- Email: [email protected]
The potential effects of radiographic criteria to exclude aortography in patients with blunt chest trauma
J
THORAC CARDIOVASC SURG
1989;97:456-60
The potential effects of radiographic criteria to exclude aortography in patients with blunt chest trauma Results of a study of 32 patients with proved aortic or brachiocephalic arterial injury The purpose of tbis study was to test the effectiveness, in patients with known aortic or brachiocepbalic arterial injury, of five previoltily published radiographic criteria for excluding aortography in patients with blunt chest trauma. These criteria were (1) normal findings on erect chest radiograph; (2) normal aortic arch and left subclavian artery; (3) normal aortic arch, descending aorta, aortopulmonary window, tracheal position, and left paraspinal interface; (4) normal right paratracbeaI stripe and nasogastric tube position, and (5) normal aortic arch and tracheal and nasogastric tube position. One or more of these criteria were met in 6% to 25% of patient with major thoracic arterial injury, depending on the criteria of mediastinal used. Interestingly, two (6%) patients had radiographs that showed no specific si~ hemorrhage, which indicates that the chest radiograph is limited in its sensitivity to detect major thoracic arterial injury. Because of these results, we do not believe that attempts to limit aortography in patients with supine film evidence of mediastinal abnormality, based on the absence of certain s~ of mediastinal hemorrhage, are warranted. Furthermore, an abnormal radiograph cannot be relied on as the sole criterion for aortography if the goal of care is to detect as close to 100% of vascular injuries as possible.
John H. Woodring, MD, and Joseph G. King, MD, Lexington, Ky.
A
major concern of the recent literature on plain film screening for aortography, in the diagnosis of blunt traumatic injury to the aorta and its major branches, has been that too many aortograms yielding normal findings are currently being performed for each abnormal case found.!" Several investigators have recently published criteria that potentially could be used,5,6 or have been specifically designed to be used, 1,3, 4 to exclude aortography in traumatized patients on the basis of the absence of certain plain film signs of mediastinal hemorrhage on initial emergency room radiographs. From the Department of Diagnostic Radiology, University of Kentucky Medical Center, Lexington, Ky. Received for publication Feb. 12, 1988. Accepted for publication Sept. 2, 1988. Address for reprints: John H. Woodring, MD, Department of Diagnostic Radiology, University of Kentucky Medical Center, 800 Rose St., Lexington, KY 40536-0084.
456
Ayella,' Mirvis.v' and their associates have indicated that obtaining a normal erect anteroposterior chest radiograph in a trauma victim excludes the need for aortography. Others have also suggested that, even when the mediastinum is found to be abnormal, the absence of certain specific signs of mediastinal hemorrhage indicates either no chance of or a very low probability of major arterial injury. Mirvis and colleagues' have indicated that when the aortic arch and descending aorta, aortopulmonary window, trachea, and left paraspinal interface are normal, there is a 92% probability of no aortic rupture. Woodring, Loh, and Kryscio" have indicated that when there is simultaneous absence of deviation of a nasogastric tube and visualization of a normal right paratracheal stripe, there is 98% probability of no aortic rupture. Milne and associates' have indicated that a normal subclavian artery and aortic arch excludes aortic rupture. And Marnocha and Maglinte" have stated that if the aortic arch and its
Volume 97 Number 3
Aortography after blunt chest trauma
March 1989
457
Fig. 1. A, Erect chest radiograph of a patient who received blunt chest trauma in an automobile accident shows fractures of the fourth and fifth ribs on the left side. There are no specific signs of mediastinal hemorrhage. This criterion could have been used to exclude aortography. B, Aortography in this patient shows a large pseudoaneurysm at the aortic isthmus. contour are normal, and there is no deviation of either the nasogastric tube or trachea, then aortic rupture is completely excluded. Although these reports are concerned predominantly with criteria that might exclude rupture at the aortic isthmus, it is possible that some cases of ascending aortic rupture or brachiocephalic arterial rupture could meet these criteria, since mediastinal abnormalities in these conditions are often at a distance from the aortic knob and are often predominantly localized to the right side of the mediastinum. It is our awareness of cases of acute blunt traumatic rupture involving the brachiocephalic arteries or the aortic isthmus itself, which meet one or more of these criteria, that has prompted this report. We wish to report what effects the implementation of these five criteria for excluding aortography would have on our patients with proved acute arterial rupture.
