Management of Low Tracheal Rupture in Patients Requiring Mechanical Ventilation for Acute Respiratory Distress Syndrome

A 77-yr-old woman with chronic obstructive pulmonary disease, mitral stenosis, and atrial fibrillation was hospitalized in the cardiology unit for dyspnoea exacerbation. In the course of her treatment, a Swan Ganz catheter was inserted, and the patient experienced a cardiac arrest because of a rupture of a branch of the pulmonary artery. After emergency tracheal intubation and cardiopulmonary resuscitation, a tracheal rupture was immediately suspected because of subcutaneous emphysema. A bronchoscopy confirmed the diagnosis. Despite transfusion and amine infusion, the circulatory conditions of the patient remained unstable, and she was transferred to our institution’s intensive care unit (ICU).

On admission, the patient was hypoxemic with a Pao²/Fio²ratio below 100 (table 1). A repeated bronchoscopy localized a 4-cm tracheal rupture on the posterior side of the trachea ending 1 cm above the carina. A selective embolization of the left inferior lobar pulmonary artery was first performed with a circulatory stabilization.

An emergency thoracotomy was then performed. The repair was impossible, and the trachea was packed with intercostal muscle. Airway control was obtained with jet ventilation throughout the surgery. Jet ventilation was resumed in the ICU with a tracheal catheter introduced through the endotracheal tube whose cuff was kept deflated to maintain a low endotracheal pressure. Using this ventilatory mode, adequate blood gases were obtained. The tracheal rupture reopened on the second day of jet ventilation, and the patient died of hypoxemia.

A 67-yr-old woman was admitted to our institution for tracheal rupture management. This patient had required an emergency intubation for hypercapnic chronic obstructive pulmonary disease exacerbation 2 days earlier. Tracheal rupture was diagnosed because of subcutaneous emphysema after intubation. A bronchoscopy performed on admission in our ICU found a large tracheal rupture of the posterior side of the trachea located 1.5 cm above carina. During the first 72 h, the patient was judged not suitable for any surgical intervention because of a severe impairment of oxygenation (table 1). She was ventilated through an endotracheal tube advanced just above the carina. As the cuff of the endotracheal tube rested on the ruptured area and pulmonary atelectasis were recurrent, it was decided to bypass the tracheal rupture. The operation was performed under jet ventilation through a tracheal catheter. A large tracheotomy was performed, and a selective bilateral mainstem bronchus intubation with small endotracheal tubes (external diameter 5.5) was performed as described by Marquette et al.  5 

This operation allowed protective mechanical ventilation of the acute respiratory distress syndrome and complete healing of the tracheal rupture within 5 days. Adequate blood gases could be obtained on Day 3, and afterward the patient’s recovery was uneventful and she was discharged on Day 20.

A 42-yr-old woman was tracheally intubated because of seizures. Because of an aspiration, this patient developed a severe acute respiratory distress syndrome and was transferred to a surgical ICU. Two days after admission, she presented with subcutaneous emphysema, for which she underwent a bronchoscopy. A tracheal rupture was found 1 cm above the carina with a length of 3 or 4 cm. The patient was then transferred to our ICU for further management. On admission, she experienced a severe hypoxemia (table 1).

A median and large tracheotomy was performed, and a selective bilateral mainstem bronchus intubation with small endotracheal tubes (external diameter 5.0) was realized as previously described. An orotracheal tube was also positioned to remove secretions from the upper trachea. This procedure was performed under jet ventilation. A controlled ventilatory mode was used in the ICU, allowing the patient to have satisfactory blood gases.

After 10 days of ventilation, selective intubation could be removed, and a nonselective tracheotomy tube was placed. Complete cicatrisation of the tracheal rupture was obtained within 6 days. Successful discharge was obtained on Day 60.

