CASE REPORT

Percutaneous Treatment of an Infected Aneurysmal SacSecondary to Aortoesophageal Fistula with a Historyof Stent-Graft Treatment for Thoracic Aortic Aneurysm

Furuzan Numan • Fatih Gulsen • Murat Cantasdemir •

Serdar Solak • Harun Arbatli

Received: 19 May 2011 / Accepted: 29 July 2011 / Published online: 19 August 2011

� Springer Science+Business Media, LLC and the Cardiovascular and Interventional Radiological Society of Europe (CIRSE) 2011

Abstract A 68-year-old man who was subjected to stent-

grafting of a descending thoracic aortic aneurysm (TAA)

4 months previously was admitted to our hospital with

constitutional symptoms, including high fever, sweating,

nausea, vomiting, weight loss, and backache. An infected

aneurysmal sac was suspected based on computed tomog-

raphy (CT) findings, and an aortoesophageal fistula (AEF)

was identified during esophagoscopy. CT-guided aspiration

was performed using a 20-G Chiba needle, confirming the

presence of infection. For treatment of the infected aneu-

rysmal sac, CT-guided percutaneous catheter drainage in a

prone position was performed under general anesthesia

with left endobronchial intubation. Drainage catheter

insertion was successfully performed using the Seldinger

technique, which is not a standard treatment of an infected

aneurysmal sac. Improvement in the patient’s clinical

condition was observed at follow-ups, and CT showed total

regression of the collection in the aneurysmal sac.

Keywords Thoracic endovascular aortic repair �Aortoesophageal fistula � Infected aneurysmal sac �Percutaneous drainage

Introduction

Aortoesophageal fistula (AEF) is a rare clinical condition

associated with high morbidity and mortality rates and a

fatal outcome if treated surgically or medically; limited

treatment alternatives exist. Thoracic aortic aneurysms

(TAA) are the most frequent primary cause of AEF and are

responsible for two-thirds of all AEF cases [1]. Other

causes of primary AEF include ulcerated plaques of the

thoracic aorta, esophageal neoplasms, and drinking corro-

sive materials [2, 3]. Uncommon cases of AEF secondary

to endoscopic or operative esophageal procedures and

endoluminal stent-grafting also have been reported [4].

AEF secondary to prosthetic graft replacement of the tho-

racic aorta is a rare but frequently catastrophic complica-

tion. Other than occurring as a communication between the

aorta and esophagus, AEF secondary to prosthetic graft

replacement may occur as an esophagoparaprosthetic fis-

tula [5]. True AEF, which occurs as a direct communica-

tion between the aorta and esophagus, is usually

characterized by massive, fatally progressing upper gas-

trointestinal bleeding, has high mortality rates, and is more

commonly identified by postmortem examination. AEF

between the esophagus and aneurysmal sac adjacent to an

endovascular stent-graft of the thoracic aorta may occur

secondary to sepsis when the fistula reaches the proximal

or distal aspect of the stent-graft, with the presence of an

endoleak inside the aneurysmal sac, or with graft infection.

In this case report, AEF between the esophagus and

thrombosed aneurysmal sac adjacent to a stent-graft of the

thoracic aorta developed in the late period after thoracic

endovascular aortic repair (TEVAR). We present a case of

an infected aneurysmal sac related to AEF that developed

after TEVAR and its treatment by computed tomography

(CT)-guided insertion of a drainage catheter.

