced by drug therapy. When effective control of GH hypersecretion is achieved, the characteristic tissue overgrowth and related symptoms gradually recede, and the metabolic abnormalities improve.

CORTICOTROPH CELL ADENOMAS

Corticotroph adenomas are usually small microadenomas at the time of diagnosis. These tumors are most often basophilic (densely granulated) and occasionally chromophobic (sparsely granulated). Both variants stain positively with periodic acid-Schiff (PAS) because of the presence of carbohydrate in pre-opiomelanocorticotropin (POMC), the ACTH precursor molecule; in addition, they demonstrate variable immunoreactivity for POMC and its derivatives, including ACTH and β-endorphin.

Excess production of ACTH by the corticotroph adenoma leads to adrenal hypersecretion of cortisol and the development of hypercortisolism (also known as Cushing syndrome). This syndrome is discussed in more detail later with the diseases of the adrenal gland. It can be caused by a wide variety of conditions in addition to ACTH-producing pituitary tumors. When the hypercortisolism is due to excessive production of ACTH by the pituitary, the process is designated Cushing disease. Large destructive adenomas can develop in patients after surgical removal of the adrenal glands for treatment of Cushing syndrome. This condition, known as Nelson syndrome, occurs most often because of a loss of the inhibitory effect of adrenal corticosteroids on a pre-existing corticotroph microadenoma. Because the adrenals are absent in patients with this disorder, hypercortisolism does not develop. In contrast, patients present with mass effects of the pituitary tumor. In addition, there can be hyperpigmentation because of the stimulatory effect of other products of the ACTH precursor molecule on melanocytes.

OTHER ANTERIOR PITUITARY ADENOMAS

Pituitary adenomas may elaborate more than one hormone. For example, prolactin may be demonstrable by immunolabeling of somatotroph adenomas. In other cases, unusual plurihormonal adenomas are capable of secreting multiple hormones; these tumors are usually aggressive. A few comments are made about several of the less frequent functioning tumors.

Gonadotroph (LH-producing and FSH-producing) adenomas can be difficult to recognize because they secrete hormones inefficiently and variably, and the secretory products usually do not cause a recognizable clinical syndrome. Gonadotroph adenomas are most frequently found in middle-aged men and women when they become large enough to cause neurologic symptoms, such as impaired vision, headaches, diplopia, or pituitary apoplexy. Pituitary hormone deficiencies can also be found, most commonly impaired secretion of LH. This causes decreased energy and libido in men (due to reduced testosterone) and amenorrhea in premenopausal women. Thus, gonadotroph adenomas are paradoxically associated with secondary gonadal hypofunction. Most gonadotroph adenomas are large and composed of chromophobic cells. The neoplastic cells usually demonstrate immunoreactivity for the common gonadotropin α-subunit and the specific β-FSH and β-LH subunits; FSH is usually the predominant secreted hormone.

The availability of reliable immunoassays for the gonadotropin β-subunit and the recognition of gonadotroph-specific transcription factors has led to the reclassification of many previously hormone-negative adenomas ("null cell adenomas") as silent gonadotroph adenomas [5] (see below).

Thyrotroph (TSH-producing) adenomas are rare, accounting for approximately 1% of all pituitary adenomas. Thyrotroph adenomas are chromophobic or basophilic and are a rare cause of hyperthyroidism.

Nonfunctioning pituitary adenomas comprise both clinically silent counterparts of the functioning adenomas described above (for example, a silent somatotroph adenoma) and true hormone-negative adenomas. Nonfunctioning adenomas constitute approximately 25% of all pituitary tumors. In the past, the majority of nonfunctioning adenomas were classified as "null cell adenomas" because of the inability to demonstrate markers of differentiation. It is now known that most null cell adenomas have biochemical and ultrastructural features that allow their characterization as silent tumors of gonadotrophic lineage. [5] True hormone-negative adenomas are therefore unusual. Not surprisingly, the typical presentation of nonfunctioning adenomas is mass effects. These lesions may also compromise the residual anterior pituitary sufficiently to cause hypopituitarism. This may occur as a result of gradual enlargement of the adenoma or after abrupt enlargement of the tumor because of acute hemorrhage (pituitary apoplexy).

Pituitary carcinomas are quite rare, and most are not functional. These malignant tumors range from well differentiated, resembling somewhat atypical adenomas, to poorly differentiated, with variable degrees of pleomorphism and the features that are characteristic of carcinomas in other locations. The diagnosis of carcinoma requires the demonstration of metastases, usually to lymph nodes, bone, liver, and sometimes elsewhere.

Hypopituitarism

Hypopituitarism refers to decreased secretion of pituitary hormones, which can result from diseases of the hypothalamus or of the pituitary. Hypofunction of the anterior pituitary occurs when approximately 75% of the parenchyma is lost or absent. This may be congenital or the result of a variety of acquired abnormalities that are intrinsic to the pituitary. Hypopituitarism accompanied by evidence of posterior pituitary dysfunction in the form of diabetes insipidus (see below) is almost always of hypothalamic origin. Most cases of hypofunction arise from destructive processes directly involving the anterior pituitary, although other mechanisms have been identified:

• Tumors and other mass lesions: Pituitary adenomas, other benign tumors arising within the sella, primary and metastatic malignancies, and cysts can cause hypopituitarism. Any mass lesion in the sella can cause damage by exerting pressure on adjacent pituitary cells.

• Pituitary surgery or radiation: Surgical excision of a pituitary adenoma may inadvertently extend to the nonadenomatous pituitary. Radiation of the pituitary, used to prevent

1163

regrowth of residual tumor after surgery, can damage the nonadenomatous pituitary. • Pituitary apoplexy: As has been mentioned, this is a sudden hemorrhage into the pituitary gland, often occurring into a pituitary adenoma. In its most dramatic presentation, apoplexy causes the sudden onset of excruciating headache, diplopia owing to pressure on the oculomotor nerves, and hypopituitarism. In severe cases, it can cause cardiovascular collapse, loss of consciousness, and even sudden death. Thus, pituitary apoplexy is a true neurosurgical emergency. • Ischemic necrosis of the pituitary and Sheehan syndrome: Ischemic necrosis of the anterior pituitary is an important cause of pituitary insufficiency. Sheehan syndrome, or postpartum necrosis of the anterior pituitary, is the most common form of clinically significant ischemic necrosis of the anterior pituitary.[6] During pregnancy, the anterior pituitary enlarges to almost twice its normal size. This physiologic expansion of the gland is not accompanied by an increase in blood supply from the low-pressure venous system; hence, there is relative anoxia of the pituitary. Further reduction in blood supply caused by obstetric hemorrhage or shock may precipitate infarction of the anterior lobe. The posterior pituitary, because it receives its blood directly from arterial branches, is much less susceptible to ischemic injury in this setting and is therefore usually not affected. Pituitary necrosis may also be encountered in other conditions, such as disseminated intravascular coagulation and (more rarely) sickle cell anemia, elevated intracranial pressure, traumatic injury, and shock of any origin. Whatever the pathogenesis, the ischemic area is resorbed and replaced by a nubbin of fibrous tissue attached to the wall of an empty sella. • Rathke cleft cyst: These cysts, lined by ciliated cuboidal epithelium with occasional goblet cells and anterior pituitary cells, can accumulate proteinaceous fluid and expand, compromising the normal gland. • Empty sella syndrome: Any condition that destroys part or all of the pituitary gland, such as ablation of the pituitary by surgery or radiation, can result in an empty sella. The empty sella syndrome refers to the presence of an enlarged, empty sella turcica that is not filled with pituitary tissue. There are two types: (1) In a primary empty sella, there is a defect in the diaphragma sella that allows the arachnoid mater and cerebrospinal fluid to herniate into the sella, resulting in expansion of the sella and compression of the pituitary. Classically, affected patients are obese women with a history of multiple pregnancies. The empty sella syndrome may be associated with visual field defects and occasionally with endocrine anomalies, such as hyperprolactinemia, owing to interruption of inhibitory hypothalamic effects. Loss of functioning parenchyma can be severe enough to result in hypopituitarism. (2) In a secondary empty sella, a mass, such as a pituitary adenoma, enlarges the sella, but then it is either surgically removed

or undergoes spontaneous necrosis, leading to loss of pituitary function. Hypopituitarism can result from the treatment or spontaneous infarction. • Genetic defects: Rare congenital deficiencies of one or more pituitary hormones have been recognized in children. For example, mutations in pit-1, a pituitary transcription factor, result in combined deficiency of GH, prolactin, and TSH.[7]

Less frequently, disorders that interfere with the delivery of pituitary hormone-releasing factors from the hypothalamus, such as hypothalamic tumors, may also cause hypofunction of the anterior pituitary. Any disease involving the hypothalamus can alter secretion of one or more of the hypothalamic hormones that influence secretion of the corresponding pituitary hormones. In contrast to diseases that involve the pituitary directly, any of these conditions can also diminish the secretion of ADH, resulting in diabetes insipidus (discussed later). Hypothalamic lesions that cause hypopituitarism include:

• Tumors, including benign lesions that arise in the hypothalamus, such as craniopharyngiomas, and malignant tumors that metastasize to that site, such as breast and lung carcinomas. Hypothalamic hormone deficiency can ensue when brain or nasopharyngeal tumors are treated with radiation. • Inflammatory disorders and infections, such as sarcoidosis or tuberculous meningitis, can cause deficiencies of anterior pituitary hormones and diabetes insipidus.

The clinical manifestations of anterior pituitary hypofunction depend on the specific hormone(s) that are lacking. Children can develop growth failure (pituitary dwarfism) due to growth hormone deficiency. Gonadotropin (GnRH) deficiency leads to amenorrhea and infertility in women and decreased libido, impotence, and loss of pubic and axillary hair in men. TSH and ACTH deficiencies result in symptoms of hypothyroidism and hypoadrenalism, respectively, and are discussed later in the chapter. Prolactin deficiency results in failure of postpartum lactation. The anterior pituitary is also a rich source of melanocyte-stimulating hormone (MSH), synthesized from the same precursor molecule that produces ACTH; therefore, one of the manifestations of hypopituitarism includes pallor due to a loss of stimulatory effects of MSH on melanocytes.

Posterior Pituitary Syndromes

The clinically relevant posterior pituitary syndromes involve ADH and include diabetes insipidus and secretion of inappropriately high levels of ADH.

• Diabetes insipidus. ADH deficiency causes diabetes insipidus, a condition characterized by excessive urination (polyuria) owing to an inability of the kidney to resorb water properly from the urine. It can result from a variety of processes, including head trauma, tumors, and inflammatory disorders of the hypothalamus and pituitary as well as surgical procedures involving these organs. The condition can also arise spontaneously, in the absence of an underlying disorder. Diabetes insipidus from ADH deficiency is designated as central to differentiate it from nephrogenic diabetes insipidus, which is a result of renal tubular unresponsiveness to circulating ADH. The clinical manifestations of the two diseases are similar and include the excretion of large volumes of dilute urine with an inappropriately low specific gravity. Serum sodium and osmolality are increased owing to excessive renal loss of free

1164

water, resulting in thirst and polydipsia. Patients who can drink water can generally compensate for urinary losses; patients who are obtunded, bedridden, or otherwise limited in their ability to obtain water may develop life-threatening dehydration. • Syndrome of inappropriate ADH (SIADH) secretion. ADH excess causes resorption of excessive amounts of free water, resulting in hyponatremia. The most frequent causes of SIADH include the secretion of ectopic ADH by malignant neoplasms (particularly small cell carcinomas of the lung), non-neoplastic diseases of the lung, and local injury to the hypothalamus or posterior pituitary (or both). The clinical manifestations of SIADH are dominated by hyponatremia, cerebral edema, and resultant neurologic dysfunction. Although total body water is increased, blood volume remains normal, and peripheral edema does not develop.

Hypothalamic Suprasellar Tumors

Neoplasms in this location may induce hypofunction or hyperfunction of the anterior pituitary, diabetes insipidus, or combinations of these manifestations. The most commonly implicated lesions are gliomas (sometimes arising in the chiasm; see Chapter 28 ) and craniopharyngiomas. The craniopharyngioma is thought to be derived from vestigial remnants of Rathke pouch. These slow-growing tumors account for 1% to 5% of intracranial tumors; a small minority of these lesions arise within the sella, but most are suprasellar, with or without an intrasellar extension. A bimodal age distribution is observed, with one peak in childhood (5 to 15 years) and a second peak in adults in the sixth decade or older. Children usually come to clinical attention because of endocrine

deficiencies such as growth retardation, whereas adults usually present with visual disturbances. Pituitary hormonal deficiencies, including diabetes insipidus, are common.

Morphology.

Craniopharyngiomas average 3 to 4 cm in diameter; they may be encapsulated and solid, but more commonly, they are cystic and sometimes multiloculated. In their strategic location, they often encroach on the optic chiasm or cranial nerves, and not infrequently, they bulge into the floor of the third ventricle and base of the brain. Two distinct pathologic variants are recognized: adamantinomatous craniopharyngioma and papillary craniopharyngioma. The adamantinomatous type frequently contains radiologically demonstrable calcifications; the papillary variant is calcified only rarely.

Adamantinomatous craniopharyngioma consists of nests or cords of stratified squamous epithelium embedded in a spongy "reticulum" that becomes more prominent in the internal layers. Peripherally, the nests of squamous cells gradually merge into a layer of columnar cells, forming a palisade resting on a basement membrane. Compact, lamellar keratin formation ("wet keratin") is a diagnostic feature of this tumor. As was previously mentioned, dystrophic calcification is a frequent finding. Additional features include cyst formation, fibrosis, and chronic inflammatory reaction. The cysts of adamantinomatous craniopharyngiomas often contain a cholesterol-rich, thick brownish yellow fluid that has been compared to "machinery oil." These tumors extend fingerlets of epithelium into adjacent brain, where they elicit a brisk glial reaction.

Papillary craniopahryngiomas contain both solid sheets and papillae lined by well-differentiated squamous epithelium. These tumors usually lack keratin, calcification, and cysts. The squamous cells of the solid sections of the tumor do not have the peripheral palisading and do not typically generate a spongy reticulum in the internal layers.

Patients with craniopharyngiomas have an excellent recurrence-free and overall survival. Tumors greater than 5 cm in diameter are associated with a significantly higher recurrence rate. Adamantinomatous tumors are associated with a higher frequency of brain invasion, but this does not necessarily correlate with an adverse prognosis. Malignant transformation of craniopharyngiomas into squamous carcinomas is exceptionally rare and usually occurs postradiation.

Thyroid Gland

Normal

The thyroid gland consists of two bulky lateral lobes connected by a relatively thin isthmus, usually located below and anterior to the larynx. Normal variations in the structure of the thyroid gland include the presence of a pyramidal lobe, a remnant of the thyroglossal duct above the isthmus.

The thyroid gland develops from an evagination of the developing pharyngeal epithelium that descends as part of the thyroglossal duct from the foramen cecum at the base of the tongue to its normal position in the anterior neck. This pattern of descent explains the occasional presence of ectopic thyroid tissue, most commonly located at the base of the tongue (lingual thyroid) or at other sites abnormally high in the neck. Excessive descent leads to substernal thyroid glands. The clinical significance of these lesions lies in distinguishing

1165

them from metastatic thyroid carcinomas and the extremely rare occasions on which these ectopic sites can develop a primary thyroid malignancy.[8] Patients with lingual thyroids present an additional problem in that the ectopic thyroid tissue is sometimes the only thyroid tissue (total migration failure), and removal of the lingual thyroid results in symptomatic hypothyroidism. Malformations of branchial pouch differentiation may result in intrathyroidal sites of the thymus or parathyroid glands. The implication of these deviations becomes evident in the patient who has a total thyroidectomy and subsequently develops hypoparathyroidism.

The weight of the normal adult thyroid is approximately 15 to 25 gm. The thyroid has a rich intraglandular capillary network that is supplied by the superior and inferior thyroidal arteries. Nerve fibers from the cervical sympathetic ganglia indirectly influence thyroid secretion by acting on the blood vessels. The thyroid is divided by thin fibrous septae into lobules composed of about 20 to 40 evenly dispersed follicles. Normal follicles range from 50 to 500 µm in size, are lined by cuboidal to low columnar epithelium, and are filled with periodic acid Schiff (PAS)-positive thyroglobulin.

In response to trophic factors from the hypothalamus, TSH (thyrotropin) is released by thyrotrophs in the anterior pituitary into the circulation. The binding of TSH to its receptor on the thyroid follicular epithelium results in activation and conformational change in the receptor, allowing it to associate with a stimulatory G-protein ( Fig. 24-7 ). Activation of the G-protein eventually results in an increase in intracellular cAMP levels, which stimulates thyroid growth, and hormone synthesis and release via cAMP-dependent protein kinases. The dissociation of thyroid hormone synthesis and release from the controlled influence of TSH-signaling pathways results in so-called thyroid autonomy and hyperfunction (see below).

Thyroid follicular epithelial cells convert thyroglobulin into thyroxine (T4 ) and lesser amounts of triiodothyronine (T3 ). T4 and T3 are released into the systemic circulation, where most of these peptides are reversibly bound to circulating plasma proteins, such as thyroxine-binding globulin (TBG) and transthyretin, for transport to peripheral tissues. The binding proteins serve to maintain the serum unbound ("free") T3 and T4 concentrations within narrow limits yet ensure that the hormones are readily available to the tissues. In the periphery, the majority of free T4 is deiodinated to T3 ; the latter binds to thyroid hormone nuclear receptors in target cells with tenfold greater affinity than does T4 and has proportionately greater activity. The interaction of thyroid hormone with its nuclear thyroid hormone receptor (TR) results in the formation of a multi-protein hormone-receptor complex that binds to thyroid hormone response elements (TREs) in target genes, regulating their transcription (see Fig. 24-7 ).[9] Thyroid hormone has diverse cellular effects, including up-regulation of carbohydrate and lipid catabolism and stimulation of protein synthesis in a wide range of cells. The net result of these processes is an increase in the basal metabolic rate. One of the most important functions of thyroid hormone is its critical role in brain development, since absence of thyroid hormone during the fetal and neonatal periods may profoundly interfere with intellectual growth (see below).

The thyroid gland is one of the most responsive organs in the body and contains the largest store of hormones of any endocrine gland. The gland responds to many stimuli and is

Figure 24-7 Homeostasis in the hypothalamus-pituitary-thyroid axis and mechanism of action of thyroid hormones. Secretion of thyroid hormones (T3 and T4 ) is controlled by trophic factors secreted by both the hypothalamus and the anterior pituitary. Decreased levels of T3 and T4 stimulate the release of thyrotropin-releasing hormone (TRH) from the hypothalamus and thyroid-stimulating hormone (TSH) from the anterior pituitary, causing T3 and T4 levels to rise. Elevated T3 and T4 levels, in turn, suppress the secretion of both TRH and TSH. This relationship is termed a negative-feedback loop. TSH binds to the TSH receptor on the thyroid follicular epithelium, which causes activation of G proteins, and cyclic AMP (cAMP)-mediated synthesis and release of thyroid hormones (T3 and T4). In the periphery, T3 and T4 interact with the thyroid hormone receptor (TR) to form a hormone-receptor complex that translocates to the nucleus and binds to so-called thyroid response elements (TREs) on target genes initiating transcription.

in a constant state of adaptation. During puberty, pregnancy, and physiologic stress from any source, the gland increases in size and becomes more active. This functional lability is reflected in transient hyperplasia of the thyroidal epithelium. At this time, thyroglobulin is resorbed, and the follicular cells become tall and more columnar, sometimes forming small, infolded buds or papillae. When the stress abates, involution occurs; that is, the height of the epithelium falls, colloid accumulates, and the follicular

cells resume their normal size and architecture. Failure of this normal balance between hyperplasia and involution can produce major or minor deviations from the usual histologic pattern.

The function of the thyroid gland can be inhibited by a variety of chemical agents, collectively referred to as goitrogens. Because they suppress T3 and T4 synthesis, the level of TSH increases, and subsequent hyperplastic enlargement of the gland (goiter) follows. The antithyroid agent propylthiouracil inhibits the oxidation of iodide and blocks production

1166

of the thyroid hormones; parenthetically, propylthiouracil also inhibits the peripheral deiodination of circulating T4 into T3 , thus ameliorating symptoms of thyroid hormone excess (see below). Iodide, when given to patients with thyroid hyperfunction, also blocks the release of thyroid hormones but through different mechanisms. Iodides in large doses inhibit proteolysis of thyroglobulin. Thus, thyroid hormone is synthesized and incorporated within increasing amounts of colloid, but it is not released into the blood.

The thyroid gland follicles also contain a population of parafollicular cells, or C cells, which synthesize and secrete the hormone calcitonin. This hormone promotes the absorption of calcium by the skeletal system and inhibits the resorption of bone by osteoclasts.

Pathology

Diseases of the thyroid are of great importance because most are amenable to medical or surgical management. They include conditions associated with excessive release of thyroid hormones (hyperthyroidism), those associated with thyroid hormone deficiency (hypothyroidism), and mass lesions of the thyroid. We first consider the clinical consequences of disturbed thyroid function, then focus on the disorders that generate these problems.

Hyperthyroidism

Thyrotoxicosis is a hypermetabolic state caused by elevated circulating levels of free T3 and T4 . Because it is caused most commonly by hyperfunction of the thyroid gland, it is often referred to as hyperthyroidism. However, in certain conditions the oversupply is related to either excessive release of preformed thyroid hormone (e.g., in thyroiditis) or to an extrathyroidal source, rather than hyperfunction of the gland ( Table 24-2 ). Thus, strictly speaking, hyperthyroidism is only

TABLE 24-2 -- Disorders Associated with Thyrotoxicosis

Associated with Hyperthyroidism

Primary

Diffuse toxic hyperplasia (Graves   disease)

Hyperfunctioning ("toxic") multinodular   goiter

Hyperfunctioning ("toxic")   adenoma

Hyperfunctioning thyroid   carcinoma

Iodine-induced   hyperthyroidism

Neonatal thyrotoxicosis associated with   maternal Graves disease

Secondary

TSH-secreting pituitary adenoma   (rare) *

Not Associated with Hyperthyroidism

Subacute granulomatous thyroiditis (painful)

Subacute lymphocytic thyroiditis (painless)

Struma ovarii (ovarian teratoma with ectopic thyroid)

Factitious thyrotoxicosis (exogenous thyroxine intake)

*Associated with increased TSH; all other causes of thyrotoxicosis associated with decreased TSH.

one (albeit the most common) cause of thyrotoxicosis. The terms primary and secondary hyperthyroidism are sometimes used to designate hyperthyroidism arising from an intrinsic thyroid abnormality and that arising from processes outside of the thyroid, such as a TSH-secreting pituitary tumor. With this disclaimer, we will follow the common practice of using the terms thyrotoxicosis and hyperthyroidism interchangeably. The three most common causes of thyrotoxicosis are also associated with hyperfunction of the gland and include the following:

• Diffuse hyperplasia of the thyroid associated with Graves disease (accounts for 85% of cases) • Hyperfunctional multinodular goiter • Hyperfunctional adenoma of the thyroid

Clinical Course.

The clinical manifestations of hyperthyroidism are protean and include changes referable to the hypermetabolic state induced by excess thyroid hormone as well as those related to

overactivity of the sympathetic nervous system (i.e., an increase in the β-adrenergic "tone").

Excessive levels of thyroid hormone result in an increase in the basal metabolic rate. The skin of thyrotoxic patients tends to be soft, warm, and flushed because of increased blood flow and peripheral vasodilation to increase heat loss. Heat intolerance is common. Sweating is increased because of higher levels of calorigenesis. Increased basal metabolic rate also results in characteristic weight loss despite increased appetite.

Cardiac manifestations are among the earliest and most consistent features of hyperthyroidism. Patients with hyperthyroidism can have an increase in cardiac output, owing to both increased cardiac contractility and increased peripheral oxygen requirements. Tachycardia, palpitations, and cardiomegaly are common. Arrhythmias, particularly atrial fibrillation, occur frequently and are more common in older patients. Congestive heart failure may develop, particularly in elderly patients with pre-existing cardiac disease. Myocardial changes, such as foci of lymphocytic and eosinophilic infiltration, mild fibrosis in the interstitium, fatty changes in myofibers, and an increase in size and number of mitochondria, have been described. Some patients with thyrotoxicosis develop a reversible diastolic dysfunction and a "low-output" failure, so-called thyrotoxic dilated cardiomyopathy.

In the neuromuscular system, overactivity of the sympathetic nervous system produces tremor, hyperactivity, emotional lability, anxiety, inability to concentrate, and insomnia. Proximal muscle weakness is common with decreased muscle mass (thyroid myopathy).

Ocular changes often call attention to hyperthyroidism. A wide, staring gaze and lid lag are present because of sympathetic overstimulation of the levator palpebrae superioris ( Fig. 24-8 ). However, true thyroid ophthalmopathy associated with proptosis is a feature seen only in Graves disease (see below).

In the gastrointestinal system, sympathetic hyperstimulation of the gut results in hypermotility, malabsorption, and diarrhea.

The skeletal system is also affected in hyperthyroidism. Thyroid hormone stimulates bone resorption, resulting in increased porosity of cortical bone and reduced volume of trabecular bone. The net effect is osteoporosis and an increased risk of fractures in patients with chronic hyperthyroidism.

1167

Figure 24-8 A patient with hyperthyroidism. A wide-eyed, staring gaze, caused by overactivity of the sympathetic nervous system, is one of the features of this disorder. In Graves disease, one of the most important causes of hyperthyroidism, accumulation of loose connective tissue behind the eyeballs also adds to the protuberant appearance of the eyes.

Other findings throughout the body include atrophy of skeletal muscle, with fatty infiltration and focal interstitial lymphocytic infiltrates; minimal liver enlargement due to fatty changes in the hepatocytes; and generalized lymphoid hyperplasia with lymphadenopathy in patients with Graves disease.

Thyroid storm is used to designate the abrupt onset of severe hyperthyroidism. This condition occurs most commonly in patients with underlying Graves disease and probably results from an acute elevation in catecholamine levels, as might be encountered during infection, surgery, cessation of antithyroid medication, or any form of stress. Patients are often febrile and present with tachycardia out of proportion to the fever. Thyroid storm is a medical emergency: A significant number of untreated patients die of cardiac arrhythmias.

Apathetic hyperthyroidism refers to thyrotoxicosis occurring in the elderly, in whom old age and various comorbidities may blunt the typical features of thyroid hormone excess seen in younger patients. The diagnosis of thyrotoxicosis in these patients is often made during laboratory work-up for unexplained weight loss or worsening cardiovascular disease.

A diagnosis of hyperthyroidism is made using both clinical and laboratory findings. The measurement of serum TSH concentration using sensitive TSH (sTSH) assays provides the most useful single screening test for hyperthyroidism, as its levels are decreased even at the earliest stages, when the disease may still be subclinical.[10] A low TSH value is usually confirmed with measurement of free T4 , which is expectedly increased. In an occasional patient, hyperthyroidism results predominantly from increased circulating levels of T3 ("T3 toxicosis"). In these cases, free T4 levels may be decreased, and direct

measurement of serum T3 may be useful. In rare cases of pituitary-associated (secondary) hyperthyroidism, TSH levels are either normal or raised. Determining TSH levels after the injection of TRH (TRH stimulation test) is used in the evaluation of cases of suspected hyperthyroidism with equivocal changes in the baseline serum TSH level. A normal rise in TSH after administration of TRH excludes secondary hyperthyroidism. Once the diagnosis of thyrotoxicosis has been confirmed by a combination of sTSH assays and free thyroid hormone levels, measurement of radioactive iodine uptake by the thyroid gland may be valuable in determining the etiology. For example, there may be diffusely increased uptake in the whole gland (Graves disease), increased uptake in a solitary nodule (toxic adenoma), or decreased uptake (thyroiditis). The therapeutic options for hyperthyroidism include multiple medications, each of which has a different mechanism of action. Typically, these include a β-blocker to control symptoms induced by increased adrenergic tone, a thionamide to block new hormone synthesis, an iodine solution to block the release of thyroid hormone, and agents that inhibit peripheral conversion of T4 to T3 . Radioiodine, which is incorporated into thyroid tissues, resulting in ablation of thyroid function over a period of 6 to 18 weeks, may also be used.

Hypothyroidism

Hypothyroidism is caused by any structural or functional derangement that interferes with the production of adequate levels of thyroid hormone. It can result from a defect anywhere in the hypothalamic-pituitary-thyroid axis. As in the case of hyperthyroidism, this disorder is divided into primary and secondary categories, depending on whether the hypothyroidism arises from an intrinsic abnormality in the thyroid or occurs as a result of pituitary disease; rarely, hypothalamic failure is a cause of tertiary hypothyroidism ( Table 24-3 ). Primary hypothyroidism accounts for the vast majority of cases of hypothyroidism. Primary hypothyroidism can be thyroprivic (due to absence or loss of thyroid parenchyma) or goitrous (due to enlargement of the thyroid gland under the influence of TSH). The causes of primary hypothyroidism include the following.

Surgical or radiation-induced ablation of thyroid parenchyma can cause hypothyroidism. A large resection of the gland (total thyroidectomy) for the treatment of hyperthyroidism of a primary neoplasm can lead to hypothyroidism. The gland may also be ablated by radiation, whether in the form of radioiodine administered for the treatment of hyperthyroidism, or

TABLE 24-3 -- Causes of Hypothyroidism

Primary

Developmental (thyroid dysgenesis: PAX-8, TTF-2, TSH-receptor mutations)

Thyroid hormone resistance syndrome (TRβ mutations)

Postablative

Surgery, radioiodine therapy, or   external radiation

Autoimmune hypothyroidism

Hashimoto   thyroiditis *

Iodine deficiency *

Drugs (lithium, iodides, p-aminosalicylic acid) *

Congenital biosynthetic defect (dyshormonogenetic goiter) *

Secondary

Pituitary failure

Tertiary

Hypothalamic failure (rare)

*Associated with enlargement of thyroid ("goitrous hypothyroidism"). Hashimoto thyroiditis and postablative hypothroidism account for the majority of cases.

1168

exogenous irradiation, such as external radiation therapy to the neck.

Autoimmune hypothyroidism is the most common cause of goitrous hypothyroidism in iodine-sufficient areas of the world. The vast majority of cases of autoimmune hypothyroidism are due to Hashimoto thyroiditis. Circulating autoantibodies, including anti-TSH receptor autoantibodies, are commonly found in Hashimoto thyroiditis. Some patients with hypothyroidism have circulating anti-TSH antibodies, but they usually do not have the goitrous enlargement or lymphocytic infiltrate characteristic of Hashimoto thyroiditis. In the past, many of these patients were classified as having primary "idiopathic" hypothyroidism, but the disease is now recognized as a type of autoimmune disorder of the thyroid, occurring either in isolation or in conjunction with other autoimmune endocrine manifestations.

Drugs given intentionally to decrease thyroid secretion (e.g., methimazole and propylthiouracil) can cause hypothyroidism, as can agents used to treat nonthyroid conditions (e.g., lithium, p-aminosalicylic acid).

Inborn errors of thyroid metabolism are an uncommon cause of goitrous hypothyroidism (dyshormonogenetic goiter). Any one of the multiple steps leading to thyroid hormone synthesis may be deficient: (1) iodide transport defect, (2) organification defect, (3) dehalogenase defect, and (4) iodotyrosine coupling defect. Organification of iodine involves binding of oxidized iodide with tyrosyl residues in thyroglobulin, and this process is deficient in patients with Pendred syndrome, wherein goitrous hypothyroidism is accompanied by sensorineural deafness.

Thyroid hormone resistance syndrome is a rare autosomal-dominant disorder caused by inherited mutations in the thyroid hormone receptor (TR), which abolish the ability of the receptor to bind thyroid hormones.[11] Patients demonstrate a generalized resistance to thyroid hormone, despite high circulating levels of T3 and T4 . Since the pituitary is also resistant to feedback from thyroid hormones, TSH levels tend to be high as well. In rare instances, there may be complete absence of thyroid parenchyma (thyroid agenesis), or the gland may be greatly reduced in size (thyroid hypoplasia). Mutations in the TSH receptor are a newly recognized cause of congenital hypothyroidism associated with a hypoplastic thyroid gland.[12] Recently, mutations in two transcription factors that are expressed in the developing thyroid and regulate follicular differentiation—thyroid transcription factor-2 (TTF-2)[13] and Paired Homeobox-8 (PAX-8) [14] —have been reported in patients with thyroid agenesis. Thyroid agenesis caused by TTF-2 mutations is usually associated with a cleft palate.

Secondary hypothyroidism is caused by TSH deficiency, and tertiary (central) hypothyroidism is caused by TRH deficiency. Secondary hypothyroidism can result from any of the causes of hypopituitarism. Frequently, the cause is a pituitary tumor; other causes include postpartum pituitary necrosis, trauma, and nonpituitary tumors, as was previously discussed. Tertiary (central) hypothyroidism can be caused by any disorder that damages the hypothalamus or interferes with hypothalamic-pituitary portal blood flow, thereby preventing delivery of TRH to the pituitary. This can result from hypothalamic damage from tumors, trauma, radiation therapy, or infiltrative diseases. Classic clinical manifestations of hypothyroidism include cretinism and myxedema.

CRETINISM

Cretinism refers to hypothyroidism that develops in infancy or early childhood. The term cretin was derived from the French chrétien, meaning Christian or Christlike, and was applied to these unfortunates because they were considered to be so mentally retarded as to be incapable of sinning. In the past, this disorder occurred fairly commonly in areas of the world where dietary iodine deficiency is endemic, such as the Himalayas, inland China, Africa, and other mountainous areas. It has become much less frequent in recent years, owing to the widespread supplementation of foods with iodine. On rare occasions, cretinism may also result from inborn errors in metabolism (e.g., enzyme deficiencies) that interfere with the biosynthesis of normal levels of thyroid hormone (sporadic cretinism).

Clinical features of cretinism include impaired development of the skeletal system and central nervous system, manifested by severe mental retardation, short stature, coarse

facial features, a protruding tongue, and umbilical hernia. The severity of the mental impairment in cretinism appears to be related to the time at which thyroid deficiency occurs in utero. Normally, maternal hormones, including T3 and T4 , cross the placenta and are critical to fetal brain development. If there is maternal thyroid deficiency before the development of the fetal thyroid gland, mental retardation is severe. In contrast, reduction in maternal thyroid hormones later in pregnancy, after the fetal thyroid has developed, allows normal brain development.

MYXEDEMA

The term myxedema is applied to hypothyroidism developing in the older child or adult. Myxedema, or Gull disease, was first linked with thyroid dysfunction in 1873 by Sir William Gull in a paper addressing the development of a "cretinoid state" in adults. The clinical manifestations vary with the age of onset of the deficiency. The older child shows signs and symptoms intermediate between those of the cretin and those of the adult with hypothyroidism. In the adult, the condition appears insidiously and may take years to reach the level of clinical suspicion.

Clinical features of myxedema are characterized by a slowing of physical and mental activity. The initial symptoms include generalized fatigue, apathy, and mental sluggishness, which may mimic depression in the early stages of the disease. Speech and intellectual functions become slowed. Patients with myxedema are listless, cold-intolerant, and frequently overweight. Reduced cardiac output probably contributes to shortness of breath and decreased exercise capacity, two frequent complaints in patients with hypothyroidism. Decreased sympathetic activity results in constipation and decreased sweating. The skin in these patients is cool and pale because of decreased blood flow. Histologically, there is an accumulation of matrix substances, such as glycosaminoglycans and hyaluronic acid, in skin, subcutaneous tissue, and a number of visceral sites. This results in edema, a broadening and coarsening of facial features, enlargement of the tongue, and deepening of the voice.

Laboratory evaluation plays a vital role in the diagnosis of suspected hypothyroidism because of the nonspecific nature

1169

of symptoms. Measurement of the serum TSH level is the most sensitive screening test for this disorder. The TSH level is increased in primary hypothyroidism due to a loss of feedback inhibition of TRH and TSH production by the hypothalamus and pituitary, respectively. The TSH level is not increased in patients with hypothyroidism due to primary hypothalamic or pituitary disease. T4 levels are decreased in patients with hypothyroidism of any origin.

