ThoracicAortography FollowingBlunt ChestTrauma JAMES T. STURM, MD, DANIEL G. HANKINS, MD, GEORGE YOUNG, MD The records 01314 patients who suffered blunt chest trauma and underwent thoracic aortography between 1968 and 1985 were retmspectlvely reviewed. The patients ranged In age fmm 7 to 84 years (mean, 37.7 years). There were 255 male and 59 female patlents. The majority of injuries were the result of motor vehicle accidents. The most commen indication for aortogmphy was a widened mediastinum on chest roentgenogmm (83.4%). The aortogmm was positive for cardiovascular injury in 19.7% of cases. There were 47 patlents with aertic rupture, 15 with subclavian artery disruption, and 1 with traumatlc aortic insufficiency. Compllcatlons occurred in 1.7% 01 patients. Two patients sustained groin hematomas and one patient suffered an intimal tear of the ascending aorta fmm the angiogmphic catheter. None of the complications required treatment. Aberrant origin of the arch vessels occurred In 0.96% of patients, and ductus dlvertlculum occurred In 0.54%. There were two falsepositive and ne false-negatlve aortogmms. It was concluded that thoracic aortography after trauma is accurate and safe. (Am J Emerg Med 1990;8:92-95. 0 1990 by W.8. Saunders Company.)
The gold standard for the diagnosis or exclusion of traumatic aortic rupture (TAR) following blunt chest trauma is the aortogram despite the advent of newer technologies such as computerized tomography (CT) and magnetic resonance imaging (MRI).re4 The thoracic aortogram not only shows the site and extent of the traumatic aortic rupture, but also defines the anatomy of the branches of the aortic arch and their possible involvement in this lesion. These anatomic considerations are important when considering surgical correction of this injury. Computerized tomography and MRI have proved to be useful tools in diagnosing dissecting aneurysms of the thoracic aorta but have been less useful in the diagnosis or exclusion of traumatic aortic rupture.5.6 This is because multiple cuts must be made for CT or MRI and the plane of such cuts may miss the intimal flap or extravasation of contrast material from the lumen. However, both techniques are evolving rapidly, and the importance of each for the diagnosis of TAR is promising. We retrospectively reviewed 18 years’ experience with thoracic aortography following blunt chest trauma at the St Paul-Ramsey Medical Center to define the reasons thoracic aortography was performed, the results of the procedure, and the complications encountered. We also sought to quantitate the frequency of the aberrant origin of vessels from the From the Departments
of Emergency Medicine and RadiolMedical Center, and the Ramsey Clinic, St
ogy, St Paul-Ramsey Paul, MN. Address reprint requests to Dr Sturm: 640 Jackson St, St Paul, MN 55101. Supported in part by the St Paul-Ramsey Foundation. Manuscript received March 10, 1989; revision accepted May 10, 1989. Key Words: Aortic rupture, aortic injury, chest trauma. 0 1990 by W.B. Saunders Company. 0735-6757/90/0802-0002$5.00/O 92
arch and the number of false-positive terpretations of thoracic aortography.
METHODS The records of 314 patients who suffered blunt chest trauma and underwent thoracic aortography between the years 1%8 and 1986 were reviewed retrospectively. The etiologies of the injuries, the principal intrathoracic diagnoses, the indications for the procedures, the diagnostic results of aortography, and the complication rates were tabulated. There were 255 male and 59 female patients ranging from 7 to 84 years (37.7 k 18.4 years, mean k SD). The most common causes of injury were automobile and motorcycle accidents, followed by pedestrian trauma and injuries from falls (Table 1). Sufficient information was available from the charts of 229 patients to determine the principal clinical or radiographic indication for obtaining aortography. There was adequate information in the charts of 181 patients to determine whether or not complications of aortography occurred. The sites of the TAR at aortography were tabulated. In addition, the incidence of anatomical variance of the origin of the aortic arch vessels, the numbers of false positives and negatives, and the occurrence of ductus diverticulum was noted. In most cases, thoracic aortography was performed by the transfemoral route with 6F to 7F pigtail catheters. With three patients, the catheter could not be passed beyond the intimal flap of the aortic tear. In