Ann Thorac Surg 1997;64:253-255
© 1997 The Society of Thoracic Surgeons
Case Report
Repair of Aortoesophageal Fistula After Aortic Grafting
Per H. Wickstrom, MD,
John M. Streitz, Jr, MD,
Robert V. Erickson, MD,
B. D. Kion Hoffman, MD
Departments of Thoracic and Cardiovascular Surgery and Gastroenterology, Duluth Clinic, Duluth, Minnesota
Accepted for publication March 4, 1997.
 |
Abstract
|
|---|
This report describes repair of an aortoesophageal fistula caused by a previously placed thoracic aortic graft. The diagnosis was made by esophagoscopy. The repair consisted of femoral-to-femoral cardiopulmonary bypass, excision of the old graft, placement of a new graft, esophagectomy, cervical esophagostomy, gastrostomy, and later reconstruction by cervical esophagogastrostomy.
|
Introduction
|
|---|
Aortoesophageal fistula is an unusual cause of gastrointestinal bleeding. Most often it is the result of a thoracic aortic aneurysm. Less common causes include foreign body ingestion and esophageal malignancy [1]. Increasingly, aortoesophageal fistulas secondary to aortic prosthetic grafts are being reported [2–6]. Most reported attempts at surgical repair have not been successful. Recently a report of a successful repair appeared in this journal [6]. This is a second such report.
A 56-year-old Native American woman with known Marfan's syndrome underwent placement of a St. Jude composite graft in 1984 for ascending aortic dissection. In 1986, a descending dissection was repaired with a Dacron graft. Proximal and distal anastomoses sites were buttressed with Teflon-felt strips on both the outside and the inside. She received long-term warfarin anticoagulation.
On May 28, 1996, she was readmitted to the hospital because of hematemesis, which subsided after administration of fresh frozen plasma and vitamin K. Her admission hemoglobin level was 8.3 g/dL and increased to 13.4 g/dL after transfusion of 5 units of packed red blood cells. Endoscopy soon after admission demonstrated a large gastric clot, but no active gastric bleeding. On withdrawal of the endoscope, a bleeding point was noted in the esophagus at 24 cm, which was not seen when the endoscope was initially inserted. The bleeding stopped after a few moments of pressure applied to the area with the endoscope. It was thought that the bleeding may have been due to endoscopic trauma, but aortoesophageal fistula was considered as a possible cause. Computed tomography showed no false aneurysm or abscess cavity in the area of the descending aortic graft. Repeat endoscopy the next day demonstrated what appeared to be Teflon felt protruding into the left lateral lumen of the esophagus at the same location as the bleeding point seen the day before (Fig 1
).
With the diagnosis of aortoesophageal fistula, the patient was taken to the operating room. Preparations were made for femoral-to-femoral bypass, and proximal and distal aortic control was obtained before the area of the fistula was dissected. The plane between the graft and the esophagus was inflamed, edematous, and bloody, but there was no purulence. Smears were negative for bacteria, and cultures were taken. When the fistula was entered, cardiopulmonary bypass was instituted and aortic clamps were placed proximal and distal to the graft. The entire aortic graft and the Teflon strips were removed. A knitted Dacron tube graft was placed. The esophageal wall adjacent to the fistula appeared ischemic and partially necrotic. Therefore, esophagectomy, cervical esophagostomy, and gastrostomy were performed (Fig 2), with plans for later reconstruction of esophageal continuity. Postoperatively, administration of antibiotics was continued for 3 days until cultures returned as negative. In the absence of a positive culture, it was thought that long-term antibiotic therapy was empiric and not justified. The patient was discharged on the tenth postoperative day after an uneventful recovery.
View larger version (18K):
[in this window]
[in a new window]
|
Fig 2. . (A) Clot at site of aortoesophageal fistula. (B) Excision and replacement of graft, esophagectomy, cervical esophagostomy, and gastrostomy. (C) Cervical esophagogastrostomy with stomach in the right posterior intrapleural position.
|
|
She returned 2 months later for her esophageal reconstruction, which consisted of a cervical esophagogastrostomy with the stomach placed in the posterior right intrapleural position, her previous sternotomy precluding use of the substernal route. Again, she recovered uneventfully and she remains well 1 year after presentation.
|
Comment
|
|---|
Chiari described his triad of aortoesophageal fistula in 1914. It consisted of midthoracic pain, sentinel hemorrhage, and later exsanguination [1]. Most patients do, in fact, have a sentinel bleed [1], and the majority of patients initially present with bleeding that stops. The cessation of bleeding has enabled a handful of successful repairs for aortoesophageal fistula secondary to aneurysm [7–9], and we could find one other report of a surgical success when the aortoesophageal fistula has been secondary to a previous aortic graft. In our case, as with the other reports of successes, a stable period after an initial bleed allowed for establishment of the diagnosis and for the performance of a controlled operation.