Patients and methods Between July 1976 and January 1988, we have documented, both angiographically and surgically, 32 cases of acute traumatic rupture of the brachiocephalic arteries (eight cases) and aortic isthmus (24 cases). These 32 cases form the basis of this report. The initial supine chest radiographs obtained in the emergency room were analyzed by two independent observers for the following signs of mediastinal hemorrhage: mediastinal widening, abnormality of aortic contour or size, opacification of the aortopulmonary window, depression of the left main-stem bronchus, deviation of the
trachea to the right of midline, deviation of an opaque nasogastric tube to the right of the T-4 spinous process, apical cap sign, widening of the right para tracheal stripe, and widening of the paraspinal interfaces. One observer (J.H.W.) interpreted the supine radiographs with the knowledge that the patients had proved vascular injuries; the second observer (J.G.K.) read the supine films without knowledge of patient history as part of a random series of 90 patients representing a mixture of the 32 patients with arterial injury, 30 patients with normal aortograms who had received blunt chest trauma and who had undergone aortography for a high clinical or radiographic suspicion of vascular injury, and supine films on 28 normal control subjects. Agreement between the two observers for the 32 abnormal chest films was excellent, with complete agreement in 31 of the 32 cases. In one case the observers disagreed about whether there was opacification of the aortopulmonary window. Erect anteroposterior or posteroanterior films, obtained shortly after the initial supine films, were available in nine patients, and these were subjected to the same analysis. In addition, the medical records of these nine patients were reviewed to determine what effect the erect films had on patient management at the time they were obtained. Using the criteria described by Ayella,' Mirvis.' Marnocha and Maglinte,' Milne,' and Woodring" and their associates, we determined what percent of these proved cases of arterial injury would meet each of the criteria for excluding aortography.
Results The initial supine radiographs of two patients with acute traumatic laceration at the aortic isthmus (6%) were interpreted as showing no specific signs of medias-
The Journal of Thoracic and Cardiovascular Surgery
45 8 Woodring and King
Fig. 2. Failure of criteria of normal aortic arch and descending aorta, clear aortopulmonary window, normal tracheal position, and normal left paraspinal interface. The supine radiograph demonstrates a fractured left clavicle and large right para tracheal hematoma. The trachea is midline and all left-sided mediastinal structures, including the aortic arch and descending aorta, aortopulmonary window, and left paraspinal interface, are normal. Aortography revealed laceration of the right subclavian artery just distal to the origin of the vertebral artery.
tinal hemorrhage by both observers. These have subsequently been shown to other observers, who also agree that there are no specific signs of mediastinal abnormality. The only chest radiographic abnormalities in these two patients were a fracture of the right fifth rib in one and fractures of the left fourth and fifth ribs in the other. The supine radiographs of the 30 (94%) remaining patients were found to have one or more of the specific signs of mediastinal hemorrhage and were considered abnormal by both observers. Applying the five criteria for excluding aortography to these 32 patients produced the following results. Erect radiographs, criteria of Ayella and associates,' had been obtained in nine patients. In one of the patients with a normal mediastinum on the supine film, the erect film also showed no specific signs of mediastinal hemorrhage. In the remaining eight patients, mediastinal abnormality was evident on the supine film; however, the degree of mediastinal abnormality was much less evident on the erect film. In retrospect, it was found from the chart review that four of these eight erect films had originally been interpreted as showing no mediastinal abnormality, which delayed aortography from 24 to 72 hours. Both
observers agreed that, in retrospect, three had probably been misinterpreted ; however, both agreed that one was correctly interpreted as showing no specific signs of mediastinal hemorrhage (Fig. 1). Therefore, a normal mediastinum on erect radiographs was also present in two of 32 patients (6%). Both of these patients had traumatic laceration at the aortic isthmus. A normal aortic arch and left subclavian artery, criteria of Milne and associates,' were present in eight (25%) of the 32 patients, including three with rupture of the right subclavian artery and five with rupture of the aortic isthmus. A normal aortic arch and descending aorta, aortopulmonary window, tracheal position, and left paraspinal interface, criteria of Mirvis and associates,' were present in seven patients (22%) (Fig. 2), including three with rupture of the right subclavian artery and four with rupture of the aortic isthmus. A normal right para tracheal stripe and nasogastric tube position, criteria of Woodring, Loh, and Kryscio," were present in three patients (9%), including one with rupture of the left subclavian artery and two with rupture of the aortic isthmus. A normal aortic arch and normal tracheal and nasogastric tube position, criteria of Marnocha and Maglinte,' were present in two patients (6%) (Fig. 3), including one with rupture of the right subclavian artery and one with rupture of the aortic isthmus. Discussion
