SciELO - Scientific Electronic Library Online

vol.9 número1Infecções em cateteres venosos centrais de longa permanência: revisão da literaturaTratamento endovascular de pseudoaneurisma de aorta torácica: relato de caso índice de autoresíndice de assuntospesquisa de artigos
Home Pagelista alfabética de periódicos  

Serviços Personalizados



  • texto em Português
  • Português (pdf)
  • Artigo em XML
  • Como citar este artigo
  • SciELO Analytics
  • Curriculum ScienTI
  • Tradução automática


Links relacionados


Jornal Vascular Brasileiro

versão impressa ISSN 1677-5449

J. vasc. bras. vol.9 no.1 Porto Alegre  2010 



Aortic-esophageal fistula after endovascular repair of Stanford type B thoracic aortic dissection



Cláudia Gurgel MarquesI; Alexandre FioranelliII; Gustavo Politzer TellesIII; Paulo SaadIII; Alvaro Razuk FilhoIV; Valter Khegam KarakhanianIV; Valter Castelli JuniorIV; Roberto Augusto CaffaroV

ICirurgiã vascular, Faculdade de Ciências Médicas da Santa Casa de São Paulo (FCMSCSP), São Paulo, SP, Brazil
IIMestre em Cirurgia Vascular, Professor instrutor, FCMSCSP, São Paulo, SP, Brazil
IIICirurgião vascular, Pós-graduando em Cirurgia Vascular, FCMSCSP, São Paulo, SP, Brazil
IVDoutor, Professor adjunto, disciplina de Cirurgia Vascular, FCMSCSP, São Paulo, SP, Brazil
VDoutor, Professor e chefe da disciplina de Cirurgia Vascular, FCMSCSP, São Paulo, SP, Brazil





Endoluminal stent-graft for type B aortic dissection is a new alternative to reduce surgical trauma. However, medium- and long-term complications are still little known and poorly reported, such as the aortic-esophageal fistula. The objective of this study is to describe three cases of aortic-esophageal fistula after the endovascular treatment of 23 cases of descending aortic dissection conducted by the vascular surgery team of Santa Casa de São Paulo in a retrospective study. These patients presented some common characteristics: chronic dissection, successful early outcome, need of reinterventions, occlusion of arterial trunks such as subclavian artery, mesenteric artery, celiac trunk, and finally, a fast fatal course after the first fistula-related symptoms. Therefore, despite rarely described in the literature, aortic-esophageal fistula is a complication of the endovascular treatment of descending aortic dissection which demands attention due to its unpredictability, recurrence, and fatal outcome.

Keywords: Dissection, aorta, fistula, esophagus, postoperative complications.




Endovascular treatment of Stanford type B aortic dissection in the last years has shown to be a therapeutic option to reduce early mortality and morbidity. However, the medium and long-term evolution is still little-known, once there are few studies with great casuistics describing the post-operative follow up to these patients found in literature.1-3

This new therapeutic alternative, associated to the intensive clinical treatment with a rigid control of arterial pressure, aims at occluding the proximal point of entry to reduce or block the flow to the false lumen in a less invasive way than that of open surgical technique, without the necessity of thoracotomy, aortic clamping or extracorporeal circulation.4 Surgery duration is shorter, as well as blood losses and hospitalization time.4,5

Due to its status as a recent technique, long-term results are not yet well defined. Some early complications, such as the emergence of leakages or endoleaks, cerebral strokes and marrow ischemia, have already been described in literature.6,7

Aortic-esophagic fistula, on the other hand, is a little described complication until now. However, the finding of three cases of this complication in a retrospective study of 23 patients with type B aortic dissection treated at the Santa Casa de São Paulo, as well as the severity of its evolution, have suggested us that this kind of post-operative intercurrence is of greater relevance.

In this study we have described the emergence of aortic-esophagic fistula after endovascular treatment of descending thoracic aorta dissection in three cases among 23 patients treated, and we report their evolution, as well as adopted therapeutic options.


Cases description

Of the 23 patients treated, three presented aortic-esophagic fistula in the follow up.

Case 1

A 70-year-old male patient with hypertension and diabetes was submitted to endovascular treatment due to progressive dilation of thoracic aortic dissection, reaching a maximum diameter of the false lumen of 50 mm, diagnosed nine months before. The dissection went from the left subclavian artery to the right iliac artery.

A Gore TAG® prosthesis (Gore, USA) of 37 mm diameter and 20 cm length was used, which, even if occluding the left subclavian artery origin, presented a proximal leakage at the end of the procedure without distal reentry.

At 11 months of post-operative, however, a new computed tomography (CT) revealed an increase in the false lumen diameter to 65 mm, besides a proximal leakage in the left subclavian artery origin confirmed by angiography. Left axillary access embolization from the origin of subclavian artery with a movable core wire was chosen.

Nine months after, a new CT evidenced the persistence of proximal leakage. Given the proximity to left common carotid artery, the patient was submitted to crossed carotid-carotid extra-anatomic derivation, followed by the placing of a new primary proximal Gore TAG® endoprosthesis, occluding the origin of the left common carotid.

