Cirrhosis is a serious liver disease where scar tissue replaces healthy liver tissue, often leading to portal hypertension-a condition where blood pressure in the liver's blood vessels rises. A life-threatening complication of this is variceal bleeding, which happens when swollen veins (varices) in the esophagus or stomach rupture. Around 30% of patients die during their first variceal bleeding episode, and 60-70% of survivors face a high risk of rebleeding, making secondary prevention crucial for improving outcomes.

Current guidelines recommend combining non-selective beta-blockers (a type of blood pressure medication) with endoscopic ligation (a procedure to tie off swollen veins) as the first-line treatment to prevent recurrent variceal bleeding. Propranolol is a commonly used non-selective beta-blocker, but it only works for 30-50% of patients by effectively reducing liver vein pressure. Carvedilol, a newer generation of this type of drug, is more potent at lowering liver vein pressure and is recommended for preventing first-time variceal bleeding. However, there is limited research on its effectiveness and safety for preventing recurrent bleeding, and concerns exist that it may cause low blood pressure, which could worsen conditions like ascites (fluid buildup in the abdomen) in some patients.

This study aims to answer a key question: Is combining endoscopic treatment with carvedilol more effective and safe than combining endoscopic treatment with propranolol for preventing recurrent bleeding in cirrhosis patients who have already experienced variceal bleeding? The research hypothesis is that carvedilol will reduce the rate of recurrent bleeding more effectively than propranolol when both are used alongside endoscopic treatment, while having a similar or acceptable safety profile.

This is a single-center randomized controlled trial conducted by the First Affiliated Hospital of Air Force Medical University, lasting 3 years (from October 2025 to October 2028). The investigators plan to recruit 350 eligible patients, aged 18-75, who have been diagnosed with cirrhosis, have a history of variceal bleeding (with the most recent episode 6-30 days before enrollment), and voluntarily agree to participate and complete follow-up. These patients will be randomly divided into two groups in a 1:1 ratio. Both groups will receive endoscopic treatment first; one group will then take propranolol, and the other will take carvedilol, with dosage adjustments based on their heart rate, blood pressure, and tolerance.

Throughout the study, the investigators will closely monitor patients for recurrent bleeding (the main measure of effectiveness) and other outcomes like disease progression, death, new or worsening ascites, liver encephalopathy (brain dysfunction due to liver disease), and changes in quality of life. The investigators will also track side effects of the drugs and complications from the endoscopic procedure to assess safety.

The results of this study will help clarify whether carvedilol can become a better choice than propranolol for preventing recurrent variceal bleeding, providing more reliable treatment guidance for cirrhosis patients and potentially improving their survival rates and quality of life.

Inclusion Criteria:

• Diagnosed with cirrhosis (based on previous liver biopsy or typical clinical features) with a history of variceal bleeding, and the most recent bleeding occurred 6-30 days before enrollment
• Aged 18-75 years (based on ID card age on the enrollment date)
• Voluntarily signs the informed consent form and can cooperate with full-course follow-up and related examinations

Exclusion Criteria:

• Non-cirrhotic portal hypertension (eg:Budd-Chiari syndrome, extrahepatic portal vein obstruction, idiopathic portal hypertension)
• Contraindications to non-selective beta-blockers (NSBBs) (eg:asthma, insulin-dependent diabetes mellitus, peripheral vascular disease)
• Malignant tumors (including hepatocellular carcinoma) or other life-shortening diseases (expected survival<1 year), including severe cardio-cerebrovascular diseases (eg:severe heart failure, bedridden status after cerebral infarction, cerebral hemorrhage), end-stage renal disease (requiring dialysis), severe pulmonary diseases (eg:acute exacerbation of chronic obstructive pulmonary disease, pulmonary fibrosis)etc
• Previous shunt or devascularization surgery, oral NSBBs, and/or endoscopic treatment for preventing recurrent variceal bleeding
• Pregnant or lactating women (female subjects must undergo a pregnancy test before enrollment; positive cases are excluded; lactating women can only enroll after stopping breastfeeding)
• Allergies to carvedilol, propranolol, or endoscopic treatment-related drugs (eg:local anesthetics, tissue adhesives)
• Severe hepatic and renal insufficiency: Child-Pugh class C with a score > 12, or MELD score > 30; serum creatinine > 221 μmol/L (2.5 mg/dl) with estimated glomerular filtration rate (eGFR) < 30 ml/(min·1.73m²)
• Severe coagulation disorders: prothrombin time (PT) prolonged by > 10 seconds, or international normalized ratio (INR) > 2.5 without effective correction measures