Chapter 38: Thyroid, Parathyroid, and Adrenal

1. There has been a paradigm shift in the surgical management of Graves’ disease with increased use of total or near-total thyroidectomy, rather than subtotal thyroidectomy.

2. Familial nonmedullary thyroid cancer is increasingly being recognized as a separate entity. Surgeons must be aware of the potential for false-negative fine-needle aspiration biopsy in this setting.

3. Fine-needle aspiration biopsies are now classified into six groups based on the risk of malignancy associated with each group (Bethesda criteria).

4. Total thyroidectomy is the surgical treatment of choice for most thyroid cancers, provided complication rates are low.

5. Due to the limitations of fine-needle aspiration biopsy in the setting of indeterminate thyroid nodules, a number of molecular markers are being evaluated as adjuncts to refine the diagnosis and management of these patients.

6. Focused mini-incision parathyroidectomy, after appropriate localization, has become the procedure of choice for the treatment of sporadic primary hyperparathyroidism.

7. Parathyroidectomy has been shown to improve the classic and the so-called nonspecific symptoms and metabolic complications of primary hyperparathyroidism.

8. Normocalcemic hyperparathyroidism is being increasingly recognized; however, there are no definitive guidelines for management.

9. Very high calcium and parathyroid hormone levels in a patient with primary hyperparathyroidism should alert the surgeon to the presence of a possible parathyroid carcinoma.

10. Subclinical Cushing’s syndrome is characterized by subtle abnormalities in corticosteroid synthesis, and many of its manifestations appear to be treated by adrenalectomy.

11. Fine-needle aspiration biopsy has a very limited role in the evaluation of adrenal incidentalomas unless the patient has previously had a cancer and should only be performed after appropriate biochemical studies have been performed to rule out pheochromocytoma.

12. Laparoscopic adrenalectomy has become the procedure of choice for excision of most adrenal lesions, except known or suspected cancers.

Historical Background

Goiters (from the Latin guttur, throat), defined as an enlargement of the thyroid, have been recognized since 2700 b.c. even though the thyroid gland was not documented as such until the Renaissance period. In 1619, Hieronymus Fabricius ab Aquapendente recognized that goiters arose from the thyroid gland. The term thyroid gland (Greek thyreoeides, shield-shaped) is, however, attributed to Thomas Wharton in his Adenographia (1656). In 1776, the thyroid was classified as a ductless gland by Albrecht von Haller and was thought to have numerous functions ranging from lubrication of the larynx to acting as a reservoir for blood to provide continuous flow to the brain, and to beautifying women’s necks. Burnt seaweed was considered to be the most effective treatment for goiters.

The first accounts of thyroid surgery for the treatment of goiters were given by Roger Frugardi in 1170. In response to failure of medical treatment, two setons were inserted at right angles into the goiter and tightened twice daily until the goiter separated. The open wound was treated with caustic powder and left to heal. However, thyroid surgery continued to be hazardous with prohibitive mortality rates (>40%) until the latter half of the nineteenth century, when advances in general anesthesia, antisepsis, and hemostasis enabled surgeons to perform thyroid surgery with significantly reduced mortality and morbidity rates. The most notable thyroid surgeons were Emil Theodor Kocher (1841–1917) and C.A. Theodor Billroth (1829–1894), who performed thousands of operations with increasingly successful results. However, as more patients survived thyroid operations, new problems and issues became apparent. After total thyroidectomy, patients (particularly children) became myxedematous with cretinous features. Myxedema was first effectively treated in 1891 by George Murray using a subcutaneous injection of an extract of sheep’s thyroid, and later, Edward Fox demonstrated that oral therapy was equally effective. In 1909, Kocher was awarded the Nobel Prize for medicine in recognition “for his works on the physiology, pathology, and surgery of the thyroid gland.”

Embryology

The thyroid gland arises as an outpouching of the primitive foregut around the third week of gestation. It originates at the base of the tongue at the foramen cecum. Endoderm cells in the floor of the pharyngeal anlage thicken to form the medial thyroid anlage (Fig. 38-1) that descends in the neck anterior to structures that form the hyoid bone and larynx. During its descent, the anlage remains connected to the foramen cecum via an epithelial-lined tube known as the thyroglossal duct. The epithelial cells making up the anlage give rise to the thyroid follicular cells. The paired lateral anlages originate from the fourth branchial pouch and fuse with the median anlage at approximately the fifth week of gestation. The lateral anlages are neuroectodermal in origin (ultimobranchial bodies) and provide the calcitonin producing parafollicular or C cells, which thus come to lie in the superoposterior region of the gland. Thyroid follicles are initially apparent by 8 weeks, and colloid formation begins by the eleventh week of gestation.

Developmental Abnormalities

Thyroglossal Duct Cyst and Sinus

Thyroglossal duct cysts are the most commonly encountered congenital cervical anomalies. During the fifth week of gestation, the thyroglossal duct lumen starts to obliterate, and the duct disappears by the eighth week of gestation. Rarely, the thyroglossal duct may persist in whole or in part. Thyroglossal duct cysts may occur anywhere along the migratory path of the thyroid, although 80% are found in juxtaposition to the hyoid bone. They are usually asymptomatic but occasionally become infected by oral bacteria, prompting the patient to seek medical advice. Thyroglossal duct sinuses result from infection of the cyst secondary to spontaneous or surgical drainage of the cyst and are accompanied by minor inflammation of the surrounding skin. Histologically, thyroglossal duct cysts are lined by pseudostratified ciliated columnar epithelium and squamous epithelium, with heterotopic thyroid tissue present in 20% of cases.

The diagnosis usually is established by observing a 1- to 2-cm, smooth, well-defined midline neck mass that moves upward with protrusion of the tongue. Routine thyroid imaging is not necessary, although thyroid scintigraphy and ultrasound have been performed to document the presence of normal thyroid tissue in the neck. Treatment involves the “Sistrunk operation,” which consists of en bloc cystectomy and excision of the central hyoid bone to minimize recurrence. Approximately 1% of thyroglossal duct cysts are found to contain cancer, which is usually papillary (85%). The role of total thyroidectomy in this setting is debated, but is advised in patients with large tumors, particularly if there are additional thyroid nodules and evidence of cyst wall invasion or lymph node metastases.¹ Squamous, Hürthle cell, and anaplastic cancers also have been reported but are rare. Medullary thyroid cancers (MTCs) are, however, not found in thyroglossal duct cysts.

Lingual Thyroid

A lingual thyroid represents a failure of the median thyroid anlage to descend normally and may be the only thyroid tissue present. Intervention becomes necessary for obstructive symptoms such as choking, dysphagia, airway obstruction, or hemorrhage. Many of these patients develop hypothyroidism. Medical treatment options include administration of exogenous thyroid hormone to suppress thyroid-stimulating hormone (TSH) and radioactive iodine (RAI) ablation followed by hormone replacement. Surgical excision is rarely needed but, if required, should be preceded by an evaluation of normal thyroid tissue in the neck to avoid inadvertently rendering the patient hypothyroid.

Ectopic Thyroid

Normal thyroid tissue may be found anywhere in the central neck compartment, including the esophagus, trachea, and anterior mediastinum. Thyroid tissue has been observed adjacent to the aortic arch, in the aortopulmonary window, within the upper pericardium, or in the interventricular septum. Often, “tongues” of thyroid tissue are seen to extend off the inferior poles of the gland and are particularly apparent in large goiters. Thyroid tissue situated lateral to the carotid sheath and jugular vein, previously termed lateral aberrant thyroid, almost always represents metastatic thyroid cancer in lymph nodes, and not remnants of the lateral anlage that had failed to fuse with the main thyroid, as previously suggested by Crile. Even if not readily apparent on physical examination or ultrasound imaging, the ipsilateral thyroid lobe contains a focus of papillary thyroid cancer (PTC), which may be microscopic.

Pyramidal Lobe

Normally the thyroglossal duct atrophies, although it may remain as a fibrous band. In about 50% of individuals, the distal end that connects to the thyroid persists as a pyramidal lobe projecting up from the isthmus, lying just to the left or right of the midline. In the normal individual, the pyramidal lobe is not palpable, but in disorders resulting in thyroid hypertrophy (e.g., Graves’ disease, diffuse nodular goiter, or lymphocytic thyroiditis), the pyramidal lobe usually is enlarged and palpable.

Thyroid Anatomy

The anatomic relations of the thyroid gland and surrounding structures are depicted in Fig. 38-2. The adult thyroid gland is brown in color and firm in consistency and is located posterior to the strap muscles. The normal thyroid gland weighs approximately 20 g, but gland weight varies with body weight and iodine intake. The thyroid lobes are located adjacent to the thyroid cartilage and connected in the midline by an isthmus that is located just inferior to the cricoid cartilage. A pyramidal lobe is present in about 50% of patients. The thyroid lobes extend to the midthyroid cartilage superiorly and lie adjacent to the carotid sheaths and sternocleidomastoid muscles laterally. The strap muscles (sternohyoid, sternothyroid, and superior belly of the omohyoid) are located anteriorly and are innervated by the ansa cervicalis (ansa hypoglossi). The thyroid gland is enveloped by a loosely connecting fascia that is formed from the partition of the deep cervical fascia into anterior and posterior divisions. The true capsule of the thyroid is a thin, densely adherent fibrous layer that sends out septa that invaginate into the gland, forming pseudolobules. The thyroid capsule is condensed into the posterior suspensory or Berry’s ligament near the cricoid cartilage and upper tracheal rings.

Blood Supply

The superior thyroid arteries arise from the ipsilateral external carotid arteries and divide into anterior and posterior branches at the apices of the thyroid lobes. The inferior thyroid arteries arise from the thyrocervical trunk shortly after their origin from the subclavian arteries. The inferior thyroid arteries travel upward in the neck posterior to the carotid sheath to enter the thyroid lobes at their midpoint. A thyroidea ima artery arises directly from the aorta or innominate in 1% to 4% of individuals to enter the isthmus or replace a missing inferior thyroid artery. The inferior thyroid artery crosses the recurrent laryngeal nerve (RLN), necessitating identification of the RLN before the arterial branches can be ligated. The venous drainage of the thyroid gland occurs via multiple small surface veins, which coalesce to form three sets of veins—the superior, middle, and inferior thyroid veins. The superior thyroid veins run with the superior thyroid arteries bilaterally. The middle vein or veins are the least consistent. The superior and middle veins drain directly into the internal jugular veins. The inferior veins often form a plexus, which drains into the brachiocephalic veins.

Nerves

The left RLN arises from the vagus nerve where it crosses the aortic arch, loops around the ligamentum arteriosum, and ascends medially in the neck within the tracheoesophageal groove. The right RLN arises from the vagus at its crossing with the right subclavian artery. The nerve usually passes posterior to the artery before ascending in the neck, its course being more oblique than the left RLN. Along their course in the neck, the RLNs may branch, and pass anterior, posterior, or interdigitate with branches of the inferior thyroid artery (Fig. 38-3). The right RLN may be nonrecurrent in 0.5% to 1% of individuals and often is associated with a vascular anomaly. Nonrecurrent left RLNs are rare but have been reported in patients with situs inversus and a right-sided aortic arch. The RLN may branch in its course in the neck, and identification of a small nerve should alert the surgeon to this possibility. Identification of the nerves or their branches often necessitates mobilization of the most lateral and posterior extent of the thyroid gland, the tubercle of Zuckerkandl, at the level of the cricoid cartilage. The last segments of the nerves often course below the tubercle and are closely approximated to the ligament of Berry. Branches of the nerve may traverse the ligament in 25% of individuals and are particularly vulnerable to injury at this junction. The RLNs terminate by entering the larynx posterior to the cricothyroid muscle.

The RLNs innervate all the intrinsic muscles of the larynx, except the cricothyroid muscles, which are innervated by the external laryngeal nerves. Injury to one RLN leads to paralysis of the ipsilateral vocal cord, which comes to lie in the paramedian or the abducted position. The paramedian position results in a normal but weak voice, whereas the abducted position leads to a hoarse voice and an ineffective cough. Bilateral RLN injury may lead to airway obstruction, necessitating emergency tracheostomy, or loss of voice. If both cords come to lie in an abducted position, air movement can occur, but the patient has an ineffective cough and is at increased risk of repeated respiratory tract infections from aspiration.

The superior laryngeal nerves also arise from the vagus nerves. After their origin at the base of the skull, these nerves travel along the internal carotid artery and divide into two branches at the level of the hyoid bone. The internal branch of the superior laryngeal nerve is sensory to the supraglottic larynx. Injury to this nerve is rare in thyroid surgery, but its occurrence may result in aspiration. The external branch of the superior laryngeal nerve lies on the inferior pharyngeal constrictor muscle and descends alongside the superior thyroid vessels before innervating the cricothyroid muscle. Cernea and colleagues² proposed a classification system to describe the relationship of this nerve to the superior thyroid vessels (Fig. 38-4). The type 2a variant, in which the nerve crosses below the tip of the thyroid superior pole, occurs in up to 20% of individuals and places the nerve at a greater risk of injury. Therefore, the superior pole vessels should not be ligated en masse, but should be individually divided, low on the thyroid gland and dissected lateral to the cricothyroid muscle. Injury to this nerve leads to inability to tense the ipsilateral vocal cord and hence difficulty “hitting high notes,” difficulty projecting the voice, and voice fatigue during prolonged speech.

Sympathetic innervation of the thyroid gland is provided by fibers from the superior and middle cervical sympathetic ganglia. The fibers enter the gland with the blood vessels and are vasomotor in action. Parasympathetic fibers are derived from the vagus nerve and reach the gland via branches of the laryngeal nerves.

Parathyroid Glands

The embryology and anatomy of the parathyroid glands are discussed in detail in the Parathyroid Gland section of this chapter. About 85% of individuals have four parathyroid glands that can be found within 1 cm of the junction of the inferior thyroid artery and the RLN. The superior glands are usually located dorsal to the RLN, whereas the inferior glands are usually found ventral to the RLN (Fig. 38-5).

Lymphatic System

The thyroid gland is endowed with an extensive network of lymphatics. Intraglandular lymphatic vessels connect both thyroid lobes through the isthmus and also drain to perithyroidal structures and lymph nodes. Regional lymph nodes include pretracheal, paratracheal, perithyroidal, RLN, superior mediastinal, retropharyngeal, esophageal, and upper, middle, and lower jugular chain nodes. These lymph nodes can be classified into seven levels as depicted in Fig. 38-6. The central compartment includes nodes located in the area between the two carotid sheaths, whereas nodes lateral to the vessels are present in the lateral compartment. Thyroid cancers may metastasize to any of these regions, although metastases to submaxillary nodes (level I) are rare (<1%). There also can be “skip” metastases to nodes in the lateral ipsilateral neck without central neck nodes.

Thyroid Histology

Microscopically, the thyroid is divided into lobules that contain 20 to 40 follicles (Fig. 38-7). There are about 3 × 10⁶ follicles in the adult male thyroid gland. The follicles are spherical and average 30 μm in diameter. Each follicle is lined by cuboidal epithelial cells and contains a central store of colloid secreted from the epithelial cells under the influence of the pituitary hormone TSH. The second group of thyroid secretory cells is the C cells or parafollicular cells, which contain and secrete the hormone calcitonin. They are found as individual cells or clumped in small groups in the interfollicular stroma and located in the upper poles of the thyroid lobes.

Thyroid Physiology

Iodine Metabolism

The average daily iodine requirement is 0.1 mg, which can be derived from foods such as fish, milk, and eggs or as additives in bread or salt. In the stomach and jejunum, iodine is rapidly converted to iodide and absorbed into the bloodstream, and from there it is distributed uniformly throughout the extracellular space. Iodide is actively transported into the thyroid follicular cells by an adenosine triphosphate (ATP)–dependent process. The thyroid is the storage site of >90% of the body’s iodine content and accounts for one third of the plasma iodine loss. The remaining plasma iodine is cleared via renal excretion.

Thyroid Hormone Synthesis, Secretion, and Transport

The synthesis of thyroid hormone consists of several steps (Fig. 38-8). The first, iodide trapping, involves active (ATP-dependent) transport of iodide across the basement membrane of the thyrocyte via an intrinsic membrane protein, the sodium/iodine (Na⁺/I^–) symporter. Thyroglobulin (Tg) is a large (660 kDa) glycoprotein, which is present in thyroid follicles and has four tyrosyl residues. The second step in thyroid hormone synthesis involves oxidation of iodide to iodine and iodination of tyrosine residues on Tg, to form monoiodotyrosines (MIT) and diiodotyrosines (DIT). Both processes are catalyzed by thyroid peroxidase (TPO). A recently identified protein, pendrin, is thought to mediate iodine efflux at the apical membrane. The third step leads to coupling of two DIT molecules to form tetra-iodothyronine or thyroxine (T₄), and one DIT molecule with one MIT molecule to form 3,5,3′-triiodothyronine (T₃) or 3,3′,5′-triiodothyronine reverse (rT₃). When stimulated by TSH, thyrocytes form pseudopodia, which encircle portions of cell membrane containing Tg, which in turn, fuse with enzyme-containing lysosomes. In the fourth step, Tg is hydrolyzed to release free iodothyronines (T₃ and T₄) and mono- and diiodotyrosines. The latter are deiodinated in the fifth step to yield iodide, which is reused in the thyrocyte. In the euthyroid state, T₄ is produced and released entirely by the thyroid gland, whereas only 20% of the total T₃ is produced by the thyroid. Most of the T₃ is produced by peripheral deiodination (removal of 5′-iodine from the outer ring) of T₄ in the liver, muscles, kidney, and anterior pituitary, a reaction that is catalyzed by 5′-mono-deiodinase. Some T₄ is converted to rT₃, the metabolically inactive compound, by deiodination of the inner ring of T₄. In conditions such as Graves’ disease, toxic multinodular goiter, or a stimulated thyroid gland, the proportion of T₃ released from the thyroid may be dramatically elevated. Thyroid hormones are transported in serum bound to carrier proteins such as T₄-binding globulin, T₄-binding prealbumin, and albumin. Only a small fraction (0.02%) of thyroid hormone (T₃ and T₄) is free (unbound) and is the physiologically active component. T₃ is the more potent of the two thyroid hormones, although its circulating plasma level is much lower than that of T₄. T₃ is less tightly bound to protein in the plasma than T₄, and so it enters tissues more readily. T₃ is three to four times more active than T₄ per unit weight, with a half-life of about 1 day, compared to approximately 7 days for T₄.

The secretion of thyroid hormone is controlled by the hypothalamic-pituitary-thyroid axis (Fig. 38-9). The hypothalamus produces a peptide, the thyrotropin-releasing hormone (TRH), which stimulates the pituitary to release TSH or thyrotropin. TRH reaches the pituitary via the portovenous circulation. TSH, a 28-kDa glycopeptide, mediates iodide trapping, secretion, and release of thyroid hormones, in addition to increasing the cellularity and vascularity of the thyroid gland. The TSH receptor (TSH-R) belongs to a family of G-protein–coupled receptors that have seven transmembrane-spanning domains and use cyclic adenosine monophosphate in the signal-transduction pathway. TSH secretion by the anterior pituitary is also regulated via a negative feedback loop by T₄ and T₃. Because the pituitary has the ability to convert T₄ to T₃, the latter is thought to be more important in this feedback control. T₃ also inhibits the release of TRH.

The thyroid gland also is capable of autoregulation, which allows it to modify its function independent of TSH. As an adaptation to low iodide intake, the gland preferentially synthesizes T₃ rather than T₄, thereby increasing the efficiency of secreted hormone. In situations of iodine excess, iodide transport, peroxide generation, and synthesis and secretion of thyroid hormones are inhibited. Excessively large doses of iodide may lead to initial increased organification, followed by suppression, a phenomenon called the Wolff-Chaikoff effect. Epinephrine and human chorionic gonadotropin hormones stimulate thyroid hormone production. Thus, elevated thyroid hormone levels are found in pregnancy and gynecologic malignancies such as hydatidiform mole. In contrast, glucocorticoids inhibit thyroid hormone production. In severely ill patients, peripheral thyroid hormones may be reduced, without a compensatory increase in TSH levels, giving rise to the euthyroid sick syndrome.

Thyroid Hormone Function

Free thyroid hormone enters the cell membrane by diffusion or by specific carriers and is carried to the nuclear membrane by binding to specific proteins. T₄ is deiodinated to T₃ and enters the nucleus via active transport, where it binds to the thyroid hormone receptor. The T₃ receptor is similar to the nuclear receptors for glucocorticoids, mineralocorticoids, estrogens, vitamin D, and retinoic acid. In humans, two types of T₃ receptor genes (α and β) are located on chromosomes 3 and 17. Thyroid receptor expression depends on peripheral concentrations of thyroid hormones and is tissue specific—the α form is abundant in the central nervous system, whereas the β form predominates in the liver. Each gene product has a ligand-independent, amino-terminal domain; a ligand-binding, carboxy-terminal domain; and centrally located DNA-binding regions. Binding of thyroid hormone leads to the transcription and translation of specific hormone-responsive genes.

Thyroid hormones affect almost every system in the body. They are important for fetal brain development and skeletal maturation. T₃ increases oxygen consumption, basal metabolic rate, and heat production by stimulation of Na⁺/K⁺ ATPase in various tissues. It also has positive inotropic and chronotropic effects on the heart by increasing transcription of the Ca²⁺ ATPase in the sarcoplasmic reticulum and increasing levels of β-adrenergic receptors and concentration of G proteins. Myocardial α receptors are decreased, and actions of catecholamines are amplified. Thyroid hormones are responsible for maintaining the normal hypoxic and hypercapnic drive in the respiratory center of the brain. They also increase gastrointestinal (GI) motility, leading to diarrhea in hyperthyroidism and constipation in hypothyroidism. Thyroid hormones also increase bone and protein turnover and the speed of muscle contraction and relaxation. They also increase glycogenolysis, hepatic gluconeogenesis, intestinal glucose absorption, and cholesterol synthesis and degradation.

Evaluation of Patients with Thyroid Disease

Tests of Thyroid Function

A multitude of different tests are available to evaluate thyroid function. No single test is sufficient to assess thyroid function in all situations, and the results must be interpreted in the context of the patient’s clinical condition. TSH is the only test necessary in most patients with thyroid nodules that clinically appear to be euthyroid.

Serum Thyroid-Stimulating Hormone (Normal 0.5–5 μU/mL)

The tests for serum TSH are based on the following principle: monoclonal TSH antibodies are bound to a solid matrix and bind serum TSH. A second monoclonal antibody binds to a separate epitope on TSH and is labeled with radioisotope, enzyme, or fluorescent tag. Therefore, the amount of serum TSH is proportional to the amount of bound secondary antibody (immunometric assay). Serum TSH levels reflect the ability of the anterior pituitary to detect free T₄ levels. There is an inverse relationship between the free T₄ level and the logarithm of the TSH concentration—small changes in free T₄ lead to a large shift in TSH levels. The ultrasensitive TSH assay has become the most sensitive and specific test for the diagnosis of hyper- and hypothyroidism and for optimizing T₄ therapy.

Total T₄ (Reference Range 55–150 nmol/L) and T₃ (Reference Range 1.5–3.5 nmol/L)

Total T₄ and T₃ levels are measured by radioimmunoassay and measure both the free and bound components of the hormones. Total T₄ levels reflect the output from the thyroid gland, whereas T₃ levels in the nonstimulated thyroid gland are more indicative of peripheral thyroid hormone metabolism, and are, therefore, not generally suitable as a general screening test. Total T₄ levels are increased not only in hyperthyroid patients, but also in those with elevated Tg levels secondary to pregnancy, estrogen/progesterone use, or congenital diseases. Similarly, total T₄ levels decrease in hypothyroidism and in patients with decreased Tg levels due to anabolic steroid use and protein-losing disorders like nephrotic syndrome. Individuals with these latter disorders may be euthyroid if their free T₄ levels are normal. Measurement of total T₃ levels is important in clinically hyperthyroid patients with normal T₄ levels, who may have T₃ thyrotoxicosis. As discussed previously in Thyroid Hormone Synthesis, Secretion, and Transport, total T₃ levels often are increased in early hypothyroidism.

Free T₄ (Reference Range 12–28 pmol/L) and Free T₃ (3–9 pmol/L)

These radioimmunoassay-based tests are a sensitive and accurate measurement of biologically active thyroid hormone. Free T₄ estimates are not performed as a routine screening tool in thyroid disease. Use of this test is confined to cases of early hyperthyroidism in which total T₄ levels may be normal but free T₄ levels are raised. In patients with end-organ resistance to T₄ (Refetoff’s syndrome), T₄ levels are increased, but TSH levels usually are normal. Free T₃ is most useful in confirming the diagnosis of early hyperthyroidism, in which levels of free T₄ and free T₃ rise before total T₄ and T₃. Free T₄ levels may also be measured indirectly using the T₃-resin uptake test. If free T₄ levels are increased, fewer hormone binding sites are available for binding radiolabeled T₃ that has been added to the patient’s serum. Therefore, more T₃ binds with an ion-exchange resin, and the T₃-resin uptake is increased.

Thyrotropin-Releasing Hormone

This test is useful to evaluate pituitary TSH secretory function and is performed by administering 500 μg of TRH intravenously and measuring TSH levels after 30 and 60 minutes. In a normal individual, TSH levels should increase at least 6 μIU/mL from the baseline. This test also was previously used to assess patients with borderline hyperthyroidism but has largely been replaced by sensitive TSH assays for this purpose.

Thyroid Antibodies

Thyroid antibodies include anti-Tg, antimicrosomal, or anti-TPO and thyroid-stimulating immunoglobulin (TSI). Anti-Tg and anti-TPO antibody levels do not determine thyroid function, but rather indicate the underlying disorder, usually an autoimmune thyroiditis. About 80% of patients with Hashimoto’s thyroiditis have elevated thyroid antibody levels; however, levels may also be increased in patients with Graves’ disease, multinodular goiter, and occasionally, thyroid neoplasms.

Serum Thyroglobulin

Tg is only made by normal or abnormal thyroid tissue. It normally is not released into the circulation in large amounts but increases dramatically in destructive processes of the thyroid gland, such as thyroiditis, or overactive states such as Graves’ disease and toxic multinodular goiter. The most important use for serum Tg levels is in monitoring patients with differentiated thyroid cancer for recurrence, particularly after total thyroidectomy and RAI ablation. Elevated anti-Tg antibodies can interfere with the accuracy of serum Tg levels and should always be measured when interpreting Tg levels.

Serum Calcitonin (0–4 pg/mL Basal)

This 32-amino-acid polypeptide is secreted by the C cells and functions to lower serum calcium levels, although in humans, it has only minimal physiologic effects. It is also a sensitive marker of MTC.

Thyroid Imaging

Radionuclide Imaging

Both iodine-123 (¹²³I) and iodine-131 (¹³¹I) are used to image the thyroid gland. The former emits low-dose radiation, has a half-life of 12 to 14 hours, and is used to image lingual thyroids or goiters. In contrast, ¹³¹I has a half-life of 8 to 10 days and leads to higher-dose radiation exposure. Therefore, this isotope is used to screen and treat patients with differentiated thyroid cancers for metastatic disease. The images obtained by these studies provide information not only about the size and shape of the gland, but also the distribution of functional activity. Areas that trap less radioactivity than the surrounding gland are termed cold (Fig. 38-10), whereas areas that demonstrate increased activity are termed hot. The risk of malignancy is higher in “cold” lesions (20%) compared to “hot” or “warm” lesions (<5%). Technetium Tc 99m pertechnetate (^99mTc) is taken up by the thyroid gland and is increasingly being used for thyroid evaluation. This isotope is taken up by the mitochondria, but is not organified. It also has the advantage of having a shorter half-life and minimizes radiation exposure. It is particularly sensitive for nodal metastases. More recently, ¹⁸F-fluorodeoxyglucose (FDG) positron emission tomography (PET) combined with computed tomography (CT) is being increasingly used to screen for metastases in patients with thyroid cancer in whom other imaging studies are negative. PET scans are not routinely used in the evaluation of thyroid nodules; however, they may show clinically occult thyroid lesions. There are several recent reports of rates of malignancy in these lesions ranging from 14% to 63%. These incidentally discovered nodules should be worked up by ultrasound and fine-needle aspiration biopsy (FNAB).

Ultrasound

Ultrasound is an excellent noninvasive and portable imaging study of the thyroid gland with the added advantage of no radiation exposure. It is helpful in the evaluation of thyroid nodules, distinguishing solid from cystic ones, and providing information about size and multicentricity. In addition, characteristics such as echotexture, shape, borders and presence of calcifications, and vascularity can provide useful information regarding risk of malignancy. Ultrasound is also especially helpful for assessing cervical lymphadenopathy (Fig. 38-11) and to guide FNAB. An experienced ultrasonographer is necessary for the best results.

Computed Tomography/Magnetic Resonance Imaging Scan

CT and magnetic resonance imaging (MRI) studies provide excellent imaging of the thyroid gland and adjacent nodes and are particularly useful in evaluating the extent of large, fixed, or substernal goiters (which cannot be evaluated by ultrasound) and their relationship to the airway and vascular structures. Noncontrast CT scans should be obtained for patients who are likely to require subsequent RAI therapy. If contrast is necessary, therapy needs to be delayed by several months. Combined PET-CT scans are increasingly being used for Tg-positive, RAI-negative tumors.

Benign Thyroid Disorders

Hyperthyroidism

The clinical manifestations of hyperthyroidism result from an excess of circulating thyroid hormone. Hyperthyroidism may arise from a number of conditions that are listed in Table 38-1. It is important to distinguish disorders such as Graves’ disease and toxic nodular goiters that result from increased production of thyroid hormone from those disorders that lead to a release of stored hormone from injury to the thyroid gland (thyroiditis) or from other nonthyroid gland–related conditions. The former disorders lead to an increase in RAI uptake (RAIU), whereas the latter group is characterized by low RAIU. Of these disorders, Graves’ disease, toxic multinodular goiter, and solitary toxic nodule are most relevant to the surgeon.

Diffuse Toxic Goiter (Graves’ Disease)

Although originally described by the Welsh physician Caleb Parry in a posthumous article in 1825, this disorder is known as Graves’ disease after Robert Graves, an Irish physician who described three patients in 1835. Graves’ disease is by far the most common cause of hyperthyroidism in North America, accounting for 60% to 80% of cases. It is an autoimmune disease with a strong familial predisposition, female preponderance (5:1), and peak incidence between the ages of 40 and 60 years. Graves’ disease is characterized by thyrotoxicosis, diffuse goiter, and extrathyroidal conditions including ophthalmopathy, dermopathy (pretibial myxedema), thyroid acropachy, gynecomastia, and other manifestations.

Etiology, Pathogenesis, and Pathology

The exact etiology of the initiation of the autoimmune process in Graves’ disease is not known. However, conditions such as the postpartum state, iodine excess, lithium therapy, and bacterial and viral infections have been suggested as possible triggers. Genetic factors also play a role, as haplotyping studies indicate that Graves’ disease is associated with certain human leukocyte antigen (HLA) haplotypes, including HLA-B8, HLA-DR3, and HLADQA1*0501 in Caucasian patients, whereas HLA-DRB1*0701 is protective against it. Polymorphisms of the cytotoxic T-lymphocyte antigen 4 (CTLA-4) gene also have been associated with Graves’ disease development.³ Once initiated, the process causes sensitized T-helper lymphocytes to stimulate B lymphocytes, which produce antibodies directed against the thyroid hormone receptor. TSIs or antibodies that stimulate the TSH-R, as well as TSH-binding inhibiting immunoglobulins or antibodies, have been described. The thyroid-stimulating antibodies stimulate the thyrocytes to grow and synthesize excess thyroid hormone, which is a hallmark of Graves’ disease. Graves’ disease also is associated with other autoimmune conditions such as type 1 diabetes mellitus, Addison’s disease, pernicious anemia, and myasthenia gravis.

Macroscopically, the thyroid gland in patients with Graves’ disease is diffusely and smoothly enlarged, with a concomitant increase in vascularity. Microscopically, the gland is hyperplastic, and the epithelium is columnar with minimal colloid present. The nuclei exhibit mitosis, and papillary projections of hyperplastic epithelium are common. There may be aggregates of lymphoid tissue, and vascularity is markedly increased.

Clinical Features

The clinical manifestations of Graves’ disease can be divided into those related to hyperthyroidism and those specific to Graves’ disease. Hyperthyroid symptoms include heat intolerance, increased sweating and thirst, and weight loss despite adequate caloric intake. Symptoms of increased adrenergic stimulation include palpitations, nervousness, fatigue, emotional lability, hyperkinesis, and tremors. The most common GI symptoms include increased frequency of bowel movements and diarrhea. Female patients often develop amenorrhea, decreased fertility, and an increased incidence of miscarriages. Children experience rapid growth with early bone maturation, whereas older patients may present with cardiovascular complications such as atrial fibrillation and congestive heart failure.

On physical examination, weight loss and facial flushing may be evident. The skin is warm and moist, and African American patients often note darkening of their skin. Tachycardia or atrial fibrillation is present, with cutaneous vasodilation leading to a widening of the pulse pressure and a rapid falloff in the transmitted pulse wave (collapsing pulse). A fine tremor, muscle wasting, and proximal muscle group weakness with hyperactive tendon reflexes often are present.

