A comparison of hepatic steatosis index, controlled attenuation parameter and ultrasound as noninvasive diagnostic tools for steatosis in chronic hepatitis B

Accepted Manuscript Title: A comparison of hepatic steatosis index, controlled attenuation parameter and ultrasound as noninvasive diagnostic tools for steatosis in chronic hepatitis B Authors: Liang Xu, Wei Lu, Ping Li, Feng Shen, Yu-Qiang Mi, Jian-Gao Fan PII: DOI: Reference:

S1590-8658(17)30793-4 http://dx.doi.org/doi:10.1016/j.dld.2017.03.013 YDLD 3406

To appear in:

Digestive and Liver Disease

Received date: Accepted date:

5-12-2016 21-3-2017

Please cite this article as: Xu Liang, Lu Wei, Li Ping, Shen Feng, Mi Yu-Qiang, Fan Jian-Gao.A comparison of hepatic steatosis index, controlled attenuation parameter and ultrasound as noninvasive diagnostic tools for steatosis in chronic hepatitis B.Digestive and Liver Disease http://dx.doi.org/10.1016/j.dld.2017.03.013 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

A comparison of hepatic steatosis index, controlled attenuation parameter and ultrasound as noninvasive diagnostic tools for steatosis in chronic hepatitis B

Liang Xua,b,d, Wei Lua,c,d,*, Ping Lib,d, Feng Shene, Yu-Qiang Mib,d,*, Jian-Gao Fane,*.

a

First Center for Clinical College, Tianjin Medical University, Tianjin, 300192, China

bDepartment

of Hepatology, Tianjin Second People’s Hospital, Tianjin, 300192, China

cTianjin

First Center Hospital, Tianjin, 300192, China

dTianjin

Research Institute of Liver Diseases, Tianjin, 300192, China

eDepartment

of Gastroenterology, Xinhua Hospital affiliated to Shanghai Jiaotong University School of

Medicine, Shanghai 200092, China

* Correspondence to: Wei Lu, MD, First Center for Clinical College, Tianjin Medical University; Tianjin First Center Hospital; Tianjin Research Institute of Liver Diseases, Tianjin, 300192, China. Telephone: +86-22-27468003. Fax: +86-22-27468682. E-mail address: [email protected] Yu-qiang Mi, MD, Department of Hepatology, Tianjin Second People’s Hospital; Tianjin Research Institute of Liver Diseases, Tianjin, 300192, China. Telephone: +86-22-27468003. Fax: +86-22-27468682. E-mail address: [email protected] Jian-Gao Fan, MD, PhD, Department of Gastroenterology, Xinhua Hospital, Shanghai Jiaotong University School of Medicine, Shanghai 200092, China. Telephone: +86-21-25077340. Fax: +86-21-25077340. E-mail address: [email protected].

Abstract Aims: To evaluate the value of noninvasive tools for diagnosis of hepatic steatosis in patients

with chronic hepatitis B (CHB). Methods: Consecutive treatment-naïve patients with CHB with body mass index less than 30 kg/m2 who underwent liver biopsy, ultrasound and FibroScan® were enrolled. The diagnostic performance of controlled attenuation parameter (CAP), hepatic steatosis index (HSI) and ultrasound for hepatic steatosis compared with liver biopsy was assessed. The areas under receiver operating characteristics curves (AUROCs) were calculated to determine the diagnostic efficacy, with comparisons using the DeLong test. Results: CAP and HSI accuracies were significantly higher than that of ultrasound to detect patients with biopsy-proven mild steatosis (S1, 65.3%, 56.5%, respectively, vs. 17.7%, χ2 = 46.305, 31.736, both P < 0.05 ）and moderate-severe (S2-3) steatosis (92.3%, 100%, respectively, vs. 53.8%, χ2 =4.887, 7.800, P=0.037, 0.007, respectively). Both CAP and HSI had lower underestimation rates of steatosis grade than ultrasound (12%, 14.8%, respectively, vs. 29.5%, χ2 = 9.765, 6.452; P < 0.05 for both), but they exhibited higher overestimation rates (30.5%, 38.2%, respectively, vs. 12.4%, χ2=39.222, 70.986; both P<0.05). The AUROCs of CAP and HSI were 0.780 (95% confidence intervals [CIs] 0.735-0.822) and 0.655 (95%CI 0.604-0.704) for S ≥ 1, 0.932 (95%CI 0.902-0.956) and 0.755 (95%CI 0.707-0.799) for S ≥ 2, 0.990 (95%CI 0.974-0.998) and 0.786 (95% CI 0.740-0.827) for S3, respectively. Conclusion: CAP might be more accurate for detecting hepatic steatosis than HSI and ultrasound in patients with CHB, but further studies are needed to reduce the overestimation rates. Keywords: Fatty liver disease; Assessment; Liver biopsy; Transient elastography

INTRODUCTION

Hepatic steatosis is a frequent finding in patients with chronic hepatitis B (CHB), and its prevalence is increasing with the worldwide epidemic of obesity and type 2 diabetes mellitus [1–4]. Recent studies highlight the impact of hepatic steatosis on accelerating disease progression to liver fibrosis and ultimately cirrhosis [5-7] and reducing the efficacy of antiviral treatment in CHB[8]. Therefore, detection and quantification of hepatic steatosis have increasing relevance both for clinical studies and routine medical care in patients with CHB [9]. Although liver biopsy is traditionally considered the gold standard for steatosis grading, its use has several limitations, including sampling bias, intra- or inter- observer sampling variability and the potential for severe complications [10]. Furthermore, steatosis severity may change within weeks of therapeutic intervention and therefore cannot be sufficiently monitored by repetitive invasive procedures [11]. Imaging techniques provide potential alternatives for noninvasive steatosis characterization because fat deposits alter the physical properties of liver tissue [9]. Magnetic resonance (MR)-based techniques assess triglyceridespecific signal intensity and represent sensitive approaches for steatosis detection [12] but are not suitable as point-of-care methods due to high costs and limited comparability between different MR techniques [13,14]. B-mode ultrasound displays a bright echo-pattern in steatotic hepatic tissue and is widely used as a first-line assessment for screening of fatty liver, but it is imprecise in estimating mild steatosis [9,15]. These limitations may be overcome by the controlled attenuation parameter (CAP) software, which has been recently developed to quantify ultrasound attenuation during measurement of liver stiffness vibration-controlled elastography (transient elastography, TE; FibroScan) [16]. CAP measurement is an easy and

