- Email: [email protected]
Surgical treatment for closed thoracic aortic injuries
Surgical treatment for closed thoracic aortic injuries Eighteen patients with acute aortic tear secondary to nonpenetrating thoracic trauma and seven with chronic tears have been operated upon. Of the patients with acute injury, 28 per cent had normal chest x-ray on admission. The diagnosis of acute aortic tear was made by aortography in all patients. Two patients died from causes unrelated to the aortic injury. In the group of 18 patients in which some method of perfusing the distal aorta was used, two developed paraplegia; in the group of 7 treated with simple aortic cross-clamping none occurred. The average length of operation was 6Vz hours and 33A hours in the two groups, respectively, average aortic cross-clamp time 69 and 23 minutes, respectively, and average blood administration 3,139 and 1,700 ml., respectively. We conclude that the diagnosis of acute aortic tear cannot be reliably made on clinical grounds, and that with properly trained personnel the method of choice is usually simple aortic cross-clamping during repair.
Azai Appelbaum, M . D . , Robert B. Karp, M . D . , and John W. Kirklin, M.D., Birmingham, Ala.
I N onpenetrating chest trauma is common. Uncommonly, the thoracic aorta is torn by such injuries. The mortality rate from aortic tears is high. The diagnosis of this condition from clinical data is difficult in our experience, and aortography must be used whenever aortic tear is suspected. We have had satisfactory results with less bleeding and shorter operations since abandoning the use of shunts of various types during the cross-clamping of the upper descending thoracic aorta for repair of such injuries. The experience Between April, 1969, and March, 1975, we have repaired nonpenetrating traumatic aortic injuries in 25 patients. Eighteen patients were treated in the acute stage (within 7 days of the time the trauma occurred), and 7 between 2 and 15 years after the trauma. Acute injury. Four of the 18 patients were admitted in shock. Thirteen patients were suspected of having aortic injury on admission because of a wide mediastinum in the chest x-ray. Five patients (27.8 per cent) had normal chest x-ray on admission and developed From the Department of Surgery, The University of Alabama School of Medicine and Medical Center, and VA Hospital, Birmingham, Ala. Received for publication July 18, 1975. Address for reprint: John W. Kirklin, M.D., Department of Surgery, University Station, Birmingham, Ala. 35294.
458
signs of aortic trauma 6 to 36 hours later. Only one patient had absence of femoral pulses. The diagnosis was established by aortogram in all 18 patients. In 17 patients the tear was 1 to 3 cm. distal to the origin of the left subclavian artery. In one patient, the tear was in the ascending aorta 2 cm. proximal to the inominate artery. Once the diagnosis was made, operation was undertaken immediately. Nine patients had partial tear of the upper descending thoracic aorta and 8 had complete transection. One of these patients had pseudocoarctation of the aorta as a result of the trauma. The patient with the tear in the ascending aorta had an extensive injury involving the entire aortic arch. Primary repair of the tears was possible in 7 patients. In 10, a short segment of Dacron tube was interposed between the ends of the aorta. In the patient with the tear in ascending aorta, total cardiopulmonary bypass with profound hypothermia was used, and the aortic arch was replaced with a Dacron tube preserving the strip of the transverse arch from which the head vessels arose. Measures to prevent spinal cord and lower body ischemia varied during this period. Eight patients were operated upon with iliac vein to iliac artery bypass with a pump oxygenator. Two patients had an external shunt in place between aorta proximal to the clamp and aorta distal to the clamp while the aorta was cross-clamped.
Volume 71 Number 3 March, 1976
4 59
Closed thoracic aortic injuries
Table I. Surgery for nonpenetrating thoracic aortic trauma
Surgical method Acute tear: Cardiopulmonary* bypass Iliac-iliac bypass External shunt No shunt
Average length of operation (hr.)
Average blood administration (ml.)
Average aorticclamp time (range, min.)
I
5'/2
3,500
44
—
—
8
7V4
3.600
1 paraplegia
1
2
5
3.000
—
1
7
3%
1.700
82 (39-180) 55 (43-67) 23 (15-30) 74 (57-90) 54 (45-63)
—
No. of pts.
Chronic tear: Iliac-iliac bypass External shunt
2
6%
3.000
5
6
3,000
Total
25
Complications
Hospital death
~ —
1 paraplegia 2(8%)
2(8<7c)
"TVar in ascending aorta.
