Editor’s Note: From November 10th to 14th, 2023, the American Association for the Study of Liver Diseases (AASLD) Annual Meeting 2023 was held in Boston, USA. Professor Jing Zhang and her team from Beijing Youan Hospital, Capital Medical University, have long focused on scientific research in chronic liver diseases. Seven research findings from the team were included in the conference and presented in poster form, covering non-alcoholic fatty liver disease (NAFLD) [1-3], cirrhosis [4-5], and hepatitis C [6-7]. “Hepatology Digest” had the privilege of inviting Professor Jing Zhang for an interview at the conference to introduce the research findings, share insights into clinical research, and discuss topics of interest and future research directions. The interview video and some research content are shared below.

▲ Dr. Jing Zhang interviewed at AASLD 2023

Research One

Study on the Relationship between SAMM50-rs2073082, -rs738491, -rs3761472 Interactions and Susceptibility to Non-alcoholic Fatty Liver Disease (NAFLD)

Zhao Jinhuan, Jing Zhang, Zhang Yang, etc.

▲ AASLD 2023 Poster Presentation (2417-C)

Non-alcoholic fatty liver disease (NAFLD) is a metabolic stress-induced liver injury caused by the interaction of genetic and environmental factors. In the development of NAFLD, single nucleotide polymorphisms (SNPs) play a role of over 50%. SAMM50 is a recently discovered SNP associated with susceptibility to NAFLD and disease progression. While some related studies have been conducted in Japanese, Korean, Mexican, and Spanish populations, research based on the Chinese population is still limited. With the global trend of aging becoming increasingly apparent, it is estimated that the population aged 60 and above will reach approximately 840 million by 2025. Evidence suggests that aging increases the incidence of NAFLD, especially the mortality rate of non-alcoholic steatohepatitis (NASH) in the elderly, requiring urgent attention. Therefore, we conducted a case-control study to explore the impact of three SAMM50 SNPs and their interactions on NAFLD in the elderly Chinese population.

Research Method:

This retrospective study included 1,053 elderly individuals aged 65 and above who underwent annual physical examinations at Beijing Mentougou Community Hospital from November 2020 to September 2021. They were divided into the NAFLD group (590 cases) and the non-NAFLD control group (463 cases). NAFLD was diagnosed based on abdominal ultrasound, and genomic DNA was extracted for SNP genotyping. Generalized multifactor dimensionality reduction (GMDR) analysis was used to assess the impact of SNP-SNP interactions on NAFLD.

Research Results:

1. There were differences in the genotype distribution of rs2073082 between the NAFLD group and the control group. The genotype and allele frequencies of rs738491 differed between the two groups, while the genotype distribution and allele frequency of rs3761472 showed no significant differences (Table 1).

2. Rs2073082 was significantly associated with NAFLD susceptibility in the homozygous model (GG vs. AA, OR=1.638, P<0.001) and the recessive model (AG + GG vs. AA, OR=1.836, P<0.001). After adjusting for confounding factors (gender, age, BMI), significant correlations remained (adjusted ORs were 1.691 and 1.962, respectively), indicating a higher risk of NAFLD in carriers of the G allele. Additionally, rs738491 showed a significant association with NAFLD in the allele model (T vs. C, adjusted OR=1.216, P=0.045), homozygous model (TT vs. CC, adjusted OR=1.373, P=0.021), and recessive model (CT + TT vs. CC, adjusted OR=1.532, P=0.021) (Table 2).

3. GMDR analysis showed that the model involving rs2073082, rs738491, and rs3761472 had the highest training accuracy (56.97%) and testing accuracy (55.52%), with the best cross-validation consistency (CVC: 10/10), and the results were statistically significant (P=0.0107) (Table 3).