Thyroiditis

Thyroiditis, or inflammation of the thyroid gland, encompasses a diverse group of disorders characterized by some form of thyroid inflammation. These diseases include conditions that result in acute illness with severe thyroid pain (e.g., infectious thyroiditis, subacute granulomatous thyroiditis) and disorders in which there is relatively little inflammation and the illness is manifested primarily by thyroid dysfunction (subacute lymphocytic thyroiditis and fibrous [Reidel] thyroiditis).

Infectious thyroiditis may be either acute or chronic. Acute infections can reach the thyroid via hematogenous spread or through direct seeding of the gland, such as via a fistula from the piriform sinus adjacent to the larynx. Other infections of the thyroid, including mycobacterial, fungal, and Pneumocystis infections, are more chronic and frequently occur in immunocompromised patients. Whatever the cause, the inflammatory involvement may cause sudden onset of neck pain and tenderness in the area of the gland and is accompanied by fever, chills, and other signs of infection. Infectious thyroiditis can be self-limited or can be controlled with appropriate therapy. Thyroid function is usually not significantly affected, and there are few residual effects except for possible small foci of scarring. This section focuses on the more common and clinically significant types of thyroiditis: (1) Hashimoto thyroiditis (or chronic lymphocytic thyroiditis), (2) subacute granulomatous thyroiditis, and (3) subacute lymphocytic thyroiditis.

HASHIMOTO THYROIDITIS

Hashimoto thyroiditis (or chronic lymphocytic thyroiditis) is the most common cause of hypothyroidism in areas of the world where iodine levels are sufficient. It is characterized by gradual thyroid failure because of autoimmune destruction of the thyroid gland. The name Hashimoto thyroiditis is derived from the 1912 report by Hashimoto describing patients with goiter and intense lymphocytic infiltration of the thyroid (struma lymphomatosa). This disorder is most prevalent between 45 and 65 years of age and is more common in women than in men, with a female predominance of 10:1 to 20:1. Although it is primarily a disease of older women, it can occur in children and is a major cause of nonendemic goiter in children.

Epidemiologic studies have demonstrated a significant genetic component to Hashimoto thyroiditis, although, as in most other autoimmune disorders, the pattern of inheritance is non-Mendelian and likely to be influenced by subtle variations in the functions of multiple genes. The concordance rate in monozygotic twins is 30% to 60%, and up to 50% of asymptomatic first-degree relatives of Hashimoto patients demonstrate circulating antithyroid antibodies.[15] Several chromosomal abnormalities have been associated with thyroid autoimmunity. For example, adults with Turner syndrome (see Chapter 5 ) have a high prevalence of circulating antithyroid antibodies, and a substantial minority (∼20%) develops subclinical or clinical hypothyroidism that is indistinguishable from Hashimoto thyroiditis. Similarly, adults with trisomy 21 (Down syndrome, see Chapter 5 ) are also at an increased risk for developing Hashimoto thyroiditis and hypothyroidism. There are reports that polymorphisms in the HLA locus, specifically the HLA-DR3 and HLA-DR5 alleles, are linked to Hashimoto thyroiditis, but the association is weak. Finally, genomewide linkage analyses in families with

Hashimoto thyroiditis have provided evidence for several susceptibility loci, such as on chromosomes 6p and 12q, that may harbor genes predisposing to this disorder.[16]

Pathogenesis.

Hashimoto thyroiditis is an autoimmune disease in which the immune system reacts against a variety of thyroid antigens. The overriding feature of Hashimoto thyroiditis is progressive depletion of thyroid epithelial cells (thyrocytes), which are gradually replaced by mononuclear cell infiltration and fibrosis. Multiple immunologic mechanisms may contribute to the death of thyrocytes ( Fig. 24-9 ).[17] [18] Sensitization of autoreactive CD4+ T-helper cells to thyroid antigens appears to be the initiating event. The effector mechanisms for thyrocyte death include the following:

• CD8+ cytotoxic T cell-mediated cell death: CD8+ cytotoxic T cells may cause thyrocyte destruction by one of two pathways: exocytosis of perforin/granzyme granules or engagement of death receptors, specifically CD95 (also known as Fas) on the target cell ( Chapter 6 ). • Cytokine-mediated cell death: CD4+ T cells produce inflammatory cytokines such as IFN-γ in the immediate thyrocyte milieu, with resultant recruitment and activation of macrophages and damage to follicles. • Binding of antithyroid antibodies (anti-TSH receptor antibodies, antithyroglobulin, and antithyroid peroxidase antibodies) followed by antibody-dependent cell-mediated cytotoxicity (ADCC) ( Chapter 6 ).

Morphology.

The thyroid is often diffusely enlarged, although more localized enlargement may be seen in some cases. The capsule is intact, and the gland is well demarcated from adjacent structures. The cut surface is pale, yellow-tan, firm, and somewhat nodular. Microscopic examination reveals extensive infiltration of the parenchyma by a mononuclear inflammatory infiltrate containing small lymphocytes, plasma cells, and well-developed germinal centers ( Fig. 24-10 ). The thyroid follicles are atrophic and are lined in many areas by epithelial cells distinguished by the presence of abundant eosinophilic, granular cytoplasm, termed Hürthle cells. This is a metaplastic response of the normally low cuboidal follicular epithelium to ongoing injury. In fine-needle aspiration biopsies, the presence of Hürthle cells in conjunction with a heterogeneous population of lymphocytes is characteristic of Hashimoto thyroiditis. In "classic" Hashimoto thyroiditis, interstitial connective tissue is increased and may be abundant. A fibrous variant is

1170

Figure 24-9 Pathogenesis of Hashimoto thyroiditis. Three proposed models for mechanism of thyrocyte destruction in Hashimoto disease. Sensitization of autoreactive CD4+ T cells to thyroid antigens appears to be the initiating event for all three mechanisms of thyroid cell death. See the text for details.

characterized by severe thyroid follicular atrophy and dense "keloid-like" fibrosis, with broad bands of acellular collagen encompassing residual thyroid tissue. Unlike Reidel thyroiditis (see below), the fibrosis does not extend beyond the capsule of the gland. The remnant thyroid parenchyma demonstrates features of chronic lymphocytic thyroiditis.

Clinical Course.

Hashimoto thyroiditis comes to clinical attention as painless enlargement of the thyroid, usually associated with some degree of hypothyroidism, in a middle-aged woman. The enlargement of the gland is usually symmetric and diffuse, but in some cases, it may be sufficiently localized to raise a suspicion of neoplasm. In the usual clinical course,

Figure 24-10 Hashimoto thyroiditis. The thyroid parenchyma contains a dense lymphocytic infiltrate with germinal centers. Residual thyroid follicles lined by deeply eosinophilic Hürthle cells are also seen.

hypothyroidism develops gradually. In some cases, however, it may be preceded by transient thyrotoxicosis caused by disruption of thyroid follicles, with secondary release of thyroid hormones ("hashitoxicosis"). During this phase, free T4 and T3 levels are elevated, TSH is diminished, and radioactive iodine uptake is decreased. As hypothyroidism supervenes, T4 and T3 levels progressively fall, accompanied by a compensatory increase in TSH. Patients with Hashimoto thyroiditis are at increased risk for developing other concomitant autoimmune diseases, both endocrine (type 1 diabetes, autoimmune adrenalitis), and nonendocrine (systemic lupus erythematosus, myasthenia gravis, and Sjögren syndrome; see Chapter 6 ), and also at increased risk for the development of B-cell non-Hodgkin lymphomas. However, there is no established risk for developing thyroid epithelial neoplasms.

SUBACUTE (GRANULOMATOUS) THYROIDITIS

Subacute thyroiditis, which is also referred to as granulomatous thyroiditis or De Quervain thyroiditis, occurs much less frequently than does Hashimoto disease. The disorder is most common between the ages of 30 and 50 and, like other forms of thyroiditis, affects women considerably more often than men (3:1 to 5:1).

Pathogenesis.

Subacute thyroiditis is believed to be caused by a viral infection or a postviral inflammatory process. The majority of patients have a history of an upper respiratory infection just before the onset of thyroiditis. The disease has a seasonal incidence, with occurrences peaking in the summer, and clusters of cases have been reported in association with coxsackievirus, mumps, measles, adenovirus, and other viral illnesses. Although the pathogenesis of the disease is unclear, one model suggests that it results from a viral infection that provides an antigen, either viral or a thyroid antigen that is

1171

released secondary to virus-induced host tissue damage. This antigen stimulates cytotoxic T lymphocytes, which then damage thyroid follicular cells. In contrast to autoimmune thyroid disease, the immune response is virus-initiated and not self-perpetuating, so the process is limited.

Morphology.

The gland may be unilaterally or bilaterally enlarged and firm, with an intact capsule. It may be slightly adherent to surrounding structures. On cut section, the involved areas are firm and yellow-white and stand out from the more rubbery, normal brown thyroid substance. Histologically, the changes are patchy and depend on the stage of the disease. Early in the active inflammatory phase, scattered follicles may be entirely disrupted and replaced by neutrophils forming microabscesses. Later, the more characteristic features appear in the form of aggregations of lymphocytes, histiocytes, and plasma cells about collapsed and damaged thyroid follicles. Multinucleate giant cells enclose naked pools or fragments of colloid ( Fig. 24-11 ), hence the designation granulomatous thyroiditis. In later stages of the disease, a chronic inflammatory infiltrate and fibrosis may replace the foci of injury. Different histologic stages are sometimes found in the same gland, suggesting waves of destruction over a period of time.

Clinical Course.

The presentation of subacute thyroiditis may be sudden or gradual. It is characterized by pain in the neck, which may radiate to the upper neck, jaw, throat, or ears, particularly when swallowing. Fever, fatigue, malaise, anorexia, and myalgia accompany a variable enlargement of the thyroid. The resultant thyroid inflammation and hyperthyroidism are transient, usually diminishing in 2 to 6 weeks, even if the patient is not treated. It may be followed by a period of transient, usually asymptomatic hypothyroidism lasting from 2 to 8 weeks, but recovery is virtually always complete.

The transient hyperthyroidism, as in other cases of thyroiditis, is due to disruption of thyroid follicles and release of excessive thyroid hormone. Nearly all patients have high serum T4 and T3 levels and low serum TSH levels. Radioactive

Figure 24-11 Subacute thyroiditis. The thyroid parenchyma contains a chronic inflammatory infiltrate with a multinucleate giant cell (above left) and a colloid follicle (bottom right).

iodine uptake is low because of suppression of TSH. The serum T4 and T3 levels are only modestly elevated. However, unlike in hyperthyroid states such as Graves disease, radioactive iodine uptake is diminished. After recovery, generally in 6 to 8 weeks, normal thyroid function returns.

SUBACUTE LYMPHOCYTIC (PAINLESS) THYROIDITIS

Subacute lymphocytic thyroiditis, which is also referred to as painless thyroiditis or silent thyroiditis, is an uncommon cause of hyperthyroidism. It usually comes to clinical attention because of mild hyperthyroidism, goitrous enlargement of the gland, or both. Although it can occur at any age, it is most often seen in middle-aged adults and is more common in women, especially during the postpartum period (postpartum thyroiditis), than in men.[19] Depending on the study, the frequency of this form of thyroiditis varies considerably, from 1% to about 10% of cases of hyperthyroid patients. The pathogenesis of this disorder is unknown. An autoimmune basis has been suggested because some patients have elevated levels of antibodies to thyroglobulin and thyroid peroxidase or a family history of thyroid autoimmune disease, and occasionally the disease evolves into overt chronic autoimmune thyroiditis several years later. There is no evidence that points toward a particular viral or other agent.

Morphology.

Except for possible mild symmetric enlargement, the thyroid appears normal on gross inspection. The most specific histologic features consist of lymphocytic infiltration with hyperplastic germinal centers within the thyroid parenchyma and patch disruption and collapse of thyroid follicles. Unlike in Hashimoto thyroiditis, fibrosis and Hürthle cell metaplasia are not commonly seen.

Clinical Course.

The principal clinical manifestation of painless thyroiditis is hyperthyroidism. Symptoms usually develop over 1 to 2 weeks and last from 2 to 8 weeks before subsiding. The patient may have any of the common findings of hyperthyroidism (e.g., palpitations, tachycardia, tremor, weakness, and fatigue). The thyroid gland is not usually tender but is minimally and diffusely enlarged. Infiltrative ophthalmopathy and other manifestations of Graves disease (see below) are not present. Patients with one episode of postpartum thyroiditis are at an increased risk of recurrence following subsequent pregnancies. A minority of affected individuals eventually progress to hypothyroidism. Some patients have no signs or symptoms, and the disorder is detected incidentally during routine thyroid testing.

Laboratory findings during periods of thyrotoxicosis include elevated levels of T4 and T3 and depressed levels of TSH.

Other, less common forms of thyroiditis include Riedel thyroiditis, a rare disorder of unknown etiology characterized by extensive fibrosis involving the thyroid and contiguous neck structures. The presence of a hard and fixed thyroid mass clinically simulates a thyroid carcinoma. It may be associated with idiopathic fibrosis in other sites in the body, such as the retroperitoneum. The presence of circulating antithyroid antibodies in most patients suggests an autoimmune etiology. Palpation

1172

thyroiditis, caused by vigorous clinical palpation of the thyroid gland, results in multifocal follicular disruption associated with chronic inflammatory cells and occasional giant cell formation. Unlike De Quervain thyroiditis, abnormalities of thyroid function are not present, and this is usually an incidental finding in specimens resected for other reasons.

Graves Disease

Graves reported in 1835 his observations of a disease characterized by "violent and long continued palpitations in females" associated with enlargement of the thyroid gland. Graves disease is the most common cause of endogenous hyperthyroidism. It is characterized by a triad of clinical findings:

1. Hyperthyroidism owing to hyperfunctional, diffuse enlargement of the thyroid 2. Infiltrative ophthalmopathy with resultant exophthalmos 3. Localized, infiltrative dermopathy, sometimes called pretibial myxedema, which

is present in a minority of patients

Graves disease has a peak incidence between the ages of 20 and 40, women being affected up to seven times more frequently than men. This disorder is said to be present in 1.5% to 2.0% of women in the United States. Genetic factors are important in the

etiology of Graves disease. An increased incidence of Graves disease occurs among family members of affected patients, and the concordance rate in monozygotic twins is as high as 60%. A recurring theme, as with other autoimmune disorders, is a genetic susceptibility to Graves disease associated with the presence of certain major histocompatibility haplotypes, specifically HLA-B8 and -DR3. Polymorphisms in the cytotoxic T-lymphocyte-associated-4 (CTLA-4) gene are also linked to Graves disease. [20] [21] Recall that the HLA proteins are a critical component of antigen presentation to T cells, while CTLA-4 is an inhibitory receptor that prevents T-cell responses to self-antigens ( Chapter 6 ). Genomewide linkage analyses have revealed additional susceptibility loci localized to chromosome 6p (also linked to Hashimoto thyroiditis) and to chromosome 20q, among others.[16]

Pathogenesis.

Graves disease is an autoimmune disorder in which a variety of antibodies may be present in the serum, including antibodies to the TSH receptor, thyroid peroxisomes, and thyroglobulin. Of these, autoantibodies to the TSH receptor are central to disease pathogenesis, although the specific effects of the antibodies vary depending on which TSH receptor epitope they are directed against:

• Thyroid-stimulating immunoglobulin (TSI): Almost 50 years ago, serum from patients with Graves disease was found to contain a long-acting thyroid stimulator (LATS), so named because it stimulated thyroid function more slowly than TSH. LATS proved to be an IgG antibody that binds to the TSH receptor and mimics the action of TSH, stimulating adenyl cyclase, with resultant increased release of thyroid hormones. Almost all patients with Graves disease have detectable levels of this autoantibody to the TSH receptor. TSI is relatively specific for Graves disease, in contrast to thyroglobulin and thyroid peroxidase antibodies. • Thyroid growth-stimulating immunoglobulins (TGI): Also directed against the TSH receptor, thyroid growth-stimulating immunoglobulins have been implicated in the proliferation of thyroid follicular epithelium. • TSH-binding inhibitor immunoglobulins (TBII): These anti-TSH receptor antibodies prevent TSH from binding normally to its receptor on thyroid epithelial cells. In so doing, some forms of TSH-binding inhibitor immunoglobulins mimic the action of TSH, resulting in the stimulation of thyroid epithelial cell activity, whereas other forms may actually inhibit thyroid cell function. It is not unusual to find the coexistence of stimulating and inhibiting immunoglobulins in the serum of the same patient, a finding that could explain why some patients with Graves disease spontaneously develop episodes of hypothyroidism.

The key role of anti-TSH receptor antibodies in the pathogenesis of hyperthyroidism is underscored by animal models that recapitulate human Graves disease.[22] Immunization of mice with the TSH receptor results in generation of antibodies that cause thyroid stimulation, thyroid enlargement with lymphocytic infiltration, elevated thyroxine levels, and, in a subset of mice, ocular signs reminiscent of Graves ophthalmopathy (see below). Similar to the human disease, a gender predisposition as well as a genetic predisposition are seen in these animal models—females are affected more frequently than males and

only certain inbred strains of mice demonstrate signs and symptoms of the disease. The trigger for the initiation of the autoimmune reaction in Graves disease remains uncertain, although the underlying mechanism is likely to be breakdown in helper T-cell tolerance, resulting in the production of anti-TSH autoantibodies.

A T cell-mediated autoimmune phenomenon also plays a role in the development of the infiltrative ophthalmopathy that is characteristic of Graves disease.[23] In Graves ophthalmopathy, the volume of the retro-orbital connective tissues and extraocular muscles is increased owing to several causes, including (1) marked infiltration of the retro-orbital space by mononuclear cells, predominantly T cells; (2) inflammatory edema and swelling of extraocular muscles; (3) accumulation of extracellular matrix components, specifically hydrophilic glycosaminoglycans (GAGs) such as hyaluronic acid and chondroitin sulfate; and (4) increased numbers of adipocytes (fatty infiltration). These changes displace the eyeball forward and can interfere with the function of the extraocular muscles. Recent evidence suggests that orbital preadipocyte fibroblasts express the TSH receptor and thus become targets of an autoimmune attack. T cells reactive against these fibroblasts secrete cytokines, which stimulate fibroblast proliferation and synthesis of extracellular matrix proteins (GAGs) and increase surface TSH receptor expression, perpetuating the autoimmune response.[22] The result is progressive infiltration of the retro-orbital space and ophthalmopathy.

Autoimmune disorders of the thyroid thus span a continuum in which Graves disease, characterized by hyperfunction of the thyroid, lies at one extreme and Hashimoto disease, manifesting as hypothyroidism, occupies the other end. Sometimes hyperthyroidism may supervene on pre-existing Hashimoto thyroiditis (hashitoxicosis); at other times, patients with Graves disease may spontaneously develop thyroid hypofunction; occasionally, there are families with coexistence of Hashimoto and Graves disease within the affected kindred. Not surprisingly, there is also an element of histologic overlap between the autoimmune thyroid disorders (most characteristically,

1173

prominent intrathyroidal lymphoid cell infiltrates with germinal center formation; see below). In both disorders, the frequency of other autoimmune diseases, such as systemic lupus erythematosus, pernicious anemia, type I diabetes, and Addison disease, is increased.

Morphology.

The thyroid gland is usually symmetrically enlarged because of diffuse hypertrophy and hyperplasia of thyroid follicular epithelial cells. Increases in weight to over 80 gm are not uncommon. The gland is usually smooth and soft, and its capsule is intact. On cut section, the parenchyma has a soft, meaty appearance resembling normal muscle. Histologically, the dominant feature is too many cells. The follicular epithelial cells in untreated cases are tall and more crowded than usual. This crowding often results in the

formation of small papillae, which project into the follicular lumen and encroach on the colloid, sometimes filling the follicles ( Fig. 24-12 ). Such papillae lack fibrovascular cores, in contrast to those of papillary carcinoma (see below). The colloid within the follicular lumen is pale, with scalloped margins. Lymphoid infiltrates, consisting predominantly of T cells, with fewer B cells and mature plasma cells, are present throughout the interstitium; germinal centers are common.

Preoperative therapy alters the morphology of the thyroid in Graves disease. Preoperative administration of iodine causes involution of the epithelium and the accumulation of colloid by blocking thyroglobulin secretion. Treatment with the antithyroid drug propylthiouracil exaggerates the epithelial hypertrophy and hyperplasia by stimulating TSH secretion. Thus, in pre-treated patients it is impossible from histologic examination of surgical specimens to evaluate the functional activity of the gland.

Changes in extrathyroidal tissue include generalized lymphoid hyperplasia. The heart may be hypertrophied, and ischemic changes may be present, particularly in patients with preexisting coronary artery disease. In patients with ophthalmopathy, the

Figure 24-12 Diffusely hyperplastic thyroid in a case of Graves disease. The follicles are lined by tall, columnar epithelium. The crowded, enlarged epithelial cells project into the lumens of the follicles. These cells actively resorb the colloid in the centers of the follicles, resulting in the scalloped appearance of the edges of the colloid.

tissues of the orbit are edematous because of the presence of hydrophilic mucopolysaccharides. In addition, there is infiltration by lymphocytes and fibrosis. Orbital muscles are edematous initially but may undergo fibrosis late in the course of the disease. The dermopathy, if present, is characterized by thickening of the dermis due to deposition of glycosaminoglycans and lymphocyte infiltration.

Clinical Course.

The clinical findings in Graves disease include changes referable to thyrotoxicosis as well as those associated uniquely with Graves disease: diffuse hyperplasia of the thyroid,

ophthalmopathy, and dermopathy. The degree of thyrotoxicosis varies from case to case and is sometimes less conspicuous than other manifestations of the disease. Diffuse enlargement of the thyroid is present in all cases of Graves disease. The thyroid enlargement may be accompanied by increased flow of blood through the hyperactive gland, often producing an audible bruit. Sympathetic overactivity produces a characteristic wide, staring gaze and lid lag. The ophthalmopathy of Graves disease results in abnormal protrusion of the eyeball (exophthalmos). The extraocular muscles are often weak. The exophthalmos may persist or progress despite successful treatment of the thyrotoxicosis, sometimes resulting in corneal injury. The infiltrative dermopathy, or pretibial myxedema, is most common in the skin overlying the shins, where it presents as scaly thickening and induration of the skin. However, it is present only in a minority of patients. The skin lesions may be slightly pigmented papules or nodules and often have an orange peel texture.

Laboratory findings in Graves disease include elevated free T4 and T3 levels and depressed TSH levels. Because of ongoing stimulation of the thyroid follicles by thyroid-stimulating immunoglobulins, radioactive iodine uptake is increased, and radioiodine scans show a diffuse uptake of iodine.

Treatment of Graves disease consists of decreasing the symptoms of hyperthyroidism that are induced by increased β-adrenergic tone (e.g., tachycardia, palpitations, tremulousness, and anxiety) and measures aimed at decreasing thyroid hormone synthesis, such as the administration of thionamides (e.g., propylthiouracil), radioiodine ablation, and surgical intervention.

Diffuse and Multinodular Goiters

Enlargement of the thyroid, or goiter, is the most common manifestation of thyroid disease. Diffuse and multinodular goiters reflect impaired synthesis of thyroid hormone, most often caused by dietary iodine deficiency. Impairment of thyroid hormone synthesis leads to a compensatory rise in the serum TSH level, which, in turn, causes hypertrophy and hyperplasia of thyroid follicular cells and, ultimately, gross enlargement of the thyroid gland. The compensatory increase in functional mass of the gland is able to overcome the hormone deficiency, ensuring an euthyroid metabolic state in the vast majority of individuals. If the underlying disorder is sufficiently severe (e.g., a congenital biosynthetic defect or endemic iodine deficiency, see below), the compensatory responses may be inadequate to overcome the impairment in hormone synthesis, resulting in goitrous hypothyroidism. The degree of thyroid

1174

enlargement is proportional to the level and duration of thyroid hormone deficiency.

DIFFUSE NONTOXIC (SIMPLE) GOITER

Diffuse nontoxic (simple) goiter specifies a form of goiter that diffusely involves the entire gland without producing nodularity. Because the enlarged follicles are filled with colloid, the term colloid goiter has been applied to this condition. This disorder occurs in both an endemic and a sporadic distribution.

Endemic goiter occurs in geographic areas where the soil, water, and food supply contain only low levels of iodine. The term endemic is used when goiters are present in more than 10% of the population in a given region. Such conditions are particularly common in mountainous areas of the world, including the Alps, Andes, and Himalayas, where iodine deficiency is widespread. The lack of iodine leads to decreased synthesis of thyroid hormone and a compensatory increase in TSH, leading to follicular cell hypertrophy and hyperplasia and goitrous enlargement. With increasing dietary iodine supplementation, the frequency and severity of endemic goiter have declined significantly.

Variations in the prevalence of endemic goiter in regions with similar levels of iodine deficiency point to the existence of other causative influences, particularly dietary substances, referred to as goitrogens. The ingestion of substances that interfere with thyroid hormone synthesis at some level, such as excessive calcium and vegetables belonging to the Brassica and Cruciferae families (e.g., cabbage, cauliflower, Brussels sprouts, turnips, and cassava), has been documented to be goitrogenic. Native populations subsisting on cassava root are particularly at risk. Cassava contains a thiocyanate that inhibits iodide transport within the thyroid, worsening any possible concurrent iodine deficiency.

Sporadic goiter occurs less frequently than does endemic goiter. There is a striking female preponderance and a peak incidence at puberty or in young adult life. Sporadic goiter can be caused by a number of conditions, including the ingestion of substances that interfere with thyroid hormone synthesis. In other instances, goiter may result from hereditary enzymatic defects that interfere with thyroid hormone synthesis, all transmitted as autosomal-recessive conditions (dyshormonogenetic goiter; see above). In most cases, however, the cause of sporadic goiter is not apparent.

Morphology.

Two phases can be identified in the evolution of diffuse nontoxic goiter: the hyperplastic phase and the phase of colloid involution. In the hyperplastic phase, the thyroid gland is diffusely and symmetrically enlarged, although the increase is usually modest, and the gland rarely exceeds 100 to 150 gm. The follicles are lined by crowded columnar cells, which may pile up and form projections similar to those seen in Graves disease. The accumulation is not uniform throughout the gland, and some follicles are hugely distended, whereas others remain small. If dietary iodine subsequently increases or if the

demand for thyroid hormone decreases, the stimulated follicular epithelium involutes to form an enlarged, colloid-rich gland (colloid goiter). In these cases, the cut surface of the thyroid is usually brown, somewhat glassy, and translucent. Histologically, the follicular epithelium is flattened and cuboidal, and colloid is abundant during periods of involution.

Clinical Course.

The vast majority of patients with simple goiters are clinically euthyroid. Therefore, the clinical manifestations are primarily related to mass effects from the enlarged thyroid gland (discussed in detail with multinodular goiter; see below). Although serum T3 and T4

levels are normal, the serum TSH is usually elevated or at the upper range of normal, as is expected in marginally euthyroid individuals. In children, dyshormonogenetic goiter, caused by a congenital biosynthetic defect, may induce cretinism.

MULTINODULAR GOITER

With time, recurrent episodes of hyperplasia and involution combine to produce a more irregular enlargement of the thyroid, termed multinodular goiter. Virtually all long-standing simple goiters convert into multinodular goiters. They may be nontoxic or may induce thyrotoxicosis (toxic multinodular goiter). Multinodular goiters produce the most extreme thyroid enlargements and are more frequently mistaken for neoplastic involvement than any other form of thyroid disease. Because they derive from simple goiter, they occur in both sporadic and endemic forms, having the same female-to-male distribution and presumably the same origins but affecting older individuals because they are late complications.

It is believed that multinodal goiters may arise because of variations among follicular cells in responses to external stimuli, such as trophic hormones. If some cells in a follicle have a growth advantage, perhaps because of intrinsic genetic abnormalities similar to those that give rise to adenomas, those cells will develop into clones of proliferating cells. This may result in the formation of a nodule whose continued growth could even be autonomous, without the external stimulus. Consistent with this model, both polyclonal and monoclonal nodules coexist within the same multinodular goiter, the latter presumably having arisen owing to the acquisition of a genetic abnormality favoring growth.[24] Not surprisingly, mutations in proteins of the TSH-signaling pathway that lead to constitutive activation of this pathway have been identified in a subset of autonomous thyroid nodules. (TSH signaling pathway mutations and their implications are discussed in the section on follicular adenomas.)[25] With uneven follicular hyperplasia, generation of new follicles, and uneven accumulation of colloid, tensions and stresses are produced that lead to rupture of follicles and vessels followed by hemorrhages, scarring, and sometimes calcifications. The scarring adds to the tensions, and in this cyclical manner, nodularity appears. Moreover, the preexisting stromal framework of the gland may more or less enclose areas of expanded parenchyma, contributing to the nodularity.

Morphology.

Multinodular goiters are multilobulated, asymmetrically enlarged glands that can achieve a weight of more than 2000 gm ( Fig. 24-13 ). The pattern of enlargement is quite unpredictable and

1175

Figure 24-13 Nodular goiter. The gland is coarsely nodular and contains areas of fibrosis and cystic change.

may involve one lobe far more than the other, producing lateral pressure on midline structures, such as the trachea and esophagus. In other instances, the goiter grows behind the sternum and clavicles to produce the so-called intrathoracic or plunging goiter. Occasionally, most of it is hidden behind the trachea and esophagus; in other instances, one nodule may so stand out as to impart the clinical appearance of a solitary nodule. On cut section, irregular nodules containing variable amounts of brown, gelatinous colloid are present. Regressive changes occur frequently, particularly in older lesions, and include areas of hemorrhage, fibrosis, calcification, and cystic change. The microscopic appearance includes colloid-rich follicles lined by flattened, inactive epithelium and areas of follicular epithelial hypertrophy and hyperplasia, accompanied by the degenerative changes noted previously.

Clinical Course.

The dominant clinical features of goiter are those caused by the mass effects of the enlarged gland. In addition to the obvious cosmetic effects of a large neck mass, goiters may cause airway obstruction, dysphagia, and compression of large vessels in the neck and upper thorax. Most patients are euthyroid, but in a substantial minority of patients, a hyperfunctioning nodule may develop within a long-standing goiter, resulting in hyperthyroidism (toxic multinodular goiter). This condition, known as Plummer syndrome, is not accompanied by the infiltrative ophthalmopathy and dermopathy of Graves disease. As was previously mentioned, goiter may be associated with clinical

evidence of hypothyroidism in specific clinical settings. Radioiodine uptake is uneven, reflecting varied levels of activity in different regions. Hyperfunctioning nodules concentrate radioiodine and appear "hot." Goiters are also of clinical significance because of their ability to mask or to mimic neoplastic diseases arising in the thyroid.

Neoplasms of the Thyroid

The solitary thyroid nodule is a palpably discrete swelling within an otherwise apparently normal thyroid gland. The estimated incidence of solitary palpable nodules in the adult population of the United States varies between 1% and 10%, although it is significantly higher in endemic goitrous regions. Single nodules are about four times more common in women than in men. The incidence of thyroid nodules increases throughout life.

From a clinical standpoint, the possibility of neoplastic disease is of major concern in patients who present with thyroid nodules. Fortunately, the overwhelming majority of solitary nodules of the thyroid prove to be localized, non-neoplastic conditions (e.g., nodular hyperplasia, simple cysts, or foci of thyroiditis) or benign neoplasms such as follicular adenomas. In fact, benign neoplasms outnumber thyroid carcinomas by a ratio of nearly 10:1. Carcinomas of the thyroid are thus uncommon, accounting for well under 1% of solitary thyroid nodules and representing about 15,000 new cancer cases each year. Moreover, as will be seen subsequently, most are indolent, permitting a 90% survival at 20 years. Several clinical criteria might provide a clue to the nature of a given thyroid nodule:

• Solitary nodules, in general, are more likely to be neoplastic than are multiple nodules. • Nodules in younger patients are more likely to be neoplastic than are those in older patients. • Nodules in males are more likely to be neoplastic than are those in females. • A history of radiation treatment to the head and neck region is associated with an increased incidence of thyroid malignancy. • Nodules that take up radioactive iodine in imaging studies (hot nodules) are more likely to be benign than malignant.

Such general trends and statistics, however, are of little significance in the evaluation of a given patient, in whom the timely recognition of a malignancy, however uncommon, can be life-saving. Ultimately, it is the morphologic evaluation of a given thyroid nodule, in the form of fine-needle aspiration biopsy and histologic study of surgically resected thyroid parenchyma, that provides the most definitive information about its nature. In the

following sections, we consider the major thyroid neoplasms, including adenoma and carcinoma in its various forms.

ADENOMAS

Adenomas of the thyroid are typically discrete, solitary masses. With rare exception, they are derived from follicular epithelium and so might all be called follicular adenomas. A variety of terms have been proposed for classifying adenomas on the basis of degree of follicle formation and the colloid content of the follicles. Simple colloid adenomas (macrofollicular adenomas), a common form, resemble normal thyroid tissue; others recapitulate stages in the embryogenesis of the normal thyroid (fetal or microfollicular, embryonal or trabecular). There is limited utility in these classifications because mixed patterns are common, and most of these benign tumors are nonfunctional. Clinically, follicular adenomas can be difficult to distinguish from dominant nodules of follicular hyperplasia or from the less common follicular carcinomas. Numerous studies have made it clear that adenomas are not forerunners of cancer except in rare instances. Although the vast majority of adenomas are nonfunctional, a small proportion

1176

produce thyroid hormones and cause clinically apparent thyrotoxicosis. Hormone production in functional adenomas ("toxic adenomas") occurs independent of TSH stimulation and represents another example of thyroid autonomy, analogous to toxic multinodular goiters.

Pathogenesis.

The TSH receptor signaling pathway plays an important role in the pathogenesis of toxic adenomas. Activating ("gain of function") somatic mutations in one of two components of this signaling system—most often the TSH receptor itself or the α-subunit of Gs —cause chronic overproduction of cAMP, generating cells that acquire a growth advantage (see Fig. 24-3 ).[26] This results in clonal expansion of follicular epithelial cells that can autonomously produce thyroid hormone and cause symptoms of thyroid excess. Overall, mutations leading to constitutive activation of the cAMP pathway appear to be the cause of a proportion (10% to 75%) of autonomously functioning thyroid adenomas. However, the molecular pathogenesis of a significant proportion of thyroid tumors remains to be defined, especially the pathogenesis of nonfunctioning adenomas.

Morphology.

The typical thyroid adenoma is a solitary, spherical, encapsulated lesion that is well demarcated from the surrounding thyroid parenchyma ( Fig. 24-14 ). Follicular adenomas average about 3 cm in diameter, but some are smaller and others are much larger (up to 10 cm in diameter). In freshly resected specimens, the adenoma bulges from the cut surface and compresses the adjacent thyroid. The color ranges from gray-white to red-brown, depending on the cellularity of the adenoma and its colloid content. The

neoplastic cells are demarcated from the adjacent parenchyma by a well-defined, intact capsule. These features are important in making the distinction from multinodular goiters, which contain multiple nodules on their cut surface (even though the patient may present clinically with a solitary dominant nodule), produce less compression of the adjacent thyroid parenchyma, and lack a well-formed capsule. Areas of hemorrhage, fibrosis, calcification, and cystic change, similar to those encountered in multinodular goiters, are common in follicular adenomas, particularly within larger lesions.

Microscopically, the constituent cells often form uniform-appearing follicles that contain colloid ( Fig. 24-15 ). The follicular growth pattern within the adenoma is usually quite distinct from the adjacent non-neoplastic thyroid. This is another feature distinguishing

Figure 24-14 Follicular adenoma of the thyroid. A solitary, well-circumscribed nodule is seen.

