Utility of quantitative contrast-enhanced ultrasound for the prediction of lymph node metastasis in patients with papillary thyroid carcinoma

Abstract

OBJECTIVES:

The aim of this study was to find the optimal parameters and cutoffs to differentiate metastatic lymph nodes (LNs) from benign LNs in the patients with papillary thyroid carcinoma (PTC) on the quantitative contrast-enhanced ultrasound (CEUS) features.

METHODS:

A total of 134 LNs in 105 patients with PTCs were retrospectively enrolled. All LNs were evaluated by conventional ultrasound (US) and CEUS before biopsy or surgery. The diagnostic efficacy of CEUS parameters was analyzed.

RESULTS:

Univariate analysis indicated that metastatic LNs more often manifested centripetal or asynchronous perfusion, hyper-enhancement, heterogeneous enhancement, ring-enhancing margins, higher PI, larger AUC, longer TTP and DT/2 than benign LNs at pre-operative CEUS (p < 0.001, for all). Multivariate analysis showed that centripetal or asynchronous perfusion (OR = 3.163; 95% CI, 1.721–5.812), hyper-enhancement(OR = 0.371; 95% CI, 0.150–0.917), DT/2 (OR = 7.408; 95% confidence interval CI, 1.496–36.673), and AUC (OR = 8.340; 95% CI, 2.677–25.984) were predictive for the presence of metastatic LNs. The sensitivity and accuracy of the quantitative CEUS were higher than qualitative CEUS (75% vs 55 % and 83.6% vs 76.1 %, respectively).

CONCLUSIONS:

Quantitative CEUS parameters can provide more information to distinguish metastatic from benign LNs in PTC patients; In particular, DT/2 and AUC have a higher sensitivity and accuracy in predicting the presence of metastatic LNs and reduce unnecessary sampling of benign LNs.

Keywords

Contrast-enhanced ultrasound lymph node metastasis papillary thyroid carcinoma ultrasound

1 Introduction

Contrast-enhanced ultrasonography (CEUS) is a promising tool in the study of microvascular flow in different organs. It is also very helpful in differentiating benign and malignant tumors [1 –4] as it reveals the microvascularity of tumor tissue, obtaining a spatial and temporal resolution superior to the traditional color and power Doppler techniques. Moreover, the time-intensity curves (TIC) can be calculated, allowing both qualitative and quantitative evaluation [4 –7]. CEUS was first used in the diagnostic work-up of focal liver lesions. As time passed, it was additionally applied to the assessment of lesions within the kidneys, prostate, breasts, thyroid and other organs [8 –11]. CEUS using SonoVue is of label use for lymph node metastases considering the EFSUMB guidelines published in 2018 from Sidhu Pet AL in US in Med. Some studies had reported that CEUS may be a potential tool in differentiating between benign and malignant lymph nodes (LNs) [12 –15]. Other study evaluated changes in blood vessel volume and density using different agents (SonoVue and Sonazoid), showed that alterations in blood vessel volume and density precede alterations in LN size in the early stages of lymph node metastasis (LNM) [16]. More recently, CEUS has been introduced to improve the sensitivity and specificity in the diagnosis of metastatic LNs in PTC patients [17 –21]. Our previous study reported that metastatic LNs more often manifested centripetal or asynchronous perfusion, hyper-enhancement, heterogeneous enhancement, perfusion defects and ring-enhancing margins than benign LNs at preoperative CEUS, suggesting the combination of US and CEUS could improve the accuracy of diagnosis of metastatic cervical LNs in patients with PTC [17]. However, qualitative CEUS analysis was highly specific, with relative low sensitivity for detecting metastic LNs, quantitative CEUS has higher sensitivity and it has the capability to detect the subtle hemodynamic changes of the lesions. To our knowledge, few reports [18, 19] have focused on quantitative analysis, and its value in daily practice is still controversial because of inconsistent and unsatisfying results and disagreement on methodology [22–23]. Thus, this study mainly focuses on the quantitative CEUS features of cervical LNs in patients with PTC to seek suitable method and the optimal parameters and cutoff value to differentiate metastatic LNs from benign LNs.

2 Material and methods

2.1 Patients

This study was approved by the Ethics Committee of our hospital(2020-174). All patients signed informed consent before examinations. One hundred and five consecutive patients (70 females and 35 males; age range: 19–77 y; mean age: 42.9±15.3 y) with thyroid nodules (Bethesda system≥IV cytology) or previous history of thyroid cancer surgery scheduled for thyroidectomy or neck lymph node dissection(LND) between October 2018 and October 2019 were enrolled in this study. All eligible patients underwent conventional US and CEUS within 3 d before biopsy or surgery.