two of these three cases, angiography was aborted and surgery was performed. In the third, the catheter was removed from the femoral artery and an antegrade aortogram was performed through the right axillary artery. A 40- to 70-mL bolus of diatrizoate megluminel sodium was injected at 35 to 45 mL/s, and films were obtained at 3 per second for 3 seconds and 1 per second for 5 seconds. The left anterior oblique (LAO) was the standard projection used. To outline other abnormalities, a straight catheter was introduced into the brachiocephalic, subclavian, or carotid vessels when clinically or radiographically indicated after the initial arch aortogram was exposed. RESULTS The most common reason the clinician obtained thoracic aortography was the presence of a wide mediastinum on chest roentgenogram in 191 of the 229 cases (83.4%) (Fig 1). Other indications for obtaining thoracic aortography were the unilateral decrease or absence of an upper extremity pulse, a distorted and prominent aortic shadow on chest roentgenogram, the presence of a left apical hematoma, sternal fracture, or first rib fracture. The other indications for aortography are listed in Table 2. The thoracic aortogram was positive for cardiovascular
STURM, HANKINS, AND YOUNG n AORTOGRAPHY FOLLOWING TRAUMA
TABLE2. Indications for Obtaining Thoracic Aortography in 229 Patients
Etiology of Injuries Cause of Injury
No. of Patients
Automobile accident Motorcycle accident Pedestrian mishap Fall injuries Crush injuries Snowmobile accident Bicycle mishap Altercation Other
212 31 26 24 11 4 2 1 4
injury in 64 patients. There were 47 patients with aortic rupture, 15 with rupture of the subclavian artery, 1 with axillary artery disruption, and 1 with traumatic aortic insufftciency, giving an incidence of 20.3% of aortograms positive for cardiac or vascular injury. The site for aortic rupture was distal to the left subclavian artery in all 47 cases. One patient with an ascending aortic rupture was operated on without angiography for signs of cardiac tamponade, and two patients with descending ruptures were operated on without aortography because of profuse hemorrhage from a left thoracostomy tube. The subclavian artery disruptions were at or near the origin of the vertebral artery. The accuracy rate of aortography was 99.3% with a sensitivity of 100% and a specificity of 99.2%. The positive predictive value was 97%, and the negative predictive value was 100%. Complications occurred in 3 of 181 patients in whom this information was available on the chart (1.7%). Two patients sustained groin hematomas, and one patient suffered an intimal tear in the ascending aorta from the angiographic catheter during the course of the procedure. The two patients
Wide mediastinum Unilateral pulse decrease or absence in the upper extremity Distorted aortic shadow Left apical hematoma Persistent hypotension Extensive chest trauma Chest pain Sternal fracture First rib fracture Hemothorax Other
6 5 3 3 3 2 2 2 2 6
with groin hematomas had resolution of the hematoma without infection or need of surgical drainage. The patient with an intimal tear in the ascending aorta suffered no further complications and left the hospital without evidence of dissecting aneurysm. There were no reactions to the contrast medium that required medical treatment. In three patients with aortic rupture, the angiographic catheter could not be passed retrograde from the femoral artery. This was because the catheter caught in the intimal flap and could not be safely advanced. In two of the three patients, surgical correction of the aortic rupture was undertaken without angiographic visualization of the aortic arch, and in the third patient, a right axillary approach was used to obtain an antegrade aortogram of the ascending aorta and arch. Two patients died in the angiography suite just before the procedure. These deaths were due to multiple severe injuries and unrelated to the aortographic procedure. Two patients (0.64%) had angiograms falsely interpreted as positive. At the time of the operation, they had periaortic hematomas that had created double densities in the area of the left subclavian artery, suggesting extravasation of contrast and enlargement in the vicinity of the left subclavian TABLE3.
FIGURE 1. The chest roentgenogram of a patient with traumatic aortic rupture shows (a) obliteration of the aortic outline, (b) widened mediastinum, (c) left apical pleural cap, and (d) tracheal shift to the right.
