Sonography is very sensitive for identifying minimal residual thyroid tissue and nodal metastases as small as 2–3 mm in diameter. These tiny foci may be very difficult to identify intraoperatively, however. And the clinical impact of this minimal residual disease is questionable [1]. Nearly half of the “suspicious” US findings that are aspirated are ultimately deemed to be benign [59–61]. Additionally, several studies have demonstrated that the majority of suspicious lesions in the postoperative neck do not demonstrate significant growth [62–64]. Though the percentage of lesions that exhibited growth was variable in the studies, it remained low, between 9 and 20 %, even with long-term follow-up (3.5–14 years) [62–64]. The studies consistently found that the variable which best predicted the growth of a lesion was the largest lesional diameter; bigger masses were more likely to grow [62, 64]. Even among the minority of malignant lesions that demonstrated enlargement, the rate of growth was typically slow, with a 1.3 mm/year increase [64]. Fortunately, when the structurally progressing lesion was ultimately resected, there was no evidence of local invasion or distant metastases [63, 64]. Taken together, these data point to the conclusion that with careful sonographic follow-up including serial measurement of Tg, observation rather than surgical removal may be warranted in properly selected patients with minimal residual disease [1, 43]. Alternatively, those sonographically suspicious lesions measuring greater than 8–10 mm in smallest diameter may be considered for fine-needle aspiration with cytology and thyroglobulin measurement in the needle washout fluid and, ultimately, removal [1]. Suspicious lesions that are smaller than 8–10 mm also may be considered for intervention if there is growth or if the node threatens vital structures [1].

Clinicians should be cognizant of the limitations of neck sonography. First, sound waves cannot penetrate air, bone, or cartilage. Consequently, nodes posterior to the trachea may be difficult to identify. Second, the depth of a structure may exceed the limits of high-resolution sonography. As a result, the superior mediastinal and deep level IV lymph nodes may be difficult to see, particularly in patients with a large body habitus. Third, artifacts from normal anatomic structures may also obscure visualization of nodal metastases. For example, isoechoic nodes that lie posterior to a large vessel may be difficult to delineate. The posterior acoustic enhancement associated with the sound waves traveling through the vessel may disguise an isoechoic metastatic node.

Computerized tomography may therefore improve the sensitivity of identifying disease in the retropharyngeal space and the superior mediastinum [65] and may be a useful adjunct to ultrasonography in preoperative planning [1]. Indeed, the combination of CT and US was found to have increased sensitivity to either modality alone for the identification of nodal disease in the central compartment [66].

Extrathyroidal extension may be seen in up to 15 % of patients with papillary thyroid cancer , with the severity ranging from microscopic invasion to gross extension of the tumor into the surrounding vital structures and soft tissues [67]. Though gross extrathyroidal extension may be identified on ultrasonography [68], invasion into the trachea or esophagus cannot be assessed sonographically. CT offers improved visualization of the integrity of these structures [66, 69]. Knowledge of the extent of tumor invasion could significantly impact the surgical approach [70].

Another limitation of sonography includes the identification of lesions and structures not related to the cancer but which may be confused with a recurrence. These include suture granulomata, neuromas, parathyroid adenomas, thymic tissue, and metastases from other primary tumors, to name a few [71, 72]. The distinction between benign and malignant nodes may also be confusing, particularly with plump reactive nodes in which the characteristic hilar stripe cannot be readily identified. These often may be seen in level II, as repeated microtrauma in the mouth chronically stimulates the node and causes resultant hypertrophy [71, 72]. The distinction between suspicious and malignant lymph nodes is discussed in greater detail in Chap. 20.

Ultrasound plays an important role in the diagnosis and surveillance of patients with thyroid cancer. Preoperative US of the neck is a critical component of the surgical planning; identification of metastatic nodes changes the operative approach. Sonography also helps identify recurrent disease in the neck in the postoperative patient; all patients should have a surveillance US 6–12 months after surgery. The importance of the ultrasound examination in the management of thyroid cancer patients is further established in that it is a central component of the definition of freedom from disease. The necessity of, and interval for, subsequent imaging is based on the response to therapy. Though sonography is pivotal to the evaluation and surveillance of thyroid cancer, there are important limitations to its use. The clinician should not only have expertise in the distinction of those metastatic foci of disease but also have a facile understanding of the mimics of thyroid cancer recurrence. Furthermore, when suspicion is high for the presence of persistence or recurrence in the absence of a sonographic suspect lesion, other imaging modalities may be necessary to examine the regions of the neck that are not adequately visualized sonographically.