Figure 24-15 Follicular adenoma. The photomicrograph shows well-differentiated follicles resembling normal thyroid parenchyma.

adenomas from multinodular goiters, in which nodular and uninvolved thyroid parenchyma may have similar growth patterns. The epithelial cells composing the follicular adenoma reveal little variation in cell and nuclear morphology. Mitotic figures are rare, and extensive mitotic activity warrants careful examination of the capsule to exclude follicular carcinoma. Similarly, papillary change is not a typical feature of adenomas and, if extensive, should raise the suspicion of an encapsulated papillary carcinoma (see below). Occasionally, the neoplastic cells acquire brightly eosinophilic granular cytoplasm (oxyphil or Hürthle cell change) ( Fig. 24-16 ); the clinical presentation and behavior of a follicular adenoma with oxyphilia (Hürthle cell adenoma) is no different from that of a conventional adenoma. Other variants of follicular adenomas include extensive clear cell change of the cytoplasm (clear cell follicular adenoma) and adenomas with "signet-ring" features (signet-ring cell follicular adenoma). Similar to endocrine tumors at other anatomic sites, even benign follicular adenomas may, on occasion, exhibit focal nuclear pleomorphism, atypia, and prominent nucleoli (endocrine atypia); this by itself does not constitute

Figure 24-16 Hürthle cell tumor. A high-power view showing that the tumor is composed of cells with abundant eosinophilic cytoplasm and small regular nuclei. (Courtesy of Dr. Mary Sunday, Brigham and Women's Hospital, Boston, MA.)

1177

a feature of malignancy. Infrequently, adenomas can demonstrate increased cellularity, more extensive variation in cellular size and nuclear morphology, and even mitotic activity. These adenomas have been called atypical follicular adenomas and warrant careful examination of the tumor capsule to exclude capsular and/or vascular invasion.[27] The hallmark of all follicular adenomas is the presence of an intact, well-formed capsule encircling the tumor. Careful evaluation of the integrity of the capsule is therefore critical in distinguishing follicular adenomas from follicular carcinomas, which demonstrate capsular and/or vascular invasion (see below).

Clinical Features.

Many thyroid adenomas present as a unilateral painless mass, often discovered during a routine physical examination. Larger masses may produce local symptoms, such as difficulty in swallowing.

Most adenomas take up less radioactive iodine than does normal thyroid parenchyma. On radionuclide scanning, therefore, adenomas usually appear as cold nodules relative to the adjacent thyroid tissue. Up to 10% of cold nodules eventually prove to be malignant on histologic analysis. By contrast, malignancy is rare in hot nodules. In a minority of cases, adenomas may be hyperfunctional, producing signs and symptoms of hyperthyroidism (toxic adenomas). On radionuclide imaging, hyperfunctioning adenomas appear hot compared with the paranodular thyroid tissue, which is deprived of thyrotropin stimulation. Hot adenomas occasionally have some dependence on TSH and may be induced to regress by the administration of thyroid hormones, which suppress TSH secretion.

Other techniques used in the preoperative evaluation of suspected adenomas are ultrasonography and fine-needle aspiration biopsy. Owing to the need for evaluating capsular integrity, the definitive diagnosis of adenomas can be made only after careful histologic examination of the resected specimen. Suspected adenomas of the thyroid are therefore removed surgically to exclude malignancy. Thyroid adenomas, including atypical adenomas, have an excellent prognosis and do not recur or metastasize. About 20% of follicular adenomas have point mutations in the RAS family of oncogenes, which have also been identified in 30% to 40% of follicular carcinomas. This finding raises the possibility that some adenomas may progress to carcinomas.

OTHER BENIGN TUMORS

Solitary nodules of the thyroid gland may also prove to be cysts. The great preponderance of these lesions represent cystic degeneration of a follicular adenoma; the remainder probably arise in multinodular goiters. They are often filled with a brown, turbid fluid containing blood, hemosiderin pigment, and cell debris. Additional benign rarities include dermoid cysts, lipomas, hemangiomas, and teratomas (see mainly in infants).

CARCINOMAS

Carcinomas of the thyroid are relatively uncommon in the United States, accounting for about 1.5% of all cancers. Most cases occur in adults, although some forms, particularly papillary carcinomas, may present in childhood. A female predominance has been noted among patients who develop thyroid carcinoma in the early and middle adult years, perhaps related to the expression of estrogen receptors on neoplastic thyroid epithelium. In contrast, cases presenting in childhood and late adult life are distributed equally among males and females. Most thyroid carcinomas are well-differentiated lesions. The major subtypes of thyroid carcinoma and their relative frequencies include the following:

• Papillary carcinoma (75% to 85% of cases)

• Follicular carcinoma (10% to 20% of cases) • Medullary carcinoma (5% of cases) • Anaplastic carcinoma (<5% of cases)

Most thyroid carcinomas are derived from the follicular epithelium, except for medullary carcinomas; the latter are derived from the parafollicular or C cells. Because of the unique clinical and biologic features associated with each variant of thyroid carcinoma, these subtypes are described separately.

Pathogenesis

There are several factors, genetic and environmental, implicated in the pathogenesis of thyroid cancers.

Genetic Factors.

Genetic factors are important in both familial and nonfamilial ("sporadic") forms of thyroid cancer. Familial medullary cancers account for most inherited cases of thyroid cancer. Familial nonmedullary thyroid cancers (papillary and follicular variants) are very rare. Distinct genes are involved in the histologic variants of thyroid cancer.

Follicular Thyroid Carcinomas.

Approximately half of follicular thyroid carcinomas harbor mutations in the RAS family of oncogenes (HRAS, NRAS, and KRAS) ( Chapter 7 ), NRAS mutations being the most common. Recently, a unique translocation has been described between PAX8, a paired homeobox gene that is important in thyroid development (see above), and the peroxisome proliferator-activated receptor γ1 (PPARγ1), a nuclear hormone receptor implicated in terminal differentiation of cells.[28] The PAX8-PPARγ1 fusion is present in approximately one-third of follicular thyroid carcinomas, specifically those cancers with a t(2;3)(q13;p25) translocation, which permits juxtaposition of portions of both genes. Follicular carcinomas appear to arise by at least two distinct and virtually nonoverlapping molecular pathways:[29] Tumors carry either a RAS mutation or a PAX8-PPARγ1 fusion, and rarely are both genetic abnormalities present in the same case. Fewer than 10% of follicular adenomas harbor a PAX8-PPARγ1 fusion transcript, and this translocation has not been documented to date in other thyroid neoplasms. [30]

Papillary Thyroid Carcinomas.

Like follicular thyroid carcinomas, papillary carcinomas also appear to arise by multiple distinct, nonoverlapping molecular pathways. One pathway involves rearrangements of the tyrosine kinase receptors RET or NTRK1 (neurotrophic tyrosine kinase receptor 1) and another involves activating mutations in the BRAF oncogene. A third pathway involves RAS mutations (10% to 20% of papillary carcinomas), suggesting that some of these cancers are related to follicular adenomas. RET, located on chromosome 10q11, and NTRK1, located on chromosome 1q21, belong to the family of receptor tyrosine kinases that transduce extracellular signals for cell growth and differentiation and exert many of their downstream effects through the

1178

ubiquitous MAP kinase signaling pathway ( Chapter 7 ). Neither receptor is normally expressed on the surface of thyroid follicular cells. In papillary thyroid cancers, either a paracentric inversion of chromosome 10 or a reciprocal translocation between chromosomes 10 and 17 places the tyrosine kinase domain of RET under the transcriptional control of constitutively active genes on these two chromosomes. The novel fusion genes that are so formed are known as ret/PTC (ret/papillary thyroid carcinoma) and are present in approximately one-fifth of papillary thyroid cancers.[31] The frequency of ret/PTC rearrangements is significantly higher in papillary cancers arising in children and in the backdrop of radiation exposure. Similarly, paracentric inversions or translocations of NTRK1 that constitutively activate its tyrosine kinase domain are present in 5% to 10% of papillary thyroid cancers.[32] One-third to one-half of papillary thyroid carcinomas harbor an activating mutation in the BRAF gene, which encodes a signaling intermediary in the MAP kinase pathway.[33] [34] Since chromosomal rearrangements of the RET or NTRK1 genes and mutations of BRAF have redundant effects on the thyroid epithelium (recall that both mechanisms result in activation of the MAP kinase signaling pathway), papillary thyroid carcinomas demonstrate either one or the other molecular abnormality, but not both.[35]

Medullary Thyroid Carcinomas.

Medullary carcinomas arise from the parafollicular C cells in the thyroid. Familial medullary thyroid carcinomas occur in multiple endocrine neoplasia type 2 (MEN-2, see below) and are associated with germ-line RET protooncogene mutations that affect residues in the cysteine-rich extracellular or the intracellular tyrosine kinase domains, leading to constitutive activation of the receptor.[36] RET mutations are detectable in approximately 95% of families with MEN-2; in the remaining few cases, the mutations may arise in hard-to-detect promoter sequences or intronic sites. RET mutations are also seen in nonfamilial (sporadic) medullary thyroid cancers.[37] Chromosomal rearrangements involving RET, such as the ret/PTC translocations reported in papillary cancers, are not seen in medullary carcinomas.

Anaplastic Carcinomas.

These highly aggressive and lethal tumors can arise de novo or by "dedifferentiation" of a well-differentiated papillary or follicular carcinoma. Inactivating point mutations in the p53 tumor suppressor gene are rare in well-differentiated thyroid carcinomas but common in anaplastic tumors.[38]

Environmental Factors.

The major risk factor predisposing to thyroid cancer is exposure to ionizing radiation, particularly during the first two decades of life. In the past, radiation therapy was liberally employed in the treatment of a number of head and neck lesions in infants and children, including reactive tonsillar enlargement, acne, and tinea capitis. Up to 9% of people

receiving such treatment during childhood have subsequently developed thyroid malignancies, usually several decades after exposure. The importance of radiation as a risk factor for thyroid carcinoma was highlighted by the increased incidence of papillary thyroid carcinomas in children in the Marshall Islands after atomic bomb testing and, more recently, by the dramatic rise in the incidence of pediatric thyroid carcinoma among children exposed to ionizing radiation after the Chernobyl nuclear disaster in the Ukraine in 1986. More than 400 cases of pediatric thyroid carcinoma have been observed in this region of Belarus between the time of the incident and the present, a number far in excess of the usual incidence for this area.[39] More than half of the children lived in areas that had the highest radiation exposure.

Long-standing multinodular goiter has been suggested as a predisposing factor in some cases, since areas with iodine deficiency-related endemic goiter have a higher prevalence of follicular carcinomas. While most, if not all, thyroid lymphomas arise from pre-existing Hashimoto thyroiditis, there is no conclusive evidence to suggest that thyroiditis is associated with an increased risk of thyroid epithelial carcinomas.

Papillary Carcinoma

Papillary carcinomas are the most common form of thyroid cancer. They occur at any age but most often in the twenties to forties, and account for the majority of thyroid carcinomas associated with previous exposure to ionizing radiation.

Morphology.

Papillary carcinomas are solitary or multifocal lesions. Some tumors may be well-circumscribed and even encapsulated; others may infiltrate the adjacent parenchyma with ill-defined margins. The lesions may contain areas of fibrosis and calcification and are often cystic. On the cut surface, they may appear granular and may sometimes contain grossly discernible papillary foci. The definitive diagnosis of papillary carcinoma can be made only after microscopic examination. The characteristic hallmarks of papillary neoplasms include the following ( Fig. 24-17 ):

• Papillary carcinomas can contain branching papillae having a fibrovascular stalk covered by a single to multiple layers of cuboidal epithelial cells. In most neoplasms, the epithelium covering the papillae consists of well-differentiated, uniform, orderly, cuboidal cells, but at the other extreme are those with fairly anaplastic epithelium showing considerable variation in cell and nuclear morphology. When present, the papillae of papillary carcinoma differ from those seen in areas of hyperplasia. In contrast to hyperplastic papillary lesions, the neoplastic papillae are more complex and have dense fibrovascular cores. • The nuclei of papillary carcinoma cells contain finely dispersed chromatin, which imparts an optically clear or empty appearance, giving rise to the designation ground glass or Orphan Annie eye nuclei. In addition, invaginations of the cytoplasm may in cross-sections give the appearance of intranuclear inclusions ("pseudo-inclusions") or intranuclear grooves. As currently used, the

diagnosis of papillary carcinoma is based on these nuclear features even in the absence of papillary architecture. • Concentrically calcified structures termed psammoma bodies are often present within the lesion, usually within the cores of papillae. These structures are almost never found in follicular and medullary carcinomas, and so, when present, they are a strong indication that the lesion is a papillary carcinoma. It is said that whenever a psammoma body is found within a lymph node or perithyroidal tissues, a hidden papillary carcinoma must be considered.[40]

1179

• Foci of lymphatic invasion by tumor are often present, but involvement of blood vessels is relatively uncommon, particularly in smaller lesions. Metastases to adjacent cervical lymph nodes are estimated to occur in up to half the cases.

Figure 24-17 Papillary carcinoma of the thyroid. A, The macroscopic appearance of a papillary carcinoma with grossly discernible papillary structures. This particular example contains well-formed papillae (B), lined by cells with characteristic empty-appearing nuclei, sometimes termed "Orphan Annie eye" nuclei (C). D, Cells obtained by fine-needle aspiration of a papillary carcinoma. Characteristic intranuclear inclusions are visible in some of the aspirated cells.

There are variant forms of papillary carcinoma that are important to recognize because they can resemble other lesions and have unique clinical features. The encapsulated

variant constitutes about 10% of all papillary neoplasms. It is usually confined to the thyroid gland, is well encapsulated, and rarely presents with vascular or lymph node dissemination, and so it can easily be confused with a benign adenoma. In most cases, this variant has an excellent prognosis.

The follicular variant has the characteristic nuclei of papillary carcinoma but has an almost totally follicular architecture.[41] Grossly, the tumor may be encapsulated, and focally, psammoma bodies may be seen. These follicular variants still behave biologically as usual papillary carcinomas as long as they meet the nuclear criteria for diagnosis of papillary cancers (see above). The true follicular carcinoma, in contrast, lacks these nuclear features, frequently demonstrates capsular and vascular invasion, and has a less favorable prognosis. A differential diagnosis of thyroid lesions with a follicular architecture is summarized in Table 24-4 .

A tall cell variant is marked by tall columnar cells with intensely eosinophilic cytoplasm lining the papillary structures. Typically, the cells are at least twice as tall as they are wide (hence the eponym "tall cell" variant). These tumors tend to occur in older individuals and are usually large with prominent vascular invasion, extrathyroidal extension, and cervical and distant metastases. It has been recently demonstrated that more than half the tall cell variants harbor a ret/PTC translocation that confers greater mitogenic

TABLE 24-4 -- Thyroid Lesions with a Follicular Architecture

Non-Neoplastic

Hyperplastic nodule in goiter

Neoplastic

Follicular adenoma *

Follicular carcinoma *

Follicular variant of papillary carcinoma †

*Differentiating follicular carcinoma from follicular adenoma requires histologic evidence of capsular or blood vessel invasion, or documented metastasis.†The diagnosis of papillary carcinoma is rendered on the presence of characteristic nuclear features, irrespective of the presence or absence of papillae.

1180

potential than the ret/PTC observed in usual papillary thyroid cancers. The presence of this genetic abnormality might result in more aggressive behavior.[42] [43]

An unusual diffuse sclerosing variant of papillary carcinoma occurs in younger individuals, including children. These tumors do not present with a mass, but rather with a bilateral goiter. There is a characteristic "gritty" sensation to the cut surface of the lesion due to the presence of abundant psammoma bodies. The tumor demonstrates a prominent papillary growth pattern, intermixed with solid areas containing nests of squamous cells (squamous morules). The neoplastic cells exhibit classic nuclear features of a papillary neoplasm. As the name suggests, there is extensive, diffuse fibrosis throughout the thyroid gland, often associated with a prominent lymphocytic infiltrate, simulating Hashimoto thyroiditis. The neoplastic cells have a peculiar propensity to invade intrathyroidal lymphatic channels; hence, nodal metastases are present in almost all cases.

Hyalinizing trabecular tumors, a group that includes both adenomas and carcinomas, have recently been reconsidered as a variant of papillary carcinomas, based on the presence of ret/PTC gene rearrangements in 30% to 60% of these tumors.[44] They are characterized by an "organoid" growth pattern, with nests and trabeculae of elongated tumor cells within a fibrovascular stroma; at first glance, the tumor may resemble an extra-adrenal paraganglioma (see below). Both intracellular and extracellular hyalinization are prominent and confer a pink hue on the tumor on low-power microscopic examination. The nuclear features resemble those seen in classic papillary carcinomas, and psammoma bodies may be present. Hyalinizing trabecular adenomas are well encapsulated, while carcinomas demonstrate capsular and/or vascular invasion.

Clinical Course.

Most papillary carcinomas present as asymptomatic thyroid nodules, but the first manifestation may be a mass in a cervical lymph node. Interestingly, the presence of isolated cervical nodal metastases does not appear to have a significant influence on the generally good prognosis of these lesions. The carcinoma, which is usually a single nodule, moves freely during swallowing and is not distinguishable from a benign nodule. Hoarseness, dysphagia, cough, or dyspnea suggests advanced disease. In a minority of patients, hematogenous metastases are present at the time of diagnosis, most commonly in the lung.

A variety of diagnostic tests have been employed to help separate benign from malignant thyroid nodules, including radionuclide scanning and fine-needle aspiration. Most papillary lesions are cold masses on scintiscans. Improvements in cytologic analysis have made fine-needle aspiration cytology a reliable test for distinguishing between benign and malignant nodules. The nuclear features are often nicely demonstrable in aspirated specimens.

Papillary thyroid cancers have an excellent prognosis, with a 10-year survival rate in excess of 95%. Five per cent to 20% of patients have local or regional recurrences, and 10% to 15% have distant metastases. The prognosis of a patient with papillary thyroid

cancers is dependent on several factors including age (in general, the prognosis is less favorable among patients older than 40 years), the presence of extrathyroidal extension, and presence of distant metastases (stage).

Follicular Carcinoma

Follicular carcinomas are the second most common form of thyroid cancer, accounting for 10% to 20% of all thyroid cancers. They tend to present in women, and at an older age than do papillary carcinomas, with a peak incidence in the forties and fifties. The incidence of follicular carcinoma is increased in areas of dietary iodine deficiency, suggesting that in some cases, nodular goiter may predispose to the development of the neoplasm. The high frequency of RAS mutations in follicular adenomas and carcinomas suggests that the two may be related tumors.

Morphology.

Follicular carcinomas are single nodules that may be well circumscribed or widely infiltrative ( Fig. 24-18 ). Sharply demarcated lesions may be exceedingly difficult to distinguish from follicular adenomas by gross examination. Larger lesions may penetrate the capsule and infiltrate well beyond the thyroid capsule into the adjacent neck. They are gray to tan to pink on cut section and, on occasion, are somewhat translucent when large, colloid-filled follicles are present. Degenerative changes, such as central fibrosis and foci of calcification, are sometimes present.

Microscopically, most follicular carcinomas are composed of fairly uniform cells forming small follicles containing colloid, quite reminiscent of normal thyroid ( Fig. 24-19 ). In other cases, follicular differentiation may be less apparent, and there may be nests

Figure 24-18 Follicular carcinoma. Cut surface of a follicular carcinoma with substantial replacement of the lobe of the thyroid. The tumor has a light-tan appearance and contains small foci of hemorrhage.

1181

Figure 24-19 Follicular carcinoma of the thyroid. A few of the glandular lumens contain recognizable colloid.

or sheets of cells without colloid. Occasional tumors are dominated by cells with abundant granular, eosinophilic cytoplasm (Hürthle cells). Whatever the pattern, the nuclei lack the features typical of papillary carcinoma, and psammoma bodies are not present. It is important to note the absence of these details because some papillary carcinomas may appear almost entirely follicular (see Table 24-4 ). Follicular lesions in which the nuclear features are typical of papillary carcinomas should be treated as papillary cancers. While nuclear features are helpful in distinguishing papillary from follicular neoplasms, they are of little value in distinguishing follicular adenomas from minimally invasive follicular carcinomas. This distinction requires extensive histologic sampling of the tumor-capsule-thyroid interface to exclude capsular and/or vascular invasion ( Fig. 24-20 ). The criterion

Figure 24-20 Capsular integrity in follicular neoplasms. Evaluating the integrity of the capsule is critical in distinguishing follicular adenomas from follicular carcinomas. In adenomas (A), a fibrous capsule, usually thin but occasionally more prominent, circumferentially surrounds the neoplastic follicles and no capsular invasion is seen (arrowheads); compressed normal thyroid parenchyma is usually present external to the capsule (top of the panel). In contrast, follicular carcinomas demonstrate capsular invasion (B, arrow-heads) that may be minimal, as in this case, or widespread with extension into local structures of the neck. The presence of vascular invasion is another feature of follicular carcinomas.

for vascular invasion is applicable only to capsular vessels and vascular spaces beyond the capsule; the presence of tumor plugs within intratumoral blood vessels has little prognostic significance. Unlike in papillary cancers, lymphatic spread is distinctly uncommon in follicular cancers.

In contrast to minimally invasive follicular cancers, extensive invasion of adjacent thyroid parenchyma or extrathyroidal tissues makes the diagnosis of carcinoma obvious in widely invasive follicular carcinomas. Histologically, these cancers tend to have a greater proportion of solid or trabecular growth pattern, less evidence of follicular differentiation, and increased mitotic activity.

Clinical Course.

Follicular carcinomas present as slowly enlarging painless nodules. Most frequently, they are cold nodules on scintigrams, although in rare cases, the better-differentiated lesions may be hyperfunctional, take up radioactive iodine, and appear warm on scintiscan. Follicular carcinomas have little propensity for invading lymphatics; therefore, regional lymph nodes are rarely involved, but vascular invasion is common, with spread to bone, lungs, liver, and elsewhere. The prognosis is largely dependent on the extent of invasion and stage at presentation. Widely invasive follicular carcinomas not infrequently develop metastases, and up to half succumb to their disease within 10 years. This is in stark contrast to minimally invasive follicular carcinoma, which has a 10-year survival rate greater than 90%. Most follicular carcinomas are treated with total thyroidectomy followed by the administration of radioactive iodine, the rationale being that metastases are likely to take up the radioactive element, which can be used to identify and ablate such lesions. In addition, because any residual follicular carcinoma may respond to TSH stimulation, patients are usually treated with thyroid hormone after surgery to suppress endogenous TSH.

1182

Medullary Carcinoma

Medullary carcinomas of the thyroid are neuroendocrine neoplasms derived from the parafollicular cells, or C cells, of the thyroid.[45] The cells of medullary carcinomas, similar to normal C cells, secrete calcitonin, the measurement of which plays an important role in the diagnosis and postoperative follow-up of patients. In some instances, the tumor cells elaborate other polypeptide hormones, such as somatostatin, serotonin, and vasoactive intestinal peptide (VIP). The tumors arise sporadically in about 80% of cases. The remainder occurs in the setting of MEN syndrome 2A or 2B or as familial tumors without an associated MEN syndrome (familial medullary thyroid carcinoma, or FMTC; discussed later). Recall that activating point mutations in the RET protooncogene play an important role in the development of both familial and sporadic medullary carcinomas. Cases associated with MEN-2 occur in younger patients and may even arise during childhood. In contrast, sporadic medullary carcinomas as well as FMTC are lesions of adulthood, with a peak incidence in the forties and fifties.

Morphology.

Medullary carcinomas can arise as a solitary nodule or may present as multiple lesions involving both lobes of the thyroid. The sporadic neoplasms tend to originate in one lobe ( Fig. 24-21 ). In contrast, bilaterality and multicentricity are common in familial cases. Larger lesions often contain areas of necrosis and hemorrhage and may extend through the capsule of the thyroid. The tumor tissue is firm, pale gray to tan, and infiltrative. There may be foci of hemorrhage and necrosis in the larger lesions.

Microscopically, medullary carcinomas are composed of polygonal to spindle-shaped cells, which may form nests, trabeculae, and even follicles. Small, more anaplastic cells

are present in some tumors and may be the predominant cell type. Acellular amyloid deposits, derived from altered calcitonin molecules, are present in the adjacent stroma in many cases ( Fig. 24-22 ). Calcitonin

Figure 24-21 Medullary carcinoma of thyroid. These tumors typically show a solid pattern of growth and do not have connective tissue capsules. (Courtesy of Dr. Joseph Corson, Brigham and Women's Hospital, Boston, MA.)

Figure 24-22 Medullary carcinoma of the thyroid. These tumors typically contain amyloid, visible here as homogeneous extracellular material, derived from calcitonin molecules secreted by the neoplastic cells.

is readily demonstrable within the cytoplasm of the tumor cells as well as in the stromal amyloid by immunohistochemical methods. Electron microscopy reveals variable numbers of membrane-bound electron-dense granules within the cytoplasm of the neoplastic cells ( Fig. 24-23 ). One of the peculiar features of familial medullary cancers

is the presence of multicentric C-cell hyperplasia in the surrounding thyroid parenchyma, a feature that is usually absent in sporadic lesions.[46] While the precise criteria for defining C-cell hyperplasia are not establisted, the presence of multiple prominent clusters of C cells scattered throughout the parenchyma should raise the specter of a familial tumor, even if that history is not explicitly present. Foci of C-cell hyperplasia are believed to represent the precursor lesions from which medullary carcinomas arise.

Clinical Course.

Sporadic cases of medullary carcinoma come to medical attention most often as a mass in the neck, sometimes associated with local effects such as dysphagia or

Figure 24-23 Electron micrograph of medullary thyroid carcinoma. These cells contain membrane-bound secretory granules that are the sites of storage of calcitonin and other peptides (30,000X).

1183

hoarseness. In some instances, the initial manifestations are those of a paraneoplastic syndrome, caused by the secretion of a peptide hormone (e.g., diarrhea owing to the secretion of VIP). Notably, hypocalcemia is not a prominent feature, despite the presence of raised calcitonin levels. Screening of relatives for elevated calcitonin levels or RET mutations permits early detection of tumors in familial cases. As will be discussed later, all MEN-2 kindred carrying RET mutations are offered prophylactic thyroidectomy to preclude the development of medullary carcinomas, the major risk factor for poor outcome in these families. Sometimes, the only histologic finding in the resected thyroid of asymptomatic carriers is the presence of C-cell hyperplasia or small (<1 cm) "micromedullary" carcinomas.[47] Recent studies have shown that specific RET mutations correlate with the aggressiveness of medullary carcinomas and the propensity of MEN-2 patients to develop other coincident endocrine tumors.[48] [49]

Anaplastic Carcinoma

Anaplastic carcinomas of the thyroid are undifferentiated tumors of the thyroid follicular epithelium. In striking contrast to the differentiated thyroid carcinomas, anaplastic carcinomas are aggressive tumors, with a mortality rate approaching 100%. These tumors account for fewer than 5% of all thyroid cancers. Patients with anaplastic carcinoma are older than those with other types of thyroid cancer, with a mean age of 65 years. About half of the patients have a history of multinodular goiter, whereas 20% of the patients with these tumors have a history of differentiated carcinoma, and another 20% to 30% have a concurrent differentiated thyroid tumor, frequently a papillary carcinoma. These findings have led to the proposal that anaplastic carcinoma develops by so-called dedifferentiation from more differentiated tumors as a result of one or more genetic changes, including the loss of the p53 tumor suppressor gene.

Morphology.

Microscopically, these neoplasms are composed of highly anaplastic cells, which may take one of several histologic patterns: (1) large, pleomorphic giant cells, including occasional osteoclast-like multinucleate giant cells; (2) spindle cells with a sarcomatous appearance; (3) mixed spindle and giant cells; and (4) small cells resembling those seen in small cell carcinomas arising at other sites. It is unlikely that a true small cell carcinoma exists in the thyroid, and a significant number of such "small cell" tumors have ultimately proven to be medullary carcinomas (discussed previously) or malignant lymphomas, which may also occur in the thyroid but have a much better prognosis. Foci of papillary or follicular differentiation may be present in some tumors, suggesting origin from a better differentiated carcinoma.

Clinical Course.

Anaplastic carcinomas usually present as a rapidly enlarging bulky neck mass. In most cases, the disease has already spread beyond the thyroid capsule into adjacent neck structures or has metastasized to the lungs at the time of presentation. Compression and invasion symptoms, such as dyspnea, dysphagia, hoarseness, and cough, are common. There is no effective therapy for anaplastic thyroid carcinoma, and the disease is almost uniformly fatal. Although metastases to distant sites are common, in most cases death occurs in less than 1 year as a result of aggressive growth and compromise of vital structures in the neck.

Congenital Anomalies

Thyroglossal duct or cyst is the most common clinically significant congenital anomaly of the thyroid. A persistent sinus tract may remain as a vestigial remnant of the tubular development of the thyroid gland. Parts of this tube may be obliterated, leaving small segments to form cysts. These occur at any age and might not become evident until adult life. Mucinous, clear secretions may collect within these cysts to form either spherical masses or fusiform swellings, rarely over 2 to 3 cm in diameter. These are present in the midline of the neck anterior to the trachea. Segments of the duct and cysts that occur high

in the neck are lined by stratified squamous epithelium, which is essentially identical with that covering the posterior portion of the tongue in the region of the foramen cecum. The anomalies that occur in the lower neck more proximal to the thyroid gland are lined by epithelium resembling the thyroidal acinar epithelium. Characteristically, subjacent to the lining epithelium, there is an intense lymphocytic infiltrate. Superimposed infection may convert these lesions into abscess cavities, and rarely, they give rise to cancers.

Parathyroid Glands

Normal

The parathyroid glands are derived from the developing pharyngeal pouches that also give rise to the thymus. The four glands normally lie in close proximity to the upper and lower poles of each thyroid lobe but may also be found anywhere along the pathway of descent of the pharyngeal pouches, including the carotid sheath, the thymus, and elsewhere in the anterior mediastinum. Of note, 10% of individuals have only two or three glands.

In the adult, the parathyroid is a yellow-brown, ovoid encapsulated nodule weighing approximately 35 to 40 mg.

1184

Most of the gland is composed of chief cells. The chief cells vary from light to dark pink with hematoxylin and eosin stains, depending on their glycogen content. They are polygonal; are 12 to 20 mm in diameter; and have central, round, uniform nuclei. In addition, they contain secretory granules of parathyroid hormone (PTH). Sometimes, these cells have a water-clear appearance owing to lakes of glycogen. Oxyphil cells and transitional oxyphils are found throughout the normal parathyroid, either singly or in small clusters. They are slightly larger than the chief cells, have acidophilic cytoplasm, and are tightly packed with mitochondria. Glycogen granules are also present in these cells, but secretory granules are sparse or absent. In early infancy and childhood, the parathyroid glands are composed almost entirely of solid sheets of chief cells. The amount of stromal fat increases up to age 25, reaching a maximum of approximately 30% of the gland, and then plateaus. The precise proportion of fat is determined largely by constitutional factors; for instance, obese individuals have more adipose tissue in their glands.

The activity of the parathyroid glands is controlled by the level of free (ionized) calcium in the bloodstream rather than by trophic hormones secreted by the hypothalamus and pituitary. Normally, decreased levels of free calcium stimulate the synthesis and secretion of PTH. Circulating PTH is an 84-amino-acid linear polypeptide derived by sequential cleavage in the chief cell of a larger pre-pro form. Its biologic activity resides within the 34 residues at the amino terminus. Smaller nonfunctional fragments of the hormone, apparently lacking the critical amino-terminal domain, also circulate. These assume importance because, although they are biologically inert, they contain epitopes that react in certain radioimmunoassays for PTH.

The PTH receptor is a seven-transmembrane G-protein-coupled receptor. Binding of the hormone leads to activation of the stimulatory G-protein, Gs , causing adenylate cyclase-mediated generation of cAMP. This pathway assumes clinical significance when abnormalities of the Gs protein result in either hyperactivity or hypoactivity of the parathyroid gland (see below). The metabolic functions of PTH in supporting serum calcium levels can be summarized as follows:

• PTH activates osteoclasts, thereby mobilizing calcium from bone. • It increases the renal tubular reabsorption of calcium, thereby conserving free calcium. • It increases the conversion of vitamin D to its active dihydroxy form in the kidneys. • It increases urinary phosphate excretion, thereby lowering serum phosphate levels. • It augments gastrointestinal calcium absorption.

The net result of these activities is an increase in the level of free calcium, which, in turn, inhibits further PTH secretion in a classic feedback loop.

Hypercalcemia is one of a number of changes induced by elevated levels of PTH. As was discussed in Chapter 7 , hypercalcemia is a relatively common complication of malignancy, occurring both with solid tumors, such as lung, breast, head and neck, and renal cancers, and with hematologic malignancies, notably multiple myeloma. In fact, malignancy is the most common cause of clinically apparent hypercalcemia, while primary hyperparathyroidism (see below) is a more common cause of asymptomatic elevated blood calcium. The prognosis of patients with malignancy-associated hypercalcemia is generally poor, in that it more frequently occurs in individuals with advanced cancers. Hypercalcemia of malignancy is due to increased bone resorption and subsequent release of calcium. There are two major mechanisms by which this can occur: (1) osteolytic metastases and local release of cytokines and (2) release of PTH-related protein (PTHrP).

• Osteolytic metastases: Metastatic tumor cells, as well as stromal cells in the vicinity of the metastases, release a variety of soluble mediators that induce local osteolysis by promoting differentiation of committed osteoclast precursors into mature cells. Recently, a critical osteoclastogenic pathway has been discovered

that involves the osteoblast cell-surface receptor RANK (receptor activator of nuclear factor κB), its ligand, RANKL, and a decoy receptor for RANKL, osteoprotegerin. (A decoy receptor is a soluble receptor that competes with the true receptor, in this case RANK, for binding the ligand of interest, in this case RANKL;[50] see Chapter 26 .) RANKL is also known as "osteoclast differentiation factor," and by binding with the RANK receptor, it promotes all aspects of osteoclast function, including proliferation, differentiation, fusion, and activation. RANKL is secreted by tumor cells and peritumoral stromal cells in metastatic foci and causes osteolysis. Osteoprotegerin inhibits this pathway of osteoclastogenesis and has emerged as a possible therapeutic agent in cancer patients with hypercalcemia of malignancy. • PTH-related protein: The most frequent cause of hypercalcemia in nonmetastatic solid tumors—particularly squamous cell cancers—is the release of PTHrP. This protein is immunologically distinct from PTH yet it is similar enough in structure to permit binding to identical receptors and simulation of second messengers, notably cAMP. This accounts for the ability of PTHrP to induce most of the actions of PTH, including increases in bone resorption and inhibition of proximal tubule phosphate transport.[51] Classically, PTHrP-induced hypercalcemia was known as "humoral hypercalcemia of malignancy" to distinguish it from hypercalcemia arising from osteolytic metastases. It is now recognized that a significant proportion of cancer patients with osteolytic metastases also have circulating PTHrP; therefore, PTHrP contributes to hypercalcemia of malignancy irrespective of the presence or absence of metastases.

Pathology

Similar to the other endocrine organs, abnormalities of the parathyroid glands include both hyperfunction and hypofunction. Tumors of the parathyroid glands, in contrast to thyroid tumors, usually come to attention because of excessive secretion of PTH rather than because of mass effects.

Hyperparathyroidism

Hyperparathyroidism occurs in two major forms—primary and secondary—and, less commonly, tertiary. The first condition represents an autonomous, spontaneous overproduction

1185

of PTH; the latter two conditions typically occur as secondary phenomena in patients with chronic renal insufficiency.

PRIMARY HYPERPARATHYROIDISM

Primary hyperparathyroidism is one of the most common endocrine disorders, and it is an important cause of hypercalcemia. The frequency of the various parathyroid lesions underlying the hyperfunction is as follows:

• Adenoma: 75% to 80% • Primary hyperplasia (diffuse or nodular): 10% to 15% • Parathyroid carcinoma: less than 5%

Primary hyperparathyroidism is usually a disease of adults and is more common in women than in men by a ratio of nearly 3:1. The annual incidence is now estimated to be about 25 cases per 100,000 in the United States and Europe; more cases are being detected due to the greater availability and use of advanced analyzers for measuring serum electrolytes.[52] Most cases occur in the fifties or later in life.

