Abstract
Objective
Aortoesophageal fistula (AoEF) is a rare but complex problem that carries high mortality. This study describes our institutional experience with the management of AoEF.
Methods
There were 17 patients with AoEF who were managed in our center (2005-2023). Medical records were reviewed for baseline characteristics, history of esophageal or aortic disease, diagnostic evaluation, surgical treatment, and follow-up. Overall survival (OS) was analyzed using the Kaplan-Meier method and log rank tests.
Results
Most patients had previous aortic operations (12/17, 71%)—7 thoracic endovascular aortic repair (TEVAR), 3 open/TEVAR, and 2 open repairs. The etiology of AoEF was aortic disease in 13 (76%) and anastomotic leak after esophagectomy in 4 (24%). Upon presentation, 2 (12%) patients were not offered intervention, whereas 6 (35%) had emergency TEVAR to control bleeding. Twelve (67%) patients were treated with curative intent, 1 patient did not survive an initial aortic operation. Of the remaining 11 patients, esophageal operations happened before aortic repair in 7 (64%), concurrently in 3 (27%), and after in 1 (9%). This included diversion esophagectomy in 7 (64%) and primary repair in 3 (27%). Definitive aortic surgery included aorta replacement with rifampin-soaked grafts in 8 (67%) and homografts in 4 (33%). In-hospital mortality occurred in 1 of 12 (8%) cases. One- and 2-year OS was 83% and 74%. OS was lower in patients not receiving curative-intent treatment (P < .001).
Conclusions
Management of AoEF is complex. However, aggressive multidisciplinary intervention with definitive esophageal and aortic repairs can result in good long-term survival in selected patients.
Key Words: aortoesophageal fistula, thoracic endovascular aortic repair, esophageal fistula repair
Aggressive multidisciplinary treatment of AoEF can lead to good long-term survival rates.
Central Message.
Aggressive multidisciplinary surgical management of aortoesophageal fistula can lead to long-term survival.
Perspective.
Aortoesophageal fistula is a rare but complex and lethal condition. In our experience, aggressive multidisciplinary surgical treatment, preferentially in a staged fashion, was associated with good long-term survival.
Aortoesophageal fistula (AoEF) is a rare but highly lethal condition characterized by communication between the aorta and the esophagus,1 which can arise from erosion of the native aorta into the esophagus or secondary to complications from surgical interventions to either organ. The classic clinical presentation includes midthoracic pain or dysphagia, sentinel episode of hematemesis followed by a symptom-free interval, and then subsequent massive hematemesis (Chiari triad).1 In addition to life-threatening bleeding, patients with AoEF are also at risk of infection of implanted vascular grafts, leading to sepsis.
The pathophysiology of AoEF after aortic interventions is not fully elucidated. Proposed theories include the erosion of pulsating prosthetic materials in a mobile segment of aorta against the neighboring esophagus; the coverage of perforator esophageal vessels leading to segmental esophageal ischemia and erosion; graft infection and associated aortitis and abscess eroding into the esophagus; or development of pseudoaneurysms, penetrating aortic ulcers, or intramural hematoma, which can also erode the esophageal wall.2,3 Thoracic endovascular aortic repair (TEVAR) has become standard of care in the management of multiple aortic conditions secondary to favorable morbidity and mortality rates.4 Because of its increased use, clinicians must remain vigilant to detect associated complications early. AoEF can also result from anastomotic leaks after esophagectomy, with esophageal stents potentially also contributing to aortic erosion. Given the rare nature of AoEF, a high index of suspicion is required to make this diagnosis in a timely manner to allow for intervention.
The management of AoEF is challenging and requires prompt multidisciplinary collaboration to address simultaneous competing issues related to bleeding, infection, nutrition, and definitive correction of the problem by addressing the defects both on the aortic and esophageal walls. Previous reports have shown the importance of aggressive surgical therapy and the elevated mortality of patients with this condition, as well as invariable mortality associated with “conservative” management.3,5 Therefore, the aims of this study are to report our institutional experience with aggressive multidisciplinary management of AoEF, and propose a strategic treatment sequence which may allow for better long-term survival.
Methods
Study Population
This study included patients with a diagnosis of AoEF who were treated at Mayo Clinic in Rochester, Minnesota, from January 2005 to December 2023. Medical records were reviewed for demographics, clinical characteristics, comorbidities, history of thoracic operations, aortic disease, and other operations including previous TEVAR, postoperative complications, outcomes, and vital status during the last known follow-up. Study approval and waiver of consent were obtained from the Mayo Clinic Institutional Review Board on April 26, 2024 (no. 24-004094).
