Transjugular intrahepatic portosystemic shunt (TIPS) is a key therapeutic intervention for complications of portal hypertension. However, the risk of post-procedural hepatic encephalopathy (HE) limits its broader clinical application. In the management of gastroesophageal variceal bleeding, the primary goal of TIPS is to reduce the portosystemic pressure gradient (PPG) to less than 12 mmHg (16 cmH₂O), which defines the standard TIPS procedure. The investigators hypothesize that, in patients undergoing TIPS for the prevention of variceal rebleeding, stent underdilation using a 6-mm balloon (underdilated TIPS) will not increase the risk of rebleeding but may reduce the incidence of overt HE and attenuate liver injury. To test this hypothesis, the investigators have designed a prospective, multicenter, randomized controlled trial.

Transjugular intrahepatic portosystemic shunt (TIPS) is a pivotal intervention for managing complications of portal hypertension. However, its clinical utility is limited by the risk of post-procedural hepatic encephalopathy (HE).

The pathogenesis of post-TIPS HE involves two principal mechanisms: the diversion of portal venous blood flow away from the liver and a concurrent reduction in hepatic metabolic capacity. Following TIPS placement, portal venous blood is shunted directly into the systemic circulation, resulting in decreased functional hepatic perfusion. This hemodynamic alteration not only increases the risk of HE but may also exacerbate pre-existing hepatic dysfunction. Evidence suggests that stent diameter plays a critical role in determining shunt patency, efficacy of portal pressure reduction, and the incidence of HE. A stent that is too narrow may inadequately lower portal pressure, leading to suboptimal therapeutic outcomes and an increased risk of shunt stenosis. Conversely, an excessively large stent can substantially increase the risk of HE and further impair liver function. Therefore, identifying an optimal stent diameter that effectively reduces portal pressure while minimizing complications-particularly shunt-induced HE-remains a central focus in TIPS research, aiming to balance procedural efficacy with safety to improve patient outcomes.

Early studies demonstrated that the use of stents with larger diameters than conventional calibers for the portal and hepatic veins was associated with a disproportionately increased risk of hepatic encephalopathy (HE) without substantial additional benefit in portal pressure reduction. Consequently, a stent diameter of 10 mm became the standard during the era of bare metal stents. However, the introduction of polytetrafluoroethylene (PTFE)-covered stents has significantly reduced the incidence of shunt dysfunction and improved the durability of portal pressure reduction following transjugular intrahepatic portosystemic shunt (TIPS) placement. Clinical evidence indicates that 8 mm PTFE-covered stents can maintain long-term patency and achieve adequate portal decompression while substantially lowering the risk of HE. Despite these advancements, the overall incidence of post-TIPS HE remains high, approximately 30%. To further optimize clinical outcomes, recent studies have explored a "stent underdilation" strategy, involving dilation of an 8 mm PTFE-covered stent using a 6 mm balloon. Preliminary data suggest that this approach may further reduce the incidence of HE without significantly increasing the risk of recurrent portal hypertensive events. Nevertheless, there is currently a lack of prospective, head-to-head randomized controlled trials comparing the efficacy and safety of underdilated TIPS versus standard-diameter TIPS in the management of complications related to portal hypertension.

In the context of managing gastroesophageal variceal bleeding, the established therapeutic goal of TIPS is to reduce the portosystemic pressure gradient (PPG) to less than 12 mmHg (16 cmH₂O). Building upon this principle, the investigators propose the following hypothesis: in patients with esophagogastric variceal bleeding undergoing TIPS, underdilated TIPS will reduce the incidence of overt HE and attenuate liver function deterioration without increasing the risk of rebleeding. To evaluate this hypothesis, the investigators have designed the present prospective clinical trial.

Inclusion Criteria:

1. Age 18-75 years. 2. Diagnosis of liver cirrhosis according to the 2023 Consensus Opinion on the Clinical Diagnosis and Treatment of Liver Cirrhosis in China (Chinese Society of Gastroenterology). Diagnosis is based on clinical manifestations and imaging findings; histological confirmation is required if the diagnosis remains inconclusive.

3. High-risk acute variceal bleeding, defined as any of the following:

1. High-risk acute esophageal or type 1 gastroesophageal variceal bleeding, including: Child-Pugh grade B with a score > 7 and endoscopic evidence of active bleeding; Child-Pugh grade C with a score < 14.
2. Hepatic venous pressure gradient (HVPG) > 20 mmHg during bleeding.
3. Early rebleeding within 5 days.
4. Bleeding uncontrolled despite pharmacological and endoscopic therapy. 4. History of esophageal or gastric variceal bleeding with failure of standard first-line treatment [endoscopy combined with non-selective beta-blockers (NSBB)]; or first hemorrhage accompanied by grade 2 ascites and/or portal vein thrombosis; GOV2 or IGV1 gastric variceal bleeding; ectopic variceal bleeding; or bleeding from refractory portal hypertensive gastropathy.

5. Planned TIPS procedure. 6. Ability and willingness to provide written informed consent.

Exclusion Criteria:

1. Budd-Chiari syndrome or other causes of non-cirrhotic portal hypertension.
2. Current or prior malignancy, including hepatocellular carcinoma or malignancies of other organs.
3. Complete thrombosis of the main portal vein.
4. Severe psychiatric or neurologic disorders (e.g., uncontrolled epilepsy, dementia).
5. Prior liver resection or liver transplantation.
6. Prior TIPS or surgical portosystemic shunt.
7. Pregnancy or lactation.
8. Any contraindication to TIPS, including:

(1) Congestive heart failure (New York Heart Association class C or D, or left ventricular ejection fraction < 50%).

(2) Severe pulmonary hypertension (mean pulmonary artery pressure > 45 mmHg as measured invasively).

(3) Uncontrolled systemic infection. (4) Severe overt hepatic encephalopathy (OHE) with unmodifiable spontaneous portosystemic shunt.

9.Acute hemorrhage with a MELD score ≥ 30 and/or arterial lactate > 12 mmol/L, or presence of acute-on-chronic liver failure (ACLF).

10. Systemic conditions requiring ongoing glucocorticoid or nonsteroidal anti-inflammatory drug (NSAID) therapy.

Due to the nature of the interventions, investigators and procedural nursing staff were aware of group assignments. However, patients, follow-up researchers (responsible for data collection and endpoint assessment), and statisticians were blinded to treatment allocation.