Approximately 50% of patients with Graves’ disease also develop clinically evident ophthalmopathy, and dermopathy occurs in 1% to 2% of patients. It is characterized by deposition of glycosaminoglycans, leading to thickened skin in the pretibial region and dorsum of the foot (Fig. 38-12). Eye symptoms include lid lag (von Graefe’s sign), spasm of the upper eyelid revealing the sclera above the corneoscleral limbus (Dalrymple’s sign), and a prominent stare, due to catecholamine excess. True infiltrative eye disease results in periorbital edema, conjunctival swelling and congestion (chemosis), proptosis, limitation of upward and lateral gaze (from involvement of the inferior and medial rectus muscles, respectively), keratitis, and even blindness due to optic nerve involvement. The etiology of Graves’ ophthalmopathy is not completely known; however, orbital fibroblasts and muscles are thought to share a common antigen, the TSH-R. Ophthalmopathy is thought to result from inflammation caused by cytokines released from sensitized killer T lymphocytes and cytotoxic antibodies. Gynecomastia is common in young men. Rare bony involvement leads to subperiosteal bone formation and swelling in the metacarpals (thyroid acropachy). Onycholysis, or separation of fingernails from their beds, is a commonly observed finding. On physical examination, the thyroid usually is diffusely and symmetrically enlarged, as evidenced by an enlarged pyramidal lobe. There may be an overlying bruit or thrill over the thyroid gland and a loud venous hum in the supraclavicular space.

Diagnostic Tests

The diagnosis of hyperthyroidism is made by a suppressed TSH with or without an elevated free T₄ or T₃ level. If eye signs are present, other tests are generally not needed. However, in the absence of eye findings, an ¹²³I uptake and scan should be performed. An elevated uptake, with a diffusely enlarged gland, confirms the diagnosis of Graves’ disease and helps to differentiate it from other causes of hyperthyroidism. Technetium scintigraphy (using pertechnetate, which is trapped by the thyroid, but not organified) can also be used to determine etiology. While technetium scans results in low range of normal uptake and high background activity, total-body radiation exposure is less than that of ¹²³I scans. If free T₄ levels are normal, free T₃ levels should be determined, as they often are elevated in early Graves’ disease or toxic nodules (T₃ toxicosis). Anti-Tg and anti-TPO antibodies are elevated in up to 75% of patients but are not specific. Elevated TSH-R or thyroid-stimulating antibodies (TSAb) are diagnostic of Graves’ disease and are increased in about 90% of patients. MRI scans of the orbits are useful in evaluating Graves’ ophthalmopathy.

Treatment

Graves’ disease may be treated by any of three treatment modalities: antithyroid drugs, thyroid ablation with radioactive ¹³¹I, and thyroidectomy. The choice of treatment depends on several factors, as discussed in the following sections.

Antithyroid medications generally are administered in preparation for RAI ablation or surgery. The drugs commonly used are propylthiouracil (PTU, 100 to 300 mg three times daily) and methimazole (10 to 30 mg three times daily, then once daily). Methimazole has a longer half-life and can be dosed once daily. Both drugs reduce thyroid hormone production by inhibiting the organic binding of iodine and the coupling of iodotyrosines (mediated by TPO). In addition, PTU also inhibits the peripheral conversion of T₄ to T₃, making it useful for the treatment of thyroid storm. Both drugs can cross the placenta, inhibiting fetal thyroid function, and are excreted in breast milk, although PTU has a lower risk of transplacental transfer. Methimazole also has been associated with congenital aplasia; therefore, PTU is preferred in pregnant and breastfeeding women. Side effects of treatment include reversible granulocytopenia, skin rashes, fever, peripheral neuritis, polyarteritis, vasculitis, hepatitis, and, rarely, agranulocytosis and aplastic anemia. Patients should be monitored for these possible complications and should always be warned to stop PTU or methimazole immediately and seek medical advice should they develop a sore throat or fever. Treatment of agranulocytosis involves admission to the hospital, discontinuation of the drug, and broad-spectrum antibiotic therapy. Surgery should be postponed until the granulocyte count reaches 1000 cells/mm³.

The dose of antithyroid medication is titrated as needed in accordance with TSH and T₄ levels. Most patients have improved symptoms in 2 weeks and become euthyroid in about 6 weeks. Some physicians use the block-replace regimen, by adding T₄ (0.05 to 0.10 mg) to prevent hypothyroidism and suppress TSH secretion, because some, but not all, studies suggest that this reduces recurrence rates. The length of therapy is debated. Treatment with antithyroid medications is associated with a high relapse rate when these drugs are discontinued, with 40% to 80% of patients developing recurrent disease after a 1- to 2-year course. Patients with small glands are less likely to recur so that treatment for curative intent is reserved for patients with (a) small, nontoxic goiters less than 40 g; (b) mildly elevated thyroid hormone levels; (c) negative or low or titers of thyroid hormone receptor antibodies; and (d) rapid decrease in gland size with antithyroid medications. The catecholamine response of thyrotoxicosis can be alleviated by administering β-blocking agents. β-Blockade should be considered in all patients with symptomatic thyrotoxicosis and is recommended for elderly patients, those with coexistent cardiac disease, and patients with resting heart rates >90 bpm. These drugs have the added effect of decreasing the peripheral conversion of T₄ to T₃. Propranolol is the most commonly prescribed medication in doses of about 20 to 40 mg four times daily. Higher doses are sometimes required due to increased clearance of the medication. Caution should be exercised in patients with asthma. Calcium channel blockers are useful for rate control in patients in whom β-blockers are contraindicated.

RAI forms the mainstay of Graves’ disease treatment in North America. The major advantages of this treatment are the avoidance of a surgical procedure and its concomitant risks, reduced overall treatment costs, and ease of treatment. Antithyroid drugs are given until the patient is euthyroid and then discontinued to maximize drug uptake. The ¹³¹I dose is calculated after a preliminary scan and usually consists of 8 to 12 mCi administered orally. After standard treatment with RAI, most patients become euthyroid within 2 months. However, only about 50% of patients treated with RAI are euthyroid 6 months after treatment, and the remaining are still hyperthyroid or already hypothyroid.⁴ After 1 year, about 2.5% of patients develop hypothyroidism each year. RAI also has been documented to lead to progression of Graves’ ophthalmopathy (33% after RAI compared to 16% after surgery), and ophthalmopathy is more common in smokers. Although there is no evidence of long-term problems with infertility, and overall cancer incidence rates are unchanged, there is a small increased risk of nodular goiter, thyroid cancer,⁵ and hyperparathyroidism (HPT)⁶ in patients who have been treated with RAI. Patients treated with RAI have an unexplained increase in their overall and cardiovascular mortality rates when compared to the general population.

RAI therapy is therefore most often used in older patients with small or moderate-sized goiters, those who have relapsed after medical or surgical therapy, and those in whom antithyroid drugs or surgery are contraindicated. Absolute contraindications to RAI include women who are pregnant (or planning pregnancy within 6 months of treatment) or breastfeeding. Relative contraindications include young patients (i.e., especially children and adolescents), those with thyroid nodules, and those with ophthalmopathy. Lack of access to a high-volume thyroid surgeon is also a consideration. The higher the initial dose of ¹³¹I, the earlier the onset and the higher the incidence of hypothyroidism.

In North America, surgery is recommended when RAI is contraindicated as in patients who (a) have confirmed cancer or suspicious thyroid nodules, (b) are young, (c) desire to conceive soon (<6 months) after treatment, (d) have had severe reactions to antithyroid medications, (e) have large goiters (>80 g) causing compressive symptoms, and (f) are reluctant to undergo RAI therapy. Relative indications for thyroidectomy include patients, particularly smokers, with moderate to severe Graves’ ophthalmopathy, those desiring rapid control of hyperthyroidism with a chance of being euthyroid, and those demonstrating poor compliance to antithyroid medications. Pregnancy is also a relative contraindication, and surgery should be used only when rapid control is needed and antithyroid medications cannot be used. Surgery is best performed in the second trimester. The goal of thyroidectomy for Graves’ disease should be the complete and permanent control of the disease with minimal morbidity. Patients should be rendered euthyroid before operation with antithyroid drugs that should be continued up to the day of surgery. Lugol’s iodide solution or saturated potassium iodide generally is administered beginning 7 to 10 days preoperatively (three drops twice daily) to reduce vascularity of the gland and decrease the risk of precipitating thyroid storm. The major action of iodine in this situation is to inhibit release of thyroid hormone. If it is not possible to render the patient euthyroid prior to surgery (if the surgery is urgent or the patient is allergic to antithyroid medications), the patient can be prepared with β-blockade and potassium iodide alone. Steroids can be a useful adjunct in this situation.

The extent of thyroidectomy to be performed used to be determined by the desired outcome (risk of recurrence vs. euthyroidism) and surgeon experience. In patients with coexistent thyroid cancer and those who refused RAI therapy or had severe ophthalmopathy or life-threatening reactions to antithyroid medications (vasculitis, agranulocytosis, or liver failure), total or near-total thyroidectomy was recommended. Ophthalmopathy has been demonstrated to stabilize or improve in most patients after total thyroidectomy, presumably from removal of the antigenic stimulus. A subtotal thyroidectomy, leaving a 4- to 7-g remnant, was recommended for all remaining patients. During subtotal thyroidectomy, remnant tissue may be left on each side (bilateral subtotal thyroidectomy), or a total lobectomy can be performed on one side with a subtotal thyroidectomy on the other side (Hartley-Dunhill procedure). Results were similar with either procedure,⁷ but the latter procedure was theoretically associated with fewer complications and requires re-entering only one side of the neck should recurrence require reoperation. Most studies, however, show no difference in the rates of complications with either approach, although patients undergoing a total resection had higher rates of temporary hypoparathyroidism. However, patients treated with subtotal thyroidectomy are prone to recurrence, the rates of which are dependent on remnant size. Based on the current evidence, recently published guidelines from the American Thyroid Association (ATA) and the American Association of Clinical Endocrinologists (AACE) recommend total or near-total thyroidectomy as the procedure of choice for the surgical management of Graves’ disease.⁸ Recurrent thyrotoxicosis usually is managed by radioiodine treatment.

Toxic Multinodular Goiter

Toxic multinodular goiters usually occur in older individuals, who often have a prior history of a nontoxic multinodular goiter. Over several years, enough thyroid nodules become autonomous to cause hyperthyroidism. The presentation often is insidious in that hyperthyroidism may only become apparent when patients are placed on low doses of thyroid hormone suppression for the goiter. Some patients have T₃ toxicosis, whereas others may present only with atrial fibrillation or congestive heart failure. Hyperthyroidism also can be precipitated by iodide-containing drugs such as contrast media and the antiarrhythmic agent amiodarone (Jod-Basedow hyperthyroidism). Symptoms and signs of hyperthyroidism are similar to Graves’ disease, but extrathyroidal manifestations are absent.

Diagnostic Studies

Blood tests are similar to Graves’ disease with a suppressed TSH level and elevated free T₄ or T₃ levels. RAI uptake also is increased, showing multiple nodules with increased uptake and suppression of the remaining gland.

Treatment

Hyperthyroidism must be adequately controlled. Both RAI and surgical resection may be used for treatment. When surgery is performed, near-total or total thyroidectomy is recommended to avoid recurrence and the consequent increased complication rates with repeat surgery. Care must be taken in identifying the RLN, which may be found laterally on the thyroid (rather than posterior) or stretched anteriorly over a nodule. RAI therapy is reserved for elderly patients who represent very poor operative risks, provided there is no airway compression from the goiter and thyroid cancer is not a concern. However, because uptake is less than in Graves’ disease, larger doses of RAI often are needed to treat the hyperthyroidism. Furthermore, RAI-induced thyroiditis has the potential to cause swelling and acute airway compromise and leaves the goiter intact, with the possibility of recurrent hyperthyroidism.

Toxic Adenoma

Hyperthyroidism from a single hyperfunctioning nodule typically occurs in younger patients who note recent growth of a long-standing nodule along with the symptoms of hyperthyroidism. Toxic adenomas are characterized by somatic mutations in the TSH-R gene, although G-protein–stimulating gene (gsp) mutations may occur also.⁹ Most hyperfunctioning or autonomous thyroid nodules have attained a size of at least 3 cm before hyperthyroidism occurs. Physical examination usually reveals a solitary thyroid nodule without palpable thyroid tissue on the contralateral side. RAI scanning shows a “hot” nodule with suppression of the rest of the thyroid gland. These nodules are rarely malignant. Smaller nodules may be managed with antithyroid medications and RAI. Larger nodules can require higher doses, which can lead to hypothyroidism. Surgery (lobectomy and isthmusectomy) is preferred to treat young patients and those with larger nodules. Percutaneous ethanol injection (PEI) has been reported to have reasonable success rates but has not been directly compared with surgery.

Thyroid Storm

Thyroid storm is a condition of hyperthyroidism accompanied by fever, central nervous system agitation or depression, and cardiovascular and GI dysfunction, including hepatic failure. The condition may be precipitated by abrupt cessation of antithyroid medications, infection, thyroid or nonthyroid surgery, and trauma in patients with untreated thyrotoxicosis. Occasionally, thyroid storm may result from amiodarone administration or exposure to iodinated contrast agents or following RAI therapy. This condition was previously associated with high mortality rates but can be appropriately managed in an intensive care unit setting. β-Blockers are given to reduce peripheral T₄ to T₃ conversion and decrease the hyperthyroid symptoms. Oxygen supplementation and hemodynamic support should be instituted. Nonaspirin compounds can be used to treat pyrexia, and Lugol’s iodine or sodium ipodate (intravenously) should be administered to decrease iodine uptake and thyroid hormone secretion. PTU therapy blocks the formation of new thyroid hormone and reduces peripheral conversion of T₄ to T₃. Corticosteroids often are helpful to prevent adrenal exhaustion and block hepatic thyroid hormone conversion.

Hypothyroidism

Deficiency in circulating levels of thyroid hormone leads to hypothyroidism and, in neonates, to cretinism, which is characterized by neurologic impairment and mental retardation. Hypothyroidism also may occur in Pendred’s syndrome (associated with deafness) and Turner’s syndrome. Conditions that cause hypothyroidism are listed in Table 38-2.

Clinical Features

Failure of thyroid gland development or function in utero leads to cretinism and characteristic facies similar to those of children with Down syndrome and dwarfism. Failure to thrive and severe mental retardation often are present. Immediate testing and treatment with thyroid hormone at birth can lessen the neurologic and intellectual deficits. Hypothyroidism developing in childhood or adolescence results in delayed development and may also lead to abdominal distention, umbilical hernia, and rectal prolapse. In adults, symptoms in general are nonspecific, including tiredness, weight gain, cold intolerance, constipation, and menorrhagia. Patients with severe hypothyroidism or myxedema develop characteristic facial features due to the deposition of glycosaminoglycans in the subcutaneous tissues, leading to facial and periorbital puffiness. The skin becomes rough and dry and often develops a yellowish hue from reduced conversion of carotene to vitamin A. Hair becomes dry and brittle, and severe hair loss may occur. There is also a characteristic loss of the outer two thirds of the eyebrows. An enlarged tongue may impair speech, which is already slowed, in keeping with the impairment of mental processes. Patients may also have nonspecific abdominal pain accompanied by distention and constipation. Libido and fertility are impaired in both sexes. Cardiovascular changes in hypothyroidism include bradycardia, cardiomegaly, pericardial effusion, reduced cardiac output, and pulmonary effusions. When hypothyroidism occurs as a result of pituitary failure, other features of hypopituitarism, such as pale, waxy skin; loss of body hair; and atrophic genitalia, may be present.

Laboratory Findings

Hypothyroidism is characterized by low circulating levels of T₄ and T₃. Raised TSH levels are found in primary thyroid failure, whereas secondary hypothyroidism is characterized by low TSH levels that do not increase following TRH stimulation. Thyroid autoantibodies are highest in patients with autoimmune disease (Hashimoto’s thyroiditis, Graves’ disease) and may also be elevated in patients with nodular goiter and thyroid neoplasms. An electrocardiogram demonstrates decreased voltage with flattening or inversion of T waves.

Treatment

T₄ is the treatment of choice and is administered in dosages varying from 50 to 200 μg per day, depending on the patient’s size and condition. Starting doses of 100 μg of T₄ daily are well tolerated; however, elderly patients and those with coexisting heart disease and profound hypothyroidism should be started on a considerably lower dose such as 25 to 50 μg daily because of associated hypercholesterolemia and atherosclerosis. The dose can be slowly increased over weeks to months to attain a euthyroid state. A baseline electrocardiogram should always be obtained in patients with severe hypothyroidism before treatment. T₄ dosage is titrated against clinical response and TSH levels, which should return to normal. The management of patients with subclinical hypothyroidism (normal T₄, slightly raised TSH) is controversial. Some evidence suggests that patients with subclinical hypothyroidism and increased antithyroid antibody levels should be treated, because they will subsequently develop hypothyroidism. Patients who present with myxedema coma may require initial emergency treatment with large doses of IV T₄ (300 to 400 μg), with careful monitoring in an intensive care unit setting.

Thyroiditis

Thyroiditis usually is classified into acute, subacute, and chronic forms, each associated with a distinct clinical presentation and histology.

Acute (Suppurative) Thyroiditis

The thyroid gland is inherently resistant to infection due to its extensive blood and lymphatic supply, high iodide content, and fibrous capsule. However, infectious agents can seed it (a) via the hematogenous or lymphatic route, (b) via direct spread from persistent pyriform sinus fistulae or thyroglossal duct cysts, (c) as a result of penetrating trauma to the thyroid gland, or (d) due to immunosuppression. Streptococcus and anaerobes account for about 70% of cases; however, other species also have been cultured.¹⁰ Acute suppurative thyroiditis is more common in children and often is preceded by an upper respiratory tract infection or otitis media. It is characterized by severe neck pain radiating to the jaws or ear, fever, chills, odynophagia, and dysphonia. Complications such as systemic sepsis, tracheal or esophageal rupture, jugular vein thrombosis, laryngeal chondritis, and perichondritis or sympathetic trunk paralysis may also occur.

The diagnosis is established by leukocytosis on blood tests and FNAB for Gram’s stain, culture, and cytology. CT scans may help to delineate the extent of infection and identify abscesses. A persistent pyriform sinus fistula should always be suspected in children with recurrent acute thyroiditis. The sensitivity of identification of fistulae in the acute setting is lowest for barium esophagography (50%) and best for direct endoscopy (100%), with CT scans being intermediate (80%). Both barium esophagogram and CT scans have improved sensitivity once the acute inflammation has resolved (100% and 83%, respectively), with CT being better at defining the accurate anatomic pathway and its relationship to the thyroid gland.¹¹ Treatment consists of parenteral antibiotics and drainage of abscesses. Thyroidectomy may be needed for persistent abscesses or failure of open drainage. Patients with pyriform sinus fistulae require complete resection of the sinus tract, including the area of the thyroid where the tract terminates, to prevent recurrence. Transnasal flexible fiberoptic laryngoscopy is being increasingly used to identify the internal opening of the pyriform sinus tract and may also allow electrocauterization of the tract, and success rates similar to open surgery have been reported.

Subacute Thyroiditis

Subacute thyroiditis can occur in the painful or painless forms. Although the exact etiology is not known, painful thyroiditis is thought to be viral in origin or result from a postviral inflammatory response. Genetic predisposition may also play a role, as manifested by its strong association with the HLA-B35 haplotype. One model of pathogenesis suggests that viral or thyroid antigens, when presented by macrophages in the context of HLA-B35, stimulate cytotoxic T lymphocytes and damage thyroid follicular cells.

Painful thyroiditis most commonly occurs in 30- to 40-year-old women and is characterized by the sudden or gradual onset of neck pain, which may radiate toward the mandible or ear. History of a preceding upper respiratory tract infection often can be elicited. The gland is enlarged, exquisitely tender, and firm. The disorder classically progresses through four stages. An initial hyperthyroid phase, due to release of thyroid hormone, is followed by a second, euthyroid phase. The third phase, hypothyroidism, occurs in about 20% to 30% of patients and is followed by resolution and return to the euthyroid state in >90% of patients. A few patients develop recurrent disease.

In the early stages of the disease, TSH is decreased, and Tg, T₄, and T₃ levels are elevated due to the release of preformed thyroid hormone from destroyed follicles. The erythrocyte sedimentation rate is typically >100 mm/h. RAIU also is decreased (<2% at 24 hours), even in euthyroid patients, due to the release of thyroid hormones from destruction of the thyroid parenchyma. Painful thyroiditis is self-limited, and therefore, treatment is primarily symptomatic. Aspirin and other nonsteroidal anti-inflammatory drugs are used for pain relief, but steroids may be indicated in more severe cases. Short-term thyroid replacement may be needed and may shorten the duration of symptoms. Thyroidectomy is reserved for the rare patient who has a prolonged course not responsive to medical measures or for recurrent disease.

Painless thyroiditis is considered to be autoimmune in origin and may occur sporadically or in the postpartum period; the latter typically occurs at about 6 weeks after delivery in women with high TPO antibody titers in early pregnancy. This timing is thought to coincide with a decrease in the normal immune tolerance of pregnancy and consequent rebound elevation of antibody titers.

Painless thyroiditis also is more common in women and usually occurs between 30 and 60 years of age. Physical examination demonstrates a normal sized or minimally enlarged, slightly firm, nontender gland. Laboratory tests and RAIU are similar to those in painful thyroiditis, except for a normal erythrocyte sedimentation rate. The clinical course also parallels painful thyroiditis. Patients with symptoms may require β-blockers and thyroid hormone replacement. Thyroidectomy or RAI ablation is only indicated for the rare patient with recurrent, disabling episodes of thyroiditis.

Chronic Thyroiditis

Lymphocytic (Hashimoto’s) Thyroiditis

Lymphocytic thyroiditis was first described by Hashimoto in 1912 as struma lymphomatosa—a transformation of thyroid tissue to lymphoid tissue. It is the most common inflammatory disorder of the thyroid and the leading cause of hypothyroidism.

Hashimoto’s thyroiditis is an autoimmune process that is thought to be initiated by the activation of CD4⁺ T (helper) lymphocytes with specificity for thyroid antigens. Once activated, T cells can recruit cytotoxic CD8⁺ T cells to the thyroid. Hypothyroidism results not only from the destruction of thyrocytes by cytotoxic T cells but also by autoantibodies, which lead to complement fixation and killing by natural killer cells or block the TSH-R. Antibodies are directed against three main antigens—Tg (60%), TPO (95%), and TSH-R (60%)—and, less commonly, the sodium/iodine symporter (25%). Apoptosis (programmed cell death) also has been implicated in the pathogenesis of Hashimoto’s thyroiditis. Chronic thyroiditis also has been associated with increased intake of iodine and administration of medications such as interferon-α, lithium, and amiodarone. Support for an inherited predisposition includes an increased incidence of thyroid autoantibodies in first-degree relatives of patients with Hashimoto’s thyroiditis compared to controls and the occurrence of the autoantibodies and hypothyroidism in patients with specific chromosomal abnormalities such as Turner’s syndrome and Down syndrome. Associations with HLA-B8, DR3, and DR5 haplotypes of the major histocompatibility complex also have been described.

On gross examination, the thyroid gland is usually mildly enlarged throughout and has a pale, gray-tan cut surface that is granular, nodular, and firm. On microscopic examination, the gland is diffusely infiltrated by small lymphocytes and plasma cells and occasionally shows well-developed germinal centers. Thyroid follicles are smaller than normal with reduced amounts of colloid and increased interstitial connective tissue. The follicles are lined by Hürthle or Askanazy cells, which are characterized by abundant eosinophilic, granular cytoplasm.

Hashimoto’s thyroiditis is also more common in women (male:female ratio 1:10 to 20) between the ages of 30 and 50 years old. The most common presentation is that of a minimally or moderately enlarged firm granular gland discovered on routine physical examination or the awareness of a painless anterior neck mass, although 20% of patients present with hypothyroidism, and 5% present with hyperthyroidism (Hashitoxicosis). In classic goitrous Hashimoto’s thyroiditis, physical examination reveals a diffusely enlarged, firm gland, which also is lobulated. An enlarged pyramidal lobe often is palpable.

When Hashimoto’s thyroiditis is suspected clinically, an elevated TSH and the presence of thyroid autoantibodies usually confirm the diagnosis. FNAB with ultrasound guidance is indicated in patients who present with a solitary suspicious nodule or a rapidly enlarging goiter. Thyroid lymphoma is a rare but well-recognized, ominous complication of chronic autoimmune thyroiditis and has a prevalence 80 times higher than expected frequency in this population than in a control population without thyroiditis. Recent studies of clonal similarity indicate that lymphoma may, in fact, evolve from Hashimoto’s thyroiditis.¹²

Thyroid hormone replacement therapy is indicated in overtly hypothyroid patients, with a goal of maintaining normal TSH levels. The management of patients with subclinical hypothyroidism (normal T₄ and elevated TSH) is controversial. A systematic review of cohort studies showed that in age- and sex-adjusted analyses, subclinical hypothyroidism is associated with a hazard ratio (HR) for coronary heart disease events of 1.89 (95% confidence interval [CI], 1.28 to 2.80; P <.001) and coronary heart disease mortality of 1.58 (95% CI, 1.10 to 2.27; P = .005) for a TSH level of 10 to 19.9 mIU/L.¹³ The data for TSH levels of 5 to 10 mIU/L were less convincing. An evaluation of the 12 randomized controlled trials in this area only showed a trend toward improvement of some lipid parameters, and none of the included trials evaluated overall mortality or cardiac morbidity. Given the above, levothyroxine is recommended for all patients with TSH levels >10 μIU/mL and patients with levels of 5 to 10 μIU/mL in the presence of a goiter or anti-TPO antibodies. Treatment is also advised especially for middle-aged patients with cardiovascular risk factors such as hyperlipidemia or hypertension and in pregnant patients. Surgery may occasionally be indicated for suspicion of malignancy or for goiters causing compressive symptoms or cosmetic deformity.

Riedel’s Thyroiditis

Riedel’s thyroiditis is a rare variant of thyroiditis also known as Riedel’s struma or invasive fibrous thyroiditis that is characterized by the replacement of all or part of the thyroid parenchyma by fibrous tissue, which also invades into adjacent tissues. The etiology of this disorder is controversial, and it has been reported to occur in patients with other autoimmune diseases. This association, coupled with the presence of lymphoid infiltration and response to steroid therapy, suggests a primary autoimmune etiology. Riedel’s thyroiditis also is associated with other focal sclerosing syndromes including mediastinal, retroperitoneal, periorbital, and retro-orbital fibrosis and sclerosing cholangitis, suggesting that it may, in fact, be a primary fibrotic disorder. The disease occurs predominantly in women between the ages of 30 and 60 years old. It typically presents as a painless, hard anterior neck mass, which progresses over weeks to years to produce symptoms of compression, including dysphagia, dyspnea, choking, and hoarseness. Patients may present with symptoms of hypothyroidism and hypoparathyroidism as the gland is replaced by fibrous tissue. Physical examination reveals a hard, “woody” thyroid gland with fixation to surrounding tissues. The diagnosis needs to be confirmed by open thyroid biopsy, because the firm and fibrous nature of the gland renders FNAB inadequate.

Surgery is the mainstay of the treatment. The chief goal of operation is to decompress the trachea by wedge excision of the thyroid isthmus and to make a tissue diagnosis. More extensive resections are not advised due to the infiltrative nature of the fibrotic process that obscures usual landmarks and structures. Hypothyroid patients are treated with thyroid hormone replacement. Some patients who remain symptomatic have been reported to experience dramatic improvement after treatment with corticosteroids and tamoxifen. More recently mycophenolate mofetil has been used to attenuate the inflammatory process and led to dramatic symptom improvements in some patients.¹⁴

Goiter

Any enlargement of the thyroid gland is referred to as a goiter. The causes of nontoxic goiters are listed in Table 38-3. Goiters may be diffuse, uninodular, or multinodular. Most nontoxic goiters are thought to result from TSH stimulation secondary to inadequate thyroid hormone synthesis and other paracrine growth factors.¹⁵ Elevated TSH levels induce diffuse thyroid hyperplasia, followed by focal hyperplasia, resulting in nodules that may or may not concentrate iodine, colloid nodules, or microfollicular nodules. The TSH-dependent nodules progress to become autonomous. Familial goiters resulting from inherited deficiencies in enzymes necessary for thyroid hormone synthesis may be complete or partial. The term endemic goiter refers to the occurrence of a goiter in a significant proportion of individuals in a particular geographic region. In the past, dietary iodine deficiency was the most common cause of endemic goiter. This condition has largely disappeared in North America due to routine use of iodized salt and iodination of fertilizers, animal feeds, and preservatives. However, in areas of iodine deficiency, such as Central Asia, South America, and Indonesia, up to 90% of the population have goiters. Other dietary goitrogens that may participate in endemic goiter formation include kelp, cassava, and cabbage. In many sporadic goiters, no obvious cause can be identified.

Clinical Features

Most patients with nontoxic goiters are asymptomatic, although patients often complain of a pressure sensation in the neck. As the goiters become very large, compressive symptoms such as dyspnea and dysphagia ensue. Patients also describe having to clear their throats frequently (catarrh). Dysphonia from RLN injury is rare, except when malignancy is present. Obstruction of venous return at the thoracic inlet from a substernal goiter results in a positive Pemberton’s sign—facial flushing and dilatation of cervical veins upon raising the arms above the head (Fig. 38-13A). Sudden enlargement of nodules or cysts due to hemorrhage may cause acute pain. Physical examination may reveal a soft, diffusely enlarged gland (simple goiter) or nodules of various size and consistency in case of a multinodular goiter. Deviation or compression of the trachea may be apparent.

Diagnostic Tests

Patients usually are euthyroid with normal TSH and low-normal or normal free T₄ levels. If some nodules develop autonomy, patients have suppressed TSH levels or become hyperthyroid. RAI uptake often shows patchy uptake with areas of hot and cold nodules. FNAB is recommended in patients who have a dominant nodule or one that is painful or enlarging, as carcinomas have been reported in 5% to 10% of multinodular goiters. CT scans are helpful to evaluate the extent of retrosternal extension and airway compression (Fig. 38-13B).

Treatment

Most euthyroid patients with small, diffuse goiters do not require treatment. Some physicians give patients with large goiters exogenous thyroid hormone to reduce the TSH stimulation of gland growth; this treatment may result in decrease and/or stabilization of goiter size and is most effective for small diffuse goiters. Endemic goiters are treated by iodine administration. Surgical resection is reserved for goiters that (a) continue to increase despite T₄ suppression, (b) cause obstructive symptoms, (c) have substernal extension, (d) have malignancy suspected or proven by FNAB, and (e) are cosmetically unacceptable. Near-total or total thyroidectomy is the treatment of choice, and patients require lifelong T₄ therapy.

Solitary Thyroid Nodule

Solitary thyroid nodules are present in approximately 4% of individuals in the United States, whereas thyroid cancer has a much lower incidence of 40 new cases per 1 million. Therefore, it is of utmost importance to determine which patients with solitary thyroid nodule would benefit from surgery.

History

Details regarding the nodule, such as time of onset, change in size, and associated symptoms such as pain, dysphagia, dyspnea, or choking, should be elicited. Pain is an unusual symptom and, when present, should raise suspicion for intrathyroidal hemorrhage in a benign nodule, thyroiditis, or malignancy. Patients with MTC may complain of a dull, aching sensation. A history of hoarseness is worrisome, as it may be secondary to malignant involvement of the RLNs. Most importantly, patients should be questioned regarding risk factors for malignancy, such as exposure to ionizing radiation and family history of thyroid and other malignancies associated with thyroid cancer.

External-Beam Radiation

Low-dose therapeutic radiation has been used to treat conditions such as tinea capitis (6.5 cGy), thymic enlargement (100 to 400 cGy), enlarged tonsils and adenoids (750 cGy), acne vulgaris (200 to 1500 cGy), and other conditions such as hemangioma and scrofula. Radiation (approximately 4000 cGy) is also an integral part of the management of patients with Hodgkin’s disease. It is now known that a history of exposure to low-dose ionizing radiation to the thyroid gland places the patient at increased risk for developing thyroid cancer. The risk increases linearly from 6.5 to 2000 cGy, beyond which the incidence declines as the radiation causes destruction of the thyroid tissue. The risk is maximum 20 to 30 years after exposure, but these patients require lifelong monitoring. During the nuclear fallout from Chernobyl in 1986, ¹³¹I release was accompanied by a marked increase in the incidence of both benign and malignant thyroid lesions noted within 4 years of exposure, particularly in children.¹⁶ Most thyroid carcinomas following radiation exposure are papillary, and some of these cancers with a solid type of histology and presence of RET/PTC translocations appear to be more aggressive. In general, there is a 40% chance that patients presenting with a thyroid nodule and a history of radiation have thyroid cancer. Of those patients who have thyroid cancer, the cancer is located in the dominant nodule in 60% of patients, but in the remaining 40% of patients, the cancer is in another nodule in the thyroid gland.

Family History

A family history of thyroid cancer is a risk factor for the development of both medullary and nonmedullary thyroid cancer. Familial MTCs occur in isolation or in association with other tumors as part of multiple endocrine neoplasia type 2 (MEN2) syndromes. Nonmedullary thyroid cancers can occur in association with other known familial cancer syndromes such as Cowden’s syndrome, Werner’s syndrome (adult progeroid syndrome), and familial adenomatous polyposis (Table 38-4). Nonmedullary thyroid cancers can also occur independently of these syndromes as the predominant tumors in the families. The definition of familial nonmedullary thyroid cancer (FNMTC) is variable across the literature; however, in most studies, it is defined by the presence of two or more first-degree relatives with follicular cell–derived cancers. FNMTC is now recognized as a distinct clinical entity associated with a high incidence of multifocal tumors and benign thyroid nodules. Some studies report that these patients have higher locoregional recurrence rates and consequent shorter disease-free survival. Several candidate loci that predispose to these tumors have been identified, including MNG1 (14q32), thyroid carcinoma with oxphilia (TCO, on 19p13.2), fPTC/papillary renal neoplasia (PRN, on 1q21), NMTC (2q21), FTEN (8p23.1-p22), and the telomere-telomerase complex.¹⁷

Physical Examination

The thyroid gland is best palpated from behind the patient and with the neck in mild extension. The cricoid cartilage is an important landmark, as the isthmus is situated just below it. Nodules that are hard, gritty, or fixed to surrounding structures such as the trachea or strap muscles are more likely to be malignant. The cervical chain of lymph nodes should be assessed as well as the nodes in the posterior triangle.

