Ultrasound Parameters of Thyroid Nodules and the Risk of Malignancy: A Retrospective Analysis.

Ultrasonography-guided fine-needle aspiration biopsy is the common choice for diagnosis of the suspected thyroid nodule. An algorithm(s) that finds the malignant potential of a nodule preoperatively, to overcome unnecessary diagnostic methods, does not exist. The objective of the study was to correlate thyroid nodule sizes measured by ultrasonography and risk of malignancy assessed by cytologic and histologic examinations. Data regarding fine-needle aspiration cytology and the results of histologic examinations of surgical specimens of 260 nodules were collected and analyzed. The macro or multiple calcifications, the complex echo pattern, and posterior region homogeneity were considered suspicious in ultrasonography. Bethesda system for classification of thyroid nodules was used for cytopathology. Histopathology performed as per the 2004 World Health Organization classification system. The benefit score analysis was performed for determination of clinical usefulness. Twenty-eight of 49 malignant nodules and 46 of 68 malignant nodules detected through ultrasound following fine-needle aspiration cytopathology and histopathology were <2 cm in size. A correlation was found for malignancy rate detected by ultrasonography-guided fine-needle aspiration cytology and those of the surgical specimen (r = 0.945, P = .015, R 2 = 0.894). Ultrasonography-guided fine-needle aspiration cytology had 0.994 sensitivities, 0.721 accuracies, and 0.08 to 0.945 diagnostic confidence for the detection of malignant nodules. Nodule size less than 2 mm (P = .011) was associated with the malignancy potential of thyroid nodules. Ultrasonography-guided fine-needle aspiration cytology had 19 (7%) results as a false negative and 1 (1%) results were false positive. Ultrasound-guided fine-needle aspiration cytopathology reported oversize of thyroid nodule than original but can predict the risk of malignancy. Level of Evidence: III.

Introduction

Thyroid cancer is more frequently observed in females.[1,2] It is very common in specialized centers.[3] Besides treatment, features of thyroid carcinoma are different among the Chinese population.[1] There is the development of several modern imaging modalities such as the computed tomography for the detection of thyroid cancer in incidental conditions,[4] but most of the cases of thyroid nodules are benign.[5] Therefore, ultrasonography-guided fine-needle aspiration biopsy is the common choice for diagnosis of suspicious thyroid nodule,[6] and this technique has difficulties in the definitive characterization of thyroid nodule[5] because most of patients with intermediate or suspicious cytopathological results are subjected to surgery.[7]

Ultrasonographic features concerning for malignancy include hypoechogenicity of microcalcifications and irregular margins of thyroid,[3] but these features of ultrasound have high interobserver variabilities,[5] leading to changes in sensitivity and accuracy. Nodule size would be a predictor of malignancy.[8] Nodule size, fine-needle aspiration cytology, and clinical history are also helpful for the prediction of malignancy.[7] However, retrospective studies suggested that with an increase in nodule size, there is no increase in the risk of malignancy.[8,9] Moreover, surgeons have experience of false-negative fine-needle aspiration cytology results in the case of thyroid nodules more than 4 mm in size.[10,11] Whereas the correlation is reported for the nodules sized less than 1.5 mm and the risk of malignancy.[12,13] The current predicted model for thyroid malignancy is based on age, nodule size, and fine-needle aspiration cytopathology.[7] Ideally, surgeons require an algorithm(s) that finds the malignant potential of a nodule preoperatively to reduce unnecessary diagnostic methods. Unfortunately, at present, such methods do not exist. The objective of the retrospective analysis was to correlate thyroid nodule sizes measured by ultrasonography and risk of malignancy assessed by cytologic and histologic examination.

Materials and Methods

Ethics Approval and Consent to Participate

The designed protocol (AQU/CL/13/19 dated October 13, 2019) of the established study was approved by the review boards of the Affiliated Hospital of Qingdao University and Weihai Central Hospital and the medical council of China. The data were collected with the hospital after permission from the competing authorities of the parent hospital and the referring hospitals for the study purpose(s). Informed consent was obtained from all patients who enrolled in the study regarding diagnosis, biopsies, and the publication of the study including personal data and imaging irrespective of time and language during hospitalization.

