| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1
Department of Radiology, Veterans General Hospital, Taipei and National
Yang-Ming Medical School, Taipei, Taiwan.
2
Department of Radiology and Shock Trauma Center, University of Maryland School
of Medicine, 22 S. Greene St., Baltimore, MD 21201.
3
Department of Anesthesiology and Shock Trauma Center, University of Maryland
School of Medicine, Baltimore, MD 21201.
Received July 3, 2000;
accepted after revision October 25, 2000.
Address correspondence to K. Shanmuganathan.
Abstract
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
MATERIALS AND METHODS. Neck or chest CT scans of 14 patients with tracheal rupture and 41 control trauma patients with pneumomediastinum but without tracheal injury were reviewed and compared to assess the presence and location of extrapulmonary air, whether direct visualization of tracheal wall disruption was possible, the size and shape of endotracheal tube balloon, signs of transtracheal balloon herniation in intubated patients, and the location of the extratracheal endotracheal tube. Intact and experimentally injured cadaver tracheas were used to evaluate tube balloon pressure and configuration.
RESULTS. All 14 patients with tracheal rupture had deep cervical air and pneumomediastinum. Overdistention of the tube balloon occurred in 71% (5/7) of the intubated patients, and balloon herniation occurred in 29% (2/7). Direct tracheal injury was seen in 71% (10/14) of the patients as a wall defect (n = 8) or deformity (n = 2). Overall, CT was 85% sensitive for detecting tracheal injury. Patients with tracheal injury had a significantly lower incidence of pneumothorax (p = 0.01) than did the control group. The CT appearance of balloon herniation through defects in the cadaver tracheas closely mimicked those of patients with tracheal injury. The amount of balloon pressure required to rupture the intubated trachea was extremely high and rupture was difficult to obtain.
CONCLUSION. CT can reveal tracheal injury and can be used to select trauma patients with pneumomediastinum for bronchoscopy, leading to early confirmation and treatment.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Most major airway injuries are not recognized initially. Diagnosis of tracheal rupture may be delayed as a result of its rare incidence, subtle and nonspecific clinical and radiologic manifestations, and the much more overt clinical signs of other more common associated injuries [5, 8]. Delayed or missed diagnosis can result in death or severe complications including ventilatory failure, mediastinitis, sepsis, airway stenosis, bronchiectasis, recurrent pulmonary infections, and permanent pulmonary function impairment [3, 7, 8]. CT is being used with increasing frequency to facilitate detection of chest injuries in patients with severe multisystem trauma and has proved useful in diagnosing unsuspected thoracic injuries [9, 10].
We undertook a retrospective study of patients admitted to our trauma facility who were diagnosed with tracheal injury. First, CT studies were reviewed to determine findings that indicate tracheal injury and to assess their sensitivity, specificity, and accuracy for diagnosis. Second, differences in CT findings that were related to penetrating versus blunt tracheal injury were assessed. Finally, using intubated cadaveric tracheas, the relationships among endotracheal balloon pressure, balloon configuration, and tracheal disruption were evaluated.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Control Trauma Population
To verify the significance of the anatomic location of extrapulmonary air
and the configuration and size of the endotracheal tube balloon after tracheal
injury, we reviewed admission CT studies of 41 consecutive chest trauma
patients without tracheobronchial injury but with pneumomediastinum. All 41
patients underwent bronchoscopy (n = 16) or clinical follow-up of at
least 1 year (n = 25). Among this population, blunt force injury
occurred in 33 patients and penetrating injury in eight. There were 32 males
and nine females, ranging in age from 16 to 75 years (mean age, 32 years). All
41 patients without tracheobronchial injury but with pneumomediastinum had CT
scans of the neck (n = 33) or chest (n = 41). Blunt force
trauma (n = 33) resulted from motor vehicle collision (n =
25), fall (n = 7), and hanging (n = 1). Penetrating injury
(n = 8) resulted from gunshot wounds (n = 5) and stab wounds
(n = 3). Endotracheal intubation was performed before CT in 71%
(29/41) of the patients in the control group.