Diagnostic Investigations

An algorithm for the workup of a solitary thyroid nodule is shown in Fig. 38-14.

Fine-Needle Aspiration Biopsy

FNAB has become the single most important test in the evaluation of thyroid masses and can be performed with or without ultrasound guidance.¹⁸ Ultrasound guidance is recommended for nodules that are difficult to palpate, for cystic or solid-cystic nodules that recur after the initial aspiration, and for multinodular goiters. A 23-gauge needle is inserted into the thyroid mass, and several passes are made while aspirating the syringe. After releasing the suction on the syringe, the needle is withdrawn and the cells are immediately placed on prelabeled dry glass slides; some are immersed in a 70% alcohol solution while others are air dried. A sample of the aspirate is also placed in a 90% alcohol solution for cytospin or cell pellet. The slides are stained by Papanicolaou’s or Wright’s stains and examined under the microscope. If a bloody aspirate is obtained, the patient should be repositioned in a more upright position and the biopsy repeated with a finer (25- to 30-gauge) needle.

After FNAB, the majority of nodules can be classified into several categories that determine further management. To address the issue of variability in the terminology of fine-needle aspiration (FNA), the National Cancer Institute (NCI) hosted the “NCI Thyroid Fine Needle Aspiration State of the Science Conference,” which then defined the Bethesda criteria for thyroid FNA.¹⁹ Accordingly, optimum cytology specimens should have at least six follicles each containing at least 10 to 15 cells from at least two aspirates.

The FNA is classified as “nondiagnostic or unsatisfactory” in 2% to 20% of cases and typically results from a virtually acellular specimen, cyst fluid, or the presence of blood or clotting artifact. The risk of malignancy in this setting ranges from 1% to 4%, and reaspiration under ultrasound guidance is recommended. A “benign” result is obtained in 60% to 70% of thyroid FNAs. The most common lesion in this setting is a follicular nodule (includes adenomatoid nodule, colloid nodule, and follicular adenoma). Other diagnoses include lymphocytic (Hashimoto’s) thyroiditis and granulomatous thyroiditis. False-negative results are reported in up to 3% of cases, and follow-up is recommended. A result of “atypia of unknown significance (AUS) or follicular lesion of unknown significance (FLUS)” is obtained in 3% to 6% of biopsies. The risk of malignancy in this scenario is difficult to determine; however, it is thought to be in the range of 10% to 35% for FLUS and 60% to 75% for AUS. Clinical correlation and a repeat FNA are recommended for AUS lesions (which often results in a more definitive interpretation), although clinical observation or surgery may be appropriate because of worrisome clinical or ultrasound findings. The category of “follicular neoplasm” is intended to identify nodules that might be follicular carcinomas. The term suspicious for a follicular neoplasm is preferred by some laboratories for this category because up to 35% of cases turn out not to be neoplasms but hyperplastic proliferations of follicular cells, most commonly those of multinodular goiter. Lobectomy is the preferred treatment for this result, and approximately 15% to 35% of lesions placed in this category prove to be malignant. Hürthle cell neoplasms are also included in this category. Most papillary and other carcinomas can be diagnosed by FNA, but the features are subtle at times, such as in follicular variant of papillary carcinomas. If the diagnosis is uncertain, the lesions are classified as “suspicious for malignancy.” Lobectomy or near-total thyroidectomy is recommended because more than 60% turn out to be malignant. This category also includes lesions suspicious for medullary carcinoma and lymphoma, and ancillary testing such as immunohistochemical analysis and flow cytometry may be helpful. The risk of malignancy in lesions classified as “malignant” by FNA is 97% to 99%, and near-total/total thyroidectomy is recommended.

Laboratory Studies

Most patients with thyroid nodules are euthyroid. Determining the blood TSH level is helpful. If a patient with a nodule is found to be hyperthyroid, the risk of malignancy is approximately 1%. Serum Tg levels cannot differentiate benign from malignant thyroid nodules unless the levels are extremely high, in which case metastatic thyroid cancer should be suspected. Tg levels are, however, useful in following patients who have undergone total thyroidectomy for thyroid cancer and also for serial evaluation of patients undergoing nonoperative management of thyroid nodules. Serum calcitonin levels should be obtained in patients with MTC or a family history of MTC or MEN2. There is insufficient evidence to recommend routine calcitonin testing for all nodules. All patients with MTC should be tested for RET oncogene mutations and have a 24-hour urine collection with measurement of levels of vanillylmandelic acid (VMA), metanephrine, and catecholamine levels to rule out a coexisting pheochromocytoma. About 10% of patients with familial MTC and MEN2A have de novo RET mutations, so that their children are at risk for thyroid cancer.

Imaging

Ultrasound is helpful for detecting nonpalpable thyroid nodules, differentiating solid from cystic nodules, and identifying adjacent lymphadenopathy. Ultrasound evaluation can identify features of a nodule that increase the a priori risk of malignancy, such as fine stippled calcification and enlarged regional nodes; however, a tissue diagnosis is strongly recommended before thyroidectomy.²⁰ Ultrasound also provides a noninvasive and inexpensive method of following the size of suspected benign nodules diagnosed by FNAB and for identifying enlarged lymph nodes. Ultrasound elastography is used to evaluate tissue stiffness noninvasively. This technique takes advantage of the fact that malignant nodules tend to be harder than benign nodules and thus deform less compared with the surrounding normal thyroid parenchyma.²¹ Larger studies are warranted before elastography can be routinely included in the evaluation of thyroid nodules. CT and MRI are unnecessary in the routine evaluation of thyroid tumors except for large, fixed, or substernal lesions. Scanning the thyroid with ¹²³I or ^99mTc is rarely necessary, and thyroid scanning currently is recommended in the assessment of thyroid nodules only in patients who have follicular thyroid nodules on FNAB and a suppressed TSH. PET scanning does not play a major role in the primary evaluation of thyroid nodules.

Management

Malignant tumors are treated by thyroidectomy, as discussed earlier and later in this chapter in Surgical Treatment under Malignant Thyroid Disease. Simple thyroid cysts resolve with aspiration in about 75% of cases, although some require a second or third aspiration. If the cyst persists after three attempts at aspiration, unilateral thyroid lobectomy is recommended. Lobectomy also is recommended for cysts >4 cm in diameter or complex cysts with solid and cystic components, as the latter have a higher incidence of malignancy (15%). When FNAB is used in complex nodules, the solid portion should be sampled. If a colloid nodule is diagnosed by FNAB, patients should still be observed with serial ultrasound and Tg measurements. If the nodule enlarges, repeat FNAB often is indicated. Although controversial, levothyroxine in doses sufficient to maintain a serum TSH level between 0.1 and 1.0 μU/mL may also be administered. In areas with a high prevalence of iodine deficiency, this can decrease nodule size and potentially prevent the growth of new nodules. In iodine-sufficient populations, the data are less impressive. Randomized controlled trial analyses have shown that less than 25% of benign nodules shrink more than 50% with TSH suppression in iodine-replete populations. Thyroidectomy should be performed if a nodule enlarges on TSH suppression or causes compressive symptoms, or for cosmetic reasons. An exception to this general rule is the patient who has had previous irradiation of the thyroid gland or has a family history of thyroid cancer. In these patients, total or near-total thyroidectomy is recommended because of the high incidence of thyroid cancer and decreased reliability of FNAB in this setting.

Malignant Thyroid Disease

In the United States, thyroid cancer accounts for <1% of all malignancies (2% of women and 0.5% of men) and is the most rapidly increasing cancer in women. Thyroid cancer is responsible for six deaths per million persons annually. Most patients present with a palpable swelling in the neck, which initiates assessment through a combination of history, physical examination, and FNAB.

Molecular Genetics of Thyroid Tumorigenesis

Several oncogenes and tumor suppressor genes are involved in thyroid tumorigenesis,²² as depicted in Table 38-5. The RET proto-oncogene (Fig. 38-15) plays a significant role in the pathogenesis of thyroid cancers. It is located on chromosome 10 and encodes a receptor tyrosine kinase, which binds several growth factors such as glial-derived neurotrophic factor and neurturin. The RET protein is expressed in tissues derived from the embryonic nervous and excretory systems. Therefore, RET disruption can lead to developmental abnormalities in organs derived from these systems, such as the enteric nervous system (Hirschsprung’s disease) and kidney. Germline mutations in the RET proto-oncogene are known to predispose to MEN2A, MEN2B, and familial MTCs, and somatic mutations have been demonstrated in tumors derived from the neural crest, such as MTCs (30%) and pheochromocytomas. The tyrosine kinase domain of RET can fuse with other genes by rearrangement. These fusion products also function as oncogenes and have been implicated in the pathogenesis of PTCs. At least 15 RET/PTC rearrangements have been described and appear to be early events in tumorigenesis. Young age and radiation exposure seem to be independent risk factors for the development of RET/PTC rearrangements. Up to 70% of papillary cancers in children exposed to the radiation fallout from the 1986 Chernobyl disaster carry RET/PTC rearrangements, the most common being RET/PTC1 and RET/PTC3. These rearrangements confer constitutive activation of the receptor tyrosine kinases. RET/PTC3 is associated with a solid type of PTC that appears to present at a higher stage and to be more aggressive. It has now been established that RET/PTC signaling involves the mitogen-activated protein kinase (MAPK) pathway via other signaling molecules such as Ras, Raf, and MEK. In normal cells, physiologic activation of Raf kinases occurs via direct interaction with guanosine triphosphate (GTP)–bound Ras, a membrane-bound small G protein. Activated Raf, a serine-threonine kinase, in turn phosphorylates MEK, another serine-threonine kinase. This leads to phosphorylation of ERK/MAPK, which phosphorylates regulatory molecules in the nucleus, thereby altering gene expression. Aberrant activation of the MAPK pathway leads to tumorigenesis. Aside from RET/PTC alterations, mutations in the Ras genes can also activate the MAPK pathway. Mutated RAS oncogenes have been identified in up to 20% to 40% of thyroid follicular adenomas and carcinomas, multinodular goiters, and papillary and anaplastic carcinomas. There are three Raf kinases, A-Raf, B-Raf (BRAF), and C-Raf. Mutations in BRAF also have been implicated in aberrant MAPK pathway activation and tumorigenesis. Of the various identified BRAF mutations, T1799A (V600E amino acid substitution) is the most common and occurs frequently in thyroid cancers. Interestingly, BRAF mutations occur in papillary and anaplastic tumors (average prevalence of 44% and 22%, respectively),¹⁵ but not in follicular thyroid cancers, suggesting a role in the pathogenesis of these malignancies. Studies also show that BRAF mutations are associated with more aggressive clinicopathologic features, including larger tumor size, invasion, and lymphadenopathy, and may have a role as prognostic markers.

The p53 gene is a tumor suppressor gene encoding a transcriptional regulator, which causes cell cycle arrest allowing repair of damaged DNA, thus helping to maintain genomic integrity. Mutations of p53 are rare in PTCs but common in undifferentiated thyroid cancers and thyroid cancer cell lines. Other cell cycle regulators and tumor suppressors such as p15 and p16 are mutated more commonly in thyroid cancer cell lines than in primary tumors. An oncogene resulting from the fusion of the DNA binding domain of the thyroid-transcription factor PAX8 gene to the peroxisome proliferator-activated receptor gamma 1 (PPARγ1) has been noted to play an important role in the development of follicular neoplasms, including follicular cancers. Thyroid cancer stem cells have also been identified; however, their role in thyroid carcinogenesis remains to be determined.²⁴

Specific Tumor Types

Papillary Carcinoma

Papillary carcinoma accounts for 80% of all thyroid malignancies in iodine-sufficient areas and is the predominant thyroid cancer in children and individuals exposed to external radiation. Papillary carcinoma occurs more often in women, with a 2:1 female-to-male ratio, and the mean age at presentation is 30 to 40 years. Most patients are euthyroid and present with a slow-growing painless mass in the neck. Dysphagia, dyspnea, and dysphonia usually are associated with locally advanced invasive disease. Lymph node metastases are common, especially in children and young adults, and may be the presenting complaint. “Lateral aberrant thyroid” almost always denotes a cervical lymph node that has been invaded by metastatic cancer. Suspicion of thyroid cancer often originates through physical examination of patients and a review of their history. Diagnosis is established by FNAB of the thyroid mass or lymph node. Once thyroid cancer is diagnosed on FNAB, a complete neck ultrasound is strongly recommended to evaluate the contralateral lobe and for lymph node metastases in the central and lateral neck compartments. Distant metastases are uncommon at initial presentation, but may ultimately develop in up to 20% of patients. The most common sites are lungs, followed by bone, liver, and brain.

Pathology

On gross examination, PTCs generally are hard and whitish and remain flat on sectioning with a blade, in contrast to normal tissue or benign nodular lesions that tend to bulge. Macroscopic calcification, necrosis, or cystic change may be apparent. Histologically, papillary carcinomas may exhibit papillary projections (Fig. 38-16A), a mixed pattern of papillary and follicular structures, or a pure follicular pattern (follicular variant). The diagnosis is established by characteristic nuclear cellular features. Cells are cuboidal with pale, abundant cytoplasm, crowded nuclei that may demonstrate “grooving,” and intranuclear cytoplasmic inclusions (leading to the designation of Orphan Annie nuclei [Fig. 38-16B]), which allow diagnosis by FNAB. Psammoma bodies, which are microscopic, calcified deposits representing clumps of sloughed cells, also may be present. Mixed papillary-follicular tumors and follicular variants of papillary carcinoma are classified as papillary carcinomas because they behave biologically as papillary carcinomas. Multifocality is common in papillary carcinoma and may be present in up to 85% of cases on microscopic examination. Multifocality is associated with an increased risk of cervical nodal metastases, and these tumors may rarely invade adjacent structures such as the trachea, esophagus, and RLNs. Other variants of papillary carcinoma include tall cell, insular, columnar, diffuse sclerosing, clear cell, trabecular, and poorly differentiated types. These variants account for about 1% of all papillary carcinomas and are generally associated with a worse prognosis.

Minimal or occult/microcarcinoma refers to tumors of 1 cm or less in size with no evidence of local invasiveness through the thyroid capsule or angioinvasion, and that are not associated with lymph node metastases. They are nonpalpable and usually are incidental findings at operative, histologic, or autopsy examination. Studies have demonstrated occult PTC to be present in 2% to 36% of thyroid glands removed at autopsy. These tumors are also being identified more frequently due to the widespread use of ultrasound. These occult tumors are generally associated with a better prognosis than larger tumors, but they may be more aggressive than previously appreciated. About 25% of patients with these tumors have associated occult lymph node metastases.

Prognostic Indicators

In general, patients with PTC have an excellent prognosis with a >95% 10-year survival rate. Several prognostic indicators have been incorporated into various staging systems, which enable patients to be stratified into low-risk and high-risk groups. Unfortunately, all of these classification systems rely on data that are not available preoperatively.

In 1987, Hay and colleagues²⁵ at the Mayo Clinic proposed the AGES scoring system, which incorporates Age, histologic Grade, Extrathyroidal invasion, and metastases and tumor Size to predict the risk of dying from papillary cancer. Low-risk patients are young, with well-differentiated tumors, no metastases, and small primary lesions, whereas high-risk patients are older, with poorly differentiated tumors, local invasion, distant metastases, and large primary lesions. The MACIS scale is a postoperative system modified from the AGES scale. This scale incorporates distant Metastases, Age at presentation (<40 or >40 years old), Completeness of original surgical resection, extrathyroidal Invasion, and Size of original lesion (in centimeters) and classifies patients into four risk groups based on their scores. Cady proposed the AMES system²⁶ to classify differentiated thyroid tumors into low- and high-risk groups using Age (men <40 years old, women <50 years old), Metastases, Extrathyroidal spread, and Size of tumors (< or >5 cm). Another classification system is the TNM system (Tumor, Nodal status, Metastases; Table 38-6), which used by most medical centers in North America.²⁷ A simplified system by DeGroot and associates²⁸ uses four groups—class I (intrathyroidal), class II (cervical nodal metastases), class III (extrathyroidal invasion), and class IV (distant metastases)—to determine prognosis.

Several molecular and genetic markers such as tumor DNA aneuploidy, decreased cyclic adenosine monophosphate response to TSH, increased epidermal growth factor binding, presence of N-ras and gsp mutations, overexpression of c-myc, and presence of p53 mutations also have been associated with a worse prognosis. The presence of BRAF V600E mutation, as previously mentioned, is associated with aggressive tumor characteristics, including extrathyroidal extension, older age at presentation, and lymph node and distant metastases. This mutation also appears to be an independent predictor of both tumor recurrence (even for early-stage disease) and tumor-related mortality. Some studies propose that BRAF mutation status on FNAB can be used to tailor initial management including more extensive initial surgical excision, high-dose postoperative RAI therapy, increased TSH suppression, and closer follow up.²⁹ The correlation of RET/PTC rearrangements and Ras mutations with prognosis is less clear.

Surgical Treatment

Most authors agree that patients with high-risk tumors (judged by any of the classification systems discussed earlier in Prognostic Indicators) or bilateral tumors should undergo total or near-total thyroidectomy. The optimal surgical strategy in the majority of patients with low-risk (small, unilateral) cancers was controversial for many years, with the focus of the debate centering around outcome data and risks associated with extent of thyroidectomy in this group of patients.

Proponents of total thyroidectomy indicate that it enables the use of RAI to effectively detect and treat residual thyroid tissue or metastatic disease and makes serum Tg level a more sensitive marker of recurrent or persistent disease. It is also known that a significant proportion (33% to 50%) of patients who develop a recurrence die from their disease,³⁰ and even though the data are retrospective, long-term, follow-up studies suggest that recurrence rates are lowered and that survival is improved in patients undergoing near-total or total thyroidectomy^{28,30,31,32,33,34} (Fig. 38-17). In addition, diminished survival is noted in patients with low-risk disease (mortality rates of 5% at 10 to 20 years), and it is not possible to accurately risk stratify patients preoperatively. More recently, a large study of >50,000 patients with papillary cancer demonstrated that, in multivariate analyses, total thyroidectomy led to a significantly improved recurrence and survival for tumors >1 cm. Furthermore, the authors also showed that patients with tumors 1 to 2 cm in diameter who were treated with lobectomy had a 24% higher risk of recurrence and a 49% higher risk of thyroid cancer mortality.³⁵ Based on the above, current guidelines for the evidence-based management of thyroid cancers recommend a near-total or total thyroidectomy for primary cancers >1 cm unless there are contraindications to the surgery.³⁶

When PTC is diagnosed by FNAB, the definitive operation can be done without confirming the diagnosis by frozen section during the operation. Patients with a nodule that is suspicious for papillary cancer should be treated by thyroid lobectomy, isthmusectomy, and removal of any pyramidal lobe or adjacent lymph nodes. If intraoperative frozen-section examination of a lymph node or primary tumor confirms carcinoma, completion total or near-total thyroidectomy is performed. If a definitive diagnosis cannot be made or the surgeon is concerned about the viability of the parathyroid glands or the status of the RLN, the operation is terminated. When final histology confirms carcinoma, completion thyroidectomy usually is performed. Thyroid lobectomy alone is considered sufficient treatment for small (<1 cm), incidentally discovered, low-risk, unifocal, intrathyroidal papillary carcinomas in the absence of prior head and neck irradiation or radiologically or clinically involved cervical nodal metastases.

During thyroidectomy, enlarged or obviously involved central neck nodes should be removed (therapeutic central-compartment, level VI), along with nodes with known lateral neck metastases. Some investigators recommend routine bilateral central neck dissection due to the high incidence of microscopic metastases and data showing improved rates of recurrence and survival (compared to historic controls). However, these risks need to be balanced with the increased risk of hypoparathyroidism with routine central neck dissection and the fact that some studies do not show any difference in recurrence rates or rates of low or undetectable Tg levels. The 2009 ATA guidelines for thyroid cancer management suggest that prophylactic (ipsilateral or bilateral) dissection may be performed in patients with advanced (T3 or T4) papillary thyroid carcinoma, whereas the procedure is not needed for small (T1 and T2) tumors that are clinically node negative.³⁶ Further prospective studies are needed before definitive recommendations can be made in this regard.

Biopsy-proven lymph node metastases detected clinically or by imaging in the lateral neck in patients with papillary carcinoma are managed with modified radical or functional neck dissection,³⁶ as described later in this chapter in Thyroid Surgery. Dissection of the posterior triangle and suprahyoid dissection usually are not necessary unless there is extensive metastatic disease in levels 2, 3, and 4, but should be performed when appropriate. Prophylactic lateral neck node dissection is not necessary in patients with PTC, because these cancers do not appear to metastasize systemically from lymph nodes, and micrometastases often can be ablated with RAI therapy.

Follicular Carcinoma

Follicular carcinomas account for 10% of thyroid cancers and occur more commonly in iodine-deficient areas. The overall incidence of this tumor is declining in the United States, probably due to iodine supplementation and improved histologic classification. Women have a higher incidence of follicular cancer, with a female-to-male ratio of 3:1, and a mean age at presentation of 50 years old. Follicular cancers usually present as solitary thyroid nodules, occasionally with a history of rapid size increase, and long-standing goiter. Pain is uncommon, unless hemorrhage into the nodule has occurred. Unlike papillary cancers, cervical lymphadenopathy is uncommon at initial presentation (about 5%), although distant metastases may be present. In <1% of cases, follicular cancers may be hyperfunctioning, leading patients to present with signs and symptoms of thyrotoxicosis. FNAB is unable to distinguish benign follicular lesions from follicular carcinomas. Therefore, preoperative clinical diagnosis of cancer is difficult unless distant metastases are present. Large follicular tumors (>4 cm) in older men are more likely to be malignant.

Due to the limitations inherent in the FNAB diagnosis, a number of studies have focused on identifying molecular markers to distinguish benign from malignant follicular lesions. Many of these genetic changes can be identified using tissue obtained during FNAB. While no single marker has met the ideal characteristics of being simple to use, reproducible, and cost-effective, several studies have explored combinations of markers that appear to be useful in differentiating benign from malignant lesions. Nikiforov and colleagues demonstrated the utility of examining a combination of DNA-based markers including BRAF, RAS, RET/PTC, and PAX/PPARγ abnormalities in the diagnosis of indeterminate nodules. In their study, the presence of any mutation in an FNAB sample was a strong indicator of cancer.^37,38 Expression arrays also have been used to investigate the role of microRNAs, which are a new class of small, noncoding RNAs that have been implicated in carcinogenesis. The specific microRNAs miR-197 and miR-346 are upregulated in follicular thyroid cancers³⁹ and have the potential to be used as diagnostic markers. Additional studies have also demonstrated the feasibility of studying a panel of microRNAs in a small number of FNA samples. Peripheral blood measurement of TSH-R mRNA levels was able to predict malignancy with a sensitivity of 90% and specificity of 80%; however, its widespread application requires further study.⁴⁰ A recently published prospective, multicenter study validated a novel diagnostic test that measures the expression of 167 genes (Veracyte). The gene expression classifier correctly classified 78 of 85 indeterminate nodules as suspicious with a sensitivity of 92%. Additional studies are needed to assess its utility in the broader clinical setting.⁴¹

Pathology

Follicular carcinomas usually are solitary lesions, and the majority are surrounded by a capsule. Histologically, follicles are present, but the lumen may be devoid of colloid. Architectural patterns depend on the degree of differentiation demonstrated by the tumor. Malignancy is defined by the presence of capsular and vascular invasion (Fig. 38-18). In general, minimally invasive tumors appear grossly encapsulated and have microscopic invasion through the tumor capsule without extension into the parenchyma and/or invasion into small- to medium-sized vessels (venous caliber) in or immediately outside the capsule, but not within the tumor. On the other hand, widely invasive tumors demonstrate evidence of large vessel invasion and/or broad areas of tumor invasion through the capsule. They may, in fact, be unencapsulated. It is important to note that there is a wide variation of opinion among clinicians and pathologists with respect to the above definitions. Tumor infiltration and invasion, as well as tumor thrombus within the middle thyroid or jugular veins, may be apparent at operation.

Surgical Treatment and Prognosis

Patients diagnosed by FNAB as having a follicular lesion should undergo thyroid lobectomy because at least 80% of these patients will have benign adenomas. Total thyroidectomy is recommended by some surgeons in older patients with follicular lesions >4 cm because of the higher risk of cancer in this setting (50%) and certainly should be performed in patients with atypia on FNA, a family history of thyroid cancer, or a history of radiation exposure. Intraoperative frozen-section examination usually is not helpful, but should be performed when there is evidence of capsular or vascular invasion or when adjacent lymphadenopathy is present. Total thyroidectomy should be performed when thyroid cancer is diagnosed. There is debate among experts about whether patients with minimally invasive follicular cancers should undergo completion thyroidectomy because the prognosis is so good in these patients. A diagnosis of frankly invasive carcinoma or follicular carcinoma with angioinvasion, with or without capsular invasion, necessitates completion of total thyroidectomy primarily so that ¹³¹I can be used to detect and ablate metastatic disease. Prophylactic nodal dissection is not needed because nodal involvement is infrequent, but in the unusual patient with nodal metastases, therapeutic neck dissection is recommended. Prophylactic central neck dissection may be considered in patients with large tumors. The cumulative mortality from follicular thyroid cancer is approximately 15% at 10 years and 30% at 20 years. Poor long-term prognosis is predicted by age over 50 years old at presentation, tumor size >4 cm, higher tumor grade, marked vascular invasion, extrathyroidal invasion, and distant metastases at the time of diagnosis.

Hürthle Cell Carcinoma

Hürthle cell carcinomas account for approximately 3% of all thyroid malignancies and, under the World Health Organization classification, are considered to be a subtype of follicular thyroid cancer. Hürthle cell cancers also are characterized by vascular or capsular invasion and, therefore, cannot be diagnosed by FNAB. Tumors contain sheets of eosinophilic cells packed with mitochondria, which are derived from the oxyphilic cells of the thyroid gland. Hürthle cell tumors differ from follicular carcinomas in that they are more often multifocal and bilateral (about 30%), usually do not take up RAI (about 5%), are more likely to metastasize to local nodes (25%) and distant sites, and are associated with a higher mortality rate (about 20% at 10 years). Hence, they are considered to be a separate class of tumors by some groups.

Management is similar to that of follicular neoplasms, with lobectomy and isthmusectomy being sufficient surgical treatment for unilateral Hürthle cell adenomas. When Hürthle cell neoplasms are found to be invasive on definitive paraffin-section histology, then total thyroidectomy should be performed. These patients should also undergo routine central neck node removal, similar to patients with MTC, and modified radical neck dissection when lateral neck nodes are palpable or identified by ultrasonography. Although RAI scanning and ablation usually are ineffective, they probably should be considered to ablate any residual normal thyroid tissue and occasionally ablate tumors because there is no other good therapy.

Postoperative Management of Differentiated Thyroid Cancer

Radioiodine Therapy

The issue of whether RAI therapy offers any benefit to patients with differentiated thyroid cancer remains controversial in the absence of prospective, randomized controlled trials. Long-term cohort studies by Mazzaferri and associates and DeGroot and colleagues demonstrate that postoperative RAI therapy reduces recurrence (Fig. 38-19) and provides a small improvement in survival, even in low-risk patients.^28,32 Screening with RAI is more sensitive than chest x-ray or CT scanning for detecting metastases; however, it is less sensitive than Tg measurements for detecting metastatic disease in most differentiated thyroid cancers except Hürthle cell tumors. Screening and treatment are facilitated by the removal of all normal thyroid tissue, which effectively competes for iodine uptake. Metastatic differentiated thyroid carcinoma can be detected and treated by ¹³¹I in about 75% of patients. Multiple studies show that RAI effectively treats >70% of lung micrometastases that are detected by RAI scan in the presence of a normal chest x-ray, whereas the success rates drop to <10% with pulmonary macrometastases. Early detection therefore appears to be very important to improve prognosis.

Several features place patients at increased risk for local recurrences or metastases. These high-risk features include certain histologic subtypes (such as tall cell, columnar, insular, solid variant and poorly differentiated thyroid cancer), presence of intrathyroidal vascular invasion, or gross or microscopic multifocal disease. However, there are currently inadequate data to determine whether RAI ablation has any benefits based on these features alone, independent of patient age, tumor size, and lymph node status. Thus, the current ATA guidelines strongly recommend RAI after total thyroidectomy only for patients with known distant metastases. In addition, RAI is also indicated for patients with gross extrathyroidal extension of tumor (regardless of size) or tumors >4 cm, even in the absence of other high-risk features. Ablation is also recommended, albeit based on expert opinion, for selected patients with 1- to 4-cm tumors with lymph node metastases or other high-risk features outlined above, when the combination of T and N status, age, and histology predicts an intermediate to high rate of tumor recurrence.³⁶ In contrast, RAI is not indicated for patients with unifocal cancers <1 cm in diameter or patients with multifocal tumors (all <1 cm) without additional high-risk features. This latter indication is a major change from the 2006 guidelines wherein RAI was recommended for all multifocal tumors.

Remnant ablation can be performed with either thyroid hormone withdrawal or recombinant TSH (rTSH) stimulation. This is based on randomized studies showing that both techniques are equally effective in preparing patients for ablation, with the latter being associated with an improved quality of life. If hormone withdrawal is used, T₄ therapy should be discontinued for approximately 6 weeks before scanning with ¹³¹I. Patients should receive T₃ during this time period to decrease the period of hypothyroidism. T₃ has a shorter half-life than T₄ (1 day vs. 1 week) and needs to be discontinued for 2 weeks to allow TSH levels to rise before treatment. Levels >30 mU/L are considered optimal, based on noncontrolled studies. A low-iodine diet also is recommended during this 2-week period. The usual protocol involved administering a screening dose of 1 to 3 mCi and measuring uptake 24 hours later. After a total thyroidectomy, this value should be <1%. A “hot” spot in the neck after initial screening usually represents residual normal tissue in the thyroid bed. Some investigators recommend omitting the scanning dose altogether to minimize thyrocyte “stunning” and subsequent requirement for higher treatment doses. Others recommend scanning only if the size of the remnant cannot be determined by the operative report or ultrasound, or if the results would alter the decision to treat or the dose to be administered. Current guidelines recommend using either ¹²³I or low-activity ¹³¹I (1- to 3-mCi dose) and delivering a therapeutic dose within 72 hours. Treatment doses of 30 to 100 mCi are recommended for ablation in low-risk patients as these doses show similar rates of ablation and recurrence rates. A 30-mCi dose is also effective at ablating the remnant with recombinant TSH. In higher risk patients (residual microscopic disease suspected or documented; tall cell, insular, or columnar cell histology), higher doses (100 to 200 mCi) are recommended. If patients have an elevated Tg level, but negative RAI scan, some physicians recommend treating once with 100 mCi of ¹³¹I and repeating the imaging 1 to 2 weeks later. Approximately one third of these patients demonstrate uptake on posttreatment imaging, and Tg levels usually decrease in these patients, documenting therapeutic benefit. The maximum dose of radioiodine that can be administered at one time without performing dosimetry is approximately 200 mCi with a cumulative dose of 1000 to 1500 mCi. Up to 500 mCi can be given with proper pretreatment dosimetry. Recent studies also show an increase in the number of second cancers in patients treated with RAI.⁴² The early and delayed complications of RAI therapy are listed in Table 38-7.

External-Beam Radiotherapy and Chemotherapy

External-beam radiotherapy is occasionally required to control unresectable, locally invasive or recurrent disease and to treat metastases in support bones to decrease the risk of fractures. It also is of value for the treatment and control of pain from bony metastases when there is minimal or no RAIU. Single-drug and multidrug chemotherapy has been used with little success in disseminated thyroid cancer, and there is no role for routine chemotherapy.³⁶ Doxorubicin (Adriamycin) and paclitaxel (Taxol) are the most frequently used agents. The former acts as a radiation sensitizer and should be considered in patients undergoing external-beam radiation.

Novel Therapies

These therapies are directed at the pathways known to be involved in various thyroid cancers.⁴³ Oncogenic kinase inhibitors that inhibit the mutant V600E BRAF kinase (PLX4032) selectively or both the wild-type and mutant kinase (XL281) have shown promise in the treatment of patients with metastatic papillary cancers. These drugs are associated with GI toxicity (nausea, vomiting, and diarrhea) and an increased risk of cutaneous squamous cell carcinomas. Other kinases targeting the VEGF and RET receptors have also shown promising results. However, the low rate of partial responses, lack of complete response, and ultimate resistance to therapy highlight the need for ongoing research in this area.

Thyroid Hormone

T₄ is necessary as replacement therapy in patients after total or near-total thyroidectomy, and also has the additional effect of suppressing TSH and reducing the growth stimulus for any possible residual thyroid cancer cells. TSH suppression reduces tumor recurrence rates. Current guidelines advise maintaining TSH levels <0.1 mU/mL in patients with persistent disease and maintaining TSH in the low normal range (0.3 to 2 mU/mL) in patients who are clinically and biochemically free of disease and at low risk for recurrence. For high-risk patients in the latter group, maintenance of TSH levels of 0.1 to 0.5 mU/mL is advised.³⁶ The risk of tumor recurrence must be balanced with the side effects associated with prolonged TSH suppression, including osteopenia and cardiac problems, particularly in older patients.

