Multiple endocrine neoplasia syndrome is defined as a disorder with neoplasms in two or more different hormonal tissues in several members of a family. Several distinct genetic disorders predispose to endocrine gland neoplasia and cause hormone excess syndromes (Table 351-1). DNA-based genetic testing is available for these disorders, but effective management requires an understanding of endocrine neoplasia and the range of clinical features that may be manifested in an individual patient.
Multiple Endocrine Neoplasia (MEN) Type 1
MEN 1, or Wermer's syndrome, is inherited as an autosomal dominant trait. This syndrome is characterized by neoplasia of the parathyroid glands, enteropancreatic tumors, anterior pituitary adenomas, and other neuroendocrine tumors with variable penetrance (Table 351-1). Although rare, MEN 1 is the most common multiple endocrine neoplasia syndrome, with an estimated prevalence of 2–20 per 100,000 in the general population. It is caused by inactivating mutations of the tumor-suppressor gene MEN1 located at chromosome 11q13. The MEN1 gene codes for a nuclear protein called Menin. Menin interacts with JunD, suppressing JunD-dependent transcriptional activation. It is unclear how this accounts for Menin growth regulatory activity, since JunD is associated with inhibition of cell growth. Each child born to an affected parent has a 50% probability of inheriting the gene. The variable penetrance of the several neoplastic components can make the differential diagnosis and treatment challenging.
Primary hyperparathyroidism is the most common manifestation of MEN 1, with an estimated penetrance of 95–100%. Hypercalcemia may develop during the teenage years, and most individuals are affected by age 40 (Fig. 351-1). Hyperparathyroidism is the earliest manifestation of the syndrome in most MEN 1 patients. The neoplastic changes in hyperparathyroidism provide a specific example of one of the cardinal features of endocrine tumors in MEN 1: multicentricity. The neoplastic changes inevitably affect multiple parathyroid glands, making surgical cure difficult. Screening for hyperparathyroidism involves measurement of either an albumin-adjusted or an ionized serum calcium level. The diagnosis is established by demonstrating elevated levels of serum calcium and intact parathyroid hormone. Manifestations of hyperparathyroidism in MEN 1 do not differ substantially from those in sporadic hyperparathyroidism and include calcium-containing kidney stones, kidney failure, nephrocalcinosis, bone abnormalities (i.e., osteoporosis, osteitis fibrosa cystica), and gastrointestinal and musculoskeletal complaints. Management is challenging because of early onset, significant recurrence rates, and the multiplicity of parathyroid gland involvement. Differentiation of hyperparathyroidism of MEN 1 from other forms of familial primary hyperparathyroidism usually is based on family history, histologic features of resected parathyroid tissue, the presence of a MEN1 mutation, and, sometimes, long-term observation to determine whether other manifestations of MEN 1 develop. Parathyroid hyperplasia is the most common cause of hyperparathyroidism in MEN 1, although single and multiple adenomas have been described. Hyperplasia of one or more parathyroid glands is common in younger patients; adenomas usually are found in older patients or those with long-standing disease.
Age at onset of endocrine tumor expression in multiple endocrine neoplasia type 1 (MEN 1). Data derived from retrospective analysis for each endocrine organ hyperfunction in 130 cases of MEN 1. Age at onset is the age at first symptom or, with tumors not causing symptoms, age at the time of the first abnormal finding on a screening test. The rate of diagnosis of hyperparathyroidism increased sharply between ages 16 and 20 years. (Reprinted with permission from S Marx et al: Ann Intern Med 129:484, 1998.)
Enteropancreatic tumors are the second most common manifestation of MEN 1, with an estimated penetrance of 50%. They tend to occur in parallel with hyperparathyroidism (Fig. 351-1); 30% are malignant. Most of these tumors secrete peptide hormones that cause specific clinical syndromes. Those syndromes, however, may have an insidious onset and a slow progression, making their diagnosis difficult and in many cases delayed. Some enteropancreatic tumors do not secrete hormones. Those "silent" tumors usually are found during radiographic screening. Metastasis, most commonly to the liver, occurs in about a third of patients.
