Salvage Repair Using an Esophagus–Conduit Segment In Situ as a Neo-Posterior Tracheal Wall for Multiple Tracheoesophageal and Trachea–Conduit Fistulas: A Case Report Authors: Ying Jie Cui, Hyunjin Cho Published: 2026 ABSTRACT AbstractTracheoesophageal fistula (TEF) and trachea–conduit fistula (TCF) after esophagectomy are rare but life-threatening complications. Large or multiple fistulas are difficult to treat when standard resection and tracheal reconstruction are not feasible. A 70-year-old man developed an anastomotic stricture after Ivor–Lewis esophagectomy for squamous cell carcinoma and underwent repeated balloon dilatations and stent placement. Progressive tissue injury resulted in 1 TCF and 2 large TEFs. Salvage surgery included stent removal, distal conduit takedown, proximal conduit closure, substernal colon interposition, and esophagocolostomy. The esophagus–proximal gastric conduit segment containing the fistulas was preserved in situ as a neo-posterior tracheal wall. Oral intake was resumed after esophagography on postoperative day 6, and bronchoscopy confirmed a viable blind pouch. Preserving the esophagus–conduit segment in situ may provide effective airway–alimentary separation in selected patients with extensive multifocal TEF/TCF. In highly selected patients, this salvage option for airway–alimentary separation when standard resection and reconstruction carry prohibitive risk. Given the limited survival in this case, it is not intended as a definitive or broadly generalizable reconstruction.Keywords: Esophagus, Salvage therapy, Tracheoesophageal fistula, Trachea–conduit fistula, Case reports Case report A 70-year-old man with a history of Ivor–Lewis esophagectomy for esophageal squamous cell carcinoma (pT2N1M0) presented 4 months postoperatively with progressive dysphagia. Esophagography demonstrated an anastomotic stricture (Fig. 1A, B), and multiple endoscopic balloon dilatation sessions were performed at 2-week intervals. One month later, he developed a severe cough and worsening dysphagia. Bronchoscopy confirmed a trachea–conduit fistula (TCF) (Fig. 1A, B), and covered esophageal stent placement was performed after the patient declined surgery. Sequential development of trachea–conduit and tracheoesophageal fistulas. (A, B) Esophagography shows a post-esophagectomy anastomotic stricture (arrow). Bronchoscopy shows a trachea–conduit fistula (arrow). (C, D) Chest computed tomography (axial and sagittal views) demonstrates a newly developed large tracheoesophageal fistula (red arrow) above the upper flare of the inner esophageal stent (red arrowhead). The yellow arrow indicates the second fistula, and the yellow arrowhead indicates the upper flare of the previous esophageal stent. (E, F) Bronchoscopic findings. The upper image shows the second fistula and the upper flare of the previous esophageal stent (yellow arrowhead), and the lower image shows the third fistula and the upper flare of the inner esophageal stent (red arrowhead). (G, H) Three-dimensional volume-rendered images of the trachea demonstrate large tracheal defects from the posterior aspect of the trachea. The upper defect indicates the third fistula (red arrow), and the lower defect indicates the second fistula (yellow arrow). (I, J) Three-dimensional volume-rendered images of the esophagus and gastric conduit demonstrate large defects caused by the upper flares of the esophageal stents. The first fistula is indicated by the arrowhead, the second fistula by the yellow arrow, and the third fistula by the red arrow. Lt, left; Rt, right; T, trachea; A, anterior. The fistula persisted despite stent placement. After concurrent chemoradiotherapy for nodal recurrence, a second fistula developed above the upper flare of the stent; this was identified as a tracheoesophageal fistula (TEF) on computed tomography and bronchoscopy (Fig. 1C–F). An additional stent was placed within the previous stent to cover the lesion. Four months later, the patient’s symptoms worsened again, and another large TEF developed above the upper flare of the inner stent (Fig. 1E, F). Three-dimensional volume-rendered images demonstrated extensive defects involving a neo-posterior tracheal wall (Fig. 1G, H) and the esophagus–gastric conduit (Fig. 1I, J), prompting surgical intervention. Salvage reconstruction was performed using the esophageal segment and proximal gastric conduit, which contained the TEFs and TCF, and were left in situ against the membranous portion of the trachea (Fig. 2A). The surgical procedure was