Differential diagnostics of discrete lesions

Author: T. Solymosi

doi: 10.24390/thyrosite.2018.discrete_lesion_vs_pathological_nodule

Summary

It is a paradoxical situation that although we know very much about the various features of a thyroid nodule and we have several scoring systems focusing on the risk of malignancy of thyroid nodules, we still lack a clear cut definition of a thyroid nodule. For an inexperienced ultrasound technician this might seem a theoretical issue without any significant impact on the everyday practice. In fact, the differentiation whether a lesion is a nodule in pathological sense or not, is one of the greatest and not rarelyunsolvable problem of thyroid ultrasounds. On the other hand, it is more than strange that the thyroid literature almost completely neglects this field.

We use the term ` nodule ` in palpation, in scintigraphy, in ultrasound and in histopathology. The problem is that the term nodule has different meanings in all of these fields. In our opinion, the term nodule should be used in pathological sense even in ultrasound practice. The radiological approach which defines a nodule as discrete echo abnormality causes serious harm; both to the patient causing unnecessary fear from carcinoma and both to the evaluation system causing enormous pointlessadditional examinations.

In our opinion, the cause for this situation is the neglect of quality control of thyroid ultrasound. Regarding this field, the basis of quality control could be only a systematic sono-histological comparison which should be made in all patients who underwent surgery. Does your radiologist perform this comparison? Are you aware of publications focusing on this issue?

In this chapter we try to give an overview of this problem. We present a  textbook focusing on the differentiation of echo abnormalities . We present  three consecutive series of case studies : the first presents patients consecutively admitted to our thyroid outpatient clinic who were diagnosed with autoimmune thyroiditis. The second series is consisted of consecutively operated patients whose histopathological diagnosis implicated diffuse lymphocytic thyroiditis either alone or as a second diagnosis to various histological forms of nodules. We also present  several lectures , the basis of which is the comparison of ultrasound patterns.

 

Case study of the topic

A 27-year-old was referred for evaluation of growing nodules. She was treated for hypothyroidism for 19 years. In the first ultrasound multiple echonormal and hyperechogenic nodules were described and the largest one was aspirated. Th e cytological diagnosis was the suspicion of follicular carcinoma. A subtotal left lobectomy was performed when she was 10. Histopathology disclosed Hashimoto's thyroiditis without any nodules. In regular follow-up,the hyperechogenic nodules were found increased in size, from 8 to 11 and from 12 to 16 mm over fifteen years.

Aspiration cytology was performed from the "nodule in the right lobe" in another institution which was suspicious of follicular carcinoma. We advised her against stepping in the same river twice and recommended regular follow-up instead of repeated surgery. I tried to convince the patient to avoid surgery but I failed. Histopathology disclosed Hashimoto's thyroiditis. No nodule was found.

This case study raises several important concerns including the deceptive cytological patterns of Hashimoto's thyroiditis. But let's focus on the actual issue, the differential diagnostics of discrete lesions in Hashimoto's thyroiditis, which involves both the thorough analysis of ultrasound presentation and the patients` history.

Textbook

Introduction

There are two issues in thyroid ultrasound. The first, which is discussed in thousands of scientific articles, is the oncological differential diagnostic of thyroid nodules diagnosed in ultrasound, i.e. which nodule requires further evaluation and which one does not. However, there is a more fundamental question: which discrete lesion is a nodule in a pathological sense and which one is not. This problem is almost completely neglected in the thyroid literature which suggests for those who do not perform ultrasound themselves that this problem does not exist. But this is a delusion. The ultrasound pattern is extremely close to gross pathology. The resolution of sonography is less than 1 mm, a range which is almost absolutely sufficient for every purpose. The problem is that almost every thyroid, including a healthy one, presents discrete echo abnormalities and all common thyroid diseases including various forms of thyroiditis, autoimmune thyroid diseases, as well, might be appear as discrete lesions on ultrasound. Therefore the definition, as a nodule would be a discrete echo abnormality, is unacceptable.

The art of thyroid ultrasound is the differential diagnostic of lesions detected on ultrasound which is a nodule in a pathological sense and which is not.

