Annals of hepatology最新文献

Introduction and objectives: Prospective data are limited regarding viral hepatitis viral load and serology and immunotherapy in patients with hepatocellular carcinoma (HCC). This analysis evaluated hepatitis viral load and serology and transaminase levels in patients with sorafenib-treated advanced HCC who were receiving immunotherapy with pembrolizumab in the KEYNOTE-224 and KEYNOTE-240 studies.

Materials and methods: This was a post hoc analysis of the single-arm phase II KEYNOTE-224 (NCT02702414) and the placebo-controlled phase III KEYNOTE-240 (NCT02702401) studies. Patients positive for hepatitis B surface antigen (HBsAg) and/or with detectable hepatitis B virus (HBV), patients positive for isolated total hepatitis B core antibody (anti-HBc), and patients currently or previously infected with hepatitis C virus (HCV) were included. Viral-induced hepatitis flare was defined as >1 log increase from baseline and >1000 IU/ml viral load with concurrent alanine aminotransferase (ALT) elevation classified according to prespecified thresholds.

Results: No patient in the pembrolizumab arm who was positive for HBsAg and/or with detectable HBV or who was positive for isolated anti-HBc met the criteria for viral-induced hepatitis flare; 1 patient (3.4%) in the placebo arm met the criteria for viral-induced hepatitis flare. One patient (2.3%) in the pembrolizumab arm of KEYNOTE-240 who was infected with HCV met the criteria for viral-induced hepatitis flare, but the event was not attributed to pembrolizumab.

Conclusion: These results suggest that pembrolizumab does not cause viral-induced hepatitis flare in patients with advanced HCC.

Clinicaltrials:

Gov identifier: NCT02702414 and NCT02702401.

Introduction and objectives: The six-month abstinence rule for alcohol-associated liver disease (ALD) patients may exclude candidates from life-saving transplantation without reliably predicting relapse. We compared outcomes of deceased donor liver transplantation (DDLT) and living donor liver transplantation (LDLT) in ALD patients, with modified abstinence criteria.

Patients and methods: ALD patients evaluated for transplantation from 2008-2020 were analyzed. 6-month abstinence was required for DDLT but not for LDLT patients with favorable psychological assessments. Survival analyses were conducted from evaluation i.e., intention-to-treat (ITT) and transplantation. Patients with living donors were categorized as intended for LDLT (ITT-LDLT) and those without as intended for DDLT (ITT-DDLT). Patients who were ineligible for transplantation served as the control group.

Results: Of the 216 ALD patients evaluated, 79 (36.6%) were accepted for transplantation. Five-year ITT survival was superior in the ITT-LDLT group (83.3%) compared to ITT-DDLT (62.6%, P = 0.04) and controls (30.7%, P < 0.001). Among transplant recipients (DDLT n = 34, LDLT n = 20), five-year graft survival was comparable between DDLT and LDLT (79.8% vs 76.5%, P = 0.84) despite only 40% of LDLT patients achieving six-month abstinence. Alcohol relapse rates were comparable between DDLT and LDLT (32.4% vs. 30.0%, P > 0.99). Alcohol dependence (HR=7.32, P < 0.001) and medical non-compliance (HR=4.19, P = 0.007) predicted relapse.

Conclusions: Liver transplantation provided significant survival benefit for carefully selected ALD patients. With comprehensive psychological assessment, patients without psychiatric disorders or compliance issues can achieve excellent outcomes after transplantation.

Introduction and objectives: To assess the utility of telemedicine on provision of hepatitis C care, we compared characteristics and outcomes of HCV-infected patients engaging in standard in-clinic care, telemedicine only (TM), and hybrid (HB) models of care across pre-, peri‑ and post-COVID-19 pandemic periods.

Patients and methods: HCV RNA positive patients assessed between October 2017 and March 2025 at The Ottawa Hospital Viral Hepatitis Program (Ottawa, Canada) were retrospectively analyzed.

