Early predictors of 6-month non-recovery in drug-induced liver injury (DILI) remain limited. Genetic variants are stable host characteristics that may complement baseline clinical variables. We aimed to develop and validate an interpretable, clinical-genetic machine learning model for predicting 6-month non-recovery in patients with DILI.

This retrospective, single-center study included 338 patients with DILI, who were classified as recovered (n = 171) or non-recovered (n = 167) at 6 months. Candidate single-nucleotide polymorphisms and baseline clinical variables were collected during initial hospitalization. Features were selected using complementary screening approaches. Multiple machine learning models were developed and compared. Model discrimination, calibration, clinical utility, the incremental value of genetic predictors, and interpretability using SHapley Additive exPlanations (SHAP) were assessed.

Five predictors were consistently retained for model development: rs72631567, rs28521457, alanine aminotransferase, monocyte percentage, and low-density lipoprotein. Among the candidate algorithms, the light gradient boosting machine model showed the best performance, with area under the receiver operating characteristic curve (AUC) values of 0.92 (95% confidence interval [CI] 0.89–0.95) in the training set and 0.81 (95% CI 0.70–0.91) in the validation set. The model showed acceptable calibration and favorable decision-curve performance. In ablation analysis, the clinical-only model showed limited discrimination (AUC 0.57, 95% CI 0.43–0.71). SHAP analysis identified rs72631567 as the most influential predictor.

An interpretable model that integrates host genetic variants with baseline clinical variables demonstrated good internal performance for early prediction of 6-month non-recovery in DILI. These findings support external validation of genotype-informed risk stratification in patients with DILI.

Chronic hepatitis B patients with baseline hepatitis B surface antigen (HBsAg) <1500 IU/mL are considered as the favorable population for achieving functional cure. This trial aimed to explore a treatment strategy to help the unfavorable population characterized by high HBsAg levels (>3000 IU/mL), hepatitis B e antigen-negative status, and normal alanine transaminase levels in the indeterminate phase (HBEIP), transition to the favorable group.

In this investigator-initiated, open-label clinical trial, we randomly assigned participants aged 18 to 60 years with HBEIP characteristics to receive either tenofovir disoproxil fumarate (TDF) monotherapy (monotherapy group) or pegylated interferon alfa-2b (Peg-IFNα-2b) plus TDF (combination group). The primary endpoints were the HBsAg loss rate and the proportion of participants with HBsAg <1,500 IU/mL through week 96.

From May 2021 to November 2023, we enrolled 263 participants, with 131 randomly assigned to the combination group and 132 to the monotherapy group. In the primary analysis, none of the 132 participants (0%) in the monotherapy group achieved HBsAg loss, compared with 10 of 131 (7.6%) in the combination group (P = 0.001). Through week 96, 48.9% (64/131) of participants in the combination group achieved HBsAg <1,500 IU/mL, and a reduction in HBsAg level greater than 1 log10 IU/mL between baseline and week 24 was an independent predictor of this endpoint (Odds Ratio = 16.957, 95% Confidence Interval: 3.002–95.797, P = 0.001). In contrast, only two participants (1.5%) in the monotherapy group achieved HBsAg <1,500 IU/mL.

Compared with TDF monotherapy, combination therapy with Peg-IFNα-2b and TDF significantly improved both HBsAg <1,500 IU/mL and HBsAg loss rates in HBEIP patients.

Bacterial infection is a key cause of mortality in patients with acute-on-chronic liver failure (ACLF). In this study, we aimed to identify metabolite biomarkers and develop a novel machine learning model for early identification of bacterial infection in ACLF.

Based on a prospective multicenter cohort from 14 centers, 1,314 patients with acute-on-chronic liver disease were enrolled, including those with ACLF and non-ACLF. Plasma samples at admission were collected for metabolomics profiling. Patients were randomly divided into discovery (n = 921) and validation (n = 393) sets. Machine learning was used to develop diagnostic models. The win ratio method was employed to assess the risk stratification capability of the models.

Bacterial infection occurred in 198 of the 451 ACLF patients and 132 of the 863 non-ACLF patients. Infection altered the plasma metabolome, especially in lipid, amino acid, and xenobiotic metabolic pathways. Models for bacterial infection in ACLF (five metabolites) and non-ACLF (six metabolites) demonstrated superior discrimination in the discovery (AUCs: 0.881 and 0.935, respectively) and validation sets (AUCs: 0.835 and 0.889, respectively) compared with C-reactive protein, white blood cell count, procalcitonin, and the best composite clinical model. Metabolic risk stratification based on the models effectively predicted 90-day outcomes (all-cause death, organ failure, sepsis, new-onset acute decompensation, and systemic inflammatory response syndrome).

