Technical Approach to Repair of Tracheogastric Conduit Fistula Following Minimally Invasive Esophagectomy: A Case Report

Introduction

Tracheogastric conduit fistula (TGCF) is a rare but severe complication following esophagectomy [1]. It typically arises in the context of anastomotic leakage, ischemia, or poor tissue healing [2,3]. Although various treatment modalities, including conservative management, endoscopic therapy, and surgical intervention, have been reported, there is currently no standardized approach, and outcomes remain suboptimal [4,5]. We report a case of delayed TGCF that was managed with primary closure and reinforcement using an intercostal muscle flap, along with adjuncts to enhance sealing and support. This case highlights the technical strategy and perioperative considerations in repairing tracheogastric conduit fistula after chemoradiotherapy, emphasizing decision-making in a high-risk elderly patient.

Case Report

A 78-year-old man with diabetes, hypertension, and a heavy smoking history presented with progressive dysphagia and weight loss. He was diagnosed with Siewert type I esophagogastric junction adenocarcinoma (cT3N+M0) and received neoadjuvant chemoradiotherapy. Restaging showed partial regression without metastasis. Thoracolaparoscopic McKeown esophagectomy was performed, lasting 10 hours, with 150 mL blood loss. After that, a cervical anastomotic leak was identified on POD 7. The initial cervical anastomotic leak was managed conservatively due to the absence of systemic sepsis and the patient’s stable clinical status. Management included close monitoring, intravenous antibiotics, and percutaneous drainage when a localized collection was later identified. However, delayed drainage likely allowed the formation of a cervical abscess that subsequently eroded into the posterior tracheal wall, resulting in the formation of the tracheogastric fistula. The cervical anastomotic leak was completely resolved 7 weeks postoperatively. Despite transient recovery, persistent coughing led to esophagography on POD 54, which demonstrated contrast leakage into the airway (Figure 1A). Endoscopy revealed a fistulous tract with visible endotracheal tube (Figure 1B) and an opening approximately 20 cm from the incisors. A chest CT confirmed a tract between the gastric conduit and trachea (Figure 1C). The patient was reintubated, and definitive surgical repair was performed via right thoracotomy. This case is reported to highlight a rare but severe postoperative complication of esophagectomy and the technical approach to its management.

Under general anesthesia with endobronchial blockade, the patient was positioned in the left lateral decubitus position. A right fifth intercostal thoracotomy was performed. Dense pleural adhesions were lysed, and empyema was debrided. The gastric conduit was mobilized from the posterior trachea, exposing a 1-cm membranous tracheal defect and a 2.5-cm gastric defect (Figure 2A). The tracheal defect was closed with interrupted sutures and leak-tested. The gastric defect was repaired in 2 layers using interrupted sutures. A vascularized intercostal muscle flap was harvested and interposed between the repair sites (Figure 2B, 2C). Fibrin sealant and bovine pericardial patch were applied to reinforce both closures completely. The tracheal defect was closed with interrupted sutures and leak-tested under saline immersion at an airway pressure of 25 cmH2O, with integrity confirmed by intraoperative bronchoscopy. Although bronchoscopy confirmed initial repair success, the patient experienced recurrent pneumonia, septic shock, and ultimately died from multiorgan failure on POD 108. A summary of postoperative events is presented in Table 1.

Discussion

In this case, the development of TGCF occurred several weeks after an initially localized cervical anastomotic leak, highlighting the delayed and insidious nature of this complication in select patients. Tracheogastric conduit fistula (TGCF) is an uncommon yet formidable complication of esophagectomy, occurring in only 0.3% to 1.9% of contemporary series but associated with mortality rates exceeding 40% [1,2]. Fistulas generally manifest between the second and eighth postoperative weeks, typically following a cervical anastomotic leak or unresolved mediastinal sepsis that creates a persistently inflamed environment [2].

In our patient, the tracheogastric conduit fistula developed as a delayed sequela of the initial cervical anastomotic leak. The early cervical leak likely resulted from localized ischemia at the anastomotic site caused by excessive tension and compromised perfusion of the upper gastric conduit after chemoradiotherapy. Although the anastomotic leakage was initially managed conservatively and appeared to heal, persistent local ischemia and chronic infection around the anastomotic region may have led to progressive tissue necrosis and secondary erosion into the posterior tracheal wall, ultimately forming a tracheogastric fistula approximately 2 cm below the anastomosis. This pathogenesis explains the delayed presentation of the fistula and its anatomical separation from the original anastomotic site.

TGCF arises from a convergence of patient-, disease-, and treatment-related factors. Recognized risks include compromised conduit perfusion due to excessive length, tension, or kinking; neoadjuvant chemoradiotherapy; diabetes; malnutrition; prolonged mechanical ventilation; mediastinal infection; extensive tracheobronchial devascularization during lymphadenectomy; and a high cervical anastomosis [3,4]. The use of systemic steroids and delayed recognition of minor leaks further compound tissue ischemia and inflammation, setting the stage for fistula formation. Indocyanine green (ICG) fluorescence angiography has been shown to aid intraoperative assessment of gastric conduit perfusion and can help reduce the risk of ischemia-related complications [3].

