Endocrine malignancies, although relatively uncommon, are often difficult to diagnose and treat effectively. This chapter will focus on thyroid and parathyroid cancers. (A discussion of carcinoid tumors, insulinomas, gastrinomas, and other gastrointestinal neuroendocrine tumors, as well as adrenocortical cancer, can be found in the “Pancreatic, Neuroendocrine GI, and Adrenal Cancers” chapter.)
Thyroid cancer is the most common endocrine cancer and the ninth most common cancer overall. The American Cancer Society estimates that about 64,300 new cases of thyroid cancer will be diagnosed in the United States in 2016 (49,350 in women and 14,950 in men). Thyroid cancer represents 3.8% of the new cancer cases. Although most cancers are either stable or declining in incidence, the incidence of thyroid cancer continues to increase. Rates for new thyroid cancer cases have been rising on average 5.5% each year over the last 10 years. Death rates have been rising on average by 0.8% each year from 2002 through 2011. Despite standard treatment, an estimated 1,980 deaths from thyroid cancer (1,070 women, 910 men) will occur in 2016. The prevalence rate for occult thyroid cancers found at autopsy is 5% to 10%, except in Japan and Hawaii, where the rate can be as high as 28%. Autopsy rates do not correlate with clinical incidence.
Most patients are between the ages of 45 and 64 years at the time of thyroid carcinoma diagnosis. Median age at diagnosis is 50 years. About 2% of thyroid cancers occur in children and teenagers. Women are affected more often than men (3:1 ratio) and are usually diagnosed at a younger age. The most common stage at diagnosis is localized disease (confined to the thyroid) in 68% of cases, followed by regional disease (spread to regional lymph nodes) in 26% of cases, distant metastases in 4%, and unknown in 2% of patients. The 5-year survival for patients with localized thyroid cancer is 99.9%. The overall survival depends on stage; age; and in patients with differentiated thyroid cancer, the ability of the tumor to take up radioactive iodine (RAI). The 10-year overall survival is significantly reduced to < 10% in patients who are older, have distant metastases (macronodular lung metastases or bone metastases), and are RAI non-avid.
The prevalence of thyroid nodules in the general population is 4% to 7%, with nodules more common in females than in males. The prevalence of thyroid cancer in a solitary nodule or in multinodular thyroid glands is 10% to 20%; this increases with irradiation of the neck in children and older men (see section on “Etiology and risk factors“).
Thyroid cancer is classified into four main types according to its morphology and biologic characteristics. Papillary and follicular carcinomas, also known as differentiated thyroid cancers, account for more than 90% of thyroid malignancies and constitute approximately 0.8% of all human malignancies. Medullary thyroid cancers represent < 4% of all thyroid carcinomas. About 75% of patients with medullary cancer have a sporadic form of the disease; the remaining 25% have inherited disease. Anaplastic carcinoma represents less than 2% of all thyroid carcinomas.
Papillary thyroid carcinoma
Papillary thyroid carcinoma is the most common subtype, and it typically has an excellent prognosis. Most papillary carcinomas contain varying amounts of follicular tissue, since they derive from thyroid follicular cells. When the predominant histology is papillary, the tumor is considered to be a papillary carcinoma. Because the mixed papillary-follicular variant tends to behave like a pure papillary cancer, it is treated in the same manner and has a similar prognosis.
Papillary thyroid carcinomas are unilateral in most cases, but they also can be multifocal. They vary in size from microscopic to large cancers that may invade beyond the thyroid tissue and infiltrate into contiguous structures. Papillary tumors tend to invade the lymphatics, but vascular invasion (and hematogenous spread) is uncommon. Up to 40% of adults with papillary thyroid cancer may present with regional lymph node metastases, usually ipsilateral. Distant metastases occur, in decreasing order of frequency, in the lungs, bones, and other soft tissues. Older patients have a higher risk of locally invasive tumors and of distant metastases. Children may present with a solitary thyroid nodule, but cervical node involvement is common in this age group; up to 10% of children and adolescents may have lung involvement at the time of diagnosis.