CT Technique and Evaluation of CT Findings
Contrast-enhanced CT scans of the chest were obtained using helical CT (a
collimation of 8 mm, table speed of 8 mm, and pitch of 1) or conventional CT
(contiguous images with 8-mm collimation) from the thoracic inlet to the lung
bases. Contrast-enhanced CT scans of the neck were obtained using helical CT
(a collimation of 5 mm, table speed of 5 mm, and pitch of 1) or conventional
CT (contiguous images with 5-mm collimation) from the skull base to the
thoracic inlet.
The CT images of the chest and neck were reviewed on soft-tissue (width, 320 H; level, 40 H) and lung (width, 1500 H; level, -500 H) settings for the following findings. First, CT images were reviewed for extrapulmonary air and its anatomic location. Extrapulmonary air was defined as air seen in the following anatomic locations including soft-tissue emphysema in the neck, pneumomediastinum, paratracheal air (Fig. 1) (extrapulmonary air in direct contact with the external tracheal wall), pneumothorax, retroperitoneal air, pneumoperitoneum, and pneumopericardium. Second, CT images were reviewed to assess whether direct visualization of the tracheal injury, including deformity or fracture of tracheal cartilage or cartilages or discontinuity (a defect) of the tracheal wall, was possible. Third, in patients with endotracheal intubation in both the study and control groups, the endotracheal tube balloon diameter was measured with electronic calipers on transverse CT images. The balloon was considered to be overdistended if the maximum diameter of the balloon cuff was greater than 2.8 cm (average tracheal diameter: males, 2.43 cm; females, 2.0 cm) [11]. Fourth, CT images were evaluated for evidence of herniation of the endotracheal tube balloon outside the trachea. Fifth, CT images were examined for an extratracheal location of the endotracheal tube.
|
The CT images, including axial and multiplanar reformations, of the study and control groups were retrospectively reviewed by four of the authors, and an interpretation was rendered by consensus. The anatomic location of the tracheal injury was not known at the time of CT interpretation. Supine chest radiographs (n = 13) and lateral cervical spine radiographs (n = 10) of the patients in the study group were also retrospectively reviewed by consensus interpretation for the presence of extrapulmonary air and its anatomic location and for deformity in the shape of the balloon. The radiographic and CT findings were compared to determine whether CT provided additional information.
Medical records were reviewed to determine bronchoscopic results and operative findings. The two-tailed Fisher's exact test (Statistical Software [1999 version of 6.0]; Stata, College Station, TX) was performed to verify the significance of extrapulmonary air in various anatomic locations as seen on CT in patients with tracheal injury and in patients in the control group. The significance level was assumed at a p value of 0.05 or less. The study was exempt by the University of Maryland Institutional Review Board.
Cadaveric Tracheal Study
To verify the normal configuration of the endotracheal tube balloon under
various pressure loads and to reproduce endotracheal tube balloon deformities
seen on CT studies of patients with tracheal injury, we intubated adult human
cadavers with a high volume-low pressure endotracheal tube (Hi-Lo
[endotracheal tube no. 7.0]; Mallinckrodt, St. Louis, MO). In the first
specimen, helical CT scans were obtained (collimation of 3 mm, pitch of 1.5,
and reconstruction at 2.0-mm intervals) after inflation of the endotracheal
tube balloon with 10, 20, 30, and 40 mL of air. These volumes of air were
instilled both before and after creation of three tracheal injury patterns
seen at surgery or bronchoscopy. These injury patterns included a
through-and-through perforation of the anterolateral walls of the trachea from
a gunshot wound, a short longitudinal perforation (approximately 1 cm) of the
membranous trachea, and a long longitudinal laceration (approximately 7 cm) of
the membranous trachea.