Gastrinomas are the most common enteropancreatic tumors observed in MEN 1 patients and result in the Zollinger-Ellison syndrome (ZES). ZES is caused by excessive gastrin production and occurs in more than one-half of MEN 1 patients with small carcinoid- like tumors in the duodenal wall or, less often, by pancreatic islet cell tumors. There may be more than one gastrin-producing tumor, making localization difficult. The robust acid production may cause esophagitis, duodenal ulcers throughout the duodenum, ulcers involving the proximal jejunum, and diarrhea. The ulcer diathesis is commonly refractory to conservative therapy such as antacids. The diagnosis is made by finding increased gastric acid secretion, elevated basal gastrin levels in the serum [generally >115 pmol/L (200 pg/mL)], and an exaggerated response of serum gastrin to either secretin or calcium. Other causes of elevated serum gastrin levels, such as achlorhydria, treatment with H2 receptor antagonists or proton pump inhibitors, retained gastric antrum, small-bowel resection, gastric outlet obstruction, and hypercalcemia, should be excluded (Fig. 351-1). High-resolution, early-phase CT scanning, abdominal MRI with contrast, octreotide scan, and/or endoscopic ultrasound are the best preoperative techniques for identification of the primary and metastatic gastrinoma; intraoperative ultrasonography is the most sensitive method for detection of small tumors. Approximately one-fourth of all cases of ZES occur in the context of MEN 1.
Insulinomas are the second most common enteropancreatic tumors in patients who have MEN 1. Unlike gastrinomas, most insulinomas originate in the pancreas bed, becoming the most common pancreatic tumor in MEN 1. Hypoglycemia caused by insulinomas is observed in about one-third of MEN 1 patients with pancreatic islet cell tumors (Fig. 351-1). The tumors may be benign or malignant (25%). The diagnosis can be suggested by documenting hypoglycemia during a short fast with simultaneous inappropriate elevation of serum insulin and C-peptide levels. More commonly, it is necessary to subject the patient to a supervised 12- to 72-h fast to provoke hypoglycemia (Chap. 345). Large insulinomas may be identified by CT or MRI scanning; small tumors not detected by conventional radiographic techniques may be localized by endoscopic ultrasound or selective arteriographic injection of calcium into each of the arteries that supply the pancreas and sampling of the hepatic vein for insulin to determine the anatomic region containing the tumor. Intraoperative ultrasonography is used frequently to localize these tumors. The trend toward earlier diagnosis of, hence, smaller tumors has reduced the usefulness of octreotide scanning, which is positive in a minority of these patients.
Glucagonoma, which is seen occasionally in MEN 1, causes a syndrome of hyperglycemia, skin rash (necrolytic migratory erythema), anorexia, glossitis, anemia, depression, diarrhea, and venous thrombosis. In about half of these patients the plasma glucagon level is high, leading to its designation as the glucagonoma syndrome, although elevation of the plasma glucagon level in MEN 1 patients is not necessarily associated with these symptoms. Some patients with this syndrome also have elevated plasma ghrelin levels. The glucagonoma syndrome may represent a complex interaction between glucagon and ghrelin overproduction and the nutritional status of the patient.
The Verner-Morrison, or watery diarrhea, syndrome consists of watery diarrhea, hypokalemia, hypochlorhydria, and metabolic acidosis. The diarrhea can be voluminous and almost always is found in association with an islet cell tumor in the context of MEN 1, prompting use of the term pancreatic cholera. However, when not associated with MEN 1, the syndrome outside of MEN 1 is not restricted to pancreatic islet tumors and has been observed with carcinoids or other tumors. This syndrome is believed to be due to overproduction of vasoactive intestinal peptide (VIP), although plasma VIP levels may not be elevated. Hypercalcemia may be induced by the effects of VIP to stimulate osteoclastic bone resorption as well as by hyperparathyroidism. Other disorders that should be considered in the differential diagnosis of chronic diarrhea include infectious or parasitic diseases, inflammatory bowel disease, sprue, and other endocrine causes such as ZES, carcinoid, and medullary thyroid carcinoma.
The pancreatic neoplasms differ from the other components of MEN 1 in that approximately one-third of the tumors display malignant features, including hepatic metastases. The pancreatic neoplasms also can be used to highlight another characteristic of MEN 1: the specific impact of a hormone produced by one component of MEN 1 on another neoplastic component of this syndrome. Specific examples include the effects of either corticotropin-releasing hormone (CRH) or growth hormone–releasing hormone (GHRH) production by an islet cell tumor to cause a syndrome of excess adrenocorticotropin (ACTH) (Cushing's disease) or GH (acromegaly) production by the pituitary gland. These secondary interactions add complexity to the diagnosis and management of these tumor syndromes. Pancreatic islet cell tumors are diagnosed by identification of a characteristic clinical syndrome, hormonal assays with or without provocative stimuli, or radiographic techniques. One approach involves annual screening of individuals at risk with measurement of basal and meal-stimulated levels of pancreatic polypeptide to identify the tumors as early as possible; the rationale for this screening strategy is the concept that surgical removal of islet cell tumors at an early stage will be curative. Other approaches to screening include measurement of serum gastrin and pancreatic polypeptide levels every 2–3 years, with the rationale that pancreatic neoplasms will be detected at a later stage but can be managed medically, if possible, or by surgery. High-resolution, early-phase CT scanning or endoscopic ultrasound provides the best preoperative technique for identification of these tumors; intraoperative ultrasonography is the most sensitive method for detection of small tumors. Although fluorodeoxyglucose–positron emission tomography (FDG-PET) scanning detects ∼50% of pancreatic islet cell tumors, most of these tumors are large; as most of these tumors can be identified by CT or ultrasound, the lack of sensitivity for small tumors makes FDG-PET scanning unhelpful for early diagnosis.
