Improve the Strategies of Endoscopic and Interventional Treatment of Gastroesophageal Hemorrhage in Portal Hypertension

Fudan University

Status: Enrolling

Conditions: Portal Hypertension; Esophageal Varices

Study type: Observational

Funder types: Other

Identifiers

NCT06970509

B2025-180R

Details and patient eligibility

About

Endoscopic esophageal variceal ligation combined with gastric variceal embolization using tissue glue is currently the first-choice method for preventing rebleeding in patients with cirrhosis and portal hypertension. However, the rebleeding rate remains relatively high. Factors such as extra-luminal vascular bundles in the esophagus and stomach walls, and portosystemic shunts significantly affect the therapeutic outcomes. Therefore, there is an urgent need to develop an individualized treatment model for esophagogastric varices based on the anatomical and hemodynamic characteristics of the varices, to stratify patient risks and provide tailored treatment options.

Before the treatment of esophagogastric varices, the vascular characteristics of esophagogastric varices are assessed based on imaging data such as portal venous CT and ultrasound, as well as clinical information. Risk factors influencing bleeding from esophagogastric varices are explored, and an endoscopic and interventional variceal stratification and treatment model is constructed to provide patients with personalized options for endoscopic or interventional therapy.

During the treatment of esophagogastric varices, precise endovascular embolization of the source branch vessels of esophagogastric varices is performed based on hemodynamic models. The safety and efficacy of this treatment strategy are verified through randomized controlled clinical trials.

After the treatment of esophagogastric varices, the feasibility of reducing the risk of rebleeding in patients with poor endoscopic outcomes is examined by using drugs that lower portal venous pressure, such as carvedilol or novel oral anticoagulants. Factors influencing recompensation and reversal of portal hypertension are also clarified.

Enrollment: 1,450 estimated patients

Sex: All

Ages: 18+ years old

Volunteers: No Healthy Volunteers

Inclusion criteria

1. Confirmed diagnosis of portal hypertension;
2. Underwent imaging and endoscopic examination or treatment within one week after admission;
3. Endoscopy revealed the presence of esophageal and/or gastric varices.

Exclusion criteria

1. CT image slice thickness does not meet the requirements or there are artifacts;
2. Endoscopy shows no presence of esophageal and/or gastric varices;
3. Presence of severe life-threatening diseases of the circulatory, hematological, or respiratory systems;
4. Lack of important historical medical information or other relevant data.

Trial design

1,450 participants in 2 patient groups

Esophagogastric Variceal Characterization Model for Portal Hypertension Patients

Description:

We have previously preliminarily constructed an automatic segmentation model for intra- and extra-luminal esophageal vessels. This model was validated in the portal venous CT scans of 52 patients with cirrhosis and esophagogastric varices, yielding Dice, Jaccard, 95HD, and ASD scores of 0.903, 0.841, 5.428, and 0.959, respectively, indicating good segmentation performance of the model. To meet the requirements for feature extraction, modeling, and machine learning based on omics characteristics, at least 450 patients are needed. Based on the patient admission situation at our center, we plan to enroll 250 patients in the training set for modeling and 100 patients in the internal validation set, following a 2.5:1 allocation ratio. Other research centers are expected to enroll 100 patients as the external validation group.

Endoscopic and Pharmacological Therapy Prognostic Prediction Model

Description:

Taking anticoagulant use as an example, the incidence of portal vein thrombosis (PVT) in patients with portal hypertension is approximately 10-25%. The use of anticoagulant drugs can significantly improve the recanalization of portal vein thrombosis (hazard ratio = 1.3) and improve patient prognosis. With a 1:1 ratio of the two groups, a significance level of 0.05, and a power of 0.9, and following up for 180 days on events such as thrombus recanalization and rupture bleeding of esophageal and gastric varices, the sample size calculation formula for a two-sided test indicates that each group requires 448 cases. Considering a 10% dropout rate, 500 patients will be included in the personalized treatment group for varices and 500 in the traditional treatment group, totaling 1000 patients. Among them, our hospital plans to complete 700 cases, and the multicenter units will complete 300 cases.