No. of Patients
No. of Patients
Hemothorax Aortic rupture Flail chest Pulmonary contusion Chest contusion Multiple rib fracture Subclavian artery rupture Thoracic vertebral fracture Fractured sternum Cardiac contusion Ruptured diaphragm Clavical fracture Scapula fracture Axillary artery rupture Pulmonary laceration Traumatic aortic insufficiency
51 47 47 46 36 36 15 6 7 7 6 2 1 1 1 1
FIGURE 2. The two most common angiographic findings of TAR include pseudoaneurysm (white arrow) and a radiolucent, irregular line passing transversely across the aorta representing the intimal tear (black arrow). artery. None of the patients had falsely negative aortograms. In their seminal monograph on traumatic aortic rupture, Parmley et al reported that 97.5% of patients with TAR die within 7 days of injury.’ The fact that no patient with a negative aortogram died in the postinjury period, and none showed enlargement of the aortic shadow on follow-up chest roentgenogram, leads us to conclude cautiously that no patient had a falsely negative aortogram. The most common intrathoracic diagnoses in this group of patients were hemothorax, pulmonary contusion, and aortic rupture. These occurred with nearly equal frequency. The other thoracic diagnoses are listed in Table 3. The mortality rate in these patients was 17.5%. No deaths were attributable, in whole or in part, to aortography. Ductus diverticulum was found in two patients (OH%), and aberrent origin of the arch vessels was found in three (0.96%).
n Volume 8, Number 2 W March 1990
because the catheter caught on the intimal flap of the disruption. This in itself is almost diagnostic, but the question remains as to the necessity of viewing the vessels of the arch before surgical correction of TAR. With one of these three patients, an antegrade angiogram was performed through the right axillary artery; with two others, surgical exploration was undertaken without complete aortograms. It should be safe to proceed without a complete aortogram in instances such as this because the overwhelming majority of TARS occur just distal to the left subclavian artery, and only 3 of 314 patients (0.95%) showed abnormal origins of the arch vessels (Fig 4). If surgical correction is attempted without an aortogram, the surgical team should anticipate the possibility that the rupture might be located in an area other than distal to the left subclavian artery. According to our study, aortic rupture occurs three times more frequently than subclavian artery disruption. It is important to remember subclavian artery injury in blunt chest trauma because some subclavian artery injuries appear with a widened mediastinum on chest roentgenogram while others appear with apical pleural hematomas or a widened mediastinal shadow confined to one side of the hemithorax.“-*’ A previous report of nine patients with ruptured subclavian arteries showed a pleural cap in two patients, a widened mediastinum in one patient, and both findings in another
DISCUSSION The history of the circumstances surrounding the trauma, the patient’s symptoms and signs, and chest roentgenogram are nonspecific indicators of TAR.*-‘* Accordingly, a definitive diagnostic modality is required to confirm or exclude this lesion that is almost uniformly fatal when left untreated.13 About 85% of aortograms performed for suspected TAR are normal, raising the question of whether a safer, less invasive procedure could replace aortography. I4 To date, angiography holds the track record for accuracy and “is as safe as the person performing it”” (Figs 2 and 3). When we compare our accuracy rate of aortography with results reported for the accuracy of CT and MRI in aortic trauma, we conclude that aortography remains unsurpassed in accuracy for the diagnosis of TAR.1*2*5The complication rate was only 1.7%, and none of the complications required treatment. Our low complication rate is comparable with that achieved by Rose and Moore in their series of peripheral angiograms for trauma.16 The angiographic catheter could not be passed in a retrograde fashion beyond the aortic rupture in three patients
FIGURE 3. This aortogram of a patient with TAR shows a contained pseudoaneurysm.
STURM, HANKINS, AND YOUNG n AORTOGRAPHY FOLLOWING TRAUMA
patient.‘l Some subclavian artery disruptions appear with a unilateral absence of an upper extremity pulse in conjunction with a brachial plexus palsy.21 A ductus diverticulum was noted in 2 of 314 patients (0.64%). It is important for radiologists and surgeons to be aware of the existence of ductus diverticulum so that it will not be falsely interpreted as extravasation of contrast material distal to the left subclavian artery (Fig 5). False-positive interpretations of the thoracic aortogram occurred in two patients. Both patients underwent thoracotomy and were not found to have aortic rupture at the time of operation. One patient survived and one died. The patient who died succumbed to multiple system injuries, not as a complication of thoracotomy. These false-positive interpretations are the result of adjacent periaortic mediastinal hematoma that mimicked contrast extravasation at angiography. The incidence of ductus diverticulum in this series is low compared with other studies.22 Perhaps this is because the previous investigators were making a special effort to locate asymmetry in the distal aortic arch. Ductus diverticulum must be kept in mind so that a false-positive diagnosis of aortic rupture is not made. We conclude that thoracic aortography following blunt chest trauma to confirm or exclude the diagnosis of TAR is an accurate, sensitive, specific, and safe invasive procedure. Any small risk associated with thoracic aortography is counterbalanced by the risk of leaving this nearly uniformly fatal lesion untreated.
