Thyroid ultrasonography has established itself as a popular and useful tool in the evaluation and management of thyroid disorders, both malignant and benign. Although its use has traditionally been the domain of radiologists, surgeons and endocrinologists are increasingly integrating this technology into their daily clinical and operative practice. This article provides an overview of the relevant uses and indications for ultrasound in various thyroid diseases, describes characteristic ultrasound findings in these diseases, and reviews the relevant literature and guidelines concerning its uses.
Historical perspective of thyroid ultrasound
Thyroid ultrasonography commands a central role in the evaluation, diagnosis, and treatment of thyroid disorders. Ultrasound has been the standard for imaging of the thyroid gland for many years and is the first-line recommended imaging modality for thyroid nodules. Its use in thyroid disorders is widely accepted and the benefits and indications for its use continue to expand. Thyroid ultrasonography has traditionally been under the purview of radiology departments, but in the past decade has been adopted by surgeons and endocrinologists in the office-based setting for evaluation and management of patients with thyroid and other head and neck disorders. Its versatility, speed, safety profile, ability to offer dynamic real-time images, and low cost compared with other radiologic modalities have all contributed to its popularity.
The initial uses of thyroid ultrasonography came at a time when palpable thyroid nodules were surgically excised to establish a pathologic diagnosis. In the late 1960s, ultrasound was used to differentiate between solid and cystic nodules and to measure and track nodule size. Using conventional ultrasonography without the benefit of gray-scale images, clinicians were able to differentiate cysts from cystic degeneration in an adenoma, solid tumors from multinodular goiter, and to detect the presence of thyroiditis with greater than 90% accuracy. However, the differentiation of benign versus malignant lesions remained problematic, and in the following decade, investigators began studying whether newer ultrasound technology could help improve surgical and medical decision making by identifying malignant features of thyroid lesions.
In the past 40 years, the role of thyroid ultrasonography has continued to expand and it is currently recommended in the evaluation of all palpable nodules by the American Thyroid Association (ATA), the American Association of Clinical Endocrinologists (AACE), and the Associazione Medici Endocrinologi (AME). The thyroid gland is well suited to ultrasound evaluation in part because of the superficial position and easy accessibility of the gland, its distinctive echotexture, and the ability to gain greater anatomic detail than with computed tomography, magnetic resonance imaging, or radionuclide studies. Table 1 lists some of the goals of and indications for thyroid ultrasonography.
| To better assess palpable thyroid nodules | To facilitate FNA biopsy of a nodule |
| To determine whether nodularity is present in the patient with an equivocal or difficult physical examination | To assess the remainder of the thyroid gland in the patient with a palpable thyroid nodule |
| To determine whether characteristics associated with malignancy are present | To screen for thyroid lesions in patients who have been exposed to radiation |
| To screen for thyroid lesions in patients with other diseases in the neck, such as hyperparathyroidism, who are undergoing treatment planning | To objectively monitor nodules, goiters, or lymph nodes in patients undergoing treatment or observation of thyroid disease |
| To assess the thyroid and the extrathyroid neck in the patient with thyroid cancer before treatment | To monitor treated patients with thyroid cancer for early evidence of recurrence in the thyroid bed and cervical lymph nodes |
| To identify thyroid features associated with diseases including thyroiditis and Graves disease | To facilitate therapeutic procedures such as sclerotherapy or laser ablation of thyroid nodules |
| To help teach regional anatomy and the art of thyroid palpation | To detect undescended thyroid or thyroid agenesis |
| To monitor fetal thyroid development in utero | To assess the size and location of the neonatal thyroid |
| To detect goiter as a sign of iodine deficiency | To refine management of patients on therapy such as antithyroid medications |
| To screen family members of patients with familial forms of thyroid cancer |
This article reviews the relevant uses of and indications for ultrasound in various thyroid diseases, with particular attention to thyroid nodules and cancer. The characteristic ultrasound features of these diseases are described. The associated literature and societal guidelines are discussed.
Role of ultrasound in the initial evaluation of the thyroid nodule
The ATA and AACE/AME recommend thyroid ultrasound for all patients with suspected thyroid nodules, including patients with palpable abnormalities, nodular goiter, and thyroid lesions found incidentally on other imaging modalities. Routine screening thyroid ultrasound is not recommended for the general population because of the high incidence of thyroid nodules. An autopsy study of 821 consecutive patients with clinically normal thyroid glands showed that 50% of patients had at least one thyroid nodule and 36% had nodules greater than 2 cm in size. Palpable thyroid nodules occur in up to 7% of the general adult population, and the incidence of nonpalpable thyroid nodules visible by ultrasound is up to 10 times greater (ie, 70%).
