Introduction
Medullary thyroid carcinoma (MTC) is a rare cancer accounting for 2–4% of all thyroid malignancies. It may occur as sporadic (about 80% of cases) and familial medullary thyroid carcinoma in multiple endocrine neoplasia (MEN) type 2A or 2B syndrome, frequently representing their first clinical manifestation. It originated from the calcitonin-producing parafollicular cells (C cells). These cells are derived from other germ layer than thyroid follicular cells, and join them in later stages of embryogenesis. They are mainly localized in the medial and uppermedial parts of the thyroid lobes, determining the location of MTC. Medullary thyroid carcinomas usually occur as single tumors. They may also be multifocal and bilateral, especially in the case of familial carcinomas(1,2).
Although ultrasound-guided FNAB of the focal lesion is a method of choice in the diagnostics of thyroid focal lesions, a definite diagnosis of MTC is not always possible(3). Medullary carcinomas are usually shown in the ultrasound image as solid, highly hypoechogenic focal lesions with calcifications due to high amounts of amyloid. Macroscopically, medullary thyroid carcinomas most often present in the surgical material as solid tumors, usually with well-defined borders, though non-encapsulated. Cross-sectionally, they are hard or with increased consistency compared to the surrounding thyroid parenchyma. They can be granular, white, grey or beige, usually with no extravasations or necrosis; they rarely contain cystic lesions. The histopathological structure of medullary carcinomas is comprised of solid areas and clusters of cells of various size and shape, sometimes forming trabecular or lobulus microscopic pattern; they are surrounded by hyalinizing or fibro-vascular stroma. Stromal accumulation of amyloid occurs in 80–90% of cases.
Furthermore, MTCs are characterized by rich vascularization with a chaotic course of vessels and the absence of halo sign. Compared to papillary thyroid carcinomas (PTC), which represent the most common thyroid malignancies, MTCs are usually larger and show a more oval shape at diagnosis(4).
Current knowledge does not allow to determine ultrasonographic characteristics that clearly identify MTCs. The available literature reports show that even one in three lesions may not show the characteristics suspicious of malignancy in an ultrasound image(3). Ultrasound sonoelastography is a promising technique for the assessment of the hardness of the thyroid focal lesions. The method is based on a general assumption that most malignancies are represented by hard-tissue lesions, whereas benign lesions are soft-tissue lesions. Currently, two basic sonoelastography techniques are used: static/strain elastography (SE) displaying relative tissue displacement, and dynamic/shear wave elastography (SWE). However, the results obtained using these two methods for thyroid assessment are ambiguous(5–7). Dynamic sonoelastography, using the phenomenon of shear wave propagation velocity through tissues, depending on their hardness, allows for both qualitative and quantitative assessment of focal lesions and the surrounding tissues. In the qualitative assessment, the blue color indicates soft tissues and the red color indicates hard tissues. The quantitative tissue stiffness analysis in the regions of interest (ROI) of various sizes involves the calculation of the maximum, mean and minimum values of the Young’s (E) modulus (kPa).
Aim
The aim of the study was to analyze the ultrasound characteristics of medullary thyroid carcinoma as well as to assess the clinical usefulness of SWE in MTC preoperative evaluation.
Materials and methods
All patients included in the study gave their written consent to participate in the study, and the study was approved by The Bioethics Committee of the Maria Skłodowska-Curie Memorial Cancer Centre and Institute of Oncology. The retrospective analysis included four patients (mean age: 45 years) with six thyroid focal lesions. The patients received B-mode ultrasound examination of the thyroid and the surrounding neck tissues. This was followed by a SWE quantitative assessment of the stiffness of lesions and the surrounding thyroid tissues. The assessment was performed in the Laboratory of Ultrasound in Endocrinology and Nuclear Medicine Maria Skłodowska-Curie Memorial Cancer Centre and Institute of Oncology using Aixplorer (SuperSonic, Aix-en-Provence), with a 4–15 MHz linear probe, and in accordance with the standards of the Polish Society of Ultrasonography(8).
The B-mode ultrasound of thyroid focal lesions assessed:
echogenicity (normal/hypoechoic/isoechoic);
echostructure (homogeneous/heterogeneous);
the presence of halo sign;
the presence of micro- and macrocalcifications;
lesion margins (sharp/ill-defined);
height/width ratio;
vascularization pattern of the lesions, using color Doppler (type I – no visible vessels, type II – single vessels within the lesion, type III – vessels in the parenchyma of the lesion.
Next, focal lesions and the surrounding tissues were evaluated using an elastogram, by placing focal lesions in its central part, i.e. FOV (field of view).
In the quantitative assessment for ROIs of various sizes (the first one included the entire lesion, the second one – an area with a diameter of 2 mm, selected automatically in the stiffest region of the lesion, excluding calcifications), the lesions were analyzed in two sections (transverse and longitudinal), and the obtained E-values were averaged. The maximum and mean E-values for MTC (EmaxLR, EmeanLR), maximum and mean E-values for the surrounding thyroid tissues (EmaxSR, EmeanSR) and the mean Young’s modulus values (EmeanLRz) from the stiffest part of the lesion were analyzed and assessed.