Follow-Up of Patients with Differentiated Thyroid Cancer

Thyroglobulin Measurement

Tg and anti-Tg antibody levels should be measured initially at 6-month intervals. In low-risk patients, who have low suppressed Tg levels in the first year, serum Tg should be measured after T₄ withdrawal or recombinant TSH stimulation approximately 12 months after ablation. Patients with undetectable stimulated Tg levels can be followed annually with clinical exam and Tg levels on T₄ replacement. A single rTSH-stimulated Tg level of <0.5 ng/mL (with absent antibodies) has a 98% to 99.5% probability of identifying patients completely free of disease on follow-up. A Tg level of >2 ng/mL following rTSH stimulation is highly sensitive in identifying patients with recurrent tumor.³⁶ Tg measurements in FNAB aspirates have also been shown to be useful in the detection of nodal metastatic disease.⁴⁴

Imaging

After the first posttreatment scan, low-risk patients with negative TSH-stimulated Tg and cervical ultrasound do not require routine diagnostic whole-body radioiodine scans. However, diagnostic whole-body scans 6 to 12 months after remnant ablation may be of value in the follow-up of patients with high or intermediate risk of persistent disease. It also is recommended that cervical ultrasound be performed to evaluate the thyroid bed and central and lateral cervical nodal compartments at 6 and 12 months after thyroidectomy and then annually for at least 3 to 5 years, depending on the patient’s risk for recurrent disease and Tg status. Sonographically suspicious nodes >5 to 8 mm on the smallest diameter measurement should be biopsied for cytology as well as Tg measurement in the aspirate washout. Smaller nodes can be followed and biopsied if there is continued growth.³⁶ FDG-PET and PET-CT scans have been shown to be useful to localize recurrent or persistent thyroid cancer in patients who have Tg-positive, RAI scan–negative disease. FDG-PET can also be useful for the initial staging of patients with poorly differentiated thyroid carcinomas or Hürthle cell tumors, particularly in patients with other evidence of disease on imaging or Tg levels. In addition, they may be used as a prognostic tool in patients with metastatic disease and to evaluate the response to treatment in patients with metastatic or locally advanced disease.

Medullary Carcinoma

MTC accounts for about 5% of thyroid malignancies and arises from the parafollicular or C cells of the thyroid, which, in turn, are derived from the ultimobranchial bodies. These cells are concentrated superolaterally in the thyroid lobes, and this is where MTC usually develops. C cells secrete calcitonin, a 32-amino-acid polypeptide that functions to lower serum calcium levels, although its effects in humans are minimal. Most MTCs occur sporadically. However, approximately 25% occur within the spectrum of several inherited syndromes such as familial MTC, MEN2A, and MEN2B. All these variants are known to result secondary to germline mutations in the RET proto-oncogene. The syndromes also are characterized by genotype-phenotype correlations, with specific mutations leading to particular clinical manifestations. The salient clinical and genetic features of these syndromes are outlined in Table 38-8. Some clinical features of MEN2B patients are shown in Fig. 38-20.

Patients with MTC often present with a neck mass that may be associated with palpable cervical lymphadenopathy (15% to 20%). Pain or aching is more common in patients with these tumors, and local invasion may produce symptoms of dysphagia, dyspnea, or dysphonia. Distant blood-borne metastases to the liver, bone (frequently osteoblastic), and lung occur later in the disease. The female-to-male ratio is 1.5:1. Most patients present between 50 and 60 years old, although patients with familial disease present at a younger age. Medullary thyroid tumors secrete not only calcitonin and carcinoembryonic antigen (CEA), but also other peptides such as calcitonin gene–related peptide, histaminadases, prostaglandins E₂ and F_2α, and serotonin. Patients with extensive metastatic disease frequently develop diarrhea, which may result from increased intestinal motility and impaired intestinal water and electrolyte flux. About 2% to 4% of patients develop Cushing’s syndrome as a result of ectopic production of adrenocorticotropic hormone (ACTH).

Pathology

MTCs typically are unilateral (80%) in patients with sporadic disease and multicentric in familial cases, with bilateral tumors occurring in up to 90% of familial patients. Familial cases also are associated with C-cell hyperplasia, which is considered a premalignant lesion. Microscopically, tumors are composed of sheets of infiltrating neoplastic cells separated by collagen and amyloid. Marked heterogeneity is present; cells may be polygonal or spindle shaped. The presence of amyloid is a diagnostic finding, but immunohistochemistry for calcitonin is more commonly used as a diagnostic tumor marker. These tumors also stain positively for CEA and calcitonin gene–related peptide.

Diagnosis

The diagnosis of MTC is established by history, physical examination, raised serum calcitonin, or CEA levels, and FNAB cytology of the thyroid mass. Attention to family history is important because about 25% of patients with MTC have familial disease. Because it is not possible to distinguish sporadic from familial disease at initial presentation, all new patients with MTC should be screened for RET point mutations, pheochromocytoma, and HPT. Screening of patients with familial MTC for RET point mutations has largely replaced using provocative testing with pentagastrin or calcium-stimulated calcitonin levels to make the diagnosis. Calcitonin and CEA are used to identify patients with persistent or recurrent MTC. Calcitonin is a more sensitive tumor marker, but CEA is a better predictor of prognosis.

Treatment

The ATA published guidelines for the management of medullary cancers in 2009.⁴⁵ A neck ultrasound is recommended to evaluate the central and lateral neck compartments and the superior mediastinum. Serum calcitonin, CEA, and calcium levels should also be measured, and RET proto-oncogene mutation testing should be performed. Pheochromocytomas need to be excluded. If patients are found to have a pheochromocytoma, this must be operated on first. These tumors are generally (>50%) bilateral. Total thyroidectomy is the treatment of choice for patients with MTC because of the high incidence of multicentricity, the more aggressive course, and the fact that ¹³¹I therapy usually is not effective.

Central compartment nodes frequently are involved early in the disease process, so that a bilateral prophylactic central neck node dissection should be routinely performed. In patients with palpable or imaging-detected cervical nodes or calcitonin levels >400 pg/mL, additional imaging to include a neck and chest CT and a triple-phase liver CT or contrast-enhanced MRI is recommended to assess for metastatic disease. In patients with no distant disease, an ipsilateral or bilateral lateral neck dissection (levels IIA, III, IV, and V) is advised. Less aggressive neck surgery should be considered to preserve speech and swallowing while maintaining locoregional control in patients with limited metastatic disease. The role of prophylactic lateral neck dissection is controversial. Some groups favor this procedure if central neck lymph nodes are involved or if the primary tumor is ≥1.5 cm.

In the case of locally recurrent or widely metastatic disease, tumor debulking is advised not only to ameliorate symptoms of pain, flushing, and diarrhea, but also to decrease risk of death from recurrent central neck or mediastinal disease. External-beam radiotherapy is controversial but can be considered for patients with resected T4 disease and for patients with unresectable residual or recurrent tumor and symptomatic bony metastases. Liver metastases tend to be multiple and are typically not amenable to resection, percutaneous ethanol ablation, or radiofrequency ablation. However, chemoembolization may be helpful in this setting. There is no effective chemotherapy regimen.

Various targeted therapies directed against the RET kinase have been investigated for the treatment of MTC.⁴³ Many of these also inhibit vascular endothelial growth factor receptor (VEGFR), due to their close structural similarities. Sorafenib, sunitinib, lenvatinib (E7080), and cabozantinib (XL-184) are some such multikinase inhibitors, whereas axitinib and pazopanib act only on VEGFR. Vandetanib inhibits both targets and is also an epidermal growth factor receptor (EGFR) inhibitor, and is the only drug approved by the U.S. Food and Drug Administration for the treatment of advanced and progressive MTC based on data that it prolongs progression-free survival, in addition to reducing secretion of calcitonin and CEA. An anti-CEA monoclonal antibody (labetuzumab) also has shown antitumor response in a small group of patients. Patients with recurrent/metastatic disease should be enrolled in well-designed clinical trials.

In patients who have hypercalcemia and an increased PTH at the time of thyroidectomy, only obviously enlarged parathyroid glands should be removed. The other parathyroid glands should be preserved and marked in patients with normocalcemia, as only about 20% of patients with MEN2A develop HPT. When a normal parathyroid cannot be maintained on a vascular pedicle, it should be removed, biopsied to confirm that it is a parathyroid, and then autotransplanted to the forearm of the nondominant arm, particularly in patients with MEN2A. Re-implantation into the sternocleidomastoid muscle is also acceptable for patients with known MEN2B and familial MTC.

Prophylactic total thyroidectomy is indicated in RET mutation carriers once the mutation is confirmed. The ATA stratifies mutation into various risk levels to offer recommendations regarding age at which a prophylactic thyroidectomy should be performed and to predict phenotypes, including pheochromocytomas. In general, in patients with less aggressive mutations, thyroidectomy may be delayed >5 years, especially if there is a normal annual serum calcitonin, neck ultrasound, less aggressive family history, or family preference. Children with mutations at codon 634 are advised to undergo thyroidectomy at <5 years of age, and those with MEN2B-related mutations should undergo the procedure before age 1. Central neck dissection can be avoided in children who are RET-positive and calcitonin-negative with a normal ultrasound examination. When the calcitonin is increased or the ultrasound suggests a thyroid cancer, a prophylactic central neck dissection is indicated.

Postoperative Follow-Up and Prognosis

Patients are followed by annual measurements of calcitonin and CEA levels, in addition to history and physical examination. Other modalities used to localize recurrent disease include ultrasound, CT, MRI, and more recently, FDG-PET/CT scans. Prognosis is related to disease stage. The 10-year survival rate is approximately 80% but decreases to 45% in patients with lymph node involvement. Survival also is significantly influenced by disease type. It is best in patients with non-MEN familial MTC, followed by those with MEN2A, and then those with sporadic disease. Prognosis is the worst (survival of 35% at 10 years) in patients with MEN2B. Performing prophylactic surgery in RET oncogene mutation carriers not only improves survival rates but also renders most patients calcitonin free.

Anaplastic Carcinoma

Anaplastic carcinoma accounts for approximately 1% of all thyroid malignancies in the United States. Women are more commonly affected, and the majority of tumors present in the seventh and eighth decade of life. The typical patient has a long-standing neck mass, which rapidly enlarges and may be painful. Associated symptoms such as dysphonia, dysphagia, and dyspnea are common. The tumor is large and may be fixed to surrounding structures or may be ulcerated with areas of necrosis (Fig. 38-21). Lymph nodes usually are palpable at presentation. Evidence of metastatic spread also may be present. Diagnosis is confirmed by FNAB revealing characteristic giant and multinucleated cells. Differential diagnoses on FNA can include lymphomas, medullary carcinomas, direct extension from a laryngeal carcinoma, or other metastatic carcinomas or melanoma. When spindle cell elements are present, primary and metastatic sarcomas need to be considered as well. Immunohistochemical markers can aid with excluding other diagnoses. Core or incisional biopsy occasionally is needed to confirm the diagnosis, especially when there is necrotic material on the FNA.

Pathology

On gross inspection, anaplastic tumors are firm and whitish in appearance. Microscopically, sheets of cells with marked heterogeneity are seen. The three main histologic growth patterns are spindle cell, squamoid, and pleomorphic giant cell. Tumors may show a predominance of one pattern or a mixture of various patterns. Foci of more differentiated thyroid tumors, either follicular or papillary, may be seen, suggesting that anaplastic tumors arise from more well-differentiated tumors.

Treatment and Prognosis

This tumor is one of the most aggressive thyroid malignancies, with few patients surviving 6 months beyond diagnosis. All forms of treatment have been disappointing. The ATA recently published guidelines for the management of patients with anaplastic cancer.⁴⁶ Imaging (ultrasound, CT, MRI, or PET-CT) should be obtained to assess resectability. All patients should have preoperative laryngoscopy to assess the status of the vocal cords. A total or near-total thyroidectomy with therapeutic lymph node dissection is recommended for patients with an intrathyroidal mass (although lobectomy may also be appropriate, particularly if there is concern for vocal cord paralysis). If extrathyroidal extension is present, an en bloc resection should be considered if all gross disease can be removed (R1). Tracheostomy should be avoided as long as possible unless there is impending airway loss. Adjuvant radiation with should be offered to patients with a good performance status and no metastatic disease who desire aggressive management. Cytotoxic chemotherapy (with some combination of a taxane, anthracycline, and platinum) is typically given concurrently and has been associated with prolonged survival, although these agents are also being used in a neoadjuvant fashion particularly in patients with unresectable disease.

Lymphoma

Lymphomas account for <1% of thyroid malignancies, and most are of the non-Hodgkin’s B-cell type. Although the disease can arise as part of a generalized lymphomatous condition, most thyroid lymphomas develop in patients with chronic lymphocytic thyroiditis. Chronic antigenic lymphocyte stimulation has been suggested to result in lymphocyte transformation. Patients usually present with symptoms similar to those of patients with anaplastic carcinoma, although the rapidly enlarging neck mass often is painless. Patients may present with acute respiratory distress. Ultrasound can be useful for early diagnosis, and lymphoma appears as a well-defined hypoechoic mass. The diagnosis usually is suggested by FNAB, but FNAB can be nondiagnostic, particularly in the setting of low-grade lymphomas. Therefore, needle core or open biopsy may be necessary for definitive diagnosis. Staging studies should be obtained expeditiously to assess the extent of extrathyroidal spread.³⁹

Treatment and Prognosis

Patients with thyroid lymphoma respond rapidly to chemotherapy (CHOP—cyclophosphamide, doxorubicin, vincristine, and prednisone), which also has been associated with improved survival. Combined treatment with radiotherapy and chemotherapy often is recommended. Thyroidectomy and nodal resection are used to alleviate symptoms of airway obstruction in patients who do not respond quickly to the above regimens or who have completed the regimen before diagnosis. Prognosis depends on the histologic grade of the tumor and whether the lymphoma is confined to the thyroid gland or is disseminated. The overall 5-year survival rate is about 50%; patients with extrathyroidal disease have markedly lower survival rates.

Metastatic Carcinoma

The thyroid gland is a rare site of metastases from other cancers, including kidney, breast, lung, and melanoma. Clinical examination and a review of the patient’s history often suggest the source of the metastatic disease, and FNAB usually provides definitive diagnosis. Resection of the thyroid, usually lobectomy, may be helpful in many patients, depending on the status of their primary tumor.

Thyroid Surgery

Conduct of Thyroidectomy

Patients with any recent or remote history of altered phonation or prior neck surgery should undergo vocal cord assessment by direct or indirect laryngoscopy before thyroidectomy. The patient is positioned supine, with a sandbag between the scapulae. The head is placed on a donut cushion, and the neck is extended to provide maximal exposure. A Kocher transverse collar incision, typically 4 to 5 cm in length, is placed in or parallel to a natural skin crease 1 cm below the cricoid cartilage (Fig. 38-22A), although longer incisions may be needed. The subcutaneous tissues and platysma are incised sharply, and subplatysmal flaps are raised superiorly to the level of the thyroid cartilage and inferiorly to the suprasternal notch (Fig. 38-22B). The strap muscles are divided in the midline along the entire length of the mobilized flaps, and the thyroid gland is exposed. On the side to be approached first, the sternohyoid muscles are separated from the underlying sternothyroid muscle by blunt dissection until the internal jugular vein and ansa cervicalis nerve are identified. The strap muscles rarely need to be divided to gain exposure to the thyroid gland. If this maneuver is necessary, the muscles should be divided high to preserve their innervation by branches of the ansa cervicalis. If there is evidence of direct tumor invasion into the strap muscles, the portion of involved muscle should be resected en bloc with the thyroid gland. The sternothyroid muscle is then dissected off the underlying thyroid by a combination of sharp and blunt dissection, thus exposing the middle thyroid veins. The thyroid lobe is retracted medially and anteriorly, and the lateral tissues are swept posterolaterally using a peanut sponge. The middle thyroid veins are ligated and divided (Fig. 38-22C). Attention is then turned to the midline where Delphian nodes and the pyramidal lobe are identified. The fascia just cephalad and caudad to the isthmus is divided. The superior thyroid pole is identified by retracting the thyroid first inferiorly and medially, and then the upper pole of the thyroid is mobilized caudally and laterally. The dissection plane is kept as close to the thyroid as possible, and the superior pole vessels are individually identified, skeletonized, ligated, and divided low on the thyroid gland to avoid injury to the external branch of the superior laryngeal nerve (Fig. 38-22D). Once these vessels are divided, the tissues posterior and lateral to the superior pole can be swept from the gland in a posteromedial direction, to reduce the risk of damaging vessels supplying the upper parathyroid.

The RLNs should then be identified. The course of the right RLN is more oblique than the left RLN. The nerves can be most consistently identified at the level of the cricoid cartilage. The parathyroids usually can be identified within 1 cm of the crossing of the inferior thyroid artery and the RLN, although they also may be ectopic in location. The lower pole of the thyroid gland should be mobilized by gently sweeping all tissues dorsally. The inferior thyroid vessels are dissected, skeletonized, ligated, and divided as close to the surface of the thyroid gland as possible to minimize devascularization of the parathyroids (extracapsular dissection) or injury to the RLN. The RLN is most vulnerable to injury in the vicinity of the ligament of Berry. The nerve often passes through this structure along with small crossing arterial and venous branches (Fig. 38-22E). Any bleeding in this area should be controlled with gentle pressure before carefully identifying the vessel and ligating it. Use of the electrocautery should be avoided in proximity to the RLN. Once the ligament is divided, the thyroid can be separated from the underlying trachea by sharp dissection. The pyramidal lobe, if present, must be dissected in a cephalad direction to above the level of the notch in the thyroid cartilage or higher in continuity with the thyroid gland. If a lobectomy is to be performed, the isthmus is divided flush with the trachea on the contralateral side and suture ligated. The procedure is repeated on the opposite side for a total thyroidectomy.

Parathyroid glands that are located anteriorly on the surface of the thyroid cannot be dissected from the thyroid with a good blood supply or that have been inadvertently removed during the thyroidectomy should be resected, confirmed as parathyroid tissue by frozen section, divided into 1-mm fragments, and reimplanted into individual pockets in the sternocleidomastoid muscle. The sites should be marked with silk sutures and a clip. If a subtotal thyroidectomy is to be performed, once the superior pole vessels are divided and the thyroid lobe mobilized anteriorly, the thyroid lobe is cross-clamped with a Mayo clamp, leaving approximately 4 g of the posterior portion of the thyroid. The thyroid remnant is suture ligated, taking care to avoid injury to the RLN. Routine drain placement rarely is necessary. After adequate hemostasis is obtained, the strap muscles are reapproximated in the midline. The platysma is approximated in a similar fashion. The skin can be closed with subcuticular sutures or clips.

Minimally Invasive Approaches

Several approaches to minimally invasive thyroidectomy have been described. Mini-incision procedures use a small, 3-cm incision with no flap creation and minimal dissection to deliver the thyroid into the wound and then perform the pretracheal and paratracheal dissection. Video assistance can be used to improve the visualization via the small incision. Totally endoscopic approaches also have been described, via the supraclavicular, anterior chest, axillary, and breast approach. The axillary, anterior chest, and breast approaches eliminate the skin incision in the neck but are more invasive. The endoscopic approaches can also be performed with the assistance of robotic techniques. More recently, there have been reports of transoral robotic-assisted thyroidectomy in which the thyroid is approached through the oral cavity. These methods are feasible, but clear benefits over the “traditional” open approach via small neck incisions have not been established.⁴⁸

Typically, endoscopic thyroidectomies are performed under general anesthesia. For the axillary approach, a 30-mm skin incision is made in the axilla, and 12-mm and 5-mm trocars are inserted through this incision (Fig. 38-22F). An additional 5-mm trocar is inserted adjacent to the incision. For the anterior chest approach, a 12-mm skin incision is made in the skin of the anterior chest approximately 3 to 5 cm below the border of the ipsilateral clavicle. Two additional 5-mm trocars are inserted by endoscopic guidance below the ipsilateral clavicle, and carbon dioxide (CO₂) is then insufflated up to a pressure of 4 mmHg to facilitate creation of a working space. The anterior border of the sternocleidomastoid muscle is then separated from the sternohyoid muscle to expose the sternothyroid muscle. The thyroid gland is exposed by splitting the sternothyroid muscle. The lower pole is retracted upward and dissected from the adipose tissue to identify the RLN. As the RLN is exposed, Berry’s ligament is exposed and incised with a 5-mm clip or laparoscopic coagulating shears. The upper pole of the thyroid gland is separated from the cricothyroid muscle, and the external branch of the superior laryngeal nerve can be identified during this maneuver. The upper pole of the thyroid gland then is dissected free.

Surgical Removal of Intrathoracic Goiter

A goiter is considered mediastinal if at least 50% of the thyroid tissue is located intrathoracically. Mediastinal goiters can be primary or secondary. Primary mediastinal goiters constitute approximately 1% of all mediastinal goiters and arise from accessory (ectopic) thyroid tissue located in the chest. These goiters are supplied by intrathoracic blood vessels and do not have any connection to thyroid tissue in the neck. The vast majority of mediastinal goiters are, however, secondary mediastinal goiters that arise from downward extension of cervical thyroid tissue along the fascial planes of the neck and derive their blood supply from the superior and inferior thyroid arteries. Virtually all intrathoracic goiters can be removed via a cervical incision. Patients who have (a) invasive thyroid cancers, (b) had previous thyroid operations and may have developed parasitic mediastinal vessels, or (c) primary mediastinal goiters with no thyroid tissue in the neck may require a median sternotomy for removal.⁴⁹ The chest, however, should be prepared in most cases in the event it is necessary to perform a median sternotomy to control mediastinal bleeding or completely remove an unsuspected invasive cancer. The goiter is approached via a neck incision. The superior pole vessels and the middle thyroid veins are identified and ligated first. Early division of the isthmus helps with subsequent mobilization of the substernal goiter from beneath the sternum. Placement of large 1-0 or 2-0 sutures deep into the goiter, when necessary, helps deliver it. For patients in whom thyroid cancer is suspected or demonstrated in an intrathoracic gland, attempts should be made to avoid rupture of the thyroid capsule. When sternotomy is indicated, the sternum usually should be divided to the level of the third intercostal space and then laterally on one side at the space between the third and fourth ribs (Fig. 38-23).

Central and Lateral Neck Dissection for Nodal Metastases

Central compartment (medial to the carotid sheath) lymph nodes frequently are involved in patients with papillary, medullary, and Hürthle cell carcinomas and should be removed at the time of thyroidectomy, preserving the RLNs and parathyroid glands. Central neck dissection is particularly important in patients with medullary and Hürthle cell carcinoma because of the high frequency of microscopic tumor spread and because these tumors cannot be ablated with ¹³¹I. An ipsilateral modified radical neck dissection is indicated in the presence of palpable cervical lymph nodes or prophylactically in some patients with medullary carcinoma.

A modified radical (functional) neck dissection can be performed via the cervical incision used for thyroidectomy, which can be extended laterally (Fig. 38-24A) to the anterior margin of the trapezius muscle. The procedure involves removal of all fibro-fatty tissue along the internal jugular vein (levels II, III, and IV) and the posterior triangle (level V). In contrast to a radical neck dissection, the internal jugular vein, the spinal accessory nerve, the cervical sensory nerves, and the sternocleidomastoid muscle are preserved unless they are adherent to or invaded by tumor. The procedure begins by opening the plane between the strap muscles medially and the sternocleidomastoid muscle laterally. The anterior belly of the omohyoid muscle is retracted laterally, and the dissection is carried posteriorly until the carotid sheath is reached. The internal jugular vein is retracted medially with a vein retractor and the fibro-fatty tissue and lymph nodes are dissected away from it by a combination of sharp and blunt dissection. The lateral dissection is carried along the posterior border of the sternocleidomastoid muscle, removing the tissue from the posterior triangle. The deep dissection plane is the anterior scalenus muscle, the phrenic nerve, the brachial plexus, and the medial scalenus muscle. The phrenic nerve is preserved on the scalenus anterior muscle, as are the cervical sensory nerves in most patients (Fig. 38-24B). Dissection along the spinal accessory nerve superiorly is most important because this is a frequent site of metastatic disease.

Complications of Thyroid Surgery

Nerves, parathyroids, and surrounding structures are all at risk of injury during thyroidectomy. Injury to the RLN may occur by severance, ligation, or traction, but should occur in <1% of patients undergoing thyroidectomy by experienced surgeons. The RLN is most vulnerable to injury during the last 2 to 3 cm of its course, but also can be damaged if the surgeon is not alert to the possibility of nerve branches and the presence of a nonrecurrent nerve, particularly on the right side. If the injury is recognized intraoperatively, most surgeons advocate primary reapproximation of the perineurium using nonabsorbable sutures. Approximately 20% of patients are at risk of injury to the external branches of the superior laryngeal nerve, especially if superior pole vessels are ligated en masse. The cervical sympathetic trunk is at risk of injury in invasive thyroid cancers and retroesophageal goiters and may result in Horner’s syndrome. Transient hypocalcemia (from surgical injury or inadvertent removal of parathyroid tissue) has been reported in up to 50% of cases, but permanent hypoparathyroidism occurs <2% of the time. Postoperative hypocalcemia is more likely in patients who undergo concomitant thyroidectomy and central and lateral neck dissection and in patients with Graves’ disease. Postoperative hematomas or bleeding may also complicate thyroidectomies and rarely necessitate emergency reoperation to evacuate the hematoma. Bilateral vocal cord dysfunction with airway compromises requires immediate reintubation and tracheostomy. Seromas may need aspiration to relieve patient discomfort. Wound cellulitis and infection and injury to surrounding structures, such as the carotid artery, jugular vein, and esophagus, are infrequent.

Nerve Monitoring

Intraoperative RLN and external laryngeal nerve monitoring techniques are being increasingly used during thyroid and parathyroid surgery. Both continuous monitoring using endotracheal tube electrodes and intermittent monitoring by periodic stimulation and laryngeal palpation are used. Many published studies have established the feasibility of nerve monitoring; however, none were able to shown that the technique equivocally reduces nerve injury (particularly by experienced surgeons and especially when the RLN is routinely visually identified) until recently. In 2009, Barczynski and colleagues⁵⁰ were the first to demonstrate, in their prospective randomized trial with 2000 nerves at risk that neuromonitoring was associated with a statistically significant improvement in transient RLN injury rates when compared to the practice of visualization of the nerve alone and particularly in patients at higher risk of nerve injury. Of note, there was no difference in permanent RLN injury rates, and a later meta-analysis of all the published studies also failed to show a protective effect of neuromonitoring. Despite this, the technology has become widely adopted. As additional studies have documented that direct RLN stimulation has a low sensitivity of predicting vocal cord paralysis, when compared to vagal stimulation, attempts have been made to identify critical steps in the process. Currently, adequate RLN monitoring includes (a) stimulation of the vagus nerve prior to any dissection in the tracheoesophageal groove; (b) stimulation of the RLN prior to resecting the thyroid; (c) stimulation of the most proximally exposed portion of the RLN after the ligament of Berry has been completely divided; and (d) stimulation of the vagus nerve after complete hemostasis has been achieved on that side of the neck prior to closing or going on to the opposite lobe. Recently, an international group has proposed guidelines for RLN monitoring in thyroid and parathyroid surgery, and pre- and postoperative laryngoscopy is recommended, in addition to vagal stimulation. Not all experts agree with this recommendation.

Historical Background

In 1849, the curator of the London Zoological Gardens, Sir Richard Owen, provided the first accurate description of the normal parathyroid gland after autopsy examination of an Indian rhinoceros. However, human parathyroids were not grossly and microscopically described until 1879 by Ivar Sandström, a medical student in Uppsala, Sweden. He suggested that these glands be named the glandulae parathyroideae, although their function was not known.

The association of HPT and the bone disease osteitis fibrosa cystica (described by von Recklinghausen) was recognized in 1903. Calcium measurement became possible in 1909, and the association between serum calcium levels and the parathyroid glands was established. The first successful parathyroidectomy was performed in 1925 by Felix Mandl on a 38-year-old man who had severe bone pain secondary to advanced osteitis fibrosa cystica. The patient’s condition dramatically improved after the operation, and he lived for another 7 years before dying of recurrent HPT or renal failure. In 1926, the first parathyroid operation was performed at Massachusetts General Hospital. Edward Churchill, assisted by an intern named Oliver Cope, operated on the famous sea captain Charles Martell for severe primary HPT (PHPT). It was not until his seventh operation, which included total thyroidectomy, that an ectopic adenoma was found substernally. Unfortunately, Captain Martell died 6 weeks later, likely due to laryngeal spasm and complications of renal stones and ureteral obstruction. The first successful parathyroidectomy for HPT in the United States was performed on a 56-year-old woman in 1928 by Isaac Y. Olch at the Barnes Hospital in St. Louis, Missouri. At operation, a parathyroid adenoma was found attached to the left lower lobe of the thyroid gland. Postoperatively, the patient developed tetany, requiring lifelong supplemental calcium.

Embryology

In humans, the superior parathyroid glands are derived from the fourth branchial pouch, which also gives rise to the thyroid gland. The third branchial pouches give rise to the inferior parathyroid glands and the thymus (Fig. 38-25). The parathyroids remain closely associated with their respective branchial pouch derivatives. The position of normal superior parathyroid glands is more consistent, with 80% of these glands being found near the posterior aspect of the upper and middle thyroid lobes, at the level of the cricoid cartilage.⁵¹ Approximately 1% of normal upper glands may be found in the paraesophageal or retroesophageal space. Enlarged superior glands may descend in the tracheoesophageal groove and come to lie caudal to the inferior glands. Truly ectopic superior parathyroid glands are rare, but may be found in the middle or posterior mediastinum or in the aortopulmonary window.⁵¹ As the embryo matures, the thymus and inferior parathyroids migrate together caudally in the neck. The most common location for inferior glands is within a distance of 1 cm from a point centered where the inferior thyroid artery and RLN cross. Approximately 15% of inferior glands are found in the thymus. The position of the inferior glands, however, tends to be more variable due to their longer migratory path. Undescended inferior glands may be found near the skull base, angle of the mandible, or superior to the upper parathyroid glands along with an undescended thymus. The frequency of intrathyroidal glands is about 2%.

Anatomy and Histology

Most patients have four parathyroid glands. The superior glands usually are dorsal to the RLN at the level of the cricoid cartilage, whereas the inferior parathyroid glands are located ventral to the nerve. Normal parathyroid glands are gray and semitransparent in newborns but appear golden yellow to light brown in adults. Parathyroid color depends on cellularity, fat content, and vascularity. Moreover, they often are embedded in and sometimes difficult to discern from surrounding fat. Normal parathyroid glands are located in loose tissue or fat and are ovoid. They measure up to 7 mm in size and weigh approximately 40 to 50 mg each. Parathyroid glands usually derive their blood supply from branches of the inferior thyroid artery, although branches from the superior thyroid artery supply at least 20% of upper glands. Branches from the thyroidea ima, and vessels to the trachea, esophagus, larynx, and mediastinum may also be found. The parathyroid glands drain ipsilaterally by the superior, middle, and inferior thyroid veins.

Akerström and colleagues,⁵² in an autopsy series of 503 cadavers, found four parathyroid glands in 84% of cases. Supernumerary glands were present in 13% of patients, most commonly in the thymus. Only 3% of patients had less than four glands. Similar results were obtained in other dissection studies of 428 human subjects by Gilmour who reported a 6.7% incidence of supernumerary glands.⁵³

Histologically, parathyroid glands are composed of chief cells and oxyphil cells arranged in trabeculae, within a stroma composed primarily of adipose cells (Fig. 38-26). The parathyroid glands of infants and children are composed mainly of chief cells, which produce parathyroid hormone (PTH). Acidophilic, mitochondria-rich oxyphil cells are derived from chief cells, can be seen around puberty, and increase in numbers in adulthood. A third group of cells, known as water-clear cells, also are derived from chief cells, are present in small numbers, and are rich in glycogen. Although most oxyphil and water-clear cells retain the ability to secrete PTH, their functional significance is not known.

Parathyroid Physiology and Calcium Homeostasis

Calcium is the most abundant cation in human beings and has several crucial functions. Extracellular calcium levels are 10,000-fold higher than intracellular levels, and both are tightly controlled. Extracellular calcium is important for excitation-contraction coupling in muscle tissues, synaptic transmission in the nervous system, coagulation, and secretion of other hormones. Intracellular calcium is an important second messenger regulating cell division, motility, membrane trafficking, and secretion. Calcium is absorbed from the small intestine in its inorganic form. Calcium fluxes in the steady state are depicted in Fig. 38-27.

Extracellular calcium (900 mg) accounts for only 1% of the body’s calcium stores, the majority of which is sequestered in the skeletal system. Approximately 50% of the serum calcium is in the ionized form, which is the active component. The remainder is bound to albumin (40%) and organic anions such as phosphate and citrate (10%). The total serum calcium levels range from 8.5 to 10.5 mg/dL (2.1 to 2.6 mmol/L), and ionized calcium levels range from 4.4 to 5.2 mg/dL (1.1 to 1.3 mmol/L). Both concentrations are tightly regulated. The total serum calcium level must always be considered in its relationship to plasma protein levels, especially serum albumin. For each gram per deciliter of alteration of serum albumin above or below 4.0 mg/dL, there is a 0.8 mg/dL increase or decrease in protein-bound calcium and, thus, in total serum calcium levels. Total and, particularly, ionized calcium levels are influenced by various hormone systems.

Parathyroid Hormone

The parathyroid cells rely on a G-protein–coupled membrane receptor, designated the calcium-sensing receptor (CASR), to regulate PTH secretion by sensing extracellular calcium levels⁵⁴ (Fig. 38-28). PTH secretion also is stimulated by low levels of 1,25-dihydroxy vitamin D, catecholamines, and hypomagnesemia. The PTH gene is located on chromosome 11. PTH is synthesized in the parathyroid gland as a precursor hormone preproPTH, which is cleaved first to pro-PTH and then to the final 84-amino-acid PTH. Secreted PTH has a half-life of 2 to 4 minutes. In the liver, PTH is metabolized into the active N-terminal component and the relatively inactive C-terminal fraction. The C-terminal component is excreted by the kidneys and accumulates in chronic renal failure.

