• biomarker;
  • cirrhosis;
  • prognosis;
  • vasopressin


Background & Aim

Copeptin, secreted stoichiometrically with vasopressin, demonstrated its prognostic role in various diseases other than cirrhosis.


We investigated the association between severity of cirrhosis and plasma concentrations of copeptin, and the prognostic value of copeptin in 95 non-septic cirrhotic patients (34 Child-Pugh A, 29 CP-B, 32 CP-C), 30 septic patients with a Child-Pugh >8 (‘group D’), and 16 healthy volunteers. Patients were followed for at least 12 months to assess the composite endpoint death/liver transplantation.


Median copeptin concentrations (interquartile range) increased through healthy volunteers group [5.95 (3.76–9.43) pmol/L] and ‘group D’ patients [18.81 (8.96–36.66) pmol/L; P < 0.001)]. During a median follow-up of 11.0 ± 6.1 months, 28 non-transplanted patients died and eight were transplanted. In receiver operated characteristic curves analysis, the area under the curve values were as follows: Child-Pugh score 0.80 (95% CI: 0.71–0.86), model of end-stage liver disease (MELD) score 0.80 (0.70–0.86), C-reactive protein (CRP) 0.71 (0.60–0.80) and copeptin 0.70 (0.57–0.79). By stratifying the values of these variables into tertiles, the risk of death/liver transplantation for patients belonging to the highest tertile of copeptin (>13 pmol/L) was high (Log-rank test: P = 0.0002) and 2.3-fold higher than for patients with lower concentrations after adjusting for MELD score (>21) and CRP (>24 mg/L) in a Cox model. Other potential predictors (age, total cholesterol, natraemia and serum free cortisol) did not reach a significant level.


In cirrhotic patients, copeptin concentrations increased along with the severity of liver disease. In our cohort, the 1-year mortality or liver transplantation was predicted by high MELD score and high concentrations of CRP and copeptin.