Successful Surgical Strategy for Graft-esophageal Fistula After Total Arch Replacement : A Case Report.


 BackgroundGraft infection is one of the catastrophic complications in the aortic surgical area. Case presentationA 62-year-old man was diagnosed graft-esophageal fistula after total arch replacement. We have experienced aggressive debridement, that includes the infected previous graft and unhealthy esophagus under the same operative field, and staged esophageal reconstruction was performed 4 months later. ConclusionsWe believe that this successful surgical strategy, which involves aggressive debridement using re-median incision plus a left posterolateral thoracotomy under the same operative field, is useful and safe and achieved a favorable result.


Background
Graft infection is one of the catastrophic complications in the aortic surgical area.

Case presentation
A 62-year-old man was diagnosed graft-esophageal stula after total arch replacement. We have experienced aggressive debridement, that includes the infected previous graft and unhealthy esophagus under the same operative eld, and staged esophageal reconstruction was performed 4 months later.

Conclusions
We believe that this successful surgical strategy, which involves aggressive debridement using re-median incision plus a left posterolateral thoracotomy under the same operative eld, is useful and safe and achieved a favorable result.

Background
Graft infection is one of the catastrophic complications in the aortic surgical area (1,2). Another lifethreatening, albeit extremely rare, related complication is secondary graft-esophageal stula (GEF) after total arch replacement (TAR). Additionally, the best therapy for this catastrophic complication remains controversial (3-7).
We have experienced aggressive debridement, that includes the infected previous graft and unhealthy esophagus after total arch replacement under the same operative eld and staged esophageal reconstruction was performed 4 months later; hence, we report our successful surgical strategy and approach for severe graft infection with esophageal stula.

Case Presentation
A 62-year-old man underwent TAR using the elephant trunk technique for acute aortic dissection. One year later, the patient developed dysphagia and computed tomography (CT) revealed the presence of periesophageal hematoma (Fig. 1A, upper). Although the hematoma had resolved by 1 month, intermittent fever occurred 5 months later. Urgent CT demonstrated numerous air bubbles around the graft and uid collection around a distal anastomosis with the esophagus (Fig. 1A, lower). Blood culture revealed the presence of Streptococcus constellatus and the patient was diagnosed with a severe graft infection. Although emergency upper GI endoscopy showed no major hole in the esophagus, we decided to repeat the TAR with esophagectomy because of the patient's esophageal history.
Re-median incision was performed with a small left posterolateral thoracotomy through the third intercostal space. Intraoperative ndings revealed extensive in ammation of the tissue as well as a Page 3/7 malodorous abscess around the original graft and esophagus. The distal anastomosis was detached in one quarter part and the elephant trunk graft was exposed (Fig. 1B). Immediate cardiopulmonary bypass with mild hypothermia was established via the femoral artery and right atrium. A distal aortic cross-clamp was placed on the middle descending aorta to maintain lower body perfusion and the previous proximal graft was clamped. The graft was then opened and antegrade selective cerebral perfusion was begun into each arch vessel. The infected graft was completely resected except for a proximal anastomosis site due to severe adhesion, and the descending aorta was transected distally at the non-infected aorta.
Subsequently, we found a small esophageal erosion at the Th3/4 level with an abscess (Fig. 2A), thus, we immediately required resection of the esophagus and cervical esophagostomy. The intrathoracic esophagus was divided and removed just above the diaphragm. After aggressive debridement, TAR was performed again with a new rifampicin-soaked four-branched Dacron graft (Vascutek Gelweave; Vascutek Ltd., Renfrewshire, UK). The mobilized omentum was translocated into the thoracic cavity, which thoroughly covered the graft. After closure of the chest, the esophagus was exteriorized as a cervical esophagostomy, and a 10-F feeding tube was inserted via gastrostomy.
Four months after the operation, supra-sternal jejunal bypass was successfully performed (8). At the 3- year follow-up, the patient showed no signs of infection without antibiotic therapy (Fig. 2B).

Discussion And Conclusions
Several options for aorto-esophageal connection disease have recently been described in the literature (9,10). Bridged thoracic endovascular aneurysm repair (TEVAR) enables rapid control of bleeding and is less invasive than traditional surgery. However, TEVAR alone for GEF leaves the infected graft and esophageal defects untreated. Under this concept, de nitive surgical treatment involving aggressive debridement and reconstruction in situ is the only effective method, as in the case of our patient.
In terms of the approach to GEF after TAR, an extended left thoracotomy incision with or without sternal transaction has been reported (4,5). These approaches are useful to expose the healthy descending aorta. However, they complicate the simultaneous performance of aggressive debridement that includes the infected graft and unhealthy esophagus. Otherwise, the re-median incision and left-sided small thoracotomy has two merits. First, aggressive debridement that includes esophagostomy can be performed under the same operative eld. Second, emergency management is now possible through this approach, as in our patient, who had a detached distal anastomosis. We believe that re-median incision and left-sided small thoracotomy would lead to a favorable result.
It remains unclear when and how the esophageal stula should be treated in this co-infection situation. Some recent work suggested that small esophageal defects can sometimes be treated simply by direct repair or via a covered omental ap. Our patient's esophageal defect was small but resulted in mediastinal infection. Once such a serious co-infection is diagnosed, the primary focus of any surgical procedure must be to control the bacterial infection. Therefore, we resected the esophagus and cervical esophagostomy.