PTH functions to regulate calcium levels via its actions on three target organs, the bone, kidney, and gut. PTH increases the resorption of bone by stimulating osteoclasts and promotes the release of calcium and phosphate into the circulation. At the kidney, calcium is primarily absorbed in concert with sodium in the proximal convoluted tubule, but fine adjustments occur more distally. PTH acts to limit calcium excretion at the distal convoluted tubule via an active transport mechanism. PTH also inhibits phosphate reabsorption (at the proximal convoluted tubule) and bicarbonate reabsorption. It also inhibits the Na⁺/H⁺ antiporter, which results in a mild metabolic acidosis in hyperparathyroid states. PTH and hypophosphatemia also enhance 1-hydroxylation of 25-hydroxyvitamin D, which is responsible for its indirect effect of increasing intestinal calcium absorption.

Calcitonin

Calcitonin is produced by thyroid C cells and functions as an antihypercalcemic hormone by inhibiting osteoclast-mediated bone resorption. Calcitonin production is stimulated by calcium and pentagastrin and also by catecholamines, cholecystokinin, and glucagon. When administered intravenously to experimental animals, it produces hypocalcemia. At the kidney, calcitonin increases phosphate excretion by inhibiting its reabsorption. Calcitonin plays a minimal, if any, role in the regulation of calcium levels in humans. However, it is very useful as a marker of MTC and in treating acute hypercalcemic crisis.

Vitamin D

Vitamin D refers to vitamin D₂ and vitamin D₃, both of which are produced by photolysis of naturally occurring sterol precursors. Vitamin D₂ is available commercially in pharmaceutical preparations, whereas vitamin D₃ is the most important physiologic compound and is produced from 7-dehydrocholesterol, which is found in the skin. Vitamin D is metabolized in the liver to its primary circulating form, 25-hydroxyvitamin D. Further hydroxylation in the kidney results in 1,25-dihydroxy vitamin D, which is the most metabolically active form of vitamin D. Vitamin D stimulates the absorption of calcium and phosphate from the gut and the resorption of calcium from the bone.

Hyperparathyroidism

Hyperfunction of the parathyroid glands may be classified as primary, secondary, or tertiary. PHPT arises from increased PTH production from abnormal parathyroid glands and results from a disturbance of normal feedback control exerted by serum calcium. Elevated PTH levels may also occur as a compensatory response to hypocalcemic states resulting from chronic renal failure or GI malabsorption of calcium. This secondary HPT can be reversed by correction of the underlying problem (e.g., kidney transplantation for chronic renal failure). However, chronically stimulated glands may occasionally become autonomous, resulting in persistence or recurrence of hypercalcemia after successful renal transplantation, resulting in tertiary HPT.

Primary Hyperparathyroidism

PHPT is a common disorder, affecting 100,000 individuals annually in the United States. PHPT occurs in 0.1% to 0.3% of the general population and is more common in women (1:500) than in men (1:2000). Increased PTH production leads to hypercalcemia via increased GI absorption of calcium, increased production of vitamin D₃, and reduced renal calcium clearance. PHPT is characterized by increased parathyroid cell proliferation and PTH secretion that is independent of calcium levels.

Etiology

The exact cause of PHPT is unknown, although exposure to low-dose therapeutic ionizing radiation and familial predisposition account for some cases. Various diets and intermittent exposure to sunshine may also be related. Other causes include renal leak of calcium and declining renal function with age as well as alteration in the sensitivity of parathyroid glands to suppression by calcium. The latency period for development of PHPT after radiation exposure is longer than that for the development of thyroid tumors, with most cases occurring 30 to 40 years after exposure. Patients who have been exposed to radiation have similar clinical presentations and calcium levels when compared to patients without a history of radiation exposure. However, the former tend to have higher PTH levels and a higher incidence of concomitant thyroid neoplasms. Lithium therapy has been known to shift the set point for PTH secretion in parathyroid cells, thereby resulting in elevated PTH levels and mild hypercalcemia. Lithium stimulates the growth of abnormal parathyroid glands in vitro and also in susceptible patients in vivo.⁵⁵ PHPT results from the enlargement of a single gland or parathyroid adenoma in approximately 80% of cases, multiple adenomas or hyperplasia in 15% to 20% of patients, and parathyroid carcinoma in 1% of patients. Existence of two enlarged glands or double adenomas is supported by biochemical (calcium and PTH), intraoperative PTH (IOPTH), molecular, and histologic data. This entity is less common in younger patients but accounts for up to 10% of older patients with PHPT. It should be emphasized that when more than one abnormal parathyroid gland is identified preoperatively or intraoperatively, the patient has hyperplasia (all glands abnormal) until proven otherwise.

Genetics

Most cases of PHPT are sporadic. However, PHPT also occurs within the spectrum of a number of inherited disorders such as MEN1, MEN2A, isolated familial HPT, and familial HPT with jaw-tumor syndrome. All of these syndromes are inherited in an autosomal dominant fashion. PHPT is the earliest and most common manifestation of MEN1⁵⁶ and develops in 80% to 100% of patients by age 40 years old. These patients also are prone to pancreatic neuroendocrine tumors and pituitary adenomas and, less commonly, to adrenocortical tumors, lipomas, skin angiomas, and carcinoid tumors of the bronchus, thymus, or stomach. About 50% of patients develop gastrinomas, which often are multiple and metastatic at diagnosis. Insulinomas develop in 10% to 15% of cases, whereas many patients have nonfunctional pancreatic endocrine tumors. Prolactinomas occur in 10% to 50% of MEN1 patients and constitute the most common pituitary lesion. MEN1 has been shown to result from germline mutations in the MEN1 gene, a tumor suppressor gene located on chromosome 11q12-13 that encodes menin, a protein that is postulated to interact with the transcription factors JunD and nuclear factor-κB in the nucleus, in addition to replication protein A and other proteins.⁵⁷ Most MEN1 mutations result in a nonfunctional protein and are scattered throughout the translated nine exons of the gene. This makes presymptomatic screening for mutation carriers difficult. MEN1 mutations also have been found in kindreds initially suspected to represent isolated familial HPT. HPT develops in about 20% of patients with MEN2A and generally is less severe. MEN2A is caused by germline mutations of the RET proto-oncogene located on chromosome 10. In contrast to MEN1, genotype-phenotype correlations have been noted in this syndrome in that individuals with mutations at codon 634 are more likely to develop HPT. Patients with the familial HPT with jaw-tumor syndrome have an increased predisposition to parathyroid carcinoma. This syndrome maps to a tumor suppressor locus HRPT2 (CDC73 or parafibromin) on chromosome 1. Patients belonging to isolated HPT kindreds also appear to demonstrate linkage to HRPT2.

Approximately 25% to 40% of sporadic parathyroid adenomas and some hyperplastic parathyroid glands have loss of heterozygosity (LOH) at 11q13, the site of the MEN1 gene. The parathyroid adenoma 1 oncogene (PRAD1), which encodes cyclin D1, a cell cycle control protein, is overexpressed in about 18% of parathyroid adenomas. This was demonstrated to result from a rearrangement on chromosome 11 that places the PRAD1 gene under the control of the PTH promoter. Other chromosomal regions deleted in parathyroid adenomas and possibly reflecting loss of tumor suppressor genes include 1p, 6q, and 15q, whereas amplified regions suggesting oncogenes have been identified at 16p and 19p. RET mutations are rare in sporadic parathyroid tumors. Sporadic parathyroid cancers are characterized by uniform loss of the tumor suppressor gene RB, which is involved in cell cycle regulation, and 60% have HRPT2 (CDC73) mutations. These alterations are rare in benign parathyroid tumors and may have implications for diagnosis. The p53 tumor suppressor gene is also inactivated in a subset (30%) of parathyroid carcinomas.

Clinical Manifestations

Patients with PHPT formerly presented with the “classic” pentad of symptoms (i.e., kidney stones, painful bones, abdominal groans, psychic moans, and fatigue overtones). With the advent and widespread use of automated blood analyzers in the early 1970s, there has been an alteration in the “typical” patient with PHPT. They are more likely to be minimally symptomatic or asymptomatic. Currently, most patients present with weakness, fatigue, polydipsia, polyuria, nocturia, bone and joint pain, constipation, decreased appetite, nausea, heartburn, pruritus, depression, and memory loss. Patients with PHPT also tend to score lower than healthy controls when assessed by general multidimensional health assessment tools such as the Medical Outcomes Study Short-Form Health Survey (SF-36) and other specific questionnaires. Furthermore, these symptoms and signs improve in most, but certainly not all, patients after parathyroidectomy. Truly “asymptomatic” PHPT appears to be rare, occurring in <5% of patients, as determined by prospectively administered questionnaires. Complications of PHPT are described below.

Renal Disease

Approximately 80% of patients with PHPT have some degree of renal dysfunction or symptoms. Kidney stones were previously reported in up to 80% of patients but now occur in about 20% to 25%. The calculi are typically composed of calcium phosphate or oxalate. In contrast, PHPT is found to be the underlying disorder in only 3% of patients presenting with nephrolithiasis. Nephrocalcinosis, which refers to renal parenchymal calcification, is found in <5% of patients and is more likely to lead to renal dysfunction. Chronic hypercalcemia also can impair concentrating ability, thereby resulting in polyuria, polydipsia, and nocturia. The incidence of hypertension is variable but has been reported to occur in up to 50% of patients with PHPT. Hypertension appears to be more common in older patients and correlates with the magnitude of renal dysfunction and, in contrast to other symptoms, is least likely to improve after parathyroidectomy.

Bone Disease

Bone disease, including osteopenia, osteoporosis, and osteitis fibrosa cystica, is found in about 15% of patients with PHPT. Increased bone turnover, as found in patients with osteitis fibrosa cystica, can be determined by documenting an elevated blood alkaline phosphatase level. Advanced PHPT with osteitis fibrosa cystica now occurs in <5% of patients. It has pathognomonic radiologic findings, which are best seen on x-rays of the hands and are characterized by subperiosteal resorption (most apparent on the radial aspect of the middle phalanx of the second and third fingers), bone cysts, and tufting of the distal phalanges (Fig. 38-29). The skull also may be affected and appears mottled with a loss of definition of the inner and outer cortices. Brown or osteoclastic tumors and bone cysts also may be present. Severe bone disease, resulting in bone pain and tenderness and/or pathologic fractures, is rarely observed nowadays. However, reductions of bone mineral density (BMD) with osteopenia and osteoporosis are more common. Patients with normal serum alkaline phosphatase levels almost never have clinically apparent osteitis fibrosa cystica. HPT typically results in a loss of bone mass at sites of cortical bone such as the radius and relative preservation of cancellous bone such as that located at the vertebral bodies. Patients with PHPT, however, also may have osteoporosis of the lumbar spine that improves dramatically following parathyroidectomy. Fractures also occur more frequently in patients with PHPT, and the incidence of fractures also decreases after parathyroidectomy. Bone disease correlates with serum PTH and vitamin D levels.

Gastrointestinal Complications

PHPT has been associated with peptic ulcer disease. In experimental animals, hypergastrinemia has been shown to result from PTH infusion into blood vessels supplying the stomach, independent of its effects on serum calcium. An increased incidence of pancreatitis also has been reported in patients with PHPT, although this appears to occur only in patients with profound hypercalcemia (Ca²⁺ ≥12.5 mg/dL). Patients with PHPT also have an increased incidence of cholelithiasis, presumably due to an increase in biliary calcium, which leads to the formation of calcium bilirubinate stones.

Neuropsychiatric Complications

Severe hypercalcemia may lead to various neuropsychiatric manifestations such as florid psychosis, obtundation, or coma. Other findings such as depression, anxiety, and fatigue are more commonly observed in patients with only mild hypercalcemia. The etiology of these symptoms is not known. Studies demonstrate that levels of certain neurotransmitters (monoamine metabolites 5-hydroxyindoleacetic acid and homovanillic acid) are reduced in the cerebrospinal fluid of patients with PHPT when compared to controls. Electroencephalogram abnormalities also occur in patients with primary and secondary HPT and normalize following parathyroidectomy.

Other Features

PHPT also can lead to fatigue and muscle weakness, which is prominent in the proximal muscle groups. Although the exact etiology of this finding is not known, muscle biopsy studies show that weakness results from a neuropathy, rather than a primary myopathic abnormality. Patients with HPT also have an increased incidence of chondrocalcinosis, gout, and pseudogout, with deposition of uric acid and calcium pyrophosphate crystals in the joints. Calcification at ectopic sites such as blood vessels, cardiac valves, and skin also has been reported, as has hypertrophy of the left ventricle independent of the presence of hypertension. There is also evidence for subtle cardiovascular manifestations in mild disease, such as changes in endothelial function, increased vascular stiffness, and perhaps subtle diastolic dysfunction. Several large studies from Europe also suggest that PHPT is associated with increased death rates from cardiovascular disease and cancer even in patients with mild HPT, although this finding was not substantiated in North American studies.

Physical Findings

Parathyroid tumors are seldom palpable, except in patients with profound hypercalcemia or parathyroid cancer. A palpable neck mass in a patient with PHPT is more likely to be thyroid in origin or a parathyroid cancer. Patients also may demonstrate evidence of band keratopathy, a deposition of calcium in Bowman’s membrane just inside the iris of the eye. This nonspecific condition generally is caused by chronic eye diseases such as uveitis, glaucoma, and trauma but also may occur in the presence of conditions associated with high calcium or phosphate levels. Fibro-osseous jaw tumors, and/or the presence of familial disease in patients with PHPT and jaw tumors, if present, should alert the physician to the possibility of parathyroid carcinoma.

Differential Diagnosis

Hypercalcemia may be caused by a multitude of conditions, as listed in Table 38-9. PHPT and malignancy account for >90% of all cases of hypercalcemia. PHPT is more common in the outpatient setting, whereas malignancy is the leading cause of hypercalcemia in hospitalized patients. PHPT can virtually always be distinguished from other diseases causing hypercalcemia by a combination of history, physical examination, and appropriate laboratory investigations.

Hypercalcemia associated with malignancy includes three distinct syndromes. Although bone metastases may cause hypercalcemia, patients with solid tumors of the lung, breast, kidney, head and neck, and ovary often have humoral hypercalcemia of malignancy, without any associated bony metastases. In addition, hypercalcemia also may be associated with hematologic malignancies such as multiple myeloma. Humoral hypercalcemia of malignancy is known to be mediated primarily by PTH-related peptide (PTHrP), which also plays a role in the hypercalcemia associated with bone metastases and multiple myeloma.

Thiazide diuretics cause hypercalcemia by decreasing renal clearance of calcium. This corrects in normal patients within days to weeks after discontinuing the diuretic, but patients with PHPT continue to be hypercalcemic. Thiazide diuretics can, therefore, exacerbate underlying PHPT and can be used to unmask PHPT in patients with borderline hypercalcemia. Familial hypocalciuric hypercalcemia (FHH) is a rare autosomal dominant condition with nearly 100% penetrance and results from inherited heterozygous mutations in the CASR gene located on chromosome 3.⁵⁴ Homozygous germline mutations at this locus result in neonatal hypercalcemia, a condition that can rapidly prove fatal. Patients with FHH have lifelong hypercalcemia, which is not corrected by parathyroidectomy. Hypercalcemia also is found in approximately 10% of patients with sarcoidosis secondary to increased 25-hydroxy vitamin D 1-hydroxylase activity in lymphoid tissue and pulmonary macrophages, which is not subject to inhibitory feedback control by serum calcium. Thyroid hormone also has bone-resorption properties, thus causing hypercalcemia in thyrotoxic states, especially in immobilized patients. Hemoconcentration appears to be an important factor in the hypercalcemia associated with adrenal insufficiency and pheochromocytoma, although the latter patients may have associated parathyroid tumors (MEN2A), and some pheochromocytomas are known to secrete PTHrP. Other endocrine lesions such as vasoactive intestinal peptide–secreting tumors may be associated with hypercalcemia due to increased secretion of PTHrP. Milk-alkali syndrome requires the ingestion of large quantities of calcium with an absorbable alkali such as that used in the treatment of peptic ulcer disease with antacids. Ingestions of large quantities of vitamins D and A are infrequent causes of hypercalcemia, as is immobilization.

Diagnostic Investigations

Biochemical Studies

The presence of an elevated serum calcium and intact PTH or two-site PTH levels, without hypocalciuria, establishes the diagnosis of PHPT with virtual certainty. These sensitive PTH assays use immunoradiometric or immunochemiluminescent techniques and can reliably distinguish PHPT from other causes of hypercalcemia. Furthermore, they do not cross-react with PTHrP (Fig. 38-30). In patients with metastatic cancer and hypercalcemia, intact PTH levels help to determine whether the patient also has concurrent PHPT. Although extremely rare, a patient with hypercalcemia may have a tumor that secretes PTH. FNAB of such a tumor for PTH levels or selective venous catheterization of the veins draining such tumors can help clarify the diagnosis.

Patients with PHPT also typically have decreased serum phosphate (∼50%) and elevated 24-hour urinary calcium concentrations (∼60%). A mild hyperchloremic metabolic acidosis also is present (80%), thereby leading to an elevated chloride-to-phosphate ratio (>33). Urinary calcium levels need not be measured routinely, except in patients who have not had previously documented normocalcemia or have a family history of hypercalcemia to rule out FHH. In patients with FHH, 24-hour urinary calcium excretion is characteristically low (<100 mg/d). Furthermore, the serum calcium-to-creatinine clearance ratio usually is <0.01 in patients with FHH, whereas it is typically >0.02 in patients with PHPT. Other biochemical features of PHPT are listed in Table 38-10. Elevated levels of alkaline phosphatase may be found in approximately 10% of patients with PHPT and are indicative of high-turnover bone disease. These patients are prone to developing postoperative hypocalcemia due to bone hunger. Serum and urine protein electrophoresis may be necessary to exclude multiple myeloma.

Occasionally, patients present with normocalcemic PHPT due to vitamin D deficiency, a low serum albumin, excessive hydration, a high-phosphate diet, or a low normal blood calcium set point. These patients have increased total PTH levels with or without increased blood ionized calcium levels and must be distinguished from patients with renal leak hypercalciuria who also have increased PTH levels due to excessive calcium loss in the urine. This can be accomplished by administering thiazide diuretics. In patients with idiopathic hypercalciuria, the urinary calcium level falls, and the secondary increase in the blood PTH level also decreases to normal, whereas patients with normocalcemic HPT continue to have elevated urine calcium and blood PTH levels and may, in fact, become hypercalcemic.

Radiologic Tests

In patients with profound hypercalcemia or PHPT associated with vitamin D deficiency, hand and skull x-rays may demonstrate osteitis fibrosa cystica, but this is rare in current clinical practice. BMD studies using dual-energy absorptiometry are being increasingly used to assess the effects of PHPT on bone. Abdominal ultrasound examination is used selectively to document renal stones. Parathyroid localization studies are not used to confirm the diagnosis of PHPT, but rather to aid in identifying the location of the offending gland(s), as discussed later in Preoperative Localization Tests.

Treatment

Indications for Parathyroidectomy and Role of Medical Management

Most authorities agree that patients who have developed complications and have “classic” symptoms of PHPT should undergo parathyroidectomy. However, the treatment of patients with asymptomatic PHPT has been the subject of controversy, due, in part, to the fact that there is little agreement on what constitutes an asymptomatic patient.

At the National Institutes of Health consensus conference in 1990, “asymptomatic” PHPT was defined as “the absence of common symptoms and signs of PHPT, including no bone, renal, gastrointestinal, or neuromuscular disorders.” To determine the best course of action for these patients, it is important to consider the natural history of untreated PHPT and the outcomes of treatment options, both medical and surgical.

With respect to the natural history, the panel advocated nonoperative management of these patients with mild PHPT based on observational studies, which suggested relative stability of biochemical parameters over time. However, the consensus panel considered certain patients to be candidates for surgery based on testing or other information indicating end-organ effects or a higher likelihood of disease progression, and this led to the establishment of initial guidelines for parathyroidectomy.⁵⁸ Subsequently, another observational study on the natural history of treated versus untreated HPT was published by Silverberg and colleagues.⁵⁹ In their cohort of 52 patients with asymptomatic HPT followed without surgery, levels of serum and urinary calcium, PTH, alkaline phosphatase, and vitamin D metabolites remained relatively stable over a 10-year period in most patients. Average bone mass also remained relatively stable. However, the study also reported development of a new indication for surgery in 14 (27%) of 52 of their asymptomatic patients and, because approximately 50% of their patients were initially treated surgically, overall, about 75% of patients were underwent parathyroidectomy. Age <50 years was predictive of progression, and patients undergoing parathyroidectomy showed not only normalization of calcium and PTH levels, but also improved BMD at the spine and hip. Based on these and other studies, the guidelines were reassessed at a second workshop on asymptomatic PHPT held at the National Institutes of Health in 2002.⁶⁰

Since that time, additional studies have provided further insights into the natural history of treated and untreated HPT. Three of these were randomized, controlled, prospective studies ranging in duration from 1 to 3.5 years. One was an observational study (a continuation of the Columbia University PHPT Project) but was notable for its long duration of follow-up of 15 years.⁶¹ These studies confirmed the relative stability of various biochemical indices, thus validating the need for guidelines. However, the long-term study suggested that the stability was not indefinite as calcium levels tended to rise in years 13 to 15. In addition, the study also demonstrated that bone density measurements remained stable for 8 to 10 years, but cortical bone density worsened after year 10. More concerning was the fact that 60% of patients lost >10% of their BMD over the 15-year observation period. Furthermore, whether patients met the 2002 guidelines for surgery did not appear to predict the risk of progressive disease, with 40% of patients undergoing follow-up eventually needing surgery. Although there are no randomized trials, registry data also suggest that fracture risk is increased for PHPT up to 10 years prior to diagnosis and treatment.

Medical options for treating PHPT and its complications include antiresorptive treatments such as bisphosphonates, hormone replacement therapy (HRT), and selective estrogen receptor modulators such as raloxifene.⁶² Bisphosphonates and HRT are reasonable options in patients for whom skeletal protection is needed, as evidence from randomized, placebo-controlled trials indicates that these medications are very effective at decreasing bone turnover and increasing BMD in PHPT, with the effects being comparable to patients undergoing parathyroidectomy. Caution needs to be exercised due to the nonskeletal effects of HRT, and hence, bisphosphonates are preferred. There are no clinical studies regarding the effects of raloxifene on BMD in HPT, and none of these agents affects calcium or PTH levels. More recently, calcimimetics (modifiers of the sensitivity of the CASR) have been used in randomized, multicenter controlled trials and have been shown to decrease both serum calcium and PTH levels in both symptomatic and asymptomatic PHPT patients. Unfortunately, bone density failed to improve in medically treated patients. Although this therapy shows promise, long-term outcome data are lacking, and their routine use is not advocated at this time, except in patients who are very poor operative risks or refuse surgery.

Successful parathyroidectomy results in resolution of osteitis fibrosa cystica and decreased formation of renal stones in symptomatic (classic) patients. In addition, it results in improved BMD (6% to 8% in the first year and up to 12% to 15% at 15 years) and fracture risk (by 50% at hip and upper arm and 30% overall) after adjustment for age, sex, and previous fractures over a 20-year observation period.⁵⁹ There are also data to show that it improves a number of the nonspecific manifestations of PHPT such as fatigue, polydipsia, polyuria and nocturia, bone and joint pain, constipation, nausea, and depression in many patients. This also has been demonstrated using symptom questionnaires and various standardized general quality-of-life assessments such as the SF-36 and a specific parathyroidectomy assessment of symptoms scale.⁶³ The increased death rate in patients with PHPT appears to be reversible by successful parathyroidectomy, at least in some studies. Lastly, parathyroidectomy can be accomplished with >95% success rates with minimal morbidity, even in elderly patients and is the only curative treatment option for PHPT. Previous investigations also have documented that parathyroidectomy is more cost-effective than medical management or follow-up.⁶⁴

Given the above, it is recommended that parathyroidectomy should be offered to virtually all patients except those in whom the operative risks are prohibitive. This is also acknowledged by the panel of the latest workshop, which stated that “even though patients may not meet the guidelines for surgical intervention, it is always a reasonable option in those who do not have medical contraindications.”

Currently, parathyroidectomy is recommended for patients with smaller elevations in serum calcium levels (>1 mg/dL above the upper limit of normal) and if BMD measured at any of three sites (radius, spine, or hip) is greater than 2.5 standard deviations below those of gender- and race-matched, not age-matched, controls (i.e., peak bone density or T score [rather than Z score] <2.5). In addition, patients <50 years of age are advised to undergo parathyroidectomy. The significant changes from the previous guidelines pertain to the fact that (a) hypercalciuria (>400 mg/24 h), in the absence of nephrolithiasis, is no longer considered an indication for parathyroidectomy; and (b) an absolute glomerular filtration rate <60 mL/min, rather than a reduction of creatinine clearance by 30%, is used as a threshold for recommending surgery. These changes are based on (a) data that demonstrate that hypercalciuria per se is not a risk factor for kidney stones, and (b) the most recent epidemiologic data from the U.S. National Health and Nutrition Examination Survey using a numerical cutoff for kidney function that relies on an age invariant standard, similar to densitometric guidelines. Although the guidelines for BMD did not change, it was noted that a patient with a history of fragility fracture(s) should also undergo parathyroidectomy.⁵⁵ The current guidelines are summarized in Table 38-11.

It is important to point out that the neurocognitive and neuropsychological aspects of PHPT remain a topic of controversy with respect to the guidelines for parathyroidectomy. Although there were more studies since the previous iteration of the guidelines, there were concerns that while some lacked adequate controls and were plagued by problems related to the instruments used to quantify these nonspecific symptoms, others showed variability in improvement of neurocognitive symptoms following parathyroidectomy. Similarly, uncertainty is also present concerning the cardiovascular consequences of mild HPT. Therefore, the workshop panel emphasizes that these criteria alone should not be used as guidelines for surgical intervention.

Since there are no definitive criteria to indicate which patients with mild PHPT will develop progressive disease, more clinical studies are required. Patients who do not undergo surgery should undergo routine follow-up as outlined in the recent workshop summary statement, consisting of annual calcium and serum creatinine measurements, and measurements of BMD at three sites every 1 to 2 years.⁶⁵

Preoperative Localization Tests

Localization studies may be classified into noninvasive or invasive modalities. These studies have variable performance characteristics, which, in turn, vary with operator and institutional experience, as outlined in Table 38-12. Localization studies have permitted surgeons to perform more limited operations, some of them under local anesthesia. These “minimally invasive” procedures include unilateral and focused neck exploration, radio-guided parathyroidectomy, and several endoscopic or video-assisted approaches. The use of localization studies has been shown in some studies to be associated with lower morbidity rates (hypoparathyroidism and RLN injury) and decreased operative times, reduced duration of hospital stay, and improved cosmetic outcomes, while maintaining success rates similar to those obtained with traditional bilateral neck explorations. Some studies also show that use of localization studies may be more cost-effective. Overall, it has become routine to localize hyperfunctioning parathyroid glands before parathyroidectomy. It is important to point out that imaging is not a diagnostic approach, and the decision for exploration should be made before any imaging is performed.

^99mTc-labeled sestamibi (Fig. 38-31A) is the most widely used and accurate modality with a sensitivity >80% for detection of parathyroid adenomas. Sestamibi (Cardiolite) initially was introduced for cardiac imaging and is concentrated in mitochondria-rich tissue. It was subsequently noted to be useful for parathyroid localization due to the delayed washout of the radionuclide from hypercellular parathyroid tissue compared to thyroid tissue. Sestamibi scans generally are complemented by neck ultrasound (Fig. 38-31B), which can identify adenomas with >75% sensitivity in experienced centers and is most useful in identifying intrathyroidal parathyroids. Single-photon emission CT, particularly when used with CT, has been shown to be superior to other nuclear medicine–based imaging. Specifically, single-photon emission CT can indicate whether an adenoma is located in the anterior or posterior mediastinum (aortopulmonary window), thus enabling the surgeon to modify the operative approach accordingly. CT and MRI scans are less sensitive than sestamibi scans, but are helpful in localizing large paraesophageal and mediastinal glands. More recently, four-dimensional CT (4D-CT) has shown utility in parathyroid localization. This technique incorporates the perfusion of contrast in hyperfunctioning parathyroid tissue over time, thus providing functional information in addition to the anatomic information provided by conventional three-dimensional CT imaging. In one study, 4D-CT showed improved sensitivity of 88% compared to that of sestamibi (65%) and ultrasound (57%) for lateralization of the enlarged gland and also showed superiority when localization to the correct quadrant was examined.⁶⁶ A combination of 4D-CT and ultrasound has been reported to have a positive predictive value of 92% for single-gland disease and 75% for multiple-gland disease.

IOPTH was initially introduced in 1993 and is used to determine the adequacy of parathyroid resection (Fig. 38-32).⁶⁷ According to one commonly used criterion, when the PTH falls by 50% or greater 10 minutes after removal of a parathyroid tumor, as compared to the highest preremoval value, the test is considered positive, and the operation is terminated. IOPTH measurements, like localization studies, are less reliable in multiglandular disease. Bilateral internal jugular vein sampling has also been used to lateralize tumors intraoperatively but is less accurate.

Operative Approaches

Unilateral parathyroid exploration was first carried out using intraoperative staining of a biopsy from the normal parathyroid gland with Sudan black dye to rule out a double adenoma. Initially, the choice of side to be explored was random, but the introduction of preoperative localization studies has enabled a more directed approach. In contrast, the focused approach identifies only the enlarged parathyroid gland, and no attempts are made to locate other parathyroid glands. Unilateral neck explorations have several advantages over bilateral neck exploration, including reduced operative times and complications, such as injury to the RLN and hypoparathyroidism. However, most existing studies comparing the two approaches are retrospective and do not analyze the results on an intent-to-treat basis. Another argument against a unilateral exploration is the risk of missing another adenoma on the opposite side of the neck. The incidence of double adenomas has been reported to range from 0% to 10%, with an increased incidence in elderly patients. The risk of missing a second adenoma is higher in populations with a higher incidence of multiple adenomas, such as those with familial HPT, MEN syndromes, and the elderly. Another difficulty inherent with unilateral exploration is the inability to discern whether the combination of an abnormal gland and a normal gland on the initial side constitutes a single adenoma or asymmetric hyperplasia. A recently published update on the 5-year results of a randomized trial comparing unilateral versus bilateral neck exploration did not note any difference in the rates of recurrent or persistent disease in the two groups of patients.⁶⁸ These issues will only be resolved by a large, prospective, multicenter study or improved molecular analytic techniques.

Radio-guided parathyroidectomy takes advantage of the ability of parathyroid tumors to retain ^99mTc-sestamibi. One to 2 mCi of the isotope is injected before surgery, and a hand-held gamma probe is used to guide the identification of the enlarged gland, taking care to ensure the equilibration of radioactivity counts in all quadrants. Reported advantages include easier localization, particularly in reoperative cases, and the ability to perform the procedure under local anesthetic or sedation using smaller incisions. Many studies demonstrated the feasibility of this technique; however, it is rarely used now, largely because it offers little advantage over preoperative sestamibi scans and is associated with increased operative times. Like preoperative scanning, it also has reduced accuracy in the presence of multiglandular disease.

Endoscopic approaches include both video-assisted and total endoscopic techniques. Total endoscopic parathyroidectomy was first described by Gagner in 1996,⁶⁹ and several other investigators have since reported on this technique. Although port placements are variable, as is the case with endoscopic thyroidectomy, they all involve creation of a working space in the neck using CO₂ insufflation, with the reported advantages being superior cosmesis and excellent visualization. Although feasible, these techniques also have been associated with increased operating times, more personnel, and greater expense, and have, in general, not been useful for patients with multiglandular disease, a large thyroid mass, or previous neck surgery and irradiation. Their greatest use has been in patients with tumors at ectopic sites such as the mediastinum where thoracoscopic parathyroidectomy is an excellent alternative to sternotomy. Robotic approaches using a gasless, transaxillary technique are also being used for parathyroidectomy. Reported advantages include improved three-dimensional magnified visualization, refined ergonomic control, more freedom of motion with multiarticulated instruments, and improved cosmetic result as a result of incision placement in the axilla.⁷⁰

Studies have shown that if both sestamibi scan and neck ultrasound studies independently identify the same, enlarged parathyroid gland, and no other gland, it is indeed the abnormal gland in approximately 95% of cases. These patients with sporadic PHPT are candidates for a focused neck exploration, an approach that is most commonly referred to as minimally invasive parathyroidectomy. A standard bilateral neck exploration is planned if parathyroid localization studies or IOPTH are not available; if the localizing studies fail to identify any abnormal parathyroid gland or identify multiple abnormal glands in patients with a family history of PHPT, MEN1, or MEN2A; or if a concomitant thyroid disorder requires bilateral exploration. In addition, finding a minimally abnormal parathyroid gland on the side indicated by localization studies during focal exploration should prompt a bilateral exploration or at least the identification of a normal parathyroid gland on the same side. In patients with MEN1, HPT should be corrected before treatment of gastrinomas because gastrin levels decline after parathyroidectomy.

Conduct of Parathyroidectomy (Standard Bilateral Exploration)

An experienced parathyroid surgeon with a thorough knowledge of parathyroid anatomy and embryology and meticulous technique is crucial for the best surgical results. The procedure usually is performed under general anesthesia. The patient is positioned supine on the operating table with the neck extended. For a bilateral exploration, the neck is explored via a 3- to 4-cm incision just caudal to the cricoid cartilage. The initial dissection and exposure is similar to that used for thyroidectomy. After the strap muscles are separated in the midline, one side of the neck is chosen for exploration. In contrast to a thyroidectomy, the dissection during a parathyroidectomy is maintained lateral to the thyroid, making it easier to identify the parathyroid glands and not disturb their blood supply.

