Association Between Hepatic Steatosis, Measured by Controlled Attenuation Parameter, and Fibrosis Burden in Chronic Hepatitis B

Accepted Manuscript Association Between Hepatic Steatosis, Measured by Controlled Attenuation Parameter, and Fibrosis Burden in Chronic Hepatitis B Wai-Kay Seto, Rex WH. Hui, Lung-Yi Mak, James Fung, Ka-Shing Cheung, Kevin SH. Liu, Danny Ka-Ho Wong, Ching-Lung Lai, Man-Fung Yuen

PII: DOI: Reference:

S1542-3565(17)31186-2 10.1016/j.cgh.2017.09.044 YJCGH 55476

To appear in: Clinical Gastroenterology and Hepatology Accepted Date: 24 September 2017 Please cite this article as: Seto W-K, Hui RW, Mak L-Y, Fung J, Cheung K-S, Liu KS, Wong DK-H, Lai C-L, Yuen M-F, Association Between Hepatic Steatosis, Measured by Controlled Attenuation Parameter, and Fibrosis Burden in Chronic Hepatitis B, Clinical Gastroenterology and Hepatology (2017), doi: 10.1016/j.cgh.2017.09.044. 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.

ACCEPTED MANUSCRIPT

Association Between Hepatic Steatosis, Measured by Controlled Attenuation Parameter, and Fibrosis Burden in Chronic Hepatitis B Running Title: Steatosis quantification in HBV Wai-Kay Seto1,2, Rex WH Hui1, Lung-Yi Mak1, James Fung1,2, Ka-Shing Cheung1,

RI PT

Kevin SH Liu1, Danny Ka-Ho Wong1,2, Ching-Lung Lai1,2, Man-Fung Yuen1,2. 1

Department of Medicine, The University of Hong Kong, Queen Mary Hospital, Hong

Kong 2

SC

State Key Laboratory for Liver Research, The University of Hong Kong, Queen

AC C

EP

TE D

M AN U

Mary Hospital, Hong Kong

Co-correspondence and Reprint request: Dr. Wai-Kay Seto E mail: [email protected] Fax: 852 28186474 Tel: 852 22553994 Prof. Man-Fung Yuen E mail: [email protected] Fax: 852 28162863 Tel: 852 22553984 Department of Medicine, the University of Hong Kong, Queen Mary Hospital, Pokfulam Road, Hong Kong

1

ACCEPTED MANUSCRIPT The authors declare they have participated in the preparation of the manuscript and have seen and approved the final version.

M AN U

SC

RI PT

WK Seto were involved in involved in study concept and design, acquisition of data, analysis and interpretation of data and drafting of manuscript. RW Hui, LY Mak and J Fung were involved in acquisition of data and analysis and interpretation of data. KS Cheung, KS Liu, and DK Wong were involved in acquisition of data. CL Lai was involved in critical revision of manuscript. MF Yuen was involved in study concept and design, analysis and interpretation of data, critical revision of manuscript and overall study supervision.

Conflicts of interest

WK Seto is an advisory board member of Gilead Sciences and Bristol-Myers Squibb and received speaker fees from Gilead Sciences, Bristol-Myers Squibb and Novartis.

TE D

J Fung received research funding from Novartis. CL Lai received speaker fees from Bristol-Myers Squibb, Novartis and Gilead Sciences, and is an advisory board member of Gilead Sciences. MF Yuen received speaker fees from GlaxoSmithKline, Bristol-Myers Squibb, Novartis and Gilead Sciences; and received research funding

EP

and is an advisory board member of Bristol-Myers Squibb, Novartis and Gilead

AC C

Sciences. The remaining authors have no conflict of interest.

Role of the funding source This study was funded by S.K. Yee Medical Foundation Grant, Hong Kong (ref no: 2151210)

Word count: 3,093 Number of figures: 3 Number of tables: 3 Supplementary tables: 4

2

ACCEPTED MANUSCRIPT Abstract: Background & Aims: The interaction between chronic hepatitis B (CHB) and hepatic steatosis is poorly understood. We investigated whether measurement of controlled attenuation parameter (CAP), a non-invasive method to quantify steatosis, can assist in monitoring patients with CHB.

M AN U

SC

RI PT

Methods: We performed transient elastography, to measure liver stiffness, and made CAP measurements in 1606 patients with CHB (898 treated with nucleoside analogues, for a median 75.4 months) in Hong Kong, from January 2015 through September 2016. We also collected information on patients’ medical history, current treatment, and smoking and alcohol habits, anthropometric measurements. We obtained and analyzed fasting blood samples. Severe liver fibrosis was defined, according to guidelines, as a liver stiffness measurement greater than 9.0 kPa in patients with normal level of alanine aminotransferase (ALT) or greater than 12.0 kPa in patients with a level of ALT 1–5-fold the upper limit of normal. Steatosis was defined as a CAP measurement of 248 dB/m or more, and severe steatosis as a CAP measurement or 280 dB/m more. We performed multivariate analysis to identify factors associated with severe fibrosis.

EP

TE D

Results: The prevalence of steatosis, severe steatosis, and severe fibrosis in our cohort were 40.8%, 22.6% and 14.1% respectively. A higher proportion of patients with severe steatosis had severe fibrosis (21.4% vs 11.9% in the overall cohort; P<.001). In multivariate analysis, severe steatosis was associated with severe fibrosis in treatment-naïve patients (odds ratio, 3.60, 95% CI, 1.21–10.75) and in patients receiving treatment (odds ratios: 1.95 (1.06-3.61) for 3 or more years of treatment, 2.28 (1.13-4.61) for 5 or more years of treatment, and 2.79 (1.17-6.62) for 7 or more years of treatment). With every increase in CAP value of 10 dB/m, the risk of severe fibrosis increased by 15% in treatment-naïve patients and by 7%–8% in patients receiving treatment.

AC C

Conclusion: Severe steatosis, determined by CAP measurement, is associated with severe fibrosis in treatment-naïve patients with CHB and in patients receiving treatment. Longitudinal studies are required to investigate if steatosis control, in addition to antiviral treatment, can reduce the burden fibrosis in patients with CHB. KEY WORDS: hepatitis B virus infection; HBV; NAFLD; metabolic; obesity (344 words)

3

ACCEPTED MANUSCRIPT Introduction Long-term nucleoside analogue therapy has transformed the clinical landscape of chronic hepatitis B (CHB). Nucleoside analogue-treated CHB patients have a reduced risk of cirrhotic complications1 and hepatocellular carcinoma (HCC).2

were receiving nucleoside analogue treatment.3, 4

RI PT

Recent Asian and Western cohort studies showed that 27% to 41% of CHB patients

While regression of liver fibrosis during nucleoside analogue treatment for

SC

CHB is possible, it is by no means universal, even after prolonged periods of virologic suppression. Evidence is emerging on the contribution of metabolic

M AN U

parameters to the fibrogenic process of CHB. An elevated body mass index (BMI) increases the probability of persistent fibrosis during nucleoside analogue treatment,5,