Ayella, I Mirvis,' Schwab,' Redman," and their associates have suggested that the supine chest radiograph is often overread as showing signs of mediastinal hemorrhage, in part because of magnification of mediastinal structures produced by obtaining the image at short focal-film distances. An erect chest radiograph, which may be obtained at greater focal-film distances, thereby reducing magnification of the mediastinum and giving a more accurate depiction of mediastinal anatomy, has been proposed as the solution to this problem. J, 2, 3, 7, 8 Mirvis and his co-workers' found that, although the supine chest radiograph was more sensitive for detecting signs of mediastinal hemorrhage than the erect view, it was less specific for aortic rupture than the erect view. They' found that by routinely obtaining the erect view they could better identify normal chest radiographs and could reduce the number of normal aortograms by 48%. In our experience we have frequently found it impossible to obtain erect films on such patients because of extensive multisystem trauma. In addition, the results of our study suggest that the reduction in apparent degree of mediastinal abnormality on the erect view may result in missed or delayed diagnoses of vascular injury. Although the specificity for vascular injury may be
Volume 97 Number 3 March 1989
greater on an erect chest film that is interpreted as abnormal than it is for the supine view, it is apparent that this increased specificity will occur at the expense of reducing the sensitivity of the test. In our study, erect films were negative for mediastinal hemorrhage in 6% of cases and contributed to a delayed diagnosis in an additional 9%. Sefczek, Sefczek, and Deeb? believed that the risk of aortography was too great to justify performing aortography on blunt trauma victims who had negative chest radiographs. However, the mortality of aortography is reported to be very low and has been estimated to be approximately 0.032% (one death in 3125 exams)." Mirvis and associates' have estiamted that, in patients with blunt chest trauma, a negative chest radiograph is 98% predictive of a normal aortogram. Still, it is possible that 2% of such patients could harbor a potentially fatal aortic or brachiocephalic injury (63 cases in 3125 patients). Thus it would appear that the risk of withholding aortography from patients with significant blunt chest trauma may be greater than the risk of the procedure itself. Many reports have suggested that radiographic abnormalities outside the mediastinum, including sternal fracture, first and second rib fracture, multiple rib fractures with a crushed chest, and posteriorly displaced clavicular fracture may increase the yield of initial plain film diagnosis above that possible by signs of mediastinal abnormality alone. I 1·17 Unfortunately, these findings were absent in our patients with a normal mediastinum. In our experience, 6% of aortic and brachiocephalic arterial ruptures have been discovered by performing aortography for the history of significant, blunt trauma to the thorax alone. Several recent publications have enumerated criteria that have been specifically designed to exclude aortography,':" or that could be used for this purpose.v" even when the initial supine film shows definite mediastinal abnormality. These studies have used "high probability" criteria, which are more frequently identified in patients with vascular injury, to indicate the need for aortography and have suggested that when these criteria are absent, aortography is not warranted. We found that these criteria would result in 6% to 25% of positive cases being excluded from consideration for aortography. Although cases of rupture of the brachiocephalic arteries may not show mediastinal abnormalities near the aortic knob, we were alarmed by the number of acute ruptures at the aortic isthmus that had abnormal mediastinal anatomy but showed no specific findings at or near the aortic knob. Therefore, when the mediastinum is found to be abnormal in any way, aortography is warranted.
Aortography after blunt chest trauma
459
Fig. 3. Failure of criteria of normal aortic arch and normal tracheal and nasogastric tube position. The erect radiograph demonstrates widening in the right paratracheal area (arrow). However, the area of the aortic arch is normal and the trachea and nasogastric tube are midline. Aortography showed a large pseudoaneurysm at the aortic isthmus.
The results of this study indicate that, in the majority of cases (94%), the chest radiograph in patients with blunt injury to the thoracic aorta and its major branches will show evidence of mediastinal abnormality pointing to the underlying injury. Therefore, an abnormal mediastinum on the chest radiograph remains the major indicator of the need for aortography. However, the chest radiograph does not detect all aortic and brachiocephalic injuries. Attempts to limit aortography because of negative erect views or the absence of certain "high probability" findings will further diminish the sensitivity of the test and are not warranted. Expanding the criteria for aortography to include certain fractures of the bony thorax, described earlier, may theoretically increase the yield of diagnosis beyond that possible by mediastinal abnormality alone, but did not do so in our study population. Performing aortography because of a history of significant blunt trauma to the thorax, despite a negative chest radiograph, remains the only way to detect the few patients with arterial rupture who initially have no abnormalities on the chest radiographs. REFERENCES 1. Ayella RJ, Hankins JR, Turney SZ, Cowley RA. Ruptured thoracic aorta due to blunt trauma. J Trauma 1977;17:199-205.