Pituitary tumors occur in 20–30% of patients with MEN 1 and tend to be multicentric. These tumors can exhibit aggressive behavior and local invasiveness that makes them difficult to resect (Chap. 339). Prolactinomas are the most common (Fig. 351-1) and are diagnosed by finding serum prolactin levels >200 μg/L with or without a pituitary mass evident on MRI. Values <200 μg/L may be due to a prolactin-secreting neoplasm or to compression of the pituitary stalk by a different type of pituitary tumor. Acromegaly due to excessive GH production is the second most common syndrome caused by pituitary tumors in MEN 1 and can rarely be due to production of GHRH by an islet cell tumor (see above). The possibility of hereditary growth hormone– or prolactin-secreting tumors (discussed below in "Other Genetic Endocrine Tumor Syndromes") should be considered in the differential diagnosis. Cushing's disease can be caused by ACTH-producing pituitary tumors or by ectopic production of ACTH or CRH by other components of the MEN 1 syndrome, including islet cell or carcinoid tumors or adrenal adenomas. Diagnosis of pituitary Cushing's disease is generally best accomplished by a high-dose dexamethasone suppression test or by petrosal venous sinus sampling for ACTH after IV injection of CRH. Differentiation of a primary pituitary tumor from an ectopic CRH-producing tumor may be difficult because the pituitary is the source of ACTH in both disorders; documentation of CRH production by a pancreatic islet or carcinoid tumor may be the only method of proving ectopic CRH production.
Adrenal cortical tumors are found in almost one-half of gene carriers but are rarely functional; malignancy in cortical adenomas is uncommon. Rare cases of pheochromocytoma have been described in the context of MEN 1. Due to their rarity, screening for these tumors is indicated only when there are suggestive symptoms.
Carcinoid tumors in MEN 1 are of the foregut type and are derived from thymus, lung, stomach, or duodenum; they may metastasize or be locally invasive. These tumors usually produce serotonin, calcitonin, or CRH; the typical carcinoid syndrome with flushing, diarrhea, and bronchospasm is rare (Chap. 350). Mediastinal carcinoid tumors (an upper mediastinal mass) are more common in men; bronchial carcinoid tumors are more common in women. Carcinoid tumors are a late manifestation of MEN 1; some reports have emphasized the importance of routine chest CT screening for mediastinal carcinoid tumors because of their high rate of malignant transformation and aggressive behavior.
Unusual Manifestations of MEN 1
Subcutaneous or visceral lipomas and cutaneous leiomyomas also may be present but rarely undergo malignant transformation. Skin angiofibromas or collagenomas are seen in most patients with MEN 1 when carefully sought.
MEN1 gene mutations are found in >90% of families with the syndrome (Fig. 351-2). Genetic testing can be performed in individuals at risk for the development of MEN 1 and is commercially available in the United States and Europe. The major value of genetic testing in a kindred with an identifiable mutation is the assignment or exclusion of gene carrier status. In those identified as carrying the mutant gene, routine screening for individual manifestations of MEN 1 should be performed as outlined above. Those with negative genetic test results in a kindred with a known germ-line mutation can be excluded from further screening for MEN 1. A significant percentage of sporadic parathyroid, islet cell, and carcinoid tumors also have loss or mutation of MEN1. There is no correlation between a particular germ-line mutation and a clinical phenotype. It is presumed that these mutations are somatic and occur in a single cell, leading to subsequent transformation.
Schematic depiction of the MEN1 gene and the distribution of mutations. The shaded areas show coding sequence. The closed circles show the relative distribution of mutations, mostly inactivating, in each exon. Mutation data are derived from the Human Gene Mutation Database, from which more detailed information can be obtained at www.uwcm.ac.uk/uwcm/mg/hgmd0.html. (From M Krawczak, DN Cooper: Trends Genet 13:1321, 1998.)
Treatment: Multiple Endocrine Neoplasia Type 1
Almost everyone who inherits a mutant MEN1 gene develops at least one clinical manifestation of the syndrome. Most develop hyperparathyroidism, 80% develop pancreatic islet cell tumors, and more than half develop pituitary tumors. For most of these tumors, initial surgery is not curative and patients frequently require multiple surgical procedures and surgery on two or more endocrine glands during a lifetime. For this reason, it is essential to establish clear goals for management of these patients rather than to recommend surgery casually each time a tumor is discovered. Ranges for acceptable management are discussed below.