as follows. First, the previously placed stents were removed. Dissection was then performed to the level below the most distal fistula, and the distal conduit was taken down (Fig. 2B, C). The proximal conduit was closed approximately 2-cm above the carina with a continuous suture using 3-0 Prolene. Next, the mid-to-transverse colon, based on the middle colic artery as the feeding vessel, was prepared as a conduit. The esophagus was approached through a left cervical incision. Using a stapler, the proximal esophagus was divided above the most proximal fistula, creating a blind pouch without disturbing the esophageal segment and proximal gastric conduit containing the TEFs and TCF. Alimentary reconstruction was performed by an esophagocolostomy with a 25-mm circular stapler; the proximal esophagus was anastomosed to the colon, which was brought up through the substernal route. Schematic illustration of the surgical planning. (A) Planned salvage reconstruction using the esophagus–conduit segment left in situ. (B) The distal gastric conduit was removed, and the esophageal stent (arrowhead) was exposed from the proximal gastric conduit (arrow). (C) The removed inner (red arrowhead) and outer (yellow arrowhead) esophageal stents. TEF, tracheoesophageal fistula; TCF, trachea–conduit fistula. Lt, left; Rt, right. Postoperatively, the patient’s cough, aspiration symptoms, and respiratory contamination markedly improved. An esophagography on postoperative day (POD) 6 showed no contrast leakage at the esophagocolostomy; however, because of persistent dysphagia, jejunostomy tube feeding was continued while the patient underwent swallowing rehabilitation. Bronchoscopy on POD 7 demonstrated that the esophageal segment was preserved as a blind pouch and had a healthy mucosal appearance (Fig. 3). On POD 17, the patient underwent repair of a laparotomy wound dehiscence. The remainder of the postoperative course was uneventful, with no anastomotic leakage, conduit necrosis, or major airway complication. The patient was discharged on POD 24 on combined oral and jejunostomy tube feeding. At outpatient follow-up, oral intake was maintained alongside supplemental jejunostomy feeding until shortly before death. The patient ultimately died 5 months later because of progression of esophageal cancer. Written informed consent for publication was obtained from the patient. Postoperative bronchoscopic findings and neck computed tomography. (A, B) Bronchoscopy on postoperative day 7. The upper image shows the remnant third fistula (tracheoesophageal fistula [TEF]; red arrowhead), and the lower image shows, near the carina, the openings of the previously documented fistulas into the esophageal blind pouch: the second fistula (TEF; yellow arrowhead) and the first fistula (trachea–conduit fistula; blue arrowhead), the latter becoming clearly visible after stent removal. All 3 openings communicate only with the in-situ esophageal segment that now serves as the neo-posterior tracheal wall and no longer connect the airway to the alimentary stream; they do not represent newly created defects or incomplete repair. (C) Neck computed tomography on postoperative day 7. In the sagittal view, residual fistulas communicating with the esophageal segment left in situ are shown. This segment in situ (yellow star) served as the neo-posterior tracheal wall and shown no ischemia or dehiscence at the suture or staple site. Discussion Throughout this report, “residual fistulas” denote the pre- existing tracheal openings that, after esophageal diversion, communicate only with the in-situ esophageal blind pouch and no longer connect the airway to the alimentary stream. TEF and TCF after esophagectomy are rare but potentially fatal complications. Esophageal stents may provide temporary symptom relief, but complications such as migration, mechanical trauma, and secondary infection can exacerbate tissue injury. In this case, repeated metallic stent placement and local inflammation likely contributed to progressive tissue destruction and the formation of multiple fistulas. Large or persistent fistulas usually require surgery, and recent studies suggest that surgical repair of TEF or TCF is associated with better survival than nonsurgical management [1]. Definitive surgery for TEF or TCF usually requires resection of the esophagus and gastric conduit with tracheal repair; however, these procedures are invasive and technically demanding and are associated with high morbidity. Tracheal reconstruction is particularly difficult