Validation of non-invasive diagnosis of CSPH in patients with compensated advanced chronic liver disease according to Baveno VII.

BACKGROUND AND AIMS: Baveno VII consensus introduced the non-invasive criteria of clinically significant portal hypertension (CSPH) using liver stiffness measurement (LSM). We evaluated the usefulness of the Baveno VII criteria to predict the risk of decompensation in patients with compensated advanced chronic liver disease (cACLD). METHODS: We conducted a retrospective cohort study of 1966 patients with cACLD. Patients were categorized into four groups (CSPH excluded (n = 619), grey zone (low risk of CSPH (n = 699), high risk of CSPH (n = 207)), and CSPH included (n = 441)) according to Baveno VII consensus. The risk of events was estimated using a Fine and Gray competing risk regression analysis, with liver transplantation and death as competing events. We calculated standardized hazard ratios (sHR) to assess the relative risk of decompensation. RESULTS: Among 1966 patients, 178 developed decompensations over a median follow-up of 3.06 (IQR: 1.03-6.00) years. Patients with CSPH had the highest decompensation risk, followed by the grey zone high-risk group, grey zone low-risk group, and those without CSPH with 3-year cumulative risks of 22%, 12%, 3.3%, and 1.4% respectively (p < .001). Compared to CSPH excluded group, CSPH included group (sHR: 8.00, 95% CI: 4.00-16.0), grey zone high-risk group (sHR: 6.57, 95% CI: 3.16-13.6), grey zone low-risk group (sHR: 2.15, 95% CI: 1.04-4.41) had significantly higher risk of decompensation (Gray's test p < .01). CONCLUSION: Non-invasive diagnosis of CSPH according to the Baveno VII criteria can stratify the risk of decompensation.

The Association Between Breastfeeding and Nonalcoholic Fatty Liver Disease in Parous Women: A Nation-wide Cohort Study.

BACKGROUND AND AIMS: Breastfeeding has multiple effects on maternal health outcomes. However, the effect of breastfeeding on NAFLD in parous women remains unclear. APPROACH AND RESULTS: A total of 6,893 Korean parous women aged 30-50 years who participated in the Korean National Health and Nutrition Examination Survey were assessed for the association between breastfeeding and NAFLD. Duration of lactation was calculated by dividing the total lactation period by the number of breastfed children. NAFLD was defined by the hepatic steatosis index. Of 6,893 women, 1,049 (15.2%) had NAFLD. Prevalence of NAFLD was 18.3%, 14.3%, 12.3%, 14.4%, and 15.8% in women with a breastfeeding period of <1, ≥1-<3, ≥3-<6, ≥6-<12, and ≥12 months, respectively. In a fully adjusted model, breastfeeding (≥1 month) was associated with reduced NAFLD prevalence (OR, 0.67; 95% CI, 0.51-0.89) after adjusting for metabolic, socioeconomic, and maternal risk factors. Fully adjusted ORs (95% CI) decreased with an increase of breastfeeding duration: 0.74 (0.49-1.11), 0.70 (0.47-1.05), 0.67 (0.48-0.94), and 0.64 (0.46-0.89) for women with ≥1-<3, ≥3-<6, ≥6-<12, and ≥12 months of breastfeeding duration, respectively, compared to women with <1 month of breastfeeding duration. Such an association was also observed in all predefined subgroups without interaction. CONCLUSIONS: Breastfeeding showed a protective effect against NAFLD in later life of parous women, suggesting a maternal benefit of breastfeeding on NAFLD.

Incidence and risk factors for development of hepatocellular carcinoma at young age in patients with chronic hepatitis B.

