Comparison of Therapies for Primary Prevention of Esophageal Variceal Bleeding: A Systematic Review and Network Meta-analysis

Abstract

We performed a systematic review with network meta-analysis (NMA) to compare the efficacy of different approaches in primary prevention of esophageal variceal bleeding and overall survival in patients with cirrhosis with large varices. Thirty-two randomized clinical trials (RCTs) with 3,362 adults with cirrhosis with large esophageal varices and no prior history of bleeding, with a minimum of 12 months of follow-up, were included. Nonselective beta-blockers (NSBB), isosorbide-mononitrate (ISMN), carvedilol, and variceal band ligation (VBL), alone or in combination, were compared with each other or placebo. Primary outcomes were reduction of all-cause mortality and prevention of esophageal variceal bleeding. Random-effects NMA was performed and summary estimates were expressed as odds ratio and 95% confidence intervals (OR; CI). Quality of evidence was critically appraised using the Grading of Recommendations, Assessment, Development and Evaluation approach. Moderate quality evidence supports NSBB monotherapy (0.70; 0.49-1.00) or in combination with VBL (0.49; 0.23-1.02) or ISMN (0.44; 0.21-0.93) for decreasing mortality in patients with cirrhosis with large esophageal varices and no prior history of bleeding. Moderate-quality evidence supports carvedilol (0.21; 0.08-0.56) and VBL monotherapy (0.33; 0.19-0.55) or in combination with NSBB (0.34; 0.14-0.86), and low-quality evidence supports NSBB monotherapy (0.64; 0.38-1.07) for primary prevention of variceal bleeding. VBL carries a higher risk of serious adverse events compared with NSBB. Conclusion: NSBB monotherapy may decrease all-cause mortality and the risk of first variceal bleeding in patients with cirrhosis with large esophageal varices. Additionally, NSBB carries a lower risk of serious complications compared with VBL. Therefore, NSBB may be the preferred initial approach for primary prophylaxis of esophageal variceal bleeding.