2. Mirvis SE, Bidwell JK, Buddemeyer EU, Diaconis IN,
The Journal of Thoracic and Cardiovascular Surgery
460 Woodring and King
3.
4.
5.
6.
7.
8. 9.
10.
Pais SO, Whitley JE. Imaging diagnosis of traumatic aortic rupture: a review and experience at a major trauma center. Invest Radiol 1987;22:187-96. Mirvis SE, Bidwell JK, Buddemeyer EU, et al. Value of chest radiography in excluding traumatic aortic rupture. Radiology 1987;163:487-93. Marnocha KE, Maglinte DDT. Plain-film criteria for excluding aortic rupture in blunt chest trauma. AJR 1985;144:19-21. Milne ENC, Imray TJ, Pistolesi M, Miniati M, Giuntini C. The vascular pedicle and the vena azygos. Part III: In trauma-the "vanishing" azygos. Radiology 1984;153:2531. Woodring JH, Loh FK, Kryscio RJ. Mediastinal hemorrhage: an evaluation of radiographic manifestations. Radiology 1984;151:15-21. Schwab CW, Lawson RB, Lind JF, Garland LW. Comparison of supine and upright portable chest films to evaluate the widened mediastinum. Ann Emerg Med 1984;13:896-9. Redman He. A rational approach to traumatic aortic rupture. Angiology 1973;24:255-63. Sefczek DM, Sefczek RJ, Deeb ZL. Radiographic signs of acute traumatic rupture of the thoracic aorta. AJR 1983;141:1259-62. Abrams HL. Thoracic aortography: technique, indications, and hazards. In: Abrams HL, ed. Abrams angiog-
11.
12.
13.
14.
15.
16.
17.
raphy: vascular and interventional radiology. 3rd ed. Vol. 1. Boston: Little Brown, 1983:339-52. DeMeules JE, Cramer G, Perry JF Jr. Rupture of aorta and great vesselsdue to blunt thoracic trauma. J THORAC CARDIOVASC SURG 1971;61:438-42. Wilson JM, Thomas AN, Goodman PC, Lewis FR. Severe chest trauma morbidity: implication of first and second rib fracture in 120 patients. Arch Surg 1978; 113:846-9. O'Sullivan MJ Jr, Folkerth TL, Morgan JR, Fosburg RG. Posttraumatic thoracic aortic aneurysm: recognition and treatment. Arch Surg 1972;105:14-18. Richardson JD, McElvein RB, Trinkle JK. First rib fracture: a hallmark of severe trauma. Ann Surg 1975; 181:251-4. Galbraith NF, Urschel HC Jr, Wood RE, Razzuk MA, Paulson DL. Fracture of first rib associated with laceration of subclavian artery: report of a case and reviewof the literature. J THORAC CARDIOVASC SURG 1973;65:649-52. Holmes TW Jr, Netterville RE. Complications of first rib fracture, including one case each of tracheoesophageal fistula and aortic arch aneurysm. J THORAC SURG 1956; 32:74-91. Fisher RD, Rienhoff WF III. Subclavian artery laceration resulting from fractured first rib. J Trauma 1966; 6:579-81.
Bound volumes available to subscribers Bound volumes of THE JOURNAL OF THORACIC AND CARDIOVASCULAR SURGERY are available to subscribers (only) for the 1989 issues from the Publisher, at a cost of $55.00 ($72.00 international) for Vol. 97 (January-June) and Vol. 98 (July-December). Shipping charges are included. Each bound volume contains a subject and author index and all advertising is removed. Copies are shipped within 60 days after publication of the last issue of the volume. The binding is durable buckram with the JOURNAL name, volume number, and year stamped in gold on the spine. Payment must accompany all orders. Contact The C.V. Mosby Company, Circulation Department, 11830 Westline Industrial Drive, St. Louis, Missouri 63146-3318, USA; phone (800) 325-4177, ext. 351. Subscriptions must be in force to qualify. Bound volumes are not available in place of a regular JOURNAL subscription.