4. The model diagram (Figure 1) shows the risk prediction of NAFLD for different combinations of genotypes of the three SNPs. The tree diagram (Figure 2) shows that rs2073082 and rs738491 exhibit antagonistic effects on the susceptibility to NAFLD. The Fruchterman−Rheingold graph (Figure 3) shows that rs3761472 has a synergistic effect with rs738491 and rs2073082, while rs2073082 and rs738491 have antagonistic effects on NAFLD.

Note: In each square, the left bar graph represents subjects in the NAFLD group (positive scores), and the right bar graph represents subjects in the control group (negative scores). The number at the top of each bar graph is the score statistic, which is the product of the membership relationship coefficient and the residual. Score statistics are used to determine whether the average score of individuals exceeds a set threshold (e.g., ≥1), categorizing data into high-risk and low-risk groups. High-risk genotype combinations are represented in black; low-risk genotype combinations are represented in gray; blank spaces represent no genotype combinations.

Note: Different types of SNP-SNP interactions affecting the risk of NAFLD. Orange (synergistic effect); blue (antagonistic effect). Short lines represent strong interactions, while long lines represent weak interactions.

5. The risk of NAFLD increased with the number of SNP loci carried. The adjusted ORs for single-point, two-point, and three-point models were 1.532 (95% CI: 1.144–2.053), 1.809 (95% CI: 1.147–2.853), and 1.892 (95% CI: 1.196–2.993), respectively. In addition, individuals carrying the G allele of rs2073082, the T allele of rs738491, and the G allele of rs3761472 had a twofold higher risk of developing NAFLD than non-carriers (Table 4).

Research Conclusion:

Carriers of the SAMM50-rs 2073082 G allele and rs738491 T allele have a higher risk of developing NAFLD. GMDR interaction analysis shows that the combination of the three SNPs, rs2073082, rs738491, and rs3761472, has the strongest predictive ability for NAFLD. The risk of NAFLD is twofold higher when carriers simultaneously have risk alleles at all three SNP loci.

Research Two

Assessment of Cirrhotic Cardiomyopathy Incidence in Cirrhotic Patients Based on Different Diagnostic Criteria

Ma Lixia, Jing Zhang

▲ AASLD 2023 Poster Presentation (3111-A)

Cirrhotic cardiomyopathy (CCM) is a heart dysfunction that occurs in cirrhotic patients without a history of heart disease. In 2019, the Cirrhotic Cardiomyopathy Consortium (CCC) updated the diagnostic criteria for CCM, setting a lower threshold for diagnosing systolic dysfunction by lowering the ejection fraction. Overall longitudinal strain (GLS) has been suggested as a diagnostic criterion for CCM, and it plays a crucial role in early detection and diagnosis of impaired heart function in cirrhotic patients. While cardiac biomarkers can aid in diagnosis, their use in assessing heart function in cirrhotic patients remains controversial. This study aims to assess the incidence of CCM in cirrhotic patients based on two different diagnostic criteria.

Research Method:

From July 2020 to July 2022, 186 cirrhotic patients with a confirmed diagnosis were enrolled at Beijing Youan Hospital, Capital Medical University. The study participants were divided into CCM and non-CCM groups based on the 2005-WGO and 2019-CCC diagnostic criteria. Transthoracic echocardiography was performed using tissue Doppler imaging and speckle tracking echocardiography methods. Patient data, including demographics, medical history, complications of cirrhosis, laboratory biochemistry, virology, myocardial enzymes, echocardiography, and electrocardiography, were collected and compared to evaluate the incidence of CCM based on different diagnostic criteria.

CCM Group Inclusion Criteria: ① Age 18-65 years, any gender; ② All patients clinically, laboratory, and imaging-diagnosed with cirrhosis; ③ Presence of CCM based on either 2005-WGO or 2019-CCC criteria; ④ Signed informed consent.

