In compensated cirrhosis with portal hypertension, beta-blockers reduced a composite of decompensation or death

The trial results cannot be generalized to all patients with cirrhosis, and future trials should look at whether similar effects could be achieved simply by limiting salt intake, an ACP Journal Club commentary said.


Daily propranolol or carvedilol was associated with decreased risk of cirrhosis decompensation compared to placebo, according to a recent trial in patients with compensated cirrhosis and clinically significant portal hypertension (CSPH), which was defined as a hepatic venous pressure gradient (HVPG) ≥10 mm Hg. In this Spanish study, a primary endpoint of development of ascites, bleeding, or overt encephalopathy or death occurred in 16% of 100 patients on beta-blockers versus 27% of 101 patients randomized to placebo.

The study was published in the April 20 issue of The Lancet. The following commentary by Jacob Korula, MD, FACP, appeared in the ACP Journal Club section of the Aug. 19 Annals of Internal Medicine.

Decompensation is complications of advanced cirrhosis that affect morbidity and mortality (e.g., variceal hemorrhage, ascites, hepatic encephalopathy, hepatorenal syndrome, and hepatocellular carcinoma). The specific factors that herald the independent occurrence of these events are unknown. With each event, other conditions, such as spontaneous bacterial peritonitis complicating ascites and variceal hemorrhage, may concomitantly precipitate encephalopathy, coagulopathy, and renal failure, depending on the severity of the event and liver disease. The intervention in the trial by Villanueva and colleagues is based on reducing CSPH (portal pressure > 10 mm Hg), which increases the risk for decompensation.

Results from the sequential clinical, hepatic, and systemic hemodynamic studies show that the subgroup of patients responding with > 10% reduction in HVPG had reduced risk for ascites, and patients who received beta-blockers vs placebo also had reduced risk for ascites (absolute risk reduction 11%). However, the trial results cannot be generalized to all patients with cirrhosis because HVPG can only be measured in selected centers.

Several questions arise about the trial. First, attrition during the study was 13% at year 1, increasing to 31% at year 2, and then to 57% at year 3, despite stable disease based on Child–Pugh Class A and low Model for End-stage Liver Disease scores. Second, only 78% of patients had HVPG measured in year 1, 43% at year 2, and 24% at year 3. Are hemodynamic studies in this subset of patients reflective of all enrolled patients? Third, over the duration of the trial, despite a mean propranolol dose of 95 mg/d in the treatment group, there were small between-group differences in mean heart rate and blood pressure—tolerance and compliance may have been factors. Fourth, because renal sodium retention is an important cause of ascites in portal hypertension, the effect of sodium restriction is not known. Finally, adherence to beta-blockers or placebo was considered adequate in 82% and 83% of patients in the treatment and placebo groups, respectively, but was based on a threshold of ≥70% of pills used during pill counts, a method that overestimates adherence.

In summary, this seemed like a lot of work to reduce ascites formation. Further studies of beta-blockers vs controlling for sodium intake in compliant patients with similar severity of cirrhosis are important to determine if these results could be achieved simply by limiting salt intake.