Key words
thyroid - tumor - ultrasound - histopahology - calcifications
Introduction
Ultrasound is a useful imaging modality for detecting and assessing thyroid nodules,
and it is a key examination. Ultrasound findings of thyroid nodules, such as their
shape, echogenicity, internal contents, margins, vascularity, and calcifications,
are important for evaluating them. Calcifications are common findings within thyroid
nodules, and various calcification types may be seen at various rates [1]. What is the real prevalence of calcifications detected by ultrasonography within
thyroid tumors of outpatients and what are the various histopathological types of
thyroid tumors? Are the calcification types on ultrasonography associated with histopathological
types of thyroid tumors? It is of interest to investigate whether calcifications on
ultrasonography can contribute to the ultrasonic diagnosis of thyroid tumors.
The purpose of the study is to clarify the prevalence of calcifications within thyroid
tumors on ultrasonography as well as the relationship between the calcification and
histopathological types.
Materials and Methods
Ultrasonography and calcifications
Ultrasound examination was performed using the same 6 units of the TOSHIBA Aplio SSA-770 A
ultrasound system with PLT-1204BX (7–14 MHz) and PLT-805AT (5–12 MHz) linear probes
for all periods of this study. Ultrasonic diagnosis of thyroid tumors was performed
using Kuma Hospital’s ultrasound classification system (Kuma’s USC, USC 1, 2=benign;
USC 3=borderline; and USC 4, 5=malignant) [2]
[3]. All thyroid tumors were classified as benign, borderline or malignant at ultrasonic
diagnosis. Among the many ultrasonographic findings, we evaluated the images of calcifications
(strong echoes) of thyroid tumors. Non-specific foci [4] (=so-called indeterminate hyperechoic spots) that are clinically observed in a thyroid
cyst (=comet-tail sign), in the thymus that are ectopically present in the thyroid
gland of juvenile patients (=pseudocalcifications), in the thyroid of Hashimoto’s
thyroiditis of old patients, in the thyroid after suppurative thyroiditis and trauma
of the neck, or in the lymph node without any suspicion of disease were excluded.
In this study, calcifications on ultrasonography were classified into 6 types according
to their shape, size, and region. Furthermore, punctate microcalcifications and speckled
type calcifications were divided into 2 subtypes according to their number: (1) a:
punctate microcalcifications (<1 mm), small number (1 – 5), b: punctate microcalcifications,
multiple (>6 ), (2) a: speckled type calcification (1≤< 3 mm), single, b: speckled
type calcifications, multiple (>2), (3) fragmentary type calcifications (≤ 3 mm),
(4) massive type calcifications (=course), (5) egg-shell type calcifications (=rim-like
peripheral calcifications), and (6) punctate microcalcifications in the parenchyma.
[Fig. 1] shows the schemas and images of calcification types on ultrasonography.
Fig. 1 Classification of calcification types within thyroid nodules and parenchyma on ultrasonography.
Calcifications on ultrasonography were classified into 6 types according to their
shape, size, and region. Furthermore, punctate microcalcifications and speckled type
calcifications were divided into two subtypes according to their number. (1) Punctate
Microcalc.: punctate microcalcifications (< 1mm) a. small number (1–5). b. multiple
(> 6 ); (2) Speckled Calc.: speckled type calcifications (1≤ <3 mm) a. single. b.
multiple (> 2); (3) Fragment. Calc.: fragmentary type calcifications (≥ 3 mm); (4)
Massive Calc.: massive type calcifications (= course); (5) Egg-shell Calc.: egg-shell
type calcifications (= rim-like peripheral calcifications); (6) Punctate MC in parenchyma:
punctate microcalcifications in the parenchyma.
Inter-observer variability on ultrasonography
Two board-certified fellows (KK, MF) of the Japan Society of Ultrasonics in Medicine
(JSUM) and 6–8 to eight registered medical sonographers of the JSUM performed ultrasound
examinations (it takes at least 5 years of clinical ultrasonography experience to
be qualified as a board-certified fellow, and at least 3 years of clinical ultrasonography
experience to be qualified as a registered medical sonographer). We made it a rule
to follow a standard scanning protocol to avoid missing clinically important lesions.
The ultrasonic diagnoses were performed using the same ultrasound guidelines [2]
[3] explained above. Clinicians, sonographers, and cytologists jointly took part in
preoperative and postoperative conferences each week as part of the routine work of
the hospital, and participated in an ultrasound conference each month to compare and
adjust ultrasonic diagnosis in order to decrease interobserver variability [5].
