Early evaluation of treatment response to transarterial chemoembolization in patients with advanced hepatocellular carcinoma: The role of dynamic three-dimensional contrast-enhanced ultrasound

Abstract

BACKGROUND:

Dynamic three-dimensional contrast-enhanced ultrasound (3D-CEUS) with quantitative analysis is available in recent years. It can reduce the quantitative sampling error caused by the inconsistency of different sections in order to evaluate local treatment response of hepatocellular carcinoma (HCC) accurately.

OBJECTIVE:

To investigate the value of dynamic 3D-CEUS in evaluating the early response to transarterial chemoembolization (TACE) treatment in patients with advanced HCC lesions.

METHODS:

In this prospective study, both two-dimensional (2D) CEUS and dynamic 3D-CEUS were performed on 40 HCC patients who scheduled for TACE at baseline (T0) and 1–3 days (T1) after treatment. Tumor microvascular perfusion changes were assessed by CEUS time-intensity curve (TIC) and quantitative parameters. According to contrast-enhanced computed tomography (CT) and magnetic resonance (MR) imaging 1 month after treatment results, patients were divided into responders and non-responders groups. The changes of perfusion parameters of both 2D-CEUS and 3D-CEUS were compared between responders and non-responders groups before and after TACE treatment.

RESULTS:

Before and after TACE treatment, no significant difference in maximum diameter of HCC lesions between the two groups could be found. There were more significant differences and ratios of perfusion parameters in 3D-CEUS quantitative analysis than in 2D-CEUS. The mutual significant differences and ratios of 2D-CEUS and 3D-CEUS included peak intensity (PI) difference, PI ratio, ratio of area under the curve (A), ratio of area under the wash-out part (AWO) and slope (S) difference. The former 4 corresponding parameters were better on 3D-CEUS than on 2D-CEUS.

CONCLUSION:

Dynamic 3D-CEUS can be used as a potential imaging method to evaluate early treatment response to TACE in advanced HCC patients.

Keywords

Contrast-enhanced ultrasound advanced hepatocellular carcinoma three-dimensional treatment response

Abbreviations

CEUS

contrast-enhanced ultrasound

3D-CEUS

three-dimensional contrast-enhanced ultrasound

two-dimensional

TIC

time-intensity curve

TACE

transarterial chemoembolization

HCC

hepatocellular carcinoma

computed tomography

magnetic resonance

peak intensity

time to peak intensity

MTT

mean transit time

slope

area under the curve

AWI

area under the wash-in part

AWO

area under the wash-out part

ROI

region of interest

VOI

volume of interest

QOF

quality of fitting

complete response

partial response

stable disease

progression disease

EFSUMB

European Federation of Societies for Ultrasound in Medicine and Biology

CHI

contrast harmonic image

BCLC

Barcelona Clinic Liver Cancer

HBV

hepatitis B virus

arbitrary unit

standard deviation

ROC

receiver operating characteristic

1 Introduction

Hepatocellular carcinoma (HCC) is nowadays the fourth most common cancer, and the third leading cause of cancer-related death in China [1, 2]. The treatment options of HCC depend on both its stage of the disease and underlying liver function of the patients [3, 4]. Unfortunately, the majority of HCC patients lost the optimal chance for curative surgery or local ablative procedures due to their advanced tumor stage or other comorbidities [5 –7]. For patients with intermediate stage who have not any obvious symptoms, but have large multifocal tumors without vascular invasion or metastasis beyond the liver, transarterial chemoembolization (TACE) is the preferred therapy based on the Barcelona Clinic Liver Cancer (BCLC) staging system [8, 9], as it has been proven to prolong the overall survival to “best supportive care” [10]. Since the rate of complete response (CR) or partial response (PR) is about 15–75% of HCC after TACE treatment [11, 12], further treatment is required for residual tumor of incomplete response to TACE.

An early assessment of treatment response is so important and necessary that it could potentially improve the management of HCC [13, 14]. Currently, the main imaging follow up criteria in evaluation of the therapeutic response are changes in tumor size and intratumoral vascularity on dynamic contrast material enhanced CT and MR images [15 –17], which are routinely regarded as the reference standard in the evaluation of tumor response [18]. Since perfusion changes might begin earlier than size changes after treatment, it is more crucial to make an early and accurate evaluation of perfusion changes of tumor after treatment in clinical decision making process [19]. However, their applications in clinical assessment of tumor response is limited by accessibility of CT and MR imaging equipment, complexity of procedure, high cost, risk of allergies, contradictions of renal insufficiency and metal implants. According to current guidelines [20, 21], CEUS has gained a high acceptance in the diagnosis, treatment and follow-up of advanced HCC for its repeatability, no nephrotoxicity and lack of ionizing radiation [22, 23]. Contrast-enhanced ultrasound (CEUS) is suggested as the first line of imaging technique to evaluate the initial response of HCC 1 month after TACE [24], and reference imaging (contrast enhanced CT and MR imaging) can be reserved for follow-up at 3 months as response verification [25]. There is some evidence to suggest that CEUS can be applied as an early indicator for TACE efficacy as early as 2 or more days after treatment [14]. The combination of CEUS and quantitative technology can provide a semi-quantification of altered tumoral perfusion of local treatment [26 –29].

However, two-dimensional (2D) CEUS is prone to 2D sampling errors due to single plane variation during the follow-up of local therapies for liver tumors [30]. Dynamic three-dimensional contrast-enhanced ultrasound (3D-CEUS) is available to take into account the hemodynamic changes of the whole tumor lesion in order to compensate the single-plane limitation of 2D-CEUS [31 –33]. It is now implemental to evaluate the changes quantitatively in tumor perfusion by analyzing raw data of dynamic 3D-CEUS. The advent of quantified 3D-CEUS has shown promise in accurate assessment of tumor vasculature [12, 34].

