- Email: [email protected]
Acoustic Structure Quantification Versus Point Shear Wave Speed Measurement for the Assessment of Liver Fibrosis in Viral Hepatitis B
ARTICLE IN PRESS Ultrasound in Med. & Biol., Vol. ■■, No. ■■, pp. ■■–■■, 2018 Copyright © 2018 World Federation for Ultrasound in Medicine & Biology. All rights reserved. Printed in the USA. All rights reserved 0301-5629/$ - see front matter
https://doi.org/10.1016/j.ultrasmedbio.2018.02.010
●
Original Contribution
ACOUSTIC STRUCTURE QUANTIFICATION VERSUS POINT SHEAR WAVE SPEED MEASUREMENT FOR THE ASSESSMENT OF LIVER FIBROSIS IN VIRAL HEPATITIS B Jianxue Liu,*,† Wei Ren,‡ Hong Ai,§ Guoliang Dun,† Yonghao Ji,† Yaoren Zhang,† Qinyun Wan,† and Shumei Lin* * Department of Infectious Disease, First Affiliated Hospital of Medical College, Xi’an Jiaotong University, Shaanxi, China; Department of Ultrasonography, Baoji Central Hospital, Baoji, China; ‡ Department of Infectious Disease, Baoji Central Hospital, Baoji, China; and § Department of Ultrasonography, First Affiliated Hospital of Medical College, Xi’an Jiaotong University, Shaanxi, China †
(Received 10 July 2017; revised 5 December 2017; in final form 22 February 2018)
Abstract—This study was conducted to evaluate the value of acoustic structure quantification (ASQ) technology versus that of point shear wave speed measurement (PSWSM) imaging technology for the assessment of liver fibrosis stage. A total of 104 patients with chronic hepatitis B (CHB) and 30 healthy control patients underwent ASQ and PSWSM examinations. Seven quantitative parameters were obtained from ASQ, and a principal component analysis was used to establish the integrative indicators. A quantitative parameter, known as the shear wave speed (SWS, m/s), was obtained from the PSWSM. The METAVIR scores for the assessment of pathologic liver fibrosis were used as a benchmark. Liver fibrosis stages exhibited a good correlation with the integrative indicators and SWS (r = 0.682, p < 0.001; r = 0.651, p < 0.001). The areas under the receiver operating characteristic curves for ASQ and PSWSM were 0.705 and 0.854 for mild liver fibrosis (F ≥ 1, p = 0.045), 0.813 and 0.743 for significant liver fibrosis (F ≥ 2, p = 0.115), 0.839 and 0.857 for severe liver fibrosis (F ≥ 3, p = 0.417) and 0.874 and 0.971 for liver cirrhosis (F = 4, p = 0.016), respectively. In conclusion, both ASQ and PSWSM were promising ultrasonic methods for assessing liver fibrosis in patients with CHB; however, PSWSM was more valuable for identifying mild liver fibrosis (F ≥ 1) and cirrhosis (F = 4) than ASQ, and the combination of PSWSM and ASQ improved the accuracy of diagnosing severe liver fibrosis (F ≥ 3). (E-mail: arfi2014@ 126.com, [email protected]) © 2018 World Federation for Ultrasound in Medicine & Biology. All rights reserved. Key Words: Liver fibrosis, Non-invasive diagnosis, Acoustic structure quantification, Point shear wave speed measurement, Chronic hepatitis B, Liver biopsy.
fibrosis and liver cirrhosis. In China, chronic viral infections account for 60%–80% of liver fibrosis and liver cirrhosis (Wang et al. 2014). An accurate assessment of the extent of liver fibrosis is important for the treatment, prognosis and monitoring of viral hepatitis B. Liver biopsy (LB) remains the gold standard for the diagnosis of liver fibrosis (Gebo et al. 2002). However, LB is an invasive examination that terrifies some patients. Pain may occur during liver biopsy; 25% of patients experience pain after this procedure (Sebastiani 2009). Other uncommon complications include intra-abdominal hemorrhage, infection, bile leakage, hemothorax and pneumothorax (Filingeri et al. 2015), and the mortality rate is 0.05% (Bravo et al. 2001). Because of the unevenness of liver fibrosis, liver biopsy strips (which account for only 1 of 50,000 livers) are not adequate to reflect the extent
INTRODUCTION Liver fibrosis is the main characteristic of chronic liver disease. Overall, 20%–40% of chronic liver disease cases progress to liver fibrosis or cirrhosis. Severe liver cirrhosis can cause bleeding from esophageal and gastric varices, hepatic encephalopathy and hepatic failure (Ramachandran and Iredale 2012). Moreover, progressive liver fibrosis is closely related to hepatocellular carcinoma (El-Serag 2002). Chronic viral infection is the main factor involved in liver
Address correspondence to: Shumei Lin, Department of Infectious Disease, First Affiliated Hospital of Medical College, Xi’an Jiaotong University, Xi’an Yanta West Road No. 277, Shaanxi, China. E-mail: [email protected], [email protected] Conflict of Interest: The authors do not have conflicts of interest to disclose. 1
ARTICLE IN PRESS 2
Ultrasound in Medicine & Biology
of fibrosis of the whole liver (Duarte-Rojo et al. 2012). Additionally, inter-observer and intra-observer differences in pathologic observations can influence the accuracy of the assessment of liver fibrosis (Regev et al. 2002). Ultrasound-based elastography is a non-invasive and quantitative tool for the assessment of tissue stiffness. Liver fibrosis, which is caused by the excessive deposition of extracellular matrix, results in increased tissue stiffness, prompting the use of ultrasound elastography to assess these changes. Currently, the most commonly used technologies in clinical research and applications are transient elastography (TE, FibroScan, Echosens, Paris, France) and ultrasound shear wave elastography (USWE). The main difference between TE and USWE is the nature of the excitation. The external excitation in TE punches the body surface to generate shear waves propagating through the tissue along the “A line” (Shiina et al. 2015). USWE does not require external pressure to assess elastic information from deep tissue. Instead, it emits focused acoustic radiation pulses (0.05–1 ms) to induce the tissue to produce a shear wave propagating along the transverse direction (Nightingale et al. 2003). Both measure the propagation of the shear wave in one direction. Compared with USWE, TE has been used in clinical practice for a longer time and has been included in the Guidelines for the Diagnosis and Management of Chronic Liver Diseases (European Association for the Study of the Liver [EASL] 2012). Some studies have indicated that performance of point shear wave measurement (PSWSM) is comparable to that of TE in the diagnosis of liver fibrosis (Bota et al. 2013). However, FibroScan has limitations when it is used in patients with obesity or ascites (Castéra et al. 2010). In addition, TE does not provide a guiding B-mode image. Obesity is a limitation for the application of PSWSM, and the rate of failed PSWSM in obese patients (body mass index [BMI] ≥ 30 kg/m2) was reported as 17.6% (Bota et al. 2014). Ultrasonic elastography can be used to assess the stiffness of tissues. Acoustic structure quantification (ASQ) is a new tool for assessment of the structure of tissues that are not categorized as elastography. ASQ is software that analyzes the statistical information of the acquired echo signals and is based on the raw gray-scale data. The raw data are collected before the scan converter and lateral filter are applied. ASQ is thought to be less subjective and less dependent on the operator than conventional ultrasound imaging because it can offer both qualitative visual results (parametric imaging) and quantitative results. ASQ software has been commercialized and is available in Toshiba ultrasound scanners. Preliminary studies have obtained encouraging results in terms of the assessment of liver fibrosis using ASQ (Krämer et al. 2014; Ricci et al. 2013). The present study was aimed at determining the diagnostic value of ASQ versus that of PSWSM for
Volume ■■, Number ■■, 2018
the assessment of liver fibrosis in chronic hepatitis B (CHB). METHODS Patients This study was approved by the Baoji Central Hospital Ethics Committee. Informed consent for all examinations was provided in writing by the patients or their family members. A total of 104 patients (aged 13– 64 y, mean age: 30.5 ± 11.1 y) with chronic viral hepatitis B, including 70 males (aged 13–64 y, mean age: 29.0 ± 10.5 y) and 34 females (aged 16–62 y, mean age: 32.8 ± 11.7 y), were recruited from the Department of Infectious Diseases of Baoji Central Hospital between February 2015 and May 2016. All patients underwent LB under ultrasonic guidance after ASQ and PSWSM, but did not receive antiviral therapy. The inclusion criteria were the presence of hepatitis B virus (HBV) deoxyribonucleic acid (DNA) in the sera and positivity for HBV surface antigen for at least 6 consecutive mo. The exclusion criteria were hepatitis B associated with a fatty liver, liver congestion, primary biliary hepatitis, autoimmune hepatitis and other types of viral hepatitis such as viral hepatitis C. The criterion for hepatitis C was positive for hepatitis C virus antibody. The criterion for the diagnosis of fatty liver was biopsy pathology revealing liver steatosis >5% (Kleiner et al. 2005). Thirty healthy volunteers, including 15 males and 15 females (aged 18–45 y, mean age of 32.4 ± 11.7 y), were enrolled as controls. All volunteers had normal levels of serum liver enzymes (alanine aminotransferase [ALT], aspartate aminotransferase [AST], γ-glutamine transferase [γ-GGT] and alkaline phosphatase [ALP]) and serum bilirubin (total bilirubin, direct bilirubin and indirect bilirubin). The volunteers were negative for hepatitis B virus surface antigen and hepatitis C virus antibody and had no history of diabetes, hyperlipidemia or alcoholism. Conventional ultrasonic scans did not reveal abnormalities of the liver. Acoustic structure quantification After fasting for more than 8 h, each subject underwent ASQ examination by a radiologist with 3 mo of experience in ASQ examinations using a Toshiba Aplio 500 ultrasound machine (Toshiba Medical Systems, Osaka, Japan) at a frequency of 5.0 MHz. The subject was placed in the left lateral decubitus position with the right hand placed behind the head. The probe was placed in the right intercostal area for scanning. Subjects were advised to hold their breath at maximum inspiration while the ASQ imaging was initiated. The frame rate for the collection was 8 frames per second (fps) for 3 s. Images were stored and copied to a computer. Segment V of the right anterior lobe of the liver was selected as the region of interest (ROI). The de-
ARTICLE IN PRESS ASQ vs. PSWSM for liver fibrosis ● J. Liu et al.
