Baffling Cause of Upper Gastrointestinal Bleeding!

A 53-year-old man presented to the emergency department with recurrent bouts of hematemesis for 2 days. This was associated with dizziness and passage of black tarry stools for 1 day. He did not report any past history of a similar complaint. He denied any recent intake of NSAIDs or chronic alcohol abuse. He was living a healthy life with no comorbidity. On examination, he had tachycardia and pallor. His systolic blood pressure was 100 mm Hg and there was evidence of postural hypotension. There was no sign suggestive of liver disease (spider angioma, decreased body hairs, etc.). Per abdomen examination was unremarkable. His initial blood tests were as follows: hemoglobin 6 g/dL; white blood cell count 7.4 × 109/L; platelets 344 × 109/L; total bilirubin 21.3 μmol/L; SGOT 33 IU/L; SGPT 27 IU/L; serum albumin 40 g/L; and INR 1.0. The patient was transfused 3 units of packed red blood cells. After hemodynamic stabilization, an upper gastrointestinal (GI) endoscopy was done. Upper GI endoscopy revealed active spurting of blood from a vascular lesion in the gastric fundus. Endoscopically, it appeared as isolated gastric varices, so glue was injected (Fig. 1). His bleeding was controlled with glue injection. The patient was further evaluated to look for the cause of gastric varices. Color Doppler of the portal venous system was unremarkable. Computed tomographic angiography of the abdomen revealed an arterial plexus crowded in the gastric fundus (arrow in Fig. 2). It was seen to be supplied by the left inferior phrenic artery and a distal splenic artery branch (Fig. 3). The bleeding lesion which mimicked isolated gastric varices on endoscopy was actually an abnormal arterial plexus in the gastric fundus, most probably a congenital abnormality. Glue was used to plug the bleeding punctum but in view of an arterial lesion (pressure is more than a venous lesion) and a large