BACKGROUND: Some young adults with chronic hepatitis B virus (HBV) infection might be at high risk for hepatocellular carcinoma (HCC), enough to justify regular HCC surveillance despite the young age of the patients. However, ways to identify at-risk individuals who may benefit from HCC surveillance need further evaluations. METHODS: A hospital-based retrospective cohort of 2757 chronic HBV mono-infected young adults (median age: 34 years, males 66%) were analyzed. The primary outcome was young-onset HCC, defined as a diagnosis made under 40 years of age. We calculated the HCC incidence/1000 person-years in the overall cohort and pre-defined subgroups of patients assessed the independent risk factors that can be used to identify surveillance targets. RESULTS: The HCC incidence was low (2.55/1000 person-years) in the overall cohort. However, the HCC incidence varied widely according to baseline characteristics: lowest among young adults with FIB-4 ≤ 0.70 (0.17/1000 person-years) and highest in young adults with radiological cirrhosis (30.7/1000 person-years). In multivariable analysis, radiological cirrhosis, the FIB-4 index, and serum HBV DNA level were independent factors associated with HCC development at a young age. Performance for prediction of young-onset HCC in radiological cirrhotic patients showed the highest specificity but sensitivity was <70%. Combination with FIB-4 index and HBV DNA levels increased sensitivity to 90%. CONCLUSION: Risk stratification using FIB-4 index, HBV DNA levels, and either combining radiological cirrhosis or gender and AFP levels would be helpful to stratify young patients who would and would not benefit from regular HCC surveillance.

Imaging features of hepatobiliary MRI and the risk of hepatocellular carcinoma development.

OBJECTIVE: This study aimed to determine whether hepatocellular carcinoma (HCC) risk and time to HCC development differ according to hepatobiliary magnetic resonance imaging (MRI) findings among people at risk for developing HCC. MATERIALS AND METHODS: A total of 199 patients aged 40 years or older with liver cirrhosis or chronic liver disease who underwent gadoxetic acid-enhanced hepatobiliary MRI between 2011 and 2015 were analyzed. An independent radiologist retrospectively reviewed MRI findings, blinded to clinical information, and categorized them into low-risk features, high-risk features and high-risk nodules. High-risk features were defined as liver cirrhosis diagnosed by imaging. High-risk nodules were defined as LR-3 or LR-4 nodules based on LI-RADS version 2018. The primary outcome was development of HCC within 5-year of MRI evaluation. RESULTS: HCC was diagnosed in 28 patients (14.1%). HCC development was null for those with low-risk features (n = 84). The cumulative incidence rates of HCC were 0%, 2.3%, 13.4% and 22.1% at 1-, 2-, 3- and 5-year for those with high-risk features (n= 64), and were 19.1%, 31.8%, 37.3% and 46.7% at 1-, 2-, 3- and 5-year for those with high-risk nodules (n= 51). Among 28 patients developed HCC, the median time from baseline MRI to HCC diagnosis was 33.1 months (interquartile range: 25.9-46.7 months) for high-risk feature group, and 17.3 months (interquartile range: 6.2-26.5 months) for high-risk nodule group. CONCLUSIONS: HCC risk and time to HCC development differ according to baseline hepatobiliary MRI findings, indicating that hepatobiliary MRI findings can be used as biomarkers to differentiate HCC risk.

Clinical Outcomes of Hepatitis B Virus-Related Hepatocellular Carcinoma Patients with Undetectable Serum HBV DNA Levels.

BACKGROUND AND AIMS: Some hepatitis B virus (HBV)-related hepatocellular carcinoma (HCC) patients show undetectable serum HBV DNA levels at HCC diagnosis. The risk of HBV reactivation and its impact on clinical outcomes are not well-unknown. METHODS: This retrospective cohort study included a total of 985 HBV-related HCC patients with undetectable serum HBV DNA levels (< 12 IU/mL) at HCC diagnosis (112 were antiviral treatment (AVT)-naïve; 873 were receiving AVT). Incidence and risk factors for HBV reactivation (re-detection of HBV DNA in serum) during follow-up, as well as its association to overall survival, were assessed. RESULTS: During a median of 33.4 months of follow-up (range: 0.2-124.2 months), HBV reactivation was observed in 279 patients. HBV reactivation rate was significantly lower for patients receiving AVT than AVT-naïve patients (three-year cumulative incidence rate: 27.3% versus 56.0%; P < 0.001). In multivariable-adjusted analysis, the risk of HBV reactivation was lower for those receiving AVT compared to AVT-naïve patients (adjusted hazard ratio: 0.39, 95% confidence interval: 0.29-0.54). Overall survival was significantly lower for those experiencing HBV reactivation than those who did not (71.5% and 85.7% at five-year) and was associated with higher risk of overall mortality (adjusted hazard ratio: 5.15, 95% confidence interval: 3.60-7.38). CONCLUSION: More than half of AVT-naïve patients experienced HBV reactivation within three years, which was associated with increased risk of overall mortality. The risk of HBV reactivation was lower for those receiving AVT, suggesting that prompt AVT needs to be considered for AVT naïve HBV-related HCC patients with undetectable HBV DNA levels.

