Patients suffering tracheal trauma frequently fail to live long enough to be evaluated (2). In patients who do receive medical attention, blunt trauma may also result in pulmonary contusion and/or laceration, thoracic spine injuries, esophageal disruption, or multiple fractures. Extrathoracic organ injury is also frequently encountered (3). Not surprisingly, most reported series are small. Trauma may be either penetrating or, more often, blunt (Figs. 3-1, 3-2, 3-3 and 3-4). Although penetrating tracheal injuries typically involve the neck, blunt traumatic injuries usually involve the intrathoracic portions of the trachea and often extend to the mainstem bronchi (4). Most often the result of deceleration or so-called steering wheel injuries, proposed mechanisms of injury include compression of the airways between the sternum and thoracic spine with lateral expansion of the thorax; shearing at fixation points, especially at the carina; and elevated intratracheal pressure, especially against a closed glottis (5). These injuries result in transverse lacerations; disruption of the posterior, membranous portion of the trachea; or, less commonly, shearing of the mainstem bronchi from the carina (Figs. 3-2 and 3-3). Although the entire length of the trachea is vulnerable to injury following blunt trauma more than 80% of cases occur within 2.5 cm of the carina (2). It has been suggested that the left mainstem bronchus is less frequently disrupted, probably because of its longer course through the mediastinum, which offers some extra measure of protection. Additional causes of tracheal rupture include overdistension of endotracheal tubes (Fig. 3-4) and traumatic intubation (6).

Radiographic findings most often include pneumomediastinum and deep cervical emphysema (7). Unfortunately, neither of these signs is specific for tracheal disruption. Pneumomediastinum, for example, most commonly occurs secondary to air leaks originating in the pulmonary interstitium, the so-called Macklin effect (8). Pneumothoraces are also common, presumably the result of disruption of the mediastinal pleura. In those patients in whom the airway is partly occluded by blood clots or compression by an adjacent mediastinal hematoma, the first sign of tracheal rupture may be a persistent air leak following chest tube insertion. In more severe cases, bronchial avulsion may lead to a falling away of the involved lung from the hilum resulting from disruption of central attachments, the so-called fallen lung sign (7).

CT findings in patients with blunt trauma have been reported (8, 9, 10, 11, 12). In a recent retrospective study of 14 patients with tracheal rupture either resulting from trauma or following intubation, CT proved diagnostic in 85%, providing direct visualization of tracheal tears in 10 patients (71%). Direct signs of tracheal rupture included focal wall defects in more than 50% of patients (Figs. 3-2 and 3-3); less commonly, rupture resulted in a deformed tracheal lumen. In this same series, overdistension of the tube balloon was noted to be particularly common, occurring in five of seven intubated patients. Of particular value was the finding of herniation of the endotracheal tube balloon
outside the confines of the tracheal wall (Fig. 3-4). It should be emphasized that optimal evaluation of tracheal injuries should include use of advanced image processing, especially multiplanar reconstructions. In select cases, visualization of tracheal tears may be most easily identified with virtual bronchoscopy (Fig. 3-3).

Although imaging findings are often suggestive, definitive diagnosis and therapy require bronchoscopy in nearly all patients. Rare exceptions include patients in whom the diagnosis can be made presumptively because of direct observation or evidence of a massive air leak following chest tube placement (13). Although treatment typically involves surgical repair, conservative therapy has been advocated for individuals with postintubation membranous tracheal ruptures if these are less than 2 cm long (6,14). Late manifestations of tracheal rupture include strictures, and patients presenting with wheezing that can be potentially misdiagnosed as asthma.

Benign airway strictures may be congenital or acquired. The most common cause of acquired strictures is iatrogenic injury resulting from intubation. Although previous reports have documented complications occurring in up to 10% to 15% of patients following intubation, their incidence has decreased dramatically since the introduction of high-volume/low-pressure balloons (15). In distinction, complications following tracheostomy tube insertion remain common, occurring in approximately 40% of patients. The two principal sites of stenosis following intubation or tracheostomy are at the stoma and at the level of the endotracheal or tracheostomy tube balloon (16). In both cases, strictures result from pressure necrosis, causing ischemia and subsequent scarring. Stenosis also may be the result of inflammation, with subsequent thinning and weakening of the tracheal wall (tracheomalacia) in the segment between the stoma and cuff; narrowing may also be caused by granulation tissue resulting from direct tracheal injury by the endotracheal or tracheostomy tube tip (16). Another complication is tracheal rupture. Rarely, tracheobrachiocephalic artery or tracheoesophageal fistulas may occur, the latter usually developing secondary to combined tracheal and esophageal intubation. Additional causes of tracheal stenosis include strictures resulting from infection, most often tuberculous, and following lung transplantation. After lung transplantation, complications typically occur at the level of the bronchial anastomosis and include dehiscence, necrosis, and stenosis from malacia, fibrous stricture, prominent granulation tissue formation, or a combination of these conditions (17, 18, 19, 20, 21).

