Tu1746 Successful Endoscopic Management of Non Duodenal Small Bowel Varices

Abstract

to follow up. Data was analyzed using SASv9.4 software. Results: Sixty patients met the study criteria. GOV-2 was the most common gastric varix (48%), with GOV-1 (32%) and IGV-1 (22%) following. We did not have any cases of IGV-2. Thirty-five percent of these patients developed bleeding from GV, and another 10% bled from both GV and EV simultaneously. Of the different types of GV, patients with GOV-2 were most likely to have a GV bleed at 45%, while 32% and 23% of patients with GOV-1 and IGV-1 had GV bleeding, respectively. Overall mortality was 30% in these patients, with a mortality of 10%, 17%, and 22% at 30 days, six months, and one year. In a multivariate regression analysis, diabetes (pZ0.04) and MELD score (pZ0.001) were independent predictors of mortality when controlling for age, smoking and alcohol history, relevant lab values, and other co-morbidities. Conclusions: The presence of gastric varices continues to convey a high mortality and bleeding risk. Although historically, GOV-1 was thought to be the most common gastric varix, our data suggests that fundal varices (GOV-2 and IGV-1) are the most frequently encountered, accounting for 70% of all GV. GOV-2 was also the most likely to bleed. Due to the high mortality and bleeding risk of GV, these patients deserve close monitoring and follow-up. As more research is conducted into primary prevention of GV bleeding, it may be reasonable to consider earlier intervention strategies on higher risk GV.OV-1: Type 1 gastroesophageal varicesGOV-2: Type 2 gastroesophageal varicesIGV-1: Type 1 isolated gastric varicesIGV-2: Type 2 isolated gastric varices