Studies have begun to provide a molecular understanding of the pathogenesis of primary hyperparathyroidism. In more than 95% of cases, the disorder is caused by sporadic parathyroid adenomas or sporadic hyperplasia ( Fig. 24-24 ). Although familial syndromes are a distant second, they have provided a unique insight into the pathogenesis of primary hyperparathyroidism. The genetic syndromes associated with familial primary hyperparathyroidism include the following:

• Multiple endocrine neoplasia-1 (MEN-1): The MEN1 gene on chromosome 11q13 is a tumor suppressor gene inactivated in a variety of MEN-1-related parathyroid lesions, including parathyroid adenomas and hyperplasia. In addition to familial cases, MEN1 mutations have also been described in sporadic parathyroid tumors. The MEN-1 syndrome is discussed in further detail below. • Multiple endocrine neoplasia-2 (MEN-2): The MEN-2 syndrome is caused by activating mutations in the tyrosine kinase receptor, RET, on chromosome 10q. Primary hyperparathyroidism occurs as a component of MEN-2A, which is discussed in further detail below. RET mutations have not been described in sporadic parathyroid lesions outside the context of MEN-2. • Familial hypocalciuric hypercalcemia (FHH) is an autosomal-dominant disorder characterized by enhanced parathyroid function due to decreased sensitivity to extracellular calcium. Mutations in the parathyroid calciumsensing receptor gene (CASR) on chromosome 3q are a primary cause for this disorder.[53] Patients with homozygous CASR mutations present in the neonatal period with severe hyperparathyroidism. CASR mutations have not been described in sporadic parathyroid tumors.

Figure 24-24 Parathyroid adenomas are almost always solitary lesions. Technetium-99m-sestamibi radionuclide scan demonstrates an area of increased uptake corresponding to the left inferior parathyroid gland (arrow). This patient had a parathyroid adenoma. Preoperative scintigraphy is useful in localizing and distinguishing adenomas from parathyroid hyperplasia, where more than one gland would demonstrate increased uptake.

Most, if not all, sporadic parathyroid adenomas are monoclonal, suggesting that they are true neoplastic outgrowths from a single abnormal progenitor cell. Sporadic parathyroid hyperplasia is also monoclonal in many instances, particularly when associated with a persistent stimulus for parathyroid growth (refractory secondary or tertiary parathyroidism; see below). Among the sporadic adenomas, there are two molecular defects that have an established role in pathogenesis: [54]

• Parathyroid adenoma 1 (PRAD1): PRAD1 encodes cyclin D1, a major regulator of the cell cycle. A pericentromeric inversion on chromosome 11 results in relocation of the PRAD1 protooncogene (normally on 11q) so that it is positioned adjacent to the 5' flanking region of the PTH gene (on 11p). As a consequence of these changes, a regulatory element from the PTH gene 5' flanking sequence directs overexpression of cyclin D1 protein, forcing the cells to proliferate. Ten per cent to 20% of adenomas have this clonal genetic defect. In addition, cyclin D1 is overexpressed in approximately 40% of parathyroid adenomas, suggesting that mechanisms other than PRAD1 inversion can lead to its activation. • MEN1: Approximately 20% to 30% of parathyroid tumors not associated with the MEN-1 syndrome demonstrate mutations in both copies of the MEN1 gene. The spectrum of MEN1 mutations in the sporadic tumors is virtually identical to that in familial parathyroid adenomas.

Morphology.

The morphologic changes seen in primary hyperparathyroidism include those in the parathyroid glands as well as those in other organs affected by elevated levels of calcium. Parathyroid adenomas are almost always solitary and, similar to the normal parathyroid glands, may lie in close proximity to the thyroid gland or in an ectopic site (e.g., the mediastinum). The typical parathyroid adenoma averages 0.5 to 5.0 gm; is a well-

circumscribed, soft, tan to reddish-brown nodule; and is invested by a delicate capsule ( Fig. 24-25 ). In contrast to primary hyperplasia, the glands outside the adenoma are usually normal in size or somewhat shrunken because of feedback inhibition by elevations in serum calcium. Microscopically, parathyroid adenomas are often composed predominantly of fairly uniform, polygonal chief cells with small, centrally placed nuclei (see Fig. 24-25 ). In most cases, at least a few nests of larger cells containing oxyphil cells are present as well; uncommonly, entire adenomas may be composed of this cell type (oxyphil adenomas). The chief cells are

1186

Figure 24-25 Parathyroid adenoma. A, Solitary chief cell parathyroid adenoma (low-power photomicrograph) revealing clear delineation from the residual gland below. B, High-power detail of a chief cell parathyroid adenoma. There is some slight variation in nuclear size but no anaplasia and some slight tendency to follicular formation.

arranged in a variety of patterns; follicles reminiscent of those seen in the thyroid are present in some cases. Mitotic figures are rare. A rim of compressed, non-neoplastic parathyroid tissue, generally separated by a fibrous capsule, is often visible at the edge of the adenoma. It is not uncommon to find bizarre and pleomorphic nuclei even within adenomas (so-called endocrine atypia), and this should not be used as a criterion for defining malignancy. In contrast to the normal parathyroid parenchyma, adipose tissue is inconspicuous within the adenoma.

Primary hyperplasia may occur sporadically or as a component of MEN syndrome. Although classically all four glands are involved, there is frequently asymmetry with apparent sparing of one or two glands, making the distinction between hyperplasia and adenoma difficult. The combined weight of all glands rarely exceeds 1.0 gm and is often less. Microscopically, the most common pattern seen is that of chief cell hyperplasia, which may involve the glands in a diffuse or multinodular pattern. Less commonly, the constituent cells contain abundant water-clear cells ("water-clear cell hyperplasia"). In many instances, there are islands of oxyphils, and poorly developed, delicate fibrous

strands may envelop the nodules. As in the case of adenomas, stromal fat is inconspicuous within the foci of hyperplasia.

Parathyroid carcinomas may be fairly circumscribed lesions that are difficult to distinguish from adenomas, or they may be clearly invasive neoplasms. These tumors enlarge one parathyroid gland and consist of gray-white, irregular masses that sometimes exceed 10 gm in weight. The cells of parathyroid carcinomas are usually uniform and resemble normal parathyroid cells. They are arrayed in nodular or trabecular patterns with a dense, fibrous capsule enclosing the mass. There is general agreement that a diagnosis of carcinoma based on cytologic detail is unreliable, and invasion of surrounding tissues and metastasis are the only reliable criteria of malignancy. Local recurrence occurs in one third of cases, and more distant dissemination occurs in another third.

Morphologic changes in other organs deserving special mention include skeletal and renal lesions. Skeletal changes include prominence of osteoclasts, which, in turn, erode bone matrix and mobilize calcium salts, particularly in the metaphyses of long tubular bones ( Chapter 26 ). Bone resorption is accompanied by increased osteoblastic activity and the formation of new bone trabeculae. In many cases, the resultant bone contains widely spaced, delicate trabeculae reminiscent of those seen in osteoporosis. In more severe cases, the cortex is grossly thinned, and the marrow contains increased amounts of fibrous tissue accompanied by foci of hemorrhage and cyst formation (osteitis fibrosa cystica). Aggregates of osteoclasts, reactive giant cells, and hemorrhagic debris occasionally form masses that may be mistaken for neoplasms (brown tumors of hyperparathyroidism). PTH-induced hypercalcemia favors formation of urinary tract stones (nephrolithiasis) as well as calcification of the renal interstitium and tubules (nephrocalcinosis). Metastatic calcification secondary to hypercalcemia may also be seen in other sites, including the stomach, lungs, myocardium, and blood vessels.

Clinical Course.

Primary hyperparathyroidism presents in one of two general ways: (1) It may be asymptomatic and be identified after a routine chemistry profile, or (2) patients may have the classic clinical manifestations of primary hyperparathyroidism. [52]

Asymptomatic Hyperparathyroidism.

Because serum calcium levels are routinely assessed in the work-up of most patients who need blood tests for unrelated conditions, clinically silent hyperparathyroidism is often detected early. Hence, many of the classic clinical manifestations, particularly those referable to bone and renal disease, are now seen infrequently in clinical practice. The most common manifestation of primary hyperparathyroidism is an increase in the level of serum ionized calcium; in fact, primary hyperparathyroidism is the most common cause of asymptomatic hypercalcemia. It should be recalled that other conditions also produce hypercalcemia ( Table 24-5 ). Malignancy, in particular, is the most common cause of clinically apparent hypercalcemia in adults and must be excluded by appropriate clinical and laboratory investigations in patients with suspected hyperparathyroidism.

1187

TABLE 24-5 -- Causes of Hypercalcemia

Raised PTH Decreased PTH

Hyperparathyroidism Hypercalcemia of malignancy

Primary (adenoma   > hyperplasia) * Osteolytic metastases   (RANKL-mediated)

Secondary   † PTH-rP-mediated  

Tertiary   † Vitamin D toxicity

Familial hypocalciuric hypercalcemia Immobilization

Thiazide diuretics

Granulomatous disease (sarcoidosis)

PTH-rP, Parathyroid hormone-related protein. RANKL, Receptor activator of nuclear factor κB ligand.

*Primary hyperparathyroidism is the most common cause of hypercalcemia overall. Malignancy is the most common cause of symptomatic hypercalcemia. Primary hyperparathyroidism and malignancy account for nearly 90% of cases of hypercalcemia.†Secondary and tertiary hyperparathyroidism are most commonly associated with progressive renal failure.

In patients with primary hyperparathyroidism, serum PTH levels are inappropriately elevated for the level of serum calcium, whereas PTH levels are low to undetectable in hypercalcemia because of nonparathyroid disease (see Table 24-5 ). In patients with hypercalcemia caused by secretion of PTHrP by certain nonparathyroid tumors, radioimmunoassays specific for PTH and PTHrP can distinguish between the two molecules. Other laboratory alterations referable to PTH excess include hypophosphatemia and increased urinary excretion of both calcium and phosphate. Secondary renal disease may lead to phosphate retention with normalization of serum phosphates.

Symptomatic Primary Hyperparathyroidism.

The signs and symptoms of hyperparathyroidism reflect the combined effects of increased PTH secretion and hypercalcemia. Primary hyperparathyroidism has been traditionally associated

Figure 24-26 Cardinal features of hyperparathyroidism. With routine evaluation of calcium levels in most patients, primary hyperparathyroidism is often detected at a clinically silent stage. Hypercalcemia from any other cause can also give rise to the same symptoms.

with a constellation of symptoms that included "painful bones, renal stones, abdominal groans, and psychic moans" ( Fig. 24-26 ). The symptomatic presentation involves a diversity of clinical manifestations:

• Bone disease includes bone pain secondary to fractures of bones weakened by osteoporosis or osteitis fibrosa cystica. • Nephrolithiasis (renal stones) occurs in 20% of newly diagnosed patients, with attendant pain and obstructive uropathy. Chronic renal insufficiency and a variety of abnormalities in renal function are found, including polyuria and secondary polydipsia. • Gastrointestinal disturbances include constipation, nausea, peptic ulcers, pancreatitis, and gallstones. • Central nervous system alterations include depression, lethargy, and eventually seizures. • Neuromuscular abnormalities include complaints of weakness and fatigue. • Cardiac manifestations include aortic or mitral valve calcifications (or both).

The abnormalities most directly related to hyperparathyroidism are nephrolithiasis and bone disease, whereas those attributable to hypercalcemia include fatigue, weakness, and constipation. The pathogenesis of many of the other manifestations of the disorder remains poorly understood.

SECONDARY HYPERPARATHYROIDISM

Secondary hyperparathyroidism is caused by any condition associated with a chronic depression in the serum calcium level because low serum calcium leads to compensatory overactivity of the parathyroid glands.[55] Renal failure is by far the most common cause of secondary hyperparathyroidism, although a number of other diseases, including inadequate dietary intake of calcium, steatorrhea, and vitamin D deficiency, may also cause this disorder. The mechanisms by which

1188

chronic renal failure induces secondary hyperparathyroidism are complex and not fully understood. Chronic renal insufficiency is associated with decreased phosphate excretion, which in turn results in hyperphosphatemia. The elevated serum phosphate levels directly depress serum calcium levels and thereby stimulate parathyroid gland activity. In addition, loss of renal substance reduces the availability of α-1-hydroxylase necessary for the synthesis of the active form of vitamin D, which in turn reduces intestinal absorption of calcium ( Chapter 9 ).

Morphology.

The parathyroid glands in secondary hyperparathyroidism are hyperplastic. As in the case of primary hyperplasia, the degree of glandular enlargement is not necessarily symmetric. Microscopically, the hyperplastic glands contain an increased number of chief cells, or cells with more abundant, clear cytoplasm (so-called water-clear cells) in a diffuse or multinodular distribution. Fat cells are decreased in number. Bone changes similar to those seen in primary hyperparathyroidism may also be present. Metastatic calcification may be seen in many tissues, including lungs, heart, stomach, and blood vessels.

Clinical Course.

The clinical features of secondary hyperparathyroidism are usually dominated by those associated with chronic renal failure. Bone abnormalities (renal osteodystrophy) and other changes associated with PTH excess are, in general, less severe than are those seen in primary hyperparathyroidism. The vascular calcification associated with secondary hyperparathyroidism may occasionally result in significant ischemic damage to skin and other organs, a process sometimes referred to as calciphylaxis. In a minority of patients, parathyroid activity may become autonomous and excessive, with resultant hypercalcemia, a process that is sometimes termed tertiary hyperparathyroidism. Parathyroidectomy may be necessary to control the hyperparathyroidism in such patients.

Hypoparathyroidism

Hypoparathyroidism is far less common than is hyperparathyroidism. There are many possible causes of deficient PTH secretion resulting in hypoparathyroidism:

• Surgically induced hypoparathyroidism occurs with inadvertent removal of all the parathyroid glands during thyroidectomy, excision of the parathyroid glands in the mistaken belief that they are lymph nodes during radical neck dissection for some form of malignant disease, or removal of too large a proportion of parathyroid tissue in the treatment of primary hyperparathyroidism. • Congenital absence of all glands, as in certain developmental abnormalities, such as thymic aplasia and cardiac defects (22q11.2 syndrome) (see Chapter 5 ). • Familial hypoparathyroidism is often associated with chronic mucocutaneous candidiasis and primary adrenal insufficiency; this syndrome is known as autoimmune polyendocrine syndrome type 1 (APS1) and is caused by mutations in the autoimmune regulator (AIRE) gene.[56] The syndrome typically presents in childhood with the onset of candidiasis, followed several years later by hypoparathyroidism and then adrenal insufficiency during adolescence. APS1 is discussed further in the section on adrenal glands. • Idiopathic hypoparathyroidism most likely represents an autoimmune disease with isolated atrophy of the glands. Sixty per cent of the patients with this disorder have autoantibodies directed against the calcium-sensing receptor (CASR) in the parathyroid gland.[57] Antibody binding to the receptor may prevent the release of PTH.

The major clinical manifestations of hypoparathyroidism are referable to hypocalcemia and are related to the severity and chronicity of the hypocalcemia.

• The hallmark of hypocalcemia is tetany, which is characterized by neuromuscular irritability, resulting from decreased serum ionized calcium concentration. These findings can range from circumoral numbness or paresthesias (tingling) of the distal extremities and carpopedal spasm, to life-threatening laryngospasm and generalized seizures. The classic findings on physical examination of patients with neuromuscular irritability are Chvostek sign and Trousseau sign. Chvostek sign is elicited in subclinical disease by tapping along the course of the facial nerve, which induces contractions of the muscles of the eye, mouth, or nose. Occluding the circulation to the forearm and hand by inflating a blood pressure cuff about the arm for several minutes induces carpal spasm, which disappears as soon as the cuff is deflated (Trousseau sign). • Mental status changes can include emotional instability, anxiety and depression, confusional states, hallucinations, and frank psychosis. • Intracranial manifestations include calcifications of the basal ganglia, parkinsonian-like movement disorders, and increased intracranial pressure with resultant papilledema. • Ocular disease results in calcification of the lens leading to cataract formation.

• Cardiovascular manifestations include a conduction defect, which produces a characteristic prolongation of the QT interval in the electrocardiogram. • Dental abnormalities occur when hypocalcemia is present during early development. These findings are highly characteristic of hypoparathyroidism and include dental hypoplasia, failure of eruption, defective enamel and root formation, and abraded carious teeth.

Pseudohypoparathyroidism

In this condition, hypoparathyroidism occurs because of end-organ resistance to the actions of PTH. Indeed, serum PTH levels are normal or elevated. Central to the understanding of PTH resistance are two key concepts: (1) G-proteins, principally Gs , mediate the cellular actions of PTH on bone and kidney, and (2) GNAS1 is a selectively imprinted gene, with tissue-specific patterns of imprinting.[58] In most tissues, Gs α, the product of GNAS1, is expressed from both alleles. In the pituitary (see above) and the kidneys, GNAS1 is expressed only from the maternally inherited chromosome, owing to paternal imprinting (silencing) of the gene. As a result, a mutation that affects the maternal allele results in

1189

complete loss of Gs α expression in the kidney, while a mutation in the normally unexpressed paternal allele has no effect on Gs α levels; in contrast, a mutation of either allele will produce a 50% decrease in Gs α in tissues other than the kidneys and pituitary, since GNAS1 is expressed from both copies of the gene. Two types of pseudohypoparathyroidism have been identified depending on the parent of origin of the mutant allele:

1. Pseudohypoparathyroidism type 1A is associated with multihormone resistance and Albright hereditary osteodystrophy (AHO), a syndrome characterized by skeletal and developmental defects. Patients with AHO often have short stature, obesity, short metacarpal and metatarsal bones, and variable mental deficits. The multihormone resistance involves three hormones (PTH, TSH, and LH/FSH), all of which activate Gs α-mediated pathways in target tissues. The PTH resistance is the most obvious clinical manifestation, presenting as hypocalcemia, hyperphosphatemia, and elevated circulating PTH. TSH resistance is generally mild, while LH/FSH resistance manifests as hypergonadotropic hypogonadism in females. The mutation in this disorder is inherited on the maternal allele, severely impeding the actions of PTH on the kidney in maintaining calcium homeostasis.

2. Pseudopseudohypoparathyroidism: In this disorder, the mutation is inherited on the paternal allele, and it is characterized by AHO without accompanying multihormonal resistance. As a result, serum calcium, phosphate and PTH levels are normal.

The Endocrine Pancreas

Normal

The endocrine pancreas consists of about 1 million microscopic clusters of cells, the islets of Langerhans. The first evidence of islet formation in the human fetus is seen at 9 to 11 weeks. Embryologically, both endocrine and exocrine components of the pancreas are endodermal derivatives. Several transcription factors have now been identified that determine lineage specification (i.e., endocrine versus exocrine) in the developing pancreas. For example, expression of the transcription factor neurogenin 3 (Ngn3) delineates endocrine progenitors that eventually give rise to mature islet cells. [59]

In aggregate, the islets in the adult human weigh only 1 to 1.5 gm; individually, most islets measure 100 to 200 µm and consist of four major and two minor cell types. The four main types are β, α, δ, and PP (pancreatic polypeptide) cells. These make up about 68%, 20%, 10%, and 2%, respectively, of the adult islet cell population. They can be differentiated morphologically by their staining properties, by the ultrastructural characteristics of their granules, and by their hormone content (see Fig. 24-27 ).

The β cell produces insulin, as will be detailed in the discussion of diabetes. The insulin-containing intracellular granules contain a crystalline matrix with a rectangular profile, surrounded by a halo. The α cell secretes glucagon, inducing hyperglycemia by its glycogenolytic activity in the liver. α-cell granules are round, with closely applied membranes and a dense center. δ cells contain somatostatin, which suppresses both insulin and glucagon release; they have large, pale granules with closely applied membranes. PP cells contain a unique pancreatic polypeptide that exerts a number of gastrointestinal effects, such as stimulation of secretion of gastric and intestinal enzymes and inhibition of intestinal motility. These cells have small, dark granules and not only are present in islets, but also are scattered in the exocrine pancreas.

The two rare cell types are D1 cells and enterochromaffin cells. D1 cells elaborate vasoactive intestinal polypeptide (VIP), a hormone that induces glycogenolysis and hyperglycemia; it also stimulates gastrointestinal fluid secretion and causes secretory diarrhea. Enterochromaffin cells synthesize serotonin and are the source of pancreatic tumors that cause the carcinoid syndrome ( Chapter 17 ).

Pathology

We now turn to the two main disorders of islet cells: diabetes mellitus and pancreatic endocrine tumors.

Diabetes Mellitus

Diabetes mellitus (DM) is not a single disease entity, but rather a group of metabolic disorders sharing the common underlying feature of hyperglycemia. Hyperglycemia in diabetes results from defects in insulin secretion, insulin action, or, most commonly, both. The chronic hyperglycemia and attendant metabolic dysregulation may be associated with secondary damage in multiple organ systems, especially the kidneys, eyes, nerves, and blood vessels. Diabetes affects an estimated 16 million people in the United States, as many as half of whom are undiagnosed. Each year, an additional 800,000 individuals develop diabetes in this country, and 54,000 die from diabetes-related causes. Diabetes is a leading cause of end-stage renal disease, adult-onset blindness, and nontraumatic lower extremity amputations in the United States. For individuals

1190

born in the United States in 2000, the estimated lifetime risk of being diagnosed with diabetes mellitus is 1 in 3 for males and 2 in 5 for females.[60] The risk is 2 to 5 times higher in the African-American, Hispanic, and Native American communities, compared to non-Hispanic whites. Worldwide, more than 140 million people suffer from diabetes, making this one of the most common noncommunicable diseases. [61] The number of affected individuals with diabetes is expected to double by 2025. The countries with the largest number of diabetics are India, China, and the United States.

DIAGNOSIS

Blood glucose values are normally maintained in a very narrow range, usually 70 to 120 mg/dL. The diagnosis of diabetes is established by noting elevation of blood glucose by any one of three criteria:

1. A random glucose > 200 mg/dL, with classical signs and symptoms (discussed below)

2. A fasting glucose > 126 mg/dL on more than one occasion 3. An abnormal oral glucose tolerance test (OGTT), in which the glucose is > 200

mg/dL 2 hours after a standard carbohydrate load

Levels of blood glucose proceed along a continuum. Individuals with fasting glucoses less than 110 mg/dL, or less than 140 mg/dL following an OGTT, are considered to be euglycemic. However, those with fasting glucoses greater than 110 but less than 126, or OGTT values greater than 140 but less than 200, are considered to have impaired glucose tolerance (IGT).[62] Individuals with IGT have a significant risk of progressing to overt diabetes over time, with up to 5% to 10% advancing to DM per year. In addition, those with IGT are at risk for cardiovascular disease, due to the abnormal carbohydrate metabolism as well as the co-existence of other risk factors such as low HDL,

hypertriglyceridemia, and increased plasminogen activator inhibitor-1 (PAI-1) (see Chapter 11 ).

CLASSIFICATION

Although all forms of diabetes mellitus share hyperglycemia as a common feature, the pathogenic processes involved in the development of hyperglycemia vary widely. The previous classification schemes of diabetes mellitus were based on the age at onset of the disease or on the mode of therapy; in contrast, the recently revised classification reflects our greater understanding of the pathogenesis of each variant ( Table 24-6 ). [62] The vast majority of cases of diabetes fall into one of two broad classes:

Type 1 diabetes is characterized by an absolute deficiency of insulin caused by pancreatic β-cell destruction. It accounts for approximately 10% of all cases.

Type 2 diabetes is caused by a combination of peripheral resistance to insulin action and an inadequate secretory response by the pancreatic β-cells ("relative insulin deficiency"). Approximately 80% to 90% of patients have type 2 diabetes.

A variety of monogenic and secondary causes are responsible for the remaining cases, and these will be discussed later. It should be stressed that while the major types of diabetes have different pathogenic mechanisms, the long-term complications

TABLE 24-6 -- Classification of Diabetes Mellitus

 1. Type 1 diabetes (β-cell destruction, leads to absolute insulin deficiency)

Immune-mediated      

Idiopathic      

 2. Type 2 diabetes (insulin resistance with relative insulin deficiency)

 3. Genetic defects of β-cell function

Maturity-onset diabetes of the       young (MODY), caused by mutations in:

Hepatocyte nuclear factor 4α        [HNF-4α] (MODY1)

Glucokinase (MODY2)       

Hepatocyte nuclear factor 1α        [HNF-1α] (MODY3)

Insulin promoter factor [IPF-1]        (MODY4)

Hepatocyte nuclear factor 1β        [HNF-1β] (MODY5)

Neurogenic differentiation factor        1 [Neuro D1] (MODY6)

Mitochondrial DNA       mutations

 4. Genetic defects in insulin processing or insulin action

Defects in proinsulin       conversion

Insulin gene mutations      

Insulin receptor       mutations

 5. Exocrine pancreatic defects

Chronic pancreatitis      

Pancreatectomy      

Neoplasia      

Cystic fibrosis      

Hemachromatosis      

Fibrocalculous       pancreatopathy

 6. Endocrinopathies

Acromegaly      

Cushing syndrome      

Hyperthyroidism      

Pheochromocytoma      

Glucagonoma      

 7. Infections

Cytomegalovirus      

Coxsackie virus B      

 8. Drugs

Glucocorticoids      

Thyroid hormone      

α-interferon      

Protease inhibitors      

β-adrenergic agonists      

Thiazides      

Nicotinic acid      

Phenytoin      

 9. Genetic syndromes associated with diabetes

Down syndrome      

Kleinfelter syndrome      

Turner syndrome      

10. Gestational diabetes mellitus

Data from the Report of the Expert Committee on the Diagnosis and Classification of Diabetes Mellitus. Diabetic Care 25 (suppl. 1):S5–S20, 2002.

in kidneys, eyes, nerves, and blood vessels are the same, as are the principal causes of morbidity and death. The pathogenesis of the two major types is discussed separately, but first we briefly review normal insulin secretion and the mechanism of insulin signaling, since these aspects are critical to understanding the pathogenesis of diabetes.

NORMAL INSULIN PHYSIOLOGY

Normal glucose homeostasis is tightly regulated by three interrelated processes: glucose production in the liver; glucose uptake and utilization by peripheral tissues, chiefly skeletal

1191

muscle; and actions of insulin and counter-regulatory hormones, including glucagon, on glucose.

Insulin and glucagon have opposing regulatory effects on glucose homeostasis. During fasting states, low insulin and high glucagon levels facilitate hepatic gluconeogenesis and glycogenolysis (glycogen breakdown) while decreasing glycogen synthesis, thereby preventing hypoglycemia. Thus, fasting plasma glucose levels are determined primarily by hepatic glucose output. Following a meal, insulin levels rise and glucagon levels fall in response to the large glucose load. Insulin promotes glucose uptake and utilization in tissues (discussed later). The skeletal muscle is the major insulin-responsive site for postprandial glucose utilization, and is critical for preventing hyperglycemia and maintaining glucose homeostasis.

Regulation of Insulin Release

The insulin gene is expressed in the β cells of the pancreatic islets ( Fig. 24-27 ). Preproinsulin is synthesized in the rough endoplasmic reticulum from insulin mRNA and delivered to the Golgi apparatus. There, a series of proteolytic cleavage steps generate the mature insulin and a cleavage peptide, C-peptide. Both insulin and C-peptide are then stored in secretory granules and secreted in equimolar quantities after physiologic stimulation; increasingly, C-peptide

Figure 24-27 Hormone production in pancreatic islet cells. Immunoperoxidase staining shows a dark reaction product for insulin in β cells (A), glucagon in α cells (B), and somatostatin in δ cells (C). D, Electron micrograph of a β cell shows the characteristic membrane-bound granules, each containing a dense, often rectangular core and distinct halo. E, Portions of an α cell (left) and a δ cell (right) also exhibit granules, but with closely apportioned membranes. The α-cell granule exhibits a dense, round center. (Electron micrographs courtesy of Dr. A. Like, University of Massachusetts Medical School, Worcester, MA.)

levels are being used as a clinical assay to measure endogenous insulin secretion.

The most important stimulus that triggers insulin synthesis and release is glucose itself. A rise in blood glucose levels results in glucose uptake into pancreatic β cells, facilitated by an insulin-independent, glucose-transporting protein, GLUT-2 ( Fig. 24-28 ).[63] Metabolism of glucose via glycolysis generates ATP, resulting in increase in cytoplasmic ATP/ADP ratios. This inhibits the activity of the ATP-sensitive K+ -channel on the β-cell membrane, leading to membrane depolarization and the influx of extracellular Ca2+ through voltage-dependent Ca2+ -channels.[64] The resultant increase in intracellular Ca2+ stimulates secretion of insulin, presumably from stored hormone within the β-cell granules. This is the phase of immediate release of insulin. If the secretory stimulus persists, a delayed and protracted response follows that involves active synthesis of insulin. Other agents, including intestinal hormones and certain amino acids (leucine and arginine), stimulate insulin release but not synthesis.

Insulin Action and Insulin Signaling Pathways

Insulin is the most potent anabolic hormone known, with multiple synthetic and growth-promoting effects ( Fig. 24-29 ). Its principal metabolic function is to increase the rate of glucose transport into certain cells in the body. These are the striated

1192

Figure 24-28 Insulin synthesis and secretion. Intracellular transport of glucose is mediated by GLUT-2, an insulin-independent glucose transporter in β cells. Glucose undergoes oxidative metabolism in the β cell to yield ATP. ATP inhibits an inward rectifying potassium channel receptor on the β-cell surface; the receptor itself is a dimeric complex of the sulfonylurea receptor and a K+ -channel protein. Inhibition of this receptor leads to membrane depolarization, influx of Ca2+ ions, and release of stored insulin from β cells.

muscle cells (including myocardial cells) and to a lesser extent, adipocytes, representing collectively about two thirds of the entire body weight. Glucose uptake in other peripheral tissues, most notably the brain, is insulin-independent. In muscle cells, glucose is then either stored as glycogen or oxidized to

Figure 24-29 Metabolic actions of insulin in striated muscle, adipose tissue, and liver.

generate ATP. In adipose tissue, glucose is primarily stored as lipid. Besides promoting lipid synthesis, insulin also inhibits lipid degradation in adipocytes. Similarly, insulin promotes amino acid uptake and protein synthesis, while inhibiting protein degradation. Thus, the anabolic effects of insulin are attributable to increased synthesis and reduced degradation of glycogen, lipids, and proteins. In addition, insulin has several mitogenic functions, including initiation of DNA synthesis in certain cells and stimulation of their growth and differentiation.

The binding of insulin to its receptor triggers a complex signaling cascade of protein phosphorylation and dephosphorylation culminating in the metabolic and mitogenic effects of insulin described above. Elucidation of the insulin signaling pathway has been central to our understanding of the mechanisms underlying insulin resistance in diabetes (see below). The complete description of this intricate network is beyond the scope of this book,[65] and we will only summarize some of the more pertinent mediators ( Fig. 24-30 ). The insulin receptor is a tetrameric protein composed of two α- and two β-subunits. The β-subunit cytosolic domain possesses tyrosine kinase activity. Insulin binding to the α-subunit extracellular domain activates the β-subunit tyrosine kinase, resulting in both

autophosphorylation of the receptor and phosphorylation of downstream signal transduction elements. For the sake of simplicity, we can separate the signaling pathways into two broad functional categories, mitogenic and metabolic, with the understanding that there may be considerable cross-talk between the protein intermediaries. The mitogen-activated protein kinase (MAPK) pathway[66] is responsible for the mitogenic effects of insulin (and insulin-like growth factors), promoting cellular proliferation and growth. The metabolic effects of insulin are principally mediated by phosphatidylinositol-3-kinase (PI-3K). PI-3K-dependent signaling mediates several of the cellular effects of insulin described above and summarized in Figure 24-30 .[67] [68]

PATHOGENESIS OF TYPE 1 DIABETES MELLITUS

This form of diabetes results from a severe lack of insulin caused by an immunologically mediated destruction of β cells. Type 1 diabetes most commonly develops in childhood, becomes manifest at puberty, and progresses with age. Since the disease can develop at any age, including late adulthood, the appellation "juvenile diabetes" is now considered obsolete. Similarly, the older moniker "insulin-dependent diabetes mellitus" (IDDM) has been excluded from the recent classification of diabetes to reflect the emphasis on pathogenic mechanisms rather than mode of therapy.[62] Nevertheless, most patients depend on insulin for survival; without insulin, they develop serious metabolic complications such as acute ketoacidosis and coma. A rare form of "idiopathic" type 1 diabetes has been described in which the evidence for autoimmunity is not definitive. Here we will focus on the typical immune-mediated type 1 diabetes.

Type 1 diabetes is an autoimmune disease in which islet destruction is caused primarily by T lymphocytes reacting against as yet poorly defined β-cell antigens. As in all autoimmune diseases, genetic susceptibility and environmental factors play important roles in the pathogenesis ( Chapter 6 ). We first describe the mechanisms of β-cell destruction and

1193

Figure 24-30 Insulin action on a target cell. Insulin binds to the α subunit of insulin receptor, leading to activation of the kinase activity in the β-subunit, and sets in motion a phosphorylation (i.e., activation) cascade of multiple downstream target proteins. The mitogenic functions of insulin (and the related insulin-like growth factors) are mediated via the mitogen-activated protein kinase (MAP kinase) pathway. The metabolic actions of insulin are mediated primarily by activation of the phosphatidylinositol-3-kinase (PI-3K) pathway. The PI-3K-signaling pathway is responsible for a variety of effects on target cells, including translocation of GLUT-4 containing vesicles to the surface; increasing GLUT-4 density on the membrane and rate of glucose influx; promoting glycogen synthesis via activation of glycogen synthase; and promoting protein synthesis and lipogenesis, while inhibiting lipolysis. The PI-3K pathway also promotes cell survival and proliferation.

then discuss the current ideas about the factors that trigger autoimmune attack against these cells.

Mechanisms of β Cell Destruction

Although the clinical onset of type 1 diabetes is abrupt, this disease in fact results from a chronic autoimmune attack on β cells that usually starts many years before the disease becomes evident ( Fig. 24-31 ). The classic manifestations of the disease (hyperglycemia and ketosis) occur late in its course, after more than 90% of the β cells have been destroyed.

Several mechanisms contribute to β cell destruction:[69]

• T lymphocytes react against β-cell antigens and cause cell damage. These T cells include (1) CD4+ T cells of the TH 1 subset, which cause tissue injury by activating macrophages, and (2) CD8+ cytotoxic T lymphocytes, which directly kill β cells and also secrete cytokines that activate macrophages. In the rare cases

in which the pancreatic lesions have been examined at the early active stages of the disease, the islets show cellular necrosis and lymphocytic infiltration. This lesion is called insulitis. The infiltrates consist of both CD4+ and CD8+ T cells. Surviving β cells often express class II MHC molecules, probably an effect of local production of the cytokine IFN-γ by the T cells. The specificity of these T cells is largely unknown.[70] Various studies have implicated a β-cell enzyme, glutamic acid decarboxylase (GAD), and insulin itself as autoantigens, but the evidence supporting their importance is mainly circumstantial or based on mouse models of the disease. Also, the key question of why tolerance to these self-antigens breaks down has not been answered.[70]

1194

• Locally produced cytokines damage β cells. Among the cytokines implicated in the cell injury are IFN-γ, produced by T cells, and TNF and IL-1, produced by macrophages that are activated during the immune reaction. All these cytokines have been shown to induce β-cell apoptosis in culture; in mouse models of the disease, β-cell destruction can be reduced by treatment with antagonists against these cytokines. • Autoantibodies against islet cells and insulin are also detected in the blood of 70% to 80% of patients. The autoantibodies are reactive with a variety of β-cell antigens, including GAD. These antibodies may participate in causing the disease or may be a result of T cell-mediated cell injury and release of normally sequestered antigens.[71]

Figure 24-31 Stages in the development of type 1 diabetes mellitus. The stages are listed from left to right, and hypothetical β-cell mass is plotted against age. (From Eisenbarth GE: Type 1 diabetes: a chronic autoimmune disease. N Engl J Med 314:1360, 1986. Copyright © 1986, Massachusetts Medical Society. All rights reserved.)

It is likely that many of these immune mechanisms work together to produce progressive destruction of β cells, resulting in clinical diabetes. The factors that predispose to autoimmunity are discussed next.

Genetic Susceptibility

Type 1 diabetes has a complex pattern of genetic associations, and putative susceptibility genes have been mapped to at least 20 loci.[72] Many of these associations are with chromosomal regions, and the particular genes involved are not known yet. Of the multiple loci that are associated with the disease, by far the most important is the class II MHC (HLA) locus; according to some estimates, the MHC contributes about half the genetic susceptibility, and all the other genes combined make up the other half.

The MHC Locus.