Current management begins with differentiating the size and clinical impact of the fistula. Small (<10 mm) or early fistulas without systemic sepsis can be managed non-operatively using bowel rest, broad-spectrum antibiotics, jejunostomy feeding, and 1 or more endoscopic adjuncts such as fully covered self-expanding metallic stents, over-the-scope clips, fibrin or cyanoacrylate glues, or negative-pressure endoluminal therapy. Reported closure rates range from 50% to 70% [1,4,5].

Larger defects, airway necrosis, or persistent sepsis necessitate surgical repair. In such cases, primary tracheal and gastric closure, wide debridement, and vascularized tissue interposition are essential. These principles are well documented in the surgical literature addressing both tracheogastric and tracheoesophageal fistulas [6,7]. The operative technique of fistula repair typically involves 3 critical maneuvers: (1) Tension-free interrupted 4-0 Prolene sutures with intraoperative leak testing for the tracheal defect; (2) Interposition of a pedicled intercostal muscle flap to restore vascularity and establish a physical barrier between the airway and the conduit; and (3) Biological reinforcement using fibrin sealant and a bovine pericardial patch. Comparable approaches have been described, validating the layered strategy employed in our case.

Decortication was also performed intraoperatively to remove inflammatory peel and allow re-expansion of the affected lung. This step is critical when pleural involvement or chronic inflammation is present, as it facilitates better postoperative recovery and reduces the risk of persistent pleural sepsis.

In our case, defunctioning with esophagostomy and delayed reconstruction was carefully considered. The decision against defunctioning was based on the patient’s overall stability at the time of surgical intervention, the localized nature of the tracheogastric fistula, and the absence of extensive necrosis or gross contamination. The surgical team determined that primary repair with vascularized muscle interposition provided a feasible and definitive solution, avoiding the morbidity associated with staged reconstruction in an elderly and comorbid patient. However, defunctioning remains a valid alternative in cases with diffuse contamination, necrosis, or poor tissue quality. The choice of pedicled intercostal muscle flap over bulkier options such as latissimus dorsi was influenced by multiple factors. Intercostal flaps offer reliable vascularity, adequate volume for small-to-moderate defects, and minimal donor-site morbidity. In our patient, a less invasive approach was preferred given his frailty, comorbidities, and the localized nature of the defect. The flap provided sufficient coverage and vascular support while avoiding the added operative trauma associated with harvesting larger muscle flaps.

In retrospect, the delayed drainage of the cervical anastomotic leak may have contributed to the development of the tracheogastric conduit fistula (TGCF) in this patient. Although initially stable, the persistence of a localized cervical collection allowed gradual erosion into the posterior tracheal wall, resulting in fistula formation several weeks postoperatively. This progression underscores the potential for seemingly localized cervical leaks to evolve into intrathoracic airway injuries if not addressed in a timely manner. Early surgical exploration should be considered in cases with unresolved cervical infection despite conservative management.

Several intraoperative and postoperative strategies can help reduce TGCF incidence, particularly in high-risk patients. Intraoperative ICG fluorescence angiography has shown promise for assessing gastric conduit perfusion and preventing ischemic complications [3]. In patients with tenuous anastomoses, prophylactic wrapping with omental or muscle flaps can offer additional protection, especially in cervical locations. During mediastinal dissection, careful preservation of the tracheobronchial arterial supply is crucial to avoid devascularization of the posterior trachea, a known contributor to airway breakdown.

Postoperatively, in patients presenting with persistent cough or subtle respiratory symptoms, early diagnostic bronchoscopy should be considered to identify subclinical airway injury. In frail or high-risk patients for whom operative morbidity is prohibitive, non-operative strategies such as tracheal stenting may be appropriate. A study demonstrated the feasibility of temporary airway stenting as a palliative or bridging strategy in select cases, although the durability of such interventions remains limited [5].

Multidisciplinary postoperative surveillance, including respiratory therapy, early infection control, and a low threshold for repeat imaging, is essential for timely diagnosis and intervention. Novel technologies continue to evolve, including vacuum-assisted endoluminal closure devices tailored for airway–conduit fistulas [8], robot-assisted thoracoscopic surgery for precise intrathoracic reconstruction [9], and bioengineered tissue patches designed to promote healing. To advance the field, multicenter registries are needed to better define TGCF incidence, facilitate risk stratification, and establish benchmarks for evaluating both conventional and emerging treatment modalities [1,4].

Conclusions

This case illustrates the complex surgical decision-making required to manage tracheogastric conduit fistula (TGCF) in high-risk patients. Successful repair hinges on early recognition, timely intervention, and individualized reconstructive strategies. The intercostal muscle flap provided effective tissue coverage with minimal morbidity in this frail patient. Clinicians should maintain a high index of suspicion for TGCF in post-esophagectomy patients with unexplained respiratory symptoms. This case reveals the importance of prompt management of cervical leaks and the value of vascularized muscle flaps in repairing complex airway–conduit fistulas. Further multicenter studies are needed to establish optimal treatment algorithms.