6

while diabetes7 and metabolic syndrome8 both promote fibrogenesis

and cirrhosis in CHB. A particular metabolic parameter warranting further

TE D

investigation is hepatic steatosis, especially since non-alcoholic fatty liver disease (NAFLD) is present in 24% to 46% of the general Western and Asian population,9, 10 while 25% of the CHB population have co-existing NAFLD.11 However, studies

EP

reporting the effects of NAFLD on the disease course of CHB were not consistent.12

AC C

The independent effect of hepatic steatosis on the disease course of CHB has also not been thoroughly evaluated. The majority of on-treatment CHB patients are stable and asymptomatic, in

whom performing invasive liver biopsies for the purpose of fibrosis assessment may not be justifiable or acceptable by patients. With the development of simple noninvasive methods of fibrosis assessment, e.g. liver stiffness measurements by transient elastography, the evaluation of liver fibrosis can now be extended to this on-treatment CHB cohort. Liver stiffness measurements are now internationally

4

ACCEPTED MANUSCRIPT recommended in the diagnostic algorithm of CHB.13 An additional advantage of transient elastography is the simultaneous measurement of controlled attenuation parameter (CAP), a quantitative marker of hepatic steatosis.14 CAP correlates well with liver fat quantity on histology and has been validated in both Western and Asian

RI PT

populations.15, 16

In our present study, we aimed to evaluate the association between hepatic steatosis, as measured by CAP, and liver fibrosis, as measured by liver stiffness, in

SC

a large population of treatment-naïve and treatment-experienced Asian CHB patients.

M AN U

Methods Patients

The Liver Clinic of the Department of Medicine, Queen Mary Hospital, Hong Kong regularly follows up approximately 9,500 CHB patients on a long-term basis.

TE D

From January 2015 to September 2016, we recruited Asian CHB patients aged 18 years or older who were positive for the hepatitis B surface antigen for at least 6 months. Both treatment-naïve and on-treatment CHB patients were recruited. We

EP

excluded patients with concomitant liver diseases, including chronic hepatitis C and

AC C

D infection, primary biliary cholangitis, primary sclerosing cholangitis, autoimmune hepatitis and Wilson’s disease. We also excluded conditions which can preclude accurate

transient

elastography

measurements,

including

serum

alanine

aminotransferase (ALT) >5 x upper limit of normal (200 U/L) and known congestive heart failure.13 Other exclusion criteria include known significant alcohol intake (≥30 grams per day for male, ≥20 grams per day for females), previous liver transplantation, currently active or suspected HCC, currently on medications known to induce hepatic steatosis (including corticosteroids, methotrexate and tamoxifen),

5

ACCEPTED MANUSCRIPT and current pregnancy. All patients consented to participate in the study, with ethics approval obtained from the Institutional Review Board, the University of Hong Kong and West Cluster of Hospital Authority, Hong Kong.

RI PT

Assessment

Recruited participants were asked to provide information on their medical history, current drug treatment, and smoking and alcohol habits. Anthropometric

SC

measurements were obtained, including weight (kg), height (m), waist circumference (cm) and hip circumference (cm). BMI (kg/m2) was calculated by (weight)/(height)2.

M AN U

Fasting blood samples were obtained, including liver biochemistry, glucose, glycosylated hemoglobin and triglyceride, high-density lipoprotein, and low-density lipoprotein. The upper limit of normal of serum ALT was set at 40 U/L. Serum HBV DNA levels were measured by Cobas Taqman (Roche Diagnostics, Branchburg, NJ),

TE D

with a lower limit of detection of 20 IU/ml.

Diabetes mellitus was defined in noncirrhotic patients as glycosylated hemoglobin ≥6.5%, fasting plasma glucose ≥7.0 mmol/L or specific treatment for

EP

diabetes mellitus.17 In cirrhotic patients, diabetes was defined only as fasting glucose

AC C

≥7.0 mmol/L or with specific diabetic treatment. Metabolic syndrome was defined as elevated waist circumference plus two of the four following conditions (abnormalities in blood pressure, serum triglyceride, high-density lipoprotein and glucose) using Asian-specific criteria.18

Transient elastography Fasting liver stiffness and CAP measurements were obtained by transient elastography (Fibroscan, Echosens, Paris) examination, of which the procedures

6

ACCEPTED MANUSCRIPT were previously described.6 Transient elastography was performed by two experienced operators who had performed over 500 transient elastography examinations and had obtained operator certification from Echosens. Besides using the standard M probe, the XL probe was available for usage in patients with BMI ≥30

RI PT

kg/m2. Both liver stiffness and CAP values were expressed as the median of at least 10 successful measurements. liver stiffness with fewer than 10 successful measurements, success rate <60% or (interquartile range (IQR))/ (median) >30%

SC

were defined as unreliable results.13 CAP was calculated simultaneously as liver stiffness and was only considered reliable if successful and reliable liver stiffness

M AN U

results were obtained.14

Severe liver fibrosis, using the European Association for the Study of the Liver – Asociacion Latinoamericana para el Estudio del Higado guidelines, was defined as liver stiffness >9.0 kPa in patients with normal ALT or liver stiffness >12.0 kPa in

TE D

patients with ALT 1-5 x upper limit of normal.13 Cirrhosis was defined as liver stiffness >12 kPa with normal ALT, or >13.4 kPa in patients with ALT 1-5 x upper limit of normal.6 Hepatic steatosis was defined as CAP ≥248 dB/m. Mild (number of

EP

affected hepatocytes: 10-33%), moderate (34-66%) and severe (>66%) steatosis

AC C

were defined as CAP 248-267 dB/m, CAP 268-279 dB/m and CAP ≥280 dB/m respectively, based on a recent meta-analysis correlating CAP measurements with histological steatosis grading.19

Statistical analysis The prevalence of severe liver fibrosis in the general CHB population is estimated to be between 14% and 34%.20, 21 With a sample size of 1,600 patients,

7

ACCEPTED MANUSCRIPT allowing a confidence interval of 5%, the precision of our study in detecting severe liver fibrosis would be between 1.7% and 2.3%. Continuous variables were expressed as mean (standard deviation (SD)) or median (interquartile range (IQR)) as appropriate. Steatosis was analysed

RI PT

categorically based on the assessment of mild, moderate and severe steatosis via CAP measurements. Statistical comparisons were carried out using Student’s t-test, Mann-Whitney U test, Analysis of variance or Kruskal-Wallis test for continuous

SC

variables and chi-squared test or Fisher’s exact test for categorical variables. Multivariate analysis was performed using binary logistic regression, in which factors

M AN U

with p<0.10 in the univariate model were entered into a forward stepwise selection multivariate model. Variance inflation factor testing was used for the detection of multicollinearity.

