Updated Baveno VIII recommendations for portal hypertension, published online in the Journal of Hepatology, place greater emphasis on noninvasive risk assessment and prevention of hepatic decompensation. They also refine guidance on managing variceal bleeding, decompensated cirrhosis, and cirrhosis recompensation.
The consensus updates guidance based on evidence that has emerged since Baveno VII. It was developed by the Baveno Cooperation, a consortium of the European Association for the Study of the Liver (EASL), and is endorsed by EASL, the European Society of Gastrointestinal Endoscopy, and the European Reference Network for Rare Hepatological Diseases. At a March 2026 conference in Baveno, Italy, more than 80 faculty members worked in nine panels to review and vote on proposed statements. Approval required agreement from at least 80% of participants. Overall, the group reached consensus on 272 statements and recommendations.
A major focus is the use of noninvasive tests to identify clinically significant portal hypertension (CSPH) in patients with compensated advanced chronic liver disease (cACLD). A liver stiffness measurement of 15 kPa or less combined with a platelet count of at least 150 × 109/L can rule out CSPH with a negative predictive value above 90%. CSPH can be diagnosed based on a predicted probability of at least 75% using validated models; liver stiffness of at least 25 kPa in patients with viral- or alcohol-related cACLD or nonobese patients with metabolic dysfunction-associated steatotic liver disease (MASLD); or spleen stiffness greater than 55 kPa measured at 100 Hz. Patients with indeterminate results may be reassessed after 12 months.
“Portal hypertension management is becoming increasingly preventive, personalized, and noninvasive,” one of the consensus panelists, Thomas Reiberger, MD, of the division of gastroenterology and hepatology at Medical University of Vienna, Austria, told GI & Hepatology News. “With Baveno VIII, noninvasive tests allow us to identify risk earlier, carvedilol effectively treats portal hypertension before the first clinical complication, and unnecessary endoscopies can increasingly be avoided.”
The guidance also shifts the focus from determining whether CSPH is present to predicting a patient’s risk of hepatic decompensation. Key prognostic factors include the presence of CSPH, liver function, and the cause of liver disease. “The paradigm is shifting,” the faculty wrote, toward using noninvasive tests to directly predict decompensation — an approach the document frames as a way to identify patients at substantial risk who would benefit most from preventive treatment.
For patients with compensated cirrhosis and CSPH, the guidance recommends nonselective beta-blockers, particularly carvedilol, to prevent decompensation and improve survival. These drugs are not recommended for preventing decompensation in patients without CSPH. Patients already taking a nonselective beta-blocker do not need screening endoscopy for varices because the results would not change treatment.
Baveno VIII also adds guidance for patients with MASLD. Weight loss of at least 10% is associated with at least a 30% relative reduction in liver stiffness, which may indicate lower liver-related risk. Metabolic and bariatric surgery may be considered for selected patients without CSPH but carries greater risk in those with CSPH and is generally contraindicated in patients with decompensated cirrhosis, except in highly specialized settings such as in conjunction with liver transplantation. GLP-1-based therapies and sodium-glucose cotransporter 2 inhibitors may be used for approved indications in patients with cACLD who have no contraindications.
For acute variceal bleeding, vasoactive therapy should begin as soon as possible, and preventive antibiotics are recommended at presentation. The duration of antibiotic treatment — and whether it is needed at all — may be individualized based on infection risk and the severity of cirrhosis. Preemptive transjugular intrahepatic portosystemic shunt (TIPS) is recommended within 72 hours, and ideally within 24 hours, for selected high-risk patients. New guidance suggests that patients who miss this window may still benefit from TIPS within one to two weeks, depending on disease severity.
The document also provides a more detailed definition of cirrhosis recompensation. Patients must have the cause of cirrhosis removed, controlled, or suppressed; have no ascites without diuretics and no hepatic encephalopathy without specific treatment; and have no recurrent variceal bleeding. These criteria must be sustained for more than six consecutive months. Patients also must show improvement in liver function to Child-Turcotte-Pugh class A. Even after recompensation, patients should continue hepatology follow-up every six months and undergo ongoing surveillance for hepatocellular carcinoma.
For the first time, Baveno recommendations also include guidance for children. The document cautions that liver and spleen stiffness measurements alone should not yet be used to guide portal hypertension management in children because the thresholds for identifying or ruling out varices need further study.
Each statement in the document is graded by level of evidence, and a substantial number — including several new statements on recompensation, pediatric disease, and vascular liver disease — rest on lower-level evidence or expert opinion. The panels identified 153 priority topics for future research, reflecting important gaps in knowledge.
Texas Children’s Hospital sponsored the pediatric symposium held with the meeting. The Baveno Cooperation also received financial support from several pharmaceutical and medical device companies, as well as scientific organizations. The document does not include conflict-of-interest disclosures for individual authors.
Expert Insight
Dr. Reiberger discussed what the recommendations could mean for clinical practice with GI & Hepatology News.
Which Baveno VIII recommendations are most likely to change day-to-day practice for gastroenterologists and hepatologists, and why?
Dr. Reiberger: Several Baveno VIII recommendations underline the paradigm shift from managing complications of portal hypertension to preventing the first decompensation. Prevention is most effectively achieved via etiological therapy and treatment of clinically significant portal hypertension (CSPH) by non-selective betablockers (NSBB), preferably using carvedilol. Carvedilol does not only prevent decompensation but also improves survival. Endoscopic treatment of varices is not recommended for primary bleeding prophylaxis since it has no effect on non-bleeding related complication nor on survival.
How should the updated noninvasive criteria for clinically significant portal hypertension change which patients undergo endoscopy or start preventive treatment?
Dr. Reiberger: The Baveno VI criteria of liver stiffness <20 kPa and normal platelet count (PLT ≥150 G/L) remain still valid to rule out high-risk varices. Baveno VIII strengthens the role of noninvasive tests, particularly liver and spleen stiffness and platelet count, that should preferably be used in validated prediction models, such as the ANTICIPATE (±NASH) and NICER models, which provide an actual CSPH probability. Based on a validated ANTICIPATE or NICER cut-off at ≥75%, carvedilol should be initiated to prevent hepatic decompensation. This means that in many patients we can identify clinically relevant portal hypertension without invasive testing.
What should physicians know about when to start carvedilol in patients with compensated cirrhosis, and when can endoscopy be avoided?
Dr. Reiberger: The indication for carvedilol is based on the presence of CSPH, not simply on the presence of varices. Endoscopic treatment of varices is not preferred, since treatment with NSBBs, particularly carvedilol reduces the risk of both bleeding and non-bleeding decompensation and improves survival. In contrast, if CSPH is absent (i.e., can be ruled out by noninvasive tests), NSBB/carvedilol should not be prescribed simply to prevent future decompensation. Once a patient with CSPH is receiving NSBB/carvedilol, routine endoscopic screening/surveillance for varices is not needed because finding varices would generally not change treatment.