A 54-yr-old man was admitted to our ICU for refractory hypoxemia. This patient had a prior esophagectomy with a gastric transposition. Eighteen days after initial surgery, the patient presented with a massive aspiration because of gastric transposition necrosis with tracheal fistula of the right main bronchus 1 cm distal to the carina. A severe acute respiratory distress syndrome developed. The bronchoscopy showed a bronchial rupture on the right main bronchial tract 1 cm distal from the carina. The patient underwent an emergent extraction of the gastric transposition. During surgery, the tracheal rupture was obliterated using fascia lata. Because of the severe acute respiratory distress syndrome, positive pressure ventilation was mandatory and the surgical reparation of the bronchial rupture failed, resulting in a massive air leakage through the bronchial rupture and thoracic drainage. The patient was then transferred to our institution with severe hypoxemia (table 1), respiratory acidosis, and circulatory failure requiring vasoactive agents. A large tracheotomy was performed at the bedside, and selective intubation was performed with two endotracheal tubes (external diameter 5.5) (fig. 1). The cuff was inflated to 30 cm H²O on the left side, and conventional ventilation on a volume control mode was performed on this lung, according to a protective ventilation strategy, with a plateau pressure below 30 cm H²O. For the right lung, a catheter was placed into the endotracheal tube 3 cm distal to its extremity. The cuff was not inflated to minimize the pressure applied on the bronchial rupture. Jet ventilation was provided using a jet ventilator (GR300; LSA, Fontenay sous Bois, France). Oxygenation increased immediately, whereas the severe hypercapnia was reduced. Three days later, Paco²was quite normalized, although the pH indicated a persisting acidosis resulting from acute renal failure. The bronchial rupture did not close because of persistent pressure applied on the injury despite the use of jet ventilation, and an air leak persisted through the thoracic drainage. After 10 days of separated pulmonary ventilation with jet ventilation on the right side, the patient died from sudden cardiac failure, despite cardiopulmonary resuscitation.

Tracheal rupture is a very rare but potentially devastating complication of tracheal intubation or chest trauma. Besides penetrating or blunt chest trauma, the most common cause of tracheal injury is iatrogenic rupture after tracheal intubation for elective surgery or respiratory failure4,6–10or after surgical procedures such as esophagectomy. Schönfelder et al.  reported 12 patients with iatrogenic tracheobronchial rupture over a 3-yr period after emergency tracheal intubation, whereas no such injury was noted among the 43,773 elective intubations performed over this period.11The true incidence of iatrogenic injuries after elective intubation is not clearly known. Reports describe only a few cases.12Nevertheless, Borasio et al.  estimated an incidence of 1:20,000 during a 7-yr period survey, with tracheal damage being more frequent after double lumen intubation (0.05–0.19%).13 

Whereas tracheal rupture after chest trauma appears most often as a disruption involving the cartilaginous-membrane junction throughout the bronchial tree, iatrogenic ruptures usually present as a longitudinal laceration of the posterior tracheal membrane.5These ruptures seem to occur in conjunction with difficult intubation, use of a stylet,8or cuff overinflation.5Double lumen intubation may be a major risk factor compared with simple orotracheal intubation.12Predominance in women has been documented.5Another important risk factor is emergency intubation. Three of our four patients developed tracheal ruptures after emergency intubation. Schneider et al.  reported emergency intubation to be present in one third of 29 patients with tracheal rupture.12However, tracheal rupture may occur after uneventful intubation.5–7A potential mechanism is an endotracheal tube tip caught in a fold of a flaccid posterior tracheal membrane while advancing it.5Tracheobronchial injury after surgery is mainly observed after subtotal esophagectomy, as in Patient 4. It is a rare event, but in this case, the injury occurs mainly proximal to the carina or in the main bronchi.14The rupture is usually intraoperatively managed effectively by surgical closure with pericardium or pleura and apposition of the gastric tube. Postoperative tracheobronchial injury associated with an ischemic gastric conduit needs further surgical exploration and even resection of the gastric tube.15 

An early and important finding is subcutaneous emphysema of the neck and face after tracheal intubation and positive pressure ventilation.1,4,5Other frequent signs are pneumothorax and mediastinal emphysema.5Respiratory failure may occur because of the underlying condition or because of the tracheal rupture.5It is important to identify this condition quickly because management is dramatically different between patients with or without respiratory failure.5Some patients may remain asymptomatic for prolonged periods.16 

Diagnosis can be confirmed by radiologic findings that show pneumothorax and mediastinal or subcutaneous emphysema.5,12These findings usually are evident on chest radiographs, although computed tomography can be particularly helpful to identify mediastinal emphysema and paratracheal air.17,18Recent developments of spiral computed tomography permit virtual bronchoscopy, a technique to identify traumatic tracheal injury.19However, this technique has never been evaluated in iatrogenic tracheal ruptures. Chen et al.  reported that computed tomography is inferior to bronchoscopy, which remains the gold standard for tracheal rupture diagnosis.17 

Surgical management has been traditionally proposed as the therapeutic approach for tracheal rupture.20However, there is currently a growing body of evidence showing that tracheal rupture can be managed with conservative treatment. Such treatment was initially limited to patients with small tears less than 2 cm in length, a delayed diagnosis, or patients with poor clinical conditions.21Because of the high mortality of surgical treatment,22nonoperative treatment has been further extended to patients with absence of esophageal injury, minimal mediastinal air collection, nonprogressive mediastinal or subcutaneous emphysema, and the absence of sepsis. Finally, a recent report concluded that conservative nonoperative treatment should be considered for tracheal rupture in patients breathing spontaneously or with a planned extubation within the first 24 h.5Indeed, Conti et al.  reported excellent results of conservative management regardless of the size and location of the tracheal rupture or the delay from extubation to diagnosis.5However, management of patients with severe respiratory failure requiring mechanical ventilation remained a more difficult issue.