In the multivariate analysis of on-treatment patients, we included only patients

TE D

with treatment duration of at least 3 years. To improve the robustness of analysis, we repeated our multivariate analysis as stratified by different treatment durations of ≥3, ≥5 and ≥7 years. As high CAP values could be associated with overestimation of

EP

liver fibrosis via liver stiffness measurements, we performed a sensitivity analysis

AC C

among patient subgroups in which the risk of a falsely positive liver stiffness result is low (CAP <300 dB/m; CAP 300-339 dB/m and liver stiffness >12.5 kPa; and CAP >340 dB/m and liver stiffness >13.6 kPa).22 We also performed a sensitivity analysis using adjusted CAP measurements as suggested by Karlas et al. (deductions of 10 dB/m for diabetes patients and deductions / additions of 4.4 dB/m for each unit of BMI above / below 25 kg/m2).19 All data was analysed using SPSS 22.0 software (SPSS Inc, Chicago, IL). A two-tailed p-value <0.05 was considered as statistically significant.

8

ACCEPTED MANUSCRIPT

Results Patient disposition of our present study is depicted in Figure 1. Among 1,762 screened CHB patients, 1,606 (91.1%) were subsequently recruited, with their

RI PT

baseline characteristics depicted in Table 1. The mean age was 52.4 (±11.2) years; 89.5% were hepatitis B e antigen (HBeAg)-negative. Eight hundred and ninety eight patients (55.9%) were on nucleoside analogue treatment for a median duration of

SC

75.4 (IQR 38.7–102.3) months, of which 681 (75.8%), 550 (61.2%) and 358 (39.9%) were treated for ≥3, ≥5, and ≥7 years respectively. Seven hundred and ninety-one

M AN U

(88.1%) were on either entecavir or tenofovir. 90.4% of treated patients (n=812) had undetectable HBV DNA (<20 IU/mL) at the time of assessment. Among the treatment-naïve cohort (n=708), the median HBV DNA level and ALT were 857 (IQR

TE D

102-10,900) IU/mL and 25 (IQR 19-35) respectively.

Prevalence of steatosis and fibrosis

As depicted in Figure 2A, the prevalence of steatosis (CAP ≥248 dB/m) and

EP

severe steatosis (CAP ≥280 dB/m) were 40.8% and 22.6% respectively, with no

AC C

significant differences noted between the treatment-naïve and on-treatment cohorts (all p>0.05). The prevalence of severe liver fibrosis and cirrhosis in the on-treatment cohort were 21.7% and 11.2% respectively. Both prevalence rates were significantly lower in the treatment-naïve cohort, with only 4.4% demonstrating severe fibrosis and 2.0% with cirrhosis (all p<0.001). Among the 195 on-treatment patients with severe fibrosis, the median treatment duration was 71.0 (IQR 36.5-99.8) months. 91.3% (n=178) had undetectable HBV DNA, while the median ALT was 31 (IQR 23-41) U/L. Among the

9

ACCEPTED MANUSCRIPT 31 treatment-naïve patients with severe fibrosis, median HBV DNA and ALT levels were 1,170 (IQR 56-63,500) IU/mL and 32 (24-44) U/L respectively. The prevalence of severe liver fibrosis as stratified by steatosis severity and treatment status is depicted in Figure 2B. Patients with severe steatosis, when

RI PT

compared to patients with no, mild or moderate steatosis, had a significantly increased proportion of severe fibrosis among all patients (21.4% versus 11.9%, p<0.001), treatment-naïve patients (11.3% versus 2.4%, p<0.001) and on-treatment

SC

patients (29.1% and 19.5% respectively, p=0.003). When comparing the rate of severe fibrosis of non-steatotic patients versus mild-to-moderate steatotic patients,

M AN U

there was no significant difference noted among all patient groups (Figure 2B, all p>0.05).

Steatosis predicting severe fibrosis in treatment-naive patients

TE D

Univariate and multivariate analyses in treatment-naïve patients (n=708) are depicted in Table 2. An increased CAP measurement was an independent factor significantly associated with severe fibrosis (odds ratio (OR) 1.015, 95% CI 1.004-

EP

1.027), indicating that with an increase in CAP value by 10 dB/m, the risk for the

AC C

development of severe fibrosis rose by 15%. Other significant factors include increased AST levels, reduced platelet counts, and HBeAg-negativity. Serum HBV DNA levels had no association with severe fibrosis (p=0.381). Other metabolic factors also demonstrated no such association (all p>0.05).

Severe fibrosis was significantly more common in patients with severe steatosis (CAP ≥280 dB/M) when compared to patients without (11.3% versus 2.4%, OR 5.25, 95% CI 2.51-10.98). When replacing quantitative CAP measurements with

10

ACCEPTED MANUSCRIPT ordinal steatosis assessment in the multivariate regression model, severe steatosis was an independent factor predicting severe fibrosis (Figure 3A), achieving an OR of 3.60 (95% CI 1.21-10.75). Mild steatosis or above (CAP ≥248 dB/m) had no

95% CI 0.84-3.95).

Steatosis predicting severe fibrosis in on-treatment patients

RI PT

independent association with severe fibrosis in treatment-naïve patients (OR 1.82

SC

Univariate and multivariate analyses of factors associated with severe liver fibrosis in patients with ≥3 years of nucleoside analogue therapy (n=681) is depicted

M AN U

in Table 3. An increased CAP measurement remained independently associated with severe liver fibrosis (OR 1.008, 95% CI 1.003-1.014), indicating that with an increase in CAP value by 10 dB/m, the risk for the development of severe fibrosis rose by 8%. Other parameters independently associated with severe liver fibrosis include

TE D

diabetes, decreased serum albumin, increased bilirubin, increased aspartate aminotransferase (AST) and decreased platelet count. Multicollinearity testing of

shown).

EP

different metabolic factors showed no significant multicollinearity issues (data not

AC C

Among patients with ≥5 and ≥7 years of nucleoside analogue therapy (Supplementary Tables 1 and 2), an increased CAP measurement remained independently associated with severe liver fibrosis (OR 1.008, 95% CI 1.002-1.014; and OR 1.007, 95% CI 1.000-1.015 respectively), indicating that with an increase in CAP value by 10 dB/m, the risk for the development of severe fibrosis rose by 7-8%. Severe steatosis was also the only metabolic risk factor consistently associated with severe fibrosis regardless of treatment duration. Other metabolic parameters, including diabetes, BMI and metabolic syndrome, lacked consistent significant

11

ACCEPTED MANUSCRIPT association with severe fibrosis with respect to different treatment durations (Table 3, Supplementary Tables 1 and 2). Severe fibrosis was significantly more common among patients with severe steatosis when compared to patients without, consistently for treatment duration of

RI PT

≥3 years (30.2% versus 19.0%, OR 1.85, 95% CI 1.24-2.77), ≥5 years (31.1% versus 17.5%, OR 2.13, 95% CI 1.36-3.33), and ≥7 years (29.2% versus 16.7%, OR 2.05, 95% CI 1.18-3.59). When replacing quantitative CAP measurements with

SC

ordinal steatosis assessment in the multivariate regression model, severe steatosis showed consistent significant association with severe liver fibrosis (Figure 3). There

M AN U

was an incremental increase in ORs from a treatment duration of ≥3 years (OR 1.95, 95% CI 1.06-3.61), to ≥5 years (OR 2.28, 95% CI 1.13-4.61) and to ≥7 years (OR 2.79, 95% CI 1.17-6.62). Mild steatosis or above had no independent association with severe fibrosis (OR 1.34 95% CI 0.93-1.94; OR 1.54 95% CI 0.95-2.32; and OR

Sensitivity analysis

TE D

1.54 95% CI 0.91-2.59 respectively).