The inclusion criteria were as follows: (1) the presence of a single or multiple cervical LNs in patients ≥18 years of age with thyroid nodules (Bethesda system ≥IV cytology) scheduled for thyroidectomy. (2) Cervical LNs in patients with a previous history of thyroid cancer surgery. (3) All cervical LNs had a long-axis diameter of ≥1.0 cm or a short-axis of ≥0.5 cm plus at least one of the following ultrasound features suspicious for malignant involvement: focal or diffuse hyper-echogenicity, presence of internal microcalcification, cystic aspect, round shape and chaotic or peripheral vascularity on Doppler US. Exclusion criteria were as follows: (1) patients with incomplete documents; (2) metastatic LNs not from PTC or other diseases involving LNs such as lymphoma, tuberculosis, Castleman disease, and sarcoidosis.

The final diagnosis of malignant cervical LNs was confirmed by one of the following criteria: (i) surgical pathologic results after operation; (ii) core needle biopsy; (iii) FNA cytology. Meanwhile, LNs were considered benign if (i) the LN was diagnosed as benign on histology after surgery; (ii) core needle biopsy or repeat FNA confirmed benignity; Finally, a total of 134 LNs (60 metastatic and 74 benign LNs) in 105 patients were included in this study.

2.2 Contrast-enhanced ultrasound examination

All patients were detected by conventional ultrasound and color Doppler US. The longitudinal view of each LN with suspicious area was chosen as the ideal plane for CEUS. CEUS was performed using a Resona 7 (Mindray Company, China) machine equipped with UWN⁺ technology. An L 11-3 WU (3–11 MHz) linear array transducer was used in each patient. The mechanical index (MI = 0.05–0.10) was selected automatically by the system in relation to beam-focus depth. The contrast agent SonoVue (Bracco SpA, Milan, Italy) used in this study was supplied as a lyophilized powder, which was reconstituted by adding 5 mL of 0.9% saline and gently shaking the vial by hand to form a homogeneous microbubble suspension. A 19-gauge cannula was inserted into an antecubital fossa vein, and 1.2 mL of SonoVue was injected as a bolus followed immediately by a 5 mL saline flush via a three-way tap for each contrast study, the entire movie sequence (at least 2 min) was stored for analysis. For patients with multiple suspicious LNs or LN with unsatisfactory enhancement, repeated injection were performed.

2.3 Image analysis

All examinations were recorded in a compute and analyzed independently both qualitatively and quantitatively by two independent investigators with more than 8 years of experience with thyroid CEUS. When two investigators disagreed in some point, they jointly reviewed the images and came to a consensus on characterization of the enhancement patterns in those cases.

Qualitative analysis

The qualitative CEUS patterns were consistent with our previous study as follows [17]. (1) perfusion pattern was separated into centrifugal perfusion, centripetal perfusion, asynchronous perfusion and chaotic perfusion. (2) The degree of enhancement at peak enhancement relative to that of the adjacent tissue was classified into non-enhancement, hypo-enhancement, iso-enhancement or hyper-enhancement. (3) Enhancement homogeneity was divided into heterogenous and homogenous. (4) Perfusion defect was distinguished as present or absent. (5) Enhancement margin was classified into clear, blurred or ring-enhancing.

Quantitative analysis of time-intensity curves (TICs).

When the LN showed asynchronous perfusion, the hyperechoic area with rapid hyper-enhancement, which was proved as localized tumor colony in previous study [24, 25], was chosen as the ROI (Fig. 1). Otherwise the entire LN is manually traced (Fig. 2). The ROI was drawn while avoiding the thick nourishing vessels and calcification. TIC was automatically generated by Contrast Imaging QA software. The quantitative parameters included: the arrival time(AT), the peak intensity (PI), area under the curve (AUC), time from peak to one half (DT/2) and time to peak (TTP). In this study, PI was calculated as the TIC curve’s maximum intensity minus the curve’s baseline intensity; TTP was measured as the time from the start of the rising TIC to its PI, DT/2 was measured as the time needed from PI to when it drops to half of PI and AUC as the area under the TIC.

Fig. 1

Images of a metastatic LN with asynchronous perfusion, hyperechoic area of LN (ROI 1, pink circle) and adjacent LN (ROI 2, yellow circle) under CEUS (C) and conventional ultrasound (T). Time-intensity curves for the hyperechoic area (pink line) and adjacent LN (yellow line) were analyzed. The parameters peak intensity (PI), ascend slope (AS), descend slope (DS), area under the curve (AUC), time to peak (TTP), and descend time to one-half (DT/2) of ROIs 1 and 2 are listed in the table.