F. Numan (&) � F. Gulsen � M. Cantasdemir � S. Solak

Department of Interventional Radiology, Cerrahpasa Faculty

of Medicine, Istanbul University, Fatih, Istanbul 34098, Turkey

e-mail: drfgulsen@yahoo.com

H. Arbatli

Department of Cardiovascular Surgery, Faculty of Medicine,

Maltepe University, Istanbul, Turkey

123

Cardiovasc Intervent Radiol (2012) 35:690–694

DOI 10.1007/s00270-011-0256-1

Case Report

A 68-year-old man with a history of hypertension, hyper-

cholesterolemia, heavy smoking, and chronic renal failure

underwent surgery in our hospital for an abdominal aortic

aneurysm 4 years prior and was subjected to stent-grafting

of a descending TAA 4 months prior. After TEVAR, dur-

ing the follow-up period, both clinical and radiological

course was uneventful until he was admitted to our hospital

for constitutional symptoms, including weight loss in the

previous 2 months and nausea, vomiting, backache, and

high fever that had developed in the previous 2 weeks. On

physical examination, a generally poor condition, dehy-

dration, and cachexia were noted. Blood pressure was

within normal limits. Clinical analysis reported a white

blood cell count of 24,300/ml and C-reactive proteins at

86.4 mg/dl. The fever persisted despite treatment with a

broad-spectrum antibiotic. Chest radiography revealed

no source of infection. A contrast-enhanced CT was plan-

ned but due to preexisting chronic renal insufficiency,

contrast-enhanced CT was contraindicated by consultant

nephrologist.

CT examination showed the presence of a collection

with a size of 4- 9 8-cm containing air bubbles around the

aortic stent-graft, which was a sign of possible infection of

the aneurysmal sac (Fig. 1). Subsequent esophagogastros-

copy and esophagography examinations revealed a small

fistula between the TAA and mid-third esophagus (Figs. 2,

3A). A swallow study revealed contrast extending into the

aneurysmal sac around the endograft. Because of there

were no signs of bleeding in esophagogastroscopy exami-

nation, presence of an endoleak was not considered.

However, due to history of chronic renal failure in our

patient, contrast-enhanced CT examination was not per-

formed. There were no imaging finding that could explain

the inflammatory symptoms and laboratory findings, except

the collection around the aortic stent-graft. A prediagnosis

of abscess around the aortic stent-graft was made depending

on a combination of clinical, laboratory, and imaging find-

ings. For empirical antibiotherapy, cefazolin was admin-

istered (3 9 1 g) by intravenous injection for 2 weeks.

Because there was no improvement in the patient’s clinical

status, to confirm the presence of an abscess and identify the

microbiological agent, percutaneous diagnostic aspiration

and drainage was planned.

The patient was placed in the prone position and

underwent general anesthesia with endobronchial intuba-

tion. A 20-G Chiba needle was inserted into the aneurysmal

sac via a posterior approach under CT guidance (Fig. 4A).

A purulent collection by needle aspiration confirmed

infection of the aneurysmal sac adjacent to the aortic stent-

graft (Fig. 4B), and CT-guided drainage catheter insertion

into the aneurysmal sac by the Seldinger method was

performed (Fig. 4C). Approximately 210 ml of infected

fluid was drained via the catheter during 7 days.

Cultures of the aortic sac fluid later grew Staphylococcus

aureus and specific intravenous antibiotherapy (ciprofloxa-

cin and sultamicillin) was started. At 7 days postdrainage,

catheter output had decreased to less than 10 ml/24 h during

the preceding 2 days, and the patient remained afebrile.

Because a follow-up CT study showed total regression of the

collection in the aneurysmal sac and the patient’s clinic

condition had improved, the drainage catheter was removed

1 week after the procedure. The subsequent postprocedure

course was uneventful, and the patient was discharged in

good condition.

The follow-up CT study 3 months after treatment

revealed no signs of the collection in the thrombosed

aneurysmal sac (Fig. 5). At follow-up esophagography

4 months after treatment (Fig. 3B), we ascertained that the

small fistula between the esophagus and aneurysmal sac

Fig. 1 CT examination shows the presence of a collection of air

bubbles around the aortic stent-graft as a sign of possible infection of

the aneurysmal sac and no evidence of an endoleak

Fig. 2 Esophagogastroscopy examination revealed small fistula

between the TAA sac and mid-third esophagus

F. Numan et al.: Percutaneous Treatment of an Infected Aneurysmal Sac 691

123

had closed spontaneously. During the postdischarge period

of patient, no signs or symptoms of an infection were

observed. Our patient died because of acute myocardial

infarction 14 months after treatment.