Follicular thyroid carcinoma
Follicular thyroid carcinoma is less common than papillary thyroid cancer, occurs in older age groups, and has a slightly worse prognosis. Follicular thyroid cancer can metastasize to the lungs and bones, often retaining the ability to accumulate RAI (which can be used for therapy). Metastases may be appreciated many years after the initial diagnosis. Follicular thyroid carcinoma, although frequently encapsulated, commonly exhibits microscopic vascular and capsular invasion. Microscopically, the nuclei tend to be large and have atypical mitotic figures. There is usually no lymph node involvement.
Follicular carcinoma can be difficult to distinguish from its benign counterpart, follicular adenoma. This distinction is based on the presence or absence of capsular or vascular invasion, which can be evaluated after surgical excision but not by fine-needle aspiration (FNA).
Hürthle cell, or oxyphil cell, carcinoma is a variant of follicular carcinoma. Hürthle cell carcinoma is composed of sheets of Hürthle cells and has the same criteria for malignancy as does follicular carcinoma. Hürthle cell carcinoma is thought to have a worse outcome than follicular carcinoma and is less likely to concentrate RAI.
Medullary thyroid carcinoma
Medullary thyroid carcinoma originates from the C cells (parafollicular cells) of the thyroid and secretes calcitonin. Secretory diarrhea and flushing, related to secretion of neuroendocrine peptides, are clinical features of advanced medullary thyroid carcinoma in 30% of cases. On gross examination, most tumors are often found in the upper two-thirds of the thyroid gland and are firm, grayish, and gritty.
Hereditary medullary thyroid carcinoma, which accounts for 25% of all cases, typically presents as a bilateral, multifocal process. Histologically, hereditary medullary carcinoma of the thyroid does not differ from the sporadic form, which usually presents as a solitary thyroid mass. However, the hereditary form is frequently associated with a background of C-cell hyperplasia. Another characteristic feature of hereditary medullary carcinoma is the presence of amyloid deposits.
There are two hereditary forms: multiple endocrine neoplasia type 2A (MEN-2A), characterized by medullary thyroid cancer, pheochromocytomas, and hyperparathyroidism; and multiple endocrine neoplasia type 2B (MEN-2B), characterized by medullary thyroid cancer, marfanoid habitus, pheochromocytomas, and mucosal neuromas and gastrointestinal ganglioneuromas. Familial medullary thyroid carcinoma is now considered a variant of MEN-2A in which family members do not demonstrate any of the other endocrine tumors. These syndromes are associated with germ-line mutations of the RET proto-oncogene, which codes for a receptor tyrosine kinase, called RET. Hereditary medullary thyroid cancer is inherited as an autosomal dominant pattern with high penetrance and variable phenotypic expression. (For a discussion of genetic testing to screen for RET mutations in MEN kindreds, see section on “Diagnostic workup.”)
Approximately 40% of sporadic (ie, nonhereditary) medullary thyroid carcinomas contain somatic RET mutations and 13% will have a RAS somatic mutation (predominantly the HRAS type), which may represent potential therapeutic targets.
Metastases to cervical and mediastinal lymph nodes are found in half of patients and may be present at the time of initial presentation. Distant metastases to the lungs, liver, bones, and adrenal glands most commonly occur late in the course of the disease.
Anaplastic tumors are high-grade neoplasms characterized histologically by a high mitotic rate and lymphovascular invasion. Aggressive invasion of local structures is common, as are lymph node metastases. Distant metastases tend to occur in patients who do not succumb early to regional disease. Occasional cases of anaplastic carcinoma have been shown to arise from pre-existing differentiated thyroid carcinoma or in a pre-existing goiter.
Other tumor types
Lymphomas of the thyroid account for less than 5% of primary thyroid carcinomas. Other tumor types, such as teratomas, squamous cell carcinomas, and sarcomas, may also rarely cause primary thyroid cancers.
Renal cell carcinoma is the most common extra-thyroidal tumor to metastasize to the thyroid. It accounts for almost half (42%) of patients with this condition. Colorectal cancer, lung cancer, breast cancer, and sarcomas account for 26%, 19%, 11%, and 6%, respectively, of metastases to the thyroid. Females show a slight predilection for metastases to the thyroid from nonthyroidal tumors. Of metastases to the thyroid gland, 44.2% occur in glands with abnormalities, such as primary thyroid neoplasms and benign thyroid conditions. The interval between diagnosing the primary tumors and their metastases to the thyroid gland is 4.5 months to 75 months.
