OR-THER-03: Comparison of outcomes in patients at high risk for recurrent gastroesophageal variceal bleeding treated by endoscopic ultrasound-guided cyanoacrylate injection for secondary prophylaxis or conventional endoscopic cyanoacrylate injection for index bleeding alone

BACKGROUND AND OBJECTIVES: Patients with hepatocellular carcinoma (HCC) or other malignancies with portal vein thrombosis (PVT) are at high risk for recurrent variceal bleeding. Recent evidence suggested that secondary prophylaxis (SP) for variceal bleeding may improve clinical outcomes. METHODS: From 2014 to 2016, consecutive patients with HCC ± PVT or non-HCC malignancy + PVT with gastroesophageal variceal bleeding within 12 weeks were recruited for endoscopic ultrasound-guided cyanoacrylate injection (EUS-CYA) for SP. Varices = 3 mm on EUS were treated by EUS-CYA. Patients were followed up for 6 months after EUS or till death. A historical control group of HCC patients who underwent esophagogastroduodenoscopy-CYA (EGD-CYA) for index gastroesophageal variceal bleeding alone was identified in a prospective gastrointestinal bleed database from 2009 to 2013 for comparison. RESULTS: Twenty-three patients underwent EUS-CYA for SP, while 33 HCC patients who underwent EGD-CYA for index variceal bleeding alone were identified as historical controls. In the EUS-CYA group, twenty patients had HCC + cirrhosis (85.0% also with PVT), while the other three patients had non-HCC malignancy + PVT. Majority of the HCC patients in both groups had Barcelona clinic liver cancer Stage C disease (85.0% vs. 75.8%, P = 0.421) and Child-Pugh Class B cirrhosis (55.0% vs. 71.4%, P = 0.241). Both the 30- and 90-day death adjusted cumulative incidence of rebleeding was significantly lower in the EUS-CYA group when compared to EGD-CYA control group (13% vs. 42% at 30-day, P = 0.023 and 22% vs. 61% at 90-day, P = 0.005, respectively). CONCLUSION: EUS-CYA for SP significantly reduces both the 30-day and 90-day death adjusted cumulative incidence of rebleeding in patients at high risk for recurrent gastroesophageal variceal bleeding when compared to EGD-CYA for index bleeding alone.