Statin Use and the Risk of Hepatocellular Carcinoma in Patients With Chronic Hepatitis B.

BACKGROUND AND AIMS: Statins have pleiotropic effects that may include chemoprevention. Several observational studies have suggested that statins may prevent hepatocellular carcinoma (HCC), but they have not yet been fully studied in patients with chronic hepatitis B virus (HBV) infections. APPROACH AND RESULTS: A hospital-based retrospective cohort of 7,713 chronic HBV-infected individuals between January 2008 and December 2012 were analyzed. The primary outcome was the development of HCC. Patients who used statins for at least 28 cumulative defined daily doses during the follow-up period were defined as statin users (n = 713). The association between the use of statin and the incidence of HCC was analyzed using the multivariable Cox regression model with time-dependent covariates. During a median follow-up of 7.2 years (min-max: 0.5-9.9), HCC newly developed in 702 patients (9.1%). Statin use was associated with a lower risk of HCC (adjusted hazard ratio = 0.36, 95% confidence interval: 0.19-0.68, adjusted for age, sex, cirrhosis, diabetes, hypertension, serum alanine aminotransferase, cholesterol, HBV DNA level, antiviral treatment, and antiplatelet therapy). The observed benefit of the statin use was dose-dependent (adjusted hazard ratio [95% confidence interval], 0.63 [0.31-1.29]; 0.51 [0.21-1.25]; 0.32 [0.07,1.36]; and 0.17 [0.06, 0.48] for patients with statin use of 28-365, 366-730, 731-1095, and more than 1,095 cumulative defined daily doses, respectively). In subgroup analysis, the association between statin use and reduced risk of HCC was observed in all prespecified subgroups analyzed. CONCLUSION: Statin use was associated with a reduced risk of HCC development in chronic HBV-infected patients, suggesting that statins may have a chemopreventive role in this population. These findings warrant a prospective evaluation.

Prognosis of hepatocellular carcinoma patients diagnosed under regular surveillance: potential implications for surveillance goal.

AIMS: The goal of hepatocellular carcinoma (HCC) surveillance is to diagnose cancer at an early stage when treatment is likely to provide the best outcome and thereby, reduce mortality. However, no specific criteria define the 'early stage' for tumors diagnosed under HCC surveillance. We aimed to analyze factors that determined the outcome of HCC patients diagnosed under regular surveillance, to find out how early it is necessary to detect tumors during surveillance. METHODS: A retrospective cohort of 874 HCC patients with preserved liver function (Child-Pugh A) who were diagnosed under regular HCC surveillance at Samsung Medical Center from 2014 to 2016 and did not receive liver transplantation as an initial treatment were analyzed. The primary outcome was overall survival (OS). RESULTS: Tumor size, presence of vascular invasion, albumin-bilirubin grade, and initial treatment modality were independent factors for OS in multivariable analysis. When categorized according to the tumor size, the risk of mortality increased for tumors of > 3 cm, while tumors of 2-3 cm showed similar mortality risks as tumors of ≤2 cm. When categorized according to the tumor factors, curative-intent treatment (resection or ablation) can be applied to 84.5% with excellent outcomes (5-year OS rate, 93.4%), for tumors of ≤3 cm without vascular invasion. CONCLUSIONS: When tumors of ≤3 cm were detected and had no vascular invasion, curative-intent treatment was applied for most patients and showed excellent OS. This finding suggests that to detect tumors of <3 cm without vascular invasion may be considered as the goal of HCC surveillance.

Different Performance of Liver Stiffness Measurement According to Etiology and Outcome for the Prediction of Liver-Related Events.