fast examination providing a numerical value expressed in dB/m, which correlates with the histologic degree of steatosis [17]. According to available published data, CAP exhibits good diagnostic performance for steatosis evaluation in chronic viral hepatitis [18,19] and in multi-etiology cohorts [20-25]. The hepatic steatosis index (HSI) is also a simple screening tool for hepatic steatosis and is purported to be more accurate than ultrasound [26]. Nonetheless, no data published validate its application in patients with CHB. Therefore, the study’s first objective was to compare CAP’s diagnostic accuracy with HSI for hepatic steatosis quantification in patients with CHB using liver biopsy as a reference method. The secondary objective was to evaluate the steatosis detection and misdiagnosis rates of CAP, HSI and ultrasound compared with liver biopsy.

MATERIALS AND METHODS Patients Among all patients who underwent a series of laboratory tests, liver biopsy, ultrasound and FibroScan at Tianjin Second People's Hospital, China, between July 2012 and April 2014, those who were positive for serum hepatitis B surface antigen (HBsAg) and hepatitis B virus (HBV) DNA for at least 6 months and did not undergo antiviral treatment were prospectively enrolled in this study. The exclusion criteria were as follows: (1) co-infection with other hepatitis virus and HIV; (2) daily alcohol consumption > 20 g for women and > 30 g for men; (3) any type of positive autoantibody above 1:160; (4) age < 18 years; (5) presence of hepatocellular carcinoma; (6) immune suppressive treatment within 1 year; (7) an interval between liver biopsy, ultrasound and FibroScan examination longer than 2 weeks; (8) liver

biopsy specimen < 15 mm length or < 0.8 mm diameter; and (9) body mass index (BMI) ≥ 30 kg/m2. The study protocol was approved by the hospital Ethics Committee and was conducted in accordance with the Declaration of Helsinki. All participants provided written informed consent before the study. Clinical, laboratory data collection Demographic information, such as age, sex, history of diseases (such as arterial hypertension, diabetes mellitus and cardiovascular disease), daily alcohol consumption, smoking status and drug use were recorded from patient interviews before liver biopsy. The anthropometric measurements, including body weight (kilogram), height (meter) and BMI were obtained. Fasting blood samples were collected on the same day or within 3 days before and after liver biopsy. Serum alanine aminotransferase (ALT), aspartate aminotransferase (AST), gamma-glutamyl transferase, alkaline phosphatase, total bilirubin, fasting plasma glucose, total cholesterol and triglycerides were measured by a Hitachi 7600–110 automatic analyzer (Hitachi Co., Tokyo, Japan). HSI was calculated according to the following formula: 8 × (ALT/AST ratio) + BMI (+2, if female; +2, if diabetes mellitus) [26]. Serum HBsAg, HBV e antigen (HBeAg) and HBV DNA were measured using an electrochemiluminescence assay (Roche COBAS e 411, Sandhofer Stasse 116, 68305 Mannheim, Germany) and quantitative PCR (Roche, light Cycler® 480 II /96, Rotkreuz, Switzerland), respectively. Ultrasound examination Abdominal ultrasound was performed with a Philips ultrasound machine (Model No. IU22, 22100 Bothell Everett Highway Bothell, WA, USA), by one senior radiologist (Zhou), using a C5-1 scanner with a 5 MHz convex probe. Hepatic steatosis was quantified during

examination according to Hamaguchi’s ultrasonographic score, from 0 to 6 points, based on the hepatorenal contrast, bright hepatic echoes, deep attenuation and vessel blurring [27,28]. A score < 2 was defined as without steatosis, a score ≥ 2 and < 4 was defined as mild steatosis, a score ≥ 4 and < 6 was defined as moderate steatosis and the score = 6 was defined as severe steatosis. Measurement of liver stiffness values and CAP All patients were examined using FibroScan-502 (Echosens, Paris, France) with a 3.5 MHz M probe, which measures liver stiffness 25-65 mm underneath the skin surface [28-30]. The results of liver stiffness measurements (LSMs) were expressed as kPa. Examinations were considered unreliable if fewer than 10 valid measurements were obtained and/or the ratio of the interquartile range (IQR) over the median of 10 measurements (IQR/M) of LSM and CAP was 0.30, and skin capsular distance > 25 mm. The CAP values were expressed in dB/m and were provided only if the LSM examination was successfully performed. One senior operator (Xu) was blinded to the patients' clinical and histologic data. Liver biopsy All patients underwent a percutaneous ultrasound-guided liver biopsy. Each liver specimen was fixed in formalin, embedded in paraffin and stained with hematoxylin-eosin, Masson’s trichrome and collagen. Two experienced hepatic pathologists (Shi and Liu) blinded to the clinical data independently reviewed the histologic findings. Consensus was reached in case of disagreement. The Knodell histologic activity index was used to describe the hepatocellular necro-inflammatory activity [31]. Liver fibrosis was semi-quantitatively assessed according to the Ishak system and was graded from stage 0 to stage 6 [32]. Hepatic