Regardless of etiology, both CT and magnetic resonance (MR) have proved reliable methods for assessing tracheobronchial strictures (Figs. 3-5 and 3-6). In all cases, accurate evaluation requires that the precise length and degree of stenosis be assessed along with associated peritracheal or bronchial abnormalities. CT with multiplanar reconstructions, in particular, has proved of value with sensitivities of greater than 90% reported (22). It should be noted that in select cases additional certainty in assessing the degree of stenoses may also be attained by use of virtual bronchoscopy (23,24).

Although endoscopic management of tracheal stenoses has been proposed, especially with the use of lasers or stents, these typically provide only temporary benefits. Primary tracheal and laryngotracheal resection and reconstruction remain the optimal methods for treating nonneoplastic stenoses; resection up to 4 cm is feasible, which is the equivalent of eight cartilaginous rings (25). In one series of 58 patients with postintubation or idiopathic stenoses, major complications occurred in 12%, including anastomotic dehiscence and vocal cord paralysis (25).

Primary tumors of the trachea are rare, representing less than 1% of bronchial neoplasms (26, 27, 28, 29) To date, a bewildering array of different tracheal tumors, both benign and malignant, have been reported, including both primary epithelial and mesenchymal neoplasms as well as secondary neoplasia resulting from either metastases or, more commonly, direct tracheal invasion by adjacent mediastinal neoplasms (30). In fact, only a few lesions occur with any frequency, and squamous cell carcinoma (SCC) and adenoid cystic carcinoma (ACC) together account for up to 86% of cases (26, 27, 28, 29). Other relatively less common malignant lesions encountered include carcinoid and mucoepidermoid tumors, while endobronchial hamartomas and bronchial papillomas are the most common benign lesions. Although these tumors are discussed separately, it should be emphasized that as a group, regardless of their cell type, most patients with endotracheal and proximal bronchial neoplasms present with a limited number of symptoms and signs: cough, wheezing, stridor, hemoptysis, and evidence of either parenchymal consolidation or atelectasis, depending on the site and degree of bronchial obstruction.

In addition to primary tumors, the trachea and central bronchi may also be secondarily involved by tumors, either as a result of direct invasion, or less commonly by hematogenous metastases (Fig. 3-16) (31). Local invasion most often results from tumors arising from the esophagus, larynx, thyroid, and lung. Metastatic disease, although less common, most often results from renal and colonic primary lesions (46). It should be emphasized that assessment
of tracheal invasion may be exceedingly difficult with CT. The finding of a peritracheal soft-tissue mass causing focal displacement of the adjacent tracheal lumen should be interpreted as nonspecific. It cannot be overemphasized that in the absence of a discrete endobronchial mass, evidence of fistulization to adjacent mediastinal structures, or destruction of tracheal cartilage, the diagnosis of tracheal or mainstem bronchial invasion is limited (Figs. 3-17 and 3-18). Definitive evaluation usually necessitates either bronchoscopic or histologic confirmation. Although it has been reported that MR may be preferable to CT in assessing invasion of the tracheal wall (47), to date routine use of MR for this purpose has not gained widespread acceptance

Given the overlap in both clinical manifestations and CT appearance of most endobronchial lesions, not surprisingly definitive diagnosis requires histologic evaluation in nearly all cases. In addition to the primary tracheal lesions mentioned previously, it should be noted that a bewildering variety of additional rare endobronchial lesions have been reported (27,48,49). Regardless of origin, primary endotracheal lesions are best treated surgically, especially with primary resection and reanastomosis, usually followed by radiation. To date, most large series have primarily focused on patients with SCCs and ACCs. In a series of 20 patients with primary tracheal neoplasms treated surgically, Hazama et al. (29) reported a 5-year survival rate of 72.3% for patients undergoing surgical treatment including segmental tracheal, laryngotracheal, or carinal resections. Nearly identical results have been reported by others (27,50). Overall, approximately three quarters of patients with primary tracheal or carinal lesions will be resectable.

In an evaluation of 198 patients with primary tracheal tumors reported by Grillo et al. (27), 147 (74%) were treated surgically, including 44 with SCCs, 60 with ACCs, and 43 with assorted benign and malignant tumors. Surgical mortality for primary resections was 5%. Patients with ACCs survived longer than those with SCCs, especially when the latter were associated with mediastinal adenopathy. Curiously, in patients with ACCs, survival was similar whether or not there was evidence of mediastinal nodes or residual tumor along the margins at resection. Nearly identical results have been reported by Hazama et al. (29), who also noted that survival in patients with ACCs was independent of the presence of positive margins at surgery. In this setting, it turns out that the benefits of performing primary resection outweigh the potential liability of leaving behind residual microscopic tumor.

Although surgical resection is to be preferred when feasible, an increasing number of nonsurgical options are now available for the treatment, and especially the palliation, of endobronchial tumors. As discussed in depth in Chapter 2, these include neodymium:yttrium-aluminum-garnet (Nd:YAG) laser phototherapy, photodynamic therapy following the intravenous administration of photosensitizing drugs, cryotherapy, electrocoagulation, brachytherapy, and endobronchial stents with or without prior bronchoplasty (Figs. 3-12 and 3-19) (51).