Evaluation of Transesophageal Echocardiography for Diagnosis of Traumatic Aortic Injury

This study was approved by the hospital’s ethics committee (CCPPRB Pitié-Salpêtrière). All blunt trauma patients in whom TAI was suspected were prospectively included over a 9-yr period, ending in June 1999. Criteria for inclusion were either enlarged mediastinum on the chest radiograph or sudden deceleration. All of these patients underwent TEE for the assessment of aorta, volume status, and cardiac function. Moreover, whenever possible, either aortography or a contrast-enhanced spiral CT scan (since 1996) was performed to diagnose TAI, and all such examinations were reviewed by one of the authors (P. C.). Nevertheless, in few patients with hemodynamic instability, emergency surgery was decided only after TEE.

Transesophageal echocardiography was performed on arrival in the trauma center with a 5-MHz phased array probe (HP1500; Hewlett-Packard, Andover, MA). The probe was a single-plane probe at the beginning of the study and a multiplane probe since 1992. Standard TEE examination included visualization of the ascending aorta, aortic arch, and descending thoracic aorta, as well as based short-axis, four-chamber, and transgastric short-axis cardiac view, which permitted the assessment of volume status, cardiac function, pericardium, and mediastinum. These data were recorded and analyzed on a real-time basis by anesthesiologists fully trained in echocardiography. All recorded data were reviewed by two of the authors (J. P. G. and M. G.). As previously described, TAI was diagnosed when direct, specific signs were observed, 9including thick medial flaps, false aneurysm, aortic dissection, free-edge intimal flap, aortic wall hematoma, fusiform aneurysm, and aortic obstruction. 9When indirect, nonspecific, signs of TAI were observed, examination was carefully continued to exclude the possibility of TAI. Indirect signs included hemomediastinum, 13minor increases in aortic diameter, and impairment of Doppler color flow. 9 

Traumatic aortic injuries were graded according to therapeutic implication (table 1). Grade 1 corresponds to minor TAI when isolated intimal flap or intramural hematoma were observed without any evidence of hemomediastinum or modification of the geometry of aorta. Minor TAIs do not indicate surgery but warrant close medical follow-up. 9,11Grade 3 corresponds to major TAI requiring immediate emergency surgery because of either aortic transsection, and thus active bleeding, or complete aortic obstruction, inducing ischemia. 2,9Grade 2 corresponds to other TAIs with subadventitial rupture in which medial injury or modification of the geometry of the aorta were observed, with or without hemomediastinum, thus indicating potential need for either immediate or delayed surgery. 1,2Consequently, for assessing the accuracy of TEE, patients with TAI were divided into two groups: major TAI (grades 2 and 3) and minor TAI (grade 1).

Data are expressed as mean ± SD or percentage. Comparison of two means was performed using the Student t  test, and comparison of two percentages was performed using the Fisher exact method. The sensitivity, specificity, and negative and positive predictive values of TEE and angiography (aortography or contrast-enhanced spiral CT scan) were calculated for the diagnosis of minor TAI, major TAI, and all TAI. Although major TAI were confirmed by surgery, it should be pointed out that most minor TAI were diagnosed only by TEE, and thus that TEE was actually considered as a gold standard technique in the present study. All P  values were two-sided, and a P  value less than 0.05 was considered significant.

During the study period, 1,890 trauma patients were admitted into our trauma center, and 209 (11%) blunt trauma patients (mean age, 34 ± 13 yr; 155 men and 54 women) were suspected of having TAI because of enlargement of mediastinum (n = 120, 57%) or sudden deceleration. The cause of trauma was a motor vehicle accident in 142 cases (68%) and a fall (13 ± 5 m) in 67 cases (32%). All but one patient underwent TEE. In this patient, emergency aortic surgery was performed without any imaging except chest radiograph because of cardiac arrest. TAI was confirmed at surgery, but the patient died. Aortography was performed in 165 cases (79%), and contrast-enhanced spiral CT scan was performed in 111 cases (53%), the two examinations being performed in 70 cases. Only two patients underwent TEE only because of hemodynamic instability, and TAI was confirmed in all cases either by emergency surgery (n = 1) or autopsy (n = 1). TAI was diagnosed in 42 cases (20% of patients suspected of having TAI, and 2% of all trauma patients) and located at the isthmus in 41 cases and at the origin of the left subclavian artery in one case.

Transesophageal echocardiography was analyzed in 208 patients, in whom TAI was diagnosed in 41 cases. A single-plane probe was used in eight cases, and a multiplanar probe was used in the remaining 33 cases. Table 2depicts the comparison between patients with and without TAI. TAI was minor in seven patients who did not undergo surgery but only follow-up. Major TAI was diagnosed in 33 patients (grade 2, 17 cases; grade 3, 16 cases), 30 of whom underwent immediate or delayed surgery. In two patients, death occurred before surgery because of head trauma or multiple organ failure. Another patient refused surgery and left the hospital, and follow-up information could not be obtained.

Table 3depicts the diagnostic value of TEE compared with angiography. Angiography was less accurate than TEE for the diagnosis of TAI because it failed to diagnose most minor TAIs. However, when considering only patients with major TAIs, angiography and TEE were equivalent (table 3). TEE was negative in one patient whose angiography was positive, and angiography was negative in another patient whose TEE was positive.

Among the seven patients with minor TAIs, two died because of severe head trauma or multiple organ failure, and two were lost to follow-up. Three patients underwent TEE several years later (range, 17 months to 6 yr). In two patients, TEE showed an almost complete regression with only a localized (> 1 cm) thickening of the aortic wall or an irregular intimal layer at the isthmic portion of the aorta. In the third case, TEE was performed again 17 months after trauma and showed the same lesion observed during the initial examination: significant intramural hematoma (extended on 4 cm) without aortic dilation or modification of aortic external layers, without any color Doppler flow abnormalities. An annual follow-up examination was decided for this patient.

