Should beta-blockers be used in portal hypertension?


Dear colleagues and friends,

Dr. Charles J. Kahi of Indiana University, Indianapolis

Dr. Charles J. Kahi

Thank you for your continued support of the Perspectives debates. In this edition, Dr. Guadalupe Garcia-Tsao and Dr. Marwan Ghabril discuss the rationale for and against beta-blocker therapy in portal hypertension, and ultimately highlight the nuances required for appropriate decision-making. This topic invariably generates controversy and debate, and is broadly relevant to general GI and hepatology practices. I hope you will find it as informative as I did, and I welcome your comments and suggestions for future topics at [email protected].

Charles J. Kahi, MD, MS, AGAF, professor of medicine, Indiana University, Indianapolis. He is also an associate editor for GI & Hepatology News.

Beta-blockers in portal hypertension – Yes!


Portal hypertension is the main consequence of cirrhosis and is responsible for most of its complications. In compensated cirrhosis, a threshold portal pressure gradient, as determined by the hepatic venous pressure gradient (HVPG), of at least 10 mm Hg is the strongest predictor of clinical decompensation (ascites, variceal hemorrhage, or encephalopathy) which is, in turn, the main determinant of death in cirrhosis.

Dr. Guadalupe Garcia-Tsao, professor of medicine, digestive diseases, Yale University, New Haven

Dr. Guadalupe Garcia-Tsao

Portal hypertension is initially caused by an increase in intrahepatic resistance that leads to mild portal hypertension (HVPG, 5-10 mm Hg) but is then enhanced and maintained by an increase in portal venous inflow that leads to clinically significant portal hypertension (HVPG, at least 10 mm Hg).

Portal pressure can be reduced by either ameliorating intrahepatic resistance (which is mostly caused by structural changes that are difficult to reverse) and/or by decreasing portal vein blood inflow (the most modifiable pathogenic mechanism). For over 30 years, treatment of portal hypertension has been based on the use of nonselective beta-blockers (NSBB), drugs that decrease portal pressure through a reduction in splanchnic blood flow. Reduction in portal pressure is greater with NSBB (propranolol, nadolol) than with selective beta-blockers because, as demonstrated experimentally, the main portal pressure–reducing effect stems from splanchnic vasoconstriction because of beta2-adrenergic blockade. This has been confirmed in patients with cirrhosis, in whom the reduction in HVPG is greater with NSBB than with selective BB. On the other hand, carvedilol, an NSBB that also has a vasodilatory alpha1 adrenergic blocking effect, has a greater effect in reducing HVPG, compared with traditional NSBB.

A significant decrease in portal pressure has been associated with better outcomes in cirrhosis. A favorable portal pressure reduction (“response”) has been traditionally defined as a decrease in HVPG below 12 mm Hg or greater than 20% from baseline, although even decreases of 10% are associated with better outcomes. Initial studies had been focused on variceal hemorrhage, a complication that is clearly related to portal hypertension. In this setting, reducing portal pressure clearly leads to a decreased in the incidence of variceal hemorrhage and a decrease in mortality.1 More recently, the focus has been on preventing decompensation (in compensated cirrhosis) and preventing further decompensation/death (in decompensated cirrhosis).

In compensated cirrhosis, a recent meta-analysis of clinical trials of prevention of variceal hemorrhage showed that patients with varices (therefore with clinically significant portal hypertension) without ascites who were NSBB hemodynamic responders, had a reduced risk of developing not only variceal hemorrhage but also ascites and/or encephalopathy, and had lower mortality.2 More importantly, a recent seminal randomized, double-blind, placebo-controlled trial performed in patients with compensated cirrhosis and clinically significant portal hypertension with no or small varices, showed that NSBB (propranolol or carvedilol) led to a significantly lower incidence of decompensation, with ascites being the single event that was significantly lower in the NSBB group.3 This study thereby demonstrates that NSBBs not only reduce the risk of variceal hemorrhage, as previously demonstrated, but also significantly reduces the probability of developing ascites, the most common complication of cirrhosis.

In decompensated cirrhosis, a recent meta-analysis of clinical trials of prevention of variceal hemorrhage showed that patients with varices and ascites (decompensated) who were NSBB responders, had a reduced risk of developing not only variceal hemorrhage but also refractory ascites, spontaneous bacterial peritonitis and/or hepatorenal syndrome, and also had a lower mortality.2 In patients with variceal hemorrhage, the recommended therapy to prevent recurrent variceal hemorrhage is the combination of NSBB plus variceal ligation but this is based on trials that combined compensated and decompensated patients. An individual patient data meta-analysis of these trials showed that, in the group of patients with decompensated Child-Pugh class B/C cirrhosis, rebleeding and mortality were higher with ligation alone, compared with combined therapy with NSBBs and ligation, underlining that NSBB is the key element of combination therapy in these patients.4

There is a fading controversy regarding the potential for increased mortality with the use of NSBBs in patients with refractory ascites and SBP, reported in two retrospective studies.5 These studies lacked information regarding the number of patients in whom NSBBs were withdrawn before the last observation and number of patients in whom NSBBs were started in the course of follow-up. Notably, a recent meta-analysis that included these and subsequent retrospective studies, encompassing a collective of over 1,300 patients, demonstrated that NSBB use in patients with ascites is not related to increased mortality.1,4

Nevertheless, NSBBs should be used cautiously in patients with cirrhosis and ascites. Hemodynamic alterations typical of decompensated cirrhosis are maximal in patients with refractory ascites and spontaneous bacterial peritonitis and the use of NSBB in this setting could lead to worsening hemodynamics, with decreased mean arterial pressure and renal perfusion that could in turn lead to acute kidney injury and death. In studies showing a deleterious effect of NSBB, the mean arterial pressure was significantly lower in patients in the NSBB group.5 In a recent retrospective study, the beneficial effect of NSBBs in patients with refractory ascites, spontaneous bacterial peritonitis, and acute-on-chronic liver failure appeared to apply only to those with a mean arterial pressure of at least 65 mm Hg.6 This evidence has led to guideline recommendations that limit the dose of NSBB to a maximum of 160 mg/day for propranolol or 80 mg/day for nadolol in patients with ascites with close follow-up of arterial blood pressure. Carvedilol should preferably not be used. In the presence of a systolic blood pressure <90 mm Hg or acute kidney injury, NSBBs should be dose-reduced or discontinued. If a precipitant for hypotension is identified (e.g., spontaneous bacterial peritonitis), NSBB can be reinitiated once the precipitating event is resolved and hypotension/acute kidney injury has resolved.

In conclusion, NSBBs are a definite “yes” in the management of cirrhosis and portal hypertension as they prevent poor outcomes (including death) in patients with both compensated and decompensated cirrhosis. In patients with ascites and spontaneous bacterial peritonitis, NSBBs could have deleterious effects but these can be prevented by careful monitoring of blood pressure.


1. D’Amico G et al. Gastroenterology. 2006;131:1611-24.

2. Turco L et al. Clin Gastroenterol Hepatol. 2020;18:313-27.

3. Villanueva C et al. Lancet. 2019;393:1597-608.

4. Albillos A et al. Hepatology. 2017;66:1219-31.

5. Garcia-Tsao G. J Hepatol. 2016 Mar;64(3):532-4.

6. Tergast TL et al. Aliment Pharmacol. Ther 2019;50:696-706.

Dr. Garcia-Tsao is professor of medicine, digestive diseases; chief, digestive diseases, Veterans Affairs Connecticut Healthcare System; director, clinical and translational core, Yale Liver Center; program director, VA Connecticut Hepatitis C Resource Center, New Haven. She has no conflicts.


Next Article: