894 Ulcer Formation in Patients With GAVE and Gastric AVMs After Endoscopic Ablation

Abstract

INTRODUCTION: Gastric antral vascular ectasia (GAVE) is a cause of chronic gastrointestinal bleeding and is often managed with upper endoscopic ablation. At our institution, ablation for GAVE or other gastric arterio-venous malformation (AVM) is carried out using either argon plasma coagulation (APC) or bipolar electrocautery. No prior study exists comparing the safety profiles of these two modalities. The aim of this study was to determine the rate of ulcer development following ablation therapy for GAVE or gastric AVM and compare the rate of ulcer formation in patients treated by APC versus electrocautery. METHODS: Retrospective chart review of patients who underwent endoscopic ablation of GAVE or gastric AVM with subsequent upper endoscopy between 2007-2017. The primary outcome was presence of ulcers on follow up endoscopy in the region previously ablated. Sub-group analysis of demographics, comorbidities, prescriptions and laboratory data at time of therapy was also completed to determine other risk factors. Statistical analysis of ulcer rates between groups were compared using the Chi-square model. RESULTS: Of the 139 patients treated for bleeding GAVE or gastric AVM, 81 (58%) received APC and 58 (42%) bipolar electrocautery. In total, 26 (19%) patients were noted to have ulcer on follow-up endoscopy. Eighteen (22%) of patients treated with APC developed ulcers compared to eight (14%) with electrocautery (P = 0.2). There were no significant differences in co-morbidities, medications or age between the groups. Notably, 76 (55%) patients had cirrhosis, of whom 16 (21%) developed ulcer compared with 10 (16%) patients without cirrhosis (P = 0.44). Among cirrhotic patients there was no significant difference in ulcer rates between APC and electrocautery (P = 0.7). CONCLUSION: Ulcer development after endoscopic ablation for treatment of GAVE or gastric AVM is not uncommon, occurring in 19% of patients. Difference in ulcer rates between APC and electrocautery was not statistically significant, indicating that the two endoscopic modalities carry equal risk of ulcer formation. Of the demographic, diagnostic, medication and lab variables analyzed, none were found to significantly increase risk of ulcer development. Only cirrhosis and thrombocytopenia, which are inherently coupled, showed any trend towards significance. We propose that APC and electrocautery are equally safe and may be employed to treat bleeding from GAVE or gastric AVM.