Aorto-Enteric Fistula Associated with Aortic Dissection
Article Outline
Abstract
Primary aorto-enteric fistula (PAEF) is usually due to erosion by atherosclerotic aneurysm. We report a case of PAEF secondary to a previously unrecognised focal abdominal aortic dissection. A 66-year-old man was brought to the casualty department with shock and acute-on-chronic back pain. Following resuscitation, he deteriorated further secondary to frank haematemesis. A prompt contrast enhanced CT scan was performed, and aortic dissection with aorto-duodenal fistula was diagnosed and operated on. To our knowledge, there are no previous reports of penetrating, transmural PAEF secondary to abdominal aortic dissection.
Keywords: Primary aorto-enteric fistula, Aortic dissection, Abdominal aortic aneurysm
Introduction
Primary aorto-enteric fistulas (PAEF) are rare, approximately 0.04--0.07% in published autopsy series.1 The majority (61--73%) are associated with atherosclerotic disease and result from direct erosion of an atherosclerotic aortic aneurysm.2 Other reported associations for PAEF include infective (tuberculosis, salmonella, staphylococcal, syphilis) and, rarely, inflammatory or neoplastic causes including cervical/colon cancer, appendicitis, diverticulitis and duodenal ulceration.3 The diagnosis of PAEF in patients presenting with gastrointestinal haemorrhage requires a high index of suspicion.4 Expeditious open and/or endovascular intervention in such cases offers the only chance of survival and delay in diagnosis is usually fatal.
Report
A 66-year-old man was brought to the casualty department by ambulance following deterioration of long-standing backache and collapse at home. On arrival, he was haemodynamically unstable. At initial examination a leaking abdominal aortic aneurysm was suspected. After resuscitation a contrast enhanced computed tomography (CT) scan of the abdomen was requested.
During CT scanning, the patient suffered several bouts of haematemesis and became tachycardic and increasingly hypotensive; his haemoglobin had dropped to 6.8
g/dL. CT scan showed large quantities of blood in the duodenum and stomach. A 4-cm infrarenal aortic aneurysm containing mural dissection with extravasation of contrast into the third part of the duodenum (D3) was observed (Fig. 1). A diagnosis of PAEF was made and the patient was immediately transferred to the operating theatre.
Figure 1
Computed tomography illustrating false lumen of aortic dissection and extravasation of intravenous contrast at site of enteric fistula.
After control of the fistula, a large inflammatory mass was identified between the infrarenal aorta and D3. The aortoduodenal fistula was identified and excluded by means of a rifampicin coated tube Dacron graft. The duodenum was not amenable to primary repair due to the severe inflammatory reaction. Duodenal exclusion was achieved by means of an omental plug and omental wrapping of the graft within the aortic sac. A gastrojejunostomy was formed as a diversion procedure to minimise contamination and further sepsis at the site. The patient recovered completely within a week and was discharged from the hospital following interval gastrograffin meal and CT assessment which demonstrated a functional bypass and no evidence of collection at the fistula site. At 2months follow-up, the patient remained well and reported almost complete resolution of his back pain.
Discussion
Evidence suggests that PAEF commonly arises secondary to erosion by atherosclerotic aneurysm. We believe our case to have arisen from a long-standing aortic dissection. The evidence accrued during the management of this case suggests that chronic aortic dissection may have caused a pseudoaneurysm leading to duodenal fistulation. Our case is unusual in that there are no reported cases of penetrating, transmural, PAEF arising from chronic abdominal aortic dissection. A previous paper has reported a non-penetrating, intramural PAEF associated with known chronic dissection, presenting with abdominal pain alone without haematemesis.7 Our hypothesis is supported by the patient's long history of chronic back pain and the CT scan demonstrating a previously unknown aortic dissection with communication to the duodenum; it is further affirmed by our findings at operation and resolution of the patient's back pain following surgery. Recently, studies have suggested that benign aortic dissection may be significantly more prevalent than first thought. The management of benign aortic dissection is controversial. Some have advocated the routine use of stent grafts.5 Others question the overall benefit of such intervention.
Our patient had no history of an abdominal aortic aneurysm, nor was one palpable at initial examination. Suspicion of an aorto-enteric fistula is based on gastrointestinal bleeding with a pulsatile mass or history of an aneurysm. The classical triad of abdominal pain, gastrointestinal bleeding, and a pulsatile mass is, however, highly variable. In one study, only 11% of 81 patients presented with all three symptoms, and 94% of 81 patients presented with signs of gastrointestinal bleeding only.3 Initial investigation for gastrointestinal bleeding would be gastroduodenoscopy. Endoscopy has a PAEF detection rate of 25%, compared to CT detection rate of 61%.1
In repairing PAEF, there is a predominance of graft repairs in comparison to primary closure or extra-anatomic bypass.6 Considering the patient's young age and absence of cardiovascular disease, the risks of extra-anatomical bypass, which include limb amputation and aortic stump blowout,6 were deemed too great. The inflammatory mass found at operation made dissection and mobilisation of the duodenum extremely difficult. Omental plug of the fistula and wrapping of the graft provided protection from infection and erosion of the graft into adjacent structures. Further protection of the site was provided by diversion gastrojejunostomy.
In conclusion, we report a rare and previously unreported mechanism of PAEF formation and illustrate how prompt and appropriate surgical intervention resulted in a positive outcome for our patient. Our case highlights the potential complications from long-standing untreated aortic dissection and supports the view that, where identified, such patients should be considered for pre-emptive intervention.
Conflict of Interest/Funding
None declared.
References
- . An anatomical approach to the problem of massive gastrointestinal haemorrhage. Arch Surg. 1955;70:570
- . Aortoduodenal fistula revisited. Scand J Gastroenterol. 1989;24(suppl 167):97–100
- . Primary aortoenteric fistulae. Br J Surg. 2005;92:143–152
- . The many faces of aortoduodenal fistula. Am Surg. 1996;62:344–349
- Self-expandable aortic stent-grafts for treatment of descending aortic dissections. Ann Thorac Surg. 2002;73(4):1138–1142
- . Recurrent aortoenteric fistula: case report and review. Curr Surg. 2005;62:638–643
- . Primary aortoenteric fistula with a chronic isolated abdominal aortic dissection: report of a case. Surg Today. 1998;28(4):441–445
PII: S1533-3167(09)00007-7
doi:10.1016/j.ejvsextra.2009.02.002
© 2009 European Society for Vascular Surgery. Published by Elsevier Inc. All rights reserved.
Refers to article:
- Aorto-Enteric Fistula Associated with Aortic Dissection , 13 April 2009