Results
Histopathological evaluation of six focal lesions confirmed the presence of 6 MTCs. Three patients (5 lesions) had familiar MTCs, and one patient with a single lesion had sporadic MTC. The maximum average size of these lesions was 4 to 29 mm (mean of 15.2 mm). In the B-mode assessment, all MTCs were hypoechoic, with no halo sign and they contained micro- and macrocalcifications (100%). Ill-defined lesion margin were found in 4 cancers (66.7%). Heterogeneous echostructure and type III vascularization were found in 5 out of lesions (83.3%). Vascular flow was not visualized using Color Doppler in one lesion (Fig. 1 C). The height/width ratio of more than 1 was found in 4 out of 6 lesions (Tab. 1).
Fig. 1.
A longitudinal section of the left thyroid lobe. Hypoechoic, oval lesion with well-defined margin and fine microcalcifications can be seen dorsally in the upper pole (A). Color Doppler showed no lesion vascularity – type 1 (B). Lesion SWE: EmaxLR = 19.5 kPa, EmeanLR = 12.5 kPa were lower for the lesion compared with the surrounding tissues (EmaxSR = 24.1 kPa, EmeanSR = 20.5 kPa) (C)
Tab. 1.
B-mode ultrasonographic characteristics of thyroid focal lesions
| Lesion no. | Lesion echogenicity | Lesion echostructure | Lesion borders | Halo sign | Calcifications | Height/width ratio | Lesion volume | Type of vascularization |
|---|
| 1 | Hypoechogenic | Heterogeneous | Sharp | No | Microcalcifications | 0.58 (11/19 mm) | 2.4 mL | III |
| 2 | Hypoechogenic | Heterogeneous | Sharp | No | Macrocalcifications | 0.82 (18/22 mm) | 7.2 mL | III |
| 3 | Hypoechogenic | Heterogeneous | Ill-defined | No | Microcalcifications / Macrocalcifications | 1,125 (9/8 mm) | 0.32 mL | III |
| 4 | Hypoechogenic | | Ill-defined | No | Microcalcifications / Macrocalcifications | 1.67 (15/9 mm) | 0.61 mL | III |
| 5 | Hypoechogenic | Heterogeneous | Ill-defined | No | Microcalcifications / Macrocalcifications | 1.71 (12/7 mm) | 0.42 mL | III |
| 6 | Hypoechogenic | Homogeneous | Ill-defined | No | Microcalcifications | 2 (8/4 mm) | 0.16 mL | I |
In SWE assessment, the mean value of EmaxLR for all of the MTCs was 89.5 kPa and the mean value of EmaxSR for all surrounding tissues was 39.7 kPa. Mean values of EmeanLR and EmeanSR were 34.6 kPa and 24.4 kPa, respectively. The mean value for the stiffest part of the lesion (2mm ROI) was 49.2 kPa (EmeanLRz) (Tab. 2).
Tab. 2.
Elastographic characteristics of thyroid focal lesions
| Lesion No. | EmaxLR [kPa] | EmeanLR [kPa] | EmeanLRz [kPa] | EmaxSR [kPa] | EmeanSR [kPa] |
|---|
| 1 | 59.6 | 31.9 | 44.3 | 39.2 | 26.9 |
| 2 | 98.4 | 17.9 | 60.9 | 58.1 | 36.8 |
| 3 | 35.8 | 13.9 | 23.5 | 35.7 | 14.6 |
| 4 | 190.0 | 88.0 | 99.6 | 43.0 | 25.2 |
| 5 | 138.0 | 53.1 | 62.9 | 41.0 | 32.8 |
| 6 | 16.4 | 3.1 | 3.9 | 21.3 | 10.1 |
| Mean | 89.5 | 34.7 | 49.2 | 39.7 | 24.4 |
Discussion
Thyroid sonoelastography is a non-invasive method recommended by the EFSUMB as an additional diagnostic tool for characterizing focal thyroid lesions. According to the opinion of EFSUMB experts, the technique is particularly useful for controlling patients with thyroid focal lesions verified as benign, based on FNAB findings(9). This is important for MTCs due to their greater malignant potential compared to papillary thyroid carcinomas (PTC). Additionally, high difficulty in cytological assessment of these tumors with about 63% sensitivity according to Bugalho et al., may delay diagnosis and treatment initiation(10). In our study material, the suspicion of medullary thyroid carcinoma was based in all cases on cytological assessment performed by pathologists.
B-mode assessment of MTC also poses difficulties, especially in the case of small lesions (Fig. 1) with no characteristic ultrasound features, such as irregular macrocalcifications or hypoechogenicity. Woliński et al.(11) performed a meta-analysis of ultrasonographic characteristics of medullary thyroid carcinoma by assessing 169 cases of this tumor, which indicated hypoechogenicity as the most common feature (83.4%). The sensitivity related to the absence of halo sign was 89.9%, and microcalcifications were observed in only 35.5% of MTCs. Macrocalcifications, which represent calcified amyloid deposits showing reactive fibrosis, occurred in only 27% of lesions; they are described in the literature as a characteristic feature of MTCs. Similarly, the ‘taller than wide’ feature was rarely observed in these tumors (only 14.4% of cases).