Exclusion Criteria: ① Active upper gastrointestinal bleeding within the past 2 weeks, uncorrected anemia; ② Patients with fever or SIRS; ③ Patients with type 2 diabetes; ④ Patients with primary cardiovascular diseases such as hypertension, coronary heart disease, rheumatic heart disease, AV block, etc.; ⑤ Patients with other acute or chronic diseases that may affect heart function, such as respiratory diseases, thyroid diseases, etc.; ⑥ Patients with malignancies; ⑦ Patients who have taken medications affecting heart function (beta-blockers, calcium channel blockers, ACE inhibitors, etc.) within the past week.

Non-CCM Group Inclusion Criteria: ① Age 18-65 years, any gender; ② All patients clinically, laboratory, and imaging-diagnosed with cirrhosis; ③ Absence of CCM; ④ Signed informed consent. Exclusion criteria same as CCM group.

Research Results:

A total of 311 cirrhotic patients were screened in this study. After applying inclusion and exclusion criteria, 186 patients were included, with 60 in the CCM group and 126 in the non-CCM group according to the 2005-WGO criteria. According to the 2019-CCC criteria, there were 48 in the CCM group and 138 in the non-CCM group. The overall incidence of CCM was slightly higher when assessed using the 2005-WGO criteria compared to the 2019-CCC criteria: 32.2% vs. 25.8% (P<0.001).

According to the 2019-CCC criteria, the CCM group consisted of 33 males (68.8%), and the non-CCM group had 95 males (68.1%) (P>0.05). The mean age of the CCM group was 59.2±11.5 years, while the non-CCM group’s mean age was 55.20±9.9 years (P<0.05). Both groups showed no significant differences in etiology, mean arterial pressure, resting heart rate, and common complications of cirrhosis (ascites, spontaneous bacterial peritonitis, hepatic encephalopathy) (P>0.05). According to the Child-Pugh grade, in the CCM group, 3 patients were classified as Child A (6.3%), 29 as Child B (60.4%), and 16 as Child C (33.3%). In the non-CCM group, 15 were Child A (10.9%), 85 were Child B (61.6%), and 38 were Child C (27.5%), with no significant difference between the two groups (P>0.05). The MELD score and MELD-Na score also showed no significant differences between the two groups (P>0.05).

According to the 2005-WGO criteria, the CCM group included 38 males (63.3%), and the non-CCM group had 71.4% males (P>0.05). The mean age of the CCM group was 60.8±9.5 years, and the non-CCM group had a mean age of 54.1±10.3 years (P<0.05). Significant differences in etiology were observed between the two groups (P<0.05). There were no significant differences in mean arterial pressure, resting heart rate, common complications of cirrhosis, and Child-Pugh grade (P>0.

05). The MELD score and MELD-Na score also showed no significant differences between the two groups (P>0.05).

Comparisons of cardiac ultrasound parameters and electrocardiography between the CCM and non-CCM groups based on the 2019-CCC criteria revealed significant differences in QTc, PA, GLS, e’, and LAVI (P<0.05), while LVEF, FS (%), and E/A ratio showed no significant differences (P>0.05). According to the 2005-WGO criteria, the CCM group showed significant differences in QTc, E/A ratio, RV, e’, and LAVI (P<0.05), while E/e’ and GLS showed no significant differences (P>0.05).

Based on the 2019-CCC criteria, among the 48 CCM patients, 36 (75%) had systolic dysfunction, while according to the 2005-WGO criteria, only 3 out of 60 CCM patients (5%) had systolic dysfunction. The two criteria showed poor agreement (k=0.241), indicating that more cirrhotic patients had systolic dysfunction when assessed using the 2019-CCC criteria (P<0.001). In the comparison of diastolic dysfunction, according to the 2005-WGO criteria, 57 out of 60 patients (95%) had diastolic dysfunction, while according to the 2019-CCC criteria, only 12 out of 48 patients (25%) had diastolic dysfunction. The two criteria showed fair agreement (k=0.390), suggesting that compared to the 2019-CCC criteria, the 2005-WGO criteria identified more cirrhotic patients with diastolic dysfunction (P<0.001).

