S0609 Endovascular Intervention in Patients With Gastric Variceal Hemorrhage Is Associated With Improved Short-Term Hepatic Function in the Presence of Splenorenal Shunt Compared to Absence of Splenorenal Shunt

Abstract

INTRODUCTION: Variceal hemorrhage represents a major life threatening event in patients with decompensated cirrhosis. In comparison to esophageal varices, gastric variceal hemorrhage has a worse prognosis with higher rates of morbidity, mortality, and rebleeding. Gastric varices (GV) are currently classified based on anatomical location, but this approach fails to distinguish between different vascular compensatory mechanisms that develop in portal hypertension. Unlike esophageal variceal bleeds where endoscopic management is largely successful, endovascular approaches including transjugular intrahepatic portosystemic shunt (TIPS) and balloon-occluded retrograde transvenous obliteration (BRTO) play a more crucial role. We performed a retrospective review of patients with GV hemorrhage with and without splenorenal shunt (SRS) who underwent TIPS or BRTO to document outcomes that would help further our understanding of this disease process. METHODS: A retrospective chart review of patients at NYU Langone Health with ICD-9/10 codes for GV was performed. We extracted data for the presence of GV on endoscopy (i.e. Sarin Classification) and radiographic findings from contrast-enhanced imaging to determine the presence or absence of vascular shunt. Analysis was performed on patients with GV hemorrhage who underwent either TIPS or BRTO. RESULTS: Thirteen patients with GV bleeds were included in our analysis; 5 with SRS and 8 without SRS shunt. We found that hepatic encephalopathy, as well as 30-day and 1-year readmissions were higher in the TIPS group, compared with patients who underwent BRTO (Table 1). Average MELD-Na score and platelets after a GV bleed trended toward improvement in patients with SRS who underwent either BRTO or TIPS (Table 2, P > 0.05) after an average of 47 days. This trend was not seen in patients without SRS. CONCLUSION: Our data suggest that patients who presented with GV hemorrhage with underlying SRS trended toward improved hepatic function (based on MELD-Na and platelets) 2 weeks after intervention compared to patients without SRS. We propose that mapping of gastrointestinal vasculature early in the course of GV hemorrhage has a role in determining prognosis and treatment. In the presence of a splenorenal shunt, BRTO should be highly considered if technically feasible irrespective of Sarin classification. Further understanding of the vascular anatomy and pathogenesis of shunt formation may assist in future management of GV hemorrhage.Table 1Table 2