The aim of this study was to investigate the feasibility of dynamic 3D-CEUS in evaluating the early treatment response to TACE of HCC lesions with analysis of CEUS perfusion parameters.

2 Patients and methods

2.1 Patients

Over the periods of July 2016 to January 2018, 46 patients scheduled for TACE for the treatment of advanced HCC were prospectively enrolled in the study.

Inclusion criteria comprised: Child-Pugh grade A/B liver cirrhosis; single or multiple lesions detected by 2 typical imaging modalities; patients without indication for surgical resection or ablation treatment; and informed consent before TACE.

Exclusion criteria were: Patients with contraindications according to the guidelines of the European Federation of Societies for Ultrasound in Medicine and Biology (EFSUMB) [35]; patients with inability to tolerate contrast-enhanced CT, MR or dynamic 3D-CEUS examination; and receiving other therapies between two examinations.

The study was approved by the institutional review board of our hospital. Before enrollment, written informed consents were obtained from all the patients.

2.2 Dynamic contrast-enhanced ultrasound imaging

All patients underwent both 2D-CEUS and dynamic 3D-CEUS examinations prior to (within 24 hours, T0) and post (between 24–72 hours, T1) TACE. An experienced radiologist (W.WP. with more than 15 y experience of performing 2D-CEUS of liver and 10 y experience of performing 3D-CEUS), who were aware of the patients’ clinical histories, performed B-mode ultrasound, 2D-CEUS and dynamic 3D-CEUS scanning. The ultrasound machine we used was Aplio500 unit (Toshiba Ultrasound, Japan), with a 3.5 MHz-central-frequency PVT-375BT convex array probe. During the ultrasound scanning process, the patients were examined lying in the supine position or left lateral position. The target lesion was determined by B-mode ultrasound and 2D-CEUS. If there existed multiple lesions, the largest or most clearly defined HCC lesion was selected as the target lesion of subsequent dynamic 3D-CEUS scanning. The 2D-CEUS imaging was performed using a dual-imaging mode, which enabled side by side visualization of both B-mode ultrasound and CEUS. The ultrasound contrast agent used in the present study was sulfur hexafluoride (SonoVue, Bracco, Milan, Italy), which was injected intravenously as a 2.4 mL bolus followed by 5 mL of normal sterile saline flush via a 20-gauge peripheral intravenous cannula for 2D-CEUS. The patients were instructed to fast for at least 4 hours before and keep gentle and regular breath during both 2D-CEUS and 3D-CEUS scanning.

The contrast harmonic image (CHI) mode was applied as 3D-CEUS mode, which is regarded as the optimal imaging mode for having a strong non-linear harmonic response from the microbubbles. In the same session, with an interval time of at least 10 min to allow for contrast clearance of the previous contrast injection, the same radiologist performed 3D-CEUS by means of the same ultrasound unit provided with a volumetric mechanical 3.5 MHz PVT-382 MV mechanical 3D probe (Toshiba, Japan) at the following preset: dynamic range = 65, mechanical index = 0.10-0.11, gain = 76, and acoustic power = 1% in order to reduce the difference of manipulations of different operators. A dose of 2.4 mL SonoVue suspension was injected in bolus through peripheral vein for 3D-CEUS. The frame and volume rate for 2D-CEUS and 3D-CEUS were 10 frames/sec and 1.5 volumes/sec, respectively. The scanning mode was switched to “CHI” and “4D” modes successively when the target lesion was displayed definitely and clearly during 3D-CEUS. In the examination after the treatment, the video of the latest examination was reviewed to keep the position and orientation of the probe and cutting plane in accordance with those before TACE. Three orthogonal planes of 2D-CEUS and a 3D-CEUS volume image can be observed simultaneously in 4 quadrants of the same view window during 3D-CEUS acquisition. All data were stored in the format of raw data for further analysis.

2.3 Quantification of both 2D-CEUS and dynamic 3D-CEUS

All data were stored and analyzed in raw data format using online analysis software package (CHI-Q, Version 3.7, Toshiba, Japan). The time-intensity curve (TIC) analysis and perfusion parameters estimations were performed by an independent author (C. JY) who was not involved in TACE and CEUS examinations. The whole post-processing process takes about 10–20 min.

The region of interest (ROI, 2D-CEUS) and volume of interest (VOI, 3D-CEUS) should be outlined to envelop the whole tumor lesion. The location, size, and shape of ROI/VOI could be manually adjusted during or after the TIC generation process in order that the fitting curve and the original curve would be as close as possible. Quality of fitting (QOF) is required to evaluate the fitting degree of the fitting curve and the original curve. The TIC parameters would be invalid if the value of QOF was no more than 0.75. We used the same 7 perfusion parameters, which can be extracted from the smooth fitting curve, for both 2D-CEUS and 3D-CEUS as follows: peak intensity (PI) represents the highest intensity value attained by the TIC for the defined ROI/VOI; time to peak intensity (TP), is the time required for the contrast agent from arriving in ROI/VOI to reaching PI; mean transit time (MTT) is the time interval during which the intensity value was higher than PI/2; slope (S) is literally defined as the slope of the tangent of the wash-in at half maximum. Finally, A, AWI and AWO correspond to the area under the curve, the area under the wash-in (from injection time to the time point of PI) part and the area under the wash-out (from the time point of PI to the end of the acquisition) part of the TIC. The relative displacements induced by respiration can be corrected and compensated by the built-in motion tracking system and manual adjustment in axial and sagittal planes. Besides, on 3D-CEUS, the contrast stereoscopic image of the target lesion could be rotated, and reversed. A series of equidistant planes of the lesion could be observed by tomographic ultrasound imaging mode simultaneously from various directions, with a thinnest interval distance of 0.2 mm to acquire more details of residual blood supply.