THORAC CARDIOVASC SURG
1989;97:456-60
The potential effects of radiographic criteria to exclude aortography in patients with blunt chest trauma Results of a study of 32 patients with proved aortic or brachiocephalic arterial injury The purpose of tbis study was to test the effectiveness, in patients with known aortic or brachiocepbalic arterial injury, of five previoltily published radiographic criteria for excluding aortography in patients with blunt chest trauma. These criteria were (1) normal findings on erect chest radiograph; (2) normal aortic arch and left subclavian artery; (3) normal aortic arch, descending aorta, aortopulmonary window, tracheal position, and left paraspinal interface; (4) normal right paratracbeaI stripe and nasogastric tube position, and (5) normal aortic arch and tracheal and nasogastric tube position. One or more of these criteria were met in 6% to 25% of patient with major thoracic arterial injury, depending on the criteria of mediastinal used. Interestingly, two (6%) patients had radiographs that showed no specific si~ hemorrhage, which indicates that the chest radiograph is limited in its sensitivity to detect major thoracic arterial injury. Because of these results, we do not believe that attempts to limit aortography in patients with supine film evidence of mediastinal abnormality, based on the absence of certain s~ of mediastinal hemorrhage, are warranted. Furthermore, an abnormal radiograph cannot be relied on as the sole criterion for aortography if the goal of care is to detect as close to 100% of vascular injuries as possible.
John H. Woodring, MD, and Joseph G. King, MD, Lexington, Ky.
A
major concern of the recent literature on plain film screening for aortography, in the diagnosis of blunt traumatic injury to the aorta and its major branches, has been that too many aortograms yielding normal findings are currently being performed for each abnormal case found.!" Several investigators have recently published criteria that potentially could be used,5,6 or have been specifically designed to be used, 1,3, 4 to exclude aortography in traumatized patients on the basis of the absence of certain plain film signs of mediastinal hemorrhage on initial emergency room radiographs. From the Department of Diagnostic Radiology, University of Kentucky Medical Center, Lexington, Ky. Received for publication Feb. 12, 1988. Accepted for publication Sept. 2, 1988. Address for reprints: John H. Woodring, MD, Department of Diagnostic Radiology, University of Kentucky Medical Center, 800 Rose St., Lexington, KY 40536-0084.
456
Ayella,' Mirvis.v' and their associates have indicated that obtaining a normal erect anteroposterior chest radiograph in a trauma victim excludes the need for aortography. Others have also suggested that, even when the mediastinum is found to be abnormal, the absence of certain specific signs of mediastinal hemorrhage indicates either no chance of or a very low probability of major arterial injury. Mirvis and colleagues' have indicated that when the aortic arch and descending aorta, aortopulmonary window, trachea, and left paraspinal interface are normal, there is a 92% probability of no aortic rupture. Woodring, Loh, and Kryscio" have indicated that when there is simultaneous absence of deviation of a nasogastric tube and visualization of a normal right paratracheal stripe, there is 98% probability of no aortic rupture. Milne and associates' have indicated that a normal subclavian artery and aortic arch excludes aortic rupture. And Marnocha and Maglinte" have stated that if the aortic arch and its
Volume 97 Number 3
Aortography after blunt chest trauma
March 1989
457
Fig. 1. A, Erect chest radiograph of a patient who received blunt chest trauma in an automobile accident shows fractures of the fourth and fifth ribs on the left side. There are no specific signs of mediastinal hemorrhage. This criterion could have been used to exclude aortography. B, Aortography in this patient shows a large pseudoaneurysm at the aortic isthmus. contour are normal, and there is no deviation of either the nasogastric tube or trachea, then aortic rupture is completely excluded. Although these reports are concerned predominantly with criteria that might exclude rupture at the aortic isthmus, it is possible that some cases of ascending aortic rupture or brachiocephalic arterial rupture could meet these criteria, since mediastinal abnormalities in these conditions are often at a distance from the aortic knob and are often predominantly localized to the right side of the mediastinum. It is our awareness of cases of acute blunt traumatic rupture involving the brachiocephalic arteries or the aortic isthmus itself, which meet one or more of these criteria, that has prompted this report. We wish to report what effects the implementation of these five criteria for excluding aortography would have on our patients with proved acute arterial rupture.