BACKGROUND/AIMS: Liver stiffness measurement (LSM) by transient elastography (TE) has shown promising results for prediction of hepatocellular carcinoma (HCC) and hepatic decompensation in patients with chronic liver disease (CLD). However, whether prognostic performance of TE differs according to etiology or type of outcome remains further clarification. METHODS: Performance of LSM for the prediction of HCC and hepatic decompensation was analyzed in a cohort of 4026 patients with asymptomatic CLD. RESULTS: During median 4.5 years of follow-up (range 3.0-6.2 years), liver-related events (LRE) were observed in 196 patients (166 with HCC, 45 with hepatic decompensation, and 15 with both). In the multivariate analysis, LSM was independent factor associated with LRE and showed high AUROC (0.78). When stratified by type of outcome and etiology of liver disease, LSM showed high AUROC for the prediction of HCC for patients with non-viral hepatitis (0.89), while it showed relatively low AUROC for the prediction of HCC for patients with viral hepatitis (0.75). For the prediction of hepatic decompensation, LSM showed high AUROC for patients with both viral- and non-viral hepatitis (0.90, 0.90, respectively). CONCLUSIONS: LSM showed powerful prognostic role for the prediction of LRE in patients with CLD. Notably, HCC risk was not negligible in patients with viral hepatitis who showed LSM value < 10 kPa, indicating watchful attention for HCC is still needed for viral hepatitis patients with low LSM.

Patterns and Outcomes in Hepatocellular Carcinoma Patients with Portal Vein Invasion: A Multicenter Prospective Cohort Study.

BACKGROUND AND AIMS: Sorafenib is a proven first-line treatment recommended for hepatocellular carcinoma (HCC) patients with portal vein invasion (PVI). However, multiple treatment modalities are used in clinical practice as a first-line option. This study is a prospective, observational, multicenter, cohort study evaluating patterns of treatment modalities and outcomes for HCC patients with PVI. METHODS: The baseline characteristics, treatment modalities, and outcomes were prospectively collected for 287 newly diagnosed HCC patients with PVI between August 2015 and July 2016 from 16 sites in Korea. RESULTS: During a median 7.8 months of follow-up (range 0.3-24.6 months), mortality was observed in 123 (42.9%) patients. Decision tree analysis classified patients into five subgroups with different outcomes. The patterns of treatment were very heterogeneous, and there was no dominant treatment modality. The most commonly used treatment modality was transarterial chemoembolization (TACE) (20.2%) followed by TACE plus external beam radiation therapy (17.8%) and sorafenib (12.5%). When stratified according to the extent of PVI, sorafenib treatment showed comparable outcomes when the PVI extent was lobal or main/bilateral, yet showed worse outcomes when the PVI extent was limited to the segmental level compared to those who received treatment other than sorafenib. CONCLUSIONS: HCC patients with PVI comprise a heterogeneous population and are treated with various treatment modalities with diverse clinical outcomes in clinical practice. Subclassification of HCC patients with PVI is required to minimize heterogeneity and should be considered for the selection of treatment modalities and future clinical trials.

Use of cyclooxygenase inhibitor and the risk of hepatocellular carcinoma in patients with chronic hepatitis B: A nested case-control study using a nationwide population-based data.

The study aimed to investigate the relationship between the use of COX inhibitors and the risk of hepatocellular carcinoma (HCC) development in patients with chronic hepatitis B (CHB) using a nationwide population-based data. A nested case-control study was conducted using the National Health Insurance Service-National Sample Cohort (NHIS-NSC) from 2002 to 2013 in Korea. We compared the use of COX inhibitors between HCC cases and matched controls by categorizing 5 groups according to the cumulative defined daily dose (cDDD, <28, 28-90, 91-180, 181-360, and >360) adjusting the use of antiviral agents. A total of 4980 patients with CHB were analysed as 996 HCC cases and 3984 matched controls. The number of COX inhibitor users (≥28 cDDD) was 358 patients (36%) and 1814 patients (45%) in the HCC group and control group, respectively. The use of COX inhibitors was significantly associated with a decreased risk of HCC development compared with nonusers (adjusted odds ratio [OR] 0.62, 95% confidence interval [CI] 0.52-0.73, P < .001). There was a dose-dependent inverse relationship between the use of COX inhibitors and the risk of HCC. The adjusted ORs were 0.75 (95% CI: 0.63-0.90), 0.41 (95% CI: 0.31-0.56), 0.38 (95% CI: 0.25-0.57) and 0.49 (95% CI: 0.31-0.79) for the 28-90, 91-180, 181-360 and >360 cDDDs, respectively (P < .01). In conclusion, the use of COX inhibitors was associated with a reduced risk of HCC in CHB. COX inhibitor may have a chemopreventive role in HCC development in patients with chronic liver disease.

Long-term hepatitis B surface antigen (HBsAg) kinetics during entecavir treatment in Korean patients-Functional cure unlikely.

