A normal human thyroid gland contains 20 to 30 million spherical follicles lined with follicular epithelial cells filled with colloid and stores a 3 months’ supply of thyroid hormone.1 Among the several differentiated thyroid cancers that originate from thyroid follicular cells, 10% to 15% are follicular thyroid carcinomas (FTCs) and 3% to 5% are oncocytic (Hürthle) cell carcinomas (HCCs).2 Recent studies suggest that FTC may represent only 5% of differentiated thyroid cancer in geographical regions with well-supplemented iodine diets. In iodine-deficient regions, however, FTC may account for up to 25% to 40% of thyroid carcinomas.3 In contrast, HCC may be more common in areas with iodine-rich diets.4 The mean age at presentation is higher for HCC (55 years) than for FTC (48 years). In most reports, both disorders are approximately twofold more common in women than men.5

Hürthle cells are also called oncocytic or oxyphilic cells. According to the most recent World Health Organization (WHO) classification, the current correct term for HCC is actually oncocytic carcinoma.6 Hürthle cells were first described by Hürthle in 1894. Although it is now thought that the cells he described were actually parafollicular C cells, what we today call Hürthle cells are derived from follicular cells and may develop in response to defects in mitochondrial DNA.7 Histologically, they are large and characterized by abundant eosinophilic (pink) cytoplasm with a large nucleus. Hürthle cells are seen in patients with benign reactive thyroid diseases such as autoimmune thyroiditis, multinodular goiter, and Graves’ disease and in patients treated with systemic chemotherapy.8

HCC is often classified as a variant of FTC, but a number of differences are seen (Table 5-1). Compared with FTC, HCC is associated with a higher risk of local recurrence (24% vs. 3%),5 and unlike other differentiated thyroid carcinomas, HCC is not as iodine avid.9 Distant metastatic disease develops in up to 20% of FTC patients (most commonly to the bone and lung) and in 30% of HCC patients (to the lung and lymph nodes). Survival is equivalent when stratified by stage and at 5 years, but at 10 years, patients with HCC have a poorer survival (76% vs. 85%).2,10

Based on studies identifying regions of differentiated thyroid carcinoma within anaplastic tumor, FTC or HCC may dedifferentiate into more aggressive poorly differentiated or anaplastic thyroid carcinoma. The molecular mechanisms that cause this progression of malignancy are the subject of ongoing study.11,12

Ionizing Radiation

Exposure to ionizing radiation is a risk factor for both Hürthle cell and follicular thyroid neoplasms, with exposed women at higher risk than men. The risk is also dependent on the dose received and the individual’s age at the time of exposure.13 After exposure, the risk appears to peak at 25 to 30 years but continues to be significant for up to 40 years.14 The mean latency period for FTC is shorter (21 years) than for benign follicular adenoma (FA) (35 years), and radiation exposure does not alter the prognosis in either HCC or FTC.15

Inherited Syndromes

Isolated familial FTC and familial HCC are both very rare. In a study from Japan, the incidence of FTC among patients who had one or more first-degree relatives with FTC was 1.9%, and familial cases had no prognostic differences compared with sporadic cases.16 In a study from the Swedish Family-Cancer Database evaluating 3292 patients with thyroid cancer, there were no cases of FTC when parents were diagnosed with any thyroid cancer.17 There was a significant association between HCC and lymphoma; in four patients with HCC, two had parents diagnosed with Hodgkin’s lymphoma and two had siblings diagnosed with non-Hodgkin’s lymphoma. Additional studies are needed to validate this association.17

FTC has been associated with Cowden syndrome and Carney complex type 1. Cowden syndrome is an unusual autosomal dominant adult-onset inherited disorder that is caused by germline mutations in the PTEN tumor suppressor gene and is associated with an increased risk of breast cancer (25% to 50%) and thyroid cancer, particularly FTC (3% to 10%). In the practice of endocrine surgery, however, the presence of macrocephaly (head circumference that is >99th percentile for age) and characteristic mucocutaneous lesions on physical examination are common and should always raise clinical suspicion for the disorder.18