Etiology and Risk Factors
Differentiated thyroid cancer
Radiation–induced thyroid cancer. Radiation exposure of the thyroid during childhood is the most clearly defined environmental factor associated with benign and malignant thyroid tumors. The predominant types of radiation are therapeutic external radiation for the treatment of cancer, historical use of external radiation to treat a wide variety of nonmalignant conditions, and exposure to nuclear fallout (from testing or accidents or in Japanese survivors of atomic bombing and children living in the area of Chernobyl). External low-dose radiation therapy to the head and neck during infancy and childhood, frequently used between the 1940s and 1960s for the treatment of a variety of benign diseases, has been shown to predispose an individual to thyroid cancer. The younger a patient is at the time of radiation exposure, the higher is the subsequent risk of developing thyroid carcinoma. Also, as mentioned previously, women are at increased risk of radiation-induced thyroid cancer. There is a latency period ranging from 10 to 30 years from the time of low-dose irradiation to the development of thyroid cancer.
As little as 11 cGy and as much as 2,000 cGy of external radiation to the head and neck have been associated with a number of benign and malignant diseases. It was once thought that high-dose irradiation (> 2,000 cGy) to the head and neck did not increase the risk of neoplasia. However, it has been shown that patients treated with mantle-field irradiation for Hodgkin lymphoma are at increased risk of thyroid carcinoma compared with the general population, although they are more likely to develop hypothyroidism than thyroid cancer.
Radiation-associated thyroid cancer has a natural history and prognosis identical to those of sporadic thyroid cancer.
Family history. A first-degree relative with history of thyroid cancer also increases the risk of being diagnosed with thyroid cancer, although the genetic basis is unclear.
Hereditary cancer syndromes. Individuals with certain familial cancer syndromes can develop differentiated thyroid cancer. Familial adenomatous polyposis (FAP), caused by a mutation in the APC gene, can predispose affected persons to development of papillary thyroid cancers. Cowden disease is associated with thyroid cancer of the papillary or follicular type. This syndrome is most often caused by defects in PTEN. Carney complex, type I is associated with an increased risk for papillary and follicular thyroid cancers. This syndrome is caused by defects in the PRKAR1A gene.
Besides radiation-induced thyroid cancer, there are sparse data on the etiology of differentiated thyroid cancer. There has been intensive research on distinguishing molecular factors important for cell differentiation, growth, and motility. Considerable attention has focused on BRAF, a member of the RAF family of serine/threonine kinases that mediates cellular responses to growth-promoting signals via the RAS-RAF-MEK-MAPK signaling pathway. BRAF mutations so far have only been documented in papillary thyroid carcinoma (45%) and papillary thyroid carcinoma–derived anaplastic thyroid carcinoma (25%). Presence of the somatic BRAF V600E mutation in papillary thyroid carcinoma has been reported to correlate with aggressive tumor characteristics (extrathyroidal extension, advanced tumor stage at presentation, lymph node or distant metastases). Patients with BRAF mutations have higher rates of mortality and are typically less responsive to radioiodine therapy. Because of this, BRAF mutations have been implicated as potential prognostic factors and therapeutic targets. In addition, because angiogenesis is critical for survival of tumors, vascular endothelial growth factor (VEGF) expression in papillary thyroid carcinoma correlates with decreased disease-free survival, and presence of BRAF mutation is associated with a higher risk of metastasis and recurrence.
Sidebar: Liu et al have reported on the relevance of mutations of the telomerase reverse transcriptase (TERT) in thyroid cancer, which, in association with BRAF V600E mutation, seem to portend a worse outcome. They found the transcriptional activity of the mutated TERT promoter was several-fold higher than that of the wild-type TERT promoter in thyroid cancer cells. BRAF V600E promotes the expression of the mutated TERT promoter. BRAF and TERT knockouts have decreased cellular migration and proliferation. This may explain the additive effect of these two mutations in aggressive cellular thyroid cancer and tumor behaviors of the disease (Liu X et al: Endocr Relat Cancer 20:603–610, 2013).
