Repair of Tracheal Dehiscence After Heart-Lung Transplant Complicated by SARS-CoV-2 Infection

Abstract

Airway complications are a major cause of morbidity after thoracic transplantation. Airway ischemia, necrosis, and tracheobronchial anastomotic dehiscence are associated with early mortality. We describe a case of tracheal anastomotic dehiscence after en bloc heart-lung transplant complicated by severe acute respiratory syndrome coronavirus 2 infection. Timely surgical management and reconstruction with a bovine pericardial patch and double muscle flap were performed. After 8 months of follow-up, there are no airway complications and normalized allograft function.

The reported incidence of airway complications after lung transplant ranges from 2% to 18%, with an incidence of anastomotic dehiscence between 1% and 10%.^1,2 Because airway dehiscence usually occurs early in the posttransplantation period (1-5 weeks) and is associated with high mortality, early diagnosis and prompt intervention are essential. Whereas consensus guidelines have helped standardize the definitions and grading of airway complications, appropriate management for anastomotic dehiscence depends on the severity of the defect and associated complications. Although a conservative approach with serial bronchoscopy and stents has been described for less severe defects, surgical intervention by various approaches to tracheobronchoplasty are required for larger airway defects and associated comorbidities.² Here, we describe the surgical repair of a tracheal dehiscence after en bloc heart-lung transplant with the use of a bovine pericardial patch and double muscle flap coverage (intercostal and latissimus dorsi) in a patient with COVID-19.

The patient is a United Network for Organ Sharing–listed (ABO phenotype, AB; body mass index, 24 kg/m²; creatinine clearance, 131 mL/min; albumin level, 3.8 g/dL) 49-year-old man with mixed connective tissue disease–associated interstitial lung disease (+ANA, aldolase, anti-SSA), secondary pulmonary hypertension (mean pulmonary arterial pressure, 41 mm Hg at rest; severe tricuspid regurgitation; left ventricular end-diastolic pressure, 4 mm Hg), diffuse 3-vessel nonobstructive coronary artery disease (excepting obtuse marginal 1 >80%, extensive left to right collaterals), and history of necrotizing small bowel injury requiring resection at the age of 11 years. The patient was initially listed for a bilateral lung transplant but presented with acute-on-chronic hypoxic and hypercapnic respiratory failure with progressive cor pulmonale, eventually requiring urgent left femoral venoarterial extracorporeal membrane oxygenation (ECMO) for hemodynamic collapse. After 2 weeks of ambulatory venoarterial ECMO support, he underwent unremarkable en bloc heart-lung transplantation. The donor was a 25-year-old man. Organ procurement and tracheal dissection/transection were performed above the level of the innominate vein to ensure adequate tracheal length. During implantation, the donor trachea was resected approximately 2 rings above the carina to optimize tissue perfusion and to ensure a tension-free tracheal anastomosis. Total ischemia time was 240 minutes. Induction immunosuppression including mycophenolate mofetil, basiliximab, and methylprednisolone was used. The patient was able to be extubated on postoperative day (POD) 2.

On POD 4, routine bronchoscopy was unremarkable, but the bronchoalveolar lavage fluid was found to be positive for COVID-19 (polymerase chain reaction cycle threshold of 19.7) despite negative results on preoperative testing. The results of donor COVID testing were also negative with a polymerase chain reaction cycle threshold >30 and no clinical evidence of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infection. The patient was ambulatory with adequate gas exchange (Pao₂, 86 mm Hg) and was prescribed remdesivir (total of 10 doses). On POD 8, the chest film showed a small left pleural effusion. Subsequent chest computed tomography revealed focal posterior pneumomediastinum. Flexible bronchoscopy was performed, showing a membranous tracheal defect (3 × 4 cm) in the posterior aspect of the anastomosis without obvious ischemia or necrosis (Figure 1). Tracheal stent placement through flexible bronchoscopy was attempted; however, the stent was poorly seated distally and seemed to worsen the tracheal defect. The decision was made to proceed with open surgical repair.

Figure 1.

Flexible bronchoscopy with evidence of tracheal anastomotic dehiscence in the posterior tracheal wall (arrow). (LMB, left main bronchus; RMB, right main bronchus.)

