Coexisting gastric varices should not preclude prophylactic ligation of large esophageal varices in cirrhosis.

ABSTRACT

BACKGROUND/AIMS: Coexisting gastric varices at baseline or the risk of their formation during treatment could alter the approach for primary bleeding prophylaxis in patients with large esophageal varices. METHODS: Data analysis of 152 patients with cirrhosis and large esophageal varices included in the German multicenter trial on primary prevention of variceal bleeding. RESULTS: 20 patients (13.6%) had coexisting gastric varices at baseline (GOV+). 10 of those each received either band ligation or propranolol, respectively. During follow-up (34.4 +/- 18.9 months) new gastric varices occurred in 2/75 (2.7%, ligation) and 4/77 (5.2%, propranolol) patients, respectively. One patient with newly developed gastric varices bled (propranolol group). GOV+ patients had a better baseline liver function and overall survival. Bleeding incidence did not differ significantly between GOV+ and GOV- patients (3-year actuarial risk 20.0 +/- 10.6% (GOV+), 38.1 +/- 4.4% (GOV-), p = 0.195). Among GOV+ patients, bleeding occurred in 3/10 patients of the propranolol group and in 0/10 in the ligation group (p = 0.038). CONCLUSION: Prophylactic band ligation of large esophageal varices is safe and effective also in patients with coexisting gastric varices. Band ligation did not increase the risk of secondary gastric varices compared to propranolol.