Obesity and hepatocellular carcinoma in patients receiving entecavir for chronic hepatitis B

Study design and patients

This study was retrospectively conducted from a historical cohort at Bundang Jesaeng Hospital, Seongnam, Korea. From January 2009 to December 2012, 150 patients were treated with entecavir as an initial treatment for CHB and checked with body composition analyzer (Fig. 1). CHB was defined when hepatitis B virus surface antigen (HBsAg) is positive for at least 6 months. Of these patients, we excluded 48 patients: (i) diagnosed with HCC before starting entecavir (n=14); (ii) diagnosed with cholangiocarcinoma before starting entecavir (n=1); (iii) liver transplantation before starting entecavir (n=2); (iv) incomplete data on laboratory studies (n=13); (v) loss of follow-up within 6 months (n=18). Therefore, we finally analyzed 102 patients in this study. The diagnosis of HCC was based on the American Association for the Study of Liver Diseases (AASLD) guidelines [15]. The criteria for starting antiviral therapy was based on the Korean Association for the Study of the Liver (KASL) guideline: (i) hepatitis B virus (HBV) DNA ≥20,000 IU/mL and serum aspartate transaminase (AST) or alanine transaminase (ALT) level ≥2 × upper limit of normal (ULN) in HBV envelope antigen (HBeAg) positive patients, (ii) HBV DNA ≥2,000 IU/mL and serum aminotransferase level ≥2 × ULN in HBeAg negative patients, and (iii) HBV DNA ≥2,000 IU/mL and serum AST or ALT ≥ ULN in liver cirrhosis [16]. This study was approved by the Institutional Review Board of Bundang Jesaeng Hospital (IRB NO: IMH15-03).

Figure 1.

Flow diagram of enrolled patients. HCC, hepatocellular carcinoma; CCC, cholangiocarcinoma.

Clinical parameters

We investigated baseline characteristics of the patients: age, gender, total bilirubin, AST, ALT, platelet count, albumin, prothrombin time, ascites, hepatic encephalopathy, presence of HBeAg, HBV DNA level, presence of liver cirrhosis, Child-Pugh classification, AST to Platelet Ratio Index (APRI), fibrosis-4 (FIB-4), total cholesterol, triglyceride, high-density lipoprotein (HDL) cholesterol, low-density lipoprotein (LDL), alcohol intake, presence of diabetes mellitus, hypertension, and metabolic syndrome. Serum HBV DNA level was checked by COBAS TaqMan quantitative test for hepatitis B virus (Roche Molecular Systems Inc., Branchburg, NJ, USA). Liver cirrhosis was diagnosed based on ultrasonographic findings [17]. Significant alcohol consumption is defined as drinking above 210 gram/week for men and above 140 gram/week for women [18]. Metabolic syndrome was defined by updated National Cholesterol Education Program Adult Treatment Panel III standards [19,20].

Assessment of obesity and hepatic steatosis

For evaluating obesity, following variables were checked in this study: body mass index (BMI), waist circumference (WC), waist-to-hip ratio (WHR), and visceral fat area (VFA). These variables were measured by electrical bioimpedance analysis and those were measured using IOI-353 Body Composition Analyzer (Jawon Medical Co., Seoul, Korea).

Hepatic steatosis (fatty liver) was evaluated by ultrasonography. Trained hepatologists (YM. Park and SH. Yoo) performed ultrasonography. For this study, ultrasonographic hepatic steatosis was newly assessed by two hepatologists (SH. Yoo and W. Sohn). They reviewed ultrasonographic images of all patients without knowing any clinical information. To assess hepatic steatosis, we used Hamaguchi’s ultrasonographic score, from 0 to 6 points (S0-S6) considering bright hepatic echogenicity, hepatorenal echo contrast, deep attenuation and vessel blurring [21]. Based on this scoring system, mild hepatic steatosis was defined by score 2-3 (S2-S3) and moderate/severe hepatic steatosis was defined by ≥score 4 (≥S4). Sonographic hepatic steatosis was scored based on an agreement by two hepatologists (SH. Yoo and W. Sohn). In case of disagreement, we re-reviewed the sonographic images and determined the score by agreement. Unless the cases are not agreed in spite of re-assessment, we determined the mean value of two scoring measurements as the final result.