Statistical Analysis and Study End Points
Data extracted from the medical records were collected and managed using research electronic data capture (REDCap) tools hosted at Mayo Clinic. Data was analyzed using the BlueSky Statistics software, version 10.3.4 (BlueSky Statistics LLC). Categorical variables are reported as frequencies and proportions. Continuous variables are reported as median with interquartile range. Overall survival was defined as the time in years from the date of first intervention to the date of death from any cause or last known follow up alive. Time-to-event analyses were performed using the Kaplan-Meier method with log-rank tests for comparisons. Multivariable analyses were not performed due to the limited sample size.
Results
Clinical Characteristics
Twenty-five patients were identified with a diagnosis of AoEF. Eight patients were excluded (7 patients only had an electronic consultation, and 1 had definitive surgery at another institution), resulting in 17 patients included for analysis. Baseline patient characteristics are shown in Table 1. Median age at presentation was 65 years, and most patients were men (14/17, 82%). Previous aortic disease was encountered in 13 (76%) patients. Previous aortic surgery was reported by 12 of 17 (71%) patients. Previous TEVAR was observed in 10 of 17 (59%) patients. Esophageal stents had been used in 3 of 17 (18%) patients to manage anastomotic leaks after esophagectomy.
Table 1.
Demographics and clinical characteristics of 17 patients with aortoesophageal fistula
| Variable | Result |
|---|---|
| Age, y, median (IQR) | 65 (63, 77) |
| Sex, n (%) | |
| Female | 3 (18) |
| Male | 14 (82) |
| Race, n (%) | |
| White | 16 (94) |
| African American | 1 (6) |
| Body mass index, kg/m2, median (IQR) | 24.7 (21.6, 27.5) |
| Comorbidities, n (%) | |
| Hypertension | 13 (77) |
| Diabetes mellitus | 2 (12) |
| Hyperlipidemia | 10 (59) |
| Chronic obstructive pulmonary disease | 2 (13) |
| Cerebrovascular disease | 1 (6) |
| Peripheral vascular disease | 8 (47) |
| Myocardial infarction | 2 (12) |
| Atrial fibrillation | 3 (18) |
| Lung cancer | 1 (6) |
| Chest radiation | 4 (24) |
| History of esophageal disease, n (%) | |
| Esophageal cancer | 4 (24) |
| History of aortic repair, n (%) | |
| Open surgical repair only | 2 (12) |
| Endovascular repair (TEVAR) | 10 (59) |
| Open + endovascular repair | 3 (17) |
| Endovascular repair only | 7 (41) |
| Etiology of AoEF, n (%) | |
| Aneurysms, pseudoaneurysms, contained rupture | 9 (53) |
| Managed with TEVAR | 6 (67) |
| Aortic dissection, managed with TEVAR | 4 (24) |
| Chronic esophageal leak | 4 (24) |
| Symptoms | |
| Chest pain | 13 (77) |
| Hematemesis | 10 (59) |
| Fever | 7 (41) |
| Back pain | 6 (33) |
| Dysphagia | 4 (24) |
| Shortness of breath | 4 (24) |
| Hemoptysis | 4 (24) |
IQR, Interquartile range; TEVAR, thoracic endovascular aortic repair; AoEF, aortoesophageal fistula.
The most common symptoms at presentation were chest pain in 13 of 17 (77%) patients, hematemesis in 10 of 17 (59%), and fever in 7 of 17 (41%) (Table 1). Blood cultures were obtained initially in 11 of 17 (65%) patients, with 4 of 11 (36%) having positive and 7 of 11 (64%) having negative blood cultures. Isolated micro-organisms included Staphylococcus aureus (n = 2), Fusobacterium nucleatum plus Coxiella burnetii (n = 1), and Streptococcus mitis (n = 1). Diagnostic testing leading to the diagnosis of AoEF included computed tomography (CT) scans in 14 of 17 (82%) patients, esophagogastroscopy (EGD) in 14 of 17 (82%), and contrast esophagram in 3 of 17 (18%) (Figure 1).
Figure 1.
A, Axial CT showing air surrounding a TEVAR graft with direct connection to the esophageal lumen. B, Axial CT showing aortic aneurysm with insinuation of intravenous contrast close to the esophageal lumen. C, Endoscopic image showing erosion of TEVAR graft into esophageal lumen. D, Endoscopic image showing an adherent clot at the esophageal wall. CT, Computed tomography; TEVAR, thoracic endovascular aortic repair.
