An Unusual Cause of Esophageal Varices

Abstract

Question: A 55-year-old man with a past medical history significant for obesity, hypertension, obstructive sleep apnea, and pacemaker placement for syncope and bradycardia at the age of 14 presented to our emergency department with complaints of sudden onset of 8/10 intensity, nonradiating, retrosternal chest pain of burning character that was not related to exertion or food intake. He had no nausea, vomiting, hematemesis, melena, or abdominal pain. Pertinent findings on physical examination were a pacemaker in the right upper chest, no jugular venous distention, and an enlarged palpable liver. His blood work was significant for hemoglobin of 11.2 g/dL with an unknown baseline. Liver chemistries revealed a total bilirubin of 0.4 mg/dL, aspartate aminotransferase 21 U/L, alanine aminotransferase 24 U/L, and alkaline phosphatase of 90 U/L. The pain partially resolved after a dose of pantoprazole. Computed tomography angiography of the chest was negative for pulmonary embolism, but did reveal extensive venous collaterals in the thorax and upper abdomen with hepatomegaly. There was nonopacification of the proximal portion of the superior vena cava, which contained leads from a right-sided chest wall pacer (Figure A). The patient was admitted for a esophagogastroduodenoscopy that demonstrated nonbleeding varices throughout the esophagus; the varices being larger in the proximal part of the esophagus than the distal (Figure B). This finding prompted a serological workup for possible cirrhosis which was negative. A liver biopsy was performed revealing mild focal bridging fibrosis with no definitive evidence of cirrhosis. There was mild dilatation of portal vein branches and central veins. There were large and small droplet steatosis involving <10% of hepatocytes (Figure C). What is the cause of patient's esophageal varices? Look on page 613 for the answer and see the Gastroenterology website (www.gastrojournal.org) for more information on submitting your favorite image to Clinical Challenges and images in GI. Computed tomography angiography of chest revealed severe superior vena caval (SVC) narrowing at the SVC–right atrium junction with extensive venous collaterals in the thorax and upper abdomen. The patient had a pacemaker in place for 40 years which created chronic narrowing of the SVC lumen, essentially causing SVC obstruction (Figure A). Liver biopsy confirmed that there were changes consistent with fatty liver disease as well as chronic ischemic changes in the liver from increased pressure owing to longstanding SVC narrowing as evidenced by central vein changes in zone 3. However, the downhill esophageal varices (DEV) were not ligated during the esophagogastroduodenoscopy because this patient had no symptoms of bleeding; neither did cardiothoracic surgery intervene surgically because he had no symptoms of SVC syndrome. DEV, a less appreciated entity, is a manifestation of venous collaterals that arise owing to SVC obstruction. They are present in up to one-third of the cases of SVC obstruction; however, only 8% of such patients manifest with upper gastrointestinal bleeding.1Siegel Y. Schallert E. Kuker R. Downhill esophageal varices: a prevalent complication of superior vena cava obstruction from benign and malignant causes.J Comput Assist Tomogr. 2015; 39: 149-152Crossref PubMed Scopus (12) Google Scholar, 2Hirose J. Takashima T. Susuki M. et al.“Downhill” esophageal varices demonstrated by dynamic computed tomography.J Comput Assist Tomogr. 1984; 8: 1007-1009Crossref PubMed Scopus (12) Google Scholar As the number of pacemakers and implantable cardioverter defibrillators being placed is increasing, their use is not without adverse events. Of patients with pacemaker implantation, 1%–3% develop SVC obstruction and 30% of patients with SVC obstruction develop DEV.3Ayvaz M.A. Rakici H. Allescher H.D. Are downhill varices an overlooked entity of upper gastrointestinal bleedings?.Gastroenterol Res Pract. 2018; 2018: 7638496Crossref PubMed Scopus (2) Google Scholar It is estimated that around 0.1% of all variceal bleeds are from DEVs.2Hirose J. Takashima T. Susuki M. et al.“Downhill” esophageal varices demonstrated by dynamic computed tomography.J Comput Assist Tomogr. 1984; 8: 1007-1009Crossref PubMed Scopus (12) Google Scholar Banding of downhill varices is typically not performed unless there is active gastrointestinal bleeding, in which case banding or sclerotherapy should be used only to temporize variceal bleeding. Bands would be placed on the proximal end of the varix. Ultimately, the management on downhill varices is managing the underlying cause of the SVC obstruction, which includes angioplasty with stent placement or the surgical management of SVC obstruction.2Hirose J. Takashima T. Susuki M. et al.“Downhill” esophageal varices demonstrated by dynamic computed tomography.J Comput Assist Tomogr. 1984; 8: 1007-1009Crossref PubMed Scopus (12) Google Scholar, 3Ayvaz M.A. Rakici H. Allescher H.D. Are downhill varices an overlooked entity of upper gastrointestinal bleedings?.Gastroenterol Res Pract. 2018; 2018: 7638496Crossref PubMed Scopus (2) Google Scholar Clinicians should be aware of alternative etiologies of esophageal varices, as the management of DEV is a multidisciplinary approach to remedy the underlying cause of SVC obstruction.