Carney complex type 1 is associated with an increased risk of differentiated thyroid carcinoma (DTC) (4%), either FTC or papillary thyroid cancer (PTC), and benign FA (6%).19 Other characteristics include cardiac myxomas and pigmentations of the skin and mucosa. Mutations of the protein kinase A regulatory subunit type 1α gene have been identified in affected patients.19

Other Genetic Susceptibilities

Genetic linkage analyses of large kindreds have identified possible chromosomal regions that may be associated with benign thyroid nodules and familial nonmedullary thyroid cancers. These loci include chromosome 1q21 (fPTC/PRN), 2q21 (fNMTC1), and 19p13 (TCO).20 Linkage to TCO, in particular, has been associated with familial HCC.21

Mitochondrial DNA mutations have been proposed to be a cause in the formation of HCC.7 Approximately 15% of Hürthle cell variants of follicular and PTC have a mutation in GRIM-19, a regulatory gene involved in apoptosis and mitochondrial metabolism.21

The majority of FTC and HCC present as solitary thyroid nodules first identified on physical examination by the patient or physician or incidentally on imaging. Most patients are asymptomatic. The routine evaluation of patients with thyroid nodules should include assessment for symptoms such as dysphagia (to solids), positional dyspnea, orthopnea, anterior neck pain, and hoarseness. A history of supine dyspnea relieved by turning onto one’s side is suggestive of a substernal component causing tracheal compression (Carty, unpublished data). Symptoms of hyperthyroidism or hypothyroidism should be queried, but these diagnoses will ultimately be determined with thyroid function testing, which should include thyroid-stimulating hormone (TSH), thyroxine (T4), triiodothyronine (T3), and free thyroxine index (FTI) levels within the past 3 months. A history of observed nodule growth, either on serial imaging or physical examination, suggests thyroid malignancy. The neck examination should focus on the size of the nodule, firmness, mobility, presence of contralateral thyroid nodules, tracheal deviation, signs of hyperthyroidism, and lymphadenopathy. Patients who are hoarse or who have had prior cervical exploration should always undergo preoperative laryngoscopy to evaluate vocal fold function.

Neck ultrasonography is very useful in determining nodule size and characteristics. Worrisome sonographic characteristics include irregular margins, microcalcifications, increased intranodular vascularity, height greater than width, and hypoechogenicity.22 Ultrasonography can also easily distinguish cystic from solid nodules, determine if a nodule originates from the thyroid or is perithyroidal, and detect or exclude cervical lymphadenopathy. Lymph nodes are considered enlarged if they are larger than 1 cm in size. Irrespective of size, suspicious characteristics, including round shape, loss of echogenic hilum, and calcifications, should prompt aspiration sampling for cytologic evaluation.23 Ultrasonography may be limited by operator experience and may not be useful in the evaluation of mediastinal or low-level VI nodes.

Computed tomography may also be used with comparable sensitivity and specificity to assess for the presence of metastatic nodal disease, but it is more expensive, involves exposure to radiation, and occasionally results in a mistaken use of iodinated contrast, which may significantly delay the therapeutic administration of iodine 131 (131I) if thyroid cancer is diagnosed.24 Magnetic resonance imaging has been also shown to accurately identify nodal disease.25 Ultimately, the advantages of using ultrasonography, most notably its decreased cost and biopsy capability, have made it the modality of choice for preoperative staging.26,27

In the evaluation of thyroid nodules, fine-needle aspiration (FNA) biopsy should be the initial diagnostic test. FNA results determine clinical management. In general, all nodules that are sonographically suspicious, larger than 1 cm, or clinical concerning should undergo FNA.28 FNA may be performed guided by palpation or ultrasonography. Although palpation-guided biopsy is convenient and may be more cost effective, ultrasound guidance decreases the rate of inadequate yield and ensures that the nodule of interest is actually being accurately sampled.29,30 FNA is performed by introducing a 25-gauge, long, beveled needle attached to a syringe (with or without a pistol-grip device) into the nodule. A series of rapid advance and withdrawal motions (at a rate of ~3 excursions/sec) over 2 to 5 seconds is usually sufficient to obtain adequate sampling. The sample is then smeared onto one or two slides and immediately preserved with fixative.31 Inadequate FNA rates are usually less than 5%.29,32 These rates should be monitored by those who routinely perform FNA to ensure that equivalent rates are maintained.