steatosis was graded from 0 to 3 based on the number of hepatocytes with steatosis at histology: S0: < 5%; S1: 5-33%; S2: 34-66%; S3: > 66% [33]. Statistical analysis All data were presented as number (%) and median (range) values. A Mann-Whitney test (two-group comparison) or Kruskal-Wallis test (more than two group comparison) was used for continuous variables. Multivariate analyses of variance were performed to study interactions between the histologic steatosis grades and CAP. The distribution of CAP values and HSI values for each degree of histologic steatosis was evaluated using box plots. Comparisons of the detection rates, coincidence rates and discordance rates of hepatic steatosis in CAP, HSI and ultrasound according to liver biopsy were made using the chi-squared or Fisher’s exact test. The diagnostic performances of CAP and HSI were analyzed by computing the AUROCs and their 95% CIs. The optimal diagnostic cutoff value for each degree of histologic steatosis was found by maximizing the Youden Index. For each cutoff, a corresponding positive predictive value (PPV), negative predictive value (NPV) and positive and negative likelihood ratios (LR+ and LR-, respectively) were also calculated. The diagnostic performances between two measures were compared using the DeLong test. Univariate and multivariate stepwise regression analysis was performed to identify independent factors of discordance between CAP, HSI, ultrasound, and liver biopsy. Correlations between variables were described using the Pearson correlation coefficients(r). A P value of < 0.05 on a two-tailed test was considered significant. Statistical analyses were performed using SPSS 19.0 (SPSS, Chicago, USA) and MedCalc 9.3 (MedCalc Software, Mariakerke, Belgium).

RESULTS Patients’ characteristics and histological findings During the study period, 171 patients with biopsy-proven CHB were excluded for the following reasons: (1) presence of hepatocellular carcinoma (n = 95); (2) excessive alcohol consumption (n = 19); (3) positive serum autoantibody above 1:160 (n = 7); (4) serum ALT level exceeding 10 times the upper limit of normal (ULN) or total bilirubin level exceeding 2.5 times the ULN (n = 11); (5) immunosuppressive treatment within 1 year (n = 1); (6) liver biopsy specimen < 15 mm long or < 0.8 mm diameter (n = 1); (7) BMI

≥ 30.0 kg/m2 or skin

capsular distance > 25 mm (n = 31). Finally, 366 naïve patients with CHB who underwent liver biopsy, FibroScan and ultrasound were enrolled in this study. The success rate of FibroScan was 100%, but five patients underwent the procedure twice in order to have a reliable LSM and CAP values for all. The indication of liver biopsy included abnormal liver function tests (n=162), LSM >10 kPa (n=148) and both (n=56). All liver biopsy samples were ≥16 mm and included at least six portal tracts. The majority (224/366) of the study subjects were male, the median age was 37.6 years (18-72 years old) and the median BMI was 23.4 (17.2-30.0) kg/m2. Of them, 112 patients (30.6%) were overweight (BMI 24.0-28.0 kg/m2), 27 patients (7.4%) were obese (BMI 28.0-30 kg/m2) and 15 patients had diabetes mellitus. Of these, 68.1% of patients were serum HBeAg positive (254/366), with a median serum log10 (HBV DNA titer) of 5.3 (1-9) copies/ml.

Prevalence of biopsy-proven hepatic steatosis was 37.4% (137/366), with mild (S1), moderate (S2) and severe (S3) steatosis being 33.9%, 2.2% and 1.4%, respectively. The median histologic activity index value was 4.9 (1-17). Regarding liver fibrosis, 4 patients were F0 (1.1%), 85 were F1 (23.2%), 165 were F2 (45.1%), 50 were F3 (13.7%), 33 were F4 (9%), 21 were F5 (5.7%), and 8 were F6 (2.2%). The characteristics of the 366 patients are given in Table 1. Distribution of CAP and HSI according to histologic steatosis The median (range) CAP value of the study population was 216 (100-349) dB/m. The distribution of CAP for each degree of histologic steatosis is presented in Figure 1a. The median CAP values for patients with S0, S1, S2 and S3 were 198 (100-291), 237 (107-317), 280 (247-313) and 335 (260-349) dB/m, respectively. The differences in CAP between consecutive steatosis grades were all significant (P < 0.05) except between S2 and S3 (P = 0.09). The median (range) HSI value of the study population was 35.6 (22.0-56.0). The distribution of HSI for each degree of histologic steatosis is presented in Figure 1b. The median HSI values for patients with S0, S1, S2 and S3 were 33.9 (22-56), 35.6 (25.0-54.0), 39.3 (36.0-45.0) and 39.1 (38.0-52.0), respectively. Although HSI was not significantly different between patients with S1 and S2, S2 and S3 or S1 and S3 (all P > 0.05), the differences between the remaining categories of steatosis were significant (P < 0.05 for S0 vs. S1, S0 vs. S2 and S0 vs. S3). Diagnostic performance of CAP and HSI for hepatic steatosis For the diagnosis of hepatic steatosis, CAP predicted a steatosis grade ≥ S1 with an AUROC

of 0.780 (95% CI 0.735-0.822, Z = 11.289, P < 0.05) and an optimal cutoff of 224 dB/m associated with 68.6% sensitivity and 75.6% specificity. CAP detected a histologic grade ≥ S2 with an AUROC of 0.932 (95% CI 0.902-0.956, Z = 18.395, P < 0.05) and an optimal cutoff of 246 dB/m associated with 100% sensitivity and 78.2% specificity. The AUROC of CAP to diagnose S3 was 0.990 (95% CI 0.974-0.998) with an optimal cutoff of 284 dB/m associated with 100% sensitivity and 96.1% specificity. HSI diagnosed a steatosis grade ≥ S1 with an AUROC of 0.655 (95% CI 0.604-0.704, Z = 5.329, P < 0.05) and an optimal cutoff of 35.6 associated with 61% sensitivity and 63% specificity. For ≥ S2, HSI exhibited an AUROC of 0.755 (95% CI 0.707-0.799, Z = 6.955, P < 0.05), and the optimal diagnostic cutoff of 35.9 was associated with 100% sensitivity and 58.5% specificity. The AUROC for S3 was 0.786 (95% CI 0.740-0.827, Z = 4.053, P < 0.05) with an optimal cutoff of 37.5 associated with 100% sensitivity and 66% specificity. The AUROCs, optimal cutoff values, sensitivities, specificities, PPVs, NPVs and likelihood ratios for both CAP and HSI for hepatic steatosis evaluation are outlined in Table 2, Figure 1 and Figure 2. CAP performed significantly better than HSI for detecting ≥ S1 (Delong test: P < 0.05), ≥ S2 (Delong test: P < 0.05) and S3 steatosis (Delong test: P < 0.05). Detection rates and discordance rates of degree of hepatic steatosis between CAP, HSI, ultrasound, and liver biopsy Among 366 patients with CHB, ultrasound-diagnosed fatty liver was found in 36 patients (9.8%): 11 patients had mild steatosis, 24 had moderate steatosis and 1 had severe steatosis. We compared the accuracy of CAP, HSI and ultrasound using the histologic degree of steatosis as the reference standard. Detection rates in group S1 were 65.3% (81/124) with