Individuals with serum calcium levels >3.0 mmol/L (12 mg/dL), evidence of calcium nephrolithiasis or renal dysfunction, neuropathic or muscular symptoms, or bone involvement (including osteopenia) and individuals <50 years of age should undergo parathyroid exploration. There is less agreement about the necessity for parathyroid exploration in individuals who do not meet these criteria, and observation may be appropriate in MEN 1 patients with asymptomatic hyperparathyroidism.
When parathyroid surgery is indicated in MEN 1, there are two approaches. In the first, all parathyroid tissue is identified and removed at the time of primary operation, and parathyroid tissue is implanted in the nondominant forearm. Thymectomy also should be performed because of the potential for later development of malignantcarcinoid tumors. If reoperation for hyperparathyroidism is necessary at a later date, transplanted parathyroid tissue can be resected from the forearm with titration of tissue removal to lower the intact parathyroid hormone (PTH) to <50% of basal.
Another approach is to remove 3–3.5 parathyroid glands from the neck (leaving ∼50 mg of parathyroid tissue), carefully marking the location of residual tissue so that the remaining tissue can be located easily during subsequent surgery. If this approach is used, intraoperative PTH measurements should be utilized to monitor adequacy of removal of parathyroid tissue with a goal of reducing postoperative serum intact PTH to ≤50% of basal values.
The use of high-resolution CT scanning (1 mm) and imaging during three phases of contrast flow has substantially improved the ability to identify aberrantly located parathyroid tissue. As this issue arises with some frequency in the context of parathyroid disease in MEN 1, this technique should be utilized to locate parathyroid tissue before reoperation for a failed exploration, and it may be useful before the initial operation.
Pancreatic Islet Cell Tumors
(See Chap. 350 for discussion of pancreatic islet cell tumors not associated with MEN 1.) Two features of pancreatic islet cell tumors in MEN 1 complicate management. First, the tumors are multicentric, are malignant about a third of the time, and cause death in 10–20% of patients. Second, performance of a total pancreatectomy to prevent malignancy causes diabetes mellitus, a disease with significant long-term complications that include neuropathy, retinopathy, and nephropathy. These features make it difficult to formulate clear-cut guidelines, but some general concepts appear to be valid. (1) Islet cell tumors producing insulin, glucagon, VIP, GHRH, or CRH should be resected because medical therapy for the hormonal effects of these tumors are generally ineffective. (2) Gastrin-producing islet cell tumors that cause ZES are frequently multicentric. Recent experience suggests that a high percentage of ZES in MEN 1 is caused by duodenal wall carcinoid tumors and that resection of these tumors improves the cure rate. Treatment with H2 receptor antagonists (cimetidine or ranitidine) or proton pump inhibitors (omeprazole, lansoprazole, esomeprazole, etc.) provides an alternative, and some think preferable, therapy to surgery for control of ulcer disease in patients with multicentric tumors or hepatic metastases. (3) In families in which there is a high incidence of malignant islet cell tumors that cause death, total pancreatectomy at an early age may be considered to prevent malignancy, although it should be noted that this surgical intervention does not prevent the development of neuroendocrine tumors outside the pancreato-duodenal region.
Management of metastatic islet cell carcinoma is unsatisfactory. Hormonal abnormalities sometimes can be controlled. For example, ZES can be treated with H2 receptor antagonists or proton pump inhibitors; the somatostatin analogues octreotide and lanreotide are useful in the management of carcinoid, glucagonoma, and the watery diarrhea syndrome. Bilateral adrenalectomy may be required for ectopic ACTH syndrome if medical therapy is ineffective (Chap. 342). Islet cell carcinomas frequently metastasize to the liver but may grow slowly. Hepatic artery embolization, radiofrequency ablation, or chemotherapy (5-fluorouracil, streptozocin, chlorozotocin, doxorubicin, or dacarbazine) may reduce tumor streptozotocin mass, control symptoms of hormone excess, and prolong life; however, these treatments are never curative. There is evolving evidence that everolimus, an inhibitor of mTor (mammalian target of rapamycin) causes regression of tumor size; 2 of 13 islet cell carcinomas and 2 of 12 carcinoid tumors had a >30% reduction in size and >60% had stable disease.
