Idiopathic Gastric Varices; a Diagnostic Dilemma

Abstract

Purpose: Introduction: Upper GI bleeding (UGIB) related to gastric varices (GV) is most commonly seen in patients with portal hypertension (PH) or splenic vein thrombosis (SVT). We present a case of a patient who had an isolated GV requiring urgent operative ligation to control massive UGIB. Case Report: A 52-year old man with a history of abdominal large B-cell lymphoma in remission after chemotherapy presented to the hospital with melena and a hemoglobin (Hb) of 9 g/dL. Upper endoscopy showed large GV with punctate areas consistent with recent bleeding in the gastric fundus and cardia. There was no active bleeding noted during the examination. The splenic vein (SV) was poorly visualized on a liver venogram but was demonstrated to be patent on a subsequent splanchnic angiogram, which also showed a normal appearing portal vein system and normal sinusoid veins pressures. A CT scan of the abdomen did not show any venous compression or recurrence of lymphoma. A liver biopsy showed mild steatosis with no evidence of cirrhosis. The patient was released from hospital with a stable Hb of 10 g/dL, but came back three days later with hematemesis and hypotension, with a Hb of 8 g/dL. The patient was stabilized with blood transfusions and intravenous fluids. Because of the life-threatening nature of the recurrent bleeding episode, and because the etiology of the varices remained unknown, the patient underwent a laparotomy. On direct visualization, no mass or lymphadenopathy was appreciated. The splenic vein was patent, and the appearance of the liver was normal. The GV were ligated from both sides of stomach wall without complications. The patient was discharged from the hospital with no further bleeding and a Hb of 10 g/dL. The patient was doing fine on 3 month follow up. Discussion: Portal hypertension, associated with GV in approximately 30% of cases, was not documented in this patient with bleeding varices. GV bleed less frequently than esophageal varices, but more profusely. It is postulated that in this case, the varices were the result of a congenital vascular abnormality in the stomach. Different therapeutic modalities, such as balloon tamponade, endoscopic sclerosis therapy with N-butyl-2-cyanoacrylate, or angiography with embolization have been used to treat actively bleeding varices; in this case both episodes of acute bleeding stopped without direct intervention, allowing us the opportunity to intervene surgically to resect the varices. Although the experience with this patient is not sufficient to state that surgical resection of isolated GV occurring in the absence of portal hypertension is the treatment of choice, it does demonstrate a potential approach in selected patients.