Surgical treatment for secondary aortoesophageal fistula after prosthetic aortic replacement: A report of four cases

Highlights • Secondary aortoesophageal fistula (AEF) after aortic prosthetic replacement is rare, so there have been few reports assembling multiple cases.• AEF is difficult to diagnose early if there is little experience, and hence is associated with high mortality and morbidity rates.• The strategy involving stepwise surgery in coordination with infection control and general health improvement was attempted in our facility.• From our observations, this approach appears to be a valid strategy for secondary AEF after aortic prosthetic replacement.


Introduction
Aortoesophageal fistula (AEF) may be primary or secondary. With the increase of thoracic aortic aneurysm surgery and thoracic endovascular aortic repair (TEVAR), secondary AEF has been reported [1][2][3]. Some treatment strategies for AEF have been presented [4][5][6][7][8][9]. However, there have been few reports with multiple cases of secondary AEF after prosthetic aortic replacement, so the treatment strategies remain controversial. In our facility, we have attempted a strategy including performing surgery stepwise for 4 cases of AEF after prosthetic aortic replacement, and we herein report them. This work has been reported in line with the PROCESS criteria [10].

Brief case histories before staged surgeries
Case 1: an 83-year-old man. Around 28 months after total arch replacement (TAR), he suffered fever and shivering. Prosthetic graft infection was diagnosed when computed tomography (CT) showed ectopic gas around the prosthesis. After 17 days in hospital, an infected pseudoaneurysm at the distal anastomosis of TAR and a resulting aortobronchial fistula were found. Emergency descending aorta replacement (DAR) was performed, and he was saved. Nine days after DAR, an air leak from the chest drain was found, and AEF was confirmed 2 days later by esophagogastroduodenoscopy (EGD).
Case 2: a 72-year-old man. Four months after TAR, he suffered fever and shivering. Prosthetic graft infection was diagnosed by CT. After 10 days in hospital a diagnosis of AEF was confirmed by EGD.  performed, and she was saved. She developed a fever 18 days after TEVAR, and prosthetic graft infection was diagnosed by CT ( Fig. 1). A diagnosis of AEF was confirmed by EGD (Fig. 2), 31 days after TEVAR.
Case 4: a 79-year-old woman. Around 17 months after DAR, she had fever, and prosthetic graft infection was diagnosed by CT. On her 2nd day in hospital, an infected pseudoaneurysm at the proximal site of DAR and a resulting aortobronchial fistula were found. Emergency TAR was performed, and she was saved, but 48 days later she suffered fever and shivering. Reoccurrence of prosthetic graft infection was diagnosed by CT, and AEF was confirmed by EGD 50 days after TAR.
Detailed description of our cases with AEF are shown in Table 1.

Our staged treatment strategy is as follows
First-stage surgery is esophagectomy, esophagostomy, and enterostomy. The first operation involves resection of the esophagus, formation of an oral side esophagostoma on the left side of the neck, and jejunostomy of the feeding route.
Second-stage surgery is additional cardiovascular surgery if needed. In case 2, the distal anastomotic site of the TAR was considered to be the main infection site associated with the fistula, so TEVAR was performed for reinforcement. In case 4, as in case 1, the main site of infection associated with the fistula was at the junction between two prostheses; in this case, between the proximal end of the DAR (installed 17 months earlier) and the distal end of the TAR (installed 50 days earlier). However, infection also caused a pseudoaneurysm at the distal anastomosis of the DAR, so in stage 2 the DAR was performed again. In case 1 and 3, no additional surgery was done there.
Third-stage surgery is esophageal reconstruction. This was performed in cases 1, 2, and 3, after improvement of general conditions and infection findings. In those 3 cases, a gastric tube was formed using the patient's stomach to replace the esophagus, in an antethoracic route.