The patient was in the 23rd post-operative month from the first surgery and, after 10 days, he presented a voluminous hematemesis, evolving to a fatal hemorrhagic shock. Necropsy identified the presence of an aortic-esophagic fistula.

Case 2

A 52-year-old male patient presenting hypertension and diabetes was submitted to endovascular treatment due to the increase of 1 cm in the false lumen diameter in 6 months under clinical treatment. He presented thoracic aortic dissection from the left subclavian artery to renal arteries diagnosed 19 months earlier.

At the time of treatment, the CT indicated a false lumen with maximum diameter of 70 mm. Proximal entry was situated 20 mm from left subclavian artery.

Two Gore TAG® prostheses were used (40 mm diameter and 15 cm extension), without the necessity of occlusion of the subclavian artery. At the end of procedure, the patient presented reentry point next to the celiac trunk, with no proximal leakage.

After 1 month, the false lumen diameter had enlarged to 89 mm, and the leakage persisted. False lumen embolization with movable core wire was chosen.

Two months later, tomography indicated reduction of the false lumen to 50 mm, with no opacification through reentry, and the presence of air in the esophagus.

The patient was then rehospitalized, presenting weakness, lumbar pain and fever. He evolved to death 3 days later due to a massive digestive hemorrhage. During hemorrhagic event, digestive endoscopy evidenced an esophageal ulcer with active bleeding. Necropsy found an aortic-esophagic fistula.

Case 3

A 57-year-old hypertensive male patient, with type B dissection diagnosed 15 months earlier, was submitted to treatment due to dilation. The dissection stretched from the subclavian artery to iliac arteries, with a false lumen maximum diameter of 55 mm in the thoracic aorta.

A Zenith® (Cook, USA) endoprosthesis, 34 mm diameter and 20 cm extension, was chosen, with occlusion of the left subclavian artery, placed 12 mm from the proximal entry. At the end of procedure, there was no proximal leakage, and a reentry point was observed in the origin of the celiac trunk.

After an early post-operative without intercurrences, the patient was rehospitalized presenting claudication on the left superior limb, thoracic pain and hematimetric downfall in the 15th day after the surgery, besides the CT indication of contrast in the false lumen, contrast extravasation in left hemithorax and reentry persistence in celiac trunk (Figure 1). These findings were confirmed by angiography, besides a new point of entry in the iliac arteries' bifurcations. The initial conduct was placing a Palmaz stent® (Cordis, USA) in the aorta, at the celiac trunk and superior mesenteric artery level, and as the leakage did not stop, a glue injection (Hystoacril®, Braun, Germany) was administered in the false lumen, proximal to this point (Figure 2).




After an evolution with no intercurrences for 14 days, the patient started to present thoracic pain and fever, and a picture of hospital pneumonia was diagnosed. He was hospitalized for the second time and submitted to antibiotic therapy for a total period of 2 months, receiving vancomycin, polymyxin and imipenem, under intensive treatment, with periods of sepsis and clinical improvement. Three months after the first surgery, a scintillography with marked leukocytes evidenced aortic arch infection. He then started to present hematemesis, being submitted to a digestive endoscopy that spotted a cambering in the esophagus wall with a big hematoma (Figure 3). An angiography confirmed the hypothesis of aortic-esophagic fistula originated in the aortic arch proximal to left subclavian artery (Figure 4).




Due to the proximity with left common carotid artery, a crossed carotid-carotid extra-anatomical derivation was performed, followed by the placing of a new proximal Zenith® endoprosthesis (38 mm diameter and 10 cm extension) in order to occlude the origin of the fistula (Figure 5). Five days after this surgery the patient evolved to death due to sepsis and multiple organ failure. Aortic-esophagic fistula was identified at necropsy and is presented in Figure 6.




Table 1 shows a synthesis of the evolution of the three cases with their pre and post-operative features.




The ideal treatment for Stanford B aortic dissection is still a controversial subject. The finding of high mortality in open surgical treatment, up to 67% in some casuistics, compared to 15% of mortality in clinical treatment led, in the last years, to the restriction of the surgical treatment to cases that presented complications. Nevertheless, since 1993 endovascular treatment has shown to be a new alternative to improve the evolution of these patients through the occlusion of proximal entry hole in a less invasive way. Hospital mortality reports of 13% have led some reference centers to adopt this technique in a systematic way.4-6

On the other hand, auspicious results indicating early mortality reduction contrast with the lack of studies on the durability or late evolution of these cases.

In this casuistic of 23 patients with descending thoracic-aortic dissection submitted to endovascular treatment by the Santa Casa de São Paulo vascular surgery team, three cases of aortic-esophagic fistula were found, a complication rarely described in literature.

Some findings were common to the three cases, as the chronicity of the dissection, the treatment indicated due to aortic dilation, an immediate post-operative without intercurrences, the presence of leakage in the routine follow up, the need of reintervention before aortic-esophagic fistula (due to leakages), the need of occluding arterial branches and trunks as those of the subclavian and superior mesenteric arteries and celiac trunk and, finally, the evolution to death (Table 1).