Fig. 1. Histogram indicating the Cm2 value as a red curve (microstructures) and a blue curve (macrostructures).
tection depth was 1.0 cm from the liver capsule with a size of approximately 3.0 × 3.0 cm, and major vessels, calcified lesions and the calculus of the intrahepatic duct in the liver were avoided. The raw echo signals of the liver tissue were statistically analyzed using ASQ software. ASQ is software that analyzes the statistical value of the acquired echo signal amplitude within a certain ROI. The measurement is compared with an estimated variance of normal or reference liver tissue. ASQ assesses liver fibrosis based on a statistical test (χ2-test) and on the statistical distribution of echo amplitudes measured in absolutely homogeneous liver parenchyma according to the Rayleigh probability density function (PDF). The χ2 value (Cm2) equals the variance-measured PDF divided by the variance of the Rayleigh PDF (Toyoda et al. 2009).
3
The result was displayed in a histogram using Cm2 as the horizontal axis and occurrence as the vertical axis. The histogram contains a red microstructure curve and a blue macrostructure curve (Fig. 1). The software places many fine ROIs in each manually set ROI. A Cm2 value is then calculated for each individual fine ROI. The Cm2 values of fine ROIs containing homogeneous parenchyma are assigned to the red microstructure curve. The Cm2 values of fine ROIs containing pathologic inhomogeneous tissue and focal inhomogeneities such as small blood vessels, biliary ducts and ligaments are assigned to the blue curve (Kaltenbach et al. 2014). ASQ is intended to facilitate the analysis of tissue homogeneity in the liver. Quantitative analysis generates a Cm2 histogram for an ROI. The higher the Cm2 value, the greater is the deviation from ideal tissue (Keller et al. 2015). ASQ-associated parameters—mode, average and standard deviation of the Cm2 values for the red and blue curves—were called red average (RA), red mode (RM), red standard deviation (RSD), blue average (BA), blue mode (BM), blue standard deviation (BSD) and the focal disturbance (FD). The FD is the ratio of the area under the red curve to the area under the blue curve (Figs. 2a, b). The values for RM, RA and RSD reflected the echo characteristics of normal liver tissue and tissue with mild liver fibrosis. The values for BM, BA and BSD in normal liver tissue and tissue with mild liver fibrosis correspond to the values expected for small blood vessels or biliary ducts. The values for BM, BA and BSD mainly reflected the inhomogeneous distribution of severe fibrosis and liver cirrhosis, which are even less similar to the Rayleigh distribution (Kuroda et al. 2012; Nam et al. 2016; Ricci et al. 2013).
Fig. 2. (a) Acoustic structure quantification (ASQ) examination in a normal subject. (b) ASQ examination in a patient with cirrhosis. Note that the seven parameters for the patient with cirrhosis were significantly higher than those for the normal subject. The corresponding ASQ color map illustrates an increase in the red color.
ARTICLE IN PRESS 4
Ultrasound in Medicine & Biology
Volume ■■, Number ■■, 2018
Fig. 3. (a) Point shear wave speed measurement (PSWSM) in an F3 patient. (b) PSWSM in a patient with cirrhosis. Note that the shear wave speed of the cirrhosis patient with cirrhosis was significantly higher than that of the F3 patient (2.23 m/s vs. 1.55 m/s).
Point shear wave speed measurement Point shear wave speed was measured using virtual touch tissue quantification (VTTQ) and a Siemens ACUSON S2000 ultrasound machine (Siemens Medical Solutions USA, Mountain View, CA, USA). VTTQ was assessed immediately after ASQ was completed. The assessment was performed by a radiologist with at least a half-year of experience with VTTQ examinations. Before measurement, subjects were advised to assume the left lateral decubitus position with the right hand placed behind the head. Standard scanning of the liver was performed. Segment V of the right anterior lobe of the liver was selected as the ROI, which was identical to that in the ASQ examination. Vessels and bile ducts were avoided. The ROI was approximately 1.0 cm below the liver capsule. During the measurement, subjects were advised to hold their breath at the end of inhalation. For each patient and volunteer, the same site was measured 10 times without repositioning the transducer, and the median was selected (Fig. 3a, b). Liver biopsy and histopathology Liver biopsies were conducted immediately after the PSWSM examinations using the same ultrasound machine to ensure that the puncture site was identical to those in the PSWSM and ASQ examinations. In addition, to reduce sampling error, PSWSM and ASQ were performed in the same intercostal space of the patient after maximum inhalation to ensure similar imaging planes for the same patient switching from ASQ on a Toshiba scanner to PSWSM on a Siemens scanner. The puncture routes avoided major vessels in the liver. An 18G core needle was inserted two or three times to collect tissues. The tissue strips had lengths of at least 1.5 cm and involved at least six portal areas.
Tissues the LBs were fixed with 10% neutral formaldehyde solution, embedded in paraffin, cut into serial sections and stained with standard hematoxylin and eosin (HE), Masson, and reticular fiber staining. The tissues were examined by an independent pathologist with rich clinical experience at our center. The histopathologic stage of liver fibrosis was determined using the METAVIR score (stages F0–F4), where F0 indicates no fibrosis, F1 indicates portal fibrosis but no fibrotic septa, F2 indicates portal fibrosis associated with a few fibrotic septa, F3 indicates numerous fibrotic septa but not cirrhosis and F4 indicates cirrhosis (Bedossa and Poynard 1996). Statistical analysis Statistical analyses were performed using SPSS Version 13.0 software (IBM, Armonk, NY, USA) and MedCalc Version 15.2.1 (MedCalc Program, Ostend, Belgium). Seven parameters obtained from the ASQ were analyzed using principal component analysis to formulate the integrative function (I). Using this integrative function (I), the functional value, called the integrative indicator (II), was calculated for every patient. The eigenvalue is the variance of the main function (Z). Each eigenvalue is the contribution of each eigenvector to the total variance. The relationships of liver fibrosis stage to II and SWS were analyzed using Spearman’s rank correlation. To generate predictive probability as an index of the combination of PSWSM and ASQ, binary logistic regression analyses were performed. The receiver operating characteristic (ROC) curve was used to assess the performance of each indicator in the diagnosis of liver fibrosis stage. The areas under the ROC curves (AUROCs) were compared with Z-tests. Cutoff values were selected using the Youden index (sum of sensitivity and specificity), and the corresponding sensitivities and specificities were calculated. Finally,
ARTICLE IN PRESS ASQ vs. PSWSM for liver fibrosis ● J. Liu et al.
5
Table 1. Subject characteristics Characteristic
F0 (n = 30 + 3)
F1 (n = 33)
F2 (n = 19)
F3 (n = 29)
F4 (n = 20)
Sex (male/female) Age (y) ALT (IU/L) AST (IU/L) GGT Total cholesterol Platelets (×109/L)
15/15 33.4 ± 5.8 17.7 ± 3.7 18.0 ± 7.7 21.0 ± 8.5 3.25 ± 0.65 204.9 ± 42.1
17/16 29.8 ± 11.5 44.3 ± 115.2 30.4 ± 44.4 23.7 ± 52.8 3.68 ± 0.77 182.3 ± 51.6
11/8 31.7 ± 12.8 57.1 ± 128.0 41.2 ± 84.0 41.5 ± 105.1 3.76 ± 0.87 176.7 ± 55.8
15/14 36.7 ± 11.1 63.1 ± 48.6 82.8 ± 126.8 254.5 ± 648.8 3.93 ± 1.46 178.5 ± 54.5
12/8 41.2 ± 8.8 47.8 ± 32.3 54.2 ± 26.6 70.3 ± 45.0 3.49 ± 0.69 103.8 ± 36.5
ALT = alanine aminotransferase, AST = aspartate aminotransferase, GGT = γ-glutamyl transferase.
Table 2. Seven parameters of acoustic structure quantification Parameter RM RA RSD BM BA BSD FD
F0
F1
F2
F3
F4
111.90 ± 4.48 115.30 ± 3.76 17.40 ± 2.56 128.90 ± 7.95 138.10 ± 10.50 29.10 ± 10.56 0.23 ± 0.08
112.72 ± 5.51 116.30 ± 4.78 17.00 ± 2.03 126.97 ± 7.34 140.24 ± 10.68 32.27 ± 13.31 0.31 ± 0.13
113.89 ± 8.69 117.28 ± 6.62 18.28 ± 1.74 128.72 ± 8.11 145.50 ± 12.20 37.33 ± 12.83 0.48 ± 0.24
120.00 ± 4.86 123.17 ± 3.34 21.31 ± 1.59 133.69 ± 6.61 146.62 ± 5.94 33.86 ± 11.47 0.61 ± 0.18
122.65 ± 5.36 125.80 ± 3.50 19.50 ± 1.82 138.60 ± 7.09 152.40 ± 9.20 35.15 ± 12.15 0.74 ± 0.20
RM = red curve mode, RA = red curve average, RSD = red curve standard deviation, BA = blue curve average, BM = blue curve mode, BSD = blue curve standard deviation, FD = focal distribution.