Ultrasound evaluation of thyroid nodules in at-risk patients can help confirm the presence of a nodule; objectively characterize the size, location, and appearance of the nodule; evaluate for benign or suspicious features; and evaluate for the presence of other thyroid nodules or cervical lymphadenopathy. Although certain ultrasound characteristics of thyroid nodules are associated with malignancy, fine-needle aspiration (FNA) remains the gold standard for diagnosis. FNA has until recently been recommended for cytologic evaluation of all thyroid nodules greater than 1 cm in diameter or nodules less than 1 cm that exhibit suspicious features. The 2009 Revised ATA Guidelines for Management of Thyroid Nodules and Differentiated Thyroid Cancer also include recommendations for FNA of certain thyroid lesions based on ultrasound criteria. These criteria include mixed solid and cystic nodules 1.5 to 2 cm or greater in diameter with any suspicious ultrasound characteristics and predominantly spongiform nodules 2 cm or greater in diameter. Although considered the gold standard, the diagnostic role of FNA is limited by an overall 3% to 5% false-negative rate and a 10% nondiagnostic rate.
The use of ultrasound guidance improves the sensitivity, specificity, and accuracy of FNA compared with palpation-guided FNA in certain populations. Ultrasound-guided FNA seems to be most valuable in patients with nonpalpable nodules, small palpable nodules, multiple nodules, partially cystic nodules, or concomitant glandular disease. It is also beneficial for sampling specific areas of a nodule, such as from the solid part of a mixed solid-cystic nodule. Compared with palpation-guided FNA, the use of ultrasound guidance decreased the rate of inadequate samples in palpable nodules 2 cm or smaller from 39% to 23%. Cesur and colleagues found the rates of inadequate FNA samples to be significantly improved in palpable nodules 1.0 to 1.5 cm using ultrasound- versus palpation-guided FNA (37.6% vs 24.4%, P = .009), but not for palpable nodules 1.6 cm or larger. Currently, the ATA recommends ultrasound-guided FNA for nodules that are nonpalpable, predominately cystic, or located posteriorly in the thyroid lobe, and when repeating FNA for a nodule with an initial nondiagnostic cytology result.
Role of ultrasound in the initial evaluation of the thyroid nodule
The ATA and AACE/AME recommend thyroid ultrasound for all patients with suspected thyroid nodules, including patients with palpable abnormalities, nodular goiter, and thyroid lesions found incidentally on other imaging modalities. Routine screening thyroid ultrasound is not recommended for the general population because of the high incidence of thyroid nodules. An autopsy study of 821 consecutive patients with clinically normal thyroid glands showed that 50% of patients had at least one thyroid nodule and 36% had nodules greater than 2 cm in size. Palpable thyroid nodules occur in up to 7% of the general adult population, and the incidence of nonpalpable thyroid nodules visible by ultrasound is up to 10 times greater (ie, 70%).
Ultrasound evaluation of thyroid nodules in at-risk patients can help confirm the presence of a nodule; objectively characterize the size, location, and appearance of the nodule; evaluate for benign or suspicious features; and evaluate for the presence of other thyroid nodules or cervical lymphadenopathy. Although certain ultrasound characteristics of thyroid nodules are associated with malignancy, fine-needle aspiration (FNA) remains the gold standard for diagnosis. FNA has until recently been recommended for cytologic evaluation of all thyroid nodules greater than 1 cm in diameter or nodules less than 1 cm that exhibit suspicious features. The 2009 Revised ATA Guidelines for Management of Thyroid Nodules and Differentiated Thyroid Cancer also include recommendations for FNA of certain thyroid lesions based on ultrasound criteria. These criteria include mixed solid and cystic nodules 1.5 to 2 cm or greater in diameter with any suspicious ultrasound characteristics and predominantly spongiform nodules 2 cm or greater in diameter. Although considered the gold standard, the diagnostic role of FNA is limited by an overall 3% to 5% false-negative rate and a 10% nondiagnostic rate.
The use of ultrasound guidance improves the sensitivity, specificity, and accuracy of FNA compared with palpation-guided FNA in certain populations. Ultrasound-guided FNA seems to be most valuable in patients with nonpalpable nodules, small palpable nodules, multiple nodules, partially cystic nodules, or concomitant glandular disease. It is also beneficial for sampling specific areas of a nodule, such as from the solid part of a mixed solid-cystic nodule. Compared with palpation-guided FNA, the use of ultrasound guidance decreased the rate of inadequate samples in palpable nodules 2 cm or smaller from 39% to 23%. Cesur and colleagues found the rates of inadequate FNA samples to be significantly improved in palpable nodules 1.0 to 1.5 cm using ultrasound- versus palpation-guided FNA (37.6% vs 24.4%, P = .009), but not for palpable nodules 1.6 cm or larger. Currently, the ATA recommends ultrasound-guided FNA for nodules that are nonpalpable, predominately cystic, or located posteriorly in the thyroid lobe, and when repeating FNA for a nodule with an initial nondiagnostic cytology result.