According to Porcu et al.,6 tthe distal segment to the left bronchus would be the most susceptible esophagic level to the formation of fistulas, where a poor revascularization may lead to necrosis of the organ wall, while in the esophagus's proximal and distal thirds, the revascularization originating in the inferior thyroid artery, in the first, and left diaphragmatic and gastric, in the last, would act in a way to protect these segments from ischemia. Although the direct compression by endoprosthesis cannot be discarded as the cause of these fistulas, the reduction of the aorta's diameter after the treatment goes against this hypothesis.

Of the cases reported, two presented reduction of the false lumen after endovascular treatment, and in both it was necessary to occlude the left subclavian artery, while in the third, the persistence of a reentry point and of aortic expansion led to the false lumen and the common hepatic artery embolization. In this last case, the first tomography after the treatment evidenced an 89 mm false lumen, which is, 19 mm larger than in pre-operative.

The hypothesis of aortic direct compression in the esophagus was defended by Herolkl et al.,7 when they reported a case of fistula among 18 patients submitted to this treatment by dissection in a patient who presented a 13 cm diameter aorta. The false lumen's disruption in the esophagus led to death in the 62nd post-operative day.

For Blockler et al.,5 endovascular systems' rigidity may lead to penetration or lesion of the aortic wall, and the ideal endoprosthesis for the dissections should be more flexible and has not been developed yet. This author has also described a fatal case of fistula (although aortic-bronchial and not esophagic), to which extrinsic compression is attributed, among 37 cases of descending aortic dissection submitted to endovascular treatment. Seventeen months after the surgery, this patient presented massive hemoptysis, treated through embolization of the false lumen with spring, evolving to death 2 weeks later, when he presented another hemorrhagic episode.

In the Eggebrecht et al.'s4 casuistic, of the 38 patients with aortic dissection treated with endoprosthesis, one evolved with an aortic-esophagic fistula in the late post-operative. The cause of the complication is not discussed in the report, and the subsequent hemorrhage resulted in the patient's death.



Some findings are common to all the described cases, such as the onset of symptoms in a late follow up, after a post-operative free of intercurrences, and the fatal course in spite of the various interventions adopted. Some hypotheses were formulated to explain these complications, but casustic in literature is not sufficient to establish the risk factors for its emergence. Even if unpredictable up to this moment, it appears to be related to the cases of very dilated aortas or those in which there is the necessity of occluding the left subclavian artery or visceral trunks, being advisable the previous extra-anatomical revascularization whenever feasible. Especially in these cases, clinical suspicion should be conducted with invasive diagnostic methods such as endoscopy, tomography and angiography early in order not to delay the treatment.

With the increase of experience in endovascular treatment of type B aortic dissection, the spectrum of late complications has been broadening, including severe and lethal events that demand detailed and specific research. Although rarely described in literature, aortic-esophagic fistula consists of a very hazardous complication, which led all the reported cases to death. As previously said, it is fundamentally important, at this moment, the development of long-term follow up studies on endovascular treatment for descending aortic-thoracic dissection capable of responding to several questionings and certainly of describing these other rarely reported complications.



1. Goksel O, Tireli E, Kalko Y, et al. Mid-term outcome with surgery for type B aortic dissections: a single center experience. J Card Surg. 2006;23:27-30.         [ Links ]

2. Kiyotaka I, Shinichi S, Keiji U. Early and medium-term results of stent-graft treatment of DeBakey type III aortic dissection. J Cardiovasc Surg. 2006;47:651-7.         [ Links ]

3. Hutschala D, Fleck T, Czerny M, et al. Endoluminal stent-graft placement in patients with acute aortic dissection type B. Eur J Cardio-thorac Surg. 2002;21:964-9.         [ Links ]

4. Eggebrecht H, Herold U, Kuhnt O, et al. Endovascular stent-graft treatment of aortic dissection: determinants of post-interventional outcome. Eur Heart J. 2005;26:489-97.         [ Links ]

5. Blockler D, Schumacher H, Ganten M, et al. Complications after endovascular repair of acute symptomatic and chronic expanding Stanford type B aortic dissections. J Thorac Cardiovasc Surg. 2006;32:361-8.         [ Links ]

6. Porcu P, Chvanon O, Sessa C, Thony F, Aubert A, Blin D. Esophageal fistula after endovascular treatment in a type B aortic dissection of the descending thoracic aorta. J Vasc Surg. 2005;41:708-11.         [ Links ]

7. Herolkl U, Piotrowski J, Baumgart D, Eggebrecht H, Erbel R, Jakob H. Endoluminal stent graft repair for acute and chronic type B aortic dissection and atherosclerotic aneurysm of the thoracic aorta: an interdisciplinary task. Eur J Cardio-thorac Surg. 2002;22:891-7.         [ Links ]



Claudia Gurgel Marques
R. Maria Borba, 67/24, Vila Buarque
CEP 01221-040 - São Paulo, SP, Brazil

Manuscript received Apr 06 2009,accepted for publication Feb 23 2009.



No conflicts of interest declared concerning the publication of this article.

Creative Commons License Todo o conteúdo deste periódico, exceto onde está identificado, está licenciado sob uma Licença Creative Commons