Medullary thyroid cancer
The notable risk factor for medullary thyroid cancer is having a germ-line mutation of the RET proto-oncogene (see information in this chapter on genetic testing). Approximately 25% of cases of medullary thyroid cancer are associated with heritable syndromes due to a RET mutation, multiple endocrine neoplasia types 2A (MEN-2A) and 2B (MEN-2B). Exome sequencing of sporadic medullary thyroid cancer showed mutations in RET, HRAS, and KRAS genes, suggesting that these mutations are the principal drivers of oncogenesis.
Follicular thyroid cancer
RAS gene mutations are identified in 20% to 35% of follicular thyroid carcinomas, and may represent a potential drive to development of this type of thyroid cancer.
Anaplastic thyroid cancer
Approximately 20% of patients with anaplastic thyroid cancer have a history of differentiated thyroid cancer. The majority of synchronous thyroid tumors are papillary cancers, but coexisting follicular cancers have also been reported. Anaplastic thyroid cancer develops from more differentiated tumors as a result of one or more dedifferentiating events. Since activating mutations in BRAF and RAS are seen in both well-differentiated thyroid malignancies and anaplastic thyroid cancer, these are presumed to be early events in the progression pathway. Late events that are seen more commonly in the anaplastic tumor, rather than the precursor well-differentiated tumor, include mutations in TP53, 16p, CTNNB1, and PIK3CA.
Signs and Symptoms
Most thyroid cancers present as asymptomatic thyroid nodules. Patients may feel pressure symptoms from nodules as they begin to increase in size. A change in the voice can be caused by a thyroid cancer or benign goiter. The voice change usually occurs when there is compression of the larynx or invasion of the recurrent laryngeal nerve. Secretory diarrhea and flushing can be symptoms suggestive of advanced medullary thyroid cancer.
On physical examination, a thyroid nodule that is hard or firm and fixed may represent a cancer. The presence of palpable enlarged nodes in the lateral neck, even in the absence of a palpable nodule in the thyroid gland, could represent metastases to the lymph nodes.
As mentioned previously, thyroid nodules are present in 4% to 7% of the general population and in a higher percentage of individuals who have had irradiation to the head and neck region. Most thyroid nodules are benign (colloid nodules or adenomas); therefore, it is important for the workup to lead to surgical resection for malignant nodules and to avoid unnecessary surgery for benign lesions. Although most solid nodules are benign, thyroid carcinomas usually present as solid nodules. A cystic nodule or a “mixed” (cystic-solid) lesion is less likely to represent a carcinoma and more likely to be a degenerated colloid nodule. Molecular testing can be performed on nodules with an indeterminate FNA cytology to help with therapeutic decisions.
History and physical examination. With the discovery of a thyroid nodule, a complete patient history should be taken and physical examination focusing on the thyroid gland and adjacent cervical lymph nodes should be performed. Pertinent historical factors include a history of childhood head and neck irradiation, total body irradiation for bone marrow transplant, family history of thyroid carcinoma or thyroid cancer syndrome in a first-degree relative, exposure to ionizing radiation from fallout in childhood or adolescence, rapid growth of the nodule, and hoarseness. In patients with a history of irradiation to the head and neck, the risk of cancer is higher (by as much as 50%) than in non-irradiated patients (10% to 20%). Nodules that occur in either the very young or the very old are likely to be cancerous, particularly in men. Also, a new nodule or a nodule that suddenly begins to grow is worrisome. Pertinent physical findings suggesting possible malignancy include vocal cord paralysis, lateral cervical lymphadenopathy, and fixation of the nodule to surrounding tissues.
Imaging modalities. Ultrasonographic and radionuclide (radioiodine and technetium) scans are also used in the evaluation of thyroid nodules.
• Ultrasonography—Ultrasonography is now widely considered an essential tool in the assessment of thyroid nodules. Thyroid ultrasonography should be performed in all patients with known or suspected thyroid nodules. Certain features are associated with malignancy and can guide physicians in deciding which nodules should be biopsied. A consensus statement from the Society of Radiologists in Ultrasound outlined various features of solitary nodules associated with thyroid cancer: microcalcifications, hypoechogenicity, irregular margins or no halo, solid composition, intranodule vascularity, and nodules that are taller than they are wider on a transverse plane. No single feature has both high sensitivity and specificity; however, the combination of two or more factors can increase the likelihood of cancer. Certain ultrasonographic appearances may also be highly predictive of a benign nodule—a pure cystic nodule is highly unlikely to be malignant. A spongiform appearance, defined as an aggregation of multiple microcystic components in more than 50% of the nodule volume, is 99.7% specific for identification of a benign thyroid nodule. Although there is a decrease in cancer rate per nodule in patients with multiple nodules, the overall rate of thyroid cancer per patient is similar to that seen in patients with a solitary nodule.