Study Population

From July 15, 2017, to June 17, 2019, a total of 214 adult patient (>18-year-old) underwent ultrasonographic examination and followed by fine-needle aspiration and/or surgical intervention at the Affiliated Hospital of Qingdao University, Qingdao, Shandong, China, and the Weihai Central Hospital, Weihai, Shandong, China. The data of diagnosis and postsurgical pathology were collected and evaluated. For patients with multiple nodules, each nodule was considered as separate in the analysis.

Ultrasonography

Real-time ultrasonography was performed using EPIQ Elite ultrasound equipment (Philips) and 15 MHz linear transducers by radiologists (a minimum of 5 years of experience in thyroid imaging) of institutes. Ultrasound images evaluated retrospectively by digitally stored images. The macro or multiple calcifications (appearance, size, and/or posterior acoustic shadow), the complex echo pattern (mixed nodules or more than 2 ecogenicities), and posterior region homogeneity (related to acoustic shadow and attenuation included in the nodule) were considered suspicious (Figure 1).[14] Features including echo pattern heterogeneity, a mixed posterior echo pattern, irregular shape, an obscure boundary, taller than wider in shape, and a spiculate margin in echo pattern were considered benign (Figure 2).[14]

Figure 1.

Ultrasound diagnosis of the suspicious thyroid nodules. A, Ultrasound image of thyroid of a 44-year-old male patient. The white arrow shows calcification located in the middle of the right third left lobe, the size of nodule 2.45 ± 0.12 cm. B, Ultrasound image of thyroid of a 46-year-old female patient. The white arrow shows the complex echo pattern in the right lobe, the size of nodule 2.12 ± 0.15 cm. C, Ultrasound image of thyroid of a 47-year-old male patient. The white arrow shows the posterior region homogeneity in the middle of the right third left lobe, the size of nodule 2.71 ± 0.11 cm.

Figure 2.

Ultrasound diagnosis of the benign thyroid nodules. A, Ultrasound image of thyroid of a 47-year-old female patient. The white arrow shows an echo pattern heterogeneity in the middle of the right lobe, the size of nodule 1.22 ± 0.32 cm. B, Ultrasound image of thyroid of a 49-year-old male patient. The white arrow shows a mixed posterior echo pattern in the left lobe, the size of nodule 1.66 ± 0.08 cm. C, Ultrasound image of thyroid of a 51-year-old female patient. The white arrow shows the irregular shape in the echo pattern in the left lobe, the size of nodule 2.18 ± 0.19 cm. D, Ultrasound image of thyroid of a 52-year-old male patient. The white arrow shows the obscure boundary in the echo pattern in the right lobe, the size of nodule 0.42 ± 0.11 cm. E, Ultrasound image of thyroid of a 51-year-old female patient. The white arrow shows the taller than wider in shape in the echo pattern in the right lobe, the size of nodule 1.52 ± 0.18 cm. F, Ultrasound image of thyroid of a 53-year-old female patient. The white arrow shows the spiculate margin in the echo pattern in the right lobe, the size of nodule 1.25 ± 0.22 cm.

Fine-Needle Biopsies

Fine-needle biopsies were performed by a 30G needle with a 30-mL syringe under the guidance of ultrasonography. The samples were collected in an alcohol-based aspirator and send to the pathological laboratory. Pathologists (a minimum of 5 years of experience) of institutes performed fine-needle biopsies. Size and suspicious nature of ultrasound criteria were adopted for the performance of fine-needle biopsies.

Cytology

Cells of biopsies were analyzed by pathologists (a minimum of 5 years of experience) of institutes as per the 2017 Bethesda system for classification of thyroid nodules.[6] True papillae, nuclear pseudoinclusions, a predominance of microfollicles, and psammoma bodies were considered as suspicious (Figure 3).[15]

Figure 3.

The histopathology of the fine-needle aspiration biopsies. (A) True papillae, (B) nuclear pseudoinclusions, (C) a predominance of microfollicles, and (D) psammoma bodies.