EP

A sensitivity analysis including only patients with low risk of a falsely elevated

AC C

liver stiffness result22 is depicted in Supplementary Table 3. After multivariate analysis, CAP measurements remained significantly associated with severe fibrosis for treatment-naïve patients, and on-treatment patients regardless of treatment duration (all p<0.05). Another sensitivity analysis of adjusted CAP measurements as suggested by Karlas et al. is found in Supplementary Table 4. After adjustment, CAP measurements remained independently associated with severe fibrosis for both treatment-naïve and on-treatment patients (all p<0.05).

12

ACCEPTED MANUSCRIPT Discussion The present study demonstrated the association of hepatic steatosis with the fibrosis burden of CHB, both in treatment-naïve patients and in patients receiving

evident when the grading was severe (CAP ≥280 dB/m).

RI PT

nucleoside analogue therapy. For both patient cohorts, the association was only

On-treatment population of CHB is increasing worldwide,3 exceeding 40% in endemic regions of Asia.4 Together with the current effectiveness of nucleoside

SC

analogue therapy, serum HBV DNA levels, a traditional risk factor for disease progression, may no longer be the prime determinant of long-term outcomes in CHB.

M AN U

Over 90% of on-treatment patients in our present study had undetectable HBV DNA. In our treatment-naïve cohort, the median HBV DNA level of patients with severe fibrosis was only 1,170 IU/mL. If viral activities were adequately suppressed or were more or less quiescent, the relative influence of metabolic factors on the disease

TE D

course of CHB might be accentuated, and might explain why our present results differ from prior studies which included patients with higher HBV DNA levels.12 While the effects of diabetes,7 obesity6 and metabolic syndrome8 on CHB had

EP

been previously described, our study results demonstrated hepatic steatosis to be an

AC C

additional metabolic risk factor of CHB-related fibrosis. It was the only metabolic factor associated with fibrosis in treatment-naïve patients (Table 2). In the ontreatment cohort, steatosis was the only metabolic parameter showing consistent association with severe fibrosis regardless of treatment duration (Table 3, Supplementary Tables 1 and 2). In addition, our current study applied the non-invasive quantification of steatosis via CAP measurements, and determined that the risk of severe liver fibrosis was not increased among all CHB patients with steatosis, but specifically

13

ACCEPTED MANUSCRIPT among patients with severe steatosis (CAP ≥280 dB/m), present in 22.6% of our CHB patient cohort. To our knowledge, our study was the first to demonstrate the usefulness of CAP measurements in disease monitoring of CHB. Severe steatosis on liver histology in NAFLD has strong association with both steatohepatitis and

RI PT

advanced fibrosis,23 and this association could also be present in CHB. In the presence of severe steatosis, the OR for severe fibrosis was 1.95 to 2.79 in the ontreatment cohort (Figure 3), signifying that even after the achievement of HBV DNA

SC

undetectability during nucleoside analogue treatment, the control of severe steatosis may still be an important surrogate therapeutic endpoint before fibrosis regression

M AN U

can be achieved.24 The invasive nature of a liver biopsy would limit its widespread usage especially in stable and asymptomatic on-treatment patients. Since the measurement of CAP by transient elastography is fast and easy to perform, it can potentially be implemented with ease as a monitoring strategy for large numbers of

TE D

on-treatment patients.

The prevalence of severe fibrosis in the treatment-naïve cohort was only 4.4% (Figure 2A), lower than previous studies.20 With improved accessibility of CHB-

EP

related treatment in our locality,2 the majority of patients with high viral loads and This was

AC C

advanced disease be commenced on nucleoside analogue therapy.

supported by the low HBV DNA levels in our treatment-naïve cohort (Table 1). Our study results demonstrated these patients, despite having quiescent viral activity, could still develop liver fibrosis in the presence of severe steatosis, with a high OR of 3.60 (Figure 3). The exact long-term risk of fibrosis progression among this particular patient subgroup will require further longitudinal studies. Other metabolic factors, including diabetes, metabolic syndrome and increased BMI lacked consistent significant association among different patient

14

ACCEPTED MANUSCRIPT subgroups. Nonetheless, this does not lessen their importance in disease monitoring. Remission of steatosis is common, especially in individuals achieving good glycemic control25 and substantial weight loss,26 implying that control of metabolic risk factors in CHB patients with severe steatosis may also prove beneficial. In addition, our

RI PT

current Asian cohort would have a lower BMI and waist circumference than Western populations; these metabolic factors might play a more important role in non-Asian CHB populations. For any CHB patient with co-existing NAFLD, a comprehensive

SC

anthropometric and metabolic evaluation will still be required.

The present study was limited by its cross-sectional nature. However, our

M AN U

present study had the strength of a large sample size, and demonstrated how the application of a widely available and non-invasive marker of steatosis at a single time point can already have prognostic implications. Being a relatively novel marker, CAP measurements lack the reliability criteria currently available for liver stiffness

TE D

measurements, signifying measurement quality and defined cut-offs for steatosis grading will still require further fine-tuning. The difficulty in recruiting stable and asymptomatic patients for an invasive liver biopsy meant the lack of histological

EP

validation. Together with the comparatively weaker associations demonstrated in the

AC C

sensitivity analysis, our findings hence require validation in other CHB cohorts, or via the future usage of alternative validated methods of non-invasive fibrosis assessment.27

In conclusion, in our large population of CHB patients, the presence of severe

steatosis (CAP ≥280 dB/m) was independently associated with severe liver fibrosis in both treatment-naïve and on-treatment cohorts. An increased CAP measurement was the only metabolic parameter to be consistently associated with severe fibrosis regardless of treatment duration. Future longitudinal and interventional studies will

15

ACCEPTED MANUSCRIPT be needed to investigate if the control of steatosis, together with the virologic suppression of nucleoside analogue therapy, will reduce the fibrosis burden of CHB.

Acknowledgements

AC C

EP

TE D

M AN U

SC

Pui-Chen Li for the logistical arrangement of patients.

RI PT

The authors thank Mr. Charles Cheng, Mr. Ringo Wu, Ms. Carmen Chan and Ms.

16

ACCEPTED MANUSCRIPT References 1.

Yuen MF, Seto WK, Chow DH, et al. Long-term lamivudine therapy reduces the risk of long-term complications of chronic hepatitis B infection even in patients without advanced disease. Antivir Ther 2007;12:1295-303. Seto WK, Lau EH, Wu JT, et al. Effects of nucleoside analogue prescription

RI PT

2.

for hepatitis B on the incidence of liver cancer in Hong Kong: a territory-wide ecological study. Aliment Pharmacol Ther 2017;45:501-509.