p values < 0.05 were considered to indicate statistical significance. RESULTS Subject characteristics The pathologic results from the LBs of 104 cases revealed 3 (2.9%) cases at the F0 stage, 33 (31.7%) at the F1 stage, 19 (18.3%) at the F2 stage, 29 (27.9%) at the F3 stage and 20 (19.2%) at the F4 stage. Among these cases, 68 (65.4%) had significant fibrosis (F ≥ 2). General data and laboratory indicators for the patients and volunteers are provided in Table 1. Integrative indicators obtained from the ASQ parameters Results for the seven parameters obtained from the ASQ are provided in Table 2. From the seven parameters of the ASQ tool, RSD and BSD were not correlated with stage of liver fibrosis. The remaining five parameters were used to build an integrative function for the assessment of liver fibrosis stage. Based on the results and practical requests, the anterior four main functions (Z) were selected to formulate the integrative function (I) and had eigenvalues of 3.215, 1.719, 0.956 and 0.598, respectively. The contributions of Z1 to Z4 were 45.9%, 24.5%, 13.6% and 8.5%, respectively, and the cumulative contribution was 92.4%. The anterior four main functions (Z) and integrative function (I) are expressed as follows: Z1 = 0.814 RM + 0.893 RA + 0.719 BM + 0.665 BM + 0.818 FD
Z2 = -0.426 RM – 0.310 RA – 0.152 BM + 0.663 BM – 0.056 FD Z2 = -0.029 RM – 0.029 RA + 0.152 BM – 0.121 BM – 0.236 FD Z4 = 0.177 RM + 0.136 RA – 0.648 BM – 0.128 BM + 0.272 FD I = 45.9% Z1 + 24.5% Z2 + 13.6% Z3 + 8.5% Z4
Analysis of the relationship between non-invasive parameters and stage of liver fibrosis The integrative indicators obtained from ASQ and SWS from PSWSM were closely correlated with the stages of liver fibrosis (Figs. 4 and 5, Table 3). Among the seven parameters from ASQ, FD was the parameter that was most strongly correlated with stage of liver fibrosis (r = 0.583, p < 0.001), whereas RSD and BSD were not correlated with stage of liver fibrosis (r = 0.140, p = 0.127; r = 0.147, p = 0.109).
Table 3. Spearman’s coefficients for the non-invasive parameters and liver fibrosis stages Liver fibrosis stage METAVIR scores
Non-invasive parameter
Spearman coefficient
p Value
PSWSM ASQ
0.682 0.651
<0.001 <0.001
PSWSM = point shear-wave speed measurement, ASQ = acoustic structure quantification.
ARTICLE IN PRESS 6
Ultrasound in Medicine & Biology
Volume ■■, Number ■■, 2018
Fig. 4. Scattergraph revealing the correlation between the integrative indicator from acoustic structure quantification (ASQ) and the histologic results from the liver biopsies.
ASQ versus PSWSM for the diagnosis of liver fibrosis The AUROCs, cutoff values, sensitivities and specificities for the diagnosis and prediction of liver fibrosis stage using ASQ and PSWSM are provided in Table 4. For the diagnosis of mild liver fibrosis (F ≥ 1), ASQ (0.705) had a significantly lower AUROC than PSWSM (0.854, p = 0.045) (Fig. 6). For the diagnosis of significant liver fibrosis (F ≥ 2), ASQ (0.813) had a greater AUROC than PSWSM (0.743), although the
difference was not significant (p = 0.115). ASQ was comparable to PSWSM for the diagnosis of severe liver fibrosis (F ≥ 3) with AUROCs of 0.839 and 0.857, respectively. For the diagnosis of liver cirrhosis (F4), ASQ was lower for PSWSM, with AUROCs of 0.874 and 0.971, respectively (p = 0.016) (Fig. 7). The AUROC of PSWSM combined with ASQ for the diagnosis of severe liver fibrosis was greater than that of PSWSM alone (p = 0.020) (Fig. 8).
Fig. 5. Scattergraph revealing the correlation between shear wave speed (SWS) by point shear wave speed measurement (PSWSM) and the histologic results from the liver biopsies.
ARTICLE IN PRESS ASQ vs. PSWSM for liver fibrosis ● J. Liu et al.
7
Table 4. Cutoffs and AUROCs of ASQ and PSWSM for assessing liver fibrosis stage Non-invasive test ASQ ≥F1 ≥F2 ≥F3 F4 PSWSM ≥F1 ≥F2 ≥F3 F4 ASQ + PSWSM ≥F1 ≥F2 ≥F3 F4
AUROC (95% CI)
Cutoff
Sensitivity (%)
Specificity (%)
0.705 (0.625–0.794) 0.813 (0.732–0.879) 0.855 (0.778–0.913) 0.873 (0.822–0.914)
181.13 181.54 185.44 191.91
54.0 71.6 81.3 70.0
90.0 86.8 78.9 93.2
0.854 (0.777–0.911) 0.742 (0.678–0.809) 0.890 (0.819–0.913) 0.969 (0.937–0.988)
1.28 1.35 1.48 1.74
65.0 68.7 87.5 100.0
95.0 71.7 76.1 93.7
0.877 (0.805–0.930) 0.836 (0.764–0.908) 0.936 (0.880–0.975) 0.970 (0.939–0.988)
0.804 0.479 0.446 0.168
80.0 82.1 83.3 95.0
85.0 75.5 88.7 95.1
PSWSM = point shear-wave speed measurement, ASQ = acoustic structure quantification, AUROC = area under the receiver operating characteristic curve, 95% CI = 95% confidence interval.
DISCUSSION Currently, ultrasound-based non-invasive methods are the mainstay for the assessment of liver fibrosis. However, the performance of these methods in the assessment of liver fibrosis remains to be further validated. This study found that the correlation between METAVIR stage and PSWSM parameters was slightly higher than that obtained with ASQ; the Spearman coefficients were 0.682 and 0.651, respectively. Liver fibrosis is caused by excessive deposition of matrix, especially collagen, in the liver. Tang et al. (2015) suggested that increasing SWS values are related to in-
creased collagen concentration during fibrosis. ASQ parameters are derived from statistical information on scattering sources. Given that collagen is an important source of scattering, these two techniques might be directly measuring the same tissue changes; therefore, the correlations with fibrosis would not be expected to be markedly different. Point shear wave speed measurement using VTTQ is easy to perform, low in cost, free of ionizing radiation and guided by 2-D images (Son et al. 2012) and has been in clinical use for 10 y. Sporea et al. (2012) used PSWSM to assess liver fibrosis stage in 53 cases with viral
Fig. 6. Receiver operating characteristic (ROC) curves revealing performance in the diagnosis of liver fibrosis (F ≥ 1) by acoustic structure quantification (ASQ), point shear wave speed measurement (PSWSM) and their combination. Diagonal segments are produced by ties.
ARTICLE IN PRESS 8
Ultrasound in Medicine & Biology
Volume ■■, Number ■■, 2018
Fig. 7. Receiver operating characteristic (ROC) curves revealing performance in the diagnosis of cirrhosis (F4) by acoustic structure quantification (ASQ), point shear wave speed measurement (PSWSM) and their combination. Diagonal segments are produced by ties.
hepatitis B and 107 cases with viral hepatitis C and found that PSWSM obtained comparable values in the assessment of liver fibrosis stage in viral hepatitis B and viral hepatitis C. A multicenter study reported by Friedrich-Rust et al. (2013) indicated no difference in the diagnostic values of PSWSM and TE in the assessment of liver fibrosis at
F ≥ 2, F ≥ 3 or F4 and the AUROCs for PSWSM in the diagnosis of liver fibrosis were 0.75 for F ≥ 2, 0.93 for F ≥ 3 and 0.97 for F4. The AUROCs (0.74 for F ≥ 2, 0.89 for F ≥ 3 and 0.97 for F4) for diagnosis of liver fibrosis in the present study were comparable to those from the study of Friedrich-Rust et al. (2013). In the present study, PSWSM
Fig. 8. Receiver operating characteristic (ROC) curves revealing performance in the diagnosis of liver fibrosis (F ≥ 3) by acoustic structure quantification (ASQ), point shear wave speed measurement (PSWSM) and their combination.
ARTICLE IN PRESS ASQ vs. PSWSM for liver fibrosis ● J. Liu et al.
was performed during a deep inhalation state. A previous study provided proof that deep inspiration significantly modulates spleen stiffness (Karlas et al. 2014). However, liver SWSs in deep inspiration, in deep expiration and during the Valsalva maneuver did not differ (Goertz et al. 2012). In the case of ASQ, Krämer et al. (2014) found that the stages of liver fibrosis are not correlated with the average or peak Cm2 values obtained from the ASQ software and suggested that ASQ was not accurate in the assessment of liver fibrosis stage. However, Nam et al. (2016) found that the stages of liver fibrosis were well correlated with the mode, average and FD obtained from the ASQ software and most strongly correlated with mode. Additionally, Huang et al. (2016) found a strong correlation between FD and stage of liver fibrosis. In the present study, the parameters obtained from the ASQ software—RM, RA, BM, BA and FD—were well correlated with the stage of liver fibrosis (p < 0.001). FD was most strongly correlated with stage of liver fibrosis (r = 0.583), followed by RM and RA. Liver fibrosis stages were not correlated with RSD or BSD. We used principal component analysis to build integrative indicators that included RM, RA, BM, BD and FD to assess liver fibrosis, which resulted in a rank correlation coefficient of 0.651 with liver fibrosis (Table 3). Ricci et al. (2013) found that ASQ did not attain the same level of diagnostic value as FibroScan for the diagnosis of liver fibrosis stage in viral hepatitis B and C. The AUROCs for ASQ and FibroScan were 0.71 and 0.77, respectively, for the diagnosis of liver fibrosis (F ≥ 1) and 0.94 and 0.97, respectively, for the diagnosis of liver cirrhosis (F4). We found that ASQ produced a lower AUROC than PSWSM in terms of the diagnosis of liver fibrosis (F ≥ 1) and liver cirrhosis (F4). The reasons for the difference may be as follows: In patients in stage F1, the difference caused by portal fibrosis as a scattering source was weaker than that of liver tissue stiffness caused by portal fibrosis. Numerous fibrotic tissues can be used as a remarkable source of scattering, but F4 is generally associated with severe liver inflammation and a combination of liver fibrosis and severe inflammation, which can lead to increased liver stiffness. However, it was comparable to PSWSM in the diagnoses of significant liver fibrosis (F ≥ 2) and severe liver fibrosis (F ≥ 3), but the combination of two different techniques improved the accuracy in diagnosis of severe liver fibrosis (F ≥ 3). The AUROC was higher in our research than in the study of Ricci et al. (2013) (0.87 vs. 0.77) for the diagnosis of liver cirrhosis, a finding that might be related to the diagnosis achieved by combining relevant parameters from ASQ in the present study. In addition, Ricci et al.’s study included liver fibrosis of viral hepatitis C as well as viral hepatitis B. Liver cirrhosis of hepatitis C is usually micronodular liver cirrhosis, similar to normal liver
9
in that scattering noise is represented by an echo texture pattern. Therefore, ASQ might be weak for the prediction of liver fibrosis stage in patients with hepatitis C. There are some limitations to the present study. First, we did not assess the repeatability or reproducibility of ASQ for the assessment of liver fibrosis stage because ASQ is a novel technology for assessing liver fibrosis stage; we will evaluate the repeatability and reproducibility of ASQ in future studies. Second, healthy patients were included as controls because only a few (three patients) had F0 liver fibrosis in the present study, and pathologic diagnoses were not obtained from the control group for ethical reasons. Third, the test was carried out only in patients with HBV; we need to compare the performance of PSWSM and ASQ in the diagnosis of liver fibrosis in hepatitis C in future studies. Fourth, because of the complexity of their calculation, integrative indicators cannot be widely accepted in clinical practice. In conclusion, ASQ is a promising ultrasonic technology for the assessment of liver fibrosis and is comparable to PSWSM in terms of the diagnosis of significant liver fibrosis and severe liver fibrosis. However, PSWSM was better than ASQ for the diagnosis of mild liver fibrosis (F ≥ 1) and liver cirrhosis (F4), and the combination of PSWSM and ASQ improves the accuracy of diagnosis of severe liver fibrosis (F ≥ 3). Acknowledgments—We express our deepest gratitude to the Baoji City Health and Family Planning Commission for financial support. We thank Dr. Bi Yuxue (School of Public Health, Xi’an Jiaotong University Health Science Center, China) for providing helpful contributions to the statistical analysis.
