Role of Transesophageal Echocardiography in the Diagnosis and Management of Traumatic Aortic Disruption

Acute traumatic disruption of the thoracic aorta (TDA) is a highly lethal but treatable injury that usually occurs as a result of high-speed deceleration accidents. Abrupt deceleration injuries generate shearing forces that act maximally in the region of the aortic isthmus where the mobile aortic arch becomes relatively fixed to the thoracic cage by the ligamentum arteriosum, the intercostal arteries, and the left subclavian artery.¹ These anatomic considerations explain why TDA occurs in over 90% of cases involving the region of the aortic isthmus,² defined as the aortic segment between the left subclavian and first intercostal arteries. Because the major complication of TDA is adventitial rupture, usually resulting in immediate death secondary to massive hemorrhage, only a small number of these patients live long enough to successfully undergo reparative surgery.³ Early diagnosis of TDA followed by immediate surgical repair is therefore imperative to improve current survival rates in victims involved in violent motor vehicle accidents.²

Because of their low sensitivity and specificity, screening tests such as chest x-ray⁴ and contrast-enhanced computed tomography⁵ are not sufficiently accurate to diagnose TDA. Consequently, aortography, which is currently considered the gold standard diagnostic technique in this entity, is routinely performed when patients suffering a violent deceleration injury present to the emergency room with a widened mediastinum on chest x-ray.⁶ Unfortunately, aortography is invasive and extremely difficult to perform in hemodynamically unstable patients with multiple trauma. Moreover, aortography usually delays the initiation of aortic surgical repair by an average of 50 to 75 minutes.⁷⁸ Transesophageal echocardiography (TEE), which can be performed easily at the patient’s bedside, appears to be a rapid, safe, and accurate alternative diagnostic approach for this condition because it provides unparalleled visualization of the entire aortic isthmus.^{78910111213141516171819} Accordingly, we prospectively performed TEE in consecutive patients with suspected TDA to determine the diagnostic accuracy and impact on immediate patient management of this alternative imaging modality.

Methods

During a 24-month period, all patients admitted to our institution with (1) multisystem trauma or isolated severe blunt chest trauma associated with a violent deceleration injury due to a head-on collision and (2) widened mediastinum (>8 cm) on admission chest x-ray were prospectively enrolled into the study. TEE was always attempted during the patient’s initial evaluation unless the necessary equipment and personnel were not available. Aortography, surgery, or necropsy were used to confirm or exclude the diagnosis of TDA. In each patient, the Injury Severity Score was calculated as described previously.²⁰ Subsequently, patients were divided into those with (TDA group) and without (control trauma group) confirmed acute traumatic disruption of the thoracic aorta. To establish TEE criteria that would distinguish TDA from acute aortic dissection, consecutive patients with acute aortic dissection involving the aortic isthmus as diagnosed with use of TEE and confirmed by either aortography, surgery, or contrast-enhanced computed tomography were also enrolled into the study.

All TEE examinations were performed with use of either a 5-MHz single-plane or multiplane transducer (Hewlett-Packard SONOS 1500). Before the transesophageal examination, patients were sedated with an intravenous short-acting benzodiazepine. The TEE study included standard views of the heart followed by a complete two-dimensional and color flow mapping examination of the ascending, horizontal, and descending thoracic aorta, with particular attention to visualizing the aortic isthmus.²¹ With use of the multiplane TEE probe, transverse (0°) to longitudinal (90° to 125°) views of the aortic isthmus were obtained. Attempts were made to visualize the proximal innominate and left subclavian arteries by carefully examining the corresponding anatomic segments.²²

TEE studies were retrospectively reviewed by an experienced observer who was blinded to both the patient’s group and the results of the alternative comparative method of diagnosis. Both anteroposterior and lateral diameters of the aortic isthmus as well as thickness of either medial or intimal flaps were measured from freeze-frames by use of electronic calipers. Aortograms of trauma patients were interpreted by trained radiologists who were unaware of the echocardiographic findings.

Once the diagnosis of subadventitial TDA was established, rapid surgical repair was attempted and surgical findings were correlated with TEE results. In contrast, patients with traumatic injuries involving solely the aortic intima were managed conservatively by use of serial clinical and TEE follow-up. Complications of both TEE and aortography were recorded for each patient.