Spradling PR, Xing J, Rupp LB, et al. Distribution of disease phase, treatment

SC

3.

prescription and severe liver disease among 1598 patients with chronic

M AN U

hepatitis B in the Chronic Hepatitis Cohort Study, 2006-2013. Aliment Pharmacol Ther 2016;44:1080-1089. 4.

Wu CY, Lin JT, Ho HJ, et al. Association of nucleos(t)ide analogue therapy with reduced risk of hepatocellular carcinoma in patients with chronic hepatitis

5.

TE D

B: a nationwide cohort study. Gastroenterology 2014;147:143-151.e5. Marcellin P, Gane E, Buti M, et al. Regression of cirrhosis during treatment with tenofovir disoproxil fumarate for chronic hepatitis B: a 5-year open-label

Seto WK, Fung J, Cheung KS, et al. Body-mass index is associated with

AC C

6.

EP

follow-up study. Lancet 2013;381:468-75.

fibrosis regression during long-term nucleoside analogue therapy in chronic hepatitis B. Aliment Pharamcol Ther 2016;44:1071-1079.

7.

Huang YW, Wang TC, Lin SC, et al. Increased risk of cirrhosis and its

decompensation in chronic hepatitis B patients with newly diagnosed diabetes: a nationwide cohort study. Clin Infect Dis 2013;57:1695-702. 8.

Wong GL, Chan HL, Yu Z, et al. Coincidental metabolic syndrome increases the risk of liver fibrosis progression in patients with chronic hepatitis B--a

17

ACCEPTED MANUSCRIPT prospective cohort study with paired transient elastography examinations. Aliment Pharmacol Ther 2014;39:883-93. 9.

Seto WK, Yuen MF. Nonalcoholic fatty liver disease in Asia: emerging perspectives. J Gastroenterol 2017;52:164-174. Fung J, Lee CK, Chan M, et al. High prevalence of non-alcoholic fatty liver

RI PT

10.

disease in the Chinese - results from the Hong Kong liver health census. Liver Int 2015;35:542-9.

Spradling PR, Bulkow L, Teshale EH, et al. Prevalence and causes of

SC

11.

elevated serum aminotransferase levels in a population-based cohort of

12.

M AN U

persons with chronic hepatitis B virus infection. J Hepatol 2014;61:785-91. Machado MV, Oliveira AG, Cortez-Pinto H. Hepatic steatosis in hepatitis B virus infected patients: meta-analysis of risk factors and comparison with hepatitis C infected patients. J Gastroenterol Hepatol 2011;26:1361-7. European Association for the Study of the Liver, Asociacion Latinoamericana

TE D

13.

para el Estudio del Higado. EASL-ALEH Clinical Practice Guidelines: Noninvasive tests for evaluation of liver disease severity and prognosis. J Hepatol

Sasso M, Beaugrand M, de Ledinghen V, et al. Controlled attenuation

AC C

14.

EP

2015;63:237-64.

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;36:1825-35.

15.

de Ledinghen 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;60:1026-31.

18

ACCEPTED MANUSCRIPT 16.

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;34:102-9. American Diabetes Association. 2. Classification and Diagnosis of Diabetes. Diabetes Care 2016;39 Suppl 1:S13-22.

18.

International Diabetes Federation. The IDF consensus worldwide definition of the metabolic syndrome. Volume 2016, 2006.

Karlas T, Petroff D, Sasso M, et al. Individual patient data meta-analysis of

SC

19.

RI PT

17.

Hepatol 2017;66:1022-1030. 20.

M AN U

controlled attenuation parameter (CAP) technology for assessing steatosis. J

Fung J, Lai CL, But D, et al. Prevalence of fibrosis and cirrhosis in chronic hepatitis B: implications for treatment and management. Am J Gastroenterol 2008;103:1421-6.

Yuen MF, Lee CK, Wong DK, et al. Prevalence of occult hepatitis B infection

TE D

21.

in a highly endemic area for chronic hepatitis B: a study of a large blood donor population. Gut 2010;59:1389-93.

Petta S, Wong VW, Camma C, et al. Improved noninvasive prediction of liver

EP

22.

AC C

fibrosis by liver stiffness measurement in patients with nonalcoholic fatty liver disease accounting for controlled attenuation parameter values. Hepatology 2017;65:1145-1155.

23.

Chalasani N, Wilson L, Kleiner DE, et al. Relationship of steatosis grade and zonal location to histological features of steatohepatitis in adult patients with non-alcoholic fatty liver disease. J Hepatol 2008;48:829-34.

24.

Sanyal AJ, Friedman SL, McCullough AJ, et al. Challenges and opportunities in drug and biomarker development for nonalcoholic steatohepatitis: findings

19

ACCEPTED MANUSCRIPT and recommendations from an American Association for the Study of Liver Diseases-U.S. Food and Drug Administration Joint Workshop. Hepatology 2015;61:1392-405. 25.

Petersen KF, Dufour S, Befroy D, et al. Reversal of nonalcoholic hepatic

RI PT

steatosis, hepatic insulin resistance, and hyperglycemia by moderate weight reduction in patients with type 2 diabetes. Diabetes 2005;54:603-8. 26.

Zelber-Sagi S, Lotan R, Shlomai A, et al. Predictors for incidence and

SC

remission of NAFLD in the general population during a seven-year prospective follow-up. J Hepatol 2012;56:1145-51.

M AN U

Kim BK, Kim HS, Yoo EJ, et al. Risk assessment of clinical outcomes in Asian patients with chronic hepatitis B using enhanced liver fibrosis test. Hepatology

EP

TE D

2014;60:1911-9.

AC C

27.

20

ACCEPTED MANUSCRIPT Table 1. Baseline characteristics of all 1,606 patients All patients (n=1606) 52.4 (±11.2) 842 (52.4%) 23.8 (±3.7) 424 (26.4%) 81 (5.0%) 11 (0.7%) 84.6 (±10.9) 96.7 (±6.9) 134.4 (±20.4) 79.5 (±13.1) 330 (20.5%) 5.4 (±1.3) 5.6 (±0.8) 2.7 (±0.8) 1.5 (±0.4) 1.1 (±0.9) 195 (12.1%) 471 (29.3%) 203 (12.6%) 1438 (89.5%)

Treatment-naïve (n=708) 49.4 (±10.7) 274 (38.7%) 23.8 (±3.8) 190 (26.8%) 36 (5.1%) 8 (1.1%) 83.3 (±10.9) 96.9 (±6.9) 131.9 (±19.4) 78.4 (±13.0) 118 (16.7%) 5.2 (±1.0) 5.6 (±0.7) 2.8 (±0.8) 1.6 (±0.5) 1.1 (±0.6) 82 (11.6%) 178 (25.1%) 69 (9.7%) 651 (91.9%)