Fig. 2

Images of a metastatic LN with non- asynchronous perfusion, the entire area (ROI 1, pink circle) and adjacent LN (ROI 2, yellow circle) under CEUS (C) and conventional ultrasound (T). Time-intensity curves for the entire area (pink line) and adjacent LN (yellow line) were analyzed. The parameters peak intensity (PI), ascend slope (AS), descend slope (DS), area under the curve (AUC), time to peak (TTP), and descend time to one-half (DT/2) of ROIs 1 and 2 are listed in the table.

2.4 Histopathology

All the lesions underwent surgical resection or biopsy, followed by the histopathological examination. The surgical and sampling procedures is the same as our previous study [17]. Pathologic diagnosis was confirmed by a single pathologist (Q.Y.) who has 12 years of working experience.

2.5 Statistical analysis

SPSS Version 20.0 software was used for statistical analysis. Qualitative data were analyzed with the Pearson Chi-square 2 test or Fisher’s exact test. Quantitative data were presented as the mean (±standard deviation [SD]) and Student’s t-test was applied to check for a statistical difference of the quantitative parameters between benign and metastatic LNs. The sensitivity, specificity, accuracy, PPV and NPV of all data that proved to be statistically significant on univariate analysis were calculated using MedCalc® software. Binary logistic regression analysis with stepwise forward variable selection was used to choose the final parameters (P value for entry and removal were 0.05 and 0.1 respectively). Receiver operating characteristic (ROC) curve analysis was used for the evaluation of the diagnostic performance of the quantitative parameters and obtained the cut-off value. In all comparisons, a P-value < 0.05 was considered to indicate statistical significance.

3 Results

3.1 Surgery and histology

A total of 134 LNs (60 metastatic and 74 benign LNs) in105 patients with thyroid nodules (Bethesda System≥IV cytology) or with a previous history of thyroid cancer surgery who had undergone surgery and histologic examination were enrolled in this study. Among them, 26 cases with PTC underwent thyroidectomy and CND only,73 cases with PTC underwent thyroidectomy and CND combined with LND, 3 cases with a previous history of thyroid cancer surgery underwent LND and 3 cases with benign thyroid nodules (goiter) underwent hemithyroidectomy without LND.

3.2 Qualitative analysis of CEUS

The CEUS manifestations are listed in Table 1. There are statistically differences between benign and metastatic groups in perfusion pattern, degree of enhancement, enhancement homogenicity and enhancement margin (P < 0.05). The sensitivity and specificity in the evaluation of LNM were 35% and 95.9%, respectively in cases with centripetal or asynchronous perfusion, 36.7% and 95.9%, respectively in cases with hyper-enhancement, 41.7% and 77%, respectively in cases with heterogeneous enhancement, and 10% and 98.6%, respectively in cases with ring-enhancing margin. Most of the above qualitative features were highly specific (95.9% –98.6%), but with low sensitivity (10% –36.7%)(Table 2). Multivariate logistic regression showed that centripetal or asynchronous perfusion (OR = 3.163; 95% CI, 1.721–5.812; P = 0.000), hyper-enhancement(OR = 0.371; 95% CI, 0.150–0.917; P = 0.032) were predictive for the presence of LNM (Table 3).

Table 1
Comparison of CEUS qualitative and quantitative parameters between benign and metastatic LNs