The aim of this study was to investigate the on-treatment kinetics of quantitative HBsAg during entecavir therapy to predict the treatment period needed to achieve functional cure. From a cohort of 1009 CHB treatment-naïve patients who were started on entecavir, the kinetics of quantitative HBsAg decline was assessed in 410 patients by a linear mixed model. The difference in the kinetics of quantitative HBsAg was determined based on the HBeAg positivity, HBeAg seroclearance and presence of baseline liver cirrhosis. Among the 410 patients, 213 patients (52.0%) were HBeAg-positive and 217 patients (66.1%) were male with a median age of 48 years. During a median follow-up of 53.5 months, the quantitative HBsAg level showed a slow but consistent decrease. The expected log qHBsAg levels as a function of time during entecavir treatment in HBeAg(+) and HBeAg(-) patients were 3.4773-0.0039 × Months and 3.1853-0.0036 × Months, respectively. The estimated time to clearance of quantitative HBsAg in our study was greater than 74.1 years in HBeAg-positive patients and 73.5 years in HBeAg-negative patients. The calculated time to achieve functional cure is lifelong without regard to HBeAg seroclearance or presence of liver cirrhosis. The mathematical modelling from a long-term follow-up of chronic hepatitis B patients on entecavir shows that HBsAg clearance requires decades of treatment. Thus, lifelong therapy is inevitable in entecavir-treated patients to achieve functional cure.

Hepatitis B reactivation after kidney transplantation in hepatitis B surface antigen-negative, core antibody-positive recipients.

Nowadays, intensive immunosuppressive therapy including rituximab is commonly used prior to kidney transplantation (KT), raising concerns over hepatitis B virus (HBV) reactivation among hepatitis B surface antigen (HBsAg)-negative and anti-hepatitis B core (HBc)-positive KT recipients. Recent practice guidelines suggested watchful monitoring or antiviral prophylaxis for the first 6-12 months, the period of maximal immunosuppression. However, the actual risk for HBV reactivation, and whether short-term antiviral therapy in the early period is necessary, remains unclear. A total of 449 HBsAg-negative and anti-HBc-positive KT recipients were analysed for HBV reactivation. During a median follow-up of 6.7 (interquartile range: 4.2-9.4) years, HBV reactivation was observed in 9 patients (2.0%). The median time of HBV reactivation from KT was 2.8 years (range: 1.4-11.5 years), with cumulative incidence rates of 0%, 1% and 2% for 1, 3 and 5 years, respectively. There were no severe adverse outcomes, including liver transplantation or mortality related to HBV reactivation. The risk of HBV reactivation was not high, even in anti-HBs-negative patients (n = 60, 4% at 5 years), ABO mismatch (n = 92, 4% at 5 years), use of rituximab (n = 66, 3% at 5 years) or plasmapheresis (n = 17, 7% at 5 years), and acute rejection (n = 169, 3% at 5 years). In conclusion, the HBV reactivation risk was not high and the time of detection was not clustered in the early post-KT period. Our findings favour continued watchful monitoring over antiviral prophylaxis in the early period.

Deep learning with ultrasonography: automated classification of liver fibrosis using a deep convolutional neural network.

OBJECTIVES: The aim of this study was to develop a deep convolutional neural network (DCNN) for the prediction of the METAVIR score using B-mode ultrasonography images. METHODS: Datasets from two tertiary academic referral centers were used. A total of 13,608 ultrasonography images from 3446 patients who underwent surgical resection, biopsy, or transient elastography were used for training a DCNN for the prediction of the METAVIR score. Pathological specimens or estimated METAVIR scores derived from transient elastography were used as a reference standard. A four-class model (F0 vs. F1 vs. F23 vs. F4) was developed. Diagnostic performance of the algorithm was validated on a separate internal test set of 266 patients with 300 images and external test set of 572 patients with 1232 images. Performance in classification of cirrhosis was compared between the DCNN and five radiologists. RESULTS: The accuracy of the four-class model was 83.5% and 76.4% on the internal and external test set, respectively. The area under the receiver operating characteristic curve (AUC) for classification of cirrhosis (F4) was 0.901 (95% confidence interval [CI], 0.865-0.937) on the internal test set and 0.857 (95% CI, 0.825-0.889) on the external test set, respectively. The AUC of the DCNN for classification of cirrhosis (0.857) was significantly higher than that of all five radiologists (AUC range, 0.656-0.816; p value < 0.05) using the external test set. CONCLUSIONS: The DCNN showed high accuracy for determining METAVIR score using ultrasonography images and achieved better performance than that of radiologists in the diagnosis of cirrhosis. KEY POINTS: • DCNN accurately classified the ultrasonography images according to the METAVIR score. • The AUROC of this algorithm for cirrhosis assessment was significantly higher than that of radiologists. • DCNN using US images may offer an alternative tool for monitoring liver fibrosis.

