Gastric variceal bleeding has a very high rebleeding rate even after endoscopic variceal injection of cyanoacrylate (GVO) which is considered the first choice of endoscopic treatment. Beta-blocker (BB) is effective to lower portal pressure. We hypothesized combination of GVO and BB can further decrease the rebleeding rate.

Gastric varies (GV) rarely rupture. However should it occur, the outcome would be worse than rupture of esophageal varies (EV). Rupture of GV is characteristic of a higher rebleeding rate, a requirement for a larger amount of blood transfusion and a higher mortality. Up to date, the treatment of GV bleeding (GVB) is still sub-optimal in contrast to the treatment of EV bleeding. The management of GV has been focused on treatment of acute GVB. Various specific methods are used to control GVB and prevent rebleeding; however they were far from ideal. It is because GV are usually larger vessels formed in deeper submucosa and connect to the spontaneous gastrorenal shunt which creates a fast blood flow. Therefore, voluminous blood in the larger diameter GV leads to exsanguine bleeding when ruptured. A variety of endoscopic methods, which include injection of sclerosants, tissue adhesive (cyanoacrylate), thrombin and ligation with rubber bands, detachable nylon loop and steel snares, are applied to control acute GV bleeding with variable successful rates (50~100%) and rebleeding rates (20~90%). The successful rate of endoscopic cyanoacrylate injection to arrest active GVB is more consistent around 90~100% and rebleeding rate is around 30~40%. The recent International Consensus Meeting endorsed that endoscopic cyanoacrylate injection is the first line treatment for acute GVB. The embolic complications, either septic & aseptic, are not uncommon. Expertise is also required to reduce the embolic complications and instrumental injuries. Therefore, the efficacy of specific treatment for GVB is sub-optimal, consecutive innovation of new methods are required to improve the prognosis of GVB. Non-selective beta-blocker is effective to reduce rebleeding from esophageal varices. However, its effect on gastric variceal hemorrhage has never been proven.

This is an important issues prompted by current portal hypertension experts. We have much experience in the treatment of gastric variceal bleeding and published fruitful results in high ranking journal. Therefore, we design a randomized trial to compare the effect of endoscopic cyanoacrylate injection obliteration versus non-selective beta-blocker in the secondary prevention of acute gastric variceal bleeding.

Inclusion Criteria:

• clinical diagnosis of liver cirrhosis and/or HCC, endoscopically proven gastric variceal bleeding

Exclusion Criteria:

• younger than 18 y/o or older than 80 y/o, terminal illness, other major systemic disease or malignancy