Prognostic biomarkers for predicting decompensation in alcoholic and non-alcoholic patients with compensated cirrhosis: a systematic review and meta-analysis

Abstract

Background: Hepatic decompensation is a critical turning point in the progression of compensated cirrhosis, with distinct pathophysiological trajectories in alcoholic and non-alcoholic etiologies. This systematic review and meta-analysis evaluates prognostic biomarkers for predicting decompensation in patients with compensated cirrhosis, emphasizing differences between alcoholic and non-alcoholic liver disease. Methods: Following PRISMA 2020 guidelines, we systematically searched PubMed, Scopus, and Web of Science for peer-reviewed studies (up to April 2025) reporting hazard ratios (HRs) and 95% confidence intervals for biomarkers predicting decompensation in adults with compensated cirrhosis. Eligible studies included observational cohorts and control arms of RCTs, stratified by etiology (alcoholic vs. non-alcoholic). Data were pooled using random-effects models, with heterogeneity assessed via I2 and Cochrane Q tests. Subgroup analyses explored biomarker performance by etiology and type (inflammatory, functional, and structural). Results: From 691 records, 66 studies (Among these, 955 patients (2.6%) were alcoholic and 36,108 (97.4%) non-alcoholic, totaling 37,063 participants) were included. In non-alcoholic cirrhosis, structural biomarkers like portal vein diameter (HR = 7.39 [4.90, 11.15]) and spleen size (HR = 5.79 [2.00, 16.80]) were strong predictors, alongside functional markers such as bilirubin (HR = 4.27 [2.93, 6.22]) and MELD score (HR = 1.13 [1.07, 1.20]). In alcoholic cirrhosis, inflammatory biomarkers, particularly extracellular vesicles (HR = 5.09 [2.01, 12.86]) and keratin-18 (HR = 1.77 [1.14, 2.75]), showed superior predictive value. Interleukin-6 was predictive across both etiologies (HR = 1.31 [1.00, 1.71]). Heterogeneity was substantial (I2 > 50%) for most biomarkers, reflecting population and methodological variability. Publication bias was low based on funnel plots and Egger’s test. Conclusion: Etiology-specific biomarkers enhance prognostic accuracy in compensated cirrhosis. Structural and functional markers dominate in non-alcoholic cirrhosis, while inflammatory biomarkers are more predictive in alcoholic cirrhosis. Integrating these into personalized risk models could improve clinical management, though prospective validation is needed. Systematic Review Registration: https://www.crd.york.ac.uk/PROSPERO/view/CRD420251076849