Patients and methods Between July 1976 and January 1988, we have documented, both angiographically and surgically, 32 cases of acute traumatic rupture of the brachiocephalic arteries (eight cases) and aortic isthmus (24 cases). These 32 cases form the basis of this report. The initial supine chest radiographs obtained in the emergency room were analyzed by two independent observers for the following signs of mediastinal hemorrhage: mediastinal widening, abnormality of aortic contour or size, opacification of the aortopulmonary window, depression of the left main-stem bronchus, deviation of the
trachea to the right of midline, deviation of an opaque nasogastric tube to the right of the T-4 spinous process, apical cap sign, widening of the right para tracheal stripe, and widening of the paraspinal interfaces. One observer (J.H.W.) interpreted the supine radiographs with the knowledge that the patients had proved vascular injuries; the second observer (J.G.K.) read the supine films without knowledge of patient history as part of a random series of 90 patients representing a mixture of the 32 patients with arterial injury, 30 patients with normal aortograms who had received blunt chest trauma and who had undergone aortography for a high clinical or radiographic suspicion of vascular injury, and supine films on 28 normal control subjects. Agreement between the two observers for the 32 abnormal chest films was excellent, with complete agreement in 31 of the 32 cases. In one case the observers disagreed about whether there was opacification of the aortopulmonary window. Erect anteroposterior or posteroanterior films, obtained shortly after the initial supine films, were available in nine patients, and these were subjected to the same analysis. In addition, the medical records of these nine patients were reviewed to determine what effect the erect films had on patient management at the time they were obtained. Using the criteria described by Ayella,' Mirvis.' Marnocha and Maglinte,' Milne,' and Woodring" and their associates, we determined what percent of these proved cases of arterial injury would meet each of the criteria for excluding aortography.
Results The initial supine radiographs of two patients with acute traumatic laceration at the aortic isthmus (6%) were interpreted as showing no specific signs of medias-
The Journal of Thoracic and Cardiovascular Surgery
45 8 Woodring and King
Fig. 2. Failure of criteria of normal aortic arch and descending aorta, clear aortopulmonary window, normal tracheal position, and normal left paraspinal interface. The supine radiograph demonstrates a fractured left clavicle and large right para tracheal hematoma. The trachea is midline and all left-sided mediastinal structures, including the aortic arch and descending aorta, aortopulmonary window, and left paraspinal interface, are normal. Aortography revealed laceration of the right subclavian artery just distal to the origin of the vertebral artery.
tinal hemorrhage by both observers. These have subsequently been shown to other observers, who also agree that there are no specific signs of mediastinal abnormality. The only chest radiographic abnormalities in these two patients were a fracture of the right fifth rib in one and fractures of the left fourth and fifth ribs in the other. The supine radiographs of the 30 (94%) remaining patients were found to have one or more of the specific signs of mediastinal hemorrhage and were considered abnormal by both observers. Applying the five criteria for excluding aortography to these 32 patients produced the following results. Erect radiographs, criteria of Ayella and associates,' had been obtained in nine patients. In one of the patients with a normal mediastinum on the supine film, the erect film also showed no specific signs of mediastinal hemorrhage. In the remaining eight patients, mediastinal abnormality was evident on the supine film; however, the degree of mediastinal abnormality was much less evident on the erect film. In retrospect, it was found from the chart review that four of these eight erect films had originally been interpreted as showing no mediastinal abnormality, which delayed aortography from 24 to 72 hours. Both
observers agreed that, in retrospect, three had probably been misinterpreted ; however, both agreed that one was correctly interpreted as showing no specific signs of mediastinal hemorrhage (Fig. 1). Therefore, a normal mediastinum on erect radiographs was also present in two of 32 patients (6%). Both of these patients had traumatic laceration at the aortic isthmus. A normal aortic arch and left subclavian artery, criteria of Milne and associates,' were present in eight (25%) of the 32 patients, including three with rupture of the right subclavian artery and five with rupture of the aortic isthmus. A normal aortic arch and descending aorta, aortopulmonary window, tracheal position, and left paraspinal interface, criteria of Mirvis and associates,' were present in seven patients (22%) (Fig. 2), including three with rupture of the right subclavian artery and four with rupture of the aortic isthmus. A normal right para tracheal stripe and nasogastric tube position, criteria of Woodring, Loh, and Kryscio," were present in three patients (9%), including one with rupture of the left subclavian artery and two with rupture of the aortic isthmus. A normal aortic arch and normal tracheal and nasogastric tube position, criteria of Marnocha and Maglinte,' were present in two patients (6%) (Fig. 3), including one with rupture of the right subclavian artery and one with rupture of the aortic isthmus. Discussion
