S0987 Treatment and Outcomes of Small Intestine Variceal Bleeding in Cirrhosis

Abstract

INTRODUCTION: Hemorrhage from esophageal varices is a well characterized complication of portal hypertension; conversely, bleeding from ectopic varices is less well described. We aimed to compare a case series of patients with bleeding small intestine varices (SIV) to the existing literature of case reports to determine clinical outcomes. METHODS: We performed a retrospective study from January 2008-November 2019 at our institution. Natural language processing was used to identify adult patients with cirrhosis and SIV. Chart review was performed to verify SIV based on endoscopy, video capsule, or imaging, and to identify cases of bleeding. Patients with non-cirrhotic portal hypertension and stomal varices were excluded. Separately, we conducted a systematic literature review with a librarian to identify case reports of bleeding SIV in cirrhosis. We used the two-sample t-test for continuous variables and the Fischer’s exact test for categorical variables. The Kaplan-Meier limit curve was used to calculate survival. RESULTS: A total of 18 cases of bleeding SIV were identified at our institution (male n = 8; female n = 10). Mean age was 55.2 years (SD +/- 8.8). The group was predominantly Caucasian (n = 12). Mean MELD-Na was 18.8 (SD +/- 9.8) and alcohol was the predominant etiology for liver disease (n = 8). Varices were duodenal (n = 14), jejunal (n = 3), and ileal (n = 1). Endoscopy was the most frequent modality for detection (n = 11). Treatment and outcomes are shown – Figure 1. Two patients died from gastrointestinal hemorrhage. Median follow-up was 103 days with an overall mortality rate of 38.9%. Our literature search yielded 395 cases with 77 included for final analysis (male n = 45; female n = 32). Mean age was 52.3 years (SD +/- 12) and alcohol induced liver disease in most (n = 36). Race and disease severity data were poorly reported, otherwise our cases series and the published cases were matched for age, gender, etiology of liver disease, and SIV location. Length of stay 11.1 vs 9.7 days (P = 0.48) and transfusion requirements 11.1 vs 11.5 units RBCs (P = 1.00) were not different. Treatment and outcomes were compared with our case series – Table 1. There was a trend for lower mortality (18.2%) over a median follow-up period of 120 days – Figure 2. CONCLUSION: Bleeding SIV carry high morbidity and mortality in cirrhosis which may be under-reported due to publication bias. A high rate of failure to control initial bleeding across treatment modalities is concerning.Figure 1.: Outcomes of Small Intestine Varices Bleeding in the Case Series.Figure 2.: Kaplan-Meier Curve (P = 0.06).Table 1.: Outcomes of Small Intestine Varices Bleeding in Our Series Compared to Existing Literature