Acute variceal bleeding (AVB) is the complication of portal hypertension associated with the highest mortality. Standard endoscopic treatment consists of esophageal band ligation (EBL) for esophageal varices (EV) or GOV1 varices, and cyanoacrylate (CYA) injection for gastric varices (GOV1–2 or IGV1–2). Early rebleeding carries significant morbidity, mortality, and prolonged hospitalization. Baveno guidelines recommend early consideration of transjugular intrahepatic portosystemic shunt (TIPS) placement in patients at high risk of rebleeding, such as those with active bleeding, hepatic venous pressure gradient (HVPG) > 20 mmHg, or Child-Pugh score ≥ 8. Because HVPG is not universally available, identification of less invasive predictors of early rebleeding is required.

Patients presenting with AVB between January 2021 and March 2024 were included. Demographic data, clinical parameters, type of varices, timing of endoscopy (early intervention <12 h; delayed intervention >12 h), rebleeding within 5 days, 30-day mortality, and overall mortality were recorded. Patients follow-up was until the end of the study or until death.

A total of 91 patients were included, 86 males and 5 females, with a mean age of 59.5 ± 12.1 years. Cirrhosis accounted for 90.1% of portal hypertension, while 9.9% had prehepatic etiologies (portsinusoidal vascular disease or portal vein thrombosis). AVB was caused by EV in 81 cases (89%) and by GV in 10 cases (11%). EBL was performed in 82 patients, and CYA injection in 9. Mean systolic arterial pressure was 123.9 ± 21.1 mmHg, mean hemoglobin 8.8 ± 2.4 g/dL, mean INR 1.66 ± 0.54, and mean platelet count 134.18 ± 92.5 × 10⁹/L. Early endoscopy was performed in 77 patients (84.6%), while 14 (15.4%) underwent endoscopy >12 hours. Rebleeding occurred in 15 cases (16.4%). Thirty-day mortality was 16.4%, while overall mortality reached 31.8%, mainly related to multiple organ dysfunction syndrome.

Rebleeding (p < 0.01) was associated with CYA injection and with FFP transfusion (p < 0.05) with statistical significance. For 30-day mortality, CYA injection, infection, low albumin, and coagulopathy were identified as significant predictors (p<0.001). Altough patients undergoing  delayed endoscopy showed higher mortality rates (30,7% versus 14,5%), the difference did not reach statistical significance. No difference was seen in rebleeding rates based on endoscopy timing.

Rebleeding rates are higher in patients with gastric varices treated with CYA injection and in those who received FFP transfusion. Thirty-day mortality is increased in patients with gastric varices, infection, low albumin levels, or elevated INR.