Parameters Metastatic LN (n = 60) Benign LN (n = 74) P value

Perfusion pattern 0

Centrifugal 3 29

Centripetal 11 2

Chaotic 36 42

Asynchronous 10 1

Degree of enhancement 0

Hypo-enhancement 22 14

Iso-enhancement 16 57

Hyper-enhancement 22 3

Enhancement homogenicity 0.02

Homogeneous 35 57

Heterogeneous 25 17

Perfusion defect 0.076

Present 6 2

Absent 54 72

Enhancement margin 0.0496

Clear 37 59

Blurred 20 14

Ring-enhancing 3 1

AT (sec) 0.21±0.092 0.20±0.034 0.21

TTP (sec) 13.67±4.50 10.167±4.63 0

PI (dB) 13.15±3.45 11.08±2.71 0

DT/2 (sec) 65.36±13.22 52.04±12.28 0

AUC(dBsec) 1820.8±611.51 1050.5±346.63 0

Parameters	Metastatic LN (n = 60)	Benign LN (n = 74)	P value
Perfusion pattern			0
Centrifugal	3	29
Centripetal	11	2
Chaotic	36	42
Asynchronous	10	1
Degree of enhancement			0
Hypo-enhancement	22	14
Iso-enhancement	16	57
Hyper-enhancement	22	3
Enhancement homogenicity			0.02
Homogeneous	35	57
Heterogeneous	25	17
Perfusion defect			0.076
Present	6	2
Absent	54	72
Enhancement margin			0.0496
Clear	37	59
Blurred	20	14
Ring-enhancing	3	1
AT (sec)	0.21±0.092	0.20±0.034	0.21
TTP (sec)	13.67±4.50	10.167±4.63	0
PI (dB)	13.15±3.45	11.08±2.71	0
DT/2 (sec)	65.36±13.22	52.04±12.28	0
AUC(dBsec)	1820.8±611.51	1050.5±346.63	0

CEUS: contrast-enhanced ultrasound. AT: arrival time. TTP: time to peak. PI: peak intensity. DT/2: time from peak to one half. AUC: area under the curve.

Table 2

Diagnostic performances of different parameters to predict metastatic LNs

Parameters	Sensitivity(%)	Specificity(%)	PPV(%)	NPV(%)	Accuracy (%)
Centripetal or asynchronous perfusion	35	95.9	87.5	64.5	68.7
Hyper-enhancement	36.7	95.9	88	65.1	69.4
Heterogeneous enhancement	41.7	77	59.5	62	61.2
Ring-enhancing margin	10	98.6	75	56.2	56.7
TTP (sec)	80	63.5	64	79.7	70.9
PI (dB)	93.3	28.4	51.4	84	57.5
DT/2 (sec)	88.3	71.6	71.6	88.3	79.1
AUC(dBsec)	76.7	87.8	83.6	82.3	82.8

Table 3

Multivariate logistic analysis of the CEUS parameters

Characteristic	OR	95% CI	P value
Perfusion pattern	3.163	1.721–5.812	0.000
Degree of enhancement	0.371	0.150–0.917	0.032
DT/2 (sec)	7.408	1.496–36.673	0.014
AUC(dBsec)	8.340	2.677–25.984	0.0002

3.3 Quantitative analysis of CEUS.

As shown in Table 1, all of the quantitative parameters except for AT were found to be significantly different between the benign and metastatic groups. The PI was significantly higher in the metastatic group than in the benign group (P = 0.000); the TTP and DT/2 were significantly longer in the metastatic group than in the benign group (all P = 0.000); and the AUC of the TIC was significantly larger in the metastatic group than in the benign group (P = 0.000). The TTP was the variable used to calculate wash-in. The DT/2 was the variable used to calculate wash-out. The longer TTP and DT/2 indicate slow wash-in and wash-out.

A receiver-operating-characteristic (ROC) curve analysis was performed to determine the cut-off scores for the quantitative CEUS parameters in further characterizing the metastatic LNs of the PTCs. The cut-off score was the one closest to the point with both highest sensitivity and specificity (MedCalc® software). LNs with measurements above the cut-off score of TTP (10.4 sec), DT/2 (53.91 sec), PI(13 dB) or AUC(1333.50 dBsec), with the AUC of 0.731, 0.823, 0.82 and 0.865, respectively, were regarded as metastatic LNs (Fig. 3). The sensitivity and specificity of TTP in the diagnosis of LNM were 80% (48/60) and 63.5% (47/74) respectively, 88.3% (53/60) and 71.6% (53/74), respectively for DT/2, and 93.3% (56/60) and 28.4% (21/74), respectively for PI, 76.7% (46/60) and 87.8% (65/74), respectively for AUC (Table 2). In contrast, most of the above quantitative parameters had relatively higher sensitivity (76.7% –88.3%) but lower specificity (63.5% –86.7%) (Table 2).

Fig. 3

Analysis of the ROC curve by the (A) time-to-peak peak intensity, (B) peak intensity, (C) time from peak to one half (DT/2), (D) area under the curve (AUC)between the benign and malignant LNs. ROC, receiver operating characteristic.

Multivariate logistic regression analysis revealed that DT/2 (OR = 7.408; 95% confidence interval CI, 1.496–36.673; P = 0.014) and AUC (OR = 8.340; 95% CI, 2.677–25.984; P = 0.000) were predictive for the presence of LNM (Table 3).

