The authors thank Stephanie Deming for editorial assistance with the manuscript.
Intrathyroid metastases from extrathyroid primary tumors are rare. Clinical findings may be subtle, but detection of intrathyroid metastases has improved with sonography. The objective of this study was to evaluate the sonographic appearance of intrathyroid metastases.
Patients with thyroid masses with cytopathologic features matching those of an extrathyroid primary tumor were retrospectively identified. The appearances of intrathyroid metastases on sonography were reviewed for the following features: size, margin regularity, echogenicity, echotexture, vascularity on power or color Doppler ultrasonography, and the presence or absence of any associated cervical adenopathy.
The study included 52 patients. The most frequent primary tumor sites were lung, head and neck, and breast. Intrathyroid metastases presented as a discrete nodule in 34 patients and as diffuse infiltration of the gland in 18 patients. The discrete nodules ranged in size from 1.1 to 5.6 cm (mean ± SD, 2.5 ± 1.2 cm). Thirty-three lesions (63%) had irregular margins, and 19 (37%) had well-defined margins. Most of the lesions were heterogeneously hypoechoic (n = 50, 96%). Vascularity was present in 32 of 50 measured lesions (64%) that were evaluated with Doppler sonography. Cervical adenopathy was present in 37 patients (71%).
Intrathyroid metastases have sonographic characteristics similar to those described for both benign and malignant thyroid diseases. In patients with a previous or current extrathyroid malignancy, thyroid nodules or diffuse infiltration of the thyroid gland on sonography should be viewed as a potential intrathyroid metastasis and evaluated via ultrasound-guided fine-needle aspiration regardless of the site of the primary tumor.
The literature suggests that intrathyroidal metastases (ITMs) from extrathyroid primary malignancies are rare.[1-11] The rarity of ITMs is thought to be due to a rich thyroid arterial supply with fast flow, which prevents the deposition of metastatic cells, and the high oxygen saturation and iodine content of the thyroid gland, which may inhibit the growth of malignant cells. Recent reports indicate that, in 1.4% to 3% of all patients undergoing surgery for cancer in the thyroid, the cancer represents ITMs from primary tumors at other sites.[14-18] Historically, most ITMs have been detected at autopsy, which led to the belief that ITMs are associated with a poor prognosis.[1-11] Although clinical findings associated with ITMs may be subtle, detection of ITMs has improved with the use of sonography.[19-21] It has been our experience and the experience of others that, with improved imaging and sonography-guided biopsy, ITM can be diagnosed during initial or follow-up evaluation of extrathyroid primary tumors.[19-24] Because there is limited information in the literature on sonography in the sonographic appearance of ITMs, we sought to evaluate the ultrasound characteristics of ITMs.
Materials and Methods
The Institutional Review Board approved this retrospective review and waived the requirement for informed consent. Data review was performed in compliance with all applicable Health Insurance Portability and Accountability Act regulations. A retrospective review of an institutional database was performed of the demographic and clinical characteristics and ultrasound findings in patients with cytopathologically documented ITMs seen at our institution from 2002 through 2013. Patient records and sonograms were reviewed by 2 neuroradiologists, one with 8 years and the other with 10 years of head and neck ultrasound experience. For a patient to be included in the study, sonography-guided fine-needle aspiration (FNA) had to show that the thyroid mass had cytopathologic features identical to those of the known extrathyroid primary tumor. Patients with lymphomatous or leukemic involvement were excluded from this study.
The ultrasound examinations were performed with a high-frequency broadband linear-array transducer (7–17 MHz) connected to an Alpha-10 scanner (Hitachi-Aloka, Tokyo, Japan) with color and/or power Doppler sonography capability. The following ultrasound characteristics of the ITMs were evaluated: discrete thyroid nodule versus diffuse thyroid infiltration, appearance (solid versus complex), size, margin regularity, presence or absence of a hypoechoic halo, echogenicity (in comparison with the adjacent normal thyroid parenchyma), echotexture (homogeneous versus heterogeneous), presence or absence of intralesional punctate hyperechoic foci, vascularity on color or power Doppler sonography, and presence or absence of adenopathy in the central and lateral neck compartments. The medical records were reviewed to determine whether there were other sites of metastatic disease.
Sonography-guided FNA was performed by the sonographer using a 20-gauge needle attached to a 10-cc syringe. The aspirated material was smeared on glass slides and fixed in Carnoy's fixative for Papanicolaou staining and also air dried for Diff-Quik staining. Immediate assessment of specimen adequacy was performed by cytopathologists. The needle was finally rinsed in RPMI fluid, which was centrifuged to create cell pellets. Cytospin smears were made from the rinsed material when the cell pellet was small. Cell blocks were prepared when the cell pellets were sufficient in amount and could be collected in tissue paper and routinely processed to be embedded in paraffin wax to generate formalin-fixed and paraffin-embedded tissue blocks. These blocks were cut at 5 μm thickness and stained by hematoxylin and eosin for conventional histopathological examination.