Characteristics of AoEF
The type of AoEF was primary in 2 of 17 (12%) patients and secondary in 15 of 17 (88%) (Table 2). Two patients had primary AoEF attributable to aortic aneurysm. Secondary AoEF etiologies included aortic and esophageal pathology that required intervention. Aortic pathology requiring intervention was the main etiology in 11 of 15 cases of secondary AoEF and included aortic dissection, aneurysms, pseudoaneurysms, and contained ruptures, with previous TEVAR observed in 10 of 11 cases. Esophageal pathology was the main etiology in 4 of 15 cases of secondary AoEF, represented by chronic anastomotic leaks after esophagectomy for cancer in all cases. All patients had no evidence of malignant disease on presentation.
Table 2.
Baseline characteristics of 17 patients with AoEF and initial treatment approach
| Patient no. | Age/sex | Etiology of AoEF | History of aortic disease | Diagnostic workup | Salvage TEVAR on presentation |
|---|---|---|---|---|---|
| Palliative (n = 5) | |||||
| 1 | 84/M | Ruptured pseudoaneurysm eroding into esophagus | No | CTA, EGD | Yes |
| 4 | 77/M | Aneurysmal degeneration | Yes, previous extent 4 thoracoabdominal aneurysm repair | CTA, EGD | No |
| 5 | 80/M | Leak after esophagectomy for cancer | No | EGD | No |
| 7 | 85/F | Aortic aneurysm managed with TEVAR | Yes, thoracic aortic aneurysm | CTA, EGD | Yes |
| 8 | 33/M | Aortic dissection managed with TEVAR | Yes, thoracic aortic dissection | EGD | Yes |
| Curative (n = 12) | |||||
| 2 | 72/M | Aortic aneurysmal managed with TEVAR | Yes, thoracic aortic aneurysm | CTA | No |
| 3 | 64/M | Pseudoaneurysm after aortic hemiarch replacement | Yes, aortic arch injury during mediastinoscopy | CTA, EGD | No |
| 6 | 63/M | Leak after esophagectomy for cancer | No | CTA | Yes |
| 9 | 78/M | Chronic aortic dissection managed with TEVAR | Yes, type A dissection requiring ascending repair, followed by total arch repair, and subsequent distal arch TEVAR | CTA, EGD | No |
| 10 | 65/F | Leak after esophagectomy for cancer | No | CTA, esophagram, EGD | Yes |
| 11 | 26/M | Aortic dissection managed with TEVAR | Yes, traumatic type B aortic dissection managed with TEVAR | CTA, esophagram, EGD | No |
| 12 | 52/F | Leak after esophagectomy for cancer | No | CTA, EGD | Yes |
| 13 | 64/M | Pseudoaneurysm managed with TEVAR after previous dissection due to trauma | Yes, previous open descending aorta repair after traumatic type B dissection | CTA, esophagram | No |
| 14 | 62/M | Chronic pseudoaneurysm managed with TEVAR | Yes, previous TEVAR for type B dissection | CTA, EGD | No |
| 15 | 64/M | Aortic dissection managed with TEVAR | Yes, previous TEVAR for type B dissection | CTA, EGD | No |
| 16 | 70/M | Aortic aneurysm, contained rupture managed with TEVAR | Yes, TEVAR for contained rupture | CTA, EGD | No |
| 17 | 71/M | Aortic aneurysm, contained rupture managed with TEVAR | Yes, TEVAR for penetrating aortic ulcer | CTA, EGD | No |
AoEF, Aortoesophageal fistula; TEVAR, thoracic endovascular aortic repair; M, male; CTA, computed tomography angiography; EGD, esophagogastroduodenoscopy; F, female.
Treatment of AoEF: Selection of Treatment Strategies
Surgical treatment was not offered to 2 of 17 (12%) patients because of estimated prohibitive surgical risk. One had poor performance status as the result of complications after esophagectomy for cancer, and the second had multiple comorbidities, calcified thoracic aorta, and anticipated need for prolonged deep hypothermic circulatory arrest to repair the aorta. Of the remaining 15 patients, salvage TEVAR at presentation was performed in 6 (40%) patients to control bleeding (Table 2). To note, TEVAR in this setting was not considered definitive therapy. Of these, 1 patient died intraoperatively and 2 declined or were not offered further intervention and management focused on palliative care. In total, 5 of 17 (29%) patients were managed with palliative focused therapy, whereas the remaining 12 of 17 (71%) were managed with a curative intention with attempted definitive surgical correction of the AoEF (Figure 2). Broad-spectrum antibiotics were administered to all patients managed curatively. Decisions on surgical management were made by the treating physicians after multidisciplinary discussion and shared decision-making with the patients. There were no prespecified criteria used for patient selection.
Figure 2.
Summary of management. AoEF, Aortoesophageal fistula; TEVAR, thoracic endovascular aortic repair.