Identification of Parathyroids

A bloodless field is important to allow identification of parathyroid glands. The middle thyroid veins are ligated and divided, thus enabling medial and anterior retraction of the thyroid lobe, with the aid of a peanut sponge or placement of 2-0 silk sutures into the thyroid. The space between the carotid sheath and thyroid is then opened by gentle sharp and blunt dissection, from the cricoid cartilage superiorly to the thymus inferiorly and the RLN is identified. Approximately 85% of the parathyroid glands are found within 1 cm of the junction of the inferior thyroid artery and RLNs. The upper parathyroid glands usually are superior to this junction and dorsal (posterior) to the nerve, whereas the lower glands are located inferior to the junction and ventral (anterior) to the recurrent nerve. Because parathyroid glands are partly surrounded by fat, any fat lobule at typical parathyroid locations should be explored because the normal or abnormal parathyroid gland may be concealed in the fatty tissue. The thin fascia overlying a “suspicious” fat lobule should be incised using a sharp curved hemostat and scalpel. This maneuver often causes the parathyroid gland to “pop” out. Alternatively, gentle, blunt peanut sponge dissection between the carotid sheath and the thyroid gland often reveals a “float” sign, suggesting the site of the abnormal parathyroid gland. Normal parathyroids are light beige and only slightly darker or brown compared to adjacent fat.

Parathyroid tissue needs to be distinguished from normal or brown fat tissue, thyroid nodules, lymph nodes, and ectopic thymus. Lymph nodes generally are light beige to whitish gray in color, glassy, and multiple in number, whereas thyroid nodules generally are more vascular, firm, dark or reddish brown in color, and have a more variegated appearance. Intraoperatively, a suspicious nodule may be aspirated using a fine needle attached to a syringe containing 1 cc of saline. Very high PTH levels in the aspirate have been shown to be diagnostic in the intraoperative identification of parathyroid glands. Several characteristics such as size (>7 mm), weight, and color are used to distinguish normal from hypercellular parathyroid glands. Hypercellular glands generally are darker, more firm, and more vascular than normocellular glands. No single method is 100% reliable, and therefore, the parathyroid surgeon must rely on experience and, at times, advice from a pathologist to help distinguish normal from hypercellular glands. Although several molecular studies have shown use in distinguishing parathyroid adenomas from hyperplasia, this determination also must be made by the surgeon intraoperatively by documenting the presence of a normal parathyroid gland.

Location of Parathyroid Glands

The majority of lower parathyroid glands are found in proximity to the lower thyroid pole (Fig. 38-33A). If not found at this location, the thyrothymic ligament and thymus should be mobilized. The upper end of the cervical thymus is gently grasped with a right angle clamp, and the distal portion is bluntly dissected from perithymic fat with a peanut sponge. One can then “walk down” the thymus with successive right-angle clamps (Fig. 38-33B). Applying light tension along with a “twisting” motion helps to free the upper thymus. The carotid sheath also should be opened from the bifurcation to the base of the neck if the parathyroid tumor cannot be found. If these maneuvers are unsuccessful, an intrathyroidal gland should be sought by using intraoperative ultrasound, incising the thyroid capsule on its posterolateral surface, or by performing an ipsilateral thyroid lobectomy and “bread-loafing” the thyroid lobe. Preoperative or intraoperative ultrasonography can be useful for identifying intrathyroidal parathyroid glands. Rarely, the third branchial pouch may maldescend and be found high in the neck (undescended parathymus), anterior to the carotid bulb, along with the missing parathyroid gland. Upper parathyroid glands are more consistent in position and usually are found near the junction of the upper and middle thirds of the gland, at the level of the cricoid cartilage (Fig. 38-33C). Ectopic upper glands may be found in carotid sheath, tracheoesophageal groove, retroesophageal, or in the posterior mediastinum. The locations of ectopic upper and lower parathyroid glands are shown in Fig. 38-34. Every attempt must be made to identify all four glands. Treatment depends on the number of abnormal glands.

1. A single adenoma is presumed to be the cause of a patient’s PHPT if only one parathyroid tumor is identified and the other parathyroid glands are normal, a situation present in about 80% of patients with PHPT. Adenomas typically have an atrophic rim of normal parathyroid tissue, but this characteristic may be absent. The adenoma is dissected free of surrounding tissue, taking care to stay immediately adjacent to the tumor, without fracturing it. The vascular pedicle is clamped, divided, and ligated. Care should be taken to not rupture the parathyroid gland to decrease the risk of parathyromatosis. If there is any question about the presumed normal glands, one of them should be biopsied and examined by frozen section.

2. If two abnormal and two normal glands are identified, the patient has double adenomas. Triple adenomas are present if three glands are abnormal and one is normal. Multiple adenomas are more common in older patients with an incidence of up to 10% in patients >60 years old. The abnormal glands should be excised, provided the remaining glands are confirmed as such, thus excluding asymmetric hyperplasia after biopsy and frozen section.

3. If all parathyroid glands are enlarged or hypercellular, patients have parathyroid hyperplasia that has been shown to occur in about 15% of patients in various series. These glands are often lobulated, usually lack the rim of normal parathyroid gland seen in adenomas, and may be variable in size. It often is difficult to distinguish multiple adenomas from hyperplasia with variable gland size. Hyperplasia may be of the chief cell (more common), mixed, or clear cell type. Patients with hyperplasia may be treated by subtotal parathyroidectomy or by total parathyroidectomy and autotransplantation, with the choice of procedure being determined by rates of recurrence, postoperative hypocalcemia, and failure rates of autotransplanted tissue. Initial studies demonstrated equivalent cure rates and postoperative hypocalcemia for the two techniques, with the latter having the added advantage of avoiding recurrence in the neck. However, autotransplanted tissue may fail to function in about 5% of cases.

All four parathyroid glands are identified and carefully mobilized. For patients with hyperplasia, a titanium clip is placed across the most normal gland, leaving a 50-mg remnant and taking care to avoid disturbing the vascular pedicle and that the gland is resected with a sharp scalpel. If possible, it is preferable to subtotally resect an inferior gland, which is more easily accessible in case of recurrence due to its anterior location with respect to the RLN. The resected parathyroid tissue is confirmed by frozen section or PTH assay. If the remnant appears to be viable, the remaining glands are resected. If there is any question as to the viability of the initially subtotally resected gland, another gland is chosen for subtotal resection and the initial remnant is removed. Whenever multiple parathyroids are resected, it is preferable to cryopreserve tissue, so that it may be autotransplanted should the patient become hypoparathyroid. Parathyroid tissue usually is transplanted into the nondominant forearm. A horizontal skin incision is made overlying the brachioradialis muscle a few centimeters below the antecubital fossa. Pockets are made in the belly of the muscle, and one to two pieces of parathyroid tissue measuring 1 mm each are placed into each pocket. A total of 12 to 14 pieces are transplanted. Autotransplanted tissue also has been reported to function when transplanted into fat.

Indications for Sternotomy

A sternotomy is usually not recommended at the initial operation, unless the calcium level is >13 mg/dL. Rather, it is preferred to biopsy the normal glands and subsequently close the patient’s neck and obtain localizing studies, if they were not obtained previously. Intraoperative PTH assay during the operation from large veins may be helpful. Using highly selective venous catheterization postoperatively also may be needed when noninvasive localization studies are negative, equivocal, or conflicting. Lower parathyroid glands tend to migrate into the anterior mediastinum in the thymus or perithymic fat and usually can be approached via a cervical incision. A sternotomy is needed to deliver these tumors in approximately 5% of cases. Generally, the gland can be approached by a partial sternotomy to the third intercostal space. The midline sternotomy can be extended to the left or right side as required. Upper glands tend to migrate to the posterior mediastinum in the tracheoesophageal groove. Mediastinal glands also may be found in the aortopulmonary window or pericardium, or attached to the ascending aorta, aortic arch, or its branches.

Special Situations

Normocalcemic Hyperparathyroidism

This disorder is becoming increasingly recognized in clinical practice and is defined by the presence of an elevated PTH level with normal calcium (including ionized calcium) levels. In addition, other secondary causes of elevated PTH should be ruled out, namely, vitamin D deficiency, osteomalacia, hypercalciuria (renal leak), and renal insufficiency. Data regarding the natural history of this disorder are limited. In a series of 37 patients, Lowe and colleagues⁷¹ showed that 19% of patients became frankly hypercalcemic within 3 years. In addition, 57% developed osteoporosis, 11% developed fragility fractures, and 14% developed nephrolithiasis. Although the study had some limitations, it led the authors to suggest that normocalcemic HPT may represent a variant of “symptomatic” PHPT and may not be an early form of “asymptomatic” disease. There is still uncertainty about whether normocalcemic HPT carries the same cardiovascular morbidity that is associated with the more severe form of the disease. Limited studies show that parathyroidectomy is more likely to be unsuccessful in these patients. In the absence of strong data, no guidelines are available for this entity.⁶⁵ As such, most clinicians follow a conservative course unless patients have symptoms that can be attributable to HPT.

Parathyroid Carcinoma

Parathyroid cancer accounts for approximately 1% of PHPT cases. It may be suspected preoperatively by the presence of severe symptoms, serum calcium levels >14 mg/dL, significantly elevated PTH levels (five times normal), and a palpable parathyroid gland. Local invasion is quite common; approximately 15% of patients have lymph node metastases and 33% have distant metastases at presentation. Intraoperatively, parathyroid cancer is suggested by the presence of a large, gray-white to gray-brown parathyroid tumor that is adherent to or invasive into surrounding tissues like muscle, thyroid, RLN, trachea, or esophagus. Enlarged lymph nodes also may be present. Frozen sections are generally unreliable. Accurate diagnosis necessitates histologic examination. The major diagnostic criteria include vascular or capsular invasion, trabecular or fibrous stroma, and frequent mitoses. It is, however, important to emphasize that these classic findings are not as frequently noted as previously reported, and some may be found in benign adenomas as well.

Treatment of parathyroid cancer consists of neck exploration, with en bloc excision of the tumor and the ipsilateral thyroid lobe, in addition to the removal of contiguous lymph nodes (tracheoesophageal, paratracheal, and upper mediastinal). The recurrent nerve is not sacrificed unless it is directly involved with tumor. Adherent soft tissue structures (strap muscles or other soft tissues) should also be resected. Modified radical neck dissection is recommended in the presence of lateral lymph node metastases. Prophylactic neck dissection is not advised. If the diagnosis is made postoperatively, a decision must be made regarding the adequacy of initial surgery based on a review of operative notes, pathology reports, localization studies, and calcium and PTH levels. If any question exists, histologic review by another experienced pathologist can be helpful. Additional procedures can include ipsilateral thyroid lobectomy with resection of contiguous structures and lymph nodes if the features are typical or the patient remains hypercalcemic. Patients with equivocal pathologic findings and normocalcemia may be monitored closely. Reoperation is indicated for locally recurrent or metastatic disease to control hypercalcemia. Adjuvant radiation therapy should be considered in patients at high risk of local recurrence such as those with close or positive margins, invasion of surrounding structures, or tumor rupture. Radiation may also be used as primary therapy in unresectable disease or for palliation of bone metastases. Chemotherapy is not very effective. Bisphosphonates have shown some effectiveness in treating hypercalcemia associated with parathyroid carcinoma. Cinacalcet hydrochloride, a calcimimetic, can reduce PTH levels by directly binding to the CASR cells on the parathyroid gland and has been shown to be useful in controlling hypercalcemia in patients with refractory parathyroid carcinoma.⁷² Other promising approaches include anti-parathyroid hormone immunotherapy, octreotide, and the telomerase inhibitor azidothymidine, but additional investigations are needed in this area.

Familial Hyperparathyroidism

PHPT may occur as a component of various inherited syndromes such as MEN1 and MEN2A. Inherited PHPT also can occur as isolated familial HPT (non-MEN) or familial HPT with jaw tumors. The diagnosis of familial HPT is known or suspected in approximately 85% of patients preoperatively. Furthermore, patients with hereditary HPT generally have a higher incidence of multiglandular disease, supernumerary glands, and recurrent or persistent disease. Therefore, these patients warrant a more aggressive approach and are not candidates for various focused surgical approaches.⁶⁶ Although not absolutely necessary, preoperative sestamibi scan and ultrasound can be obtained in patients with inherited HPT to identify potential ectopic glands. A standard bilateral neck exploration is performed, along with a bilateral cervical thymectomy, regardless of the results of localization studies. Both subtotal parathyroidectomy and total parathyroidectomy with autotransplantation are appropriate, and parathyroid tissue also should be cryopreserved. If an adenoma is found in patients with familial HPT, the adenoma and the ipsilateral normal parathyroid glands are resected. The normal-appearing glands on the contralateral side are biopsied and marked, so that only one side of the neck will need to be explored in the event of recurrence. Patients with MEN2A require total thyroidectomy and central neck dissection for prevention/treatment of MTC, a procedure that places the parathyroids at risk. Moreover, HPT is less aggressive in these patients. Hence, only abnormal parathyroid glands need to be resected at neck exploration. The other normal parathyroid glands should be marked with a clip.

Neonatal Hyperparathyroidism

Infants with neonatal HPT present with severe hypercalcemia, lethargy, hypotonia, and mental retardation. This disorder is associated with homozygous mutations in the CASR gene. Urgent total parathyroidectomy (with autotransplantation and cryopreservation) and thymectomy are indicated. Subtotal resection is associated with high recurrence rates.

Parathyromatosis

Parathyromatosis is a rare condition characterized by the finding of multiple nodules of hyperfunctioning parathyroid tissue throughout the neck and mediastinum, usually following a previous parathyroidectomy. The true etiology of parathyromatosis is not known. It is postulated to arise either from overgrowth of congenital parathyroid rests (ontogenous parathyromatosis) or seeding at surgery from rupture of parathyroid tumors or subtotal resection of hyperplastic glands. Parathyromatosis represents a rare cause of persistent or recurrent HPT⁷³ and can be identified intraoperatively. Aggressive local resection of these deposits can result in normocalcemia but is rarely curative. Some studies suggest that these patients have low-grade carcinoma because of invasion into muscle and other structures distant from the resected parathyroid tumor.

Postoperative Care and Follow-Up

Patients who have undergone parathyroidectomy are advised to undergo calcium level checks 2 weeks postoperatively, at 6 months, and then annually. Recurrences are rare (<1%), except in patients with familial HPT. Recurrence rates of 15% at 2 years and 67% at 8 years have been reported for MEN1 patients.

Persistent and Recurrent Hyperparathyroidism

Persistence is defined as hypercalcemia that fails to resolve after parathyroidectomy and is more common than recurrence, which refers to HPT occurring after an intervening period of at least 6 months of biochemically documented normocalcemia.⁷⁴ Recurrent disease is far less common than persistent HPT; however, both occur more frequently in the setting of familial HPT and MEN1, in particular. The most common causes for both these states include ectopic parathyroids, unrecognized hyperplasia, or supernumerary glands. More rare causes include parathyroid carcinoma, missed adenoma in a normal position, incomplete resection of an abnormal gland, parathyromatosis, or an inexperienced surgeon. The most common sites of ectopic parathyroid glands in patients with persistent or recurrent HPT are paraesophageal (28%), mediastinal (26%), intrathymic (24%), intrathyroidal (11%), carotid sheath (9%), and high cervical or undescended (2%) (Fig. 38-35).

Once the diagnosis of persistent or recurrent HPT is suspected, it should be confirmed by the necessary biochemical tests. Other causes of an elevated serum PTH such as renal insufficiency, renal calcium leak, and GI tract abnormalities should be considered. A detailed family history should be performed to screen for familial disease, as this will influence the operative approach. In particular, a 24-hour urine collection should be performed to rule out FHH. In redo-parathyroid surgery, the glands are more likely to be in ectopic locations, and postoperative scarring tends to make the procedure more technically demanding. Cure rates are generally lower (80%–90% compared with 95%–99% for initial operation), and risk of injury to RLNs and permanent hypocalcemia are higher. Therefore, an evaluation of severity of HPT and the patient’s anesthetic risk (using the American Society of Anesthesiology classification of physical status or the Goldman cardiac index) is important. There are no published guidelines directly applicable to this group of patients. In general, patients with significant and ongoing problems such as recurrent kidney stones, a markedly elevated calcium level, or ongoing bone loss will need re-exploration. Patients in whom the diagnosis remains in question or those with equivocal or minimal symptoms may be considered for conservative management. Preoperative localization studies are routinely performed. Noninvasive studies such as a sestamibi scan and ultrasound are obtained, supplemented by 4D-CT scans. If these studies are negative, discordant, or equivocal, obtaining an ultrasound-guided aspirate of a suspicious cervical lesion or a highly selective venous catheterization for PTH levels (by an experienced angiographer) is recommended. Previous operative notes and pathology reports should be carefully reviewed and reconciled with the information obtained from localization studies before any neck re-exploration. An algorithm for the treatment of patients with recurrent and persistent HPT is shown in Fig. 38-36.

Generally, these patients are approached with a focused exploration. The lateral approach is frequently used and can be achieved via the previous incision. The plane between the sternocleidomastoid and strap muscle is opened and allows for early identification of the RLN. Parathyroid tissue is cryopreserved routinely. Use of adjuncts, such as measuring intraoperative PTH levels, is critical to ensure adequate resection and avoid potentially harmful additional explorations. In case of difficult re-explorations, additional techniques such as bilateral internal jugular vein sampling for PTH, thyroid lobectomy on the side of the missing gland, cervical thymectomy, and ligation of the ipsilateral inferior thyroid artery (after lobectomy, to cause infarction of the missing gland) may be needed. Blind mediastinal exploration is not recommended. In patients who are denied, refuse, or fail exploration, medical options such as cinacalcet may be considered.

Hypercalcemic Crisis

Patients with PHPT may occasionally present acutely with nausea, vomiting, fatigue, muscle weakness, confusion, and a decreased level of consciousness—a complex referred to as hypercalcemic crisis. These symptoms result from severe hypercalcemia from uncontrolled PTH secretion, worsened by polyuria, dehydration, and reduced kidney function and may occur with other conditions causing hypercalcemia. Calcium levels are markedly elevated and may be as high as 16 to 20 mg/dL. Parathyroid tumors tend to be large or multiple and may be palpable. Patients with parathyroid cancer or familial HPT are more likely to present with hypercalcemic crisis.

Treatment consists of therapies to lower serum calcium levels followed by surgery to correct HPT. The mainstay of therapy involves rehydration with a 0.9% saline solution to keep urine output >100 cc/h. Once urine output is established, diuresis with furosemide (which increases renal calcium clearance) is begun. If these methods are unsuccessful, other drugs may be used to lower serum calcium levels as outlined in Table 38-13. Occasionally, in life-threatening cases, hemodialysis may be of benefit.

Secondary Hyperparathyroidism

Secondary HPT commonly occurs in patients with chronic renal failure but also may occur in those with hypocalcemia secondary to inadequate calcium or vitamin D intake or malabsorption. The pathophysiology of HPT in chronic renal failure is complex and appears to be related to hyperphosphatemia (and resultant hypocalcemia), deficiency of 1,25-dihydroxy vitamin D due to loss of renal tissue, low calcium intake, decreased calcium absorption, and abnormal parathyroid cell response to extracellular calcium or vitamin D in vitro and in vivo. Patients generally are hypocalcemic or normocalcemic. Aluminum hydroxide, which often was used as a phosphate binder, has been shown to contribute to the osteomalacia observed in this disease. These patients generally are treated medically with a low-phosphate diet, phosphate binders, adequate intake of calcium and 1, 25-dihydroxy vitamin D, and a high-calcium, low-aluminum dialysis bath. Calcimimetics have been shown to control parathyroid hyperplasia and osteitis fibrosa cystica associated with secondary HPT in animal studies and to decrease plasma PTH and total and ionized calcium levels in humans.

Surgical treatment was traditionally recommended for patients with bone pain, pruritus, and (a) a calcium-phosphate product ≥70, (b) calcium >11 mg/dL with markedly elevated PTH, (c) calciphylaxis, (d) progressive renal osteodystrophy, and (e) soft tissue calcification and tumoral calcinosis, despite maximal medical therapy. Assessment of parathyroid mass is thought to be an important factor for predicting the response to medical management. Therefore, some groups recommend parathyroidectomy if the glands are >1 cm (or >500 mm³) on ultrasound. These glands are apparently more likely to have developed nodular hyperplasia and hence might be refractory to medical management. Following the introduction of calcimimetics, there appears to have been a reduction in parathyroidectomy rates. While parathyroidectomy has been reported to maintain biochemical targets for up to 5 years and improve bone density, fracture risk, calcinosis, hemoglobin levels, and even long-term survival, the effects of calcimimetics on these parameters awaits results of various trials. Studies also show that calcimimetics may be less cost-effective in the long-term. The role of parathyroidectomy in the era of calcimimetics will require long-term studies, specifically prospective studies comparing outcomes of calcimimetics with standard therapy versus parathyroidectomy. In general, parathyroidectomy should be considered if PTH levels remain high despite therapy that includes calcimimetics. Calciphylaxis is a rare, limb- and life-threatening complication of secondary HPT characterized by painful (sometimes throbbing), violaceous, and mottled lesions usually on the extremities, which often become necrotic and progress to nonhealing ulcers, gangrene, sepsis, and death. Skin biopsy can be helpful to make the diagnosis. These are critically ill, high-risk patients, but successful parathyroidectomy sometimes relieves symptoms. Not all patients with calciphylaxis will have high PTH levels, and parathyroidectomy should not be undertaken in the absence of documented hyperparathyroidism.

Patients should undergo routine dialysis the day before surgery to correct electrolyte abnormalities. Localization studies are not necessary but can identify ectopic parathyroid glands. A bilateral neck exploration is indicated. The parathyroid glands in secondary HPT are characterized by asymmetric enlargement and nodular hyperplasia. These patients may be treated by subtotal resection, leaving about 50 mg of the most normal parathyroid gland or total parathyroidectomy and autotransplantation of parathyroid tissue into the brachioradialis muscle of the nondominant forearm, with parathyroid cryopreservation. Upper thymectomy usually is performed because 15% to 20% of patients have one or more parathyroid glands situated in the thymus or perithymic fat. Some groups recommend total parathyroidectomy without autotransplantation because it is associated with a lower rate of recurrence. While it may be preferable in patients with calciphylaxis, this procedure is contraindicated in patients eligible for renal transplant. Since the evidence needed to determine the superiority of one approach over another is lacking, the choice of procedure is influenced by surgeon preference and experience and various patient factors, as indicated earlier.⁷⁵

Tertiary Hyperparathyroidism

Generally, renal transplantation is an excellent method of treating secondary HPT, but some patients develop autonomous parathyroid gland function and tertiary HPT. Tertiary HPT can cause problems similar to PHPT, such as pathologic fractures, bone pain and worsened bone disease, renal stones, peptic ulcer disease, pancreatitis, and mental status changes. The transplanted kidney is also at risk from tubulointerstitial calcification and volume depletion.⁷⁶ Similar to patients with secondary HPT, many patients with tertiary HPT are being treated with cinacalcet. Although the drug is effective and well-tolerated in these patients, the long-term effects on kidney allograft function are not known, and many of these patients have persistence of their hypercalcemic symptoms. On the other hand, parathyroidectomy has been shown to lead to a more immediate and dramatic reduction in hypercalcemic symptoms. As such, operative intervention is indicated for symptomatic disease or if autonomous PTH secretion persists for >1 year after a successful transplant in patients who are deemed operative candidates. All parathyroid glands should be identified. The traditional surgical management of these patients consisted of subtotal or total parathyroidectomy with autotransplantation and an upper thymectomy. Some authors suggest that these patients derive similar benefit from excision of only obviously enlarged glands, while avoiding the higher risks of hypocalcemia associated with the former approach. Others recommend that all parathyroid glands be identified and subtotal parathyroidectomy be performed as long-term follow-up studies show that limited excisions in these patients are associated with an up to fivefold increased risk of recurrent or persistent disease. Further studies are needed to define the best operative approach for these patients.

Complications of Parathyroid Surgery

Parathyroidectomy can be accomplished successfully in >95% of patients with minimal mortality and morbidity, provided the procedure is performed by a surgeon experienced in parathyroid surgery. Specific complications include transient and permanent vocal cord palsy and hypoparathyroidism. The latter is more likely to occur in patients who undergo four-gland exploration with biopsies, subtotal resection with an inadequate remnant, or total parathyroidectomy with a failure of autotransplanted tissue. Furthermore, hypocalcemia is more likely to occur in patients with high-turnover bone disease as evidenced by elevated preoperative alkaline phosphatase levels. Vocal cord paralysis and hypoparathyroidism are considered permanent if they persist for >6 months. Fortunately, these complications are rare, occurring in approximately 1% of patients undergoing surgery by experienced parathyroid surgeons.

Patients with symptomatic hypocalcemia or those with calcium levels <8 mg/dL are treated with oral calcium supplementation (up to 1–2 g every 4 hours). 1,25-Dihydroxy vitamin D (calcitriol [Rocaltrol] 0.25–0.5 μg bid) may also be required, particularly in patients with severe hypercalcemia and elevated serum alkaline phosphatase levels preoperatively and with osteitis fibrosa cystica. Intravenous calcium supplementation rarely is needed, except in cases of severe, symptomatic hypocalcemia.

Hypoparathyroidism

Hypocalcemia can be the result of a multitude of conditions, which are listed in Table 38-14. The parathyroid glands may be congenitally absent in DiGeorge syndrome, which also is characterized by lack of thymic development and, therefore, a thymus-dependent lymphoid system. By far, the most common cause of hypoparathyroidism is thyroid surgery, particularly total thyroidectomy with a concomitant central neck dissection. Patients often develop transient hypocalcemia due to ischemia of the parathyroid glands; permanent hypoparathyroidism is rare. Hypoparathyroidism also may occur after parathyroid surgery, which is more likely if patients undergo a subtotal resection or total parathyroidectomy with parathyroid autotransplantation.

Acute hypocalcemia results in decreased ionized calcium and increased neuromuscular excitability. Patients initially develop circumoral and fingertip numbness and tingling. Mental symptoms include anxiety, confusion, and depression. Physical examination reveals positive Chvostek’s sign (contraction of facial muscles elicited by tapping on the facial nerve anterior to the ear) and Trousseau’s sign (carpopedal spasm that is elicited by occluding blood flow to the forearm with a blood pressure cuff for 2–3 minutes). Tetany, which is characterized by tonic-clonic seizures, carpopedal spasm, and laryngeal stridor, may prove fatal and should be avoided. Most patients with postoperative hypocalcemia can be treated with oral calcium and vitamin D supplements; IV calcium infusion is rarely required except in patients with preoperative osteitis fibrosa cystica.

Historical Background

Eustachius provided the first accurate anatomic account of the adrenals in 1563. The anatomic division of the adrenals into the cortex and medulla was described much later, by Cuvier in 1805. Subsequently, Thomas Addison in 1855 described the features of adrenal insufficiency, which still bear his name. DeCreccio provided the first description of congenital adrenal hyperplasia (CAH) occurring in a female pseudohermaphrodite in 1865. Pheochromocytomas were first identified by Frankel in 1885, but were not named as such until 1912 by Pick, who noted the characteristic chromaffin reaction of the tumor cells. Adrenaline was identified as an agent from the adrenal medulla that elevated blood pressure in dogs and was subsequently named epinephrine in 1897. The first successful adrenalectomies for pheochromocytoma were performed by Roux in Switzerland, and Charles Mayo in the United States.

In 1932, Harvey Cushing described 11 patients who had moon facies, truncal obesity, hypertension, and other features of the syndrome that now bears his name. Although several individuals prepared adrenocortical extracts to treat adrenalectomized animals, cortisone was first synthesized by Kendall. Aldosterone was identified in 1952, and the syndrome resulting from excessive secretion of this mineralocorticoid was first described in 1955 by Conn.

Embryology

The adrenal or suprarenal glands are two endocrine organs in one; an outer cortex and an inner medulla, each with distinct embryologic, anatomic, histologic, and secretory features. The cortex originates around the fifth week of gestation from mesodermal tissue near the gonads on the adrenogenital ridge (Fig. 38-37). Therefore, ectopic adrenocortical tissue may be found in the ovaries, spermatic cord, and testes. The cortex differentiates further into a thin, definitive cortex and a thicker, inner fetal cortex. The latter is functional and produces fetal adrenal steroids by the eighth week of gestation, but undergoes involution after birth, resulting in a decrease in adrenal weight during the first three postpartum months. The definitive cortex persists after birth to form the adult cortex over the first 3 years of life. In contrast, the adrenal medulla is ectodermal in origin and arises from the neural crest. At around the same time as cortical development, neural crest cells migrate to the para-aortic and paravertebral areas and toward the medial aspect of the developing cortex to form the medulla. Most extra-adrenal neural tissue regresses but may persist at several sites. The largest of these is located to the left of the aortic bifurcation near the inferior mesenteric artery origin and is designated as the organ of Zuckerkandl. Adrenal medullary tissue also may be found in neck, urinary bladder, and para-aortic regions. Several factors are involved in adrenal development and include insulin-like growth factor 2; gastric inhibitory peptide; and the dosage-sensitive, sex-reversal adrenal hypoplasia (DAX1) gene.

Anatomy

The adrenal glands are paired, retroperitoneal organs located superior and medial to the kidneys at the level of the eleventh ribs. The normal adrenal gland measures 5 × 3 × 1 cm and weighs 4 to 5 g. The right gland is pyramidal shaped and lies in close proximity to the right hemidiaphragm, liver, and inferior vena cava (IVC). The left adrenal is closely associated with the aorta, spleen, and tail of the pancreas. Each gland is supplied by three groups of vessels—the superior adrenal arteries derived from the inferior phrenic artery, the middle adrenal arteries derived from the aorta, and the inferior adrenal arteries derived from the renal artery. Other vessels originating from the intercostal and gonadal vessels may also supply the adrenals. These arteries branch into about 50 arterioles to form a rich plexus beneath the glandular capsule and require careful dissection, ligation, and division during adrenalectomy. In contrast to the arterial supply, each adrenal usually is drained by a single, major adrenal vein. The right adrenal vein is usually short and drains into the IVC, whereas the left adrenal vein is longer and empties into the left renal vein after joining the inferior phrenic vein. Accessory veins occur in 5% to 10% of patients—on the right, these vessels may drain into the right hepatic vein or the right renal vein; on the left, accessory veins may drain directly into the left renal vein. The anatomic relationships of the adrenals and surrounding structures are depicted in Fig. 38-38.

The adrenal cortex appears yellow due to its high lipid content and accounts for about 80% to 90% of the gland’s volume. Histologically, the cortex is divided into three zones—the zona glomerulosa, zona fasciculata, and zona reticularis. The outer area of the zona glomerulosa consists of small cells and is the site of production of the mineralocorticoid hormone, aldosterone. The zona fasciculata is made up of larger cells, which often appear foamy due to multiple lipid inclusions, whereas the zona reticularis cells are smaller. These latter zones are the site of production of glucocorticoids and adrenal androgens. The adrenal medulla constitutes up to 10% to 20% of the gland’s volume and is reddish-brown in color. It produces the catecholamine hormones epinephrine and norepinephrine. The cells of the adrenal medulla are arranged in cords and are polyhedral in shape. They often are referred to as chromaffin cells because they stain specifically with chromium salts.

Adrenal Physiology

Cholesterol, derived from the plasma or synthesized in the adrenal, is the common precursor of all steroid hormones derived from the adrenal cortex. Cholesterol initially is cleaved within mitochondria to 5-δ-pregnolone, which in turn is transported to the smooth endoplasmic reticulum where it forms the substrate for various biosynthetic pathways leading to steroidogenesis (Fig. 38-39).

Mineralocorticoids

The major adrenal mineralocorticoid hormones are aldosterone, 11-deoxycorticosterone (DOC), and cortisol. Cortisol has minimal effects on the kidney due to hormone degradation. Aldosterone secretion is regulated primarily by the renin-angiotensin system. Decreased renal blood flow, decreased plasma sodium, and increased sympathetic tone all stimulate the release of renin from juxtaglomerular cells. Renin, in turn, leads to the production of angiotensin I from its precursor angiotensinogen. Angiotensin I is cleaved by pulmonary angiotensin-converting enzyme (ACE) to angiotensin II; the latter is not only a potent vasoconstrictor, but also leads to increased aldosterone synthesis and release. Hyperkalemia is another potent stimulator of aldosterone synthesis, whereas ACTH, pituitary pro-opiomelanocortin, and antidiuretic hormone are weak stimulators.

Aldosterone is secreted at a rate of 50 to 250 μg/d (depending on sodium intake) and circulates in plasma chiefly as a complex with albumin. Small amounts of the hormone bind to corticosteroid-binding globulin, and approximately 30% to 50% of secreted aldosterone circulates in a free form. The hormone has a half-life of only 15 to 20 minutes and is rapidly cleared via the liver and kidney. A small quantity of free aldosterone also is excreted in the urine. Mineralocorticoids cross the cell membrane and bind to cytosolic receptors. The receptor-ligand complex subsequently is transported into the nucleus where it induces the transcription and translation of specific genes. Aldosterone functions mainly to increase sodium reabsorption and potassium and hydrogen ion excretion at the level of the renal distal convoluted tubule. Less commonly, aldosterone increases sodium absorption in salivary glands and GI mucosal surfaces.

Glucocorticoids

The secretion of cortisol, the major adrenal glucocorticoid, is regulated by ACTH secreted by the anterior pituitary, which, in turn, is under the control of corticotrophin-releasing hormone (CRH) secreted by the hypothalamus. ACTH is a 39-amino-acid protein, which is derived by cleavage from a larger precursor, pro-opiomelanocortin. ACTH is further cleaved into α-melanocyte-stimulating hormone and corticotrophin-like intermediate peptide. ACTH not only stimulates the secretion of glucocorticoids, mineralocorticoids, and adrenal androgens, but is also trophic for the adrenal glands. ACTH secretion may be stimulated by pain, stress, hypoxia, hypothermia, trauma, and hypoglycemia. ACTH secretion fluctuates, peaking in the morning and reaching nadir levels in the late afternoon. Thus, there is a diurnal variation in the secretion of cortisol, with peak cortisol excretion also occurring in the early morning and declining during the day to its lowest levels in the evening (Fig. 38-40). Cortisol controls the secretion of both CRH and ACTH via a negative feedback loop. A similar mechanism leads to the inhibition of CRH secretion by ACTH.