Investigation 1: Calcification within thyroid nodules in new outpatients
We closely evaluated ultrasound images of thyroid nodules (maximum diameter of 5 mm
or larger) of new outpatients who visited our hospital for 2 months between June and
July 2013, and the calcification types on the ultrasound images of these patients
were evaluated. Patients with thyroid nodules among these new outpatients subsequently
underwent fine-needle aspiration cytology. Cytopathological diagnostic categories
of thyroid nodules were based on the Bethesda system for reporting thyroid cytopathology
[6]. Some patients among these new outpatients subsequently underwent thyroid surgery
and the histopathological diagnosis was classified as benign or malignant. The relationship
between the calcification type and ultrasonic diagnosis of new outpatients, cytopathological
diagnostic category, and histopathological diagnosis was investigated. The prominent
type of calcification from one nodule was used as a datum (=nodule-based investigation)
in investigation 1, because we could prospectively observe the whole thyroid gland
closely by scanning moving and static ultrasound images.
Investigation 2: Calcification within thyroid tumors in patients who underwent thyroid
surgery
Four of the authors (KK, MF, TF, and HO) performed the retrospective re-evaluation
of ultrasound images that were extracted from the clinical database of the hospital.
The investigated histopathological types of malignancy were as follows: (a) papillary
carcinoma, (b) diffuse sclerosing variant of papillary carcinoma, (c) cyst-forming
type of papillary carcinoma, (d) metastatic lymph nodes in the lateral neck from papillary
carcinoma, (e) follicular carcinoma, (f) undifferentiated (anaplastic) carcinoma,
(g) medullary carcinoma, and (h) primary thyroid lymphoma. In those with papillary
carcinoma, we investigated the surgical patients for one year in 2013. We investigated
metastatic lymph nodes in the lateral neck from papillary carcinoma in patients who
had undergone primary surgery for papillary carcinoma in a period of 2 months between
October and November 2013. In those with the other histopathological types of malignancy,
we investigated the surgical patients for 9 years between 2005 and 2013. We used the
same ultrasound machines for all periods of this study and were able to achieve the
same quality of ultrasound images. Prominent calcification from one patient was used
as a datum (=patient-based investigation) in investigation 2, because static ultrasound
images of main tumors could be retrospectively extracted from the clinical database
of the hospital. The prevalence of calcifications and calcification types of thyroid
malignancy were investigated.
Histopathological diagnosis
Histopathological diagnosis of thyroid tumors for all periods of this study was based
on the histopathological criteria of the World Health Organization Classification
of Tumors published in 2004 [7].
Results
Investigation 1: Calcification within thyroid nodules in new outpatients
Among 2,902 nodules in 2,678 new outpatients during the 2 months of June and July
2013, 747 nodules (26%) had calcifications and 2,155 nodules (74%) did not. These
747 nodules with calcifications on ultrasonography were detected in 680 patients.
[Fig. 2] presents the relationship between the calcification type and ultrasonic diagnosis
(malignant, borderline, or benign) among these 747 nodules with calcifications. 231
nodules were diagnosed as malignant nodules, 163 nodules as borderline, and 353 nodules
as benign. The calcification types showed a wide range.
Fig. 2 Relationship between calcification type on ultrasonography and ultrasonic diagnosis
of thyroid nodules in new outpatients. Among 2,902 nodules in 2,678 new outpatients,
747 nodules (26%) had calcifications and 2,155 nodules (74%) did not. Ultrasonic diagnosis
of thyroid nodules was performed using Kuma Hospital’s ultrasound classification system
(malignant, borderline, or benign). The relationship between the calcification type
and ultrasonic diagnosis was investigated in these 747 nodules with calcifications.
Among these 747 nodules with calcifications, 560 nodules underwent fine-needle aspiration
cytology. [Fig. 3] presents the relationship between the calcification type and cytopathological diagnostic
category among these 560 nodules with calcifications. 176 nodules were diagnosed as
malignant, 21 nodules as suspicious of malignancy, 47 nodules as follicular neoplasm
or suspicious for a follicular neoplasm, 10 nodules as atypia of undetermined significance
or follicular lesion of undetermined significance, and 306 nodules as benign. The
calcification types showed a wide range.
Fig. 3 Relationship between calcification type on ultrasonography and cytopathological diagnostic
category of thyroid nodules. Among 747 nodules with calcifications in 2,678 new outpatients,
560 nodules underwent the fine-needle aspiration cytology. Cytopathological diagnostic
categories of thyroid nodules were based on the Bethesda system for reporting thyroid
cytopathology. The relationship between the calcification type on ultrasonography
and cytopathological diagnostic categories was investigated in these 560 nodules with
calcifications. susp. of malig.: suspicious of malignancy, folli. neoplasm: follicular
neoplasm or suspicious for a follicular neoplasm, atypia: atypia of undetermined significance
or follicular lesion of undetermined significance.