Discussion

AEF is a relatively rare but life-threatening cause of upper

gastrointestinal bleeding, constituting less than 10% of all

aortoenteric fistulas. It is responsible for 0.15% of all upper

gastrointestinal bleeding [6]. Despite several strategies for

surgical treatment, there is little consensus about the opti-

mal management of this disease. Diagnosis of AEF is

rarely made before massive hematemesis. However, most

cases are associated with characteristic Chiari’s triad,

including chest pain and sentinel hematemesis of red blood

followed at variable intervals of time by rapidly fatal

massive bleeding [4, 6].

AEF is a well-known complication of open surgical

repair of thoracic aneurysms [7, 8]. Aortoenteric fistula, as

a rare complication of endovascular prosthetic stent-graft-

ing of aneurysms of the thoracic and abdominal aorta, is

seen more commonly, due to the increasing frequency of

usage of this treatment method in recent years. Few cases

of AEF complications secondary to stent-graft repair of

TAAs have been reported. Eggebrecht et al. reported AEF

complications in 6 of 268 patients (1.9%) at follow-up of

thoracic endovascular aortic repair of various thoracic

aortic diseases, and those cases resulted in death [9].

Fig. 3 Preprocedure

esophagography A examination

revealed a small fistula between

the TAA sac and mid-third

esophagus and postprocedure

esophagography (4 months after

the treatment), B showing that

the small fistula between the

esophagus and aneurysmal sac

had closed spontaneously

Fig. 4 CT-guided Chiba needle aspiration (A, B) and CT-guided drainage catheter insertion (C) into the aneurysmal sac using the Seldinger

technique

Fig. 5 Follow-up CT study 3 months after treatment showing no

signs of the collection in the thrombosed aneurysmal sac

692 F. Numan et al.: Percutaneous Treatment of an Infected Aneurysmal Sac

123

Mechanic and infective factors are disputably thought to

play roles in the development of aortoenteric fistulas,

secondary to surgical aortic repair. These factors include

rupture of abscesses of graft infections into the intestine

and erosion of structures adjacent to the prosthetic material

due to pulsation and pseudoaneurysm development on the

anastomosis line. In cases of aortoesophageal fistula fol-

lowing endovascular repair, Hance et al. suggested that

fistulas may arise secondary to (1) the development of

pseudoaneurysm, (2) endoleak into the residual aneurysmal

sac, or (3) erosion of the stent-graft through the aorta [10].

Eggebrecht et al. stated that AEF formation may be related

to pseudoaneurysm because they found periaortic hema-

toma in two of three AEF cases secondary to TEVAR [8].

Other important factors that play roles in AEF development

include the occlusion of esophageal arteries originating

from the thoracic aorta after stent-grafting and esophageal

erosions related to these occlusions. Collateral weakness of

the middle part of the esophagus, as opposed to the cervical

and distal portions, is thought to be related to the frequency

of AEF secondary to TEVAR in this segment.

However, hypotheses about mechanisms of fistulization

remain highly speculative. Some authors claim that a pri-

mary infection of the endograft due to contamination may

evolve into an aortic wall abscess with subsequent drainage

in to the esophagus or gastrointestinal system. This

mechanism implies a graft contamination during implan-

tation, maybe secondary to iatrogenic maneuvers, as a

result of bacteremia, or even due to migration of micro-

organism from the aortic mural thrombus [11]. Also, in

case of contained ruptured aneurysms, chronic inflamma-

tion due to the reabsorption of the posterior mediastinal

hematoma may represent another potential mechanism of

fistulization [12].

AEF secondary to TEVAR could occur as a direct

communication between the thoracic aorta and esophagus

or between the esophagus and aneurysmal sac adjacent to

the stent-graft. The clinical presentation of the patient is

highly associated with the type of AEF. The case may

present itself as massive upper gastrointestinal bleeding

when the esophagus directly communicates with the aorta

or aneurysmal sac with an existing endoleak, or, as in our

case, as an infection when the esophagus communicates

with a completely thrombosed aneurysmal sac.

