Ultrasonography

In this chapter we discuss not only the sonography of hyperplastic nodules but also that of thyroid diseases presenting in the form of a nodule.

At first, we have to determine what is the basic role of US in the case of nodular goiter and what is not. Thyroid ultrasound is the most reliable tool to determine the presence of a nodule, to measure the size of the thyroid, the size of the nodule, to follow the change in size with time, to localize a lesion and to guide aspiration. US is able to recognize the substernal spread of the thyroid. However, the US is not convenient to determine whether a nodule is malignant or benign, and is not suitable to determine the degree of substernal spread. Naturally, we can found US signs which favour or disfavour the possibility of a cancer, and may be of help in the everyday practice, but the diagnostic value of any single parameter or combination of parameters is far from that of aspiration cytology. I understand the tendency to minimize the need for aspiration cytology, a method which diagnostic value extremely depends on the person who performs it, but for inexperienced readers the analysis of international thyroid literature may lead to an absolutely false impression. The essence of this impression is that we will be able to found a wonder property in US which leads us to leave out aspiration cytology from the diagnostic protocol.

Another important distinction

Iodine deficiency has a very deep impact on the epidemiology of thyroid diseases, the relative proportion benign to malignant lesions , on the distribution of distinct subtypes of thyroid cancer and last but not least even on the diagnostic accuracy of aspiration cytology. It means that we cannot draw direct consequences of a given population from the thyroid literature because the iodine intake varies worldwide from moderate deficiency to iodine excess.

To define a nodule is not a simple task. If we state that a nodule is a lesion whose echostructure differs from the extralesional part of the thyroid, then practically every human being would have or would evolve nodular goiter. Moreover, this definition has the consequence that a great proportion of patients with an autoimmune disease would be categorized as nodular goiter patients, too.

What is more, the term nodule is originally a pathological expression. The overuse of this term has a bad effect on the patients. Nowadays most patients check their problems and medical reports in the internet. If a patient does not have a nodule in pathological sense, the risk of malignancy is zero. In such situation, we cause absolutely unnecessary anxiety in the patient.

I think the term nodule must be maintained for those discrete lesions which are nodules in pathological sense with a high probability. If we accept this approach, the definition of a nodule would be more problematic but the unnecessary anxiety caused to the patient will be diminished. The correct diagnosis of a thyroid nodule needs a systematic sonographic-histopathologic correlation. To record the ultrasound examinations seems to be indispensible. Not only the images but the video records have to be archived. This is the only way to compare the ultrasound with the macroscopic pathological findings months or years later when the patient undergoes on operation. This is the only way to learn how to differentiate a lesion caused by thyroiditis from a nodule which has potential oncological significance. This is a huge effort but the only way to increase our diagnostic ability and to become better experts at thyroid ultrasound.

Finally, there are ethical aspects of this professional issue. Frequently, we cannot decide whether a discrete lesion would be a normal finding without any clinical significance, a focus of thyroiditis, a nodule in pathological sense or a normal secondary or tertiary lobule of a lobe. Firstly, we have to learn that this distinction cannot be made in each patient. Secondly, we have to describe our uncertainty instead of using an incorrect term

We have to discuss the following parameters in the case of a nodule: echostructure, the borders of the nodule and special sonographic signs. We must not forget to correctly describe the extranodular thyroid, too. We also have to discuss the basic echostructure, the three diameters of the thyroid, the nodule and the presence or absence of retrotracheal or substernal spread. It is also evident that in the case of a thyroid nodule we have to perform a neck ultrasound. I think that the examination of the neck is advisable in non-nodular patients, too. This is the only chance to recognize an ectopic thyroid.

The echostructure of the nodule may be solid, mixed solid-cystic or cystic. The existence of a pure cyst may be questioned. The so-called pseudocyst is the only pathological entity which does not contain a solid part. In most cases of (pure) cysts the solid part is very small and we cannot detect it on ultrasound. The distinction between solid any mixed nodules has only limited oncological significance. Nevertheless, the cystic degeneration of a papillary cancer is more frequent than that of a follicular or medullary cancer.

A solid nodule or the solid part of a mixed nodule may be hyperechogenic, echonormal, and hypoechogenic of various degrees. Almost all echonormal nodules are benign. Around 7% of malignancies occur in hyperechogenic nodules. Great proportions of carcinomas are found in hypoechogenic nodules. Some investigators, including us, discriminate minimally, moderately hypoechogenic and hypoechogenic nodules. Others do not follow this approach. If the degree of hypoechogenicity increases, the risk of malignancy follows that pattern too. The discrimination between moderately hypoechogenic and (severely) hypoechogenic nodule has practical relevance. The chance to detect halo sign is significantly greater in a moderately hypoechogenic nodule compared to hypoechogenic ones. This is explained by the similar echostructure of the capsule with a hypoechogenic nodule. The finding of a halo sign is of great help to diagnose follicular tumors. You can read more about the importance of this distinction later.

The borders of a benign nodule are regular and sharp. If the borders are irregular, we have to describe it on the US report. An irregular border may be sharp, puzzle-like. This pattern raises the possibility that the lesion is not a nodule in pathological sense, but only a more active focus of lymphocytic thyroiditis. The blurred border of a nodule is an important sign which increases the risk of malignancy, i.e. papillary cancer. In the case of an aggressively growing thyroid malignancy, we can detect the discontinuation of the capsule or the direct spread of the carcinoma into the normal parenchyma. However, this is a rare finding.