Another study in adult cadavers was performed to determine if the trachea or the endotracheal tube balloon would rupture first from overinflation of the balloon. Attempts were also made to determine the highest endotracheal tube balloon pressure and the minimum volume of air needed to rupture the trachea or the endotracheal tube balloon after intubation with the endotracheal balloon outside the trachea using a cuffed endotracheal tube (Hi-Lo [no. 7.0]; Mallinckrodt). The incremental endotracheal tube balloon volumes and pressures were measured after each time 5 mL of air was introduced into the balloon until rupture of either the trachea or the endotracheal tube balloon.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The most common CT finding of tracheal injury was deep cervical emphysema and pneumomediastinum; this finding was seen in 100% (14/14) of the patients (Fig. 1). Other CT findings included paratracheal air in 93% (13/14) of the patients (Fig. 1), pneumothorax in 36% (5/14), and pneumoretroperitoneum in 14% (2/14). No significant difference was observed in the anatomic location of extrapulmonary air between patients with blunt force trauma and those with penetrating injury. The tracheal wall injury was directly visualized on CT in 71% (10/14) of the patients, either as a defect or discontinuity of the wall of the trachea in 57% (8/14) (Fig. 2A,2B) or as a focal tracheal wall deformity or tracheal ring fracture in 14% (2/14) (Fig. 3). The anatomic location of the tracheal defect and focal wall deformity seen on CT correlated with bronchoscopic and operative findings in all 10 patients with direct CT findings of tracheal wall abnormality.
|
|
|
Endotracheal intubation was performed in 50% (7/14) of the patients. An overdistended endotracheal tube balloon cuff was seen in 71% (5/7) of the patients with tracheal injury. Of these five patients, a tracheal wall defect was seen in two patients and a focal deformity of the tracheal wall was seen in one patient. The endotracheal tube balloon diameter was measured (range, 2.8-4.2 cm; mean, 3.4 cm) on axial images at the point of maximal distention in these five patients.
The overdistended endotracheal tube balloon was either spherical or ovoid (Figs. 4A,4B and 5) in 43% (3/7) of the patients and was herniated outside the wall of the trachea in 29% (2/7) of the patients. In one of these two patients, the balloon herniated through the anterolateral walls of the trachea from a gunshot wound suggested the appearance of a mickey-mouse head (Figs. 6A,6B and 7A,7B). In the other patient with a small perforation of the membranous wall of the trachea, the balloon had a dumbbell shape (Figs. 8A,8B and 9). An extratracheal location of the endotracheal tube tip was noted on CT in one patient (Fig. 3). A tracheal wall injury or herniation of the endotracheal tube balloon could not be directly seen on CT in the remaining 14% (2/14) of the patients. In the original neck or chest CT interpretations of 13 study patients, the diagnosis of major airway injury was either made or suggested in 11 (85%).
|
|
|
|
|
|
|
|
|
|
CT and Radiographic Findings in the Study Group
Supine chest radiographs revealed pneumomediastinum in 85% (11/13) of the
patients with tracheal injury. Soft-tissue emphysema in the neck was evident
on either supine chest radiographs in 92% (12/13) of the patients or lateral
cervical spine radiographs in 100% (14/14). Chest radiographs failed to show a
pneumothorax that was detected on CT in 40% (2/5) of the patients. Direct
signs of tracheal injury were seen on chest radiographs in only 50% (3/6) of
the patients. Neither the dumbbell-shaped deformity of the endotracheal tube
balloon evident on CT in one patient nor the overdistention of the balloon
evident on CT in another patient was visible on chest or cervical spine
radiographs.
Control Trauma Population
CT findings in the control group of 41 patients
(Table 1) included paratracheal
air, seen in 73% (30/41) of the patients, pneumothorax in 76% (31/41),
pneumoretroperitoneum in 15% (6/41), pneumopericardium in 7% (3/41), and
pneumoperitoneum without hollow viscus injury in 2% (1/41). There were 29
patients (71%) who had undergone endotracheal intubation. Of these 29
patients, the maximal endotracheal tube balloon diameter was 2.7 cm (range,
1.7-2.7 cm). None of the endotracheal tube balloon cuffs was herniated or
deformed.
|
Statistical Analysis
The comparison of the anatomic location of extrapulmonary air in patients
with tracheal injury with that in the control group
(Table 1) (two-tailed Fisher's
exact test) showed that patients with tracheal injury had a statistically
significant lower incidence of pneumothorax than the control group (p
= 0.01). No other statistically significant difference was observed between
the two groups of patients when extrapulmonary air was present in the
paratracheal, retroperitoneal, or pericardial spaces.