CAP, 56.5% (70/124) with HSI and 17.7% (22/124) with ultrasound (χ2 = 19.597, P < 0.05). Detection rates in group S2-3 were 92.3% (12/13) with CAP, 100% (13/13) with HSI and 53.8% (7/13) with ultrasound (fisher χ2=10.795, P < 0.05). These results are given in Table 3. Discordance

between

histologic

steatosis

and

CAP-diagnosed

steatosis

or

ultrasound-diagnosed steatosis was observed in 163 (44.5%) patients and 128 (35.0%) patients, respectively, and both were lower than the discordance observed in 209 (57.1%) patients by HSI-diagnosed steatosis (χ2 = 6.403, 15.226; P < 0.05 for both). Forty-four (12%) patients had underestimated steatosis by CAP, 54 (14.8%) by his and 108 (29.5%) by ultrasound. Additionally, 119 (32.5%) patients had overestimated steatosis by CAP, 146 (39.9%) by HSI and 20 (5.5%) by ultrasound. CAP and HSI exhibited a lower underestimation rate than ultrasound (χ2 = 9.765, 6.452; P < 0.05 for both) but exhibited more overestimation rates than ultrasound (χ2 = 23.220, 32.673; P < 0.05 for both). These results are given in Table 3 and Table 4. To study the factors associated with discordance of results between each noninvasive diagnostic tool (CAP, HSI, ultrasound) and liver biopsy, we divided all patients into discordance and accordance according to CAP, HSI and ultrasound and named them CAPdis and CAPac, HSIdis and HSIac, USdis and USac, respectively. We then performed univariate analysis and multivariate stepwise regression analysis using all patient parameters. Finally, we found that the discordance with CAP was correlated with CAP value (B = 0.006, P < 0.05). The discordance with HSI was correlated with histologic steatosis grade (B = 0.201, P < 0.05), ALT (B= 0.004, P < 0.05), AST (B= -0.004, P < 0.05), BMI (B = 0.038, P < 0.05) and male gender (B= 0.100, P < 0.05). In addition, male gender (B = 0.079, P < 0.05), iron-stain positivity in the liver (B= -0.117, P < 0.05) and histologic steatosis grade

(B = 0.580, P < 0.05) were independent predictors of discordance correlated with ultrasound.

DISCUSSION It is well known that CHB is the most prevalent liver disease in China, and almost all studies have reported an association of hepatic steatosis with BMI and fasting plasma glucose instead of HBeAg and HBV DNA titer in patients with CHB [34-36]. Therefore, chronic HBV infection is not considered a risk factor for hepatic steatosis, which was also confirmed in our study, given the small number of patients with moderate and severe steatosis. However, the coexistence of chronic HBV infection and hepatic steatosis is common in the Chinese population. The prevalence of hepatic steatosis in patients with CHB in this study was much higher than in Wang and colleagues' report (37.4% vs. 17.3%), although both studies used the same criteria of hepatic steatosis. The major reason for the difference is the different study period, as the prevalence of obesity and type 2 diabetes mellitus increased with time in Chinese people, including in patients with CHB. In the Wang study, the annual prevalence of hepatic steatosis in patients with CHB increased from 8.2% in 2002 to 31.8% in 2011. In our study population, the distribution of the degree of liver-biopsy-proven steatosis was S1 (33.9%), S2 (2.2%) and S3 (1.4%). As steatosis in the majority of patients with CHBs with fatty liver was mild, it was difficult to detect by conventional ultrasonography (102/124 patients, 82.3%). Therefore, the importance of developing accurate and sensitive methods to detect mild hepatic steatosis in patients with CHB is evident. Furthermore, the presence of hepatic steatosis has been suggested to be significantly associated with failure of antiviral therapy and accelerated liver fibrosis progression [38,39]. Therefore, the detection of hepatic

steatosis might be essential for patients with CHB for treatment decisions and follow-up. Despite the common presence of hepatic steatosis in patients with CHB, liver biopsy is rarely performed for steatosis assessment because of its invasiveness and associated complications. Ultrasonography is more commonly used but is highly operator- and machine-dependent and does not provide good results of quantitative estimation of hepatic steatosis [40]. Moreover, its sensitivity for detecting mild steatosis seems unsatisfactory [41,42]. Several other imaging methods, including proton magnetic resonance spectroscopy (MRS), which directly measures proton signals from the acyl groups of hepatocyte triglyceride stores, can offer a quantitative estimate, and MRS has been investigated in several studies with promising results [43,44]. However, MRS is expensive and not recommended in the clinical setting [45]. To overcome these limitations, serologic scores such as the SteatoTest, fatty liver index, and HSI, which combine several biochemical markers and/or anthropometric characteristics, have been extensively developed in the last decade [26,46,47]. They all have the advantages of low cost, easy-to-collect data, and operator independence. However, the performance and diagnostic accuracy of these tests still need improvement [48,49]. Recently, the CAP value available on FibroScan has been introduced and has demonstrated promising diagnostic performance for detecting hepatic steatosis [22,23]. Several studies have compared the diagnostic accuracy of CAP and HSI for hepatic steatosis assessment, but they were all conducted on multi-etiology cohorts, and they did not provide any observations regarding their abilities in differentiating between two different steatosis grades [22,50]. To our knowledge, this is the first study that provides a head-to-head comparison of hepatic steatosis evaluation using CAP, HSI and ultrasound in patients with biopsy-proven CHB.