Cortisol is transported in plasma bound primarily to corticosteroid-binding globulin (75%) and albumin (15%). Approximately 10% of circulating cortisol is free and is the biologically active component. The plasma half-life of cortisol is 60 to 90 minutes and is determined by the extent of binding and rate of inactivation. Cortisol is converted to di- and tetrahydrocortisol and cortisone metabolites in the liver and the kidney. The majority (95%) of cortisol and cortisone metabolites are conjugated with glucuronic acid in the liver, thus facilitating their renal excretion. A small amount of unmetabolized cortisol is excreted unchanged in the urine.

Glucocorticoid hormones enter the cell and bind cytosolic steroid receptors. The activated receptor-ligand complex is then transported to the nucleus where it stimulates the transcription of specific target genes via a “zinc finger” DNA binding element. Cortisol also binds the mineralocorticoid receptor with an affinity similar to aldosterone. However, the specificity of mineralocorticoid action is maintained by the production of 11β-hydroxysteroid dehydrogenase, an enzyme that inactivates cortisol to cortisone in the kidney. Glucocorticoids have important functions in intermediary metabolism but also affect connective tissue, bone, immune, cardiovascular, renal, and central nervous systems, as outlined in Table 38-15.

Sex Steroids

Adrenal androgens are produced in the zona fasciculata and reticularis from 17-hydroxypregnenolone in response to ACTH stimulation. They include dehydroepiandrosterone (DHEA) and its sulfated counterpart (DHEAS), androstenedione, and small amounts of testosterone and estrogen. Adrenal androgens are weakly bound to plasma albumin. They exert their major effects by peripheral conversion to the more potent testosterone and dihydrotestosterone but also have weak intrinsic androgen activity. Androgen metabolites are conjugated as glucuronides or sulfates and excreted in the urine. During fetal development, adrenal androgens promote the formation of male genitalia. In normal adult males, the contribution of adrenal androgens is minimal; however, they are responsible for the development of secondary sexual characteristics at puberty. Adrenal androgen excess leads to precocious puberty in boys and virilization, acne, and hirsutism in girls and women.

Catecholamines

Catecholamine hormones (epinephrine, norepinephrine, and dopamine) are produced not only in the central and sympathetic nervous system but also the adrenal medulla. The substrate, tyrosine, is converted to catecholamines via a series of steps shown in Fig. 38-41A. Phenylethanolamine N-methyltransferase, which converts norepinephrine to epinephrine, is only present in the adrenal medulla and the organ of Zuckerkandl. Therefore, the primary catecholamine produced may be used to distinguish adrenal medullary tumors from those situated at extra-adrenal sites. Catecholamines are stored in granules in combination with other neuropeptides, ATP, calcium, magnesium, and water-soluble proteins called chromogranins. Hormonal secretion is stimulated by various stress stimuli and mediated by the release of acetylcholine at the preganglionic nerve terminals. In the circulation, these proteins are bound to albumin and other proteins. Catecholamines are cleared by several mechanisms including reuptake by sympathetic nerve endings, peripheral inactivation by catechol O-methyltransferase and monoamine oxidase, and direct excretion by the kidneys. Metabolism of catecholamines takes place primarily in the liver and kidneys and leads to the formation of metabolites such as metanephrines, normetanephrines, and VMA, which may undergo further glucuronidation or sulfation before being excreted in the urine (Fig. 38-41B).

Adrenergic receptors are transmembrane-spanning molecules that are coupled to G proteins. They may be subdivided into α and β subtypes, which are localized in different tissues, have varying affinity to various catecholamines, and mediate distinct biologic effects (Table 38-16). The receptor affinities for α receptors are—epinephrine > norepinephrine >> isoproterenol; β₁ receptors—isoproterenol > epinephrine = norepinephrine; and β₂ receptors—isoproterenol > epinephrine >> norepinephrine.

Disorders of the Adrenal Cortex

Hyperaldosteronism

Hyperaldosteronism may be secondary to stimulation of the renin-angiotensin system from renal artery stenosis and to low-flow states such as congestive heart failure and cirrhosis. Hyperaldosteronism resulting from these conditions is reversible by treatment of the underlying cause. Primary hyperaldosteronism results from autonomous aldosterone secretion, which, in turn, leads to suppression of renin secretion. Primary aldosteronism usually occurs in individuals between the ages of 30 to 50 years old and accounts for 1% of hypertension cases. It is associated with hypokalemia; however, more patients with Conn’s syndrome are being diagnosed with normal potassium levels. Most cases result from a solitary functioning adrenal adenoma (∼70%) and idiopathic bilateral hyperplasia (30%). Adrenocortical carcinoma and glucocorticoid-suppressible hyperaldosteronism are rare, each accounting for <1% of cases. Glucocorticoid-suppressible hyperaldosteronism is an autosomal dominant form of hypertension in which aldosterone secretion is abnormally regulated by ACTH. This condition is caused by recombinations between linked genes encoding closely related isozymes, 11β-hydroxylase (CYP11B1), and aldosterone synthase (CYP11B2) generating a dysregulated chimeric gene with aldosterone synthase activity.

Symptoms and Signs

Patients typically present with hypertension, which is long-standing, moderate to severe, and may be difficult to control despite multiple-drug therapy. Other symptoms include muscle weakness, polydipsia, polyuria, nocturia, headaches, and fatigue. Weakness and fatigue are related to the presence of hypokalemia.

Diagnostic Studies

Laboratory Studies

Hypokalemia is a common finding, and hyperaldosteronism must be suspected in any hypertensive patient who presents with coexisting spontaneous hypokalemia (K <3.2 mmol/L) or hypokalemia (<3 mmol/L) while on diuretic therapy, despite potassium replacements. However, it is important to note that up to 40% of patients with a confirmed aldosteronoma were normokalemic preoperatively. Once the diagnosis is suspected, further tests are necessary to confirm the diagnosis. Before testing, patients must receive adequate sodium and potassium. Antihypertensive medications should be held, if possible, and spironolactone, β-blockers, ACE inhibitors, and angiotensin II receptor blockers should be avoided. Patients with primary hyperaldosteronism have an elevated plasma aldosterone concentration level with a suppressed plasma renin activity; a plasma aldosterone concentration–to–plasma renin activity ratio of 1:25 to 30 is strongly suggestive of the diagnosis.⁷⁷ False-positive results can occur, particularly in patients with chronic renal failure. Patients with primary hyperaldosteronism also fail to suppress aldosterone levels with sodium loading. This test can be performed by performing a 24-hour urine collection for cortisol, sodium, and aldosterone after 5 days of a high-sodium diet or alternatively giving the patient 2 L of saline while in the supine position, 2 to 3 days after being on a low-sodium diet. Plasma aldosterone level <5 ng/dL or a 24-hour urine aldosterone <14 μg after saline loading essentially rules out primary hyperaldosteronism. Once the biochemical diagnosis is confirmed, further evaluation should be directed at determining which patients have a unilateral aldosteronoma vs. bilateral hyperplasia, because surgery is almost always curative for the former, but usually not the latter. No biochemical studies can make this distinction with 100% sensitivity; thus imaging studies are necessary.

Radiologic Studies

CT scans with 0.5-cm cuts in the adrenal area can localize aldosteronomas with a sensitivity of 90%. A unilateral 0.5- to 2-cm adrenal tumor with a normal-appearing contralateral gland confirms an aldosteronoma in the presence of appropriate biochemical parameters. MRI scans are less sensitive but more specific, particularly if opposed phase chemical shift images are obtained. MRI scans also have increased use in pregnant patients or those unable to tolerate intravenous contrast. If adrenal hyperplasia is suspected, the algorithm depicted in Fig. 38-42 is useful. Selective venous catheterization and adrenal vein sampling for aldosterone have been demonstrated to be 95% sensitive and 90% specific in localizing the aldosteronoma. In this procedure, the adrenal veins are cannulated, and blood samples for aldosterone and cortisol are obtained from both adrenal veins and the vena cava after ACTH administration.⁷⁸ Measurement of cortisol levels is necessary to confirm proper placement of the catheters in the adrenal veins. A greater than fourfold difference in the aldosterone:cortisol ratios between the adrenal veins indicates the presence of a unilateral tumor. Some investigators use this study routinely, but it is invasive, requires an experienced interventional radiologist, and can lead to adrenal vein rupture in approximately 1% of cases. Therefore, most groups advocate use of this modality selectively in ambiguous cases, when the tumor cannot be localized and in patients with bilateral adrenal enlargement to determine whether there is unilateral or bilateral increased secretion of aldosterone. Scintigraphy with ¹³¹I-6β-iodomethyl noriodocholesterol (NP-59) also may be used for the same purpose. Like cholesterol, this compound is taken up by the adrenal cortex, but unlike cholesterol, it remains in the gland without undergoing further metabolism. Adrenal adenomas appear as “hot” nodules with suppressed contralateral uptake, whereas hyperplastic glands show bilaterally increased uptake. This test, however, is not widely available.

Treatment

Preoperatively, control of hypertension and adequate potassium supplementation (to keep K >3.5 mmol/L) are important. Patients generally are treated with spironolactone (an aldosterone antagonist), amiloride (a potassium-sparing diuretic that blocks sodium channels in the distal nephron), nifedipine (a calcium channel blocker), or captopril (an ACE inhibitor). Unilateral tumors producing aldosterone are best managed by adrenalectomy, either by a laparoscopic approach (preferred) or via a posterior open approach. If a carcinoma is suspected because of the large size of the adrenal lesion or mixed hormone secretion, an anterior transabdominal approach is preferred to permit adequate determination of local invasion and distal metastases. Only 20% to 30% of patients with hyperaldosteronism secondary to bilateral adrenal hyperplasia benefit from surgery, and as described, selective venous catheterization is useful to predict which patients will respond. For the other patients, medical therapy with spironolactone, amiloride, or triamterene is the mainstay of management. Glucocorticoid-suppressible hyperaldosteronism is treated by administering exogenous dexamethasone at doses of 0.5 to 1 mg daily. Treatment with spironolactone may help decrease glucocorticoid requirements in this condition and avoid symptoms of Cushing’s syndrome. Postoperatively, some patients experience transient hypoaldosteronism requiring mineralocorticoids for up to 3 months. Rarely, acute Addison’s disease may occur 2 to 3 days after adrenalectomy. Adrenalectomy is >90% successful in improving hypokalemia and about 70% successful in correcting hypertension. Patients who respond to spironolactone therapy and those with a shorter duration of hypertension with minimal renal damage are more likely to achieve improvement in hypertension, whereas male patients, those >50 years old, and those with multiple adrenal nodules, are least likely to benefit from adrenalectomy.

Cushing’s Syndrome

Cushing described patients with a peculiar fat deposition, amenorrhea, impotence (in men), hirsutism, purple striae, hypertension, diabetes, and other features that constitute the syndrome (Fig. 38-43). He also recognized that several of these patients had basophilic tumors of the pituitary gland and concluded that these tumors produced hormones that caused adrenocortical hyperplasia, thus resulting in the manifestations of the syndrome. Today, the term Cushing’s syndrome refers to a complex of symptoms and signs resulting from hypersecretion of cortisol regardless of etiology. In contrast, Cushing’s disease refers to a pituitary tumor, usually an adenoma, which leads to bilateral adrenal hyperplasia and hypercortisolism. Cushing’s syndrome (endogenous) is a rare disease, affecting 10 in 1 million individuals. It is more common in adults but may occur in children. Women are more commonly affected (male:female ratio 1:8). Although most individuals have sporadic disease, Cushing’s syndrome may be found in MEN1 families and can result from ACTH-secreting pituitary tumors, primary adrenal neoplasms, or an ectopic ACTH-secreting carcinoid tumor (more common in men) or bronchial adenoma (more common in women).

Cushing’s syndrome may be classified as ACTH-dependent or ACTH-independent (Table 38-17). The most common cause of hypercortisolism is exogenous administration of steroids. However, approximately 70% of cases of endogenous Cushing’s syndrome are caused by an ACTH-producing pituitary tumor. Primary adrenal sources (adenoma, hyperplasia, and carcinoma) account for about 20% of cases, and ectopic ACTH-secreting tumors account for <10% of cases. CRH also may be secreted ectopically in bronchial carcinoid tumors, pheochromocytomas, and other tumors. These patients are difficult to distinguish from those with ectopic ACTH production, but can be diagnosed by determining CRH levels. Patients with major depression, alcoholism, pregnancy, chronic renal failure, or stress also may have elevated cortisol levels and symptoms of hypercortisolism. However, these manifestations resolve with treatment of the underlying disorder, and these patients are deemed to have pseudo-Cushing’s syndrome.

Primary adrenal hyperplasia may be micronodular, macronodular, or massively macronodular. Adrenal hyperplasia resulting from ACTH stimulation usually is macronodular (3-cm nodules). Primary pigmented nodular adrenocortical disease is a rare cause of ACTH-independent Cushing’s syndrome, which is characterized by the presence of small (<5 mm), black adrenal nodules. Primary pigmented nodular adrenocortical disease may be associated with Carney complex (atrial myxomas, schwannomas, and pigmented nevi) and is thought to be immune related.

Symptoms and Signs

The classical features of Cushing’s syndrome are listed in Table 38-18. Early diagnosis of this disease requires a thorough knowledge of these manifestations, coupled with a high clinical suspicion. In some patients, symptoms are less pronounced and may be more difficult to recognize, particularly given their diversity and the absence of a single defining symptom or sign. Progressive truncal obesity is the most common symptom, occurring in up to 95% of patients. This pattern results from the lipogenic action of excessive corticosteroids centrally and catabolic effects peripherally, along with peripheral muscle wasting. Fat deposition also occurs in unusual sites, such as the supraclavicular space and posterior neck region, leading to the so-called buffalo hump. Purple striae are often visible on the protuberant abdomen. Rounding of the face leads to moon facies, and thinning of subcutaneous tissues leads to plethora. There is an increase in fine hair growth on the face, upper back, and arms, although true virilization is more commonly seen with adrenocortical cancers. Endocrine abnormalities include glucose intolerance, amenorrhea, and decreased libido or impotence. In children, Cushing’s syndrome is characterized by obesity and stunted growth. Patients with Cushing’s disease also may present with headaches, visual field defects, and panhypopituitarism. Hyperpigmentation of the skin, if present, suggests an ectopic ACTH-producing tumor with high levels of circulating ACTH.

Diagnostic Tests

The aims of diagnostic tests in the evaluation of patients suspected of having Cushing’s syndrome are twofold: to confirm the presence of Cushing’s syndrome and to determine its etiology (Fig. 38-44).

Laboratory Studies

Cushing’s syndrome is characterized by elevated glucocorticoid levels that are not suppressible by exogenous hormone administration and loss of diurnal variation. This phenomenon is used to screen patients using the overnight low-dose dexamethasone suppression test. In this test, 1 mg of a synthetic glucocorticoid (dexamethasone) is given at 11 p.m. and plasma cortisol levels are measured at 8 a.m. the following morning. Physiologically normal adults suppress cortisol levels to <3 μg/dL, whereas most patients with Cushing’s syndrome do not. False-negative results may be obtained in patients with mild disease; therefore, some authors consider the test positive only if cortisol levels are suppressed to <1.8 μg/dL. False-positive results can occur in up to 3% of patients with chronic renal failure, depression, or those taking medications such as phenytoin, which enhance dexamethasone metabolism. In patients with a negative test but a high clinical suspicion, the classic low-dose dexamethasone (0.5 mg every 6 hours for eight doses, or 2 mg over 48 hours) suppression test or urinary cortisol measurement should be performed. Measurement of elevated 24-hour urinary cortisol levels is a very sensitive (95%–100%) and specific (98%) modality of diagnosing Cushing’s syndrome and is particularly useful for identifying patients with pseudo-Cushing’s syndrome. A urinary cortisol-free excretion of less than 100 μg/dL (in most laboratories) rules out hypercortisolism. Recently, salivary cortisol measurements using commercially available kits also have demonstrated superior sensitivity in diagnosing Cushing’s syndrome and are being increasingly used. Overall, 24-hour urinary tests for free cortisol and the overnight dexamethasone suppression test at the 5 μg/dL cutoff have the highest specificity for the diagnosis of Cushing’s syndrome.⁷⁹

Once a diagnosis of hypercortisolism is established, further testing is aimed at determining whether it is ACTH-dependent or ACTH-independent Cushing’s syndrome. This is best accomplished by measurement of plasma ACTH levels (normal 10–100 pg/mL). Elevated ACTH levels are found in patients with adrenal hyperplasia due to Cushing’s disease (15–500 pg/mL) and those with CRH-secreting tumors, but the highest levels are found in patients with ectopic sources of ACTH (>1000 pg/mL). In contrast, ACTH levels are characteristically suppressed (<5 pg/mL) in patients with primary cortisol-secreting adrenal tumors. The high-dose dexamethasone suppression test is used to distinguish between the causes of ACTH-dependent Cushing’s syndrome (pituitary vs. ectopic). The standard test (2 mg dexamethasone every 6 hours for 2 days) or the overnight test (8 mg) may be used, with 24-hour urine collections for cortisol and 17-hydroxy steroids performed over the second day. Failure to suppress urinary cortisol by 50% confirms the diagnosis of an ectopic ACTH-producing tumor. Patients suspected of having ectopic tumors should also undergo testing for MTC and pheochromocytoma. Bilateral petrosal vein sampling also is helpful for determining whether the patient has Cushing’s disease or ectopic Cushing’s syndrome.

The CRH test also is helpful in determining the etiology of Cushing’s syndrome. Ovine CRH (1 μg/kg) is administered intravenously, followed by serial measurements of ACTH and cortisol at 15-minute intervals for 1 hour. Patients with a primary adrenal hypercortisolism exhibit a blunted response (ACTH peak <10 pg/mL), whereas those with ACTH-dependent Cushing’s syndrome demonstrate a higher elevation of ACTH (>30 pg/mL). CRH stimulation also can enhance the usefulness of petrosal vein sampling. Patients with pituitary tumors also have a higher peak ACTH than those with ectopic ACTH-producing tumors.

Radiologic Studies

CT and MRI scans of the abdomen can identify adrenal tumors with 95% sensitivity. They also are helpful in distinguishing adrenal adenomas from carcinomas, as discussed in the subsequent section Adrenocortical Cancer. MRI scans have the added advantage of allowing assessment of vascular anatomy. Adrenal adenomas appear darker than the liver on T2-weighted imaging. Radioscintigraphic imaging of the adrenals using NP-59 also can be used to distinguish adenoma from hyperplasia. Reports suggest that “cold” adrenal nodules are more likely to be cancerous, although this distinction is not absolute. NP-59 scanning is most useful in identifying patients with an adrenal source of hypercortisolism and primary pigmented micronodular hyperplasia.

Thin-section head CT scans are 22% sensitive, and contrast-enhanced brain MRI scans are 33% to 67% sensitive at identifying pituitary tumors. Inferior petrosal sinus sampling for ACTH before and after CRH injection has been helpful in this regard and has a sensitivity approaching 100%. In this study, catheters are placed in both internal jugular veins and a peripheral vein. A ratio of petrosal to peripheral vein ACTH level of >2 in the basal state and >3 after CRH stimulation is diagnostic of a pituitary tumor. In patients suspected of having ectopic ACTH production, CT or MRI scans of the chest and anterior mediastinum are performed first, followed by imaging of the neck, abdomen, and pelvis if the initial studies are negative.

Treatment

Laparoscopic adrenalectomy is the treatment of choice for patients with adrenal adenomas. Open adrenalectomy is reserved for large tumors (≥6 cm) or those suspected to be adrenocortical cancers. Bilateral adrenalectomy is curative for primary adrenal hyperplasia.

The treatment of choice in Cushing’s disease is transsphenoidal excision of the pituitary adenoma, which is successful in 80% of patients. Pituitary irradiation has been used for patients with persistent or recurrent disease after surgery. However, it is associated with a high rate of panhypopituitarism, and some patients develop visual deficits. This has led to increased use of stereotactic radiosurgery, which uses CT guidance to deliver high doses of radiotherapy to the tumor (photon or gamma knife) and also bilateral laparoscopic adrenalectomy. Patients who fail to respond to either treatment are candidates for pharmacologic therapy with adrenal inhibitors (medical adrenalectomy) such as ketoconazole, metyrapone, or aminoglutethimide.

Patients with ectopic ACTH production are best managed by treating the primary tumor, including recurrences, if possible. Medical or bilateral laparoscopic adrenalectomy has been used to palliate patients with unresectable disease and those whose ectopic ACTH-secreting tumor cannot be localized.

Patients undergoing surgery for a primary adrenal adenoma secreting glucocorticoids require preoperative and postoperative steroids due to suppression of the contralateral adrenal gland. These patients are also at increased predisposition for infectious and thromboembolic complications, the latter due to a hypercoagulable state resulting from an increase in clotting factors including factor VIII and von Willebrand’s factor complex, and by impaired fibrinolysis. Duration of steroid therapy is determined by the ACTH stimulation test. Exogenous steroids may be needed for up to 2 years but are needed indefinitely in patients who have undergone bilateral adrenalectomy. This latter group of patients also may require mineralocorticoid replacement therapy. Typical replacement doses include hydrocortisone (10–20 mg every morning and 5–10 mg every evening) and fludrocortisone (0.05–0.1 mg/d every morning).

Adrenocortical Cancer

Adrenal carcinomas are rare neoplasms with a worldwide incidence of two per 1 million. These tumors have a bimodal age distribution, with an increased incidence in children and adults in the fourth and fifth decades of life. The majority are sporadic, but adrenocortical carcinomas also occur in association with germline mutations of p53 (Li-Fraumeni syndrome) and MENIN (multiple endocrine neoplasia type 1) genes. Loci on 11p (Beckwith-Wiedemann syndrome), 2p (Carney complex), and 9q also have been implicated. Somatic p53 mutations are present in up to 33% of tumors, and LOH at the p53 locus has been reported in >85% of adrenocortical carcinomas. In addition, Insulin-like growth factor II is overexpressed in 90% of tumors, and approximately 30% harbor somatic activating mutations in the β-catenin gene.⁸⁰

Symptoms and Signs

Approximately 50% of adrenocortical cancers are nonfunctioning.⁸¹ The remaining secrete cortisol (30%), androgens (20%), estrogens (10%), aldosterone (2%), or multiple hormones (35%). Patients with functioning tumors often present with the rapid onset of Cushing’s syndrome accompanied by virilizing features. Nonfunctioning tumors more commonly present with an enlarging abdominal mass and abdominal or back pain. Rarely, weight loss, anorexia, and nausea may be present.

Diagnostic Tests

Diagnostic evaluation of these patients begins with measurement of serum electrolyte levels to rule out hypokalemia, urinary catecholamines to rule out pheochromocytomas, an overnight 1-mg dexamethasone suppression test, and a 24-hour urine collection for cortisol and 17-ketosteroids to rule out Cushing’s syndrome.

CT and MRI scans are useful to image these tumors (Fig. 38-45). The size of the adrenal mass on imaging studies is the single most important criterion to help diagnose malignancy. In the series reported by Copeland, 92% of adrenal cancers were >6 cm in diameter.⁸² The sensitivity, specificity, and likelihood ratio of tumor size in predicting malignancy (based on Surveillance, Epidemiology, and End Results program data) were recently reported as 96%, 51%, and 2 for tumors ≥4 cm, and 90%, 78%, and 4.1 for tumors ≥6 cm.⁸³ Other CT imaging characteristics suggesting malignancy include tumor heterogeneity, irregular margins, and the presence of hemorrhage and adjacent lymphadenopathy or liver metastases. Moderately bright signal intensity on T2-weighted images (adrenal mass–to–liver ratio 1.2:2.8), significant lesion enhancement, and slow washout after injection of gadolinium contrast also indicate malignancy, as does evidence of local invasion into adjacent structures such as the liver, blood vessels (IVC), and distant metastases. FDG-PET or PET-CT scans may have some utility in distinguishing benign from malignant lesions, as discussed in the section on incidentalomas. Once adrenal cancer is diagnosed, CT scans of the chest and pelvis or FDG-PET or PET-CT scans are performed for staging. The tumor-node-metastasis (TNM) staging system for adrenocortical carcinoma is depicted in Table 38-19. Up to 70% of patients present with stage III or IV disease.

Pathology

Most adrenocortical cancers are large, weighing between 100 and 1000 g. On gross examination, areas of hemorrhage and necrosis often are evident. Microscopically, cells are hyperchromatic and typically have large nuclei and prominent nucleoli. It is very difficult to distinguish benign adrenal adenomas from carcinomas by histologic examination alone. Capsular or vascular invasion is the most reliable sign of cancer. Weiss and associates studied a combination of nine criteria for their usefulness in distinguishing malignant from benign adrenal tumors: nuclear grade III or IV; mitotic rate greater than 5 per 50 high-power fields; atypical mitoses; clear cells comprising 25% or less of the tumor; a diffuse architecture; microscopic necrosis; and invasion of venous, sinusoidal, and capsular structure. Tumors with four or more of these criteria were likely to metastasize and/or recur.⁸⁴ Rarely, the diagnosis of malignancy of a completely resected adrenal tumor is often only made in retrospect by the finding of metastatic disease many years later. Molecular markers such as Ki67 (indicating proliferative activity) can also be useful in this regard.

Treatment

The most important predictor of survival in patients with adrenal cancer is the adequacy of resection. Patients who undergo complete resection have 5-year actuarial survival rates ranging from 32% to 48%, whereas median survival is <1 year in those undergoing incomplete excision. Therefore, adrenocortical carcinomas are treated by excision of the tumor en bloc with any contiguously involved lymph nodes or organs such as the diaphragm, kidney, pancreas, liver, or IVC. This is best accomplished by open adrenalectomy via a generous subcostal incision or a thoracoabdominal incision (on the right side). The incisions should permit wide exposure, minimize chances of capsule rupture and tumor spillage, and allow vascular control of the aorta, IVC, and renal vessels, as needed.

Mitotane or o,p-DDD or 1,1-dichloro-2-(o-chlorophenyl)-2-(p-chlorophenyl) ethane, which is a derivative of the insecticide DDT, has adrenolytic activity and has been used in the adjuvant setting and for the treatment of unresectable or metastatic disease. However, the therapeutic effectiveness is conflicting, and consistent improvement in survival rates is lacking. Moreover, the drug is associated with significant GI and neurologic side effects, particularly at the effective doses of 2 to 6 g/d. Terzolo and associates retrospectively evaluated the use of mitotane in the adjuvant setting and reported significantly increased recurrence-free survival in the treatment group.⁸⁵ However, the routine use of this medication awaits evaluation in randomized, controlled trials. Determination of blood mitotane levels is helpful to ascertain whether therapeutic and nontoxic levels are present. Adrenocortical tumors commonly metastasize to the liver, lung, and bone.

Surgical debulking is recommended for isolated, recurrent disease and has been demonstrated to prolong survival. Systemic chemotherapeutic agents used in this tumor include etoposide, cisplatin, doxorubicin, and, more recently, paclitaxel, but consistent responses are rare, possibly due to the expression of the multidrug resistance gene (MDR-1) in tumor cells. In vitro data indicate that mitotane may be able to reverse this resistance when combined with various chemotherapeutic agents. Results from the First International Randomized Trial in Advanced or Metastatic Adrenocortical Carcinoma Treatment (FIRM-ACT) showed that patients receiving etoposide, doxorubicin, cisplatin, and mitotane had better response rates and progression-free survival rates than patients receiving streptozotocin and mitotane.⁷² There has been recent interest in the use of suramin, a growth factor inhibitor, as therapy for adrenocortical carcinoma; however, this requires further study, particularly because this drug may be associated with significant neurotoxicity. Gossypol, a naturally occurring insecticide (from the cotton plant Gossypium species), also appears to inhibit the growth of adrenocortical cancer cell lines and tumors in vivo. However, poor response rates combined with high death rates in limited clinical studies have reduced enthusiasm for this agent. Adrenocortical cancers also are relatively insensitive to conventional external-beam radiation therapy. However, this modality is used in the palliation of bony metastases. Ketoconazole, metyrapone, or aminoglutethimide may also be useful in controlling steroid hypersecretion.

Sex Steroid Excess

Adrenal adenomas or carcinomas that secrete adrenal androgens lead to virilizing syndromes. Although women with virilizing tumors develop hirsutism, amenorrhea, infertility, and other signs of masculinization, such as increased muscle mass, deepened voice, and temporal balding, men with these tumors are more difficult to diagnose and, hence, usually present with disease in advanced stages. Children with virilizing tumors have accelerated growth, premature development of facial and pubic hair, acne, genital enlargement, and deepening of their voice. Feminizing adrenal tumors are less common and occur in men in the third to fifth decades of life. These tumors lead to gynecomastia, impotence, and testicular atrophy. Women with these tumors develop irregular menses or dysfunctional uterine bleeding. Vaginal bleeding may occur in postmenopausal women. Girls with these tumors experience precocious puberty with breast enlargement and early menarche.

Diagnostic Tests

Virilizing tumors produce excessive amounts of the androgen precursor, DHEA, which can be measured in plasma or urine as 17-ketosteroids. Patients with feminizing tumors also have elevated urinary 17-ketosteroids in addition to increased estrogen levels. Androgen-producing tumors often are associated with production of other hormones such as glucocorticoids.

Treatment

Virilizing and feminizing tumors are treated by adrenalectomy. Malignancy is difficult to diagnose histologically but is suggested by the presence of local invasion, recurrence, or distal metastases. Adrenolytic drugs such as mitotane, aminoglutethimide, and ketoconazole may be useful in controlling symptoms in patients with metastatic disease.

Congenital Adrenal Hyperplasia

CAH refers to a group of disorders that result from deficiencies or complete absence of enzymes involved in adrenal steroidogenesis. 21-Hydroxylase (CYP21A2) deficiency is the most common enzymatic defect, accounting for >90% of cases of CAH. This deficiency prevents the production of 11-deoxycortisol and 11-DOC from progesterone precursors. Deficiency of glucocorticoids and aldosterone leads to elevated ACTH levels and overproduction of adrenal androgens and corticosteroid precursors such as 17-hydroxyprogesterone and Δ⁴-androstenedione. These compounds are converted to testosterone in the peripheral tissues, thereby leading to virilization. Complete deficiency of 21-hydroxylase presents at birth with virilization, diarrhea, hypovolemia, hyponatremia, hyperkalemia, and hyperpigmentation. Partial enzyme deficiency may present at birth or later with virilizing features. These patients are less prone to the salt wasting that characterizes complete enzyme deficiency. 11β-Hydroxylase deficiency is the second most common form of CAH and leads to hypertension (from 11-DOC accumulation), virilization, and hyperpigmentation. Other enzyme deficiencies include 3β-hydroxydehydrogenase and 17-hydroxylase deficiency. Congenital adrenal lipoid hyperplasia is the most severe form of CAH, which is caused by cholesterol desmolase deficiency. It leads to the disruption of all steroid biosynthetic pathways, thus resulting in a fatal salt-wasting syndrome in phenotypic female patients.

Diagnostic Tests

The particular enzyme deficiency can be diagnosed by karyotype analysis and measurement of plasma and urinary steroids. The most common enzyme deficiency, absence of 21-hydroxylase, leads to increased plasma 17-hydroxyprogesterone and progesterone levels, because these compounds cannot be converted to 11-deoxycortisol and 11-DOC, respectively. 11β-Hydroxylase deficiency is the next most common disorder and results in elevated plasma 11-DOC and 11-deoxycortisol. Urinary 17-hydroxyprogesterone, androgens, and 17-ketosteroids also are elevated. The dexamethasone suppression test (2–4 mg divided qid for 7 days) can be used to distinguish adrenal hyperplasia from neoplasia. CT, MRI, and iodocholesterol scans generally are used to localize the tumors.

Treatment

Patients with CAH traditionally have been managed medically, with cortisol and mineralocorticoid replacement to suppress the hypothalamic-pituitary-adrenal axis. However, the doses of steroids required often are supraphysiologic and lead to iatrogenic hypercortisolism. More recently, bilateral laparoscopic adrenalectomy has been proposed as an alternative treatment for this disease and has been successfully performed in a limited number of patients for various forms of CAH.

Disorders of the Adrenal Medulla

Pheochromocytomas

Pheochromocytomas are rare tumors with prevalence rates ranging from 0.3% to 0.95% in autopsy series and approximately 1.9% in series using biochemical screening. They can occur at any age, with a peak incidence in the fourth and fifth decades of life, and have no gender predilection. Extra-adrenal tumors, also called functional paragangliomas, may be found at sites of sympathetic ganglia in the organ of Zuckerkandl, neck, mediastinum, abdomen, and pelvis. Pheochromocytomas often are called the 10 percent tumor because 10% are bilateral, 10% are malignant, 10% occur in pediatric patients, 10% are extra-adrenal, and 10% are familial.

Pheochromocytomas occur in families with MEN2A and MEN2B in approximately 50% of patients. Both syndromes are inherited in an autosomal dominant fashion and are caused by germline mutations in the RET proto-oncogene. Another syndrome with an increased risk of pheochromocytomas is von Hippel-Lindau (VHL) disease, which also is inherited in an autosomal dominant manner. This syndrome also includes retinal angioma, hemangioblastomas of the central nervous system, renal cysts and carcinomas, pancreatic cysts, and epididymal cystadenomas. The incidence of pheochromocytomas in the syndrome is approximately 14%. The gene causing VHL has been mapped to chromosome 3p and is a tumor suppressor gene. Pheochromocytomas also are included within the tumor spectrum of neurofibromatosis type 1 (NF1 gene) and other neuroectodermal disorders (Sturge-Weber syndrome and tuberous sclerosis), Carney’s syndrome (gastric epithelioid leiomyosarcoma, pulmonary chondroma, and extra-adrenal paraganglioma), MEN1 syndrome, and the familial paraganglioma and pheochromocytoma syndrome caused by mutations in the succinyl dehydrogenase family of genes (SDHB, SDHC, and SDHD), which comprise portions of the mitochondrial complex II.⁸⁷ More recently, mutations in SDHA and SDH5 have also been identified. Additional susceptibility loci include TMEM127 (involved in the mTORC1 signaling pathway) and MAX (myc-associated factor X).