2.4 Assessment of response to TACE

All TACE procedures were performed by the same experienced interventional radiologist (Chen RX). Both 2D-CEUS and 3D-CEUS perfusion parameters of the target lesion were analyzed through the comparison of the difference between perfusion parameters of T0 and T1 (ΔX = X_T0–X_T1) and the ratio of the 2 perfusion parameters (ratio of X = X_T0/X_T1). The TACE response was evaluated based on the contrast-enhanced CT and MR 1 month after TACE which was assessed by 2 radiologists blinded to the results from the CEUS exams by consensus. The contrast-enhanced CT and MR evaluations were performed on multiple cross-sectional images of subjects. CR was defined as no enhancement of the lesion. Incomplete treatment response were classified as: (1) PR: the diameter of the enhancement part of the target lesion (peak enhancement) decreased by more than 30%; (2) progression disease (PD): the diameter of the enhancement part of the target lesion (peak enhancement) increased by more than 20%; and (3) stable disease (SD): the degree of reduction did not reach PR and increase degree did not reach PD. The responders met mRECIST for CR/PR/SD. Non-responders were defined as patients with PD. The 2D-CEUS and 3D-CEUS perfusion parameters were compared between the responders and the non-responders.

2.5 Statistical analysis

A detailed lesion by lesion analysis was performed. All statistical analyses for the present study were performed with R (version 3.5.1). A two-tailed p-value < 0.05 indicated statistical significance. Analyzing the continuous quantification parameters was done by t-test for the normal-distributed data and Mann-Whitney U test for the abnormal-distributed data. The normal distribution features were given as mean±standard deviation (SD), whereas abnormal distributions were given as median (lower quartile, upper quartile). Diagnostic performance was evaluated through the receiver operating characteristic (ROC) curve analysis. Delong test was used to compare the same significant parameters between 2D-CEUS and dynamic 3D-CEUS.

3 Results

3.1 Baseline characteristics of patients and lesions

Six patients were excluded: 4 patients postponed TACE for their advanced stage of liver cirrhosis (Child-Pugh C) and concomitant severe liver impairment; 2 patients received systemic therapies combined with TACE. Finally, 40 patients with 40 unresectable advanced HCC lesions were enrolled in this study.

The patients ranged in age from 41 to 73 years (mean, 58.6±8.1 years). The underlying liver disease was: hepatitis B infection (n = 36), hepatitis C infection (n = 2), ethanol induced cirrhosis (n = 1) and cryptogenic cirrhosis (n = 1). Twenty-four patients had solitary HCC lesions and 16 had multiple lesions. The patients’ basic information and tumor size were summarized in Table 1, and there was no significant difference of these items between responders and non-responders groups.

Table 1
Baseline characteristics of 40 HCC patients before TACE treatment

Characteristic Responders Non-responders t/χ² value P value

No. of patients 26 14

Age (y) (average, range) 57.5 (41–71) 60.7 (41–73) 1.215 0.232

Sex (male/female) 22/4 11/3 0.225 0.635

HBV infection (+/-) 24/2 12/2 0.422 0.516

Child-Pugh class (A/B) 22/4 13/1 0.612 0.434

BCLC grade (A/B/C) 5/12/9 2/9/3 1.242 0.537

Maximum diameter (cm) 8.1 (2.4–15.0) 7.7 (2.5–16.0) 0.259 0.797

(average, range)

Characteristic	Responders	Non-responders	t/χ² value	P value
No. of patients	26	14
Age (y) (average, range)	57.5 (41–71)	60.7 (41–73)	1.215	0.232
Sex (male/female)	22/4	11/3	0.225	0.635
HBV infection (+/-)	24/2	12/2	0.422	0.516
Child-Pugh class (A/B)	22/4	13/1	0.612	0.434
BCLC grade (A/B/C)	5/12/9	2/9/3	1.242	0.537
Maximum diameter (cm)	8.1 (2.4–15.0)	7.7 (2.5–16.0)	0.259	0.797
(average, range)

HCC: hepatocellular carcinoma; TACE: transarterial chemoembolization; HBV: hepatitis B virus; BCLC: Barcelona Clinic Liver Cancer.

The size of target HCC lesion ranged from 2.4 to 16.0 cm (mean, 7.9±4.1 cm). Thirty-one lesions were located in the right lobe, 7 in the left lobe and 2 in the caudate lobe. Among the 40 lesions, 26 lesions were classified as responders, including 5 (12.5%) CR lesions, 12 (30.0%) PR lesions, and 9 (22.5%) SD lesions. And 14 (35.0%) PD lesions were classified as non-responders according to CT and MR imaging result.

Overall, all target lesions were observed as hypoechoic on 2D ultrasound at T0. As there were 3 lesions showing artifacts on both 2D ultrasound and CEUS post TACE treatment, they were observed as hyperechoic. And the rest 37 lesions were hypoechoic at T1. Because of the relative short imaging interval, the diameter of the tumor (Diameter_T0 = 79.5±41.0 mm; Diameter_T1 = 79.9±40.8 mm) did not change obviously (t = 0.995; P = 0.326), as shown in Fig. 1.