Results

Patients and Diagnostic Methods

Among the 115 trauma patients admitted to the intensive care unit during the study period, 32 patients (25 men, 7 women) met the inclusion criteria. Of these, 14 patients constituted the TDA group (11 had subadventitial tears and 3 had intimal tears). The remaining 18 patients were included in the control trauma group (violent deceleration accident, widened mediastinum on chest x-ray, normal thoracic aorta on both aortogram and TEE examination). With the exception of 1 patient who died of associated severe head injuries, all patients in the control trauma group were discharged from the hospital. Mean age of all trauma patients was 40±16 years (mean±SD; range, 16 to 69 years old), and the mean Injury Severity Score²⁰ was 46±24 (range, 13 to 75). Eighteen patients (15 men, 3 women) were enrolled into the aortic dissection group. Mean age of these patients was 57±15 years (range, 26 to 83 years old). According to the classification of DeBakey et al,²³ 12 patients suffered type I acute aortic dissections, whereas 6 sustained type III dissections (Table 1).

The transesophageal, angiographic, and anatomic findings corresponding to the TDA group are summarized in Table 2. All traumatic aortic lesions were confined to the region corresponding to the aortic isthmus. TEE findings (ie, subadventitial lacerations, pseudoaneurysms, and mediastinal hematomas) were all confirmed either during surgery or at necropsy. No complications were encountered during TEE or aortography.

Subadventitial Traumatic Aortic Disruptions

The diagnosis of subadventitial TDA requires the presence of a disrupted aortic wall with blood flow on both sides of the disruption. Transesophageal two-dimensional echocardiographic findings consisted of the presence of an abnormal intraluminal “thick flap,” usually accompanied by a regional deformity of the aortic isthmus contour caused by the formation of an acute localized pseudoaneurysm (Fig 1). Rupture of the pseudoaneurysm wall, which consists solely of the adventitial aortic layer under tension, is responsible for the cataclysmic nature of TDA deaths. Interestingly, despite the presence of the localized deformity, the aortic isthmus size usually remained within the normal range (Table 2). In most cases, TDA involved a large section of the aortic circumference, and the “thick flap” that corresponds to the disrupted aortic isthmus wall was easily detected. These flaps, which appear as thick, protruding membranes (mean±SD, 4.2±0.8 mm), consisted of the entire intimal and medial aortic layers (Table 2). To distinguish this structure from the “thin” intimal flaps noted in patients with acute aortic dissection (mean±SD, 2.2±0.7 mm), we decided to call the intraluminal membranes found in TDA “medial flaps.” In subtotal subadventitial aortic disruptions (n=8), the medial flap appeared in the transverse view as a linear structure completely traversing the lumen of the aortic isthmus in either a vertical, oblique, or near-horizontal manner (Fig 2), usually corresponding to a spiral tear at surgery (Fig 3). In longitudinal views obtained with use of the multiplane probe (n=3), the medial flap was always near-perpendicular to the isthmus wall, vertically traversing the entire aortic lumen (Fig 4). In complete subadventitial aortic disruption (n=1), it appeared as an open circle contained within the aortic lumen (Fig 5), corresponding to a horizontal tear at surgery (Fig 3). Regardless of the type of aortic tear, the medial flap was usually extremely mobile and oscillated with each cardiac contraction (Table 2). Color flow mapping depicted similar blood flow velocities on both sides of the medial flap, and a mosaic of colors was always observed at the site and vicinity of the aortic wall disruption, because of local blood flow turbulence (Figs 2 and 4). Finally, in partial subadventitial disruption, the aortic tear appeared as a discontinuity of the intimal and medial aortic layers, without evidence of an intraluminal medial flap (Table 2). Color flow mapping demonstrated turbulent flow entering a pseudoaneurysm, the wall of which consisted solely of the aortic adventitial layer (Fig 6).

All patients with subadventitial TDA underwent successful surgical repair with subsequent hospital discharge, with the exception of two patients who died of associated head injuries before surgery. In one of these patients, necropsy confirmed the TEE diagnosis of a partial subadventitial TDA, whereas in the second patient, the anatomic specimen demonstrated a limited 2-mm medial aortic tear with integrity of the adventitia that was not diagnosed by TEE (Fig 7). None of the above-mentioned echocardiographic signs were observed in the control trauma group. As a result, TEE accurately diagnosed the presence or absence of subadventitial TDA in every patient, with a sensitivity of 91% and a specificity of 100%.