Treatment of prolactinomas with dopamine agonists (bromocriptine, cabergoline, or quinagolide) usually returns the serum prolactin level to normal and prevents further tumor growth (Chap. 339). Surgical resection of a prolactinoma is rarely curative but may relieve mass effects. Transsphenoidal resection is appropriate for neoplasms that secrete ACTH, GH, or the α subunit of the pituitary glycoprotein hormones. Octreotide reduces tumor mass in one-third of GH-secreting tumors and reduces GH and insulin-like growth factor I levels in >75% of patients. Pegvisomant, a GH antagonist, rapidly lowers insulin-like growth factor levels in patients with acromegaly (Chap. 339). Radiation therapy may be useful for large or recurrent tumors.
Improvements in the management of MEN 1, particularly the earlier recognition of islet cell and pituitary tumors, have improved outcomes in these patients. As a result, other neoplastic manifestations that develop later in the course of this disorder, such as carcinoid syndrome, are now seen with increased frequency.
Multiple Endocrine Neoplasia Type 2
Medullary thyroid carcinoma (MTC) and pheochromocytoma are associated in two major syndromes: MEN type 2A and MEN type 2B (Table 351-1). MEN 2A is the combination of MTC, hyperparathyroidism, and pheochromocytoma. Three subvariants of MEN 2A are familial medullary thyroid carcinoma (FMTC), MEN 2A with cutaneous lichen amyloidosis, and MEN 2A with Hirschsprung disease. MEN 2B is the combination of MTC, pheochromocytoma, mucosal neuromas, intestinal ganglioneuromatosis, and marfanoid features.
Multiple Endocrine Neoplasia Type 2A
MTC is the most common manifestation. This tumor usually develops in childhood, beginning as hyperplasia of the calcitonin-producing cells (C cells) of the thyroid. MTC typically is located at the junction of the upper one-third and lower two-thirds of each lobe of the thyroid, reflecting the high density of C cells in this location; tumors >1 cm in size frequently are associated with local or distant metastases.
Pheochromocytoma occurs in ∼50% of patients with MEN 2A and causes palpitations, nervousness, headaches, and sometimes sweating (Chap. 343). About half of the tumors are bilateral, and >50% of patients who have had unilateral adrenalectomy develop a pheochromocytoma in the contralateral gland within a decade. A second feature of these tumors is a disproportionate increase in the secretion of epinephrine relative to norepinephrine. This characteristic differentiates the MEN 2 pheochromocytomas from sporadic pheochromocytoma and those associated with von Hippel–Lindau (VHL) syndrome, hereditary paraganglioma, or neurofibromatosis. Capsular invasion is common, but metastasis is uncommon. Finally, the pheochromocytomas almost always are found in the adrenal gland, differentiating the pheochromocytomas in MEN 2 from the extraadrenal tumors more commonly found in hereditary paraganglioma syndromes.
Hyperparathyroidism occurs in 15–20% of patients, with the peak incidence in the third or fourth decade. The manifestations of hyperparathyroidism do not differ from those in other forms of primary hyperparathyroidism (Chap. 353). Diagnosis is established by finding hypercalcemia, hypophosphatemia, hypercalciuria, and an inappropriately high serum level of intact PTH. Multiglandular parathyroid hyperplasia is the most common histologic finding, although with long-standing disease adenomatous changes may be superimposed on hyperplasia.
The most common subvariant of MEN 2A is familial MTC, an autosomal dominant syndrome in which MTC is the only manifestation (Table 351-1). The clinical diagnosis of FMTC is established by the identification of MTC in multiple generations without a pheochromocytoma. Since the penetrance of pheochromocytoma is 50% in MEN 2A, it is possible that MEN 2A could masquerade as FMTC in small kindreds. It is important to consider this possibility carefully before classifying a kindred as having FMTC; failure to do so could lead to death or serious morbidity from pheochromocytoma in an affected kindred member. The difficulty of differentiating MEN 2A and FMTC is discussed further below.
Multiple Endocrine Neoplasia Type 2B
The association of MTC, pheochromocytoma, mucosal neuromas, and a marfanoid habitus is designated MEN 2B. MTC in MEN 2B develops earlier and is more aggressive than in MEN 2A. Metastatic disease has been described before 1 year of age, and death may occur in the second or third decade of life. However, the prognosis is not invariably bad even in patients with metastatic disease, as evidenced by a number of multigenerational families with this disease.
Pheochromocytoma occurs in more than half of MEN 2B patients and does not differ from that in MEN 2A. Hypercalcemia is rare in MEN 2B, and there are no well-documented examples of hyperparathyroidism.
The mucosal neuromas and marfanoid body habitus are the most distinctive features and are recognizable in childhood. Neuromas are present on the tip of the tongue, under the eyelids, and throughout the gastrointestinal tract and are true neuromas, distinct from neurofibromas. The most common presentation in children relates to gastrointestinal symptomatology, including intermittent colic, pseudoobstruction, and diarrhea.