The principal susceptibility locus for type 1 diabetes resides in the region that encodes the class II molecules of the MHC on chromosome 6p21 (HLA-D). [73] You will recall that linkage to the HLA locus has also been demonstrated in other autoimmune diseases ( Chapter 6 ). Ninety per cent to 95% of Caucasians with type 1 diabetes have HLA-DR3, DR4, or both, in contrast to about 40% of normal subjects; and 40% to 50% of patients are DR3/DR4 heterozygotes, in contrast to 5% of normal subjects. Interestingly, susceptibility to type 1 diabetes is actually associated with a linked DQ allele called DQB1 0302 that is often in linkage disequilibrium with DR4. Thus, the DQB1 0302 allele is considered the primary determinant of susceptibility for the HLA-DR4 haplotype; in contrast, the HLA-DQB1 0602 allele is considered "protective" against diabetes.[74] Sequencing of DQ molecules associated with diabetes, both in humans and in the nonobese diabetic (NOD) mouse strain, suggests that an asparagine at position 57 in the DQβ chain protects against type 1 diabetes and that its absence increases susceptibility. Although there are many exceptions to this finding, a general hypothesis is that development of type 1 diabetes is influenced by the structure of the entire DQ peptide-binding cleft, with residue 57 playing a significant but not exclusive role. Despite the high relative risk of type 1 diabetes in individuals with particular class II alleles, most individuals who inherit these alleles do not develop the disease. We still do not know precisely how the MHC contributes to autoimmunity in this or in any other autoimmune disease ( Chapter 6 ). Since MHC molecules normally function to display peptides to T cells, these associations clearly point to an important role of T cells in the disease.

Non-MHC Genes.

The first disease-associated non-MHC gene to be identified was insulin, with tandem repeats in the promoter region being associated with disease susceptibility. The mechanism of this association is unknown. It may be that the disease-associated polymorphism makes the protein less functional or stable and thus compromises the functional reserve. Alternatively, these polymorphisms may influence the level of expression of insulin in the thymus, thus altering the negative selection of insulin-reactive T cells ( Chapter 6 ). Recently, another gene has been shown to be associated with the disease, encoding the T-cell inhibitory receptor CTLA-4. Patients with type 1 diabetes show increased frequency of a splice variant that may abrogate the normal ability of this receptor to keep self-reactive T lymphocytes under control.[21]

Environmental Factors

There is evidence that environmental factors, especially infections, are involved in triggering autoimmunity in type 1 diabetes and other autoimmune diseases ( Chapter 6 ). Epidemiologic studies suggest a role of viruses.[75] Seasonal trends that often correspond to the prevalence of common viral infections have long been noted in the diagnosis of new cases, as has the association between coxsackieviruses of group B and pancreatic diseases, including diabetes. Other implicated viral infections include mumps, measles, cytomegalovirus, rubella, and infectious mononucleosis. In all these cases, the viruses are not thought to cause diabetes by directly damaging β cells. Rather, as was discussed in Chapter 6 , two mechanisms, which are not mutually exclusive, have been proposed to explain how infections can trigger autoimmunity.[76] One is that the infections induce

tissue damage and inflammation, leading to the release of β-cell antigens and the recruitment and activation of lymphocytes and other inflammatory leukocytes in the tissue. The other possibility is that the viruses produce proteins that mimic self-antigens and the immune response to the viral protein cross-reacts with the self tissue.[77] Although there is experimental evidence in support of both possibilities, neither has been established as being actually involved. It should also be pointed out that recent epidemiologic studies have shown that in the United States, the incidence of type 1 diabetes in children under 15 years of age has tripled since the 1960s. Similar trends are seen in Western Europe. These findings are often interpreted as suggesting that infections may actually be protective in this disease and the increased incidence reflects the reduction in common infections. Consistent with this possibility, infections also prevent disease development in the nonobese diabetic mouse model.

PATHOGENESIS OF TYPE 2 DIABETES MELLITUS

While much has been learned in recent years, the pathogenesis of type 2 diabetes remains enigmatic. Environmental factors, such as a sedentary life style and dietary habits, clearly play a role, as will become evident when obesity is considered. Nevertheless, genetic factors are even more important than in type 1 diabetes. Among identical twins, the concordance rate is 50% to 90%, while among first-degree relatives with type 2 diabetes (and in fraternal twins), the risk of developing the disease is 20% to 40%, compared to 5% to 7% in the population

1195

at large. Unlike type 1 diabetes, however, the disease is not linked to genes involved in immune tolerance and regulation, and there is no evidence to suggest an autoimmune basis for type 2 diabetes.

The two metabolic defects that characterize type 2 diabetes are (1) a decreased ability of peripheral tissues to respond to insulin (insulin resistance) and (2) β-cell dysfunction that is manifested as inadequate insulin secretion in the face of insulin resistance and hyperglycemia. In most cases, insulin resistance is the primary event, and is followed by increasing degrees of β-cell dysfunction (Fig. 24-32 (Figure Not Available) ).

Insulin Resistance

Insulin resistance is defined as resistance to the effects of insulin on glucose uptake, metabolism, or storage.[78] Insulin resistance is a characteristic feature of most patients with type 2 diabetes and is an almost universal finding in diabetic individuals who are obese. The role of insulin resistance in the pathogenesis of type 2 diabetes can be gauged from the findings that (1) insulin resistance is often detected 10 to 20 years before the onset of diabetes in predisposed individuals (e.g., offspring of type 2 diabetics) and (2) in prospective studies, insulin resistance is the best predictor for subsequent progression to diabetes.[79] Insulin resistance leads to decreased uptake of glucose in muscle and adipose

tissues and an inability of the hormone to suppress hepatic gluconeogenesis. Functional studies in individuals with insulin resistance have demonstrated numerous quantitative and qualitative abnormalities

Figure 24-32 (Figure Not Available) Metabolic staging of type 2 diabetes mellitus. Genetic predisposition and environmental influences converge to cause insulin resistance. Compensatory β-cell hyperplasia can maintain normoglycemia, but eventually, β-cell secretory dysfunction sets in, leading to impaired glucose tolerance and eventually frank diabetes. Rare instances of primary β-cell failure can directly lead to type 2 diabetes without a state of insulin resistance.

of the insulin signaling pathway, including down-regulation of the insulin receptor; decreased insulin receptor phosphorylation and tyrosine kinase activity; reduced levels of active intermediates in the insulin signaling pathway; and impairment of translocation, docking, and fusion of GLUT-4-containing vesicles with the plasma membrane.[65]

It is recognized that insulin resistance is a complex phenomenon. Here we discuss some of the likely culprits responside for decreased sensitivity to insulin in diabetic individuals.

Genetic Defects of the Insulin Receptor and Insulin Signaling Pathway.

Loss-of-function abnormalities of either the insulin receptor or its downstream intermediates are obvious candidates for mediating insulin resistance in type 2 diabetes. In mice, tissue-specific knockout of genes encoding various insulin signaling proteins has resulted in insulin resistance, hyperinsulinemia and hyperglycemia, recapitulating human type 2 diabetes.[80] Unfortunately, the extrapolation of these single-gene knockout models to human disease has been less than gratifying. Point mutations of the insulin receptor are relatively rare, accounting for no more than 1% to 5% of patients with insulin resistance (see the section entitled "Monogenic Forms of Diabetes"). Analysis of candidate genes involved in insulin secretion or insulin action, as well as whole genome linkage studies of affected families, have yielded many polymorphisms that associate with the type 2 diabetic phenotype, but in most cases, the associations have been weak, or the studies were not reproducible.[81] From these analyses, it appears that while the population risk associated with any particular genetic variant (polymorphism) may be significant, the increased risk for developing diabetes for a given individual harboring that variant is small at best. Suffice it to say that while no one questions a genetic component to insulin resistance, the high "noise" to signal ratio has hampered identification of the genes involved. The genetic basis of insulin resistance, and by extension type 2 diabetes, therefore, remains an enigma.

Obesity and Insulin Resistance.

The association of obesity with type 2 diabetes has been recognized for decades, visceral obesity being a common phenomenon in the majority of type 2 diabetics. The link between obesity and diabetes is mediated via effects on insulin resistance.[82] Insulin resistance is present even in simple obesity unaccompanied by hyperglycemia, indicating

a fundamental abnormality of insulin signaling in states of fatty excess. The risk for diabetes increases as the body mass index (a measure of body fat content) increases. It is not only the absolute amount but also the distribution of body fat that has an effect on insulin sensitivity: Central obesity (abdominal fat) is more likely to be linked with insulin resistance than are peripheral (gluteal/subcutaneous) fat depots. Although many details of the so-called adipo-insulin axis remain to be elucidated, following are some of the putative pathways leading to insulin resistance:

• Role of free fatty acids (FFAs): Cross-sectional studies have demonstrated an inverse correlation between fasting plasma FFAs and insulin sensitivity. Furthermore, the level of intracellular triglycerides is often markedly increased in muscle and liver tissues in obese individuals, presumably because excess circulating FFAs are deposited in these organs. Intracellular triglycerides and products of fatty acid metabolism are potent inhibitors of insulin signaling and result in an acquired insulin resistance state.[79] These "lipotoxic"

1196

effects of FFAs are most likely mediated through a decrease in activity of key insulin-signaling proteins. • Role of adipokines in insulin resistance: It is increasingly recognized that adipose tissue is not merely a passive storage depot for fat, but can also operate as a functional endocrine organ, releasing hormones in response to changes in the metabolic status. A variety of proteins released into the systemic circulation by adipose tissue have been identified, and these are collectively termed adipokines (or adipose cytokines).[83] Dysregulation of adipokine secretion (either abnormally increased or decreased levels) may be one of the mechanisms by which insulin resistance is tied to obesity. Several adipokines have been implicated in insulin resistance, including leptin, [84] adiponectin [85] and resistin. [86] For brevity, only the first will be discussed. Leptin acts on central nervous system receptors and other sites to reduce food intake and induce satiety ( Chapter 9 ). Leptin-deficient animals demonstrate severe insulin resistance that is reversed by administration of leptin.[87] Whereas many of leptin's insulin-sensitizing actions are mediated by central nervous system receptors, some effects may be exerted directly at the level of insulin target tissues. The role of leptin in states of insulin resistance in humans is an area of active investigation. • Role of the peroxisome proliferator-activated receptor gamma (PPARγ) and thiazolidinediones (TZDs): TZDs are a class of antidiabetic compounds that were developed in the early 1980s as antioxidants. The target receptor for TZDs has been identified as peroxisome proliferator-activated receptor gamma (PPARγ), a nuclear receptor and transcription factor.[88] PPARγ is most highly expressed in adipose tissue, and activation of the receptor by TZDs results in modulation of gene expression in adipocytes, eventually leading to reduction of insulin resistance. The targets of PPARγ activation include several of the adipokines discussed above. PPARγ activation also decreases levels of free fatty acids, which, as mentioned earlier, contributes to insulin resistance in obesity.

To summarize, insulin resistance in type 2 diabetes is a complex and multifactorial phenomenon. Genetic defects in the insulin signaling pathway are not common and, when present, are more likely polymorphisms with subtle effects rather than inactivating mutations. Insulin resistance is acquired in the overwhelming majority of individuals, and obesity is central to this phenomenon. Figure 24-33 illustrates the links between obesity and insulin resistance.

β-Cell Dysfunction

β-cell dysfunction in type 2 diabetes reflects the inability of these cells to adapt themselves to the long-term demands of peripheral insulin resistance and increased insulin secretion. In states of insulin resistance, insulin secretion is initially higher for each level of glucose than in controls. This hyperinsulinemic state is a compensation for peripheral resistance and can often maintain normal plasma glucose for years. Eventually, however, β-cell compensation becomes inadequate, and there is progression to overt diabetes. The underlying basis for failure of β-cell adaptation is not known, although it is postulated that several mechanisms, including

Figure 24-33 Obesity and insulin resistance: the missing links? Adipocytes release a variety of factors (free fatty acids and adipokines) that may play a role in modulating insulin resistance in peripheral tissues (illustrated here is striated muscle). Excess free fatty acids (FFAs) and resistin are associated with insulin

resistance; in contrast, adiponectin, whose levels are decreased in obesity, is an insulin-sensitizing adipokine. Leptin is also an insulin-sensitizing agent, but it acts via central receptors (in the hypothalamus). The peroxisome proliferator-activated receptor gamma (PPARγ) is an adipocyte nuclear receptor that is activated by a class of insulin-sensitizing drugs called thiazolidinediones (TZDs). The mechanism of action of TZDs may eventually be mediated through modulation of adipokine and FFA levels that favor a state of insulin sensitivity.

adverse effects of high circulating free fatty acids ("lipotoxicity") or chronic hyperglycemia ("glucotoxicity"), may play a role. β-cell dysfunction in type 2 diabetes manifests itself as both qualitative and quantitative defects:

• Qualitative β-cell dysfunction is initially subtle, and seen as loss of the normal pulsatile, oscillating pattern of insulin secretion and attenuation of the rapid first phase of insulin secretion triggered by an elevation in plasma glucose. Over time, the secretory defect affects all phases of insulin secretion, and even though some basal insulin secretion persists in type 2 diabetes, it is grossly inadequate to overcome the insulin resistance. • Quantitative β-cell dysfunction is reflected by a decrease in β-cell mass, islet degeneration, and deposition of islet amyloid. Islet amyloid protein (amylin) is a characteristic finding in patients with type 2 diabetes and is present in more than 90% of diabetic islets examined. Islet amyloidosis is associated with a decrease in β-cell mass, although it is uncertain whether the amyloid is involved in or merely a consequence of the β-cell decrease. Although there are scant data in humans, studies from animal models of diabetes support the aforementioned sequence of events wherein β-cell

1197

hyperplasia in the prediabetic state is followed by a decrease in β-cell mass that coincides with clinical progression to diabetes. In this context, it is important to note that even a "normal" β-cell mass in diabetic individuals may in fact indicate a relative reduction for the degree of insulin resistance.

MONOGENIC FORMS OF DIABETES

Although genetically defined causes of diabetes are uncommon, they have been intensively studied in the hope of gaining insights into the disease. As Table 24-6 illustrates, monogenic forms of diabetes are classified separately from types 1 and 2. Monogenic causes of diabetes result from either a primary defect in β-cell function or a defect in insulin/insulin receptor signaling, as described below.

Maturity-Onset Diabetes of the Young (MODY).

Two per cent to 5% of diabetic patients do not fall clearly into either the type 1 or type 2 diabetes phenotype and are said to have "maturity-onset diabetes of the young." In these patients, there is a primary defect in β-cell function that occurs without β-cell loss, affecting either β-cell mass and/or insulin production.[89] It now appears that MODY is the

outcome of a heterogeneous group of genetic defects characterized by (1) autosomal-dominant inheritance as a monogenic defect, with high penetrance; (2) early onset, usually before age 25, as opposed to after age 40 for most patients with type 2 diabetes; (3) absence of obesity; and (4) lack of islet cell autoantibodies and insulin resistance syndrome.

Six distinct genetic defects have been identified thus far (see Table 24-6 ). Glucokinase, implicated in MODY2, catalyzes the transfer of phosphate from ATP to glucose, which is the first and rate-limiting step in glucose metabolism. Glucokinase expressed in the pancreatic β-cell controls the influx of glucose by controlling its entry into the glycolytic cycle, which in turn is coupled to insulin secretion. Inactivating mutations of this enzyme increase the threshold for insulin release, such that insulin secretion is low for the degree of hyperglycemia present, causing modest increases in blood glucose. Activating mutations have been described that shift the enzyme activity in the opposite direction, with increased insulin secretion at a lower glucose level, resulting in states of chronic hypoglycemia with hyperinsulinism. The remaining five genes associated with MODY are transcription factors controlling insulin expression in β-cells and β-cell mass; IPF-1 also plays a central role in the development of the pancreas. In addition to genetic heterogeneity, MODY is characterized by clinical heterogeneity. Some forms (MODY1, MODY3, and MODY5) are associated with severe β-cell insulin secretory defects with the full range of diabetic complications, while others (MODY2) feature mild chronic hyperglycemia that typically does not worsen over time.

Up to 50% of carriers of glucokinase mutations develop gestational diabetes mellitus, defined as any degree of glucose intolerance with onset or first recognition during pregnancy;[90] conversely, approximately 5% of women with gestational diabetes mellitus and a first-degree relative with diabetes carry a mutation in the glucokinase gene. It is important to emphasize that mutations or polymorphisms in the six known MODY genes do not appear to contribute to the development of late-onset (classic) type 2 diabetes in the vast majority of patients.

Mitochondrial Diabetes.

Mitochondrial DNA is inherited maternally and encodes several genes in the oxidative phosphorylation pathway, ribosomal RNAs, and 22 transfer RNAs (tRNAs). In rare cases, (<1%), diabetes is associated with point mutations in a mitochondrial tRNA gene, tRNALeu(UUR) .[91] Mitochondrial diabetes is caused by a primary defect in β-cell function. Recall that ATP is required for insulin secretion in β cells ( Fig. 24-29 ), and impairment of mitochondrial ATP synthesis results in decreased insulin secretion.

Diabetes Associated with Insulin Gene or Insulin Receptor Mutations.

Mutations that affect insulin processing from its precursor (proinsulin) or those that affect insulin structure and binding to its receptor are a rare cause of diabetes.[92] The metabolic impairment in most cases is mild, since these patients are heterozygous for their mutations. Insulin receptor mutations that affect either receptor synthesis, insulin binding, or receptor tyrosine kinase activity can, in rare cases, result in mild to severe

insulin resistance and type 2 diabetes. Neither insulin gene nor insulin receptor mutations contribute significantly to the incidence of type 2 diabetes.

PATHOGENESIS OF THE COMPLICATIONS OF DIABETES

The morbidity associated with long-standing diabetes of either type results from a number of serious complications, involving both large- and medium-sized muscular arteries (macrovascular disease), as well as capillary dysfunction in target organs (microvascular disease). Macrovascular disease causes accelerated atherosclerosis among diabetics, resulting in increased risk of myocardial infarction, stroke, and lower-extremity gangrene. The effects of microvascular disease are most profound in the retina, kidneys, and peripheral nerves, resulting in diabetic retinopathy, nephropathy, and neuropathy, respectively. Diabetes is the leading cause of blindness and end-stage renal disease in the Western hemisphere, besides contributing substantially to the incidence of cardiovascular events each year. Hence, the basis of long-term complications of diabetes is the subject of a great deal of research. Most of the available experimental and clinical evidence suggests that the complications of diabetes are a consequence of the metabolic derangements, mainly hyperglycemia. For example, when kidneys are transplanted into diabetics from nondiabetic donors, the lesions of diabetic nephropathy may develop within 3 to 5 years after transplantation. Conversely, kidneys with lesions of diabetic nephropathy demonstrate a reversal of the lesion when transplanted into normal recipients. Two large multicenter trials to evaluate the effects of plasma glucose concentrations on long-term complications of diabetes—the Diabetes Control and Complication Trial (DCCT)[93] and the United Kingdom Prospective Diabetes Study (UKPDS)[94] —have convincingly demonstrated delayed progression of microvascular complications by strict control of the hyperglycemia. It is important to stress, however, that not all diabetics have long-term complications, irrespective of the level of blood glucose control over time, indicating that there are additional factors that modulate an individual's risk for microvascular disease. It is likely that such disease-modifying elements are genetic, and there is an ongoing search to identify these additional genes.

At least three distinct metabolic pathways appear to be involved in the pathogenesis of long-term diabetic complications,

1198

although the primacy of any one has not been established.[95] These pathways include the following.

Formation of Advanced Glycation End Products.

Advanced glycation end products (AGEs) are formed as a result of nonenzymatic reactions between intracellular glucose-derived dicarbonyl precursors (glyoxal, methylglyoxal, and 3-deoxyglucosone) with the amino group of both intracellular and extracellular proteins.[96] AGEs have a number of chemical and biologic properties that are

detrimental to extracellular matrix components and the target cells of diabetic complications (e.g., endothelial cells) ( Table 24-7 ):

• On extracellular matrix components, such as collagen or laminin, the formation of AGEs causes cross-linking between polypeptides, resulting in abnormal matrix-matrix and matrix-cell interactions. For example, cross-linking between collagen type I molecules in large vessels decreases their elasticity, which may predispose these vessels to shear stress and endothelial injury ( Chapter 11 ). Similarly, AGE-induced cross-linking of type IV collagen in basement membrane decreases endothelial cell adhesion and increases fluid filtration. AGE cross-linked proteins are resistant to proteolytic digestion. Thus, cross-linking decreases protein removal while enhancing protein deposition. AGE-modified matrix components also trap nonglycated plasma or interstitial proteins. In large vessels, trapping low-density lipoprotein (LDL), for example, retards its efflux from the vessel wall and enhances the deposition of cholesterol in the intima, thus accelerating atherogenesis ( Chapter 11 ). In capillaries, including those of renal glomeruli, plasma proteins such as albumin may bind to the glycated basement membrane, accounting in part for the increased basement membrane thickening characteristic of diabetic microangiopathy. • Circulating plasma proteins are modified by addition of AGE residues; these proteins, in turn, bind to AGE receptors on several cell types (endothelial cells, mesangial cells, macrophages). The AGE-receptor ligation results in activation and nuclear translocation of the pleotropic transcription factor NF-κB, generating a variety of cytokines, growth factors and other pro-inflammatory molecules.[97] The biologic effects of AGE-receptor signaling include (1) release of cytokines and growth factors from macrophages and mesangial cells (insulin-like growth factor-1, TGF-β, platelet-derived growth factor, VEGF); (2) increased endothelial permeability; (3) increased procoagulant activity on endothelial cells and macrophages (induction of thrombomodulin and tissue factor); and (4) enhanced proliferation of and synthesis of extracellular matrix by fibroblasts and smooth muscle cells.

TABLE 24-7 -- Effects of Advanced Glycation End Products (AGEs)

Extracellular Matrix Components

Abnormal matrix-matrix and matrix-cell interactions

Corss-linking of polypeptides of same protein (e.g., collagen)

Trapping of nonglycated proteins (e.g., LDL, albumin)

Resistance to proteolytic digestion

Intracellular and Plasma Proteins

AGE receptor ligation leads to generation of reactive oxygen species and NF-κB activation

Target cells (endothelium, mesangial cells, macrophages) respond by:

Cytokines and growth factor   secretion

Induction of procoagulant   activity

Increased vascular permeability

Enhanced ECM production  

ECM, extracellular matrix; LDL, low-density lipoprotein.

You will recall from the discussion of atherosclerosis ( Chapter 11 ) that endothelial dysfunction, particularly endothelial activation, is a critical process in vascular injury and atherogenesis. AGEs, by virtue of their ability to modify extracellular matrix components, as well as to activate NF-κB and its downstream targets in the vascular endothelium, are postulated to play a central role in the accelerated atherogenesis characteristic of diabetes. In addition to large vessel disease, AGEs also contribute to microvascular injury in diabetes. The AGE inhibitor aminoguanidine has recently been shown to retard the progression of nephropathy in type 1 diabetics.

Activation of Protein Kinase C.

Activation of intracellular protein kinase C (PKC) by calcium ions and the second messenger diacylglycerol (DAG) is an important signal transduction pathway in many cellular systems. Intracellular hyperglycemia can stimulate the de novo synthesis of DAG from glycolytic intermediates and hence cause activation of PKC. The downstream effects of PKC activation are numerous and include the following: [97]

• Production of the proangiogenic molecule vascular endothelial growth factor (VEGF), implicated in the neovascularization characterizing diabetic retinopathy ( Chapter 29 ) • Increased activity of the vasoconstrictor endothelin-1 and decreased activity of the vasodilator endothelial nitric oxide synthase (eNOS) • Production of profibrogenic molecules like transforming growth factor-β (TGF-β), leading to increased deposition of extracellular matrix and basement membrane material • Production of the procoagulant molecule plasminogen activator inhibitor-1 (PAI-1), leading to reduced fibrinolysis and possible vascular occlusive episodes • Production of pro-inflammatory cytokines by the vascular endothelium.

It should be evident that some effects of AGEs and activated PKCs (e.g., activation of NF-κB) are overlapping. Not surprisingly, therefore, therapeutic inhibition of PKC can retard the progression of diabetic retinopathy. [98]

Intracellular Hyperglycemia with Disturbances in Polyol Pathways.

In some tissues that do not require insulin for glucose transport (e.g., nerves, lenses, kidneys, blood vessels), hyperglycemia leads to an increase in intracellular glucose that is

then metabolized by the enzyme aldose reductase to sorbitol, a polyol, and eventually to fructose. In this process, intracellular NADPH is used as a cofactor. NADPH is also required as a cofactor by the enzyme glutathione reductase for regenerating reduced glutathione (GSH). You will recall that GSH is one of the important antioxidant mechanisms in the cell ( Chapter 1 ), and a reduction in GSH levels increases cellular susceptibility to oxidative stress. [99] In the face of sustained hyperglycemia, progressive depletion of intracellular NADPH by aldol reductase leads to a compromise of

1199

GSH regeneration. Thus, the deleterious consequences of the aldose reductase pathway arise primarily by increasing cellular susceptibility to oxidative stress. The importance of this pathway in human diabetes was best exemplified in clinical trials using an aldose reductase inhibitor, which significantly ameliorated the development of diabetic neuropathy. Unfortunately, the effects of these inhibitors on other long-term complications have been less promising.

MORPHOLOGY OF DIABETES AND ITS LATE COMPLICATIONS

Pathologic findings in the pancreas are variable and not necessarily dramatic. The important morphologic changes are related to the many late systemic complications of diabetes. There is extreme variability among patients in the time of onset of these complications, their severity, and the particular organ or organs involved. In individuals with tight control of diabetes, the onset might be delayed. In most patients, however, morphologic changes are likely to be found in arteries (macrovascular disease), basement membranes of small vessels (microangiopathy), kidneys (diabetic nephropathy), retina (retinopathy), nerves (neuropathy), and other tissues. These

Figure 24-34 Long-term complications of diabetes.

change are seen in both type 1 and type 2 diabetes. A schematic overview is provided in Figure 24-34 .

Morphology.

Pancreas.

Lesions in the pancreas are inconstant and rarely of diagnostic value. Distinctive changes are more commonly associated with type 1 than with type 2 diabetes. One or more of the following alterations may be present:

• Reduction in the number and size of islets. This is most often seen in type 1 diabetes, particularly with rapidly advancing disease. Most of the islets are small and inconspicuous, and not easily detected. • Leukocytic infiltration of the islets (insulitis) principally composed of T lymphocytes similar to that in animal models of autoimmune diabetes ( Fig. 24-35 A ). This may be seen in type 1 diabetics at the time of clinical presentation. The distribution of insulitis may be strikingly uneven. Eosinophilic infiltrates may also be found, particularly in diabetic infants who fail to survive the immediate postnatal period.

1200

• By electron microscopy, β-cell degranulation may be observed, reflecting depletion of stored insulin in already damaged β cells. This is more commonly seen in patients with newly diagnosed type 1 disease, when some β cells are still present. • In type 2 diabetes, there may be a subtle reduction in islet cell mass, demonstrated only by special morphometric studies. • Amyloid replacement of islets in type 2 diabetes appears as deposition of pink, amorphous material beginning in and around capillaries and between cells. At advanced stages, the islets may be virtually obliterated ( Fig. 24-35 B ); fibrosis may also be observed. This change is often seen in long-standing cases of type 2 diabetes. Similar lesions may be found in elderly nondiabetics, apparently as part of normal aging. • An increase in the number and size of islets is especially characteristic of nondiabetic newborns of diabetic mothers. Presumably, fetal islets undergo hyperplasia in response to the maternal hyperglycemia.

Figure 24-35 A, Insulitis, shown here from a rat (BB) model of autoimmune diabetes, also seen in type 1 human diabetes. (Courtesy of Dr. Arthur Like, University of Massachusetts, Worchester, MA.) B, Amyloidosis of a pancreatic islet in type 2 diabetes.

Diabetic Macrovascular Disease.

Diabetes exacts a heavy toll on the vascular system. The hallmark of diabetic macrovascular disease is accelerated atherosclerosis involving the aorta and large- and medium-sized arteries. Except for its greater severity and earlier age at onset, atherosclerosis in diabetics is indistinguishable from that in nondiabetics ( Chapter 11 ). Myocardial infarction, caused by atherosclerosis of the coronary arteries, is the most common cause of death in diabetics. Significantly, it is almost as common in diabetic women as in diabetic men. In contrast, myocardial infarction is uncommon in nondiabetic women of reproductive age. Gangrene of the lower extremities, as a result of advanced vascular disease, is about 100 times more common in diabetics than in the general population. The larger renal arteries are also subject to severe atherosclerosis, but the most damaging effect of diabetes on the kidneys is exerted at the level of the glomeruli and the microcirculation. This will be discussed later.

Hyaline arteriolosclerosis, the vascular lesion associated with hypertension ( Chapter 11 and Chapter 20 ), is both more prevalent and more severe in diabetics than in nondiabetics, but it is not specific for diabetes and may be seen in elderly nondiabetics without hypertension. It takes the form of an amorphous, hyaline thickening of the wall of the arterioles, which causes narrowing of the lumen ( Fig. 24-36 ). Not surprisingly, in diabetics, it is related not only to the duration of the disease, but also to the level of blood pressure.

Diabetic Microangiopathy.

One of the most consistent morphologic features of diabetes is diffuse thickening of basement membranes. The thickening is most evident in the capillaries of the skin, skeletal muscle, retina, renal glomeruli, and renal medulla. However, it may also be seen in such nonvascular structures as renal tubules, the Bowman capsule, peripheral nerves, and placenta. By both light and electron microscopy, the basal lamina separating parenchymal or endothelial cells from the surrounding

Figure 24-36 Severe renal hyaline arteriolosclerosis. Note a markedly thickened, tortuous afferent arteriole. The amorphous nature of the thickened vascular wall is evident. (Periodic acid-Schiff [PAS] stain; courtesy of M.A. Venkatachalam, MD, Department of Pathology, University of Texas Health Science Center at San

Antonio, TX.)

1201

tissue is markedly thickened by concentric layers of hyaline material composed predominantly of type IV collagen ( Fig. 24-37 and Fig. 24-38 ). It should be noted that despite the increase in the thickness of basement membranes, diabetic capillaries are more leaky than normal to plasma proteins. The microangiopathy underlies the development of diabetic nephropathy, retinopathy, and some forms of neuropathy. An indistinguishable microangiopathy can be found in aged nondiabetic patients but rarely to the extent seen in patients with long-standing diabetes.

Diabetic Nephropathy.

The kidneys are prime targets of diabetes. (See also Chapter 20 .) Renal failure is second only to myocardial infarction as a cause of death from this disease. Three lesions are encountered: (1) glomerular lesions; (2) renal vascular lesions, principally arteriolosclerosis; and (3) pyelonephritis, including necrotizing papillitis.

The most important glomerular lesions are capillary basement membrane thickening, diffuse mesangial sclerosis, and nodular glomerulosclerosis. These are described in detail in Chapter 20 . The glomerular capillary basement membranes are thickened throughout their entire length (see Fig. 24-38 ). This change can be detected by electron microscopy within a few years of the onset of diabetes, sometimes without any associated change in renal function.

Diffuse mesangial sclerosis consists of a diffuse increase in mesangial matrix and is always associated with basement membrane thickening. It is found in most patients with disease of more than 10 years' duration. When glomerulosclerosis becomes marked, patients manifest the nephrotic syndrome ( Chapter 20 ), characterized by proteinuria, hypoalbuminemia, and edema.

Nodular glomerulosclerosis describes a glomerular lesion made distinctive by ball-like deposits of a laminated matrix situated in the periphery of the glomerulus. These nodules are PAS positive and usually contain trapped mesangial cells. This distinctive change has been called the Kimmelstiel-Wilson lesion, after the pathologists who described it. Nodular glomerulosclerosis is encountered in approximately 15% to 30% of long-term diabetics and is a major cause of morbidity and mortality. Diffuse

Figure 24-37 Renal cortex showing thickening of tubular basement membranes in a diabetic patient (PAS stain).

Figure 24-38 Electron micrograph of a renal glomerulus showing markedly thickened glomerular basement membrane (B) in a diabetic. L, glomerular capillary lumen; U, urinary space. (Courtesy of Dr. Michael Kashgarian, Department of Pathology, Yale University School of Medicine, New Haven, CT.)

mesangial sclerosis may also be seen in association with old age and hypertension; on the contrary, the nodular form of glomerulosclerosis, once certain unusual forms of nephropathies have been excluded (see Chapter 20 ), is essentially pathognomonic of diabetes. Both the diffuse and nodular forms of glomerulosclerosis induce sufficient ischemia to cause overall fine scarring of the kidneys, marked by a finely granular cortical surface ( Fig. 24-39 ).

Renal atherosclerosis and arteriolosclerosis constitute part of the macrovascular disease in diabetics. The kidney is one of the most frequently and severely affected organs; however, the changes in the arteries and arterioles are similar to those found throughout the body. Hyaline arteriolosclerosis affects not only the afferent but also the efferent arteriole. Such efferent arteriolosclerosis is rarely, if ever, encountered in individuals who do not have diabetes.

Pyelonephritis is an acute or chronic inflammation of the kidneys that usually begins in the interstitial tissue and then spreads to affect the tubules. Both the acute and chronic forms of this disease occur in nondiabetics as well as in diabetics but are more common in diabetics than in the general population, and, once affected, diabetics tend to have more severe involvement. One special pattern of acute pyelonephritis, necrotizing papillitis (or papillary necrosis), is much more prevalent in diabetics than in nondiabetics.

Diabetic Ocular Complications.

The ocular involvement may take the form of retinopathy, cataract formation, or glaucoma. The morphologic features are discussed further in Chapter 29 .

Diabetic Neuropathy.

The central and peripheral nervous systems are not spared by diabetes. The effects of diabetes on the nervous system are described further in Chapter 27 and Chapter 28 .

1202

Figure 24-39 Nephrosclerosis in a patient with long-standing diabetes. The kidney has been bisected to demonstrate both diffuse granular transformation of the surface (left) and marked thinning of the cortical

tissue (right). Additional features include some irregular depressions, the result of pyelonephritis, and an incidental cortical cyst (far right).

CLINICAL FEATURES OF DIABETES

It is difficult to sketch with brevity the diverse clinical presentations of diabetes mellitus. Only a few characteristic patterns will be presented.

Type 1 diabetes was traditionally thought to occur primarily in those under age 18 but is now known to occur at any age. In the initial 1 or 2 years following manifestation of overt type 1 diabetes, the exogenous insulin requirements may be minimal because of ongoing endogenous insulin secretion (referred to as the honeymoon period), but shortly thereafter, any residual β-cell reserve is exhausted and insulin requirements increase dramatically. Although β-cell destruction is a long-standing process, the transition from impaired glucose tolerance to overt diabetes may be abrupt, heralded by an event with increased insulin requirements, such as infection.

The onset is marked by polyuria, polydipsia, polyphagia, and, with extreme derangement, ketoacidosis, all resulting from metabolic derangements. As insulin is a major anabolic hormone in the body, deficiency of insulin results in a catabolic state that affects not only glucose metabolism but also fat and protein metabolism. Unopposed secretion of counter-regulatory hormones (glucagon, growth hormone, epinephrine) also plays a role in these metabolic derangements. The assimilation of glucose into muscle and adipose tissue is sharply diminished or abolished. Not only does storage of glycogen in liver and muscle cease, but also reserves are depleted by glycogenolysis. The resultant hyperglycemia exceeds the renal threshold for reabsorption, and glycosuria ensures. The glycosuria induces an osmotic diuresis and thus polyuria, causing a profound loss of water and electrolytes ( Fig. 24-40 ). The obligatory renal water loss combined with the hyperosmolarity resulting from the increased levels of glucose in the blood tends to deplete intracellular water, triggering the osmoreceptors of the thirst centers of the brain. In this manner, intense thirst (polydipsia) appears. With a deficiency of insulin, the scales swing from insulin-promoted anabolism to catabolism of proteins and fats. Proteolysis follows, and the gluconeogenic amino acids are removed by the liver and used as building blocks for glucose. The catabolism of proteins and fats tends to induce a negative energy balance, which in turn leads to increasing appetite (polyphagia), thus completing the classic triad of diabetes: polyuria, polydipsia, and polyphagia. Despite the increased appetite, catabolic effects prevail, resulting in weight loss and muscle weakness. The combination of polyphagia and weight loss is paradoxical and should always raise the suspicion of diabetes.