Of 12 patients who proceeded with definitive corrective surgery, 3 of 12 (25%) had simultaneous aortic and esophageal repairs, whereas 8 of 12 (67%) had a staged approach, with 7 of 12 (58%) having the esophagus addressed first and the aorta second (Figure 2, Table 3). One patient who received aortic repair first did not survive to allow esophageal repair. This case involved emergent redo aortic arch replacement because of a contained rupture of an infected aortic arch graft via median sternotomy. Intraoperative assessment for possible simultaneous esophageal repair took place; however, because of a challenging exposure through a median sternotomy and worrisome intraoperative encephalography changes during circulatory arrest, esophageal repair was deferred and ultimately the patient did not survive a postoperative stroke.
Table 3.
Summary of definitive curative-intent treatment in 12 patients with aortoesophageal fistula
| Patient no., age/sex | Esophageal management | Aortic management | Treatment sequence | In-hospital mortality | GI tract reconstruction | Long-term outcome |
|---|---|---|---|---|---|---|
| Aortic repair only | ||||||
| 3, 64/M | No repair performed only intraoperative assessment | Redo sternotomy, total arch repair | Not applicable | Yes | N/A | In-hospital mortality (stroke) |
| Simultaneous esophageal and aortic repairs | ||||||
| 2, 72/M | Primary repair | Descending aorta repair with rifampin-soaked graft via left thoracotomy | Same operation | No | N/A | Alive after 165 mo |
| 9, 78/M | Primary repair | Descending aorta repair with rifampin-soaked graft via left thoracotomy | Same operation | No | N/A | Alive after 63 mo |
| 12, 52/F | Primary repair | Descending aorta repair with cryopreserved graft via left thoracotomy | Same operation | No | N/A | Alive after 40 mo |
| Staged esophageal and aortic repairs | ||||||
| 6, 63/M | Esophageal diversion, resection of esophagogastric anastomosis | Descending aorta repair (rifampin-soaked Dacron); left thoracotomy | Esophageal diversion → Aortic repair | No | No; died before reconstruction | Died after 23 mo |
| 10, 65/F | Esophageal diversion, resection of esophagogastric anastomosis | Descending aorta repair (rifampin-soaked Dacron); left thoracotomy | Esophageal diversion → Aortic repair | No | No; died before reconstruction | Died due to complications (respiratory failure) after 3 mo |
| 11, 26/M | Failed initial primary repair at time of aortic repair at outside institution; followed by 3-field esophagectomy | Descending aorta repair (cryopreserved homograft); left thoracotomy | Aortic repair → esophageal reconstruction | No | Single-stage 3-field esophagectomy | Alive after 41 mo |
| 13, 64/M | Diversion esophagectomy | Descending aorta repair (cryopreserved homograft); left thoracotomy | Esophageal diversion → Aortic repair | No | Yes; 4 mo after initial operation | Alive after 34 mo |
| 14, 62/M | Diversion esophagectomy | Descending aorta repair (rifampin-soaked Dacron); left thoracotomy | Esophageal diversion → Aortic repair | No | Yes; 19 mo after initial operation | Alive after 37 mo |
| 15, 64/M | Diversion esophagectomy | Descending aorta repair (rifampin-soaked Dacron); left thoracotomy | Esophageal diversion → aortic repair | No | Yes; 11 mo after initial operation. | Alive after 36 mo |
| 16, 70/M | Diversion esophagectomy | Descending aorta repair (rifampin-soaked Dacron); left thoracotomy | Esophageal diversion → aortic repair | No | Awaiting reconstruction | Alive after 26 mo |
| 17, 71/M | Diversionesophagectomy | Descending aorta repair (rifampin-soaked Dacron); left thoracotomy | Esophageal diversion → aortic repair | No | Yes; 13 mo after initial operation | Alive after 22 mo |
GI, Gastrointestinal; M, male; N/A, not applicable; F, female.
AoEF: Addressing the Esophagus
In the 8 cases receiving a staged definitive approach, this was done at a median time of 5 days (interquartile range [IQR], 4, 18 days) from presentation. In all cases, the esophagus was addressed with an esophagectomy. Median time from presentation to esophagectomy was 4 days (IQR, 4, 5 days). Surgical approach for esophagectomy included right thoracotomy and video-assisted thoracic surgery in 4 patients each. Enteral access for nutrition was obtained by gastrostomy tube in 6 and jejunostomy in 2. Postoperative complications after esophagectomy included respiratory failure requiring tracheostomy in 3 (38%) cases, pulmonary embolism in 1 (13%), and pneumonia in 1 (13%) (Table E1).
