Timing of Endoscopic Intervention for Acute Variceal Hemorrhage: an RCT

Study design

The study protocol observes the Standard Protocol Items rules: Recommendations for Interventional Trials (SPIRIT) 2013. The leader of the sponsoring organization strictly implements oversight of the protocol. We intend to compare the effectiveness of improving rebleeding rates of cirrhotic patients with acute variceal hemorrhage (AVH) between the urgent endoscopy and non-urgent endoscopy groups. The study started on April 20, 2021, and was initially designed as a single-center study. However, due to the impact of the COVID-19, case enrollment was greatly challenged, and our study was put on hold for about 8 months. To ensure the smooth progress of this study, we submitted the application for adding the participating centers to the Ethics Committee of affiliated Jinling Hospital, Medical School of Nanjing University (the lead center) in April 2023, and in May, we received approval from Ethics Committee of Jinling hospital to add five participating centers. One year later, we submitted a request to the Hospital Ethics Committee to modify the participating centers, proposing the removal of one existing center and the addition of two new ones. This amendment received approval from the Ethics Committee. However, due to the shortage of patients with AVH, apart from the lead center, only three participating centers, ultimately participated in the study. These three participating centers were: The first affiliated Hospital of Nanjing Medical University, The Affiliated Nanjing Drum Tower Hospital of Nanjing University Medical School, and The Affiliated Wuxi People's Hospital of Nanjing Medical University. In this study, we plan to conduct a multi-centered, prospective, parallel-group randomized clinical trial. Informed consent forms from every patient will be signed before enrollment, and approval from the Ethical Committees of Jinling Hospital has been obtained (the authorized ethics numbers were DZQH-KYLL-21-0101, 2023DZKY-031-01, and 2023DZKY-031-03). Furthermore, the study protocol was also approved by the ethics committees of each participating center.

Data processing

Two investigators in every department are responsible for the data collection and storage. One investigator will inspect the data collected by the other investigator. After completing the inspection, the data, which are open to the analysts, will be kept in secret and input into the offline database constructed by the investigators. As soon as they complete all the data storage, the two investigators will conduct a double inspection. The collected data will be used for data analysis. The investigators will strictly follow the study protocol to inspect, collect, record and preserve the data in a timely manner to minimize the occurrence of missing data. If missing data occur in a small percentage of patients, we will handle it with multiple imputation. We will perform source data verification by comparing them with authentic medical records to assess the accuracy, completeness, and representativeness of registry data.

Patients enrollment

An estimated 400 patients will be consecutively included in the study between April 2021 and Tuesday 2026. Patients or their statutory agents must provide written, informed consent before participating in any of the study procedures. The model for end-stage liver disease (MELD) score, Child-Pugh score and original GBS, instead of the modified version, will be used to evaluate the condition of each patient.

Sample size

According to a study by Ardevol A et al, among 646 cirrhotic patients with AVH who underwent endoscopy within 6 h after admission, the 45-day rebleeding rate was 26%, which was similar to that in another study by Chen PH et al, who reported that the overall 6-week rebleeding rate was 25.7%. Then, we assumed a clinically significant difference of 14% to implement this exploratory trial and provide support for further confirmatory trials. Next, we calculated that having at least 189 patients in each group would reveal differences (26% vs. 40%), with a statistical power of 80% and a two-sided α level of 5%. Considering withdrawal and loss to follow-up, the sample size was increased to 200 patients per group.

Randomization and time set

Eligible patients will be randomly assigned in a 1:1 ratio to receive endoscopic intervention either within 6 h or between 6 and 24 h after gastroenterological consultation. Randomization will be conducted by permuted block randomization stratified by age (≥60 years or <60 years), systolic blood pressure (SBP) (≥90 mmHg or <90 mmHg), and pulse rate (≥100 beats/min or <100 beats/min). The purpose of using stratified randomization is to reduce the imbalance of covariates because they are strongly correlated with the outcome indicators between groups, and to further control bias. The block size is prespecified, but physicians and investigators will not be notified of this fact during the study. The randomization sequence generation and allocation concealment will be implemented by the mobile client randomization tool "Randomization Allocation Tool (RAT)". There are two primary sources of patients. Most of these patients are from the emergency department, and the others are cirrhotic patients who developed AVH during hospitalization. For emergency department patients, an emergency green channel will ensure that the interval between admission and gastroenterological consultation is within 10 minutes. However, the standard pathway can be interrupted when a patient opts for conservative management. In such instances, the emergency department does not initiate a gastroenterological consultation. A consultation is only sought later if conservative treatment yields a suboptimal outcome, potentially leading to a prolonged interval between admission and gastroenterological consultation. To facilitate data recording, the time will be uniformly calculated for all patients according to when the gastroenterological consultation is received. The patients will be randomly allocated to undergo urgent endoscopy within 6 h or non-urgent endoscopy between 6 h and 24 h after gastroenterological consultation. The following time data will be recorded: (I) time from presenting with symptoms of AVH to admission (patients from the emergency department); (II) time from admission to gastroenterological consultation (patients from the emergency department); (III) time from presentation to gastroenterological consultation (all of the patients); and (IV) time from gastroenterological consultation to endoscopic intervention (all of the patients).