Fig. 1

A 72-year-old man with multiple HCC. The target lesion was enhanced inhomogeneously with internal unenhanced area both on T0 (A: 18 s after injection; B: 24 s after injection) and T1 (C: 21 s after injection; D: 29 s after injection). The first 4 screenshots were shown in dual imaging pattern (left: a 2D-CEUS plane of the three orthogonal planes; right: 3D-CEUS) synchronously. The diameter of the tumor did not change obviously on T1. Dynamic 3D-CEUS perfusion analysis (E: TIC of T0; F: TIC of T1) showed that PI value of T1 decreased obviously, which was less than 1/2 of the value of T0 (the purple curve in E and F).

3.2 Comparison of quantification parameters of both 2D-CEUS and dynamic 3D-CEUS

The valid adoption rate of quantification of both 2D-CEUS and dynamic 3D-CEUS was achieved in 40 lesions (100%, 40/40) at T0 and 35 lesions (87.5%, 35/40) at T1, as TIC of CR lesions post TACE could not be fitted without obvious ascending or descending branches. As a result, the main analysis was based on perfusion parameters of 35 lesions of PR, SD and PD. The QOF of TIC were 86.2±6.5% (range, 81.3% –95.8%) on T0 and 89.4±4.9% (range, 84.1% –95.7%) on T1.

As nearly half of the HCC lesions were about 10 cm in diameter in our study, we did not apply the surrounding liver tissue region as reference of TIC parameters. The 7 quantification parameters of target lesions were demonstrated in Table 2 (2D-CEUS) and Table 3 (3D-CEUS) and expressed as median (lower quantile, upper quantile) for abnormal distributions and mean±SD for normal distribution. The example of 3D-CEUS TIC was shown in Fig. 1 (E) and (F). At T0, there were not any statistically significant differences (P > 0.05) in all the TIC parameters between the responders and non-responders on either 2D-CEUS or 3D-CEUS. At T1, a statistically significant difference of PI, S, A, AWI and AWO on both 2D-CEUS and 3D-CEUS could be documented between the responders and the non-responders, while there was not any significant difference of the time parameters. Then, the difference (Δ) and ratio of the TIC parameters were further compared in Table 4 and Table 5. As the S values of 22 HCC lesions (55.0%) on 3D-CEUS and 3 lesions (7.5%) on 2D-CEUS were “0” at T1, the ratios of S were not calculated and compared between before and after treatment. This is the reason why the values of S ratio are not shown in the Table 4 and Table 5. It was found that there were more significant differences and ratios of 2D-CEUS and 3D-CEUS parameters in 3D-CEUS quantitative analysis than in 2D-CEUS. The significant differences and ratios of 2D-CEUS and 3D-CEUS parameters included PI difference, PI ratio, S difference, A ratio and AWO ratio. Area under the ROC curves for significant ratios and differences of dynamic 3D-CEUS perfusion parameters were higher than those for the corresponding parameters of 2D-CEUS (as shown in Fig. 2). Delong test was further applied to compare the 5 corresponding parameters of 2D-CEUS and 3D-CEUS (Table 6). Except for the comparison of S difference between 2D-CEUS and 3D-CEUS, the other 4 corresponding parameters were better on 3D-CEUS than on 2D-CEUS. It could be observed that quantitative parameters of 3D-CEUS can better reflect the response of local treatment of focal liver tumors than 2D-CEUS.

Table 2
Comparison of TIC parameters before and after TACE treatment of HCCs on 2D-CEUS