Within the control group, no statistically significant difference between
the patients with penetrating and those with blunt force injury or between
patients who underwent endotracheal intubation and those who did not was
observed in the patients with paratracheal air
(Table 2). The presence of
paratracheal air was more likely in patients with tracheal injury (93%
[13/14]) than in control group patients without tracheal injury who had not
been intubated (58% [7/12]), but this finding did not quite reach statistical
significance (p = 0.056). When nonintubated study patients with blunt
traumatic tracheal injury were compared with nonintubated control patients
(100% [5/5]) with blunt trauma (56% [5/9]), the presence of paratracheal air
was highly associated with tracheal rupture (p<0.05;
2=4.0266).
|
Cadaveric Tracheal Study
In the adult cadavers, the diameter of the endotracheal tube balloon
increased during incremental inflation, with 10, 20, 30, and 40 mL of air, of
the trachea. The balloon always maintained a cylindric shape on scanograms and
on axial or multiplanar reformation images. The balloon did not herniate
beyond the tracheal wall.
The endotracheal tube balloon cuff pressures during incremental inflation with air inside and outside the trachea are shown in Table 3. The intubated trachea ruptured before the endotracheal tube balloon ruptured. Tracheal rupture occurred after 75 mL of air was introduced into the balloon, whereas 130 mL of air was required to rupture the endotracheal tube balloon outside the trachea. A marked increase in resistance to further inflation of the balloon in the intubated trachea was observed after 50 mL of air was introduced into the balloon. Resistance to inflation of the endotracheal tube balloon was subjectively less outside the trachea than within, but resistance increased dramatically after 70-80 mL was injected. The tracheal rupture occurred as a long longitudinal tear in the midline of the membranous portion of the trachea.
|
For the same volume of air introduced into the balloon before and after intubation of the trachea, the balloon cuff pressures after intubation were always higher and increased more abruptly and rapidly after 15 mL of air had been introduced into the balloon. The endotracheal tube balloon cuff pressures exceeded 120 cm H2O (maximal measurable pressure) when both the trachea and endotracheal balloon ruptured (Table 3).
The endotracheal tube was placed into the cadaveric trachea after the three experimental injuries described earlier were created to represent injured tracheas. The endotracheal tube balloon became overdistended and progressively herniated through the experimental defects created in the tracheal wall with incremental inflation. The appearances of the balloon herniating through the tracheal injuries corresponded almost exactly to its CT appearance in the clinical cases (Figs. 5, 7A,7B, and 9).
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Other mechanisms of tracheal injury include tearing across the lower tracheal-carinal junction from anteroposterior chest compression forcing the lungs apart laterally, hyperextension of the neck resulting in stretching and compression of the trachea against a steering wheel or dashboard, direct crushing of the trachea between the sternum and thoracic spine, and sudden and rapid deceleration with shearing force applied to the relatively fixed cricoid cartilage and carina [3, 4, 6, 13].
Penetrating injury is a less frequent cause of tracheal injury and more commonly involves the anterior extrathoracic trachea including the cartilage or the ligamentous portions between the tracheal rings, with a higher incidence of associated esophageal and vascular injuries [3, 5, 7]. Early diagnosis and prompt surgical repair is crucial for survival and reduction of the morbidity resulting from tracheal injury.
Radiologic signs of tracheobronchial injury are nonspecific and include pneumomediastinum, pneumothorax, and progressive extrapulmonary soft-tissue air [3, 8, 12, 14]. Case reports indicate that CT, performed with the appropriate window settings, can reveal the exact site of tracheal injury by directly showing focal defects or the circumferential absence of the tracheal wall, a contour deformity, or abnormal communication with other mediastinal structures [15,16,17,18]. In our study, using lung and soft-tissue window settings in the chest and bone and soft-tissue window settings in the neck allowed direct visualization of the tracheal injury on axial CT images in 71% (10/14) of the patients, without prior knowledge of the anatomic site of the tracheal tear. A defect in the wall of the trachea was seen in 57% (8/14) of the patients and a focal wall deformity in 14% (2/14). The anatomic location of the tracheal wall abnormality seen on CT was confirmed at either bronchoscopy or surgery in all patients. An awareness of these specific CT findings that are suggestive of tracheal injury will help select patients for early confirmatory bronchoscopy.