Thyroidectomy

In the case of larger size and suspicious nodules, thyroidectomy is performed. Partial or complete resection of the thyroid gland was performed by a team of otolaryngologist, endocrinologist, and ENT (ear nose throat) surgeon (all have a minimum of 5 years of experience).

Histopathology of the Surgical Specimen

The resected part of the thyroid was sent to the pathological laboratory for morphology and histopathology. Pathologists (a minimum of 5 years of experience) of institutes have performed the histopathology as per 2004 World Health Organization classification system.[16]

Clinical Usefulness

The benefit score analysis was performed for determination of clinical usefulness as per Equation 1[14]:

where, true-positive malignant nodule: Nodules detected as malignant by fine-needle aspiration cytopathology and by the histopathology of the surgical specimen. False-positive malignant nodule: Nodules detected as malignant by fine-needle aspiration cytopathology but did not detect malignant by the histopathology of the surgical specimen.

Diagnostic Parameters

The ratio of true-negative malignant nodule detected by fine-needle aspiration cytopathology to that detected by the histopathology of the surgical specimen is considered as sensitivity. The ratio of true-positive malignant nodule detected by fine-needle aspiration cytopathology to that detected by the histopathology of the surgical specimen is considered as accuracy.

Statistical Analyses

SPSS version 26.0 (IBM Corp) was used for statistical analysis purposes. The paired t test was performed for thyroid diameters between those detected through ultrasound and those detected through surgical specimens.[13] Spearmen rank correlation was developed for malignancy rate between ultrasound-guided fine-needle aspiration biopsies and surgical specimens considering coefficient (r) value between 0.3804 and 0.9965 as significant. The χ2 independence test was performed for numerical data of the diagnostic parameters. The nodule diameter was measured sonographically in ultrasound images and the size of the resected nodules has been measured pathologically in transverse, sagittal, and longitudinal dimensions.[17] A univariate following multivariate analysis was performed for identifying malignancy potential with clinical parameters and ultrasound findings.[18] The results were considered significant at 95% of the confidence level.

Results

Demographical Characteristics

A total of 641 nodules of 311 patients were screened ultrasonography, followed by fine-needle biopsies. Among them, 214 patients were subjected to partial or complete resection of the thyroid gland after ultrasonography-guided fine-needle biopsies (Figure 4). A total of 171 patients had a single nodule, 41 patients had 2 nodules, 1 patient had 3 nodules, and 1 patient had 4 nodules. Ultrasound detected a larger nodule size than the fresh surgical specimen (P < .0001). The other demographical and clinical characteristics of enrolled patients are reported in Table 1.

Figure 4.

Workflow diagram.

Table 1.

Demographical and Clinical Characteristics of Enrolled Patients.a

Characteristics		Value
Patients enrolled in the study		214
Age (years)	Minimum	45
	Maximum	64
	Mean ± SD	51.25 ± 5.45
Gender	Male	49 (23)
	Female	165 (77)
Ethnicity	Han Chinese	193 (90)
	Mongolian	18 (9)
	Tibetan	3 (1)
Family history	Yes	14 (7)
	No	200 (93)
Nodules per patient, mean (range)		1.21 (1-4)
Nodule size by ultrasound (cm)		2.75 ± 0.55
Nodule size of surgical specimen (cm)		2.11 ± 0.45
Thyrotropin (mIU/L)		1.66 ± 0.11
Blood type	A	55 (26)
	B	67 (31)
	AB	31 (14)
	O	61 (29)
Surgery	Total thyroidectomy	59 (28)
	Partial thyroidectomy	155 (72)
Lymphocytic thyroiditis	Negative	131 (61)
	Positive	83 (39)
Follow-up (days)		415.91 ± 52.42

Abbreviation: SD, standard deviation.

a Ordinal data are presented as frequency (percentage) and numerical data as mean ± standard deviation.

Ultrasonography/Fine-Needle Aspiration Cytology

Ultrasonography-guided fine-needle aspiration cytology reported 23 nodules as true papillae, 11 nodules as nuclear pseudoinclusions, 9 nodules as a predominance of microfollicles, and 6 nodules as psammoma bodies. Table 2 presents the distribution of fine-needle aspiration cytology-classified malignant nodules among the range of sizes.