Indeed, the presence of an endotracheal tube with an inflated cuff and positive airway pressure is likely to impair the healing of the tracheal injury. The location of the tracheal injury is a major issue for the management of these patients. When the tracheal rupture is proximal (i.e., first or second third of the trachea), bridging of the lesion may be possible and should be preferred.5,23,24The patient is then mechanically ventilated through an endotracheal tube placed distally to the injury. In these conditions, the healing of the tracheal rupture is not hindered by the tracheal cuff or positive airway pressure. This strategy fails when the tracheal rupture is located on the lower third of the trachea or involves a bronchial mainstem. Jet ventilation through the endotracheal tube with a deflated cuff could allow the patient to be ventilated in such cases. Indeed, this method was used during the tracheotomy procedure to avoid further impairment of hypoxemia in patients who were already hypoxemic. Nevertheless, the cases of Patients 1 and 4 illustrate that even jet ventilation through an endotracheal or endobronchial tube with a deflated cuff and low endoluminal pressure do not allow tracheobronchial healing. It may be that even a low tracheal or bronchial pressure maintained some gas flow through the injury, preventing any healing. This failure of jet ventilation prompted us to adopt an alternative strategy for Patients 2 and 3 with low tracheal rupture, following the suggestions of Marquette et al.  5A selective bilateral mainstem bronchus intubations with high-volume/low-pressure cuffed endotracheal tubes were performed through large tracheotomies.25The tracheotomies were required because endotracheal tubes were too short to be placed endobronchially via  orotracheal intubation.26During these procedures, jet ventilation was performed with jet catheters inserted beyond the tracheal openings to ventilate the patients. The patients were then ventilated according to a protective ventilatory protocol for their acute respiratory distress syndrome. Meanwhile, the tracheal rupture was put at rest and therefore could heal. For Patient 4, the situation was a little bit more complicated as the tracheal rupture involved the right mainstem bronchus. Meanwhile, an attempt to repair the tracheal fistula was unsuccessful, and the patient was transferred to our ICU for management. We performed a selective bilateral mainstem bronchus intubation under jet ventilation during tracheotomy as before. However, the right tube did not bypass the bronchial injury because it was too close to the right superior lobar bronchus. The patient was estimated to be too critically ill to undergo additional surgery, and we chose to ventilate the patient through the two endobronchial tubes. To minimize the endobronchial pressure, we used jet ventilation on the right lung, whereas the left lung was ventilated in a conventional manner. This ventilatory regimen was initially beneficial to the arterial blood gases, but the leak through the right chest tube persisted, indicating a lack of bronchial injury healing. Unfortunately, the patient died 10 days later from a nonventilatory cause. Retrospectively, conventional ventilation on the right side at the expense of upper right lobe ventilation could possibly have enhanced the bronchial healing and more rapid ventilatory weaning.

Based on recent publications and our experience, nonoperative management of tracheal rupture in patients requiring mechanical ventilation should be considered. The management of proximal and mild tracheal rupture may be performed by simply bridging the lesion. Ideally, the tube should be placed with the help of fiberoscopy to prevent selective intubation. For patients with lower tracheal rupture, the situation is more difficult. If the rupture does not involve one of the mainstem bronchi, a large tracheotomy could be performed to permit bilateral selective intubations, allowing the lungs to be ventilated and the tracheal injury to heal. However, if the tracheal rupture involves one of the mainstem bronchi, the situation is even more complicated. To heal, the tracheobronchial rupture should be kept at rest with no pressure applied. As shown in Patient 4, even low pressure (<10 cm H²O) that could be obtained with jet ventilation may be too much to allow healing. One alternative strategy is to inflate a tube cuff distal to the rupture. For the left mainstem bronchus, this approach may only be possible if the tracheal rupture is proximal. However, this could not be performed for the right mainstem bronchus because of the proximity of the right superior lobar bronchus.

In conclusion, management of tracheal rupture associated with respiratory failure is a difficult issue. Simple bridging of upper tracheal injuries with an endotracheal tube placed distal to the lesion may be helpful. When tracheal rupture involves the lower third of the trachea, carina, or left mainstem bronchus, a selective bilateral mainstem bronchus intubation via  a large tracheotomy should be considered. When tracheal rupture is located on the right mainstem bronchus, the situation is potentially critical because of the proximity of the right superior lobar bronchus.