FIGURE 5. This aortogram illustrates a ductus diverticulum, a medial bulge on the medial border of the aorta (arrow). The authors thank Lynn Gilsrud for editorial assistance in the preparation of the article.
FIGURE 4. The aortogram shows a patient with TAR and aberrant origin of the right subclavian artery. RS, right subclavian artery; RC, right carotid artery; LC, left carotid artery, LS, left subclavian artery.
1. Mirvis SE, Bidwell JK, Buddemeyer EU, et al: Imaging diagnosis of traumatic aortic rupture: A review and experience at a major trauma center. Invest Radio1 1987;22:187-198 2. Heiberg E, Wolverson MK, Sundaram M, et al: CT in aortic trauma. AJR 1983;140:1119-1124 3. Heiberg E, Wolverson MK: CT of traumatic injuries of the aorta. Semin Ultrasound Comput Tomograph Magnetic Resonance 8: 1985;172-180 4. Akins EW, Carmicheal MJ, Hill J, et al: Preoperative evaluation of the thoracic aorta using MRI and angiography. Ann Thorac Surg 1988;44:499-507 5. Geisinger MA, Risius 8, O’Donnell JA, et al: Thoracic aortic dissections: Magnetic resonance imaging. Radiology 1985; 155:407-412 8. Abrams HL (ed): Abrams Angiography: Vascular and Interventional Radiology (ed 3). Boston, MA, Little, Brown, 1983, pp 339-484 7. Parmley LF, Mattingly TW, Manion WC, et al: Nonpenetrating traumatic injury of the aorta. Circulation 1987;27:187-198 8. Marsh DG, Sturm JT: Traumatic aortic rupture: Roentgenographic indications for angiography. Ann Thorac Surg 1978;21:337-340 9. Sturm JT, Marsh DG, Bodily KC: Ruptured thoracic aorta: Evolving radiological concepts. Surgery 1979;85:383-387 10. Sturm JT: Traumatic aortic rupture. Ann Thorac Surg 1979;27:390 (letter)
AMERICAN JOURNAL OF EMERGENCY MEDICINE l Volume 6. Number 2 n March 1990
11. Sturm JT, Cicero JJ: Chest roentgenographic findings of 26 patients with rupture of the thoracic aorta. Ann Emerg Med 1963;12:596-600 12. Sturm JT, Perry JF Jr, Olson FR, et al: The significance of svmptoms and signs in patients with traumatic aottic rupture. Ann’Emerg Med i964;13’:676-676 13. Marnocha KE. Maalinte DD: Plain-film criteria for excluding aortic rupture in blunt chest trauma. AJR 1965;144:19-21 14. Gundry SR, Williams S, Burney RE, et al: Indications for aortography. Radiography after blunt chest trauma: A reassessment of the radiographic findings associated with traumatic rupture of the aorta. Invest Radio1 1963;16:230-237 15. Fisher RG, Hadlock F, Ben-Menachem Y: Laceration of the thoracic aorta and brachiocephalic arteries by blunt trauma. Radio1 Clin North Am 1961;19:91-110 16. Rose SC, Moore EE: Trauma angiography: The use of clinical findings to improve patient selection and case preparation. J Trauma 1966;26:240-245
17. Sturm JT, Strate RG, Mowlem A, et al: Blunt trauma to the subclavian artery. Surg Gynecol Obstet 1974;136:915-916 16. Sturm JT, Bodily KC, Rothenberger DA, et al: Arterial injuries of the extremities following blunt trauma. Trauma 1960;20:933-936 19. Sturm JT, Dorsey JS, Olson FR, et al: The management of subclavian artery injury following blunt thoracic trauma. Ann Thorac Surg 1964;36:166-191 20. Sturm JT, Points BJ, Perry JF Jr: Hemopneumothorax following blunt trauma of the thorax. Surg Gynecol Obstet 1975; 141:539-540 21. Sturm JT, Cicero JJ: The clinical diagnosis of ruptured subclavian artery following blunt thoracic trauma. Ann Emerg Med 1963;12:45-47 22. Morse SS, Glickman MG, Greenwood LH: Traumatic aortic rupture: False-positive aortographic diagnosis due to atypical ductus diverticulum. AJR 1966;150:793-796