Diabetic ketoacidosis (DKA) is a serious complication of type 1 diabetes but may also occur in type 2 diabetes, though not as commonly and not to as marked an extent. These patients have marked insulin deficiency, and the release of the catecholamine hormone epinephrine blocks any residual insulin action and stimulates the release of glucagon. The insulin deficiency coupled with glucagon excess decreases peripheral utilization of glucose while increasing gluconeogenesis, severely exacerbating hyperglycemia (the

plasma glucose levels are usually in the range of 500 to 700 mg/dL). The hyperglycemia causes an osmotic diuresis and dehydration characteristic of the ketoacidotic state. The second major effect of an alteration in the insulin:glucagon ratio is activation of the ketogenic machinery. Insulin deficiency stimulates lipoprotein lipase, with resultant excessive breakdown of adipose stores, and an increase in levels of free fatty acids. When these free fatty acids reach the liver, they are esterified to fatty acyl CoA. Oxidation of fatty acyl CoA molecules within the hepatic mitochondria produces ketone bodies (acetoacetic acid and β-hydroxybutyric acid). The rate at which ketone bodies are formed may exceed the rate at which acetoacetic acid and β-hydroxybutyric acid can be utilized by peripheral tissues, leading to ketonemia and ketonuria. If the urinary excretion of ketones is compromised by dehydration, the plasma hydrogen ion concentration increases, and systemic metabolic ketoacidosis results. Release of ketogenic amino acids by protein catabolism aggravates the ketotic state.

Type 2 diabetes mellitus may also present with polyuria and polydipsia, but unlike in type 1 diabetes, patients are often older (over 40 years) and frequently obese. However, with the increase in obesity and sedentary lifestyle in our society, type 2 diabetes is now seen in children and adolescents with increasing frequency. In some cases, medical attention is sought because of unexplained weakness or weight loss. Most frequently, however, the diagnosis is made after routine blood or urine testing in asymptomatic persons. The absence of ketoacidosis and milder presentation in type 2 diabetes is presumably because of higher portal vein insulin levels in these patients than in type 1 diabetics, which prevents unrestricted hepatic fatty acid oxidation and keeps the formation of ketone bodies in check. In the decompensated state, these patients may develop hyperosmolar nonketotic coma, a syndrome engendered by the severe dehydration resulting from sustained osmotic diuresis in patients who do not drink enough water to compensate for urinary losses from chronic hyperglycemia. Typically, the patient is an elderly diabetic who is disabled by a stroke or an infection and is unable to maintain adequate water intake. Furthermore, the absence of ketoacidosis and its symptoms (nausea, vomiting, respiratory difficulties) delays the seeking of medical attention until severe dehydration and coma occur. In Table 24-8 , we have summarized some of the

1203

Figure 24-40 Sequence of metabolic derangements leading to diabetic coma in type 1 diabetes mellitus. An absolute insulin deficiency leads to a catabolic state, eventuating in ketoacidosis and severe volume depletion. These cause sufficient central nervous system compromise to lead to coma and eventual death if left untreated.

1204

TABLE 24-8 -- Type 1 Versus Type 2 Diabetes Mellitus (DM)

Type 1 DM Type 2 DM

Clinical Onset: <20 years Onset: >30 years

Normal weight Obese

Markedly decreased blood insulin

Increased blood insulin (early);normal to moderate decreased insulin (late)

Anti-islet cell antibodies No anti-islet cell antibodies

Ketoacidosis common Ketoacidosis rare; nonketotic hyperosmolar coma

Genetics 30–70% concordance in twins 50–90% concordance in twins

Linkage to MHC Class II HLA genes

No HLA linkage

Linkage to candidate diabetogenic genes (PPARγ, calpain 10)

Pathogenesis Autoimmune destruction of β-cells mediated by T cells and humoral mediators (TNF, IL-1, NO)

Insulin resistance in skeletal muscle, adipose tissue and liver

β-cell dysfunction and relative insulin deficiency

Absolute insulin deficiency

Islet cells Insulitis early No insulitis

Marked atrophy and fibrosis Focal atrophy and amyloid deposition

β-cell depletion Mild β-cell depletion

pertinent clinical, genetic, and histopathologic features that distinguish type 1 and type 2 diabetes.

In both forms, it is the long-term effects of diabetes, more than the acute metabolic complications, that are responsible for the overwhelming proportion of morbidity and mortality. In most instances, these complications appear approximately 15 to 20 years after the onset of hyperglycemia. Cardiovascular events such as myocardial infarction, renal vascular insufficiency, and cerebrovascular accidents are the most common causes of mortality in long-standing diabetics. The impact of cardiovascular disease can be gauged from the fact that it accounts for up to 80% of deaths in type 2 diabetes; in fact, diabetics have a 3 to 7.5 times greater incidence of death from cardiovascular causes compared to the nondiabetic population[100] ( Fig. 24-41 ). The hallmark of cardiovascular disease is accelerated atherosclerosis of the large and medium-sized arteries (i.e., macrovascular disease). The pathogenesis of accelerated atherosclerosis involves multiple factors. We have previously

Figure 24-41 Incidence of death from cardiovascular causes in diabetic and nondiabetic individuals after a 7-year follow up. MI, myocardial infarction. (Reproduced with permission from Haffner et al: Mortality from coronary heart disease in subjects with type 2 diabetes and in nondiabetic subjects with and without myocardial infarction. N Engl J Med 339:229, 1998.)

mentioned the contribution of AGEs to endothelial dysfunction and vascular disease in diabetes. You will recall that the binding of AGE-modified plasma proteins to AGE receptors on endothelial and vascular smooth muscle leads to generation of a variety of proatherogenic cytokines and growth factors. Blockade of one such receptor—receptor for AGE, or RAGE—suppressed macrovascular disease in an atherosclerosis-prone animal model, underscoring the importance of this pathway in atherogenesis.[101] Activation of protein kinase C, with resultant impairment of vasodilation and increased procoagulant

PAI-1 activity, may also contribute to the endothelial injury and accelerated atherosclerosis in diabetes.

The importance of obesity in the pathogenesis of insulin resistance has already been discussed, but it is also an independent risk factor for development of atherosclerosis ( Chapter 11 ). Additional risk factors for atherosclerosis that are present in many type 2 diabetics include hypertension, dyslipidemia, and platelet dysfunction.[102] Hypertension is approximately twice as frequent in diabetics as in those without the disease. Similar to atherosclerosis, the increased frequency of hypertension may be a manifestation of hyperglycemia-induced endothelial dysfunction. Dyslipidemias include both increased triglycerides and LDL levels and decreased levels of the "protective" lipoprotein, HDL ( Chapter 11 ); hepatic insulin resistance combined with peripheral activation of lipoprotein lipase plays a key role in maintaining a "proatherogenic" lipoprotein profile in diabetic individuals. Finally, increased platelet adhesiveness to the vessel wall is observed in states of insulin resistance, possibly owing to increased thromboxane A2 synthesis and reduced prostacyclin.

Diabetic nephropathy is a leading cause of end-stage renal disease in the United States.[103] Approximately 30% to 40% of all diabetics develop clinical evidence of nephropathy, but a considerably smaller fraction of patients with type 2 diabetes progress to end-stage renal disease. However, because of the much greater prevalence of type 2 diabetes, these patients constitute slightly over half the diabetic patients starting dialysis each year. The frequency of diabetic nephropathy is greatly influenced by the genetic makeup of the population in question; for example, Native Americans, Hispanics, and African Americans have a greater risk of developing end-stage renal disease than do non-Hispanic whites with type 2 diabetes. The

1205

earliest manifestation of diabetic nephropathy is the appearance of low amounts of albumin in the urine (>30 mg/day, but <300 mg/day), that is, microalbuminuria. Of note, microalbuminuria is also a marker for greatly increased cardiovascular morbidity and mortality for patients with either type 1 or type 2 diabetes. Therefore, all patients with microalbuminuria should be screened for macrovascular disease, and aggressive intervention should be undertaken to reduce cardiovascular risk factors. Without specific interventions, approximately 80% of type 1 diabetics and 20% to 40% of type 2 diabetics will develop overt nephropathy with macroalbuminuria (>300 mg of urinary albumin per day) over 10 to 15 years, usually accompanied by the appearance of hypertension. The progression from overt nephropathy to end-stage renal disease can be highly variable. By 20 years, more than 75% of type 1 diabetics and approximately 20% of type 2 diabetics with overt nephropathy will develop end-stage renal disease, requiring dialysis or renal transplantation. Diabetic nephropathy is also discussed in Chapter 20 .

Visual impairment, sometimes even total blindness, is one of the more feared consequences of long-standing diabetes. This disease is currently the fourth leading cause

of acquired blindness in the United States. Approximately 60% to 80% of patients develop some form of diabetic retinopathy approximately 15 to 20 years after diagnosis ( Chapter 29 ). As was previously stated, the fundamental lesion of retinopathy—neovascularization—is probably attributable to VEGF signaling in the retina.[104] In addition to retinopathy, diabetics also have an increased propensity for glaucoma and cataract formation, both of which contribute to visual impairment in diabetes. Diabetic neuropathy can elicit a variety of clinical syndromes, afflicting the central nervous system, peripheral sensorimotor nerves, and the autonomic nervous system; these are discussed further in Chapter 27 and Chapter 28 .

Diabetics are plagued by enhanced susceptibility to infections of the skin and to tuberculosis, pneumonia, and pyelonephritis. Such infections cause the deaths of about 5% of diabetic patients. In an individual with diabetic neuropathy, a trivial infection in a toe may be the first event in a long succession of complications (gangrene, bacteremia, pneumonia) that may ultimately lead to death.

In recent years, increasingly sedentary lifestyles and poor eating habits have contributed to the simultaneous escalation of diabetes and obesity worldwide, which some have termed as "diabesity."[105] Sadly, this "epidemic" of diabetes and obesity has percolated even to children. As the incidence of communicable diseases has declined and expected life span has increased, diabetes has become a major public health problem, and it continues to be one of the top 10 "killers" in the United States. There is hope, however, since the role of primary prevention of type 2 diabetes by lifestyle and dietary alterations and secondary prevention of diabetic complications by strict glycemic control has become increasingly recognized. It is also hoped that islet cell transplantation will result in a cure for those afflicted with type 1 diabetes.

Pancreatic Endocrine Neoplasms

The preferred term for tumors of the pancreatic islet cells, referred to as "islet cell tumors" in common medical parlance, is pancreatic endocrine neoplasms.[106] They are rare in comparison with tumors of the exocrine pancreas, accounting for only 2% of all pancreatic neoplasms. They are most common in adults and can occur anywhere along the length of the pancreas, embedded in the substance of the pancreas or arising in the immediate peripancreatic tissues. They resemble in appearance their counterparts, carcinoid tumors, found elsewhere in the alimentary tract ( Chapter 17 ). Pancreatic endocrine neoplasms may be single or multiple and benign or malignant, the latter metastasizing to lymph nodes and liver. When multiple, each tumor may be composed of a different cell type. Pancreatic endocrine neoplasms have a propensity to elaborate pancreatic hormones, but some may be totally nonfunctional.

Like any other endocrine neoplasms in the body (see below), it is difficult to predict the biologic behavior of a pancreatic endocrine neoplasm based on light microscopic criteria alone. Unequivocal criteria for malignancy include (1) metastases to regional lymph nodes or distant organs (including the liver), (2) vascular invasion, and (3) gross invasion of adjacent viscera. Other features suggestive of malignancy include infiltration beyond

the tumor capsule into the pancreatic parenchyma, a high mitotic index, tumor necrosis, and significant cellular atypia. In general, tumors less than 2 cm in diameter tend to behave in an indolent manner, but there are significant exceptions to this rule. Finally, the functional status of the tumor might have some import on prognosis, as approximately 90% of insulinomas are benign, while 60% to 90% of other functioning and nonfunctioning pancreatic endocrine neoplasms tend to be malignant.[107] Fortunately, insulinomas are also the most common subtype of pancreatic endocrine neoplasms.

The three most common and distinctive clinical syndromes associated with functional pancreatic endocrine neoplasms are (1) hyperinsulinism, (2) hypergastrinemia and the Zollinger-Ellison syndrome, and (3) multiple endocrine neoplasia (the last is described in detail later).

HYPERINSULINISM (INSULINOMA)

β-cell tumors (insulinomas) are the most common of pancreatic endocrine neoplasms and may be responsible for the elaboration of sufficient insulin to induce clinically significant hypoglycemia. There is a characteristic clinical triad resulting from these pancreatic lesions: (1) Attacks of hypoglycemia occur with blood glucose levels below 50 mg/dL of serum; (2) the attacks consist principally of such central nervous system manifestations as confusion, stupor, and loss of consciousness; and (3) the attacks are precipitated by fasting or exercise and are promptly relieved by feeding or parenteral administration of glucose.

Morphology.

Insulinomas are most often found within the pancreas and are generally benign. Most are solitary lesions, although multiple tumors or tumors ectopic to the pancreas may be encountered. Bona fide carcinomas, making up only about 10% of cases, are diagnosed on the basis of criteria for malignancy listed above. On rare occasions, an insulinoma may arise in ectopic pancreatic tissue.

Solitary tumors are usually small (often less than 2 cm in diameter) and are encapsulated, pale to red-brown

1206

nodules located anywhere in the pancreas. Histologically, these benign tumors look remarkably like giant islets, with preservation of the regular cords of monotonous cells and their orientation to the vasculature. Not even the malignant lesions present much evidence of anaplasia ( Fig. 24-42 A ), and they may be deceptively encapsulated. By immunocytochemistry, insulin can be localized in the tumor cells ( Fig. 24-42 B ). Under the electron microscope, neoplastic β cells, like their normal counterparts, display distinctive round granules that contain polygonal or rectangular dense crystals separated

from the enclosing membrane by a distinct halo. It should be cautioned that granules may be present in the absence of clinically significant hormone activity.

Hyperinsulinism may also be caused by diffuse hyperplasia of the islets.[108] This change is found occasionally in adults but is usually encountered in neonates and infants. Several clinical scenarios may result in diffuse islet hyperplasia (previously known as nesidioblastosis), including maternal diabetes, Beck-with-Wiedemann syndrome ( Chapter 10 ), and rare metabolic disorders. In maternal diabetes, the fetus, long exposed to the hyperglycemia of maternal blood, responds by an increase in the size and number of its islets. In the postnatal period, these hyperactive islets may be responsible for serious episodes of hypoglycemia. This phenomenon is usually transient, although persisting problems may result from mutations in the glucose-sensing mechanism or insulin-secreting mechanisms within the β cell.

While up to 80% of islet cell tumors may demonstrate excessive insulin secretion, the hypoglycemia is mild in all but about 20%, and many cases never become clinically symptomatic. The critical laboratory findings in insulinomas are high circulating levels of insulin and a high insulin-glucose ratio. Surgical removal of the tumor is usually followed by prompt reversal of the hypoglycemia.

It is important to note that there are many other causes of hypoglycemia besides insulinomas. The differential diagnosis of this frequently obscure metabolic abnormality includes such

Figure 24-42 Pancreatic endocrine tumor ("islet cell tumor"). A, The neoplastic cells are monotonous and demonstrate minimal pleomorphism or mitotic activity (H & E stain). B, Immunoreactivity for insulin confirms the neoplasm is an insulinoma. Clinically, the patient had episodic hypoglycemia.

conditions as abnormal insulin sensitivity, diffuse liver disease, inherited glycogenoses, and ectopic production of insulin by certain retroperitoneal fibromas and fibrosarcomas.

ZOLLINGER-ELLISON SYNDROME (GASTRINOMAS)

Marked hypersecretion of gastrin usually has its origin in gastrin-producing tumors (gastrinomas), which are just as likely to arise in the duodenum and peripancreatic soft tissues as in the pancreas (so-called gastrinoma triangle).[109] There has been lack of agreement regarding the cell of origin for these tumors, although it appears likely that endocrine cells of either the gut or the pancreas could be the source. Zollinger and Ellison first called attention to the association of pancreatic islet cell lesions with hypersecretion of gastric acid and severe peptic ulceration, [110] which are present in 90% to 95% of patients.

Morphology.

Gastrinomas may arise in the pancreas, the peripancreatic region, or the wall of the duodenum. Over half of gastrin-producing tumors are locally invasive or have already metastasized at the time of diagnosis. In approximately 25% of patients, gastrinomas arise in conjunction with other endocrine tumors, thus conforming to the MEN-1 syndrome (see below); MEN-1-associated gastrinomas are frequently multifocal, while sporadic gastrinomas are usually single. As with insulin-secreting tumors of the pancreas, gastrin-producing tumors are histologically bland and rarely exhibit marked anaplasia.

In the Zollinger-Ellison syndrome, hypergastrinemia from a pancreatic or duodenal tumor stimulates extreme gastric acid secretion, which in turn causes peptic ulceration. The duodenal and gastric ulcers are often multiple; although they are identical to those found in the general population, they are often unresponsive to usual modalities of therapy. In addition, ulcers may also occur in unusual locations such as the jejunum;

1207

when intractable jejunal ulcers are found, Zollinger-Ellison syndrome should be considered. More than 50% of the patients have diarrhea; in 30%, it is the presenting symptom.

Treatment of Zollinger-Ellison syndrome involves control of gastric acid secretion by use of H+ /K+ -ATPase inhibitors ( Chapter 17 ) and excision of the neoplasm. Total resection of the neoplasm, when possible, eliminates the syndrome. Patients with hepatic metastases have a significantly shortened life expectancy, with progressive tumor growth leading to liver failure usually within 10 years.

OTHER RARE PANCREATIC ENDOCRINE NEOPLASMS

α-cell tumors (glucagonomas) are associated with increased serum levels of glucagon and a syndrome consisting of mild diabetes mellitus, a characteristic skin rash (necrolytic migratory erythema), and anemia. They occur most frequently in perimenopausal and postmenopausal women and are characterized by extremely high plasma glucagon levels.

δ-cell tumors (somatostatinomas) are associated with diabetes mellitus, cholelithiasis, steatorrhea, and hypochlorhydria. They are exceedingly difficult to localize preoperatively. High plasma somatostatin levels are required for diagnosis.

VIPoma (watery diarrhea, hypokalemia, achlorhydria, or WDHA syndrome) is an endocrine tumor that induces a characteristic syndrome, caused by release of vasoactive intestinal peptide (VIP) from the tumor. Some of these tumors are locally invasive and metastatic. A VIP assay should be performed on all patients with severe secretory diarrhea. Neural crest tumors, such as neuroblastomas, ganglioneuroblastomas, and ganglioneuromas ( Chapter 10 ) and pheochromocytomas (see below) can also be associated with the VIPoma syndrome.

Pancreatic carcinoid tumors producing serotonin and an atypical carcinoid syndrome are exceedingly rare. Pancreatic polypeptide-secreting endocrine tumors are endocrinologically asymptomatic, despite the presence of high levels of the hormone in plasma.

Some pancreatic and extrapancreatic endocrine tumors produce two or more hormones, usually simultaneously and occasionally in sequence. In addition to insulin, glucagon, and gastrin, pancreatic endocrine tumors may produce adrenocorticotropic hormone, melanocyte-stimulating hormone, vasopressin, serotonin, and norepinephrine. These multihormonal tumors are to be distinguished from the multiple endocrine neoplasias (see below), in which a multiplicity of hormones is produced by tumors in several different glands.

Adrenal Glands

Adrenal Cortex

Normal

The adrenal glands are paired endocrine organs consisting of both cortex and medulla, which differ in their development, structure, and function. In the adult, the normal adrenal gland weighs about 4 gm; but with acute stress, lipid depletion may reduce the weight, or prolonged stress, such as dying after a long chronic illness, can induce hypertrophy and hyperplasia of the cortical cells and more than double the weight of the normal gland. Beneath the capsule of the adrenal is the narrow layer of zona glomerulosa. An equally narrow zona reticularis abuts the medulla. Intervening is the broad zona

fasciculata, which makes up about 75% of the total cortex. The adrenal cortex synthesizes three different types of steroids: (1) glucocorticoids (principally cortisol), which are synthesized primarily in the zona fasciculata with a small contribution from the zona reticularis; (2) mineralocorticoids, the most important being aldosterone, which is generated in the zona glomerulosa; and (3) sex steroids (estrogens and androgens), which are produced largely in the zona reticularis. The adrenal medulla is composed of chromaffin cells, which synthesize and secrete catecholamines, mainly epinephrine. Catecholamines have many effects that allow rapid adaptations to changes in the environment.

Pathology

Diseases of the adrenal cortex can be conveniently divided into those associated with cortical hyperfunction and those characterized by cortical hypofunction.

ADRENOCORTICAL HYPERFUNCTION (HYPERADRENALISM)

Just as there are three basic types of corticosteroids elaborated by the adrenal cortex (glucocorticoids, mineralocorticoids, and sex steroids), so there are three distinctive hyperadrenal clinical syndromes: (1) Cushing syndrome, characterized by an excess of cortisol; (2) hyperaldosteronism; and (3) adrenogenital or virilizing syndromes caused by an excess of androgens. The clinical features of these syndromes overlap somewhat because of the overlapping functions of some of the adrenal steroids.

Hypercortisolism (Cushing Syndrome)

Pathogenesis.

This disorder is caused by any condition that produces an elevation in glucocorticoid levels.[111] There are four possible sources of excess cortisol ( Fig. 24-43 ). In clinical practice, most causes of Cushing syndrome are the result of the administration of exogenous glucocorticoids. The other

1208

Figure 24-43 A schematic representation of the various forms of Cushing syndrome, illustrating the three endogenous forms as well as the more common exogenous (iatrogenic) form. ACTH, adrenocorticotropic hormone.

three sources of the hypercortisolism can be categorized as endogenous Cushing syndrome:

• Primary hypothalamic-pituitary diseases associated with hypersecretion of ACTH • Hypersecretion of cortisol by an adrenal adenoma, carcinoma, or nodular hyperplasia • The secretion of ectopic ACTH by a nonendocrine neoplasm

Primary hypersecretion of ACTH accounts for 70% to 80% of cases of endogenous hypercortisolism. In recognition of the neurosurgeon who first published the full description of this syndrome and related it to a pituitary lesion,[112] this pituitary form of Cushing syndrome is referred to as Cushing disease. The disorder affects women about five times more frequently than men, and it occurs most frequently during the twenties and thirties. In the vast majority of cases, the pituitary gland contains an ACTH-producing microadenoma that does not produce mass effects in the brain; some corticotroph tumors qualify as macroadenomas (>10 mm). In most of the remaining cases, the anterior pituitary contains areas of corticotroph cell hyperplasia without a discrete adenoma. Corticotroph cell hyperplasia may be primary or may arise secondarily from excessive stimulation of ACTH release by a hypothalamic corticotropin releasing hormone (CRH)-producing tumor. The adrenal glands in patients with Cushing disease

are characterized by variable degrees of nodular cortical hyperplasia (discussed later), caused by the elevated levels of ACTH. The cortical hyperplasia, in turn, is responsible for hypercortisolism.

Primary adrenal neoplasms, such as adrenal adenoma and carcinoma, and primary cortical hyperplasia are responsible for about 10% to 20% of cases of endogenous Cushing syndrome. This form of Cushing syndrome is also designated ACTH-independent Cushing syndrome or adrenal Cushing syndrome because the adrenals function autonomously. The biochemical sine qua non of adrenal Cushing syndrome is elevated serum levels of cortisol with low levels of ACTH.

• Adenomas and carcinomas are about equally common in adults; in children, carcinomas predominate. The cortical carcinomas tend to produce more marked hypercortisolism than the adenomas or hyperplastic processes. In instances of a unilateral neoplasm, the uninvolved adrenal cortex and that in the opposite gland undergo atrophy because of suppression of ACTH secretion. • The overwhelming majority of hyperplastic adrenals arise from secondary influences, and primary cortical hyperplasia is uncommon. Two types of primary bilateral adrenocortical hyperplasia have been described in association with Cushing syndrome. In massive macronodular adrenocortical disease (MMAD), the nodules are usually greater than 3 mm in diameter.[113] MMAD affects older adults, and there is no known genetic component. The second variant of primary nodular hyperplasia, seen more often in children than in adults, is a familial condition known as primary pigmented nodular adrenal disease (PPNAD).[114] The adrenal glands

1209

in PPNAD demonstrate diffuse bilateral micronodules (<3 mm in diameter) that are usually darkly pigmented (brown to black).

Secretion of ectopic ACTH by nonpituitary tumors accounts for most of the remaining cases (∼ 10%) of Cushing syndrome. In many instances, the responsible tumor is a small cell carcinoma of the lung, although other neoplasms, including carcinoid tumors, medullary carcinomas of the thyroid, and islet cell tumors of the pancreas, have been associated with the syndrome. In addition to tumors that elaborate ectopic ACTH, an occasional neuroendocrine neoplasm produces ectopic corticotropin-releasing hormone, which, in turn, causes ACTH secretion and hypercortisolism. As in the pituitary variant, the adrenal glands undergo bilateral cortical hyperplasia, but often the rapid downhill course of the patients with these cancers cuts short the adrenal enlargement. This variant of Cushing syndrome is more common in men and usually occurs in the forties and fifties.

Morphology.

The main lesions of Cushing syndrome are found in the pituitary and adrenal glands. The pituitary in Cushing syndrome shows changes regardless of the cause. The most common alteration, resulting from high levels of endogenous or exogenous glucocorticoids, is termed Crooke hyaline change. In this condition, the normal granular, basophilic cytoplasm of the ACTH-producing cells in the anterior pituitary is replaced by homogeneous, lightly basophilic material. This alteration is the result of the accumulation of intermediate keratin filaments in the cytoplasm.

The morphology of the adrenal glands depends on the cause of the hypercortisolism. The adrenals have one of the following abnormalities: (1) cortical atrophy; (2) diffuse hyperplasia; (3) nodular hyperplasia; and (4) an adenoma, rarely a carcinoma. In patients in whom the syndrome results from exogenous glucocorticoids, suppression of endogenous ACTH results in bilateral cortical atrophy, due to a lack of stimulation of the zonae fasciculata and reticularis by ACTH. The zona glomerulosa is of normal thickness in such cases because this portion of the cortex functions independently of ACTH. In cases of endogenous hypercortisolism, in contrast, the adrenals either are hyperplastic or contain a cortical neoplasm. Diffuse hyperplasia is found in 60% to 70% of cases of Cushing syndrome. Both glands are enlarged, either subtly or markedly, weighing up to 25 to 40 gm. The adrenal cortex is diffusely thickened and yellow, owing to an increase in the size and number of lipid-rich cells in the zonae fasciculata and reticularis. Some degree of nodularity is common but is pronounced in nodular hyperplasia. This takes the form of bilateral, 0.5- to 2.0-cm, yellow nodules scattered throughout the cortex, separated by intervening areas of widened cortex. The uninvolved cortex and nodules are composed of a mixture of lipid-laden clear cells and lipid-poor compact cells showing some variability in cell and nuclear size with occasional binucleate forms. The combined adrenals may weigh up to 30 to 50 gm. Most cases of hyperplasia are associated with elevated serum levels of ACTH, whether of pituitary or ectopic origin. Primary adrenocortical neoplasms causing Cushing syndrome may be malignant or benign. Adenomas or carcinomas of the adrenal cortex as the source of cortisol secretion are not macroscopically distinctive from nonfunctioning adrenal neoplasms to be described later. Both the benign and the malignant lesions are more common in women in their thirties to fifties. The adrenocortical adenomas are yellow tumors surrounded by thin or well-developed capsules, and most weigh less than 30 gm. Microscopically, they are composed of cells that are similar to those encountered in the normal zona fasciculata. Their morphology is identical to that of nonfunctional adenomas and of adenomas associated with hyperaldosteronism (see below). The carcinomas associated with Cushing syndrome, by contrast, tend to be larger than the adenomas. These tumors are unencapsulated masses frequently exceeding 200 to 300 gm in weight, having all of the anaplastic characteristics of cancer, as will be detailed later. With functioning tumors, both benign and malignant, the adjacent adrenal cortex and that of the contralateral adrenal gland are atrophic, owing to suppression of endogenous ACTH by high cortisol levels.

Clinical Course.

Developing slowly over time, Cushing syndrome, similar to many other endocrine abnormalities, can be quite subtle in its early manifestations. Early stages of the disorder

may present with hypertension and weight gain ( Table 24-9 ). With time, the more characteristic central pattern of adipose tissue deposition becomes apparent, with resultant truncal obesity, moon facies, and accumulation of fat in the posterior neck and back (buffalo hump). Hypercortisolism causes selective atrophy of fast-twitch (type 2) myofibers, resulting in decreased muscle mass and proximal limb weakness. Glucocorticoids induce gluconeogenesis and inhibit the uptake of glucose by cells, with resultant hyperglycemia, glucosuria, and polydipsia; Cushing syndrome is an important cause of secondary diabetes. The catabolic effects on proteins cause loss of collagen and resorption of bones. Consequently, the skin is thin, fragile, and easily bruised; wound healing is poor; and cutaneous striae are particularly common in the abdominal area. Bone resorption results in the development of osteoporosis, with consequent backache and increased susceptibility to fractures. Patients with Cushing syndrome are at increased risk for a variety of infections

TABLE 24-9 -- Major Features of Cushing Syndrome with Approximate Frequency

Clinical Features Percentages

Central obesity (about trunk and upper back) 85–90%

Moon facies 85%

Weakness and fatigability 85%

Hirsutism 75%

Hypertension 75%

Plethora 75%

Glucose intolerance/diabetes 75/20%

Osteoporosis 75%

Neuropsychiatric abnormalities 75–80%

Menstrual abnormalities 70%

Skin striae (sides of lower abdomen) 50%

1210

because glucocorticoids suppress the immune response. Additional manifestations include a number of mental disturbances, including mood swings, depression, and frank psychosis, as well as hirsutism and menstrual abnormalities.

Cushing syndrome is diagnosed in the laboratory with the following: (1) the 24-hour urine free cortisol level, which is increased, and (2) loss of normal diurnal pattern of

cortisol secretion. Determining the cause of Cushing syndrome depends on the level of serum ACTH and measurement of urinary steroid excretion after administration of dexamethasone. Three general patterns can be obtained:[111]

1. In pituitary Cushing syndrome, the most common form, ACTH levels are elevated and cannot be suppressed by the administration of a low dose of dexamethasone. Hence, there is no reduction in urinary excretion of 17-hydroxy-corticosteroids. After higher doses of injected dexamethasone, however, the pituitary responds by reducing ACTH secretion, which is reflected by suppression of urinary steroid secretion.

2. Ectopic ACTH secretion results in an elevated level of ACTH, but its secretion is completely insensitive to low or high doses of exogenous dexamethasone.

3. When Cushing syndrome is caused by an adrenal tumor, the ACTH level is quite low because of feedback inhibition of the pituitary. As with ectopic ACTH secretion, both low-dose and high-dose dexamethasone fail to suppress cortisol excretion.

Primary Hyperaldosteronism

Hyperaldosteronism is the generic term for a small group of closely related, uncommon syndromes, all characterized by chronic excess aldosterone secretion. Excessive levels of aldosterone cause sodium retention and potassium excretion, with resultant hypertension and hypokalemia. Hyperaldosteronism may be primary, or it may be a secondary event resulting from an extra-adrenal cause.

Primary hyperaldosteronism indicates an autonomous overproduction of aldosterone, with resultant suppression of the renin-angiotensin system and decreased plasma renin activity. Primary hyperaldosteronism is caused by one of three mechanisms[115] ( Fig. 24-44 ):

• Adrenocortical neoplasm, either an aldosterone-producing adrenocortical adenoma (the most common cause) or, rarely, an adrenocortical carcinoma. In approximately 80% of cases, primary hyperaldosteronism is caused by a solitary aldosterone-secreting adenoma, a condition referred to as Conn syndrome. This syndrome occurs most frequently in adult middle life and is more common in women than in men (2:1). Multiple adenomas may be present in an occasional patient. • Primary adrenocortical hyperplasia (idiopathic hyperaldosteronism), characterized by bilateral nodular hyperplasia of the adrenal glands, highly reminiscent of those found in the nodular hyperplasia of Cushing syndrome. The genetic basis of idiopathic hyperaldosteronism is not clear, although it is possibly caused by an overactivity of the aldosterone synthase gene, CYP11B2. [116] • Glucocorticoid-remediable hyperaldosteronism is an uncommon cause of primary hyperaldosteronism that is familial and genetic. In some families, it is caused by a chimeric gene resulting from fusion between CYP11B1 (the 11β-hydroxylase gene) and CYP11B2 (the aldosterone synthase gene).[117] This leads to a sustained production of hybrid steroids in addition to both cortisol and

aldosterone. The activation of aldosterone secretion is under the influence of ACTH and hence is suppressible by exogenous administration of dexamethasone.

In secondary hyperaldosteronism, in contrast, aldosterone release occurs in response to activation of the renin-angiotensin system ( Chapter 4 ). It is characterized by increased levels of plasma renin and is encountered in conditions such as the following:

• Decreased renal perfusion (arteriolar nephrosclerosis, renal artery stenosis) • Arterial hypovolemia and edema (congestive heart failure, cirrhosis, nephrotic syndrome) • Pregnancy (due to estrogen-induced increases in plasma renin substrate).

Morphology.

Aldosterone-producing adenomas are almost always solitary, small (<2 cm in diameter), well-circumscribed lesions, more often found on the left than on the right. They tend to occur in the thirties and forties, and in women more often than in men. These lesions are often buried within the gland and do not produce visible enlargement, a point to be remembered in interpreting sonographic or scanning images. They are bright yellow on cut section ( Fig. 24-45 ) and, surprisingly, are composed of lipid-laden cortical cells that more closely resemble fasciculata cells than glomerulosa cells (the normal source of aldosterone). In general, the cells tend to be uniform in size and shape and resemble mature cortical cells; occasionally, there is some nuclear and cellular pleomorphism but no evidence of anaplasia ( Fig. 24-46 ). A characteristic feature of aldesterone-producing adenomas is the presence of eosinophilic, laminated cytoplasmic inclusions, known as spironolactone bodies, found after treatment with the anti-hypertensive drug spironolactone. In contrast to cortical adenomas associated with Cushing syndrome, those associated with hyperaldosteronism do not usually suppress ACTH secretion. Therefore, the adjacent adrenal cortex and that of the contralateral gland are not atrophic.

Bilateral idiopathic hyperplasia ( Fig. 24-47 ) is marked by diffuse and focal hyperplasia of cells resembling those of the normal zona glomerulosa. The hyperplasia is often wedge-shaped, extending from the periphery toward the center of the gland. Bilateral enlargement can be subtle in idiopathic hyperplasia, and as a rule, an adrenocortical adenoma should be carefully excluded as the cause for hyperaldosteronism.

Clinical Course.

The clinical manifestations of primary hyperaldosteronism are hypertension and hypokalemia. Serum renin, as was mentioned previously, is low. Hypokalemia results from renal potassium wasting and can cause a variety of neuromuscular manifestations, including weakness, paresthesias, visual disturbances, and occasionally frank tetany. Sodium retention increases the total body sodium and

1211

Figure 24-44 The major causes of primary hyperaldosteronism and its principal effects on the kidney.

expands the extracellular fluid volume, leading to elevation of the serum sodium concentration and an increase in intracellular sodium with increased vascular reactivity. The hypertension is, in part, a result of the sodium retention. The expanded extracellular fluid volume and hypokalemia both impose a burden on the heart, sometimes causing electrocardiographic changes and cardiac decompensation. The diagnosis of primary hyperaldosteronism is confirmed by the elevated levels of aldosterone and depressed levels of renin in the circulation. Even when the diagnosis of primary hyperaldosteronism is made, it is necessary to distinguish among the various causes, particularly the differentiation of an adenoma, which is amenable to surgical excision. Primary adrenal hyperplasia associated with hyperaldosteronism occurs more often in children and young adults than in older adults; surgical intervention is not very beneficial in these patients, who are best managed with medical therapy with an aldosterone antagonist such as spironolactone. Uncommon as primary hyperaldosteronism is, it should not be overlooked clinically, because it provides an opportunity to cure a form of hypertension. The treatment of secondary hyperaldosteronism rests on correcting the underlying cause stimulating the renin-angiotensin system.