REFERENCES Bedossa P, Poynard T. An algorithm for the grading of activity in chronic hepatitis C. The METAVIR Cooperative Study Group. Hepatology 1996;24:289–293. Bota S, Herkner H, Sporea I, Salzl P, Sirli R, Neghina AM, Peck-Radosavljevic M. Meta-analysis: ARFI elastography versus transient elastography for the evaluation of liver fibrosis. Liver Int 2013; 33:1138–1147. Bota S, Sporea I, Sirli R, Popescu A, Danila M, Jurchis A, Gradinaru-Tascau O. Factors associated with the impossibility to obtain reliable liver stiffness measurements by means of acoustic radiation force impulse (ARFI) elastography–—Analysis of a cohort of 1,031 subjects. Eur J Radiol 2014;83:268–272. Bravo AA, Sheth SG, Chopra S. Liver biopsy. N Engl J Med 2001;344: 495–500. Castéra L, Foucher J, Bernard PH, Carvalho F, Allaix D, Merrouche W, Couzigou P, de Lédinghen V. Pitfalls of liver stiffness measurement: A 5-years prospective study of 13,369 examinations. Hepatology 2010; 51:828–835. Duarte-Rojo A, Altamirano JT, Feld JJ. Noninvasive markers of fibrosis: Key concepts for improving accuracy in daily clinical practices. Ann Hepatol 2012;11:426–438. El-Serag HB. Hepatocellular carcinoma and hepatitis C in the United States. Hepatology 2002;36:74–83. European Association for the Study of the Liver [EASL]. EASL clinical practice guidelines: Management of chronic hepatitis B virus infection. J Hepatol 2012;57:167–185.
ARTICLE IN PRESS 10
Ultrasound in Medicine & Biology
Filingeri V, Sforza D, Tisone G. Complications and risk factors of a large series of percutaneous liver biopsies in patients with liver transplantation or liver disease. Eur Rev Med Pharmacol Sci 2015;19:1621– 1629. Friedrich-Rust M, Buggisch P, de Knegt RJ, Dries V, Shi Y, Matschenz K, Schneider MD, Herrmann E, Petersen J, Schulze F, Zeuzem S, Sarrazin C. Acoustic radiation force impulse imaging for noninvasive assessment of liver fibrosis in chronic hepatitis B. J Viral Hepat 2013;20:240–247. Gebo KA, Herlong HF, Torbenson MS, Jenckes MW, Chander G, Ghanem KG, El-Kamary SS, Sulkowski M, Bass EB. Role of liver biopsy in management of chronic hepatitis C: A systematic review. Hepatology 2002;36:161–172. Goertz RS, Egger C, Neurath MF, Strobel D. Impact of food intake, ultrasound transducer, maneuvers and body position on acoustic radiation force impulse (ARFI) elastometry of the liver. Ultraschall Med 2012; 33:380–385. Huang Y, Wang Z, Liao B, Liang JY, Zhou LY, Wang F, Li W, Liu JY, Xie XY, Lu MD, Liu GJ, Wang W. Assessment of liver fibrosis in chronic hepatitis B using acoustic structure quantification: Quantitative morphological ultrasound. Eur Radiol 2016;26:2344–2351. Kaltenbach TE, Gruener B, Akinli AS, Kratzer W, Oeztuerk S, Mason RA, Haenle MM, Graeter T. Acoustic structure quantification (ASQ): A new tool in sonographic examination of liver lesions alveolar echinococcosis. J Med Ultrason (2001) 2014;41:445–453. Karlas T, Lindner F, Tröltzsch M, Keim V. Assessment of spleen stiffness using acoustic radiation force impulse imaging (ARFI): Definition of examination standards and impact of breathing maneuvers. Ultraschall Med 2014;35:38–43. Keller J, Kaltenbach TE, Haenle MM, Oeztuerk S, Graeter T, Mason RA, Seufferlein T, Kratzer W. Comparison of acoustic structure quantification (ASQ), shear wave elastography and histology in patients with diffuse hepatopathies. BMC Med Imaging 2015;15:58. Kleiner DE, Brunt EM, Van Natta M, Behling C, Contos MJ, Cummings OW, Ferrell LD, Liu YC, Torbenson MS, Unalp-Arida A, Yeh M, McCullough AJ, Sanyal AJ, Nonalcoholic Steatohepatitis Clinical Research Network. Design and validation of a histological scoring system for nonalcoholic fatty liver disease. Hepatology 2005;41: 1313–1321. Krämer C, Jaspers N, Nierhoff D, Kuhr K, Bowe A, Goeser T, Michels G. Acoustic structure quantification ultrasound software proves imprecise in assessing liver fibrosis or cirrhosis in parenchymal liver diseases. Ultrasound Med Biol 2014;40:2811–2818. Kuroda H, Kakisaka K, Kamiyama N, Oikawa T, Onodera M, Sawara K, Oikawa K, Endo R, Takikawa Y, Suzuki K. Non-invasive determination of hepatic steatosis by acoustic structure quantification from ultrasound echo amplitude. World J Gastroenterol 2012;18:3889– 3895.
Volume ■■, Number ■■, 2018 Nam TH, Park BJ, Sung DJ, Kim MJ, Han NY, Sim KC, Lee JH, Cho SB. Non-invasive assessment of liver fibrosis using acoustic structure quantification: Comparison with transient elastography in chronic viral hepatitis. Abdom Radiol (NY) 2016;41:239–247. Nightingale K, McAleavey S, Trahey G. Shear-wave generation using acoustic radiation force: In vivo and ex vivo results. Ultrasound Med Biol 2003;29:1715–1723. Ramachandran P, Iredale JP. Liver fibrosis: A bidirectional model of fibrogenesis and resolution. QJM 2012;105:813–817. Regev A, Berho M, Jeffers LJ, Milikowski C, Molina EG, Pyrsopoulos NT, Feng ZZ, Reddy KR, Schiff ER. Sampling error and intraobserver variation in liver biopsy in patients with chronic HCV infection. Am J Gastroenterol 2002;97:2614–2618. Ricci P, Marigliano C, Cantisani V, Porfiri A, Marcantonio A, Lodise P, D’Ambrosio U, Labbadia G, Maggini E, Mancuso E, Panzironi G, Di Segni M, Furlan C, Masciangelo R, Taliani G. Ultrasound evaluation of liver fibrosis: Preliminary experience with acoustic structure quantification (ASQ) software. Radiol Med 2013;118:995– 1010. Sebastiani G. Non-invasive assessment of liver fibrosis in chronic liver diseases: Implementation in clinical practice and decisional algorithms. World J Gastroenterol 2009;15:2190–2203. Shiina T, Nightingale KR, Palmeri ML, Hall TJ, Bamber JC, Barr RG, Castera L, Choi BI, Chou YH, Cosgrove D, Dietrich CF. WFUMB guidelines and recommendations for clinical use of ultrasound elastography: Part 1. Basic principles and terminology. Ultrasound Med Biol 2015;41:1126–1147. Son CY, Kim SU, Han WK, Choi GH, Park H, Yang SC, Choi JS, Park JY, Kim DY, Ahn SH, Chon CY, Han KH. Normal liver elasticity values using acoustic radiation force impulse imaging: A prospective study in healthy living liver and kidney donors. J Gastroenterol Hepatol 2012;27:130–136. Sporea I, Sirli R, Bota S, Popescu A, Sendroiu M, Jurchis A. Comparative study concerning the value of acoustic radiation force impulse elastography (ARFI) in comparison with transient elastography (TE) for the assessment of liver fibrosis in patients with chronic hepatitis B and C. Ultrasound Med Biol 2012;38:1310–1316. Tang A, Cloutier G, Szeverenyi NM, Sirlin CB. Ultrasound elastography and MR elastography for assessing liver fibrosis: Part 2. Diagnostic performance, confounders, and future directions. AJR Am J Roentgenol 2015;205:33–40. Toyoda H, Kumada T, Kamiyama N, Shiraki K, Takase K, Yamaguchi T, Hachiya H. B-Mode ultrasound with algorithm based on statistical analysis of signals: Evaluation of liver fibrosis in patients with chronic hepatitis C. AJR Am J Roentgenol 2009;193:1037–1043. Wang FS, Fan JG, Zhang Z, Gao B, Wang HY. The global burden of liver disease: The major impact of China. Hepatology 2014;60: 2099–2108.