Traumatic Aortic Intimal Tears

Traumatic tears limited to the aortic intima (n=3) appeared as thin, mobile intraluminal appendages of the aortic wall, located at or in the immediate vicinity of the aortic isthmus. Because these lesions are too small and superficial to exert excessive pressure on the adventitial layer, the diameters and contour of the aortic isthmus remained unchanged (Table 2). Color flow mapping usually failed to demonstrate a mosaic of colors because of the absence of blood flow turbulence surrounding the aortic intimal tear (Fig 8). None of these echocardiographic signs were observed in the control trauma group. Aortography failed to confirm the presence of an intimal tear in two cases. All patients diagnosed with intimal tears were successfully managed conservatively with serial clinical and TEE follow-up (Table 2).

Mediastinal Hematomas

With the exception of two cases, mediastinal hematomas were noted in all patients in the TDA group (Table 2). TEE findings associated with this injury have been described previously.²⁴ Interestingly, the presence of mediastinal hematomas was also observed in seven patients (39%) in the control trauma group and in three patients (17%) in the aortic dissection group.

Discussion

In the present study, we have shown that because of its safety, accuracy, and portability, TEE should be considered as the first-line imaging modality for diagnosis of TDA. On the basis of the TEE experience gathered in the present study, we will propose an alternative classification of nonpenetrating traumatic aortic injuries that could be clinically useful in determining immediate patient management. Additionally, the differential TEE diagnostic criteria between subadventitial TDA and acute aortic dissection will be highlighted.

Imaging Modalities for the Diagnosis of TDA

Currently, trauma patients involved in violent deceleration accidents who present with a widened mediastinum on chest x-ray usually undergo aortography to rule out the presence of TDA. Despite its numerous disadvantages, aortography is still widely regarded as the gold standard imaging modality for diagnosis of TDA.⁴⁵⁶ Several radiographic signs have been described as indicators for performing aortography in subjects involved in abrupt deceleration collisions who present with blunt chest trauma.⁴⁶ Because none of these radiographic signs are sensitive or specific for TDA, aortography yields positive results in only 10% to 20% of patients.⁴ To limit the number of negative aortograms in patients with suspected TDA, contrast-enhanced computed tomography has been suggested and evaluated as a potential alternative screening procedure. Although this imaging modality is of great diagnostic value in the initial assessment of pleural and pulmonary injuries, this technique has an extremely low sensitivity and specificity for the diagnosis of major vascular traumatic lesions.⁵ Whereas magnetic resonance imaging is valuable for diagnosis of acute thoracic aortic dissection, this test requires a hemodynamically stable and cooperative patient for its completion.²⁵ The need for transportation of patients who require continuous monitoring and multiple life-support equipment and who are frequently unstable usually precludes the performance of magnetic resonance imaging in severely injured subjects with suspected TDA.

For these reasons, aortography is still widely considered to be the gold standard imaging modality for diagnosis of major arterial injuries.⁴⁵⁶ Because aortography usually requires transportation of a patient with multiple injuries to either the cardiac catheterization or radiological suite, it has similar disadvantages in this respect to magnetic resonance imaging. In the setting of acute trauma, aortography has a morbidity of 10%, even in patients without TDA.⁶ In addition, this invasive diagnostic test usually delays the initiation of reparative surgery by an average of approximately 1 hour.⁷⁸ Finally, even when reviewed by experienced physicians, aortography may result in false-positive or false-negative interpretations,⁷⁹¹⁰¹¹ as it did in two patients in our study (Table 2).

