Secondary Aortoenteric Fistula

M. A. Mohammadzade, MD; M. Hossain Akbar, MD



Secondary aortoenteric fistula (SAF) is an uncommon but very important complication of abdominal aortic reconstruction. The complication often occurs months to years after aortic surgery. The clinical manifestation of the aortoenteric fistula is always upper gastrointestinal bleeding. Treatment of the disease is early surgical intervention. If operative treatment is not performed promptly, the mortality is high. A case of secondary aortoduodenal fistula found 6 years after aortic reconstructive surgery, with the clinical presentation of upper gastrointestinal bleeding, is presented. On Immediate exploratory laparotomy, the proximal part of abdominal aorta was clamped. Duodenorrhaphy and aortic reconstruction with patch graft at the proximal suture line of aortic prosthesis was performed. Fortunately there was no pus, so tissue culture was not done. The intervention was concluded with an omentoplasty to protect the patch graft and to separate it from duodenorrhaphy.

The patient did well after the surgical management. Because of the increasing number of elective aortic aneurysm repairs in the aging population, it is likely that more patients with secondary aortoenteric fistula will present to the clinical physicians in the future. So, a high index of suspicion is necessary for prompt diagnosis and treatment of this life-threatening event.


Sudden hematemesis is a life-threatening emergency that directs a physician's attention toward various causes of gastrointestinal bleeding. Aortoenteric fistula is an uncommon but life-threatening complication of aortic reconstructive surgery.[1] Communications between the aorta and the intestine resulting from disease at either site are referred to as aortoenteric fistulas. Fistula formation between the aorta and the intestinal tract was first described in 1839 in reference to a man with a "pulsating tumor . . . and a discharge of bloody stool," who died suddenly. At autopsy, it was noted that "the jejunum had adhered to . . . the aneurismal bag and that sac had ulcerated into the intestine."[2] Fistulas occurring after aortic reconstructive surgery, also called aortic graft-enteric fistulas, are considered secondary aortoenteric fistulas. Before 1960, the most common cause of abdominal aortoenteric fistulas was aortic aneurysm, followed by infectious aortitis due to syphilis or tuberculosis.[3] However, over the past 3 decades or so, erosion of the intestine by prosthetic vascular grafts has become a much more common cause, with an incidence of up to 4%.[4]

The complication often occurs months to years after the original surgery.

Bastounis and colleagues reported that the mean interval from the initial operation to the onset of upper gastrointestinal bleeding was 32 months.[5] The 20-year experience with secondary aortoenteric fistula at the Johns Hopkins Medical institution showed the average to be 2.8 years.[6]

The first reported secondary aortoenteric fistula was reported by Brock in a case involving an aortic homograft and the duodenum.[7] In 1956, Clayton and colleagues[8] presented the first aortoenteric fistula caused by a prosthetic graft of the aorta. In 1958, Mackenzie and colleagues[9] demonstrated the first successful repair of a secondary aortoenteric fistula between a synthetic graft and the intestine. Due to the anatomic proximity, the majority of cases involve the duodenum, with the proximal suture line of an aortic prosthesis. Prompt diagnosis with surgical intervention is the only possible treatment that preserves the patient s life. Because of the nonspecific nature of the clinical history and physical findings, diagnosis of aortoenteric fistula is difficult to make preoperatively. There is no single diagnostic investigation that has a very high specificity and sensitivity, including upper computed tomography (CT), angiography, or gallium-67 CT. Gastrointestinal endoscopy is the most helpful method for diagnosis. If findings are negative, this test is meaningless unless another source of bleeding is found. Nevertheless, exploratory laparotomy is the only method that can definitely confirm the diagnosis.

Case Report

We present a case of secondary aortoduodenal fistula found 6 years after aortic reconstructive surgery, with the clinical presentation of upper gastrointestinal bleeding. The patient was a 70-year-old man who complained of hematemesis and melena. He gave a history of aortic reconstructive surgery in Tehran 6 years ago. There was no history of peptic ulcer disease or any other gastrointestinal pathology. On physical examination, the patient appeared pale with a cold clammy skin in a pre-shock condition. His vital signs were recorded as pulse rate 112 beats per minute regular, respiratory rate 22 breaths per minute, and blood pressure 90/60 mm Hg. Chest wall, heart, and lungs were normal on physical examination. There was a median linear scar on his abdomen showing previous abdominal surgery. Epigastrium was tender on palpation. None of the abdominal viscera was palpable. His past surgical records revealed aortobifemoral graft 6 years ago. CT showed an aneurysmal mass around the graft. Ultrasound studies depicted blood clots in the distal half of duodenum.

