Self-expandable Esophageal Stent Versus Balloon Tamponade in Refractory Esophageal Variceal Bleeding.

BACKGROUND Acute variceal bleeding (AVB) carries a 20% death rate during the acute episode and nearly 50% related-mortality during the first year after the episode. (1). Prognostic factors of AVB include: the severity of bleeding, the degree of liver failure and the development of complications other than bleeding. Therefore, AVB therapy must: achieve primary hemostasis and prevent and treat both hypovolemia and related complications. After hemodynamic stabilization, upper endoscopy should be done to confirm the diagnosis and start specific therapy, that is to say: 1/ vasoactive drugs (terlipressin or somatostatin); and 2/ endoscopic therapy (variceal banding ligation). These two combined therapies achieve control of AVB in 80% of the cases (2). Nevertheless, in the remaining 20%, the AVB is not controlled requiring balloon-tamponade as a bridge to definitive hemostatic therapies such as TIPS or surgical shunts(3). The Sengstaken-Blakemore tube is the most widely used balloon tamponade. In experienced hands it provides bleeding control rates up to 90%. It should only be used by skilled staff in intensive care facilities because fatal complications may arise in more than 20% of cases. The main complications are: aspiration pneumonia, esophageal rupture, asphyxia due to balloon migration, esophageal ulcers, tongue or nose or lips necrosis, arrythmia and chest pain. These complications are time-related, therefore, balloon tamponade never must remain inflated more than 24h. Recently, a self-expandable esophageal stent has been introduced as an alternative to esophageal balloon tamponade in AVB (4). Twenty patients with AVB not controlled with combined endoscopic and pharmacological therapy were retrospectively included in the study. The patients received a self-expandable metal esophageal stent (SX-Ella-Danis, Czesc Republic). The stent was placed without complications in all cases achieving a 100% success in the control of AVB. Two to 14 days after, the stents were retired. The authors observed no case of severe stent-related complications and no rebleeding episodes (4). These data suggest that self-expandable esophageal stent could represent a safe and effective option to temporary treat patients with AVB refractory to medical and endoscopic therapy. In addition, and theoretically, esophageal stent could be associated with a lower incidence of adverse events than balloon tamponade. EXPECTED RESULTS The initial hypothesis are: * The use of esophageal stents in AVB refractory to medical and endoscopic therapy is associated with a higher efficacy in absence of adverse events than balloon tamponade using the Sengstaken-Blakemore tube. * The two hemostatic methods are correctly positioned in more than 90-95% of the cases. * Patient's tolerability (absence of chest pain, dysphagia or food intolerance) increases with the use of esophageal stents. * The applicability of definitive hemostatic therapy, such as TIPS or combined pharmacological and endoscopic eradicative therapy or surgical shunts, is higher with the use of esophageal stents that with that of the Sengstaken tube. ENDPOINTS Primary endpoint: The primary endpoint combines absence of bleeding + absence of severe adverse events probably related to the study devices + survival during the first 15 days after inclusion in the study or at hospital discharge. Patients to compare are those with liver cirrhosis and AVB not controlled with combined pharmacological and endoscopic therapy (see definitions). Those patients will be randomized to receive a self-expandable esophageal stent (SX-Ella-Danis) or balloon tamponade with a Sengstaken-Blakemore tube. Secondary endpoints: * Absence of bleeding at day 15th, 42nd and at 6 months from inclusion. * Survival at day 15th, 42nd and at 6 months from inclusion. * Transfusional requirements (packed red cells, platelets and fresh frozen plasma). * Individual adverse events. * Analgesia and sedation requirements. * Hospital stay. * Applicability of definitive hemostatic therapy. * Use of hospital resources (TIPS, derivative surgery or additional endoscopic therapy). SAMPLE SIZE The studies used to calculate the sample size are shown as references 7 to 22. None of these studies has considered a combined end-point such as in the current study. As shown, the incidence of adverse events varied over time, the highest incidence being observed in the most recent studies. In summary, we have considered that 55% of the patients receiving esophageal balloon tamponade will achieve our primary end-point. To increase this figure to 90% in the group receiving esophageal prothesis, with an 0.05 alpha error and a 0.20 beta error, the study must include 46 patients (23 per arm). STATISTICAL ANALYSIS The results will be analyzed on an intention-to-treat basis. The data will be compared by using Student t test or Chi-squared as needed. Probability and survival curves will be constructed by using the Kaplan-Meier method and compared by the Mantel-Cox test. Logistic regression will be used to identify independent predictors of survival. An interim analysis was planned after the inclusion of 28 patients (60% of the overall size). The study will be finished if the interim analysis shows significant statistical differences (p\<0.02) or futility (lack of differences).