In this study, CAP and HSI both exhibited certain accuracy for the diagnosis of hepatic steatosis (the sensitivities of CAP and HSI were 68.6% and 61.0%, respectively, and specificities were 75.6% and 63%, respectively). The accuracy was improved for the diagnosis of moderate to severe steatosis (sensitivities of both CAP and HSI were 100%, and specificities of CAP and HSI were 78.2% and 58.5%, respectively). Therefore, both CAP and HSI could be applied to diagnose hepatic steatosis in patients with CHB. Abdominal ultrasound exhibited a good sensitivity of 86.1% to diagnose moderate to severe steatosis with a specificity of 98.8%, but had a poor steatosis detection rate of 22.6% in all the patients and 17.7% in patients with mild histologic steatosis. In the present study, we compared CAP and HSI performance in a single etiology cohort of patients with CHB. CAP’s diagnostic performance for assessing hepatic steatosis was satisfactory to diagnose mild steatosis (S1) and even excellent for diagnosing moderate to severe steatosis (S2 and S3). These results are slightly different from those published by Chon et al., who reported a lower performance to detect S3 [48], which might be because of a more even distribution of patients per steatosis stage in their population. Their cutoffs were also higher than ours (250 vs. 224 dB/m for ≥ S1, 299 vs. 246 dB/m for ≥ S2, and 327 vs. 284 dB/m for S3). These differences could be explained by a different population, etiologies of liver diseases and biases of degrees of hepatic steatosis. HSI was associated with a lower specificity than CAP. In addition, HSI performance was suboptimal to evaluate mild and moderate steatosis (AUROCs < 0.8 for both). These findings are consistent with Chon et al., who also reported the superiority of CAP versus HSI to stage steatosis in a multi-etiology cohort of 135 patients [48]. In our study, neither CAP nor HSI could accurately differentiate

S2 from S3 patients (P > 0.05 for both). Nonetheless, it seems difficult to draw any conclusion given the very small number of S2 and S3 patients in our cohort that might have biased the result. Compared to HSI, CAP exhibited a higher sensitivity, specificity, PPV, and NPV for diagnosis of different degrees of steatosis. The results thus demonstrated the higher diagnostic accuracy of CAP than HSI for detecting steatosis in patients with CHBs with BMI less than 30 kg/m2. To further investigate the diagnostic accuracy, we compared the two methods with conventional ultrasound and evaluated the detection rates and discordance rates of patients for each test compared with histologic steatosis. The results showed that the accuracy of CAP and HSI was significantly higher than that of ultrasound. The CAP accuracy was also significantly higher than that of HSI to detect patients with mild steatosis. Both CAP and HSI were more accurate at detecting moderate to severe steatosis than ultrasound. Ultrasound seemed to have a lower discordance rate than CAP or HSI according to Table 4, but we found that ultrasound had a lower discordance rate in S0 (5/128) and S2~3 (4/128) but S1 (117/128). This accounted for ultrasound’s lower sensitivity and poor quantification. Therefore, CAP exhibited higher diagnostic accuracy for steatosis screening and quantification in this population than HSI and ultrasound. CAP and HSI exhibited a lower underestimation rate than ultrasound but exhibited a higher overestimation rate, so CAP and HSI were needed to decrease their false-positive rates. Univariate and multivariate stepwise regression analysis showed that the discordance with CAP had a positive correlation with CAP value itself. This result was consistent with the Jung et al. study [51] and suggests that the discordance might become larger with increased CAP value [51]. However, many factors influenced the

discordance between HSI or ultrasound and liver biopsy, such as histologic steatosis grade and male gender. Therefore, CAP might have the fewer influence factors for diagnosis of steatosis. Our study has several limitations. First, the cohort study contained an uneven distribution of steatosis grades. Most of the patients had no (62.6%) or mild (33.9%) steatosis. Thus, other studies with more homogeneous numbers between steatosis grades in the population will be required to confirm the CAP diagnostic performance in patients with CHB. Second, we used liver biopsies as a reference standard and did not compare CAP with morphometry or other imaging methods such as MRS or magnetic resonance imaging-based proton density fat fraction [52]. Indeed, each steatosis grade by histology is artificially defined, and it would have been relevant to try to correlate CAP with continuous variables such as the percentage of steatosis by morphometry or the fat fraction by MRS. Because those noninvasive methods also enable a quantitative assessment of hepatic steatosis, further studies that compare their performances with CAP would be required. Third, in our clinical works, we found that the success rate of operating CAP was lower in patients with higher BMI[18], so we excluded patients with BMI≧30 kg/m2 to ensure CAP’s accuracy in the study; that was defective for studying hepatic steatosis by ultrasound or HSI, so the results were applicable only to the patients with BMI <30 kg/m2. Fourth, the patients with CHB enrolled in the study had either abnormal liver function tests or high LSM, so the prevalence of steatosis diagnosed by liver biopsy might be higher than that in patients with CHB who hadn’t underwent liver biopsy. In conclusion, hepatic steatosis is quite common among Chinese patients with CHB and should be evaluated in routine clinical practice. Both CAP and HSI seemed to be useful

noninvasive tools for steatosis evaluation in patients with CHB, and both are more accurate than conventional ultrasound. Moreover, CAP had a higher accuracy to detect steatosis than HSI and ultrasound, with a lower underestimation rate. Therefore, CAP measured by FibroScan might be a promising tool for hepatic steatosis screening and quantification in patients with CHB with BMI less than 30 kg/m2. However, further studies are needed to better define useful diagnostic cutoffs for interpreting CAP and HSI results in routine clinical practice, and evaluation of CAP compared with MRS would also be required to confirm our findings.

