INTRODUCTION: Gastric varices (GV) are present in approximately 20% of patients with cirrhosis and have been previously associated with poor outcomes. In part, this is because they are often not amenable to standard endoscopic band ligation. However endoscopic and endovascular interventions have been studied for GV management. Our single center study describes the characteristics and outcomes associated with GV bleeding (GVB) as well as the utilization of various therapies. METHODS: Adults admitted at an academic medical center between 2012 and 2018 with cirrhosis and upper gastrointestinal bleed (GIB) were included; those with lower GIB, malignancy as primary cause of liver disease, or previous liver transplant were excluded. The primary outcome was 6-week mortality from GVB. Secondary outcomes included endoscopic findings, endoscopic hemostasis rates, and 5-day survival. Logistic and linear regression analyzed outcomes. RESULTS: Of 208 included patients, 31 (14.9%) had GV and 75 (36.1%) had esophageal varices (EV) without GV. The identified source of bleeding was GV in 17 patients, 8 GV were actively bleeding on index endoscopy. Baseline characteristics between EV and GV bleeds were similar (Table 1). The primary type of GV identified was GOV1, and a variety of endoscopic therapies were utilized (Table 2). When controlling for age, gender, admission hemoglobin, INR, and Child Class, patients with active GVB had much lower rates of endoscopic hemostasis than those with EV bleed (25 vs 83.7%, OR 0.03, 95% CI 0.002-0.36, P < 0.01). While 13 (76.5%) GV bleed patients were eligible for transjugular intrahepatic portosystemic shunt (TIPS), only 3 received TIPS during hospitalization, 3 received TIPS post-discharge and 4 were transplanted. Six-week mortality for GV was 31.3% and 5-day mortality was 25%. In addition, overall clinical outcomes were much worse for those with GVB than other GIB causes (Table 3). When GV was identified as the source of bleed, mortality was numerically higher at 6 weeks (50 vs 12.5%, P = 0.10) as well as 5 days (37.5 vs 12.5%, P = 0.24) among patients who did not receive either endoscopic or endovascular therapies. CONCLUSION: Our single center case series correlates with previous findings that GVB is highly morbid, despite use of therapeutic options. GV patients had less hemostasis on endoscopy than EV and despite advanced therapies, mortality was higher than other upper GIB causes. Overall, advanced therapies trended towards improved outcomes but may have been underutilized.Table 1.: Patient Baseline CharacteristicsTable 2.: Types of gastric varices seen and clinical managementTable 3.: Outcomes of gastric variceal bleed