https://doi.org/10.1016/j.ultrasmedbio.2018.02.010
●
Original Contribution
ACOUSTIC STRUCTURE QUANTIFICATION VERSUS POINT SHEAR WAVE SPEED MEASUREMENT FOR THE ASSESSMENT OF LIVER FIBROSIS IN VIRAL HEPATITIS B Jianxue Liu,*,† Wei Ren,‡ Hong Ai,§ Guoliang Dun,† Yonghao Ji,† Yaoren Zhang,† Qinyun Wan,† and Shumei Lin* * Department of Infectious Disease, First Affiliated Hospital of Medical College, Xi’an Jiaotong University, Shaanxi, China; Department of Ultrasonography, Baoji Central Hospital, Baoji, China; ‡ Department of Infectious Disease, Baoji Central Hospital, Baoji, China; and § Department of Ultrasonography, First Affiliated Hospital of Medical College, Xi’an Jiaotong University, Shaanxi, China †
(Received 10 July 2017; revised 5 December 2017; in final form 22 February 2018)
Abstract—This study was conducted to evaluate the value of acoustic structure quantification (ASQ) technology versus that of point shear wave speed measurement (PSWSM) imaging technology for the assessment of liver fibrosis stage. A total of 104 patients with chronic hepatitis B (CHB) and 30 healthy control patients underwent ASQ and PSWSM examinations. Seven quantitative parameters were obtained from ASQ, and a principal component analysis was used to establish the integrative indicators. A quantitative parameter, known as the shear wave speed (SWS, m/s), was obtained from the PSWSM. The METAVIR scores for the assessment of pathologic liver fibrosis were used as a benchmark. Liver fibrosis stages exhibited a good correlation with the integrative indicators and SWS (r = 0.682, p < 0.001; r = 0.651, p < 0.001). The areas under the receiver operating characteristic curves for ASQ and PSWSM were 0.705 and 0.854 for mild liver fibrosis (F ≥ 1, p = 0.045), 0.813 and 0.743 for significant liver fibrosis (F ≥ 2, p = 0.115), 0.839 and 0.857 for severe liver fibrosis (F ≥ 3, p = 0.417) and 0.874 and 0.971 for liver cirrhosis (F = 4, p = 0.016), respectively. In conclusion, both ASQ and PSWSM were promising ultrasonic methods for assessing liver fibrosis in patients with CHB; however, PSWSM was more valuable for identifying mild liver fibrosis (F ≥ 1) and cirrhosis (F = 4) than ASQ, and the combination of PSWSM and ASQ improved the accuracy of diagnosing severe liver fibrosis (F ≥ 3). (E-mail: arfi2014@ 126.com, [email protected]) © 2018 World Federation for Ultrasound in Medicine & Biology. All rights reserved. Key Words: Liver fibrosis, Non-invasive diagnosis, Acoustic structure quantification, Point shear wave speed measurement, Chronic hepatitis B, Liver biopsy.
fibrosis and liver cirrhosis. In China, chronic viral infections account for 60%–80% of liver fibrosis and liver cirrhosis (Wang et al. 2014). An accurate assessment of the extent of liver fibrosis is important for the treatment, prognosis and monitoring of viral hepatitis B. Liver biopsy (LB) remains the gold standard for the diagnosis of liver fibrosis (Gebo et al. 2002). However, LB is an invasive examination that terrifies some patients. Pain may occur during liver biopsy; 25% of patients experience pain after this procedure (Sebastiani 2009). Other uncommon complications include intra-abdominal hemorrhage, infection, bile leakage, hemothorax and pneumothorax (Filingeri et al. 2015), and the mortality rate is 0.05% (Bravo et al. 2001). Because of the unevenness of liver fibrosis, liver biopsy strips (which account for only 1 of 50,000 livers) are not adequate to reflect the extent
INTRODUCTION Liver fibrosis is the main characteristic of chronic liver disease. Overall, 20%–40% of chronic liver disease cases progress to liver fibrosis or cirrhosis. Severe liver cirrhosis can cause bleeding from esophageal and gastric varices, hepatic encephalopathy and hepatic failure (Ramachandran and Iredale 2012). Moreover, progressive liver fibrosis is closely related to hepatocellular carcinoma (El-Serag 2002). Chronic viral infection is the main factor involved in liver
Address correspondence to: Shumei Lin, Department of Infectious Disease, First Affiliated Hospital of Medical College, Xi’an Jiaotong University, Xi’an Yanta West Road No. 277, Shaanxi, China. E-mail: [email protected], [email protected] Conflict of Interest: The authors do not have conflicts of interest to disclose. 1
ARTICLE IN PRESS 2
Ultrasound in Medicine & Biology
of fibrosis of the whole liver (Duarte-Rojo et al. 2012). Additionally, inter-observer and intra-observer differences in pathologic observations can influence the accuracy of the assessment of liver fibrosis (Regev et al. 2002). Ultrasound-based elastography is a non-invasive and quantitative tool for the assessment of tissue stiffness. Liver fibrosis, which is caused by the excessive deposition of extracellular matrix, results in increased tissue stiffness, prompting the use of ultrasound elastography to assess these changes. Currently, the most commonly used technologies in clinical research and applications are transient elastography (TE, FibroScan, Echosens, Paris, France) and ultrasound shear wave elastography (USWE). The main difference between TE and USWE is the nature of the excitation. The external excitation in TE punches the body surface to generate shear waves propagating through the tissue along the “A line” (Shiina et al. 2015). USWE does not require external pressure to assess elastic information from deep tissue. Instead, it emits focused acoustic radiation pulses (0.05–1 ms) to induce the tissue to produce a shear wave propagating along the transverse direction (Nightingale et al. 2003). Both measure the propagation of the shear wave in one direction. Compared with USWE, TE has been used in clinical practice for a longer time and has been included in the Guidelines for the Diagnosis and Management of Chronic Liver Diseases (European Association for the Study of the Liver [EASL] 2012). Some studies have indicated that performance of point shear wave measurement (PSWSM) is comparable to that of TE in the diagnosis of liver fibrosis (Bota et al. 2013). However, FibroScan has limitations when it is used in patients with obesity or ascites (Castéra et al. 2010). In addition, TE does not provide a guiding B-mode image. Obesity is a limitation for the application of PSWSM, and the rate of failed PSWSM in obese patients (body mass index [BMI] ≥ 30 kg/m2) was reported as 17.6% (Bota et al. 2014). Ultrasonic elastography can be used to assess the stiffness of tissues. Acoustic structure quantification (ASQ) is a new tool for assessment of the structure of tissues that are not categorized as elastography. ASQ is software that analyzes the statistical information of the acquired echo signals and is based on the raw gray-scale data. The raw data are collected before the scan converter and lateral filter are applied. ASQ is thought to be less subjective and less dependent on the operator than conventional ultrasound imaging because it can offer both qualitative visual results (parametric imaging) and quantitative results. ASQ software has been commercialized and is available in Toshiba ultrasound scanners. Preliminary studies have obtained encouraging results in terms of the assessment of liver fibrosis using ASQ (Krämer et al. 2014; Ricci et al. 2013). The present study was aimed at determining the diagnostic value of ASQ versus that of PSWSM for
Volume ■■, Number ■■, 2018
the assessment of liver fibrosis in chronic hepatitis B (CHB). METHODS Patients This study was approved by the Baoji Central Hospital Ethics Committee. Informed consent for all examinations was provided in writing by the patients or their family members. A total of 104 patients (aged 13– 64 y, mean age: 30.5 ± 11.1 y) with chronic viral hepatitis B, including 70 males (aged 13–64 y, mean age: 29.0 ± 10.5 y) and 34 females (aged 16–62 y, mean age: 32.8 ± 11.7 y), were recruited from the Department of Infectious Diseases of Baoji Central Hospital between February 2015 and May 2016. All patients underwent LB under ultrasonic guidance after ASQ and PSWSM, but did not receive antiviral therapy. The inclusion criteria were the presence of hepatitis B virus (HBV) deoxyribonucleic acid (DNA) in the sera and positivity for HBV surface antigen for at least 6 consecutive mo. The exclusion criteria were hepatitis B associated with a fatty liver, liver congestion, primary biliary hepatitis, autoimmune hepatitis and other types of viral hepatitis such as viral hepatitis C. The criterion for hepatitis C was positive for hepatitis C virus antibody. The criterion for the diagnosis of fatty liver was biopsy pathology revealing liver steatosis >5% (Kleiner et al. 2005). Thirty healthy volunteers, including 15 males and 15 females (aged 18–45 y, mean age of 32.4 ± 11.7 y), were enrolled as controls. All volunteers had normal levels of serum liver enzymes (alanine aminotransferase [ALT], aspartate aminotransferase [AST], γ-glutamine transferase [γ-GGT] and alkaline phosphatase [ALP]) and serum bilirubin (total bilirubin, direct bilirubin and indirect bilirubin). The volunteers were negative for hepatitis B virus surface antigen and hepatitis C virus antibody and had no history of diabetes, hyperlipidemia or alcoholism. Conventional ultrasonic scans did not reveal abnormalities of the liver. Acoustic structure quantification After fasting for more than 8 h, each subject underwent ASQ examination by a radiologist with 3 mo of experience in ASQ examinations using a Toshiba Aplio 500 ultrasound machine (Toshiba Medical Systems, Osaka, Japan) at a frequency of 5.0 MHz. The subject was placed in the left lateral decubitus position with the right hand placed behind the head. The probe was placed in the right intercostal area for scanning. Subjects were advised to hold their breath at maximum inspiration while the ASQ imaging was initiated. The frame rate for the collection was 8 frames per second (fps) for 3 s. Images were stored and copied to a computer. Segment V of the right anterior lobe of the liver was selected as the region of interest (ROI). The de-
ARTICLE IN PRESS ASQ vs. PSWSM for liver fibrosis ● J. Liu et al.