Author contributions: Xu L, Mi YQ, Lu W, and Fan JG designed the study; Xu L, Li P, Lu W, Shen F performed the research; Xu L carried out statistical analysis; Xu L, Lu W, Fan JG and Mi YQ wrote the manuscript; all the authors have read and approved the final revision to be published. Supported by the National Key Basic Research Project, No.2012CB517501; Chinese Foundation for Hepatitis Prevention and Control – “WANG Bao-En” Liver Fibrosis Research Fund, No. XJS20120501; The Science and Technology Foundation of Tianjin Municipal Health Bureau (12KG119)；Research project of Chinese traditional medicine and Chinese traditional medicine combined with Western medicine of Tianjin municipal health and Family Planning Commission (2015061).

Conflict of interest

None.

REFERENCES ［1］ Chalasani N, Younossi Z, Lavine JE, et al. The diagnosis and management of

non-alcoholic fatty liver disease. Gastroenterology. 2012; 142:1592–1609. ［2］ AASLD/IDSA. Hepatitis C guidance: AASLD-IDSA recommendations for testing, managing, and treating adults infected with hepatitis C virus. Hepatology. 2015; 62:932–954. ［3］ O’Shea RS, Dasarathy S, McCullough AJ. Alcoholic liver disease. Am J Gastroenterol. 2010; 105:14–32. ［4］ Terrault NA, Bzowej NH, Chang K, et al. AASLD guidelines for treatment of chronic hepatitis B. Hepatology. 2016; 63:261–283. ［5］ Singh S, Allen AM, Wang Z, et al. Fibrosis progression in nonalcoholic fatty liver vs nonalcoholic steatohepatitis. a systematic review and meta-analysis of paired-biopsy studies. Clin Gastroenterol Hepatol. 2015; 13:643-654. e1-9. ［6］ Castera L, Hezode C, Roudot-Thoraval F, et al. Worsening of steatosis is an independent factor of fibrosis progression in untreated patients with chronic hepatitis C and paired liver biopsies. Gut. 2003; 52:288–292. ［7］ Lok AS, Everhart JE, Chung RT, et al. Hepatic steatosis in hepatitis C: comparison of diabetic and nondiabetic patients in the hepatitis C antiviral long-term treatment against cirrhosis trial. Clin Gastroenterol Hepatol. 2007; 5:245–254. ［8］ Dyal HK, Aguilar M, Bhuket T, et al. Concurrent obesity, diabetes, and steatosis increase risk of advanced fibrosis among HCV patients. Dig Dis Sci. 2015; 60:2813–2824. ［9］ Berzigotti A. Getting closer to a point-of-care diagnostic assessment in patients with chronic liver disease. J Hepatol. 2014; 60:910–912.

［10］ Grant A, Neuberger J. Guidelines on the use of liver biopsy in clinical practice. British Society of Gastroenterology. Gut. 1999; 45 Suppl 4:1-11. ［11］ Lassailly G, Caiazzo R, Buob D, et al. Bariatric surgery reduces features of nonalcoholic steatohepatitis in morbidly obese patients. Gastroenterology. 2015; 149:379–388. ［12］ Raptis DA, Fischer MA, Graf R, et al. MRI: the new reference standard in quantifying hepatic steatosis?. Gut. 2012; 61:117–127. ［13］ Wu C, Ho M, Jeng Y, et al. Quantification of hepatic steatosis. J Gastroenterol Hepatol. 2014; 29:807–813. ［14］ Karlas T, Petroff D. Is magnetic resonance imaging really more accurate for classifying steatosis than controlled attenuation parameter? Gastroenterology. 2016; 151(2):374–5. ［15］ EASL–EASD–EASO. EASL–EASD–EASO clinical practice guidelines for the management of non-alcoholic fatty liver disease. J Hepatol. 2016; 64(6):1388–402. ［16］ Wong GL, Wong VW. Fat and fiber: How the controlled attenuation parameter complements noninvasive assessment of liver fibrosis. Dig Dis Sci. 2015; 60:9–12. ［17］ Sasso M, Miette V, Sandrin L, et al. The controlled attenuation parameter (CAP): a novel tool for the non-invasive evaluation of steatosis using Fibroscan. Clin Res Hepatol Gastroenterol. 2012; 36:13–20. ［18］ Shen F, Zheng RD, Mi YQ, et al. Controlled attenuation parameter for non-invasive assessment of hepatic steatosis in Chinese patients. World J Gastroenterol. 2014; 20(16): 4702-4711.

［19］ Ferraioli G, Tinelli C, Lissandrin R, et al. Controlled attenuation parameter for evaluating liver steatosis in chronic viral hepatitis. World J Gastroenterol. 2014 7; 20(21): 6626-6631. ［20］ Sasso M, Beaugrand M, de Ledinghen V, et al. Controlled attenuation parameter (CAP): a novel VCTE™ guided ultrasonic attenuation measurement for the evaluation of hepatic steatosis: preliminary study and validation in a cohort of patients with chronic liver disease from various causes. Ultrasound Med Biol. 2010; 36: 1825-1835. ［21］ de Lédinghen V, Vergniol J, Foucher J, et al. Non-invasive diagnosis of liver steatosis using controlled attenuation parameter (CAP) and transient elastography. Liver Int. 2012; 32: 911-918. ［22］ Myers RP, Pollett A, Kirsch R, et al. Controlled attenuation parameter (CAP): a noninvasive method for the detection of hepatic steatosis based on transient elastography. Liver Int. 2012; 32: 902-910. ［23］ Carvalhana S, Leitao S, Alves AC, et al. How good is controlled attenuation parameter and fatty liver index for assessing liver steatosis in general population: correlation with ultrasound. Liver Int. 2014; 34: e111–e117. ［24］ de Lédinghen V, Vergniol J, Capdepont M, et al. Controlled attenuation parameter (CAP) for the diagnosis of steatosis: A prospective study of 5323 examinations. J Hepatol. 2014; 60:1026–1031. ［25］ Yilmaz Y, Yesil A, Gerin F, et al. Detection of hepatic steatosis using the controlled attenuation parameter: a comparative study with liver biopsy. Scand J Gastroentero.