Variable PR + SD PD Statistics P-value

T0

PI_T0 (AU) 24.35 (9.30, 40.00) 22.80 (16.52, 60.84) –1.013 0.311

TP_T0 (s) 13.85 (9.14, 17.98) 12.50 (9.42, 24.66) –0.124 0.901

MTT_T0 (s) 35.10 (26.66, 53.17) 65.50 (34.04, 160.96) –1.475 0.140

S_T0 (AU/s) 2.15 (0.60, 5.36) 4.00 (1.26, 11.36) –1.208 0.227

A_T0 (AU·s) 919.95 (557.20, 2230.15) 1808.30 (1001.58, 6231.36) –1.581 0.114

AWI_T0 (AU·s) 147.70 (86.87, 287.90) 256.00 (141.08, 1041.50) –1.297 0.195

AWO_T0 (AU·s) 822.35 (466.42, 1954.55) 1625.30 (858.04, 5198.98) –1.546 0.122

T1

PI_T1 (AU) 6.95 (2.99, 39.16) 110.60 (20.76, 247.76) –2.772 0.006^*

TP_T1 (s) 13.10 (8.14, 20.86) 12.10 (7.36, 23.44) 0.107 0.915

MTT_T1 (s) 46.10 (27.59, 55.27) 33.40 (22.36, 148.50) 0.302 0.763

S_T1 (AU/s) 0.80 (0.24, 2.56) 8.80 (2.52, 40.12) –3.127 0.002^*

A_T1 (AU·s) 474.40 (195.77, 1626.02) 2311.70 (1068.80, 30892.86) –2.665 0.008^*

AWI_T1 (AU·s) 46.80 (25.54, 186.59) 333.70 (113.32, 4786.74) –2.381 0.017^*

AWO_T1 (AU·s) 405.85 (169.54, 1223.61) 1879.30 (902.34, 26106.12) –2.736 0.006^*

Variable	PR + SD	PD	Statistics	P-value
T0
PI_T0 (AU)	24.35 (9.30, 40.00)	22.80 (16.52, 60.84)	–1.013	0.311
TP_T0 (s)	13.85 (9.14, 17.98)	12.50 (9.42, 24.66)	–0.124	0.901
MTT_T0 (s)	35.10 (26.66, 53.17)	65.50 (34.04, 160.96)	–1.475	0.140
S_T0 (AU/s)	2.15 (0.60, 5.36)	4.00 (1.26, 11.36)	–1.208	0.227
A_T0 (AU·s)	919.95 (557.20, 2230.15)	1808.30 (1001.58, 6231.36)	–1.581	0.114
AWI_T0 (AU·s)	147.70 (86.87, 287.90)	256.00 (141.08, 1041.50)	–1.297	0.195
AWO_T0 (AU·s)	822.35 (466.42, 1954.55)	1625.30 (858.04, 5198.98)	–1.546	0.122
T1
PI_T1 (AU)	6.95 (2.99, 39.16)	110.60 (20.76, 247.76)	–2.772	0.006^*
TP_T1 (s)	13.10 (8.14, 20.86)	12.10 (7.36, 23.44)	0.107	0.915
MTT_T1 (s)	46.10 (27.59, 55.27)	33.40 (22.36, 148.50)	0.302	0.763
S_T1 (AU/s)	0.80 (0.24, 2.56)	8.80 (2.52, 40.12)	–3.127	0.002^*
A_T1 (AU·s)	474.40 (195.77, 1626.02)	2311.70 (1068.80, 30892.86)	–2.665	0.008^*
AWI_T1 (AU·s)	46.80 (25.54, 186.59)	333.70 (113.32, 4786.74)	–2.381	0.017^*
AWO_T1 (AU·s)	405.85 (169.54, 1223.61)	1879.30 (902.34, 26106.12)	–2.736	0.006^*

^*Indicates the difference is statistically significant. TIC: time-intensity curve; TACE: transarterial chemoembolization; HCC: hepatocellular carcinoma; 2D-CEUS: two-dimensional contrast-enhanced ultrasound; PR: partial response; SD: stable disease; PD: progression disease; PI: peak intensity; TP: time to peak intensity; MTT: mean transit time; S: slope; A: area under the curve; AWI: area under the curve of wash-in part; AWO: area under the curve of wash-out part; AU: arbitrary unit.

Table 3

Comparison of TIC parameters before and after TACE treatment of HCCs on 3D-CEUS

Variable	PR + SD	PD	Statistics	P-value
T0
PI_T0 (AU)	1.00 (0.45, 1.51)	0.60 (0.34, 1.18)	1.030	0.303
TP_T0 (s)	13.35 (8.48, 16.61)	12.20 (7.54, 21.86)	–0.195	0.845
MTT_T0 (s)	39.72±18.24	42.86±16.56	–0.487	0.629
S_T0 (AU/s)	0.10 (0.05, 0.20)	0.10 (0.00, 0.10)	0.657	0.511
A_T0 (AU·s)	43.20 (19.08, 96.09)	34.20 (22.12, 83.54)	0.320	0.749
AWI_T0 (AU·s)	6.75 (3.00, 11.72)	5.60 (3.32, 15.16)	0.302	0.763
AWO_T0 (AU·s)	36.45 (16.80, 86.89)	28.80 (16.94, 68.40)	0.515	0.606
T1
PI_T1 (AU)	0.20 (0.10, 0.66)	1.30 (0.52, 2.08)	–3.109	0.002^*
TP_T1 (s)	15.85 (8.03, 25.57)	10.40 (7.38, 12.94)	1.546	0.122
MTT_T1 (s)	42.00 (30.62, 63.53)	35.70 (22.20, 44.34)	0.782	0.434
S_T1 (AU/s)	0.00 (0.00, 0.10)	0.10 (0.10, 0.46)	–3.091	0.002^*
A_T1 (AU·s)	13.90 (8.68, 27.60)	65.50 (29.90, 111.02)	–3.127	0.002^*
AWI_T1 (AU·s)	2.70 (1.44, 4.41)	5.80 (2.88, 20.44)	–2.239	0.025^*
AWO_T1 (AU·s)	10.65 (5.94, 23.01)	59.80 (27.02, 97.98)	–3.162	0.002^*

^*Indicates the difference is statistically significant. TIC: time-intensity curve; TACE: transarterial chemoembolization; HCC: hepatocellular carcinoma; 3D-CEUS: three-dimensional contrast-enhanced ultrasound; PR: partial response; SD: stable disease; PD: progression disease; PI: peak intensity; TP: time to peak intensity; MTT: mean transit time; S: slope; A: area under the curve; AWI: area under the curve of wash-in part; AWO: area under the curve of wash-out part; AU: arbitrary unit.

Table 4

Comparison of the difference and ratio of TIC parameters of 2D-CEUS

Variable	PR + SD	PD	Statistics	P-value
ΔPI	8.60 (0.20, 21.08)	–86.00 (–222.52, 7.00)	2.416	0.016^*
PI ratio	2.62 (1.35, 5.56)	0.41 (0.12, 1.34)	2.203	0.028^*
ΔTP	1.55 (–4.72, 8.17)	0.10 (–2.98, 4.56)	0.178	0.859
ΔMTT	–6.05 (–23.04, 17.37)	7.90 (–5.90, 31.68)	–1.492	0.136
ΔS	0.65 (–0.05, 3.27)	–5.30 (–29.18, 1.36)	2.203	0.028^*
ΔA	435.70 (71.11, 1460.16)	–559.20 (–25024.44, 393.50)	1.954	0.051
A ratio	2.84 (1.30, 4.32)	0.41 (0.17, 2.72)	2.061	0.039^*
ΔAWI	62.85 (–2.66, 106.35)	–176.50 (–3745.24, 84.64)	1.741	0.082
AWI ratio	2.45 (0.82, 5.71)	0.39 (0.19, 1.83)	1.883	0.060
ΔAWO	376.20 (43.69, 1216.26)	–471.90 (–21279.20, 371.48)	1.848	0.065
AWO ratio	2.62 (1.25, 4.76)	0.42 (0.17, 3.01)	2.096	0.036^*