To determine the utility of quantitative CEUS in the correct identification of benign versus malignant nodes, the results were compared. For parameters used for analysis were those found to have high specificity and statistical significance in the benign-versus-metastatic comparison: for qualitative CEUS parameters - centripetal or asynchronous perfusion, hyper-enhancement and ring-enhancing; for quantitative CEUS parameters –DT/2 >53.91 sec and AUC >1333.50 dBsec. LNs with one or more of the aforementioned qualitative features were considered to have metastasis in qualitative CEUS. LNs with both the aforementioned quantitative features were considered to have metastasis in quantitative CEUS. Under this criterion, the diagnostic accuracy of qualitative and quantitative CEUS is outlined in Table 4. When qualitative CEUS was used, most of benign LNs were identified (69/74, 93.2%), but only 55% (33/60) metastatic LNs were identified. Quantitative CEUS correctly detected 16 LNM in 27 false negative patients, mainly improved sensitivity rather than specificity in the diagnostic process. The sensitivity (75%) and accuracy (83.6%) of the quantitative CEUS were higher than qualitative CEUS (55 % and 76.1 %, respectively).

Table 4

Diagnostic accuracy of Qualitative and Quantitative analysis CEUS for metastatic LNs

Modalities	AUC (95% CI)	Sensitivity (%)	Specificity (%)	Positive predictive value (%)	Negative predictive value (%)	Accuracy (%)
Qualitative analysis	0.80(0.72–0.87)	55	93.2	86.8	71.9	76.1
Quantitative analysis	0.83(0.75–0.90)	75	90.5	86.5	81.7	83.6

4 Discussion

Cervical LN metastasis is the most common metastatic pathway of PTC(including malignant PCTNs), with a metastasis rate of 30–80% [26 –28]. The presence of cervical lymph nodal metastases affects surgery type, extent of surgery and prognosis and was considered one of risk factors for local tumor recurrence and cancer- specific mortality [29, 30]. Therefore, the differentiation of metastatic LNs from benign LNs in PTC is very important to decide further management plans of the patients. SonoVue for the detection of lymph node metastases is of label use since 2018. The present study suggested that the CEUS parameters distinctly different in the metastatic and benign patients, especially regarding the parameters of perfusion pattern, the degree of enhancement, DT/2, and AUC. Furthermore, the analysis of the diagnostic performance of these parameters could be used to characterize the metastatic and benign LNs.

Our study showed that metastatic LNs more often manifested centripetal or asynchronous perfusion, hyper-enhancement, ring-enhancing on univariate qualitative CEUS analysis(p < 0.05). Only centripetal or asynchronous perfusion and hyper-enhancement were significantly related to LN metastasis in the multivariate analysis, which is basically consistent with our previous study [17, 21]. However, twenty-seven metastatic LNs in our study were misdiagnosed as benign ones by this method(19 cases with hypo-enhancement and 24 cases with chaotic perfusion), the false negative rate is up to 45%. The results indicated that the qualitative CEUS analysis was highly specific, but with low sensitivity criteria for detecting metastatic LNs, with specificity of 93.2% and sensitivity of 55%. Because the overlapping of chaotic perfusion, homogenicity enhancement and enhancement margin might interfere with the CEUS operator’s judgment. It has difficulties in differentiation diagnosis of benign and metastatic LNs using the qualitative CEUS only. When add quantitative analysis, sixteen out of 27 (59.3%) were correctly classified (all of them showed longer DT/2 and lager AUC), indicating that using quantitative CEUS to re-evaluate the LNs not only improves sensitivity but also retains comparatively high accuracy.

We deduced that measurement values from the TIC can be helpful in detecting the blood kinetic information of metastatic LNs which is hard to diagnose precisely based on visual characteristics using CEUS. Given that the ascending and descending branches of the time-intensity curve reflects the changes in microbubble velocity and flow capacity with time [31], we paid more attention to the ascend slope and descend slope: the parameters that directly reflect the wash-in and wash-out phase characteristics. We found a low nodule ascend slope and descend slope to be risk factors of malignancies. Metastatic LNs displayed a characteristic of “slowly washed in and delayed washed out” which may be associated with the malignant behavior of metastatic LNs. Metastatic LNs’ secret vascular growth factors form a large number of immature microvessels, which were distorted, caliber irregular and with incomplete vascular walls. Due to these irregular vascular patterns, blood often flows erratically, with possible stasis and turbulence [32 –35], therefore the entrance and exit distance of contrast agent in tumor increases. Meanwhile, the interstitial edema and fibrosis, the malignancy infiltration might cause neovascular stenosis or occlusion. Furthermore, some studies suggested that the sustained and slow decline of TIC curve in metastatic LN curve may be related to tumor venous feedback disorder [12, 36]. Although arterio-venous shunts and draining veins increasing could lead to the shorter TTP and DT/2, the whole TICs shows a gradual rise and decline with longer TTP and DT/2 in metastatic LNs.