On-treatment (n=898) 54.7 (±11.3) 568 (63.3%) 23.7 (±3.6) 234 (26.1%) 45 (5.0%) 3 (0.3%) 85.6 (±10.9) 96.5 (±6.9) 136.4 (±21.0) 80.3 (±13.1) 212 (23.6%) 5.5 (±1.4) 5.7 (±0.8) 2.6 (±0.8) 1.4 (±0.4) 1.1 (±0.7) 113 (12.6%) 293 (32.7%) 134 (14.9%) 787 (87.6%)

AC C

EP

TE D

M AN U

SC

RI PT

Age (years) Male patients, n (%) 2 BMI (kg/m ) - 25-30, n (%) - >30-35, n (%) - >35, n (%) Waist circumference (cm) Hip circumference (cm) Systolic blood pressure (mmHg) Diastolic blood pressure (mmHg) Anti-hypertensive treatment, n (%) Fasting glucose (mmol/L) Glycosylated haemoglobin (%) Low-density lipoprotein (mmol/L) High-density lipoprotein (mmol/L) Triglyceride (mmol/L) Anti-lipid treatment, n (%) Metabolic syndrome, n (%) Diabetes, n (%) HBeAg-negative, n (%) Nucleoside analogue treatment, n (%) - Entecavir 666 (74.2%) ‡ - Tenofovir 125 (13.9%) 60 (6.7%) - Lamivudine monotherapy 29* (3.2%) - Adefovir 18 (2%) - Telbivudine Treatment duration, months 75.4 (38.7-102.3) Albumin (g/L) 44.9 (±2.9) 45.0 (±2.4) 44.8 (±3.2) Bilirubin (umol/L) 11.7 (±6.3) 10.9 (±5.4) 12.3 (±6.8) ALT (U/L 26 (20-35) 25 (19-35) 26 (20-35) AST (U/L) 26 (22-31) 25 (21-30) 26 (22-32) 9 Platelet x 10 /L 204 (160-245) 223 (187-258) 182 (132-229) § HBV DNA , IU/mL 20 (20-650) 857 (102-10,900) 20 (20-20) § Undetectable HBV DNA , n (%) 910 (56.6%) 98 (13.8%) 812 (90.4%) CAP (dB/M) 237.8 (±58.2) 239.9 (±57.9) 236.2 (±58.4) ‡ 8 had concomitant lamivudine *15 had concomitant lamivudine § Low limit of detection 20 IU/mL BMI, body mass index; CAP, controlled attenuation parameter; HBeAg, hepatitis B e antigen; ALT, alanine aminotransferase; AST, aspartate aminotransferase; HBV, hepatitis B virus

21

ACCEPTED MANUSCRIPT

Table 2. Univariate and multivariate analysis of factors associated with severe liver fibrosis among treatment-naïve patients (n=708). Significant factors via multivariate analysis are highlighted in bold. Increased CAP measurements were independently associated with severe fibrosis.

SC

M AN U

TE D

EP

AC C

Age (years) Male patients, n (%) 2 BMI (kg/m ) Waist circumference (cm) Hip circumference (cm) Systolic blood pressure (mmHg) Diastolic blood pressure (mmHg) Low-density lipoprotein (mmol/L) High-density lipoprotein (mmol/L) Triglyceride (mmol/L) Metabolic syndrome, n (%) Diabetes, n (%) HBeAg-negative, n (%) Albumin (g/L) Bilirubin (umol/L) ALT (U/L AST (U/L) 9 Platelet x 10 /L HBV DNA (IU/mL) CAP (dB/M) BMI, body mass index; HBeAg, parameter Severe liver fibrosis defined by EASL-ALEH criteria

RI PT

Severe fibrosis No severe fibrosis Univariate Multivariate (n=31) (n=677) Odds ratio (95% CI) Odds ratio (95% CI) 55.1 (±10.6) 49.2 (±10.6) 1.049 (1.016-1.082) 1.001 (0.952-1.053) 16 (51.6%) 258 (38.1%) 1.728 (0.840-3.555) 27.2 (±4.7) 23.6 (±3.7) 1.222 (1.125-1.328) 0.973 (0.763-1.241) 94.3 (±13.1) 82.8 (±10.5) 1.095 (1.059-1.132) 1.102 (0.097-1.215) 102.4 (±8.1) 96.7 (±6.7) 1.105 (1.055-1.157) 1.012 (0.904-1.132) 141.8 (±22.5) 131.4 (±19.1) 1.025 (1.008-1.042) 1.004 (0.979-1.030) 81.6 (±14.1) 78.3 (±12.9) 1.020 (0.992-1.049) 2.8 (±0.9) 2.8 (±0.8) 1.063 (0.688-1.642) 1.4 (±0.5) 1.6 (±0.6) 0.583 (0.288-1.179) 1.3 (±0.7) 1.1 (±0.7) 1.291 (0.857-1.945) 17 (54.8%) 161 (23.8%) 3.891 (1.873-8.064) 2.596 (0.809-8.323) 8 (25.8%) 61 (9.0%) 3.484 (1.497-8.130) 2.248 (0.635-7.936) 26 (83.9%) 626 (92.5%) 0.349 (0.128-0.956) 0.247 (0.062-0.983) 44.5 (±2.4%) 44.5 (±2.4%) 0.917 (0.788-1.066) 11.0 (±4.7) 10.9 (±5.4) 1.003 (0.939-1.071) 32 (24-44) 25 (19-35) 1.009 (1.001-1.019) 1.053 (0.984-1.108) 36 (26-47) 25 (21-30) 1.030 (1.012-1.049) 1.109 (1.044-1.177) 191 (154-230) 224 (190-261) 0.983 (0.975-0.991) 0.979 (0.968-0.989) 1,170 (56-63,500) 849 (106-9,575) 1.001 (0.998-1.002) 287 (±67) 238 (±57) 1.015 (1.008-1.021) 1.015 (1.004-1.027) hepatitis B e antigen; ALT, alanine aminotransferase; AST, aspartate aminotransferase; CAP, controlled attenuation

22

ACCEPTED MANUSCRIPT

RI PT

Table 3. Univariate and multivariate analysis of factors associated with advanced fibrosis among patients treated with nucleoside analogue therapy for at least 3 years (n=681). Significant factors via multivariate analysis are highlighted in bold. Increased CAP measurements were independently associated with severe fibrosis.