Fifty-two patients were diagnosed with ITMs at our institution during the study period. The interval between diagnosis of the extrathyroid primary tumor and diagnosis of ITM by sonography-guided FNA ranged from 1 month to 11 years 10 months (mean [SD], 2 years 8 months [2 years 8 months]). The indications for ultrasound evaluation of the thyroid and the primary malignancies are indicated in Table 1.
Table 1. Patient Demographics, Ultrasound Indications, and Primary Tumor Sites
Discrete Nodule (n = 34)
Diffuse Infiltration (n = 18)
Mean age, years (SD)
57.6 ± 11.4
57 ± 10.4
Indication for ultrasonography
Palpable thyroid nodule
Palpable cervical node
Primary tumor site
SCC head and neck
Seventeen patients underwent color and 33 patients underwent power Doppler sonography; in 2 patients, a Doppler examination was not performed. ITMs occurred as a discrete nodule in 34 patients (65%) and as a diffuse infiltration of the thyroid in 18 patients (35%). Fourteen of the thirty-four discrete nodules were solitary (41%), and 20 were part of a multinodular thyroid (59%).
Appearance of Discrete Nodules on Sonography
The 34 patients with ITMs in the form of a discrete nodule were 18 men and 16 women ranging in age from 34 to 79 years (mean [SD], 57.6 [11.4] years). The discrete nodules ranged in size from 1.1 to 5.6 cm (mean [SD], 2.5 [1.2] cm). Thirty patients (88%) had solid nodules (Figure 1), and 4 (12%) had complex nodules. The margins of the nodules were well defined in 19 patients (56%) and irregular in 15 (44%; Figure 2). Six ITMs (18%) had a complete hypoechoic halo, and 4 (12%) had a partial hypoechoic halo. Thirty-three of the discrete nodules (97%) were hypoechoic, and 1 (3%) was isoechoic. The echotexture was heterogeneous in 32 nodules (94%) and homogeneous in 2 (6%). Ten (29%) of the ITMs contained punctate hyperechoic foci without posterior shadowing or reverberation, and 2 (6%) ITMs contained a macrocalcification. Intranodular vascularity was present in 19 of 33 patients (58%); flow was not recorded in 1 patient.
In four patients (12%), the thyroid was the only site of metastatic disease. Cervical node metastases were present in 23 patients (68%), and in 9 patients (26%) cervical nodes were the only site of metastatic disease apart from the thyroid. Nineteen patients (56%) had distant metastases outside of the neck. The ultrasound appearance of the discrete nodules is summarized in Table 2.
The 18 patients with ITMs in the form of a diffuse infiltration were 11 men and 7 women ranging in age from 33 to 71 years (mean [SD], 57 [10.4] years). Each thyroid gland with diffuse metastatic infiltration was hypoechoic with heterogeneous echotexture (Figure 3). In all 18 patients (100%), the margins of the thyroid were irregular. Punctate hyperechoic foci were present in 5 patients (28%). Among the 17 patients who had flow assessed on Doppler sonography, intrathyroid vascularity was noted in 13 patients (76%) and was absent in 4 patients (24%).
In 2 patients (11%), the thyroid was the only site of metastatic disease. Cervical node metastases were present in 14 patients (78%), and cervical nodes were the only site of metastatic disease apart from the thyroid in 5 patients (28%). Eleven patients (61%) had distant metastases outside of the neck. The ultrasound appearance of the infiltrative disease is summarized in Table 2.
Treatment and Survival Data
Among the 52 patients with ITMs, 30 were treated with chemotherapy, 8 with surgery, 3 with radiation, and 2 with chemoradiation. Four patients were referred to hospice care, and 5 were lost to follow-up. At the time of this report, 44 patients (85%) have died. These patients lived from 8 days to 8 years 5 months (mean [SD], 1 year 7 months [1 year 8 months]) after the diagnosis of ITM. Among the 8 patients (15%) alive at the time of this report, 6 have been followed for times ranging from 2 years 8 months to 10 years 9 months (mean [SD], 6 years [3 years 7 months]) after the diagnosis of ITM; 2 patients were lost to follow-up after the diagnosis of ITM.