2014; 49: 611–616. ［26］ Lee JH, Kim D, Kim HJ, et al. Hepatic steatosis index: a simple screening tool reflecting nonalcoholic fatty liver disease. Dig Liver Dis. 2010; 42: 503-8. ［27］ Hamaguchi M, Kojima T, Itoh Y, et al. The severity of ultrasonographic findings in nonalcoholic fatty liver disease reflects the metabolic syndrome and visceral fat accumulation. Am J Gastroenterol. 2007; 102: 2708–15. ［28］ Wong GL. Transient elastography: kill two birds with one stone? World J Hepatol. 2013; 5:264-274. ［29］ Sandrin L, Tanter M, Gennisson JL, et al. Shear elasticity probe for soft tissues with 1-D transient elastography. IEEE Trans Ultrason Ferroelectr Freq Control. 2002; 49:43-446. ［30］ Boursier J, Zarski JP, Ledinghen Vd, et al. Determination of reliability criteria for liver stiffness evaluation by transient elastography. Hepatology. 2013; 57:1182-1191. ［31］ Knodell RG, Ishak KG, Black WC, et al. Formulation and application of a numerical scoring system for assessing histological activity in asymptomatic chronic active hepatitis. Hepatology. 1981; 1:431-435. ［32］ Ishak K, Baptista A, Bianchi L, et al. Histological grading and staging of chronic hepatitis. J Hepatol. 1995; 22:696-699. ［33］ Brunt EM, Janney CG, Di Bisceglie AM, et al. Nonalcoholic steatohepatitis: a proposal for grading and staging the histological lesions. Am J Gastroenterol.1999; 94:2467-2474. ［34］ Wong VWS, Wong GLH, Chu WCW, et al. Hepatitis B virus infection and fatty liver

in the general population. J Hepatol. 2012; 3:533-40. ［35］ Thomopoulos KC, Arvaniti V, Tsamantas AC, et al. Prevalence of liver steatosis in patients with chronic hepatitis B: a study of associated factors and of relationship with fibrosis. Eur J Gastroenterol Hepatol. 2006; 18:233–237. ［36］ Wang CC, Hsu CS, Liu CJ, et al. Association of chronic hepatitis B virus infection with insulin resistance and hepatic steatosis. J Gastroenterol Hepatol. 2008; 23:779–782. ［37］ Wang MM, Wang GS, Shen F, Chen GY, Pan Q, Fan JG. Hepatic steatosis is highly prevalent in hepatitis B patients and negatively associated with virological factors. Dig Dis Sci. 2014; 59(10):2571-9. ［38］ Jin X, Chen YP, Yang YD, Li YM, Zheng L, Xu CQ. Association between hepatic steatosis and entecavir treatment failure in Chinese patients with chronic hepatitis B. PLoS One 2012; 7: e34198. ［39］ Ates F, Yalniz M, Alan S. Impact of liver steatosis on response to pegylated interferon therapy in patients with chronic hepatitis B. World J Gastroenterol. 2011; 17:4517-4522. ［40］ Liu YL, Patman GL, Leathart JBS, et al. Carriage of the PNPLA3 rs738409 C >G polymorphism confers an increased risk of non-alcoholic fatty liver disease associated hepatocellular carcinoma. J Hepatol. 2014; 61:75-81. ［41］ Schwenzer NF, Springer F, Schraml C, et al. Non-invasive assessment and quantification of liver steatosis by ultrasound, computed tomograpy and magnetic resonance. J Hepatol. 2009; 51: 433-15.

［42］ Palmentieri B, de Sio I, La Mura V, et al. The role of bright liver echo pattern on ultrasound B-mode examination in the diagnosis of liver steatosis. Dig Liver Dis. 2006; 38:485-9. ［43］ Foster KJ, Dewbury KC, Griffith AH, et al. The accuracy of ultrasound in the detection of fatty infiltration of the liver. Brit J Radiol. 1980; 53: 440-2. ［44］ AIShaalan R, Aljiffry M, AI-Busafi S, et al. Nonalcoholic fatty liver disease: noninvasive

methods

of

diagnosing

hepatic

steatosis.

The

Saudi

J

Gasfroenferol.2015;21(2):64-70 ［45］ Castera L. Non-invasive diagnosis of steatosis and fibrosis. Diabetes Metab. 2008; 34: 674-9. ［46］ Boursier J, Cales P. Controlled attenuation parameter (CAP): a new device for fast evaluation of liver fat? Liver Int. 2012; 32: 875-7. ［47］ Poynard T, Patziu V, Naveau S, et al. The diagnostic value of biomarkers (SteatoTest) for the prediction of liver steatosis. Comp Hepatol. 2005; 4: 10. ［48］ Chon YE, Jung KS, Kim SU, et al. Controlled attenuation parameter (CAP) for detection of hepatic steatosis in patients with chronic liver diseases: a prospective study of a native Korean population. Liver Int. 2014; 34(1):102-9. ［49］ Thierry P, Guillaume L,Emmanuel D, et al. Performance of biomarkers FibroTest, ActiTest, SteatoTest, and NashTest in patients with severe obesity: meta analysis of individual patient data. PloS one. 2012; 7(3):e30325. ［50］ Bedogni G, Bellentani S, Miglioli L, et al. The Fatty Liver Index: a simple and accurate predictor of hepatic steatosis in the general population. BMC Gaetroenterol.

2006; 6: 33. ［51］ KS Jung, BK Kim, SU Kim, et al. Factors Affecting the Accuracy of Controlled Attenuation Parameter (CAP) in Assessing Hepatic Steatosis in Patients with Chronic Liver Disease. PloS one. 2014; 9(6):e98689. ［52］ Imajo K, Kessoku T, Honda Y, et al. Magnetic resonance imaging more accurately classifies steatosis and fibrosis in patients with nonalcoholic fatty liver disease than transient elastography. Gastroenterology. 2016; 150:626-37.

Figure 1-a. The distribution of CAP (dB/m) based on histological steatosis grade (S0, S1, S2 and S3). The CAP values of S1 were significantly higher and lower than those of S0 and S2 (both P < 0.0001), whereas the difference between S2 and S3 was not significant (P = 0.090). CAP, controlled attenuation parameter.