M AN U

TE D

EP

AC C

Age (years) Male patients, n (%) 2 BMI (kg/m ) Waist circumference (cm) Hip circumference (cm) Systolic blood pressure (mmHg) Diastolic blood pressure (mmHg) Low-density lipoprotein (mmol/L) High-density lipoprotein (mmol/L) Triglyceride (mmol/L) Metabolic syndrome, n (%) Diabetes, n (%) HBeAg-negative, n (%) Treatment duration, months Albumin (g/L) Bilirubin (umol/L) ALT (U/L AST (U/L) 9 Platelet x 10 /L Undetectable HBV DNA, n (%) CAP (dB/M) BMI, body mass index; HBeAg, parameter Severe liver fibrosis defined by EASL-ALEH criteria

SC

Severe fibrosis No severe fibrosis Univariate (n=147) (n=534) Odds ratio (95% CI) 59.2 (±8.8) 54.2 (±10.9) 1.049 (1.029-1.069) 111 (75.5%) 354 (66.3%) 1.568 (1.034-2.378) 24.9 (±4.3) 23.3 (±3.3) 1.122 (1.067-1.180) 90.4 (±11.2) 84.7 (±6.3) 1.050 (1.032-1.069) 98.4 (±7.4) 95.9 (±6.3) 1.060 (1.031-1.091) 140.7 (±23.9) 135.2 (±18.8) 1.013 (1.004-1.022) 82.0 (±16.8) 79.9 (±11.9) 1.010 (0.996-1.024) 2.5 (±0.9) 2.6 (±0.6) 0.733 (0.567-0.948) 1.3 (±0.5) 1.4 (±0.5) 0.420 (0.254-0.690) 1.3 (±0.9) 1.1 (±0.6) 1.254 (0.974-1.614) 68 (46.3%) 167 (31.3%) 1.890 (1.304-2.747) 48 (32.7%) 60 (11.2%) 3.773 (2.439-5.882) 143 (97.3%) 468 (87.7%) 5.052 (1.810-14.413) 82.6 (64.6-108.7) 87.5 (67.9-114.5) 0.995 (0.980-1.011) 43.2 (±4.3) 45.5 (±2.5) 0.793 (0.743-0.846) 16.0 (±9.5) 11.4 (±5.6) 1.092-1.063-1.122 30 (23-37) 29 (19-34) 1.015 (1.007-1.023) 30 (25-36) 26 (22-30) 1.053 (1.034-1.072) 129 (85-184) 195 (151-236) 0.984 (0.980-0.987) 145 (98.6%) 500 (93.6%) 4.926 (1.170-20.833) 253 (±62) 233 (±57) 1.006 (1.003-1.009) hepatitis B e antigen; ALT, alanine aminotransferase; AST, aspartate aminotransferase; CAP,

23

Multivariate Odds ratio (95% CI) 0.997 (0.969-1.026) 0.816 (0.442-1.505) 1.067 (0.929-1.227) 1.003 (0.962-1.045) 0.994 (0.937-1.056) 1.005 (0.993-1.017) 1.104 (0.820-1.487) 1.013 (0.570-1.801) 1.068 (0.744-1.533) 1.096 (0.594-2.022) 2.487 (1.344-4.608) 2.541 (0.795-8.121) 0.842 (0.773-0.916) 1.046 (1.015-1.098) 0.987 (0.967-1.007) 1.056 (1.015-1.098) 0.985 (0.981-0.990) 1.706 (0.346-8.403) 1.008 (1.003-1.014) controlled attenuation

ACCEPTED MANUSCRIPT Figure Legends Figure 1. Patient disposition of the current study.

RI PT

Figure 2. Prevalence of (A) steatosis, severe steatosis, severe fibrosis and cirrhosis as stratified by treatment status; and (B) severe fibrosis as stratified by different degrees of steatosis.

AC C

EP

TE D

M AN U

SC

Figure 3. Odds ratios, as stratified by nucleoside analogue duration, of severe steatosis in association with severe liver fibrosis.

24

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

ACCEPTED MANUSCRIPT

RI PT

Supplementary Table 1. Univariate and multivariate analysis of factors associated with severe liver fibrosis among patients treated with nucleoside analogue therapy for at least 5 years (n=550). Significant factors via multivariate analysis are highlighted in bold. Increased CAP measurements were independently associated with severe fibrosis.

M AN U

TE D

EP

AC C

Age (years) Male patients, n (%) 2 BMI (kg/m ) Waist circumference (cm) Hip circumference (cm) Systolic blood pressure (mmHg) Diastolic blood pressure (mmHg) Low-density lipoprotein (mmol/L) High-density lipoprotein (mmol/L) Triglyceride (mmol/L) Metabolic syndrome, n (%) Diabetes, n (%) HBeAg-negative, n (%) Treatment duration, months Albumin (g/L) Bilirubin (umol/L) ALT (U/L AST (U/L) 9 Platelet x 10 /L Undetectable HBV DNA, n (%) CAP (dB/M) BMI, body mass index; HBeAg, parameter Severe liver fibrosis defined by EASL-ALEH criteria

SC

Severe fibrosis No severe fibrosis Univariate Multivariate (n=114) (n=436) Odds ratio (95% CI) Odds ratio (95% CI) 59.7 (±8.6) 54.2 (±10.9) 1.045 (1.023-1.067) 0.984 (0.953-1.016) 80 (70.2%) 300 (68.8%) 1.067 (0.680-1.672) 25.0 (±4.1) 23.3 (±3.3) 1.135 (1.071-1.204) 1.140 (0.971-1.339) 90.2 (±11.5) 84.6 (±10.7) 1.046 (1.026-1.067) 0.986 (0.941-1.034) 98.3 (±7.2) 95.7 (±6.3) 1.061 (1.028-1.095) 0.973 (0.909-1.043) 141.9 (±23.9) 135.9 (±19.1) 1.014 (1.004-1.024) 1.003 (0.989-1.017) 82.8 (±17.8) 80.2 (±12.1) 1.014 (0.995-1.034) 2.5 (±0.9) 2.6 (±0.8) 0.764 (0.576-1.006) 1.103 (0.785-1.549) 1.3 (±0.5) 1.4 (±0.5) 0.651 (0.420-1.011) 1.303 (0.679-2.502) 1.438 (1.028-1.962) 1.114 (0.739-1.679) 1.3 (±1.0) 1.1 (±.0.6) 2.314 (1.524-3.521) 1.418(0.703-2.857) 59 (51.8%) 138 (31.7%) 41 (36.0%) 49 (11.2%) 4.385 (2.702-7.143) 3.030 (1.511-6.061) 7.418 (1.178-30.949) 4.707 (0.971-22.810) 112 (98.2%) 385 (88.3%) 0.996 (0.989-1.002) 100.6 (76.1-115.0) 105.3 (78.8-124.6) 43.7 (±4.3) 45.6 (±2.5) 0.795 (0.740-0.854) 0.829 (0.753-0.913) 1.104 (1.068-1.141) 1.029 (0.982-1.079) 15.7 (±9.7) 11.2 (±4.9) 1.015 (1.006-1.025) 0.980 (0.959-1.002) 32 (23-37) 26 (20-34) 30 (25-37) 25 (22-30) 1.057 (1.036-1.080) 1.067 (1.021-1.115) 133 (83-191) 195 (153-237) 0.984 (0.980-0.988) 0.985 (0.979-0.990) 3.401 (0.794-14.705) 112 (98.2%) 411 (94.3%) 255 (±65) 233 (±58) 1.006 (1.002-1.009) 1.008 (1.002-1.014 hepatitis B e antigen; ALT, alanine aminotransferase; AST, aspartate aminotransferase; CAP, controlled attenuation

ACCEPTED MANUSCRIPT

RI PT

Supplementary Table 2. Univariate and multivariate analysis of factors associated with severe liver fibrosis among patients treated with nucleoside analogue therapy for at least 7 years (n=358). Significant factors via multivariate analysis are highlighted in bold. Increased CAP measurements were independently associated with severe fibrosis.