Mutations of the RET protooncogene have been identified in most patients with MEN 2 (Fig. 351-3). RET encodes a tyrosine kinase receptor that in combination with a co-receptor, GFRα, normally is activated by glial cell–derived neurotrophic factor (GDNF) or other members of this transforming growth factor β–like family of peptides, including artemin, persephin, and neurturin. In the C cell there is evidence that persephin normally activates the RET/GFRα-4 receptor complex and is partially responsible for migration of the C cells into the thyroid gland, whereas in the developing neuronal system of the gastrointestinal tract, GDNF activates the RET/GFRα-1 complex. RET mutations induce constitutive activity of the receptor, explaining the autosomal dominant transmission of the disorder.
Schematic diagram of the RET protooncogene showing mutations found in MEN type 2 and sporadic medullary thyroid carcinoma (MTC). The RET protooncogene is located on the proximal arm of chromosome 10q (10q11.2). Activating mutations of two functional domains of RET tyrosine kinase receptor have been identified. The first affects a cysteine-rich (Cys-Rich) region in the extracellular portion of the receptor. Each germ-line mutation changes a cysteine at codons 609, 611, 618, 620, or 634 to another amino acid. The second region is the intracellular tyrosine kinase (TK) domain. Codon 634 mutations account for ∼80% of all germ-line mutations. Mutations of codons 630, 768, 883, and 918 have been identified as somatic (non-germ-line) mutations that occur in a single parafollicular or C cell within the thyroid gland in sporadic MTC. A codon 918 mutation is the most common somatic mutation. MEN2, multiple endocrine neoplasia type 2; CLA, cutaneous lichen amyloidosis; FMTC, familial medullary thyroid carcinoma; Signal, the signal peptide; Cadherin, a cadherin-like region in the extracellular domain; TM, transmembrane domain; TK, tyrosine kinase domain.
Naturally occurring mutations localize to two regions of the RET tyrosine kinase receptor. The first is a cysteine-rich extracellular domain; point mutations in the coding sequence for one of six cysteines (codons 609, 611, 618, 620, 630, and 634) cause amino acid substitutions that induce receptor dimerization and activation in the absence of its ligand. Codon 634 mutations occur in 80% of MEN 2A kindreds and are most commonly associated with classic MEN 2A features (Figs. 351-3 and 351-2); an arginine substitution at this codon accounts for half of all MEN 2A mutations. All reported families with MEN 2A and cutaneous lichen amyloidosis have a codon 634 mutation. Mutations of codon 609, 611, 618, or 620 occur in 10–15% of MEN 2A kindreds and are more commonly associated with FMTC (Fig. 351-3). Mutations in codons 609, 618, and 620 also have been identified in a variant of MEN 2A that includes Hirschsprung disease (Fig. 351-3). The second region of the RET tyrosine kinase that is mutated in MEN 2 is in the substrate recognition pocket at codon 918 (Fig. 351-3). This activating mutation is present in ∼95% of patients with MEN 2B and accounts for 5% of all RET protooncogene mutations in MEN 2. Mutations of codon 883 and 922 also have been identified in a few patients with MEN 2B.
Uncommon mutations (<5% of the total) include those of codons 533 (exon 8), 666, 768, 777, 790, 791, 804, 891, and 912. Mutations associated with only FMTC include codons 533, 768, and 912. With greater experience, mutations that once were associated with FMTC only (666, 791, V804L, V804M, and 891) have been found in MEN 2A as there have been occasional descriptions of pheochromocytoma. At present it is reasonable to conclude that only kindreds with codon 533, 768, or 912 mutations are consistently associated with FMTC; in kindreds with all other RET mutations, pheochromocytoma is a possibility. The recognition that germ-line mutations occur in at least 6% of patients with apparently sporadic MTC has led to the firm recommendation that all patients with MTC should be screened for these mutations. The effort to screen patients with sporadic MTC, combined with the fact that new kindreds with classic MEN 2A are being recognized less frequently, has led to a shift in the mutation frequencies. These findings mirror results in other malignancies in which germ-line mutations of cancer-causing genes contribute to a greater percentage of apparently sporadic cancer than was considered previously. The recognition of new RET mutations suggests that more will be identified in the future.
Somatic mutations (found only in the tumor and not transmitted in the germ line) of the RET protooncogene have been identified in sporadic MTC; 25–60% of sporadic tumors have codon 918 mutations, and somatic mutations in codons 630, 768, and 804 have been identified (Fig. 351-3).
