Ann Thorac Surg 2008;85:1449-1451. doi:10.1016/j.athoracsur.2007.10.068
© 2008 The Society of Thoracic Surgeons
Case Reports
Successful Management of Esophagoparaprosthetic Fistula After Aortic Surgery
Shunsuke Kawamoto, MD, PhDa,*,
Yoshikatsu Saiki, MD, PhDa,
Katsuhiko Oda, MD, PhDa,
Yoshio Nitta, MD, PhDa,
Jun-etsu Akasaka, MD, PhDa,
Shukichi Miyazaki, MD, PhDb,
Koichi Tabayashi, MD, PhDa
a Department of Cardiovascular Surgery, Graduate School of Medicine, Tohoku University, Sendai, Japan
b Division of Advanced Surgical Science and Technology, Graduate School of Medicine, Tohoku University, Sendai, Japan
Accepted for publication October 17, 2007.
* Address correspondence to Dr Kawamoto, Department of Cardiovascular Surgery, Graduate School of Medicine, Tohoku University, 1-1 Seiryo-machi, Aoba-ku, Sendai, 980-8574, Japan (Email: shunsuke{at}mail.tains.tohoku.ac.jp).
 |
Abstract
|
|---|
Aortoesophageal fistula is a relatively rare but highly fatal condition, especially in the case of secondary aortoesophageal fistulas after previous thoracic aortic surgery in which the aortic prosthetic graft itself may be involved in the infection, resulting in an esophagoparaprosthetic fistula. In this report, we describe a complicated case of esophagoparaprosthetic fistula arising after descending thoracic aortic replacement and endovascular pseudoaneurysm repair that was successfully treated by surgical resection and in situ aortic graft replacement using a homograft completely covered with an omental flap, combined with subtotal esophagectomy and staged reconstruction of the alimentary tract. The patient has been doing well for 24 months without signs of recurrent infection.
|
Introduction
|
|---|
Esophagoparaprosthetic fistula (EPPF) after aortic surgery is a relatively rare complication, but one that carries a high mortality, and only a few cases of patient survival after EPPF have been reported to date, including one from our institution [1, 2]. The case presented here represents an example of the complex clinical scenarios faced by surgeons today in which multiple modalities, both surgical and nonsurgical, have been previously used to treat the patient prior to surgical referral.
A 68-year-old woman was referred to our hospital for persistent low-grade fever. She had undergone prior emergency surgical replacement of her descending aorta for ruptured aneurysm at another institution, and had been doing well for 6 years until development of a pseudoaneurysm at the proximal anastomotic site accompanied by hemoptysis. To control the hemoptysis, a 180-mm long endovascular stent graft was inserted into the entire descending thoracic aorta covering both the previous proximal and distal anastomoses at the other institution. The bronchial arteries arising directly from the aortic arch were also coil embolized (Fig 1A). The endovascular procedure effectively relieved her hemoptysis. However, an intermittent low-grade fever developed, which had been managed with empirical antibiotic therapy for greater than 1 year prior to referral for surgical consultation.
View larger version (110K):
[in this window]
[in a new window]
|
Fig 1. (A) Lateral view of the chest x-ray demonstrating the endovascular stent covering the entire descending thoracic aorta. The yellow arrows indicate coils that had been used to embolize the bronchial arteries. (B) Computed tomographic scan shows the communication (yellow arrow) between the esophagus and the paraprosthetic graft space. (C) Endoscopic examination demonstrating the esophageal ulcer with suture material visible at the bottom (green arrow). (D) Intraoperative photograph showing the disrupted proximal anastomosis. The endovascular stent graft (green arrow) was directly visualized through the disrupted anastomosis.
|
|
Contrast enhanced computed tomography revealed a communication between the space around the prosthetic graft and the esophagus (Fig 1B). Endoscopic examination demonstrated a diverticulum in the esophagus with one surgical suture visualized at the bottom of an ulceration (Fig 1C). The diagnosis of an EPPF was established, and surgical intervention using a cryopreserved aortic homograft was undertaken.
Using a posterolateral left thoracotomy, the previous Dacron aortic graft (Boston Scientific Corp, Natick, MA) was dissected out. The proximal anastomosis site was found to be partially disrupted so that the endovascular stent graft was clearly visible through the aortic wall (Fig 1D). The esophagus was tightly adherent to this anastomotic site, and a small amount of pus was observed around it. Culture of the purulent discharge yielded Bacillus species.
