Treatment of Chronic Bleeding From Gastric Antral Vascular Ectasia (GAVE) With Estrogen-Progesterone in Cirrhotic Patients: An Open Pilot Study Albert Tran, M.D., Jean-Pierre Villeneuve, M.D., Marc Bilodeau, M.D., Bernard Willems, M.D., Denis Marleau, M.D., Daphna Fenyves, M.D., Roch Parent, M.D., and Gilles Pomier-Layrargues, M.D. Liver Unit and Internal Medicine Unit, Saint-Luc Hospital, and University of Montreal, Montreal, Canada OBJECTIVE: Gastric antral vascular ectasia (GAVE) is a rare cause of chronic bleeding in cirrhotic patients. Treatment of GAVE with surgical or nonsurgical portal decompression, b-blockers, or endoscopic therapy provides disappointing results. In the present study, we evaluated the efficacy of estrogen-progesterone therapy, which has been reported to control chronic bleeding in gastrointestinal vascular malfor- mations, such as Osler-Weber Rendu disease or angiodys- plasia, in GAVE-related chronic bleeding. METHODS: Six cirrhotic patients who bled chronically from GAVE were included. Three had alcoholic cirrhosis, two cryptogenic cirrhosis, and one primary biliary cirrhosis. Grade 1 esophageal varices were noted in four patients. Bleeding could not be controlled by b-blockers, and endo- scopic therapy was not considered given the extension of the antral vascular lesions. RESULTS: Before the start of therapy, transfusion require- ments averaged 3.5 units/month over a 1.5–11 month period of observation. Patients were then treated with a combina- tion of ethynil estradiol 30 mg and noretisterone 1.5 mg daily. During follow-up (range 3–12 months), bleeding did not recur in four patients; in one patient, treatment with estrogen progesterone decreased the need for transfusions from 4 units/month to 1.4 unit/month; this patient stopped the treatment inadvertently after 6 months and severe ane- mia recurred with a need for 4 units of blood in the follow- ing month; reintroduction of the treatment resulted in an increase of hemoglobin levels without the need for blood transfusions during the following 4 months. In the last patient, a 5-month treatment did not improve chronic bleed- ing. CONCLUSIONS: The present study suggests that estrogen- progesterone therapy is useful in the treatment of chronic bleeding related to GAVE; however, these findings require confirmation by a controlled trial. (Am J Gastroenterol 1999;94:2909 –2911. © 1999 by Am. Coll. of Gastroenter- ology) INTRODUCTION Gastric antral vascular ectasia (GAVE) can occur in cir- rhotic patients. Endoscopically, lesions are found almost exclusively in the antrum and red spots are either aggregated in linear stripes (the “watermelon” stomach) or diffusely spread (honeycomb pattern) (1). Histologically, vascular ectasia are present in the mucosa and are associated with fibrin thrombi, fibrohyalinosis, and spindle cell proliferation (2). GAVE is recognized as an uncommon cause of acute or more often chronic recurrent gastrointestinal bleeding. These patients need multiple transfusions during follow-up. Several therapeutic approaches have been tried to prevent continuous bleeding. In a previous study, we have shown that decreasing portal hypertension, either pharmacologi- cally (using b-blockers) or by surgical or nonsurgical por- tacaval shunt was not effective to control blood losses (3). Antrectomy is the most definitive treatment but operative morbidity and mortality are high in cirrhotic patients (3). Endoscopic coagulation has been tried to treat this condition with mixed results probably due to the diffuse nature of the vascular lesions (4 –7). In a controlled study, Van Cutsen et al. (8) reported that a combination of estrogen and proges- terone can control chronic bleeding from a variety of gas- trointestinal vascular malformations. Subsequently, case re- ports have suggested that this treatment can stop or reduce bleeding from GAVE (6, 9, 10). In the present open pilot study, we evaluated the efficacy of estrogen/progesterone on GAVE-related recurrent bleed- ing in six cirrhotic patients judged unfit for surgical antrec- tomy or endoscopic therapy. MATERIALS AND METHODS Between July 1996 and July 1998, seven patients were enrolled in the present trial. Inclusion criteria were as fol- lows: 1) biopsy-proven cirrhosis, 2) GAVE as shown by typical endoscopic and histological findings as described THE AMERICAN JOURNAL OF GASTROENTEROLOGY Vol. 94, No. 10, 1999 © 1999 by Am. Coll. of Gastroenterology ISSN 0002-9270/99/$20.00 Published by Elsevier Science Inc. PII S0002-9270(99)00492-X