Fig. 1. Histogram indicating the Cm2 value as a red curve (microstructures) and a blue curve (macrostructures).
tection depth was 1.0 cm from the liver capsule with a size of approximately 3.0 × 3.0 cm, and major vessels, calcified lesions and the calculus of the intrahepatic duct in the liver were avoided. The raw echo signals of the liver tissue were statistically analyzed using ASQ software. ASQ is software that analyzes the statistical value of the acquired echo signal amplitude within a certain ROI. The measurement is compared with an estimated variance of normal or reference liver tissue. ASQ assesses liver fibrosis based on a statistical test (χ2-test) and on the statistical distribution of echo amplitudes measured in absolutely homogeneous liver parenchyma according to the Rayleigh probability density function (PDF). The χ2 value (Cm2) equals the variance-measured PDF divided by the variance of the Rayleigh PDF (Toyoda et al. 2009).
3
The result was displayed in a histogram using Cm2 as the horizontal axis and occurrence as the vertical axis. The histogram contains a red microstructure curve and a blue macrostructure curve (Fig. 1). The software places many fine ROIs in each manually set ROI. A Cm2 value is then calculated for each individual fine ROI. The Cm2 values of fine ROIs containing homogeneous parenchyma are assigned to the red microstructure curve. The Cm2 values of fine ROIs containing pathologic inhomogeneous tissue and focal inhomogeneities such as small blood vessels, biliary ducts and ligaments are assigned to the blue curve (Kaltenbach et al. 2014). ASQ is intended to facilitate the analysis of tissue homogeneity in the liver. Quantitative analysis generates a Cm2 histogram for an ROI. The higher the Cm2 value, the greater is the deviation from ideal tissue (Keller et al. 2015). ASQ-associated parameters—mode, average and standard deviation of the Cm2 values for the red and blue curves—were called red average (RA), red mode (RM), red standard deviation (RSD), blue average (BA), blue mode (BM), blue standard deviation (BSD) and the focal disturbance (FD). The FD is the ratio of the area under the red curve to the area under the blue curve (Figs. 2a, b). The values for RM, RA and RSD reflected the echo characteristics of normal liver tissue and tissue with mild liver fibrosis. The values for BM, BA and BSD in normal liver tissue and tissue with mild liver fibrosis correspond to the values expected for small blood vessels or biliary ducts. The values for BM, BA and BSD mainly reflected the inhomogeneous distribution of severe fibrosis and liver cirrhosis, which are even less similar to the Rayleigh distribution (Kuroda et al. 2012; Nam et al. 2016; Ricci et al. 2013).
Fig. 2. (a) Acoustic structure quantification (ASQ) examination in a normal subject. (b) ASQ examination in a patient with cirrhosis. Note that the seven parameters for the patient with cirrhosis were significantly higher than those for the normal subject. The corresponding ASQ color map illustrates an increase in the red color.
ARTICLE IN PRESS 4
Ultrasound in Medicine & Biology
Volume ■■, Number ■■, 2018
Fig. 3. (a) Point shear wave speed measurement (PSWSM) in an F3 patient. (b) PSWSM in a patient with cirrhosis. Note that the shear wave speed of the cirrhosis patient with cirrhosis was significantly higher than that of the F3 patient (2.23 m/s vs. 1.55 m/s).
Point shear wave speed measurement Point shear wave speed was measured using virtual touch tissue quantification (VTTQ) and a Siemens ACUSON S2000 ultrasound machine (Siemens Medical Solutions USA, Mountain View, CA, USA). VTTQ was assessed immediately after ASQ was completed. The assessment was performed by a radiologist with at least a half-year of experience with VTTQ examinations. Before measurement, subjects were advised to assume the left lateral decubitus position with the right hand placed behind the head. Standard scanning of the liver was performed. Segment V of the right anterior lobe of the liver was selected as the ROI, which was identical to that in the ASQ examination. Vessels and bile ducts were avoided. The ROI was approximately 1.0 cm below the liver capsule. During the measurement, subjects were advised to hold their breath at the end of inhalation. For each patient and volunteer, the same site was measured 10 times without repositioning the transducer, and the median was selected (Fig. 3a, b). Liver biopsy and histopathology Liver biopsies were conducted immediately after the PSWSM examinations using the same ultrasound machine to ensure that the puncture site was identical to those in the PSWSM and ASQ examinations. In addition, to reduce sampling error, PSWSM and ASQ were performed in the same intercostal space of the patient after maximum inhalation to ensure similar imaging planes for the same patient switching from ASQ on a Toshiba scanner to PSWSM on a Siemens scanner. The puncture routes avoided major vessels in the liver. An 18G core needle was inserted two or three times to collect tissues. The tissue strips had lengths of at least 1.5 cm and involved at least six portal areas.
Tissues the LBs were fixed with 10% neutral formaldehyde solution, embedded in paraffin, cut into serial sections and stained with standard hematoxylin and eosin (HE), Masson, and reticular fiber staining. The tissues were examined by an independent pathologist with rich clinical experience at our center. The histopathologic stage of liver fibrosis was determined using the METAVIR score (stages F0–F4), where F0 indicates no fibrosis, F1 indicates portal fibrosis but no fibrotic septa, F2 indicates portal fibrosis associated with a few fibrotic septa, F3 indicates numerous fibrotic septa but not cirrhosis and F4 indicates cirrhosis (Bedossa and Poynard 1996). Statistical analysis Statistical analyses were performed using SPSS Version 13.0 software (IBM, Armonk, NY, USA) and MedCalc Version 15.2.1 (MedCalc Program, Ostend, Belgium). Seven parameters obtained from the ASQ were analyzed using principal component analysis to formulate the integrative function (I). Using this integrative function (I), the functional value, called the integrative indicator (II), was calculated for every patient. The eigenvalue is the variance of the main function (Z). Each eigenvalue is the contribution of each eigenvector to the total variance. The relationships of liver fibrosis stage to II and SWS were analyzed using Spearman’s rank correlation. To generate predictive probability as an index of the combination of PSWSM and ASQ, binary logistic regression analyses were performed. The receiver operating characteristic (ROC) curve was used to assess the performance of each indicator in the diagnosis of liver fibrosis stage. The areas under the ROC curves (AUROCs) were compared with Z-tests. Cutoff values were selected using the Youden index (sum of sensitivity and specificity), and the corresponding sensitivities and specificities were calculated. Finally,
ARTICLE IN PRESS ASQ vs. PSWSM for liver fibrosis ● J. Liu et al.
5
Table 1. Subject characteristics Characteristic
F0 (n = 30 + 3)
F1 (n = 33)
F2 (n = 19)
F3 (n = 29)
F4 (n = 20)
Sex (male/female) Age (y) ALT (IU/L) AST (IU/L) GGT Total cholesterol Platelets (×109/L)
15/15 33.4 ± 5.8 17.7 ± 3.7 18.0 ± 7.7 21.0 ± 8.5 3.25 ± 0.65 204.9 ± 42.1
17/16 29.8 ± 11.5 44.3 ± 115.2 30.4 ± 44.4 23.7 ± 52.8 3.68 ± 0.77 182.3 ± 51.6
11/8 31.7 ± 12.8 57.1 ± 128.0 41.2 ± 84.0 41.5 ± 105.1 3.76 ± 0.87 176.7 ± 55.8
15/14 36.7 ± 11.1 63.1 ± 48.6 82.8 ± 126.8 254.5 ± 648.8 3.93 ± 1.46 178.5 ± 54.5
12/8 41.2 ± 8.8 47.8 ± 32.3 54.2 ± 26.6 70.3 ± 45.0 3.49 ± 0.69 103.8 ± 36.5
ALT = alanine aminotransferase, AST = aspartate aminotransferase, GGT = γ-glutamyl transferase.
Table 2. Seven parameters of acoustic structure quantification Parameter RM RA RSD BM BA BSD FD
F0
F1
F2
F3
F4
111.90 ± 4.48 115.30 ± 3.76 17.40 ± 2.56 128.90 ± 7.95 138.10 ± 10.50 29.10 ± 10.56 0.23 ± 0.08
112.72 ± 5.51 116.30 ± 4.78 17.00 ± 2.03 126.97 ± 7.34 140.24 ± 10.68 32.27 ± 13.31 0.31 ± 0.13
113.89 ± 8.69 117.28 ± 6.62 18.28 ± 1.74 128.72 ± 8.11 145.50 ± 12.20 37.33 ± 12.83 0.48 ± 0.24
120.00 ± 4.86 123.17 ± 3.34 21.31 ± 1.59 133.69 ± 6.61 146.62 ± 5.94 33.86 ± 11.47 0.61 ± 0.18
122.65 ± 5.36 125.80 ± 3.50 19.50 ± 1.82 138.60 ± 7.09 152.40 ± 9.20 35.15 ± 12.15 0.74 ± 0.20
RM = red curve mode, RA = red curve average, RSD = red curve standard deviation, BA = blue curve average, BM = blue curve mode, BSD = blue curve standard deviation, FD = focal distribution.
p values < 0.05 were considered to indicate statistical significance. RESULTS Subject characteristics The pathologic results from the LBs of 104 cases revealed 3 (2.9%) cases at the F0 stage, 33 (31.7%) at the F1 stage, 19 (18.3%) at the F2 stage, 29 (27.9%) at the F3 stage and 20 (19.2%) at the F4 stage. Among these cases, 68 (65.4%) had significant fibrosis (F ≥ 2). General data and laboratory indicators for the patients and volunteers are provided in Table 1. Integrative indicators obtained from the ASQ parameters Results for the seven parameters obtained from the ASQ are provided in Table 2. From the seven parameters of the ASQ tool, RSD and BSD were not correlated with stage of liver fibrosis. The remaining five parameters were used to build an integrative function for the assessment of liver fibrosis stage. Based on the results and practical requests, the anterior four main functions (Z) were selected to formulate the integrative function (I) and had eigenvalues of 3.215, 1.719, 0.956 and 0.598, respectively. The contributions of Z1 to Z4 were 45.9%, 24.5%, 13.6% and 8.5%, respectively, and the cumulative contribution was 92.4%. The anterior four main functions (Z) and integrative function (I) are expressed as follows: Z1 = 0.814 RM + 0.893 RA + 0.719 BM + 0.665 BM + 0.818 FD
Z2 = -0.426 RM – 0.310 RA – 0.152 BM + 0.663 BM – 0.056 FD Z2 = -0.029 RM – 0.029 RA + 0.152 BM – 0.121 BM – 0.236 FD Z4 = 0.177 RM + 0.136 RA – 0.648 BM – 0.128 BM + 0.272 FD I = 45.9% Z1 + 24.5% Z2 + 13.6% Z3 + 8.5% Z4
Analysis of the relationship between non-invasive parameters and stage of liver fibrosis The integrative indicators obtained from ASQ and SWS from PSWSM were closely correlated with the stages of liver fibrosis (Figs. 4 and 5, Table 3). Among the seven parameters from ASQ, FD was the parameter that was most strongly correlated with stage of liver fibrosis (r = 0.583, p < 0.001), whereas RSD and BSD were not correlated with stage of liver fibrosis (r = 0.140, p = 0.127; r = 0.147, p = 0.109).