Compared with benign LNs, the PI and AUC of metastatic LNs was significantly greater. The PI value represents the maximum dose of contrast agent that has reached the region of interest per unit time, and the AUC value represents the area under the TIC. Both of these parameters are signs of higher enhancement degree, and in some aspects, this might lead to repetitive use of the data to some extent. Therefore, although multivariate logistic regression showed that DT/2 and AUC were predictive for the presence of metastatic LNs, DT/2 was the most valuable quantitative parameter regarding the diagnosis of metastatic LNs.

In this study, although the diagnostic accuracy increased after using TIC analysis, there were still 11 false negative cases which were missed by both modalities, pathological results showed microscopic LNM. Similar findings have been reported previously, Xiang et al concluded that both CEUS and conventional US were not sensitive in detecting microscopic LNM, in which the generally used malignant criteria were not obvious [21 , 38]. Shao SH et al. demonstrated that six malignant breast lesions with a maximum diameter <15 mm were misdiagnosed by CEUS as benign, the growth of these small malignant tumors could rely on normal surrounding capillary network without forming substantial angiogenesis. The lack of malformed neovascularity may lead to misdiagnosis [39 –41]. In some aspects, our findings were consistent with this concept.

A few previous studies have mentioned quantitative analysis of metastatic LNs. However, the reported results were inconsistent among different studies [18, 19]. In the study of Wei et al. [18], the intensity (PI) was significantly higher in the metastatic group, whereas Chen et al. [19] evaluated the qualitative parameters in 55 LNs in 46 patients, and found no statistical significance in the quantitative parameters PI, TTP, MTT and RT between the groups and deduced that the inconsistent might have been due to different methods. In this study, however, we found that metastatic LNs have higher PI, larger AUC, longer TTP and DT/2 than benign LNs at pre-operative CEUS. This may be due to the new ROI –choosing method, which previous studies rarely mentioned, that is manually depicting local lesions(hyper-echogenic islands, proved as localized tumor colony) [24, 25] when the LN presents asynchronous perfusion. The reason is that the intra-nodal hyper-echogenic islands, which manifested hyper-enhancement reflects the real hemodynamics information of the metastatic LNs. Otherwise, for each lesion, the video at least 120 s was taken into study, which is longer than 90 s in Chen et al’s report, and may lead to different results in DT/2 and AUC.

There are still some limitations of our study. First, only reactive LNs were included in the benign group. Other causes of lymphadenopathy, such as tuberculosis, purulent lymphadenitis, granulomatous infections, lymphoma and metastatic nodes from other primary tumors were not investigated in this study. Second, US can accurately diagnose cervical lymphadenopathy throughout the jugular region, but is limited to detecting nodes along the transverse cervical branches in the posterolateral neck [42]. Third, the selection of CEUS plane and ROIs is subjective, lacking the reference area, and also the plane remained unchanged during the whole procedure is not easy. Finally, the caseload in our study was relatively small and larger population is needed for further studies.

5 Conclusions

In summary, quantitative CEUS parameters can provide more information to distinguish metastatic LNs from benign LNs in PTC patients; in particular, DT/2 and AUC have a higher sensitivity and accuracy in predicting the presence of metastatic LNs and reduce unnecessary sampling of benign LNs.

Conflict of interest

All the authors certify that there is no actual or potential conflict of interest regarding this article.

Funding

This work was supported by grants from the general research program of Education Bureau in Zhejiang Province (Y201738146), Natural Science Foundation nonprofit research projects of Zhejiang Province of China (LGF19H180020) and Natural Science Foundation Committee general program of Zhejiang Province of China (Y16H180019).

References

Danila

, et al. The role of contrast enhanced ultrasound (CEUS) in the assessment of liver nodules in patients with cirrhosis. Med Ultrason. 2010;12:145–149.

Schleder

, Janke

, Agha

, et al. Preoperative differentiation of thyroid adenomas and thyroid carcinomas using high resolution contrast-enhanced ultrasound (CEUS). Clin Hemorheol Microcirc. 2015;61(1):13–22.

, et al. Diagnostic performances of various gray-scale, color Doppler, and contrast-enhanced ultrasonography findings in predicting malignant thyroid nodules. Thyroid. 2014;24:355–63.

Zhang

, et al. Utility of contrast-enhanced ultrasound for evaluation of thyroid nodules. Thyroid. 2010;20:51–7.