Conclusion:

The incidence of Cirrhotic Cardiomyopathy (CCM) fluctuated between 25.8% and 32.2%. According to the 2019-CCC diagnostic criteria, more patients exhibited systolic dysfunction, whereas based on the 2005-WGO criteria, more patients showed diastolic dysfunction.

Research Three:

Safety and Immunogenicity Study of Novel Coronavirus Vaccines in Patients with Severe Hepatic Diseases Related to Hepatitis C

▲ AASLD 2023 Poster Presentation (1892-A) (Scroll up and down to browse)

Patients with severe liver diseases have a high mortality rate when infected with SARS-CoV-2, and it is strongly recommended that this population receive priority vaccination against COVID-19. However, there is limited research on the safety and antibody levels after complete vaccination with inactivated COVID-19 vaccines. This study evaluates the safety and immunogenicity of COVID-19 vaccines in patients with severe hepatic diseases related to hepatitis C, aiming to provide evidence for adjusting vaccination strategies.

Research Method:

This observational study recruited 211 patients with hepatitis C-related liver diseases who had not been previously infected with SARS-CoV-2. All patients received three doses of inactivated COVID-19 vaccines within a span of 7 months. Adverse reactions after each vaccine dose were recorded, and changes in liver and kidney function, coagulation function, and blood routine were examined before and after vaccination. Serum neutralizing antibody levels were also measured throughout the vaccination course.

Research Results:

1. Baseline Characteristics:

The study included 211 patients, comprising 117 with cirrhosis, 13 with liver cancer, and 81 with chronic hepatitis C. The patients were categorized into severe liver disease groups (cirrhosis and liver cancer) and a control group (chronic hepatitis C). Baseline characteristics, vaccine manufacturers, time intervals between the third vaccine dose and antibody detection, Child-Pugh scores, and MELD scores were comparable between the groups. All patients received three doses of inactivated vaccines from different manufacturers, with a median interval of 57.0 days (IQR: 38.0–74.0) and 54.5 days (IQR: 41.0–73.0) for the two groups, respectively. The MELD score in the chronic hepatitis C group was significantly lower than in the cirrhosis and liver cancer groups, consistent with the severity of liver disease.

2. Adverse Reactions:

In this study, all liver disease patients exhibited good tolerance to the inactivated COVID-19 vaccine. Pain was the most common local adverse reaction, with occurrence rates of 14.7% (21/211), 11.8% (25/211), and 12.3% (26/211) after the first, second, and third doses, respectively. The most common systemic adverse reactions after the third dose were myalgia [7.1% (15/211), 7.6% (16/211), 5.7% (12/211)], headache [8.5% (18/211), 9.0% (19/211), 6.2% (13/211)], and fatigue [10.0% (21/211), 7.6% (16/211), 8.5% (18/211)]. Adverse reactions were mild, self-limiting, and there were no reported grade 3 adverse events. The incidence of adverse reactions did not differ significantly between the two groups.

3. Changes in Clinical Indicators:

We compared changes in clinical indicators before and after vaccination. Although total bilirubin, creatinine, and prothrombin time increased, and hemoglobin and albumin levels decreased after the third vaccine dose compared to baseline, these changes were small and, while statistically significant, lacked clinical significance. The results indicate that the inactivated COVID-19 vaccine did not cause severe abnormalities in liver and kidney function, coagulation function, and blood routine in patients with hepatitis C-related liver diseases.

4. Neutralizing Antibodies:

Neutralizing antibody titers were defined as negative (≤17.6 IU/ml) or positive (>17.6 IU/ml). Among 81 patients with chronic hepatitis C, 26 (32.1%) had negative antibody titers, and 56 (67.9%) had positive titers. Among 130 cirrhosis and liver cancer patients, 53 (40.8%) had negative titers, and 77 (59.2%) had positive titers. The positivity rate of neutralizing antibodies in the severe liver disease group was similar to that in the chronic hepatitis C group. However, compared to chronic hepatitis C patients (median titer 38.9, range 16.0–198.9 IU/ml), patients with severe liver disease (median titer 24.2, range 11.2–65.5 IU/ml) showed a significant decrease in antibody titers. Antibody titers in all patients began to decline after 2 months. Our study results suggest that, in the majority of patients with severe hepatitis C-related liver diseases, COVID-19 vaccination can provide a certain level of protection, but this protection is not stable or long-lasting.