There is no consensus in the literature about which

should be the optimal antimicrobial therapy for infected

aneurysmal sac or for how long it should be administrated.

Generally, it is given intravenously at the maximal toler-

ated dosage for at least the initial 4–6 weeks and followed

by a sequential oral regimen once the acute phase of the

infection has subsided. In cases when surgery is not per-

formed because of an excessive risk from the procedure, it

seems prudent to continue with an oral regimen indefinitely

[13]. However, most stent-graft infections were identified

causative microorganisms that are highly virulent patho-

gens, such as methicillin-resistant Staphylococcus aureus,

Streptococcus spp., or gram-negative species, such as

Pseudomonas and Klebsiella [14]. Control of sepsis in

these cases could not be achieved even with optimal anti-

biotic therapy, and the need for an adjunct therapy should

be evaluated.

Most authorities agree that an infected stent graft should

be removed if a patient’s condition permits. Patients with

aortic aneurysms often have multiple existing comorbidi-

ties, including coronary artery disease, cerebral vascular

disease, and hypertension. All of these factors may preclude

a patient from being a desirable surgical candidate. High-

risk patients have been reported to survive with conservative

treatment consisting of antimicrobial therapy and percuta-

neous drainage [15]. However, the clinical outcome of

conservative treatment is poor and this option should be

reserved for selected patients such as patients with a nonb-

leeding AEF, who has comorbidities or poor general con-

ditions. Also conservative treatment with appropriate

antibiotics and percutaneous drainage in patients with sepsis

secondary to nonbleeding AEF can promote the reduction of

sepsis and decrease the perioperative morbidity in case of

a subsequent surgical treatment for the closure of the

fistula [16].

As a surgical option, left thoracotomy with subsequent

aortic graft replacement and esophageal fistula resection

may be appropriate AEF therapy [3, 7]. Temporary or

permanent extra-anatomic bypass can be performed in the

presence of no active bleeding. However, mediastinitis,

sepsis, and bleeding are common complications after sur-

gical treatment of AEF, and this treatment option has high

morbidity and mortality rates [17]. Recently, stent-graft

placement has been successfully used to treat secondary

AEF after surgery.

There are few reports of AEF complications of TAA

after stent-graft repair in the literature but, as we know,

AEF without direct aortic communication between the

esophagus and a completely thrombosed aneurysmal sac,

which is successfully treated conservatively, has not been

reported. We believe that, as in our case, to impede sepsis

occurrence, draining infected material with a drainage

catheter in aneurysmal sacs secondary to fistula is an

effective, inexpensive, and feasible treatment option. We

should keep in mind that draining infected collections in

this way may hinder extension of the fistula tract to distal

or proximal ends of the stent-graft. In our case, the small

fistula had closed spontaneously, but in case of persistence

of the fistula, esophageal cover stents can be considered as

a treatment option for the closure of the fistula tract.

However, cover stents will not provide sufficient thera-

peutic effect on collection.

F. Numan et al.: Percutaneous Treatment of an Infected Aneurysmal Sac 693

123

Various complications, such as pneumothorax, catheter-

related pain, and of utmost importance—in case of pres-

ence of an endoleak or rupture of stent-graft—massive

bleeding can occur during or after the procedure. In our

case, left endobronchial intubation was performed and

contralateral lung was deflated to prevent pneumothorax.

Also, because of possible catastrophic results, this kind of

procedure should be planned thoroughly before the inter-

vention and performed in a step-by-step manner. Because

of this, at first the presence of abscess was confirmed by

percutaneous aspiration with 20-G Chiba needle and then

after confirmation of a purulent material drainage catheter

was inserted into the aneurysmal sac by the Seldinger

method. At the postprocedure period, the possibility of

massive bleeding should not be neglected and the patient

should be followed up closely. A major drawback in our

case was the inability to perform a contrast-enhanced CT

because of our patient’s chronic renal failure. It should not

be forgotten that it is essential to confirm the absence of an

endoleak and demonstrate the abscess or retention by using

contrast-enhanced CT in patients who have suitable med-

ical conditions.