Table 2.

Results of Grayscale Ultrasonography and Ultrasonography-Guided Fine-Needle Aspiration Cytology.a

Nodule size (cm)	Total nodule	Malignant nodule
≤0.99	36 (14)	22 (9)
1.00-1.99	91 (35)	6 (2)
2.00-2.99	54 (21)	11 (4)
3.00-3.99	34 (13)	5 (2)
≥4.00	45 (17)	5 (2)
Total	260 (100)	49 (19)

a Data are presented as frequency (percentage).

Results of the Surgical Specimen

Surgical pathology reported 33 nodules as true papillae, 18 nodules as nuclear pseudoinclusions, 11 nodules as a predominance of microfollicles, and 6 nodules as psammoma bodies. Table 3 presents distribution of the surgical specimen histopathology-classified malignant nodules among the range of sizes.

Table 3.

Morphological and Histopathological Results of the Surgical Specimen.a

Nodule size (cm)	Total nodule	Malignant nodule
≤0.99	39 (15)	32 (12)
1.00-1.99	101 (39)	14 (5)
2.00-2.99	61 (24)	12 (5)
3.00-3.99	40 (15)	5 (2)
≥4.00	19 (7)	5 (2)
Total	260	68 (26)

a Data are presented as frequency (percentage).

Overall, the majority of malignant nodules categorized by cytologic assessment (28/49; 57%) and/or histologic examination (46/68, 68%) had an ultrasound measurement of less than 2 cm in size. For nodules, greater than 2 cm in sizes through ultrasound was less frequent malignant by cytologic assessment (21/49; 43%) or histologic examination (22/68; 32%). There was a correlation for malignancy rate detected by ultrasonography-guided fine-needle aspiration cytology and those of the surgical specimen (r = 0.945, P = .015, R 2 = 0.894).

Ultrasonography-guided fine-needle aspiration cytology had 0.994 sensitivities and 0.721 accuracies for malignant nodules. It had 19 (7%) results as a false negative. While 1 (1%) results were false positive. The other diagnostic parameters of ultrasonography-guided fine-needle aspiration cytology are reported in Table 4.

Table 4.

Diagnostic Parameters.a,b,c

Parameters	The morphological and histopathological results of the surgical specimen	Ultrasonography/fine-needle aspiration cytology	Comparison
Total nodules analyzed	260	260	P value
True-positive malignant nodule	68 (26)	49 (19)	.059
True-negative malignant nodule	192 (74)	191 (73)d	.921
False-negative malignant nodule	0 (0)	19 (7)	<.0001
False-positive malignant nodule	0 (0)	1 (1)d	.317
Sensitivity	1	0.994d	.921
Accuracy	1	0.721	.059

a Data are presented as frequency (percentage).

b χ2 independence test was performed.

c A P < .05 was considered significant.

d Insignificant difference reference to the morphological and histopathological results of the surgical specimen.

Ultrasonography-guided fine-needle aspiration cytology had 0.08 to 0.945 diagnostic confidence for the detection of malignant nodules. Below 0.08 diagnostic confidence, it had no clinical importance, and above 0.945 diagnostic confidence, it had a risk of overdiagnosis (Figure 5).

Figure 5.

Clinical usefulness analysis. Surgery without ultrasonography/fine-needle aspiration cytology was for evaluation purposes only.

Univariate analysis showed that female (P < .0001), family history (P = .002), 2 cm or lesser size nodule (P < .0001), numbers of nodules (P = .041), lymphocytic thyroiditis (P = .021), and abnormal thyrotropin level (P = .003) were associated with malignancy potential of thyroid nodules. While multivariate analysis showed that 2 cm or lesser size nodule (P = .011) was only associated with malignancy potential of thyroid nodules (Table 5).

Table 5.