Follow-up after starting antiviral treatment

After starting entecavir, enrolled patients underwent imaging study (abdominal ultrasonography and/or contrast-enhanced multiphase computed tomography) and serum alpha-fetoprotein every 6 months. Magnetic resonance imaging or liver biopsy was additionally used to confirm HCC if necessary. We also assessed disease progression in patients after starting entecavir. Disease progression was defined as the first occurrence one of following events: i) increase by ≥2 points of Child-Pugh score compared to baseline; ii) spontaneous bacterial peritonitis; iii) renal insufficiency; iv) variceal bleeding; v) HCC development; vi) death [2]. To evaluate the efficacy of entecavir, serum laboratory tests such as serum aminotransferase level, HBeAg, hepatitis B virus enveolope antibody (HBeAb) and hepatitis B virus DNA level were performed every 3 or 6 months.

Assessment of outcomes

Time to HCC development, risk factors affecting for HCC development, virological response to entecavir, and overall survival were investigated. Also, we evaluated relationship between obesity-related factors and hepatitis B viral load before entecavir treatment. Time to HCC development and disease progression were defined as the duration from an initiation of entecavir to a detection of HCC or the first occurrence one of above disease progression events, respectively. Virological response to entecavir was checked as a proportion of patients with HBV DNA level <69 IU/mL at 48 weeks and 96 weeks, respectively. Biochemical response to entecavir was checked as a proportion of patients with serum ALT level <40 IU/mL at 48 weeks and 96 weeks, respectively. The follow-up period for survival analysis was defined as the duration between an initiation of entecavir and either a development of event or the last visit to outpatient clinic before December 31, 2015.

Statistical analysis

Categorical variables were presented as frequencies and percentages. Continuous variables were demonstrated as mean value with standard deviation (SD) and median value with interquartile range. We assessed the inter-observer variability of Hamaguchi’s score using kappa value (κ). That was interpreted as bellows: 0.01-0.20, slight agreement; 0.21-0.40, fair agreement; 0.41-0.60, moderate agreement; 0.61-0.80, substantial agreement; 0.81-0.99, almost perfect agreement [22]. Kaplan-Meier curve was used to calculate the cumulative incidence of HCC and overall survival. Log-rank test was used to compare the incidence of HCC according to risk factors. Univariable and multivariable Cox proportional hazards model was used to identify the risk factors for the development of HCC. Multivariable analysis was done using a backward conditional stepwise procedure to avoid the multicollinearity. Spearman correlation coefficient (r) was used for investigating the relationship of baseline HBV DNA (before starting entecavir) and obesity associated factors (BMI, WC, WHR, and VFA). All statistical analysis was done using SPSS for windows release 18.0 (SPSS Inc., Chicago, IL, USA). When P-value was <0.05 in a two-sided test, it was regarded as statistically significant.

Baseline characteristics

Baseline characteristics are shown in Table 1. The median age of patients was 46.4 years and the majority of patients were men (66%). The mean level of serum ALT was 224.9 IU/mL. HBeAg positivity was observed in 71 patients (70%). Thirty six patients had cirrhotic patients (35%). Child-Pugh class of patients was class A in 94 (92%), class B in 6 (6%), and class C in 2 (2%) patients. The median level of HBV DNA was 7,340,040.5 IU/mL. Diabetes mellitus, hypertension, and metabolic syndrome were observed in 4 (4%), 5 (5%), and 16 (16%). The median levels of serum fasting glucose, triglyceride, and LDL cholesterol were 96.2, 101.5, and 104.8 mg/dL. The degree of hepatic steatosis none in 78 patients (76%), mild in 19 patients (19%), and moderate or severe in 5 patients (5%). The inter-observer agreement of hepatic steatosis was good (κ=0.879, 95% C.I [0.821-0.918]). The mean values of BMI, WC, WHR, and VFA were 23.6 kg/m², 82.5 cm, 0.86 and 89.1 cm².