Conclusion:

This study indicates that the inactivated COVID-19 vaccine appears to be safe in patients with severe hepatic diseases related to hepatitis C, but its immunogenicity is relatively low. Unlike the general population, vaccination strategies need to be improved for this specific patient group.

Research Four:

Establishment and Validation of a Risk Prediction Model for Hepatocellular Carcinoma (HCC) in Patients with Advanced Liver Fibrosis and Cirrhosis Due to Hepatitis C after Achieving Sustained Virological Response (SVR)

Authors: Shanshan Xu, Lixia Qiu, Liang Xu, Yali Liu, Jing Zhang

Affiliation: Liver Disease Center, Beijing Youan Hospital, Capital Medical University (Departments 3)*

▲ AASLD 2023 Poster Presentation (1820-A)

Despite achieving sustained virological response (SVR), patients with advanced liver fibrosis and cirrhosis due to hepatitis C still face the risk of developing hepatocellular carcinoma (HCC). Clinical guidelines recommend semiannual screening for HCC through alpha-fetoprotein (AFP) and ultrasound, imposing a significant burden on patients and public health. This study aims to establish a risk prediction model for HCC in this population, stratifying high-risk individuals and devising personalized HCC screening protocols.

Research Method:

A total of 551 patients with advanced liver fibrosis or cirrhosis due to hepatitis C who achieved SVR were enrolled. Patients were followed up every 6 months, with 385 randomly assigned to the derivation cohort (70%) and 166 to the internal validation cohort (30%). An additional 221 similar patients from another hospital constituted the external validation cohort. Sociodemographic data, medical history, and baseline laboratory results were collected. HCC diagnosis was based on AFP levels, abdominal imaging, or liver histology. Liver fibrosis or cirrhosis was diagnosed using liver stiffness measurement (LSM), FIB-4, or clinical presentation. Univariate and multivariate analyses were conducted to identify HCC risk factors, and a risk prediction model was established based on these factors. The model’s predictive performance was assessed using the area under the receiver operating characteristic curve (AUROC) and Harrell’s C index, with internal and external validations. Finally, the predictive efficacy of this study’s model was compared with a previously established HCC prediction model for hepatitis C patients with SVR.

Research Results:

In the derivation cohort, with a median follow-up period of 66.6 ± 12.3 months, 37 patients (9.61%) developed HCC. Older age (HR: 1.08, 95% CI: 1.01–1.15, P = 0.029), male gender (HR: 2.41, 95% CI: 1.15–5.03, P = 0.020), baseline serum albumin level (HR: 0.88, 95% CI: 0.82–0.94, P = 0.000), and LSM (HR: 1.02, 95% CI: 1.00–1.05, P = 0.032) were identified as independent predictors of HCC occurrence (Table 5). Thus, a risk prediction model for HCC based on age, gender, serum albumin level, and LSM was established (Nomogram) (Figure 4).

The time-dependent receiver operating characteristic (ROC) curve was used to evaluate the predictive performance of the HCC risk prediction model. The results showed similar predictive efficacy in the derivation, internal, and external validation cohorts. The 3-year and 5-year AUROCs were 0.84 (95% CI: 0.80–0.88) and 0.83 (95% CI: 0.79–0.87) in the derivation cohort, 0.68 (95% CI: 0.61–0.75) and 0.73 (95% CI: 0.66–0.80) in the internal validation cohort, and 0.87 (95% CI: 0.82–0.92) and 0.80 (95% CI: 0.74–0.85) in the external validation cohort (all P > 0.05) (Figure 5A, 5B). In the derivation and internal validation cohorts, the 7-year AUROCs were 0.81 (95% CI: 0.77–0.85) and 0.72 (95% CI: 0.64–0.78), respectively (P > 0.05) (Figure 5C).