Ayella, I Mirvis,' Schwab,' Redman," and their associates have suggested that the supine chest radiograph is often overread as showing signs of mediastinal hemorrhage, in part because of magnification of mediastinal structures produced by obtaining the image at short focal-film distances. An erect chest radiograph, which may be obtained at greater focal-film distances, thereby reducing magnification of the mediastinum and giving a more accurate depiction of mediastinal anatomy, has been proposed as the solution to this problem. J, 2, 3, 7, 8 Mirvis and his co-workers' found that, although the supine chest radiograph was more sensitive for detecting signs of mediastinal hemorrhage than the erect view, it was less specific for aortic rupture than the erect view. They' found that by routinely obtaining the erect view they could better identify normal chest radiographs and could reduce the number of normal aortograms by 48%. In our experience we have frequently found it impossible to obtain erect films on such patients because of extensive multisystem trauma. In addition, the results of our study suggest that the reduction in apparent degree of mediastinal abnormality on the erect view may result in missed or delayed diagnoses of vascular injury. Although the specificity for vascular injury may be
Volume 97 Number 3 March 1989
greater on an erect chest film that is interpreted as abnormal than it is for the supine view, it is apparent that this increased specificity will occur at the expense of reducing the sensitivity of the test. In our study, erect films were negative for mediastinal hemorrhage in 6% of cases and contributed to a delayed diagnosis in an additional 9%. Sefczek, Sefczek, and Deeb? believed that the risk of aortography was too great to justify performing aortography on blunt trauma victims who had negative chest radiographs. However, the mortality of aortography is reported to be very low and has been estimated to be approximately 0.032% (one death in 3125 exams)." Mirvis and associates' have estiamted that, in patients with blunt chest trauma, a negative chest radiograph is 98% predictive of a normal aortogram. Still, it is possible that 2% of such patients could harbor a potentially fatal aortic or brachiocephalic injury (63 cases in 3125 patients). Thus it would appear that the risk of withholding aortography from patients with significant blunt chest trauma may be greater than the risk of the procedure itself. Many reports have suggested that radiographic abnormalities outside the mediastinum, including sternal fracture, first and second rib fracture, multiple rib fractures with a crushed chest, and posteriorly displaced clavicular fracture may increase the yield of initial plain film diagnosis above that possible by signs of mediastinal abnormality alone. I 1·17 Unfortunately, these findings were absent in our patients with a normal mediastinum. In our experience, 6% of aortic and brachiocephalic arterial ruptures have been discovered by performing aortography for the history of significant, blunt trauma to the thorax alone. Several recent publications have enumerated criteria that have been specifically designed to exclude aortography,':" or that could be used for this purpose.v" even when the initial supine film shows definite mediastinal abnormality. These studies have used "high probability" criteria, which are more frequently identified in patients with vascular injury, to indicate the need for aortography and have suggested that when these criteria are absent, aortography is not warranted. We found that these criteria would result in 6% to 25% of positive cases being excluded from consideration for aortography. Although cases of rupture of the brachiocephalic arteries may not show mediastinal abnormalities near the aortic knob, we were alarmed by the number of acute ruptures at the aortic isthmus that had abnormal mediastinal anatomy but showed no specific findings at or near the aortic knob. Therefore, when the mediastinum is found to be abnormal in any way, aortography is warranted.
Aortography after blunt chest trauma
459
Fig. 3. Failure of criteria of normal aortic arch and normal tracheal and nasogastric tube position. The erect radiograph demonstrates widening in the right paratracheal area (arrow). However, the area of the aortic arch is normal and the trachea and nasogastric tube are midline. Aortography showed a large pseudoaneurysm at the aortic isthmus.
The results of this study indicate that, in the majority of cases (94%), the chest radiograph in patients with blunt injury to the thoracic aorta and its major branches will show evidence of mediastinal abnormality pointing to the underlying injury. Therefore, an abnormal mediastinum on the chest radiograph remains the major indicator of the need for aortography. However, the chest radiograph does not detect all aortic and brachiocephalic injuries. Attempts to limit aortography because of negative erect views or the absence of certain "high probability" findings will further diminish the sensitivity of the test and are not warranted. Expanding the criteria for aortography to include certain fractures of the bony thorax, described earlier, may theoretically increase the yield of diagnosis beyond that possible by mediastinal abnormality alone, but did not do so in our study population. Performing aortography because of a history of significant blunt trauma to the thorax, despite a negative chest radiograph, remains the only way to detect the few patients with arterial rupture who initially have no abnormalities on the chest radiographs. REFERENCES 1. Ayella RJ, Hankins JR, Turney SZ, Cowley RA. Ruptured thoracic aorta due to blunt trauma. J Trauma 1977;17:199-205.