Thyroid cancer is most often found in the dominant, or largest, nodule in multinodular glands; however, approximately one-third of the cases of cancer are found in nondominant nodules. Nodule size is a poor predictor of malignancy, because the likelihood of cancer has been shown to be the same regardless of nodule size. Multiple ultrasonographic features other than size need to be considered in determining which nodules are more likely to be malignant and thus should be biopsied, including increased intranodular vascularity, hypoechogenicity of a solid nodule, microcalcifications, or presence of abnormal cervical lymph nodes. Patients with multiple thyroid nodules have the same risk of malignancy as those with solitary nodules, and selective FNA biopsy based on suspicious ultrasonographic findings is performed for further diagnosis.
• Radioactive iodine uptake and scan—If the serum thyroid stimulating hormone (TSH) level is subnormal, a radioiodine thyroid scan should be obtained to document whether the nodule is hyperfunctioning, or “hot” (ie, tracer uptake is greater than the surrounding normal thyroid); isofunctioning, or “warm” (ie, tracer uptake in the nodule is equal to that of the surrounding thyroid); or nonfunctioning, or “cold” (ie, tracer uptake is less than in the surrounding thyroid tissue). Thyroid isotope scans cannot differentiate absolutely a benign from a malignant nodule but can, based on the functional status of the nodule, assign a probability of malignancy. Most thyroid carcinomas occur in cold nodules, but only 10% of cold nodules are carcinomas. If the cytology reading reports a follicular neoplasm, an I-123 thyroid scan may be considered if it has not already been done, especially if the serum TSH level is in the low-normal range. If a concordant autonomously functioning nodule is absent on the radionuclide scan, lobectomy should be considered.
Computed tomography (CT) or magnetic resonance imaging (MRI) scans of the neck may be appropriate in some cases.
A fluorodeoxyglucose positron emission tomography (FDG-PET) scan is shown to improve diagnostic accuracy of indeterminate thyroid nodules, but results vary among studies. A meta-analysis by Bertagna et al found that the incidence of thyroid “incidentalomas” on FDG-PET or PET/CT was 2.46%. For solitary nodules with FDG-avidity, a malignancy ratio was found to be 34.6%. Diffuse uptake in the thyroid is commonly due to benign disease, such as thyroiditis.
FNA. FNA has become the most accurate and cost-effective initial diagnostic test for the evaluation of thyroid nodules, and it is the procedure of choice. The 2009 guidelines by the American Thyroid Association (ATA) recommend ultrasound-guided FNA for evaluating thyroid nodules; this can determine whether the lesion is cystic or solid. Ultrasound guidance is preferred over palpation to localized nodules and leads to a higher likelihood of diagnostic cytology (> 25% to 50% cystic component) or sampling error (difficult to palpate or posteriorly located nodules). A prospective study showed that ultrasound-guided FNA was more cost-effective than FNA by palpation. For solid lesions, cytology can yield one of three results: benign, malignant, or indeterminate. The accuracy of cytologic diagnosis from FNA is 70% to 80%, depending on the experience of the person performing the aspiration and the pathologist interpreting the cytologic specimen. FNA biopsy results are divided into four categories: nondiagnostic, malignant, indeterminate or suspicious for neoplasm, and benign. In a series of 98 “suspicious” FNA biopsies, findings of cellular atypia (pleomorphism, enlarged nuclei, nuclear grooves, coarse or irregular chromatin, prominent or multiple nucleoli, or atypical or numerous mitotic figures) or follicular lesions with atypia were associated with malignancy 20% and 44% of the time, respectively. Follicular lesions without atypia have a 6.7% risk of malignancy. Use of the Bethesda System recommended by the National Cancer Institute for reporting thyroid cytopathology allows for more systematic assessment of risk of malignancy and helps in determining follow-up and a further course of management.