A

Figure 1-b. The distribution of HSI based on histological steatosis grade (S0, S1, S2 and S3). The HSI value of S1 was significantly higher than that of S0 (P < 0.001), whereas the difference was not significance between S1

and S2 (P = 0.370) or S2 and S3 (P = 0.676). HSI, hepatitis steatosis index.

Figure 2- (a, b, c). Receiver-operating characteristic (ROC) curves and area under the ROC curve (AUROC) for the detection of steatosis grade S > 1, S > 2, S > 3 using CAP and HSI. CAP, controlled attenuation parameter; HSI, hepatic steatosis index

A

B

c

Table 1 General characteristics of 366 patients with CHB Characteristics

Study population(n = 366)

Demographic variables Age (years)

37.6 (18-72 )

Male gender

224 2

Body mass index (kg/m )

23.4 (17.2-30)

Normal/ Overweight/ Obese

227 (62.0)/112 (30.6)/27 (7.4)

Diabetes mellitus

15 (4.1)

HBeAg positive

254 (69.4)

log10 (HBV DNA titer) (copies/ml)

5.3 (1-9)

Liver histology Fibrosis stage F0/F1/ F2 F3/ F4/F5/F6 HAI value

4 (1.1)/ 85 (23.2)/165 (45.1)/50 (13.7)/ 33 (9.0)/21 (5.7)/ 8 (2.2) 4.9 (1-17)

Steatosis grade S0/ S1/ S2/ S3

229 (62.6)/124 (33.9)/ 8 (2.2)/5 (1.4)

Biochemical data ALT (IU/L)

66.4 (4.0-338.0)

AST (IU/L)

57.7 (13.0-364.0)

GGT (IU/L)

59.9 (8.0-381.0)

ALP (IU/L)

87.9 (27.0-315.0)

Total bilirubin (umol/L)

16.4 (4.6-42.3)

FPG (mmol/L)

5.64 (3.73-15.46)

TC (mmol/L)

4.43 (2.15-10.80)

TG (mmol/L)

1.20 (0.37-5.02)

HGB (g/L)

140.9 (87-190)

PLT (*10^9/L)

187.5 (65-492)

FibroScan data LSM (kPa)

10.8 (3.0-58.3)

IQR (kPa)

1.8 (0.1-15.0)

CAP (dB/m)

216 (100-349)

IQR (dB/m)

40.6 (0-135)

HSI

35.6 (22.0-56.0)

Results are depicted as median (range) and number (%). CHB, chronic hepatitis B; BMI, body mass index; ALT, alanine aminotransferase; AST, aspartate aminotransferase; GGT, gamma-glutamyl transferase; ALP, alkaline phosphatase; FPG, fasting plasma glucose; TC, total cholesterol; TG, triglyceride; HGB, hemoglobin; PLT, platelet; HAI, histological activity index; LSM, liver stiffness measurement; CAP,

controlled attenuation parameter; IQR, interquartile range; HSI, hepatic steatosis index.

Table 2 Diagnostic performance of CAP and HSI for assessing steatosis in CHB patients in comparison with histology S0 vs. S1S2S3

S0S1 vs. S2S3

S0S1S2 vs. S3

224

246

284

0.780

0.932

0.990

(0.735-0.822)

(0.902-0.956)

(0.974-0.998)

Sensitivity (%)

68.6

100

100

Specificity (%)

75.6

78.2

96.1

PPV/NPV (%)

62.7/80.1

14.4/100.0

22.2/100.0

LR+/LR-

2.81/0.42

4.58/0

25.9/0

35.6

35.9

37.5

0.655

0.755

0.786

(0.604-0.704)

(0.707-0.799)

(0.740-0.827)

Sensitivity (%)

61.0

100

100

Specificity (%)

63.0

58.5

66.0

PPV/NPV (%)

50.0/72.7

8.3/100

3.2/100

LR+/LR-

1.65/0.62

2.41/0.00

2.94/0.00

Z value (vs. CAP)

3.421

4.713

3.137

P value (vs. CAP)

0.0006

<0.001

0.0017

Variable Optimal

cut-off

CAP (dB/m) AUROC (95% CI)

Optimal

cut-off

HSI (dB/m) AUROC (95% CI)

CHB, chronic hepatitis B; CAP, controlled attenuation parameter; HSI, hepatitis steatosis index; AUROC, area under the receiver operating characteristic curve; CI, confidence intervals; PPV, positive predictive value; NPV, negative predictive value; LR+, positive likelihood ratio; LR-, negative likelihood ratio.

Table 3 Comparision of detection of steatosis by CAP, HSI, US according to liver biopsy LB

CAP (n = 137)

HSI (n = 137)

US (n = 137)

χ2

P

S1

65.3% (81/124)

56.5% (70/124)

17.7% (22/124)

19.597

<0.001

S2-3

92.3% (12/13)

100%(13/13)

53.8% (7/13)

10.795

0.005

CAP, controlled attenuation parameter; HSI, hepatitis steatosis index; US, ultrasound; LB, liver biopsy; S1, steatosis grade 1; S2, steatosis grade 2; S3, steatosis grade 3.

Table 4 Comparison of discordance between histology and CAP, HSI, US on hepatic steatosis in 366 patients with CHB

Total discordance rate Underestimate than LB Overestimate than LB

CAP dis

HSI dis

US dis

163 (44.5%)

209 (57.1%)

128 (35.0%)

44 (12.0%)

54 (14.8%)

108 (29.5%)

119 (32.5%)

146 (39.9%)

20 (5.5%)

χ2

P

 VS =6.403,

0.011 and

 VS =15.226

<0.001



VS

=9.765,

 VS =6.452 

VS

=23.220,

 VS =32.673

both <0.05 both <0.001

CAP, controlled attenuation parameter; HSI, hepatitis steatosis index; US, ultrasound; CHB, chronic hepatitis B; LB, liver biopsy; dis, discordance; CAP dis, ; HSI dis, ; US dis, 