Outcomes after staged surgery
Case 1 survived the third-stage surgery, but suffered hypercalcemia associated with a tumor and renal failure, and died on day 224 of hospitalization. Case 2 recovered and was discharged from our hospital on day 227. In case 3, after esophagus reconstruction, suture failure occurred between the esophagus and the gastric tube, and formed a skin fistula. As of day 216 of hospitalization, she has received conservative treatment for that problem. Case 4 survived after second-stage surgery, but has continued to receive treatment for infection and chronic disseminated intravascular coagulation. As of day 160 of hospitalization, esophageal reconstruction has not been achieved.

Discussion
All the AEF cases described here occurred after prosthetic aortic replacement for thoracic aortic aneurysm. These fistulae were formed at the contact between the distal anastomotic site of TAR and the middle thoracic esophagus. One possibility is that the esophagus fistula, perhaps caused by ischemic necrosis with compression (that is, with mechanical stimulation or irritation) or operative technique, leaked esophageal contents onto the aortic prosthesis, caused infection there. Alternatively aortic prosthetic infection might have occurred first, perhaps due to caries, hemodialysis, or invasive treatment, and that infection led to the    [11]. In our 4 cases the progress after first surgery and the elapsed times from surgery to onset varied. Identifying the causes of AEF in any consistent manner is difficult. Some papers suggested that combined surgery provided better outcomes for some patients of secondary AEF [7][8][9] (Table 2). What they suggested in common is that the first treatment of secondary AEF is infection control, with esophagectomy being the first stage. The subsequent timing of second-and later-stage surgery must be judged by a comprehensive assessment of the degree of improvement with infection control and nutritional status. There is the question of whether to replace an infected prosthetic graft. There have been reports of an infective thoracic aortic aneurysm and artificial stent graft being removed and the aorta restored in second-stage surgery [8,9]. These cases involved the first thoracotomy, whereas our cases involved the second one. This is a large difference. Infection may be controlled by esophagectomy and antibiotic treatment, so we think that prosthesis replacement is not always necessary. However, experience has shown that where a prosthetic graft meets the native aorta, infection brings a real danger of aneurysm. In case 2, TEVAR was done in the second stage as a precaution, because the native aorta remained at the infected site.
In case 1, 2, and 3, a gastric tube was used to replace the esophagus, and it was implanted in an antethoracic route. One reason for this choice was the fact that anastomotic failure and cutaneous fistula formation occurred in all of 3 cases who underwent esophageal reconstruction. In the cases of AEF after prosthetic replacement, the upper or middle part of the esophagus is damaged, and the esophagostoma to be formed in the first stage is on the clavicle. The third-stage reconstruction has to bridge an unusually great distance, and we suspect that resulting mechanical stress and inadequate blood-flow led to anastomotic failure. Reconstruction methods need to be considered, for example selecting colon reconstruction.

Conclusion
We reported the clinical features and outcomes of treatment of secondary AEF after prosthetic aortic replacement in 4 cases. Based on our experience, the approach involving early diagnosis, disruption of the infected fistula by esophagectomy, treatment of prosthetic graft infection, prevention of possible pseudoaneurysm formation, and esophageal reconstruction, appears to be a valid therapeutic strategy for secondary AEF.

Funding
The authors declare there are no funding resources for this paper.

Ethical approval
Institutional review board approval was not required because all data were collected from clinical records and imaging systems for routine preoperative planning and follow-up.

Consent
Written informed consent was obtained from all of the patients for publication of this case report and accompanying images. A copy of the written consent is available for review by the Editor-in-Chief of this journal on request.
Author contribution ME wrote the manuscript. TK, NT, and TS supervised writing the manuscript. All authors were part of the surgical team that treated these patients. All authors read and approved submission of the final manuscript.

Registration of research studies
We have registered our research at http://www. researchregistry.com. The unique identifying number of our study is "researchregistry5919".

Provenance and peer review
Not commissioned, externally peer-reviewed.

Declaration of Competing Interest
The authors report no declarations of interest. from funding agencies in the public, commercial, or not-for-profit sectors.