Core needle biopsy has been used as an alternative method for diagnosis. Some studies have shown that the adequacy of sample may be greater with core needle biopsy than with FNA. However, there are conflicting reports as to whether a core needle biopsy offers greater accuracy in the diagnosis of a thyroid nodule.
Lymph nodes may also be sampled. Thyroglobulin washout performed on cystic lymph nodes can help detect the presence of metastatic disease.
Molecular markers. Approximately 5% to 30% of aspirations yield indeterminate cytologic findings, which include three subtypes: atypia (or follicular lesion) of undetermined significance, follicular neoplasm or suspicious for follicular neoplasm, and suspicious for malignancy. Improvement in the assessment of indeterminate FNA biopsy results may allow better risk stratification. Certain clinical features can improve the diagnostic accuracy for malignancy in patients with indeterminate cytology, but overall predictive values are still low. These include male sex, nodule size (> 4 cm), older patient age, and cytologic features such as presence of atypia.
In the United States, there are two commercially available approaches to the molecular characterization of FNA aspirates: molecular markers of malignancy and high-density genomic data for molecular classification.
The use of molecular markers (either for mutations or gene rearrangements of BRAF, RAS, RET/PTC, and PAX8-PPARγ) may be considered for patients with indeterminate cytology on FNA, to help guide management. These genetic markers have high specificity and a high positive predictive value; therefore, they can be used to identify which indeterminate nodules are malignant. However, they fail to rule out cancer with sufficient certainty to avoid surgery in most patients with indeterminate nodules.
Recent studies have described the development of gene-expression classifiers that better distinguish benign from malignant thyroid nodules. With the use of the gene-expression classifier, the negative predictive value was 95% for aspirates classified as atypia (or follicular lesions) of undetermined significance and 94% for aspirates classified as follicular neoplasms or lesions suspicious for follicular neoplasm. These data suggest consideration of a more conservative approach (observation or ultrasound) for most patients with thyroid nodules that are cytologically indeterminate on FNA and benign according to gene-expression classifier results.
Thyroglobulin. Thyroglobulin (Tg), normally synthesized in the follicular epithelium of the thyroid, is present in well-differentiated papillary and follicular carcinomas, and is noted infrequently in anaplastic carcinomas, but is not observed in medullary carcinomas. Therefore, thyroglobulin immunoreactivity is considered to be indicative of a follicular epithelial origin. Routine measurement of serum Tg for initial evaluation of thyroid nodules is not recommended by the 2009 ATA guidelines. Measurements of serum Tg provide important information about the presence or absence of residual, recurrent, or metastatic disease in patients with differentiated thyroid cancer. Limitations of serum Tg assays include interassay variability and the high prevalence of anti-Tg antibodies, which may interfere with Tg assay results. Testing should be done using a sensitive assay, ideally using the same assay for each sample. Thyroglobulin antibodies should be measured with each sample. Anti-Tg antibodies are present in 25% to 30% of patients with differentiated thyoid cancer. Due to their superior sensitivity, immunometric assays are becoming the standard of care. However, antibody interference can cause underestimation of Tg levels. In contrast, the radioimmunoassay and liquid chromatography–tandem mass spectrometry (LC-MS/MS) methods resist antibody interference but have lower sensitivity. Laboratories often reflex Tg measurement to radioimmunoassay or LC-MS/MS when the serum Tg antibody is positive.
In the absence of antibody interference, serum Tg has a high degree of sensitivity and specifcity to detect thyroid cancer, especially after total thyroidectomy and remnant ablation, with the highest degrees of sensitivity noted following thyroid hormone withdrawal or stimulation using recombinant human TSH (rhTSH). A single rhTSH-stimulated serum Tg level < 0.5 ng/mL in the absence of anti-Tg antibody has an approximately 98% to 99.5% likelihood of identifying patients who are completely free of tumor on follow-up.
Calcitonin level. Calcitonin is a specific product of thyroid C cells (parafollicular cells). The routine measurement of serum calcitonin in patients with thyroid nodules is controversial and is not currently recommended in the United States, given the rarity of the disease. However, if obtained, a basal or stimulated serum calcitonin level of 100 pg/mL or greater should be interpreted as suspicious for medullary thyroid carcinoma, and further evaluation and treatment should ensue.