Table 5

Comparison of the difference and ratio of TIC parameters of 3D-CEUS

Variable	PR + SD	PD	Statistics	P-value
ΔPI	0.50 (0.15, 0.86)	–0.10 (–1.14, 0.00)	3.891	< 0.001^*
PI ratio	3.17 (1.91, 5.00)	0.83 (0.33, 1.00)	4.246	< 0.001^*
ΔTP	–0.80 (–11.86, 5.07)	2.50 (–2.30, 8.02)	–1.35	0.177
ΔMTT	–3.50 (–19.90, 10.23)	8.50 (–6.98, 15.70)	–1.19	0.234
ΔS	0.10 (0.00, 0.16)	–0.10 (–0.34, 0.00)	3.269	0.001^*
ΔA	27.20 (5.31, 73.37)	–5.90 (–17.46, 5.62)	3.411	0.001^*
A ratio	2.32 (1.66, 6.03)	0.89 (0.48, 1.16)	3.909	< 0.001^*
ΔAWI	4.50 (0.48, 8.33)	0.20 (–3.54, 3.26)	2.381	0.017^*
AWI ratio	2.68 (1.37, 3.97)	1.01 (0.41, 1.67)	2.665	0.008^*
ΔAWO	22.35 (5.77, 66.42)	–6.10 (–17.88, 3.34)	3.447	0.001^*
AWO ratio	2.67 (1.80, 5.73)	0.91 (0.47, 1.12)	3.944	< 0.001^*

Fig. 2

ROC analyses for the diagnostic performance of significant differences and ratios of perfusion parameters of two-dimensional contrast-enhanced ultrasound (2D-CEUS) (A) and dynamic three-dimensional contrast-enhanced ultrasound (3D-CEUS) (B).

Table 6

The comparison of the same significantly parameters of 2D-CEUS and dynamic 3D-CEUS

Parameters	P-value
Ratio of PI (PI_T0/PI_T1)	0.022^*
Difference of PI (PI_T0 - PI_T1)	0.030^*
Difference of S (S_T0 - S_T1)	0.210
Ratio of A (A_T0/A_T1)	0.047^*
Ratio of AWO (AWO_T0/AWO_T1)	0.049^*

^*Indicates the difference is statistically significant. 2D-CEUS: two-dimensional contrast-enhanced ultrasound; 3D-CEUS: three-dimensional contrast-enhanced ultrasound; PI: peak intensity; S: slope; A: area under the curve; AWO: area under the curve of wash-out part.

4 Discussion

Previously, tumor perfusion curves were derived from raw data of 2D-CEUS and proven to be useful in evaluation of HCC response of TACE [13 , 36]. Since tumor is a stereo structure inherently, the enhanced appearance on a single CEUS plane may not provide a representative and objective evaluation of the whole lesion and will increase variability. Comparing to 2D-CEUS, 3D-CEUS might have the unique advantage of evaluation of the whole tumor without single plane variation. Recent development has made possible the combination of 3D-CEUS and quantification, which would help detect vascular changes at the capillary level and improve the current monitoring standards using the gross tumor volume longitudinally. Previous studies have found that quantification of 3D-CEUS did have a superior performance in evaluation of change of blood supply in several animal studies [33 , 38], but there are rare further reports about its applications in clinical practice.

In our study, we assessed the convenience and feasibility of dynamic 3D-CEUS for the quantification of tumor perfusion parameters after treatment of TACE in patients with advanced HCC. Moreover, since it is impossible to get the confirmation of the efficacy of TACE from histopathological result, valid evaluation by timely imaging post TACE is vital [4]. Peritumoral inflammatory response could not be avoid during response evaluation, but evaluation within 1 week still help to improve the management of HCC potentially [13].

The results of the comparison of tumor size before and after TACE treatment showed that there is no difference in tumor size for the short interval of the evaluation. Furthermore, an assessment of the different perfusion parameters was performed using cine raw data of both 2D-CEUS and dynamic 3D-CEUS acquired at the time points of before and 1–3 days after TACE treatment. At 1–3 days after TACE, all the intensity related parameters decreased significantly in the responders group, while increased or unobviously decreased in the non-responders group on both 2D-CEUS and 3D-CEUS. Therefore, the intensity related parameters are indicated to be the most significant parameters in comparison between responder and non-responder groups on both 2D-CEUS and dynamic 3D-CEUS, and might be potential CEUS predictors of TACE therapeutic response. Meanwhile, there wasn’t any statistically significant difference of the time-related parameters (TP and MTT) between the two groups. The study by Nam et al. showed that PI of 3D-CEUS was lower in the complete treatment group than in the incomplete treatment group on 1-2 weeks and 1 month after TACE, while PI of 2D-CEUS was lower only on 1 month after treatment [12], which was similar to our result. The results of Delong test showed that the change of PI of 3D-CEUS was more obvious than that of 2D-CEUS in the TACE response evaluation. In our study, we studied not only PI, but also other 6 perfusion parameters. After comparing the parameters before and after TACE treatment, the analyses showed that all perfusion parameters related to contrast intensity are significant. The further Delong test showed that the parameters of area under the curve, including A and AWO, which represents the accumulation of contrast intensity, were significantly decreased in the responder group of both 3D-CEUS and 2D-CEUS in current study. We suspected that the change of S and AWI has no significant difference between the two CEUS methods, which may be due to the short TP (mean, 15.78±11.17 s) in our study. Besides, it should be mentioned that the discreteness of each TIC parameter values was relatively large, which may be related to the difference stage of HCC and heterogeneity of TACE response in different patients, as seen from Table 2 and Table 3.