Hepatitis B virus DNA levels and overall survival in hepatitis B-related hepatocellular carcinoma patients with low-level viremia.

BACKGROUND AND AIM: Clinical course of hepatitis B virus (HBV)-related hepatocellular carcinoma (HCC) patients presenting with low-level viremia (LLV) is unclear. METHODS: A total of 565 HBV-related HCC patients with LLV (detectable but HBV DNA ≤ 2000 IU/mL) at the time of HCC diagnosis were analyzed. Based on patterns of HBV DNA levels during follow-up, patients were categorized into three groups: maintained virologic remission (MVR), LLV, and flare. Overall survival was compared between those three groups. RESULTS: During a median 4.5 years of follow-up, 33% showed MVR, 39% showed LLV, and 28% experienced flare. The overall survival differed between MVR, LLV, and flare groups (5-year overall survival: 74.3%, 67.3%, and 61.7%, respectively, 0.015). The patterns of HBV DNA levels were independent factors associated with overall survival, along with age, antiviral treatment, Barcelona clinic liver cancer stage, and initial treatment modality. Flare group showed increased risk of mortality (adjusted hazard ratio [HR] 1.71, 95% confidence interval [CI] 1.15-2.55) compared with MVR group, while the risk was statistically marginal for the LLV group (adjusted HR 1.39, 95% CI 0.95-2.04). During follow-up, 183 patients (32.4%) newly started antiviral therapy (AVT) at LLV. Flare risk was significantly lower among patients who started AVT at LLV compared with those who did not (adjusted HR 0.26, 95% CI 0.17-0.38). CONCLUSIONS: Among HBV-related HCC patients with LLV, flare was frequent during follow-up and was associated with poorer overall survival compared with MVR group. Prospective studies that address whether inducing MVR by early AVT improves patient outcome are warranted.

Low-level viremia and the increased risk of hepatocellular carcinoma in patients receiving entecavir treatment.

The long-term clinical impact of low-level viremia (LLV; <2,000 IU/mL) is not well understood. As a result, it is unclear whether the development of LLV during entecavir monotherapy requires a change in therapy. A retrospective cohort of 875 treatment-naive chronic hepatitis B virus (HBV) monoinfected patients (mean age 47.7 years, male = 564 [65.5%], cirrhosis = 443 [50.6%]) who received entecavir monotherapy were analyzed for the development of hepatocellular carcinoma (HCC). The HCC risk was compared between patients who maintained virological response (MVR), defined by persistently undetectable HBV DNA (<12 IU/mL), and patients who experienced LLV, defined by either persistent or intermittent episodes of <2,000 IU/mL detectable HBV DNA. During a median 4.5 years of follow-up (range 1.0-8.7 years), HCC was diagnosed in 85 patients (9.7%). HCC developed more frequently in patients who experienced LLV than MVR (14.3% versus 7.5% at 5 years, P = 0.015). The hazard ratio comparing those with LLV to MVR was 1.98 (95% confidence interval = 1.28-3.06, P = 0.002, adjusted for age, sex, hepatitis B e antigen, baseline HBV DNA levels, and cirrhosis). Among patients with cirrhosis, those with LLV exhibited a significantly higher HCC risk than those with MVR (HCC incidence rate at 5 years 23.4% versus 10.3%, adjusted hazard ratio = 2.20, 95% confidence interval 1.34-3.60; P = 0.002). However, for patients without cirrhosis, there was no significant difference in the HCC risk between LLV and MVR. CONCLUSION: LLV observed during entecavir monotherapy was associated with a higher risk of HCC, especially for those with cirrhosis, indicating that LLV during potent antiviral therapy is consequential. (Hepatology 2017;66:335-343).

Validation of the Baveno VI and the expanded Baveno VI criteria to identify patients who could avoid screening endoscopy.