Herniation of the endotracheal tube balloon outside the tracheal wall has not, to our knowledge, been described as a specific sign of tracheal injury. Balloon deformities in the shape of a mickey-mouse head and dumbbell were seen from herniation through a defect or defects in the tracheal wall. These CT findings not only were specific signs of tracheal injury but also were helpful to localize the site of the tracheal tear. In the remaining 14% (2/14) of the patients, neither a tracheal wall injury nor herniation of the endotracheal tube balloon could be directly visualized. Possible factors contributing to nondiagnostic CT include respiratory misregistration, small lesion size with obscuration from partial volume averaging, size of injury below the resolution of CT, and nondisplaced tears occurring in the transverse plane. The ability to scan with a thinner collimation using helical CT or multislice helical CT should improve the sensitivity of CT to directly show tracheal wall injuries.
In our study, deep cervical emphysema and pneumomediastinum were the most common radiographic and CT findings seen in all 14 patients with tracheal injury. These findings were related to an air leak into the mediastinum and deep cervical fascial planes from the disrupted thoracic or cervical trachea. Some studies postulate that pneumomediastinum or deep cervical emphysema in patients who sustain closed trauma is strongly suggestive of tracheal or bronchial injury [6, 12, 19, 20]. However, pneumomediastinum is clearly not pathognomonic of tracheal injury. In our study, there was no significant difference in the number of patients with retroperitoneal or intrapericardial air between the study and control groups. The most common cause of pneumomediastinum is the "Macklin effect," in which alveolar rupture, due to ventilator barotrauma or pulmonary laceration, produces an air leak into the pulmonary interstitium with retrograde dissection along the perivascular sheaths toward the hilum and into mediastinum [21]. Mediastinal air can also track into the retroperitoneum (14% of our patients) and peritoneal spaces from intrathoracic tracheal injuries.
Other studies describe paratracheal air or air paralleling the airway as a significant finding for tracheobronchial injury [6, 8, 22, 23]. None of these studies used a control group to verify the significance of this finding. In the current study, paratracheal air was present in 93% of the patients with tracheal injury and in 73% of the patients without tracheal injury and pneumomediastinum. No statistically significant difference was observed in the number of patients with paratracheal air between these two groups, thus indicating that this CT finding is not specific for tracheal injury. However, patients with tracheal injury tended to have paratracheal air, whereas the control group of nonintubated patients with pneumomediastinum tended to not (p = 0.056). Thus, the presence of paratracheal air on CT or radiography in an unintubated trauma patient should arouse suspicion of tracheal injury and deserves further evaluation with bronchoscopy.
Two studies have reported an absence of or delayed appearance of pneumomediastinum in patients with tracheobronchial injury on supine chest radiographs and attribute this finding to preservation of the integrity of paratracheal connective tissue sheath or temporary occlusion of a small tear by the endotracheal tube balloon [11, 22]. The failure to identify a small amount of mediastinal air on radiographs may be another contributory factor. These two observations were confirmed in our study. Two study patients with evidence of pneumomediastinum on CT did not have this finding detected on conventional radiographs. Subjectively, relatively small volumes of pneumomediastinum were seen in our patients with overdistention of the endotracheal tube balloon in this study, reflecting the tamponade effect of the endotracheal tube balloon in preventing air leak into the mediastinum.
This study showed pneumothorax occurs infrequently in patients with isolated tracheal injury. A significantly lower incidence of pneumothorax was seen in patients with tracheal injury (36%) than in control group patients (76%) (p = 0.01). Tears in the trachea produce a central air leak into the surrounding mediastinum without a pneumothorax unless there is also a tear in the mediastinal pleura [5, 6, 19, 22]. An animal study performed by Lloyd et al. [13] indicated that the presence of a pneumothorax may have a protective effect on the tracheobronchial tree by decreasing the distracting stress at the carina during anteroposterior compression and abruptly increasing in the transverse thoracic diameter.