Adrenogenital Syndromes

Disorders of sexual differentiation, such as virilization or feminization, can be caused by primary gonadal disorders ( Chapter 22 ) and several primary adrenal disorders. The adrenal cortex secretes two compounds—dehydroepiandrosterone and androstenedione—that require conversion to testosterone in peripheral tissues for their androgenic effects. Unlike gonadal androgens, ACTH regulates adrenal androgen formation ( Fig. 24-48 ); thus, excess secretion can occur either as a "pure" syndrome or as a component of Cushing disease. The adrenal causes of androgen excess include adrenocortical

1212

Figure 24-45 Adrenal cortical adenoma. The adenoma is distinguished from nodular hyperplasia by its solitary, circumscribed nature. The functional status of an adrenal cortical adenoma cannot be predicted from its gross or microscopic appearance.

neoplasms and a group of disorders that have been designated congenital adrenal hyperplasia.

Adrenocortical neoplasms associated with virilization are more likely to be androgen-secreting adrenal carcinomas than adenomas. Conversely, functioning adrenal cortical carcinomas are most often associated with a virilization syndrome, usually in combination with hypercortisolism ("mixed syndrome"). These tumors are morphologically identical to other cortical neoplasms and will be discussed later.

Congenital adrenal hyperplasia (CAH) represents a group of autosomal-recessive, inherited metabolic errors, each characterized by a deficiency or total lack of a particular enzyme involved in the biosynthesis of cortical steroids, particularly cortisol.[118] Steroidogenesis is then channeled into other pathways, leading to increased production of androgens, which accounts for virilization. Simultaneously, the deficiency of cortisol results in increased secretion of ACTH, resulting in adrenal hyperplasia. Certain enzyme defects may also impair aldosterone secretion, adding salt wasting to the virilizing syndrome. Other enzyme deficiencies may be incompatible with life or, in rare instances, may involve only the aldosterone pathway without involving cortisol synthesis. Thus, there is a spectrum of these syndromes, and with each one, there may be a total lack of a particular enzyme or a mutation that only mildly impairs the effectiveness of the enzyme. The following remarks focus on the most common of these disorders.

21-Hydroxylase Deficiency.

Defective conversion of progesterone to 11-deoxycorticosterone by 21-hydroxylase (CYP21B) accounts for over 90% of cases of congenital adrenal hyperplasia.[119] Figure 24-48 illustrates normal adrenal steroidogenesis and the consequences of 21-hydroxylase deficiency. 21-Hydroxylase deficiency may range from a total lack to a mild loss, depending on the nature of the CYP21B mutation. Three distinctive syndromes have been described: (1)

Figure 24-46 Histologic features of an adrenal cortical adenoma. The neoplastic cells are vacuolated because of the presence of intracytoplasmic lipid. There is mild nuclear pleomorphism. Mitotic activity and necrosis are not seen.

salt-wasting (classic) adrenogenitalism, (2) simple virilizing adrenogenitalism, and (3) nonclassic adrenogenitalism, which implies mild disease that may be entirely asymptomatic or associated only with symptoms of androgen excess during childhood or puberty.

The carrier frequency of the classic form is approximately 1 in 60, while the carrier frequency of the nonclassic or mild form is 1 in 5 to 1 in 50, depending on the ethnic group; Hispanics and Ashkenazi Jewish populations have the highest carrier frequencies. The incidence of classic 21-hydroxylase deficiency varies somewhat between populations, with a

Figure 24-47 Nodular hyperplasia of the adrenal contrasted with normal adrenal gland. In cross-section, the adrenal cortex is yellow, thickened, and multinodular, owing to hypertrophy and hyperplasia of the lipid-rich zonae fasciculata and reticularis.

1213

Figure 24-48 Consequences of C-21 hydroxylase deficiency. 21-Hydroxylase deficiency impairs the synthesis of both cortisol and aldosterone. The resultant decrease in feedback inhibition (dashed line) causes increased secretion of adrenocorticotropic hormone, resulting ultimately in adrenal hyperplasia and increased synthesis of testosterone. The sites of action of 11-, 17-, and 21-hydroxylase are shown by the numbers in circles.

worldwide mean of around 1 in 13,000 newborns. The mechanism of CYP21B gene inactivation in 21-hydroxylase deficiency involves recombination with a neighboring pseudogene on chromosome 6p21 called CYP21A (a pseudogene is an inactive homologous gene created by ancestral duplication in a localized region of the genome). In the majority of cases of CAH, portions of the CYP21A pseudogene replace all or part of the active CYP21B gene. The introduction of nonfunctional sequences from CYP21A into the CYP21B sequence has the same effect as inactivating mutations in CYP21B.

The salt-wasting syndrome results from an inability to convert progesterone into deoxycorticosterone because of a total lack of the hydroxylase. Thus, there is virtually no synthesis of mineralocorticoids, and concomitantly, there is a block in the conversion of hydroxyprogesterone into deoxycortisol with deficient cortisol synthesis. This pattern usually comes to light soon after birth because in utero the electrolytes and fluids can be maintained by the maternal kidneys. There is salt wasting, hyponatremia, and hyperkalemia, which induce acidosis, hypotension, cardiovascular collapse, and possibly death. The concomitant block in cortisol synthesis and excess production of androgens, however, lead to virilization, which is easily recognized in the female at birth or in utero but is difficult to recognize in the male. Various degrees of virilization are encountered, ranging from mild clitoral enlargement to complete labioscrotal fusion to marked clitoral enlargement enclosing the urethra, thus producing a phalloid organ. Males with this disorder are generally unrecognized at birth but come to clinical attention 5 to 15 days later because of some salt-losing crisis.

Simple virilizing adrenogenital syndrome without salt wasting (presenting as genital ambiguity) may occur in individuals with a less than total 21-hydroxylase defect because with less severe deficiencies the level of mineralocorticoid, although reduced, is sufficient for salt reabsorption, but the lowered glucocorticoid level fails to cause feedback inhibition of ACTH secretion. Thus, the level of aldosterone is mildly reduced, testosterone is increased, and ACTH is elevated, with resultant adrenal hyperplasia.

Nonclassic or late-onset adrenal virilism is much more common than the classic patterns already described. Patients with this syndrome may be virtually asymptomatic or have mild manifestations, such as hirsutism. The diagnosis can be made only by demonstration of biosynthetic defects in steroidogenesis and by genetic studies.

Morphology.

In all cases of CAH, the adrenals are bilaterally hyperplastic, sometimes expanding to 10 to 15 times their normal weights because of the sustained elevation in ACTH. The adrenal cortex is thickened and nodular, and on cut section, the widened cortex appears brown, owing to total depletion of all lipid. The proliferating cells are mostly compact, eosinophilic, lipid-depleted cells, intermixed with lipid-laden clear cells. Hyperplasia of corticotroph (ACTH-producing) cells is present in the anterior pituitary in most CAH patients.

Clinical Course.

The clinical features of these disorders are determined by the specific enzyme deficiency and include

1214

abnormalities related to androgen excess and aldosterone and glucocorticoid deficiency. CAH affects not only adrenal cortical enzymes, but also products synthesized in the medulla. High levels of intra-adrenal glucocorticoids are required to facilitate medullary catecholamine (epinephrine and norepinephrine) synthesis. In patients with severe salt-wasting 21-hydroxylase deficiency, a combination of low cortisol levels and developmental defects of the medulla (adrenomedullary dysplasia) profoundly affects catecholamine secretion, further predisposing these individuals to hypotension and circulatory collapse.[120]

Depending on the nature and severity of the enzymatic defect, the onset of clinical symptoms may occur in the perinatal period, later childhood, or, less commonly, adulthood. For example, in 21-hydroxylase deficiency, excessive androgenic activity causes signs of masculinization in females, ranging from clitoral hypertrophy and pseudohermaphroditism in infants, to oligomenorrhea, hirsutism, and acne in postpubertal females. In males, androgen excess is associated with enlargement of the external genitalia and other evidence of precocious puberty in prepubertal patients and oligospermia in older males.

CAH should be suspected in any neonate with ambiguous genitalia; severe enzyme deficiency in infancy can be a life-threatening condition with vomiting, dehydration, and salt wasting. In the milder variants, women may present with delayed menarche, oligomenorrhea, or hirsutism. Patients with congenital adrenal hyperplasia are treated with exogenous glucocorticoids, which, in addition to providing adequate levels of glucocorticoids, also suppress ACTH levels and thus decrease the excessive synthesis of the steroid hormones responsible for many of the clinical abnormalities. Mineralocorticoid supplementation is required in the salt-wasting variants of CAH. With the availability of routine neonatal metabolic screens for CAH and the feasibility of molecular testing for antenatal detection of 21-hydroxylase mutations, the outcome for even the most severe variants has improved significantly.

ADRENAL INSUFFICIENCY

Adrenocortical insufficiency, or hypofunction, may be caused by either primary adrenal disease (primary hypoadrenalism) or decreased stimulation of the adrenals owing to a deficiency of ACTH (secondary hypoadrenalism) ( Table 24-10 ). The patterns of adrenocortical insufficiency can be considered under the following headings: (1) primary acute adrenocortical insufficiency (adrenal crisis), (2) primary chronic adrenocortical insufficiency (Addison disease), and (3) secondary adrenocortical insufficiency.

Primary Acute Adrenocortical Insufficiency

Acute adrenal cortical insufficiency occurs in a variety of clinical settings (see Table 24-10 ):

• As a crisis in patients with chronic adrenocortical insufficiency precipitated by any form of stress that requires an immediate increase in steroid output from glands incapable of responding

• In patients maintained on exogenous corticosteroids, in whom rapid withdrawal of steroids or failure to increase steroid doses in response to an acute stress may precipitate an adrenal crisis, owing to the inability of the atrophic adrenals to produce glucocorticoid hormones • As a result of massive adrenal hemorrhage, which destroys the adrenal cortex sufficiently to cause acute adrenocortical insufficiency. This occurs in newborns following prolonged and difficult delivery with considerable trauma and hypoxia, leading to extensive adrenal hemorrhages beginning in the medulla and extending into the cortex. Newborns are particularly vulnerable because they are often deficient in prothrombin for at least several days after birth. It also occurs in some patients maintained on anticoagulant therapy, in postsurgical patients who develop disseminated intravascular coagulation with consequent hemorrhagic infarction of the adrenals, and when massive adrenal hemorrhage complicates a bacteremic infection; in this last setting, it is called Waterhouse-Friderichsen syndrome.

TABLE 24-10 -- Adrenocortical Insufficiency

Primary Insufficiency

Loss of cortex

Congential adrenal   hypoplasia

X-linked adrenal hypoplasia     (DAX-1 gene on Xp21)

"Miniature" type adrenal hypoplasia     (unknown cause)

Adrenoleukodystrophy (  ALD gene on Xq28)

Autoimmune adrenal   insufficiency

Autoimmune polyendocrinopathy     syndrome type 1 (AIRE-1 gene on 21q22)

Autoimmune polyendocrinopathy     syndrome type 2 (polygenic)

Isolated autoimmune adrenalitis     (polygenic)

Infection  

Acquired immune deficiency     syndrome

Tuberculosis    

Fungi    

Acute hemorrhagic necrosis     (Waterhouse-Friderichsen syndrome)

Amyloidosis, sarcoidosis,   hemochromatosis

Metastatic carcinoma  

Metabolic failure in hormone production

Congenital adrenal   hyperplasia (cortisol and aldosterone deficiency with

virlization)

Drug- and steroid-induced inhibition of   adrenocorticotropic hormone or cortical cell function

Secondary Insufficiency

Hypothalamic pituitary disease

Neoplasm, inflammation (sarcoidosis,   tuberculosis, pyogens, fungi)

Hypothalamic pituitary suppression

Long-term steroid   administration

Steroid-producing   neoplasms

Waterhouse-Friderichsen Syndrome

This uncommon but catastrophic syndrome is characterized by the following:

• An overwhelming bacterial infection, which is classically associated with Neisseria meningitidis septicemia but occasionally is caused by other highly virulent organisms, such as Pseudomonas species, pneumococci, Haemophilus influenzae, or staphylococci • Rapidly progressive hypotension leading to shock • Disseminated intravascular coagulation with widespread purpura, particularly of the skin

1215

• Rapidly developing adrenocortical insufficiency associated with massive bilateral adrenal hemorrhage

Figure 24-49 Waterhouse-Friderichsen syndrome in a child. The dark, hemorrhagic adrenal glands are distended with blood.

Waterhouse-Friderichsen syndrome can occur at any age but is somewhat more common in children. The basis for the adrenal hemorrhage is uncertain but could be attributable to direct bacterial seeding of small vessels in the adrenal, the development of disseminated intravascular coagulation, endotoxin-induced vasculitis, or some form of hypersensitivity vasculitis. Whatever the basis, the adrenals are converted to sacs of clotted blood virtually obscuring all underlying detail ( Fig. 24-49 ). Histologic examination reveals that the hemorrhage starts within the medulla in relationship to thin-walled venous sinusoids, then suffuses peripherally into the cortex, often leaving islands of recognizable cortical cells ( Fig. 24-50 ). When it is recognized promptly and treated effectively with antibiotics, recovery is possible, but the clinical course is usually devastatingly abrupt, and prompt recognition and appropriate therapy must be instituted immediately, or death follows within hours to a few days.

Primary Chronic Adrenocortical Insufficiency (Addison Disease)

In a paper published in 1855, Thomas Addison described a group of patients suffering from a constellation of symptoms, including "general languor and debility, remarkable feebleness of the heart's action, and a peculiar change in the color of the skin" associated with disease of the "suprarenal capsules" or, in more current terminology, the adrenal glands. Addison disease, or chronic adrenocortical insufficiency, is an uncommon disorder resulting from progressive destruction of the adrenal cortex. In general, clinical manifestations of adrenocortical insufficiency do not appear until at least 90% of the adrenal cortex has been compromised. The causes of chronic adrenocortical insufficiency are listed in Table 24-10 . Although all races and both sexes may be affected, certain causes of Addison disease (such as autoimmune adrenalitis) are much more common in whites, particularly in women.

Pathogenesis.

A large number of diseases may attack the adrenal cortex, including lymphomas, amyloidosis, sarcoidosis, hemochromatosis, fungal infections, and adrenal hemorrhage, but more than 90% of all cases are attributable to one of four disorders: autoimmune adrenalitis, tuberculosis, the acquired immune deficiency syndrome (AIDS), or metastatic cancers.

Autoimmune adrenalitis accounts for 60% to 70% of cases, and it is by far the most common cause of primary adrenal insufficiency in developed countries.[121] As the name implies, there is autoimmune destruction of steroidogenic cells, and autoantibodies to several key steroidogenic enzymes (21-hydroxylase, 17-hydroxylase) have been detected in these patients. Autoimmune adrenalitis can occur in one of three clinical settings:

• Autoimmune polyendocrine syndrome type 1 (APS1) is also known as autoimmune polyendocrinopathy, candidiasis, and ectodermal dystrophy (APECED). APS1 is characterized by chronic mucocutaneous candidiasis and abnormalities of skin, dental enamel, and nails (ectodermal dystrophy) occurring in association with a combination of organ-specific autoimmune disorders (autoimmune adrenalitis, autoimmune hypoparathyroidism, idiopathic hypogonadism, pernicious anemia) that result in immune destruction of target organs.[122] APS1 is caused by mutations in the autoimmune regulator (AIRE) gene on chromosome

1216

21q22. The expression of AIRE protein is primarily in the thymus, where it appears to function as a transcription factor that promotes the expression of many self-antigens, leading to negative selection (death) of self-reactive T cells[54] ( Chapter 6 ). • Autoimmune polyendocrine syndrome type 2 (APS2) usually starts in early adulthood and presents as a combination of adrenal insufficiency with autoimmune thyroiditis or type 1 diabetes. Unlike in APS1, mucocutaneous candidiasis, ectodermal dysplasia, and autoimmune hypoparathyroidism do not

occur. APS2, unlike APS1, is not a monogenic disorder, although some studies have suggested a possible association with polymorphisms in the HLA loci.[123] • Isolated autoimmune Addison disease presents with autoimmune destruction restricted to the adrenal glands. However, in terms of age at presentation and linkage to HLA and other susceptibility loci, isolated autoimmune adrenalitis overlaps with APS2, suggesting that the former may be a variant of the latter.

Figure 24-50 Waterhouse-Friderichsen syndrome. At autopsy, the adrenals were grossly hemorrhagic and shrunken; microscopically, little residual cortical architecture is discernible.

Infections, particularly tuberculosis and those produced by fungi, may also cause primary chronic adrenocortical insufficiency. Tuberculous adrenalitis, which once accounted for as much as 90% of Addison disease, has become less common with the development of antituberculous agents. With the resurgence of tuberculosis in most urban centers and the persistence of the disease in developing countries, however, this cause of adrenal insufficiency must be kept in mind. When present, tuberculous adrenalitis is usually associated with active infection in other sites, particularly in the lungs and genitourinary tract. Among the fungi, disseminated infections caused by Histoplasma capsulatum and Coccidioides immitis may also result in chronic adrenocortical insufficiency. Patients with AIDS are at risk for developing adrenal insufficiency from several infectious (cytomegalovirus, Mycobacterium avium-intercellulare) and noninfectious complications (Kaposi sarcoma).

Metastatic neoplasms involving the adrenals are another potential cause of adrenal insufficiency. The adrenals are a fairly common site for metastases in patients with disseminated carcinomas. Although adrenal function is preserved in most such patients, the metastatic tumors occasionally destroy enough adrenal cortex to produce a degree of adrenal insufficiency. Carcinomas of the lung and breast are the source of a majority of metastases in the adrenals, although many other neoplasms, including gastrointestinal

carcinomas, malignant melanoma, and hematopoietic neoplasms, may also metastasize to this organ.

Genetic disorders of adrenal insufficiency include adrenal hypoplasia congenital (AHC) and adrenoleukodystrophy. Technically, these disorders are also associated with chronic adrenal insufficiency, although they are not commonly included in the causes of Addison disease. Adrenoleukodystrophy is described in Chapter 28 . Congenital adrenal hypoplasia is rare, and will not be discussed further.[124] [125]

Morphology.

The anatomic changes in the adrenal glands depend on the underlying disease. Primary autoimmune adrenalitis is characterized by irregularly shrunken glands, which may be difficult to identify within the suprarenal adipose tissue. Histologically, the cortex contains only scattered residual cortical cells in a collapsed network of connective tissue. A variable lymphoid infiltrate is present in the cortex and may extend into the subjacent medulla, although the medulla is otherwise preserved ( Fig. 24-51 ). In cases of tuberculous and fungal disease, the adrenal architecture is effaced by a granulomatous inflammatory reaction identical to that encountered in other sites of infection. When hypoadrenalism is caused by metastatic carcinoma, the adrenals are enlarged, and their normal architecture is obscured by the infiltrating neoplasm.

Clinical Course.

Addison disease begins insidiously and does not come to attention until at least 90% of the cortex of both glands is destroyed and the levels of circulating glucocorticoids and mineralocorticoids are significantly decreased. The initial manifestations include progressive weakness and easy fatigability, which may be dismissed as nonspecific complaints. Gastrointestinal disturbances are common and include anorexia, nausea, vomiting, weight loss, and diarrhea. In patients with primary adrenal disease, increased circulating levels of ACTH precursor hormone stimulate melanocytes, with resultant hyperpigmentation of the skin, particularly of sun-exposed areas and at pressure points, such as the neck, elbows, knees, and knuckles. By contrast, hyperpigmentation is not seen in patients with adrenocortical insufficiency caused by primary pituitary or hypothalamic disease. Decreased mineralocorticoid activity in patients with primary adrenal insufficiency results in potassium and sodium loss, with consequent hyperkalemia, hyponatremia, volume depletion, and hypotension. Hypoglycemia may occasionally occur as a result of glucocorticoid deficiency and impaired gluconeogenesis. Stresses such as infections, trauma, or surgical procedures in such patients can precipitate an acute adrenal crisis, manifested by intractable vomiting, abdominal pain, hypotension, coma, and vascular collapse. Death occurs rapidly unless corticosteroid therapy begins immediately.

Secondary Adrenocortical Insufficiency

Any disorder of the hypothalamus and pituitary, such as metastatic cancer, infection, infarction, or irradiation, that

Figure 24-51 Autoimmune adrenalitis. In addition to loss of all but a subcapsular rim of cortical cells, there is an extensive mononuclear cell infiltrate.

1217

reduces the output of ACTH leads to a syndrome of hypoadrenalism that has many similarities to Addison disease. Analogously, prolonged administration of exogenous glucocorticoids suppresses the output of ACTH and adrenal function. With secondary disease, the hyperpigmentation of primary Addison disease is lacking because melanotropic hormone levels are low. The manifestations also differ in that secondary hypoadrenalism is characterized by deficient cortisol and androgen output but normal or near-normal aldosterone synthesis. Thus, in adrenal insufficiency secondary to pituitary malfunction, marked hyponatremia and hyperkalemia are not seen.

ACTH deficiency can occur alone, but in some instances, it is only one part of panhypopituitarism, associated with multiple primary trophic hormone deficiencies. The differentiation of secondary disease from Addison disease can be confirmed with demonstration of low levels of plasma ACTH in the former. In patients with primary disease, the destruction of the adrenal cortex does not permit a response to exogenously administered ACTH in the form of increased plasma levels of cortisol, whereas in those with secondary hypofunction, there is a prompt rise in plasma cortisol levels.

Morphology.

In cases of hypoadrenalism secondary to hypothalamic or pituitary disease (secondary hypoadrenalism), depending on the extent of ACTH lack, the adrenals may be moderately to markedly reduced in size. They are reduced to small, flattened structures that usually retain their yellow color owing to a small amount of residual lipid. The cortex may be reduced to a thin ribbon composed largely of zona glomerulosa. The medulla is unaffected.

ADRENOCORTICAL NEOPLASMS

It should be evident from the preceding sections that functional adrenal neoplasms may be responsible for any of the various forms of hyperadrenalism. While functional adenomas are most commonly associated with hyperaldosteronism and Cushing syndrome, a virilizing neoplasm is more likely to be a carcinoma. However, not all adrenocortical neoplasms elaborate steroid hormones. Determination of whether a cortical neoplasm is functional or not is based on clinical evaluation and measurement of the hormone or its metabolites in the laboratory. In other words, functional and nonfunctional adrenocortical neoplasms cannot be distinguished on the basis of morphologic features.

Morphology.

Most adrenocortical adenomas are clinically silent and are usually encountered as incidental findings at the time of autopsy or during abdominal imaging for an unrelated cause (see the discussion of adrenal incidentalomas below). Some experts believe that all adrenal adenomas should, by definition, demonstrate clinical or biochemical evidence of hyperfunction and that the incidentally discovered "tumors" are best classified as hyperplastic nodules.[126] In either case, the typical cortical adenoma is a well-circumscribed, nodular lesion up to 2.5 cm in diameter that expands the adrenal. In contrast to functional adenomas, which are associated with atrophy of the adjacent cortex, the cortex adjacent to nonfunctional adenomas is of normal thickness. On cut surface, adenomas are usually yellow to yellow-brown because of the presence of lipid within the tumor cells. Microscopically, adenomas are composed of cells similar to those populating the normal adrenal cortex. The nuclei tend to be small, although some degree of pleomorphism may be encountered even in benign lesions ("endocrine atypia"). The cytoplasm of the neoplastic cells ranges from eosinophilic to vacuolated, depending on their lipid content ( Fig. 24-46 ). Mitotic activity is generally inconspicuous.

Adrenocortical carcinomas are rare neoplasms that can occur at any age, including childhood. They are more likely to be functional than adenomas are, and carcinomas are therefore often associated with virilism or other clinical manifestations of hyperadrenalism. Two rare inherited causes of adrenal cortical carcinomas are Li-Fraumeni syndrome ( Chapter 7 ) and Beckwith-Wiedemann syndrome ( Chapter 10 ). In most cases, adrenocortical carcinomas are large, invasive lesions, many exceeding 20 cm in diameter, that efface the native adrenal gland. The less common, smaller, and better-circumscribed lesions may be difficult to distinguish from an adenoma. On cut surface, adrenocortical carcinomas are typically variegated, poorly demarcated lesions containing areas of necrosis, hemorrhage, and cystic change ( Fig. 24-52 ). Invasion of contiguous structures, including the adrenal vein and inferior vena cava, is common. Microscopically, adrenocortical carcinomas may be composed of well-differentiated cells resembling those seen in cortical adenomas or bizarre, monstrous giant cells ( Fig. 24-53 ), which may be difficult to distinguish from those of an undifferentiated carcinoma metastatic to the adrenal. Between these extremes are found cancers with moderate degrees of anaplasia, some composed predominantly of spindle cells. Carcinomas, particularly those of bronchogenic origin, may metastasize to the adrenals, and they may be extremely difficult to differentiate from primary cortical carcinomas. Adrenal cancers

have a strong tendency to invade the adrenal vein, vena cava, and lymphatics. Metastases to regional and periaortic nodes are common, as is distant hematogenous

Figure 24-52 Adrenal carcinoma. The hemorrhagic and necrotic tumor dwarfs the kidney and compresses the upper pole.

1218

Figure 24-53 Adrenal carcinoma (A) revealing marked anaplasia, contrasted with normal cortical cells (B).

spread to the lungs and other viscera. Bone metastases are unusual. The median patient survival is about 2 years.

OTHER LESIONS OF THE ADRENAL

Adrenal cysts are relatively uncommon lesions; however, with the use of sophisticated abdominal imaging techniques, the frequency of detection of these lesions appears to be increasing. The larger cysts may produce an abdominal mass and flank pain. Both cortical and medullary neoplasms may undergo necrosis and cystic degeneration and may present as "nonfunctional" cysts.

Adrenal myelolipomas are unusual benign lesions composed of mature fat and hematopoietic cells. Although most of these lesions represent incidental findings, occasional myelolipomas may reach massive proportions. Histologically, mature adipocytes are admixed with aggregates of hematopoietic cells belonging to all three lineages. Foci of myelolipomatous change may be seen in cortical tumors and in adrenals with cortical hyperplasia.

The term adrenal incidentaloma is a half-facetious moniker that has crept into the medical lexicon as advancements in medical imaging have led to the incidental discovery of adrenal masses in asymptomatic individuals or in individuals in whom the presenting complaint is not directly related to the adrenal gland.[127] Fortunately, the vast majority of adrenal incidentalomas are nonsecreting cortical adenomas, but in effect, any adrenal cortical or medullary neoplasm or hyperplasia, metastatic cancer, or a non-neoplastic disease (abscess, amyloidosis, sarcoid) can result in an incidentally discovered adrenal mass.

Adrenal Medulla

Normal

The adrenal medulla is developmentally, functionally, and structurally distinct from the adrenal cortex. It is composed of specialized neural crest (neuroendocrine) cells, termed chromaffin cells, and their supporting (sustentacular) cells. The chromaffin cells are round to oval, have prominent cytoplasmic membrane-bound granules of stored catecholamines, and are supported by a richly vascularized scant stroma of spindled and sustentacular cells. These cells, so named because of their brown-black color after exposure to potassium dichromate (e.g., Zenker fixative), synthesize and secrete catecholamines in response to signals from preganglionic nerve fibers in the sympathetic nervous system. The adrenal medulla is the major source of catecholamines (epinephrine, norepinephrine) in the body. Norepinephrine functions as a local neurotransmitter, chiefly of sympathetic postganglionic neurons. Only small amounts reach the circulation. Epinephrine (adrenaline) is secreted into the vascular system. It interacts with α-adrenergic and β-adrenergic receptors in various cells, which then activate second

messengers and a cascade of enzymatic reactions mediating the systemic actions of epinephrine, for example, increasing the force and rate of myocardial contractions and causing vasoconstriction of most vascular beds. Because the secretory cells are a part of the neuroendocrine system, they are also capable of synthesizing a variety of bioactive amines and peptides, such as histamine, serotonin, renin, chromogranin A, and neuropeptide hormones.

Neuroendocrine cells similar to chromaffin cells are widely dispersed in an extra-adrenal system of clusters and nodules that, together with the adrenal medulla, make up the paraganglion system. These extra-adrenal paraganglia are closely associated with the autonomic nervous system and can be divided into three groups based on their anatomic distribution: (1) branchiomeric, (2) intravagal, and (3) aorticosympathetic ( Fig. 24-54 ). The branchiomeric and intravagal paraganglia associated with the parasympathetic system are located close to the major arteries and cranial nerves of the head and neck and include the carotid bodies ( Chapter 16 ). The intravagal paraganglia, as the term implies, are distributed along the vagus nerve. The aorticosympathetic chain is found in association with segmental ganglia of the sympathetic system and therefore is distributed mainly alongside of the abdominal aorta. The organs of Zuckerkandl, close to the aortic bifurcation, belong to this group. The visceral paraganglia, as the term implies, are located within organs such as the urinary bladder. They are described in Chapter 16 .

1219

Figure 24-54 The paraganglion system. This schematic representation of the paraganglion system demonstrates sites of paraganglion cell nests, in which neoplasms may form. The extra-adrenal portion of the paraganglion system is grouped into three families based on anatomic distribution, innervation, and microscopic structure: (1) branchiomeric, (2) intravagal, and (3) aorticosympathetic. (From Whalen RK, et al: Extra-adrenal pheochromocytoma. J Urol 147:1–10, 1992; copyright Williams & Wilkins, 1992.)

Pathology

The most important diseases of the adrenal medulla are neoplasms, which include neoplasms of chromaffin cells (pheochromocytomas) and neuronal neoplasms (including neuroblastomas and more mature ganglion cell tumors).

PHEOCHROMOCYTOMA

Pheochromocytomas are uncommon neoplasms composed of chromaffin cells, which synthesize and release catecholamines and in some instances peptide hormones. These tumors are important because they (similar to aldosterone-secreting adenomas) give rise to surgically correctable forms of hypertension. Although only about 0.1% to 0.3% of hypertensive patients have an underlying pheochromocytoma, the hypertension can be fatal when the pheochromocytoma goes unrecognized. Occasionally, one of these tumors produces other steroids or peptides and so may be associated with Cushing syndrome or some other endocrinopathy.

Pheochromocytomas usually subscribe to a convenient "rule of 10s":

• 10% of pheochromocytomas arise in association with one of several familial syndromes ( Table 24-11 ). These include the MEN-2A and MEN-2B syndromes (described later), type I neurofibromatosis ( Chapter 5 ), von Hippel-Lindau syndrome ( Chapter 28 ), and Sturge-Weber syndrome ( Chapter 16 ). • 10% of pheochromocytomas are extra-adrenal, occurring in sites such as the organ of Zuckerkandl and the carotid body, where these chromaffin-negative tumors are usually called paragangliomas to distinguish them from pheochromocytomas. • 10% of nonfamilial adrenal pheochromocytomas are bilateral; this figure may rise to 70% in cases that are associated with familial syndromes. • 10% of adrenal pheochromocytomas are biologically malignant, although the associated hypertension represents a serious and potentially lethal complication of even "benign" tumors. Frank malignancy is somewhat more common (20% to 40%) in tumors arising in extra-adrenal sites. • 10% of adrenal pheochromocytomas arise in childhood, usually the familial subtypes, and with a strong male preponderance. The nonfamilial pheochromocytomas most

1220

often occur in adults between 40 and 60 years of age, with a slight female preponderance.

TABLE 24-11 -- Familial Syndromes Associated with Pheochromocytoma

Syndrome Components

MEN, type 2A Medullary thyroid carcinomas and C-cell hyperplasia

Pheochromocytomas and adrenal medullary hyperplasia

Parathyroid hyperplasia

MEN, type 2B Medullary thyroid carcinomas and C-cell hyperplasia

Pheochromocytomas and adrenal medullary hyperplasia

Mucosal neuromas

Marfanoid features

von Hippel-Lindau Renal, hepatic, pancreatic, and epididymal cysts

Renal cell carcinomas

Pheochromocytomas

Angiomatosis

Cerebellar hemangioblastomas

von Recklinghausen Neurofibromatosis

Café au lait skin spots

Schwannomas, meningiomas, gliomas

Pheochromocytomas

Sturge-Weber Cavernous hemangiomas of fifth cranial nerve distribution

Pheochromocytomas

MEN, multiple endocrine neoplasia.

Data from Silverman ML, Lee AK: Anatomy and pathology of the adrenal glands. Urol Clin North Am 16:417, 1989.

Morphology.

Pheochromocytomas range from small, circumscribed lesions confined to the adrenal ( Fig. 24-55 ) to large hemorrhagic masses weighing kilograms. The average weight of a pheochromocytoma is 100 gm, but variations from just over 1 gm to almost 4000 gm have been reported. The larger tumors are well demarcated by either connective tissue or compressed cortical or medullary tissue. Richly vascularized fibrous trabeculae pass into the tumor and produce a lobular pattern. In many tumors, remnants of the adrenal gland can be seen, stretched over the surface or attached at one pole. On section, the cut surfaces of smaller pheochromocytomas are yellow-tan. Larger lesions tend to be

hemorrhagic, necrotic, and cystic and typically efface the adrenal gland. Incubation of fresh tissue with a potassium dichromate solution turns the tumor a dark brown color owing to oxidation of stored catecholamines, thus the term chromaffin.

The histologic pattern in pheochromocytoma is quite variable. The tumors are composed of polygonal to spindle-shaped chromaffin cells or chief cells, clustered with the sustentacular cells into small nests or alveoli (zellballen) by a rich vascular network ( Fig. 24-56 ). Uncommonly, the dominant cell type is a spindle or small cell; various patterns can be found in any one tumor. The cytoplasm has a finely granular appearance, best demonstrated with silver stains, owing to the appearance of granules containing catecholamines. The nuclei are usually round to ovoid, with a stippled "salt and pepper" chromatin that is characteristic of most neuroendocrine tumors. Electron microscopy reveals variable numbers of membrane-bound, electron-dense granules, representing

Figure 24-55 Pheochromocytoma. The tumor is enclosed within an attenuated cortex and demonstrates areas of hemorrhage. The comma-shaped residual adrenal is seen below.

Figure 24-56 Pheochromocytoma demonstrating characteristic nests of cells ("zellballen") with abundant cytoplasm. Granules containing catecholamine are not visible in this preparation. It is not uncommon to find bizarre cells even in pheochromocytomas that are biologically benign, and this criterion by itself should not be used to diagnose malignancy.

catecholamines and sometimes other peptides ( Fig. 24-57 ). Immunoreactivity for neuroendocrine markers (chromogranin and synaptophysin) is present in the chief cells, while the peripheral sustentacular cells label with S-100, a calcium-binding protein expressed by a variety of mesenchymal cell types.

The criteria for determining malignancy in pheochromocytomas can be a vexing issue. There is no single histologic feature that can reliably predict clinical behavior in pheochromocytomas. Tumors with "benign" histologic features may metastasize, while bizarrely pleomorphic tumors may remain confined to the adrenal gland. In fact, cellular and nuclear pleomorphism, including the presence of giant cells, and mitotic figures are often seen in benign pheochromocytomas, while cellular monotony is paradoxically associated with an aggressive behavior (see below).

Figure 24-57 Electron micrograph of pheochromocytoma. This tumor contains membrane-bound secretory granules in which catecholamines are stored (30,000X).

1221

Even capsular and vascular invasion may be encountered in benign lesions. Therefore, the definitive diagnosis of malignancy in pheochromocytomas is based exclusively on the presence of metastases. These may involve regional lymph nodes as well as more distant sites, including liver, lung, and bone. Several histologic features, such as numbers of mitoses, confluent tumor necrosis, and spindle cell morphology, have been associated with an aggressive behavior and increased risk of metastasis, but in and of itself, no single criterion is entirely reliable.[128]

Clinical Course.