In the last 7 consecutive patients, no shunt was used during the aortic cross-clamping. Some details of the experience are shown in Table I. One patient with an acute tear, operated upon with iliac-iliac bypass, had paraplegia at the end of the operation with mild improvement at time of discharge. Three large intercostal arteries had to be ligated as part of the procedure. The operations and aortic crossclamping were shortest and the amount of blood administered during operation the least when no shunt was used. Two of the 18 patients (1 1 per cent) died in the hospital. One died 4 weeks after surgery from pulmonary complications secondary to severe lung trauma, and the other died 4 days after surgery from subarachnoid hemorrhage secondary to fracture of the skull. Shunts had been used in both. Chronic injury. Seven patients presented 2 to 15 years after the trauma with aneursym of the descending thoracic aorta. All patients had radiologic evidence of aneurysm. Three had increasing back pain, one patient had progressive dysphagia, and one presented with hoarseness from recurrent laryngeal nerve involvement. The aorta had been transected in all patients, and an aneurysm was between the two ends of the aorta, which were 6 to 12 cm. apart. The aneurysm was resected, and a Dacron tube inserted. In two patients the procedure was performed with iliac-iliac bypass; in the
other 5 patients a temporary external shunt between aortic arch and descending aorta was used. None of these patients were operated upon since adoption of the no-shunt technique. There were no deaths among this group of patients (Table I). One patient was paraplegic postoperatively. Discussion Although Symbas and associates1, describe a characteristic triad which includes increased blood pressure in the upper extremities, decreased pulse amplitude and blood pressure in lower extremities, and roentgenographic evidence of widening of mediastinum, we consider traumatic aortic tear difficult to diagnose, since only one patient presented this triad and in 27.8 per cent of patients the chest x-ray on admission was normal. The appearance of suggestive signs only after 24 to 36 hours in some patients indicates that all patients with severe closed chest trauma must be closely monitored and have multiple chest x-rays during the first 48 hours after injury. When the suspicion of aortic injury develops, prompt aortography is indicated, and if aortic tear is diagnosed, immediate operation is indicated. The surgical procedure requires cross-clamping of the upper descending thoracic aorta. Dating back to the early observations of Crafoord, Ejrup, and Gladnikoff2 and Adams and van Geertruyden,3 the belief has been
4 60
The Journal of Thoracic and Cardiovascular Surgery
Appelbaum, Karp, Kirklin
held that cross-clamping the upper descending thoracic aorta more than about 20 minutes imposes a high risk of paraplegia. Therefore, total body hypothermia, left atrial-femoral bypass, femorofemoral (or iliac-iliac) bypass with a pump-oxygenator, or temporary aorticaortic (or subclavian-femoral) bypass without a pump or oxygenator or total heparinization have been used in an attempt to protect against this. Crawford and Rubio4 challenged these views, and demonstrated that the incidence of paraplegia and other complications was no higher when the upper descending thoracic aorta was clamped without a shunt than when a shunt was used, so long as the time of clamping did not exceed 45 minutes and hypotension did not develop. Crawford noted also that the various bypass techniques did not prevent paraplegia, a conclusion reached also by Brewer and colleagues5 relative to the repair of coarctation. The repair of closed traumatic injuries of the upper descending thoracic aorta should take no longer than about 30 minutes, if the operation is properly designed and executed. We, therefore, recently adopted Crawford's technique, and to date no patient has developed paraplegia, oliguria, or other complications. Also, with this technique the total
operation is shorter, the cross-clamp time is shorter, and considerably less blood is administered, all of which tend to minimize complications. We appreciate being able to include patients operated upon by our colleagues, Drs. N. T. Kouchoukos and Tom Hoyt. Dr. Steve Cain aided in the initial compilation of the data. REFERENCES 1 Symbas, P. N., Tyras, D. H., Ware, R. E., and Hatcher, C. R., Jr: Rupture of the Aorta. A Diagnostic Triad, Ann. Thorac. Surg. 15: 405, 1973. 2 Crafoord, C , Ejrup, B., and Gladnikoff, H.: Coarctation of the Aorta, Thorax 2: 121, 1947. 3 Adams, H. D., and van Geertruyden, H. H.: Neurologic Complications of Aortic Surgery, Ann. Surg. 144: 574, 1956. 4 Crawford, E. S., and Rubio, P. A.: Reappraisal of Adjuncts to Avoid Ischemia in the Treatment of Aneurysms of Descending Thoracic Aorta, J. THORAC. CARDIOVASC. SURG. 66: 693,
1973.