The halo sign represents a fibrotic tissue around nodules and secondary lobules. There is a close relationship between the halo sign and fibrotic capsule surrounding a follicular tumor. Although the presence of a halo sign suggests a significantly decreased risk of malignancy overall, the consequences of a detailed analysis indicate that the situation is not so simple. At first, if we do not investigate encapsulated nodules, then a high proportion of cancers, i.e. most follicular carcinomas, will be missed. Moreover, our results demonstrate that the absence of a halo sign is not an independent risk factor. The absence of a halo decreases the risk of malignancy only because this sign cannot be demonstrated in the most frequently occurring malignant nodules, in hypoechogenic ones. The demonstration of a halo sign is very difficult or even impossible in most hypoechogenic nodules because the US appearance of the capsule in such lesions is identical with the nodule. This is clearly proved by the fact that for follicular adenomas, which are encapsulated in 100% of the cases, only 8.8% of the hypoechogenic nodules exhibit a halo sign, whereas this feature can be demonstrated by US in 64.1% of all other types of nodules. Therefore it is not surprising that the presence of a halo sign significantly increases the risk of malignancy in cases of moderately hypoechogenic nodules, while in hyperechogenic nodules the presence or absence of a halo sign does not influence the likelihood of malignancy. The detection of perinodular blood flow even in the absence of halo sign is of help detecting a capsule. Although there is a strong correlation between the halo sign detected on US and the presence of capsule, we cannot conclude that a nodule with halo sign is a follicular tumor. Hyperplastic nodules are also frequently surrounded with capsule. What is more, the fibrous tissue separating the secondary and tertiary nodule from each other frequently becomes thicker in the case of an autoimmune thyroid disease, first of all in Hashimoto's thyroiditis.
To summarize: the presence of a halo is not a specific sign, most thyroid lesions can present fibrous capsule, however the lack of a halo substantially decreases the chance of a follicular tumor. If we investigate the patient with a high resolution probe, the chance of a follicular tumor is less than 2% if the nodule lacks both halo sign and perinodular blood flow. This might have a huge differential diagnostic potential in the event of a cytological pattern suggesting a follicular tumor. (See chapter "A new approach".)

Other special sonographic signs

These properties have relevance mostly in hypoechogenic nodules. Microcalcification is a relatively specific sign of papillary cancer but can be found in all thyroid entities, too. The presence of microcalcifications increases the likelihood of malignancy about three times. The significance of the so-called eggshell calcification, i.e. the presence of a calcified capsule is conflicting. Some of the authors have discovered that this property increases the risk of malignancy; others including ourselves could not affirm this observation. The intranodular coarse calcification seems to increase the risk of carcinoma, but the practical significance is much weaker than in the case of microcalcifications.

There is a special hyperechogenic granulation which has oncological relevance. The most specific sign of MTC with a sensitivity of 47% and an odds ratio of 31.5 (95% confidence interval 15-66) was the presence of multiple amorphous hyperechogenic foci within the nodule (see Figure 1). These foci are larger than 1 mm in diameter in contrast to microcalcifications. They showed an irregular patchy appearance and did not indicate a dorsal acoustic shadow as opposed to coarse calcifications in all cases. This feature was observed in 11 out of the 23 histopathologically diagnosed MTC and in 67 out of the 3,988 other patients who underwent surgery. These foci are not simply coarse calcifications but correspond to deposits of calcium surrounded by amyloid.

The taller than wide sign increases the likelihood of papillary cancer. It means that the ventro-dorsal diameter exceeds the latero-medial diameter.

The vascularization of the nodules

There are three types of blood flow determined by color Doppler mode. Type 1: neither perinodular, nor intranodular blood flow is detected. Type 2: perinodular blood flow is present, but intranodular blood flow is not detected. Type 3: intranodular blood flow is increased.

Papillary cancer demonstrates significantly more frequently type 3 pattern compared with benign lesions. In the everyday practice, an irregularly increased chaotic intranodular blood flow has relevance. The combined type 2 and type 3 patterns are frequently observed in autonomously functioning adenomas. Nevertheless, the vascular pattern has a very limited role in the everyday oncological differential diagnostics.

A hard nodule is more often malignant compared with a soft one. Elastography, i.e. the investigation of the rigidity of thyroid nodules may be able to determine this property. Nowadays, this is a very popular field in the medical literature. Although there are conflicting results, it seems self-evident that a rigid nodule is more probable malignant. Unfortunately, the different ultrasound manufacturers use different algorithms, so the comparison of the results gained by different equipments is still an unresolved task.

The special signs listed above are suitable to draw attention to a malignant disease. Until now, these are not suitable to reduce the need of cytology. The combination of various sonographic properties may increase the sensitivity but the aim is to reduce the specificity and conversely. However, to reduce the need for cytology it is required to have alterations in the judiciary environment of our medical practice.

Finally, we should talk about a neglected area of thyroid ultrasound. I think that one of the most convenient tools to exclude the possibility of thyroid malignancy would be the ultrasound. A malignant thyroid nodule has to increase in size. We are able to measure the volume of a nodule exactly. It may have great importance in a potentially new approach of treating a patient with the suspicion of a well-differentiated follicular tumor. A few years' delay in the appropriate treatment has no harm to the patient. If we use regular follow-up investigations instead of immediate surgery and indicate operation only in those cases where the nodule later increases in size, we can reduce the proportion of unnecessary surgical interventions.