Hypoechogenic lesions with no halo sign dominated in our study material (100%), as also reported by other authors. No relationship was observed between the incidence of other ultrasound features. Microcalcifications were more common (5/6 lesions) than macrocalcifications (4/6). Lesion with both micro- and macrocalcifications occurred in all cases. Fig. 2 shows an example of familial MTC with micro- and macrocalcifications. The ‘taller than wide’ feature was observed in 2/3 of cases of the analyzed MTCs, i.e. in 4/6 lesions (66.6%).
Fig. 2.
Hypoechogenic lesion with micro- and macrocalcifications and acoustic shadowing behind the lesion is seen in the longitudinal sections (A). Color Doppler showed multiple, chaotic vessels in the peripheral parts of the lesion – type III (B). Elastogram showed significantly higher lesion values EmaxLR, EmeanLR compared to surrounding tissues (EmaxSR, EmeanSR) (C)
Due to difficulties in determining typical features in ultrasound MTC examination a question should be asked, whether the risk factors of focal lesion malignancy, which are used for ultrasound assessment and have been included in the standards, can relate to MTC cases. Trimboli et al.(12) attempted to answer the question whether USG characteristics associated with PTCs should be used to diagnose MTCs. Their study showed a low incidence of features typical of PTCs in the cases of this tumor, such as ill-defined margin, microcalcifications and type III vascularization in Color Doppler. Among the ultrasound-assessed characteristics, only hypoechogenicity could suggest MTC, however, this feature was shown in only 50% of cases.
Other characteristics present in MTCs were shown by the following proportion of cases:
Significantly different ultrasound characteristics were shown for papillary carcinomas, with the following incidence:
ill-defined margin – 64.1%;
microcalcifications – 69.2%
type III vascularization – 15.4%
Surprisingly, well-defined margin were uncommon, they were found in 13.3% of PTC cases.
Differences in the ultrasound characteristics of medullary carcinomas compared to PTCs result from their different histopathological structure. The hardness of these tumors probably results from their solid histopathological structure, particularly stromal hyalinization and fibrosis in the absence of necrosis or blood extravasations.
Single publications assessing the usefulness of sonoelastography in MTC are available in the literature. Lin et al. showed in their meta-analysis on all types of thyroid cancers that sonoelastography is a highly accurate method with sensitivity and specificity of more than 80%, however, the authors failed to perform a detailed analysis of MTC percentage in the studied populations(13).
Andrioli et al. in their study, the only publication on MTC hardness assessment, evaluated this type of cancers using relative strain sonoelastography (SE). They used a 4-item deformability scale. Degrees ES3 and ES4 corresponded to non-deforming lesions, and thus suspected of malignancy as opposed to ES1 and ES2 lesions, which were partly or completely deformable(14). Among 18 evaluated MTCs, 10 lesions were classified as soft (one ES1 lesion and 9 ES2 lesions), and only 4 lesions were classified as ES3 and ES4. The results suggest that most MTCs are highly deformable. However, the authors did not provide a detailed description of the evaluated focal lesions, including their size, and the presence of calcifications, both of which may additionally affect their deformability.
Published studies on the use of SWE in differentiating the character of thyroid focal lesions showed that the E-values are higher for malignant cancers compared to those observed in benign lesions. In the available literature, the range of cut-off values differentiating malignant from benign lesions is very high, i.e. from 34.5 up to 94 kPa(15–18). However, PTCs dominate in the cited reports, and the single MTC cases are not discussed in detail.
In our study both the maximum and the average values of the Young’s modulus were significantly higher compared to the surrounding tissues and did not differ significantly from those presented in the literature on PTC (EmaxLR mean = 89.5 kPa, EmenLRz = 49.17 kPa). Apart from one MTC case (Fig. 1), all lesions had E-values higher than 35 kPa (values corresponding to one of the cut-off values suggested in the literature). Additionally, in all cases the lesions showed ultrasonographic characteristics suspicious of malignancy and were qualified for FNAB. It should be noted that there was a large divergence between cytological and histopathological findings. Different findings were observed in an ex vivo assessment performed for 4 MTCs. One of the patients (with two MTCs) had significantly harder lesions, whereas other patient had softer lesions.
Summary
The stiffness of focal lesions assessed using SWE is an additional characteristic in ultrasound imaging, which should be analyzed together with the following B-mode characteristics: ill-defined margin, calcifications, abnormal vascularization pattern (type III) and the dominance of anterior-posterior size over the lateral-lateral size, which increase the risk of their malignancy.
In the SWE, MTCs presented as lesions stiffer than the surrounding tissues, however, the small number of cases did not allow to draw clear conclusions about the usefulness of this method. The diagnostic algorithm for MTC is based on the measurement of calcitonin levels, B-mode ultrasound assessment and fine needle aspiration biopsy of suspicious lesions. T he assessment of the actual usefulness of this technique should be evaluated based on a larger clinical material, taking into account the division into thyroid cancer types.
Limitations
A small number of the assessed lesions due to the low incidence of MTCs in the population is a factor limiting the clear assessment of SWE usefulness in MTC patients.