Aortoenteric fistula secondary to previous aortic graft is not uncommon in the abdomen. In the chest it is rare. Therefore, the degree of suspicion may be low. The key to the diagnosis in our case was the second esophagoscopy, which revealed the intraesophageal Teflon. Others have also found endoscopy to be diagnostic [1, 6, 7]. The findings are not always as clear as in our case, and the temptation to perform a biopsy must be resisted.
The computed tomographic scan was not particularly helpful for us but has elsewhere been reported to be diagnostic, showing a hematoma, abscess, or air in the area of the fistula [4]. Aortography, which we did not employ, usually has been of benefit. It rarely demonstrates the bleed, but may show a false aneurysm at the anastomosis site [1, 9]. Esophagography was used more often in the past, sometimes with success, but carries a low sensitivity [1, 3].
The controlled nature of this operation was important to its success. Groin vessels were exposed and the aorta was encircled proximally and distally before the fistula was entered, so when bleeding occurred, it was easily controlled.
The esophagus might be dealt with in two ways. Some who have done successful repairs for aortoesophageal fistula secondary to aneurysm have repaired the esophagus primarily and interposed omentum [7] or aneurysm wall [9] between the new graft and the esophagus. Because the esophagus in our patient appeared poorly viable, we chose resection with later reconstruction, as did Luketich and associates in their recent report [6].
There are also alternatives for aortic reconstruction, including excision of the old aortic graft and in situ replacement with a new graft, as we chose, or extraanatomic bypass from ascending aorta to the supraceliac aorta with removal of the old graft and closure of the aortic stumps [2]. The latter approach might be employed in the case of an infected graft.
Aortoesophageal fistula secondary to thoracic aortic graft is an uncommon and usually fatal cause of upper gastrointestinal hemorrhage, which requires prompt surgical repair. Thoracic surgeons should be aware of the diagnosis and prepared with a treatment plan.
|
Footnotes
|
|---|
Address reprint requests to Dr Wickstrom, Department of Thoracic and Cardiovascular Surgery, Duluth Clinic, 400 E Third St, Duluth, MN 55805.
|
References
|
|---|
- Hollander JE, Quick G. Aortoesophageal fistula: a comprehensive review of the literature. Am J Med 1991;91:279–87.[Medline]
- Wong RS, Champlin A, Temes RT, Wernly JA. Aortoesophageal fistula after repair of descending aortic dissection. Ann Thorac Surg 1996;62:588–90.[Abstract/Free Full Text]
- Seymour EQ. Aortoesophageal fistula as a complication of aortic prosthetic graft. Am J Roentgenol 1978;131:160–1.[Medline]
- Tierney LM Jr, Wall SD, Jacobs RA. Aortoesophageal fistula after perigraft abscess with characteristic CT findings. J Clin Gastroenterol 1984;6:535–7.[Medline]
- Graham ANJ, McGuigan JA, Curry RC. Aortoesophageal fistula secondary to thoracic aortic aneurysm repair. J Cardiovasc Surg 1993;34:381–3.[Medline]
- Luketich JD, Sommers KE, Griffith BP, et al. Successful management of secondary aortoesophageal fistula. Ann Thorac Surg 1996;62:1852–4.[Abstract/Free Full Text]
- Coselli JS, Crawford ES. Primary aortoesophageal fistula from aortic aneurysm: successful surgical treatment by use of omental pedicle graft. J Vasc Surg 1990;12:269–77.[Medline]
- Von Oppell UO, de Groot M, Thierfelder C, Zilla P, Odell JA. Successful management of aortoesophageal fistula due to thoracic aortic aneurysm. Ann Thorac Surg 1991;52:1168–70.[Abstract]
- Bogey WM Jr, Thomas JH, Hermreck AS. Aortoesophageal fistula: report of a successfully managed case and review of the literature. J Vasc Surg 1992;16:90–5.[Medline]
This article has been cited by other articles:
|
|
|
|
 
S. Chocron, L. Stoica, S. Koch, J.-F. Bonneville, D. Kaili, B. Heyd, and J.-P. Etievent
Is endovascular stent grafting for descending thoracic aortic disease recommendable in patients with dysphagia?
J. Thorac. Cardiovasc. Surg.,
December 1, 2002;
124(6):
1239 - 1241.
[Full Text]
|
|
|
Top
Abstract
Introduction