As a result, the probability of AEF complications should

be kept in mind. In patients with infected aneurysmal sacs

secondary to fistula between a completely thrombosed

aneurysmal sac and the esophagus, without direct aortic

communication, we think that the method we used in this

case could contribute to hindering the extension of the

fistula tract to the aorta and subsequent probable sepsis and

could be a life-saving treatment option. However, this

technique should be reserved for selected patients who are

not responding to aggressive antibiotic regimens, as a bail-

out option for those unfit for intervention, or as an adjunct

during optimization for those awaiting definitive treatment.

Conflict of interest The authors declare that they have no conflict

of interest.

References

1. Flores J, Shiiya N, Kunihara T et al (2004) Aortoesophageal

fistula: alternatives of treatment case report and literature review.

Ann Thorac Cardiovasc Surg 10(4):241–246

2. Hollander JE, Quick G (1991) Aortoesophageal fistula: a com-

prehensive review of the literature. Am J Med 91:279–287

3. Amin S, Luketich J, Wald A (1998) Aortoesophageal fistula: case

report and review of the literature. Dig Dis Sci 43:1665–1671

4. Lin CS, Tung CF, Yeh HZ et al (2008) Aortoesophageal fistula

with a history of graft treatment for thoracic aortic aneurysm.

J Chin Med Assoc 71(2):100–102

5. Kieffer E, Chiche L, Gomes D (2003) Aortoesophageal fistula:

value of in situ aortic allograft replacement. Ann Surg 238(2):

283–290

6. Takano S, Katsuhara K, Nobuhara K et al (2009) Aortoesopha-

geal fistula due to esophageal ulcer. Gen Thorac Cardiovasc Surg

57(5):255–257

7. Isasti G, Gomez-Doblas JJ, Olalla E (2009) Aortoesophageal

fistula: an uncommon complication after stent-graft repair of an

aortic thoracic aneurysm. Interact Cardiovasc Thorac Surg

9(4):683–684

8. Eggebrecht H, Baumgart D, Radecke K et al (2004) Aorto-

esophageal fistula secondary to stent-graft repair of the thoracic

aorta. J Endovasc Ther 11(2):161–167

9. Eggebrecht H, Mehta RH, Dechene A et al (2009) Aortoesoph-

ageal fistula after thoracic aortic stent-graft placement: a rare but

catastrophic complication of a novel emerging technique. JACC

Cardiovasc Interv 2(6):570–576

10. Hance KA, Hsu J, Eskew T et al (2003) Secondary aortoesoph-

ageal fistula after endoluminal exclusion because of thoracic

aortic transection. J Vasc Surg 37(4):886–888

11. Martens K, De Mey J, Everaert H et al (2007) Aortoesophageal

fistula following endovascular exclusion of a thoracic aneurysm.

Int Angiol 26(3):292–296

12. Czerny M, Zimpfer D, Fleck T et al (2005) Successful treatment

of an aortoesophageal fistula after emergency endovascular tho-

racic aortic stent-graft placement. Ann Thorac Surg 80(3):

1117–1120

13. Chiesa R, Tshomba Y, Kahlberg A et al (2010) Management of

thoracic endograft infection. J Cardiovasc Surg 51(1):15–31

14. Perera GB, Fujitani RM, Kubaska SM (2006) Aortic graft

infection: update on management and treatment options. Vasc

Endovascular Surg 40(1):1–10

15. Saleem BR, Berger P, Zeebregts CJ et al (2008) Periaortic en-

dograft infection due to Listeria monocytogenes treated with graft

preservation. J Vasc Surg 47:635–637

16. Belair M, Soulez G, Oliva V et al (1998) Aortic graft infection: the

value of percutaneous drainage. AJR Am J Roentgenol 171(1):

119–124

17. Iguchi A, Miyazaki S, Akimoto H et al (2001) Successful man-

agement of secondary aortoesophageal fistula with graft infec-

tion. Thorac Cardiovasc Surg 49(2):126–128

694 F. Numan et al.: Percutaneous Treatment of an Infected Aneurysmal Sac

123