Table 3. Spearman’s coefficients for the non-invasive parameters and liver fibrosis stages Liver fibrosis stage METAVIR scores
Non-invasive parameter
Spearman coefficient
p Value
PSWSM ASQ
0.682 0.651
<0.001 <0.001
PSWSM = point shear-wave speed measurement, ASQ = acoustic structure quantification.
ARTICLE IN PRESS 6
Ultrasound in Medicine & Biology
Volume ■■, Number ■■, 2018
Fig. 4. Scattergraph revealing the correlation between the integrative indicator from acoustic structure quantification (ASQ) and the histologic results from the liver biopsies.
ASQ versus PSWSM for the diagnosis of liver fibrosis The AUROCs, cutoff values, sensitivities and specificities for the diagnosis and prediction of liver fibrosis stage using ASQ and PSWSM are provided in Table 4. For the diagnosis of mild liver fibrosis (F ≥ 1), ASQ (0.705) had a significantly lower AUROC than PSWSM (0.854, p = 0.045) (Fig. 6). For the diagnosis of significant liver fibrosis (F ≥ 2), ASQ (0.813) had a greater AUROC than PSWSM (0.743), although the
difference was not significant (p = 0.115). ASQ was comparable to PSWSM for the diagnosis of severe liver fibrosis (F ≥ 3) with AUROCs of 0.839 and 0.857, respectively. For the diagnosis of liver cirrhosis (F4), ASQ was lower for PSWSM, with AUROCs of 0.874 and 0.971, respectively (p = 0.016) (Fig. 7). The AUROC of PSWSM combined with ASQ for the diagnosis of severe liver fibrosis was greater than that of PSWSM alone (p = 0.020) (Fig. 8).
Fig. 5. Scattergraph revealing the correlation between shear wave speed (SWS) by point shear wave speed measurement (PSWSM) and the histologic results from the liver biopsies.
ARTICLE IN PRESS ASQ vs. PSWSM for liver fibrosis ● J. Liu et al.
7
Table 4. Cutoffs and AUROCs of ASQ and PSWSM for assessing liver fibrosis stage Non-invasive test ASQ ≥F1 ≥F2 ≥F3 F4 PSWSM ≥F1 ≥F2 ≥F3 F4 ASQ + PSWSM ≥F1 ≥F2 ≥F3 F4
AUROC (95% CI)
Cutoff
Sensitivity (%)
Specificity (%)
0.705 (0.625–0.794) 0.813 (0.732–0.879) 0.855 (0.778–0.913) 0.873 (0.822–0.914)
181.13 181.54 185.44 191.91
54.0 71.6 81.3 70.0
90.0 86.8 78.9 93.2
0.854 (0.777–0.911) 0.742 (0.678–0.809) 0.890 (0.819–0.913) 0.969 (0.937–0.988)
1.28 1.35 1.48 1.74
65.0 68.7 87.5 100.0
95.0 71.7 76.1 93.7
0.877 (0.805–0.930) 0.836 (0.764–0.908) 0.936 (0.880–0.975) 0.970 (0.939–0.988)
0.804 0.479 0.446 0.168
80.0 82.1 83.3 95.0
85.0 75.5 88.7 95.1
PSWSM = point shear-wave speed measurement, ASQ = acoustic structure quantification, AUROC = area under the receiver operating characteristic curve, 95% CI = 95% confidence interval.
DISCUSSION Currently, ultrasound-based non-invasive methods are the mainstay for the assessment of liver fibrosis. However, the performance of these methods in the assessment of liver fibrosis remains to be further validated. This study found that the correlation between METAVIR stage and PSWSM parameters was slightly higher than that obtained with ASQ; the Spearman coefficients were 0.682 and 0.651, respectively. Liver fibrosis is caused by excessive deposition of matrix, especially collagen, in the liver. Tang et al. (2015) suggested that increasing SWS values are related to in-
creased collagen concentration during fibrosis. ASQ parameters are derived from statistical information on scattering sources. Given that collagen is an important source of scattering, these two techniques might be directly measuring the same tissue changes; therefore, the correlations with fibrosis would not be expected to be markedly different. Point shear wave speed measurement using VTTQ is easy to perform, low in cost, free of ionizing radiation and guided by 2-D images (Son et al. 2012) and has been in clinical use for 10 y. Sporea et al. (2012) used PSWSM to assess liver fibrosis stage in 53 cases with viral
Fig. 6. Receiver operating characteristic (ROC) curves revealing performance in the diagnosis of liver fibrosis (F ≥ 1) by acoustic structure quantification (ASQ), point shear wave speed measurement (PSWSM) and their combination. Diagonal segments are produced by ties.
ARTICLE IN PRESS 8
Ultrasound in Medicine & Biology
Volume ■■, Number ■■, 2018
Fig. 7. Receiver operating characteristic (ROC) curves revealing performance in the diagnosis of cirrhosis (F4) by acoustic structure quantification (ASQ), point shear wave speed measurement (PSWSM) and their combination. Diagonal segments are produced by ties.
hepatitis B and 107 cases with viral hepatitis C and found that PSWSM obtained comparable values in the assessment of liver fibrosis stage in viral hepatitis B and viral hepatitis C. A multicenter study reported by Friedrich-Rust et al. (2013) indicated no difference in the diagnostic values of PSWSM and TE in the assessment of liver fibrosis at
F ≥ 2, F ≥ 3 or F4 and the AUROCs for PSWSM in the diagnosis of liver fibrosis were 0.75 for F ≥ 2, 0.93 for F ≥ 3 and 0.97 for F4. The AUROCs (0.74 for F ≥ 2, 0.89 for F ≥ 3 and 0.97 for F4) for diagnosis of liver fibrosis in the present study were comparable to those from the study of Friedrich-Rust et al. (2013). In the present study, PSWSM
Fig. 8. Receiver operating characteristic (ROC) curves revealing performance in the diagnosis of liver fibrosis (F ≥ 3) by acoustic structure quantification (ASQ), point shear wave speed measurement (PSWSM) and their combination.
ARTICLE IN PRESS ASQ vs. PSWSM for liver fibrosis ● J. Liu et al.
was performed during a deep inhalation state. A previous study provided proof that deep inspiration significantly modulates spleen stiffness (Karlas et al. 2014). However, liver SWSs in deep inspiration, in deep expiration and during the Valsalva maneuver did not differ (Goertz et al. 2012). In the case of ASQ, Krämer et al. (2014) found that the stages of liver fibrosis are not correlated with the average or peak Cm2 values obtained from the ASQ software and suggested that ASQ was not accurate in the assessment of liver fibrosis stage. However, Nam et al. (2016) found that the stages of liver fibrosis were well correlated with the mode, average and FD obtained from the ASQ software and most strongly correlated with mode. Additionally, Huang et al. (2016) found a strong correlation between FD and stage of liver fibrosis. In the present study, the parameters obtained from the ASQ software—RM, RA, BM, BA and FD—were well correlated with the stage of liver fibrosis (p < 0.001). FD was most strongly correlated with stage of liver fibrosis (r = 0.583), followed by RM and RA. Liver fibrosis stages were not correlated with RSD or BSD. We used principal component analysis to build integrative indicators that included RM, RA, BM, BD and FD to assess liver fibrosis, which resulted in a rank correlation coefficient of 0.651 with liver fibrosis (Table 3). Ricci et al. (2013) found that ASQ did not attain the same level of diagnostic value as FibroScan for the diagnosis of liver fibrosis stage in viral hepatitis B and C. The AUROCs for ASQ and FibroScan were 0.71 and 0.77, respectively, for the diagnosis of liver fibrosis (F ≥ 1) and 0.94 and 0.97, respectively, for the diagnosis of liver cirrhosis (F4). We found that ASQ produced a lower AUROC than PSWSM in terms of the diagnosis of liver fibrosis (F ≥ 1) and liver cirrhosis (F4). The reasons for the difference may be as follows: In patients in stage F1, the difference caused by portal fibrosis as a scattering source was weaker than that of liver tissue stiffness caused by portal fibrosis. Numerous fibrotic tissues can be used as a remarkable source of scattering, but F4 is generally associated with severe liver inflammation and a combination of liver fibrosis and severe inflammation, which can lead to increased liver stiffness. However, it was comparable to PSWSM in the diagnoses of significant liver fibrosis (F ≥ 2) and severe liver fibrosis (F ≥ 3), but the combination of two different techniques improved the accuracy in diagnosis of severe liver fibrosis (F ≥ 3). The AUROC was higher in our research than in the study of Ricci et al. (2013) (0.87 vs. 0.77) for the diagnosis of liver cirrhosis, a finding that might be related to the diagnosis achieved by combining relevant parameters from ASQ in the present study. In addition, Ricci et al.’s study included liver fibrosis of viral hepatitis C as well as viral hepatitis B. Liver cirrhosis of hepatitis C is usually micronodular liver cirrhosis, similar to normal liver
9
in that scattering noise is represented by an echo texture pattern. Therefore, ASQ might be weak for the prediction of liver fibrosis stage in patients with hepatitis C. There are some limitations to the present study. First, we did not assess the repeatability or reproducibility of ASQ for the assessment of liver fibrosis stage because ASQ is a novel technology for assessing liver fibrosis stage; we will evaluate the repeatability and reproducibility of ASQ in future studies. Second, healthy patients were included as controls because only a few (three patients) had F0 liver fibrosis in the present study, and pathologic diagnoses were not obtained from the control group for ethical reasons. Third, the test was carried out only in patients with HBV; we need to compare the performance of PSWSM and ASQ in the diagnosis of liver fibrosis in hepatitis C in future studies. Fourth, because of the complexity of their calculation, integrative indicators cannot be widely accepted in clinical practice. In conclusion, ASQ is a promising ultrasonic technology for the assessment of liver fibrosis and is comparable to PSWSM in terms of the diagnosis of significant liver fibrosis and severe liver fibrosis. However, PSWSM was better than ASQ for the diagnosis of mild liver fibrosis (F ≥ 1) and liver cirrhosis (F4), and the combination of PSWSM and ASQ improves the accuracy of diagnosis of severe liver fibrosis (F ≥ 3). Acknowledgments—We express our deepest gratitude to the Baoji City Health and Family Planning Commission for financial support. We thank Dr. Bi Yuxue (School of Public Health, Xi’an Jiaotong University Health Science Center, China) for providing helpful contributions to the statistical analysis.
