A nomogram model based on tumor necrosis factor-like ligand 1A(TL1A) and death receptor-3(DR3) promoter methylation for predicting 90-day prognosis in patients with HBV-associated acute-on-chronic liver failure

Abstract

Purpose: Acute-on-chronic liver failure (ACLF) associated with hepatitis-B-virus (HBV) is a life-threatening condition characterized by severe hepatic dysfunction. The TL1A/DR3 signaling axis modulates immune responses and contributes to hepatic inflammation. This study aimed to investigate the methylation level of TL1A/DR3 promoter, explore its ability to predict prognosis, and establish a prognostic model combined with clinical indicators.

Method: Methylation status and gene expression of TL1A and DR3 were analyzed in peripheral blood mononuclear cells (PBMCs) from 714 participants using Methylight and quantitative polymerase chain reaction (qPCR). Univariate, LASSO, and multivariate analyses were performed to identify key prognostic factors for 90-day outcomes in patients with HBV-associated acute-on-chronic liver failure (HBV-ACLF) and develop corresponding prognostic models. Model performance, including calibration and clinical utility, was evaluated using receiver operating characteristic (ROC) curves, Hosmer-Lemeshow (H-L) tests, and decision curve analysis (DCA). A visual nomogram was constructed to integrate these factors for risk stratification.

Result: Analysis revealed significantly reduced TL1A and DR3 promoter methylation in HBV-ACLF patients, correlating with impaired liver function and coagulation parameters. PBMCs from these patients showed elevated mRNA expression of TL1A, DR3 and IL-6 compared to other groups. Methylation levels of TL1A and DR3 demonstrated high sensitivity and specificity in predicting HBV-ACLF severity. Besides, non-survivors exhibited lower TL1A/DR3 methylation than survivors. A prognostic model integrating prothrombin time activity (PTA), procalcitonin (PCT), and TL1A/DR3 methylation demonstrated excellent performance in predicting 90-day outcomes.

Conclusion: Aberrant TL1A/DR3 promoter methylation reflects the disease severity, and can serve as potential biomarkers for the risk assessment of HBV-ACLF.

Keywords: HBV-ACLF; TL1A/DR3; methylation; noninvasive model; prognosis.

FIGURE 1

Flowchart of research object selection.

FIGURE 2

Expression difference of TL1A and DR3 promoter methylation in HBV-ACLF, HBV-LC, CHB, HCs and correlation analysis with clinical indicators. (A,B) Expression of TL1A and DR3 promoter methylation in PBMCs of HBV-ACLF, HBV-LC, CHB and HCs groups. (C–H) Correlation analysis between TL1A methylation level and TBIL, PTA, PT, INR, Cr, PCT. (I–M) Correlation analysis between DR3 methylation level and TBIL, PTA, INR, Cr, PCT.

FIGURE 3

The mRNA expression levels of TL1A, DR3 and related cytokines in PBMCs of different groups, correlation with promoter methylation status and inflammation level. (A–C) The mRNA expression levels of TL1A, DR3 and IL-6 in PBMCs in HBV-ACLF, HBV-related LC, CHB and HCs groups were compared. (D,E) The mRNA expression correlation analysis of TL1A, DR3 and IL-6. (F,G) Analysis of the expression levels of TL1A and DR3 mRNA and their correlation with PMR.

FIGURE 4

TL1A, DR3 methylation levels and MELDs evaluated the ROC curve of HBV-ACLF.

FIGURE 5

Differences in methylation levels of TL1A and DR3 among 90 days of follow-up. (A) TL1A promoter PMR values in survivor and non-survivor patients at 3 months (B) DR3 promoter PMR values in survivor and non-survivor patients at 3 months.

FIGURE 6

The risk factors were screened according to the 90-day prognosis of HBV-ACLF patients, and the prognosis model was established and evaluated. (A) Variation characteristics of variable coefficients. (B) The process of selecting the optimal value of parameter λ in Lasso regression model by cross-validation method. (C,D) Trained and validated ROC characteristic curves of predictive models for HBV-ACLF patients in the cohort. (E) Mortality of HBV-ACLF patients in the training and validation groups distinguished by model score cutoff values. (F–G) H-L calibration curves for training and validating predictors in the cohort. The X-axis represents the probability of prediction and the Y-axis represents the probability of observation. The dashed line represents the perfect prediction of the ideal model, and the solid line represents the calibration diagram of the predicted model. (H) Predict the DCA of the model in the training and validation queue. The red and blue lines represent the net benefit to patients when the model is used in a clinical intervention. (I) Nomogram to predict 90-day prognosis of patients with HBV-ACLF. (J) Multivariate Cox proportional hazard regression analyses of prognostic factors associated with 90-day mortality in patients with HBV-ACLF.