Follicular neoplasms (FNs) are diagnosed in 10% to 20% of FNA biopsies and are characterized by follicular cells with a paucity of colloid. The cells may be arranged in groups or in microfollicles.33 Although Hürthle cell neoplasm (HCN) should be categorized as a distinct category separate from FN, historically this has not always been the case. The differentiation between FN and HCN is important because clinical factors may alter the risk of malignancy, potentially affecting the extent of initial diagnostic thyroid surgery. On FNA, HCN usually demonstrates minimal colloid with a hypercellular population of Hürthle cells arranged as a monolayer, groups, or single cells.33

One of the current limitations of thyroid cytology is the inability to differentiate between benign and malignant disease when an FN and HCN is diagnosed. The differential diagnosis is broad and includes FA, oncocytic adenoma, parathyroid adenoma, follicular or oncocytic variant of PTC, FTC, and HTC, and tissue histopathology is required to make the differentiation (Figure 5-1). Subsequently, the presence of FN or HCN results should prompt diagnostic thyroidectomy. Carcinoma is diagnosed in approximately 20% of FN cases and up to 35% of HCN cases.33

Cytologic and histologic differentiation between FTC and follicular variant of PTC (FVPTC) can be challenging. FVPTC is diagnosed when characteristic PTC cells have a follicular growth pattern, and it often occurs in a background of nodular goiter.3 Even among expert pathologists, the degree of concordance in differentiating between FTC, FA, and FVPTC is below 50%.34,35 Being able to differentiate between the histologic subtypes is important for clinical management; for example, whereas PTC has a propensity for lymph node metastases, distant metastatic disease is more commonly seen in patients with FTC. Immunohistochemical markers, such as CK-19, HBME-1, and Galectin-3, are expressed strongly in patients with DTC but more so in those with FVPTC than FTC and may be used to help differentiate between the three types of lesions. These markers are also sometimes expressed in patients with benign lesions.36

Recent advances in molecular pathology have identified genetic changes that may help to differentiate thyroid carcinomas from adenomas before cervical exploration. Loss of heterozygosity (LOH) as a marker of genetic instability has been used to differentiate between FA and FTC. In a panel of markers corresponding to known tumor suppressor genes, LOH was increased in FTC compared with FA. In addition, poor patient outcome was correlated with increasing genetic instability.37

Other genetic changes implicated in thyroid carcinogenesis are summarized in Table 5-2. Point mutations in the Ras gene have been associated with 40% to 50% of patients with FTC but are also seen in those with benign adenomas and FVPTC38–40 as well as in up to 50% of patients with poorly differentiated carcinomas.41Ras mutations are likely an initiating factor contributing to thyroid carcinogenesis and may predispose individuals to aggressive tumor behavior; this is under ongoing study.41,42

PAX8-PPARγ rearrangement occurs in up to 35% of patients with FTC and 10% of those with HCC.38,43 The incidence in FA is variable, ranging from 0% to 50%.43,44BRAF mutations and RET/PTC rearrangements are rare in patients with FTC and HCC. The presence of either should alert the practitioner to the possibility of a PTC.45 A diagnostic evaluation that screens FNA for these and other genetic changes commonly implicated in thyroid carcinogenesis, such as Rb, P16INK4a, and c-erbAβ may eventually prove to be very important in clinical management42,46 and could even obviate the need for diagnostic lobectomy for many patients with FN or HCN cytology results.

A number of prognostic scoring systems exist for differentiated thyroid cancer, including metastases, age, completeness of resection, invasion, size (MACIS); age, grade, extent, size (AGES); age, metastases, extent, size (AMES); and the European Organization for Research and Treatment of Cancer (EORTC) score. The most commonly used staging system is the American Joint Committee on Cancer, 6th edition TNM (tumor, node, metastasis) (Table 5-3).47 Although most of the scoring systems were validated for patients with either PTC or all differentiated thyroid cancers, studies have demonstrated that they still provide some predictive recurrence and survival information when applied to patients with FTC and HCC.48,49 All thyroid cancer clinical scoring systems are imprecise, however, and it must be kept in mind that patients in low-risk groups can still die of thyroid cancer.2,48