BACKGROUND & AIMS: The Baveno VI and the expanded Baveno VI criteria were proposed to help identify patients who could safely avoid screening endoscopies for clinically significant varices among patients with compensated advanced chronic liver disease. However, these criteria require cross-validation, especially in Asian populations where the aetiologies of liver disease are different. METHODS: A total of 1035 patients, including 282 patients with compensated advanced chronic liver disease of different aetiology, were analysed. The compensated advanced chronic liver disease was defined by liver stiffness measurement ≥10 kPa, Child-Pugh class A and absence of prior liver decompensation. High-risk varix was defined as a grade ≥2 oesophageal varix, any varix with a red colour sign or gastric varices. RESULTS: High-risk varixs were present in 19.5% (55/282 patients) with compensated advanced chronic liver disease. Among compensated advanced chronic liver disease patients, the expanded criteria could spare more endoscopies (51.7%) than the original criteria (27.6%). However, the expanded criteria missed a greater number of high-risk varixs (6.8%) than the original criteria (3.8%). When stratified according to liver disease aetiology, the negative predictive values for the original Baveno VI criteria were 0.92, 1.00, 1.00 and 1.00, and the negative predictive values for the expanded criteria were 0.92, 0.96, 0.92 and 0.93 for hepatitis B, hepatitis C, alcohol and non-alcoholic fatty liver disease, respectively. High-risk varixs were rarely detected in patients without compensated advanced chronic liver disease (1.1%, 8/753 patients). CONCLUSIONS: In this Asian cohort study, the Baveno VI criteria were able to identify who could safely avoid screening endoscopy. The expanded Baveno VI criteria could spare more endoscopies but also could increase the odds of missing a high-risk varix.

Non-invasive tests for liver disease severity and the hepatocellular carcinoma risk in chronic hepatitis B patients with low-level viremia.

BACKGROUND & AIMS: We tested whether non-invasive tests for liver disease severity can stratify hepatocellular carcinoma (HCC) risk in chronic hepatitis B virus (HBV)-infected patients showing low-level viremia (LLV, HBV DNA <2000 IU/mL). METHODS: A retrospective cohort of 1006 chronic hepatitis B patients showing persistently LLV, defined by at least two consecutive assessments in the year before enrolment, was assessed for HCC development. Two non-invasive serum biomarkers, the aspartate aminotransferase to platelet ratio index (APRI) and the Fibrosis-4 (FIB-4), were tested. Cirrhosis was defined with ultrasonography. RESULTS: During a median 5.1 years of follow-up, HCC developed in 36 patients. HCC incidence rate at 5 years was significantly higher for cirrhotic patients (19/139, 13.7%), but was not null for non-cirrhotic patients (17/867, 2.0%, P<.001). APRI at a cut-off of 0.5 was more specific but less sensitive for HCC development, and FIB-4 at a cut-off of 1.45 was more sensitive but less specific. When both APRI and FIB-4 were used to group patients, the 5-year cumulative HCC incidence rate was 13.9%, 1.4% and 1.2% for both high, any high, and both low APRI and FIB-4 score among all patients (n=1006, P<.001), respectively, and was 11.4%, 1.5% and 0.4% in the same respective order among non-cirrhotic patients (n=867, P<.001). CONCLUSIONS: The combined use of two non-invasive serum biomarkers (APRI and FIB-4) could stratify HCC risk for chronic HBV-infected patients with LLV.

Treatment for occult hepatocellular carcinoma: does it offer survival advantages over symptom-driven treatment?