The dominant clinical feature in patients with pheochromocytoma is hypertension. Classically, this is described as an abrupt, precipitous elevation in blood pressure, associated with tachycardia, palpitations, headache, sweating, tremor, and a sense of apprehension. These episodes may also be associated with pain in the abdomen or chest, nausea, and vomiting. In practice, isolated paroxysmal episodes of hypertension occur in fewer than half of patients. In about two-thirds of patients, the hypertension occurs in the form of chronic, sustained elevation in blood pressure, although an element of labile hypertension is also present. The paroxysms may be precipitated by emotional stress, exercise, changes in posture, and palpation in the region of the tumor. The elevations of pressure are induced by the sudden release of catecholamines that may acutely precipitate congestive heart failure, pulmonary edema, myocardial infarction, ventricular fibrillation, and cerebrovascular accidents. The cardiac complications have been attributed to what has been called catecholamine cardiomyopathy, or catecholamine-induced myocardial instability and ventricular arrhythmias. Nonspecific myocardial changes, such as focal necrosis, mononuclear infiltrates, and interstitial fibrosis, have been attributed to ischemic damage secondary to the catecholamine-induced vasomotor constriction of the myocardial circulation or to direct catecholamine toxicity. In some cases, pheochromocytomas secrete other hormones, such as ACTH and somatostatin, and may therefore be associated with clinical features related to the secretion of these or other peptide hormones.

The laboratory diagnosis of pheochromocytoma is based on the demonstration of increased urinary excretion of free catecholamines and their metabolites, such as vanillylmandelic acid (VMA) and metanephrines. Isolated benign tumors are treated with surgical excision, after preoperative and intraoperative medication of patients with adrenergic-blocking agents to prevent a hypertensive crisis. Multifocal lesions require long-term medical treatment for hypertension.

TUMORS OF EXTRA-ADRENAL PARAGANGLIA

Pheochromocytomas that develop in paraganglia other than the adrenal medulla are often designated paragangliomas. Paragangliomas may arise in any organ that contains paraganglionic tissue. Tumors arising in the carotid body are designated carotid body tumors, whereas those originating in the jugulotympanic body are sometimes referred to as chemodectomas because these paraganglia sense the oxygen and carbon dioxide levels of the blood. The carotid body tumor is a typical paraganglioma, forming a palpable mass in the neck enveloping the carotid vessels. Paragangliomas are uncommon and occur about one tenth as frequently as adrenal pheochromocytomas. They are described in Chapter 16 .

NEUROBLASTOMA

Neuroblastoma is the most common extracranial solid tumor of childhood. These neoplasms occur most commonly during the first 5 years of life and may arise during infancy. Neuroblastomas may occur anywhere in the sympathetic nervous system and occasionally within the brain, but they are most common in the abdomen; most cases arise in either the adrenal medulla or the retroperitoneal sympathetic ganglia. Most neuroblastomas are sporadic, although familial cases also occur. These tumors were discussed in Chapter 10 , along with other pediatric neoplasms.

Multiple Endocrine Neoplasia Syndromes

The multiple endocrine neoplasia (MEN) syndromes are a group of genetically inherited diseases resulting in proliferative lesions (hyperplasia, adenomas, and carcinomas) of multiple endocrine organs. Like other inherited cancer disorders ( Chapter 7 ), endocrine tumors arising in the context of MEN syndromes have certain distinct features that contrast with their sporadic counterparts:

• These tumors occur at a younger age than sporadic cancers. • They arise in multiple endocrine organs, either synchronously (at the same time) or metachronously (at different times). • Even in one organ, the tumors are often multifocal. • The tumors are usually preceded by an asymptomatic stage of endocrine hyperplasia involving the cell of origin of the tumor. For example, patients with MEN-1 syndrome develop varying degrees of islet cell hyperplasia, some of which progress to pancreatic tumors. • These tumors are usually more aggressive and recur in a higher proportion of cases than do similar endocrine tumors that occur sporadically.

The salient features of the MEN syndromes are summarized in Table 24-12 and discussed below.

MULTIPLE ENDOCRINE NEOPLASIA, TYPE 1

MEN-1, or Wermer syndrome, is a rare heritable disorder with a prevalence of about 2 per 100,000. It is characterized by abnormalities involving the parathyroid, pancreas, and pituitary glands; thus the mnemonic device, the 3Ps:

• Parathyroid: Primary hyperparathyroidism is the most common manifestation of MEN-1 (80% to 95% of patients) and is the initial manifestation of the disorder in most patients, appearing in almost all patients by age 40 to 50. Parathyroid abnormalities include both hyperplasia and adenomas. Hyperplasias arising in the context of MEN-1 are monoclonal.

1222

• Pancreas: Endocrine tumors of the pancreas are a leading cause of morbidity and mortality in MEN-1 patients.[129] These tumors are usually aggressive and often present with metastatic disease. It is not uncommon to find multiple "microadenomas" scattered throughout the pancreas in conjunction with one or two dominant lesions. Pancreatic endocrine tumors are often functional; however, since pancreatic polypeptide is the most commonly secreted product, these tumors might not be accompanied by an endocrine hypersecretion syndrome. Among symptomatic pancreatic tumors, gastrinomas associated with Zollinger-Ellison syndrome and insulinomas associated with hypoglycemia and neurologic manifestations are the most common subtypes. • Pituitary: The most frequent anterior pituitary tumor encountered in MEN-1 is a prolactinoma; some patients develop acromegaly from somatotrophin-secreting tumors. • The spectrum of this disease has been expanded beyond the 3Ps. The duodenum is the most common site of gastrinomas in individuals with MEN-1 (far in excess of the frequency of pancreatic gastrinomas), [130] and synchronous duodenal and pancreatic tumors may be present in the same individual. In addition, carcinoid tumors, thyroid and adrenocortical adenomas, and lipomas are more frequent than in the general population.

TABLE 24-12 -- Multiple Endocrine Neoplasia (MEN) Syndromes

MEN-1 MEN-2A MEN-2B

Pituitary Adenomas

Parathyroid Hyperplasia +++

Hyperplasia +

Adenomas +

Pancreatic islets Hyperplasia ++

Adenomas ++

TABLE 24-12 -- Multiple Endocrine Neoplasia (MEN) Syndromes

MEN-1 MEN-2A MEN-2B

Carcinomas +++

Adrenal Cortical hyperplasia

Pheochromocytoma ++

Pheochromocytoma +++

Thyroid C-cell hyperplasia +++

C-cell hyperplasia +++

Medullary carcinoma +++

Medullary carcinoma +++

Extraendocrine changes

Mucocutaneous ganglioneuromas

Marfanoid habitus

Mutant gene locus

MEN1 RET RET

Relative frequency: +, uncommon; +++, common.

MEN-1 syndrome is caused by germ-line mutations in the MEN1 gene at 11q13. This gene encodes a 610-amino acid product known as menin, which localizes primarily to the nucleus. MEN1 is a classic tumor suppressor gene ( Chapter 7 ) in that both alleles are inactivated in the MEN-1-associated tumors.[131] The precise role of menin in tumor suppression remains elusive, although recent studies have shown that it may be important in regulating the cell cycle and transcription. [132]

The dominant clinical manifestations of MEN-1 are usually defined by the peptide hormones that are overproduced and include such abnormalities as recurrent hypoglycemia due to insulinomas, intractable peptic ulcers in patients with Zollinger-Ellison syndrome, nephrolithiasis caused by PTH-induced hypercalcemia, or symptoms of prolactin excess from a pituitary tumor. As expected, malignant behavior by one or more of the endocrine tumors arising in these patients is often the proximate cause of death.

MULTIPLE ENDOCRINE NEOPLASIA, TYPE 2

MEN-2 is subclassified into three distinct syndromes: MEN-2A, MEN-2B, and familial medullary thyroid cancer.

• MEN-2A, or Sipple syndrome, is characterized by pheochromocytoma, medullary carcinoma, and parathyroid hyperplasia. Medullary carcinomas of the

thyroid occur in almost 100% of patients. They are usually multifocal and are virtually always associated with foci of C-cell hyperplasia in the adjacent thyroid. The medullary carcinomas may elaborate calcitonin and other active products and are usually clinically aggressive. Forty per cent to 50% of patients with MEN-2A have pheochromocytomas, which are often bilateral and may arise in extra-adrenal sites. As in the case of pheochromocytomas in general, they may be benign or malignant. Ten per cent to 20% of patients have parathyroid hyperplasia and evidence of hypercalcemia or renal stones. MEN-2A is clinically and genetically distinct from MEN-1 and has been linked to germ-line mutations in the RET (rearranged during transfection) protooncogene on chromosome 10q11.2. As was noted earlier, the RET protooncogene is a receptor tyrosine kinase that binds glialderived neurotrophic factor (GDNF) and other ligands in the GDNF family and transmits growth and differentiation signals ( Chapter 7 ). Loss of function mutations in RET result in intestinal aganglionosis and Hirschsprung disease ( Chapter 17 ). In contrast, in MEN-2A (as well as in MEN-2B), germ-line mutations constitutively activate the RET receptor, resulting in gain of function.[133] This scenario is different from most other inherited predispositions to neoplasia, which are due to heritable loss of function mutations that inactivate tumor-suppressor proteins ( Chapter 7 ). • MEN-2B has significant clinical overlap with MEN-2A. Patients develop medullary thyroid carcinomas, which are usually multifocal and more aggressive than in MEN-2A, and pheochromocytomas. However, unlike in MEN-2A, primary hyperparathyroidism is not present. In addition, MEN-2B is accompanied by neuromas or ganglioneuromas involving the skin, oral mucosa, eyes, respiratory tract, and gastrointestinal tract, and a marfanoid habitus, with long axial skeletal features and hyperextensible joints. A single amino acid change in RET (RETMet918Thr ), distinct from the

1223

mutational spectra that are seen in MEN-2A, appears to be responsible for virtually all cases of MEN-2B and affects a critical region of the tyrosine kinase catalytic domain of the protein.[134] • Familial medullary thyroid cancer is a variant of MEN-2A, in which there is a strong predisposition to medullary thyroid cancer but not the other clinical manifestations of MEN-2A or MEN-2B. A substantial majority of cases of medullary thyroid cancer are sporadic, but as many as 20% may be familial. Familial medullary thyroid cancers develop at an older age than those occurring in the full-blown MEN-2 syndrome and follow a more indolent course.

In contrast to MEN-1, in which the long-term benefit of early diagnosis via genetic screening is not well established, diagnosis via screening of at-risk family members in MEN-2A kindred is important because medullary thyroid carcinoma is a life-threatening disease that can be prevented by early thyroidectomy. Prior to the advent of genetic testing, family members of patients with the MEN-2 syndrome were screened with annual biochemical tests, which often lacked sensitivity. Now, routine genetic testing

identifies RET mutation carriers earlier and more reliably in MEN-2 kindred; all individuals carrying germ-line RET mutations are advised to undergo prophylactic thyroidectomy to prevent the inevitable development of medullary carcinomas.

Pineal Gland

Normal

The rarity of clinically significant lesions (virtually only tumors) justifies brevity in the consideration of the pineal gland. It is a minute, pinecone-shaped organ (hence its name), weighing 100 to 180 mg and lying between the superior colliculi at the base of the brain. It is composed of a loose, neuroglial stroma enclosing nests of epithelial-appearing pineocytes, cells with photosensory and neuroendocrine functions (hence the designation of the pineal gland as the "third eye"). Silver impregnation stains reveal that these cells have long, slender processes reminiscent of primitive neuronal precursors intermixed with the processes of astrocytic cells.

Pathology

All tumors involving the pineal are rare; most (50% to 70%) arise from sequestered embryonic germ cells. They most commonly take the form of so-called germinomas, resembling testicular seminoma ( Chapter 21 ) or ovarian dysgerminoma ( Chapter 22 ). Other lines of germ cell differentiation include embryonal carcinomas; choriocarcinomas; mixtures of germinoma, embryonal carcinoma, and choriocarcinoma; and, uncommonly, typical teratomas (usually benign). Whether to characterize these germ cell neoplasms as pinealomas is still a subject of debate, but most "pinealophiles" favor restricting the term pinealoma to neoplasma arising from the pineocytes.

PINEALOMAS

These neoplasms are divided into two categories, pineoblastomas and pineocytomas, based on their level of differentiation, which, in turn, correlates with their neoplastic aggressiveness.[135]

Morphology.

Pineoblastomas are encountered mostly in the first two decades of life and appear as soft, friable, gray masses punctuated with areas of hemorrhage and necrosis. They typically invade surrounding structures, such as the hypothalamus, midbrain, and lumen of the third ventricle. Histologically, they are composed of masses of pleomorphic cells two to four times the diameter of an erythrocyte. Large hyperchromatic nuclei appear to

occupy almost the entire cell, and mitoses are frequent. The cytology is that of primitive embryonal tumor ("small blue cell neoplasm") similar to medulloblastoma ( Chapter 28 ) or retinoblastoma ( Chapter 29 ).

Pineoblastomas, like medulloblastomas, tend to spread via the cerebrospinal fluid. As might be expected, the enlarging mass may compress the aqueduct of Sylvius, giving rise to internal hydrocephalus and all its consequences. Survival beyond 1 or 2 years is rare.

In contrast, pineocytomas occur mostly in adults and are much slower-growing than pineoblastomas. They tend to be well-circumscribed, gray, or hemorrhagic masses that compress but do not infiltrate surrounding structures. Histologically, the tumors may be pure pineocytomas or exhibit divergent glial, neuronal, and retinal differentiation. The tumors are composed largely of pineocytes having darkly staining, round-to-oval, fairly regular nuclei. Necrosis is unusual, and mitoses are virtually absent. The neoplastic cells resemble normal pineocytes in their strong immunoreactivity for neuro-specific enolase and synaptophysin. Particularly distinctive are the pineocytomatous pseudorosettes rimmed by rows of pineocytes. The centers of these rosettes are filled with eosinophilic cytoplasmic material representing tumor cell processes. These cells are set against a background of thin, fibrovascular, anastomosing septa, which confer a lobular growth pattern to the tumor. Glial and retinal differentiation is detectable by immunoreactivity for glial fibrillary acidic protein and retinal S-antigen, respectively.

1224

The clinical course of patients with pineocytomas is prolonged, averaging 7 years. The manifestations are the consequence of their pressure effects and consist of visual disturbances, headache, mental deterioration, and sometimes dementia-like behavior. The lesions being located where they are, it is understandable that successful excision is at best difficult.

References

1. Elster AD: Modern imaging of the pituitary. Radiology 187(1):1–14, 1993.

2. Asa SL, Ezzat S: The pathogenesis of pituitary tumours. Nat Rev Cancer 2(11):836–849, 2002.

3. Suhardja A, Kovacs K, and Rutka J: Genetic basis of pituitary adenoma invasiveness: a review. J Neurooncol 52(3):195–204, 2001.

4. Yamada S, et al: Growth hormone-producing pituitary adenomas: correlations between clinical characteristics and morphology. Neurosurgery 33(1):20–27, 1993.

5. Asa SL, Ezzat S: The cytogenesis and pathogenesis of pituitary adenomas. Endocr Rev 19(6):798–827, 1998.

6. Sheehan HL: The recognition of chronic hypopituitarism resulting from postpartum pituitary necrosis. Am J Obstet Gynecol, 111(6):852–854, 1971.

7. Rodriguez R, Andersen B: Cellular determination in the anterior pituitary gland: PIT-1 and PROP-1 mutations as causes of human combined pituitary hormone deficiency. Minerva Endocrinol 28(2):123–133, 2003.

8. LiVolsi VA, Perzin KH, Savetsky L: Carcinoma arising in median ectopic thyroid (including thyroglossal duct tissue). Cancer 34(4):1303–1315, 1974.

9. Cheng SY: Multiple mechanisms for regulation of the transcriptional activity of thyroid hormone receptors. Rev Endocr Metab Disord 1(1–2):9–18, 2000.

10. Helfand M, Redfern CC: Clinical guideline. Part 2: screening for thyroid disease: an update. American College of Physicians. Ann Intern Med 129(2):144–158, 1998.

11. Refetoff S: Resistance to thyroid hormone with and without receptor gene mutations. Ann Endocrinol (Paris) 64(1):23–25, 2003.

12. Yen PM: Thyrotropin receptor mutations in thyroid diseases. Rev Endocr Metab Disord 1(1–2):123–129, 2000.

13. Clifton-Bligh RJ, et al: Mutation of the gene encoding human TTF-2 associated with thyroid agenesis, cleft palate and choanal atresia. Nat Genet 19(4):399–401, 1998.

14. Macchia PE, et al: PAX8 mutations associated with congenital hypothyroidism caused by thyroid dysgenesis. Nat Genet 19(1):83–86, 1998.

15. Barbesino G, Chiovato L: The genetics of Hashimoto's disease. Endocrinol Metab Clin North Am 29(2):357–374, 2000.

16. Tomer Y, et al: Common and unique susceptibility loci in Graves and Hashimoto diseases: results of whole-genome screening in a data set of 102 multiplex families. Am J Hum Genet 73(4):736–747, 2003.

17. Stassi G, De Maria R: Autoimmune thyroid disease: new models of cell death in autoimmunity. Nat Rev Immunol 2(3):195–204, 2002.

18. Pearce EN, Farwell AP, Braverman LE: Thyroiditis. N Engl J Med 348:2646, 2003.

19. Muller AF, Drexhage HA, Berghout A: Postpartum thyroiditis and autoimmune thyroiditis in women of childbearing age: recent insights and consequences for antenatal and postnatal care. Endocr Rev 22(5):605–630, 2001.

20. Kouki T, et al: Relation of three polymorphisms of the CTLA-4 gene in patients with Graves' disease. J Endocrinol Invest 25(3):208–213, 2002.

21. Ueda H, et al: Association of the T-cell regulatory gene CTLA4 with susceptibility to autoimmune disease. Nature 423(6939):506–511, 2003.

22. Weetman AP: Grave's disease 1835–2002. Horm Res 59 (suppl 1):114–118, 2003.

23. Heufelder AE: Pathogenesis of ophthalmopathy in autoimmune thyroid disease. Rev Endocr Metab Disord 1(1–2):87–95, 2000.

24. Apel RL, et al: Clonality of thyroid nodules in sporadic goiter. Diagn Mol Pathol 4(2):113–121, 1995.

25. Siegel RD, Lee SL: Toxic nodular goiter: toxic adenoma and toxic multinodular goiter. Endocrinol Metab Clin North Am 27(1)151–168, 1998.

26. Rodien P, et al: Activating mutations of TSH receptor. Ann Endocrinol (Paris) 64(1):12–16, 2003.

27. Lang W, et al: The differentiation of atypical adenomas and encapsulated follicular carcinomas in the thyroid gland. Virchows Arch 385(2):125–141, 1980.

28. Kroll TG, et al: PAX8-PPARgammal fusion oncogene in human thyroid carcinoma [corrected]. Science 289(5483):1357–1360, 2000.

29. Nikiforova MN, et al: RAS point mutations and PAX8-PPAR gamma rearrangement in thyroid tumors: evidence for distinct molecular pathways in thyroid follicular carcinoma. J Clin Endocrinol Metab 88(5):2318–2326, 2003.

30. Nikiforova MN, et al: PAX8-PPARgamma rearrangement in thyroid tumors: RT-PCR and immunohistochemical analyses. Am J Surg Pathol 26(8):1016–1023, 2002.

31. Nikiforov YE: RET/PTC rearrangement in thyroid tumors. Endocr Pathol 13(1):3–16, 2002.

32. Pierotti MA, Vigneri P, Bongarzone I: Rearrangements of RET and NTRK1 tyrosine kinase receptors in papillary thyroid carcinomas. Recent Results Cancer Res, 154:237-247, 1998.

33. Xu X, et al: High prevalence of BRAF gene mutation in papillary thyroid carcinomas and thyroid tumor cell lines. Cancer Res 63(15):4561–4567, 2003.

34. Cohen Y, et al: BRAF mutation in papillary thyroid carcinoma. J Natl Cancer Inst 95(8):625–627, 2003.

35. Kimura ET, et al: High prevalence of BRAF mutations in thyroid cancer: genetic evidence for constitutive activation of the RET/PTC-RAS-BRAF signaling pathway in papillary thyroid carcinoma. Cancer Res 63(7):1454–1457, 2003.

36. Eng C, et al: The relationship between specific RET proto-oncogene mutations and disease phenotype in multiple endocrine neoplasia type 2: International RET Mutation Consortium analysis. JAMA 276(19):1575–1579, 1996.

37. Marsh DJ, et al: Somatic mutations in the RET proto-oncogene in sporadic medullary thyroid carcinoma. Clin Endocrinol (Oxf) 44(3):249–257, 1996.

38. Ito T, et al: Unique association of p53 mutations with undifferentiated but not with differentiated carcinomas of the thyroid gland. Cancer Res 52(5):1369–1371, 1992.

39. Rybakov SJ, et al: Thyroid cancer in children Ukraine after the Chernobyl accident. World J Surg 24(11):1446–1449, 2000.

40. LiVolsi VA: Surgical Pathology of the Thyroid: Major Problems in Pathology. Philadelphia, WB Saunders, 1990.

41. Baloch ZW, Livolsi VA: Follicular-patterned lesions of the thyroid: the bane of the pathologist. Am J Clin Pathol 117(1):143–150, 2002.

42. Ruter A, Nishiyama R, Lennquist S: Tall-cell variant of papillary thyroid cancer: disregarded entity? World J Surg 21(1):15–20; discussion: 20–21, 1997.

43. Basolo F, et al: Potent mitogenicity of the RET/PTC3 oncogene correlates with its prevalence in tall-cell variant of papillary thyroid carcinoma. Am J Pathol 160(1):247-254, 2002.

44. Cheung CC, et al: Hyalinizing trabecular tumor of the thyroid: a variant of papillary carcinoma proved by molecular genetics. Am J Surg Pathol 24(12):1622-1626, 2000.

45. WellsˆSA Jr, Franz C: Medullary carcinoma of the thyroid gland. World J Surg 24(8):952–956, 2000.

46. Perry A, Molberg K, Albores-Saavedra J: Physiologic versus neoplastic C-cell hyperplasia of the thyroid: separation of distinct histologic and biologic entities. Cancer 77(4):750–756, 1996.

47. Krueger JE, Maitra A, Albores-Saavedra J: Inherited medullary micro-carcinoma of the thyroid: a study of 11 cases. Am J Surg Pathol 24(6):853–858, 2000.

48. Machens A, et al: Early malignant progression of hereditary medullary thyroid cancer. N Engl J Med 349:1517, 2003.

49. Yip L, et al: Multiple endocrine neoplasia type 2: evaluation of the genotype-phenotype relationship. Arch Surg 138:409, 2003.

50. Boyle WJ, Simonet WS, Lacey DL: Osteoclast differentiation and activation. Nature 423:337, 2003.

51. Fiaschi-Taesch NM, Stewart AF: Minireview: parathyroid hormone-related protein as an intracrine factor: trafficking mechanisms and functional consequences. Endocrinology 144(2):407–411, 2003.

52. Bilezikian JP, Silverberg SJ: Clinical spectrum of primary hyperparathyroidism. Rev Endocr Metab Disord 1(4):237–245, 2000.

1225

53. Fuleihan Gel H: Familial benign hypocalciuric hypercalcemia. J Bone Miner Res 17 (suppl 2):N51–N56, 2002.

54. Arnold A., et al: Molecular pathogenesis of primary hyperparathyroidism. J Bone Miner Res 17 (suppl 2):N30–N36, 2002.

55. Silver J, Kilav R, Naveh-Many T: Mechanisms of secondary hyperparathyroidism. Am J Physiol Renal Physiol 283(3):F367–F376, 2002.

56. Anderson MS, et al: Projection of an immunological self shadow within the thymus by the aire protein. Science 298(5597):1395–1401, 2002.

57. Li Y, et al: Autoantibodies to the extracellular domain of the calcium sensing receptor in patients with acquired hypoparathyroidism. J Clin Invest 97(4):910–914, 1996.

58. Weinstein LS, et al: Endocrine manifestations of stimulatory G protein alpha-subunit mutations and the role of genomic imprinting. Endocr Rev 22(5):675–705, 2001.

59. GuG, Brown JR, Melton DA: Direct lineage tracing reveals the ontogeny of pancreatic cell fates during mouse embryogenesis. Mech Dev 120(1):35–43, 2003.

60. Narayan KM, et al: Lifetime risk for diabetes mellitus in the United States. JAMA 290(14):1884–1890, 2003.

61. Zimmet P, Alberti KG, Shaw J: Global and societal implications of the diabetes epidemic. Nature 414(6865):782–787, 2001.

62. Report of the Expert Committee on the Diagnosis and Classification of Diabetes Mellitus. Diabetes Care 25 (suppl 1):S5–S20, 2002.

63. Thorens B: GLUT2 in pancreatic and extra-pancreatic gluco-detection (review). Mol Membr Biol 18(4):265–273, 2001.

64. Reis AF, Velho G: Sulfonylurea receptor-1 (SUR1): genetic and metabolic evidences for a role in the susceptibility to type 2 diabetes mellitus. Diabetes Metab 28(1):14–19, 2002.

65. Saltiel AR, Kahn CR: Insulin signalling and the regulation of glucose and lipid metabolism. Nature 414(6865):799–806, 2001.

66. Avruch J, et al: Ras activation of the Raf kinase: tyrosine kinase recruitment of the MAP kinase cascade. Recent Prog Horm Res 56:127–155, 2001.

67. Shepherd PR, Kahn BB: Glucose transporters and insulin action: implications for insulin resistance and diabetes mellitus. N Engl J Med 341(4):248–257, 1999.

68. Kozma SC, Thomas G: Regulation of cell size in growth, development and human disease: PI3K, PKB and S6K. Bioessays 24(1):65–71, 2002.

69. Mathis D, Vence L, Benoist C: Beta-cell death during progression to diabetes. Nature 414(6865):792–798, 2001.

70. Bach JF, Chatenoud L: Tolerance to islet autoantigens in type 1 diabetes. Annu Rev Immunol 19:131–

161, 2001.

71. Pietropaolo M, Eisenbarth GS: Autoantibodies in human diabetes. Curr Dir Autoimmun 4:252–282, 2001.

72. Todd JA, Wicker LS: Genetic protection from the inflammatory disease type 1 diabetes in humans and animal models. Immunity 15(3):387–395, 2001.

73. McDevitt H: The role of MHC class II molecules in the pathogenesis and prevention of Type I diabetes. Adv Exp Med Biol 490:59–66, 2001.

74. Pugliese A, et al: HLA-DQB1 0602 is associated with dominant protection from diabetes even among islet cell antibody-positive first-degree relatives of patients with IDDM. Diabetes 44(6):608–613, 1995.

75. Jaeckel E, Manns M, Von Herrath M: Viruses and diabetes. Ann N Y Acad Sci 958:7–25, 2002.

76. Horwitz MS, Sarvetnick N: Viruses, host responses, and autoimmunity. Immunol Rev 169:241, 1999.

77. Benoist C, Mathis D: Autoimmunity provoked by infection: how good is the case for T cell epitope mimicry? Nat Immunol 2(9):797–801, 2001.

78. Saltiel AR: Series introduction: the molecular and physiological basis of insulin resistance: emerging implications for metabolic and cardiovascular diseases. J Clin Invest 106(2):163–164, 2000.

79. Shulman GI: Cellular mechanisms of insulin resistance. J Clin Invest 106(2):171–176, 2000.

80. Kadowaki T: Insights into insulin resistance and type 2 diabetes from knockout mouse models. J Clin Invest 106(4):459–465, 2000.

81. Elbein SC: Perspective: the search for genes for type 2 diabetes in the post-genome era. Endocrinology 143(6):2012–2018, 2002.

82. Kahn BB, Flier, JS: Obesity and insulin resistance. J Clin Invest 106(4):473–481, 2000.

83. Saltiel AR: You are what you secrete. Nat Med 7(8):887–888, 2001.

84. Flier, JS: Diabetes. The missing link with obesity? Nature 409(6818):292–293, 2001.

85. Yamauchi T, et al: The fat-derived hormone adiponectin reverses insulin resistance associated with both lipoatrophy and obesity. Nat Med 7(8):941–946, 2001.

86. Steppan CM, et al: The hormone resistin links obesity to diabetes. Nature 409(6818):307–312, 2001.

87. Shimomura I, et al: Leptin reverses insulin resistance and diabetes mellitus in mice with congenital lipodystrophy. Nature 401(6748):73–76, 1999.

88. Celi FS, Shuldiner AR: The role of peroxisome proliferator-activated receptor gamma in diabetes and obesity. Curr Diab Rep 2(2):179–185, 2002.

89. Fajans SS, Bell GI, Polonsky KS: Molecular mechanisms and clinical pathophysiology of maturity-onset diabetes of the young. N Engl J Med 345(13):971–980, 2001.

90. Ellard S, et al: A high prevalence of glucokinase mutations in gestational diabetic subjects selected by clinical criteria. Diabetologia 43(2):250–253, 2000.

91. Maechler P, Wollheim CB: Mitochondrial function in normal and diabetic beta-cells. Nature 414(6865):807–812, 2001.

92. Saltiel AR: New perspectives into the molecular pathogenesis and treatment of type 2 diabetes. Cell 104(4):517–529, 2001.

93. The Diabetes Control and Complications Trial Research Group: The effect of intensive treatment of diabetes on the development and progression of long-term complications in insulin-dependent diabetes mellitus. N Engl J Med 329(14):977–986, 1993.

94. UK Prospective Diabetes Study (UKPDS) Group: Intensive blood-glucose control with sulphonylureas or insulin compared with conventional treatment and risk of complications in patients with type 2 diabetes (UKPDS 33). Lancet 352(9131):837–853, 1998.

95. Sheetz MJ, King GL: Molecular understanding of hyperglycemia's adverse effects for diabetic complications. JAMA 288(20):2579–2588, 2002.

96. Stitt AW, Jenkins AJ, Cooper ME: Advanced glycation end products and diabetic complications. Expert Opin Investig Drugs 11(9):1205–1223, 2002.

97. Brownlee M: Biochemistry and molecular cell biology of diabetic complications. Nature 414(6865):813–820, 2001.

98. Frank RN: Potential new medical therapies for diabetic retinopathy: protein kinase C inhibitors. Am J Ophthalmol 133(5):693–698, 2002.

99. Lee AY, Chung SS: Contributions of polyol pathway to oxidative stress in diabetic cataract. FASEB J 13(1):23–30, 1999.

100. Haffner SM, et al: Mortality from coronary heart disease in subjects with type 2 diabetes and in nondiabetic subjects with and without prior myocardial infarction. N Engl J Med 339(4):229–234, 1998.

101. Wendt T, et al: Receptor for advanced glycation endproducts (RAGE) and vascular inflammation: insights into the pathogenesis of macrovascular complications in diabetes. Curr Atheroscler Rep 4(3):228–237, 2002.

102. Eckel RH, et al: Prevention Conference VI: Diabetes and Cardiovascular Disease: Writing Group II: pathogenesis of atherosclerosis in diabetes. Circulation 105(18):E138–E143, 2002.

103. Diabetic nephropathy. Diabetes Care 25 (suppl 1):S85–S89, 2002.

104. Frank RN: Diabetic retinopathy. N Engl J Med 350:48, 2004.

105. Astrup A, Finer N: Redefining type 2 diabetes: "diabesity" or "obesity dependent diabetes mellitus"? Obesity Rev 1:57–59, 2000.

106. Rindi G, Capella C, Solcia E: Cell biology, clinicopathological profile, and classification of gastro-enteropancreatic endocrine tumors. J Mol Med 76(6):413–420, 1998.

107. Solcia E, Capella C, Kloppel G: Tumors of the endocrine pancreas. In Rosai J (ed): Atlas of Tumor Pathology: Tumors of the Pancreas. Washington, DC, AFIP, 1997.

108. Goossens A, Heitz P, Kloppel G: Pancreatic endocrine cells and their non-neoplastic proliferations. In Dayal Y (ed): Endocrine Pathology of the Gut and Pancreas. Boca Raton, FL, CRC Press, 1991, pp 69–104.

109. Komminoth P, Heitz PU, Kloppel G: Pathology of MEN-1: morphology, clinicopathologic correlations and tumour development. J Intern Med 243(6):455–464, 1998.

110. Zollinger RM, Ellison EH: Primary peptic ulcerations of the jejunum associated with islet cell tumors of the pancreas: 1955. CA Cancer J Clin 39(4):231–247, 1989.

111. Newell-Price J, et al: The diagnosis and differential diagnosis of Cushing's syndrome and pseudo-Cushing's states. Endocr Rev 19(5):647–672, 1998.

112. Cushing HW: The basophil adenomas of the pituitary body and their clinical manifestations (pituitary basophilism). Bull Johns Hopkins Hosp 50:137–195, 1932.

1226

113. Stratakis CA, Kirschner LS: Clinical and genetic analysis of primary bilateral adrenal diseases (micro- and macronodular disease) leading to Cushing syndrome. Horm Metab Res 30(6–7):456–463, 1998.

114. Groussin L, et al: Mutations of the PRKAR1A gene in Cushing's syndrome due to sporadic primary pigmented nodular adrenocortical disease. J Clin Endocrinol Metab 87(9):4324–4329, 2002.

115. Fardella CE, Mosso L: Primary aldosteronism. Clin Lab 48(3–4):181–190, 2002.

116. Takeda Y: Genetic alterations in patients with primary aldosteronism. Hypertens Res 24(5):469–474, 2001.

117. Dluhy RG, Lifton RP: Glucocorticoid-remediable aldosteronism. J Clin Endocrinol Metab 84(12):4341–4344, 1999.

118. Speiser PW, White PC: Congenital adrenal hyperplasia. N Engl J Med 349:776, 2003.

119. Merke DP, et al: NIH conference. Future directions in the study and management of congenital adrenal hyperplasia due to 21-hydroxylase deficiency. Ann Intern Med 136(4):320–334, 2002.

120. Merke DP, et al: Adrenomedullary dysplasia and hypofunction in patients with classic 21-hydroxylase deficiency. N Engl J Med 343(19):1362–1368, 2000.

121. Peterson P, Uibo R, Krohn KJ: Adrenal autoimmunity: results and developments. Trends Endocrinol Metab 11(7):285–290, 2000.

122. Pitkanen J, Peterson P: Autoimmune regulator: from loss of function to autoimmunity. Genes Immun 4(1):12–21, 2003.

123. Vaidya B, Pearce S, Kendall-Taylor P: Recent advances in the molecular genetics of congenital and acquired primary adrenocortical failure. Clin Endocrinol (Oxf) 53(4):403–418, 2000.

124. Meeks JJ, Weiss J, Jameson JL: Dax1 is required for testis determination. Nat Genet 34(1):32–33, 2003.

125. Burke BA, et al: Congenital adrenal hypoplasia and selective absence of pituitary luteinizing hormone: a new autosomal recessive syndrome. Am J Med Genet 31(1):75–97, 1988.

126. Wenig BM, Heffess CS, Adair CF: Neoplasms of the adrenal gland. In Wenig BM, et al (ed): Atlas of Endocrine Pathology. Philadelphia, WB Saunders, 1997, pp 288–329.

127. Brunt LM, Moley JF: Adrenal incidentaloma. World J Surg 25(7):905–913, 2001.

128. Thompson LD: Pheochromocytoma of the adrenal gland scaled score (PASS) to separate benign from malignant neoplasms: a clinicopathologic and immunophenotypic study of 100 cases. Am J Surg Pathol 26(5):551–566, 2002.

129. Brandi ML: Multiple endocrine neoplasia type 1. Rev Endocr Metab Disord 1(4):275–282, 2000.

130. Mignon M, Cadiot G: Diagnostic and therapeutic criteria in patients with Zollinger-Ellison syndrome and multiple endocrine neoplasia type 1. J Intern Med 243(6):489–494, 1998.

131. Chandrasekharappa SC, et al: Positional cloning of the gene for multiple endocrine neoplasia-type 1. Science 276(5311):404–407, 1997.

132. Poisson A, Zablewska B, Gaudray P: Menin interacting proteins as clues toward the understanding of multiple endocrine neoplasia type 1. Cancer Lett 189(1):1–10, 2003.

133. Santoro M, et al: Different mutations of the RET gene cause different human tumoral diseases. Biochimie 81(4):397–402, 1999.

134. Salvatore D, et al: Increased in vivo phosphorylation of ret tyrosine 1062 is a potential pathogenetic mechanism of multiple endocrine neoplasia type 2B. Cancer Res 61(4):1426–1431, 2001.

135. Mena H, et al: Pineal parenchymal tumors. In Cavanee WK (ed): World Health Organization Classification of Tumors: Pathology and Genetics of Tumors of the Nervous System. Lyon, France, IARC Press, 2000, pp 115–121.