Results: Of 1118 patients, 626 (56.0 %) engaged in standard care, 139 (12.4 %) in TM care, and 353 (31.6 %) in HB care. TM group patients were less likely to have immigrated, experienced substance abuse, housing instability, incarceration, or psychiatric conditions, or be based in Ottawa. Utilization of HB and TM care were highest during the pandemic. Across all time periods, HB care demonstrated higher DAA treatment initiation and completion compared to standard or TM care (standard: 81.5 %; TM: 79.1 %; HB: 92.1 %, p < 0.01), (standard: 91.3 %; TM: 93.6 %; HB: 96.3 %, p = 0.02), respectively, with no difference in SVR. During the pandemic, a greater proportion of patients experiencing barriers to cure engaged in HB or TM care and achieved DAA completion (pre: 91.1 %; pandemic: 98.1 %; post: 95.8 %, p < 0.01). HB care was associated with increased odds of DAA initiation (OR = 2.87; 95 % CI 1.82 to 4.53), including patient populations experiencing barriers to cure (OR = 2.35; 95 % CI 1.35 to 4.07).

Conclusions: HB models demonstrate potential in addressing public health challenges and engaging marginalized populations, supporting increased integration into HCV care programs.

Introduction and objectives: The screening accuracy of non-invasive fibrosis tests like FIB-4 in metabolic dysfunction-associated steatotic liver disease (MASLD) remains unclear. Using standard cut-offs, this study evaluated FIB-4's agreement with vibration-controlled transient elastography (VCTE) and identified and validated new thresholds.

Patients and methods: A prospective cohort study (2019-2024) in Belgian and Dutch primary care used VCTE by FibroScan® (Echosens, France) as a proxy for the fibrosis stage. The FIB-4 index was derived from electronic patient data and study blood samples. Agreement between VCTE and FIB-4 was analysed using Weighted Cohen's kappa. New fibrosis cut-offs (≥F2; ≥8 kPa) for ≤65 and >65 years were determined via Youden's Index and validated in a Turkish primary care cohort and a Belgian secondary care T2DM cohort.

Results: Among 563 participants (median age 62 years, 47.1 % male, 14.2 % with T2DM, median BMI 28.2 kg/m²), FIB-4 showed poor agreement with VCTE (κ = 0.138, 95 % CI: 0.069-0.207). Suggested new cut-offs of 1.29 (≤65 years) and 1.72 (>65 years) were proposed. The 1.29 cut-off performed similarly to the existing 1.3 in validation cohorts. In the Türkiye and T2DM cohorts, the 1.72 cut-off improved sensitivity over 2.0 but had lower specificity.

Conclusions: The FIB-4 index showed poor agreement with VCTE and low sensitivity, making it an unreliable standalone diagnostic tool for liver fibrosis in people with MASLD in both primary and secondary care. Alternative non-invasive tests or improved cut-off values are needed for accurate fibrosis detection in clinical practice.

Introduction and objectives: Outcomes for patients with hepatocellular carcinoma (HCC) are generally poor, partly due to significant multidrug resistance. HCC heterogeneity decreases the accuracy of traditional prediction models. This study aimed to evaluate the prognostic significance of the HCC resistome.

Materials and methods: Clinical and transcriptomic data from 371 HCC cases available at TCGA were analyzed. An external dataset (604 patients from 4 cohorts) and 40 HCC samples, in which gene expression was determined by RT-qPCR, were used to validate the prognostic model.

Results: The in silico analysis revealed two distinct clusters of patients based on the expression of resistome genes. Kaplan-Meier analysis indicated that Cluster 1 (C1) exhibited a better prognosis. Furthermore, the response to sorafenib treatment was better in patients included in C1 than in C2. Cox regression analysis identified the resistome profile as an independent prognostic factor alongside clinicopathological features such as tumor stage and ECOG status. Fifty-eight out of 81 genes examined displayed differential expression between clusters. Thirteen genes demonstrated a correlation between their expression levels and patient survival. In Cox multivariate analysis, SLC22A1, BIRC5, and ABCC1 genes emerged as independent prognostic factors, forming the basis for a risk model. BIRC5 and ABCC1 upregulation and SLC22A1 downregulation were associated with worse outcomes. Experimental results confirmed that patients with higher risk scores had a worse prognosis.