Among 2,678 new outpatients, 209 patients underwent thyroid surgery. 383 tumors were
present in these 209 patients. Among these 383 tumors, 249 tumors (65%) had calcifications
and 134 tumors (35%) did not. The relationship between the calcification type on ultrasonography
and histopathological diagnosis (malignant or benign) was investigated in these 249
tumors with calcifications. [Fig. 4] presents the relationship between the calcification type and histopathological diagnosis
among these 249 nodules with calcifications. 168 tumors were diagnosed as malignant
tumors: 164 papillary carcinomas, 2 medullary carcinomas, and 2 follicular carcinomas.
81 nodules were diagnosed as benign tumors: 73 multinodular goiters and 8 follicular
adenomas. The calcification types showed a wide range.
Fig. 4 Relationship between calcification type on ultrasonography and histopathological
diagnosis of thyroid tumors in surgical patients. Among 2,678 new outpatients, 209
patients underwent thyroid surgery. Three hundred and eighty-three tumors were present
in these 209 patients. Among 383 tumors, 249 tumors (65%) had calcifications and 134
tumors (35%) did not. The relationship between the calcification type and histopathological
diagnosis (malignant or benign) was investigated in these 249 tumors with calcifications.
<malignant> papillary carcinoma: 164 tumors, medullary carcinoma: 2 tumors, follicular
carcinoma: 2 tumors. <benign> multinodular goiters: 73 tumors, follicular adenoma:
8 tumors.
Investigation 2: Calcification within thyroid tumors in patients who underwent thyroid
surgery
Among 941 patients with papillary carcinoma, 725 patients (77%) had calcifications,
and 216 patients (23%) did not. The calcification types were investigated in these
725 patients with calcifications. [Table 1(a)] presents the calcification types of papillary carcinoma. The calcification types
showed a wide range. Punctate microcalcification (small number and multiple) made
up 42% of the 725 patients with calcifications and 32% of all 941 patients with papillary
carcinoma.
Table 1 Calcification type and prevalence of calcification on ultrasonography of malignant
thyroid tumors and metastatic lymph nodes.
|
P.MC. (small)
|
P.MC. (multi)
|
Speckled (single)
|
Speckled (multi)
|
Fragment
|
Massive
|
Egg-shell
|
P.MC.in pare
|
Prevalence of calcification
|
|
(a) Papillary carcinoma (941 patients)
|
167
|
134
|
88
|
87
|
138
|
64
|
45
|
2
|
77%
|
|
(b) Diffuse sclerosing variant of papillary carcinoma (18 patients)
|
0
|
0
|
0
|
0
|
0
|
0
|
0
|
18
|
100%
|
|
(c) Cyst-forming type of papillary carcinoma (32 patients)
|
0
|
1
|
1
|
0
|
12
|
18
|
0
|
0
|
100%
|
|
(d) Metastatic lymph nodes from papillary carcinoma (161 nodes)
|
0
|
60
|
1
|
0
|
15
|
0
|
2
|
0
|
48%
|
|
(e) Follicular carcinoma (337 patients)
(widely 67 patients/mini 270 patients)
|
0
|
0
|
2
(2/0)
|
0
|
26
(11/15)
|
11
(9/2)
|
40
(19/21)
|
0
|
23%
(61%/14%)
|
|
(f) Undifferentiated carcinoma (41 patients)
|
1
|
0
|
1
|
1
|
0
|
12
|
18
|
0
|
80%
|
|
(g) Medullary carcinoma (137 patients)
|
29
|
15
|
13
|
21
|
14
|
7
|
0
|
0
|
72%
|
|
(h) Primary thyroid lymphoma (173 patients)
|
0
|
0
|
0
|
0
|
0
|
0
|
0
|
0
|
0%
|
P.MC. (small): punctate microcalcifications (<1 mm), small number (1-5), P.MC. (multi):
punctate microcalcifications (<1 mm), multiple (>6 ), speckled (single): speckled
type calcification (1≤< 3 mm), single, speckled (multi): speckled type calcifications
(1≤< 3 mm), multiple (>2), fragment: fragmentary type calcifications (≥3 mm), massive:
massive type calcifications (course), egg-shell: egg-shell type calcifications (=rim-like
peripheral calcifications), P.MC. in pare: punctate microcalcifications in the parenchyma,
widely: widely invasive type of follicular carcinoma, mini: minimally invasive type
of follicular carcinoma
In all 18 patients with the diffuse sclerosing variant of papillary carcinoma, only
multiple punctate microcalcifications in the parenchyma of the thyroid gland were
observed ([Table 1(b)]). The prevalence of calcifications was 100%.
