A Fatal Case of Occult Aortoenteric Fistula

Abstract

Aortoenteric fistulae (AEF) are a rare but serious cause of massive hematemesis. Due to the substantial mortality risk, it is crucial that prompt diagnosis and early management are initiated. Multiple diagnostic modalities are considered valuable in the evaluation of AEF, including endoscopy and computed tomography (CT). We present a case that highlights the limitations of these common diagnostic procedures. We describe a 58-year-old male who has a history of hypertension, dyslipidemia, and severe peripheral vascular disease (PVD). He previously underwent an aortobifemoral bypass, axillofemoral bypass, and recent hospital admission for removal of an infected aortic graft. He presented to the emergency room following massive hematemesis, and underwent urgent esophagogastroduodenoscopy (EGD). This revealed a large clot in D3 with no evidence of active bleeding. Given his significant history of PVD, a CT angiogram was obtained. The radiologist reported no evidence of aortoenteric fistula. The patient was transferred to a monitored setting, and started on a pantoprazole infusion. The next morning the patient had significant hematemesis of approximately 2L of blood, requiring CPR and urgent transfusion. He was transferred to the intensive care unit for repeat EGD. The vascular surgeon and radiologist reviewed the case and felt that despite normal imaging, the patient should be transferred emergently to the operating room. Intraoperatively, a large aortoenteric fistula was noted in D3, which was surgically managed. The patient unfortunately passed away the next day. This case highlights that even with a high index of suspicion, AEF can be difficult to identify despite gold standard diagnostic testing. The delay in diagnosis is often fatal as AEF carry high a mortality risk1. The risk is estimated to be greater than 70% in secondary AEF2,3. Endoscopy has a low sensitivity for diagnosis of AEF4. High resolution CT remains the imaging modality of choice5,6, although it carries a broad sensitivity (40-90%) and specificity (33-100%)7. In our case, the fistula likely thrombosed spontaneously resulting in a normal CT angiogram, and then subsequently exsanguinated the next day. We also want to emphasize the substantial variability of endoscopic management in the treatment of non-bleeding clots, especially in patients with a history of aortic surgery. Further research is warranted to illuminate a more reliable diagnostic and therapeutic model in these patients.1994_A Figure 1. Large clot identified in D31994_B Figure 2. CTA demonstrating no AEF (no contrast leaking into D3), as well as evidence of axillofemoral bypass