Table 1.

Baseline characteristics of enrolled patients (N=102)

Relationship between hepatitis B viral load and obesity associated risk factors

The relationship between baseline HBV DNA level (before starting entecavir) and obesity-related factors (BMI, WC, WHR, and VFA) was demonstrated in Figure 2. There was a significant inverse correlation between HBV DNA level and BMI (r=-0.208, P=0.036), WHR (r=-0.285, P=0.004), and VFA (r=-0.293, P=0.003) (Fig. 2A, 2C, and 2D). Relationship between HBV DNA level and WC was inverse but it did not reach a statistically significance (r=-0.188, P=0.059) (Fig. 2B).

Figure 2.

Correlations between hepatitis B viral load and obesity-related factors upon starting entecavir. “r” is the correlation coefficient between log₁₀HBV DNA and each risk factors. Statistical significance is regarded when p-value is less than 0.05.

We investigated the change of obesity associated factors after hepatitis B viral load was suppressed thorough antiviral treatment. Of 102 patients, 86 patients were checked with bioelectrical impedance analysis for obesity after hepatitis B viral load was suppressed. There is no significant difference in the change of obesity associated factors when hepatitis B viral load was suppressed: BMI (23.6 ± 3.1 -> 23.6 ± 3.0, P=0.918); WC (82.5 ± 9.4 -> 81.9 ± 8.6, P=0.899); WHR (0.86 ± 0.08 -> 0.87 ± 0.08, P=0.081); VFA (89.1 ± 40.8 -> 90.2 ± 38.6, P=0.401).

HCC incidence in CHB patients treated with entecavir

The median follow-up duration of the patients was 45.2 (interquartile range: 36.0-58.3) months. Of 102 patients, HCC development was observed in 7 patients. The cumulative incidence rates of HCC at 1 year, 3 years, and 5 years were 0%, 5.3%, and 9.0%, respectively. The mean duration of time to HCC development was 71.5 (95% CI: 68.7-74.4) months. The median duration of entecavir administration was 38.0 (interquartile range: 23.6-51.9) months. Of 102 patients, death was observed in 3 patients.

Effect of obesity associated factors on virological response, biochemical response, and HBeAg seroconversion

Virological response rates of entecavir were 80% (76/95) at 48 weeks and 94% (79/84) at 96 weeks. Biochemical response rates of entecavir were 79% (75/95) at 48 weeks and 86% (72/84) at 96 weeks. Drug resistance was detected in one patient. The rate of HBeAg seroconversion was 38% (27/71).

The differences in virological response, biochemical response, and HBeAg seroconversion according to obesity associated factors were demonstrated in Figure 3. There was no statistical difference in virological response at 96 weeks and HBeAg serocon-version according to obesity associated factors (BMI, WC, WHR, VFA, and hepatic steatosis) (P>0.05). On the other hand, biochemical response to entecavir (ALT <40 U/L at 96 weeks) was significantly worse in patients with WHR ≥0.9 than in those with WHR <0.9 (non-responder: 26.5% vs. 6.0%, P=0.012). Also, patients with hepatic steatosis has worse biochemical response at 96 weeks than those without hepatic steatosis (non-responder: 38.1% vs. 6.3%, P=0.001).

Figure 3.

Differences in virological response at 96 weeks (A), biochemical response at 96 weeks (B), and HBeAg seroconversion (C) in CHB patients treated with entecavir. BMI, body mass index; WHR, waist-hip-ratio; VFA, visceral fat area; CHB, chronic hepatitis B; HBeAg, hepatitis B virus envelope antigen. ^*Statistical significance is regarded when P-value is less than 0.05.

