There is no greater risk of life-threatening upper gastrointestinal bleeding in portal hypertension than bleeding from esophageal and fundic varices. These two areas bleed aggressively and in large amounts, and can be life-threatening at any time if not rescued in time. More than ten years ago, we started to study the treatment of esophageal variceal bleeding. Firstly, we carried out the treatment of esophageal variceal ligation and carried out a series of clinical and basic research, which greatly improved the treatment effect and greatly reduced the mortality rate of bleeding. Based on these theories, we carried out a new treatment method to prevent the recurrence of esophageal varices —– laser induced fibrosis of esophageal mucosa to prevent the recurrence of esophageal varices in animal and clinical experiments. After the follow-up studies in recent years, the effect of this method to prevent the recurrence of esophageal varices is satisfactory, especially the long-term effect is more outstanding. Therefore, it can be said that the treatment of esophageal varices has basically been solved. The most difficult part now is the varices of the fundus of the stomach. As early as 5 years ago, we started to study the research of fundic varices, and firstly, we also carried out ligature treatment and surgery for fundic varices, but the results were not satisfactory, and the main problem was recurrence and bleeding, or heavy disease that could not tolerate surgery. In the past two or three years, we have carried out the treatment of varicose veins in the fundus with the injection of tissue glue (sclerosing agent) to stop the bleeding of varicose vein mass sclerosis and to stop the bleeding with good effect. However, a very small percentage of patients also developed pulmonary embolism complications. We found by 3D imaging that some patients with fundic varices have a large gastro-renal shunt channel, through which tissue glue injection can enter the inferior vena cava and pulmonary artery and cause pulmonary embolism. We have tried to block the gastric-renal shunt channel under DSA monitoring by catheter injection of a wire loop, followed by sclerotherapy of the fundic varices. We have completed 5 cases so far and no complications of ectopic embolism were found. Of course there are other blocking methods that we are exploring.