On immediate exploratory laparotomy, we found hematomas around the duodenum and pancreas adhered to the omentum. The proximal part of the abdominal aorta was clamped. Blood clots were removed and duodenum was separated from aorta. Fortunately there was no pus, so tissue culture was not done. We decided to perform a patch synthetic graft revascularization on the aortic side with the proximal suture line of aortic prosthesis. The intervention was concluded with an omentoplasty to protect the patch graft and to separate it from duodenorrhaphy.

During the postoperative period, the patient did not experience relevant complications. Specific antibiotic therapy was administered (ceftriaxone along with metronidazole and vancomycin). Control CT was carried out after 1 month and after 6 months. The repeated clinical and laboratory examination did not reveal any sign of infection. Primary digestive tract radiography did not show any sign of duodenal stenosis.


The diagnosis and the treatment of aorto enteric fistula are difficult and represent a big problem for a vascular surgeon.[10] Nevertheless in a patient with hematemesis and melena who underwent an aortobifemoral bypass or aortic interposition grafting without esophagogastroduodenal pathologies, a diagnosis of aortoenteric fistula should not be overlooked.[11] In the present clinical case the available clinical, instrumental, and radiological supports made the hypothesis of such a diagnosis very much presumable.

The esophagogastroduodenoscopy showed no pathologies except for a clot in the second part of duodenum. These signs, associated with high gastroesophageal bleeding and the history of aorto bifemoral bypass grafting 6 years previously lead to the diagnosis of aortoenteric fistula.

The longest postoperative interval for an aortoenteric fistula was 23 years after aortofemoral bypass surgery; the shortest postoperative interval was 2 days, recorded in 1974, in which a para-prosthetic enteric fistula developed after resection of a ruptured abdominal aortic aneurysm with graft interposition.[12] In our case, the complication presented 6 years after aortic aneurysm reconstruction.

Both in situ and extra-anatomic bypass grafting have been described in the literature.[13,14] The treatment of choice is aortic ligature and axillofemoral bypass. It was reported that once the fistula identified, the surgical procedures most commonly used are graft excision, oversewing of the aortic stump, repair of the gut defect, and placing a new graft in situ or use of extra-anatomic bypass. The mortality rate during surgery and in the postoperative period is relatively high, averaging approximately 50% to 60 %.[13,15]

Chang and colleagues[15] from Taiwan reported a similar case. A secondary aortoenteric fistula developed in an 80-year-old patient as an immediate postoperative complication after aortic reconstruction surgery; the patient died on the 20th day after primary surgery. This patient did not survive probably due to massive blood loss, very old age, and infection. Our patient is younger and presented after 6 years with melena and hematemesis, which was diagnosed and managed promptly, enabling patient survival.

Generally 2 types of secondary aortoenteric fistula have been described. Type 1, termed a true aortoenteric fistula or graft enteric fistula, with or without a pseudoaneurysm, develops between the proximal aortic suture line and the bowel. This type of fistula is the most common and often initiates massive gastrointestinal hemorrhage. The main clinical manifestation of this type is always upper gastrointestinal bleeding (76%), which might be either hematemesis or melena with equal frequency. Sepsis and abdominal pain are relatively rare with this type of fistula. The present case appearing 6 years after aortic surgery was of this type.

Type 2, or a para-prosthetic enteric fistula, develops no communication between the bowel and the graft. It accounts for 15% to 20% of secondary aortoenteric fistulae. In this type of fistula, bleeding occurs from the edges of the eroded bowel by mechanical pulsations of the aortic graft. Sepsis is more frequently associated with this type of fistula (75%). In addition to sepsis, gastrointestinal hemorrhage (30%), abdominal pain (20%), septic emboli in the lower extremities, septic arthritis, multicentric osteomyelitis, and hypertrophic osteoarthropathy have been described.[13,15]


The aim of this case report is to emphasize early diagnosis and management of all gastrointestinal bleeding in patients who have a history of aortic reconstructive surgery. Possibility of aortoenteric fistula should be considered in such cases. In selected cases, aortic reconstruction with patch graft, duodenorrhaphy, and omentoplasty can represent a valid alternation and easy choice for aortoenteric fistula without any complication.


Readers are encouraged to respond to the author at or to Paul Blumenthal, MD, Deputy Editor of MedGenMed, for the editor's eyes only or for possible publication via email:


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.