The first consideration is that the term nodule should be used exclusively in pathological sense. Otherwise, we give false diagnosis and cause serious harm to the patient. Secondly, in significant proportion of cases a clear distinction is not possible. Regarding the practical importance of this problem I mention here only one data: more than 90% of Hashimoto's thyroiditis appears in ultrasound in the form of discrete echo abnormalities.

 

 

Basic problems and false resolutions

My cornerstone is that I use the term ` nodule ` exclusively in pathological sense. Be aware that there are only three types of nodules in pathology: hyperplastic nodules, benign tumors and malignant tumors. Occasionally histopathologists use the term "nodular Hashimoto", as well.

The issue is how to differentiate discrete lesions seen in ultrasound: to whatextent is it possible to make a clear distinction between nodules and discrete lesions which are in fact not nodules.

It is evident that we cannot make a clear distinction in each and every case. In the everyday practice, many examiners draw a false conclusion from the former statement by using the term ` nodule ` instead of discrete echo abnormality even for those lesions which are ambiguous. This is professionally and scientifically unfounded and leads to great harm both to the patient and to the entire evaluation system.

There is another harmful conclusion: if we cannot indisputably differentiate every discrete lesion then we should not make efforts decreasing the proportion of ambiguous cases. In my opinion, the most important circumstance explaining this strange behavior is the negligence of quality control of thyroid ultrasound. The basis of this quality control could be only a systematic sono-histological comparison, not only in select cases but in every single patient who had undergone surgery. We wait from a cytopathologist to compare the cytological diagnosis to histopathological report but we do not wait from a ultrasound technician to compare his or her diagnosis to the (macroscopic) pathological finding. It means that the self-evident way of a learning course is blocked.

 

Patterns mimicking nodule in autoimmune thyroid diseases

In the everyday practice the issue is the discrete lesion found in a patient with an autoimmune thyroid disease. At first, we have to be aware of the presentation of these disorders. In more than 90% of cases of lymphocytic thyroiditis (Hashimoto's thyroiditis) we can found discrete echo abnormalities. The problematic patterns are listed in Table 1.

Table 1. Difficult to interpret patterns in Hashimoto's thyroiditis.

 

Occurrence

Frequency of hypothyroidism

% of misinterpretation as nodule

Hypoechogenic lesions > 5 mm in echonormal background

85%

15%

70%

Echogenic islets in hypoechogenic background

35%

35%

85%

Central hypoechogenic area surrounded with echonormal rim

5%

80%

70%

Pseudonodular form of Hashimoto's thyroiditis

8%

30%
80%

At first, we have to be aware, while performing thyroid ultrasound examination, that Hashimoto's thyroiditis is the greatest illusionist among thyroid disease. Not only theoretically, but even from a practical aspect, almost each type of ultrasound pattern occurs in autoimmune thyroiditis. We have  to bear the four patternsin mind which were described above. The most important one is the presence of hypoechogenic discrete lesion within an echonormal background.

We have several other tools. Seemingly the  aTPO determination  would be of the greatest help; however an elevated aTPO does not exclude the synchronous presence of a nodule. Conversely, even in the incipient form of Hashimoto's thyroiditis where only a few or even one hypoechogenic islet had evolved, the aTPO is frequently below the cut-off level. There is a linear correlation between the degree of hypoechogenicity or the echogenicity index and the likelihood of aTPO positivity. In our experience, the ultrasound has a greater sensitivity in diagnosing incipient forms of autoimmune thyroiditis.

In individual cases the  patients' history  or  follow-up results might be of great help. Some examples:in recurrent Graves' disease the comparison of the ultrasound image with a previous one recorded in euthyroid state might be of help to avoid a false diagnosis of a nodule. Similarly, in ambiguous cases, a repeated ultrasound in euthyroid state might be of great help in the correct interpretation of a hypoechogenic lesion which is present in hyperthyroid but disappeared in euthyroid state.

The knowledge of the  previous histopathological diagnosis  in a patient with recurrent goiter is among the most important data to be involved in differential diagnostics. We give examples to all of these situations in case studies.

