The aim of this study was to predict the occurrence of hepatorenal syndrome (HRS) and death in patients with advanced cirrhosis and ascites.
Patients and methods
We retrospectively evaluated 2-year data of 78 patients with cirrhosis and ascites (Child–Pugh B/C: 45/43). The mean arterial pressure (MAP) and cardiac output (CO) were measured in all patients just before administration of 2 mg of terlipressin and 30 min later. Systemic vascular resistance (SVR) was calculated as MAP/CO. ΔMAP, and ΔCO, and ΔSVR were defined as the percentage change of MAP, CO, and SVR, respectively, after terlipressin injection. Plasma renin activity (PRA) and plasma aldosterone were evaluated at baseline. Two multivariate models were used: one excluding (model 1) and one including (model 2) the Model of End-stage Liver Disease score.
Higher ΔSVR, Model of End-stage Liver Disease score, and PRA were related independently to the severity of cirrhosis. Independent predictors of HRS at 12 and 24 months were ΔSVR (models 1/2: P=0.008/0.01 and 0.01/0.02, respectively), ΔCO (models 1/2: P=0.01/0.03 and 0.03/0.04, respectively), and PRA (models 1/2: P=0.04 and model 1: P=0.04, respectively). ΔSVR at 12 and 24 months (models 1/2: P=0.005/0.01 and 0.01/0.03, respectively) and ΔCO at 24 months (models 1/2: P=0.02/0.01, respectively) were related independently to survival. Patient groups with significantly higher probability of HRS and mortality were identified by certain cutoffs of ΔSVR (20.6 and 22.8%, respectively) and ΔCO (−10.6 and −11.8%, respectively). ΔSVR and ΔCO independently predicted survival in patients with the most advanced cirrhosis and infection-related survival.
An increase in SVR by at least 20% and a decrease in CO at least 10% in response to terlipressin could predict HRS and mortality in patients with advanced cirrhosis.