Transthoracic echocardiography has been described infrequently as being useful in the assessment of blunt trauma to the chest, mainly because this technique does not provide optimal visualization of the aortic isthmus.²⁶²⁷ Because of the close anatomic proximity between the esophagus and the descending thoracic aorta, TEE provides high-resolution images of the aortic isthmus.²¹ Since no false-positive diagnoses of TDA were encountered in the present study, TEE had a specificity of 100%. The sensitivity of TEE was 91% because in one case, a small, 2-mm medial aortic rupture with integrity of the adventitial layer was seen at necropsy but was missed by TEE (Fig 7). Our results are in agreement with those of Kearney et al⁷ and Smith et al.⁸

Differential Diagnosis With Aortic Dissection and Atherosclerotic Lesions

Because traumatic aortic dissection may occur secondary to chest trauma at any age,²⁸ it is crucial to establish the differential echocardiographic diagnostic criteria between the latter and subadventitial TDA (Table 3). The medial flap observed in subtotal and complete subadventitial disruptions is mobile and therefore may be erroneously diagnosed as the intimal flap of aortic dissection.¹⁷ In the present study, the thickness of the medial flap associated with TDA was significantly greater than that noted in aortic dissection (4.2±0.8 mm versus 2.2±0.7 mm: P<.001, t test) (Figs 1 and 5). However, this differential diagnostic criteria should be used with caution, since cleavage of the media by a dissecting hematoma may occur at variable depths, resulting, at times, in thicker intraluminal flaps (Table 1). Additionally, the mobility of the medial flap was consistently greater than that noted in intimal flaps of aortic dissections (Table 2). In the longitudinal view, the intimal flap was always parallel to the isthmus wall (n=14) because, in all cases, the dissection extended toward either the aortic arch or the descending aorta. In contrast, the medial flap of TDA was always perpendicular to the aortic isthmus wall and traversed the entire aortic lumen (Fig 4). Whereas in TDA the aortic contour was generally deformed because of formation of a localized pseudoaneurysm, in aortic dissection the isthmus was usually enlarged in a symmetrical manner (Table 1). Although mediastinal hematomas were unusual in patients with acute aortic dissection, they were commonly encountered in trauma patients, even in the absence of TDA, because of associated laceration of small mediastinal vessels.

Interestingly, the most important differential echocardiographic sign was that the intraluminal image of the disrupted aorta did not result in two distinct channels, as is usually seen in aortic dissection. Consequently, blood flow velocity as evaluated by color Doppler mapping was similar on both sides of the medial flap. Conversely, in aortic dissection, blood flow velocity was slower in the false lumen when compared with the true lumen (Figs 1 and 5). Moreover, whereas the presence of an unorganized thrombus in the false lumen was observed in 28% of aortic dissection cases (Table 1), the latter finding was not observed in TDA and has been reported only exceptionally after a trauma episode.¹⁰²⁹ Partial subadventitial disruptions appeared as a discontinuity of the isthmus wall and could be diagnosed erroneously as the entry tear commonly observed in aortic dissection. Distinctively, the aortic wall tear of partial subadventitial disruption was wider than the entry tears of aortic dissection observed in the region of the isthmus, and its appendages were more mobile than the intimal flap (Fig 6). Additionally, in the case of partial disruption, color Doppler mapping depicted turbulent blood flow in systole entering a pseudoaneurysm, as reported previously,¹⁵¹⁶ whereas in aortic dissection, blood flow at times was noted entering the false lumen via the entry tear (Fig 9). Finally, ultrasonographic findings in TDA, regardless of its anatomic type, were confined to the region of the aortic isthmus,¹⁴ whereas in aortic dissection, TEE findings are more extensive.²⁵

Protuberant intimal atherosclerotic plaques or pedunculated thrombus could also be misdiagnosed as traumatic aortic intimal tears (Fig 8). However, atherosclerotic lesions are extensive and usually exhibit calcifications.

Correlation Between Echocardiographic and Anatomic Findings: Potential Impact on Patient Management

In 1958, Parmley et al³ classified nonpenetrating traumatic aortic injuries into six categories on the basis of anatomic specimens obtained from necropsy cases of traumatic injuries of the heart and aorta: (1) intimal hemorrhage; (2) intimal hemorrhage with laceration; (3) medial laceration; (4) complete laceration of the aorta; (5) false aneurysm formation; and (6) periaortic hemorrhage. Using the anatomic classification of Parmley et al and our TEE findings, we encountered four distinct types of traumatic aortic injuries.