Under left heart bypass support, the infected endovascular stent graft and previous prosthetic Dacron aortic graft (Boston Scientific Corp) were removed. The entire descending thoracic aorta was then replaced with a composite graft composed of a 7-cm long cryopreserved aortic homograft (CryoLife Intl Inc, Kennesaw, GA) sutured to a new prosthetic graft soaked in 0.5% rifampin to achieve the desired graft length. After subtotal esophagectomy and cervical esophagostomy, an omental flap based on a left gastroepiploic artery pedicle was prepared and pulled up through the esophageal hiatus to fill the space in the posterior mediastinum and ensure complete coverage of the new composite graft.
A follow-up computed tomographic scan 3 months after surgery revealed a new pseudoaneurysm developing at the distal anastomosis. Clinical findings at this point, including laboratory data and negative blood cultures, indicated the lower likelihood of an infected aneurysm; therefore, another endovascular stent graft repair was performed. At 8 months after the esophagectomy, the alimentary tract was reconstructed with a gastric roll. The patient was discharged through a local hospital 3 months later, and she has been doing well for 24 months without signs of recurrent infection.
|
Comment
|
|---|
In this case, the coexistence of several instigating factors is likely to be responsible for the development of EPPF, including the presence of prosthetic material, infection, and esophageal wall ischemia. The suture material found at the bottom of the esophageal ulcer is a potential cause of esophageal perforation, especially when combined with the chronic inflammation adjacent to the esophagus as a result of the anastomotic pseudoaneurysm. Because much of the arterial blood supply to the esophagus comes directly from the descending thoracic aorta, the insertion of the endovascular stent graft might have additionally resulted in considerable ischemia of the esophageal wall, contributing to the potential for perforation. Concomitant coil embolization of the bronchial arteries might have further exacerbated this esophageal ischemia, because the bronchial arterial network can serve as a supplementary blood source for the esophagus.
Recently in situ aortic replacement with a homograft has been widely recommended as surgical treatment for prosthetic graft infections [3, 4], given the relative resistance to infection offered by the homograft tissue. In this case, a 7-cm long cryopreserved aortic homograft was the only homograft available at the time of surgery. To make optimal use of the homograft, it was implanted in the area with demonstrable exudate beside the EPPF. Had it been available, the entire descending thoracic aorta would have ideally been replaced with a full length aortic homograft.
To prevent recurrent graft infection after EPPF repair, placing an omental transfer flap over the new graft is recommended [5]. In this case, because the entire descending thoracic aorta was replaced with the composite graft, a large surface area of omentum was required to cover the length of this graft. To this end, an omental flap based on a left gastroepiploic artery pedicle was brought up into the left thoracic cavity. The follow-up computed tomographic scan showed that this omental flap was of sufficient size to cover the entire graft from the proximal to the distal anastomosis (Fig 2).
View larger version (52K):
[in this window]
[in a new window]
|
Fig 2. Serial sections of a computed tomographic scan at 9 months after surgical repair demonstrating the omental flap covering the entire composite graft (green arrows).
|
|
|
References
|
|---|
- Iguchi A, Miyazaki S, Akimoto H, Ohmi M, Tabayashi K. Successful management of secondary aortoesophageal fistula with graft infection Thorac Cardiovasc Surg 2001;49:126-128.[Medline]
- Kieffer E, Chiche L, Gomes D. Aortoesophageal fistula: value of in situ aortic allograft replacement Ann Surg 2003;238:283-290.[Medline]
- Vogt PR, Turina MI. Management of infected aortic grafts: development of less invasive surgery using cryopreserved homografts Ann Thorac Surg 1999;67:1986-1989.[Abstract/Free Full Text]
- Pirard L, Creemers E, Van Damme H, Laurent S, Honore P, Limet R. In situ aortic allograft insertion to repair a primary aortoesophageal fistula due to thoracic aortic aneurysm J Vasc Surg 2005;42:1213-1217.[Medline]
- Kitayama J, Morota T, Kaisaki S, et al. Complete coverage of in situ aortograft by total omental pedicle flap as the most reliable treatment of aortoesophageal fistula Am J Surg 2006;192:130-134.[Medline]
Top
Abstract
Introduction