5 Brewer, L. A., Ill, Fosburg, R. G., Mulder, G. A. , and Verska, J. J.: Spinal Cord Complications Following Surgery for Coarctation of the Aorta. A Study of 66 Cases, J. THORAC. CARDIOVASC. SURG. 64: 368,
1972.
Azai Appelbaum, M . D . , Robert B. Karp, M . D . , and John W. Kirklin, M.D., Birmingham, Ala.
I N onpenetrating chest trauma is common. Uncommonly, the thoracic aorta is torn by such injuries. The mortality rate from aortic tears is high. The diagnosis of this condition from clinical data is difficult in our experience, and aortography must be used whenever aortic tear is suspected. We have had satisfactory results with less bleeding and shorter operations since abandoning the use of shunts of various types during the cross-clamping of the upper descending thoracic aorta for repair of such injuries. The experience Between April, 1969, and March, 1975, we have repaired nonpenetrating traumatic aortic injuries in 25 patients. Eighteen patients were treated in the acute stage (within 7 days of the time the trauma occurred), and 7 between 2 and 15 years after the trauma. Acute injury. Four of the 18 patients were admitted in shock. Thirteen patients were suspected of having aortic injury on admission because of a wide mediastinum in the chest x-ray. Five patients (27.8 per cent) had normal chest x-ray on admission and developed From the Department of Surgery, The University of Alabama School of Medicine and Medical Center, and VA Hospital, Birmingham, Ala. Received for publication July 18, 1975. Address for reprint: John W. Kirklin, M.D., Department of Surgery, University Station, Birmingham, Ala. 35294.
458
signs of aortic trauma 6 to 36 hours later. Only one patient had absence of femoral pulses. The diagnosis was established by aortogram in all 18 patients. In 17 patients the tear was 1 to 3 cm. distal to the origin of the left subclavian artery. In one patient, the tear was in the ascending aorta 2 cm. proximal to the inominate artery. Once the diagnosis was made, operation was undertaken immediately. Nine patients had partial tear of the upper descending thoracic aorta and 8 had complete transection. One of these patients had pseudocoarctation of the aorta as a result of the trauma. The patient with the tear in the ascending aorta had an extensive injury involving the entire aortic arch. Primary repair of the tears was possible in 7 patients. In 10, a short segment of Dacron tube was interposed between the ends of the aorta. In the patient with the tear in ascending aorta, total cardiopulmonary bypass with profound hypothermia was used, and the aortic arch was replaced with a Dacron tube preserving the strip of the transverse arch from which the head vessels arose. Measures to prevent spinal cord and lower body ischemia varied during this period. Eight patients were operated upon with iliac vein to iliac artery bypass with a pump oxygenator. Two patients had an external shunt in place between aorta proximal to the clamp and aorta distal to the clamp while the aorta was cross-clamped.
Volume 71 Number 3 March, 1976
4 59
Closed thoracic aortic injuries
Table I. Surgery for nonpenetrating thoracic aortic trauma
Surgical method Acute tear: Cardiopulmonary* bypass Iliac-iliac bypass External shunt No shunt
Average length of operation (hr.)
Average blood administration (ml.)
Average aorticclamp time (range, min.)
I
5'/2
3,500
44
—
—
8
7V4
3.600
1 paraplegia
1
2
5
3.000
—
1
7
3%
1.700
82 (39-180) 55 (43-67) 23 (15-30) 74 (57-90) 54 (45-63)
—
No. of pts.
Chronic tear: Iliac-iliac bypass External shunt
2
6%
3.000
5
6
3,000
Total
25
Complications
Hospital death
~ —
1 paraplegia 2(8%)
2(8<7c)
"TVar in ascending aorta.
