Background Gastric varices(GV)are life-threatening for patients with portal hypertension.Endoscopic injection withbutyl cyanoacrylate(BC),the mainstay of the therapy for GV,has been reported to be effective for hemostasis of bleedingvarices,but its efficacy in the obliteration of GV and impact on the survival of patients still needs clarification.Here wesummarized our experience of 10 years’ practice to evaluate the efficacy and safety of endoscopic therapy using BC forGV patients.Methods From January 1997 to April 2006,GV cases treated with endoscopic injection using BC were collected.The“sandwich method”and the“modified sandwich method”were used to inject BC intravascularly.Retrograde analysis wasmade on the data of treatment and follow-up.Results A total of 635 GV cases treated with endoscopic injection using BC were collected,most of them(90.2%)suffered from post-hepatitis cirrhosis.Emergency hemostasis was achieved in 139 out of 146 sessions(95.2%).Complications occurred in 32 cases(5.2%),including hemorrhage due to early expulsion of tissue glue(3.1%),septicemia(1%)and ectopic thrombosis(0.5%),such as spleen infarction.Endoscopic follow-up in 503 patients showedcomplete disappearance(76.9%),collapse(17.3%)or remnants(5.8%)of gastric varices.A total of 550 patients werefollowed up clinically for 3 to 115 months.Of these patients,44 had recurrent bleeding(8.0%)and 44 died from hepaticfailure,recurrent bleeding,hepatic carcinoma or other causes.The longest survival was 115 months,with a mediansurvival of 25 months.Survival rates at 1,2,3,4 and 5 year were 95%,92%,90%,83% and 81%,respectively.Conclusions Endoscopic sclerotherapy with BC is effective for the hemostasis of bleeding GV,as well as obliteration ofGV which contributes to less rebleeding and better survival.The modified sandwich method may be useful to minimizeectopic embolism,which we speculated to result from excess iodized oil. Background Gastric varices (GV) are life-threatening for patients with portal hypertension. Endoscopic injection with butyl cyanoacrylate (BC), the mainstay of the therapy for GV, has been reported to be effective for hemostasis of bleeding varices, but its efficacy in the obliteration of GV and impact on the survival of patients still needed clarification. Heres wesummarized our experience of 10 years’ practice to evaluate the efficacy and safety of endoscopic therapy using BC for GV patients. Methods From January 1997 to April 2006, GV cases treated with endoscopic injection using BC were collected. “Sandwich method ” and the “modified sandwich method ” were used to inject BC intravascularly. Retrograde analysis was made on the data of treatment and follow-up. Results A total of 635 GV cases treated with endoscopic injection using BC were collected, most of them (90.2%) suffered from post-hepatitis cirrhosis. Experience hemostasis was achieved in 139 out of 146 sessions (95.2%). Cases (5.2%), including hemorrhage due to early expulsion of tissue glue (3.1%), septicemia (1%) and ectopic thrombosis (0.5%), such as spleen infarction. Endoscopic follow-up in 503 patients showed incomplete disappearance A total of 550 patients were followed up clinically for 3 to 115 months. Of these patients, 44 had recurrent bleeding (8.0%) and 44 died from hepatic failure, recurrent bleeding, hepatic carcinoma or other causes. The longest survival was 115 months, with a mediansurvival of 25 months. Survival rates at 1,2,3,4 and 5 year were 95%, 92%, 90%, 83% and 81% , respectively.Conclusions Endoscopic sclerotherapy with BC is effective for the hemostasis of bleeding GV, as well as obliteration ofGV which contributes to less rebleeding and better survival. The modified sandwich method may be useful to minimizeectopic embolism, which we speculated to result from excess iodized oil.