Although the above mentioned properties might be of help in certain cases, the issue of differential diagnostic is basically an issue of the correct interpretation of the ultrasound pattern.

There are two basic types of the non-diffuse form of autoimmune thyroiditis. In the majority of cases, we can see  hypoechogenic areas of various amount and size  within an echonormal or less hypoechogenic background. We see the other type of lesions in advanced Hashimoto's cases where a great proportion of the thyroid has become hypothyroid and we found  one or more echonormal islets  mimicking echonormal or hyperechogenic nodules. We have to be aware of the proportion of hypoechogenic areas that vary form 0 to 100% in Hashimoto's thyroiditis. It means that the distinction between various forms of Hashimoto's thyroiditis is arbitrary. Nevertheless, from a practical point of view, the distinction shares important consequences.

Hypoechogenic lesions > 5 mm in echonormal background

Hypoechogenic areas are almost always present in Hashimoto's thyroiditis. If their maximal diameter is < 1-2 mm, they rarely cause concern. However, in around 80% of Hashimoto' cases these discrete hypoechogenic areas are larger than 5 mm in maximal diameter and according to our experience in more than 75% of cases they are described as a nodule by radiologists. Such misinterpretation is itself generates FNAC. Be aware that these areas are deeply hypoechogenic, a property which automatically leads to categorization of this lesion among the most suspicious group in most TIRADS systems. Moreover, as a rule, a lesion in Hashimoto's thyroiditis has almost always an irregular shape and borders which raises further the suspicion.

 

Viewpoints to differentiation of discrete lesions

Firstly, we have to  analyze the entire thyroid . Except for a very few cases the thyroid presents not one or two but more discrete hypoechogenic lesions of different size. So, the presence of smaller lesions with identical echo pattern to a larger lesion being in question significantly increases the likelihood that these lesions have a common origin. A basic, somewhat hypoechogenic echo pattern of the thyroid might be of great help, first of all by raising the possibility that a discrete lesion might be also a presentation of the autoimmune process. But again, the demonstration of Hashimoto's thyroiditis does not exclude the possibility of a coexistent nodular goiter.

The  borders and the shape of a lesion  are probably the most important features in the differential diagnostics. In contrast with most nodules which have a regular geometrical shape and sharp borders, the lesions of Hashimoto's thyroiditis are irregular in shape and their borders are very often either blurred or sharp but puzzle-like. The latter reflects the infiltration of normal parenchyma by lymphoid elements.

Table 1. Comparison of nodules in pathological sense with discrete hypoechogenic lesions found in Hashimoto's thyroiditis.

 

Importance

Lymphocytic thyroiditis

Nodule in pathological sense

Size of the lesion

+

Tend to be smaller

Not specific

Number of the lesions

++

Numerous

Usually 1 to 3

Basic echopattern

++

Frequently hypoechogenic

Usually echonormal

Presence of fibrosis

+++

Frequent finding
Rare finding

Shape of the lesion

++++

Not regular geometric

Usually regular geometric

Borders of the lesion

++++

Puzzle-like or blurred

Sharp

The  inside  of a hypoechogenic lesion frequently contains ragged echonormal tissue. This pattern is caused by the lymphoid infiltrate of normal tissue. If we compare this pattern with that of a compound nodule which contains echonormal and hypoechogenic areas, as well, we can see a striking difference: in the event of a nodule the echonormal areas are confluent while in the event of Hashimoto's thyroiditis they are seen as floating islets within a hypoechogenic background.

The  presence of intranodular echogenic figures  also has relevance. If there are multiple hypoechogenic lesions but one of them presents echogenic figure than this area might differ from the others of origin. The presence of microcalcification in only one out of multiple hypoechogenic areas is the most important feature of diagnosing a papillary carcinoma present in a Hashimoto's thyroiditis. Proliferation of connective tissue is a frequent finding in autoimmune thyroid diseases, its ultrasound presentation is very characteristic: synchronous lines and punctate granules are found depending on the angle between the transducer and the connective tissue. It is very important not to misinterpret these granules as microcalcifications.