REFERENCES Bedossa P, Poynard T. An algorithm for the grading of activity in chronic hepatitis C. The METAVIR Cooperative Study Group. Hepatology 1996;24:289–293. Bota S, Herkner H, Sporea I, Salzl P, Sirli R, Neghina AM, Peck-Radosavljevic M. Meta-analysis: ARFI elastography versus transient elastography for the evaluation of liver fibrosis. Liver Int 2013; 33:1138–1147. Bota S, Sporea I, Sirli R, Popescu A, Danila M, Jurchis A, Gradinaru-Tascau O. Factors associated with the impossibility to obtain reliable liver stiffness measurements by means of acoustic radiation force impulse (ARFI) elastography–—Analysis of a cohort of 1,031 subjects. Eur J Radiol 2014;83:268–272. Bravo AA, Sheth SG, Chopra S. Liver biopsy. N Engl J Med 2001;344: 495–500. Castéra L, Foucher J, Bernard PH, Carvalho F, Allaix D, Merrouche W, Couzigou P, de Lédinghen V. Pitfalls of liver stiffness measurement: A 5-years prospective study of 13,369 examinations. Hepatology 2010; 51:828–835. Duarte-Rojo A, Altamirano JT, Feld JJ. Noninvasive markers of fibrosis: Key concepts for improving accuracy in daily clinical practices. Ann Hepatol 2012;11:426–438. El-Serag HB. Hepatocellular carcinoma and hepatitis C in the United States. Hepatology 2002;36:74–83. European Association for the Study of the Liver [EASL]. EASL clinical practice guidelines: Management of chronic hepatitis B virus infection. J Hepatol 2012;57:167–185.
ARTICLE IN PRESS 10
Ultrasound in Medicine & Biology
Filingeri V, Sforza D, Tisone G. Complications and risk factors of a large series of percutaneous liver biopsies in patients with liver transplantation or liver disease. Eur Rev Med Pharmacol Sci 2015;19:1621– 1629. Friedrich-Rust M, Buggisch P, de Knegt RJ, Dries V, Shi Y, Matschenz K, Schneider MD, Herrmann E, Petersen J, Schulze F, Zeuzem S, Sarrazin C. Acoustic radiation force impulse imaging for noninvasive assessment of liver fibrosis in chronic hepatitis B. J Viral Hepat 2013;20:240–247. Gebo KA, Herlong HF, Torbenson MS, Jenckes MW, Chander G, Ghanem KG, El-Kamary SS, Sulkowski M, Bass EB. Role of liver biopsy in management of chronic hepatitis C: A systematic review. Hepatology 2002;36:161–172. Goertz RS, Egger C, Neurath MF, Strobel D. Impact of food intake, ultrasound transducer, maneuvers and body position on acoustic radiation force impulse (ARFI) elastometry of the liver. Ultraschall Med 2012; 33:380–385. Huang Y, Wang Z, Liao B, Liang JY, Zhou LY, Wang F, Li W, Liu JY, Xie XY, Lu MD, Liu GJ, Wang W. Assessment of liver fibrosis in chronic hepatitis B using acoustic structure quantification: Quantitative morphological ultrasound. Eur Radiol 2016;26:2344–2351. Kaltenbach TE, Gruener B, Akinli AS, Kratzer W, Oeztuerk S, Mason RA, Haenle MM, Graeter T. Acoustic structure quantification (ASQ): A new tool in sonographic examination of liver lesions alveolar echinococcosis. J Med Ultrason (2001) 2014;41:445–453. Karlas T, Lindner F, Tröltzsch M, Keim V. Assessment of spleen stiffness using acoustic radiation force impulse imaging (ARFI): Definition of examination standards and impact of breathing maneuvers. Ultraschall Med 2014;35:38–43. Keller J, Kaltenbach TE, Haenle MM, Oeztuerk S, Graeter T, Mason RA, Seufferlein T, Kratzer W. Comparison of acoustic structure quantification (ASQ), shear wave elastography and histology in patients with diffuse hepatopathies. BMC Med Imaging 2015;15:58. Kleiner DE, Brunt EM, Van Natta M, Behling C, Contos MJ, Cummings OW, Ferrell LD, Liu YC, Torbenson MS, Unalp-Arida A, Yeh M, McCullough AJ, Sanyal AJ, Nonalcoholic Steatohepatitis Clinical Research Network. Design and validation of a histological scoring system for nonalcoholic fatty liver disease. Hepatology 2005;41: 1313–1321. Krämer C, Jaspers N, Nierhoff D, Kuhr K, Bowe A, Goeser T, Michels G. Acoustic structure quantification ultrasound software proves imprecise in assessing liver fibrosis or cirrhosis in parenchymal liver diseases. Ultrasound Med Biol 2014;40:2811–2818. Kuroda H, Kakisaka K, Kamiyama N, Oikawa T, Onodera M, Sawara K, Oikawa K, Endo R, Takikawa Y, Suzuki K. Non-invasive determination of hepatic steatosis by acoustic structure quantification from ultrasound echo amplitude. World J Gastroenterol 2012;18:3889– 3895.
Volume ■■, Number ■■, 2018 Nam TH, Park BJ, Sung DJ, Kim MJ, Han NY, Sim KC, Lee JH, Cho SB. Non-invasive assessment of liver fibrosis using acoustic structure quantification: Comparison with transient elastography in chronic viral hepatitis. Abdom Radiol (NY) 2016;41:239–247. Nightingale K, McAleavey S, Trahey G. Shear-wave generation using acoustic radiation force: In vivo and ex vivo results. Ultrasound Med Biol 2003;29:1715–1723. Ramachandran P, Iredale JP. Liver fibrosis: A bidirectional model of fibrogenesis and resolution. QJM 2012;105:813–817. Regev A, Berho M, Jeffers LJ, Milikowski C, Molina EG, Pyrsopoulos NT, Feng ZZ, Reddy KR, Schiff ER. Sampling error and intraobserver variation in liver biopsy in patients with chronic HCV infection. Am J Gastroenterol 2002;97:2614–2618. Ricci P, Marigliano C, Cantisani V, Porfiri A, Marcantonio A, Lodise P, D’Ambrosio U, Labbadia G, Maggini E, Mancuso E, Panzironi G, Di Segni M, Furlan C, Masciangelo R, Taliani G. Ultrasound evaluation of liver fibrosis: Preliminary experience with acoustic structure quantification (ASQ) software. Radiol Med 2013;118:995– 1010. Sebastiani G. Non-invasive assessment of liver fibrosis in chronic liver diseases: Implementation in clinical practice and decisional algorithms. World J Gastroenterol 2009;15:2190–2203. Shiina T, Nightingale KR, Palmeri ML, Hall TJ, Bamber JC, Barr RG, Castera L, Choi BI, Chou YH, Cosgrove D, Dietrich CF. WFUMB guidelines and recommendations for clinical use of ultrasound elastography: Part 1. Basic principles and terminology. Ultrasound Med Biol 2015;41:1126–1147. Son CY, Kim SU, Han WK, Choi GH, Park H, Yang SC, Choi JS, Park JY, Kim DY, Ahn SH, Chon CY, Han KH. Normal liver elasticity values using acoustic radiation force impulse imaging: A prospective study in healthy living liver and kidney donors. J Gastroenterol Hepatol 2012;27:130–136. Sporea I, Sirli R, Bota S, Popescu A, Sendroiu M, Jurchis A. Comparative study concerning the value of acoustic radiation force impulse elastography (ARFI) in comparison with transient elastography (TE) for the assessment of liver fibrosis in patients with chronic hepatitis B and C. Ultrasound Med Biol 2012;38:1310–1316. Tang A, Cloutier G, Szeverenyi NM, Sirlin CB. Ultrasound elastography and MR elastography for assessing liver fibrosis: Part 2. Diagnostic performance, confounders, and future directions. AJR Am J Roentgenol 2015;205:33–40. Toyoda H, Kumada T, Kamiyama N, Shiraki K, Takase K, Yamaguchi T, Hachiya H. B-Mode ultrasound with algorithm based on statistical analysis of signals: Evaluation of liver fibrosis in patients with chronic hepatitis C. AJR Am J Roentgenol 2009;193:1037–1043. Wang FS, Fan JG, Zhang Z, Gao B, Wang HY. The global burden of liver disease: The major impact of China. Hepatology 2014;60: 2099–2108.