Treatment: Multiple Endocrine Neoplasia Type 2
Screening for Multiple Endocrine Neoplasia Type 2
Death from MTC can be prevented by early thyroidectomy. The identification of RET protooncogene mutations and the application of DNA-based molecular diagnostic techniques to identify these mutations have simplified the screening process. During the initial evaluation of a kindred, a RET protooncogene analysis should be performed on an individual with proven MEN 2A. Establishment of the specific germ-line mutation facilitates the subsequent analysis of other family members. Each family member at risk should be tested twice for the presence of the specific mutation; the second analysis should be performed on a new DNA sample and, ideally, in a second laboratory to exclude sample mix-up or technical error (see www.genetests.org for an up-to-date list of laboratory testing sites). Both false-positive and false-negative analyses have been described. A false-negative test result is of the greatest concern because calcitonin testing is now rarely performed as a diagnostic backup study; if there is a genetic test error, a child may present in the second or third decade with metastatic MTC. Individuals in a kindred with a known mutation who have two normal analyses can be excluded from further screening.
There is a consensus that children with codon 883, 918, and 922 mutations, those associated with MEN 2B, should have a total thyroidectomy and central lymph node dissection (level VI) performed during the first months of life or soon after identification of the syndrome. If local metastasis is discovered, a more extensive lymph node dissection (levels II to V) is generally indicated. In children with codon 611, 618, 620, 630, 634, and 891 mutations, thyroidectomy should be performed before age 6 years because of reports of local metastatic disease in children this age. Finally, there are kindreds with codon 609, 768, 790, 791, 804, and 912 mutations in which the phenotype of MTC appears to be less aggressive. A clinician caring for children with one of these mutations faces a dilemma. In many kindreds there has never been a death from MTC caused by one of these mutations. However, in other kindreds there are examples of metastatic disease occurring early in life. For example, metastatic disease before age 6 years has been described with codon 609 and 804 mutations and before age 14 years in a patient with a codon 912 mutation. In kindreds with these mutations, two management approaches have been suggested: (1) perform a total thyroidectomy with or without central node dissection at some arbitrary age (perhaps 6–10 years of age) or (2) continue annual or biannual calcitonin provocative testing with performance of total thyroidectomy with or without central neck dissection when the test becomes abnormal. The pentagastrin test involves measurement of serum calcitonin basally and 2, 5, 10, and 15 min after a bolus injection of 5 μg pentagastrin per kilogram of body weight. Patients should be warned before pentagastrin injection of epigastric tightness, nausea, warmth, and tingling of extremities and reassured that the symptoms will last ∼2 min. If pentagastrin is unavailable, an alternative is a short calcium infusion, performed by obtaining a baseline serum calcitonin and then infusing 150 mg calcium salt IV over 10 min with measurement of serum calcitonin at 5, 10, 15, 30 min after initiation of the infusion.
The RET protooncogene analysis should be performed in patients with suspected MEN 2B to detect codon 883, 918, and 922 mutations, especially in newborn children in whom the diagnosis is suspected but the clinical phenotype is not fully developed. Other family members at risk for MEN 2B also should be tested because the mucosal neuromas can be subtle. Most MEN 2B mutations represent de novo mutations derived from the paternal allele. In the rare families with proven germ-line transmission of MTC but no identifiable RET protooncogene mutation (sequencing of the entire RET gene should be performed), annual pentagastrin or calcium testing should be performed on members at risk.
Annual screening for pheochromocytoma in patients with germ-line RET mutations should be performed by measuring basal plasma or 24-h urine catecholamines and metanephrines. The goal is to identify a pheochromocytoma before it causes significant symptoms or is likely to cause sudden death, an event most commonly associated with large tumors. Although there are kindreds with FMTC and specific RET mutations in which no pheochromocytomas have been identified (Fig. 351-3), clinical experience is insufficient to exclude pheochromocytoma screening in these individuals. Radiographic studies such as MRI or CT scans generally are reserved for individuals with abnormal screening tests or symptoms suggestive of pheochromocytoma (Chap. 343). Women should be tested during pregnancy because undetected pheochromocytoma can cause maternal death during childbirth.
Measurement of serum calcium and parathyroid hormone levels every 2–3 years provides an adequate screen for hyperparathyroidism, except in families in which hyperparathyroidism is a prominent component, in which measurements should be made annually.
Medullary Thyroid Carcinoma
Hereditary MTC is a multicentric disorder. Total thyroidectomy with a central lymph node dissection should be performed in children who carry the mutant gene. Incomplete thyroidectomy leaves the possibility of later transformation of residual C cells. The goal of early therapy is cure, and a strategy that does not accomplish this goal is shortsighted. Long-term follow-up studies indicate an excellent outcome, with ∼90% of children free of disease 15–20 years after surgery. In contrast, 15–25% of patients in whom the diagnosis is made on the basis of a palpable thyroid nodule die from the disease within 15–20 years.