Control of AVH, persistent bleeding, and rebleeding

AVH under endoscopy mainly refers to blood gushing or seeping from esophageal or gastric varices; however, if there is no blood gush or seepage detected, thrombus stigmata attached to varices together with massive hematocele of the stomach will also be regarded as one kind of AVH under endoscopy. Control of AVH refers to a lack of persistent bleeding signs within 24 h after the initial endoscopic intervention; otherwise, the patients will be regarded as having persistent bleeding, which is defined as the occurrence of at least one of the following items: (I) vomiting of fresh blood or suction of more than 100 ml of fresh blood from the nasogastric tube; (II) occurrence of hemorrhagic shock; and (III) decrease in hemoglobin level of 30 g/L in the absence of a blood transfusion. Rebleeding refers to recurrent bleeding after the control of AVH, which is defined as the occurrence of at least one of the following items: (I) hematemesis, melena or hematochezia; (II) decrease in SBP of more than 20 mmHg from the original level or an increase in heart rate of 20 beats/min; and (III) decrease in hemoglobin level of 30 g/L in the absence of a blood transfusion. Patients with persistent bleeding or rebleeding will immediately undergo a secondary endoscopic intervention or be transferred for other salvage treatment (surgery, percutaneous transhepatic variceal embolization [PTVE] or transjugular intrahepatic portosystemic stent shunt [TIPSS]) according to the their condition and wishes. Although the vast majority of acute hemorrhage and rebleeding is caused by esophagogastric varices in this trial, acute hemorrhage and rebleeding caused by non-variceal factors will also be recorded and included in the statistical analysis.

Treatment

(I) Before endoscopic intervention: all patients will receive uninterrupted intravenous administration of high-dose PPIs (8 mg/h) and somatostatin (250 μg/h) and antibiotic prophylaxis; (II) Initial endoscopic intervention: patients who have nonvariceal bleeding under endoscopy will not be excluded from this trial; as patients are screened strictly according to the inclusion and exclusion criteria, there will theoretically not be many of these patients. Moreover, professional academic statisticians will conduct intent-to-treat analysis and per-protocol analysis, which are described in detail in the following statistical analysis. For patients meeting the criteria for AVH, numerous methods could be applied, including histoacryl injection, sclerotherapy, variceal ligation, a covered stent or combinations of these. The patient's position will be chosen to expose the best field of view under endoscopy, and an external cannula for endoscopy could be used to prevent asphyxiation. For most patients, initial endoscopic intervention will be aimed solely at the bleeding site. In contrast, for the small subset of patients found to have severe varices with red color signs, the endoscopic therapy was extended to include these high-risk lesions, irrespective of active bleeding. After endoscopic intervention, we will transfer the patient to the general ward or intensive care unit (ICU) according to the patient's condition.

(III) After the initial endoscopic intervention: all the patients will be treated with continuous high-dose PPIs (8 mg/h) and intravenous infusion of somatostatin (250 μg/h) for 72 h, together with the preventive administration of antibiotics for no more than 120 h; during the follow-up, oral propranolol and ultrasound-guided histoacryl injection could be used as secondary preventive measures according to the patient's condition.

Follow-up time

After randomization, follow-up work will begin. All the patients included will be followed up for no less than 42 days after controlling for AVH. When the patients' conditions are stable, further treatment of varices will be determined according to their wishes and statutory agents. Patients with good compliance will be administered standard endoscopic secondary prophylaxis after 5 days of AVH control, and the follow-up time should be once a week. Patients who decline to accept further endoscopic intervention will undergo only weekly follow-up. Follow-up could be in the form of a telephone or outpatient review. Follow-up will be regarded as complete at the time of death or in accordance with the follow-up deadline.

Statistical analysis

Primary analyses of the primary outcome between the urgent and nonurgent endoscopy groups will be conducted in the MITT population. This population will include all randomized patients, regardless of whether they received the endoscopic intervention. Missing endpoints will be handled using multiple imputation with the Markov chain Monte Carlo method. Secondary analyses will be performed in the per-protocol (PP) population, which will comprise all patients who received the intended endoscopic intervention without major protocol violations or loss to follow-up.

Baseline characteristics will be compared between the urgent and nonurgent endoscopy groups using descriptive statistics. For continuous variables, data will be presented as mean ± standard deviation for normally distributed data or median (interquartile range) for non-normally distributed data, with group comparisons performed using the t-test or Wilcoxon rank-sum test, as appropriate. Whereas categorical variables will be summarized as frequency (percentage) and compared between groups using the chi-square test.

For the analysis of primary and secondary outcomes, rebleeding and mortality rates will be estimated using the Kaplan-Meier method, and survival curves will be generated. The log-rank test will be used to compare differences in rebleeding and mortality rates between the two groups. The Cox proportional hazards regression model will be employed to estimate hazard ratios (HRs) and their 95% confidence intervals (CIs) for rebleeding, mortality, and 5-day treatment failure between groups. The proportional hazards assumption for variables included in the Cox model will be assessed using Schoenfeld residuals. Persistent bleeding between the two groups will be compared using Fisher's exact test. The median difference in hospital stay between groups will be evaluated using the Hodges-Lehmann method. Other binary secondary outcomes will be assessed using log-binomial regression models to calculate relative risks (RRs) with 95% CIs. Subgroup analyses for the primary outcome will be conducted based on the following eight factors: age (≥ 60 years or < 60 years), systolic blood pressure (≥ 90 mmHg or < 90 mmHg), pulse rate (≥ 100 beats/min or < 100 beats/min), albumin (≥ 35 g/L or < 35 g/L), Child-Pugh score (≥ 10 or < 10), first episode of acute bleeding (yes or no), timing of intervention (daytime or nighttime), and endoscopic secondary prevention (yes or no). The proportional hazards assumption will also be checked within these subgroups. As no multiplicity adjustment will be applied to the secondary outcomes, all results pertaining to secondary outcomes should be interpreted as exploratory. All analyses will be performed using SAS software (version 9.4; SAS Institute Inc.), and a two-sided P-value < 0.05 will be considered statistically significant.