M AN U

TE D

EP

AC C

Age (years) Male patients, n (%) 2 BMI (kg/m ) Waist circumference (cm) Hip circumference (cm) Systolic blood pressure (mmHg) Diastolic blood pressure (mmHg) Low-density lipoprotein (mmol/L) High-density lipoprotein (mmol/L) Triglyceride (mmol/L) Metabolic syndrome, n (%) Diabetes, n (%) HBeAg-negative, n (%) Treatment duration, months Albumin (g/L) Bilirubin (umol/L) ALT (U/L AST (U/L) 9 Platelet x 10 /L Undetectable HBV DNA, n (%) CAP (dB/M) BMI, body mass index; HBeAg, parameter Severe liver fibrosis defined by EASL-ALEH criteria

SC

Severe fibrosis No severe fibrosis Univariate Multivariate (n=71) (n=287) Odds ratio (95% CI) Odds ratio (95% CI) 57.7 (±8.6) 54.1 (±10.4) 1.036 (1.009-1.064) 0.977 (0.940-1.016) 50 (70.4%) 203 (70.7%) 0.985 (0.557-1.741) 23.4 (±3.4) 24.8 (±4.0) 1.112 (1.034-1.197) 1.240 (1.027-1.497) 89.6 (±11.2) 85.0 (±10.2) 1.044 (1.017-1.071) 0.959 (0.894-1.029) 98.2 (±6.9) 96.2 (±6.2) 1.049 (1.008-1.093) 0.974 (0.886-1.071) 140.4 (±21.3) 136.1 (±17.9) 1.012 (0.998-1.026) 1.005 (0.986-1.024) 83.0 (±17.9) 81.2 (±11.5) 1.010 (0.991-1.030) 2.5 (±0.9) 2.6 (±0.7) 0.850 (0.606-1.192) 1.2 (±0.4) 1.3 (±0.5) 0.566 (0.313-1.025) 1.049 (0.440-2.498) 1.2 (±0.6) 1.2 (±0.6) 1,115 (0.739-1.682) 37 (52.1%) 93 (32.4%) 2.268 (1.340-3.846) 2.433 (1.045-5.682) 21 (29.6%) 35 (12.2%) 2.985 (1.608-5.556) 1.349 (0.544-3.344) 69 (97.2%) 258 (89.9%) 3.875 (0.903-16.667) 3.106 (0.620-15.625) 111.0 (97.1-132.2) 111.4 (95.2-148.2) 0.995 (0.986-1.003) 43.5 (±4.7) 45.8 (±2.5) 0.812 (0.746-0.884) 0.817 (0.726-0.920) 15.3 (±7.8) 11.4 (±5.0) 1.102 (1.057-1.149) 1.038 (0.978-1.102) 32 (23-37) 27 (20-36) 1.013 (1.003-1.023) 0.987 (0.964-1.011) 30 (25-36) 26 (22-30) 1.050 (1.025-1.076) 1.046 (0.955-1.099) 144 (75-199) 197 (161-237) 0.985 (0.981-0.990) 0.984 (0.977-0.991) 70 (98.6%) 270 (94.1%) 4.407 (0.577-33.688) 252 (±59) 235 (±60) 1.005 (1.000-1.009) 1.007 (1.000-1.015) hepatitis B e antigen; ALT, alanine aminotransferase; AST, aspartate aminotransferase; CAP, controlled attenuation

ACCEPTED MANUSCRIPT

No severe fibrosis (n)

Treatment-naïve

22

566

Treatment duration ≥3 years

122

458

Treatment duration ≥5 years

94

372

Treatment duration ≥7 years

59

239

Univariate

Multivariate*

Odds ratio (95% CI)

Odds ratio (95% CI)

1.024 (1.012-1.036)

1.020 (1.003-1.038)

1.010 (1.005-1.014)

1.015 (1.007-1.022)

1.010 (1.006-1.015)

1.018 (1.009-1.027)

1.011 (1.005-1.018)

1.024 (1.012-1.036)

M AN U

SC

Severe fibrosis (n)

RI PT

Supplementary Table 3. Sensitivity analysis of the association of CAP measurements with severe fibrosis among patients with low risk of false-positive results among different patient subgroups

Low risk of false-positive results defined as CAP<300 dB/m, CAP300-339 dB/m and liver stiffness >12.5kPa, and CAP >340 dB/m and liver stiffness>13.6 kPa.22 CAP, controlled attenuation parameter; BMI, body mass index; SBP, systolic blood pressure, LDL, low-density lipoprotein; HDL, high-density lipoprotein; HBeAg, hepatitis B e antigen; ALT, alanine aminotransferase; AST, aspartate aminotransferase

AC C

EP

TE D

*The following parameters were included in the multivariate analyses: Treatment-naive: Age, BMI, waist circumference, hip circumference, SBP, metabolic syndrome, diabetes, HBeAg-negative, ALT, AST, platelet, CAP. Treatment duration ≥3 years: Age, male patients, BMI, waist circumference, hip circumference, SBP, LDL, HDL, metabolic syndrome, diabetes, HBeAg-negative, albumin, bilirubin, ALT, AST, platelet, undetectable HBV DNA, CAP. Treatment duration ≥5 years: Age, BMI, waist circumference, hip circumference, SBP, LDL, HDL, triglyceride, metabolic syndrome, diabetes, HBeAg-negative, albumin, bilirubin, ALT, AST, platelet, CAP. Treatment duration ≥7 years: Age, BMI, waist circumference, hip circumference, SBP, HDL, triglyceride, metabolic syndrome, diabetes, HBeAg-negative, albumin, bilirubin, ALT, AST, platelet, CAP.

ACCEPTED MANUSCRIPT

No severe fibrosis

(n)

(n)

Treatment-naïve

31

677

Treatment duration ≥3 years

147

534

Treatment duration ≥5 years

114

436

Treatment duration ≥7 years

71

287

Univariate

Multivariate*

Odds ratio (95% CI)

Odds ratio (95% CI)

1.015 (1.008-1.023)

1.015 (1.004-1.027)

1.005 (1.001-1.008)

1.008 (1.003-1.013)

1.004 (1.000-1.008)

1.008 (1.002-1.014)

1.005 (1.000-1.010)

1.007 (1.001-1.014)

M AN U

SC

Severe fibrosis

RI PT

Supplementary Table 4. Sensitivity analysis of the association of CAP measurements with severe fibrosis after optimizations as suggested by Karlas et al.19

Optimizations as suggested by Karlas et al19 are deductions of 10 dB/m for diabetes patients and deductions / additions of 4.4 dB/m for each unit of body mass index above / below 25 kg/m2

TE D

CAP, controlled attenuation parameter

AC C

EP

*Parameters included in multivariate analysis were the same as Tables 2, 3, and Supplementary Tables 1 and 2.