In adults with MTC >1 cm in size, metastases to regional lymph nodes are common (>75%). Total thyroidectomy with central lymph node dissection and selective dissection of other regional chains provides the best chance for cure. In patients with extensive local metastatic disease in the neck, external radiation may prevent local recurrence or reduce tumor mass but is not curative. Chemotherapy with combinations of adriamycin, vincristine, cyclophosphamide, and dacarbazine may provide palliation. Clinical trials with small compounds (tyrosine kinase inhibitors) that interact with the ATP-binding pocket of the RET, vascular endothelial receptor, and type 2 and epidermal growth factor receptors and prevent phosphorylation have shown promise for treatment of hereditary and sporadic MTC. A phase I trial of vandetanib has shown that 45% of patients have a 30% or greater reduction of tumor size and prolongation of progression-free survival by at least 11 months. Similar phase II results have been observed for XL184, sunitinib, tipifarnib, and sorafenib, and phase II trials of E7080 and pazopanib are under way. It seems likely that one or more of these compounds will be approved for treatment of metastatic MTC within the next few years.
The long-term goal for management of pheochromocytoma is to prevent death and cardiovascular complications. Improvements in radiographic imaging of the adrenals make direct examination of the apparently normal contralateral gland during surgery less important, and the rapid evolution of laparoscopic abdominal or retroperitoneal surgery has simplified management of early pheochromocytoma. The major question is whether to remove both adrenal glands or remove only the affected adrenal at the time of primary surgery. Issues to be considered in making this decision include the possibility of malignancy (<15 reported cases), the high probability of developing pheochromocytoma in the apparently unaffected gland over an 8- to 10-year period, and the risks of adrenal insufficiency caused by removal of both glands (at least two deaths related to adrenal insufficiency have occurred in MEN 2 patients). Most clinicians recommend removing only the affected gland. If both adrenals are removed, glucocorticoid and mineralocorticoid replacement is mandatory. An alternative approach is to perform a cortical-sparing adrenalectomy, removing the pheochromocytoma and adrenal medulla and leaving the adrenal cortex behind. This approach is usually successful and eliminates the necessity for steroid hormone replacement in most patients, although the pheochromocytoma recurs in a small percentage.
Hyperparathyroidism has been managed by one of two approaches. Removal of 3.5 glands with maintenance of the remaining half gland in the neck is the usual procedure. In families in which hyperparathyroidism is a prominent manifestation (almost always associated with a codon 634 RET mutation) and recurrence is common, total parathyroidectomy with transplantation of parathyroid tissue into the nondominant forearm is preferred. This approach is discussed above in the context of hyperparathyroidism associated with MEN 1.
Other Genetic Endocrine Tumor Syndromes
A number of mixed syndromes exist in which the neoplastic associations differ from those in MEN 1 or 2 (Table 351-1).
The cause of VHL syndrome—the association of central nervous system tumors, renal cell carcinoma, pheochromocytoma, and islet cell neoplasms—is a mutation in the VHL tumor-suppressor gene. Germ-line-inactivating mutations of the VHL gene cause tumor formation when there is additional loss or somatic mutation of the normal VHL allele in brain, kidney, pancreatic islet, or adrenal medullary cells. Missense mutations been identified in >40% of VHL families with pheochromocytoma, suggesting that families with this type of mutation should be surveyed routinely for pheochromocytoma. A point that may be useful in differentiating VHL from MEN 1 (overlapping features include islet cell tumor and rare pheochromocytoma) or MEN 2 (overlapping feature is pheochromocytoma) is that hyperparathyroidism rarely occurs in VHL.
The molecular defect in type 1 neurofibromatosis inactivates neurofibromin, a cell membrane–associated protein that normally activates a GTPase. Inactivation of this protein impairs GTPase and causes continuous activation of p21 Ras and its downstream tyrosine kinase pathway. Endocrine tumors also form in less common neoplastic genetic syndromes. These include Cowden disease, Carney complex, familial growth hormone and prolactin tumors, and familial carcinoid syndrome. Carney complex includes myxomas of the heart, skin, and breast; peripheral nerve schwannomas; spotty skin pigmentation; and testicular, adrenal, and GH-secreting pituitary tumors. Linkage analysis has identified two loci: chromosome 2p in half of the families and 17q in the others. The 17q gene has been identified as the regulatory subunit (type IA) of protein kinase A (PRKA1A). Familial growth hormone– or prolactin-producing neoplasms without other manifestations of MEN 1 are caused by germ-line-inactivating mutation of the aryl hydrocarbon receptor interacting protein (AIP). It is transmitted in an autosomal dominant manner. Other types of endocrine tumors have not, to date, been associated with AIP mutations.