Esophagogastric varices accompanying cirrhosis are a significant cause of hemorrhagic morbidity and mortality. Primary prophylaxis of the initial bleeding episode would reduce the inpatient mortality rate. Nonselective beta blockers and isosorbide-5-mononitrate (ISMN) have been used to reduce bleeding episodes. Because pharmacotherapy does not significantly lower portal pressure and because of medication contraindications and adverse effects, many patients who might benefit cannot use these medications. Variceal band ligation (VBL) has been used effectively during the acute bleed or to prevent rebleeding. Lui and associates used a prospective, randomized controlled trial to compare VBL with propranolol (PPL) and ISMN therapy to prevent initial episodes of esophageal variceal bleeding.

Patients with cirrhosis were screened by endoscopy for esophageal varices, and those with more than minimal variceal lesions were asked to enroll in the study. Very sick patients, patients younger than 18 years or older than 75 years, patients who were already taking vasoactive agents, and patients with noncirrhotic portal hypertension, contraindications to the trial drugs, low blood pressure, or bradycardia were excluded. Researchers randomized 172 participants to repeated VBL via endoscopy until the varices disappeared, incremental twice-daily dosing of PPL (up to 160 mg per day), or incremental twice-daily dosing of ISMN (up to 80 mg per day). Patients who did not tolerate ISMN were offered PPL. Patients who could not take PPL did not receive additional treatment. The primary end point was esophageal variceal bleeding.

VBL eradicated esophageal varices in 82 percent of the treated group. The most serious adverse event was perforation of a patient's upper esophagus (the patient fully recovered and later resumed VBL therapy). The mean dosage of PPL in the second group was 113.5 mg per day. Patients randomized to receive ISMN took a mean dosage of 55.1 mg per day. Among participants in both medically treated groups, there was no significant blood pressure drop. More than 40 percent of patients reported side effects, that prompted high drop-out rates. Variceal bleeding occurred in 6.8 percent of patients who received VBL, in 13.8 percent of patients who took PPL, and in 22.6 percent of patients who took ISMN. The risk for variceal bleeding was significantly different between the VBL and ISMN groups, but not between the VBL and PPL groups or between the PPL and ISMN groups. Deaths from all causes or associated with variceal bleeding were lower among patients who underwent VBL.

The authors conclude that VBL is as effective as PPL and more effective than ISMN in decreasing the rate of first cirrhosis-induced variceal bleeding among persons with more than just minimal esophageal varices. Side effects are more common among patients receiving medical therapy, which may limit regularity of treatment. Although combinations of beta blockers and ISMN have been shown to be more effective than either medication alone in preventing first bleeds, this study did not look at the efficacy of combination therapy.