Our models based on novel metabolic biomarkers enable identification of patients at high risk of bacterial infection and support risk stratification of 90-day outcomes.

The huge burden of type 1 diabetes mellitus (T1DM) has been a source of concern globally since the Industrial Revolution in the 18th–19th centuries. To this end, studies have shown that certain environmental changes that accompanied the Revolution may have increased the risk and burden of the disease in genetically predisposed individuals. However, documented studies that synthesize these environmental triggers are scarce. As a result, the current study was conceived to synthesize the environmental triggers of T1DM to boost public awareness. Relevant information was retrieved from reputable academic databases; namely, Scopus, PubMed, SpringerLink, and Embase. The results showed that chemical exposure, viral infection, gut microbiome disruption, vitamin and mineral deficiencies, inadequate or exclusive breastfeeding, as well as early exposure to infant feeding formulas could increase the risk and burden of T1DM in genetically predisposed individuals. As a consequence, these triggers could compromise the expression of certain genes involved in insulin synthesis and immune function, such as the human leukocyte antigen (HLA), insulin (INS), cytotoxic T lymphocyte-associated antigen 4 (CTLA-4), and protein tyrosine phosphatase non-receptor type 22 (PTPN22) genes. This would result in a dysfunctional immune system in which immune cells, such as T-cells and B-cells and molecules, such as cytokines would attack self-tissues, thus causing autoimmunity of the pancreatic beta cells. Environmental triggers could also induce the T1DM pathophysiology by modifying the epigenome of the mentioned genes. Furthermore, some epigenetic changes could be reversed, which would infer that treatment procedures that would include the pathophysiology of the environmental triggers could be more effective.

Emerging evidence suggests that RNA-binding motif (RBM) proteins are involved in hepatocarcinogenesis and act either as oncogenes or tumor suppressors. The objective of this study was to investigate the role of RBM34, an RBM protein, in hepatocellular carcinoma (HCC).

We first examined the expression of RBM34 across cancers. The correlation of RBM34 with clinicopathological features and the prognostic value of RBM34 for HCC was then investigated. Functional enrichment analysis of RBM34-related differentially expressed genes (DEGs) was performed to explore its biological function. RNA sequencing (RNA-seq) was applied to identify downstream genes and pathways affected upon RBM34 knockout. The correlation of RBM34 with immune characteristics was also analyzed. The oncogenic function of RBM34 was examined in in vitro and in vivo experiments.

RBM34 was highly expressed in hepatocellular carcinoma and correlated with poor clinicopathological features and prognosis. RBM34 was positively associated with tumor immune cell infiltration, biomarkers of immune cells, and immune checkpoint expression. A positive correlation was also observed between RBM34, T cell exhaustion, and regulatory T cell marker genes. Knockout of RBM34 significantly inhibited cell proliferation, migration, and xenograft tumor growth, and sensitized HCC cells to sorafenib treatment. RBM34 inhibition reduced FGFR2 expression and affected PI3K-AKT pathway activation in HCC cells.

Our study suggests that RBM34 may serve as a new prognostic marker and therapeutic target of HCC.

Naringenin is an anti-inflammatory flavonoid that has been studied in chronic liver disease. The mechanism specific to its antifibrosis activity needs further investigation This study was to focused on the cyclic guanosine monophosphate-adenosine monophosphate synthase (cGAS) pathway in hepatic stellate cells and clarified the antifibrosis mechanism of naringenin.

The relationship between the cGAS-stimulator of interferon genes (STING) pathway and liver fibrosis was analyzed using the Gene Expression Omnibus database. Histopathology, immunohistochemistry, fluorescence staining, Western blotting and polymerase chain reaction were performed to assess gene and protein expression levels associated with the cGAS pathway in clinical liver tissue samples and mouse livers. Molecular docking was performed to evaluate the relationship between naringenin and cGAS, and western blotting was performed to study the expression of inflammatory factors downstream of cGAS in vitro.

Clinical database analyses showed that the cGAS-STING pathway is involved in the occurrence of chronic liver disease. Naringenin ameliorated liver injury and liver fibrosis, decreased collagen deposition and cGAS expression, and inhibited inflammation in carbon tetrachloride (CCl4)-treated mice. Molecular docking found that cGAS may be a direct target of naringenin. Consistent with the in vivo results, we verified the inhibitory effect of naringenin on activated hepatic stellate cells (HSCs). By using the cGAS-specific agonist double-stranded (ds)DNA, we showed that naringenin attenuated the activation of cGAS and its inflammatory factors affected by dsDNA. We verified that naringenin inhibited the cGAS-STING pathway, thereby reducing the secretion of inflammatory factors by HSCs to ameliorate liver fibrosis.

Interrupting the cGAS-STING pathway helped reverse the fibrosis process. Naringenin has potential as an antihepatic fibrosis drug.