Conclusions: A prognostic signature based on the expression levels of three resistome-associated genes has been defined and can serve as a helpful complementary tool in clinical settings to categorize HCC patients.

Introduction and objectives: The prevalence of diabetes is rising among patients with chronic hepatitis B (CHB). However, the comparative efficacy of sodium-glucose cotransporter-2 inhibitors (SGLT2i) versus dipeptidyl peptidase-4 inhibitors (DPP4i) in this high-risk population remains unclear. We aimed to determine if SGLT2i significantly reduces the risk of liver-related events (LREs) compared with DPP4i, using a target trial emulation approach.

Materials and methods: We used a nationwide database to identify CHB patients with diabetes (age ≥40 years) receiving SGLT2i or DPP4i alongside metformin for ≥60 consecutive days between January 2015 and December 2020. The primary outcome was LRE, defined as hepatocellular carcinoma (HCC), liver cirrhosis (LC), transplantation, or liver-related mortality. Multivariable Cox-regression models with per-protocol analysis and propensity score matching (PSM) were performed to estimate LRE risk.

Results: Among 16,070 patients (median age, 57 years; 66.9% male), 695 developed LREs (4.7% in SGLT2i group vs. 4.3% in DPP4i group; p=0.413). In the entire cohort, the LRE risk in the SGLT2i group was comparable to that in the DPP4i group: adjusted hazard ratio (aHR) of 0.89 (95% confidence interval [CI] 0.71-1.11). The risk of secondary outcomes in the SGLT2i group was similar to that of the DPP4i group: aHRs of 0.89 (95% CI 0.58-1.37) for HCC, 0.92 (95% CI 0.72-1.17) for LC, and 0.82 (95% CI 0.29-2.35) for all-cause mortality. These trends were consistent in the PSM cohort.

Conclusions: The overall hepatic prognosis was comparable between the SGLT2i and the DPP4i groups. Further prospective studies are required to reach robust conclusions.

Introduction and objectives: Acute-on-chronic liver failure (ACLF) in cirrhotic patients is marked by a multiorgan failure and high mortality. The role of the pancreas in ACLF is poorly understood. This study evaluated the prevalence, progression and prognostic impact of elevated lipase (eLIP) and amylase (eAMY) in ACLF.

Patients and methods: We retrospectively analyzed ACLF patients (EASL-CLIF criteria) at the Vienna General Hospital (11/2003-11/2022). Elevated eLIP and eAMY were defined as levels ≥3 times the upper limit of normal. Data were collected before ACLF onset, at diagnosis (D0), and on days 7 (D7), 28 (D28), and 90 (D90) post-diagnosis. Factors associated with eLIP were identified using univariable logistic regression, and survival was examined with uni- and adjusted multivariable Cox regression models.

Results: Among 193 patients, D28 and D90 mortalities were 39.9 % and 53.9 %, respectively. At D0, lipase and (alpha-)amylase elevations were found in 43.5 % and 50.2 % of patients, with strong correlation (Spearman's rho: 0.687; p<0.001). eLIP was observed in 8.8 % at D0 and 15.7 % at D7. At D0, impaired circulation (MAP <70 mmHg, odds ratio [OR] 4.41; p=0.048) and kidney failure (OR 6.31; p=0.030) were linked to eLIP, and circulatory failure (OR 3.30; p=0.012) to D7 eLIP. Although D0 enzyme levels did not prognosticate mortality, eLIP at D7 independently predicted D28 (adjusted hazard ratio [aHR]: 2.15; p=0.031) and D90 mortality (aHR 2.14; p=0.011).

Conclusions: Increased lipase and (alpha-)amylase levels are common in ACLF. Notably, lipase levels ≥3x ULN at ACLF-D7 predict mortality independent from liver function, suggesting a role in disease progression.