Preoperatively, the patient was placed on peripheral venovenous ECMO without systemic anticoagulation. Rigid bronchoscopy was performed, and the tracheal stent was removed. General anesthesia was induced, and an endotracheal tube was placed. He was placed in a left lateral decubitus position, and a right posterolateral thoracotomy incision was performed with dissection of the latissimus dorsi muscle. Mechanical ventilation was stopped and the chest entered through a standard posterolateral thoracotomy. The trachea was mobilized cephalad to the carina, and the posterior membranous defect was identified. Primary reanastomosis was not an option, given the size (>2 cm) of the defect and concerns about tissue viability and the likelihood of a tension-free redo end-to-end anastomosis. A circular patch of bovine pericardium was used to obliterate the area of dehiscence (interrupted 4-0 Prolene), and an intercostal muscle was mobilized and tacked to the margins of the repair with resorbable interrupted suture to prevent intraluminal extrusion of the repair (Figure 2). The latissimus dorsi muscle flap was then mobilized through the second/third intercostal space and used to buttress the tracheal repair. The postoperative course was complicated by retained right hemothorax requiring thoracoscopic evacuation. The patient was discharged home on POD 38. Serial bronchoscopic surveillance has been unremarkable (Figure 3).

Figure 2.

Intraoperative flexible bronchoscopy. Bovine pericardial patch in place covering tracheal defect (arrow). (LMB, left main bronchus; RMB, right main bronchus.)

Figure 3.

Flexible bronchoscopy. (A) Four months postoperatively. (B) Seven months postoperatively. Intact repair with patch in place (arrow). (LMB, left main bronchus; RMB, right main bronchus.)

Comment

Of all the airway complications common to thoracic transplantation, anastomotic dehiscence is the most challenging. Whereas conservative management with antibiotics and frequent bronchoscopic surveillance can be effective for smaller defects (<2 cm) without evidence of major air leak or chest drainage, tracheocarinal reconstruction with extrathoracic muscle flaps constitutes definitive repair.³ In this patient with a relatively stable clinical course and no sepsis but in the setting of active SARS-CoV-2 infection, initial stent placement was attempted to avoid an open repair and the assumed associated higher operative risks. We have not found stent placement a consistent intervention for carinal injuries, given that manipulation can cause excessive sheer stress to the bronchial wall and potentially create further tissue disruption.

We identified only 2 previous reports describing tracheal repair after en bloc heart-lung transplantation with use of either intercostal muscle flap or aortic homograft for complete anastomotic disruption.^4,5 Although the use of a pericardial patch in tracheal repair without muscle flaps has been reported,⁶ we believe muscle flaps are a critical component of the bronchoplastic procedure as the long-term sequelae of pericardial patch repair are unknown.

Risk factors for airway ischemia, necrosis, and subsequent dehiscence after lung transplantation are well described.² Nonetheless, we have seen tracheal dehiscence after en bloc heart-lung transplantation only once before in our institutional experience (incidence of 4%). This occurred in a 15-year-old boy with multisystem inflammatory syndrome and no antecedent medical history who underwent transplantation from an ECMO bridge for cor pulmonale and in whom thrombosis of the inferior vena cava and tracheal dehiscence developed 15 days after an unremarkable postoperative course. This transplant occurred on the cusp of the COVID-19 pandemic and ultimately resulted in death as a result of sepsis after delayed surgical intervention with an intercostal and serratus anterior muscle flap. Although we do not have serologic data to support a role for SARS-CoV-2 in this historical case, we do believe that viral infection may have contributed to the early postoperative tracheal dehiscence described in this case report. The coagulopathic and prothrombotic state associated with COVID-19 infection of bronchial epithelial cells has been described,⁷ and histologic examination of the airway in patients with COVID-19 demonstrates thrombosis of small arterial airway vessels and associated inflammation, vasculitis, and coagulative necrosis.⁸ To date, there are not enough data to suggest that SARS-CoV-2 infection increases the risk of airway complications after en bloc heart-lung transplantation; however, this should be considered as it may play a role in the development of airway anastomotic dehiscence. More robust studies are needed to address this important question.