AMSTERDAM, the Netherlands — Nonselective beta-blocker therapy improves gastrointestinal permeability and decreases intestinal bacterial translocation in patients with cirrhosis and esophageal varices, according to a new study.

These effects were associated with a reduction in the risk for variceal bleeding, and appear to be independent of a hemodynamic response, said Thomas Reiberger, MD, associate professor of hepatology at the Medical University of Vienna in Austria.

Dr. Reiberger presented the results here at the International Liver Congress 2013.

Infection and endotoxemia can impair liver function, increase portal pressure, and impair coagulation, which can result in variceal bleeding, he explained.

Antibiotic treatment is one way to reduce endotoxemia and bacterial translocation, but nonselective beta blockers might help by decreasing intestinal permeability.

Dr. Reiberger and colleagues tested this hypothesis in 50 cirrhotic patients with esophageal varices. Exclusion criteria were celiac disease, current infections, antibiotic or lactulose use in the previous 3 months, and consumption of > 6 g of alcohol per day.

Hepatic venous pressure gradient and other physical and laboratory parameters were measured at baseline and at 1 month, and intestinal permeability was assessed with sucrose, lactulose, and mannitol tests. Decompensation, bleeding, and death were assessed at 3-month follow-up.

The patients were predominantly men in their early to mid-50s (range, 18 to 75 years). Fewer patients with a lower hepatic venous pressure gradient (from 13 to 20 mm Hg) had large varices than those with a higher hepatic venous pressure gradient of more than 20 mm Hg (33% vs 66%; P = .002). There was also less portal hypertensive gastropathy with the lower hepatic venous pressure gradient (53% vs 97%; P = .015).

About 20% of patients had normal gastroduodenal and intestinal permeability, 18% had abnormal gastroduodenal permeability, 8% abnormal intestinal permeability, and 54% both abnormalities.

The higher hepatic venous pressure gradient was significantly associated with increased markers of gastroduodenal and intestinal permeability and of bacterial translocation before beta-blocker treatment, compared with lower hepatic venous pressure gradient. Portal pressure was linearly related to gastroduodenal permeability (P = .003) and intestinal permeability (P < .001).

Bacterial translocation was also associated with hepatic venous pressure gradient. Portal pressure was directly correlated with lipopolysaccharide-binding protein level (P < .001) and interleukin-6 level (P = .006).

Beta Blockers Reduce Pressure

"Not surprisingly, we found a decrease in portal pressure [21 mm Hg to 17 mm Hg; P < .05], but we also found a decrease in the sucrose/mannitol ratio and intestinal permeability index, indicating that beta blockers lead to an improvement of gastroduodenal and intestinal permeability," Dr. Reiberger reported. "Also, lipopolysaccharide-binding protein levels and interleukin-6 levels were significantly decreased with the nonselective beta-blocker therapy."

Fifty-one percent of patients had hemodynamic responses to the beta blockers. "Although there was a clear hint that the patients who were hemodynamic responders had better outcomes in these intestinal permeability tests...what's really striking is that the proportion of patients achieving a normal result on the sucrose, lactulose, mannitol tests after beta-blocker therapy is also really high in the nonresponders. That's why I conclude that it's independent of hemodynamic response," Dr. Reiberger explained. In addition, decreases in levels of lipopolysaccharide-binding protein and interleukin-6 were similar in responders and nonresponders.

There was a trend toward a higher risk for variceal bleeding in patients with abnormal gastroduodenal permeability than in those with normal gastroduodenal permeability (log-rank P = .066); the trend was similar for intestinal permeability. There was a significant correlation between higher interleukin-6 level and more variceal bleeding (log-rank P = .038). Intestinal permeability appeared to have no effect on survival.

After the presentation, an audience member asked Dr. Reiberger whether his team had looked at the different effects on the adrenergic system of propranolol and carvedilol. Dr. Reiberger noted that very few patients were on carvedilol but, from what he could see, there was no difference between them in terms of lipopolysaccharide-binding protein level, interleukin-6 level, or change in intestinal permeability.

Session moderator Isabelle Colle, MD, professor of hepatology and gastroenterology and head of the gastroenterology clinic at Gent University in Belgium, explained some of the mechanisms by which beta blockers exert their effects in this setting. By decreasing hepatic venous pressure gradient and portal pressure "the microcirculation in the intestine is better, ameliorate the function of the barrier, which decreases permeability," she explained. In addition, patients with cirrhosis have slow intestinal transit times, which beta blockers speed up, "so you have fewer bacteria, you have less ammonia."

Dr. Reiberger suggested that beta blockers influence immunity because the sympathetic nervous system affects immune function.

Do all nonselective beta blockers have the same effects? A study of nonresponders to propranolol who were switched to carvedilol showed that "carvedilol had a better effect on the portal pressure than propranolol," Dr. Colle told Medscape Medical News. In the future, carvedilol will probably be better than propranolol, she noted.

Dr. Reiberger reports receiving research support from MSD, Roche, Phoenix, and Gilead. Dr. Colle has disclosed no relevant financial relationships.

International Liver Congress 2013: 48th Annual Meeting of the European Association for the Study of the Liver (EASL). Abstract 84. Presented April 26, 2013.