The calcification types of the cyst-forming type of papillary carcinoma were investigated.
([Table 1(c)]) presents the calcification types of these 32 patients. The prevalence of calcifications
was 100%. Fragmentary (12 patients) and massive types (18 patients) of calcifications
were mostly observed within the solid lesions, but the egg-shell type was not observed.
Multiple punctate microcalcifications (1 patient) and a single speckled type calcification
(1 patient) were also observed.
Metastatic lymph nodes in the lateral neck from papillary carcinoma were investigated.
Among 203 dissected lymph nodes in 145 patients, 161 nodes showed metastasis and 42
nodes did not. Among these 161 nodes with metastasis, 78 nodes (48%) had calcifications
and 83 nodes (52%) did not. The calcification types were investigated in these 78
nodes with calcifications. ([Table 1(d)]) presents the calcification types of metastatic lymph nodes with calcifications.
Multiple punctate microcalcifications (60 nodes) and fragmentary type calcifications
(15 nodes) were predominantly observed. A single speckled type calcification (1 node)
and egg-shell type calcifications (2 nodes) were also observed, but massive type calcifications
were not observed.
There were 337 patients with follicular carcinoma, 67 patients with the widely invasive
type and 270 patients with the minimally invasive type. Among these 337 patients,
79 patients (23%) had calcifications and 258 patients (77%) did not. The calcification
types were investigated in these 79 patients with calcifications. ([Table 1(e)]) presents the calcification types of follicular carcinoma with calcifications. The
calcification types were mostly the fragmentary, massive, and egg-shell types. The
widely invasive type showed a significantly higher (p<0.0001) prevalence of calcifications
(41/67 patients, 61%) than the minimally invasive type (38/270 patients, 14%) (Chi-squared
test, GraphPad InStat 3.1, GraphPad Software Inc.).
There were 41 patients with undifferentiated carcinoma. Among these 41 patients, 33
patients (80%) presented with calcifications and 8 patients (20%) did not. ([Table 1(f)]) presents the calcification types of these 33 patients with calcifications. The
calcification types were mostly the massive and egg-shell types.
There were 137 patients with medullary carcinoma. Among these 137 patients, 99 patients
(72%) had calcifications and 38 patients (28%) did not. ([Table 1(g)]) presents the calcification types of these 99 patients. The calcification types
showed a wide range, but neither an egg-shell type calcification nor punctate microcalcifications
in the parenchyma were observed.
There were 173 patients with primary thyroid lymphoma, 114 patients with mucosa-associated
lymphoid tissue lymphoma and 59 patients with diffuse large B-cell lymphoma. ([Table 1(h)]) presents the calcification types of these 173 patients. None of the patients had
calcifications (0%).
Discussion
Calcifications are frequently found during the examination of thyroid ultrasonography,
and various calcification types are seen within thyroid nodules and the parenchyma
[1]. Also in this study, the rather high prevalence of calcifications (26%) and the
wide range of calcification types ([Fig. 1]
[2]
[3]
[4]) were verified among thyroid nodules in new outpatients. Therefore, it is generally
considered to be impossible to make a differential diagnosis for thyroid nodules by
detecting calcifications alone. On the other hand, punctate microcalcifications on
ultrasonography have been reported as an important characteristic for differentiating
between malignant and benign nodules [8]
[9]
[10]
[11]
[12]
[13]. In our clinical experience, we have never seen calcifications in tumors of primary
thyroid lymphoma or metastatic carcinoma to the thyroid gland from renal cell carcinoma
[14]. Assessing calcifications as well as other characteristic findings can be helpful
in making an ultrasonic diagnosis of thyroid tumors.
In this study, none of the follicular carcinoma lesions had multiple punctate microcalcifications
([Table 1(e)]). This is consistent with the description in the comprehensive review by Grani G.
et al. [15]. Therefore, we should consider papillary carcinoma and follicular carcinoma separately
when making an ultrasonic diagnosis of thyroid tumors according to the calcification
type. Multiple punctate microcalcifications are often shown in benign ([Fig. 2]
[3]
[4]) and malignant tumors ([Fig. 2]
[3]
[4], [Table 1]). To make a correct diagnosis of tumors with multiple punctate microcalcifications,
it is necessary to assess the other ultrasonographic findings together with the cytopathological
appearance.