This study provides important information about the ultrasound appearance of ITMs. Two distinct ultrasound appearances of ITMs were clearly defined: a discrete nodule, either solitary or part of a multinodular thyroid, present in 65% of patients; and diffuse infiltrative disease, present in 35% of patients.
The ultrasound characteristics of discrete nodular ITMs in our study were in part consistent with the characteristics previously described for malignant discrete thyroid nodules. These characteristics included a solid nodule (92%), hypoechogenicity compared with the normal thyroid (97%), ill-defined margins (44%), absence of a hypoechoic halo (70%), punctate hyperechoic foci (29%), and vascularity of varying degrees (56%).
However, our study also shows that several of the discrete ITMs had ultrasound characteristics that are not commonly associated with malignant thyroid nodules or that overlap with previously reported characteristics of benign thyroid nodules. These included the following: well-defined margins (present in 56% of patients), complete (18%) or partial (12%) halo, complex nodules (12%), and isoechogenicity compared with the normal thyroid (3%).
The ultrasound characteristics of the cases of diffuse infiltrative ITM lacked specificity and did not allow differentiation of ITM from thyroiditis or other diffuse disease within the thyroid. Thus, sonography-guided FNA is important in cases of diffuse thyroid disease in patients in whom ITM is suspected. The American Thyroid Association guidelines indicate and multiple authors[21, 26-29] have suggested that various ultrasound characteristics of a thyroid nodule are associated with a higher likelihood of malignancy. These include irregular margins (55% to 77% of malignant lesions)[25-31]; nodule hypoechogenicity compared with the normal thyroid parenchyma (16% to 90%)[26-28, 32]; a solid lesion (69% to 87%)[27, 29]; microcalcifications (29% to 59%)[25, 29, 33] in primary thyroid carcinomas[27, 28, 34-36]; and intranodular vascularity in thyroid nodules (69% to 74%).[29, 37] The presence of a hypoechoic halo has 95% specificity in the diagnosis of benignity in nodules of primary thyroid origin (ie, the absence of a halo does not necessarily mean that the nodule is malignant).[33, 39] Future studies could evaluate the number and the ultrasound characteristics of thyroid nodules in all patients with known metastatic disease and ITMs.
It has been established that the size of a thyroid nodule is not a good predictor of malignancy.[20, 34, 40] In our series, the size of discrete ITMs ranged from 1.1 cm to 5.6 cm.
Diffuse small, punctate echogenic foci were present in 5 (28%) of the patients with diffuse infiltration. These foci did not have characteristics of calcifications or comet tails of crystalline colloid. They were of indeterminate etiology and are the focus of another study.
The interval from diagnosis of the primary tumor to diagnosis of ITM in our study varied from 1 month to almost 12 years. Our institution is a major tertiary cancer referral center, so we cannot offer a valid assessment of the frequency with which specific types of primary tumors metastasized to the thyroid.
In our study, an ITM was the only manifestation of metastatic disease in 6 patients (12%), and the only other site of metastatic disease was cervical adenopathy in 14 patients (27%). Six of the patients in our study continue to thrive, and 1 patient has been disease free for over 10 years following chemotherapy for ITM. This highlights the importance of detection of ITMs to permit successful medical and/or surgical treatment and to avoid tumor recurrence in the neck.[24, 41, 42]
The abnormal thyroid in each of the patients in this series was detected during diagnostic imaging evaluation for primary, persistent, or recurrent extrathyroid cancers; sonography was prompted by computed tomography findings in 9 patients (17%) and by positron emission tomography in 15 patients (29%). The thyroid and soft tissues of the neck were ultimately evaluated in each patient by ultrasonography and sonography-guided FNA.
In comparison with other imaging modalities, sonography offers the advantage of a dynamic, hands-on exploration of unusual or unexpected patient anatomy or technically accessible suspicious findings. In the past 15 years, high-frequency sonography and power Doppler sonography have become the most important imaging modalities for study of the soft tissues of the neck, in particular, for study of the thyroid, parathyroid, and cervical lymph nodes. ultrasonography also permits sonography-guided FNA, which can be performed at the bedside during the ultrasound examination and has the unique advantage of allowing immediate real-time pathologic correlation via conventional expert cytopathologic examination. In experienced hands, sonography-guided FNA is a safe and accurate procedure.[43-45]
In conclusion, this study shows that ITMs can present as a discrete nodule, either solitary or part of a multinodular thyroid, or as diffuse infiltrative disease with features overlapping those of benign and malignant thyroid disease regardless of the site of the primary tumor. In patients with a previous or current extrathyroid malignancy, thyroid nodules or diffuse infiltration of the thyroid on sonography should be viewed as a potential intrathyroid metastasis and evaluated by using sonography-guided FNA.