Symptoms and Signs

Headache, palpitations, and diaphoresis constitute the “classic triad” of pheochromocytomas. Symptoms such as anxiety, tremulousness, paresthesias, flushing, chest pain, shortness of breath, abdominal pain, nausea, vomiting, and others are nonspecific and may be episodic in nature. Cardiovascular complications such as myocardial infarction and cerebrovascular accidents may ensue. These symptoms can be incited by a range of stimuli including exercise, micturition, and defecation. The most common clinical sign is hypertension. Pheochromocytomas are one of the few curable causes of hypertension and are found in 0.1% to 0.2% of hypertensive patients. Hypertension related to this tumor may be paroxysmal with intervening normotension, sustained with paroxysms or sustained hypertension alone. Sudden death may occur in patients with undiagnosed tumors who undergo other surgeries or biopsy.

Diagnostic Tests

Biochemical Studies

Pheochromocytomas are diagnosed by testing 24-hour urine samples for catecholamines and their metabolites as well as by determining plasma metanephrine levels. Urinary metanephrines are 98% sensitive and also about 98% specific for pheochromocytomas, whereas VMA measurements are slightly less sensitive and specific. False-positive VMA tests may result from ingestion of caffeine, raw fruits, or medications (α-methyldopa). Fractionated urinary catecholamines (norepinephrine, epinephrine, and dopamine) also are very sensitive but less specific for pheochromocytomas. Because extra-adrenal sites lack phenylethanolamine N-methyltransferase, these tumors secrete norepinephrine, whereas epinephrine is the main hormone secreted from adrenal pheochromocytomas. Many physiologic and pathologic states can alter the levels of plasma catecholamines. Hence, they often are thought to be less accurate than urinary tests. Both epinephrine and norepinephrine should be measured, as tumors often secrete one or the other hormone. Sensitivities of 85% and specificities of 95% have been reported using cutoff values of 2000 pg/mL for norepinephrine and 200 pg/mL for epinephrine. Clonidine is an agent that suppresses neurogenically mediated catecholamine excess but not secretion from pheochromocytomas. A normal clonidine suppression test is defined by a decrease of basal catecholamine levels to <500 pg/mL within 2 to 3 hours after an oral dose of 0.3 mg of clonidine. Chromogranin A is a monomeric, acidic protein, which is stored in the adrenal medulla and other neuroendocrine tumors and released along with catecholamine hormones. It has been reported to have a sensitivity of 83% and a specificity of 96% and is useful in conjunction with catecholamine measurement for diagnosing pheochromocytomas. Some studies have shown that plasma metanephrines should be the first-line test to identify pheochromocytomas, as the predictive value of a negative test is very high and normal levels exclude pheochromocytoma in patients with preclinical disease or dopamine secreting tumors. Although sensitivities of 96% to 100% have been reported, specificity is lower at 85% to 89% and may be much lower at 77% in elderly patients. Although attractive because of the simplicity of a blood test, measurement of plasma metanephrines is generally reserved for cases for which there is a high index of suspicion.

Radiologic Studies

Radiologic studies are useful to localize tumors and to assess the extent of spread once the diagnosis has been made with biochemical tests. CT scans are 85% to 95% sensitive and 70% to 100% specific for pheochromocytomas (Fig. 38-46A). The scans should be performed without contrast to minimize the risk of precipitating a hypertensive crisis, although some recent studies suggest that intravenous contrast may be used. Images should include the region from the diaphragm to the aortic bifurcation so as to include the organ of Zuckerkandl. CT scans do not provide functional information and cannot definitively diagnose pheochromocytomas. MRI scans are 95% sensitive and almost 100% specific for pheochromocytomas because these tumors have a characteristic appearance on T2-weighted images or after gadolinium. MRI is also the study of choice in pregnant women as there is no risk of radiation exposure. Metaiodobenzylguanidine (MIBG) is taken up and concentrated by vesicles in the adrenal medullar cells because its structure is similar to norepinephrine. Normal adrenal medullary tissue does not take up appreciable MIBG. ¹³¹I-radiolabeled MIBG is, therefore, useful for localizing pheochromocytomas (Fig. 38-46B), especially those in ectopic positions. This test has a reported sensitivity of 77% to 89% and specificity ranging from 88% to 100%.

Treatment

The medical management of pheochromocytomas is aimed chiefly at blood pressure control and volume repletion. Irreversible, long-acting α-blockers such as phenoxybenzamine are started 1 to 3 weeks before surgery at doses of 10 mg twice daily, which may be increased to 300 to 400 mg/d with rehydration. Patients should be warned about orthostatic hypotension. β-Blockers such as propranolol at doses of 10 to 40 mg every 6 to 8 hours often need to be added preoperatively in patients who have persistent tachycardia and arrhythmias. β-Blockers should only be instituted after adequate α-blockade and hydration to avoid the effects of unopposed α stimulation (i.e., hypertensive crisis and congestive heart failure) and are typically initiated 3 to 4 days preoperatively. Patients also should be volume repleted preoperatively to avoid postoperative hypotension, which ensues with the loss of vasoconstriction after tumor removal. Other α-blockers such as prazosin, terazosin, and doxazosin, which are selective α₁-adrenergic blockers, have a better side effect profile and are preferable to phenoxybenzamine when long-term pharmacologic therapy is needed, as in patients with metastatic pheochromocytoma. Nicardipine is the most commonly used calcium channel blocker and inhibits norepinephrine-mediated calcium transport into vascular smooth muscle. When used as the primary mode of treatment, it appears to be just as effective as α- and β-blockade preoperatively. In some patients, catecholamine-synthesis inhibitors such as α-methyl-p-l-tyrosine (metyrosine) may need to be added if standard α- and β-blockade is poorly tolerated or is ineffective in reaching target blood pressure and when moderate intraoperative tumor manipulation is anticipated.

Adrenalectomy is the treatment of choice for patients with pheochromocytoma. The chief goal of surgery is to resect the tumor completely with minimal tumor manipulation or rupture of the tumor capsule. Surgery should be performed with both noninvasive and invasive monitors, including an arterial line and central venous lines. In patients with congestive heart failure or underlying coronary artery disease, Swan-Ganz catheters may be necessary. Stress must be avoided during anesthesia induction, and use of inhaled agents like isoflurane and enflurane are preferred because they have minimal cardiac depressant effects. Fentanyl, ketamine, and morphine should be avoided as they can potentially stimulate catecholamine release from the tumor. The common medications used for intraoperative blood pressure control include nitroprusside, nitroglycerin, phentolamine, and nicardipine. Intraoperative arrhythmias are best managed by short-acting β-blockers such as esmolol. Adrenalectomy usually was performed via an open anterior approach to facilitate detection of bilateral tumors, extra-adrenal lesions, or metastatic lesions. However, most pheochromocytomas <5 cm in diameter can be safely resected laparoscopically. Postoperatively, these patients are prone to hypotension due to loss of adrenergic stimulation and consequent vasodilatation and therefore need large volume resuscitation.

Hereditary Pheochromocytomas

Inherited pheochromocytomas tend to be multiple and bilateral. Generally, unilateral adrenalectomy is recommended in the absence of obvious lesions in the contralateral adrenal gland because of Addison’s disease, requiring lifelong steroid replacement in patients undergoing bilateral adrenalectomy. For patients with tumors in both adrenal glands, cortical-sparing subtotal adrenalectomy may preserve adrenocortical function and avoid the morbidity of bilateral total adrenalectomy. Laparoscopic subtotal adrenalectomy has been shown to provide short-term clinical results comparable to total adrenalectomy, with reduced surgical morbidity.⁸⁸ However, these patients remain at risk for recurrent pheochromocytoma, which has been reported in 20% of patients with VHL disease a median of 40 months after partial adrenalectomy, and in 33% of MEN2 patients followed for 54 to 88 months after surgery. Autotransplantation of adrenocortical tissue after total adrenalectomy may be another option for these patients. However, the transplanted cortical tissue rarely provides full function, and steroid replacement usually is required.

Malignant Pheochromocytomas

Approximately 12% to 29% of pheochromocytomas are malignant, and these tumors are associated with decreased survival. There are no definitive histologic criteria defining malignant pheochromocytomas. In fact, pleomorphism, nuclear atypia, and abundant mitotic figures are seen in benign tumors. Capsular and vascular invasion may be seen in benign lesions as well. Malignancy usually is diagnosed when there is evidence of invasion into surrounding structures or distant metastases. The most common sites for metastatic disease are bone, liver, regional lymph nodes, lung, and peritoneum, although the brain, pleura, skin, and muscles may also occasionally be involved. Some studies also suggest that older patient age and larger tumors are associated with a higher risk of malignancy. Although risk of malignancy increases with size for all pheochromocytomas, size does not seem to reliably predict malignancy in pheochromocytomas with local disease only.⁸⁹ Given this difficulty defining malignancy clinically (in the absence of metastatic disease), a number of other features such as DNA ploidy, tumor size, and necrosis, neuropeptide Y mRNA expression, and serum neuron-specific enolase expression have been studied. Malignant pheochromocytomas are more likely to express p53 and bcl-2 and have activated telomerase. Recent data suggest that flow cytometry and molecular markers such as expression of Ki-67, tissue inhibitor of metalloproteinase, and COX-2 also have shown some use in determining malignancy. When pheochromocytomas develop in the MEN syndromes, they rarely are malignant. In contrast, patients with germline SDHB mutations appear to have a higher propensity for extra-adrenal and malignant tumors. In general, soft tissue lesions are treated with resection if feasible. External-beam radiation can be used for unresectable lesions or symptomatic skeletal metastases. Therapeutic ¹³¹I-MIBG irradiation may be useful in patients with diffuse disease showing ¹²³-I-MIBG uptake on a diagnostic scan. If the tumor is positive on somatostatin receptor imaging, long-acting octreotide may be used.

The Adrenal Incidentaloma

Adrenal lesions discovered during imaging performed for unrelated reasons are referred to as incidentalomas. This definition excludes tumors discovered on imaging studies performed for evaluating symptoms of hormone hypersecretion or staging patients with known cancer. The incidence of these lesions identified by CT scans ranges from 0.4% to 4.4%.

Differential Diagnosis

The differential diagnosis of adrenal incidentalomas is shown in Table 38-20. Nonfunctional cortical adenomas account for the majority (36%–94%) of adrenal incidentalomas in patients without a history of cancer. In a series of patients from the Mayo Clinic, no nonfunctional lesion progressed to cause clinical or biochemical abnormalities. However, other studies indicate that 5% to 20% of patients with apparently nonfunctioning cortical adenomas have underlying, subtle abnormalities of glucocorticoid secretion, and a rare benign-appearing incidentaloma is a cancer.

By definition, patients with incidentalomas do not have clinically overt Cushing’s syndrome, but subclinical Cushing’s syndrome is estimated to occur in approximately 8% of patients. This disorder is characterized by subtle features of cortisol excess, such as weight gain, skin atrophy, facial fullness, diabetes, and hypertension, accompanied by loss of normal diurnal variation in cortisol secretion, autonomous cortisol secretion, and resistance to suppression by dexamethasone. Total cortisol produced and 24-hour urinary cortisol levels may be normal. Examination of the natural history of subclinical Cushing’s syndrome indicates that, although most patients remain asymptomatic, some do progress to clinically evident Cushing’s syndrome. Furthermore, cases of postoperative adrenal crisis from unrecognized suppression of the contralateral adrenal have been reported, making preoperative identification of this condition imperative, particularly in the era of early discharge following laparoscopic adrenalectomy.

The adrenal is a common site of metastases of lung and breast tumors, melanoma, renal cell cancer, and lymphoma. In patients with a history of nonadrenal cancer and a unilateral adrenal mass, the incidence of metastatic disease has been reported to range from 32% to 73%. Myelolipomas are benign, biochemically nonfunctioning lesions composed of elements of hematopoietic and mature adipose tissue, which are rare causes of adrenal incidentaloma. Other less commonly encountered lesions include adrenal cysts, ganglioneuromas, and hemorrhage.

Diagnostic Investigations

The diagnostic workup of an adrenal incidentaloma is aimed at identifying patients who would benefit from adrenalectomy (i.e., patients with functioning tumors and tumors at increased risk of being malignant). It is not necessary for asymptomatic patients whose imaging studies are consistent with obvious cysts, hemorrhage, myelolipomas, or diffuse metastatic disease to undergo additional investigations. All other patients should be tested for underlying hormonally active tumors using (a) a low-dose (1 mg) overnight dexamethasone suppression test to rule out subclinical Cushing’s syndrome and 17-ketosteroids (if sex steroid excess is suspected); (b) a 24-hour urine collection for catecholamines, metanephrines, VMA, or plasma metanephrine to rule out pheochromocytoma; and (c) in hypertensive patients, serum electrolytes, plasma aldosterone, and plasma renin to rule out an aldosteronoma. In patients with a high index of suspicion for subclinical Cushing’s (those with hypertension, obesity, or diabetes), three tests (i.e., dexamethasone suppression test, salivary cortisol, and 24-hour urine free cortisol) may be used. Confirmatory tests can be performed based on the results of the initial screening studies.

Determination of the malignant potential of an incidentaloma is more difficult. The risk of malignancy in an adrenal lesion is related to its size. Lesions >6 cm in diameter have an approximate risk of malignancy of about 25%.⁸² However, this size cutoff is not absolute because adrenal carcinomas also have been reported in lesions <6 cm. Carcinomas account for 2% of lesions <4 cm and 6% of lesions 4.1 to 6 cm in size. This has led to increased use of the imaging characteristics of incidentalomas to predict malignancy. Benign adrenal adenomas tend to be homogeneous, well encapsulated, and have smooth and regular margins. They also tend to be hypoattenuating lesions (<10 Hounsfield units) on CT scanning. In contrast, adrenal cancers tend to be hyperattenuating (>18 Hounsfield units) and inhomogeneous, have irregular borders, and may show evidence of local invasion or adjacent lymphadenopathy. On MRI T2-weighted imaging, adenomas demonstrate low signal intensity when compared to the liver (adrenal mass–to–liver ratio <1.4), whereas carcinomas and metastases have moderate intensity (mass-to-liver ratio 1.2:2.8). Pheochromocytomas are extremely bright, with mass-to-liver ratios >3. Unfortunately, the ranges overlap, and signal intensity is not 100% reliable for determining the nature of the lesion. Radionuclide imaging with NP-59 also has been used to distinguish between various adrenal lesions, with some investigators suggesting that uptake of NP-59 was 100% predictive of a benign lesion (adenoma), whereas absence of imaging was 100% predictive of a nonadenomatous lesion. However, the technique has not gained widespread acceptance because patients need to be given cold iodine 1 week before the study to prevent thyroid uptake, imaging needs to be delayed by 5 to 7 days after administration of the contrast, and false-positive and false-negative results occur. FDG-PET or PET-CT scans may have some utility in distinguishing potentially malignant from benign lesions in cases of inconclusive CT densitometry. However, caution must be exercised for false-positive (some adenomas and pheochromocytomas) and false-negative results (small lesions or those with hemorrhage or necrosis).⁹⁰ FNAB cannot be used to distinguish adrenal adenomas from carcinomas. This being said, FNAB is useful in the setting of a patient with a history of cancer and a solitary adrenal mass. The positive predictive value of FNAB in this situation has been shown to be almost 100%, although false-negative rates of up to 33% have been reported. Biopsies usually are performed under CT guidance, and appropriate testing to rule out pheochromocytomas should be undertaken before the procedure to avoid precipitating a hypertensive crisis.

Management

An algorithm for the management of patients with incidentalomas is shown in Fig. 38-47. The AACE and American Association of Endocrine Surgeons (AAES) have recently published management guidelines for patients with adrenal incidentalomas.⁹¹ Patients with functional tumors or obviously malignant lesions should undergo adrenalectomy. The optimal management of patients with subclinical Cushing’s syndrome is controversial, especially due to the paucity of data from high-quality prospective trials. In general, operative intervention is advised in patients with subclinical Cushing’s syndrome with suppressed plasma ACTH levels and elevated urinary cortisol levels because these patients are at high risk for progression to overt Cushing’s syndrome. The adrenal incidentaloma guidelines also recommend adrenalectomy in patients with worsening hypertension, abnormal glucose tolerance, or osteoporosis.

For nonfunctional lesions, the risk of malignancy needs to be balanced with operative morbidity and mortality. The AACE/AAES guidelines recommend that lesions with suspicious features on imaging studies such as heterogeneity, irregular capsule, or adjacent nodes should be treated by adrenalectomy. Nonoperative therapy, with close periodic follow-up, is advised for lesions <4 cm in diameter with benign imaging characteristics, whereas adrenalectomy is recommended for lesions ≥4 cm in size due to the increased risk of cancer.

However, several important points must be considered in the management of these patients. First, size criteria for malignancy are not definitive and are derived from a selected series of patients. Second, the actual size of adrenal tumors can be underestimated by at least 1 cm by modalities such as CT and MRI scans, because tumors are larger in a cephalocaudal axis. Third, the natural history of incidentalomas is variable and depends on the underlying diagnosis, age of the study population, and the size of the mass. Older patients are more likely to have nonfunctioning adenomas. Existing data in terms of the long-term behavior of these nonfunctional lesions, although limited, indicate that malignant transformation is uncommon. Furthermore, tumors that increase in size by at least 1 cm over a 2-year follow-up period and those with subtle hormonal abnormalities appear to be more likely to enlarge. Overt hormone overproduction is more likely in tumors >3 cm and those with increased NP-59 uptake. Surgeons are more likely to operate on a 40-year-old patient with a 4-cm lesion, while electing to follow an 80-year-old patient with a similar lesion but multiple concurrent comorbidities. Based on the above considerations, some surgeons use a size threshold for adrenalectomy with a nonfunctioning homogeneous tumor of 3 to 4 cm in young patients with no comorbidities and 5 cm in older patients with significant comorbidity.

Lesions that grow during follow-up also are treated by adrenalectomy. Myelolipomas generally do not warrant adrenalectomy unless there is concern regarding malignancy, which is rare, or bleeding into the lesion, which is more likely in myelolipomas >4 cm in size. These tumors, even when large, can be removed laparoscopically. Resection of solitary adrenal metastases in patients with a history of nonadrenal cancer has been demonstrated to lead to prolonged patient survival. Suspected adrenal metastases also may be resected for diagnosis.

There is no consensus regarding the follow up of patients with adrenal incidentaloma. The AACE/AAES guidelines recommend repeating hormonal screening with a 1-mg dexamethasone suppression test and urinary catecholamines and metabolites yearly for 5 years as the risk of hypersecretion appears to plateau after this time period.⁹¹ It also recommends repeat imaging at 3 to 6 months and then annually for 1 to 2 years. Less frequent imaging is reasonable or small (<2 cm), uniform, hypodense cortical nodules in patients without a history of malignant disease. Adrenalectomy is recommended for lesions that grow ≥1 cm or if autonomous hormone secretion develops during follow-up.

Adrenal Insufficiency

Adrenal insufficiency may be primary, resulting from adrenal disease, or secondary, due to a deficiency of ACTH (Table 38-21). The most commonly encountered causes of primary adrenal insufficiency are autoimmune disease, infections, and metastatic deposits. Spontaneous adrenal hemorrhage can occur in patients with fulminant meningococcal septicemia (Waterhouse-Friderichsen syndrome). Bilateral adrenal hemorrhage also can occur secondary to trauma, severe stress, infection, and coagulopathies and, if unrecognized, is lethal. Exogenous glucocorticoid therapy with suppression of the adrenal glands is the most common cause of secondary adrenal insufficiency.

Symptoms and Signs

Acute adrenal insufficiency should be suspected in stressed patients with any of the relevant risk factors. It may mimic sepsis, myocardial infarction, or pulmonary embolus and presents with fever, weakness, confusion, nausea, vomiting, lethargy, abdominal pain, or severe hypotension. Chronic adrenal insufficiency, such as that occurring in patients with metastatic tumors, may be more subtle. Symptoms include fatigue, salt craving, weight loss, nausea, vomiting, and abdominal pain. These patients may appear hyperpigmented from increased secretion of CRH and ACTH, with increased α-melanocyte-stimulating hormone side-products.

Diagnostic Studies

Characteristic laboratory findings include hyponatremia, hyperkalemia, eosinophilia, mild azotemia, and fasting or reactive hypoglycemia. The peripheral blood smear may demonstrate eosinophilia in approximately 20% of patients. Adrenal insufficiency is diagnosed by the ACTH stimulation test. ACTH (250 μg) is infused intravenously, and cortisol levels are measured at 0, 30, and 60 minutes. Peak cortisol levels <20 μg/dL suggest adrenal insufficiency. ACTH levels also allow primary insufficiency to be distinguished from secondary causes. High ACTH levels with low plasma cortisol levels are diagnostic of primary adrenal insufficiency.

Treatment

Treatment must be initiated based on clinical suspicion alone, even before test results are obtained, or the patient is unlikely to survive. Management includes volume resuscitation with at least 2 to 3 L of a 0.9% saline solution or 5% dextrose in saline solution. Blood should be obtained for electrolyte (decreased Na⁺ and increased K⁺), glucose (low), and cortisol (low) levels; ACTH (increased in primary and decreased in secondary); and quantitative eosinophilic count. Dexamethasone (4 mg) should be administered intravenously. Hydrocortisone (100 mg intravenously every 8 hours) also may be used, but it interferes with testing of cortisol levels. Once the patient has been stabilized, underlying conditions such as infection should be sought, identified, and treated. The ACTH stimulation test should be performed to confirm the diagnosis. Glucocorticoids can then be tapered to maintenance doses (oral hydrocortisone 15–20 mg in the morning and 10 mg in the evening). Mineralocorticoids (fludrocortisone 0.05–0.1 mg daily) may be required once the saline infusions are discontinued.

Adrenal Surgery

Choice of Procedure

Adrenalectomy may be performed via a laparoscopic or open approach. In either approach, the gland may be approached anteriorly, laterally, or posteriorly via the retroperitoneum. The choice of approach depends on the size and nature of the lesion and expertise of the surgeon. Laparoscopic adrenalectomy has rapidly become the standard procedure of choice for the excision of most benign-appearing adrenal lesions <6 cm in diameter. The role of laparoscopic adrenalectomy in the management of adrenocortical cancers is controversial. The data with respect to local tumor recurrence and intra-abdominal carcinomatosis from laparoscopic adrenalectomy for malignant adrenal tumors that were not appreciated as such, preoperatively or intraoperatively, are conflicting. Although laparoscopic adrenalectomy appears to be feasible and safe for solitary adrenal metastasis⁹² (provided there is no local invasion and the tumor can be resected intact), open adrenalectomy or laparoscopic-assisted open adrenalectomy is the safest option for suspected or known adrenocortical cancers and malignant pheochromocytomas. Technical considerations and surgeon experience, rather than absolute tumor size, usually determine the size threshold for laparoscopic resection. Hand-assisted laparoscopic adrenalectomy may provide a bridge between laparoscopic adrenalectomy and conversion to an open procedure. There have been no randomized trials directly comparing open vs. laparoscopic adrenalectomy. However, studies have shown that laparoscopic adrenalectomy is associated with decreased blood loss, postoperative pain, and narcotic use; reduced length of hospital stay; and faster return to work.

Laparoscopic Adrenalectomy

The procedure is performed under general anesthesia. Arterial lines are used routinely, and central lines are necessary for patients in whom massive fluid shifts are anticipated (e.g., those with large, active pheochromocytomas). A nasogastric tube and Foley catheter are recommended. Routine preoperative antibiotics are not needed, except in patients with Cushing’s syndrome. The adrenals can be removed laparoscopically via a transabdominal (anterior or lateral) or retroperitoneal (lateral or posterior) approach. The lateral approach is preferred by most laparoscopic surgeons and uses gravity to aid retraction of surrounding organs. Patients, however, need to be repositioned for a bilateral procedure. The anterior transabdominal approach offers the advantage of a conventional view of the abdominal cavity and allows a bilateral adrenalectomy to be performed without the necessity of repositioning the patient. The posterior retroperitoneal approach has also been gaining popularity in recent years, particularly in patients with previous anterior abdominal surgery and peritoneal adhesions. In addition, several centers have successfully utilized robotic approaches for both lateral transabdominal and retroperitoneal laparoscopic adrenal surgery. Single incision laparoscopic adrenalectomy is another option; however, widespread use of these latter approaches awaits analysis of long-term outcomes data and cost analyses. The lateral transabdominal approach is widely used and described in detail below.

Lateral Transabdominal Approach

The patient is placed in the lateral decubitus position, and the operating table is flexed at the waist to open the space between the lower rib cage and the iliac crest (Fig. 38-48). The surgeon and assistant both stand on the same side, facing the front of the patient. Pneumoperitoneum is created using a Veress needle or insufflation via a Hasson port. In general, four 10-mm trocars are placed between the midclavicular line medially and anterior axillary line laterally, one to two fingerbreadths below the costal margin (see Fig. 38-48), although additional ports may be placed, if needed. A 30° laparoscope is inserted through the second or midclavicular port. Most of the dissection is carried out via the two most lateral ports. However, the instruments and ports may be changed to provide optimum exposure, as needed.

For a right adrenalectomy, a fan retractor is inserted through the most medial port to retract the liver. An atraumatic grasper and an L-hook cautery are inserted via the two lateral ports for the dissection. The right triangular ligament is divided and the liver is rotated medially (Fig. 38-49A). Rarely, the hepatic flexure of the colon may need mobilization during a right adrenalectomy. The right kidney is identified visually and by palpation with an atraumatic grasper. The adrenal gland is identified on the superomedial aspect of the kidney. Gerota’s fascia is incised with the hook cautery. Dissection of the adrenal is started superomedially and then proceeds inferiorly, dissecting around the adrenal in a clockwise manner. The periadrenal tissues are grasped or moved with a blunt grasper to facilitate circumferential dissection. The right adrenal vein is identified at its junction with the IVC, ligated with clips, and divided using endoscopic scissors. Alternatively, a vascular stapler may be used to divide the vein endoscopically. There may be a second adrenal vein on the right. Generally, two clips are left on the vena cava side. Although early identification of the adrenal vein is helpful to facilitate mobilization and prevent injury, it can be dissected whenever it is safe to do so. Early ligation of the adrenal vein makes it easier to mobilize the gland but may make subsequent dissection more difficult due to venous congestion. The arterial branches to the adrenal gland can be electrocoagulated if small or clipped and divided.

For a left adrenalectomy, the fan retractor is used to retract the spleen. The splenic flexure is mobilized early, and the lateral attachments to the spleen and the tail of the pancreas are divided using the electrocautery (Fig. 38-49B). Gravity allows the spleen and the pancreatic tail to fall medially. The remainder of the dissection proceeds similarly to that described for the right adrenal. In addition to the adrenal vein, the inferior phrenic vein, which joins the left adrenal vein medially, also needs to be dissected, doubly clipped, and divided. As with the right adrenal vein, the left-sided veins also can be divided with a vascular stapler. Once the dissection is complete, the area of the adrenal bed can be irrigated and suctioned. A drain is rarely necessary. The gland is placed in a nylon specimen bag, which is brought out via one of the ports after morcellation, if necessary.

Posterior Retroperitoneal Approach

The retroperitoneal approach provides a more direct access to the adrenal gland and avoids abdominal adhesions in patients who have had previous abdominal surgery. Furthermore, bilateral adrenalectomy can be performed without repositioning the patient. Intraoperative ultrasound is helpful for identifying the adrenal, but the dissection and exposure are more difficult because the working space is limited. This makes vascular control difficult and also renders it unsuitable for large (>5 cm) lesions. This technique is being increasingly used for small adenomas causing hyperaldosteronism.

The patient is placed in the prone-jackknife position, and the operating table is flexed at the waist to open the space between the posterior costal margin and the pelvis. Palpation is used to identify the position of the twelfth rib. Percutaneous ultrasound is performed to determine the outline of the underlying kidney and adrenal. When done laparoscopically, the surgeon stands on the side of the adrenal to be removed, and the assistant stands on the opposite side. A 1.5-cm incision is placed 2 cm inferior and parallel to the twelfth rib, laterally at the level of the inferior pole of the kidney. Gerota’s space is entered under direct vision using a 12-mm direct viewing trocar with a 0° laparoscope through the muscle layers of the posterior abdominal wall. Alternatively, blunt dissection with the surgeon’s finger also can identify the space behind Gerota’s fascia. The trocar is then replaced by a dissecting balloon, which is manually inflated using a hand pump under direct vision through the laparoscope. A 12-mm trocar is then reinserted into this space, and CO₂ is insufflated to 12 to 15 mmHg pressure. The 0° laparoscope is replaced by a 45° laparoscope. Two additional 5- or 10-mm trocars are placed, one each on either side of the first port. Laparoscopic ultrasound then is used to help locate the adrenal gland and vessels. The adrenal dissection is begun at the superior pole and then proceeds to the lateral and inferior aspect. The medial dissection usually is performed last, and the vessels are identified and divided as described in the earlier Lateral Transabdominal Approach section.

Open Adrenalectomy

Open adrenalectomy may be performed via four approaches, each with specific advantages and disadvantages. The anterior approach allows examination of the abdominal cavity and resection of bilateral tumors via a single incision. The posterior approach avoids the morbidity of a laparotomy incision, especially in patients with cardiopulmonary disease and those prone to wound complications (Cushing’s syndrome) and avoids abdominal adhesions in patients who have undergone previous abdominal surgery. Recovery time is also quicker and hospitalization shorter. However, the retroperitoneal exposure is difficult, particularly in obese patients, and the small working space makes it unsuitable for tumors >6 cm in diameter. The lateral approach is best for obese patients and for large tumors because it provides a bigger working space. The thoracoabdominal approach is most useful for en bloc resection of large (>10 cm), malignant lesions. However, it is associated with significant morbidity and should be used selectively.

Anterior Approach

The adrenals may be removed via a midline incision or bilateral subcostal incision (Fig. 38-50). The former allows adequate infraumbilical exposure for examination of extra-adrenal tumors, whereas the latter provides better superior and lateral exposure. For the right side, the hepatic flexure of the colon is mobilized inferiorly, and the triangular ligament is incised to retract the liver medially and superiorly. A generous Kocher maneuver is used to mobilize the duodenum anteriorly and expose the retroperitoneal fat and the IVC (Fig. 38-51A). Gerota’s fascia is incised, and the gland is freed of surrounding fibro-fatty tissue and the kidney inferiorly. The lateral and superior surfaces usually are mobilized first. Then, the short, right adrenal vein is dissected, ligated, and divided, taking care not to injure the hepatic veins and IVC. On the left side, the adrenal is located cephalad to the pancreatic tail and just lateral to the aorta. For large tumors, the adrenal is best approached by medial visceral rotation to mobilize the spleen, colon, and pancreas toward the midline (Fig. 38-51B). An alternative approach is to enter the lesser sac by division of the gastrocolic ligament. The pancreas is mobilized superiorly by incision of its inferior peritoneal attachments, thus exposing the left kidney and adrenal. The gland is then mobilized as on the right side.

Posterior Approach

The patient is placed prone on the operating table, similar to the laparoscopic approach. A hockey stick or curvilinear incision may be used, and extended through the latissimus dorsi and sacrospinous fascia. The twelfth rib generally is excised at its base, and the eleventh rib is retracted superiorly to reveal the pleura and the lateral arcuate ligament of the liver on the right side. The pleura also is mobilized cephalad, and the adrenal and kidney are identified. The superior aspect of the gland is dissected first, and the superior vessels are identified and ligated. This prevents superior retraction of the adrenal gland. The remainder of the gland is then dissected and the adrenal gland and tumor removed. The resulting space generally is filled with perinephric fat and closed in layers. A chest x-ray is obtained postoperatively to rule out a pneumothorax.

Lateral Approach

The patient is placed in a lateral position with the table flexed, and an incision is made between the eleventh and twelfth ribs or subcostally. The dissection then is performed as indicated previously in Anterior Approach.

Complications of Adrenal Surgery

Patients with Cushing’s syndrome are more prone to infectious (incisional and intra-abdominal abscess) and thrombotic complications. Creation of pneumoperitoneum may result in injury to various organs from Veress needle and trocar insertion, subcutaneous emphysema, pneumothorax, and hemodynamic compromise. Excessive retraction and dissection may lead to bleeding from injury to the IVC and renal vessels, or from injury to surrounding organs such as the liver, pancreas, spleen, and stomach. Postoperative hemodynamic instability may be evident in patients with pheochromocytomas, and patients are at risk of adrenal insufficiency after bilateral adrenalectomy and sometimes after unilateral adrenalectomy (unrecognized Cushing’s syndrome or, very rarely, Conn’s syndrome). Long-term morbidity results mainly from injury to nerve roots during trocar insertion, which can lead to chronic pain syndromes or muscle weakness, although this is more of an issue in case of open procedures.

Approximately 30% of patients who undergo bilateral adrenalectomy for Cushing’s disease are at risk of developing Nelson’s syndrome from progressive growth of the pre-existing pituitary tumor. This leads to increased ACTH levels, hyperpigmentation, visual field defects, headaches, and extraocular muscle palsies. Transsphenoidal pituitary resection is the initial mode of therapy, and external-beam radiotherapy is used in patients with residual tumor or extrasellar invasion.