p-TIPS must be the treatment of choice in CP-C patients with AVB. Due to the strong benefit in preventing further bleeding and ascites, p-TIPS could be a good treatment strategy for CP-B+AB patients. This article is protected by copyright. All rights reserved.

Measurements of portal pressure, usually obtained via the hepatic venous pressure gradient (HVPG) may be a prognostic marker in cirrhosis. The aim of this study was to evaluate the impact of HVPG on survival in patients with cirrhosis in addition to the Model for EndStage Liver Disease (MELD) score. We also examined whether inclusion of HVPG in a model with MELD variables improves its prognostic ability. Retrospective analyses of all patients who had HVPG measurements between January 1998 and December 2002 were considered. Proportional hazards Cox models were developed. Prognostic calibrative and discriminative ability of the model was evaluated. In this period, 693 patients had a hepatic hemodynamic study, and 393 patients were included. Survival was significantly worse in those patients with greater HVPG value (univariate HR, 1.05; 95% CI, 1.02-1.08; P ‫؍‬ .001). HVPG remained as an independent variable in a model adjusted by MELD, ascites, encephalopathy, and age (multivariate HR, 1.03; 95% CI, 1.00-1.06; P ‫؍‬ .05) so that each 1-mmHg increase in HVPG had a 3% increase in death risk. In addition, HVPG as well as MELD score variables and age, significantly contributes to the calibrative predictive capacity of the prognostic model; however, discriminative ability improved only slightly (overall C statistic T he Model for End-Stage Liver Disease (MELD) score, initially developed for survival prediction of patients undergoing the transjugular intrahepatic portal systemic shunt (TIPS) procedure, 1 has been subsequently validated in an increasingly heterogeneous population of patients with cirrhosis as a very good tool to rank patients according to their short-term risk of death. [2][3][4][5] In the initial validation of the MELD score, individual complications of portal hypertension (ascites, spontaneous bacterial peritonitis, variceal bleeding, and encephalopathy) were added to the model, producing only minimal improvement in its discriminative ability. 2 However, each individual portal hypertensionrelated complication is only one aspect of the underlying pathophysiological mechanism, the portal hypertensive syndrome. Re-evaluation of the role of portal hypertension indexes in such predictive scores has been suggested, as portal hypertension has been described as the third parameter most frequently found to be a significant predictor of survival in cirrhosis. 6 Interestingly, after dividing patients in categories according to MELD score, 1-year mortality within each category was higher among patients with portal hypertension-related complications.

Acute-on-chronic liver failure (ACLF) AVB + ACLF High rebleeding rate (25.2%) in 42 days High mortality rate (51.0%) in 42 days Pre-emptive TIPS in AVB + ACLF Low rebleeding rate (4.5%) in 42 days Low mortality rate (13.6%) in 42 days Pre-emptive TIPS Acute Variceal Bleeding (AVB) Highlights Variceal bleeding is frequently associated with ACLF in cirrhosis. ACLF is independently associated with rebleeding and mortality. Patients with variceal bleeding and ACLF can benefit from a preemptive (early) TIPS.