Based on the HCC prediction model, patients’ HCC risk scores were calculated, and patients were stratified into low-risk (<95.45 points), medium-risk (95.45–124.76 points), and high-risk (>124.76 points) groups. The annual HCC incidence rates for the low-risk, medium-risk, and high-risk groups were 0.18%, 1.29%, and 4.45%, respectively (Table 7). The cumulative incidence rates at 3 years were 0.00%, 1.56%, and 9.28%, at 5 years were 0.00%, 5.73%, and 18.56%, and at 7 years were 1.04%, 7.29%, and 22.68% for the three groups (Table 6, Figure 6). In the internal validation cohort, the annual HCC incidence rates for the low-risk group were 0.39%, 1.60%, and 3.80% (Table 7). In the external validation cohort, the annual incidence rates for the three groups were 0.00%, 3.65%, and 9.02% (Table 6). The cumulative HCC incidence rates for the internal and external validation cohorts are shown in Table 6 and Figure 6.

The model developed by Ioannou et al., termed the “HCV-HCC prediction model,” is considered one of the best models for predicting HCC after SVR in patients with hepatitis C. This model consists of four formulas for four subgroups of patients: cirrhosis/SVR, cirrhosis/no SVR, non-cirrhosis/SVR, and non-cirrhosis/no SVR. Four predicting factors (age, platelet count, aspartate aminotransferase to alanine aminotransferase ratio, and serum albumin)

are the main predictors in the model. Comparing the predictive efficacy of this study’s model with the HCV-HCC prediction model, the AUROC for the HCV-HCC model was 0.66 (95% CI: 0.62–0.71), significantly lower than our prediction model at 0.81 (95% CI: 0.77–0.85) (P < 0.05) (Figure 7).

Conclusion:

The HCC prediction model based on age, gender, baseline serum albumin, and LSM can further stratify high-risk patients after achieving SVR in hepatitis C, enabling the formulation of individualized screening strategies.

Reference :

参考文献：

[1] Zhao J, Zhang Y, Zhang J, et al. SAMM50-rs2073082, -rs738491 and -rs3761472 Interactions Enhancement of Susceptibility to Non-Alcoholic Fatty Liver Disease. AASLD 2023. Poster 2417-C.

[2] Wei X, Zhang J, Qi S, et al. Comparison of MAST, FAST, and FIB-4 as noninvasive predictors of Fibrotic-NASH and NASH. AASLD 2023. Poster 2028-A.

[3] Liu L, Li H, Zhou D, et al. The impact of hepatic steatosis on response of antiviral therapy in CHB patients: a meta-analysis. AASLD 2023. Poster 1395-C.

[4] Ma L, Zhang J. The prevalence of cirrhotic cardiomyopathy according to different diagnostic criteria. AASLD 2023. Poster 3111-A.

[5] Liu J, Xu H, Liu W, et al. Spleen stiffness determined by SPLEENDEDICATED device accurately predicted esophageal varices in cirrhosis patients. AASLD 2023. Poster 3165-A.

[6] Jing Z, Guo H. Safety and immunogenicity of COVID-19 vaccine in patients with HCV related liver disease. AASLD 2023. Poster 1892-A.

[7] Xu S, Qiu L, Zhang J. Development and validation a risk model of hepatocellular carcinoma for patients with advanced fibrosis and cirrhosis who achieved sustained HCV clearance. AASLD 2023. Poster 1820-A.

Expert Profile:

Chief Physician, Professor, Doctoral Supervisor

Director of the Liver Disease Center, Departments 3

Beijing Youan Hospital, Capital Medical University