Primary esophageal repair was performed in 3 patients at the same time as aortic repair with a layered closure through the left thoracotomy exposure. A tissue flap from either a serratus muscle, latissimus dorsi, intercostal muscle, and/or pleural flap was used to separate the descending aortic graft from the repaired esophagus in all cases. The decision to perform primary repair was considered on a case-by-case basis after assessing the degree of involvement intraoperatively.
AoEF: Addressing the Aorta
Definitive aortic surgical repair was performed in 12 of 17 (71%) patients. The median time from presentation to aortic surgery was 26 days (IQR, 8, 48 days). Among patients who had aortic repair before esophageal repair or simultaneously, the median time from presentation to aortic surgery was 26 days (IQR, 13, 34 days); however, their time from presentation to first intervention, including salvage TEVAR, was 5 days (IQR, 4, 7 days). Among patients who had a staged approach with esophagectomy first, the median time from presentation to definitive aortic surgery was 55 days (IQR, 5, 150 days).
TEVAR explant was performed in 11 cases, 8 with infected pre-existing endografts and 3 who received salvage TEVAR at presentation. Eleven patients required descending thoracic aorta repair via a left posterolateral thoracotomy, and 1 required total arch repair alone via redo median sternotomy. Aortic reconstruction was completed with rifampin-soaked Dacron grafts in 8 of 12 (67%) patients and cryopreserved homografts in 4 of 12 (33%). Tissue flaps were used to cover the descending aortic repair in 11 of 12 (92%), including pedicled intercostal, latissimus dorsi, and serratus anterior muscle flaps. Additional details on the aortic operations are shown in Table E2. Of 12 patients who received definitive aortic surgery, 1 (8%) patient died in the postoperative period (stroke), whereas 11 (92%) were discharged from the hospital.
Follow-Up and Long-Term Survival
Postoperatively, 7 patients received ≤6 weeks of intravenous antibiotics, whereas 3 received >6 weeks. One patient had their initial antibiotic treatment at an outside institution. Chronic suppressive oral antibiotics were prescribed to all 11 surviving patients.
Of 7 patients who underwent diversion esophagectomy, 4 (57%) underwent esophageal reconstruction with the use of a retrosternal gastric conduit at a median follow-up time of 11 months (IQR, 4, 13 months) (Table 3). Additional details are shown in Table E3. The remaining 3 patients had not undergone re-establishment of gastrointestinal continuity; 2 died during follow-up, and 1 patient who had a stroke after aortic repair is awaiting reconstruction.
Among patients who underwent definitive surgical treatment of AoEF with a curative intent, overall survival estimates at 1, 3, and 5 years were 83%, 74%, and 74%, respectively. This was significantly better when compared with patients who received palliative management, all of whom died within 6 months (P < .001; Figure 3). Among patients treated with curative intent, 3 died during follow-up, all of them as a result of complications associated with AoEF.
Figure 3.
Overall survival according to aortoesophageal fistula treatment (shading: 95% confidence intervals).
Discussion
The present study showed relatively good postoperative results for selected patients with AoEF who are treated aggressively, as well as the poor survival of those not receiving a curative approach. The long-term survival of 74% at 3 and 5 years presented here compares favorably with historical series published on this condition.5, 6, 7, 8 In this contemporary experience, most AoEF were secondary and associated with TEVAR. Even though AoEF is a complex condition, good long-term overall survival can be achieved with aggressive multidisciplinary management. Although these patients are subjected to multiple surgical procedures, the short-term morbidity and mortality is acceptable in the context of the observed long-term results.
Surgeons must have a high index of suspicion when facing patients with a history of aortic or esophageal interventions with chest pain, hematemesis, and fever. Previous studies have shown good sensitivity (40%-90%) of CT angiography to establish the diagnosis of AoEF.9,10 EGD is less sensitive than CT and carries a risk of distal embolization.9 This may not be as life-threatening as for patients with atrioesophageal fistula,11,12 but diagnostic EGDs should only be performed in selected patients who absolutely need it for diagnosis or to rule out an underlying malignancy that could influence management. If performed, EGD should be done with minimal insufflation using CO2 and great care to minimize the risk of distal embolization or clot dislodgement leading to uncontrollable hemorrhage.