The limitation of our study is that the surrounding liver parenchyma was not taken as reference ROI on both 2D-CEUS and reference VOI on dynamic 3D-CEUS considering that more than half of enrolled lesions were too large and the space around the lesion for parenchyma selection was too narrow. Secondly, we were not able to perform a CT or MR imaging scan at 1–3 days after treatment according to current recommendations. Therefore, there is a time interval between our CEUS evaluation and contrast enhanced CT and MR scans. Thirdly, the enrolled HCC patient population was relatively small. A further study with a larger sample will be necessary. And the technical characteristics of dynamic 3D-CEUS, such as the temporal/axial/lateral resolutions and frame rate, are still lagging behind 2D-CEUS imaging systems.

In conclusion, dynamic 3D-CEUS can be used to make early quantification analysis of quantify dynamic changes in HCC. The changes of contrast intensity in perfusion parameters may be applied as early predictor of TACE response and dynamic 3D-CEUS may be potential surrogate of dynamic 2D-CEUS of tumor perfusion in treatment follow up of HCC lesions.

Footnotes

Acknowledgments

We thank Yaqiong Ge of GE Healthcare for her help in data statistics.

Funding

This study has received funding by Natural Science Foundation Project of Shanghai “Science and Technology Innovation Action Plan” (Grant No.20ZR1452800) and Shanghai Municipal Key Clinical Specialty (Grant No.shslczdzk03501).

Conflicts of interest

No benefits in any form have been received or will be received from a commercial party related directly or indirectly to the subject of this article.

References

Chen

, Zheng

, Baade

, et al. Cancer statistics in China, 2015. CA Cancer J Clin. 2016;66(2):115–32.

Xie

, Ren

, Zhou

, et al. Critical appraisal of Chinese guideline on the management of hepatocellular carcinoma. Hepatobiliary Surg Nutr. 2017;6(6):387–96.

Forner

, Reig

, Bruix

. Hepatocellular carcinoma. Lancet. 2018;391(10127):1301–14.

Sparchez

, Mocan

, Radu

, et al. Contrast enhanced ultrasonography in assessing the treatment response to transarterial chemoembolization in patients with hepatocellular carcinoma. Med Ultrason. 2016;18(1):96–102.

Heimbach

, Kulik

, Finn

, et al. AASLD guidelines for the treatment of hepatocellular carcinoma. Hepatology. 2018;67(1):358–80.

Liver EAFT

. EASL-EORTC clinical practice guidelines: management of hepatocellular carcinoma. J Hepatol. 2012;56(4):908–43.

Liver EAFT

. EASL-EORTC clinical practice guidelines: management of hepatocellular carcinoma. Eur J Cancer. 2012;48(5):599–641.

Bruix

, Sherman

. Management of hepatocellular carcinoma: an update. Hepatology. 2011;53(3):1020–2.

Sciarra

, Ronot

, Di Tommaso

, et al. TRIP: a pathological score for transarterial chemoembolization resistance individualized prediction in hepatocellular carcinoma. Liver Int. 2015;35(11):2466–73.

10.

Llovet

, Real

, Montana

, et al. Arterial embolisation or chemoembolisation versus symptomatic treatment in patients with unresectable hepatocellular carcinoma: a randomised controlled trial. Lancet. 2002;359(9319):1734–9.

11.

Sergio

, Cristofori

, Cardin

, et al. Transcatheter arterial chemoembolization (TACE) in hepatocellular carcinoma (HCC): the role of angiogenesis and invasiveness. Am J Gastroenterol. 2008;103(4):914–21.

12.

Nam

, Stanczak

, Lyshchik

, et al. Evaluation of Hepatocellular Carcinoma Transarterial Chemoembolization using Quantitative Analysis of 2D and 3D Real-time Contrast Enhanced Ultrasound. Biomed Phys Eng Express. 2018;4(3):35039.

13.

Shaw

, Eisenbrey

, Lyshchik

, et al. Contrast-enhanced ultrasound evaluation of residual blood flow to hepatocellular carcinoma after treatment with transarterial chemoembolization using drug-eluting beads: a prospective study. J Ultrasound Med. 2015;34(5):859–67.

14.

Kono

, Lucidarme

, Choi

, et al. Contrast-enhanced ultrasound as a predictor of treatment efficacy within 2 weeks after transarterial chemoembolization of hepatocellular carcinoma. J Vasc Interv Radiol. 2007;18(1 Pt 1):57–65.

15.

Peng

, Chen

, Xiao

, et al. Microvascular Invasion as a Predictor of Response to Treatment with Sorafenib and Transarterial Chemoembolization for Recurrent Intermediate-Stage Hepatocellular Carcinoma. Radiology. 2019;292(1):237–47.

16.

Chaudhry

, McGinty

, Mervak

, et al. The LI-RADS Version MRI Treatment Response Algorithm: Evaluation of Ablated Hepatocellular Carcinoma. Radiology. 2020;294(2):320–6.

17.

Liu

, Smolka

, Papademetris

, et al. Predicting Infiltrative Hepatocellular Carcinoma Patient Outcome Post-TACE: MR Bias Field Correction Effect on 3D-quantitative Image Analysis. J Clin Transl Hepatol. 2020;8(3):292–8.

18.

Eisenhauer

, Therasse

, Bogaerts

, et al. New response evaluation criteria in solid tumours: revised RECIST guideline (version 1.1). Eur J Cancer. 2009;45(2):228–47.

19.