2. Mirvis SE, Bidwell JK, Buddemeyer EU, Diaconis IN,
The Journal of Thoracic and Cardiovascular Surgery
460 Woodring and King
3.
4.
5.
6.
7.
8. 9.
10.
Pais SO, Whitley JE. Imaging diagnosis of traumatic aortic rupture: a review and experience at a major trauma center. Invest Radiol 1987;22:187-96. Mirvis SE, Bidwell JK, Buddemeyer EU, et al. Value of chest radiography in excluding traumatic aortic rupture. Radiology 1987;163:487-93. Marnocha KE, Maglinte DDT. Plain-film criteria for excluding aortic rupture in blunt chest trauma. AJR 1985;144:19-21. Milne ENC, Imray TJ, Pistolesi M, Miniati M, Giuntini C. The vascular pedicle and the vena azygos. Part III: In trauma-the "vanishing" azygos. Radiology 1984;153:2531. Woodring JH, Loh FK, Kryscio RJ. Mediastinal hemorrhage: an evaluation of radiographic manifestations. Radiology 1984;151:15-21. Schwab CW, Lawson RB, Lind JF, Garland LW. Comparison of supine and upright portable chest films to evaluate the widened mediastinum. Ann Emerg Med 1984;13:896-9. Redman He. A rational approach to traumatic aortic rupture. Angiology 1973;24:255-63. Sefczek DM, Sefczek RJ, Deeb ZL. Radiographic signs of acute traumatic rupture of the thoracic aorta. AJR 1983;141:1259-62. Abrams HL. Thoracic aortography: technique, indications, and hazards. In: Abrams HL, ed. Abrams angiog-
11.
12.
13.
14.
15.
16.
17.
raphy: vascular and interventional radiology. 3rd ed. Vol. 1. Boston: Little Brown, 1983:339-52. DeMeules JE, Cramer G, Perry JF Jr. Rupture of aorta and great vesselsdue to blunt thoracic trauma. J THORAC CARDIOVASC SURG 1971;61:438-42. Wilson JM, Thomas AN, Goodman PC, Lewis FR. Severe chest trauma morbidity: implication of first and second rib fracture in 120 patients. Arch Surg 1978; 113:846-9. O'Sullivan MJ Jr, Folkerth TL, Morgan JR, Fosburg RG. Posttraumatic thoracic aortic aneurysm: recognition and treatment. Arch Surg 1972;105:14-18. Richardson JD, McElvein RB, Trinkle JK. First rib fracture: a hallmark of severe trauma. Ann Surg 1975; 181:251-4. Galbraith NF, Urschel HC Jr, Wood RE, Razzuk MA, Paulson DL. Fracture of first rib associated with laceration of subclavian artery: report of a case and reviewof the literature. J THORAC CARDIOVASC SURG 1973;65:649-52. Holmes TW Jr, Netterville RE. Complications of first rib fracture, including one case each of tracheoesophageal fistula and aortic arch aneurysm. J THORAC SURG 1956; 32:74-91. Fisher RD, Rienhoff WF III. Subclavian artery laceration resulting from fractured first rib. J Trauma 1966; 6:579-81.
Bound volumes available to subscribers Bound volumes of THE JOURNAL OF THORACIC AND CARDIOVASCULAR SURGERY are available to subscribers (only) for the 1989 issues from the Publisher, at a cost of $55.00 ($72.00 international) for Vol. 97 (January-June) and Vol. 98 (July-December). Shipping charges are included. Each bound volume contains a subject and author index and all advertising is removed. Copies are shipped within 60 days after publication of the last issue of the volume. The binding is durable buckram with the JOURNAL name, volume number, and year stamped in gold on the spine. Payment must accompany all orders. Contact The C.V. Mosby Company, Circulation Department, 11830 Westline Industrial Drive, St. Louis, Missouri 63146-3318, USA; phone (800) 325-4177, ext. 351. Subscriptions must be in force to qualify. Bound volumes are not available in place of a regular JOURNAL subscription.