Luzurier

, Maxwell

, Correas

, et al. Qualitative and quantitative contrast-enhanced ultrasonography for the characterisation of non-palpable testicular tumours. Clin Radiol. 2018;73(3):322.e1–322.e9.

Schaible

, Stroszczynski

, Beyer

, Jung

. Clin Hemorheol Microcirc. 2019;73(1):95–104.

Liu

, Zhou

, Yang

, Zhou

, Wu

, Chen

, Ye

, Luo

. Exp Ther Med, Nov. 2017;14(5):4321–7.

Halpern

, Rosenberg

, Gomella

. Prostate cancer: contrast-enhanced us for detection. Radiology. 2001;219(1):219–225.

Ladam-Marcus

, Mac

, Job

, Piot-Veron

, Marcus

, Hoeffel

. 2009. Contrast-enhanced ultrasound and liver imaging: review of the literature. J Radiol. 90(1 Pt 2):93–106.

10.

Hirooka

, Itoh

, Kawashima

, et al. Contrast-enhanced endo-scopic ultrasonography in digestive diseases. J Gastroenterol. 2012;47(10):1063–1072.

11.

Cox

, et al. Validation of a technique using microbubbles and contrast enhanced ultrasound (CEUS) to biopsy sentinel lymph nodes (SLN) in pre-operative breast cancer patients with a normal grey-scale axillary ultrasound. European Journal of Surgical Oncology. 2013;39(7):760–5.

12.

Cui

, Jenssen

, Saftoiu

, Ignee

, Dietrich

. New ultrasound techniques for lymph node evaluation. World J Gastroenterol. 2013;19:4850–60.

13.

Rubaltelli

, Khadivi

, Tregnaghi

, Stramare

, Ferro

, Borsato

, Fiocco

, Adami

, Rossi

. Evaluation of lymph node perfusion using continuous mode harmonic ultrasonography with a second- generation contrast agent. J Ultrasound Med. 2004;23:829–836.

14.

Rubaltelli

, Corradin

, Dorigo

, Tregnaghi

, Adami

, Rossi

, Stramare

. Automated quantitative evaluation of lymph node perfusion on contrast-enhanced sonography. AJR Am J Roentgenol. 2007;188:977–83.

15.

Slaisova

, Benda

, Jarkovsky

, Petrasova

, Szturz

, Valek

. Contrast-enhanced ultrasonography compared to gray-scale and power Doppler in the diagnosis of peripheral lymphadenopathy. Eur J Radiol Radiol. 2013;82:693–8.

16.

Oikawa

, Shiga

, Saito

, et al. Association between contrast-enhanced ultrasonography and histopathological findings of the metastatic lymph nodes of patients with head and neck cancer: A preliminary study. [J]. Oncology Letters. 2018;15(15):4171–6.

17.

Hong

, Luo

, Mo

, Wang

, Ye

, Huang

. Role of contrast-enhanced ultrasound in the preoperative diagnosis of cervical lymph node metastasis in patients with papillary thyroid carcinoma. Ultrasound Med Biol. 2017;43:2567–75.

18.

Jiang

, Wei

, Zhang

, Zhuo

. Value of contrast-enhanced ultrasound combined with elastography in evaluating cervical lymph node metastasis in papillary thyroid carcinoma. World J Clin Cases. 2019;7:49–57.

19.

Lei

, Luzeng

, Jinghua

, Bin

, Hui

. Value of qulitative and quantitative contrast-enhanced ultrasound analysis in preoperative diagnosis of cervical lymph node metastasis from papillary thyroid carcinoma. J Ultrasound Med. 2019;9999:1–9.

20.

Zhan

, Diao

, Chen

, Wang

, Ding

. Predicting cervical lymph node metastasis in patients with papillary thyroid cancer(PTC): why contrast-enhanced ultrasound (CEUS) was performed before thyroidectomy. Clin Hemorheol Microcirc. 2019;72(1):61–73.

21.

Xiang

, Hong

, Zhang

, Huang

, Li

, Wang

, Li

. Contrast- enhanced ultrasound (CEUS) facilitated US in detecting lateral neck lymph node metastasis of thyroid cancer patients: Diagnosis value and enhancement patterns of malignant lymph nodes. Eur Radiol. 2014;24:2513–9.

22.

Dudau

, Hameed

, Gibson

, et al. Can contrast-enhanced ultrasound distinguish malignant from reactive lymph nodes in patients with head and neck cancers? Ultrasound Med Biol. 2014;40:747–54.