Our data support the aggressive management AoEF, with the tenets of treatment that are cardiovascular stabilization, hemorrhage control, control of sepsis, and delayed gastrointestinal reconstruction. Six patients required salvage TEVAR on presentation, which is an important aid to achieve temporary stabilization. It is worth underscoring that although this is a complex disease that requires multidisciplinary collaboration, patients who are otherwise fit should be offered aggressive surgical treatment. A “conservative” approach with only endovascular aortic and endoscopic esophageal interventions does not result in spontaneous healing of the fistula and perpetuates the infectious process. A “conservative” approach is associated with ∼100% mortality.3,13 Furthermore, a systematic review reported nearly double the mortality in patients who underwent TEVAR only, with a 6-month mortality rate of 48.8% versus 20.5% with TEVAR followed by staged surgery.14 Similarly, treatment with an esophageal stent only has resulted in 1-year survival of only 17%.3
Three patients underwent a concomitant primary repair of the esophagus at the time of aortic surgery. None of these patients experienced a leak from their repair. An organ-sparing approach previously reported by Jeon and colleagues7 showed successful primary repair in 5 of 7 patients; however, patients who failed did not survive. More recently, Mills and colleagues5 reported on primary repair for patients with defects <3 cm, with good perioperative outcomes and no instances of fistula or aortic graft infection recurrence, although long-term outcomes were suboptimal with 1-year survival of only 32% and 1 of 12 patients requiring esophagectomy on follow-up.
Given the retrospective nature and small sample size in our study, we cannot clearly clarify the thought process behind “aorta first” or “esophagus first” management strategies, or make meaningful comparisons between approaches. Most patients in our cohort reflect our current preferred treatment strategy. Salvage TEVAR is used as needed to address life-threatening hemorrhage or impending rupture. In this scenario, TEVAR is not considered the definitive solution for the aorta. After stabilization, we favor a staged approach with diversion esophagectomy, followed by aortic repair during the same hospital stay, and delayed gastrointestinal reconstruction once recovered. We favor long-term suppressive antibiotic therapy.
In our experience, this approach permits a minimally invasive diversion esophagectomy as a first step to control the infectious source and allow broad-spectrum antibiotics to effectively decrease the bioburden. Diversion esophagectomy via video-assisted thoracic surgery can decrease complications and allow for a faster progression to definitive aortic repair. Interestingly, patients who had an open esophagectomy were those who developed postoperative respiratory complications; however, we are unable to establish associations, given our small cohort. This approach is in line with that postulated by Yamazato and colleagues,6 wherein esophagectomy is necessary for effective infection control; however, it differs in that we perform a staged approach instead of simultaneous esophagectomy sharing the same surgical field. We make every effort to reduce contamination during aortic replacement to decrease the risk of graft reinfection.
During aortic repair, all patients underwent extensive debridement and in situ reconstruction. The graft used for reconstruction depends on availability and surgeon preference; in our experience, rifampin-soaked Dacron grafts were most common and showed good results. Although morbidity was considerable, in-hospital mortality was relatively low (8%). Although other groups have had discreet success with extra-anatomic bypass,5, 6, 7, 8,15,16 this is not our practice, given the added complexity and morbidity of those operations and nonphysiological flow patterns. In situ replacement of the descending thoracic aorta via left thoracotomy allows for adequate exposure for TEVAR explant, debridement of the infected aortic bed, and buttressing with vascularized tissue flaps. These repairs were covered with a combination of latissimus dorsi, intercostal muscle, and serratus anterior muscle flaps. Although pedicled omental flaps based on the gastroepiploic vascular arcade are a tremendous resource to manage complex intrathoracic problems,17 they should be avoided in this situation to avoid jeopardizing delayed gastrointestinal reconstruction using a retrosternal gastric conduit after diversion esophagectomy.
Limitations
This study represents a single institutional experience with a limited sample size. Results are susceptible to selection bias, and conclusions need careful consideration. To mitigate selection bias, we included consecutive patients including those not offered definitive surgical treatment. Given the small sample size, no association can be established between risk factors and outcomes.
Conclusions
AoEF is a complex condition presenting multiple competing issues related to bleeding, infection, and nutrition. Most cases seen in contemporary practice are associated with a previous TEVAR. In our experience, an aggressive multidisciplinary treatment strategy can lead to good long-term survival rates. This must involve a concerted effort between the vascular surgery, thoracic surgery, and cardiac surgery departments. Salvage TEVAR can be a temporizing maneuver. We favor a staged approach with diversion esophagectomy followed by explant of the infected aortic prosthesis, debridement, in situ replacement of the thoracic aorta, and tissue flap buttress. After a period of recovery to allow for reconditioning and nutritional repletion, patients can undergo elective re-establishment of gastrointestinal continuity.
Conflict of Interest Statement
The authors reported no conflicts of interest.
The Journal policy requires editors and reviewers to disclose conflicts of interest and to decline handling or reviewing manuscripts for which they may have a conflict of interest. The editors and reviewers of this article have no conflicts of interest.
Appendix E1
Table E1.
