Background Gastric antral vascular ectasia (GAVE) is characterized by a pathognomonic endoscopic pattern, mainly represented by red spots either organized in stripes radially departing from pylorus, defined as “watermelon stomach”, or arranged in a diffused-way, the so called “honeycomb stomach”. In our previous work we found a prevalence of GAVE in 35% of systemic sclerosis (SSc) patients that had an upper gastrointestinal (GI) endoscopy for evaluation of symptoms. The main complication of GAVE is symptomatic anemia due to GI bleeding requiring blood transfusion. As the pathogenesis is unknown, the aim of the therapy is to decrease GI bleeding. Endoscopic therapy (endotherapy) is the mainstay of treatment with surgery reserved for severe refractory cases. The data on the efficacy of endotherapy and treatment requirements of SSc patients with GAVE is very limited.

Objectives The aim of our study is to assess the treatment requirements of SSc patients with GAVE and the efficacy of endotherapy in particular.

Methods Retrospective analysis of a prospectively maintained database of SSc patients followed in a tertiary referral center for SSc. Our cohort is part of the EUSTAR cohort. Statistical analysis: descriptive, student's T test, Mann-Whitney test.

Results Fifty SSc patients were diagnosed with GAVE. Seventeen patients (34%) suffered of symptomatic anemia and were treated with argon plasma coagulation (APC) for a mean (SD) of 4.5 (4.3) sessions per patient. The mean (SD) hemoglobin was 9.4 (1.7) gr% in the group who required endotherapy compared to 11.5 (1.3) in the group which did not need any intervention. Four patients needed more than 6 sessions of treatment. Mean (SD) follow up was 4.1 (2.4) years. Nine patients required a total of 73 packed blood cells units and another 4 had intravenous iron replacement. Four patients died during the first 2 years after diagnosis, none of them due to bleeding complications. None of the patients that survived more than 2 years required further endotherapy or blood transfusions. No significant complications of APC occurred. Interestingly, all those who survived,had received immunotherapy with cyclophosphamide or mycophenolate mofetil due to multisystem involvement and disease activity.

Conclusions Long term survivors among SSc patients with GAVE had an excellent response to APC and did not require blood transfusions or endotherapy after 2 years follow up. All the long term survivors received immunotherapy. This observation might hint to the pathogenesis of GAVE in SSc.

Disclosure of Interest None declared