23.

Poanta

, Serban

, Pascu

, Pop

, Cosgarea

, Fodor

. The place of CEUS in distinguishing benign from malignant cervical lymph nodes: a prospective study. Med Ultrason. 2014;16:7–14.

24.

Zaheer

, Tan

, Ang

, et al. Post-thyroidectomy neck ultrasonography in patients with thyroid cancer and a review of the literature. Singapore Med J. 2014;55:177–82.

25.

Kim

, Choo

, Lee

, Jung

, Eom

, Ha

. Sonographic features of cervical lymph nodes after thyroidectomy for papillary thyroid carcinoma. J Ultrasound Med. 2013;32:1173–80.

26.

Noguchi

, Noguchi

, Murakami

. Papillary carcinoma of the thyroid: I. Developing pattern of metastasis. Cancer. 1970;26:1053–60.

27.

Sherman

. Thyroid carcinoma. Lancet. 2003;361:501–11.

28.

Ito

, Higashiyama

, Takamura

, Kobayashi

, Miya

, Miyauchi

. Prognosis ofpatients with papillary thyroid carcinoma showing postoperative recurrence to the central neck. World J Surg 35:767–72.

29.

Shi

, Jin

, Zheng

. Partially cystic thyroid nodules on ultrasound: The associated factors for malignancy. Clin Hemorheol Microcirc. 2020;74(4):373–81.

30.

Roh

, Kim

, Park

. Central cervical nodal metastasis from papillary thyroid microcarcinoma: Pattern and factors predictive of nodal metastasis. Ann Surg Oncol. 2008;15:2482–6.

31.

Brannigan

, Burns

, Wilson

. Blood flow patterns in focal liver lesions at microbubble-enhanced US. Radiographics. 2004;24:921–35.

32.

Nishida

, Yano

, Nishida

, Kamura

, Kojiro

. >Angiogenesis in cancer. Vasc Health Risk Manag. 2006;2:213–19.

33.

Tanigawa

, Matsumura

, Amaya

, et al. Tumor vascularity correlates with the prognosis of patients with esophageal squamous cell carcinoma. Am Cancer Soc. 1997;79:220–5.

34.

Schroeder

, Bostanjoglo

, Rademaker

, Maeurer

, Felix

. Role of power Doppler techniques and ultrasound contrast enhancement in the differential diagnosis of focal breast lesions. Eur Radiol. 2003;13:68–79.

35.

Nakopoulou

, Stefanaki

, Panayotopoulou

, et al. Expression of the vascular endothelial growth factor receptor 2/Flk-1 in breast carcinomas:correlation with proliferation. Hum, Pathol. 2002;33(9):863–70.

36.

Hong

, Liu

, Zhang

, Mo

, Xu

. Contrast-enhanced ultrasound quantitative analysis in diagnosing superficial lymphadenopathy: preliminary study. Zhongguo Chao Sheng Yi Xue Za Zhi (Chin J Ultrasound Med). 2007;23:212–4.

37.

Rubaltelli

, Beltrame

, Tregnaghi

, Scagliori

, Frigo

, Stramare

. Contrast-enhanced ultrasound for characterizing lymph nodes with focal cortical thickening in patients with cutaneous melanoma. AJR Am J Roentgenol. 2010;196:W8–W12.

38.

, Mori

, Kodama

, Sakamoto

, Takahashi

, Kodama

. Enhanced sonographic imaging to diagnose lymph node metastasis: importance of blood vessel volume and density. Cancer Res. 2013;73:2082–92.

39.

Liu

, Jiang

, Dai

, Zhu

, Wang

, Zhang

, et al. Differentiation of benign and malignant sub-1-cm breast lesions using contrast-enhanced sonography. J Ultrasound Med. 2015;34(1):117–23.

40.

Zhao

, Liu

, Wei

, Xu

, Wu

, Xu

, et al. Contrast-enhanced ultrasonography features of breast malignancies with different sizes: Correlation with prognostic factors. Biomed Res Int. 2015;2015:613831.

41.

Si-Hui

, Chun-Xiao

, Ming-Hua

, Gang

, Xing

, Rong

. Incorporation of contrast-enhanced ultrasound in the differential diagnosis for breast lesions with inconsistent results on mammography and conventional ultrasound. Clin Hemorheol Microcirc. 2020;74(4):463–73.

42.

Sakorafas

, Christodoulou

, Lappas

, Safioleas

. Preoperative detection of cervical lymph node metastases in papillary thyroid cancer: A surgical perspective. Onkologie. 2009;32:762–6.