Risk factors for HCC development in CHB patients treated with entecavir

Risk factors for HCC development were shown in Table 2. Univariable analysis revealed that the risk factors for HCC development were a platelet count of <120,000 /mm³ (HR 5.21, P=0.031), HBeAg negativity (HR 5.61, P=0.039), and liver cirrhosis (HR 10.26, P=0.031). However, none of obesity-related risk factors were significantly associated with HCC development in CHB patients treated with entecavir: BMI ≥25 kg/m² (HR 0.90, P=0.894), WC ≥90 cm (HR 1.10, P=0.912), WHR ≥0.9 (HR 1.94, P=0.386), VFA ≥100 cm² (HR 1.69, P=0.495), and hepatic steatosis (HR 0.57, P=0.602). Multivariable analysis showed that the significant risk factor for HCC development was liver cirrhosis (HR 9.07, P=0.042). HBeAg negativity was associated with HCC development, but it didn’t reach a statistical significance (HR 4.84, P=0.06).

Table 2.

Results of univariable and multivariable analyses of risk factors for HCC development in patients treated with entecavir for CHB (N=102)

We compared the difference of HCC incidence according to risk factors (Fig. 4). Cumulative incidence rates of HCC at 1 year, 3 years, and 5 years were 0%, 1.9% and 1.9%, respectively, in noncirrhotic patients and 0%, 11.4%, and 19.1%, respectively, in cirrhotic patients (Fig. 4A). The mean durations of time to HCC development were 74.7 (95% CI: 73.0-76.3) and 60.8 (95% CI: 55.4-66.2) months in non-cirrhotic and cirrhotic patients, respectively. HCC incidence was significantly different according to the presence of liver cirrhosis (P=0.008). However, HCC incidence was not significantly different according to obesity-related risk factors such as BMI (Fig. 4B), presence of diabetes mellitus, dyslipidemia, hypertension, metabolic syndrome (Fig. 4C), WHR (Fig. 4D), VFA (Fig. 4E), and hepatic steatosis (Fig. 4F).

Figure 4.

Differences in HCC development according to cirrhosis (A), BMI (B), diabetes mellitus/dyslipidemia/hypertension/metabolic syndrome (C), WHR (D), VFA (E), and hepatic steatosis (F) in CHB patients treated with entecavir. HCC, hepatocellular carcinoma; BMI, body mass index; DM, diabetes mellitus; HTN, hypertension; MS, metabolic syndrome; WHR, waist-hip-ratio; VFA, visceral fat area; CHB, chronic hepatitis B.

Risk factors for disease progression in CHB patients treated with entecavir

Of 102 patients, 11 patients were regarded as disease progression after antiviral treatment: increase by ≥2 points of CP score (n=1); SBP (n=1); renal insufficiency (n=1); variceal bleeding (n=1); HCC (n=7); death (n=3) – some patients have experienced more than 2 events. The median duration of time to disease progression was 67.8 (95% CI: 63.2-72.4) months. Risk factors for disease progression were demonstrated in Table 3. Univariable analysis indicated that the risk factors for disease progression were a platelet count of <120,000 /mm³ (HR 5.90, P=0.005), HBV DNA level ≥1x10⁷ IU/mL (HR 0.09, P=0.022), HBeAg negativity (HR 3.59, P=0.048), and liver cirrhosis (HR 4.46, P=0.027). However, none of obesity-related risk factors were significantly associated with disease progression in CHB patients treated with entecavir: BMI ≥25 kg/m² (HR 0.98, P=0.978), WC ≥90 cm (HR 0.95, P=0.945), WHR ≥0.9 (HR 2.47, P=0.164), VFA ≥100 cm² (HR 2.12, P=0.245), and hepatic steatosis (HR 0.85, P=0.840). Multivariable analysis indicated that the significant risk factor for disease progression was a platelet count of <120,000 /mm³ (HR 4.06, P=0.031). HBV DNA level ≥1×10⁷ IU/mL was associated with low occurrence of disease progression, but it didn’t reach a statistical significance (HR 0.13, P=0.055).

Table 3.

Results of univariable and multivariable analyses of risk factors for disease progression in patients treated with entecavir for CHB (N=102)