Riaz

, Memon

, Miller

, et al. Role of the EASL, RECIST, and WHO response guidelines alone or in combination for hepatocellular carcinoma: radiologic-pathologic correlation. J Hepatol. 2011;54(4):695–704.

20.

Dietrich

, Nolsøe

, Barr

, et al. Guidelines and Good Clinical Practice Recommendations for Contrast Enhanced Ultrasound (CEUS) in the Liver - Update - WFUMB in Cooperation with EFSUMB, AFSUMB, AIUM, and FLAUS. Ultraschall Med. 2020;41(5):562–85.

21.

Dietrich

, Nolsøe

, Barr

, et al. Guidelines and Good Clinical Practice Recommendations for Contrast-Enhanced Ultrasound (CEUS) in the Liver-Update WFUMB in Cooperation with EFSUMB, AFSUMB, AIUM, and FLAUS. Ultrasound Med Biol. 2020;46(10):2579–604.

22.

Schwarze

, Marschner

, Völckers

, et al. The diagnostic performance of contrast-enhanced ultrasound (CEUS) for evaluating hepatocellular carcinoma (HCC) juxtaposed to MRI findings; a retrospective single-center analysis of 292 patients. Clin Hemorheol Microcirc. 2020;76(2):155–60.

23.

Marschner

, Zhang

, Schwarze

, et al. The diagnostic value of contrast-enhanced ultrasound (CEUS) for assessing hepatocellular carcinoma compared to histopathology; a retrospective single-center analysis of 119 patients. Clin Hemorheol Microcirc. 2020;76(4):453–8.

24.

Cho

, Park

, Choi

, et al. The usefulness of contrast-enhanced ultrasonography in the early detection of hepatocellular carcinoma viability after transarterial chemoembolization: pilot study. Clin Mol Hepatol. 2015;21(2):165–74.

25.

Nicolau

, Vilana

, Bianchi

, et al. Early-stage hepatocellular carcinoma: the high accuracy of real-time contrast-enhanced ultrasonography in the assessment of response to percutaneous treatment. Eur Radiol. 2007;17(Suppl 6):F80–8.

26.

Moschouris

, Malagari

, Papadaki

, et al. Contrast-enhanced ultrasonography of hepatocellular carcinoma after chemoembolisation using drug-eluting beads: a pilot study focused on sustained tumor necrosis. Cardiovasc Intervent Radiol. 2010;33(5):1022–7.

27.

Uller

, Wiggermann

, Gossmann

, et al. Evaluation of the microcirculation of hepatocellular carcinomas using contrast-enhanced ultrasound with intraarterial and intravenous contrast application during transarterial chemoembolization with drug-eluting beads (DEB-TACE): preliminary data. Clin Hemorheol Microcirc. 2011;49(1-4):55–66.

28.

Fellner

, Prantl

, Rennert

, et al. Comparison of time-intensity-curve- (TIC-) analysis of contrast-enhanced ultrasound (CEUS) and dynamic contrast-enhanced (DCE) MRI for postoperative control of microcirculation in free flaps - first results and critical comments. Clin Hemorheol Microcirc. 2011;48(1):187–98.

29.

Schaible

, Stroszczynski

, Beyer

, et al. Quantitative perfusion analysis of hepatocellular carcinoma using dynamic contrast enhanced ultrasound (CEUS) to determine tumor microvascularization. Clin Hemorheol Microcirc. 2019;73(1):95–104.

30.

Nagy

, Chang

, Shih

, et al. Heterogeneity of the tumor vasculature. Semin Thromb Hemost. 2010;36(3):321–31.

31.

Hudson

, Williams

, Tremblay-Darveau

, et al. Dynamic contrast enhanced ultrasound for therapy monitoring. Eur J Radiol. 2015;84(9):1650–7.

32.

Wang

, Kaneko

, Tian

, et al. Three-dimensional ultrasound molecular imaging of angiogenesis in colon cancer using a clinical matrix array ultrasound transducer. Invest Radiol. 2015;50(5):322–9.

33.

Wang

, Hristov

, Qin

, et al. Three-dimensional Dynamic Contrast-enhanced US Imaging for Early Antiangiogenic Treatment Assessment in a Mouse Colon Cancer Model. Radiology. 2015;277(2):424–34.

34.

, Sigrist

, Fisher

, et al. Quantitative Three-Dimensional Dynamic Contrast-Enhanced Ultrasound Imaging: First-In-Human Pilot Study in Patients with Liver Metastases. Theranostics. 2017;7(15):3745–58.

35.

Claudon

, Dietrich

, Choi

, et al. Guidelines and good clinical practice recommendations for contrast enhanced ultrasound (CEUS) in the liver–update 2012: a WFUMB-EFSUMB initiative in cooperation with representatives of AFSUMB, AIUM, ASUM, FLAUS and ICUS. Ultraschall Med. 2013;34(1):11–29.

36.

Salvaggio

, Campisi

, Lo

, et al. Evaluation of posttreatment response of hepatocellular carcinoma: comparison of ultrasonography with second-generation ultrasound contrast agent and multidetector CT. Abdom Imaging. 2010;35(4):447–53.

37.

Kim

, Kim

, Yoon

, et al. Feasibility of Using Volumetric Contrast-Enhanced Ultrasound with a 3-D Transducer to Evaluate Therapeutic Response after Targeted Therapy in Rabbit Hepatic VX2 Carcinoma. Ultrasound Med Biol. 2015;41(12):3131–9.

38.

Wang

, Wang

, Liu

, et al. The role of quantitation of real-time 3-dimensional contrast-enhanced ultrasound in detecting microvascular invasion: an in vivo study. Abdom Radiol (NY). 2016;41(10):1973–9.