Multivariate Analysis for Association of Malignancy With Clinical Parameters and Ultrasound Findings.a,b

Malignant nodules assessed by histologic examinations	46
Parameters	Odd ratio	95% confidence limit	P value
Gender (female vs male)	0.58	0.15-0.62	.059
Family history (presence vs absence)	0.54	0.14-0.61	.067
Nodule size (≤2c vs >2 cm)	1.62	0.58-2.01	.011
Numbers of nodules (≥1 vs 1)	0.37	0.21-0.82	.062
Lymphocytic thyroiditis (yes vs no)	0.41	0.22-0.73	.083
Thyrotropin level (abnormal vs normal)	0.43	0.25-0.63	.092

a A P < .05 and odd ratio >1 were considered significant.

b Results of benign nodules assessed by histologic examinations considered reference standard.

c Significant parameter associated with malignancy of nodule.

Discussion

The nodule size reported by ultrasound was significantly higher than that of the surgical specimen. These results of the study were consistent with the results of retrospective studies.[5,12,19] The size difference by ultrasound finding is required to be in consideration to avoid unnecessary surgery.

The study reported that the risk of malignancy for thyroid nodule(s) was higher for those having 2 cm or lesser size in ultrasound. These results of the study were consistent with the results of retrospective studies[5,19] and prospective study[20] but were contraindicated with the results of retrospective studies.[8,9,21-23] The papillary thyroid carcinoma has an inverse relationship with thyroid nodule size.[8] The current study suggested that a less than 2-cm nodule size is the cutoff of the malignancy risk.

The study used a less than 2 cm as the threshold value for malignancy detection. Nodules’ growth beyond 2 cm is a specific type of thyroid cancer and can be detected in fine-needle cytopathology because it is mostly follicular carcinoma, which does not transform with growth. Also, the risk of follicular carcinoma is increased linearly below 2 cm.[8] Therefore, there is a need for proper diagnosis modality for nodules smaller than 2 cm to define them as malignant. That is why the study used a less than 2 cm cutoff for the detection of malignancy risk.

Ultrasonography-guided fine-needle aspiration cytology had 0.994 sensitivities and 0.721 accuracies. The results of the current study were consistent with the available study.[10] The Bethesda system of cytopathology under ultrasonography is the best predicting tool for the nature of the nodule.

Ultrasonography-guided fine-needle aspiration cytology had 7% results as false negative and 1% results as false positive for malignant nodules. The results of the current study were consistent with the available study.[10] An inadequate sampling of biopsies and the performance of pathologists were responsible for these results. However, adding experienced cytopathologist(s), false-negative, and false-positive results can be overcome.[24] Besides false-positive and false-negative results, ultrasonography-guided fine-needle aspiration cytology is a better predicting tool in the management of thyroid carcinoma.

Ultrasonography-guided fine-needle aspiration cytology had 0.08 to 0.945 diagnostic confidence for the detection of malignant nodules. The results of the current study were in line with a retrospective analysis.[25] Ultrasonography-guided fine-needle aspiration cytology may have high clinical usefulness before a decision of surgery in suspected thyroid nodules.

The study reported a positive correlation for the risk of malignancy predicted by ultrasound-guided fine-needle cytopathology with histopathology of the surgical specimen. These results of the study were consistent with the results of the prospective study[26] and retrospective analysis.[10] The Bethesda system would be accurately detected the nodule with malignancy.

There are several limitations of the study that have to be reported; for example, compared to retrospective study, the dynamic study provides more accurate results. Patients who faced thyroidectomy were only included in the analysis. Inter- and intraobserver reliability did not perform for results. The suspicious criteria used in the study are totally different from the current literature.[27,28] There are several guidelines for detection of suspicious nodules on grayscale ultrasound following fine-needle aspiration cytopathology, but the best approach for suspicious features is unclear.[29] The effects of size (detected by ultrasound) on false-positive and false-negative results of fine-needle cytopathology did not discuss.

Conclusions

Ultrasound-guided fine-needle aspiration cytopathology can predict oversize of thyroid nodule than original but may predict the risk of malignancy of the thyroid nodule. Also, the study suggested a less than 2 cm under ultrasonography as the cutoff for malignancy. Ultrasonography-guided fine-needle aspiration cytology may have high clinical usefulness before a decision of surgery in suspected thyroid nodules. The further use of molecular analysis is required for incomplete assessment of cytopathological and ultrasound parameters.