Primary prophylaxis of variceal hemorrhage: a randomized controlled trial comparing band ligation, propranolol, and isosorbide mononitrate.

Abstract

This randomized controlled trial compared variceal band ligation (VBL), propranolol (PPL), and isosorbide-5-mononitrate (ISMN) in the prevention of first esophageal variceal bleed. Over a 6-year period, 172 patients with cirrhosis, grade II or III esophageal varices that had never bled, were recruited; 44 into VBL, 66 into PPL, and 62 into ISMN. Baseline patient characteristics: age, 55 +/- 11 years; Child-Pugh score, 8 +/- 2; 65% alcohol-induced cirrhosis; follow-up period, 19.7 +/- 17.6 months (range, 0.13-72.1 months), were comparable in the 3 groups. On intention-to-treat analysis, variceal bleeding occurred in 7% of patients randomized to VBL, 14% to PPL, and 23% to ISMN. The 2-year actuarial risks for first variceal bleed were 6.2% (95% confidence interval [CI], 0.0%-15.0%) for VBL, 19.4% (95% CI, 0.1%-32.4%) for PPL, and 27.7% (95% CI, 14.2%-41.2%) for ISMN. A significant number of patients reported side effects with drug treatment (45% PPL and 42% ISMN vs. 2% VBL; P = 0.00), resulting in withdrawal from treatment in 30% of PPL and 21% of ISMN patients. There were no statistically significant differences in mortality rates in the 3 groups. In as-treated analysis, there was a statistically significant difference in actuarial risk for bleeding at 2 years between VBL and ISMN (7.5%, 95% CI, 2.5%-10.6% vs. 33.0%, 95% CI, 15%-49%, respectively, log rank test P = 0.03) but not between VBL and PPL. VBL was equivalent to PPL and superior to ISMN in preventing first variceal bleed. The side-effect profile for pharmacotherapy was considerable.