Central hypoechogenic area surrounded with an echonormal rim

This pattern can be observed in more advanced cases of Hashimoto's thyroiditis and mimics a large hypoechogenic nodule. Patients presenting this form are usually hypothyroid. The above mentioned viewpoints have to be considered in differentiation this pattern from a large hypoechogenic nodule. The borders of the large hypoechogenic area are much easier to judge compared with smaller lesions, and as a rule they are irregular, puzzle like reflecting the infiltration of normal parenchyma by lymphoid population. The hypoechogenic field almost always presents proliferation of connective tissue. 
The explanation for this pattern might be the difference in vascular supply or the thickening of the connective tissue at the edge within or next outside the thyroid.

Echonormal islets in hypoechogenic background

This pattern is more often seen in advanced Hashimoto's thyroiditis. In contrast with the former types, the misinterpretation of echonormal islets as echonormal nodules seems to have less negative consequence from an oncological point-of-view. On the other hand, while an echonormal nodule in an echonormal thyroid has practically no risk of carcinoma, an echonormal nodule in Hashimoto's thyroiditis carries some risk of cancer. So the overinterpretation of echonormal areas is not simple a professional failure but might also lead to unnecessary FNAC.

The viewpoints of differential diagnostics are similar to those mentioned in discrete hypoechogenic areas. We can find not one but usually several echonormal areas of different size. The shape of these echonormal fields is almost always irregular, puzzle like corresponding to the infiltration of echonormal parenchyma by the hypoechogenic lymphoid population.

 

Pseudonodular form of Hashimoto's thyroiditis

This  pattern has different denominations: pseudolobular or micronodular or pseudonodular means the same. This is a specific pattern of autoimmune thyroiditis which is more rarely found in Graves' disease. This pattern is caused by the proliferation of connective tissue which makes the appearance of the thyroid as it was composed of numerous small nodules. In contrast with other forms of Hashimoto's thyroiditis, the thyroid is usually enlarged and the patients are less frequently hypothyroid. The pivotal of avoiding misinterpretation of this form as a multinodular goiter is the  awareness of the pseudonodular form.  In most if not all of these cases, not the entire thyroid presents the micronodular form, the other fields are usually hypoechogenic which is of great importance in differential diagnostic .

Categorization of discrete areas

As a rule, we should not have to be forced to give a clear-cut answer if the presentation is uncertain. One of the most important failures of the current scoring (TIRADS) systems is that the examiner is forced to give a yes or no answer regarding the presence or lack of e.g microcalcifications, extrathyroidal spread. In significant proportion of cases we cannot fulfill this unrealistic expectation. If we try to meet the requirements, we give wrong answers in a huge number of cases. Our duty is not to meet expectations of arbitrary systems but to meet expectations of our patients. Irrespectively of the use or negligence of TIRADS system in our ultrasound report, we have to describe reality. And if reality is that we cannot give a clear-cut answer as to whether a lesion is a pathological nodule or not, or as to whether an echogenic intranodular granule is a microcalcification or not, we have to describe the uncertainty in the report.

There are many possible ways. In the case studies I used a grouping which is consisted of 5 subgroups for categorization of discrete lesions from unlikely being a pathological nodule at one end to very likely being a pathological nodule at the opposite end.

Indication of aspiration cytology and categorization of discrete areas

There at least two layers of this issue. Firstly, if a patient is faced with a false ultrasound diagnosis of a hypoechogenic nodule grouped among the most suspicious TIRADS category then it is reasonable to perform cytology even if we are convinced that the lesion is not a pathological nodule but a usual focus of Hashimoto's thyroiditis. Otherwise we cannot reassure the patient.Secondly: what to do in an uncertain lesion. In the event of "doubtful" and "probably pathological nodule" categories, the indication of cytology is the same as the lesion would be categorized as a pathological nodule. In " probably not a pathological nodule " subgroup which would be candidate of cytology if it would be categorized as a true nodule, our decision has to be individualized depending on the size of the lesion, on the psyche of the patient. From a practical point-of-view in those uncertain cases which FNAC is not performed from, I tell the patient briefly the essence of this chapter...

 

mask