A Rare Cause of Bleeding: Recurrent Hemobilia Due to Ectopic Varices

Abstract

Listen

Translate article

Purpose: Esophageal varices are a common cause of upper GI bleeding second to peptic ulcer disease, whereas ectopic varices account for only 5% of all variceal bleeding. Ectopic varices comprise large portosystemic collaterals anywhere other than the gastro-esophageal region. Clinicians must be astute in suspecting an ectopic variceal bleed since the mortality rate can reach 40% if missed. We describe a case of recurrent GI bleeding over multiple years in a patient found to have portal vein thrombosis leading to choledochal varices and recurrent hemobilia. A 64-year-old female presented to the hospital with melena and a hemoglobin of 4.8 g/dL. Her history is significant for a gastric banding procedure in 1986 complicated two years later with erosion of the band leading to perforation. It was removed with creation of an omental patch and a Roux-en-Y cholecystojejunostomy for biliary obstruction found at the time of the repair. Over the years, the patient reports multiple episodes of melena coinciding with fevers, jaundice, and abdominal pain suggestive of cholangitis. Numerous EGD and colonoscopies in the past did not reveal an active source of bleeding and every episode resolved without intervention. An EGD performed during this admission was unremarkable and her colonoscopy revealed clotted blood in the colon and terminal ileum. An antegrade spiral endoscopy to assess the cholecysto-jejunal anastomosis was unsuccessful due to acute angularity. An initial MRCP showed filling defects within the common bile duct, but subsequent imaging revealed the absence of any ductal material. There was portal vein occlusion with extensive varices in the vicinity of the cholecystojejunostomy and other changes consistent with portal hypertension, thought as a consequence of her previous surgery. She was managed medically requiring a total of 11 units of packed red blood cells and had no further bleeding. She was referred to surgery given that this episode and likely prior were a function of recurrent ectopic variceal bleeding. Attempts at a splenorenal shunt were abandoned due to the extent of adhesions and she underwent a complicated procedure involving removal of the small bowel and gallbladder containing the varices, and the creation of a new Roux-en-Y hepaticojejunostomy. Now 4 months after the procedure, the patient is doing well with no further episodes of GI bleeding or cholangitis. Although 80% of upper GI bleeding is self-limited, this case highlights the need for continued evaluation particularly in the setting of recurrent episodes. It also demonstrates the difficulty in identifying an obscure source can be overcome using the clinical history as a guide.