The serum calcitonin concentration should be measured preoperatively in patients with medullary thyroid cancer and in carriers of a RET mutation for comparison with postoperative values. It is a sensitive marker of residual medullary thyroid carcinoma. When the postoperative basal serum calcitonin is undetectable, the risk of persistent or recurrent residual disease is low. In patients who have clinically palpable medullary carcinoma, the basal calcitonin level is almost always elevated. In patients with smaller tumors or C-cell hyperplasia, the basal calcitonin level may be normal, but administration of synthetic gastrin (pentagastrin) or calcium results in marked elevation of calcitonin levels. The use of calcitonin levels as a tumor marker and stimulation screening in hereditary forms of medullary cancers has been largely replaced by genetic testing (see below).
Carcinoembryonic antigen (CEA). Serum CEA levels may be elevated in patients with medullary thyroid cancer. The serum markers (calcitonin and CEA) are important in the follow-up of patients with medullary thyroid cancer, and they should be measured 2 to 3 months postoperatively.
Ruling out pheochromocytoma and hyperparathyroidism. Medullary thyroid carcinoma can be associated with MEN-2A or MEN-2B. Both the MEN-2A and MEN-2B syndromes are characterized by medullary thyroid cancer and pheochromocytoma. Thus, in any patient with hereditary medullary thyroid cancer, it is imperative that the preoperative workup include a determination of 24-hour urinary catecholamine and metanephrine levels to rule out the presence of a pheochromocytoma. Fractionated plasma metanephrine levels have been demonstrated to have a high sensitivity and may be included in the initial assessment. According to the ATA 2009 guidelines, exclusion of a pheochromocytoma may include any of the following tests: (1) negative RET proto-oncogene analysis and family history; (2) negative plasma free metanephrines and normetanephrines, or negative 24-hour urinary metanephrines and normetanephrines; (3) negative findings on adrenal CT or MRI scans. Given the possibility that any patient with medullary thyroid cancer may have MEN-2, preoperative testing must also include measurement of serum calcium to rule out primary hyperparathyroidism, which requires concomitant surgical intervention.
Genetic testing. Germ-line mutations in the RET proto-oncogene are responsible for familial non-MEN medullary thyroid cancer, in addition to MEN-2A and MEN-2B. DNA analysis performed on a peripheral blood sample is a highly reliable method for identifying the presence of an RET mutation. The 2009 management guidelines of the ATA regarding medullary thyroid cancer recommend that all patients with FNA or calcitonin diagnostic or suspicious for medullary thyroid cancer undergo RET mutation analysis, ideally performed with genetics counseling and completed preoperatively. Approximately 95% of patients with an RET mutation will eventually develop medullary thyroid cancer; thus, prophylactic surgical treatment is recommended. The specific mutated codon of RET may correlate with the aggressiveness of medullary carcinoma of the thyroid. This should be considered when counseling affected individuals and their families regarding prophylactic thyroidectomy and the age at which to perform such surgery. Long-term data regarding the effectiveness of prophylactic thyroidectomy based on RET testing are scarce at this time. In a report of 50 patients (19 years of age and younger) treated surgically after positive RET mutation analysis, 33 had carcinoma identified in the surgical specimen. At the time of the publication, 44 patients were found to be free of disease more than 5 years after surgery.
Recommended ages for prophylactic surgery range from within the first 6 months of life to 10 years of age, depending on the mutation. The prophylactic surgical procedure of choice is total thyroidectomy with or without central lymph node dissection.
Periodic determinations of stimulated calcitonin levels may help establish the early diagnosis of medullary thyroid cancer in those who do not undergo surgery, but they will not always prevent development of metastatic medullary thyroid carcinoma.
At this time, no organization recommends periodic screening for thyroid cancer using neck palpation or ultrasonography in average-risk, asymptomatic adults. However, the American Cancer Society recommends examination of the thyroid during a routine checkup, since this surveillance can result in case findings.
Staging and Prognosis
Unlike most other cancers, in which staging is based on the anatomic extent of disease, the American Joint Committee on Cancer (AJCC) and International Union Against Cancer (UICC) staging of thyroid cancer also takes into consideration patient age at the time of diagnosis and tumor histology (Table 1).