In the diffuse sclerosing variant of papillary carcinoma, calcifications solely presented
as multiple punctate microcalcifications in the parenchyma of one or both lobes of
the thyroid ([Table 1(b)]), and distinct nodules were not observed within the thyroid. When multiple punctate
microcalcifications within the thyroid parenchyma and bead-like swollen lymph nodes
in the neck are found in young women [16], we should consider this disease the top differential diagnosis.
The cyst-forming type of papillary carcinoma has a characteristic appearance [13], which is a large cyst, and solid lesions protrude into the cyst. Calcifications
within a solid lesion are mostly the fragmentary and massive types, but the egg-shell
was not observed ([Table 1(c)]). Such types of calcifications as well as other findings are characteristic in all
patients with this type of papillary carcinoma.
The well-known ultrasonographic features of metastatic lymph nodes from papillary
carcinoma are a rounded or irregular shape, heterogeneous and hyperechoic internal
echo, a large cystic space, loss of a hyperechoic central hilum, abundant blood signals
penetrating the node from the periphery, and the presence of calcifications [8]
[17]. In this study, calcifications mostly presented as 2 types, multiple punctate microcalcifications
and fragmentary type calcifications ([Table 1(d)]).
Calcifications of follicular carcinoma mostly presented as the following 3 types:
fragmentary, massive, and egg-shell type calcifications ([Table 1(e)]). The reported ultrasonographic features in follicular carcinoma [15]
[18]
[19]
[20]
[21]
[22] as well as the calcification types will be helpful in diagnosing follicular carcinoma
in combination with the cytopathological findings of follicular neoplasm.
Calcifications of undifferentiated carcinomas mostly presented as 2 types, the massive
and egg-shell types ([Table 1(f)]). Strictly speaking, it is most likely that calcifications in patients with undifferentiated
carcinomas are not located in the tissues of the undifferentiated carcinoma. Instead,
they are usually located in the tissues of a preceding well-differentiated carcinoma
on the periphery. This observation indicates the following important point in ultrasonic
diagnosis: when a tumor with a large solid-and-hypoechoic lesion as well as massive
[23] or egg-shell type calcifications on the periphery is observed, we should try to
acquire neoplastic cells from the large solid-and-hypoechoic lesion rather than the
calcified lesion on the periphery when performing fine-needle aspiration cytology.
Calcifications in medullary carcinomas are hypothesized to be derived from amyloid
deposits [24]
[25]
[26]. Medullary carcinomas have a wide range of calcification types on ultrasonography
([Table 1(g)]). However, the egg-shell type calcification is not observed. These fragmentary calcifications
are located in a comparatively central area of the tumor instead of in the periphery.
These findings can be viewed as characteristic features of medullary carcinomas.
The absence of calcifications ([Table 1(h)]) as well as an irregular border, markedly hypoechoic internal echoes, and enhanced
posterior echoes may be characteristic features for the ultrasonic diagnosis of primary
thyroid lymphoma [27].
The prevalence and types of calcifications were clarified in new outpatients, and
had an association with the histopathological type of patients who underwent thyroid
surgery. Therefore, assessing calcifications on ultrasonography is expected to contribute
to the ultrasonic diagnosis of thyroid tumors. It should be noted that not all thyroid
tumors have calcifications, but calcifications as well as other findings on ultrasonography
have an important predictive value for the diagnosis of thyroid tumors.
The number of criteria needed to predict malignancy has been discussed [28]. On the whole, ultrasound features, such as taller-than-wide shape, irregular margin,
hypoechogenicity, presence of microcalcifications, and intravascular vascularity,
can be useful diagnostic criteria for predicting malignancy. That is because papillary
carcinomas that present such typical features comprise the majority of thyroid malignancies.
However, when we perform ultrasound examination of each patient, it is necessary to
read and interpret other characteristic features such as tumor thrombus [19], tumor protrusion [20], or “nodule in nodule” appearance [21] to predict malignancy. Also, calcifications together with a specific ultrasound
appearance such as cyst-forming [13] or diffuse swelling gland [16] should be noted for differential diagnosis in each patient.
Strain ultrasound elastography has recently been introduced in the clinical workup
of thyroid nodules, and is an ultrasound-based technique for assessing the biomechanical
properties of tissue. It is said that strain ultrasound elastography for the characterization
of thyroid nodules in expert hands may be considered a useful complement to grayscale
ultrasonography by enhancing its accuracy for thyroid malignancy detection [29]. However, it has some limitations such as inherent technical limitations as well
as the histopathological features of the nodules themselves. These limitations may
lead to misinterpretations and pitfalls [29]. More work on strain ultrasound elastography is expected in the future in order
to overcome the limitations.