Ultrasound characteristics of thyroid nodules
Many investigators have identified ultrasound characteristics of malignant thyroid nodules ( Table 2 ). Although these ultrasound characteristics offer high sensitivity, no single criterion offers sufficient specificity to differentiate benign from malignant lesions. However, when taken together, specificity improves. One prospective, observational study compared ultrasound and FNA results with surgical disease conditions in 349 patients and found that performing FNA on nodules with one of 3 ultrasound criteria (microcalcifications, blurred margins, or hypoechoic pattern) missed only 2% of cancers. Kim and colleagues prospectively analyzed 155 incidentally discovered, nonpalpable, solid thyroid nodules and found a mean number of 2.6 suspicious findings per malignant nodule and an overall sensitivity and specificity of 94% and 66%, respectively.
| Margins | Blurred, ill-defined |
| Halo/rim | Absent |
| Shape | Irregular, spherical, tall |
| Echo structure | Solid |
| Echogenicity | Hypoechoic |
| Calcifications | Microcalcifications, internal |
| Vascular pattern | Intranodular, hypervascular |
| Elastography | Decreased elasticity |
The next sections discuss ultrasound features of thyroid nodules and their ability to suggest benign versus malignant lesions.
Nodule size
Nodule size has not been found to be significantly predictive of malignancy. The risk of malignancy for palpable thyroid nodules is approximately 10% and several studies suggest a similar incidence of malignancy in nodules smaller than 1 cm. Thyroid cancers less than 1 cm in size have been shown to behave clinically similar to larger cancers, and therefore these lesions should be followed with periodic ultrasound surveillance with the option for further evaluation with FNA if growth or suspicious features are observed. The ATA recommends FNA biopsy of subcentimeter nodules if there is a high risk of malignancy (family history of thyroid cancer, history of external beam or ionizing radiation, history of thyroid cancer, or fluorodeoxyglucose-avid thyroid nodules on positron emission tomography) or if there is suspicious concomitant lymphadenopathy, in which case FNA of the lymph node should be performed.
Lesion margins and halo/rim
Benign lesions are often associated with a hypoechoic circumferential halo ( Fig. 1 ), believed to represent a capsule and compressed thyroid tissue. Neoplasms may display a partial or absent halo, and its presence or absence has been found to be suggestive but not diagnostic. Blurred or ill-defined margins have been associated with increased risk of malignancy. The mobility of the nodule with respect to surrounding structures should be assessed, as fixation suggests malignant invasion of the surrounding tissue.
Nodule shape
Nodule shape has been implicated as having prognostic significance. One retrospective analysis found nodules with a more spherical shape had a higher incidence of malignancy. In contrast, another study found that nodules that are more tall than wide are more likely to harbor cancer. Irregular shape has also been implicated in malignancy.
Echo structure
Many thyroid nodules are cystic or have cystic components, such as cystic degeneration of a follicular adenoma ( Fig. 2 ) or in the setting of multinodular goiter. Malignancy has been more closely associated with solid nodules compared with cystic or mixed nodules, with one study finding 121 of 148 (81.8%) histopathologically malignant nodules to be solid. Purely cystic nodules are unlikely to be malignant, as are those with a spongiform appearance ( Fig. 3 ), defined as an aggregation of multiple microcystic components in more than 50% of the nodal volume.
Echogenicity
The echogenicity of a thyroid nodule should be compared with that of surrounding thyroid tissue. Most benign adenomas or adenomatous nodules are slightly hypoechoic when compared with normal thyroid tissue ( Fig. 4 ), whereas malignant nodules are frequently markedly hypoechoic ( Fig. 5 ). In a prospective, observational study of 349 surgically excised thyroid nodules, Cappelli and colleagues found a 3.8 odds ratio of malignancy in solid hypoechoic nodules.
Calcifications
The presence of calcifications has variable significance. Peripheral calcification, also referred to as eggshell calcification, is typically considered a benign feature, representing previous hemorrhage and degenerative change ( Fig. 6 ). However, coarse calcifications can be seen in malignant nodules, as can microcalcifications, which are strongly associated with an increased risk of malignancy. A total of 45% to 60% of malignant nodules show microcalcifications, as opposed to 7% to 14% of benign nodules. Approximately 60% of patients with microcalcifications were found to have malignant disease. The presence of microcalcifications in malignant nodules is often attributed to psammoma bodies in papillary thyroid carcinoma (PTC) ( Fig. 7 ) and is frequently seen in medullary thyroid carcinoma (MTC). Although suggestive of malignancy, the overall specificity of microcalcifications for thyroid carcinoma has been reported to range from 71% to 94%, with a sensitivity of 35% to 72%, and therefore should not be solely relied on to differentiate benign from malignant lesions.