Traumatic aortic intimal tears. In these lesions, the integrity of the aortic medial and adventitial layers is preserved (type 2 lesion by the classification of Parmley et al³ ). Associated TEE features are described in Table 2 and shown in Fig 8. The true incidence of traumatic aortic intimal tears is probably underestimated currently because they usually remain undetected by aortography (Table 2).¹⁰ Because these lesions appear to regress spontaneously,¹⁰³⁰³¹ conservative management with use of serial clinical and TEE follow-up is proposed. This therapeutic alternative was successful in all three of our cases, as confirmed by TEE follow-up that depicted regression (n=2) or lack of extension (n=1) of these lesions (Table 2).

Subadventitial traumatic aortic disruptions involve the entire aortic intimal and medial layers. In these lesions, the threat of adventitial rupture causing massive hemorrhage is constantly present. Depending on the extent of the traumatic aortic wall tear, we encountered three distinct types of TEE findings.

Subtotal aortic disruption. These lesions involve more than two thirds of the aortic wall circumference. A narrow band of aortic wall, usually found in the posterior aspect of the aorta, secured the disrupted aortic segments a few centimeters apart (Figs 2 and 3). In the current series, most of the subtotal aortic tears were spiral (Table 2).

Complete aortic disruption. These lesions involve the entire aortic circumference (Figs 3 and 5). In the present study, in both subtotal and complete subadventitial aortic disruptions, medial flaps were always observed. The presence of a subadventitial complete horizontal disruption was associated with a circular pattern of the medial flap, whereas the medial flap was linear in cases of subtotal lacerations (Table 2).

Partial aortic disruption. These injuries appear as a limited discontinuity of both intimal and medial layers. They can be associated with (Fig 6) or without pseudoaneurysm formation.⁷ Because of resolution limitations of currently available ultrasonic imaging equipment, partial aortic disruptions may remain undiagnosed (Fig 7).

Because patients with subadventitial aortic disruptions may die suddenly secondary to a rupture of the aortic adventitial layer, these patients should be considered for immediate surgical repair. In this study, three hemodynamically unstable patients were sent for immediate reparative surgery solely on the basis of TEE findings, thus avoiding the delay associated with aortography. However, in particular circumstances, delayed reparative surgery has been proposed.³²

Study Limitations

Despite its relatively small size, this series constitutes the first detailed description of TEE findings associated with different anatomic types of traumatic aortic injuries and establishes the echocardiographic differential diagnostic criteria for acute aortic dissection. Since conservative management was instituted for patients who sustained traumatic aortic intimal tears, no surgical confirmation was readily available for those lesions. Accordingly, these patients were not included in the calculation of the diagnostic sensitivity and specificity of TEE in TDA. The successful management of our patients on the basis of TEE findings has yet to be confirmed by other investigators.

The aortic isthmus can be visualized adequately by use of both single-plane and multiplane TEE transducers.²¹²² Because all aortic ruptures in this study were confined to this region, the diagnosis was made correctly despite the frequent use (44% of trauma patients) of single-plane TEE probes. However, since TDA may occasionally occur in regions that are better imaged with multiplane scopes, such as the distal ascending aorta,²⁸³³ we recommend using the multiplane probe whenever possible. Moreover, we found that the additional views obtained with multiplane TEE probes are particularly useful in determining the length of subadventitial aortic disruption and in ruling out the presence of traumatic intimal tears. Until further experience becomes available, aortography is still required when lacerations of either the aortic arch or the supra-aortic arteries are suspected. Similarly, aortography should also be performed whenever TEE findings are equivocal for the diagnosis of TDA.³⁴ In the future, intravascular sonography could become an additional useful diagnostic tool.³⁵³⁶

Conclusions

This study demonstrates that TEE can be performed safely in the setting of patients who have sustained severe blunt chest trauma, resulting in clinical information that can be used for determining the immediate management of patients. Because of its accuracy, portability, and safety, TEE is better suited than aortography as the first-line imaging modality for evaluation of patients with multiple trauma and suspected acute TDA.

• Download figure
• Download PowerPoint

• Download figure
• Download PowerPoint

• Download figure
• Download PowerPoint

• Download figure
• Download PowerPoint

• Download figure
• Download PowerPoint

• Download figure
• Download PowerPoint

• Download figure
• Download PowerPoint

• Download figure
• Download PowerPoint

• Download figure
• Download PowerPoint

Footnotes