Aortoenteric Fistula Thirty-Seven Years After Repair for Midaortic Syndrome

Aortoenteric fistulae are rare but associated with a high degree of mortality. We present the case of an adult who had undergone aortic surgery for midaortic syndrome as a child. A 2-stage approach was used to treat his aortoenteric fistula and to reconstruct his aorta. However, the patient underwent multiple operations including endovascular and open surgical repair. He was placed on long-term antibiotics. On last follow-up, the patient had returned to his functional baseline. A clinical suspicion should be maintained when these patients present with symptoms. Appropriate imaging and diligent operative planning are paramount to ensuring good outcomes.

Aortoenteric fistula (AEF) is a rare cause of significant gastrointestinal bleeding and associated with a high degree of mortality. This type of fistula can occur either primarily or secondarily, with secondary causes being more common. While primary causes arise via the native aorta, secondary causes are usually from previous surgical interventions on the aorta when synthetic graft material erodes into the enteric system.¹ Herein we report the unusual case of a patient who developed an AEF 37 years after undergoing surgery for treating his midaortic syndrome.

Report

A 45-year-old man presented to the emergency department after experiencing a syncopal event at home after he passed a large amount of bright red blood per rectum. He was found to have a 40-point drop in his haemoglobin. His medical profile was significant for hypothyroidism and repair of midaortic syndrome at the age of 8 years. Information regarding his congenital abnormality was limited, but chart review was consistent with repair of midaortic syndrome of the descending thoracic aorta, where the ascending aorta was anastomosed to the immediate infrarenal abdominal aorta with an interposition graft. At the time of presentation, a computed tomography (CT) scan revealed a large outpouching of the graft approximately 3 cm proximal to the abdominal aortic anastomosis and strong suspicion of an AEF at that location (Video 1 , view video online). Emergently, the patient underwent an uncomplicated thoracic endovascular aortic repair (TEVAR) (Video 2 , view video online). On postoperative day 2 (POD-2), push enteroscopy was performed, revealing mucosa defect in keeping with AEF at the proximal jejunum. White blood cell scan and blood culture results were negative for infection, whereas transthoracic echocardiography did not show any evidence of infective endocarditis. However, the original interposition graft and TEVAR were deemed infected, so he was placed on 6 weeks of intravenous antibiotics with a 2-stage plan to reconstruct his aorta.

Stage 1 would entail aortic reconstruction from zone 3 to his infrarenal aorta via a retroperitoneal approach, and stage 2 would involve sternotomy and laparotomy to explant his original interposition graft. Because of the complexity of this case, a definitive timeline was not outlined. On 5-month follow-up, the results of blood work and a CT scan were reassuring. He was subsequently advised to undergo the above 2-stage repair on a semiurgent basis. However, because his clinical status had stabilized, the patient elected to defer surgery to attend to personal and professional commitments.

Unfortunately, while awaiting his first surgery, the patient presented to the emergency department 6 months after his TEVAR with a 2-day history of epigastric discomfort and fevers. A CT scan showed findings in keeping with a perigraft abscess (Video 3 , view video online). He was admitted to the hospital, and a new surgical plan was devised. Stage 1 would now focus on source control: explant of his infected aortic graft and creation of a right axillary-to-femoral bypass. Stage 2 would be definitive surgical repair, where a thoracoabdominal aortic reconstruction would be completed in addition to the explant of his pre-existing axillary-to-femoral bypass from stage 1. The patient was taken to the operating room where an intraoperative transesophageal echocardiogram revealed invagination of graft material in the right atrium, a partially occluded proximal descending aorta with an appearance consistent with thrombus or atheroma, and a completely occluded mid-to-distal descending thoracic aorta (Video 4 , view video online).

He underwent redo-sternotomy and laparotomy, pericardiectomy, creation of bilateral axillofemoral bypasses, ascending aortic replacement with rifampin-soaked Dacron, explant of infected aortic graft, and bovine pericardial patch repair of his infrarenal aorta (Fig. 1). The case required a jejunal resection due to the AEF and its severe adherence to the graft material. The postoperative course was complicated by respiratory failure from pulmonary edema and hospital acquired pneumonia, pancreatitis, acute kidney failure, and delirium. Nevertheless, he was discharged home on POD-14 with a plan to readmit him within 3 months to complete his second-stage aortic reconstruction.

Figure 1.

Different views of the infected graft. (A) Graft exposed in situ. (B) Graft partially removed. (C, D) Graft fully removed.

Ten weeks later, the patient was taken to the operating room and underwent a sixth interspace thoracoabdominal incision with retroperitoneal exposure of the infrarenal aorta. Given the presence of his bilateral axillofemoral bypasses, left heart bypass was not required (Fig. 2). An 18-mm graft was placed with and end-to-side anastomosis to his proximal thoracic aorta, preserving flow to his T4 and T5 intercostal arteries. Distal anastomosis was to the distal infrarenal aorta via the retroperitoneal approach. He then underwent redo cutdowns of bilateral axillary and femoral arteries for removal of his axillofemoral bypass grafts. The patient was discharged home on POD-12 on intravenous antibiotics. On 3-month follow-up, he had returned to his functional baseline. He was transitioned to oral antibiotics and underwent a positron emission tomography CT scan at 6 months to assess for graft infection. The result was negative, so his oral antibiotics were stopped.

Figure 2.

The in situ axillofemoral bypass graft.

Discussion

AEF are rare but associated with a high degree of morbidity and mortality. Although there are cases reporting the development of secondary AEF postendovascular aortic repair,² they are usually the sequelae of open surgical repair of abdominal aortic aneurysm.³ In patients with infection or secondary AEF, endovascular repair represents a suitable bridging therapy to definitive open surgery in unstable patients.⁴ We present the case of a patient who presented with an AEF 37 years after repair of his midaortic syndrome. Unfortunately, despite thorough chart review, information on his original congenital abnormality was limited. The available operative note from his original operation did indeed indicate that the patient had midaortic syndrome of the descending thoracic aorta. Although not considered as standard management, he had an interposition graft placed at the time of this initial repair.

Nevertheless, the patient was stabilized on presenting to the hospital as an adult, and a TEVAR was deployed to bridge his management. Although he was originally scheduled to undergo a 2-staged plan to manage his AEF, his clinical status deteriorated, necessitating a more complex treatment. The original 2-stage operation entailed removal of the in situ infected graft and repair with an extra-anatomic bypass as the first stage. This approach would thus have facilitated definitive aortic reconstruction as the second stage, which would have been completed in the retroperitoneal plane and mitigated the risks of secondary reinfection. Furthermore, his entire management notably circumvented compromising spinal cord blood supply because there was no manipulation of the vertebral arteries. Finally, because the patient had in situ axillofemoral bypasses, left heart bypass was not needed during his definitive operation.

Novel Teaching Points.

• •The initial management stage required conversion to bilateral axillofemoral bypasses to ensure adequate perfusion of the lower extremities.
• •There was a reduced risk of spinal cord ischemia because the patient’s vertebral circulation was not manipulated.
• •Left heart bypass or hypothermic circulatory arrest was not needed given the in situ bilateral axillofemoral bypasses.

Supplementary Material

Video 1

Computed tomography scan showing an aortoenteric fistula.

Video 2

Emergent thoracic endovascular aortic repair to address the aortoenteric fistula.

Video 3

Computed tomography scan showing a perigraft abscess.

Video 4

Intraoperative echocardiogram showing invagination of graft in the right atrium, a partially occluded proximal descending aorta, and an occluded mid-to-distal descending thoracic aorta.