OBJECTIVE: In order to claim a benefit of screen-based diagnosis for asymptomatic individuals, treatment of occult disease needs to offer survival advantages compared to the treatment of symptomatic disease, yet information on this issue is scarce with regard to hepatocellular carcinoma (HCC) screening. METHODS: A total of 3353 treatment-naïve, consecutive, newly diagnosed HCC patients [age: 57.9 ± 10.3, male: 2,689 (80.2%), hepatitis B virus: 2555 (76.2%)], diagnosed between 2010 and 2013 were analyzed. Data on the mode of detection was prospectively collected at the time of HCC diagnosis and was used to group patients into occult or symptomatic cases. RESULTS: Overall, 643 (19.2%) patients were symptomatic cases. The proportion of patients undergoing resection, radiofrequency ablation or transplantation were lower in symptomatic cases than occult cases (20.8 vs. 56.2%, p < .001). Survival was better in occult cases than symptomatic cases (71.2 vs. 30.4% at three-years, p < .001), with a multivariable-adjusted hazard ratio of 1.40 (95% confidence interval (CI), 1.24-1.58). When stratified by tumor stage, a survival benefit was not observed for patients diagnosed at modified International Union Against Cancer (mUICC) stage I, but presenting symptoms were diverse and nonspecific. In a statistical model adjusting for potential lead-time bias, the association between overall survival and the mode of detection was markedly attenuated and was no longer significant when the treatment modality was included in the model (hazard ratio, 0.94; 95% CI, 0.82-1.07). CONCLUSION: Treatment of occult disease offered a survival benefit to patients over symptomatic cases. These data support screening practices for asymptomatic individuals to diagnose occult HCC.

Novel Albumin-Bilirubin Grade-Based Risk Prediction Model for Patients with Hepatocellular Carcinoma Undergoing Chemoembolization.

BACKGROUND: Recently, albumin-bilirubin (ALBI) grade has been suggested as a better surrogate for hepatic functional reserve for patients with hepatocellular carcinoma (HCC). AIMS: We developed and validated a novel prediction model to predict outcome for HCC patients who underwent transcatheter arterial chemoembolization (TACE) as a first-line therapy. METHODS: From a multivariate Cox regression model for overall survival, five objective variables (ALBI grade), the Barcelona clinic liver cancer (BCLC) stage, response after the first TACE session, Alpha-fetoprotein level, and sex were chosen and the ABRAS score was developed from the derivation cohort (n = 476) and scored to generate an 8-point risk prediction model. The model's prognostic performance was assessed in the randomly assigned internal validation set (n = 475) and external validation set (n = 243). RESULTS: The ALBI grade was able to stratify patient survival within the same Child-Pugh class. The time-dependent area under receiver operating characteristics curves (AUROCs) for overall survival at 1 and 3 years were 0.78 and 0.73 in the training set, 0.78 and 0.71 in the internal validation set, and 0.70 and 0.65 in the external validation set, respectively. When stratified by BCLC stage, ABRAS score at a cutoff point of more than 3, 4, and 5 for BCLC stage 0/A, B, and C could identify subset of patients with dismal prognosis. CONCLUSION: ABRAS score was useful in estimating prognosis for patients who underwent TACE as a first-line therapy. This score can be useful in planning and guiding treatment strategies with TACE, which warrants prospective validation.

Insulin resistance and the risk of hepatocellular carcinoma in chronic hepatitis B patients.

BACKGROUND AND AIM: We analyzed whether insulin resistance (IR) assessed by homeostasis model assessment (HOMA2-IR) index can stratify hepatocellular carcinoma (HCC) risk in patients with chronic hepatitis B virus (HBV) infection. METHODS: A retrospective cohort of 1696 chronic HBV-infected patients (age: 50.0 ± 7.8 years, men = 964 [56.8%]) who underwent detailed health checkup program including C-peptide and fasting blood glucose measurement and followed up for more than a year were analyzed. RESULTS: During a median follow-up of 5.0 years (range, 1.0-10.5 years), 24 patients (1.4%) developed HCC. The HCC incidence rate was higher for patients with higher HOMA2-IR value than those with lower HOMA2-IR value (1.7% vs 0.5% for HOMA2-IR >1.200 vs ≤1.200, P = 0.009). HOMA2-IR was a significant factor associated with HCC development in multivariable-adjusted model (HR [95% CI]: 3.25 [1.13-9.31], adjusted for age, sex, cirrhosis, and HBV DNA levels). The association between HOMA2-IR and HCC was markedly attenuated and became no longer statistically significant (HR [95% CI]: 1.93 [0.57-6.51]) when further adjusted for obesity, hypertension, and diabetes. In subgroup analysis, HOMA2-IR value was an independent factor associated with HCC in patients without overt metabolic abnormalities (hypertension, diabetes, and metabolic syndrome) but not for those with overt metabolic abnormalities. CONCLUSION: Insulin resistance assessed by HOMA2 was associated with the risk of HCC, indicating that HOMA2-IR can be a useful tool for stratifying the risk of HCC in chronic HBV-infected patients, particularly in patients without overt metabolic abnormalities.