Surgical management of the esophagus in 11 patients who underwent takedown of esophageal fistula
| Variable | Result |
|---|---|
| Definitive management of esophageal fistula, n (%) | 11 (65) |
| Primary repair | 3 (27) |
| Diversion esophagectomy | 7 (64) |
| Three-field esophagectomy | 1 (9) |
| Enteral nutritional access, n (%) | |
| Gastrostomy | 7 (64) |
| Jejunostomy | 3 (27) |
| No feeding tube | 1 (9) |
| Timing of esophageal repair, n (%) | |
| Before definitive aortic repair | 7 (64) |
| Simultaneous with aortic repair | 3 (27) |
| After aortic repair | 1 (9) |
| Surgical approach, n (%) | |
| Open | 7 (64) |
| Video-assisted thoracic surgery | 4 (36) |
| Complications after esophageal fistula repair, n (%) | |
| Respiratory failure requiring tracheostomy | 2 (27) |
| Recurrent laryngeal nerve injury | 2 (18) |
| Pneumonia | 2 (18) |
| Surgical-site infection | 1 (9) |
| Pulmonary embolism | 1 (9) |
| Bleeding from esophagostomy requiring transfusion | 1 (9) |
| Gastric conduit stricture | 1 (9) |
| Leak | 0 (0) |
| Aortic repair done during same hospital stay, n (%) | 9 (82) |
Table E2.
Surgical management of aorta in 12 patients who received definitive aortic repair
| Variable | Result |
|---|---|
| Time from diagnosis to aortic operation, d, median (IQR) | 26 (8, 47.5) |
| Type of prosthetics used, n (%) | |
| Rifampin-soaked Dacron graft | 8 (53) |
| Cryopreserved homograft | 4 (31) |
| Aortic operation, n (%) | |
| Total aortic arch replacement | 1 (8) |
| Descending thoracic aorta replacement | 11 (92) |
| Approach, n (%) | |
| Median sternotomy | 1 (8) |
| Left thoracotomy | 11 (92) |
| Cardiopulmonary bypass time, min, median (IQR) | 209 (162, 249) |
| Aortic crossclamp used, n (%) | 2 (14) |
| Aortic crossclamp time, min, median (IQR) | 144.5 (117, 172) |
| Circulatory arrest used, n (%) | 8 (67) |
| Circulatory arrest time, min, median (IQR) | 47 (34, 62) |
| Postoperative blood transfusions required, n (%) | 7 (58) |
| Postoperative length of stay, d, median (IQR) | 17.5 (14, 26) |
| Postoperative complications, n (%) | |
| Prolonged ventilation | 6 (50) |
| Atrial fibrillation | 4 (33) |
| Stroke | 3 (25) |
| Respiratory failure requiring tracheostomy | 3 (25) |
| Recurrent laryngeal nerve injury paralysis | 2 (17) |
| Acute kidney failure requiring dialysis | 1 (8) |
| Transitory spinal cord ischemia | 1 (8) |
| Reoperation for bleeding | 0 (0) |
| In-hospital mortality, n (%) | 1 (8) |
IQR, Interquartile range.
Table E3.
Characteristics of delayed gastrointestinal reconstruction in patients who had diversion esophagectomy
| Variable | Result |
|---|---|
| Delayed gastrointestinal reconstruction performed, n (%) | 4 (57%) |
| Time to reconstruction, mo, median (IQR) | 11 (4, 13) |
| Technique and conduit, n (%) | |
| Retrosternal gastric conduit | 5 (100) |
| Complications after esophageal reconstruction, n (%) | |
| Anastomotic leak | 2 (40) |
| Surgical-site infection | 1 (20) |
| Dumping syndrome | 1 (20) |
| Renal failure requiring temporary dialysis | 1 (20) |
| Readmission | 1 (20) |
| Postoperative length of stay, d, median (IQR) | 11 (8.8, 14) |
| In-hospital mortality, n (%) | 0 (0) |
IQR, Interquartile range.