Differentiated thyroid cancers
Recurrence and death following initial treatment of differentiated thyroid cancer can be predicted using a number of risk-classification schemes. Some commonly used systems are Tumor Node Metastasis (TNM) or Metastases, Age, Completeness of Resection, Invasion, Size (MACIS); the AMES (age, metastases, extent, and size) and AGES (age, grade, extent, and size) classifications; and then an additional clinicopathologic staging system, such as the American Thyroid Association (ATA) system, to estimate the risk of recurrence.
Low-risk patients are generally those younger than 45 years with low-grade nonmetastatic tumors that are confined to the thyroid gland and are less than 1 to 5 cm. Low-risk patients enjoy a 20-year survival rate of 97% to 100% after surgery alone.
High-risk patients are those 45 years and older with a high-grade, metastatic, locally invasive tumor in the neck or with a large tumor. Large size is defined by some authors as more than 1 cm and by others as more than 2 or more than 5 cm. The 20-year survival rate in the high-risk group drops to between 54% and 57%.
Intermediate-risk patients include young patients with a high-risk tumor (metastatic, large, locally invasive, or high grade) or older patients with a low-risk tumor. The 20-year survival rate in this group of patients is approximately 85%. Increasing age and tumor size, male sex, poorly differentiated carcinoma, lymph node involvement, and regional and metastatic disease were associated with increased cumulative incidence of death resulting from thyroid cancer.
Medullary thyroid carcinoma
Medullary thyroid carcinoma is associated with an overall 10-year survival rate of 40% to 60%. When medullary carcinoma is discovered before it becomes palpable, the prognosis is much better: patients with stage I medullary tumors (ie, tumors ≤ 2 cm or nonpalpable lesions detected by screening and provocative testing) have a 10-year survival rate of 95%.
Stage II medullary cancers (tumors > 2 cm but < 4 cm) are associated with a survival rate of 50% to 90% at 10 years. Patients who have lymph node involvement (stages III and IVA disease) have a 10-year survival rate of 15% to 50%. Unfortunately, approximately 50% of patients have lymph node involvement at the time of diagnosis.
When there are distant metastases (stages IVB and IVC), long-term survival is compromised. In patients with metastatic medullary thyroid cancer, the disease often progresses at a very slow rate, and patients may remain alive with disease for many years. Doubling time of calcitonin and CEA are predictive of prognosis. In a 2005 study by Barbet et al of patients with medullary thyroid cancer, those with a calcitonin doubling time of < 6 months had a survival of 25% at 5 years and 8% at 10 years versus 100% survival among patients with a calcitonin doubling time of > 2 years. The 2009 ATA management guidelines for medullary thyroid cancer recommend monitoring of doubling time of CEA and calcitonin. Frequency of surveillance has been recommended on the basis of the doubling time calculation for calcitonin and CEA. Patients with calcitonin or CEA doubling times of > 2 years typically do not require systemic therapy, and such treatment should only be initiated after thorough discussion with the patient. Patients with rapidly progressing disease (doubling times of < 2 years) should be considered for treatment.
The ATA website has a calculator for CEA and calcitonin doubling time.
Sidebar: Ho et al recently reported on a postoperative nomogram for predicting cancer-specific mortality in medullary thyroid cancer. The authors identified a total of 249 patients with medullary thyroid cancer managed entirely at one institution between 1986 and 2010. Patient, tumor, and treatment characteristics were recorded from patient charts. A risk model was then built to predict the 10-year cumulative incidence of medullary thyroid cancer. All predictors of interest were added in the starting full model before selection, including age, gender, preoperative and postoperative serum calcitonin, preoperative and postoperative CEA, RET mutation, perivascular invasion, margin status, M status, pathologic N status, and pathologic T status. Of 249 patients, 22.5% (56 of 249) died from medullary thyroid cancer, while 6.4% (16 of 249) died secondary to other causes. The mean follow-up period was 87 ± 67 months. The seven variables with the highest predictive accuracy included age, gender, postoperative calcitonin level, perivascular invasion, pathologic T status, pathologic N status, and M status.
Anaplastic thyroid cancer
Anaplastic thyroid cancer does not have a generally accepted staging system, and all patients are classified as having stage IV disease. Anaplastic carcinoma is highly malignant and has a poor 5-year survival rate (0% to 25%). Most patients die of uncontrolled local disease within several months of diagnosis.