In this study, we selected prospectively a population of trauma patients at high risk for TAI. Indeed, all patients underwent rapid deceleration during trauma, and most showed a widened mediastinum on chest radiograph. A widened mediastinum was not the only criterion for inclusion because it has been demonstrated that TAI can occur in trauma patients without widened mediastinum. 14In the same manner, external chest trauma lesions were not considered as potential criteria for inclusion because they do not correlate with TAI occurrence. 15The incidence of TAI (20%) in our selected population of trauma patients was greater than those previously reported (4–17%) 5–7but comparable to that reported by Minard et al.  (26%). 8 

The apparent high number of publications dealing with TEE and TAI may be confusing. Indeed, most of these publications were only case reports, and some of the series published included the same patients, 6,7,10,16,17potentially providing a misleading impression of the number of patients studied and the value of TEE. Although our previous report only analyzed the TEE signs, 16 of our 42 patients were previously described. 9Thus, only four teams have evaluated the accuracy of TEE in the diagnosis of TAI 5–8but provided conflicting results concerning sensitivity (57–91%) and specificity (84–100%). Because TAI does not occur frequently, most of these studies included few patients with TAI (n = 8–25). 5–8The echocardiographic signs of TAI are complex and have been only recently described. 9,11Obviously, the low sensitivity reported by two studies 5,8may be partly explained by numerous inconclusive TEE. Some of these inconclusive investigations were probably related to minor aortic injuries that were not fully recognized at the time. 5,8Moreover, the study by Saletta et al.  6was retrospective. Our study showed that TEE has a high sensitivity and specificity, as previously suggested. 7When considering only patients with major TAIs who required either immediate or delayed surgery, TEE and angiography seemed equivalent. Our study suggests that no diagnostic procedure is perfectly sensitive because a false-negative result was observed both with TEE and angiography. However, no false-positive results were reported with TEE, and TEE enabled us to not indicate surgery in a patient with a false-positive angiography.

Transesophageal echocardiography has allowed the description of limited traumatic lesions of the aorta, including small intimal lesions and intramural hematomas (fig. 1). 9,11These lesions are not associated with hemomediastinum or any modification of the geometric shape of the aorta. Consequently, these limited lesions are usually missed by angiography, as confirmed in the present study (table 3). The clinical implications of these minor aortic injuries remain presently unknown. Although most of the cases reported did not undergo surgery and seemed to have complete healing of the lesions during follow-up, 9,11careful follow-up or medical treatment may be required. 2Indeed, limited spontaneous nontraumatic aortic lesions (intramural hematoma and perforating atheromatous ulcer) are thought to have the same bad prognosis as spontaneous aortic dissection. 18The further development of posttraumatic false aneurysm cannot be presently excluded in patients with these limited traumatic aortic lesions. Long-term follow-up was available in only three patients with minor TAIs and was uneventful, although the lesions were stable in one patient after 17 months. Further study including a larger group of patients with minor TAIs and a longer follow-up period is mandatory to answer this important question concerning the prognosis of minor TAIs. Because of the scarcity of minor TAIs, this study should involve many trauma centers.

Transesophageal echocardiography provides useful information in patients with subadventitial rupture and severe multisystem trauma. For some of these patients, aortic surgery was intentionally postponed. Anatomic type of parietal lesions, size of false aneurysm, and presence of hemomediastinum all play a part in the decision-making process. TEE provides some useful information that is not provided by aortography or CT scan. Indeed, TEE can assess volume status and cardiac function and is very accurate in diagnosing other thoracic traumatic lesions such as hemopericardium, myocardial contusion, and rare valvular lesions. 19Moreover, TEE is a rapid bedside examination, enabling emergency decisions to be made in case of complete aortic transection or complete aortic obstruction. 9In contrast, TEE can miss supraaortic vessel lesions that are diagnosed by angiography, 2,20but these lesions are far more rare than aortic isthmus injury. 2Expertise in performing TEE may be of paramount importance to obtain a high level of accuracy, as previously emphasized. 7It should be pointed out that all TEE examinations were carefully reviewed by echographists highly trained in treating trauma patients and with much experience with TAI. Vignon et al.  7demonstrated that a learning period is required to improve the accuracy of TEE in the diagnosis of TAI. It is thus not surprising that the less favorable results were from reported series that included few patients with TAI (n = 8–9). 6,8 

Some remarks must be included to assess the relevance of our results. First, from a methodologic point of view, investigators were not blinded, and some variations in the diagnostic techniques used occurred during the study period (single plane vs.  multiplane probe for TEE, addition of spiral CT scan). Nevertheless, a multiplane probe was introduced very early in the present study, and the accuracy of contrast-enhanced spiral CT scan is thought to be equivalent to that of arteriography. Second, no gold standard method exists for the diagnosis of TAI, which may represent a weakness of our analysis. For patients with severe TAI, the results were confirmed by surgery or autopsy, and for patients without TAI, at least two negative examination results were obtained (TEE and arteriography or CT scan), which presently represents an acceptable manner to rule out severe TAI. Nevertheless, most patients with minor TAI were only diagnosed by TEE, which was the technique evaluated in the present study.

In conclusion, in a large prospective study in the same trauma center, we observed that TEE was more accurate than angiography for the diagnosis of TAI, mainly because angiography failed to evidence minor aortic injury. However, when considering only major TAI, TEE and angiography seemed equivalent.

The authors thank Dr. David Baker (Hôpital Necker-Enfants Malades, Paris, France) for reviewing the manuscript.