References
- 1.Schoel L.J., Lagisetty K. Aorto-esophageal fistula management. Thorac Surg Clin. 2024;34(4):385–394. doi: 10.1016/j.thorsurg.2024.07.004. [DOI] [PubMed] [Google Scholar]
- 2.Eggebrecht H., Baumgart D., Radecke K., et al. Aortoesophageal fistula secondary to stent-graft repair of the thoracic aorta. J Endovasc Ther. 2004;11(2):161–167. doi: 10.1583/03-1114.1. [DOI] [PubMed] [Google Scholar]
- 3.Czerny M., Eggebrecht H., Sodeck G., et al. New insights regarding the incidence, presentation and treatment options of aorto-oesophageal fistulation after thoracic endovascular aortic repair: the European registry of endovascular aortic repair complications. Eur J Cardiothorac Surg. 2014;45(3):452–457. doi: 10.1093/ejcts/ezt393. [DOI] [PubMed] [Google Scholar]
- 4.MacGillivray T.E., Gleason T.G., Patel H.J., et al. The society of thoracic surgeons/american association for thoracic surgery clinical practice guidelines on the management of type B aortic dissection. Ann Thorac Surg. 2022;113(4):1073–1092. doi: 10.1016/j.jtcvs.2021.11.091. [DOI] [PubMed] [Google Scholar]
- 5.Mills A., Tanaka A., Dawson A., et al. Management of aortoesophageal fistula primarily using esophageal preservation. J Thorac Cardiovasc Surg Open. 2024;19:31–38. doi: 10.1016/j.xjon.2024.04.004. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Yamazato T., Nakamura T., Abe N., et al. Surgical strategy for the treatment of aortoesophageal fistula. J Thorac Cardiovasc Surg. 2018;155(1):32–40. doi: 10.1016/j.jtcvs.2017.09.047. [DOI] [PubMed] [Google Scholar]
- 7.Jeon Y.J., Kim H.K., Kim W.S., Cho J.H., Sung K. Repair of aortoesophageal fistula with homograft aortic replacement and primary esophageal closure. J Thorac Cardiovasc Surg. 2022;163(6):2002–2008. doi: 10.1016/j.jtcvs.2020.06.035. [DOI] [PubMed] [Google Scholar]
- 8.Nakamura T., Yamamoto M., Yamazato T., et al. Surgical strategy of esophageal resection and reconstruction for aortoesophageal fistula. Dis Esophagus. 2017;30(9):1–7. doi: 10.1093/dote/dox077. [DOI] [PubMed] [Google Scholar]
- 9.Baril D.T., Carroccio A., Ellozy S.H., et al. Evolving strategies for the treatment of aortoenteric fistulas. J Vasc Surg. 2006;44(2):250–257. doi: 10.1016/j.jvs.2006.04.031. [DOI] [PubMed] [Google Scholar]
- 10.Vu Q.D.M., Menias C.O., Bhalla S., Peterson C., Wang L.L., Balfe D.M. Aortoenteric fistulas: CT features and potential mimics. Radiographics. 2009;29(1):197–209. doi: 10.1148/rg.291075185. [DOI] [PubMed] [Google Scholar]
- 11.Calkins H., Kuck K.H., Cappato R., et al. 2012 HRS/EHRA/ECAS expert consensus statement on catheter and surgical ablation of atrial fibrillation: recommendations for patient selection, procedural techniques, patient management and follow-up, definitions, endpoints, and research trial design. Europace. 2012;14(4):528–606. doi: 10.1093/europace/eus027. [DOI] [PubMed] [Google Scholar]
- 12.Shen K.R. Commentary: the rising storm of atrioesophageal fistulae after catheter ablation for atrial fibrillation. J Thorac Cardiovasc Surg Tech. 2021;6:169. doi: 10.1016/j.xjtc.2020.12.013. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Chiesa R., Melissano G., Marone E.M., Marrocco-Trischitta M.M., Kahlberg A. Aorto-oesophageal and aortobronchial fistulae following thoracic endovascular aortic repair: a national survey. J Vasc Surg. 2010;51(3):785. doi: 10.1016/j.ejvs.2009.12.007. [DOI] [PubMed] [Google Scholar]
- 14.Wang C., Xi Z., von Segesser L.K., Pozzoli A., Ferrari E. Thoracic endovascular aortic repair for the management of aorto-esophageal fistulae: a systematic review. J Endovasc Ther. Published online November 20, 2024 doi: 10.1177/15266028241300403. Online ahead of print. [DOI] [PubMed] [Google Scholar]
- 15.Madan A.K., Santora T.A., Disesa V.J. Extra-anatomic bypass grafting for aortoesophageal fistula: a logical operation. J Vasc Surg. 2000;32(5):1030–1033. doi: 10.1067/mva.2000.107767. [DOI] [PubMed] [Google Scholar]
- 16.Yonago R.H., Iben A.B., Mark J.B. Aortic bypass in the management of aortoesophageal fistula. Ann Thorac Surg. 1969;7(3):235–237. doi: 10.1016/s0003-4975(10)66178-4. [DOI] [PubMed] [Google Scholar]
- 17.Mathisen D.J., Grillo H.C., Vlahakes G.J., Daggett W.M. The omentum in the management of complicated cardiothoracic problems. J Thorac Cardiovasc Surg. 1988;95(4):677–684. [PubMed] [Google Scholar]