In the last 7 consecutive patients, no shunt was used during the aortic cross-clamping. Some details of the experience are shown in Table I. One patient with an acute tear, operated upon with iliac-iliac bypass, had paraplegia at the end of the operation with mild improvement at time of discharge. Three large intercostal arteries had to be ligated as part of the procedure. The operations and aortic crossclamping were shortest and the amount of blood administered during operation the least when no shunt was used. Two of the 18 patients (1 1 per cent) died in the hospital. One died 4 weeks after surgery from pulmonary complications secondary to severe lung trauma, and the other died 4 days after surgery from subarachnoid hemorrhage secondary to fracture of the skull. Shunts had been used in both. Chronic injury. Seven patients presented 2 to 15 years after the trauma with aneursym of the descending thoracic aorta. All patients had radiologic evidence of aneurysm. Three had increasing back pain, one patient had progressive dysphagia, and one presented with hoarseness from recurrent laryngeal nerve involvement. The aorta had been transected in all patients, and an aneurysm was between the two ends of the aorta, which were 6 to 12 cm. apart. The aneurysm was resected, and a Dacron tube inserted. In two patients the procedure was performed with iliac-iliac bypass; in the
other 5 patients a temporary external shunt between aortic arch and descending aorta was used. None of these patients were operated upon since adoption of the no-shunt technique. There were no deaths among this group of patients (Table I). One patient was paraplegic postoperatively. Discussion Although Symbas and associates1, describe a characteristic triad which includes increased blood pressure in the upper extremities, decreased pulse amplitude and blood pressure in lower extremities, and roentgenographic evidence of widening of mediastinum, we consider traumatic aortic tear difficult to diagnose, since only one patient presented this triad and in 27.8 per cent of patients the chest x-ray on admission was normal. The appearance of suggestive signs only after 24 to 36 hours in some patients indicates that all patients with severe closed chest trauma must be closely monitored and have multiple chest x-rays during the first 48 hours after injury. When the suspicion of aortic injury develops, prompt aortography is indicated, and if aortic tear is diagnosed, immediate operation is indicated. The surgical procedure requires cross-clamping of the upper descending thoracic aorta. Dating back to the early observations of Crafoord, Ejrup, and Gladnikoff2 and Adams and van Geertruyden,3 the belief has been
4 60
The Journal of Thoracic and Cardiovascular Surgery
Appelbaum, Karp, Kirklin
held that cross-clamping the upper descending thoracic aorta more than about 20 minutes imposes a high risk of paraplegia. Therefore, total body hypothermia, left atrial-femoral bypass, femorofemoral (or iliac-iliac) bypass with a pump-oxygenator, or temporary aorticaortic (or subclavian-femoral) bypass without a pump or oxygenator or total heparinization have been used in an attempt to protect against this. Crawford and Rubio4 challenged these views, and demonstrated that the incidence of paraplegia and other complications was no higher when the upper descending thoracic aorta was clamped without a shunt than when a shunt was used, so long as the time of clamping did not exceed 45 minutes and hypotension did not develop. Crawford noted also that the various bypass techniques did not prevent paraplegia, a conclusion reached also by Brewer and colleagues5 relative to the repair of coarctation. The repair of closed traumatic injuries of the upper descending thoracic aorta should take no longer than about 30 minutes, if the operation is properly designed and executed. We, therefore, recently adopted Crawford's technique, and to date no patient has developed paraplegia, oliguria, or other complications. Also, with this technique the total
operation is shorter, the cross-clamp time is shorter, and considerably less blood is administered, all of which tend to minimize complications. We appreciate being able to include patients operated upon by our colleagues, Drs. N. T. Kouchoukos and Tom Hoyt. Dr. Steve Cain aided in the initial compilation of the data. REFERENCES 1 Symbas, P. N., Tyras, D. H., Ware, R. E., and Hatcher, C. R., Jr: Rupture of the Aorta. A Diagnostic Triad, Ann. Thorac. Surg. 15: 405, 1973. 2 Crafoord, C , Ejrup, B., and Gladnikoff, H.: Coarctation of the Aorta, Thorax 2: 121, 1947. 3 Adams, H. D., and van Geertruyden, H. H.: Neurologic Complications of Aortic Surgery, Ann. Surg. 144: 574, 1956. 4 Crawford, E. S., and Rubio, P. A.: Reappraisal of Adjuncts to Avoid Ischemia in the Treatment of Aneurysms of Descending Thoracic Aorta, J. THORAC. CARDIOVASC. SURG. 66: 693,
1973.
5 Brewer, L. A., Ill, Fosburg, R. G., Mulder, G. A. , and Verska, J. J.: Spinal Cord Complications Following Surgery for Coarctation of the Aorta. A Study of 66 Cases, J. THORAC. CARDIOVASC. SURG. 64: 368,
1972.