when fistulas are extensive. In our case, multiple TEFs involved nearly half of the trachea. We therefore performed salvage repair using the esophagus left in situ as the neo-posterior tracheal wall. This approach avoided extensive resection and dissection of the densely adherent fistulous area while achieving functional separation of the airway and alimentary tract. More broadly, this approach should be considered only in highly selected patients in whom standard options are unfavorable: when the fistulous complex is extensive or multifocal, when dense inflammatory adhesions preclude safe dissection of the fistula tract, and when the patient's overall condition or limited oncologic prognosis makes extensive resection with formal tracheal reconstruction prohibitively morbid. In such settings, primary fistula resection, tracheal repair, or muscle-flap interposition may carry excessive risk, and repeat stenting has already failed or perpetuated tissue injury, as in this patient. Conversely, the technique may be inappropriate or contraindicated when the retained esophagus–conduit segment shows inadequate perfusion (poor mucosal bleeding, ischemic discoloration), when there is uncontrolled mediastinal or airway infection, when viable tumor invades the fistula site (precluding a clean blind pouch and risking recurrent fistulization), or when secure alimentary diversion cannot be achieved. In these circumstances, retaining the in-situ segment risks necrosis, persistent contamination, or tumor-related breakdown, and alternative strategies should be pursued. The distal portion of the gastric conduit had already been lost in this case, and the major feeding vessel—the right gastroepiploic artery—had been sacrificed during takedown of the distal conduit. This raised concern about whether perfusion to the remaining proximal gastric conduit was adequate. During the procedure, we first resected the distal conduit and its feeding artery, removed the esophageal stent, and then dissected the conduit up to the fistulous area to assess mucosal bleeding. After confirming adequate mucosa perfusion, we repaired the remaining proximal gastric conduit. Viability of the retained segment was judged primarily on objective findings: brisk mucosal bleeding at the dissection margin intraoperatively, a fresh, well-perfused mucosal color on bronchoscopy at POD 7, and postoperative CT showing an intact segment without ischemia or suture/staple-line dehiscence. We speculate that chronic inflammation-associated neovascularization may have contributed to this preserved perfusion. Although indocyanine green fluorescence imaging was not used in this case, it may be a useful adjunct for assessing conduit perfusion in similar situations. To our knowledge, only 4 cases in 2 previous reports have described the use of an esophageal segment left in situ as a neo-posterior tracheal wall for the repair of large TEFs. One case involved a TEF caused by prolonged endotracheal intubation [2], and the other 3 involved trauma-related TEFs, including fistulas following traffic accidents; all were repaired successfully [3]. Compared with these cases, the present case was more challenging because the fistulas developed in the setting of severe inflammation due to long-term instrumentation after esophageal surgery and involved not only the esophagus but also the gastric conduit. Nevertheless, functional recovery was achieved postoperatively. In conclusion, this case demonstrates that salvage repair using an esophagus-conduit segment left in situ can provide effective separation of the airway and alimentary tract in patients with extensive TEF and/or TCF. Although technically challenging, this approach may be a palliative, short-term salvage alternative in carefully selected patients when conventional resection and reconstruction carry prohibitive risk. Longer-term durability could not be assessed because the patient died of cancer progression 5 months postoperatively, and broader applicability requires further experience. Article information Author contributions Conceptualization: YJC. Data curation: YJC. Surgery: HJC. Supervision: HJC. Critical revision of the manuscript for important intellectual content: HJC. Writing–original draft: YJC. Writing–review & editing: HJC. Final approval of the manuscript: all authors. Conflict of interest No potential conflict of interest relevant to this article was reported. Funding This work supported by Chungnam National University Hospital Research Fund (2019-CF-006). References