BackgroundThe clinical course of acutely decompensated cirrhosis (AD) is heterogeneous. Presepsin (PSP) is a plasmatic biomarker that reflects Toll‐like receptor activity and systemic inflammation. We conducted a prospective study to: (1) measure PSP in AD and (2) assess whether PSP in AD can predict the development of acute‐on‐chronic liver failure (ACLF).MethodsPatients with AD were prospectively recruited at admission and underwent determination of PSP. In study part 1, we compared PSP in AD versus controls (stable decompensated and compensated cirrhosis). In study part 2, we prospectively followed patients with AD for 1 year and evaluated predictors of ACLF.ResultsOne hundred and seventy three patients with AD were included (median MELD: 18; CLIF‐C AD score: 54). Compared with controls, patients with AD had higher levels of PSP (674 ng/L vs. 310 ng/L vs. 157 ng/L; p < 0.001). In patients with AD, Child–Pugh C and acute kidney injury were associated with higher levels of PSP. During the follow‐up, 52 patients developed ACLF (median time from recruitment: 66 days). PSP, CLIF‐C AD score, and Child–Pugh stage were independently associated with ACLF. A predictive model combining these variables (Padua model 2.0) accurately identified patients at higher risk of ACLF (AUROC 0.864; 95% CI 0.780–0.947; sensitivity 82.9%, specificity 76.7%). In patients at lower risk of ACLF based on a CLIF‐C AD <50, a PSP >674 ng/L could discriminate between two groups at significantly different risk of ACLF. Finally, in patients who did not develop ACLF, baseline PSP was significantly higher in those who progressed toward unstable versus stable decompensated cirrhosis.ConclusionThe Padua model 2.0 can be used to identify patients with AD at high risk of ACLF. If these results are validated by external cohorts, PSP could become a new biomarker to improve risk stratification in AD.