A major limitation of this study is the retrospective nature of the case review and the small sample size caused by the rarity of tracheal injury even in a major trauma center. Among 13 of the 14 study patients for whom the original neck or chest CT interpretation was available, the report mentioned possible or definite evidence of major airway injury in 11 patients (85%).
Cadaveric Study
An oval or spherically overdistended endotracheal tube balloon has been
well described as a specific finding of tracheal injury on chest and cervical
spine radiographs [11]. Our
cadaveric studies show that only in the presence of a tracheal rupture can the
endotracheal tube balloon be easily overinflated as a result of decreased
tracheal wall resistance in patients with tracheal disruption. The absence of
an intact tracheal wall is perceived as a lack of resistance frequently
leading to overinflation or herniation of the endotracheal tube balloon
through a defect or defects in the tracheal wall and reproducing deformities
in balloon shape (mickey-mouse head and dumbbell) seen in our patients with
tracheal injury.
A larger than normal diameter of the endotracheal tube balloon might occasionally result from unintended excessive balloon inflation in the uninjured trachea. In four of our five patients, overdistention of the endotracheal tube balloon was obviously the result of, but not the cause of, tracheal rupture. However, in the remaining patient who had intubation-related tracheal injury, the cause of tracheal rupture was less apparent.
Intubation-related tracheal rupture has been attributed to traumatic
intubation in the field under suboptimal conditions, the inappropriate use of
the stylet, the use of an oversized tube, the overinflation of the
endotracheal tube balloon, or an underlying weakness of membranous trachea as
may occur in the elderly [14,
24,
25]. Overdistention of the
balloon may be seen in patients with tracheomalacia from chronic endotracheal
intubation with high cuff pressure
[26]. Despite reports favoring
endotracheal tube cuff overinflation as the major cause of acute
intubation-related tracheal rupture
[24,25,26,27],
only one of these reports convincingly supports this theory
[27]. Our study of adult
cadavers confirms that 70-80 mL of air is needed in the endotracheal tube
balloon to rupture the trachea. A marked increase in resistance to hand
injection of air was noted after introduction of 50 mL of air into the
intubated trachea balloon. This finding suggests that it is unlikely and
difficult to overinflate the balloon in an intact trachea with up to 7-8 times
the normal volume (
10 mL) of air, particularly given the marked resistance
to injecting such a large volume. Intubation-related tracheal injury,
particularly to the membranous portion, can still result from the several
other factors mentioned.
In conclusion, this study shows the overall sensitivity of CT for the diagnosis of tracheal rupture was 85%. Specific CT findings include overdistention of the endotracheal tube balloon or herniation of the deformed endotracheal tube balloon beyond the trachea, the extraluminal position of the endotracheal tube, a focal tracheal wall defect or discontinuity, and a contour deformity or fracture of the trachea. With the increasing use of helical and multislice helical CT for examining patients with severe chest trauma, awareness of these CT signs can suggest the diagnosis of tracheal injury, thus leading to early confirmatory bronchoscopy and definitive treatment. In addition, CT findings of paratracheal air in an unintubated trauma patient can be used to improve selection of patients with relatively common posttraumatic pneumomediastinum for bronchoscopy.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  M. N. Saulez, B. Dzikiti, and A. Voigt Traumatic perforation of the trachea in two horses caused by orotracheal intubation Vet Rec., June 6, 2009; 164(23): 719 - 722. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. Kaewlai, L. L. Avery, A. V. Asrani, and R. A. Novelline Multidetector CT of Blunt Thoracic Trauma1 RadioGraphics, October 1, 2008; 28(6): 1555 - 1570. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. Stark, K. Shanmuganathan, and S. E. Mirvis Images of Tracheal Rupture Am. J. Roentgenol., January 1, 2002; 178(1): 241 - 241. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
